Masters of Roman Prose (Anthologie Stylistique Commentaires (Michael - Von - Albrecht) PDF

ARCA
Classical and Medieval Texts, Papers and Monographs
23
General Editors: Francis Cairns, Robin Seager, Frederick Williams

Assistant Editors: Neil Adkin, Sandra Cairns
ISSN 0309-5541
FROM CATO TO APULEIUS
Interpretative Studies
MICHAEL VON ALBRECHT
Translated by NEIL ADKIN
X FRANCIS CAIRNS
Published by Francis Cairns (Publications) Ltd
c/o The University, Leeds, LS2 9JT, Great Britain
-)
Die Meister Romische Prosa first published by

Lothar Stiehm Verlag, 1979
Copyright ©Verlag Lambert Schneider
English translation first published by

Francis Cairns (Publications), 1989
Copyright© Neil Adkin, 1989
All rights reserved. No part of this publication may be repro-

duced, stored in a retrieval system, or transmitted, in any form or
by any means, electronic, mechanical, photocopying, recording,
or otherwise, without the prior permission of the Publisher.
British Library Cataloguing in Publication Data
Albrecht, Michael von

Masters of Roman prose from Cato to Apuleius:
interpretative studies.-(ARCA classical and medieval
texts, papers and monographs; ISSN 0309-5541; v. 23)
1. Latin literature to ca 500 - Critical studies
I. Title II. Series III. Meister riimischer prosa
von Cato bis Apuleius. English
870.9 '001
ISBN 0-905205-72-3
Printed in Great Britain by

Redwood Burn Ltd, Trowbridge, Wiltshire
.- ~~ '
CONTENTS
Preface lX
1 The beginnings of literary prose: M. Porcius Cato

(234-149 B.C.)
I On farming. Preface
1 Introduction, 2. 2 Sentence structure, 4. 3 Verbal repetition, 5.
4 Accumulation of synonyms, 7. 5 Overall structure, 8. 6. Summary, 8.
II Speech in the Senate for the Rhodians (167 B.C.) 9
1 The problem, 11. 2 Cato's special qualities and his method of
argumentation in the speech, 11. 3 Accumulation of synonyms and
alliteration, 14. 4 Verbal repetition, 15. 5 Word order, 17. 6 Conclusion 18.
III A Roman Leonidas 21
1 Introduction, 23. 2 Form and content, 26.
2 Two great orators: C. Gracchus (154-121 B.C.)

and Cicero (106-43 B.C.)
C. Gracchus: From the speech De legibus promulgatis
(122 B.C.) 33
Cicero against Verres 33
1 The problem, 35. 2 Attitude to language: Latinitas - mundities, 38.
3 Narrative technique: Brevi! as, 40. 4 Rationality: Acutum, 44. 5 Style and
emotion, 47. 6 Conclusion 51.
3 Caesar (100-44 B.C.)

I Funeral oration for Julia (69 B.C.) 54
II Reflection and rapid action (Gall. 7 ,27) 59
1 Objective style, 59. 2 Functional approach, 61. 3 Overall structure:
economy of means, 62. 4 Facultas dicendi imperatoria, 63. 5 Can dour or
pose? 64.
4 Sallust (b. 86 B.C.)

I Ambition 68
1 Phonetics, vocabulary, syntax, 68. 2 General survey: sentence structure
and sequence of thought, 72. 3 Symmetry and asymmetry, 74. 4 Catonian
and Sallustian elements, 76.
II Triumph through treachery 78
1 Form and content, 78. 2 Centripetal style, 81. 3 Dramatic qualities, 83.
4 Structural intent, 84. 5 Psychology and authorial interpretation, 85.
5 Sullan and Augustan historiography: Claudius

Quadrigarius and Livy
Q. Claudius Quadrigarius (Sullan period) 86
Livy (59 B.C.-17 A.D.) 87
A. Comparison of content 90
B. Comparison of language and style 93
1 Claudius Quadrigarius, 93. 2 Livy, 96.
C. Narrative structure 99
1 Claudius Quadrigarius, 99. 2 Livy, 100.
6 Two philosophical texts

I Cicero (106-43 B.C.): Earthly glory and true
immortality 102
1 Sequence of thought, 103. 2 Sentence connection, 104. 3 Multiplicity in
sentence structure, 104. 4 Emphatic positions in the sentence, 105.
5 Vocabulary, 105. 6 Form and content, 106. 7 Conclusion, 110.
II Seneca (d. 65 A.D.): On the value of time 112
1 Form and the sequence of ideas, 113.2 Vocabulary, 119.3 Metaphorical
language, 119. 4 Sentence connection, 120. 5 Brilliance; wit; 'aggressive'
style, 120. 6 Rhetorical modes of thought, 121. 7 Seneca - an anti-
Cicero? 123.
7 Petronius (d. 66 A.D.)

Table talk from the 'Satyricon' 125
A. Language and style 126
1 Specific points: vulgarism and hypercorrection, 126.2 Vocabulary, 127.
127. 3 Metaphorical language; 'elevated' and 'humble' elements of
style, 128. 4 Elliptical expression and implicit meaning, 130. 5 Formulaic
elements 131. '
B. Structure and sentence connection 131
8 Tacitus (cos. 97 A.D.): A speech of the Emperor
Claudius. Original and literary recasting
Senatus consultum Claudianum (oratio Claudii) de iure
honorum Callis dando, 48 A.D. 136
Tac. ann. 11,24 140
A. Overall structure and sequence of ideas 141
1 Claudius, 141. 2 Tacitus, 142. 3 Comparison, 143.
B. Language and style 147
1 Claudius, 147. 2 Tacitus, 152. 3 Stylistic comparison, 156.
9 The Younger Pliny (cos. 100 A.D.)

A writer's success in the hunt 160
10 Apuleius (b. c. 125 A.D.)
An abortive bid for salvation 167
1 Narrative structure, 168. 2 A donkey's standpoint, 169. 3 Clarity of
expression, 171. 4 Highlights, 173. 5 Detachment and a higher level
of communication, 175.
List of abbreviations 177

Index of selected passages 183
Word and subject index, by Wilfried Stroh 185
Preface
The texts in this book are drawn from four centuries, extending from
Rome's rise to world dominion in the struggle with Carthage up to
her late flowering under the Antonines. For all their artistic
detachment they still reflect many changes: vigour and freshness at
the beginning; sharp conflict in the age of the Gracchi; the annalistic
method of Sulla's day; Caesar's career as a military commander;
intellectual brilliance in the death throes of the Republic; Augustan
reform and the idealized vision of a great past; the voice of the
preacher and the scepticism of the novelist set against the reality of
Nero's regime; under Trajan, a sigh of relief, belated candour and the
cosy world of the man of letters; and then finally Second Sophistic
and cosmopolitan humour against a background of mysticism.
The range is also wide in terms of space. Rome's writers hardly
ever came from the capital but from Italy, Gaul, Spain and Africa.
The genres too are marked by great diversity: from the rhetorical
treatise, formal historiography, the less formal commentarius, and
the philosophical dialogue, to the stylized letter and novel.
In making a selection from such abundant and varied material, it
has been necessary to leave out technical writers, specimens of legal
and official language, unliterary letters, and, above all, christian
literature: each of these areas deserves separate treatment. Within the
book's self-imposed limits, variety has been the aim. Besides
generally known authors, it is also intended that some who nowadays
are not so much read should be given their due. These include Cato,
who as Rome's first prose writer has been surprisingly neglected in
literary and linguistic studies, as well as Gaius Gracchus, one of
Rome's greatest orators. Readers who feel Cicero is not adequately
represented with just one philosophical text and two passages from
the speeches, should consult the writer's detailed treatment of
Cicero's language and style in Pauly-Wissowa's Rea/encyclopaedie
(Suppl. XIII). Of course, however much one tries to achieve a
representative cross-section, every selection is bound to be subjective
to a certain degree; and so one has to say with Quintilian: 1 sunt et alii
script ores boni, sed nos genera degustamus, non bib/iothecas excutimus.
The intention has not been to produce an alternative to a literary
1
Inst. 10, 1, 104.
ix
X PREFACE
history, or to present the development of Latin prose style in its

entirety. Neither is it intended to rival Norden's monumental 'Antike
Kunstprosa' or A. D. Leeman's 'Orationis Ratio', both of which are
sometimes based more on ancient literary theory than on actual
interpretation.
The aim of this book is more modest and more practical: its
purpose is to illustrate the wide possibilities of Roman prose artistry
by means of texts that are important or characteristic as regards form
and content. Only on the basis of specific examples can one hope to
gain new insights.
Particular attention is given to language and style,Z and especially
to the border areas between linguistics and literary criticism: here
belong questions of rhetorical and stylistic technique, syntax, and
narrative structure. It is frequently possible to reveal the inner
workings of rhetorical technique and let it have its rightful place in
ancient culture - not as a lifeless assemblage of precepts, but on the
contrary as a creative way to structure language and thought. Where
on the other hand narrative texts are concerned, the modern question
of narrative structure as a 'large-scale syntax' 3 is pursued further. For
all their convenience, labels are of course less important than the
individual character of each text. There are dangers in the sort of
determinism that thinks only in terms of generic and period styles, no
less than in jumping to hasty conclusions on the basis of a
psychologizing approach. Between these two extremes it is constantly
necessary to look out for the free play of personality and the creative
uniqueness of the work of art. Naturally one must also keep in mind
the subject-matter and the social background.
Prose as a form of literary representation appealed to one side of
Roman nature simply because it is inherently factual. Other ten-
dencies made greater refinement possible. From Rome comes the
element of social psychology, which extends from interpersonal
communication to politics. Roman too is the way individuality is
stylized behind the mask of the work. And most Roman of all is the
strict sense of formal, orchestrated construction.
Here it is not proposed to examine the history of the language in
terms of the history of ideas. Bearing in mind the importance
attaching to artistic form and the distance that exists between an
2 On the concept of 'style' see now W. Ax Sprachstil in der lat. Philo Iogie Beitr. zur
Altertumswiss. I (Hildesheim/New York 1976).

3 See the present writer's discussion in Ovid (Darmstadt 1968) pp.451-467. On 'text
syntax' cf. A. Scherer Handbuch der lateinischen Syntax (Heidelberg 1975) pp.104-112.
PREFACE xi
experience and its formulation, one has to treat each case indi-
vidually. All the same the interpretations given can be a stimulus to
tracing the way in which many key words are at first used
unreflectingly, and later elevated to a higher plane of meaning;
before long the content is lost and the words are treated ironically;
and then finally they are upgraded once again as the result of some
new philosophical attitude or religious mood.
Various methods of interpretation have been chosen, so as to
accommodate each individual text. It is hoped that this deliberate
multiplicity will be seen as an encouragement to study the artistry of
the Latin language. 4
The book is dedicated to my students.
Michael von Albrecht
4 For a thorough discussion of some chapters of this book and a rich bibliography,
see G. Calboli Nota di aggiornamento a Eduard Norden, La pros ad' arte antica dal VI
secolo all' eta della rinascenza (Rome 1986) pp. 971-1185.
1
The beginnings of literary prose
M. Porcius Cato (234-149 B.C.)
I. On farming
Preface 1
Est interdum praestare mercaturis rem quaerere, nisi tam periculosum
sit, et itemfenerari, si tam !zonestum sit. maiores nostri sic habuerunt et
ita in legibus posiverunt: furem dupli condemnari, feneratorem qua-
drupli. quanta peiorem civem existimarint feneratorem quam fur em,
hinc licet existimare. 2. et virum bonum quom laudabant, ita laudabant:
bonum agricolam bonumque colonum; 3. amplissime laudari existi-
mabatur, qui ita laudabatur. mercatorem autem strenuum studiosumque
rei quaerendae existimo, verum, ut supra dixi, periculosum et calami-
tosum. 4. at ex agricolis et viri fortissimi et milites strenuissimi
gignuntur, maximeque pius quaestus stabilissimusque consequitur
minimeque invidiosus, minimeque male cogitantes sunt qui in eo studio
occupati sunt. nunc, ut ad rem redeam, quod promisi institutum,
principium hoc erit.
It is true 2 that it is sometimes better to acquire wealth through commercial
transactions -if only it were not so hazardous -and equally to lendmoney-
if only it were so moral. That is what our forefathers thought 3 and they laid it
down in their laws as follows: the thief should be condemned to pay double
the amount, the usurer to pay four times as much. From this can be gauged
how much worse in their opinion the usurer was as a fellow-citizen than the
thief. When they praised a good man, they praised him thus: 'A good farmer
1
Text in M. Catonis De agricultura rec. A. Mazzarino (Leipzig 1962). Cf. now also
Caton, De!' agriculture texte etabli, trad. et commente parR. Goujard (Paris 1975); M.
Porcius Cato, Vom Landbau, Fragment e. Aile erhaltenen Schriften lat.-dt. hrsg. von 0.
Schonberger (Munich 1980). Monograph by F. Della Corte Catone censor e. La vitae Ia
fortuna (Turin 1949, Florence 1969 2).
2 On the translation cf. also P. Thielscher Des M. Cato Belehrung iiber die
Landwirtschaft (Berlin 1963); F. Klingner Romische Geisteswelt (Munich 1956 3) p.56

(= 1965 5 p.54; originally in Die Antike 10 [1934]pp.254f.). Commentators here explain
est as being equal to est, ut. In the present writer's opinion it has an affirmative nuance
here ('yes, granted'), containing a hidden admission that indirectly prepares the way
for the opposite idea later (farming is better). SoT. Birt PhW 35 (!915) pp.922-928.
3
On sic habuere cf. Sail. lug. 114.
2 THE BEGINNINGS OF LITERARY PROSE
and a good husbandman'. Whoever was thus praised received in their

opinion the most splendid praise. A businessman however is in my view
energetic and keen on making money, but, as I said, exposed to danger and
misadventure. But farmers produce the bravest soldiers, and they have a
source of income that is the most moral and most enduring and arouses the
least hostility, and the people who are engaged in this activity incline least to
evil thoughts. But to get back to the point, to the project that was promised. 4
This will be the beginning.
1. Introduction
'Prose offered him a more favourable basis [than poetry] and so he
used all his unique versatility and energy to create a prose literature
in his native language.' 5 Mommsen was followed by Norden 6 in the
perception that Cato is not only an opponent of the Greeks, but also
their student. By converting a general moral judgment of Cicero's
into a comment on style, Norden sees Cato's language as an
expression of his personality.' In a comparison with the 'much more
developed' style of his historical writing, 8 he describes the diction of
the manual on agriculture as still 'quite rough'. One should of course
go further and differentiate with~n De agricultura itself between the
style of the proem and that of the actual prescriptions.
Leo 9 recognized that greater care had been given to the style of the
praefatio; he speaks of its 'deliberate redundance and in its repetitions
and contrasts something approaching figures of speech'. In addition
Kappelmacher 10 noted that the entire prologue is divided into
4 So already Birt (see p.l n.2). See also Kroner-Szantyr ThLL 7, I, 13 ( 1962) 1994,3.
A different interpretation is given in Thielscher Joe. cit. (after Friedrich: institutum as

genitive). Different again is K. Buchner Beobachtungen iiber Vers und Gedankengang
bei Lukrez Hermes Einzelschriften I (Berlin 1936) pp.34-38 (after Wunsch: supine).
Klingner leaves the words untranslated.
5 T. Mommsen Romische Geschichte I (Berlin 1874 6) p.937. On De agricu/tura ib.
pp.929f. 6
Die antike Kunstprosa I (1898; Leipzig/Berlin 1923 4 ) p.l64.
7 Cic. rep. 2, I. There has never been any doubt about the correct interpretation of
this passage (cf. e.g. the translation of K. Buchner [Zurich 1960 2] p.l71: 'a life in
complete agreement with his words'). Norden's interpretation is all the more
surprising, in that he denies the validity of the principle 'Le style c'est l'homme' for
antiquity ib. p.l2.
8 Loc. cit. p.l65. In addition Norden takes into consideration the style of the
speeches (legal style, Greek rhetoric, poetic color of the ecphraseis). On the general
difference between De agricu/tura and the other works ofCato, see also S.V.F. Waite
'A computer assisted study of the style of Cato the Elder with reference to Sallust and
Livy' HSCP 74 (1970) pp.348-349. Cf. also S.V.F. Waite 'Approaches to the analysis
of Latin prose applied to Cato, Sallust, and Livy' Revue de /'Organis. Internal. pour
/'Etude des Langues Anciennes par Ordinateur 2 (1970) pp.91-120.
9 Geschichte der romischen Literatur I (Berlin 1913) p.274.
10 'Zum Stil Catos in De re rustica' WSt 43 (1922/23) pp.l68-172. Kappelmacher
leaves it open whether Cato's arrangement here is deliberate or not (p.l71).

M. PORCIUS CATO
corresponding parts in the manner of old-Italic sacral language,

while A.D. Leeman looked in our text for the principles of Greek
rhetoric, 11 starting above all from the ideas of periculum and
honestum~~hese two aspects are treated in Cato's prologue in the
sequencf abba. The chiastic structure had already been observed
here by Biichner. 13 1t is said by Leeman to have come easier to ancient
sensibility than the arrangement a b a b, which seems more natural to
us; here the train of thought starts up twice. Accordingly the ring
composition cannot be used as an argument for the influence of
Greek rhetoric; Leeman himself admits this. It might be added by
way of clarification that this feature certainly is an archaic mode of
construction, even if it does not have to be regarded as the no·rmal
ancient practice. 14 Leeman thinks the theme of the praefatio is
developed according to the TEAtKa KE(jHiA.ata of the genus delibe-
rativum. Cato emphasizes the utilitas of farming, its freedom from
periculum and its value as honestum. This would make the prologue
of the manual a suasio in miniature, in which the reader is exhorted to
take an interest in agriculture, the traditional activity of the Roman.
Fuhrmann expresses reservations about applying Hellenistic
categories to Cato's prologue. 15 Janson 16 however agrees with
Leeman. Cato may well have been familiar with Hellenistic rhetoric. 17
Even so the aspects mentioned above are simply imposed by the
subject: because Cato's remarks are so profoundly functional, the
philologist has difficulty here in distinguishing art from nature.
Klingner sees proof of Greek influence in the mere fact that Cato
starts from different forms of earning a living (agriculture, trade,
financial dealings). 18 One thing seems certain: it is too much to say
11 Orationis ratio I (Amsterdam !963) pp.2!-24. esp. 22f.

12 Utilitas in duas partes in civili consultatione dividitur: tutam hones/am. Tuta est
quae conficit ins/antis aut consequentis periculi vitationem qualibet ratione (Rhet. ad
11
Her. 3,2,3). Cited above p.2 n.4.
14
The lay-out of early inscriptions can also be used as an illustration of this archaic
mode of thought, cf. E. Zinn Schlangenschrift (1950/51). Now in: E. Pfohl (ed.) Das
Alphabet (Darmstadt 1968) pp.293-320.
15 Gnomon 38 ( 1966) pp.356-364, esp. 360. Cf. also M. Fuhrmann Das systematische
Lehrbuch. Ein Beitrag zur Geschichte der Wissenschaften in der Antike (Gottingen 1960)
pp.l57-159.
16
Latin Prose Prefaces Studia Latina Stockholmiensia 13 (Stockholm 1964) p.84
n.2.
17 On the influence of Greek rhetoric cf. also Norden, cited above p.2 n.6. But (e.g.)
the double trochee in the clausula can also be due to the tradition of Roman sacral
language (A. Kappelmacher, cited above p.2 n.IO).
" F. Klingner Joe. cit. 1956 1 p.56; 1965 5 p.54 (originally in Die Antike 10 [1934]
p.254): 'There Cato recommends and praises agriculture by contrasting it with other
that the form of the praefatio is actually determined by Hellenistic

rhetoric; rather its ring c;omposition is archaic. Archaism is also
visible in the language: 19paxataxis!0 redundance, accumulation of
synonyms, 21 repetition of words, 22 and alliteration. 23
Janson 24 tries to take account of the plurality of themes in the
prologue by distinguishing between the ethical and the economic
argument. In Janson's view it is difficult to reconcile the ethical
argument with the fact that Cato allows for a Hellenistic plantation
economy, 25 in which the landowner who lives in the city has virtually
no direct experience of the moral influence of farm-work. 26 Thus
while the moral arguments presupposed conditions that no longer
existed, the economic arguments, recommending an advantageous
form of investment, may have seriously interested members of Cato's
class. In this way Janson arrives at a result that does more to confirm
Kienast's 27 'Hellenistic' picture of Cato than that given by Klingner, 28
even though Janson tries to do justice to both approaches. 29 Janson's
argument is of course more concerned with the content of the
praefatio than with its form. A fresh examination of the stylistic form
of this important prologue appears therefore to be justified.
2. Sentence structure
A tendency to parallel sentence structure is visible in almost all parts
of the text. 30 It is however noticeable that the trend towards
parallelism is not the only one. Alongside it is a second feature that
occupations; this is an idea that is also found in Panaitios and Musonios and is
moreover related to other Greek introductions, where the value of a body of
knowledge is recommended.'
" R. Till Die Sprache Catos Philologus Supplementband 28, Heft 2 (Leipzig 1935).
20
Till p.7. On the archaic 'additive' style cf. H. Frankel 'Eine Stileigentiimlichkeit
der friihgriechischen Literatur' NGG ( 1924) pp.63-103 and 105-127, now in H. Frankel
Wege und Formen fhlhgriechischen Denkens (Munich 1955) pp.40-96.
" For verbal duplication in Cato (with many examples) see Till pp.27f.
22 21 24
Cf. Till p.26 on anaphora. Cf. Till p.25. See p.3 n.l6.
25
Cato agr. 12-13; 18-22. On this cf. T. Frank Rome and Italy of the Republic(vol. I
of An Economic Survey of Ancient Rome) (Paterson, New Jersey 1933; repr. 1959)
pp.171f. Frank presents the texts clearly, but provides them with only very short
explanations.
26 Cf. Cato 2, I. In chapter 5 (3,2 with different division) the owner is urged to build
a good villa rustica on his property. The purpose is of course to induce him to go there
more often.
27
D. Kienast Cato der Zensor (Heidelberg 1954) pp.88; 116; 134 and often. Cf. now
also K.D. White 'Roman Agricultural Writers I: Yarra and his predecessors' ANRWI
4 ( 1973) pp.439-497. Considerable knowledge of Greek language and science is also
posited by S. Boscherini Linguae scienza greca ne! 'De agri cu!tura' di Catone (Rome
29
1970). " Cited above p.l n.2. Janson p.87 n.12.
10
· On this compare A. Kappelmacher (cited p.2 n.10).
M. PORCIUS CATO
can be observed again and again: this is the tendency for the secondl
part of the sentence to turn out shorter than the first; it does so /
contrary to the law of increasing cola. This is what gives the__.j
impression of ruggedness that comes from a very individual Catonian
rhythm. 31 A further consequence is that a telling remark is always
kept in reserve for the end of the sentence:
- si tam honestum sit.
- feneratorem quadrupli.
- bonum agricolam bonumque co/anum.
- periculosum et calamitosum.
- minimeque male cogitantes sunt qui in eo studio occupati sunt.
In one case a phrase is inserted earlier as a kind of dramatic pause
that reinforces the effect of the final words: verum, ut supra dixi,
periculosum et calamitosum. In other cases a stressed demonstrative
gives the end of the sentence its punch:
- hinc licet existimare.
- qui ita laudabatur.
- principium hoc erit.
(Cf. also si tam honestum sit).
The whole text is characterized by an alertness for the meaning of
the initial and final position in the sentence. It seems to be a principle
of Catonian construction to fill the final position with concise and
meaningful words, and to combine this with reduced length in the
second half of the sentence.
3. Verbal repetition
Alongside this specifically Catonian way of filling out the sentence, it
is also striking that on many occasions words are_!~Pe(tted~ All of
these cases cannot however be ascribed indifferently to a striving
after intensification (as is attempted by Leeman p.22). Rather the
function of the repetitions is in part thematic and structural, and in
part one that has yet to be characterized. The structural importance
of the repetition is clear in the case ofjurem andfeneratorem, where
the same words are required on technical grounds. No emphasis on
the other hand is involved in the recurrence of the verb existimare in
as many as four consecutive constructions, without any recognizable
rhetorical intention. It is natural to think first of negligence in an area
the author regards as unimportant; this would be the counterpart to
31
E. Lindholm Stilistische Studien zur Erweiterung der Satzglieder im Lateinischen
(Lund 1931) p.52 notices only two examples of'rhetorical decrease' in Cato: Orig. 4,3
Jordan(= 79 P.) and 7, I Jordan(= 108 P.). He has missed the importance of the device
in De agr. ('but they are lacking in De agr.').
meticulousness in crucial areas. Cato devotes attention to the

important points in every sentence; however he clearly makes no
attempt to give careful shape to the other parts as well. Existimare is a
convenient verb, which suggests itself four times in succession as the
most obvious choice in different contexts. Cato would no doubt have
thought it petty to seek variety here, since this has no importance for
the clarity of the discourse. Only here, in an area that is unimportant,
does one notice what has been called (not quite correctly) the egestas
linguae LatinaeY
The point of repeating existimare is not intensification. Another
type of repetition is involved here: an attempt must now be made to
define it. Existimare has the advantage of being inconspicuous;
consequently it does not distract attention from the more important
parts of the discourse. The same is true of the fourfold laudare in the
sentence: et virum bonum cum laudabant, ita laudabant: bonum
agricolam bonumque colonum; amplissime laudari existimabatur, qui
ita laudabatur. It is not the verb laudare that is being stressed:
attention is directed entirely to the adjacent words (amplissime; ita).
The repetition thus serves here to give greater prominence to the
1
other words in the sentence. This is a stylistic device that strikes us as
1 'natural' rather than 'rhetorical'. It certainly achieves the clarity that
\ is its ultimate purpose.

In this kind of verbal repetition lies an element of 'oral style',
which should not be confused with a low stylistic standard. In oral
communication too there are considerable variations of stylistic
level, extending from everyday speech to epic and prayer. In man:~~
respects Cato's literary activity represents a first step towards written
language, and it still bears traces of oral style; this is particularly so in
the elevated mode. The recurrence in unstressed position of un-
varying or relatively colourless words can be seen as a fresh
component to be set alongside those oral characteristics of Cato's
32 On the stylistic value of existimare see B. Axelson Unpoetische Worter (Lund
1945) p.64; in this case the range of synonyms was in fact smaller than might be
supposed. On the 'poverty' of Latin see J. Marouzeau Quelques aspects de laformation
du latin litteraire (Paris 1949) pp.109f. 'En revanche, Ia lecture d'un ouvrage comme
celui de Caton revele !'indigence de Ia langue quand il s'agit d'enoncer les jugements ou
meme seulement d'exprimer des manieres d'etre: tout ce qui est remarquable par Ia
grandeur est qualifie de magnus; par Ia qualite, de bonus: on compte jusqu'a une
vingtaine d'exemples de bonus et bene dans les seuls chapitres 1 a 4 duDe agricultura.'
(cf. ib. p.96 on indigence in Cato's language and ib. 93ff., the chapter on 'indigence
originel!e', referring to Marouzeau 'Patrii sermonis egestas' Eranos 45 [1947] pp.22-24;
'Essai sur Ia distinction des styles' RPh 45 [1921] pp.149-193).
M. PORCIUS CATO 7
diction that have been observed so far (e.g. ut supra dixi, ut ad rem
redeam).
From the work Cato dedicated to his son Marcus comes the
famous sentence: rem tene, verba sequentur. It stands in opposition to
contemporary Greek rhetoric, in which the doctrine of the choice of
words constituted a major part of theory and practice. 33 It is
significant that this view of Cato's is close to a passage in Alcidamas,
who was an aavocate of improvisation and extempore speaking and
therefore attached weight to ideas and arrangement and not to every
single word. 34 This contact can be seen as evidence of the oral traits in
Cato's style; there is no need to assume direct dependence. Even
where Cato composed in a stylized manner, his diction continued to
be determined by the spoken word. This is hardly surprising in the
first writer of Latin prose; but remarkably it does not seem to have
been thought through in all its implications. The distinction
attempted here between 'oral style' and 'artlessness' will also be of
importance for understanding the style of the Origines.
4. Accumulation of synonyms
Several times in this text two different expressions with related
meaning are used to denote the same thing; in these cases the second
phrase can define and expand the first: sic habuerunt et ita in legibus
posiverunt; bonum agricolam bonumque colonum; strenuum studi-
osumque rei quaerendae; periculosum et calamitosum; et virifortissimi
et milites strenuissimi.
Here we see a new side of Cato's style, which is usually given less
attention. The reader needs to appreciate not only Catonian brevitas,
but also Catonian ubertas.
Accumulation of synonyms results from a conscious stylistic
intent. It occurs among the most diverse peoples, especially in the
sphere of religion and in officiallanguage; 35 and Cato himself could
draw inspiration from the old-Italic prayers he knew well and liked to
quote, as well as from official texts. The use of this solemn stylistic
33
Aristotle rhet. 3,1 (1403 b 15-18) says the opposite ofCato (not only 'what', but
also 'how' is very important).
34
Alcidamas rr. crocp. 18. Leo (cited above p.2 n.9) p.280.
35
In official language synonyms serve to 'exhaust' the possibilities of interpretation
and application. H. Kronasser 'Nugae Catonianae' WSt 79 (1966) pp.298-304,
distinguishes between terse legal language (for experts) and the redundant official
language of the Curia (e.g. SC. de Bacch.), which was widely disseminated through
decrees of the Senate (jixum aes). For literature on synonym accumulation see below in
the notes to the speech for the Rhodians (p.l4).
THE BEGINNINGS OF LITERARY PROSE
device clearly indicates that the introduction to De agricultura is

deliberately stylized.
5. Overall structure
The overall structure of the introduction has been examined most
recently by A.D. Leeman (above p.3), who has shown that the train
of ideas results from the division of the standpoint utile into tutum
and honestum, and that this sequence of standpoints is arranged
chiastically. Whether Greek theory is involved here on the conceptual
level or not, this structure with its archaic ring composition is so well
thought out and convincing that it does credit to Cato's Italic sense of
form.
6. Summary
The introduction to Cato's manual gives the impression of being
heavily oriented towards its subject: rem tene, verba sequentur.
Particular stylistic features, like indifference to the repetition of
unobtrusive words, indicate the influence of oral language. However
it is very often solemn oral language, and not the everyday kind, that
is involved.
On the other hand the text shows clear signs of deliberate
stylization: two-element groups of synonyms, a striving after paral-
lelism, exploitation of the important positions (beginning and end) in
the sentence, and artistic overall structure.
The proem has its own distinctive traits, not only in the charac-
teristic combination of artlessness and art, of the 'oral manner' and
deliberate stylization, but above all in the distinctive rhythm, which
by reducing the size of the last important colon achieves the
characteristically Catonian effects of brevity and ruggedness.
M. PORCIUS CATO 9
II. Speech in the Senate for the Rhodians 36

(167 B.C.)
163. Scio so/ere plerisque hominibus rebus secundis atque prolixis atque
prosperis animum excel/ere, atque superbiam atque ferociam augescere
atque crescere. quo mihi nunc magnae curae est, quod haec res tam
secunde processit, ne quid in consulendo advorsi eveniat, quod nostras
secundas res confutet, neve haec laetitia nimis luxuriose eveniat.
advorsae res edomant et docent, quid opus siet facto, secundae res
laetitia transvorsum trudere solent a recte consulendo atque intelle-
gendo. quo maiore opere dico suadeoque, uti haec res aliquot dies
proferatur, dum ex tanto gaudio in potestatem nostram redeamus.
164. at que ego quidem arbitror Rodienses noluisse nos ita depugnare,
uti depugnatum est, neque regem Persen vinci. sed non Rodienses modo
id noluere, sed multos populos atque multas nationes idem noluisse
arbitror atque haut scio an partim eorum fuerint qui non fwstrae
contumeliae causa id noluerint evenire: sed enim id metuere, si nemo
esset homo quem vereremur, quidquid luberet faceremus, ne sub solo
imperio nostro in servitute nostra essent. libertatis suae causa in ea
sententia fuisse arbitror. atque Rodienses tamen Persen pub/ice
numquam adiuvere. cogitate, quanto nos inter nos privatim cautius
facimus, nam unusquisque nostrum, si quis advorsus rem suam quidfieri
arbitratur, summa vi contra nititur, ne advorsus eam fiat; quod illi
tamen perpessi.
165. ea nunc derepente tanta beneficia ultro citroque, tantam
amicitiam relinquemus? quod illos dicimus voluisse facere, id nos
priores facere occupabimus?
166. qui acerrime advorsus eos dicit, ita dicit: hostes voluisse fieri.
ecquis est tandem, qui vestrorum, quod ad sese attineat, aequum
censeat poenas dare ob eam rem, quod arguatur male fa cere voluisse?
nemo, opinor; nam ego, quod ad me attinet, nolim.
167. quid nunc? ecqua tandem lex est tam acerba, quae dicat: si quis
illud fa cere voluerit, mille minus dimidium familiae multa esto; si quis
plus quingenta iugera habere voluerit, tanta poena esto; si quis maiorem
pecuum numerum habere voluerit, tantum damnas esto? atque nos
36 Text in Malcovati Oratorum Romanorum Fragmenta (Turin 1976 4 ) fr. 163-171.
The last two fragments (170f.) are not literal quotations, and are therefore no use for
an analysis of style. The latest edition oft he text, M. Po rei Catonis orationum reliquiae
lntroduzione, testa critico e commento filologico a cura di M.T. Sblendorio Cugusi
(Paravia 1982), reads in 167 impune (instead of impoene) and in 168 Rodiensibus nunc
aberit. This edition contains a detailed commentary.
omnia plura habere volumus, et id nobis impoene est.

168. sed si honorem non aequum est haberi ob earn rem, quod bene
facere voluisse quis dicit, neque fecit tam en, Rodiensibus oberit, quod
non male fecerunt, sed quia voluisse dicuntur facere?
169. Rodiensis superbos esse aiunt id obiectantes quod mihi et liberis
meis minime dici velim. sint sane superbi. quid id ad nos attinet? idne
irascimini, si quis superbior est quam nos?
163. I know that most people in favourable and comfortable and happy
circumstances become over-confident and that arrogance and truculency
grow and flourish. 37 Therefore I am now very much concerned, since this
enterprise proved such a success, that in the deliberation nothing detrimental
should come out, which can destroy our good fortune, and that the present
joy should not run riot too much. Adverse circumstances render docile and
teach what needs to be done, favourable circumstances often through joy
push a person crossways away from right consideration and understanding.
All the more urgently do I advise and move 38 that this question be postponed
for a few days, until we get control of ourselves again after such rapture. [It is
possible that this first fragment was followed by the idea that it was useful for
Rome to maintain peace with Rhodes, especially as a war would be too
expensive and time-consuming.]
164. In my view at least the Rhodians did not want us to bring the war to an
end in the way that it was brought to an end, nor that King Perseus should be
defeated. Yet not only the Rhodians did not want that, but many peoples and
many races did not want the same thing either, I believe. Why, perhaps there
were many among them who did not want this outcome, but not so as to
cause u.s offence. Rather they were scared (if there were no one 39 we needed to
be afraid of, we should do as we wanted): if we alone ruled over them, they
would be our slaves. For their freedom's sake they took this view, I believe.
And even so the Rhodians have never supported Perseus officially. Consider:
how much more carefully do we deal with each other in private life! After all,
if each one of us thinks something is being done against his interests, he
resists with all his might,~to stop anything happening contrary to them. But
that is what they have put up with nonetheless. [Although it was not in their
interests, out of loyalty the Rhodians have done nothing against Rome.
Forty years ofuncloudedfriendship should not be wiped out now through a
crazy war:]
165. Such great mutual benefits, such close friendship - shall we now
suddenly leave them behind us? What we say the Rhodians wanted to do -
37
The following translations and paraphrases were compared: D. Kienast Joe. cit.
pp.ll9-122; 0. Ribbeck Neues Schweizer Museum 1 (1861) pp.15-17; F. Leo Joe. cit.
pp.475f.
38
A technical term, cf. ThU 3,4 (1909) 792-796 (Hoppe); so also Sen. apoco!. 11,4.
39
Cf. R. Hakamies 'Homo dans Ia langue de Caton' Neuphi!o!. Mitteilungen (Bull.
Soc. Neophilol.; Helsinki) 49 (1948) pp.194-196. 'Pleonastic' use in Cato: homini ieiuno
(agr. 157,8); generic use of the singular: ubi subtus homo ('one', French 'on') ambulare
possit (agr. 48,1-2).
M. PORCIUS CATO II
shall we be the first actually to do it? [The basic idea of the passage is
developed by Kienast in relation to the balance of power: Carthage really
was a dangerous enemy of Rome. And yet Scipio advocated its preservation,
to stop the Romans becoming too arrogant and to maintain an incentive for
their effectiveness. From the paraphrase in Appian Pun. 65 (Cato fragment
170 Malcovati 2), it is clear that Cato used this argument in the speech for the
Rhodians. Should one behave differently in the case of Rhodes, which can
now no longer be dangerous? Here fragment 166 follows:]
166. Whoever speaks most sharply against them, speaks thus: they wanted
to become enemies. Is there anyone among you 40 then, who for his part
would think it fair to be punished because he is accused of having wanted to
do wrong? No one, I think. Assuredly, I for my part would not like it.
167. Now, to continue. Is there a law then so hard that it would prescribe:
'If anyone has wanted to do this or that, the penalty is to be a thousand asses
less than half his property'? 'If anyone has wanted to possess more than 500
acres, the penalty is to be such and such an amount'? 'If anyone has wanted to
have more cattle, the fine is to be such and such'? Of course we want to have
more of everything, and we can do that without being punished.
168. But if it is not right to honour someone for saying he wanted to do
good, without actually doing it: are the Rhodians then to suffer because they
have not done wrong, but because it is alleged they wanted to do so?
169. People say the Rhodians are arrogant, and that is a criticism I should
certainly not want to have made against me and my children. Then let them
be arrogant. What business is that of ours? Or are you angry if someone is
more arrogant than we are?
1. The problem
The problem of Greek rhetorical influence on Cato's speeches has
been discussed since Gellius, about whom more will be said in the
following section. Norden and Leeman are inclined to posit such
influence, while Leo and others reject this view. 41 Questions about
the text's structure are less speculative. A substantial contribution to
the understanding of archaic structure in the speech for the Rhodians
has been made by H. Fankhanel. 42 In what follows, these two
problems will be examined afresh in relation to the text. At the same
time attention will be directed to Cato's special qualities.
2. Cato's special qualities and his method of argumentation
in the speech
All the fragments of the present speech that are preserved verbatim
come from Au! us Gellius, 43 combined with brief summaries of their
content. Gellius is arguing against Cicero's freedman, Tiro, who had
40
Vestrorum here = vestrum.
41
The evidence is presented in the final section of this chapter, below pp.l8ff.
42
Verb und Satz in der /ateinischen Prosa bis Sa/lust Neue deutsche Forschungen 8
43
(Berlin 1938). Gell. 6,3.
criticized Cato's speech. 44 Tiro's criticisms are worth mentioning,

because the contrast with Cicero's day can be used to highlight
Cato's special characteristics.
Tiro's first objection is directed against Cato's anxiety that in their
joy at the victory the senators could have lost their level-headedness,
and might no longer be capable of rational judgment. According to
Tiro, an advocate should start by trying to win the judges over with
words of respect. Insults and imperious threats are out of place here.
Gellius replies to this by pointing out that Cato is not defending the
Rhodians like an advocate in court, but is giving advice in his
capacity as senator, consular and former censor, on what he thinks
should be done for the benefit of the state. Auctoritas allows the ex-
censor to adopt a tone that speakers in Cicero's day would have
hesitated to use in front of the Senate.
If Tiro's first objection makes it possible to appreciate indirectly
the censor's stature, then his second criticism leads to a better
understanding of the way ideas are developed within the speech.
When Cato concedes that the Rhodians were opposed to Rome, this
in Tiro's view is not a defence but an admission of guilt. Gellius
advances two different arguments in response. Firstly, Cato is not
conceding that the Rhodians did not want a Roman victory, but he is
putting forward this idea only as his private opinion. 45 Secondly, and
most importantly, Cato has according to Gellius not only avoided
making a mistake here, but has actually performed a masterstroke.
For thanks to his candour he gains credit for himself with all parties,
and turns to the Rhodians' advantage a fact which appears to militate
against them: if the Rhodians did not support Perseus, although this
would have been in their interests, then for this proof of friendship
they deserve the sympathy of the Romans.
Tiro's second objection thus affords an insight into the speech's
invention, and also reveals two ofCato's main characteristics: on one
hand his down-to-earth frankness, 46 and at the same time his ability
to remain versatile and flexible in argument and to turn what seems
44 Gellius is largely dependent on Tiro for the quotations from Cato; however he
did have the text in front of him, as is shown e.g. by the additional material at 6,3,49f.:
Verba adeo ipsa ponemus Catonis, quoniam Tiro ea praetermisit (but perhaps this
quotation too comes only from another intermediate source?).
45 Gell. 6,3,24: quod erat procul dubio opinionis suae professio, non Rhodiensium
culpae confessio.
3 5
4 6 On this cf. F. Klingner Joe. cit. (1956 ) p.66; (1965 ) p.64: 'In this passage Cato is
uncompromisingly down-to-earth, and yet the psychological effect is exceptionally
sure'. (=Die Antike 10 [1934] p.262).
M. PORCIUS CATO 13
to be the weakest part of the case into a strong-point.

Thirdly Tiro takes exception to the following enthymeme: quod
illos dicimus voluisse facere, id nos priores facere occupabimus?
According to Tiro the only logical answer would be: occupabimus
certe; nam si non occupaverimus, opprimemur. Against this Gellius
objects that life is not like a gladiatorial contest. The gladiator has the
choice of either forestalling his opponent or being killed. In ordinary
life however there is no such inexorable pressure to initiate a wrong
for fear of having it done to oneself. He adds that this kind of attitude
also conflicts with the gentle nature of the Roman people. For the
senators liked to be reminded of their clemency, especially when a
relatively harmless enemy was involved like the Rhodians. (In the
case of Carthage we know47 of course that Cato himself supported
the opposite principle.)
Tiro's third objection too reveals unintentionally that Cato is not
concerned here with any principles or even with a continuous line of
argument, but that he has adapted his remarks with tactical skill to fit
the situation.
While Tiro's last attack 1 as on Cato's excessively moral stance (he
failed to notice that adopting it in this case was harmless and even
helpful), in what follows he takes the opposite line and complains
that Cato uses arguments that are not honourable enough and far too
bold. He maintains that such a tactic is inappropriate for a great man
like Cato, and is crafty, deceptive and sophistic. This relates to the
assertion that the Rhodians should not be attacked, because
malicious intent alone is not an adequate ground for punishment. As
Gellius himself admits, Cato has shifted the balance considerably
with the example from civil law: it is not a sin to desire more land or
more cattle. In this way the anti-Roman attitude of the Rhodians is
made to seem innocuous. At any rate Gellius shows that Cato has
prepared his contestable line of argument with skill, and hidden its
weak points.
So Tiro's final criticism also throws an indirect light on Cato's
special qualities: the main aim of the speech is to represent the case of
the non-hostile Rhodians as being just or at least pardonable. To this
aim everything else is subordinated, including unity of argument: at
one moment Cato declares that the Rhodians have neither waged war
against Rome nor wanted to do so; at another he says one should
47
On Cato's famous utterance see S. Chabert 'Le delenda Carthago' Annates de
l'Univ. de Grenoble 25 (!913) pp.49-61. Diod. 34, fr. 32,3; App. Lib. 69; Plut. Cato 27;
Veil. I, 13; Plin. nat. 15, 74; Tertull. ad nat. 2, 16; Flor. 2, 15,4; Ps. Aur. Viet. vir. ill. 47 ,8.
judge only deeds, and that the will alone cannot be punished; but at
another he calls for indulgence, as if he admitted that they were
guilty. 48 Here too we see that Cato's reasoning in the speech is not to
be understood in terms of strict logic, but that what is involved is a
kind of 'close combat' style, in which all means to achieving the
desired end are acceptable. 49
3. Accumulation of synonyms and alliteration
Having examined the sequence of ideas in the speech, we shall turn
next to a consideration of its style. The first fragment is marked by a
solemn redundance, such as was not to be found in the preface of De
agricultura. In that text relatively sparing use was made of the
accumulation of synonyms; here however it is used with archaic
unconcern to produce a violent fortissimo: rebus secundis atque
prolixis atque prosperis; animum excel/ere atque superbiam atque
ferociam augescere atque crescere; 50 a recte consulendo atque intelle-
gendo. This accumulation can only be observed at the beginning
(fragment 163). In fragment 164 by contrast we find only one phrase
of the kind (multos populos atque multas nationes). This stylistic
device is also less frequent in the subsequent fragments (165-169).
Accordingly the speech to the Rhodians too shows a difference of
style between the beginning and the other parts. The heavy archaic
ornamentation is reserved for the opening. 51
48
Gell. 6,3,47.
49
Cf. Gell. 6,3,52. See R. Marache La critique litteraire de langue latine et le
developpement du goiit archafsant au 1/e siecle de notre ere (Rennes 1952) p.286 'il fait
fleche de tout bois'.
50 On such accumulations of synonyms see Till pp.27f. (examples), 69-72 (with
particular reference to neologisms resulting from synonym accumulation). Early

examples in 0. Altenburg 'De sermone pedestri Italorum vetustissimo' Jahrbiicher fur
classische Philologie Supplementband 24 (Leipzig 1898) pp.485-534, esp. sect. 1: De
copia verborum, 487-495 (a collection of material that leaves the reader to draw the
finer distinctions). Till p.22 is right to stress that synonym accumulation is due less to
Greek influence (as Norden Kunstprosa I p.l67 and Hofmann Syntax p.825 believed)
than to old-Italic sacral prose; cf. E. Lindholm p.54 (a carefully balanced judgment);
esp. also p.59 (emphasis on sacral style as the prerequisite for Roman artistic prose).
Modern research on the importance of sacral language for synonym accumulation is
also taken into acount now by A. Szantyr Lateinische Syntax und Stilistik (a revision of
J.B. Hofmann's work) (Munich 1965) pp.786-790. On the style of the speeches
(compared with agr. and orig.) see M.T. Sblendorio 'Note sullo stile dell'oratoria
catoniana' Annali della Facolta di Lettere e Filosofia della Universita di Cagliari 34
(1971) pp.5-32.
51 As a parallel to these accumulations of synonyms A.D. Leeman p.46 adduces
Cato fragment 23 Malcovati 1 (= 21 Malcovati 4) egoque iam pridem cognovi atque

intellexi at que arbitror rem publicam curare industrie summum periculum esse. He looks
for the origin of this mode of expression in the language of administration, cf. SC de
M. PORCIUS CATO 15
At the beginning of the speech there is also a predominance of

phonetic devices. An example is alliteration, 52 which plays a less
prominent role in Greek literature. It is true that this does also occur
in the later parts. 5 3 Even more resonant than alliteration is homoio-
teleuton: augescere atque crescere; docent ... solent. More examples
could be cited.
What is the purpose behind this accumulation of stylistic devices in
the first part of the introduction? Here the aim is to stress the
exaltatio_n of the moment. The exaggerations are meant to be noticed
and to encourage reflection. At the end of the first section stands the
terse and dryly worded motion: a truly Catonian anticlimax. The
archaic solemnity of the beginning is, so to speak, an illusionistic
front for the down-to-earth conclusion.
4. Verbal repetition
There is a tendency to avoid recurrence of the same words in the first
fragment ( 163), in which synonyms are accumulated. Even so it is still
possible to observe in it the archaic device of intensification through
repetition of a word just used. However this procedure is restricted
here to a central idea, which thus acquires thematic importance: quod
haec res tam secunde processit ... quod nostras secundas res confutet ... ;
secundae res laetitia transvorsum trudere solent. This important word
alternates with the opposite idea: ne quid in consulen do advorsi eveniat
... ; advorsae res edomant et docent. Both themes are essential for an
understanding of the introduction; apparently it has not yet been
noticed that in their arrangement Cato has opted for a similar sort of
ring composition as in the preface to De agricultura. Thus an
examination of these verbal repetitions, with their decisive structural
role, makes it possible to understand the organization of the first
fragment. 54 Nor should the recurrent atque be forgotten; the word is
used to connect synonyms, and is characteristic of the elevated
style. 55 Later in the speech however atque serves chiefly to combine
Bacchanalibus 13: neve ... coniourase neve comvovise neque conspondise neve con-
promesise velet.
52 Prolixis atque prosperis; neve haec laetitia nimis luxuriose eveniat; edomant et
docent; transvorsum trudere.

53 Cf. fragments 164 to 169; esp. 168 honorem ... haberi; 169 sint sane superbi; idne
irascimini.
54 The recurrence of the word laetitia is offset by the synonym gaudio at the end,
and by the related but more negative concepts superbiam atque ferociam at the
beginning.
55 H.C. Elmer '-Que, et, atquein the Inscriptions of the Republic, in Terence, and in
Cato' AJPh 8 (1887) pp.292-328. Et predominates in De agricultura (Elmer pp.307f.).

On at que in Cato see 0. Schoendoerffer De genu ina Catonis de agricu!tura libriforma, I:
clauses, 56 while et stands between connected words: 169 mihi et liberis

me is.
In the other fragments verbal repetition proves to be one of the
chief ways of developing the thought: 165 quod illos dicimus voluisse
facere, id nos priores facere occupabimus? 168: sed si honorem non
aequum est haberi ob eam rem, quod bene facere voluisse quis dicit,
neque fecit tam en, Rodiensibus oberit, quod non male fecerunt, sed quia
voluisse dicuntur facere? 169: Rodiensis superbos esse aiunt id
obiectantes quod mihi et liberis meis minime dici velim. sint sane
superbi. quid id ad nos attinet? idne irascimini, si quis superbior est
quam nos?
In fragment 164 the recurrent noluisse is the vehicle of the principle
idea: atque ego quidem arbitror Rodienses noluisse nos ita depugnare,
uti depugnatum est, neque regem Persen vinci. sed non Rodienses modo
id noluere, sed multos populos atque multas nationes idem noluisse
arbitror atque haut scio an partim eorum fuerint qui non nostrae
contumeliae causa id noluerint evenire.
Besides developing the thought in a surprising new light, verbal
repetition from fragment 164 onwards also functions as a cross-
reference or serves to suggest a situation indirectly: 164 ita depugnare
uti depugnatum est; siquis advorsus rem suam quid fieri arbitrantur,
summa vi contra nititur, ne advorsus earn fiat. Here the end of the
clause is not particularly stressed; rather it is the word standing first
in the clause on each occasion (ita, summa vi) that remains to the fore.
Even a relatively neutral and colourless word can recur: qui
acerrime advorsus eos dicit, ita dicit; hastes voluissefieri. Through the
repetition of dicit attention is directed here to the adjacent words:
acerrime ... ita. Again initial position and emphasis are linked. 5 7 As
has already been established, the 'solemn oral style' of such
repetitions is not to be confused with colloquialism. 58
It is hard to see what use there would be in using specific rhetorical
terminology to label the verbal repetition observed here. As we have
De syntaxi Catonis (Diss. Konigsberg 1885) p.51. On the elevated tone of atque see E.
Liifstedt Syntactica II (Lund 1933) p.341. In the speeches ofCato at que is particularly
frequent; polite fun is made of this stylistic peculiarity by Fran to p.34 van den Hout.
Cf. Leo p.286, drawing attention to the importance of at que for the difference of genre.
56 An exception is multos populos at que multas nationes (beginning of fr. 164; an
important passage).
57
Cf. also the repeated facere in fr. 165; here the adjacent words voluisse and
occupabimus are emphasized by their initial and final position. In fr. 168 the infinitives
(facere) are placed unemphatically, whereas the indicatives (fecit, fecerunt) are
58
stressed. Cf. above pp. 6f.
M. PORCIUS CATO 17
seen, and as we shall see again with increasing clarity, the most
diverse forms of repetition subserve the structure of the text and
cannot be detached from it as extraneous ornaments. It is not
therefore necessary to call on Greek theory in order to explain the
style of the speech for the Rhodians. Rather the recurrences are
explicable with greater logic and consistency in terms of the
architectonics of the text itself. This confirms that we should exploit
the interpretative possibilities inherent in the text itself before
moving on to literary theory. Even if Cato was familiar with Greek
rhetorical textbooks, he already knew how to apply the figures of
speech on the grounds of his experience as a Roman orator, 59 long
before he could give them names.
5. Word order
a) Additive sequence as an archaic feature of style
In the very first sentence the main verb is not put behind its
dependent infinitives but at the beginning: Scio so/ere plerisque
hominibus ... animum excel!ere ... Similarly later on: at que ego quidem
arbitror Rodienses noluisse nos ita depugnare ... In this way long
sentences can be formed, which do not become obscure. But above
(~~in the first sentence also has a certain emphasis. 60
b) The special role of initial position
Alongside the normal word order quod arguatur male facere voluisse,
we find: sed quia voluisse dicuntur facere, where stress is laid on
voluisse. 61 Cf. sentence 1 rebus secundis ... ; later however in antithesis
(and therefore emphatic) at the beginning of the clause: advorsae res
... secundae res ... This strong thrust at the beginning ofthe sentence
in fact seems typical ofCato: libertatis suae causa in ea sententiafuisse
arbitror; qui acerrime advorsus eos dicit, ita dicit.
c) Separation
Words belonging together are separated and thereby acquire em-
phasis: hostes voluisse fieri. Likewise: nunc derepente ... relinquemus
and in the same sentence ea ... tanta beneficia. Here it is not artistic
59
This alone can be the historical core of Plutarch's rather 'pedagogic' sounding
remark that Cato had early become accustomed to train his speaking ability 'like a
second body and an almost equally indispensable tool for the active, aspiring man'
(Plut. Cato 2).
60
This is confirmed by the numerous imitations in Sallust (so Fankhi.inel p.213); cf.
epist. ad Caes. 2 init., Leeman p.165. Elsewhere Cato puts verbs of emotion (mirari
twice) at the beginning (oral. 98 M 4 , 22 M 4 ). More often verbs of movement go to the
front (in Homer on the other hand verbs of perception): Fankhi.inel pp.217ff.
61
On this see Fankhi.inel p.l81.
separation that is involved but the 'natural' kind, such as is admitted

by every language with free word order. It originates in the instinctive
drive to fill the first and last position in the sentence (or in the colon)
with important words.
Chiastic positioning can accentuate the antithesis: 165 voluisse
facere ... facere occupabimus? Even more effective is the chiasmus in
fragment 169: Rodiensis superbos ... superbior quam nos. On
fragment 168 see the next section.
d) Ring composition
In fragment 168 an example of ring composition is created through
the alternation fa cere- fecit,fecerunt- facere. In fact in the speech
for the Rhodians the initial and final position of conjunctions and
verbs has the function of articulating the structure:
atque ego quidem arbitror Rodienses noluisse nos ita depugnare, uti
depugnatum est, neque regem Persen vinci,
sed non Rodienses modo id noluere, sed multos populos at que multas
nationes idem noluisse arbitror.
atque haut scio an partim eorumfuerint, qui non nostrae contumeliae causa
id noluerint evenire.
sed enim id metuere, si nemo esset homo, quem vereremur, quidquid
luberet faceremus, ne sub solo imperio nostro in servitute nostra
essent, Iibert at is suae causa in ea sent entia fuisse arbitror.
atque Rodienses tamen -.
The section 62 consists of two parts, each of which begins with
atque; both are subdivided within themselves by sed. The verb
arbitror appears at the beginning of the first segment and at the end of
both the first and the second. In addition noluisse stands in the first
sentence in second and penultimate position; this is the case in the
second sentence with fuerint and fuisse. Thus on each occasion the
verbal echoes close the ring: this is characteristic of early linguistic
practice. In Cato's sentence structure we find that archaic ring
composition is carried through with Italic clarity.
6. Conclusion
Can we therefore agree with Gellius when he says that in the speech
for the Rhodians Cato has used all the devices of rhetoric? According
to Leo 63 Gellius' account supplies indirect evidence 'that the speech
did not have an arrangement according to the demands of Greek
rhetoric with a logical presentation of the issues of the case'. 64
Whether the actual details of composition were determined by Greek
62 Cf. Fankhiinel pp.180f. 63
On Cato's speeches Joe. cit. pp.283-290.
64
Leo p.286.
M. PORCIUS CATO 19
stylistic theory remains for Leo 65 very doubtful. Thus he has not been
convinced by the Greek elements in Cato adduced by Nor den. 66 Even
today H. Kronasser 67 denies Greek influence on Cato's language;
and it is only with some scepticism that R. Till presents his own
observations of Greek elements present in Cato. 68 Till is right to
follow Leo in raising the question whether later rhetoric did not try to
force the early art of Cato into the pattern familiar to itself. 69
A.D. Leeman examines the beginning of the speech for the
Rhodians from the point of view of Greek rhetoric. 70 He is prepared
(perhaps excessively so) to accept Gellius' rhetorical categories of
thought. He agrees with the classification of the speech as political
suasio in terms of the genus deliberativum. In the exordium he finds
the aspects of utile, tutum and honestum combined. This allows him
to reject Tiro's criticism: Tiro sees the matter from the point of view
of tutum, but Cato sees it from that of honestum. This is certainly a
correct interpretation; but did a practitioner need to learn from a
textbook of rhetoric the rather obvious and commonplace conside-
ration of whether a thing is harmless or useful or proper?
Like Gellius, Leeman speaks of 'enthymemata'. 71 In fr. 167 he
recognizes a complexio 72 and in fr. 169 paradoxon combined with
traductio. In this connection our own interpretation has suggested
greater caution. The contextual function of verbal repetitions shows
that it is not a question of extraneous ornament. What then is the
point of applying labels? Leeman himself, for all his use of rhetorical
terminology, is still cautious in approaching the question of the
influence on Cato of Greek rhetorical technique. He is however
prepared to include Greek rhetorical works among the books at
which Cato 'had a look'; 73 here one has to agree. Yet over and above
this he also finds Greek political theory in the combination of
£u5aq.wvia, u~pt~ and -rpuqn]. 74 Kienast had already drawn attention
to points of contact with Polybius regarding the theory of the state. 75
As personal acquaintance is also involved here, this view is open to
few objections; however the thought is so obvious that in this case
there is really no need for a Greek source.
65 66
Ib. p.286. Loc. cit. pp.l65-169.
67
WSt 79 (1966) pp.298-304. However Kronasser only takes into consideration De
agricultura.
68 69
Loc. cit. pp.21-25. Till p.22; cf. already Leo p.286.
70 71
Loc. cit. pp.45-49. Fr. 168 Malcovati\ Leeman p.46.
72
Repetition of the first and last word of the sentence.
73 74
Cf. Cato's remark p.77,1-6 Jordan. Polybius 6,57,5f.
75
Loc. cit. pp.l!Off.
Perhaps the question of Greek influences on Cato has been

sometimes overrated. At all events, the present interpretation
enables us to reach a more decisive conclusion than before: when
Gellius asserts that Cato wields all the weapons of rhetoric, 76 this
means no more (to use an apt phrase of Leo) than when the same is
said of Homer. 77
As far as the style of the speech is concerned, we have recognized
its archaic character in the 'additive' manner of connecting clauses,
in the ring composition, and in the accumulations of synonyms in
particular sections, especially at the beginning, where Cato uses
powerful linguistic means to describe the overflowing joy which
impedes proper reflection. We shall encounter something similar in
the Origines, where in a majestic passage Cato contrasts the hero cult
of the Greeks with the simplicity of the Romans. In both cases the
rich ornament is functional, and is applied with a certain tacit irony.
Cato uses exaggerated stylization to reduce the joyful mood to
absurdity; he then follows it with a down-to-earth statement.
This brings us to something specific to Cato. The pattern observed
in the preface to De agricultura, which in opposition to 'Behaghel's
law' has a lengthier clause precede a short one, is also present in the
speech for the Rhodians. 78 Cato's quick-wittedness/ 9 which was
(doubtless even more effective in the actual Senate debate, is thus also
'\ perceptible in the peculiar phrasing of the finished speech: first a
\. detailed, deliberately roundabout preparation, then a sure, lightning-
! fast blow.
77
76Gel!. 6,3,52. Leo p.286.
78Note the terse, matter-of-fact end clauses: 163 quo maiore opere dico suadeoque
.... 164/ibertatis suae causa in ea sententiafuisse arbitror .... quod i/li tamen perpessi.
Cf. also the ends of 166, 167, 168, 169. This aspect seems more important to the present
writer than the controversial question of prose rhythm: its presence in Cato is both
asserted (E. Fraenkel Leseproben aus Reden Ciceros und Catos = Sussidi eruditi 22
[Rome 1968] pp.l25-128), and denied (A. Primmer 'Der Prosarhythmus in Catos
Reden' in: D. Ableitinger and H. Gugel [edd.] Festschrift fur Karl Vretska zum 70.
Geburtstag [Heidelberg 1970] pp.174-!80). Fraenkel is right to find colon formation;
the use of periods is not fully developed. The present writer accepts Primmer's
reservations about the extent to which Cato's practice is deliberate.
79 Urbanus homo erit cuius multa bene dicta responsaque erunt (Domitius Marsus
following Cato, in Quint. inst. 6,3, 105; cf. E. de Saint-Denis 'Evolution semantique de
urbanus-urbanitas' Latomus 3 [1939] pp.5-24, esp. 10-15).
M. PORCIUS CATO 21
III. A Roman Leonidas 80

Pulcrum, dii bani, facinus Graecarumque facundiarum magniloquentia
condignum M. Cato libris originum de Q. Caedicio tribuna militum
scriptum reliquit.
id profecto est ad hanc ferme sententiam:
imperator Poenus in terra Sicilia bello Carthaginiensi prima obviam
Romano exercitu progreditur, colles locosque idoneos prior occupat.
milites Romani, uti res nata est, in locum insinuant fraudi et perniciei
obnoxium. tribunus ad consulem venit, ostendit exitium de loci
importunitate et hostium circumstantia maturum. 'censeo' inquit 'si
rem servare vis, faciundum, ut quadringentos aliquos mi!ites ad
verrucam illam' -sic enim Cato locum editum asperumque appellat-
'ire iubeas, eamque uti occupent, imperes horterisque; hastes profecto
ubi id viderint, fortissimus quisque et promptissimus ad occursandum
pugnandumque in eos praevertentur unoque illo negotio sese a!ligabunt,
atque illi omnes quadringenti procul dubio obtruncabuntur. tunc interea
occupatis in ea caede hostibus tempus exercitus ex hoc loco educendi
habebis. alia nisi haec salutis via nulla est.' consul tribuna respondit
consultum istuc quidem atque 81 providens sibi viderier; 'sed istos' in quit
'milites quadringentos ad eum locum in hostium cuneos quisnam erit,
qui ducat?' 'si alium' inquit tribunus 'neminem reperis, me licet ad hoc
periculum utare; ego hanc tibi et rei publicae animam do.' consul
tribuna gratias laudesque agit. tribunus et quadringenti ad moriendum
proficiscuntur. hastes eorum audaciam demirantur; quorsum ire per-
gant, in expectando sunt. sed ubi apparuit ad eam verrucam occu-
pandam iter intendere, mittit adversum il!os imperator Carthaginiensis
peditatum equitatumque, quos in exercitu viros habuit strenuissimos.
Romani mi!ites circumveniuntur, circumventi repugnant; fit proelium
diu anceps. tandem superat multitudo. quadringenti omnes cum uno
perfossi gladiis aut missilibus operti cadunt. consul interibi, dum ea
pugna pugnatur, 82 se in locos tutos atque editos subducit.
sed quod illi tribuna, duci militum quadringentorum, divinitus in eo
proe!io usu venit, non iam nostris, sed ipsius Catonis verbis subiecimus:
dii immortales tribuna militum fortunam ex virtute eius dedere. nam ita
evenit: cum saucius multifariam ibi factus esset, tamen volnus capiti
nullum evenit, eumque inter mortuos defetigatum volneribus atque,
quod sanguen eius defluxerat, cognovere. eum sustulere, isque con-
valuit, saepeque postil!a operam rei publicae fortem atque strenuam
80 Text of Cato in Gellius 3,7 (= HRR I' [1914] fr. 83; p.78 Peter).
82
81 The text follows Bergk here (see Peter's apparatus). Bergk's text.
perhibuit 83 il!oque facto, quod illos milites subduxit, exercitum ceterum

servavit. sed idem benefactum quo in loco ponas, nimium interest.
Leoni des La co, qui simile apud Thermopylas fecit, propter eius virtutes
omnis Graecia gloriam atque gratiam praecipuam claritudinis incli-
tissimae decoravere monumentis; signis, statuis, elogiis, historiis
aliisque rebus gratissimum id eius factum habuere; at tribuna militum
parva laus pro factis relicta qui idem fecerat at que rem servaverat.
hanc Q. Caedici tribuni virtutem M. Cato tali suo testimonio
decoravit. Claudius autem Quadrigarius annali tertia non Caedicio
nomen fuisse ait, sed Laberio.
The Punic commander advances in the country of Sicily in the first
Cathaginian war to meet the Roman army; he is the first to occupy the hills
and the favourable positions. The Roman soldiers, as is natural in the
situation, enter the area, which was open to trickery and destruction. A
tribune comes to the consul and explains to him that annihilation is
imminent, as the place is unsuitable and the enemy are on all sides. 'In my
view', he said, 'if you want to save the situation, then you should order about
four hundred soldiers to that wart'- for that is what Cato calls a position that
is elevated and hard of access - 'and issue a command calling on them to
occupy it; as soon as the enemy see this, all the bravest and keenest fighters
will be diverted to them and commit themselves to this task alone. Those four
hundred will doubtless all be cut down. In the meantime, as long as the
enemy are involved in this fight to the death, you will have time to lead the
army out of its present position. There is no other way to safety.' The consul
replied to the tribune that this seemed to him well thought-out and far-
sighted; 'but', he said, 'who will be prepared to lead the four hundred soldiers
there right into the enemy formations?' 'If, answered the tribune, 'you find
r:to one else, you may use me for the test; I give my life to you and the state.'
The consul thanks the tribune and praises him. The tribune and the four
hundred men set off to die. The enemy are very surprised at their boldness
and wait to see where they will go. But as soon as it became clear that the aim
of their march was to occupy the 'wart', the Carthaginian commander sends
infantry and cavalry against them, the bravest men he had in his army. The
Roman soldiers are surrounded, whereupon they resist; a battle develops,
which long remains undecided. Finally superior numbers prevail. Pierced
with swords or covered with missiles, the four hundred all fall with the one.
In the meantime, while this battle is being fought, the consul withdraws
secretly with his men into safe and high-lying positions.
But what happened in the battle by divine dispensation to that tribune, the
leader of the four hundred soldiers, is now told not in our own words, but in
Cato's own:, ,-The immortal gods granted the military tribune a piece ofluck
in keeping with his bravery. For the story ended as follows. Although he was
wounded there many times, nonetheless his head remained unscathed and he
was recognized among the dead, exhausted by wounds and loss of blood. He
- 83 -Quicherat's praehibuit is preferable.

M. PORCIUS CATO 23
was picked up, and he recovered again and afterwards served the state
frequently through brave and bold deeds. By diverting those soldiers he
saved the rest of the army. But in the evaluation of the same good deed there
is a very big difference, depending on the place. 84 The Spartan Leonidas
accomplished a similar thing at Thermopylae - and because of his
achievements all Greece has bestowed on him gratitude and honours in
abundance and exalted him with memorials of his glorious deed; with
pictures, statues, honorific inscriptions, historical accounts and in other
ways they gave expression to their profound gratitude for this deed of his; but
for the military tribune there remained only small praise for his deeds,
although he had after all done the same thing and rescued the situation for
Rome.
1. Introduction
Norden 85 calls the style of Cato's historical work 'curt, rough,
powerful' and 'much more developed than that of the manual'.
According to Leo, 86 in the Origines we should not expect very
profound Greek influence as to style, apart from the fact that the
subject matter had already been given a Greek form and that writing
a book was in itself something Greek. His observation that the style
of the Origines is more elevated than that of De agricultura is
important. 87 Starting from literary theory, A.D. Leeman 88 is much
more reserved than Leo in his judgment on the historical and literary
value of the Origines, 89 and he puts the Greek influence higher. It was
Cato's principle 'to look at Greek books, but not to study them
thoroughly' (inspicere, non perdiscere). He borrowed from them in a
spirit of 'dissimulation and rivalry', which is also typical of the
Scipionic circle (p.70).
Leeman's stylistic judgment on the Origines is challenging: 'This is
an interesting sample of Cato's historical style or rather of his lack of
an historical style' (p. 71 ). The basic characteristic of the text under
consideration is in his eyes 'an unpretentious colloquialism' (p.71).
Since we cannot agree with Leeman on this point, we must go into
rather more detail here. He criticizes endless repetitions of the cases
84
Cf. 0. Ribbeck Neues Schweizer Museum I (1861) p.27: 'But the same
achievement, set in this place or that, how different it looks!'
85
Loc. cit. p.165. There is a new overall appraisal of the Origines in W. Kierdorf
'Catos "Origines" und die Anfiinge der romischen Geschichtsschreibung' Chiron !0
(1980) pp.205-224. 86 Loc. cit. pp.290-300.
87 88
On the style see esp. p.299. Loc. cit. pp.67-88.
89
Leeman Joe. cit. p.68: 'It is however not surprising, that a Roman amateur
historian of Cato's type generally showed his abilities better in his oratorical
achievements, which belonged to his negotia, than in his efforts in the difficult art of
historiography.'
of is and of -que, which co-ordinate all kinds of secondary ideas. The

former phenomenon is well known from Plautus, the latter we have
already encountered in De agricultura (praef). There are also loose
relative connections like quod ... dejluxerat and qui ... fecit. Even so
Leeman notices the sudden rise in stylistic level in the section about
Leonidas, with paronomasia (gloriam - gratiam) and 'epic' words
(claritudinis inclitissimae; however it is hard to see how the word
claritudo is supposed to fit epic metre). He notes further the use of the
ending -ere, which belongs to a higher stylistic level (cf. Sallust). As
regards content, Leeman finds here a reflection of the Greek theory
of the relationship between virtus (cipEcrJ) and fortuna (1'UXTJ). Thus
for him the comparison between the unknown Roman and Leonidas
presupposes a belief in the usefulness and necessity of a nationalistic
Roman historiography. 90
Leeman points to a common factor between Cato and Polybius.
Polybius, who had moderate stylistic pretensions, seems in his
pragmatic historiography to champion the Catonian principle rem
tene, verba sequentur against the literary and stylistic ambitions of
Hellenistic historians. In this connection Polybius gives the im-
pression of being very Roman. However Leeman's statement that
Cato does not set any kind of stylistic standard for Roman
historiography is misleading. 91 He even concludes from the fragments
that in the Origines Cato aimed at much less careful stylization than
in the speeches. Nonetheless he notes 'a savour of honesty, directness
I and severity' (p.71). But then how is one to account for the
l enthusiastic remarks of Cicero and for Sallust's 92 imitation of Cato?
In the last analysis, Leeman's treatment of the style of the Origines
represents a step backwards as against Leo. By denying the existence
of a historical style in Cato, Leeman comes into conflict with the
facts. It is sufficient to recall that a tendency towards archaism is
apparent in the very first sentence of the Origines: Cato uses the
pronominal form ques, which by his day had already died out in
everyday speech and in less elevated literary genres. Similarly the
90 Sail. Catil. 8,2-3: Atheniensium res gestae, sicuti ego aestumo, satis amplae
magnificaeque fuere, verum aliquanto minores !amen quam Jam a feruntur. sed quia
provenere ibi scriptorum magna ingenia, per terrarum orbem A theniensium facta pro
maxumis celebrantur.
91 It is true there is no sign that Cato had any far-reaching influence on the
republican historians (W.O. Lebek Verba Prisca Hypomnemata 25 [Gottingen 1970]

pp.210ff.). We find deliberate archaizing in Sallust: Cato influenced his style.
92 Cicero Brut. 85,294; a more qualified estimate de or at. 2,12,53 and leg. 1,6. Sallust
calls Cato the most eloquent of the Romans (hist. I, fr. 4, p.4 Maurenbrecher).
M. PORCIUS CATO 25
widespread use in the Origines of the conjunction atque, 93 which in

De agricultura is less frequent compared with et, indicates a striving
after 'elevated' style. Nor can it be denied that in particular passages
there is an accumulation of archaic and poetic words. In the
paratactic sentence structure too it is possible to see more than just a
!colloquial feature. How could Cato, the very first Latin prose writer
! worth mentioning, be expected to have written throughout in
periods? For this form of prose expression the language had first to
'-be made flexible, and in this direction Cato did a great deal.
Furthermore a careful distinction must, as we have said, be drawn
between 'colloquial' and 'oral' elements of style. For example, the
language and sentence structure of early Greek epic show charac-
teristics that are in no way colloquial, but in many respects 'oral'. The
same is true of the language of Roman law and also ofCato's literary
language, to whose Roman roots we have drawn attention several
times already. It is no surprise if his sentence structure is in many of
its aspects still reminiscent of its origin in the oral language. The
details of Leeman's case too are insecure and open to question. 94 The
only evidence he mentions for the assignment of particular passages
of the Caedicius story to a lower stylistic level is the term verruca
('hummock'). But Quintilian's 95 opinion that this expression was too
humble is not of course necessarily valid for the time of Cato. If one
looks for evidence from Early Latin, one comes upon a verse cited by
Quintilian. Leeman does not tell us that this verse comes from a
tragedy, which is decisive for assessing the stylistic level of verruca in
the Early Latin period. On closer inspection therefore the very
passage of Quintilian adduced by Leeman is evidence not for, but
quite decisively against his view.
In Leeman's opinion, Cato achieved a more careful stylization
in the speeches than in the Origines. This is inherently improbable,
for prior to including his speeches in the Origines Cato surely
revised them for this purpose. The new setting was therefore
93 Leo p.299. On the archaisms in Cato see now the balanced treatment of G. Prugni
'Per un riesame degli arcaismi Catoniani (rileggendo il Till)' Quaderni dell'Istituto di

Filologia Latina dell' Universitii di Padova 2 (1972) pp.25-36.
94 By way of comparison Leeman (p. 70) adduces Cato agr. 1,3 naves ambulant. This
is probably not justified, because it ignores the basic meaning of ambulare. Here
ancient literary criticism, which is after all relatively late, should not be allowed to
distort our view.
95 Quint. 8,3,48 cites verruca in a verse of tragedy as an example of humilitas
(naturally a vitium). A further allusion to the same line of poetry occurs at 8,6, 14.
There is no mention of Cato in either passage.
more ambitious. 96
2. Form and content
Luckily we know one story from Cato's Origines very accurately.
Gellius has preserved it for us partly verbatim and partly in a reliable
paraphrase: 97 Pisani 98 for example treats the paraphrase as if it were
as good as identical with the text of Cato itself. We make some small
qualifications here, but the vocabulary is so 'Catonian' that the
stylistic fidelity of the retelling can hardly be doubted. Above all
Gellius has retained the order of the facts and the outline of the story;
the Catonian narrative structure emerges from his report in all
clarity. 99
In particular the term verruca ('wart') for a hillock is striking; 100
Cato certainly does not avoid drastic expressions, as long as they are
apt. Typical for Cato grammatically and as an item of vocabulary, is
the formation strenuissimos, 101 which we have already encountered in
De agricultura. Among stylistic devices, one should stress particularly
the accumulation of synonyms or related ideas: fraudi et perniciei
obnoxium; de loci importunitate et hostium circumstantia; imperes
horterisque; fortissimus quisque et promptissimus ad occursandum
pugnandumque; gratias laudesque agit; in locos tutos atque editos;
fortem atque strenuam; gloriam atque gratiam; signis statuis elogiis
historiis. 102 Besides cases of hendiadys, we have here the 'exhaustive
duplications' 103 of the old official and sacral language. They give to
the diction a certain formality and detached solemnity. There are
further instances where pteonastt(>,abundance contributes to the
96
Hyperbaton is also treated differently in the stricter style of the Origines than in
De agr. According to Fankhanel pp.184f., the thematic separation of sentence parts
that is characteristic of oral discourse has been preserved in De agr., while in the
Origines (apart from the speeches) hyper baton seems to be restricted to the accusative
of pronouns.
97
Cato orig. 83 (HRR I' [1914] pp.78-81 P. =Gel!. 3,7}.
98
V. Pisani Storia della lingua latina I ( = Manuale storico della lingua latina I) (Turin
1962) pp.223f.
99
The discussion will confirm that he also preserves many details from Cato.
100
See above p.25.
101 On such degrees of comparison in Cato see· Till pp. 94f.
102
Perhaps we should add to these the passage that Bergk corrects thus: consul tum
istuc quidem at que providens (see the apparatus in Peter HRR I p. 79).
103
This is a useful term coined by H. Haffter Untersuchungen zur altlateinischen
Dichtersprache Problemata Heft lO (Berlin 1934) chapter 3 'exhaustive expressions',
esp. p. 77. On the origin of this device in Roman legal and sacral language, see E.
Lindholm (cited n.3l) pp.27-42; on Cato pp.46f.; on sacral language pp.57 to 59; cf.
also Haffter pp. 8lf.
M. PORCIUS CATO 27
dignity of the account: this is the case with terra Sicilia; an archaic
fullness appears as well in interibi, dum ea pugna pugnatur. In this last
passage there is also figura etymologica. 104 Related to it is the device
of repeating a verb by means of its participle: Romani milites
circumveniuntur, circumventi repugnant. 105 In this way it is made clear
how one event develops from another. The use of the stylistic device
is therefore strictly functional.
The state of anticipation is made plain by the gerundial con-
struction in~gxp_ec;ta!J:dO sunt, which recalls the English 'progressive'
form. 106 Hyperbaton 'is not very frequent, but all the more expressive:
alia nisi haec salutis via nulla est; quisnam erit, qui ducat. In both cases
the emphatic words are moved to the beginning and end of the clause
and so determine its structure. Closer attention should be given to the
axial symmetry of the sentence ego hanc tibi et rei publicae animam do.
Here the word order shows a tendency to put the pronouns together
which can already be observed elsewhere in Cato. 107 Furthermore the
vivid and gesture-like designation of the first person by means of the
pronoun hie in this passage should be regarded as a primitive trait,
104 Cf. bonas preees preeorin Cato agr. 134,2(twice), cf. 139; 134,3 bonas preees bene
preeatus sum.
105 This stylistic device (traduetio, ploke) is frequent later in poetry, e.g. Ov. met.
1,33 etc. The language of the historians retains it: cf. Sisenna fr. 27 Romanos ...
prate/ant - protelatos perseeuntur.
106 Could the locution have escaped the Italian commentator Pisani, because it can
be turned effortlessly into Italian? This use of the gerund, which in Early Latin is in fact
commoner than the gerundive, can also be evidenced from Plautus and Terence. Cf.
C.E. Bennett Syntax of Early Latin I (Boston 1910) p.452 (mere. 218 quid hie in
lamentando pereo; Hee. 38 in experiundo ut essem). The ablative of the gerund with a
preposition is somewhat rare in classical authors, though it does occur frequently in
Sallust; cf. L. Constans De sermone Sallustiano (Paris 1880) p.227; e.g. lug. 6, I pleraque
tempora in venando agere. So in this respect too Sallust is following Early Latin. Cf.
also Hofmann-Szantyr p.380. The collection of examples in Kuhner-Stegmann II, I
pp. 753f. shows the limits of classical usage. P. Aalto Untersuchungen fiber das
lateinisehe Gerundium und Gerundivum Annales Academiae Scientiarum Fennicae: s.
B, vol. 62,3 (Helsinki 1949) has been so far unavailable to the present writer. J. Cortese
Grammatica Catoniana (Savona 1882) is no use; cf. also Vincenz Bulhart Thl..L 7, I
(1934-1964) 784,76-785,16 (likewise without separation of gerund and gerundive). A.
Draeger Historische Syntax der lateinischen Sprache II (Leipzig 1878) p.816: 'In
classical and later authors ad is quite normal, as is in with the ablative of the gerundive,
though it is rare with the ablative of the gerund'. There follow examples from Cicero.
lb. p.817: 'No doubt Gel!. 3,7,12 is archaic: in expectando sunt. Similarly with the
gerundive 3, 1,4: iamdiu in eo ipso quaerendo Jui. For the former is quoted from Cato
orig. 4, 7: in expeetando sunt.' Thus Draeger makes two passages out of one. Draeger is
clearly unfamiliar with the passages from the comic poets cited in Bennett.
107 A.W. Ahlberg 'De traiectionis figura ab antiquissimis prosae scriptoribus
Latinis adhibita' Eranos II (1911) pp.89f., esp. 90, n.1; Fankhanel p.l85.
which lends forcefulness to the account. 108

The evaluative final section, which Gellius quotes verbatim from
Cato, deserves a more precise stylistic examination. In contrast to
Pisani, who breaks off his linguistic interpretation at the very point
where Gellius' paraphrase passes into the original words ofCato, the
present interpretation will give due attention to this part as well.
Having thus far stressed particular linguistic and stylistic traits,
insofar as they can still be identified in Gellius, we now turn to the
structure of the report as a whole.
In its structure this account is both simple and convincing. The eye
is drawn first to the periphery: the Carthaginians occupy the hill. In
this way the limits of the scene of action are marked out. Then our
attention turns to the centre: the Romans march in. Thus the field of
vision narrows from the heights to the valley floor. Finally attention
is concentrated on an even smaller circle: the tribune and the consul.
Cato's literary instinct reveals itself in this functional progression
from the exterior to the interior. The movements of the two armies
have ceased. The danger of the situation is characterized by means of
an attributive phrase placed emphatically at the end: (in locum
insinuant) fraudi et perniciei obnoxium. This sonorous conclusion
provides a syntactic pause, and as regards content is a precondition
for what happens next.
The first person to emerge as an active participant 109 is the Roman
tribune. He carries the main action: tribunus ad consulem venit. His
name appears nowhere in our text. 110 We know from other sources
that throughout his entire historical work Cato does not refer to the
Romans by name, but only by their official title. This is a clear
indication that the glory of an individual is not meant to belong to
himself, nor just to his gens, 111 but to the whole res publica. Because
108
On the stylistic value of this usage in Greek tragedy cf. W. Schadewaldt
Sophokles, Konig Oedipus (Berlin/Frankfurt 1955) p.93: 'Throughout (we) have kept
the naive and gesture-like "this man" in place of our colourless, personal "I"; thus we
have left untouched much that strikes us as strange and uncivilized in our alienation
from immediate and primitive experience'. On the Latin cf. Hofmann-Szantyr p.l80.
109
The imperator Poenus remains explicitly in the background (col/is locosque
idoneos ... occupat).
110
How does Gellius know the name? Perhaps in this special case Cato had
abandoned his practice after all. On the uncertainty as to the name cf. Frontin. I ,5, 15
(4,5,10). Important for the whole subject isS. Lundstrom Vermeintliche Glosseme in
den Tusculanen (Uppsala 1964) pp.323-328: Cic. Tusc. I, 10 I belongs in this Catonian
context.
111
Clan-consciousness is deeply rooted in Rome; as a homo novus Cato certainly
also experienced the negative sides of this attitude. Later on the very different idea of
commitment to the state by an individual was widespread: it perhaps owes more to
M. PORCIUS CATO 29
the name is not mentioned, the tribune becomes a representative of

Roman behaviour. The action flows entirely from the tribune's
decision to go to his superior and lay before him a plan for saving the
Romans. Willingness to sacrifice oneself is a basic characteristic of
Roman religion (cf. the practice of devotio). The word for word
citation of the tribune's speech reveals a resourceful and determined
mind. The verbal forms are symptomatic: future, gerundive, sub-
junctive. 112 Attention is no longer directed to the two speakers, but to
the ideas in the speech. This represents a further step in the
movement from exterior to interior, which has been maintained from
the start.
The consul's reply is at first given indirectly (as with the beginning
of the tribune's speech); 113 not until the second part do we hear his
words directly. It is no accident that this is the sentence that produces
the tribune's heroic answer. The consul's gratitude and praise round
off this first section and anticipate the glorification of the hero at the
end of the whole account. Both at the beginning and end of the
section considered so far, there is a decision of the tribune. The
formal breaks are therefore identical with those in content. The
structure is logical and serves a purpose for all its simplicity.
The second part of the account, which Gellius has probably
shortened more drastically, begins with a laconic forecast that is
certainly Catonian: tribunus et quadringenti ad moriendum profi-
ciscuntur. The inevitable outcome is anticipated in the gerund. For
the first time since the start of the account, attention then switches
_ from the departure of the Romans to their opponents. We have a
description not only of the outward behaviour of the enemy, but also
of their feelings in view of the Romans' bravery. Cato is thus able to
create an effective background for the Romans' unusual enterprise
by including the opposite side and their reactions. The passage has
another function as well. This is to increase the tension, by moving
from initial surprise, reflected in an expressive composite verb
~demirantur), 114 to deliberate restraint, which is graphically expressed
Cato's influence than is commonly realized.

112
Faciundum; ut ... iubeas; (ut) imperes horterisque, uti occupent;praevertentur; sese
alligabunt; obtruncabuntur; habebis.
113
Ostendit exitium de loci importunitate et hostium circumstantia maturum.
114
The word is mainly early and colloquial. Expressive compounds are in fact
characteristic of the early period of Latin. W. Krause Deutsche Literaturzeitung 45
(1924) p.l594 finds that 'intensive de' is mostly combined with verbs that express
durative action; in these cases de indicates that the action is carried through from
beginning to end. Compounds with dein Cato are assembled by Till p.59.
in a delightful construction: in expectando sunt. Here therefore the art

of retardation can also be seen in stylistic terms. In a temporal
subordinate clause the enemy then realize what is happening, and in
the main clause their reaction follows. The best of them go to meet
the four hundred. The Romans are surrounded and resist desperately.
Again a sentence follows where everything is left open, so that the
effect is one of retardation: fit proelium diu anceps. Then comes the
statement that superior numbers prevail. With the death of the four
hundred the ring structure is completed. The announcement at the
beginning of the second part (ad moriendum proficiscuntur) has been
fulfilled.
We do not know how drastically Gellius has abbreviated and
reorganized here (are the historic presents his? 115 ). What is clear is
that he has preserved faithfully the alternation of quick progress
(with typically Catonia4 asyndeton) and emphatic retardation (its
genuineness is guaranteed by the archaic peculiarities of grammar).
A further artistic device that is both simple and effective is the way
the narrative standpoint changes and the Romans' bravery is
reflected in the reaction of the enemy observing it.
We are on even firmer ground from 19 onwards, where Gellius
quotes Cato's own words. In discussing this section it is possible to
achieve an even closer connection between consideration of details
and an examination of the overall form. The first sentence sum-
marizes like a heading the content of those that follow: di immortales
tribuna militum fortunam ex virtute eius dedere. The thematic
character of the opening sentence also attracted our notice at the
beginning of the first and second parts. This is now confirmation that
the paraphrase in the first two parts was reliable. The transition to
what follows is made with nam ita evenit. The account then moves
forward in main clauses that are mostly strung together without
connection and whose verbs are in the perfect (this makes Gellius'
elegant historic presents rather suspicious in retrospect). Subor-
dinate clauses appear only at the beginning and end of the section.
Both have an important function. Adversative subordination occurs
115 In the historical fragments there is a definite predominance of the historic
perfect. Fr. 87 is an example to the contrary; in 138 the context is uncertain. Given the
scantiness of the material, the question can hardly be answered. The report in fr. 29
Male.' is in the present; Till p.l7 shows that this is a special case in other respects as
well (poeticisms), but he ignores the striking historic presents. Since the historic
present is familiar to Plautus, it is permissible to assume that Cato also used it more
extensively than the fragments reveal. In fact Cato even employs the historic infinitive
(fr. 79). The vividness of Cato's style perhaps points to wider use.
M. PORCIUS CA TO 31
at the beginning, where in regard to content the surprising turn of

events and the contrast with what goes before is all-important. Cato's
use of constructional devices is lavish or sparing according to the
desired effect. This is what makes his style functional. Further
confirmation is provided by the second subordinate clause at the end:
quod illos milites subduxit, exercitum ceterum servavit. That this is
meant to be the end of the section is made clear in terms of content by
the retrospective viewpoint and in formal terms by the use of a
subordinate clause.
The narrative is over and Cato passes on to reflection. The
problem is once again summarized at the beginning in the form of a
heading: sed idem benefactum quo in loco ponas nimium interest. The
reflection consists of two parts, the second being contrasted with the
first by means of at; otherwise it is an independent main clause. The
first part refers to the Greek Leonidas, the second to the Roman
tribune. Both the contrasting parts are distinguished in the first place
by different length. Here Cato inverts the 'law of increasing
members' at a decisive point: he makes the weightier second member
considerably shorter than the first; attention was drawn earlier to this
peculiarity of Catonian style. But size is not the only difference. In
vocabulary and stylistic level the two opposed parts differ conside-
rably. In the first part the vocabulary is carefully chosen. Cla;it~do,
which is prominent thanks to its archaic formation, is given the
poetic adjective inclitus, which furthermore occurs in the superlative.
In addition there is the duplication gloriam atque gratiam (paro-
nomasia, in which alliteration is combined with homoioteleuton).
This accumulation of stylistic devices continues in the remarkable
succession signis statuis elogiis historiis, 116 involving four different
words with an increasing number of syllables. Finally we have one
further superlative (gratissimum) and a hyperbaton. 117 Such is the
wealth of artistic devices and the abundance of intensifying ex-
pressions118 that against this background the simplicity of language
in the following sentence is bound to come as a revelation: at tribuno
militum parva laus pro factis relicta, qui idem fecerat atque rem
servaverat. Again a relative clause summarizes retrospectively the
hero's achievement. Here the language has the grandiose simplicity
116
Cf. the enumeration of female attire and adornment at orig. 113.
117
Gloriam atque gratiam praecipuam claritudinis inclitissimae decoravere manu-
mentis (on this see Fankhanel p.l62).
118
Similar is orig. 63 in maximum decus atque in excelsissimam claritudinem
sublimavit (p. 73 Peter')
of an inscription.
It is particularly clear from the final part of Cato 's account, which
is preserved verbatim, that his treatment of language and style is in
the highest degree functional. Where a succession of events is
involved, there is no attempt to achieve variation or contrasts by
artificial means, as long as this is not required by the subject. But as
soon as it is a question of accentuating particular parts of a sentence
and highlighting relationships or contrasts, Cato has at his disposal
all the logical particles and all the means for connecting clauses
hypotactically. The colourfulness of archaic language and syntax
serves the purpose of retardation, where this is required by the
subject and where the way events are related needs to be made clear.
If it is a question of setting Roman sobriety against the strident hero-
worship of the Greeks, then an inscription-like baldness is placed in
deliberate contrast beside the heavy splendour of archaic ubertas.
This effective anticlimax shows once again the value of the specifi-
cally Catonian rhythm we have observed, which in opposition to
Behaghel's law devastatingly tops a full-blown sentence with a
smaller unit. On the other hand a psychological subtlety, such as the
reflection of Roman daring in the enemy's surprise, seems more like
an unintended side product, arising from the functional quality of the
narrative. Cato has said what he wanted to say, and he has said it
each time in a form appropriate to the subject. He had the words at
his disposal, and they could be made to fit the varied but always
subject-oriented lines of his thought. If what he wrote often seemsl
artistic, this effect was never the writer's main aim. The fact that it is!
perceptible and actually turned out to be uncommonly fruitful to.
later writers, belongs to the secrets of Cato's greatness; to the man
himself this kind of glory doubtless did not seem particularly:
desirable.
2
Two great orators
C. Gracchus (154-121 B.C.) and Cicero (106-43 B.C.)
C. Gracchus
From the speech De legibus promulgatis (122 B.C.)!
Nuper Teanum Sidicinum consul venit. uxor ::ius dixit se in balneis
virilibus lavari vel/e. quaestori Sidicino M. Ma,'io datum est negotium,
uti balneis exigerentur, qui lavabantur. uxor renuntiat viro parum cito
sibi balneas traditas esse et parum lautasfuisse. idcirco palus destitutus
est in foro, eo que adductus suae civitatis nobi/issimus homo M. M arius.
vestimenta detracta sunt, virgis caesus est. Caleni, ubi id audierunt,
edixerunt, ne quis in balneis lavisse vellet, cum magistratus Romanus ibi
esset. Ferentini ob eandem causam praetor noster quaestores abripi
iussit: alter se de muro deiecit, alter prensus et virgis caesus est.
Not long ago the consul came to Sidicinian Teanum. 2 His wife announced
that she wanted to use the men's bath. The Sidicinian quaestor Marcus
Marius was instructed to remove the public from the bath. The wife reports
to her husband that the bath was not handed over to her fast enough and
wasn't clean enough. A stake was therefore set up in the forum. The most
distinguished man of his city, Marcus Marius, was taken there. His clothes
were torn off and he was flogged. When the people of Cales heard about this,
they issued an edict to the effect that, when a Roman official was there, no
native could use the bath. In Ferentinum our praetor had the quaestors
dragged off for the same reason; one threw himself from the wall, and the
other was seized and flogged with rods.
Cicero against Verres 3
Ipse inflammatus see/ere et furore in forum venit; ardebant oculi, toto
ex ore crudelitas eminebat. exspectabant omnes, quo tandem pro-
gressurus aut quidnam acturus esset, cum repente hominem proripi
Malcovati 4 , pp.l91f., fr. 48.
2 Teanum Sidicinum is situated in Campania under Mons Massicus, where Via
Latina and Via Appia meet. Teanum, Cales and Ferentinum were municipia with
Italic rights (H. Nissen ltalische Landeskunde [Berlin 1902] II 2 pp.693, 694, 653).
3
Cic. Verr. II 5,62,161- 63, 163; text of G. Peterson (Oxford 1907, 1917 2). Cf. now
also L. Piacente 'Cic. Verr. II 5,162' Quaderni dell'Istituto di Linguae Letteratura
Latina (Univ. di Roma, Facolta di Magistero) I (1979) pp.89-94. Piacente demon-
strates that the repetition of the word crux (162) is due to conjecture. The present
writer regards this conjecture as necessary, especially since inquam follows.
33
34 TWO GREAT ORATORS
at que in foro media nudari ac deligari et virgas expediri iubet. clamabat

il!e miser se civem esse Romanum, municipem Consanum; meruisse
cum L. Raecio, splendidissimo equite Romano, qui Panhormi negoti-
aretur, ex quo haec Verres scire posset. ftUm iste: se comperisse eum 1..-
speculandi causa in Siciliam a ducibus fugitivorum esse missum; cuius
rei neque index neque vestigium aliquod neque suspicio cuiquam esset
u/la; deinde iubet undique hominem vehementissime verberari. 162 .
.1- (Caedebatur virgis in media foro Messanae civis Romanus, iudices, cum
interea nul/us gemitus, nulla vox alia illius miseri inter dolorem
strepitumque 4 plagarum audiebatur, nisi haec: 'civis Romanus sum!'
hac se commemoratione civitatis omnia verbera depulsurum cruci-
atumque a corpore deiecturum arbitrabatur; is non modo hoc non
perfecit, ut virgarum vim deprecaretur, sed cum imploraret saepius
usurparetque nomen civitatis, crux, crux in quam, infelici et aerumnoso,
qui numquam istam pestem viderat, comparabatur.
63.163. o nomen dulce libertatis! o ius eximium nostrae civitatis! o lex
Porcia legesque Semproniae! o graviter desiderata et aliquando reddita
plebi Romanae tribunicia potestas! hucine tandem omnia reciderunt, ut
civis Roman us in provincia populi Romani, in oppido foederatorum, ab
eo qui beneficia populi Romani fascis et securis haberet, deligatus in
foro virgis caederetur? quid? cum ignes ardentesque laminae ceterique
cruciatus admovebantur, site illius acerba imploratio et vox miserabilis
non inhibebat, ne civium quidem Romanorum, qui tum aderant, fletu et
gemitu maximo commovebare? in crucem tu agere ausus es quemquam,
qui se civem Romanum esse diceret?
He came into the forum burning with rage and lusting for blood. His eyes
blazed; cruelty was written all over his face. Everyone was eager to see which
way he would finally turn and what he would do -when suddenly he had a
man dragged forward, stripped in the middle of the forum, and tied up, and
the rods prepared. The poor man shouted repeatedly that he was a Roman
citizen from the municipium Cosa, and that he had served with L. Raecius, a
highly respected Roman knight, who was in business at Panormus, and could
give Verres confirmation of the fact. To this Verres replies that he has heard
he was sent to Sicily as a spy by the leaders of the runaways -though there
was no one to accuse him, nothing definite to go on, and not the slightest
suspicion in anyone. Then he has the man beaten in the most violent way
possible from all sides. 162. A Roman citizen was flogged with rods in the
middle of the forum of Messina, men of the jury, while there was no groan,
no other word of the poor man to be heard in the midst of the painful swish of
the blows than this: 'I am a Roman citizen'. By this reference to his citizen
rights he thought he could ward off every blow of the rods and shield himself
4
strepitumque codd. nonn. Gellii (10,3,12); crepitumque codd. Cic.
C. GRACCHUS AND CICERO 35
from torture. But it was not enough that he failed to prevent violent
scourging by entreaties; worse still, as he pleaded more and more and
appealed to his citizen rights, the cross, the cross, I say, was got ready for the
unhappy man in his affliction, who had never yet set eyes on that awful thing.
163. 0 sweet name of freedom! 0 wonderful privilege of being a Roman
citizen! 0 Porcian law and Sempronian laws! 0 tribunician power, ardently
desired and at last granted to the people of Rome! Has then all this lapsed so
far into decay, that a Roman citizen in a province of the Roman people, in an
allied city, can be bound and beaten with rods in the market-place by the man
to whom the Roman people has entrusted the symbols of authority? Well!
When he was tormented with fire, glowing metal and the other forms of
torture - if his bitter entreaties and plaintive voice did not stop you then,
were you not even moved by the pathetic tears and groans of the Roman
citizens present? You dared to hand someone over to be crucified, who said
he was a Roman citizen?
1. The problem
There has been agreement since antiquity that Gaius Gracchus was
no less important as an orator than as a statesman; 5 however there
has been no such agreement about the particular quality of his
oratorical style. For Mommsen 6 the 'flaming words' of the speeches
preserve 'the passionate earnestness, the noble bearing and the tragic
fate of this lofty nature in a faithful mirror'. Central to his
interpretation is the 'terrible passion of his heart', which made
Gracchus 'the first orator that Rome ever had'. 7 The 'sober' parts of
the speeches are also interpreted from the standpoint of passion: 'For
all his mastery of oratory, he was himself often mastered by anger, so
that the brilliant speaker's flow of words became clouded or
halting'. 8 Such traits are 'the faithful reflection of his political actions
and ordeals'. 9
While a historian like Mommsen understands the speeches as
direct evidence of personality, philologists on the other hand have
established a more detached approach by looking at them in the
context of literary history and determining how far they are
conditioned by factors of this kind. Norden started from the
principle that 'style in antiquity was not the man himself, but a
5 The following go beyond Mommsen: E. Meyer 'Untersuchungen zur Geschichte
der Gracchen' Kleine Schriften I 1 (Halle 1910) pp.383-439, I' (Halle 1924) pp.363-398;
R. v. Pohlmann 'Zur Geschichte der Gracchenzeit' SB Miinchen (1907) pp.443ff.; F.
Munzer RE 2 A 2 (1923) 1375ff. and 1409ff.; A. Heuss Romische Geschichte
(Braunschweig 1960) pp.144-148 and 553f. (lit.).
6
Romische Geschichte pp.454f.
7 lb. p.104. The next sentence 'without it we should probably be ab1eto include him
among the foremost statesmen of all times' recalls Cic. Brut. 125f.
8 lb. p.I04. 9
lb. p.104.
garment that he could change as he pleased'. 10 He demonstrated

Gracchus' dependence on his Asianist teachers 11 and thus opened the
way to an understanding in terms of literary history, although in so
doing he went against his own maxim and still looked for the man in
the style: 'The passionate temperament ... of this man of genius
inevitably found in agitated Asianic eloquence a welcome means of
giving fitting expression to his ideas'. 12
Leo 13 however recognizes the opposition between Gracchus'
temperament and the Asianic manner: this certainly had some effect,
but could not make much difference to the orator. Leo is aware that
the preserved fragments do not confirm the general conception of
Gracchus' passionate gravity. Without repeating Mommsen's in-
genious theory of anger rendering speechless, he sees here an accident
of transmission.
Hapke 14 was the first to reject the cliche of the passionate and
demagogic orator and to stress his factual style of argument. 15 This
introduces an approach which it is worthwhile to pursue further.
What is the basis for the 'emotional' image of Gracchus? In
Tacitus' view, which is taken over by most moderns, Gracchus' style
is richer than Cato's. 16 Plutarch uses similar epithets and establishes
at all points a contrast with the plainer and quieter diction of his
brother Tiberius. 17 By its very consistency this contrast makes one
suspect it of being artificial, like so much else in the 'comparative'
parts of Plutarch. 18 Since his command of Latin was slight, 19 he had
° Kunstprosa p.l2.
1
11
lb. pp.l71-173; for a fuller discussion with regard to prose rhythm cf. Leo
pp.508ff.
12 In the following sentence there is a shift of emphasis. The discussion is no longer
concerned with style, but with external aspects of delivery: 'We hear of his sensational
13
actio .. .' (Norden p.l71). Leo p.308.
14 N. Hapke C. Semproni Gracchi oratoris Romanifragmenta (Diss. Munich 1915).
15 Welcome confirmation from a historical point of view is provided by Ernst
Meyer Romischer Staat und Staatsgedanke (Darmstadt 1961') p.303: 'Purely dema-
gogic motions ... are not among them [sc. those of C. Gracchus]; on the other hand the
nobility used purely demagogic means to overthrow the inconvenient tribune.'
16 Tac. dial. 18 Catoni seni comparatus C. Gracchus plenior et uberior; sic Graccho
politior et ornatior Crassus; sic utroque distinctior et ufbanior et altior Cicero. There is a
similar appraisal in Norden p.l69 and Leeman p.56. Here Tacitus is influenced by Cic.
Brut. 125: Noli enim putare quemquam, Brute, pleniorem aut uberiorem ad dicendum
17
juisse. Plutarch Tib. Gracchus 2.
1' On Plutarch's limits as a historian cf. K. Ziegler RE 21 (1951) 910. Too much
emphasis on moral questions can make Plutarch distort facts and alter their bearing.
One example is his moralizing interpretation of the voice-trainer, who sets the pitch for
Gracchus with his tuning pipe (cf. the instructive list of sources inN. Hapke [quoted
above n.l4] pp.36-38).
19 Plut. v. Demosth. 2,2ff. On this cf. K. Ziegler RE 21 (1951) 926f.
only indirect knowledge of Gracchus' speeches. 20 Besides, the whole

paragraph shows such a strong deductive tendency in proceeding
from the opposed temperaments of the two brothers, that the
particular comments on points of style can only be viewed with
reservations as historical evidence.
The same is true ofTacitus, who did not study Gracchus' speeches,
but gives a simplified version of a passage in Cicero. In the Dialogus
the remark is part of a schematic survey of the technical progress of
Roman oratory and it cannot in the last analysis claim to say
anything individual about C. Gracchus.
Thus in spite of Plutarch and Tacitus it remains unclear how and to
what extent Gracchus' strong emotion, which is undeniably present,
has entered the style of his speeches. This question entails a historical
problem. Have Plutarch and more recent scholars 21 perhaps drawn a
distorted picture of Gracchus, in that, while they were quite right to
point to the emotional aspect, in so doing they neglected other
aspects of this many-sided personality?
The wide-spread view of C. Gracchus' stylistic ubertas stands in
opposition to the opinion of Marouzeau, who cites him as a typical
example of the poverty (egestas) of archaic Latin. 22 For him of course
Gracchus is not a real person but a stage in a historical development.
In the final analysis therefore both Plutarch and Marouzeau start
with a general conception and reach their particular perceptions by
deduction from it. Thus in accordance with their premisses each
comes to the opposite result. Here the text is scarcely more than a
'pretexte'.
A more accurate picture can be obtained by paying attention to the
different modes of expression and stylistic levels in Gracchus, as
Leeman has shownY In what follows we shall make the texts our
starting-point and attempt (partly by comparison with Cicero) to
20
On Plutarch's citations from Gracchus' speeches seeN. Hapke pp.13-19. Cf.
Pohlmann p.445 (with lit.), who posits a historical work as intermediary.
21
Cf. also Val. Max. 8,10,1 (jlagrantissimo ingenio); Tac. dial. 26 (C. Gracchi
impetum); Apul. apol. 95 (impetum); Gel!. 10,3 (fortis ac vehemens); Pronto p.132 van
den Hout (contionatur ... Gracchus turbulente; ... tumultuatur Gracchus); Claud. Mam.
epist. 2 p.206 Engelbrecht (Gracchus ad acrimoniam ... capessendam usui). There is also
little attempt at discrimination in A. Heuss p.l44: 'The passion that animated him was
like a volcano.'
22
J. Marouzeau Eranos 45 (1947) pp.22-24. He gives a more subtle analysis in RPh
45 (1921) pp.l66-!68: Gracchus is at home in 'two styles' (168). Quint. ins!. 12,10,10is
lukewarm (while acknowledging the brevity). Cf. Sen. epist. 114,13. Plin. epist. 1,20
speaks of Gracchus' orationes circumcisae. Criticism is expressed by Gel!. 10,3,15 and
23
Sen. epist. 114,13. Loc. cit. pp.56-58.
reach an individual portrait of Gracchus the orator. This time the

special character of our sample text calls for some methodological
detours. If it has hitherto been thought more banal than it is, this is
due largely to the fact that it has been considered in isolation.
Consequently it will require some patience to make clear the
intellectual landscape to which it belongs and to illustrate it from
different angles by comparing other texts; in this way it can have its
proper impact. The following sections deal in succession with
attitude to language, narrative method, rationality and emotion. 24
2. Attitude to language: Latinitas - mundities
a) Choice of words
Certain words are repeated in our text without any evident rhetorical
purpose: in balneis, balneis, balneas, in balneis; lavari, lavabantur.
Another narrative of Gracchus presents a similar picture: we quote it
here for comparison: 25
Quanta libido quantaque intemperantia sit hominum adulescentium,
unum exemplum vobis ostendam. his ann is paucis ex Asia missus est, qui
per id tempus magistratum non ceperat, homo adulescens pro legato. 26 is
in lecticaferebatur. ei obviam bubulcus de plebe Venusina advenit et per
iocum, cum ignoraret, qui jerretur, rogavit, num mortuumferrent. ubi id
audivit, lecticam iussit deponi, struppis, quibus lectica deligata erat,
usque adeo verberari iussit, dum animam efflavit.
I want to show you by means of an example to what lengths the
wantonness and intemperance of young people goes. A few years ago a
young man who at that time had not yet held office was sent in place of
an ambassador from Asia. He had them carry him in a litter. An ox-
herd met him, a simple man from Venusia, and asked in jest (for he did
not know who was being carried) if they were carrying a corpse. When
the young man heard this, he had the litter put down and ordered the
ox-herd to be beaten with the straps of the litter until he gave up the
ghost.
In this text also we observe unrhetorical verbal repetitions of the
same kind: ferebatur, ferretur,ferrent; lectica, lecticam, lectica; iussit,
iussit; per id tempus, per iocum. Marouzeau stresses how much more
artistic Cicero's handling of vocabulary isY However one .needs to
be careful with phrases like 'colloquial carelessness', since even
24
The Latin headings are based on Gell. 10,3,4: brevitas sane et venus/as et
25
mundities orationis est. Fr. 49 Malcovati 4 •
26 On the constitutional background cf. T. Mommsen Romisches Staatsrecht
(Leipzig 1887) II p.681,3 'This does not seem to refer to a delegate of the Senate, but of
a Roman official present in Asia ... pro legato denotes the purpose of the journey'.
27
RPh 45 (1921) p.167. Cf. the deliberate variation in Cicero: in foro media- in
media foro (the latter more emphatic: 'En pleine place publique').
Quintilian 28 rejects as affectation the excessive search for synonyms

merely for the sake of variety. Here we may be sure that Gracchus is
keeping to the proprietas verborum.
b) Sentence connection
In fragment 49 the demonstrative pronoun is connects the sentences:
is ferebatur ... ei obviam advenit ... ubi id audivit. In fragment 48 (our
main text) asyndeton is frequent: quaestori ... uxor ... vestimenta ...
alter. Alongside this there is also connection by means of demon-
stratives: eius ... idcirco ... ubi id audierunt ... ob eandem causam.
Participial constructions are significantly absent. But does that allow
us to conclude: 'La construction dans Gracchus est uniforme et
banale'? 29
c) Appreciation
The language of Gracchan narrative is pure, clear and precise. The
origin of such Latinitas30 can be traced back to the man's early
biography. Gaius grew up -even more so than his brother- under
the supervision of his mother Cornelia: according to Cicero's
evidence (who was likewise a purist) she ensured that her sons grew
up in the healthy ambience of an uncorrupted mother-tongue and she
took personal charge of their upbringing and education 31 (we still
possess a letter of this important woman). 32
It is true that Gracchus' disciplined attitude to language and the
effects of this on his style have put offlater readers, who expected in
certain contexts a richer and more elevated tone. Compared to
Cicero's account with its artistic form and emotional colouring, 33
28
Quint. inst. 10,1,7, cf. 8,3,51.
29
J. Miuouzeau RPh 45 (1921) p.l67.
30
On Latin it as in general cf. J. Marouzeau Quelques aspects de laformation du latin
litteraire (Paris 1949) pp. 7-25 (Latinitas- Urbanitas- Rusticitas).
31
Cicero Brut. 104: Nam et Carbonis et Gracchi habemus orationes nondum satis
splendidas verbis, sed acutas prudentiaeque plenissimas. fuit Gracchus diligentia
Corneliae matris apuero doctus et Graecis litteris eruditus. nam semper habuit exquisitos
e Graecia magistros, in eis iam adolescens Diophanem Mytilenaeum, Graeciae
temporibus illis disertissimum. sed ei breve tempus ingenii augendi et declarandifuit. Cic.
Brut. 210: on the importance of usus domesticus ... Sed magni interest quos quisque
audiat quotidie domi, quibuscum loquatur a puero, quemadmodum patres, paedagogi,
matres etiam loquantur. 211: Legimus epistolas Corneliae matris Gracchorum: apparet
filios non tam in gremio educatos quam in sermone matris. On the importance of
Cornelia cf. also Tac. dial. 28,9; Quint. inst. 1,1,6; Plut. Tib. Gr. 1,8; F. Munzer REIV
1592-1595.
32
It is transmitted at the end of Nepos' life of Atticus. Nepos had evidently cited it
in his work De inlustribus vir is. Cf. HRR II pp.38-40 Peter. Leo translated this letter in
the appendix to his literary history (p.479).
33
See below pp.4l; 47ff.
Gracchus' report does indeed sound plain and commonplace:

Gellius finds the diction 'comic' ,34 i.e. close to everyday speech, 35 and
dispensing with tragic ornamentation. 36 This need not of course
entail any lack of gravitas, which in Cicero's view Gracchus did
possess. 37 It is not therefore permissible to limit Cicero's idea of
gravitas to the elevated style of the Verrines passage.
The term 'comic' is less helpful in defining the tone of Gracchan
narrative than the aesthetic notion of mundities. 38 Belonging etymG-
logically to the root *meu- ('wash'), it describes the aesthetic effect
achieved through linguistic purity (Latinitas).
3. Narrative technique: 39 Brevitas
In the matter of linguistic purity we found no basic difference
between Gracchus and Cicero. It is otherwise in regard to brevitas. 40
34 Gell. 10,3,4. He finds more gravitas in Cicero.
35
Cf. Don. Ter. Hec. 611 KffiJltKii\ xapaK-rfjpt et usu cotidiano.
36
Cf. Gloss. Plac. 5,56, II comoedia est quae res privatarum et humilium personarum
comprehendit non tam alto ut tragoedia sti/o, sed mediocri et dulci. .
37 Cic. Brut. 125: genere toto gravis (about C. Gracchus). On the genus grave cf. Cic.
or at. 96-99 (on this cf. W. Kroll), esp. 97 huius eloquentiae est tract are animas ... haec ...
inserit novas opiniones, eve/lit insitas. Plutarch Tib. Gr. 2,3 also calls C. Gracchus
yqavffil-u\vo~ (= JlEyaf.onpEnTj~; cf. R. Jeuckens Plutarch von Chaeronea und die
Rhetorik [Diss. Strassburg 1907] p.l77; Hapke p. 34).
38 For the meaning cf. Lat. lautus. Since the concept is surprisingly not treated by P.
Mantei! Beau et laid en latin (Paris 1964), some examples may be cited: Cic;. or. 79:
removebitur omnis ins ignis ornatus ... elegantia modo et munditia remanebit, sermo purus
erit et Latinus; Quint. 8,3,87: quaedam velut e tenui diligentia circa proprietatem
significationemque munditiae; Gel!. I ,23, I (on Cato) cum multa ... venustate at que luce
atque munditia verborum; 10,24,2 (Augustus) munditiarum ... patris sui in sermonibus
sect at or.
39 For a general treatment of the subject in Cicero see also now R.C. McClintock
Cicero's Narrative Technique in the Judicial Speeches (Diss. Chapel Hill 1975;
Microfilm: Dissertation Abstracts 36 [1975] 3672A); D. Berger Cicero als Erziihler,
forensische und literarische Strategien in den Gerichtsreden Europ. Hochschulschriften
15,12 (Frankfurt/Bern/Las Vegas 1978); M. Fuhrmann 'Narrative Techniken in
Ciceros zweiter Rede gegen Verres' Der altsprachliche Unterricht 26 (1980, Heft 3)
pp.5-17; J. Blansdorf 'Erzahlende, argumentierende und diskursive Prosa' Wiirz-
burger Jahrbiicher N.F. 4 (1978) pp.107ff.
40 On brevity as a characteristic of the Latin language cf. Plut. Cato maior 12,7 (on
Cato) 8auJJ.cicrat &f. q>T]crt -rou~ 'A8T]vaiou~ -ro -rcixo~ au-rou Kai -ri]v 6~1hT]'tU -rfi~
cppcicr£ffi~· a yap au-ro~ e~f.q>Ep£ ~PU)CEffi~, 'tOY EPJlT]Vf.a JlUKpiii~ Kai &ta nof.Aiiiv
unayyf.AAEtv· 'tO&' OAOV Ot£cr8at 'tU pTjJlU'ta -roi~ JlEV "Ef.AT]crtV uno )CElAEffiV, wi~ &£
'Pffi)J.Uiot~ uno Kap&ia~ cpf.pw8at. Plut. Caesar 50,3 Kat 'tfi~ JlUXTJ~ 'talhT]~ 'tTJV
6~U'tTJ'ta Kai -ro -raxo~ uvayyf.AAffiV Ei~ 'PWJlTJV, np6~ nva -riiiv cpiAffiV 'AJJ.civnov,
€ypaiJIE -rpEi~ Af.~Et~. "TEf.8ov, d&ov, 6viKT]cra." 'PffiJlatmi &£ ai AE~Et~ Ei~ OJ.!Otov
anof.Tjyoucrat crxfiJJ.a pTjJlU'tO~, OUK uni8avov 'tTJV ~paxuf.oyiav ilxoucrtv. In this trait
(as in much else) Roman and Stoic meet. The latter regarded cruv-roJJ.ia as one of the
chiefsty1istic virtues (see Leeman p.39 with n.Sl; reference to SVF[ed. H. von Arnim,
Leipzig 1903] III p.214,16). For a similar view cf. Quint. 4,54,68; Cic. inv. 1,32; Rut.
Lup. 2,8. Plin. epist. 1,20,1-4 is critical of brevitas (including that of Gracchus).
Analysis of the Cicero text will show indirectly how concisely

Gracchus writes.
In fr. 48 Gracchus strings the facts together asyndetically. Variety
is provided by the change of tense in the wife's reaction: uxor
renuntiat. Here the historic present forms a deliberate contrast to the
surrounding perfects. 41 Use of idcirco as a sentence link introduces
the erection of the stake and puts strong emphasis on the triviality of
the motive. In an otherwise asyndetic style an adverb like this
acquires .~rucJpral significance. The whole narrative is dominated by
balancing pairs. This is accentuated by the fact that sentences begin
in the same way: sentence 1 and 3 start with uxor; cf.later the proper
nouns Caleni - Ferentini and in the last sentence alter - alter. The
alliteration has ·a similar function: vestimenta - virgis. The overall
structure falls into three parts: background (2 X 2 sentences); main
event (2 X 2 short sentences); consequences (2longer sentences). The
absence of any sort of emotional comment is noteworthy.
At the beginning of the Cicero passage asyndetic connection of
sentences predominates, as in Gracchus. A new element is the
representation of psychology in terms of physiognomy: toto ex ore
crudelitas eminebat. A translation such as 'cruelty stood written on
his brow' would be far too colourless. Psychology is mirrored by
emotional adjectives and participles (inflammatus see/ere et furore;
illius miseri; o nomen dulce; o ius eximium; o graviter desiderata ...
tribunicia potestas; acerba imploratio et vox miserabilis;fletu gemitu-
que maximo) and also by psychological abstracts (see/ere et furore;
crudelitas). Whereas Gracchus is content to use only 'proper' terms,
as befits the style of the report, Cicero employs more expressive
verbs: 42
Gracchus Cicero
adductus ... Marius proripi
vestimenta detracta sunt nudari
Let us pass now to the overall structure. Whereas Gracchus simply
reports, Cicero is able to make his hearers visualize what happens by
breaking it up into smaller units that follow each other step bystep. 43
Here one of the most important devices is the imperfect 44 and the
41
This is meant to be emphatic, and perhaps also to characterize what happens as
the outcome of the preceding events.
42
Cf. J. Marouzeau RPh 45 (1921) p.l67.
43
In Cicero the arrangement of tenses in the narrative operates on several planes,
producing an impression of perspective. In Gracchus there is just one plane.
44
Gellius had already noted the function of the imperfect cgrrectly (10,3,12).
periphrastic future subjunctive: exspectabant omnes, quo tandem

progressurus aut quidnam acturus esset. Through these verbal forms
an expectant tension is created, which is suddenly relaxed in the
following cum repente.
In Gracchus the action is over very soon after it has begun. Cicero
on the other hand achieves an effective slowing down, for example
through 'the preparation of the rods. Whereas Gracchus deprives
himself of the chance of powerful visualization by using the dry
perfect caesus est, Cicero employs the imperfect caedebatur in
expressive initial position. 45 By means of diutina repraesentatio, as
Gellius calls it, Cicero is able to build up a scene: general silence,
crash of the whip strokes ... and against this background from the
mouth of the tortured man the words ring out: 'I am a Roman
citizen'. In this way Cicero lets what is outrageous about the event
become dramatically audible in the action itself, whereas Gracchus is
content with the simple statement that the most distinguished man of
his city is involved.
The evocative imperfect appears once again in the preparation of
the cross, accompanied by an expressive repetition: 46 crux, crux,
inquam, ... comparabatur. Generally speaking, verbal repetitions in
the Cicero text have an intensificatory effect, as for example the
thematically recurrent civis Romanuslcivitas and populus Romanus,
as well as the polysyndeton with neque and the anaphoric o.
Accordingly Cicero does not merely have emotion present sub-
consciously, but lets it appear in the actual text (this is clear from the
use of emotional adjectives and psychological abstracts and from the
whole of the lengthy commiseratio appended to the account)Y He
makes the hearer visualize the events dramatically (there are various
means to this end: choice of expressive verbs, emphatic initial
position, creation of a background full of tension or anticipation by
using the imperfect, which in Gracchus is completely absent at this
point, and finally the skilful use of retardation in breaking up a single
overall event into individual phases, which as they follow each other
give rise to a dramatic progression). 48
45 J. Marouzeau L' ordre des mots dans Ia phrase Ia tine val. II (Paris 1938) p. 71. Also
in Cicero: ardebant oculi ... ; exspectabant omnes ....

46 In the repetition of the word crux Piacente (quoted above p.33 n.3) sees the work
of a humanist; but surely the ensuing inquam presupposes a repetition of crux.

47 On this cf. Gel!. 10,3,14: haec M. Tullius atrociter. graviter, apte copioseque
miseratus est.
48 Gell. 10,3, 7/8 stresses the sub oculos subiectio and describes the effect on the reader
of this text of Cicero in the following terms: Animum hercle meum, cum illaM. Ciceronis
!ego, imago quaedam et sonus verberum et vocum et eiulationum circumplectitur.
There is a great temptation (and it is one to which the majority of

interpretations have in fact succumbed) to play Cicero off against
Gracchus, whether as an absolute stylistic norm or as representing a
more mature level historically. These views are basically just as one-
sided as was the wilful attempt of many archaizers to set Gracchus
above Cicero. 49 We can see for example how dangerous the idea is
that Gracchus did 'not yet' have this or that device at his disposal, if
we look at the following text of Cato, which displays a rich range of
emotion and is nonetheless earlier than Gracchus: 50
Dixit a decemviris parum bene sibi cibaria curat a esse. Iussit vestimenta
detrahi atque jlagro caedi. decemviros Bruttiani 51 verberavere, videre
multi mortales. quis hanc contumeliam, quis hoc imperium, quis hanc
servitutemferre potest? nemo hoc rex ausus estfacere: eanefieri bonis,
bono genere gnatis, boni consultis? ubi societas? ubi fides maiorum?
insignitas iniurias, plagas, verbera, vibices, eos dolores atque carni-
ficinas per dedecus at que maximam contumeliam, inspectantibus popu-
laribus suis atque multis mortalibus, te facere ausum esse? set quantum
luctum, quantum gemitum, quid lacrimarum, quantum fletum factum
audivi! servi iniurias nimis aegre ferunt: quid illos, bono genere gnatos,
magna virtute praeditos, opinamini animi habuisse atque habituros, dum
vivent?52
He said he had not been properly supplied with provisions by the
decemviri. He ordered them to be stripped of their clothing and
whipped. Decemviri flogged by beadles! Many people saw it. Who can
endure this outrage, this misuse of authority, this servitude? No king
dared do this. Can this happen to respectable, right-thinking people of
good family? What has become of the alliance? What of the pledge
given by our ancestors? You dared to inflict glaring injustices, blows,
beatings, weals, pain and torment in shame and utmost indignity
before the eyes of their countrymen and many people! But how great
was the sorrow, how great the lamentation, what abundance of tears,
how mighty the sobbing, as I heard! Even slaves feel enormous
resentment about unjust treatment. How do you think those people
must have felt, who were of good family and had done great services,
and how will they still feel, as long as they live? 53
49
Gellius is perceptive enough to dissociate himself explicitly from such eccentric
views (10,3, 15). Cf. also Sen. epist. 114,13: multi ex alieno saeculo petunt verba,
duodecim tabu/as locuntur. Gracchus illis et Crassus et Curio nimis culti et recentes sunt.
Cf. W. Soltau NJbb 9 (1906) p.26 n.l.
5° Cato fr. IX J. = fr. 58 Malcovati 4 •
51 The Bruttians performed these tasks as a punishment for their support of
Hannibal in the Second Punic War.

52 On the rhythmical structure of the passage cf. A. W. de Groot La prose metrique
des anciens (Paris 1926) pp.44f. ('periodes arrondies, membres symetriques et souvent
isochrones, mais pas de metrique').
53 Translation partly after 0. Ribbeck, in Neues Schweizer Museum I (1861) p.l2.
By comparison with Gracchus the word order in Cato is freer. He

is also aware, as Cicero was later, of the expressive initial position of
the verb: 54 videre multi mortales. 55 The text shows that the Censor
does not strive for brevity at any price, but that he too loves ubertas.
There is an anticipation of Ciceronian miseratio with anaphora,
emotional substantives and adjectives; but Cato's sentences are
short-winded in comparison, and the technique of dramatic climax is
absent. 56
Comparison with Cicero and Cato thus enables us to make two
negative statements about Gracchus:
1) He does not really narrate in a visual and dramatic way; he does
not build up an effective climax like Cicero.
2) He does not employ miseratio here, although Cato is already
familiar with it.
We have therefore to speak in terms of deliberate intent at least as
regards the second point. 57 This means we can discard the view that
there is no more to Gracchus than primitiveness. Accordingly we are
justified in adopting a positive approach to the question of Gracchus'
artistic principles in the narrative before us.
4. Rationality: Acutum
The structure of the Gracchan narrative is, as we have established,
strictly rational. Balancing pairs predominate, accentuated several
times by parallelism. The transparency of the structure, combined
with the hard language of facts, gives the tone a cutting, unmasking
quality. In rationality such as this lies the charismatic element that
distinguishes Gracchus as an 'intellectual'.
Brevitas is closely related to acutum 58 (o~(rcrg;), which is based on
the notion of a short, sharp thrust-weapon. In terms of content, ideas
are compressed into a very small space; 59 in moral terms, the
54 On this in general cf. J. Marouzeau L' ordre des mots ... passim, esp. pp.49ff.
55
The remarks of Fankhiinel (above p.ll n.42) p.230 about consular reports as the
model for such positioning are not convincing; on the other hand the structural
function of such a change of position is clear.
56 C. Gracchus is considerably more restrained than Cato e.g. in the accumulation
of synonyms; so here too he shows a refined taste. Cf. the cases of ubertas
demonstrated by Hapke, loc. cit. p.40, almost all of which are elegantly unobtrusive:
sapientia atque virtute; commoda et rem publicam; bonam existimationem atque
honorem; pretium et praemium; eadem loco atque ordine; sumptus atque pecunias.
57 There is some support for this view in the fact that miseratio is ascribed mainly to
Tiberius Gracchus in our tradition (cf. Plut. Tib. Gr. 2), whereas Gaius' style was felt to
be more virile.
58 Cf. e.g. Quint. 6,3,45: acutior est ilia atque velocior in urbanitate brevitas.
59 The proem to Eustathius' commentary on the Odyssey explains 6~UTT]~ as
expression denotes an earnest and dignified attitude (gravitas); and

as to form, it often approaches the maxim or occurs with some other
kind of point. 60 Accordingly it is no accident that antithesis is the
figure of speech which appears most often in the fragments. 61 Here
are some examples. Pessimi Tiberium fratrem meum optimum inter-
fecerunt:62 what an apt reversal of the title vir optimus, 63 claimed as
hereditary for the murderer N asica, alongside the slogans boni and
mali cives! Gracchus plays with the closely related contrast of boni
and improbi64 in the following fragment: abesse non potest, quin
eiusdem hominis sit probos improbare, qui improbos probet. 65 We owe
the quotation to Cicero, who does however suggest an improvement:
qui improbos probet probos improbare. 66 This gives rise to an even
sharper point and above all to a clausula. In the Ciceronian version
the word order is more involved and rather more artificial than is to
be expected in Gracchus. 67
Hellenistic training is evident in the following sentence: quae vos
cupide per hosce annos adpetistis atque voluistis, ea si temere
repudiaritis, abesse non pot est quin aut olim cupide adpetisse aut nunc
cupide repudiasse dicamini. 68 The period is carefully handled. The
first half contains 32 syllables, the second 31; we observe in it two
sections each beginning with aut and each of 10 syllables. 69 Norden 70
refers to Gorgias and !socrates in relation to this period. However for
once he is less critical than an ancient authority/ 1 who discovers a
VOlW<iTOJV ~a8UT11<; l:v l:mrroA.a~OU01J arrAOTllTl. (Ed. Rom. 1379 = ed. G. Stallbaum,
Leipzig 1825, T. I p.2). Cf. also J.C. Ernesti Lexicon Technologiae Graecorum
Rhetoricae (Leipzig 1795, repr. Hildesheim 1962) s.v. 6~0Tlls·
60
B.R. Voss Der pointierte Stil des Tacitus (Munster 1963) examines neither the
ancient concepts nor the modern one ('punch line'). On the concept of 'epigrammatic
point' see H. Lausberg Handbuch der literarischen Rhetorik II (Munich 1960) p.933.
61 The correct interpretation in Hapke pp.38f., with examples. Hapke misunder-
stands Leo p.309, I 'parallelism and antithesis are less prominent, but very studied, in
the sentences at Gell. II, 10,4'.
62
Fr. 17 Malcovati 4 • The fragment comes from a recommendation of the motion of
the tribune Carbo, which was intended to allow re-election to the tribunate ( 131 B.C.).
63
F. Munzer RE 2 A 4 (1923) 1380.
64
improbus =rerum nov arum cupidus cf. ThLL 7 ( 1934-1964) 690,30; 36; 40; 42; 68ff.
" Fr. 24 Malcovati 4
66
It is not clear to the present writer what is to be gained from labelling this stylistic
figure commutatio (Leeman p.57).
67
There are also sharp epigrammatic points to fragments 28, 43, 58 and 60,
Malcovati 4 " Fr. 32 Malcovati .
4
69
Norden Kunstprosa I p.l72. In the proem to the speech De legibus promulgatis
Gracchus observes the Asianic rhythms, especially the double trochee (Leo p.309, 2).
Cf. also Hapke p.59. 'Metric' and 'non-metric' passages in C. Gracchus are
70
distinguished by A.W. de Groot pp.46f. lb.
71
Titus Castricius ap. Gell. 11,13.
tautology here: the point does indeed suffer from having the decisive
adverbs already appear in the first half of the sentence. A translation
makes the weakness visible: 'If you now blindly scorn what
throughout these years you have eagerly desired and wanted, then it
will inevitably be said of you that you either desired it once in blind
greed or now blindly scorned it'. In translation the sentence loses its
rigorous structure and with it its effect. For the sake of the overall
design Gracchus has here accepted a tautology. However the
following intensification is in every respect a success: pueritia tua
adulescentiae inhonestamentum fuit, adulescentia senectuti dedeco-
ramentum, senectus rei publicae jlagitium. 72 This sentence became one
of the standard examples of a good climax, and to this circumstance
we owe its preservation. 73
From the narrative whose purpose is to unmask, it is only one step
to the pitiless candour and logic of fragment 44. Its sequence of
thought is as follows: Everyone wants something from you; none of
us works for nothing; neither do 1: I want honour from you; whoever
speaks against the law under discussion, does not want honour from
you, but money from Nicomedes; whoever speaks for the law, does
not want honour from you either, but money from Mithridates;
whoever is silent, is the worst: he let himself be bribed by both.
We have attempted to understand acutum in Gracchus' style, with
respect to both form and content, as a symptom of his marked
rationality. We also established en passant that the same trait is
reflected in the application of Greek rhetorical technique. 74 However
for Gracchus there is nothing strange in either. On the one hand the
striving for pointed formulations is the refinement of a genuinely
Italic tendency; on the other the rhythmic structure of the sentence
gives scope not just to Greek theory, but to the 'architectonic' trend
of the Latin language. The purity of the linguistic substrate is
matched by the lucidity o:fGracchan diction, whiCh gives to acuium a
character different from what it has in Cato. 75
' '
Fr. 43 Malcovati 4 from the speech against L. Calpurnius Piso Frugi of 123 B.C.
72
73
Isidore orig. 2,21,4.
74
Leeman sees such traits in Cato; scepticism is expressed by M. Fuhrmann
Gnomon 38 (1966) p.360.
75 Cicero's attitude is not to be thought of as less rational; it is merely that alongside
elevation and irony there is also humour, and besides analytic rationality there is also
the specifically artistic kind. See the end of the chapter.
5. Style and emotion

a) Word order
In contrast to Cato and Cicero it is only on rare occasions that
Gracchus' word order makes use of inversion. 76 The verbs stand
mostly at the ends of the sentences, in line with normal practice and
without particular emphasis. If however one looks at the other parts
of the sentence, a different picture emerges. 77 In fragment 48 the
proper name Marcus Marius is twice given an important place; the
same thing happens with the place names Teanum Sidicinum and
Ferentini. Gracchus felt that proper names need to be specially
stressed (which is what Goethe recommended to actors) 78 and so he
moved the names to the beginning or end of the sentence. Initial and
final positions are not particularly noticeable during silent reading:
only oral delivery reveals their full power. 79 In the sentence eoque
adductus suae civitatis nobilissimus homo M. Marius the emphasis is
due to the unusual final position of the subject and to the attribute
nobilissimus, 80 especially as Gracchus is otherwise sparing in his use
of adjectives. 81
In fragment 49 the decisive element in the sentence is each time
stressed by its position at the end: firstly the subject, which is
introduced to the audience with a certain undertone of irony as homo
adulescens pro legato; secondly the jest of the drover: num mortuum
ferrent; and finally the shocking conclusion of the scene: dum animam
efflavit. On this view even the seemingly colourless idcirco at the
beginning of the sentence acquires importance ('for this reason and
no other, for this ridiculous reason'). In contrast therefore to
Marouzeau, who thinks inversion is absent in Gracchus, we have to
conclude that skilful exploitation of initial and final position in the
sentence imparts liveliness and grace to our text.
76 77
J. Marouzeau L'ordredes mots ... II p.71. Marouzeau failed to notice this.
78
Goethe Regeln fur Schauspieler (1803) §13 (WA 40, 143) 'in general a stronger
emphasis than usual must be placed on proper names in pronunciation, because such a
name has to be especially noticeable to the listener'. Also important is §27 (WA 40,
l50f.) (proper names should be pronounced more clearly and with a special tone of
voice, to arouse the hearer's imagination).
79
The importance of final position is recognized by Quint. 9,4,29. Cf. ib. 67 (though
in connection with sentence rhythm): initia clausulaeque plurimum momenti habent,
quotiens incipit sensus aut desinit. The end is even more noticeable than the beginning
(cf. 9,4,63).
80
On the almost invariable forward position of such adjectives in early prose see A.
Reckzey Uber grammatische und rhetorische Stellung des Adjektivums bei den
Annalisten, Cato und Sallust (Programm Berlin 1888) p.29.
81 When he uses them, he does so all the more effectively: fr. 17 pessimi- optimum;
27 postremissimum nequissimumque.
b) Delivery
When Gellius criticizes the absence of emotional appeals, he may not
be paying enough attention to the fine nuances in word order, which
do make a difference even within the limits of this simple language.
He thinks that emotion, hate and irony are not given sufficient
expression here. 82 Yet could the orator not put them into the tone of
his delivery? Historical evidence confirms what is already suggested
by the word order. Gracchus was a master of delivery and actually
had recourse to powerful non-literary devices to give his words
emphasis. Plutarch depicts his lively performance in contrast to his
brother's steady manner. 83 Cicero puts C. Gracchus in the same class
as Demosthenes with regard to delivery. 84 In Gaius' case it is
permissible to speak of an art of 'register', for we should bear in mind
a fact which strikes us as unusual nowadays in an orator: he always
had a man beside him who set the tone with a tuning-pipe, whenever
his master spoke too deeply or with too much violence. 85
According to Cicero's evidence, one of the most elevated passages
in Gracchus clearly owed its effect on the audience more to skilful
delivery than to the mere words of the text: 86 'quo me miser conferam?
quo vertam? in Capitoliumne? at fratris sanguine madet. 87 an domum?
matremne ut miseram lamentantem videam et abiectam?' 88
82 Gell. 10,3,4: In tam atroci re ac tam misera atque maesta iniuriae publicae
contestatione ecquid est, quod aut amp liter insigniterque aut lacrimose at que miseranter
aut multa copiosaque invidia gravique et penetrabili querimonia dixerit? brevitas sane et
venustas et mundities orationis est, qualis haberi ferme in comoediarum festivitatibus
solet ... ; ib. 13 on Cicero: complorationem deinde tam acerbae rei et odium in Verrem
detestationemque aput civis Romanos inpense atque acriter atque injlammanter facit.
83 Plut. Tib. et C. Gracchus 2.
84 Cic. de orat. 3,214.
85 Cic. de orat. 3,224. 227. This is thesourceofQuint. inst. 1,10,27; Val. Max. 8, 10, 1;
Gell. 1,11,10ff. The story is reinterpreted in moral terms and misunderstood by

Plutarch Tib. Gr. 2,4; mar. 456 A; Cass. Dio fr. 85,2. Of fundamental importance is R.
Buttner Porcius Licinus (Leipzig 1893) pp.80ff. Cf. also NordenKunstprosai p.57, with
the important reference to L. Cresollius Vacationes autumnales, sive de perfecta
oratoris actione et pronunciatione libri III (Paris 1620) p.499.
86 Fr. 61 Malcovati 4 ; Cic. de orat. 3,214.
87 The transmission is uncertain at this point: sanguine madet M: sanguinem (-ne P 2 )
redundat L. Quint. inst. 11,3,115 abbreviates to: adfratris sanguinem. This is the source
of C. Iulius Victor (p.443 Halm). redundat is the word to be expected in such a context
in Cicero. madet is no doubt right. The word is not typically poetic (e.g. Cato agr. 85).
The present passage is strangely cited in ThiL 8 (1936-1966) 33,29 s.v. madere as a
piece of Cicero (!)without reference to Gracchus (W. Richter).
88 The moving words are assigned to a speech made by C. Gracchus in the last days
of his life (121 B.C.). (Thus Malcovati ad Joe., though without pinning herself down to
the last day, like Hapke [p.90].) This strikes the present writer as plausible, but not
absolutely certain. May not madet even indicate that the murder ofTiberius is not all
It has long been recognized that there are parallels to this passage
in earlier and later literature. 89 The links with Euripides are closer
than with Ennius. Like Euripides, Gracchus puts the objection
immediately after each question and so obtains an arrangement that
is lively and varied, but also very clear. Norden has dismissed the idea
of direct dependence on Demosthenes 90 (for it is unlikely that
Gracchus would have developed such tremendous grandeur from
Demosthenes' trivialization 91 ). The best explanation therefore is that
Gracchus is drawing on the tradition of the Greek schools. The
striking similarity with Euripides can in the present writer's opinion
be explained most simply by the fact that on mnemonic grounds
Greek rhetors liked to exemplify their teachings with quotations
from the poets.
If we ask ourselves why the passage in Gracchus moves us, what
comes to mind first are significantly not stylistic considerations, but
on the one hand the oppressive situation in which the words were
spoken, and on the other the masterly delivery attested by Cicero,
which held even the opposition in its spell. 92
The special quality of the style becomes clear on comparison with
later parallels, from which we select only Cicero Pro Murena
41 ,88f.: 93
Si, quod luppiter omen avert at, hunc vestris sententiis adflixeritis, quo
se miser vertet? domumne? ut earn imaginem clarissimi viri, parentis sui,
quam paucis ante diebus laureatam in sua gratulatione conspexit,
eandem deform at am ignominia lugentemque videat? an ad matrem, quae
misera modo consulem osculatafilium suum nunc cruciatur et sollicita
est, ne eundem paulo post spoliatum omni dignitate conspiciat? sed quid
that far back in the past? This argument could perhaps be used to support the view of
K.W. Piderit (on de or. 3,214) that.the speech was made shortly after Tiberius' death.
One would then of course have to rebut Hapke's counter-arguments (p.88).
89
Eurip. Med. 502-505 vuv rroi Tparrw).!at; rroTEpa rrpo<; rraTpo<; OO).!OU<;; I oG<; croi
rrpo8oucra Kat rraTpav aqHKO).!T]V; I il rrpo<; mA.aiva<; ITEA.ta8a<;; KUAiii<; y' iiv ouv I
8£~atVTO ).!' otKot<; ffiv rraT£pa KUTEKTavov. Enn. trag. 231 R.: quo nunc me vortam?
quod iter incipiam ingredi? Domum paternamne? anne ad Peliaefilias? Demosth. or. 28
(= KUTU 'A<po~ou ~ ') 18 rroi 8' iiv TparrOt).!E8a ... d<; TU urrOKEl).!EVa TOt<; 8avdcramv;
a
aHa TWV Grro8E).!EV(J)V i:crTiV. A.A.' d<; TU rrEptOVT' at'niiiv; ana TO\JTOU yiyvnat ... On
this cf. Norden Kunstprosa I Nachtrage pp.13f. (to p.l71); ib. examples and
bibliography. Malcovati ad Joe. has missed these important addenda. Here Norden
provides more abundant material than the article ofM. Bonnet REA 8 (1906) pp.40-46,
cited by Malcovati.
90
Leeman pp.56f. seems to assume direct influence from Demosthenes in spite of
91
Norden. Norden ib.
92
Cic. de or. 3,214: Quae sic ab illo esse acta constabat oculis. voce, gestu, inimici ut
lacrimas tenere non possent.
93
Text of A.C. Clark (Oxford 1905).
ego 94 matrem aut domum appello, quem nova poena legis et domo et
parente et omnium suorum consuetudine conspectuque privat? ibit igitur
in exsilium miser? quo? ad Orientisne partis, in quibus annos multos
legatus fuit, exercitus duxit, res maximas gessit? at habet magnum
dolor em, unde cum honore decesseris, eadem cum ignominia reverti. an
se in contrariam partem terrarum abdet, ut Gallia Transalpina, quem
nuper summp cum imperio libentissime viderit, eundem lugentem,
maerentem, exsulem videat? in ea porro provincia quo animo C.
Murenam, fratrem suum, aspiciet?
If you - Jupiter forbid it! - crush this man [Murena] with your
judgment, where then will the poor wretch turn? To his home, where
he will have to see how the image of his illustrious father is now
shamefully dishonoured and grieving, which but a few days ago, when
people congratulated him, he saw crowned with laurel? Or to his
mother, the poor woman, who lately kissed her son as consul and is
now tormented by the thought of seeing him shortly stripped of all
dignity? But why do I mention his mother and his house, when the
law's new penalty robs him of house and mother and the sight and
society of all his family? So will the poor man go into exile? Where? To
the Orient, where he was many years legate, led armies and achieved
great things? But it is very painful to return with shame to a place one
has departed from with honour. Or will he hide himself at the other
end of the world, so that Transalpine Gaul, which was recently so
pleased to have him there as commander-in-chief, should now see him
again as one sorrowing, grief-stricken and homeless? Besides, what
will be his feelings when in this province he looks his brother C.
Murena in the face?
As to detail, one should note how Cicero works out the contrast
between past and present on four occasions (with the image of his
father, his mother, the Orient and the Occident). What determines
the overall structure is that the first dilemma is immediately followed
by a second, which surpasses the previous one. The miseratio is not
limited to the father's house and the mother, but on another level also
embraces the entire globe (Orient and Occident). We saw a similar
technique in the account of the mistreatment of the Roman citizen,
that we looked at earlier. In both passages a climax and a powerful
impression of variety are produced by the method of resolution into
individual elements and by an arrangement that allows what is more
important to develop out of what is less so. 95
Diametrically opposed to this is the passage from Ennius; it
imparts brilliance to its subject, not by breaking it up, but by a tight
9
Ego codd.; eius Clark.
4
95
It is another matter whether the emphasis has increased with the expansion
(Leeman p.57 concedes an increase in ubertas, but not in vis).
juxtaposition of opposites. Gracchus' balanced and yet brief formu-

lation stands midway between the condensed manner ofEnnius with
its epigrammatic suggestiveness and Cicero's unfolding climax. If
one ignores the non-literary elements (the situation and the manner
of delivery), then here too the stylistic effect depends chiefly on the
economical use of resources: pure Latinity, clear antitheses, ex-
pansiveness only insofar as it is needed for understanding, and
colourfulness only as required for the effect on the audience. Even in
this passage, which is among the most emotional in Gracchus, a
marked rationality 96 is evident in the clarity of the arrangement and
the economy97 of the means.
6. Conclusion
Norden has shown that for Roman orators, as for the poets, it was
first and foremost the Hellenistic style, and not that of classical
Greece, which was the standard. 98 What Gracchus learnt from
Diophanes of Mytilene 99 or Menelaus of Marathus 100 wasj11_its
delicacy and polish more likely to appeal to the Italic sense of form
tfian to furnish Roman gravitas and Gracchan passion101 with an
appropriate linguistic receptacle. 102 In the carefully fashioned sen-
tences Gracchus' temperament could only appear under the surface.
It had to find an additional outlet in an intense actio involving stance
and gesture, and this was what struck the audience most about
Gracchus the orator. 103 It is surely clear to any attentive reader of the
fragments that in the brief and markedly simple formulations there is
a build-up of emotion. This gives rise to a peculiar sort of nervous,
ironic tension, such as was observable particularly in the word order
of narratives that appear quite plain and factual. The man with the
96
Following Norden, E. Meyer (quoted p.35 n.5) pp.368f. saw here 'wild passion
and Asianic rhetoric'. Leo too was heavily influenced by Norden's view; however he
implicitly admits that the fragments are only rarely characterized by an elevated
manner (Geschichte der romischen Literatur p.309). To begin with, Meyer had
discussed the fragments objectively; however he changed his mind under Norden's
influence (cf. p.368).
97
One should note for example the economical use of anaphora. Gracchus also
admits deliberate verbal repetitions elsewhere, e.g. with greater frequency in fr. 44.
98
Norden Kunstprosa I p.169.
99
Cf. Cic. Brut. 104.
100
According to Cic. Brut. 100 this Phoenician helped Gracchus compose his
speeches.
101
See p.37 n.2l.
102
A somewhat different view in Norden p.17l.
103
Cic. de or. 3,2!3f. Quint. ins!. 11,3,8ff. ll5ff. Iul. Viet. p.443,2 Halm. Plut. Tib. et
C. Gr. 2,2. Cass. Dio fr. 85,2.
tuning-pipe will have been there to prevent the voice being over-
strained, for it was all too easy for Gracchus' repressed dynamism to
vent itself in vocal over-exertion. In this self-imposed rational
control we find an outward reflection of the combination of strong
emotion and keen intellect that constitutes the charm of what
remains of Gracchus' speeches. We have to recognize in conclusion
that in his case the tension between violent feeling and disciplined
style is not due to incompetence or to the inadequacy of the Latin of
1 the day, but that it is part of the man's nature.
From this point of view it is necessary to subject both the topos of
the 'passionate' Gracchus and the views on the ubertas or egestas of
his style to a more balanced and subtle examination. Can we really
set Gracchus against Cicero and vice versa? Is it a case of primitive
strength versus decadence or artistry versus crudity?
All these antitheses are misguided. As Leeman 104 saw, Gracchus'
style is varied. What we have been able to demonstrate beyond this
pointed rather to the spirit and manner in which the means are
employed than to a one-sided selection.
This is even more true of Cicero. The artistic narratio from de
supp/iciis cannot be regarded as the only type of Ciceronian
narrative. Here the elevated tone is justified because a particularly
serious case is involved. The artistic elaboration may also have
something to do with the character of a speech intended only for
reading. Otherwise Cicero too is aware that plainness can increase
the credibility of a narratio. Even so artistic a speech as the one for
Milo tells the events with marked simplicity. 105
Similar traits are also visible in these men's attitude to language.
With Gracchus, who according to Munzer 106 never ceased 'being the
great lord', a natural feeling for style, as refined by education, 107 is
part of his character as a grand seigneur; with Cicero it has become
second nature through study and self-discipline.
Cicero's superior artistic perfection is explicable in terms not only
of the stricter requirements of a changed world, but also of his
stronger literary bent. Yet may not the conversion of powerful feeling
'\into literary form also bespeak a nature that, while perhaps less
forceful, is more interested in compromise? What we have here is not
104
Cited above p.37
Quint. inst. 4,2,57[. callidissima simp!icitatis imitatio.
10 5
106 RE 2 A 2 (1923) 1397. Cf. also L. Homo Nouvelle histoire romaine (Paris 1941)
p.185.
107 Cf. Cic. Brut. 2.
glowing passion under the surface of an icy rationality, but emotion

elevated to literary form and an artistic intellectuality that con-
sciously pervades the realm of sentiment. 108
108 The ability to speak simply on lofty subjects, which we have again come to value
since the 18th and 19th centuries, belonged not only to Gracchus, but also to Cicero;
cf. the chapter on rep. below. Cicero has an infallible sense of what is appropriate in
each situation (aptum).
3
Caesar (100-44 B.C.)
I. Funeral Oration for Julia (69 B.C.)

u
\Amitae m:a~J'!!!r:_e maternum genus ab regibus ortum, pate.~num cum
diirznmortalibus _co'!£~t. nam ab Anco Marcia /Sy,J1) Ma!f_j~
Reges, qlf2_!lomin{jj:jl mater; a Venere Iulii, cuius gentis famili~
nostra. /iE-Jrgo in ·genere et sanctitas regum, qui plurimum il;tex_
homines ftllent,. et caerimonia deorum, quorum ipsi in potestate \su~tj
reges 1 \or i)v.t,Y·' ItA\), 1-/
• _) - (-t-.Y)
Of my aunt Julia the maternallineage 2 is descended from kings; her father's

is related to the immortal gods. For from Ancus Marcius come the Marcii
Reges- this was the name her mother had; and from Venus come the Julians,
the clan to which our family belongs. Thus there is in her lineage both the
inviolable dignity of kings, who are the most powerful among men, and also
the sanctity of gods, in whose power stand even kings.
Caesar- 'the greatest of mortals': 3 such was the view taken not so
very long ago in rare unanimity by democrats like Mommsen, 4
aristocrats like Gundolf, 5 and monarchs like Napoleon I and III. 6
This figure drew almost religious tones from the century that liked to
think of itself as more sober-minded than any. Here we attempt to
1 Text in ORF 4 fr. 29, p.390. Other fragments from the speeches ib. pp.383-397; cf.
also the third volume of the Teubner edition by A. Klotz (Leipzig 1927, repr. Stuttgart
1966), esp. p.l75. On the funeral oration for Julia see most recently W. Kierdorf
Laudatio funebris, Interpretationen und Untersuchungen zur romischen Leichenrede
(Meisenheim am Glan 1980) pp.ll4f. On the genre as a whole see D. Flach 'Antike
Grabreden als Geschichtsquelle' in: R. Lenz (ed.) Leichenpredigten als Quelle
historischer Wissenschaften (Cologne/Vienna 1975) pp.l-35.
2
In the present text we should note the distinction between genus (lineage as
descent) and gens (lineage as clan).
3 J. Burckhardt Gesamtausgabe (Berlin/Leipzig 1930-34) vol. VII p.237.
4 Romische Geschichte (cited above p.2 n.5).
' F. Gundolf Caesar, Gesch. s. Ruhms (Berlin 1924, repr. Darmstadt 1968).
6 Precis des guerres de Cesar par Napoleon, ecrit par M. Marchand a l'ile Sainte-
Helene sous Ia dicteede I' empereur (Paris 1836); Comment aires de Cesar, suivis du precis
des guerres de Jules Cesar par Napoleon (Paris 1872). Napoleon III Histoire de Jules
Cesar (Paris 1865/66); E.T., 2 vols. (London 1865/66). ,
54
CAESAR 55
approach it from a more human side that is not always appreciated

and that enables something of Caesar's personal charisma to be felt.
Let us first consider him as an orator.
The passage from the funeral oration for Julia owes its appeal to
the clarity of its structure and sequence of thought. 7 The basic theme
is in itself twofold: on the mother's side descent from kings, on the
father's side from gods. First of all the idea is expressed in general
terms. Here the law of increasing members is observed: the greater
length of the second colon is ensured through amplification of the
short substantive diis by means of the adjective inmortalibus. The
combination is normal (we have already met it in Cato) and so the
stylistic device is unobtrusive: this is the sort of elegance that Caesar
loves. The next sentence, which is introduced by nam, substantiates
and puts in concrete terms the statement that up to now has only been
made generally. The sequence is the same as in the first senteQc,e.~Qn
the conceptual level therefore Caesar prefers the parallelism a b a~~ii5>
the order abba observed in Cato: this is no longer the archaic style.
The main clause is followed each time by a relative clause. Again
the second of these subordinate clauses (cuius gentis familia est
nostra) is more ample than the first (quo nominefuit mater). A small
counterweight is provided by the shortness of the second part of the
main clause (a Venere Iulii).
With ergo, the third sentence draws the conclusion from what goes
before: again on the conceptual level it exhibits the structure a b. As
compared with the beginning, it adds the dignified substantives
sanctitas and caerimonia 8 and thereby creates a climax. At the same
time it picks up the structure of the second part with its relative
clauses. The novel content is that a hierarchy is expressly established
between kings and gods. In this way the tripartite structure ends with
a convincing combination of the themes that up to now have figured
separately.
Thus the structure of our text displays a threefold parallelism. The
conceptual link between the three elements is clearly marked by
means oflogical particles. The progression is made apparent through
an increase in the size (14: 19: 21 words) and weight of each
sentence. In the second sentence relative clauses are added, and in the
7 According to Kierdorf (cited above p.54 n.l) p.60 funeral orations of the archaic
and classical period did not have a prologue; consequently the text is to be seen as the
beginning of the speech.
8 Sanctitas occurs only here in Caesar; apart from this passage caerimonia occurs
only Gall. 7 ,2,2.

56 CAESAR
third abstract substantives and a conclusion that links all the themes.
In points of detail too the cola and commata are carefully formed.
Here we give in brackets at every pause the last rhythm heard: Amitae
meae Iuliae ( _v __ v_ double cretic) maternum genus ab regibus or tum
(-vv_v heroic clausula), paternum cum diis inmortalibus comunctum
"est (----double spondee). nam ab Anco Marcia sunt Marcii 9 Reges
1.-v ___ cretic with spondee), quo nominefuit mater (vvv--v fourth
paean with trochee), a Venere Iulii (-v v v_ v- first paean and cretic),
cuius gent is familia est nostra (v v v -- v fourth paean and trochee). est
ergo in genere et sanctitas regum (-V--V cretic and trochee), qui
plurimum inter homines pollent (vvv ---fourth paean and spondee),
et caerimonia deorum (- v v v- v first paean and trochee), quorum ipsi in
potestate sunt reges (-v ___ cretic and spondee).
j \ The most frequent type to appear at the end of the colon is v..:= vv~_:::
I (cretic and trochee); here the cretic can be replaced by the paeans
1
with the equivalent metrical value (first and fourth paean). Accor-
dingly the clausula types clausulas esse and esse videatur predominate,
which are also well-known to us from Cicero. In second place comes
the double cretic (or paeon and cretic); this too is a type familiar to
Cicero (clausulas fecimus). Among these very uniform rhythms the
heroic clausula (regibus ortum) holds a special place: thanks to the
epic rhythm, it impresses the noble lineage upon the ear.
Rhythmically, the !audatio is closer to Cicero's speeches than to
Caesar's own commentarii. 10 Differences of genre play an important
role in antiquity. In the speeches Caesar does not forgo ornament.
The tone is loftier, more solemn, and more musical. The inversion
of the normal word order for the sake of the clausula rhythm is
obvious e.g. in the passage familia est nostra. The vocabulary too is
more carefully chosen (ortum; sanctitas; caerimonia). Thus the
Caesar that confronts us in the speeches is in many respects an
unknown one. 11 Plutarch 12 calls him the second best orator of Rome.
9
With Marcii and lulii the question of pronunciation arises. It is true that in the
classical period the spelling -i for -ii is more common; however this is far from being
uniform practice. In Cicero -ii is very often confirmed by the prose rhythm. So in this
passage of Caesar we cannot alter the manuscript reading.
10
However in the direct speeches at Gall. 7, 76 (Critognatus) and civ. 2,32 (Curio)
there is significantly no avoidance of oratorical rhythms (this is not noticed by Norden
Kunstprosa II p.939); the point is rightly understood by L. Holtz C. Julius Caesar quo
usus sit in orationibus dicendi genere. (Diss. J en a 1913) pp. 30-40.
11
Leeman p.l58 seems to underestimate the differences between the speeches and
the commentarii. According to G. Kennedy The Art ofRhetoric in the Roman World 300
B. C. -A.D. 300 (Princeton, N.J. 1972) p.284, the speeches that have come down to us
are 'simple in composition and diction', and their style is influenced by the theory of
12
analogy. Plut. Caes. 3.
CAESAR 57
In the Anticatones, 13 a reply to Cicero's praise of Cato theY ounger,

Caesar asks that the words of a soldier should not be measured
against the art of a skilful orator, who after all has plenty of leisure:
here we have a truly Caesarian understatement.
In Cicero's judgment 14 the speeches of Caesar fulfil on the one
hand the requirement of elegantia, i.e. choice of the proper word.
Here Caesar conforms to the Atticist ideal (though one is not
therefore justified in labelling him an Atticist ). Caesar will never have
followed the bad habit of searching for synonyms just to avoid
repeating a word. This can also be observed in the present text. 15
However Cicero's further remarks take us beyond strict Atticism and
show that as an orator Caesar is not content with mere correctness:
'And when Caesar then adds to the apt choice of words the
ornaments of rhetoric, he seems to me as it were to set up well-
painted pictures in good lighting'. 16 Cicero's praise thus confirms
that in his speeches Caesar did not speak in the manner of the
commentarii.
Observation of the rhythmical ornament 17 now enables us to
answer more precisely the question whether Caesar can be described
as an 'Atticist' in the full sense of the word. Here we do not mean
grammatical Atticism (the requirement of linguistic correctness -
EAAT]VWJ.!O~, latinitas; this of course goes without saying). What is
meant is rhetorical Atticism: this was a trend that only assumed a
fully distinct outline in first century B.C. Rome and chose Lysias and
Thucydides as stylistic models. Calvus, one of the main represen-
tatives of this trend, was Caesar's enemy: this would not of course be
a decisive argument against Caesar's being an Atticist. Whereas
Norden takes Caesar's Atticism for granted, 18 more recent scholars
like J.F. D'Alton 19 and A.D. Leeman 20 are rightly cautious. In the
present speech (and in the others as welJ2 1) the rhythms alone show by
their closeness to Cicero that Caesar cannot be described as a strict
Atticist.
13 14
Ib. =Caesar ed. Klotz III p.l88. Brut. 72,252.
15
See below pp.58 and 60.
16
Brut. 75,261: cum ad hanc elegantiam verborum Latinorum adiungit ilia aratoria
ornament a dicendi, tum videtur tamquam tabu/as bene pi etas con!ocare in bono lumine.
17 Cf. also K. Deichgriiber 'Elegantia Caesaris' Gymnasium 57 (1950) pp.l12-123;
see also now D. Rasmussen (ed.) Caesar (chapters by several authors) (Darmstadt
1967) pp.208-223. Caesar also observes the clausulae in De analogia (E. Lofstedt
Syntactica II pp.307-311; Cic. Brut. 72,253).
18 Die romische Literatur (Leipzig 1954 5 p.39.
)
19
Roman Literary Theory and Criticism (New York 1931, repr. 1962) p.254.
20 21
Loc. cit. p.156. Holtz pp.24-26.
58 CAESAR
We can advance one step further, if we note the way the rhythms
are distributed. As we have seen, Caesar uses mainly a single
rhythmical pattern, which he varies only slightly. This picture of the
clausula rhythm fits in well with the symmetry in sentence con-
struction that we observed earlier. There is much about it that recalls I
I,
the almost alarming balance of Cicero's early speeches, above all Pro
Quinctio. Both great authors later freed themselves from this initial
(hyper-correctness. So here we cannot interpret the same traits as
1
'Atticist' in Caesar and as 'Asianist' in Cicero, especially of course as

both were students of Molo of Rhodes. 22
Nonetheless even this early utterance bears unmistakably Caesa-
rian traits. In regard to content alone, the speech is a document of
intellectual history. Here already, in Caesar's early period, occurs the
appeal to divine descent. 23 So Caesar not only ended up asJ! gQd, but
also began as a son of the gods. Our attempt to approach him from a
'human' side has thus confronted us unintentionally with heroic
proportions. 24 Let us continue nonetheless. Caesar's characteristic
qualities are manifested in the)anguage as well: impressive vocabu-
lary is used to denote powe1((pol/ent23'besides potestas) and dignity
(sanctitas, caerimonia); but the-crucial words are not varied: reges
(four tilnes), di (twice), genus (twice). Variety is the aim only where it
is practically justified (cf. quo nomine with cuius gentis; in the first
passage the name is in fact particularly important: Marcii Reges). So
with 'geometrical' elegance of form goes a marked sense of dignity.
What we can infer from language and style is confirmed by the direct
impression that Caesar's oratory made on his contemporaries.
Eyewitnesses speak of his dignified bearing and at the same time of
an arresting spontaneity in his word_§, that belied his abundant
experience: splendidam quandam minimeque veteratoriam rationem
dicendi tenet, voce motu forma etiam magnificam et generosam
quodammodo. 26 Deichgdiber 27 correctly recognized in our text the
disciplined quality of Caesar's style, even if in the present writer's
view the structure still betrays signs of stiffness.
22
Suet. Jul. 4,1; Holtz passim.
21
The common topos 'mention of ancestors' is given a content that is characteristic
of the speaker. Kierdorf (cited above p.54 n.l) pp.ll6 and 135 thinks the aim is
propaganda.
24
Cf. also Hor. carm. 3,1,5f.
" On the archaic solemnity of pol!eo and its stylistic value cf. R. Syme Sallust
(Berkeley and Los Angeles 1964) p.306. (In Sallust the word increases in frequency.)
" Cic. Brut. 75.261.
27
Cited p.57 n.l7.
CAESAR 59
II. Ref1ection and rapid action (Gall. 7,27) 28

Postero die 29 Caesar promota turri
I
perfectisque operibus, ·quae facere
instituerat, magna coorto imbri non inutilem hanc ad capiendum
consilium tempestatem arbitratus, quod paulo incautius custodias in
muro dispositas videbat, suos quoque languidius in opere versari iussit
et quid fieri vellet ostendit, legionibusque intra vineas in occulto
expeditis, cohortatus ut aliquando pro tantis laboribusfructum victoriae
perciperent, iis qui primi murum ascendissent, praemia proposuit
militibusque signum dedit. illi suqito ex omnibus partibus evolaverunt
murumque celeriter compleverunt.
On the following day (a tower had been moved forward and the proposed
trenchworks completed) heavy rain set in, weather which Caesar considered
particularly suitable for taking a decision; since he saw that the watch posts
were distributed rather too carelessly on the wall, he had his own men as well
do the trenchwork mote casually and issued his instructions. He made the
legions ready for battle in secret within the 'bowers'; 30 he urged them at long
last to reap the fruits of victory for so much effort, promised rewards for
those who should be first to scale the wall, and gave the soldiers the signal for
battle. They flew out suddenly from all sides and quickly occupied the wall.
1. Objective style
Thestyle of Caesar's commentarii is profoundly 'objective' and is
thus remote from that of the speeches. It is important to point to the
influence of the official language of senatorial records. One of its
features is the1epetition of a main clause word in the relative clause
(e.g. diem, quo die); this is commoner at the beginning of the Bellum
Gallicum than later. 31 In the present text we find only a refined
reminiscence of this style: perfectisque operibus, quae facere insti-
tuerat. Repetitions of this kind occurred in Cato to a much greater
extent, but there the reason for them was different: they arose from
28
Text of 0. See! (Leipzig 1961). Commentary and bibliography in F. Kraner-W.
Dittenberger-H. Meuse!, with an epilogue and bibliographical supplements by H.
Oppermann, 3 vols. (Berlin 1961 19 ). The passage is I!OW discussed by H.A. Gartner
.. Beobachtungen zu Bauelementen in der antiken Historiographie, bes. bei Livius und
·caesar Historia Einzelschriften, Heft 25 (Wiesbaden 1975) pp. 75-78,96. In general see
G. Pascucci 'Interpretazione linguistica e stilistica del Cesare autentico' ANRW I 3
( 1973) pp.488-522; J. Kroymann 'Caesar und das Corpus Caesarianum in der neueren
Forschung. Gesamtbibliographie 1945-1970 (1972)' ib. I 3 (1973) pp.457-487.
29
On postero die and similar 'natural' connectives see J.-P. Chausserie-Lapree
L'expression narrative chez les historiens latins. Histoire d'un style (Paris 1969) pp.24-
28, 29-32.
30 On the question of emendation see below p.67.
31 On this cf. also now H. Haffter and E. Romisch Caesars Commentarii de bello
Gallico Interpretationen - didaktische Uberlegungen (Heidelberg 1971) p.14.

60 CAESAR
the closeness of Cato's prose to an elevated oral languageY In

Caesar on the other hand a repetition such as the one just described is -
an ingredient of the commentarii style. 33
Another feature of the official style is frequent use ofthe ablative
absolute (which is much rarer e.g. in Sallust, Livy and Tacitus). 34
This standard element in prayers of thanksgiving and in the
triumphal inscriptions of victorious generals is already parodied by
Plautus: hostibus vic tis civibus sa/vis re placida pacibus perfectis I bello
exstincto're bene gesta integra exercitu et praesidiis (Per sa 753-755; cf.
Amph. 188f.). 35 The formulaic repetition has a certain solemnity
about it; however the ablative absolute as such is quite down-to-earth
and belongs to the language of military reports. 36 A further
characteristic of this style is oratio obliqua: though there is only a hint
of it in the present text, in general it is more popular with Caesar than
with other historians and it has a parallel in the official language of
the senatus consultum de Bacchanalibus. 37
Another aspect of the objectivity of Caesar's language and style is
his striving for elegantia, 38 which is to be understood as proprietas
verborum: thus it is not a search for a particularly choice form of
expression, but for aptness. Consequently in this text he is not afraid
of using the word~-;rus ~s often as necessary' and he does not look
for alternative wording just for the sake of variety. When he repeats
words, he is generally content with slight variations: perfectis -
facere; operibus - in opere. On the other hand the striving after
proprietas leads in this text to such characteristic forms as incautius 39
and languidius, 40 which through the parallelism evoke and sustain
32Which is to be distinguished from colloquial language, see above pp. 6f. and 16.
33On the repetition of colourless words in Caesar cf. also now P.T. Eden 'Caesar's
Style. Inheritance Versus Intelligence' Glatt a 40 (1962) pp. 74-117, esp. 83ff. The
present writer cannot accept Eden's equation of the commentarii style with that of the
annalist Claudius Quadrigarius; see the chapter on Livy below.
34
Leeman p.176, who in this writer's view correctly assesses the stylistic value of the
ablat. absol. in Caesar.
35
Cf. E. Fraenkel Plautinisches im Plautus (Berlin 1922) p.236; Element! Plautini in
Plauto (Firenze 1960) p.228, with additions pp.428f. (copious examples and good
discussion).
6
3 Cf. also E. Laughton The Participle in Cicero (Oxford 1964) p.151 and often. On
the ab1at. absol. as an 'artificial' connective see Chausserie-Lapree pp.109-124.
37
Leeman ib.
38 On Caesar's elegantia and his purism see also now Eden pp.97-106.
39 The comparative of incaute is not attested before Caesar (0. Prinz ThiL 7, 1,6
[1939] 852,56). Cicero knows only the comparative of the adjective (ib. 850, 73).
4° Cic. Att. 7,3, 11 is the only other classical instance. Later on it occurs Sen. dial.
6,9,2; benef 2,17,4; Curt. 4, 16,4; Plin. nat. 37,92; 37,94 (communication ofW. Ehlers,
ThLL Munich).
CAESAR 61
one another.
The positioning of the finite verb in this text is rational and
completely in line with Latin tradition. It stands invariably at the
end: this is in contrast to Cicero, where the emphatic initial position
and the logical middle position also play a role in the speeches and
philosophical works respectively_41 On the other hand Caesar does
vary the word order with the ablative absolute: in this text and
elsewhere he likes to put the v~rbai element in front. 42 When the
verbal element stands first, the emphasis is on functional integration
into the sentence; at the same time this achieves a certain animation
in Caesar's characteristically discreet manner, though with him it
happens so often that any one passage should not be over-
interpreted. In the same way Caesar also puts the gerundive in front
in our text: ad capiendum consilium; by means of this inversion, due
emphasis is given to an idea which in its context is rather surprising.
It is noteworthy that Caesar's address to the troops even in its
indirect form is very different from the objective style of the
surrounding text. Here we find words with emotional overtones like
aliquandtr or tantus (pro tan tis laboribus) and a well-worn but effective
metaphor: fructum victoriae perciperent. Accordingly Caesar only
employs an elevated tone when it serves his purpose. In this he is like
Cato; b-ut in line with the commentarii style and his own unadorned
manner, he is much more sparing in his use of the relevant stylistic
devices.
2. Functional approach
We have already noted that the forward position of the participles in
many of the ablative absolutes in this text emphasizes the functional
) relationship to the whole sentence. The same is true of a feature that
1is relatively infrequent in Cicero: a substantive that first stands alone
\in the ablative absolute is referred to again in the same sentence in a
different case: magna coorto imbri ... hanc ... tempestatem arbitratus.
0. Weise 43 explains this in terms of the striving for emphasis. One
41
B.J. Parten Die Stellungsgesetze des verbum finitum bei Cicero und ihre
psychologischen Grundlagen (Diss. Cologne 1922). B. Borecky 'Beobachtungen iiber
das Verbindungsglied und die Wortfolge bei Caesar und Livius' in: I. Fischer (ed.)
Actes de Ia Xlle Conference Internationale d'Etudes Classiques Eirene (Cluj 1972)
(Bucharest and Amsterdam 1975) pp.339-347.
42
In the first eleven chapters of the seventh book final position of the participle (his
rebus agitatis) is about as frequent as initial position (in our text: promota turri
perfectisque operibus) and intermediate position (in our text: magna coorto imbri); so
final position of the substantive is about twice as frequent as initial position.
43
0. Weise Charakteristik der lateinischen Sprache (Leipzig and Berlin 19094 )
p.l56.
62 CAESAR
should add that in our case it results from the attempt to analyze the
event in all its details. As an outward circumstance the start of the
rain is the cause of everything that follows; this factor must therefore
be isolated. Here internal logic has led to a construction that is
grammatically somewhat illogical.
1 This same causative factor is further isolated by the two cases of
! hyperbaton: magna coorto imbri is a self-enclosed entity where what
L is important receives emphasis by being placed on the outside. In the
words that come next the same tendency to create functional units
leads to the following hyperbaton: non inutilem hanc ad capiendum
consilium tempestatem arbitratus. This word order is mainly due to
considerations that were not aesthetic, but functional. We also
observe shortly afterwards how an event acquires independence by
means of hyperbaton: legionibusque intra vine as in occulto expeditis.
However Caesar's functional approach is seen most impressively in
the overall structure of our passage, and to this we now turn.
3. Overall structure: economy of means
I. Subject: Caesar
Ablat. absol.: what was dealt with by others or
happened by chance.
Predicative participle: what Caesar himself thinks.
Subordinate clause: what he observes (reason for what
follows). 44
Two main verbs: Caesar's instructions (content: infin.
and subordinate clause).
Ablat. absol.: external movement of the troops.
Predicative participle: Caesar's speech of exhortation.
Subordinate clause: content of the exhortation.
Two main verbs: promise of rewards (with relative
clause) and signal for battle.
II. Subject: the troops
Two main verbs: rapid execution of orders.
Here too, as in the funeral oration for Julia, we find an astonishing
feeling for parallelism and symmetry; this time however there is more
animation inside the framework. On the other hand Caesar's report
possesses the objectivity of a Catonian narrative, but combines it
with an incomparably more artistic structure. Cato gains a sense of
depth and perspective only from the subject and the scene of action,
44 At the same time the reader is told of the strategic position (on this in general see
H. Montgomery 'Caesar und die Grenzen - Information und Propaganda in den

Commentarii de bello Gallico' SO 49 [1973] pp.57-92, esp. 74).
CAESAR 63
whereas Caesar does so conceptually as well. His logical mind uses

linguistic resources very Sl]btly and deploys them functionally like
\ soldiers in a strict order of rank. The ablative absolutes express the
J \more or less automatic execution of orders and the occurrence of
1
Fxternal events, while the predicative participles are reserved for the
thoughts and speech of the general, and the main verbs for his
diecisive commands.
l.. The distribution of the subjects in this text is significant. As a man
of action 45 Caesar dominates the complicated two-fold structure
which forms the main part of our passage and which takes us from his
understanding _of the situation via the planning up to the issuing of
his orders; on the other hand the troops are the subject of the final
sentence, which communicates the rapid execution. 46
The arrangement is made clear by rhythmical means as well. At
lesser caesuras the double trochee appears several times: arbitratus;
dispositas videbat; expeditis. However at the end of the sentence the
double cretic produces a very striking caesura: militibusque szgnum
dedit. In the second sentence evolaverunt creates the clausula (of the
type: clausulas esse) and compleveru~t gives a double spondee. Even if
these rhythms came about unintentionally, their occurrence at
structurally important caesuras is further proof of Caesar's sure
instinct in using every device functionally. The cadence effect
produced by such rhythms is less directly perceptible for us than for
the Romans, who always read aloud and thereby experienced the
rhythmic quality of a text as real sound.
4. Facultas dicendi imperatoria 47
One is struck by the ·considerable difference in the length of the two
sentences. What is the significance of this? In terms of content, the
first sentence covers all deliberations and preparations up to the
1 \ signal for battle, and the short second sentence contains the
I surprising and successful attack. The one action develops from the
other. The first sentence forms the background for the second. The
first contains no fewer than four ablative absolutes, two predicative
45 The way Caesar brings himself and his name into the account is examined by
E.D. Kollmann 'Die Macht des Namens. Beobachtungen zum "unpersi:inlichen" Stil
Caesars' Studii Clasice 17 (1977) pp.45-60.
46 P.T. Eden lac. cit. sees a general affinity between Caesar and Claudius
Quadrigarius, which in the present writer:s view does not exist in a crucial area: a clear,
rational structure is lacking in the annalist (see also the chapter on Livy below). Eden
also misses Claudius' own artistic qualities.
47 Franta ad Verum 2,1,8 p.ll7 van den Hout.
64 CAESAR
participles and five subordinate clauses. By contrast the second

sentence has only two main verbs, which are connected by -que. Two
adverbs are added to indicate suddenness and speed. The verbs
themselves are also notably lively and colourful: complere 48 empha-
sizes the completeness of the success, and evolare its swiftness. 49
Overall we find here the same surprising sequence 'long-short', which
was also characteristic of Cato.
Thus the literary presentation is adapted to the subject itself in a
way that is as simple as it is effective. The thoroughness of the
preparations and deliberations appears here as one of the conditions
for Caesar's fast and successful action. In these two sentences the
secret behind Caesar's victories has, so to speak, assumed archetypal
linguistic form. Here we can actually see what Herder once
formulated theoretically as follows: 'Caesar's ease of victory is also
recognizable in his style.' 5° Quintilian had already expressed himself
in similar terms (10,1,114): tanta in eo vis est, id acumen, ea co~jo"-,
ut ilium eodem animo dixisse, quo bellavit, appareat. )~,;· 0
, 5. Candour or pose?
i Caesar perfected his natural stylistic gifts through studies to which he
/ received early encouragement: his mother Aurelia was an important
j i and educated woman. Tacitus 51 mentions her in the same breath with
'\ Cornelia, the mother of the Gracchi, and he stresses that these
1.women took personal charge of the upbringing and education of
1
~their children, which was not always the case in noble Roman circles.
Also of importance was Caesar's uncle, C. Julius Caesar Strabo: he
was an educated and witty man, who appears as an expert on humour
in Cicero's masterpiece on the orator. 52 It was Strabo who instilled in
Caesar the taste for linguistic purism and who thereby had a decisive
influence on his style, even though he himselflacked the concentrated
energy of his nephew. 53 The grammaticus who taught Caesar was the
48 camp/ere murum is noted as a peculiarity of the seventh book (G. lhm 'Die
stilistische Eigenart des 7. Buches von Caesars Bellum Gallicum' Philologus Suppl. 6
[1892] pp.767-777, esp. 769). Cf. also later civ. 3,81,1.
49 eva/are occurs in Caesar only here and 3,28,3.
50 Vom Einjluss der Regierung auf die Wissenschaften und der Wissenschaften auf die
Regierung chap. 3,25 (= Suphan vol. IX p.333). Cf. also M. Spilman 'Cumulative
sentence building in Latin historical narrative' Univ. of California Publications in
Class. Philology II (1930-1933) pp.l53-247, esp. 241: 'Caesar's writing affords the
most important illustration of the cumulative - complex sentence.'
51 Dial. 28,5f. 52 Cf. Cic. de oral. 2,23,98; 54,216-71, 291.
53 Cf. Mar. Viet. GL 6,8 on the correct spelling and pronunciation of Tecmessa.
However Caesar could not learn an energetic style of oratory from his uncle; cf. Cic.
Brut. 48, 177; see also Suet. Jul. 55,3.
CAESAR 65
famous M. AI!tonius Gnipho, who had received his education in

Alexandria. Caesar also studied on Rhodes under the same teacher as
Cicero, Apollonius Molo. 54
The contradiction in Caesar's works between apparent artlessness
and actual perfection was already felt in antiquity. No less an
authority than Cicero has expressed this in terms that cannot be
bettered. 5 5 In his eyes Caesar's commentarii deserve 'strong appro-
bation; for they are bare, upright and graceful, without any
rhetorical ornament, as if they wore no garment. But while he wanted
to make available to others material for their historical accounts, he
has perhaps done a favour to people without taste, who will want to
frizz them into shape with the curling-tongs; at all events he has
frightened reasonable people away from writing; for in historio-
graphy there is nothing more attractive than pure, lucid brevity.'
Caesar's general Hirtius, 56 a pupil of Cicero's, expresses himself in
much the same way. Both Cicero and Hirtius assume that the
commentarii are meant as a collection of material for historians. 5 7
This view also finds support elsewhere in ancient historiographical
theory. 58 Caesar certainly uses the conventional form of the com-
mentarius in an unconventional way; at all events he gives it literary
status. 59
54
Cf. p.58 n.22 above.
55 Valde quidem probandos; nudi enim sunt, recti et venusti, omni ornatu orationis
tamquam veste detract a. sed dum voluit alios habere par at a, unde sumerent qui vellent
scribere historiam, inept is gratumfortasse fecit, qui volent ilia calamistris inurere, sanos
quidem homines a scribendo deterruit; nihil est enim in historia pur a et inlustri brevitate
dulcius. (Brut. 75,262).
56 Gall. 8, praef. 4-6: constat enim inter omnes nihil tam operose ab aliis esse
perfectum, quod non horum elegantia commentariorum superetur. qui sunt editi ne
scientia tantarum rerum scriptoribus deesset, adeoque probantur omnium iudicio, ut
praerepta, non praebitafacultas scriptoribus videatur. cuius tamen rei maior nostra quam
reliquorum est admiratio; ceteri enim quam bene atque emendate, nos etiam quam facile
atque celeriter eos perfecerit scimus.
57
On the other hand the attempt has been made to derive the commentarius from a
purely Roman tradition of keeping 'official books' (F. Bomer 'Der commentarius'
Hermes 81 [1953] pp.210-250). Stylistic arguments can also be adduced in support of
this (Leeman p.176).
58 Lucian (De hist. conscr. 48) distinguishes three stages in the composition of a
historical work: 1. collection of material, 2. preliminary formulation in a 'hypomnema'

(commentarius), 3. artistic presentation. Sulla and Cicero had written such commen-
tarii prior to Caesar.
59 H. Oppermann 'Caesars Stil' (it does not deal with stylistic matters) NJbb 7 (1931)
pp.lll-125. Deichgraber Joe. cit. points out that Caesar nowhere speaks explicitly of
commentarii; however it is not clear what other name can be given to the work in Latin.
On literary art in Caesar's commentarii see now Eden pp.l07-117, though he stresses
emotion more than rationality, which in the present writer's view is at least as
66 CAESAR
This brings us to the difficult problem of self-presentation, which

we cannot evade, especially as the present text reveals much about
Caesar's character. Here we can only touch briefly on the preliminary
question of his credibility and of the objectivity of his standpoint.
Can Caesar be called a historian? To the ancient way of thinking, the
mostly unepic and unrhetorical style could be a reason for not
assigning the work to historiography. 60 In modern times the criterion
here is not so much style as bias. Mommsen 61 sees in the commentarii
an 'occasional and tendentious work drafted in the form of a military
report', but not a 'historical work in the proper sense of the word'. M.
Rambaud 62 then tried to demonstrate Caesar's tendentiousness on a
large scale. By contrast J.H. Collins 63 stressed that the most effective
form of propaganda is the truth. In fact it is almost exclusively
successes that Caesar has to report, and the few failures he describes
are not glossed over.
Nonetheless the great formal rig our of Caesar's account cannot be
equated with historical objectivity. 64 Is he always as objective as he
likes to appear? Even Mommsen felt a little uneasy about the later
account of the civil war, which puts so much emphasis on Caesar's
important. It has long been recognized that in Caesar emotion becomes more apparent
in the course of time. and also that the commentarii style in the first book of the Bellum
Galli cum is very conservative. On the Bellum Galli cum as a work of literature in which
the material has been deliberately arranged, see H. Montgomery (cited above p.62
n.44) p. 74. On Caesar's development as a writer see Haffter-Romisch (cited above
p.59 n.31) p.l4. A shift from commentarii style to historiographic style is denied by W.
Gorier 'Die Veranderung des Erzahlerstandpunktes in Caesars Bellum Gallicum'
Poetica 8 ( 1976) pp. 95-119, esp. 95-98 (rather it is a change in perspective from a
personal view to an authorial standpoint).
6
° For all his recognition of Caesar's merits, Cicero himself had a different
historiographical ideal (Herodotus, Theopompus, !socrates; Leeman pp.l68-197).
Caesar naturally has a complete grasp of the basic principles ofhistoriography (on this
see most recently H.A. Gartner [cited above p.59 n.28]). Yet in other historians the
epic and rhetorical elements are far more prominent.
61
R6mische Geschichte III p.616.
62
L'art de Ia dfformation historique dans !es commentaires de Cesar Annales de
l'Universite de Lyon, Lettres 3,23 (Paris 1953).
63
Propaganda, ethics and psychological assumptions in Caesar's writings (type-
written diss. Frankfurt/M. 1952); cf. also later reviews of this author in Gnomon. I d.
'Caesar as political propagandist' ANR WI 1 ( 1972) pp. 922-966 (self-presentation, not
self-defence: 940f.). That Caesar had to defend his policy in Gaul is also doubted by H.
Montgomery (cited above p.62 n.44) pp.80f.; however he thinks the aim could have
been propaganda in the political struggle with Pompey (ib. 82). Cf. also F.-H.
Mutschler Erziihlstil und Propaganda in Caesars Kommentarien (Heidelberg 1975).
64
It is no solution of the problem to say that 'factional history is all there is' (H.
Frankel 'Uber philologische Interpretation am Beispiel von Caesars Gallischem
Krieg' NJbb 9 [1933] pp.26-41, esp. 39); however in many respects Frankel anticipates
Collins' position.
CAESAR 67
magnamm1ty and love of peace; however he is able to spare his

declared favourite by appealing here to general human experience:
'in Caesar's soul, as in every other, the time of hope was a purer and
fresher one than that of fulfilment'. 65
What can our own interpretation contribute to this debate? In the
text before us Caesar's generalship is revealed in exemplary fashion.
Did the author have this effect in mind? The question can never be
answered with complete certainty; however it does seem to be
supported by two minor points that Meusel 66 was discerning enough
to notice, even though in the present writer's opinion he drew the
wrong conclusion from his observations when he changed Caesar's
text. Caesar has the legions make themselves ready for battle 'in
secret in the bowers'. Meuse! realizes that there would not have been
enough space in the bowers for all the legions, and pronounces the
words spurious. He also finds ex omnibus partibus suspicious, for the
city was only accessible from one side. Even if we think Meusel's
criticism is pedantic, it must be admitted that Caesar has expressed
himself in a way that is at least vague and misleading as regards the
facts. But even on the assumption that there was something actually
false in Caesar, the text still ought not to be changed. For Caesar
achieves two things by means of the additions that have caused
criticism. On the one hand mention of the bowers illustrates in an
exemplary manner the secrecy of his preparations; on the other ex
omnibus partibus illustrates the sudden as well as sweeping character
of the attack. Caesar cannot be said to have downgraded historical
truth; however, the 'higher truth' of self-presentation has led him to
interpret it in his own particular way and through slight exaggeration
to invest it with general significance. 67
What is true of the content however applies also to the mathe-
matical clarity and functional precision of Caesar's style: in both we
are justified in seeing not only candour, but also a pose.
65 66
R6mische Geschichte III p.616. Critical appendix, p.578.
67 Minor exaggerations for the sake of emphasizing his own speed and stratagems
are also noted by T. Feller Caesars Commentarien iiber den Gallischen Krieg und die
kunstmiissige Geschichtsschreibung (Diss. Breslau 1929). Such traits should not
however be overstressed (as Feller himself appreciates). Caesar is not presenting the
facts so much as his consilia (H. Oppermann in: D. Rasmussen [ed.] Caesar
[Darmstadt 1967] p. 522 with a reference to Hirt. Gall. 8 praef. 7). In any case the 7th
book of the Bellum Gallicum is in many ways already close to the more emotional style
of the Bellum Civile (K. Barwick Caesars Bellum Civile Berichte iiber die Verh. der
Sachs. Akad. der Wiss. zu Leipzig, phil.-hist. Kl. 99,1 [Berlin 195l]pp.l70f., 174). Our
text can be seen in this respect as an interesting transitional stage between an austere
and a more varied manner.
4
Sallust (b. 86 B.C.)
I. Ambition 1
Ambitio multos mortalis falsos fieri subegit, aliud clausum in pectore,
aliud in lingua promptum habere, amicitias inimicitiasque non ex re, sed
ex commodo aestumare, magisque voltum quam ingenium bonum
habere. haec primo paulatim crescere, interdum vindicari; post ubi
contagio quasi pestilentia invasit, civitas inmutata, imperium ex
iustissimo atque optumo crudele intolerandumque factum.
Ambition forced many people to become false, to lock up in their hearts
something other than they confessed with their tongues, to judge friendship
and enmity not objectively, but according to advantage, and to have a decent
face rather than a decent character. This attitude grew at first only gradually;
now and then measures were still taken to stop it; but then when the infection
spread like a plague, the citizen-body changed, and an extremely just and
good government became a cruel and intolerable one.
1. Phonetics, vocabulary, syntax

Let us begin with details. Forms like aestumare and optumo are
striking phonetically. Whereas Caesar makes the i-spelling the rule
and Cicero also generally keeps to it, Sallust preserves the more old-
/ / fashioned spelling. 2 So already in the treatment of orthography we
I notice an archaizing trend.
' In the choice of vocabulary too there are some unusual features.
Although the combination multi mortales is also found occasionally
in Cicero, 3 even ancient readers object when morta/es is used in prose
1
Sal!. Catil. 10,5f. Text of A. Kurfess (Leipzig 1957 3 ; repr. 1981). On the translation
cf. also W. Schone Sallust, Die Verschworung des Catilina lat.-dt. (Munich 1941).
Commentary by K. Vretska C. Sallustius Crispus, De Catilinae Coniuratione 2 vols.
(Heidelberg 1976) (on our passage I pp.216-219). A review of the most recent
scholarship in C. Neumeister 'Neue Tendenzen und Ergebnisse der Sallustforschung
(1961-1981)' Gymnasium 93 (1986) pp.51-68.
2
We should not assume any difference in pronunciation. It is generally thought
that the sound resembled German ii.
3
On meaning and stylisticvaluecf. J.P. KrebsAntibarbarus 7th ed. by J.H. Schmalz
(Basel 1905; repr. 1962) vol. II pp.104f.: 'Sallust and others use mortales simply for
homines "studio quodam gravitatis et magnificentiae" without any stress on the basic
idea.'
68
SALLUST 69
to mean 'men'. Here we can compare Gellius 13,29: while a passage

of Claudius Quadrigarius is being read out, one of the listeners takes
exception to the expression mortalibus muftis: pro 'hominibus muftis'
inepte frigideque in historia nimisque id poetice dixisse. Fran to on the
other hand defends the expression: 'multorum hominum' appellatio
intra modicum quoque numerum cohiberi atque includi potest, 'multi'
autem 'mortales' nescio quo pacto et quodam sensu inenarrabili omne
jere genus, quod in civitate est et ordinum et aetatum et sexus
conprehendunt; quod scilicet Quadrigarius, ita ut res erat, ingentem
atque promiscam multitudinem volens ostendere 'cum muftis morta-
libus' Metellum in Capitolium venisse dixit EJ.HpanK6:mpov quam si
'cum muftis hominibus' dixisset. 4 In fact the statement that mortales
has a wider meaning than homines applies even more to Sallust than
to Quadrigarius. However the slight difference of meaning does not
seem sufficient to the present writer to explain Sallust's usage fully.
For our passage (in contrast to Cicero) is not an isolated case, where
the use of a special word could be explained on semantic grounds.
Sallust simply has a general preference for mortales in the sense of
homines. Could this 'unnatural' shift, which finds many parallels in
Sallust's language, 5 be due to reasons that are not semantic but
stylistic? The key is again provided by Gellius. The quotation he
preserves from Quadrigarius shows that with the choice of this word
Sallust is following a historiographical tradition. We can add a
quotation from Cato. 6 What stylistic nuance does the use of this
word have for the historian? The alliterative combination ofwords 7
4
Gel!. 13,29,4.
' Sallust likes to change or replace current expressions; cf. R. Syme Sa/lust
(Berkeley and Los Angeles 1964) e.g. p.263, and also K. Latte in: V. Posch! (ed.) Sa/lust
Wege der Forschung 94 (Darmstadt 1981') pp.401-460, esp. 410ff. Sallust loves
Catonian words (see below pp. 76f. and n.6); to these should be added the ordinary-
looking existumo, whose frequency in Sallust is left unexplained by Syme (p.307; on
this cf. our discussion of the preface to Cato's agr. ).
6
Cf. in addition Cato oral. fr. 58 Malcovati 4 : videre multi mortales. However for
Sallust the influence of Sisenna will have been even stronger (fr. 123 Peter); this is
pointed out by G. Bruennert De Sallustio imitatore Catonis Sisennae aliorumque
veterum historicorum Romanorum (Diss. Jena 1873) p.23; here he also collects the
passages where Sallust uses mortales without the addition of multi (on Sisenna's
influence cf. now also W. Richter 'Der Manierismus des Sallust und die Sprache der
romischen Historiographie' ANR WI 3 [1973] pp. 755-780, esp. 774-778). Lebek p.255
(cited p.24 n.91) has made a case for thinking that in Claudius Quadrigarius the
expression multi mort ales lacks any suggestion of solemnity; this is no longer true of
Sallust.
7
On alliteration cf. E. Skard Ennius und Sa/lust ius (Oslo 1933) pp. 72-75; id. 'Zur
sprachlichen Entwicklung des Sallust' SO 39 (1964) pp.13-37 (28-35 for collection of
material and remarks on the use of alliteration).
70 SALLUST
calls to mind epic. It is in fact found there as far back as Naevius. 8 The
word morta!es is therefore chosen mainly on stylistic grounds. Here
the pursuit of OEJ.lV<hrv; is linked very closely with Sallust's
conception of genre.
A similar procedure can be observed in the contrast9 of pectus and
lingua, where both words go beyond the object itself and to our
modern way of thinking stand in place of more abstract notions
('heart', 'speech'). Skard has shown that Sallust's figurative use of the
word pectus has its origin in the high poetry of tragedy and epic. 10
Here too the choice of vocabulary reveals the same trend towards
poetic language. Only patient investigation can alert us to each
particular instance. On the whole Leeman is right when he stresses
that Sallust is much more 'poetic' than a modern reader can sense
directly. 11 While there is no reason to doubt the value of ancient
literary theory for our understanding here (it is stressed by Leeman in
this connection), attention should be drawn to the perhaps even
greater importance of detailed lexical and stylistic observation.
Parallels from other authors do not of course explain the special
force that each word gains in its particular context in Sallust. 12 What
would be an acquired mannerism in an average writer, leads very
often in his case to a rediscovery of the original meaning of each
word. This is tantamount to saying that Sallust's pictorial language
often enthralls the reader with the directness of poetry. Here we
perhaps have the source of the serious misconception held by the
great classical scholar Wolfflin, who wanted to interpret Sallust's
language as 'the vulgar Latin of democrats'. 13 The 'secondary
8 Naev. fr. 5,1 Morel = 6,1 Str.; mortales also occurs in Enn. ann. 23; 409 V.
Transmission of epic elements by earlier historians is now also posited by E. Skard

Sa/lust und seine Vorgiinger SO fasc. supplet. 15 (Oslo 1956).
9
On antithesis as a Sallustian pattern of thought cf. Latte pp.402; 405.
10 E. Skard Sa/lust und seine Vorgiinger p.47. E. Norden Ennius und Vergilius
(Berlin/Leipzig 1915) p.54 n.1 tries to show that in Cat. 10,5 we have an inversion of a
trochaic tetrameter aliud habet in lingua promptum, aliud clausum in pectore, and thus a
direct quotation of poetry. Further conjectures as to poetic models are given in
Vretska ad Joe.
" I p.184. On the closeness of historiography to poetry cf. Quint. inst. 10,1,31. On
the stylistic value of poetic words cf. Cic. de orat. 3,38, 153; orat. 24,28. Sallust's
assumptions about style are thus the exact opposite of Caesar's (Caesar de anal. ap.
Gell. 1,10,4 =Caesar ed. Klotz III pp.178f.).
12
One example is the plural amicitiae, which is supported in our text by inimicitiae.
For the plural of other abstracts cf. S.L. FighieraLa linguae /agrammatica di C. Crispo
Sal/ustio (Savona 1896) pp.91-97.
13 'Bemerkungen iiber das Vu1garlatein' Philo/ogus 34 (1876) pp.137-165, esp. 146f.
Cf. W. Kroll 'Die Sprache des Sallust' Glotta 15 (1927) pp.280-305, esp. 280 with n.3.
SALLUST 7!
directness' that is the product of an elevated literary language has

been confused here with unrefined 'primary directness'.
Subegit too is striking as an item of vocabulary. The commentators
generally explain it as a Sallustian synonym for cogo. 14 Use of the
prefix sub- does however seem to have an intensifying effect as
against normal cogere: the appeal to the hearer's imagination is
stronger. This effect is related to that of the concrete and visual words
we lookea at in the previous section. Each of these devices gives the
language increased spontaneity and vigour. At the same time these
r__qualities are combined with a refined elegance.
1 Greater liveliness is also the effect produced by the historic
~- infinitive, which should really be called the descriptive infinitive:
haec primo paula tim crescere, interdum vindicari. Sallust is especially
fond of the construction: his partiality also increases significantly
from the Catiline to the Jugurtha. 15 We have already seen that the use
of the historic infinitive goes right back to Cato. 16 The rarer passive
form in our text (vindicari) is attested in the Latin historiansY The
way Sallust assimilates this traditional feature of the historiogra-
phical style shows that he did not take over the descriptive infinitive
as something rigid and fixed, but that there was a natural affinity
between the writer's mode of expression and his thought: this is
especially clear if we bear in mind the increased use of the device in
the later work. An average writer would probably have used the
imperfect in our passage. What does Sallust achieve by means of the
infinitives? The sentence becomes both more concise and more
1
general; above all however the infinitival form detaches it from the
surrounding account. This produces the impression of unconnected-
ness that Sallust seeks to achieve with the most diverse means, and
that has made a substantial contribution to his reputation for
14 A more precise definition is given by F. Kritz in his ed. of Sallust (Leipzig 1856)
ad Joe.: subegit non plane idem est, quod coegit, sed eo adegit, ut fierent; cf. Catil. 51, 18;
lug. 24,2; 31,4; 44,4. The word occurs with this sense in comedy and later in Livy,
Tacitus and Curti us (cf. e.g. OLD s.v.); the infinitive construction is also found in Livy
and Vergil (Liv. 9,41,5; cf. Weissenborn-Mi.iller on Liv. 1,5,6; 5,24,11; Verg. Aen.
5,794.)
" Viz. almost twofold; on the question cf. now B. Hessen Der historische lnfinitiv im
Wandel der Darstellungstechnik Sallusts (Frankfurt/M. etc. 1984), summary p.l47. In
the Catiline the historic infinitive still serves to intensify a word, whereas in the
Jugurtha we see it becoming a mannerism.
16 Cf. above p.30 n.ll5.
17 Claud. Quadr. fr. 45 P 2 noctu multa do mum dimitti; Sisenna 120 milites ... civitate
donari. For examples from Sallust cf. Fighiera pp.l84f.

72 SAL LUST
'brevity'. 18
If we turn now to sentence structure, we find that this is most
profitably considered in connection with the sequence of ideas. We
ought therefore to follow the text.
2. General survey: sentence structure and sequence of thought
l Two sentences earlier Sallust pointed to the fatal consequences of
1 ambition and greed; however he then treats the two points in reverse
order. 19 This sort of chiastic arrangement of ideas is familiar to us
from Cato, who is important for Sallust in other respects as well. The
passage selected opens with the treatment of the second point. The
first sentence begins with the statement that ambition has made
many mortals dishonest (this is not a gnomic perfect, but indicates a
specific historical event). The thought is first of all expressed in
general terms and then developed from three points of view. The first
is the contradiction between attitude and words; this is set out
antithetically: aliud clausum in pectore, aliud in lingua promptum
habere. If one ignores aliud with its structural function, the arrange-
ment is chiastic. We shall have to come back later to the importance
of chiasmus in Sallust. The second aspect is social behaviour and the
principles governing it. The arrangement of this section is strictly
parallel: amicitias inimicitiasque non ex re, sed ex commodo aestumare.
The second element here is each time longer than the first. Thirdly
there appears in somewhat modified form the contrast mentioned at
the beginning between outside and inside: magisque voltum quam
ingenium bonum habere. Here again, as in the previous sentence, there
is parallel arrangement of two contrasting words.
If we ask how the parts of the sentence we have looked at are
connected, we find that they follow each other asyndetically. 20 The
third part ends with the same word as the first (habere); the centre
18 The ancient testimonia on Sallust's brevitas are in Leeman II p.435 n.86. J.C.
Scaliger Poetices libri septem (Lyon 1561) p.l96 remarks very appositely that we can
only speak of brevitas in relation to Sallust's style, whereas his presentation of causes
and background and his psychological motivation show a marked amplitude
(opulentus). Scaliger gives an accurate characterization of Sallust's style: non
circumducit, non interponit; asyndetis utitur, omittit verba ... The historic infinitive
might also be mentioned in this connection; often the abrupt effect is further
accentuated by the asyndetic juxtaposition of several infinitives (as in the present
passage). Perhaps one can add that Sallust also aims for extreme brevity when stating
facts; however he becomes fuller when he is dealing with certain external features that
reveal some important inner truth.
19
Announcement of the themes (10,3) a) imperi cupido, b) pecuniae cupido.
Treatment: 10,4: b; 10,5-6: a; 11,1-2: a; 11,3: b.
20 On asyndeton in Sallust cf. also p.Sl n.50.
SALLUST 73
part too closes with an infinitive (aestumare). In each of the three parts
there is an antithesis; however the words that articulate it are different:
[?rst aliud- aliud, then non- sed, and finally magis- quam. The word
order too provides variety. In the first part it is chiastic (apart from
the parallel arrangement of aliud - a!iud). In the second and third,
parallelism predominates: in the second part the more creditable
aspect is put first (non ex re), but in the third it comes last. Thus it is
not only -que, but also the positioning of the significant words that
combines the second part with the third to form a unity and sets this
in independent contrast to the first part, which likewise has an internal
chiastic arrangement. So the basic idea moves in a shorter and a
longer progression according to the law of increasing members. The
second of these elements is again subdivided, this time according to
the principle of rhetorical 'decrease' which we also noticed in Cato:
Sallust finishes with a relatively short and impressive clause, which
vividly summarizes the whole sentence: magisque voltum quam in-
genium bonum habere. The thought has thus returned in the archaic
manner to its starting point, though in a more concrete form.
In the following sentence haec will cover both avaritia and ambitio;
we may compare sentence 3 earlier in the chapter: igitur primo imperi,
deinde pecuniae cupido crevit; ea quasi materies omnium malorum
fuere. Sallust now picks up this last sentence and gives a closer
definition of the kind of growth meant. Stress is therefore put on the
adverb paulatim. Despite the parallelism the addition that follows
does not introduce a contrast to the whole idea, but only a small
qualification. The real contrast is delayed (post, cf. primo ). But before
the main idea is expressed, there is further retardation through a
temporal subordinate clause, which compares the spread of the evil
to that of an infectious disease. The conclusion first states the change
in general terms (civitas inmutata) and only then is the collapse finally
described by means of two pairs of adjectives. The formulation is
again antithetic; significantly two superlatives are capped by two
positives. The reason for this is not just a general aversion towards
the superlative on Sallust's part, nor simply an attempt to avoid too
close an echo of 5,9 (ut paulatim inmutata ex pulcherruma atque
optuma pessuma acflagitiosissumafacta sit); rather it is due in the first
place to an awareness that the positive can sometimes have a stronger
impact than the superlative. Accordingly Sallust chooses very strong
adjectives here, the second of which cannot be intensified: crudele
intolerandumque. 21
21 Cf. ThiL 7,2,1 (1956) 24,59f.
74 SALLUST
3. Symmetry and asymmetry

What is the result of this examination of the sentence structure? Like
the whole of Sallust's work, our text is marked by the interpenetration
of two structural principles, one symmetrical and the other asym-
metrical. In Sallust's case we must give each of these two principles its
due and not try to play off one against the other. 22 Parallelism and
chiasmus are symmetrical structures. Several times in our text we
have found parallel construction. Provided the principle of incon-
cinnity is not given undue importance, a careful reading of Sallust
will always yield numerous cases of sentence parallelism. 23 Chiasmus
appears to be the opposite of parallelism, but in some respects it is
analogous, for here too symmetry is involved. It will be sufficient to
recall one striking example: aliud clausum in pectore, aliud in lingua
promptum. Individual studies have revealed that with a/ius - a/ius
chiasmus is typical of Sallust. 24 Doubts have been expressed whether
this involves deliberate employment of an artistic device; however it
is certain that there is increased use of chiasmus in the Jugurtha as
compared with the Catiline. 25 As with the historic infinitive, there is
no reason why deliberate technical use (which is what we have here)
should be incompatible with using the device simply because it is
appropriate. That Sallust also arranges quite large sections chiasti-
cally26 is characteristic of his thought and his style of presentation.
Latte 27 has noticed 'how in Sallust purely linguistic and formal
antithesis has its correlative in a peculiar feature of the train of
thought, which casts the listener restlessly to and fro between
opposites'. When words are grouped chiastically, the tension
between contrasting words is increased by their position. 28 This is
true regardless of whether the structural feature involved is classified
technically as a rhetorical device or not: the question is largely one of
terminology.
Let us now turn to the cases of asymmetry. This trait is in fact.
particularly striking in our text and we are quite justified in
22
Cf. A.D. Leeman in his review of Syme's book on Sallust in Gnomon 39 (1967)
p.61, against Syme p.265 and referring to Orationis ratio pp.356ff. A similar view had
been taken earlier by A.W. Ahlberg Prolegomena in Sallustium (Giiteborg 1911)
pp.l73-182 and by C. Zander Eurythmia II/(Leipzig 1914) pp.265ff.
23 Konrad Meyer Die Wort- und Satzstellung bei Saltust (Programm Magdeburg
1880) p.l6 rightly sees parallelism as one of the 'main principles' behind Sallust's word
order.
24
R.B. Steele Chiasmus in Sallust, Caesar, Tacitus (Diss. Northfield, Minn. 1890).
The most important points about chiasmus are well summarized by Kroll p.300.
25 Steele p.13. 26 27
Cf. Meyer pp.2lf. Lac. cit. p.405.
28
Opposita iuxta se posita magis exsplendescunt. Kiihner-Gerth p.602 §607,2.
SALLUST 75
subjecting it to a closer examination than the cases of symmetry. 29 In

this respect the second sentence enables us to make a whole series of
interesting observations. To begin with, the connection with the
preceding sentence is unexpected, if not actually misleading. Up to
now we have been dealing with a series of infinitives, which were all
dependent on the main verb subegit. Two more infinitives now
follow: surprisingly however they are independent. In conjunction
with the reduced length of the sentence, this produces a quicker pace.
So with great subtlety Sallust appears to be continuing the series of
infinitives, while adding a new function to this construction in the
manner of an 'enharmonic change'. A transition occurs that forces
the reader to make a mental readjustment: this is precisely what
Sallust seeks to achieve by the break in symmetry.
However what follows is also rich in syntactical surprises. The
parallelism of
prima paulatim crescere interdum vindicari
does not express a contrast. 30 The antithesis is postponed and only set
out in a much longer final sentence: post ubi contagia quasi pestilentia
invasit, civitas inmutata, imperium ex iustissumo atque optumo crudele
intolerandumque factum. This final sentence does not pick up the
construction already begun (historic infinitive); instead this is carried
on by an elliptical passive that presents the idea from a new angle:
inmutata, factum. Here too the pursuit of variety is observable on a
small scale: two superlatives (iustissumo atque optumo) 31 are con-
fronted and capped by two positives: crudele intolerandumque. The
importance of this phenomenon has already been discussed. The
variation even extends to the conjunctions: on the one hand at que, on
the other only -que. The different choice of language makes the
reader think. At the same time the various forms each express a
different shade of meaning. The historic infinitive denotes a state that
is thought of as continuing, and thereby provides a background for
the following sentence. This sentence describes a change, and for this
it uses the perfect: the change is thus seen as completed. So in both
cases ofsyntactic variation the change in form corresponds to a new
emphasis or shift in the content. Asymmetry thus enables Sallust to
29
On Sallust's inconcinnity cf. Latte p.410 (on Latte see H. Dah1mann Gnomon 12
[1936] pp.255-260).
30
On the uniqueness of this passage in regard to the historic infinitive cf. Hessen
p.61.
31
Such duplications are typical of Early Latin. On this cf. Latte p.407, with a
psychological explanation based on Sallust's character.
76 SAL LUST
·express the finest shades of meaning.

4. Catonian and Sallustian elements
In several cases we have been able to explain lexical and syntactic
peculiarities by referring to the language of Catonian historiography.
Sallust's language is shaped to a large extent by Cato's achievement
and personality. Cato's sententious 32 power and brevitas influence
Sallust, just as his ubertas influences Cicero. Sallust loves Catonian
phrases and structures. His habit of 'breaking off is also Catonian
(cf. Sen. epist. 114,17); sometimes this is more a case of stylistic
brevitas than of the concision associated with Caesar, though
occasionally it surpasses even Thucydidean succinctness. 33 With this
comes a new tone, that we need to characterize by way of conclusion.
If we turn to the content of our text, we are again confronted with
Cato. We have already indicated how Sallust idealizes the distant
Roman past in his use of superlatives. For him ambitio along with
avaritia is a destructive power that he compares to an infectious
disease. Even so, ambitio remains closer to virtus, since it spurs men
on to great achievements (cf. 11,1 ). 34 In the text under consideraton
Sallust's attitude is 'censorial', and it points to the prototype of the
Roman censor, Cato the Elder. It is no accident that Sallust's basic
idea has already been encountered in Cato's speech for the Rhodians.
The destruction of Carthage ushered in the decline of Roman virtus,
because thenceforth Rome was without a worthy opponent. 35 Cato
had not however limited himself to moralizing; and in the same way
Sallust is not content simply with edifying speeches. He enquires into
the causes of Rome's greatness and decline; and his ethical ideas need
to be seen in this context and subordinated to the higher principle of a
search for historical truth. 36 Sallust sees Cato as stylist, as censor, and
as historian; for him tradition is embodied in Cato's person (so we
cannot play down his imitation ofCato by seeing it simply in terms of
32 Cf. aliud clausum in pectore, aliud in lingua promptum habere; non ex re, sed ex
commodo; magisque voltum quam ingenium bonum habere.

33 Sen. contr. 9,1,13f. =Sail. hist. I, fr. 49,24 Kritz= 55,24 Maurenbrecher.
34 On the problem of ambitio cf. L. Schmiidderich 'Zur Staats- und Gesellschafts-
theorie in Sallusts Catilina' Der altsprachliche Unterricht 18,3 ( 1975) pp. 65-70, esp. 68:
'For while theoretically speaking ambitio may belong to the adiaphora (nam gloriam,
honorem, imperium bonus et ignavus aeque sibi exoptant 11,2), and while it may in fact
be propius virtutem (11,1), it is still for Sallust more ambitio mala (4,2) and a vitium
(II, I), since it turns out increasingly to be a misguided activity.'
35 Cf. Sallust Catil. 10,1; Cato HRR I 83 Peter (citation of Scipio).
36 Cf. K. Buchner Sallust (Heidelberg 1960) passim; on the way in which the
structure 'embodies truth', pp.252f.

SALLUST 77
literary theory and saying that it merely enabled him to compete with
Thucydides)Y When Sallust calls Cato the 'most eloquent Roman' 38
at the start of his Histories, he means this in all seriousness. Allusions
to Cato are found throughout his works. 39 However the essential
point is that Sallust is trying, in his capacity as a stylist, as an observer
of the basic Roman values, and as a historian, to create something
analogous in his own time. Yet Sallust does not restrict himself to
blind imitation of Cato.
If we turn from the content back to Sallust's style, we can see in
detail how the devices of early Latin historiography are made to
express ideas that are modern, kaleidoscopic and extremely subtle.
Sallust is no longer archaic, 40 nor is he over-enthusiastic in his
archaizing. In an age that had already seen the classical style of
Cicero, he turns archaism into a way of giving apt and novel
expression to thought that is sophisticated and capable of surprising
turns as well as imperceptible nuances and shifts. The result is
Sallust's unmistakable and highly personal style; here individual
words keep appearing in a new light owing to the unusual method of
combination, and they thereby acquire a lustre that is close to
poetry. 41 The following text will show what effect this has on the
narrative structure.
37
Thus Leeman p.l82.
3
' Quoted above p.24 n.92.
39
On Sallust's imitation of Cato see Augustus ap. Suet. Oct. 86; Pollio ap. Suet.
gramm. 10; E. Skard Sallust und seine Vorgiinger pp. 75-107; S.L. Fighiera pp.llf., 13-
5; F. Deltour De Sallustio Catonis imitatore (Paris 1859).
40 In some cases Sallust's thought patterns are close to archaic techniques of
composition (Fankhanel p.269, although he exaggerates the 'primitive' features in

Sallust; he sees a loss of power in the Jugurtha). Sallust is not himself archaic; rather he
archaizes because this is intrinsically appropriate. Besides genuinely archaic style and
superficial archaizing, there is also 'genuine archaizing': this is what Sallust achieves.
He evolves his archaizing style gradually; see e.g. E. Lofstedt Synt. II pp.290f., 412f.,
422. Leeman p.186 has some very good observations on the way Sallust spiritualizes
old-fashioned Roman concepts.
41
Evidence of his 'poetic' dignity is also to be found in the avoidance of a good
many current political slogans such as: optimates, popu/ares, consensus, auctoritas,
dignitas, felicitas, gravitas (Syme pp.256f.).
.78 SALLUST
II. Triumph through treachery 42

Sed nocte ea, quae proxumafuit ante diem conloquio decretum, Maurus
adhibitis amicis ac statim inmutata voluntate remotis ceteris dicitur
secum ipse multum agitavisse, voltu <colore motu> corporis pariter
atque animo varius; quae scilicet ita tacente ipso occulta pectoris
patefecisse, tam en postremo Sullam adcersi iubet et ex illius sententia
Numidae insidias tendit. deinde ubi dies advenit et ei nuntiatum est
Iugurtham haud procul abesse, cum paucis amicis et quae store nostro
quasi obvius honoris causa procedit in tumulum facillumum visu
insidiantibus, eodem Numida cum plerisque necessariis suis inermis, uti
dictum erat, adcedit, ac statim signo dato undique simul ex insidiis
invaditur. ceteri obtruncati, Iugurtha Sullae vinctus traditur et ab eo ad
Marium deductus est.
But in the night immediately before the day that was fixed for the parley, the
Moor summoned his confidants, immediately altered his decision again, and
removed the others; then he is said to have debated long within himself, while
expression, complexion and gestures changed according to the momentary
state of his mind; this of course revealed, despite his silence, the secret
thoughts in his breast: Finally he has Sulla sent for after all and in accordance
with the Roman's plan he sets an ambush for the Numidian. As soon as day
broke and he learned that Jugurtha was no longer far away, he sets out with a
few friends and the Roman quaestor, as if he wanted to go and meetJugurtha
in order to honour him, and he arrives at a hill that was very clearly visible to
the men in the ambush. The Numidian comes to the same place with very
many of his confidants, who carried no weapons 43 as arranged, and at a sign
he44 is immediately ambushed from all sides at once. The others were cut
down; Jugurtha is handed over to Sulla in chains and brought by him to
Marius.
1. Form and content

The preceding section closed with the words: illi pariter /aeti ac spei
bonae p/eni esse. Following the gay mood of this conclusion, our
passage opens with sed: 45 this is typical of Sallust's delight in
contrast. First the word nocte indicates the time; then a relative
clause adds that it is the night immediately before the date fixed. The
additional information contains a reference to the future. The
imminence of the parley is further emphasized by the adjective
42 Sal!. lug. 113,3-7. Commentary: C. Sa/lust ius Crispus, Bellum lugurthinum
erlautert und mit einer Einleitung versehen von E. Koestermann (Heidelberg 1971) (on
our passage pp.384- 386).
43 Inermis will here be from inermus.
44 Sallust construes invadere with the ace.: Fighiera pp.127f.
45 On sed cf. K. Buchner Sallust p.139.
SALLUST 79
proxuma. This addition, which from a logical point of view would be

superfluous, shows that Sallust wants to increase the tension. After
the first part of the sentence has in this way related the scene to the
coming negotiation, Sallust inserts a 'retardation': this is the
vacillation of the Moorish prince before his decision is made. Here
we close in progressively on the psychological action. It is revealed
first in the capricious behaviour of the Moorish king towards his
confidants, whom he summons, only to send them away again
immediately. His indecision is thus at first delightfully reflected in the
movements of this group of people. Now the king is alone with his
restless thoughts. The author closes in on him: his mental state is now
expressed through features, deportment and gesture. Here too
psychology is not yet represented directly, but in bodily reflexes.
However the last stage is decisive: quae scilicet ita tacente ipso
occulta pectoris patefecisse. Some editors have deleted this last
sentence. 46 Yet it is not a superfluous gloss, but the necessary final
phase of the movement observed so far; we have now closed in
completely and are no longer concerned with observation of
externally visible gestures or crowd movements, but with what
happens in the mind itself (occult a pectoris). The importance this has
for Sallust is shown by his choice of wording: we have already
established that pectus is a poetic expression. The syntax too is
carefully chosen: the construction occulta pectoris is reminiscent of
Greek. Sallust employed such 'Grecisms' with increasing frequency
in the course of his development; 47 his style owes much of its dignity
to this bold handling of syntax. The position too is emphatic: occulta
pectoris stands impressively at the end. This is the nub and goal of the
centripetal movement developed in the first sentence: 48 night -
conflicting instructions to people round about - facial expression -
silence and inmost thoughts. The sentence thus develops one single
motif on different levels: vacillation between opposite poles. At the
same time the account itself is permeated with opposites. They result
partly from the subject-matter: adhibitis ami cis I remotis ceteris, and
partly from the special character of Sallust's account: voltu <colore
motu> corporis pariter at que animo varius; here the contrast of corpus
and animus underlines the author's approach in seeking to represent
46 Rudolf Jacobs C. Sallusti Crispi de Bello lugurthino Liber 11th revised ed. by H.
Wirz (Berlin 1922) p.l48 (following H.R. Dietsch).

47
Cf. Lofstedt Syntactica II pp.290-294; 412ff. Such constructions consisting of
substantival neuters with genitives are not found in Sallust before the Jugurtha.
48 Richter p. 765 identifies a similar case of narrowing perspective; this time it
involves the scene as the Romans react to the shameful contract with Jugurtha ( ch. 30).
80 SALLUST
inner feelings by means of outward demeanour. Then in the last part

of the sentence Sallust brings out the contrast between the silence of
the Moor and the expressiveness of his features, which reveal what is
going on secretly inside him. The 'dark foil' tacente ipso stands
directly in front of the principal idea occulta pectoris and sets it off
brilliantly.
From the standpoint of dramatic development (which is our first
concern) the artistic portrayal of inner vacillation holds back the
action. In sentence 4 on the other hand Sallust uses characteristic
brevitas to report the final taking of the decision and the move that is
crucial for the next day: tamen postremo Sullam adcersi iubet et ex
i/lius sententia Numidae insidias tendit. The sentence is considerably
shorter than the previous one. The impression produced by the
relative lengths is underscored by the adverbs tamen and postremo:
\ the decision is late in coming, but when it does, it is final. The verbs
, are in the historic present (iubet; tendit): the reader has a close-up
-view of the events that take place in the foreground, whereas
previously what happened secretly in the mind had been moved into
the background by a distancing dicitur. The brief fourth sentence
contains more action than the long third one. Two decisions of the
Moor are summed up here and co-ordinated by a simple et. This
markedly plain structure contrasts with the complexity of the third
sentence; here we have a shift that reveals Sallustian velocitas. 49
In the fifth sentence the scene changes. The new action is
connected to what precedes by means of deinde. The transition from
night to day and the approach of Jugurtha are not conveyed in
independent sentences. Instead the account moves quickly via a
temporal clause mentioning these details to the departure of the
Numidian prince and the Roman quaestor. What is striking about
the main clause itself is that at the end there is a reference to the men
in the ambush (the last word of the sentence is insidiantibus). The
action thereby extends beyond the bounds of the sentence. Thus
immediately before the Numidian's arrival we have a word that from
the start sets the following scene in an unusual light. Yet inside the
sentence too the action is thrown into relief by skilfully anticipated
details: quasi obvius honoris causa comes before procedit in tumulum;
similarly tacente ipso was put as a foil in front of occulta pectoris
patefecisse. This preparatory sentence ends at the most exciting
point. Thereafter the story gathers pace at an astonishing rate. The
49 Cf. Quint. 10, 1, 102.
SALLUST 8!
first half of the sixth sentence, which is linked by eadem to the scene
of the fifth, still shows some fullness of detail; it is important to stress
that the Numidian's attendants were unarmed. The insertion of uti
dictum erat again underlines the treachery of the attack and at the
same time creates the necessary pause after the word inermis.
However after that the events come in quick succession. Ac links the
coming action closely with what precedes. The expressive adverb
statim depicts the suddenness of the assault and undique and simul its
concentration. The signal to attack does not have a main verb of its
own, but is put at the front in an ablative absolute. This construction
too reflects the action's uniformity and its rapid execution. No
r further quickening of pace seems to be possible. yet up to this point
I Sallust has kept in reserve the device of asyndeton which we know
Lfrom Cato, 50 and now he actually heightens the effect by means of
ellipse: ceteri obtruncati, Iugurtha Sullae vinctus traditur et ab eo ad
Marium deductus est. Here we have a greater concentration of action
than in sentences 5 and 6. Another streamlining device is the
participle: vinctus traditur. If for all its speed this last sentence still has
an air of finality, this is due partly to the threefold structure which
follows the law of increasing members, and partly to the change of
tense: the concluding perfect deductus est replaces the lively presents
and has a distancing effect due to its factual and down-to-earth
character. 51
We shall attempt to bring out the distinctive features of the text
under consideration by examining it under the following headings:
centripetal style, dramatic qualities, structural intent, psychology
and authorial interpretation.
2. Centripetal style
When we inspect chapter 113 of the Bellum Jugurthinum, we get the
impression that Sallust's narrative has very little to do with Cato. For
this reason alone it made sense to start by viewing the passage in its
entirety. If our attention is just limited to single words, we can
establish outside influences. However when a narrative sequence is
considered as a whole, the individual qualities of the author come to
rlight.
I
This dense narrative style, with many details expressed as
participles, has very little in common with Cato;_S_yntacti<edevices in
5° Caesar also uses it to produce an effective increase in pace at the end: Gall. 7,89,3
mittuntur de his rebus ad Caesarem !egati, iubet arma tradi, principes produci. ipse in
munitione pro castris consedit; eo duces producuntur. Vercingetorix deditur, arma
proiciuntur. Sallust was receptive to this aspect of Caesar's style.
51
Cf. Koestermann ad Joe.
82 SALLUST
Sallust are more subtle. However in the progressive narrowing of

perspective we observed a pattern of thought similar to Cato's.
The same centripetal tendency is also noticeable on a small scale.
In Sallust an important means of concentration are the participial
constructions, which Cicero tends to avoid in 'rhetorical' passages. 5 2
By contrast our text ofSallust has three ablative absolutes right at the
' beginning, two of which are compressed into each other. This
produces an unclassical concentration. The action is certainly
reported; however because it is in the ablative absolute, it seems to be
over no sooner than it has been mentioned. Here concision creates
energy.
Still more characteristic is a predicative participle: Iugurtha Sullae
vinctus traditur. The chaining is subsumed into the principal action
and not separated from it by a main verb of its own. 53 Here it is
helpful to compare the narrative of Cicero discussed in our chapter
on C. Gracchus. 54 When Cicero wanted to lend forcefulness to a
scene of flogging, he divided up the overall action into a series of
individual phases: arrest, chaining, preparation of rods etc. This led
through step-by-step intensification to an 'expansive' visualization
that appeals to the emotions. In contrast to such an analytical and
'centrifugal' presentation, Sallust's treatment is marked by com-
pression and concentration. An orator tries to influence the will of
his hearers by touching their hearts; his aims are therefore quite
different from those of the historian, who for all his empathy is still
concerned first and foremost with observation and knowledge.
Seneca the Elder55 once said of Sallust that nothing could be
removed from his text without the sense being impaired. And
Friedrich Nietzsche declared: 56 'My sense of style, and in particular
of the epigrammatic style, awakes almost instantly on contact with
Sallust. He is compressed and austere, and underneath there is as
much significant meaning as possible: he has a cold scorn of fine
words, and also of fine feelings.' In its exaggeration this statement is
more characteristic of Nietzsche than of Sallust. Yet the centripetal
trend ofSallust's style has been correctly understood here. This does
not however lead to an aversion against fine words and fine feelings;
52
Laughton p.156, cf. 142.
Cf. in addition Koestermann ad loc.: 'The two antagonists are pointedly
53
juxtaposed. We can imagine how they eyed each other.'

54
See above pp.33ff. and 41f.
55 At ex Sa!lusti sententia nihil demi sine detriment a sensus palest (Sen. contr. 9,1, 13),
with reference to a single passage.

56 Der Wille zur Macht val. 10 of the complete ed. (Leipzig 1922) p.342.
SALLUST 83
rather it provides a rational counterpoise .to the generation of

emotion and thought by specifically rhetorical and descriptive
means.
The same tendency is at work when Sallust as a historian rejects
Ciceronian rhythms, 57 avoids a whole series of fashionable political
catchwords 58 known to us from Cicero, and within the historio-
graphical tradition itself does not adopt Cicero's stylistic ideal based
on Herodotus and !socrates, but feels more attracted to Thucydides
and Cato. Even in word order 59 he is more conservative than Cicero:
this was of course incumbent on the historian in Rome. 60
3. Dramatic qualities
In Sallust the capacity for dramatic presentation is more highly
developed than in Cato; it reaches perfection in the Jugurtha. 61 This is
reflected in the style. In contrast to Cato, one is struck by the
abundance of conjunctions and adverbs that indicate temporal and
other relations: sed, scilicet, tamen, postremo, deinde. Generally
speaking the sentences are not placed alongside each other, but are
usually connected artistically, whether by means of the particles just
mentioned or by relatives (quae; cf. eodem). This technique is
maintained until shortly before the end. Only then does asyndeton
appear. Because it has been kept in reserve, it becomes a speeding-up
device.
The artistic use of different tenses also helps to achieve subtle
nuances. The whole story is told in the present after the manner of
Roman epic, unlike our passage from Cato 62 (at least Cato's
authentic text). With the switch of tense to the perfect, the animated
account of the capture ceases.
Sentences are accordingly combined in a way that is always
57
On Sallust's rhythms seeR. Ullmann SO 3 (1925) pp.65-75; J. Perret REL 40
(1962) pp.49f. (mainly 'even' feet: spondees, dactyls, anapaests). The differences
between Sallust and Cicero are explicable in terms of genre, though not exclusively so
(the rhythms in Tacitus show a closer affinity to Cicero: cf. Ullmann ib.). Cf. also
Vretska pp.l94f. on the epic clausulae in Cat. I Of. (text 1). There is now a monograph
on the subject by H. Alili The Prose Rhythm of Sa!lust and Livy Acta Univ. Stockholm.,
Studia Lat. 24 (Stockholm 1979).
58
See above p. 77 n.41.
59
Cf. Norden Kunstprosa p.203,1; A.W. Ahlberg Eranos II (1911) pp.88-106,
especially 95-106 (rather more freedom in the Histories). For a subtle analysis cf. W.
Kroll Glotta 15 (1927) pp.280-305, especially 299.
60
With Livy we shall notice a shift in many respects.
61
Cf. K. Buchner Der Aujbau von Sallusts Bellum lugurthinum Hermes Einzel-
schriften 9 (Wiesbaden 1953) p.89 (greater importance ofHellenistictechniques in the
Jugurtha; for comparison with the Catiline cf. pp. 77-84).
62
See above p.30.
84 SALLUST
adapted to the content and pace of the narrative. Subordinate clauses

and participial constructions play a much greater role than in Cato.
Both of them enable the writer to condense bigger units and to define
precisely the relation of individual events to each other. Sallust's
linguistic resources are thus more flexible.
Our detailed examination has revealed Sallust's dramatic art. We
must emphasize that the impression of 'brevity' and speed is not so
much the result of simply using particular stylistic devices (as for
example asyndeton); rather it derives from the calculated and subtle
application of these devices in conjunction with broader 'retarding'
sections. 63
4. Structural intent.
More than just dramatic instinct is involved in the methods discussed
above of combining sentences and ideas and in the meaningful
deployment of these methods within the text: we are dealing here
with a clear structural intent. A discerning connoisseur of Early
Latin like Fronto 64 admires Sallust's archaisms and figures of speech,
but the structure is what he admires most: Sallust writes structe. He is
not after asymmetry at any price. Our analysis of the construction of
two texts has revealed that Sallust's 'inconcinnity' ought not to be
given undue emphasis; 65 rather we saw that symmetrical and
' asymmetrical principles of design are equally balanced. We have also
established that the choice of words in these texts is due to deliberate
policy. A glance at the development of his language and style
confirms this. 66 Ceterum appears three times in the Catiline, but
about fifty times in the Jugurtha, and also repeatedly in the Histories.
The dative plural quis for quibus is found only once in the Catiline,
whereas in the Jugurtha and the Histories it is frequent. The ablative
of comparison occurs twice in the Catiline, but far more often in the
Jugurtha and the Histories. Advorsus is not found in the Catiline,
though it appears eighteen times in the Jugurtha, and is also quite
63 Cf. Richter p. 759 on Cat. 14,2: 'Sallust not only accumulates asyndeta within the
same narrow compass in different parts of the sentence: he also combines this device
with others that are intended to give the style a distinctive preciosity rather than
concision and pace; in fact they are meant to make the style seem positively overladen,
without forgoing the claim to brevitas.' On the shift to mannerism cf. pp. 760 and 769.
64 Franta 2,48 Haines; 132,1 van den Hout; 114,3 Naber; cf. 2,158f. Haines; lOOvan
den Hout; 107 Naber.

65 Leeman pp.356ff. shows that Tacitus often writes much less smoothly than
Sallust.
66 C[. Lofstedt II pp.290f.; A. Kunze Sallustiana 1-3 (Leipzig 1892-1898), especially
3 (1897f.); Syme pp.305-312 and 240-273.

SALLUST 85
common in the Histories. While the Catiline still says propter, in the
Jugurtha the more distinguished ob takes over (obis also preferred by
Tacitus). It was only gradually therefore that Sallust developed his
seria et severa oratio. 67 Consequently we are justified in regarding the
stylistic richness of our text as the result of a deliberate choice.
5. Psychology and authorial interpretation.
Up to now we have been considering passages from the standpoint of
their retarding function; we now have to approach these passages
from another angle, so as to do justice to their content as well as to
.their structural role. By means of parentheses, predicative adjectives,
and forward use of participles, Sallust suggests the hidden intentions
of the actors. In the combination of empathy (as in the syntactically
complex ablative absolutes and parentheses 68 ) and distancing (di-
citur), we find a subtle illumination of psychology that is something
new in comparison with Cato, and with Caesar as well.
Interpretation is important to Sallust and he does not let it stand in
isolation. On the contrary, the linguistic devices we have dealt with
allow it to become an inseparable part of the historical narrative,
which it invests with the somewhat restless chiaroscuro effect
characteristic of Sallust's style. His narrative is not 'two-dimensional',
like Cato's, but achieves a depth that is due to something more than
the dramatic technique of retarding or increasing the pace. Sallust's
style differs from Caesar's clear and sharp delineation by virtue of its
greater animation and colour; it differs from the language of Cicero
through its aversion to theatrical gestures.
67
Gel!. 17,18 (following Varro). In Sallust we occasionally find a trend in the
opposite direction (Lebek [quoted p. n.91] p.314).
68
See particularly the psychologically sophisticated construction of scilicet fol-
lowed by ace. and infin. (Hofmann-Szantyr p.359: 'restricted ... to Early Latin and
archaizing authors'); this comes after varius, which seems (owing to its three short
syllables) to float in the air. The author adds his own interpretation with the subtly
discriminating quasi in the combination quasi obvius; the personal construction obvius
is more elegant than obviam (cf. Lofstedt II p.412 on lug. 84,1 multus atqueferox
instare). Clarification from a future perspective is provided by insidiantibus. With this
is connected Sallust's extended use of the future participle; cf. Skard (1933) p.68, n.l;
Fighiera p.207.
5
Sullan and Augustan historiography
Claudius Quadrigarius and Livy
The Texts:
Q. Claudius Quadrigarius (Sullan period) 1
cum interim Gallus quidam nudus praeter scutum et gladios duos
torque atque armillis decoratus processit, quiet viribus et magnitudine
et adulescentia simulque virtute ceteris antistabat. is maxime proelio
commoto atque utrisque summa studio pugnantibus manu significare
coepit utrisq~Je, quiescerent. pugnae facta pausa est. extemplo silentio
facto cum voce maxima conclamat, si quis secum depugnare vellet, uti
prodiret. nemo audebat propter magnitudinem atque inmanitatem
facies. deinde Gallus inridere coepit atque linguam exertare. id subito
perdolitum est cuidam Tito Manlio, summa genere gnato, tantum
jlagitium civitati adcidere, e tanto exercitu neminem prodire. is, ut dico,
processit neque passus est virtutem Romanam ab Gallo turpiter
spoliari. scuta pedestri et gladio Hispanico cinctus contra Gallum
constitit. metu magna ea congressio in ipso ponti utroque exercitu
inspectante facta est. ita, ut ante dixi, constiterunt: Gallus sua
disciplina scuta proiecto cantabundus; Man/ius, animo magis quam arte
confisus, scuta scutum percussit atque statum Galli conturbavit. dum se
Gallus iterum eadem pacta constituere studet, Manlius iterum scuta
scutum percutit at que de loco hominem iterum deiecit; eo pacta ei sub
Gallicum gladium successit atque Hispanico pectus hausit; deinde
continuo umerum dextrum eadem concessu incidit neque recessit
usquam, donee subvertit, ne Gallus impetum icti haberet. ubi eum
evertit, caput praecidit, torquem detraxit eamque sanguinulentam sibi
in collum inponit. quo ex facto ipse posterique eius Torquati sunt
cognominati.
1 Fr. lOb, pp.207-210, Peter 2 = Gellius 9,13,7-19. On the character oflater annalistic
writing and on Claudius Quadrigarius in particular cf. E. Badian Latin Historians

(London 1966) pp.l8-20 and D. Timpe 'Erwagungen zur jiingeren Annalistik' Antike
und Abendland 25 (1979) pp.97-119. On the development of the portrait in Roman
historiography cf. M. Rambaud 'Recherches sur le portrait dans l'historiographie
romaine' Les Etudes Classiques 38 (1970) pp.417-447 and W. Richter 'Charakter-
zeichnung und Regie bei Livius' in: E. Lefevre-E. Olshausen (edd.) Livius. Werk und
Rezeption. Festschrift fur E. Burck zum 80. Geburtstag (Munich 1983) pp.59-80.
CLAUDIUS QUADRIGARIUS AND LIVY 87
( ... )when in the meantime a Gaul came forward, naked 2 except for a shield
and two swords, and adorned with necklet and bracelets, who surpassed the
rest in strength, size, youthful energy and manly virtue. As the battle was just
at its height and both parties fought most heatedly, he began to signal with
his hand to both that they should stop. The fighting was interrupted. After all
was quiet, he at once shouts with a very loud voice that if anyone wants to
fight it out with him, he should come forward. No one dared to on account of
his enormous size and his barbaric appearance. Then the Gaul began to scoff
and to stick out his tongue. Suddenly that cut to the quick a certain T.
Manlius, who was of very noble family, to think that such dishonour should
befall the citizenry, and that nobody should come forward from so big an
army. As was said, he stepped forward and did not allow Roman bravery to
be shamefully compromised by a Gaul. Girt with an infantry shield and
Spanish sword, he confronted the Gaul. Accompanied by general anxiety,
the encounter took place on the middle of the bridge before the eyes of both
armies. They took their stands thus, as I have said already: the Gaul, as he
had been taught, held his shield out in front and sang; Manlius relied more on
his courage than on his training, struck shield against shield, and upset the
Gaul's stance. While the Gaul is trying to take up his stand as before,
Manlius again strikes shield against shield, and drives the man once more
from his position. In this way he slipped under his opponent's Gallic sword
and with his Spanish one cut open his chest. Then with the same stroke he
immediately pierced his right shoulder and did not yield an inch, until he had
brought him down, so the Gaul had no chance to strike a blow. As soon as he
had laid him flat, he cut off his head, pulled the necklet off, and put the
bloody spoil around his neck. On this account he and his descendants
obtained the surname Torquatus.
Livy 3 (59 B.C.-17 A.D.)

Dictator cum tumultus Gallici causa iustitium edixisset, omnes iuniores
sacramento adegit ingentique exercitu ab urbe profectus in citeriore
ripa Anienis castra posuit. pons in medio erat, neutris rumpentibus ne
timoris indicium esset. proelia de occupando ponte ere bra erant, nee qui
potirentur incertis viribus satis discerni poterat. tum eximia corporis
magnitudine in vacuum pontem Gallus processit et quantum maxima
voce potuit 'quem nunc' in quit 'Roma virum fortissimum habet,
procedat agedum ad pugnam, ut noster duorum eventus ostendat utra
gens bello sit melior.' diu inter primores iuvenum 4 Romanorum
2 I.e. without armour.

3
7,9,6- 7,10,14. Text of C.F. Walters and R.S. Conway, vol. 2 (Oxford 1919);
concordance by D.W. Packard A Concordance to Livy 4 vols. (Cambridge, Mass.
1968).
4
On Livy's treatment of the iuventus as a warrior caste of the same age-group in
archaic society cf. J.P. Neraudau 'L'exploit de Titus Manlius Torquatus (Tite-Live
7, 9, 6-1 0). Retlexion sur Ia 'iuventus' archalque chez Tite-Live' in: Melanges offerts aJ.
Heurgon (Paris 1976) II pp.685-694.
88 SULLAN AND AUGUSTAN HISTORIOGRAPHY
silentium fuit, cum et abnuere certamen vererentur et praecipuam

sortem periculi petere nollent; tum T. Manlius L. filius, qui patrem a
vexatione tribunicia vindicaverat, ex statione ad dictatorem pergit;
'iniussu tuo' inquit, 'imperator, extra ordinem numquam pugnaverim,
non si certam victoriam videam: si tu permittis, volo ego illi beluae
ostendere, quando adeo ferox praesultat hostium signis, me ex ea
familia ortum quae Gallorum agmen ex rupe Tarpeia deiecit.' tum
dictator 'macte virtute' inquit 'ac pietate in patrem patriamque, T.
Manti, esto. perge et nomen Romanum invictum iuvantibus dis praesta.'
armant inde iuvenem a equates; pedestre scutum cap it, Hispano cingitur
gladio ad propiorem habili pugnam. armatum adornatumque adversus
Gallum stolide laetum et - quoniam id quoque memoria dignum
antiquis visum est - linguam etiam ab inrisu exserentem producunt.
recipiunt inde se ad stationem; et duo in medio armati spectacu/i magis
more quam lege belli destituuntur, nequaquam visu ac specie aesti-
mantibus pares. corpus alteri magnitudine eximium, versicolori veste
pictisque et auro caelatis refulgens armis; media in altero militaris
statura modicaque in armis habilibus magis quam decoris species; non
cantus, non exsultatio armorumque agitatio vana sed pectus animorum
iraeque tacitae plenum; omnemferociam in discrimen ipsum certaminis
distulerat. ubi constitere inter duas acies tot circa mortalium animis spe
metuque pendentibus, Gallus velut moles superne imminens proiecto
laeva scuto in advenientis arma hostis vanum caesim cum ingenti sonitu
ensem deiecit; Romanus mucrone subrecto, cum scuto scutum imum
perculisset totoque corpore interior periculo volnerisfactus insinuasset
se inter corpus armaque, uno alteroque subinde ictu ventrem atque
inguina hausit5 et in spatium ingens ruentem porrexit hostem. iacentis
inde corpus ab omni alia vexatione intactum uno torque spoliavit, quem
respersum cruore collo circumdedit suo. defixerat pavor cum admi-
ratione Gallos: Romani a/acres ab statione obviam militi suo progressi,
gratulantes laudantesque ad dictatorem perducunt. inter carminum
prope modo incondita quaedam militariter ioculantes Torquati cog-
nomen auditum; celebratum deinde posteris etiamfamiliae honorifuit.
dictator coronam auream addidit donum mirisque pro contione earn
pugnam laudibus tulit.
When on account of the Gallic onset the dictator had ordered a stop to the
administration of justice, he made all those capable of bearing arms swear
5
W.D. Lebek Verba prisca. Die Anfiinge des Archaisierens in der lateinischen
Beredsamkeit und Geschichtsschreibung Hypomnemata 25 (Gottingen 1970 [Diss.
Cologne 1964]) p.235 supposes that Livy was following Claudius Quadrigarius in
choosing this expression.
the oath of allegiance, left the city with a mighty army, and set up camp on
the bank on this side of the Anio. In the middle stood a bridge, which neither
side demolished, so as to show no signs of fear. Numerous battles were
fought for possession of the bridge and, as the relative strengths were
uncertain, there was no way of telling who would take control of it. At that
point a Gaul of unusual size stepped onto the unoccupied bridge and shouted
as loud as he could: 'Come on then! The bravest man that Rome can now
boast, let him come out to battle, so that the outcome of our duel may show
which nation is better in war.' For long there was silence among the best of
the Roman troops, since they were afraid to refuse the trial of strength and at
the same time did not want to plunge themselves of their own accord into an
exceedingly dangerous fate. FinallyT. Manlius, son of Lucius, who had freed
his father from the harassment of a tribune, 6 goes from his post to the
dictator. 'Commander', he says, 'without your order I should never wish to
fight out of turn, even if! saw certain victory before me. If you permit, I want
to show that monster, since it dances about so defiantly before the enemy
standards, that I come from a family which threw the army of the Gauls
down the Tarpeian rock.' To this the dictator replied: 'All credit to you for
your bravery and sense of duty towards father and fatherland, T. Manlius.
Go on and prove with the gods' help that what bears the name of Rome is
invincible.' Then the other young men arm the youthful hero. He takes an
infantry shield and girds himself with a Spanish sword well-suited for close
combat. In full military dress they take him out to meet the Gaul, who
foolishly exults and (since to the ancients this too seemed worthy of mention)
even sticks out his tongue in mockery. Then they withdraw to their post and
the two armed men remain alone in the middle, more in the manner of a
public spectacle than of war. To the critical eyes of the spectators, they
presented a very different sight. One was of unusually large physique, and
resplendent with multi-coloured garment and painted weapons worked in
gold. The other was of moderate size like a true soldier, and his weapons,
being more practical than pretty, imparted only modest splendour to his
appearance. Here was no singing, no dancing, no vain shaking of weapons,
but a breast full of courage and silent anger. He had saved up all his wild
defiance for the decisive combat itself. As soon as they had taken their places
between the two fronts, while round about so many human hearts hovered
between fear and hope, the Gaul pushed his shield forward with his left hand
and, like a rock that tilts over, let a futile sword stroke come slashing down
with a mighty crash on the armour of the approaching enemy. The Roman
raised his dagger, struck his shield against the bottommost edge ofthe Gaul's
shield, and, at great risk of being wounded, slipped with his whole body into
the space between the enemy's body and his armour, then, striking
repeatedly, he cut open his belly and loins, and brought down the enemy,
who fell full length to the ground. He left the fallen corpse otherwise
untouched, and robbed it only of the necklet, which he put, still blood-
spattered, round his neck. Horror and admiration rooted the Gauls to the
spot. The Romans come joyfully from their posts to meet their hero, shower
him with congratulations and praise, and bring him to the dictator. Among
6
Cf. Liv. 7,4.
his comrades, who in soldier fashion came out with some rough and joking
songs, was heard the name Torquatus. Thereafter it spread to his descendants
as well and brought honour to the family. The dictator added the gift of a
golden crown and praised the combat in the highest terms before the
assembly.
A. Comparison of content
What are the features of Claudius Quadrigarius' narrative that do
not recur in Livy? As Livy's account is on the whole more detailed,
the omissions are all the more significant. The Gaul is not made to
appear naked, but dressed in a colourful garment. Livy not only
omits the beheading, but goes out of his way to stress that Manlius
left the enemy untouched. 7 Other primitive features are slightly
modified: with Livy, sticking the tongue out requires an apologetic
rider. He is also less explicit in stating that the hesitation of the
Romans on being challenged was a disgrace to the citizen-body. We
shall meet similar aesthetic and ethical considerations when dealing
with Livy's additions.
He simplifies the description of the fight: Manlius uses his trick
only once and not twice. As a result there is room for a futile sword
thrust by the Gaul; however we lose Quadrigarius' careful contrast
between the Gaul's acquired behaviour-pattern with its stereotype
repetitiveness (sua disciplina) and the Roman's courage and creative
presence of mind. We shall see next what Livy puts in place of this
interpretation.
What does Livy add? 8 His most important addition is a whole
7 That the necklet is nonetheless blood-spattered should not be taken too seriously,
even though Livy only speaks of injuries to the lower part of the body. It is however
typical.
8 Even if Livy is in part following another tradition here, the decision not to use
Claudius would be characteristic of his approach, since he uses him often elsewhere.
For use of another source see B. Sypniewska 'De Claudii Quadrigarii fragmentis ab A.
Gellio traditis quaestiones selectae' in: Charisteria Casimiro de Morawski (Cracow
1922) pp.l49-179, esp. 177ff.; M. Zimmerer Der Annalist Qu. Claudius Quadrigarius
(Diss. Munich 1937) p.l42 n.37; A. KlotzPhW 43 (1923) p.l035; K. Buchner Romische
Literaturgeschichte (Stuttgart 1957) p.363 (= K. Buchner 'Interpretation' in: E. Burck
[ed.] Wege zu Livius Wege d. Forschung 132 [Darmstadt 1967]p.380); and Neraudau
p.685. However P.G. Walsh Livy (Cambridge 1961) pp. 71, 151, 187 etc. ascribes the
divergencies from Claudius without hesitation to Livy. As well as being more
economical, this hypothesis is in the present case perfectly in keeping with Livy's
attitude. The stylistic comparison that concerns us here is in any case independent of
the question of sources. For a brief discussion of our texts see R. Heinze Die
Augusteische Kultur (Leipzig 1933 2 ; repr. Darmstadt 1960) pp.97-102 (= R. Heinze
CLAUDIUS QUADRIGAR!US AND LIVY 91
scene. Whereas in Claudius the hero's decision is followed immedi-

ately by his stepping forward and by the duel, Livy inserts the
question to the superior and his good wishes. In regard to content,
Roman disciplina is thereby exemplified, and at the same time the act
of the individual is formally designated as a vicarious feat on behalf
of everyone, and furthermore as a demonstration of the divine care
which makes Rome invincible. The formal significance of this
inserted scene will be discussed later.
However Livy places yet another scene before the duel: the arming
of the hero by the other young men. This 'dramatic chorus' then
accompanies him onto the battle-ground also, which is explicitly
compared to a stage. There is further picturesque action by this
group at the end of the combat. Livy thus aims on the one hand at
greater tension through retardation, and on the other at dramatic
vividness 9 by inclusion of animated groups.
Claudius had already captured the oppressive atmosphere before
the fight by speaking of the spectators' anxiety. Livy adds hope to
anxiety, and so elevates the fear into a state of expectant suspense.
First however he compares the appearance of the two heroes; this too
is done from the spectators' viewpoint (aestimantibus). By setting the
noisy barbarian 10 against the silently determined representative of
the civilized nation, this synkrisis follows an epic topos that
ultimately goes back to the start of the third book of the Iliad. In it
Livy turns all externalities into a mirror of inward character: size,
colourful garment and shining armour on one side, and on the other
the average stature of a soldier and practical weapons. Whereas
Claudius had drawn a comparison in terms of national psychology
'Interpretation' in: Wege zu Livius pp.378-379); K. Buchner Rom. Literaturgesch.
pp.361-365 (= Wege zuLiviuspp.380-382); A.D. Leeman Orationisratio I pp.80-81; W.
Richter (cited above n.1) pp. 59-80. The present writer's lectures prompted a book by
his student W. Schibel: Sprachbehandlung u. Darstellungsweise in rom Prosa: Claud.
Quadr., Liv., A. Gel/ius (Amsterdam 1971).
9
I. Borzsak 'Spectaculum. Ein Motiv der tragischen Geschichtsschreibung bei
Livius und Tacitus' Acta Classica Debrecen. 9 (1973) p.60 calls the passage from Livy
in comparison to Claudius Quadrigarius 'a vivid spectaculum that is in every respect
more rounded and convincing'. In addition he notes that Livy himself draws attention
to the theatrical quality of the duel ( 10,6 duo in media armati spectaculi magis more,
quam lege belli destituuntur). P. Steinmetz 'Eine Darstellungsform des Livius'
Gymnasium 79 (1972) pp.191-208 calls this kind of account 'picture-description' (200),
and again 'description of a clearly visualized picture' (204). From a different angle J.P.
Neraudau (cited above n.4) investigates Livy's tendency to create stereotypes.
10
Cf. also belua in Manlius' speech. The description of the enemy as a monster
occurs only in the speeches of the first decade and is characteristically also absent from
Claudius Quadrigarius; cf. B. Laggner Untersuchungen zur Topologie in den Reden der
ersten und dritten Dekade des livianischen Geschichtswerkes (Diss. Graz 1972) p.21.
by confronting stereotype reactions with versatility and determi-

nation, Livy replaces this with an ethical approach that centres on the
checking of passion and the greater concentration of strength
resulting from it (jerox 11 is said of the Gaul,ferociam distulerat about
the Roman). The consequent gain in typological significance is offset
by a loss in detailed observation.
In Livy the fight itself is an encounter between two principles.
Accordingly there is a strictly parallel arrangement with first a
tremendous reverse for the Gaul, set off in epic manner by a simile,
and then Manlius' victorious counter-attack. At this point the hero is
significantly just called Romanus. By letting the Gaul attack first,
Livy underlines the Roman's correctness; on the other hand
Claudius' remark animo magis quam arte confisus will have struck
him as putting too much emphasis on Manlius' lack of training. Livy
has separated the actual comparison from the description of the
fight; yet even the description is used to give sharper focus to a
contrast he has raised to typological proportions.
The characteristic feature of Livy's account is that he subsumes all
elements into the action and allows them to develop out of it. 12 The
basic ideas that concern him are often introduced indirectly by being
put into the mouths of his heroes or reflected in the spectators'
minds. Accordingly he makes the name Torquatus come from the
joking songs of the comrades that escort Manlius, while Quadrigarius
is satisfied with just a dry statement. Fuller involvement of the
dictator and the hero's companions invests the action with a deeper
human interest. The highest authority in the state is consulted prior
to the exploit, and also at the end it is the dictator who rewards
Manlius and praises him. 13
Concern for a smoothly flowing narrative, such as we note for
example in the actions of the 'chorus', is combined with a technique
of retardation and of building up tension over long stretches. Livy's
11 Onferox andferocia cf. J.H. Michel 'La folie avant Foucault; furor et ferocia'
Antiquite C/assique 50 (1981) pp.517-525, esp. 522ff.

12 He also refers to a heroic exploit of Manlius related a little earlier (cf. 7,4). In
Manlius' speech before the dictator a historical exemplum is brought into play: that of
Manlius Capitolinus. Claudius' general reference to nobility (summa genere gnatus) is
significantly turned by Livy into a vivid visualization. An instructive comparison with
the account ofValerias Antias is made by T. Koves-Zulauf 'Zweikampfdarstellungen
in der romischen Annalistik' in: Actes du VIle Congres de Ia FIEC I (Budapest 1983;
pub!. 1984) pp.447-451.
13 Richter p. 71 and n.37 thinks that reward and decoration with a golden crown are
Livy's own invention, although even he has to admit that both are normal practice as a
concluding official act.
CLAUDIUS QUADRIGAR!US AND LIVY 93
streamlining of the actual description of the fight is one aspect of his

artistic curtailment of those elements that are interesting only from a
material point of view.
B. Comparison of language and style

1. Claudius Quadrigarius
At first sight Quadrigarius' account seems very simple. Clauses arej
connected by means of demonstratives, relatives and parataxis, !,
insofar asthere is any connection at all. Word order is largely fixed: \
the sente~ce gets its unity from the initial position of the subject and o
the ~nal position of the predicate. Dependent clauses that are
subordinated to each other follow one another on the same principle
as in official and legal language. 14 On the other hand Quadrigarius is
able to detach particular elements that he thinks important from the
overall sentence and put them in front of the subject or after the
predicate. This sets them off as independent units. In accordance
with a familiar law of psychology, the unity of each sentence or
separate section is underlined by the important words at the first and
last positions in it, and at the centre as well with longer structures.
Claudius sets himself relatively narrow limits: on the one hand the
final position of the verb, and on the other the need for analytic
linearity in order to avoid confusion in longer sentences. Yet within
these limitations he succeeds in making the most natuml word order
also the one that is most convincing artistically.
The choice of vocabulary is appropriate and exact, and yet free oL
coarseness. One notes an ,unobtrusive concern with variation: 15 \
proelio - pugnantibus- pugnae- depugnare- congressio; conturbavit- J
subvertit - evertit; disciplina - ars. Where exactly the same event is
repeated, there is at least a change of tense (percussit - percutit).
Claudius goes beyond purely factual reporting and highlights
emotion through initial position. In an explanatory addition he pays
tribute to Manlius' exploit as a vindication of Roman virtus. Through
further abstracts he underlines the differences between the acquired
behaviour-pattern of the Gaul and the Roman's courage and creative
presence of mind. The abstracts used in this connection throw light
on each other thanks to antithesis (sua disciplina- animo magis quam
14
Siquis secum depugnare vellet, uti prodiret.
15
On the other hand the repeated facta est and facto (K. Biichner Romische
Literaturgeschichte p.361) is not significant.
arte confisus) and make a discreet contribution to the author's

personal interpretation.
Slight traces of an oral 16 style are to be found in the parenthetic
expressions already mentioned: ut dico; ut ante dixi. We know that
previous scholarship was wrong to regard these as a mark of
inelegance; on the contrary they are used in a structurally significant
way 17 to creatc(hyperbatoi"t:)18 Archaic elements are also present in
our text ( cf. the accusative duo, the genitive facies, the ablative ponti
and the genitive icti, the formation cantabundus, the construction of
quiescerent without ut and the ethical dative ei sub .. . gladium
successit). 19 Yet the overall impression is not one of deliberate
archaizing. Nor have the 'poetical' features, which were heavily
stressed by W olfflin, 20 decisively affected the general tone (on scuto
scutum percussit cf. note 11; on pectus hausit 21 cf. Homer Iliad 14,517;
there is also an epic ring to summo genere gnatus with its alliterative 22
stress on the etymological relationship). The frequency of atque, 23
1 which was already preferred by Cato in elevated passages, is clear
/ evidence that this is the language of historiography.
There are few 'rhetorical' devices in our text. We do however find
16 On 'oral' features in Claudius Quadrigarius cf. E. Wolfflin 'Die Sprache des
Claudius Quadrigarius' Archiv fiir lat. Lexikogr. 15 (1908) pp.10-22, esp. 11. While
Wolfflin recognizes (as he failed to do with Sallust) that here the historian is not
writing in a deliberately 'vulgar' manner, he still does not distinguish clearly enough
between genus tenue and features belonging to the 'elevated' style like scuta scutum
percussit. This expression belongs to the sphere of the 'solemn oral style' that we have
characterized on several occasions already.
17 A one-sided view in Norden I p.l76: 'generally quite artless style'; similarly J.
Marouzeau RPh 45 (1921) p.l60.

18 is, ut dico, processit (hyperbaton in an important clause). So with ita, ut ante dixi,
constiterunt.
19 On the archaisms in Claudius Quadrigarius cf. Wolfflin pp.12-15 and (better)
Zimmerer pp. 90-104. Lebek pp.238f. is (rightly) sceptical. There is also an archaic
flavour to the use of the compound in the same sense as the uncompounded form
(conclamare of one person); the same is true of the rather ponderous way of expressing
temporal relationships: cum interim; deinde continuo (however we should not follow
Zimmerer p.111 in speaking of 'tautology' here). The abl. abs. utrisque pugnantibus
despite the subsequent dat. utrisque is also noteworthy. (The use of dative predicative
participles is not very popular in prose before Cicero).
20
Loc. cit. passim.
21 Cf. the passages quoted in Brink ThLL 6,3, 14 (1938) 2573,61-2574,2, where the
examples that all come from elevated prose and poetry are oddly ascribed to sermo
castrensis. Nearer the truth is P. Maas Archiv fiir lat. Lexikogr. 12 (1902) p.546,
although he gives no examples. On origin and use of the word cf. E. Pianezzola 'Haurio
= ferio, perfodio. Un calco omerico mediato dagli scolii' in: Scritti in onore di C. Diana
(Bologna 1975) pp.311-323 and Lebek pp.235f.
22 Cf. also in our text concessu incidit neque recessit.
23
Cf. n.55.
the beginnings of an antithetic parallelism between the two heroes, 24

although the second element here is surprisingly continued as an
independent sentence. The contrast too between 'Gallic' and 'Spanish'
sword is not unintended, even though it will have sounded less
artificial to a Roman than to us. 25 The unobtrusiveness of rhetoric in
our text is also evidenced by the repetition of words from a preceding
sentence, as in donee subvertit- ubi eum evertit, or neminem prodire-
is ... processit; the device is close to traductio (ploke). It should be
noted how Quadrigarius tries to temper the archaic heaviness of the
repetition through slight variation.
The overall impression is of a style that does not aspire to exclusive
detachment from ordinary language, but attempts to combine
dignity with lucidity and correct usage; given the stage of develop-
ment that prose had then reached, this was a considerable achieve-
ment. 26 The pursuit of gravitas is seen for example in the absolute use
of quiescerent. 27 Striking evidence for the combination of power and
perspicuity is provided by the sentence structure we have identified:
this is both self-contained and at the same time sets off the emphatic
parts. While it is true that Claudius' narrative technique is linear and
one-track, he still has a feeling for effective construction; thus he is
able to accentuate dramatic moments by making the sentences
especially short. 28 Klotz 29 notes that Quadrigarius postpones men-
tion of certain external facts until they are relevant for the narrative.
In our text this is the case with the retrospective statement in the
ablative absolute that in the meantime the battle had progressed
considerably. Similarly the bridge is only mentioned before the duel,
and not when the Gaul first appears. 30 The concern with national and
24 Gallus sua disciplina scuto proiecto cantabundus; Manlius, animo magis quam arte
confisus.
25 Sub Gallicum gladium successit atque Hispanico pectus hausit.
26 The characterizations in Gellius are one-sided (13,29,2 vir modesti at que puri ac
prope cotidiani sermonis and 9,13,4 purissime atque inlustrissime simplicique et

incompta orationis antiquae suavitate). The artistry is underestimated by P.T. Eden
Glotta 40 (1962) pp. 78-81. The 'strong formal interest' is rightly stressed by Zimmerer
p.87. Borzsak p.60 emphasizes the 'unkemptness' of his style and sees in this not so
much the artistic aims of Hellenistic historiography as the primitive reality of old
Rome.
27
Zimmerer p.l09.
28 E.g. pugnae facta pausa est. A certain expertise in the build-up of tension is also
commended by W. Richter (cited above p.69 n.6) pp. 772f. He also notices that
Quadrigarius has moved a long way from Cato in the direction of sermo purus.
29 'Der Annalist Q. Claudius Quadrigarius' RhM 91 (1942) pp.268-285, esp. 269.
30 The fact remains striking, even if the bridge was mentioned before the start of our
text.
mass psychology also reveals a literary purposeY

2. Livy
On a linguistic and stylistic level the text of Livy is rich and many-
sided. In vocabulary and expression we note repeatedly how he
refines or tones down his source. The drastic linguam exertare is
replaced by just linguam exserere (and is further modified by an
apologetic parenthesis added in advance). 32 The young Romans'
reluctance to expose themselves to danger is expressed in a discreet
but rather artificial periphrasis: praecipuam sortem periculi petere
nollent. A certain remoteness from everyday language is also to be
observed in a vexatione tribunicia ('from harassment by a tribune of
the people'). It is also in keeping with Livy's detached manner to use
a noun phrase to sum up a situation: exsultatio armorumque agitatio.
However detachment and restraint are only one side of Livy's
', vocabulary. His efforts to achieve lively visualization on the other
hand often bring him close to poetry. Thus praesultare is given in the
dictionaries as occurring only in our passage. The poetic value of
pectus and mortales has already been discussed in our chapter on
Sallust. 33 The opening recalls epic: pons in media erat; 34 the com-
r-parison of the Gaul 'to an overhanging rock has the same effect. 35
' Two other features that bring to mind the liveliness of epic are the
large number of direct speeches in Livy (these are completely absent
in the Claudius passage), and the effort to achieve continuity in the
narrative (here we should mention the accompanying role of the
comrades before and after the battle, and above all the arming
scene). 36 While the rather abundant expression dis crimen ipsum
certaminis suits the style of writing described above, in which noun
31 In exornatio Quadrigarius advanced beyond Valerius Antias according to Klotz
p.269. There is no satisfactory treatment in Norden and Marouzeau (cited above p.94
n.l7). Striking in this connection is the altogether favourable depiction of the enemy
Gaul. The negative aspect of the noun adulescentia is avoided by the close connection
with virtus, and all other unfavourable characteristics in the Gaul are absent; cf.
Richter (cited n.l) p.65 and n.20.
32 See Lebek p.234 on linguam exert are and linguam.exserere. The archaic cum voce
maxima is also corrected: A.H. McDonald 'The Style of Livy' JRS 47 (1957) pp.l55-
172, esp. 168.
33
See above pp. 68ff.
34 See the present writer's Die Parenthese in Ovids Metamorphosen und ihre
dichterische Funktion Spudasmata 7 (Hildesheim 1964) p.182 with nn.225 and 226.
35 Vergil uses moles of Polyphemus A en. 3,656f.; for the rock in epic similes cf. A en.
7, 586-590, Iliad 15,618-621.

36 On the attempt to achieve epic continuity cf. F. Mehmel Virgil und Apollonius
Rhodius Hamburger Arbeiten 1 (1940).

phrases are used to achieve detachment, it does also have an epic

root, cf. e.g. Iliad 6,346 UVEJ..lOlO 9u~:AA.a. 37
A further constituent of the elevated style are archaic devices such
as duplication 38 (armatum adornatumque), religious formulae (macte
... esto; iuvantibus dis), and alliteration 39 (pie tate in patrem patriam-
que); Livy uses this last one impressively, but not to excess.
Livy also has examples of the old-established use of traductio to
combine sentences (armant- armatum). However he uses this device
to achieve a special effect, because it no longer connects the next
sentence, but the next but one. This very obvious method of
connection is thus given a functional application as a bridge over
inserted material. 40 The device can be varied by using a word with
related meaning instead of a simple repetition: porrexit- iacentis. In
this way Livy gives an impression of familiarity and novelty at the
same time.
Livy's strong interest in moral philosophy is reflected in his
vocabulary. The Gaul's impulsive behaviour is indicated by the
adjective ferox and the ana~ Af:YOJ..lEVOV praesu/tat. On the other
hand the rationally controlled restraint of the Roman is expressed
through transformation of the same words into nouns: non exsultatio
... ; omnem ferociam ... distulerat. The choice oflanguage reflects the
achievement of an inner detachment.
Claudius had mentioned virtus; to this Livy adds the further term
pietas and also the idea of Roman discip/ina and dementia. The fact
that these last two words are not actually used and that pietas and
virtus occur only in direct speech (in the mouth of the dictator) is an
indication that for all his anthropological and ethical interpretation
Livy still tries to express as little as possible in abstract terms and as
much as possible through real action. In Claudius virtus had been
attributed, not just to the Romans, but also to the Gaul in the sense of
'manly strength' along with vires and adulescentia; Livy however
restricts virtus to the Romans and attempts to give it a moral basis in
37 Bello melior and related expressions belong to the elevated language of poetry
and historiography; cf. E. Skard (cited above p. 70 n.8) p.25, and also ThLL 2,9 (1906)
2098,2ff. (Sinko). On the basis of Lucilius 972 bello bonus (with alliteration; on this see
Marx), Sail. lug. 13,2 bello melior and repeated instances in Vergil, can we point to
Ennius?
38 Cf. also gratulantes laudantesque. On the accumulation of participles see Enn.
ann. 103 V., Naev. 5 Str. Initially, prose is reluctant to accumulate present participles.
39 Cf. also armatum adornatumque; versicolori veste; armorumque agitatio; adveni-
entis arma (here the alliterative words are artificially juxtaposed; so with spectaculi
magis more; respersum cruore collo circumdedit).
40 Similarly (towards the end of the text) ad dictatorem ... dictator.
discipline, curbing of passion before the fight, and moderation in

victory.
The dictator and the gods also play a new role in Livy. In Quadri-
garius they had remained right in the background. He was concerned
with narrative and only occasionally offerred an interpretation, and
then in abstract terms. Livy tries through his account to let the action
interpret itself. Many of the features of his vocabulary that we have
dealt with are ultimately related to this endeavour.
Let us now turn to word order. Livy goes further than Claudius in
isolating important parts of the sentence: tum eximia corporis
magnitudine in vacuum pontem Gallus41 processit. Here we have initial
position of both the emphatic element and within it of the expressive
adjective. And the same thing happens in final position: utra gens
bello sit melior. Similarly emphasis (and nobility of expression) is the
aim when normal word order is reversed in the terms used for the
weapons: pedestre scutum; Hispano ... gladio. 42
Self-contained structures are created when the main idea encloses
the reasoning behind it, which is given in the ablative: nee qui poti-
rentur incertis viribus sa tis discerni poterat. In Livy as in Cicero, stress
on the penultimate position in the sentence can produce 'fading'
conclusions, that contribute to the peaceful tone of the narrative. 43 If
beginning and end are accentuated, this kind of word order is often
combined with asyndetic sentence connection: proe/ia ... ere bra; diu
.. . silentium. In other cases there is a meeting of analogous or
contrasting words where sentences join; so for example auditum;
celebratum, or the two contrasting movements of the 'chorus'
producunt; recipiunt ... sese. In what follows, the way sentences are
connected will help us to appreciate the narrative structure. 44
41 On Livy's practice in putting the subject late in the sentence see L. Kiihnast Die
Hauptpunkte der livianischen Syntax (Berlin 1872) pp.310f.

42
The normal order occurs e.g. 6,8,6 scuta pedestri; cf. 43,6,6 scuta equestria;
31,34,4; 38,21,13 g/adio Hispaniensi. The initial position is naturally justified, if the
particular type of weapon needs to be stressed. As there is going to be a hand-to-hand
combat, this is the case with Hispano gladio. We ·should also note the frequent
separation of a noun from its attribute: Hispano cingitur g/adio ad propiorem habili
pugnam.
43 Cf. adiuvantibus dis praesta; iustitium edixisset; sacramento adegit; castra posuit;
indicium esset; crebra erant; discerni poterat; Gallus processit; silentium fuit; ad
dictatorem pergit; victoriam videam; beluae ostendere; hostium signis; ex rupe Tarpeia
deiecit; laudibus tulit etc.
44 On the need for a 'large-scale syntax' in narrative contexts see the present writer's
contribution 'Zur Funktion der Tempora in Ovids elegischer Erziihlung' in: Ovid
Wege der Forschung 92 (Darmstadt 1968) pp.451-467.
C. Narrative structure
1. Claudius Quadrigarius
In Claudius Quadrigarius we observe a whole series of devices for
sentence connection, which are made to subserve the continuity of
the narrative and the emphasis on turning-points. A demonstrative
or relative is often used in the archaic manner (is; 45 id; eo pacto; quo ex
facto). In addition temporal adverbs are important (extemplo, deinde,
subito). Asyndeton is used to good effect on several occasions, as in
the account of the pause in battle (pugnae facta pausa est). In fact a
number of the sentences in our text that are attached asyndetically
have a similar retarding function. 46
On the other hand it is demonstratives that introduce the
description of how the Gaul first gave his signal, how Manlius came
forward, and how he got close to his opponent. This device is not
intended therefore to make particular stages of the action stand out
from the overall context, and to set them in relation to each other; the
aim is simply to connect an event closely with something that
happens immediately before it. Thus sentence connection in Quadri-
garius is not meant to highlight a larger narrative structure, but first
and foremost to clarify the relationship to each other of immediately
adjacent sentences. This of course is also true of the temporal
adverbs, even if on occasion an effective structural element is saved
up for the end: the victory over the Gaul is introduced with donee.
However Claudius does not exploit the dramatic effect of having the
main action in a subordinate clause, as a later writer would have
done; instead he destroys it with a further subordinate clause. Before
re-arranging this second clause, 47 one should bear in mind that
Claudius is more interested here in presenting and explaining the
battle tactics of the Roman 48 than in building up a big dramatic
climax. That Quadrigarius is more concerned with the immediate
context than with anything wider is also shown by his use of tenses 49
and by the fact that an antithesis which is at first made to run parallel
gradually develops in its second half into an independent new
45
On the other hand Livy actually avoids is on one occasion: quem nunc ... Rama
virum fartissimum habet, pracedat ...
46 As in the account of the Romans' vacillation (nema audebat ... ),and also later:
scuta pedestri ... ; metu magna ...

47
Sypniewska Joe. cit.
48
Ne Gallus impetum icti haberet.
49
See below p.IOI n.54.
sentence. 50 Where two sentences join, he often succeeds in producing

brilliant effects of contrast. Besides the asyndeta mentioned already,
we should note particularly the form he gives to a turning-point in the
whole narrative: neminem prodire. is ... processit. What a telling
contrast between neminem and is, and between insight (infinitive)
and action (indicative)! To this we should add the powerful
hyperbaton produced in the second half by ut dico.
Thus while Quadrigarius has a distinct feeling for drama and
contrast and can even build up an impressive climax, his syntactic
and stylistic devices merely serve as indicators for the immediate
context: they are hardly meant to highlight larger structural relation-
ships.
2. Livy
The preceding action is marked by the objective stylization in the
historic perfect. 5 1 The same is true of the conclusion, which gives the
origin of the name Torquatus and the dictator's tribute. The
background to the main action is indicated by sentences in the
imperfect (pons .. . era!; proelia .. . erant; nee .. . poterat). The
introduction of each of the two heroes is strongly accentuated by
means of tum. 5 2 The main action begins in the historic perfect with
the appearance of the Gaul. On the other hand Manlius' approach to
the dictator and the subsequent arming scene, as well as the action of
/the companions before and after the duel, are all characterized as
{ marginal episodes by the historic present. The comparison between
'the two heroes that precedes the duel is also in the present, although it
gradually shifts to an elliptical form of expression and ends in a
'background tense' 53 (distulerat: pluperfect), which creates a sense of
psychological depth.
Against this foil the main action starts, after further preparation in
the form of an introductory temporal clause (ubi constitere ... ) and an
ablative absolute that captures the atmosphere. Livy then describes
the action of the heroes in parallel construction; here stance and
50 Gallus sua discip/ina scuta proiecto cantabundus;·Manlius, animo magis quam arte
confisus, scuta scutum percussit ...

51
Dictator ... castra posuit.
" Et has an important structural function before the comparison between the
two heroes (et duo ... ). Otherwise Livy avoids heavy paratactic sentence connection.
Inde is unobtrusively put second in the sentence: armant inde ... recipiunt inde ...
iacentis in de ...
53 Terminology of H. Weinrich Tempus; besprochene und erzahlte Welt (Stuttgart
1964).
gesture are always characterized in the ablative absolute and the

decisive blow is reported in the historic perfect, which is then retained
until the victor puts on the necklet.
The Gauls' dismay is again described in a background tense
(pluperfect). The action of the comrades is in the historic present, r
which is suitable for plain description. Finally the naming and tribute
are expressed in the perfect, which is appropriate for objective and
weighty statements. 54
Our examination has shown that Livy uses paratactic means 55 of~l
sentence connection to accentuate the overall structure, and that :
participial constructions 56 and subordinate clauses 57 (which he uses
with more variety than Claudius) all form part of a firmly constructed
whole owing its depth of perspective to deliberate resumption or
suppression of specific linguistic and stylistic devices. The narrative
style of Claudius Quadrigarius is rich in contrast, but flat; Livy
~urpasses it with the three-dimensional effect of a design which works
through careful gradations and resumptive links.
54
Claudius on the other hand uses changes of tense to achieve a sudden 'close-up'
effect: detraxit ... inponit (sequel); dum ... studet, percutit (repeated action; at the same
time variation of the precedingpercussit);facta ... est ... conclamat ... audebat. Cf. J.P.
Chausserie-Lapree (cited above p.59 n.29) pp.393ff.
55 On the frequency of et, -que and ac/atque cf. J.B. Solodow The copulative
particles in Livy (Diss. Harvard 1971; summary in HSCP 76 [1972] pp.303-305).

56 We do not of course intend to belittle Claudius' achievement in extending the use
of the ablative absolute (Wolfflin Joe. cit.).

57
The short speech of Livy's 'barbarian' alone contains no fewer than three
subordinate clauses.
6
Two philosophical texts
I. Cicero (106-43 B.C.):

Earthly glory and true immortality!
Quocirca si reditum in hunc locum desperaveris, in quo omnia sunt
magnis et praestantibus viris, quanti tande~f!! ist~hominum gloria, "'f\
quae pertinere vix ad unius anni partem (exigutijfljotes£, lgitur a/te"lv
spectare si voles atque hanc sedem et aeterfzanrdomum~eque
te sermonibus vulgi dedideris nee in raet!J:iis humanis spem posueris
rerum tuarum; suis te oportet n ecebri ipsa virtus trahat ad verum
decus; quid de te alii lqquan~si videant, sed /oquentur tamen.
Sermo autem omnis ille'~!J:Z!!!!ii~ cingitur his regionum, quas vides,
nee umquam de ullo perennis fuit et obruitur hominum interitu et
oblivione posteritatis extinguitur.' -~
Quae cum dixisset, 'Ego v.ero', inquam, 'Africane, siquidem bene
meritis de patria quasi/li'ln'e~" ad cael£1aditum patet, quamquam a
pueritia vestigiis ingresstis patr!1itt~i~ decori vestro non defui, nunc
tam en tanto praemio exposito fnitj}Jm'IJ!to vigilantius. 'Et i/le: 'Tu vera
enitere et sic habeto, non esse te marta/em, sed corpus hoc; nee enim tu
is es, quem forma ista dec/a rat, sed mens cuiusque is est quisque, non ea
figura, quae digito demonstrari potest. Deum te igitur scito esse,
siquidem est deus, qui viget, qui sen tit, qui meminit, qui providet, qui
tam regit et moderatur et movet id corpus, cui praepositus est, quam
hunc mundum ille princeps deus; et ut mundum ex quadam parte
marta/em ipse deus aeternus, sic fragile corpus animus sempiternus
movet.'
' ... In case therefore one despairs of returning to this place, where everything
belongs to great and outstanding men, - what then is this human glory
actually worth, that can barely cover a small portion of a single world-year?
So if you want to look upwards and see this dwelling-place, this immortal
house, then do not rely on the talk of the common people, and do not set your
personal hopes on human rewards. Virtue itself must lead you by its own
1 Cic. rep. 6,23,25 - 24,26. Text of K. Ziegler (Leipzig 19605 ). Commentary by A.
Ronconi Somnium Scipionis, Introduzione e commento (Florence 1961).

CICERO 103
charms to true honour. What others say about you is their business - they
will of course say something. But that talk is completely enclosed by the
narrowness of the area you see here; it never lasted long for anyone; and it is
covered up by men's deaths and effaced in the oblivion of posterity.'
When he had spoken these words, I said: 'Truly, Africanus, if something
like a path to heaven's gate is really open to those who have done service to
their fatherland, though from youth up I trod in my father's and your
footsteps and tried hard to emulate your distinction, now that such a high
reward is set before my eyes, I shall make an even more untiring effort.' He
spoke: 'Do but make an effort and believe firmly that it is not you who are
mortal, but this body. For you are not the person that this shape represents;
rather it is the mind of each man that constitutes his being, and not the
external appearance that can be pointed to with a finger. Know then that you
are a god, as truly as it is a god who has vital energy, who feels, who
remembers, who foresees, who so directs and leads and moves this body he is
put in charge of, as the highest God does this world. And just as the eternal
God himself moves the world, which to a certain extent is mortal, so the
i,mperishable mind moves the frail body.'
1. Sequence of thought
The opening of our text forms the conclusion of the Elder Scipio's
remarks on the futility of glory. 2 It was immediately preceded by the
argument that in view of the greatness of the world-year even the
most enduring glory is perishable. Cicero assumes that a person has
abandoned hope of returning to heaven. Earthly glory would be all
he now had; but this would be worth very little. With igitur the
conclusion is then drawn from this that Scipio should direct his gaze
upwards and disregard the talk of the crowd and human honours.
Virtue 3 must be pursued for its own sake and it will thereby lead men
to the true heavenly reward. On the other hand what people say is
2
A.D. Leeman Gloria (Diss. Leiden; Rotterdam 1949) is more important in our
context than the well-known work of U. Knoche 'Der rom. Ruhmesgedanke'
Philologus 89 (1934) pp.l02-124 (now in U. Knoche Vom Selbstverstiindnis der Romer
[Heidelberg 1962] pp.l3-30). On the concepts of glory and immortality of mind in the
Somnium Scipionis cf. K. Buchner Somnium Scipionis. Quellen, Gestalt, Sinn Hermes-
Einzelschriften 36 (Wiesbaden 1976); G. Wojaczek 'OPriA EITIETHMHE. Zur
philosophischen Initiation im Somnium Scipionis' Wiirzburger Jahrbiicher 9 (1983)
p.l24; D. Bayer 'Der Traum des Scipio' Anregung 30 (1984) p.423; W. Olbrich 'Ciceros
"Somnium Scipionis"- episch und dramatisch' Anregung 30 (1984) p. 98. On the same
topic in Cicero's work as a whole cf. K. Buchheit 'Ciceros Triumph des Geistes' in: B.
Kytzler (ed.) Ciceros literarische Leistung Wege der Forschung 240 (Darmstadt 1973)
pp.489-514, esp. 512ff.
3
On Cicero's concept of virtus cf. P. Kuklica 'Ciceros Begriff virtus und dessen
Interpretation' Graecolatina et Orientalia 7-8 (1975-76) pp.3-32, esp. 18. On the
reward of immortality for the statesman as a Roman and Ciceronian addition to an
otherwise Platonic conception cf. E. Berti II De republica e if pensiero politico classico
(Padua 1963) pp.33f. and Buchner pp.73-81.
104 TWO PHILOSOPHICAL TEXTS
ephemeral and therefore a matter of indifference.

This exhortation is followed by the vow of theY ounger African us
to redouble his previous efforts. It is only after this ethical decision
that the philosophical doctrine of the divine element in man and of
immortality is presented. 4
J
1
.
2. Sentence connection
The sentences are linked to each other by means of particles:
quocirea; igitur; autem; et; nee enim; sed; igitur; et. However the
principle of careful sentence connection is violated at a significant
point: igitur a/te speetare si voles atque hane sedem et aeternam
domum eontueri, neque te sermonibus vulgi dedideris nee in praemiis
humanis spem posueris rerum tuarum; suis te oportet inleeebris ipsa
virtus trahat ad verum decus; quid de te alii /oquantur, ipsi videant, sed
loquentur tamen. The asyndeton in tuarum; suis is adversative; suis
receives strong emphasis owing to the initial position. On the other
hand the asyndeton in the next sentence has more the character of a
supplement. Cicero stresses vigorously the important concept that
virtue is to be pursued for its own sake and that the talk of the
common people is without substance. There is no opportunity for
such a 'change of register' in those authors who invariably write
asyndetically.
3. Multiplicity in sentence structure
The sentences in our text are of many different types. The first main
clause takes the form of a question; the second contains a prohibition.
Later on there are several imperative sentences. Those containing
statements are also varied: the future (loquentur; enitar) is used as
well as the present. The subordinate clauses too are marked by
diversity: conditional clauses (two cases of si; two examples of the
half causal siquidem), several relative clauses, including one with
anaphora of the pronoun, a dependent statement clause (trahat), a
dependent interrogative clause (quid ... /oquantur), a concessive
clause (quamquam), and a comparative clause (ut). For a short text
this is quite a collection.
On the other hand the rarity of participial constructions is
noteworthy. The only ablative absolute comes in the words of the
4 On the train of thought and the interpretation of the text cf. Buchner pp.41-46 and
88-92 and 0. Schonberger 'Ciceros "Somnium Scipionis" als exemplarische Lektiire

und Einfiihrung in die Philosophie' Anregung 30 (1984) pp.93-96. The interpretation
offered by Wojaczek is interesting; he explains the whole Somnium as the different
stages of a mystery initiation: this is accepted by Bayer.
CICERO 105
Younger African us: tanto praemio exposito. The phrase serves to give
the vow an impressive brevity. We shall have to return to this passage
when we deal with vocabulary and phraseology.
4. Emphatic positions in the sentence
The position of the verb is very flexible. On several occasions it yields
its place at the end of the sentence to other elements, 5 which are
thereby given emphasis: in quo omnia sunt magnis et praestantibus
viris; quanti tandem est ista hominum gloria; spem posueris rerum
tuarum; suis te oportet inlecebris ipsa virtus trahat ad verum decus;
angustiis cingitur his regionum; enitar multo vigilantius; non esse te
mortalem; qui tam regit et moderatur et movet id corpus, cui
praepositus est, quam hunc mundum ille princeps deus. Here the
forward position of the first group of verbs involves chiastic
structure. This is even clearer at the end of 25: et obruitur hominum
interitu et oblivione posteritatis extinguitur.
The beginning and end of the sentence are often stressed: quanti
tandem est ista hominum gloria? However there are also significant
passages in which the emphasis is put on the centre: quae pertinere vix
ad unius anni partem exiguam potest; igitur alte spectare si voles at que
hanc sedem et aeternam domum contueri; spem posueris rerum
tuarum; sed loquentur tamen; limes ad caeli aditum patet; sic fragile
corpus animus sempiternus movet.
With Cicero the main emphasis on content is not always at the end
of the sentence. The penultimate position seems if anything more ,
important, whereas the end itself fades away. 6 Cicero's aim is not ·
therefore an emphatic ending at any cost, but rather harmonic
balance.
5. Vocabulary
If we turn from the text's overall impression of grandeur to
individual words, it is astonishing to find so little in the vocabulary
that is unusual. 7
5
On verb position in the middle of the sentence cf. the present writer's 'M. Tullius
Cicero, Sprache und Stil' RE Supplement band 13 (1973) 1293.
6
In this connection it should also be recalled that in the formation of clausula
rhythm the penultimate position appears more important than the last.
7
E.g. though contueri (cf. [Gudeman] ThiL 4,4 [1908] 794, 14-795,50) is admittedly
more frequent in Cicero's philosophic works than in his speeches, it occurs even in
Plautus; it can thus be regarded as perhaps a rather less common word, but on no
account as an affectation. However the comparative adverb vigilantius is first attested
in our passage and is otherwise very rare (Phil. 11, 10; Tert. anim. 34,5; Aug. civ. 21,27
p.551,20; references supplied by W. Ehlers, ThlL, Munich). Buchner p.43 n.40 also
6. Form and content

Recognition that the vocabulary of our text is unremarkable has
further implications. It is all the more important, since the content
would lead one initially to anticipate a more select vocabulary.
Accordingly we must look for artistry elsewhere: it is to be sought,
not in the·material, but in its use and arrangement.
a) Sublimity through simplicity
The effect of sublimity can be due to nothing more than Cicero's
choice of a particularly ordinary expression or of a simple con-
struction. This is the case on a number of occasions in the very first
sentence. reditum desperaveris is the commonest and most laconic
construction with this verb: Cicero forgoes the accusative and
infinitive, which would also be conceivable here (for example te
redire posse), and thus achieves a more austere dignity. In the
following relative clause he again chooses the simplest expression:
esse ('be at the disposal of, belong').
b) An archaic element: duplication
The monumental simplicity we have described does however combine
with an archaic 8 element, viz. duplication: magnis et praestantibus
viris. The main point here is the restraint Cicero shows in the use of
archaism. For him the principal means of attaining dignity is not old-
fashionedness, but linguistic purity. The ensuing question about the
value of glory is again-formulated in the most concise way possible:
quanti tandem est?
c) Artistic use of analytic word order
Only towards the end of the sentence does the style become more
diffuse: ista hominum gloria (both attributes qualify the value of glory
adduces Boethius cons. 3,12,16 (iam enim, ut arbitror, vigilantius ad cernenda vera
oculis deducis). Nor is one particularly struck by the verb vigere, applied here to deity;
Cicero uses it in his philosophic works far more frequently than elsewhere. (The verb is
found in the speeches only in the middle and final periods, while there are also sporadic
occurrences in the letters and rhetorical works, but significantly not in inv.) The
metaphorical use in the philosophic works is rather bold; in our passage we can
virtually assume the basic meaning('to be biologically strong'). The adjectivefragilis is
found only once in the speeches (Mil. 42), though more often in the philosophic works;
in the letters it is completely absent, and in the rhetorical works it occurs only at de
orat. 3, 7,2. Apart from the philosophic works all the instances are therefore close in
time to rep. In the philosophic works the word is also found later.
8 See in general E. Bn!guet 'Les archalsmes dans le De re publica de Ciceron' in:
Hommages aJ. Bayet Coil. Latomus 70 (Brussels !964) pp.l22-131. On thefrequency

of archaisms in Cicero's philosophic works cf. the present writer Joe. cit. 1253. On the
archaisms in the stylistically 'elevated' sections of Cicero, to which our passage also
belongs, cf. ib. 1255-1257 and 1260.
CICERO 107
in different ways), quae pertinere vix ad unius anni partem exiguam

potest. It should be noted how the scope of glory progressively
shrinks in the final section from one word to the next. After the
neutral pertinere, the adverb vix registers a first basic reservation.
Then follows the exaggerated formulation unius anni ('one year',
instead of 'one world-year'). However even this was still saying too
much: the next word makes a further qualification (partem). Thus it
is only part of a year. But this is still not the end of the limitations.
exiguam rounds off the descending sequence: only a small part of a
single year. Here English word order is the opposite of Cicero's and
so it does not allow us to imitate the sequence of thought with the
same naturalness. Thus in the sentence under consideration it is only
towards the end that we find a rather more exuberant style; but here
too the artistry is much less conspicuous than it would be in English,
with its different rules of word order.
d) Plainness and emphasis
Here we see two opposite stylistic principles at work. A markedly
plain form of expression produces an effect of dignity. On the other
hand adjectives and adverbs carry a certain emphasis, which can be
further accentuated by position (quanti stands at the beginning,
exiguam in penultimate position). Emphasis combines with dignity in
the archaic duplication (magnis et praestantibus viris).
e) Combination of words
The following sentence also has no surprises in vocabulary, though it
does have in the way words are combined. alte spectare is concise and
at the same time not quite at home in prose; it is true that it recurs in
Tusc. 1,34,82, but otherwise Cicero says magna spectare. 9 Even so,
the more elevated expression in the philosophical works is based on
everyday life: infam. 10,26,3 we read alte ascendere. There is likewise
nothing remarkable about the individual words aetemus and domus,
though there is about their combination, which cannot be found
anywhere in the speeches. 10 Apparently colourless words are made to
stand out through combination. The collocation praemia humana is
also alien to the speeches. hominum gloria in the sense of 'earthly
glory' occurs in Cicero only in our passage.
f) Spiritualization of basic Roman elements
Our linguistic findings show that here Cicero is g1vmg unusual
combinations to words that derive from ordinary Roman life and is
thus making it possible to spiritualize the basic concepts behind
9
Phil. I, 12,29. 10
But cf. Tusc. I ,49, 118; nat. dear. 3, 16,41.
them. 11 The combination verum decus is another indication of how

traditional values are being transformed. In the speeches there is no
reason to distinguish 'true honour' or 'true glory' from what these
words denote in ordinary usage. The subject of the sentence under
consideration is ipsa virtus. Normally the Romans are reluctant to
turn an abstract concept into an agent (te trahat). Here Cicero feels
less constraint, than e.g. Caesar: the orator 12 had a sense of the
enlivening effect that comes from personification, and philosophy
confirmed him in this tendency.
g) Verbal repetition and variation of expression
In the sentence we are examining there is a case of verbal repetition
(quid de te alii loquantur, ipsi videant, sed loquentur tamen): the
repeated word is stressed; in the last clause there is a conceit. 13 Later
on a word is repeated for the sake of clarification: deum te igitur sci to
esse, siquidem est deus, qui .. . tam regit .. . corpus .. . quam hunc
mundum ille princeps deus. In the same way corpus recurs several
times and also, though a little further apart, mortalis, sermo,
praemium, decus. All the words mentioned are thematically impor-
tant. However Cicero avoids excessive rigidity: he varies between
gloria, decus and praemium (these are not of course synonyms but
related concepts), between corpus, forma andfigura, between mens
and animus, between mortalis and fragi!is, between aeternus and
sempiternus, between hominum, vulgi, humanis, alii, posteritas, and
between nouns (sermones, sermo) and verbs (loquantur, loqueniur).
Repetition of the same word with a different function was avoided
less in antiquity than it is nowadays. We can cite the following
example, which is no problem even for modern sensibility: ipsi
videant; ... quas vides. A third type of verbal repetition is specifically
rhetorical. Fivefold anaphora occurs in the almost hymnic sentence
that establishes the divinity of the human mind by reference to that of
the universe.
We noted a repetition with a similar psychological power in the
11 Internalization through the context is also to be observed in a colourless word
like res. In Cato's speech for the Rhodians it had denoted the material interest of the
individual; here rerum tuarum refers to man's higher interests in life. inlecebra is used
elsewhere in Cicero's speeches and philosophic works only in connection with vice;
here Cicero applies it to virtue. The metaphor is made possible by the antithesis 'vice-
virtue'; in this way the semblance of familiarity that Cicero always strives to achieve is
preserved here as well.
12 Cf. the numerous cases in which virtus is in the nominative; they are noted by
Merguet in his lexicon to the speeches s.v.

13 Cf. rep. 6,20,22 ipsi autem qui de nobis loquuntur, quam loquentur diu?
CICERO 109
conversation between the Scipios. The solemn promise of the

Younger ( enitar multo vigilantius) is taken up by the Elder in
confirmation and admonition (tu vero enitere): here we have the
objective echo of a subjective affirmation, and in it the same word
appears in a different light. This device too combines naturalness
with poetic effect (it is no surprise to find the same technique in
Roman and modern European poetry 14), and here it is also ennobled
by the atmosphere of ritual dialogue. The same tone appropriate to
the paterfamilias is also heard in the formula sic habeto. 15 The second
imperative itself has a solemn ring. However the combination of
habere with sic is also linked to the idea of ancient Roman tradition:
maiores nostri sic habuerunt was what we found back in Cato. The
expression is used in a similar way by Sallust. 16
What is astonishing of course is that, while from the stylistic angle
this conversation between father and son is marked by old-fashioned
Roman dignity, Cicero gives it a 'Greek' content: deity is active in the
individual, as in the universe. Thus in the setting typically Roman
resp~ct for ancestors and for the paterfamilias and his teachings is
given a spiritual dimension.
h) Antithesis, parallelism, chiasmus
Platonic doctrine appears in antithetic form: non esse te mortalem,
sed corpus hoc. nee enim tu is es, quem forma ista dec/a rat, sed mens
cuiusque, is est quisque, non eafigura, quae digito demonstrari potest.
Antithesis is characteristic of the rational style of philosophic
exposition; however in the fifties, when De rep. was written, Cicero
uses the formula non - sed often in the speeches as well. 17 Working
with contrasts is of course particularly important in speeches, where
it contributes substantially to the emotional effect. Accordingly in
the antithetic formulation there is a fusion of rational and emotional
elements.
The enumeration of the ways God operates in the world is
arranged anaphorically, as befits the style of hymnic predication.
Here however the writing is less elevated owing to the use of the third
person. An impression of rationality is also produced by the
'logically' rising sequence: biology- sensation- intellectual activity. 18
14
Cf. the present writer's Die Parenthese ... pp. 94f.
15
Cf. Buchner p.43 n.41.
16
See above p.l n.3.
17
Cf. P. Parzinger Beitriige zur Kenntnis der Entwicklung des Ciceronischen Stils, I.
Tei!. Programm des K. Humanistischen Gymnasiums Landshut fiir das Schuljahr
1910/11 (Landshut 1911) p.12.
18
Buchner p.43 also notes the climax.
The analogy between man and the world is presented in mirror form:
deus ... qui tam regit et moderatur et movet id corpus, cui praepositus
est, quam hunc mundum ille princeps deus. In the final sentence, with
its emphasis on motion, chiasmus is replaced by parallelism: et ut
mundum ex quadam parte mortalem ipse deus aeternus, sic fragile
corpus animus sempiternus movet.
The parallelism is not however schematic, and is broken up by a
slight counter-movement: mundum ... marta/em -fragile corpus. In
the same way rigid symmetry was avoided in the antitheses, where it
was a case of distinguishing between the body and the real man. In
this the main factor was the change of subject: nee enim tu is es, quem
forma ista dec/a rat, sed mens cuiusque, is est quisque, non ea figura,
quae digit a demonstrari potest. In his mature period Cicero knew how
to avoid the monotony of mathematical precision. 19
i) Verbs and nouns
In the section we are considering, the impression of simplicity comes
largely from the use of verbs (which was originally a characteristic of
Latin) and from the aversion to abstractions: sed mens cuiusque, is est
quisque. In a translation the English-speaker can hardly avoid talking
about 'person', 'individual', or even 'ego'. However in Latin the
plainness of the language underlines the character of the speech as a
supernatural revelation: it is not so much 'scientific' as 'religious' in
tone.
On the other hand we also notice 'noun style', above all in the final
sentence of 25: sermo autem omnis ille et angustiis cingitur his
regionum, quas vides, nee umquam de ullo perennis fuit et obruitur
hominum interitu et oblivione posteritatis extinguitur. Here the use of
nouns to fix the meaning goes partly beyond what would be expected
in English: angustiae and oblivio stress the central idea and Cicero
therefore expresses them as nouns. 20
7. Conclusion
Because on the whole Cicero connects his sentences very carefully, he
has asyndeton at his disposal as an intensificatory device for
emphasizing important ideas. With writers who make asyndeton the
rule, this opportunity is lost.
19 His early speeches are still partly characterized by excessive concern for
symmetry.
20 In English the occurrence of the main idea in the form of an adjective is less
noticeable: 'the talk is always restricted to the narrow area that you see'; 'it disappears
with forgetful posterity'.
CICERO Ill
It accords with the character of a philosophic work 21 that the

sentences generally merge into each other smoothly. The fact that in)'
our text the penultimate position in the sentence is often weightier j
than the last also helps to avoid an aggressively strident tone.!
Similarly the large number of subordinate clauses and their variety of
form points to a calm periodic style. The infrequency of participial
constructions fits this-generally relaxed mood: its violation by an
ablative absolute at an important point is all the more effective.
The elegant unobtrusiveness of the vocabulary and the refined
simplicity of the constructions make every combination of words
that is in some degree unusual stand out all the more. Accordingly,
since plainness alone contributes very substantially to the dignity of
the passage, only a few extra stylistic devices are needed to heighten
the tone. Thus we have a single case of archaic duplication, or the
expression sic habeto evoking the atmosphere of instruction in the
mos maiorum by the paterfamilias, or anaphora bringing us into the
'area of religious language.
Philosophic teaching readily assumes antithetic form; at the same
time however there is no anxious overemphasis on symmetry.
Terminological rigidity is broken up as far as possible by natural
variation in expression, though without allowing avoidance of
repetition to become obsessive.
Perhaps the most important aspect is that a Roman has entered the
realm of philosophy, as we see from the language and style of our
text. It is something new in Rome to express the traditional concept
of glory by the collocation gloria hominum, to reduce it in this way to
a special case, and on the other hand to internalize the related word
decus by the addition verum decus. In the formula sic habeto the
auctoritas of a paterfamilias and respect for the mos maiorum prepare
the way for the exposition of Platonic teaching. The first major
philosophic work in Latin prose still reveals by its linguistic form the
spontaneous assumptions and emotions - political, moral and
religious - to which philosophical thought in Rome could relate.
Through spiritualization of traditional Roman concepts Cicero has
made a significant contribution towards developing the Latin
language into a tool of philosophy. 22
21
Note the different stylistic approach in the passages from the speeches discussed
inch. 2. On the many-sidedness of Cicero's style see L. LaurandEtudes sur Ia langue et
le style des discours de Ciceron 1-III(Paris 1907; 4th ed. 1936-40) and the present writer
Joe. cit. (p.l05 n.5) 1237-1347.
22
No discerning critic will blame him for leaving to later writers the introduction of
many abstract concepts. Nor can Cicero's artistic restraint in this area be taken to
II. Seneca (d. 65 A.D.):

On the value of time 23
Seneca Lucilio suo salutem.
Ita fac, mi Lucili: vjUi!i£a te tibi, et tempus quod adhuc aut aufere-
batur aut subripiebatur aut excidebat collige et serva. Persuade tibi
hoc sic esse ut scribo: quaedam tempora eripiuntur nobis, qua edam
subducuntur, quaedam efjluunt. Turpissima tamen est iactura quae
·per neglegentiamfit. Et si volueris adtendere, magna pars vitae r::la- _
]2i1ur male agentibus, maxima nihil agentTbus, tot a vita aliud agen-
tibus. 24 Quem mihi dabis qui aliquod pretium tempori ponat, qui diem
aestimet, qui intellegat se cotidie mori? In hoc enif!l fallimur, quod
mortem prospicimus: magna pars eius iam praeterft,~quidquid a eta tis
retro est mors tenet. Fac ergo, mi Lucili, quodfacere te scribis, omnes
horas cf!JJ]21e_ctere; sic fiet ut minus ex crastino pendeas, si hodierno
3- manum inieceris. Dum differtur vita transcurrit. Omnia, Lucili,
. aliena sunt, tempus tantum nostrum est; in huius rei uniusfugacis ac
~:.'"'\
c;(>\§ lubJ:ii;__ae possessionem natura nos misit, ex qua expel/it quicumque
vult. Et tanta stultitia mortalium est ut quae minima et vilissima
sunt, certe reparabilia, inputari sibi cum inpetravere patiantur,
nemo se iudicet quicquam debere qui tempus accepit, cum.interim
hoc unum est quod ne gratus quidem potest reddere.
Interrogabis fortasse quid ego faciam qui tibi ista praecipio.
Fatebor ill:~nue: quod apud luxuriosum sed diligentem evenit, ratio
mihi constat 7iipensae. Non possum dicere nihil perdere, sed quid
perdam et quare et quemadmodum dicam; causas paupertatis meae
reddam. Sed evenit mihi quod plerisque non suo vitio ad inopiam
show that the Latin language was quite incapable of expressing theoretical ideas (R.
Poncelet Ciceron traducteur de Plat on. L' expression de Ia pensee complexe en latin
c/assique [Paris 1957] gives a one-sided view).
23 Sen. epist. 1,1. Text ofL.D. Reynolds (Oxford 1965). School commentary by G.
Hess L. Annaei Senecae ad Lucilium epistulae morales selectae 2 parts (text and
commentary) (Gotha 1890). Latin-French edition with brief notes by F. Prechac and
H. No blot, 5 vols (Paris 1945-64). For a general treatment of the literary letter in
antiquity cf. G. Maurach Der Bau von Senecas Epistulae morales Bibliothek der
klassischen Altertumswissenschaft 30 (Heidelberg 1970) pp.181-206. For a com-
parison of the letters of !socrates and Seneca cf. M. Lebel 'A propos des lettres
d'Isocrate et des lettres de Seneque le philosophe' in: Melanges offerts en hommage au
reverend pere E. Gareau (Quebec 1982) pp. 79-89.
24 On the similarity of thought and wording to Horace carm. II,l4,1-2 cf. J.F.
Berthet 'Seneque lecteur d'Horace d'apres les lettres it Lucilius' Latomus 38 (1979)
pp.940-954.
SENECA 113
redactis: omnes ignoscunt, nemo succurrit. Quid ergo est? non puto
pauperem cui quantulumcumque superest
sat est; tu tam en malo " 7 !/
,,._..,
serves tua, et bono tempore incipies. Nam ut visum est maioribus
nostris, 'sera parsimonia in fundo est', non enim tantum minimum in
imo sed pessimum remanet. Vale.
Seneca sends'greetings to his dear Lucilius.
That's the way, my dear Lueilius: claim your life for yourself and keep
your time carefully together, which up to now was either taken away or
secretly stolen or it slipped out of your hands. Be convinced that it is as I say:
somehours are torn from us, some are removed secretly, some slip away
from us. However loss through carelessness is the most shameful. And if you
will pay attention: a great part of our life escapes from us, while we do evil;
the greatest part, while we do nothing; and our whole life, while we do
something else. Can you give me the name of anyone who attaches real value
to time, who knows how to appreciate the day, who realizes that he dies
daily? For we deceive ourselves in seeing death before us: a great part of it is
already past. All of our life that lies behind us, is held by death. So do, my
dear Lucilius, what you are already doing, as you say in your letter: take
J1ossession of every hour. You will thus be less dependent on tomorrow, if
you have laid hands on today. While life is postponed, it hurries past.
Everything, my dear Lucilius, belongs to others, only time is ours. Nature
has put us in possession of this one fading and transient thing, from which
anyone can oust us, if he wants to. And so great is the folly of men, that they
allow even the smallest and most worthless things, that can certainly be
replaced, to be reckoned as their debt, when they have obtained them from
someone; but no one believes he owes anything, who has obtained time,
although this is the only thing he cannot give back, even if he is grateful.
Perhaps you will ask what I, who give you these instructions, am doing. I
shall admit it to you openly: what happens with a man who lives on a grand
scale, but is still cons'.::ientious- the book-keeping on my spending is in order.
I cannot claim that I lose nothing, but I shall say what I lose, and why and
how; I shall give account of the causes of my poverty. But ultimately I am in
the same state as most people who have been reduced to penury through no
fault of their own: everyone shows sympathy, but no one comes to help. So
what does it all mean? I do not regard someone as poor who is satisfied with
the little he has left. But I would rather that you keep what is yours; and you
will start at the right time, for, as our ancestors said, 'Thrift comes too late,
when the bottom has been reached.' For not just the least, but also the worst,
is left right at the bottom. Farewell.
1. Form and the sequence of ideas

Ita fac, mi Lucili: vindica te tibi, et tempus quod adhut::' aut auferebatur
aut subripiebatur aut excidebat collige et serva. The start ofthis letter
is so lively and brisk that Hess's commentary is sure it is a real letter.
Such a view is very naive; but Seneca obviously wanted to create this
impression. 25 The reader has the feeling of being drawn from the very
first sentence into a conversation between friends. Quite often Seneca
pretends to have something in mind that Lucilius is supposed to have
said or written. 26
The language is meant to have a direct impact, yet at the same time
is very artistic. The short commata and cola at the beginning become
progressively longer according to the law of increasing members: Ita
fac, I mi Lucili, I vindica te tibi. Whereas the first half of the sentence
has a linear structure, the second shows a circular form: et tempus ...
collige et serva. This important concept forms a solid frame round the
description of the way things have been up to now: quod adhuc aut
auferebatur aut subripiebatur aut excidebat. (Here we have a threefold
climax, in which the responsibility for the loss increases from the first
to the third verb.) The relative clause is ringed by the exhortation to
'keep one's time together' now: this key idea is reflected in the self-
enclosed framework.
This first sentence is also self-contained as a whole. Two impe-
ratives open it, and two imperatives round it off. Here Seneca also
pays attention to the rhythmic clausulae (vindicate t{bi: double cretic;
-batlir a"ttt exciiebat: cretic and double trochee; collige et serva: cretic
and trochee); this is done f~r more carefully than could be expected in
a real letter. We have here a special form of artistic prose, 27 that is
25 For an affirmative answer to the question whether the correspondence was real
cf. P. Grima! Seneca. Macht und Ohnmacht des Geistes Impulse der Forschung 24
(Darmstadt 1978) pp.155-164 and 315-327. A real correspondence is also posited by
W.G. Muller 'Der Brief als Spiegel der Seele. Zur Geschichte eines Topos der
Epistolartheorie von der Antike bis Samuel Richardson' Antike und Abendland 26
( 1980) pp.l39-142. Scepticism is expressed by A. Stiickelberger 'Seneca: Der Brief als
Mittel der personlichen Auseinandersetzung mit der Philosophie' Didactica class.
Gandensia 20 (1980) pp.133-148, esp. 135f., and K. Abel 'Das Problem der Faktizitat
der Senecanischen Korrespondenz' Hermes 109 (1981) pp.472-499; the latter contra-
dicts himself here, because in his review of G. Maurach Der Bau von Senecas Epistulae
morales in Gymnasium 79 (1972) pp.83-86, he regrets that Maurach 'follows those who
think they can pronounce the correspondence a mere literary fiction, a game to be
played in the mind of the man ofletters'. B.L. Hijmans Inlaboratus et facilis. Aspects of
structure in some letters ofSeneca Mnemosyne Suppl. 38 (Lei den 1976) p.l34, keeps an
open mind, as does Lebel pp. 78ff., who fully grasped the essay character of the letters
to Lucilius and stressed this.
26 Cf. also in our letter §2. An instructive example is treated by A.D. Leeman 'The
Epistolary Form of Sen. ep. 102' Mnemosyne ser. 4,4 (1951) pp.175-181. The
atmosphere of friendly familiarity at the start of ep. 1,1 is also stressed by Maurach
pp.25f.
27 Use of the term 'essays' is suggested by the comparison with Montaigne (so
Bacon, quoted in the introduction to the Latin-English edition by R.M. Gummere, I

[1917] p.X). On the question see also H. Cancik Untersuchungen zu Senecas epistulae
morales Spudasmata 18 (Hildesheim 1967) pp.91-10 1, referring to the examination of
SENECA 115
intended to hold and appeal to the reader in a novel way.

Persuade tibi hoc sic esse ut scribo: quaedam tempora eripiuntur
nobis, quaedam subducuntur, quaedam effluunt. After the main theme
has been established in the first sentence, the basic idea is now
developed in detail in an expanding progression. Before the decision
to use one's own time economically can be put into effect, it is
necessary to know where the lost hours actually go. So first Lucilius
should again make himself thoroughly familiar with the three
possibilities indicated: loss of time ·through force of external
circumstances, time stolen without being noticed, and loss of time
through one's own carelessness. The second sentence, which seems
only to reiterate a part of the first, in fact marks a new beginning: this
is the first step towards the implementation of time-saving through
clear recognition of the possibilities of losing it. In contrast to the
first sentence a longer clause is always followed here by a shorter one.
In this way Catonian intensity and brevity are achieved. 28
Turpissima tamen est iactura quae per neg/egentiamfit. Just as the
second sentence had singled out one aspect of the first and presented
it again on a larger scale, the third emphasizes one particular aspect
of the second. Because the most significant words are placed at the
beginning and end (turpissima ... per neglegentiamfit), the statement
is made to stand out. Attention is now concentrated on the last and
worst form of time loss. This is now expanded further and again
subdivided.
Et si volueris adtendere, magna pars vitae e/abitur male agentibus,
maxima nihil agentibus, tot a vita aliud agentibus. 29 Carelessness is
attacked from the three standpoints of doing evil, doing nothing and
doing other things. Here excessive rigidity is skilfully avoided by the
polite introduction: 'And if you will pay proper attention'. Seneca
involves the addressee in the train· of thought as a partner in the
conversation. The sentence is in three parts; each section ends with
the same word; and there is a climax in the content. 30
the concept of the 'essay' and its history by H. Friedrich Montaigne (Bern 1949)
pp.419-425. However Cancik is right to see that in Montaigne there is no pedagogic
intent and no reference to a real or imaginary addressee. Montaigne rejects this
'utilitarian' aspect, while Seneca emphasizes it.
28 The first sentence also contained similar cases of rhetorical 'decrease': subripie-
batur aut excidebat and col!ige et serva. On the whole however ascending sequence
predominated.
29 Plato had given the following definition of right action: -ru atn:ou 7tpaHEtV Kai J.lll
1lOAU7tpayJ.lovdv (Rep. IV 433a).

30 The sequence maxima- magna- tot a, given by the MSS, was wrongly retained by
Beltrami.
These preliminary reflections, which treat the reasons for time loss
with increasing concentration, bring us to the end of this section.
They are followed by a new start: Quem mihi dabis qui aliquod pretium
tempori ponat, qui diem aestimet, qui intellegat se cotidie mori?
Seneca no longer has in view the time we lose but the time that we
{possess. rhus the train of thought turns away from negative
j \\preliminaries to something positive. In a threefold anaphora, in
which the middle element is the shortest, the reader has it drummed
into him that scarcely anyone knows how to value properly the time
allotted to him: In hoc enim fallimur, quod mortem prospicimus:
magna pars eius iam praeterit; quidquid aetatis retro est, mors tenet.
Once more Seneca picks up an idea in the previous sentence and
enlarges on it; so again we have the same technique of concentration
and elaboration. And again there is a descending arrangement: in the
last two sentences the second part is expressed in a particularly
concise and striking way. Yet in Seneca brevity is not an inherent
feature of the style in general, but only a characteristic of the
individual sentence. Overall he is quite happy to admit repetitions.
Most importantly, he picks up and particularizes one aspect of the
preceding sentence in the next one.
Fac ergo, mi Luci/i, quodfacere te scribis, omnes horas conplectere.
Seneca comes back to the start of the letter. Fac itself recalls.the first
sentence (ita fac), as do the address to Lucilius and the recommen-
dation to keep track of time (conplectere corresponds to collige et
serva). After the situation regarding lost and allotted time has been
clarified on both the negative and positive sides, the exhortation to
make good use of one's time can then be given on a higher level. The
imperatival aspect of the first sentence is now picked up and
continued. Again we should note how politely and tactfully the
admonition to a friend is worded, in much the same way as the
commanders in epic praise the martial spirit that is already animating
their warriors and thereby reinforce it. Alongside the urbanity of
precepts that masquerade as mere confirmation of what the other
person is already doing in any case, the lively contact with the
interlocutor also shows in the address to Lucilius and in the hint of a
letter that Seneca is supposed to have received from his friend. The
injunction itself exhibits a small but significant deviation from the
first sentence. There is no longer any mention of 'time', but of 'all
hours'. This achieves a more specific and concrete level, which seems
to take account of the points made previously.
SENECA 117
Sic fiet ut minus ex crastino pendeas, si hodierno manum inieceris.

Dum differtur vita transcurrit.
The subject of time-saving is now dealt with in greater detail. The
first point is that nothing should be postponed. This conclusion is
only reached after two sentences in which the treatment becomes
increasingly concentrated and specific. Again the most concise
wording comes at the end.
Omnia, Lucili, aliena sunt, tempus tantum nostrum est; in huius rei
unius fuga cis ac lubricae possessionem natura nos misit, ex qua expel/it
quicumque vult. Et tanta stultitia mortalium est ut quae minima et
vilissima sunt, certe reparabilia, inputari sibi cum inpetravere patiantur,
nemo se iudicet quicquam debere qui tempus accepit, cum interim hoc
unum est quod ne gratus quidem potest reddere.
Attention is first drawn to the necessity of saving time, and then to
the difficulty of doing so. Seneca now turns to the question of
practical implementation. Firstly, time must be used economically,
because it is our most personal possession. 31 Secondly, time is
something valuable, but also something very hard to hold on to; the
second section of the passage deals with this difficulty. 32 The
phrasing is deliberately pointed, not to say paradoxical. Here again
Seneca develops further and particularizes certain aspects of the
previous sentence.
Interrogabisfortasse quid ego faciam qui tibi ista praecipio. Fatebor
ingenue: quod apud luxuriosum sed diligentem even it, ratio mihi constat
inpensae. Non possum dicere nihil perdere, sed quid perdam et quare et
quemadmodum dicam; causas paupertatis meae reddam.
This brings Seneca to a piece of practical advice. He delivers it
tactfully by telling how he does it himself, and admitting that he still
wastes a good many hours. However he is able to give account of
every moment. This key idea is not expressly st.ated until the last
sentence. It is preceded by a classification of Seneca's 'book-keeping'
according to quid, quare and quemadmodum. Part of the special
31 According to Stoic teaching only our own soul is under our control. Here of
course Seneca is not thinking of time in general, but of the time we need for the moral
perfecting of our nature. To this extent his statement is no contradiction of Stoic
philosophy. On the Stoic idea of time cf. V. GoldschmidtLe systeme stoi'cien et !'idee de
temps Bibliotheque d'histoire de Ia philosophie (4th ed.; Paris 1979).
32 On the topos of possessing and losing time in Seneca's work see J. Moreau
'Seneque et le prix du temps' Bulletin de !'Association Guillaume Bude (1969) pp.ll9-

124 and P. Grima! 'Place et role du temps dans Ia philosophie de Seneque' REA 70
(1968) pp.92-109; the latter has also discovered Epicurean ideas in Seneca regarding
the question of time.
character of the epistulae morales is that the philosopher makes no

secret of his own imperfection and tries to win the listener by means
of just this honesty (jatebor ingenue). The involvement of the
addressee through an imaginary question at the beginning (inter-
rogabis fortasse) is also to be seen in this context.
Sed evenit mihi quod plerisque non suo vitio ad inopiam redactis:
omnes ignoscunt, nemo succurrit. Quid ergo est? non puto pauperem,
cui quantulumcumque superest, sat est. 33 Tu tamen malo serves tua, et
bono tempore incipies. Nam ut visum est maioribus nostris, 'sera
parsimonia in fundo est': non enim tantum minimum in imo, sed
pessimum remanet. Vale.
The sentence beginning with sed is to some degree a concluding
remark on the previous one. The idea that loss of time can never be
made good is developed once again. Here too the most concise and
pointed wording comes at the end.
Quid ergo est? introduces the final section, which addresses the
question of when to start putting Seneca's suggestions into practice.
The exhortation to begin early stands at the end in the style of a
rhetorical peroration. It is emphasized by a vivid proverb.
In conclusion we should once again draw attention to the
technique of first indicating an idea in general terms, and then
developing it in detail (often in three stages). The structure of the
letter is clear and carefully thought-out. 34 We have no reason to
suppose that this is a real letter. Nonetheless Seneca constantly
brings in the addressee. This impression of 'dialogue' is a good
indication of the considerable literary and human quality of the
epistulae morales. Among the basic characteristics of Seneca's style
we can therefore include a disciplined train of thought (perhaps this
has not been given enough attention in Seneca), liveliness in
presentation, and an urbane manner towards the addressee. 35
33
The MS Q offers the following text: superest: de homine moderato sat est.
Beltrami's conjecture is tempting: da hominem moderatum: sat est. However it is more
likely to be a gloss that has found its way into the text. See also B. Axelson Neue
Senecastudien (Lund 1939) p.IOO.
34 The associative and irrational element in Seneca is overemphasized by E.
Albertini La composition dans les ouvrages philosophiques de Seneque Bib!. des Ecoles
Frans;aises d'Athenes et de Rome 127 (Paris 1923) e.g. p.299.
35
On the interpretation of the passage as a whole cf. J. Bliinsdorf-E. Breckel Das
Paradoxon der Zeit. Zeitbesitz und Zeitverlust in Senecas Epistulae morales und De
brevitate vitae Heidelberger Texte 13 (Pii:itz-Wiirzburg 1983) pp.18-24; Maurach
pp.25-28; Moreau pp.119-124; Grima! (loc. cit. p.ll7 n.32) pp.92-94.
SENECA 119
2. Vocabulary
Let us now look at the vocabulary of our text. The first two sentences
alone are evidence of a deliberate search for variation. Seneca repeats
and elaborates the idea without using the same words twice: eripere is
substituted for auferre, subducere for subripere, and effluere for
excidere. In the third sentence effluere is replaced by the periphrasis
iactura, quae per neglegentiam fit; in the fourth the same idea is
expressed by elabi. This gives an entire semantic field in four
sentences. None of the texts we have looked at so far exhibited such a
systematic attention to variety. We recognize whom Quintilian had
in mind when he dismissed the search for synonyms as useless
affectation. 36 At all events Seneca's vocabulary reveals a stylistic
approach that is the opposite of Caesar's elegantia, 37 and that also
goes far beyond Cicero's moderate concern for variation. The same
thing can be demonstrated with other semantic fields in our letter: 38
this richness of language is meant to be noticed and admired.
3. Metaphors
It is not just the variety of the vocabulary that needs to be
appreciated, but also its vividness and colour. Metaphors arise
chiefly from the following areas of Roman life: the law ( vindica te
tibi: 39 internalization of a juridical act as a metaphor for the
philosopher asserting a claim to his own person; the manus iniectio in
2 is similar; cf. also ex possessione expel/ere for the theft of time); the
area of personal obligation and the concepts connected with it
(inputare, debere and the un-Ciceronian reparabilia); and finally
finance and property (ratio mihi constat inpensae, luxuriosus, diligens,
36
See above p.39 n.28.
37
On this concept, which should not be confused with our idea of 'elegance', see
above p.57.
38 Semantic field 'time': tempus, tempora, pars vitae, vita, tempus, dies, cotidie, aetas,
hora, crastinum, hodiernum, vita, tempus, tempus. This is the total for §1-3; in §4 and 5
the interest is due to the avoidance of any terminology relating to time, since by now it
is obvious what is meant. Only at the end do we find in a rather different sense bono
tempore ('in good time'). -'Keeping time together': tempus ... collige et serva; omnes
horas conplectere; tu tamen malo serves tua; si hodierno manum inieceris; vindicate tibi.
The semantic field 'disappearance' (of time) is further supplemented in what follows
by two adjectives: huius rei ... fugacis ac lubricae. Cf. Hijmans p.89. The idea of
'robbing' time is supplemented by the legal term expellere ('oust from possession'). In.
the same way Seneca deploys a host of expressions in our passage from the spheres
'poverty', 'wealth', 'thrift', 'extravagance', 'indebtedness'.
39 On the meaning and origin of vindica te tibi cf. Bliinsdorf pp.l9f. and Maurach
p.26 n.5.
parsimonia, serves tua). 40 This is also the source of the proverb sera
parsimonia in fundo est. Thus Seneca develops his moral philosophy
in close linguistic and conceptual contact with the realities of Roman
life. It is therefore only logical for him to refer at the end to ancestors:
ut visum est maioribus nostris.
When he spiritualizes typically Roman concepts of law or exploits
the Romans' strong sense of obligation and in particular the frugal
qualities of the paterfamilias who keeps careful account of his
expenditure, Seneca is converting the facts of social psychology into
a metaphor for what happens in the mind of the individual. The
withdrawal into the private self41 under the emperors is still expressed
in terms of the traditional language of personal relationships.
4. Sentence connection
Seneca's sentence connection is noticeably different from Cicero's.
Asyndeton has almost become the norm, and particles and con-
/ !1 junctions like et, tam en, ergo, enim, nam are distinctly rare. Ergo puts
a stress on the end (2 and 5). Nam also is an important signal: it
introduces the final proverb. 42
At first one is tempted to connect the use of asyndeton with the
epistolary style. There are however a number of points that tell
against this: firstly, the letters with didactic content in Cicero show
the same care in sentence connection as his speeches; secondly,
Seneca writes the same way in his other works as he does here; and
finally, the letters to Lucilius are fictitious ones. 43 Thus it is not an
external difference of genre that is involved, but a genuine divergence
in stylistic approach.
5. Brilliance; wit; 'aggressive' style
When we move from Cicero to Seneca, there are two features that
strike us about the rhythm. On the one hand we note by and large the
same clausulae 44 as are also preferred by Cicero, while on the other
40 Horace characterizes the average Roman with the phrase 'rem poteris servare
tuam' (usefulness of the subject arithmetic); Hor. ars 329.

41 However this is not to underestimate the importance friendship has for Seneca.
Of course moral guidance that is published presupposes self-instruction.

42 Cf. Hes. op. 369; A. Otto Die Sprichworter und sprichwortlichen Redensarten der
Romer (Leipzig 1890) p.l49. Even so we can believe Seneca's statement that the
proverb was also current in Rome; there is really no need to suppose that he is
concealing a plagiarism from Hesiod.
43
See p.ll4 n.25.
44 At the end of the cola we mainly find cretics (often double), the type 'clausulas
esse', double trochees and double spondees. For a general treatment of prose rhythm
in Seneca cf. B. Axelson Senecastudien, Kritische Bemerkungen zu Senecas Naturales
SENECA 121
the colometry is more abrupt: in place of long periods we find small

independent segments. Seneca's style can be compared to a string of
pearls. The individual short sentence achieves brilliance in a variety
of ways. There is an attempt to make every sentence end as far as
possible in an epigrammatic point. Accordingly there can be no
question of any 'fading' conclusions such as we observed in Cicero. A
climax may lead up to the end ( eripiuntur - subducuntur - effluunt;
male- nihil- aliud; pretium tempori ponat- diem aestimet- intellegat
se cotidie mori). We can see that the picture undergoes an inward
development in these climaxes just by looking at the series of verbs
that denote the various types of loss: violent robbery, secret theft,
and simple losing (here the personal responsibility is greatest). Each
of the three verbs conveys a rich descriptive content, and this is what
matters to Seneca. His style is pervaded by evocative imagination.
Antithesis can also prepare the way for the pointed conclusion (sic
fiet, ut minus ex crastino pendeas, si hodierno manum inieceris). Here
the artificial sequence should be noted. It would have been normal to
start with the conditional clause and end with the result. However
Seneca emphatically moves the decisive act to the end.
An effective setting also comes first in the following sentence:
omnia, Lucili, aliena sunt, tempus tantum nostrum est (there is also
complete parallelism here thanks to the insertion ofthe vocative after
omnia). The same effect is produced by the intensificatory pattern
non so/urn -sed etiam: non enim tan tum minimum in imo sed pessimum
remanet. In contrast to the reflective style of Cicero's dialogues,
which leave the reader free to deal with the matter objectively, Seneca
would like to stir people up, and actually 'convert' them. His style is
one that seeks to win people; almost every sentence is a maxim. To
him rounded periods would have appeared a blunt weapon. 45
6. Rhetorical modes of thought
The difference between the style of Cicero's philosophical works and
those of Seneca cannot really be explained by saying that in the
philosophical works Seneca writes more 'rhetorically'; rather it is due
quaestiones (Lund 1933) pp.7ff. and Hijrnans pp.117ff. and 138, where the author
succeeds in working out the structural principle behind Seneca's first letter ('Chinese
box').
45 The 'pointed' style was already in fashion with speakers in the Augustan period,
as the passages collected by the Elder Seneca show. However Seneca was not uncritical
of the fashionable rhetoric that went back beyond Asianisrn ultimately to Gorgias (cf.
also section 7 below).
to the way he uses rhetorical modes of thought and the end to which
he puts them. It is not enough to conclude that Seneca blurs the
distinction Cicero had observed between the urbanity of the
philosophical style and the passion of oratory. Our text can give us a
clearer understanding here. Let us take the climax for example:
Lucilius should in Seneca's view immerse himself in the thought
(persuade hoc tibi), not that he is losing time (that would be too
abstract and have no practical consequences), but that some hours
are torn violently from him, some are secretly removed, and some
slip away unnoticed. Hence the thought is divided up, and the first
two stages of the division eliminate those influences that are not in
our power. The third type of loss is the most shameful (turpissima)
because it is our own fault. In this way Seneca uses a rhetorical climax
to perform a philosophical self-examination. He again goes over the
final point in an analogous rhetorical sequence: magna pars vitae
elabitur male agentibus, maxima nihil agentibus, tota vita aliud
agentibus. Seneca makes use of climax not only for self-examination,
but also to fill his mind constructively with particular ideas: quem
mihi dabis qui aliquo¢ pretium tempori ponat (first stage: general
formulation), qui diem aestimet (second stage: time is expressed in
more definite terms), qui intellegat se cotidie mori? (third stage:
limitation to the individual and use of paradoxical language to
provoke further reflection). Besides this pattern of three stages, we
should also include here the types of two-stage contrast examined
above. Connected with this is the way Seneca creates a background
for his exhortation to Lucilius by honestly describing his own
behaviour (4: ego; 5: tu tamen malo serves tua, et bono tempore
incipies).
The influencing of the addressee is not done haphazardly, but in
logical steps whose order is determined by rhetorical modes of
thought. Seneca pretends to be writing a letter, and so to be giving his
friend advice on introspection and life-style in a personal tone. 46
Here he is not content with calm contemplation, but tries to activate
his reader. A rhetorically grouped sequence of reflections and ideas
serves as a means to this end. 47 He thus enlists rhetoric in the service
of philosophic guidance (which was always easy in view of the close
46 Cf. Abel (cited p.ll4 n.25) pp.473-475 and 482-485.

47
W. Trillitzsch Senecas Beweisfiihrung (Berlin 1962) p.135 stresses that Seneca
rejects the abstract syllogism as an effective proof.
SENECA 123
verbal reference of ancient thought) and also of philosophic self-

education,48 since he writes his letters not just for Lucilius but for a
reading public. The first of these achievements made him one of the
fathers of the Christian sermon, and through the second, as
Rabbow 49 has shown, he served as a model for Christian intellectual
training down to Ignatius Loyola and beyond. So it is not true to say
that Seneca 'ruined' philosophy with rhetoric; rather he placed
rhetoric at the service of practical philosophy as a method of verbal
influence upon oneself and others. His contribution however is not
so much the principle itself as its lively, urbane and brilliant
application.
7. Seneca- an anti-Cicero?
It would certainly be a distortion to set the 'pastor and propagandist'
Seneca against the 'urbane and contemplative' Cicero (who after all
also proclaims Roman attitudes). Quintilian, who in spirit if not in
letter felt himself committed to Cicero, thought Seneca's style full of
talent, but tasteless and dangerous to youth. 50 As a teacher of
oratory, he found Seneca's pointed manner, _il_is_ a"ersion to long
periods and also his disrespectful remarks about Cicero as distasteful ./
as the vogue Seneca enjoyed among the youth of the day. Quintilian's
judgment is biased and in need of correction; after all we must
distinguish between the excesses of the fashionable rhetoricians of his
day and the style of Seneca himself. It is instructive that Seneca is
rejected not only by a classicizing eclectic like Quintilian but also by
.tbe archaizers, who do-ITfornearly the opposite reason: for them his
vocabulary is not unusual enough. 51
Because Seneca wants to influence an audience, he does in fact try
48 For this reason it is a mistake to see Seneca's rhetoric as something external, cf.
K. Abel Bauformen in Senecas Dialogen (Heidelberg 1967) p.13 'By tracing the
influence of contemporary rhetoric on the structure of the Senecan dialogue, Grima!
might be said to approach the work from the outside.' This is a modern way of
thinking.
49 P. Rabbow Seelenfiihrung, Methodik der Exerzitien in der Antike (Munich 1954);
(despite the edifying title this is a work of strict scholarship).

50 In Quintilian's survey of literary history a pretence of objectivity barely masks
the obvious distaste; cf. especially velles eum suo ingenio dixisse, alieno iudicio
(10, 1, 130); multa etiam admiranda sunt, eligere modo curae sit, quodutinam ipsefecisset
(131).
51 Gellius 12,2,1 (criticism also in Fronto l49f. van den Hout). It should be noted
that elegantia in the archaizers no longer signifies 'apt style', but 'unusual style'. On the
avoidance of archaisms in Seneca cf. Axelson ( 1933; cited p.l20 n.44) p. 96; ( 1939; cited
p.118 n.33) p.11.
to avoid everything that is too exaggerated and unusual. 52 He aims at

the 'elevated' style, which transcends conventionality not through
extravagance but through wealth of ideas and directness. 5 3
J \ It is therefore inaccurate to see Seneca merely as a successor of the
. Stoic-Cynic diatribe; for the nobility of his style distinguishes him
from it. He shares with this genre the dialogue manner and
missionary tone, but he avoids its vulgarity. When Quintilian
reproaches him for not curbing his talent through severe self-
criticism, this sounds just as narrow-minded as his ambivalent
judgment on Ovid. 54
Is Quintilian's aversion to these two geniuses perhaps due
ultimately to the fact that too much of their personality seemed to
have entered their style? Had Seneca possessed only ability and
imagination, and not in addition a more severe taste than Quintilian
was prepared to admit, he would not have become a 'second
J j founder' 55 of Latin prose and, more significantly still, of the
I European tradition of the essay. 56
5 2 An authority like Axelson gives the following assessment ( 1939 p.ll ): 'Despite
some licences (e.g. in the use of tenses) it can be said that Seneca's language presents an
overall picture of the greatest grammatical correctness and in several respects is more
correct than for instance Cicero's.' Nevertheless we read in our letter: non possum
dicere nihil perdere. The parallels adduced in Reynold's apparatus to the passage show
that here the absence of me is not due to corruption. It is surely difficult to regard this
construction as more correct than the Ciceronian. However if we leave aside the
polemical exaggeration, there is still much truth in Axelson's statement. Cf. also F.I.
Merchant 'Seneca the Phil. and his Theory of Style' AJPh 26 (1905) pp.44-59.
53 In many respects therefore Seneca meets the requirements of the author of the
nEpi UIJIOU~ (Guillemin, cited below n.55). Of interest in this connection is the way
Seneca characterizes his own epistolary style, for in letter 75 he says: Qualis sermo meus
esset, si una sederemus aut ambularemus, inlaboratus et facilis, tales esse epistulas meas
volo, quae nihil habent accersitum nee fictum (75, 1).
54 Ovidii Medea videtur mihi ostendere, quantum ille vir praestare potuerit, si ingenio
suo imperare quam indulgere maluisset (10,1,98).
55 A.M. Guillemin 'Seneque, second fondateur de Ia prose latine' REL 35 (1957)
pp.265-284.
56 Cf. Cancik Joe. cit. pp.91-101.
7
Petronius (d. 66 A.D.)
Table talk from the 'Satyricon' 1

'Nunc populus est domi leones, foras vulpes. quod ad me attinet, iam
pannos meos comedi, et si perseverat haec annona, casu/as meas
vendam. quid enim futurum est, si nee dii nee homines huius coloniae
miserentur? ita meos fruniscar, ut ego puto omnia ilia a diibus fieri.
nemo enim caelum caelum putat, nemo ieiunium servat, nemo Iovem
pili facit, sed omnes opertis oculis bona sua computant. ante a stolatae
ibant nudis pedibus in clivum, passis capillis, mentibus puris, et Iovem
aquam exorabant. itaque statim urceaJim plovebat: aut tunc aut
numquam: et omnes redibant udi tamquam mures. itaque dii pedes
lanatos habent, quia nos religiosi non sumus. agri iacent -' 'oro te'
inquit Echion centonarius 'me/ius loquere. "modo sic, modo sic" inquit
rusticus; varium porcum perdiderat. quod hodie non est, eras erit: sic
vita truditur. non mehercules patria melior dici potest, si homines
haberet. sed laborat hoc tempore, nee haec sola. non debemus delicati
esse, ubique medius caelus est. tu si aliubifueris, dices hie porcos coctos
ambulare. et ecce habituri sumus munus excellente in triduo die festa;
familia non lanisticia, sed plurimi Iiberti. et Titus noster magnum
animum habet et est caldicerebrius: aut hoc aut illud, erit quid utique.
nam illi domesticus sum, non est mixcix. ferrum optimum daturus est,
sine fuga, carnarium in medio, ut amphitheater videat. et habet unde:
relictum est illi sestertium trecenties, decessit illius pater. male! ut
quadringenta impendat, non sentiet patrimonium illius, et sempiterno
nominabitur .. .'
1 Petron. 44,14- 45,6. Text of Konrad Muller (Munich 1961 1); cited from the 2nd
ed. (1965), which also contains a translation by W. Ehlers; the revised 3rd ed. has now
appeared (Munich 1983), though without changes to our passage. There are
commentaries by L. Friedlaender (Leipzig 19062; reprint Amsterdam 1960), E.V.
Marmorale (Florence 1948), P. Perrochat (Paris 1939; 2nd ed. 1952) and a lexicon by I.
Segebade and E. Lommatzsch (Leipzig 1898). Of fundamental importance for
Petronius' language are now H. Peters mann Petrons urbane Pros a Sitzungsberichte der
Akademie ... Wien, phil.-hist. Klasse 323 (Vienna 1977) and A. Dell'Era Problemi di
linguae stile in Petronio (Rome 1970).
125
126 PETRONIUS
'Nowadays people are lions at home, and foxes outside. As for me, I've
already eaten my rags, and if the present high price of corn continues, I'll sell
my shacks. For what'll happen, if neither gods nor men take pity on this
colony? 2 As sure as I want to enjoy my family, I think all of that comes from
the gods. For nobody takes heaven as heaven seriously any more, nobody
keeps to the fast days, nobody cares even a hair about Jupiter, but everyone
wears blinkers and tots up their assets. The ladies used to go barefoot in
ankle-length robes to the Capitol with loose hair and pure minds and pray to
Jupiter for water. And so it rained buckets on the spot -then or never-, and
all came home wet as mice. That's why the gods sneak about as if with
woollen slippers/ because we've no respect for them. The fields languish-'
'Please!', said Echion the maker of rags for the fire-brigade, 'do talk about
something nicer. "Now this way, now that", said the farmer; he'd lost a
spotted pig. What doesn't happen today, will happen tomorrow: that's how
life pushes onwards. By Hercules, you can't imagine a better home town- if
only it had real men! But that's what's lacking at the moment, and not only
with us. We mustn't be so spoiled: everywhere heaven is just as far away. If
you've been anywhere else, you'll say that here the pigs walk around roasted.
And watch out, on the holiday in three days' time 4 we'll behaving a first-rate
show here. The team isn't made up of professional gladiators, but mainly of
freedmen. And our Titus is generous and a hot-head: one thing or the other;
at any rate there'll be something. I'm very close to him, you see; he doesn't
shilly-shally. He'll provide cold steel, with no defaulting, meat warehouse in
full view, so the whole amphitheatre can see it. And he has the means. Three
hundred times a hundred thousand is what he inherited, his father died. Very
sad. Even if he lays out four hundred thousand on it, that won't hurt his
inheritance, and he'll be talked about for ever ... '
A. Language and style

1. Specific points: vulgarism and hypercorrection
Even the outward form of the words has a colourfulness that matches
the surroundings. Fruniscar is known otherwise from Early Latin and
inscriptions. 5 The form diibus (or dibus, diibus) is also attested several
times epigraphically, but occurs only here in literature. 6 The ending
excellente is meant in Sommer's view 7 to distinguish the adjective (cf.
2 Presumably Puteoli cf. J.P. Sullivan The Satyricon ofPetronius. A Literary Study
(London 1968) p.47.

3 Interpreted differently by A. Otto Archiv fiir lat. Lexikogr. 3 (1886) p.209 'they
have their feet bound so to speak'.

4 Cf. Bulhart ThLL 7,1 (1938) 778, 15ff. A different interpretation is given by
Friedlaender and Marmorale ad Joe.

5 See Vollmer ThLL 6,6 (1923) 1422,58- 1423,17. Cf. esp. CIL 5, 7453,12 = CE 1578
quite talem carui ecce modo frunitus sexdecim annis castitate et amore tui. Onfrui cf.
ThLL ib. 1427,56ff.
6
See Gudeman ThLL 5,1,4 (1912) 886, 37ff.
7 F. Sommer Handbuch der lateinischen Laut- und Formenlehre p.453.
PETRONIUS 127
66,3) from the participle (cf. sequens ferculum ib. 66,3); however
sequens too is used adjectivally in the sentence concerned. We may
assume that excellente has been picked up from everyday speech,
especially as such a hackneyed term of praise will often have been
heard in particular connection with vinum. 8 The ending accords with
the colloquial trend towards 'normalization' of forms that seem
irregular. Phonetically speaking, plovebat is also vulgar; 9 we can
compare poveri- pueri in Pompei (vis a glide sound; u is dissimilated
to o). 10 The inserted v is found in all Romance-speaking areas.U
The masculine form amphitheater is interesting. The neuters that
later die out in the Romance languages often appear already in
masculine form in the 'vulgar' parts ofPetronius' novel, 12 e.g. caelus
(39,5; 45,3),/atus (42,5),/ericulus (39,4), vinus (41,12). Inscriptions
give a similar pictureY Besides analogy, the fact that -m and -s
ceased to be pronounced is of importance here. The resulting
uncertainty can lead on the other hand to delightful cases of
hypercorrectness: litterae thesaurum est (46,8).
This brings us to the psychological basis for the artistic use of
vulgarisms. Masculine in place of neuter accords with Petronius'
pursuit of liveliness and colour. 14 In terms of social psychology on
the other hand, hypercorrectness in the language characterizes the
parvenu.
2. Vocabulary
The Bahuvrihi compound caldicerebrius ('one whose brain is hot')
occurs once again in Petroni us (58,4), if Jahn's conjecture is right. 15 1t
is surely misguided to think that such compounds must always be
8
For exce/lens of commodities see 'K.-M.' ThlL 5,2,8 (1937) 1216, 58ff.
9
This is presumably a later development rather than an archaic form.
10
Cf. Perrochat ad Joe.
11
Italian piovere, Spanish !lover, French pleuvoir, Portuguese chover, cf. also
Meyer-Lubke REW 6610.
12
On occasion already in Early Latin (caelus Enn. Ann. 546). However such cases
should perhaps be kept distinct from the later development. Cf. Sommer, Joe. cit.
p.320. A. Stefenelli Die Volkssprache im Werk des Petron Wiener romanistische
Arbeiten I (1962) pp.60f. is not much use.
13
Cf. Sommer ib.
14
This is intended to explain the artistic effect of these traits rather than their origin.
As to syntactical vulgarisms, we should note the construction of fruniscor with the
accusative and the use ofjoras instead ofjoris. We can compare the growing use of in
with the accusative, which is also attested in Late Latin; cf. also Petronius 42,3 fui enim
hodie in funus (explicable in psychological terms through the idea of 'movement
towards'; on the replacement of ire etc. by esse see J.B. Hofmann Lateinische
Umgangssprache p.I66, with bibliography). On foras cf. also Stefenelli pp. 86f.
15
The transmitted text at 58,4 is ca/dus cicer eius.
128 PETRONIUS
Grecisms, 16 since the inscriptions from Pompei supply cases that are
in some degree comparable, 17 and Friedlaender produces evidence
for similar expressions from Italian dialects (p.263). 18 Noun com-
pounds are found elsewhere in Petronius; moreover it is freedmen
who use the mY We should not be surprised if the expressions
involved are always very colourful and sometimes tinged with
emotion. It was precisely this emotional and non-analytic character
that made it hard for them to enter classical prose. 20
The drastic adverb urceatim ('in buckets') belongs in the same
class; apparently it is attested only in this passage. From a phonetic
and a semantic point of view there is some doubt about mixcix
(miscix? 21 ). Centonarius is a word that otherwise occurs in literature
only in Laberius, though it often appears in inscriptions and
sometimes also in legal contexts. Linguistically it belongs to the
numerous terms ending in -arius that denote occupations; we meet
them above all in epigraphic sources. 22 Lanisticia is attested only
here; the formation is part of a group (-aceus; -acius; -icius; -ucius)
that on the whole did not develop until Late Latin. As with the
previous word, we are dealing here with a technical term.
3. Metaphorical language; 'elevated' and 'humble' elements
of style
Metaphors have an even stronger impact. Thus the technical term
carnarium ('meat store-room') stands for the fact that the gladiators
who are unable to fight await the death-blow in the ring and are not
carried into the spoliarium. The proverbial expressions are hardly less
16
Cf. Ernout RPh 22 (1948) p.214 on mundicors,pravicordius, suaviludius, univiria,
benemorius, oridurius.
17
fulbunguis is a wholly analogous formation. The following compounds given by
A. Maiuri (Naples 1945 ed.), excursus 2,235, are a little different: culibonia, seribibi,
piscicapi.
18
Cf. also Dares p.l6,21 Meister, on Diomede: cerebra calido.
19 fulcipedia 75,6 ('high-heeled princess' Ehlers), larifuga 57,3 ('tramp').
20
Isolated experiments of this sort by Cicero, who was an admirer of Ennius and
perhaps found some support for them in the colloquial speech of his time, were
rejected by Seneca with derision (Sen. ap. Gell. 12,2,6 on suaviloquens and
breviloquentia).
21 Friedlaender p.263 refers to Paul. Fest. p.l23,7 Muller = 110 Lindsay:
miscelliones appellantur, qui non sunt certae sententiae, sed variorum mixtorumque
iudiciorum sunt. The spelling x for s is a common error; hence there is no problem
about adopting it. On the formation cf. Ernout-Meillet and Perro chat; a conservative
and sceptical view in Walde-Hofmann II p.95 (not ruling out initially onomatopoeic
gemination, cf. Hofmann Umgangssprache p.61).
22 In Petronius cf. also formations like petauristarius (60,2 and often) and the
classical tabellarius (79,6).

PETRONIUS 129
drastic: dices hie porcos coctos ambulare (cf. also ubique medius caelus
est). The exemplary tale of the farmer who lost his spotted pig is also
meant to have a metaphorical and allegorical sense. 23 Similar is nunc
populus est domi leones, foras vulpes. By contrast udi tam quam mures
is not a metaphor, but a simile.
The diminutive casu/as is also typical of colloquial speech; it is not
meant to denote the smallness of the houses, but the owner's
attachment to them. 24 Preference for the humbler expression is also
to be found in iam pannos meos comedi. The disproportion that
begins here between the object and the words used to describe it
gradually grows wider. In the sentence nemo lovem pili fa cit greatest
and smallest are pointedly juxtaposed, while there is an elevated ring
to the anaphoric repetition of nemo. When it comes to the procession
of the matrons to the Capitol, the expression rises to monumental
grandeur (antea 25 itself sounds distinguished; so also the structural
repetition nudis pedibus ... , passis capillis, mentibus puris- always a
pointer to the elevated style, and here with chiasmus anci a rhythmic
clausula at the end). The verb exorare too is quite the reverse of
humble; it is in fact entirely appropriate to the religious context. 26
/ \Sim1d·1~~ly aqua for 'rahin' is ?o: to be seen as vudlgar : but asfthe ritual
wor . However t e art1st1ca11y constructe c11max ( rom the
procession, reflected in the linear form of expression, to the ritual
prayer) is followed by a come-down in both style and content: ita que
statim urceatim plovebat; aut tunc aut numquam: et omnes redibant udi
tamquam mures. Here we have an accumulation of the characteristics
of the humble style: drastic metaphors and similes, vulgar phonetics,
and elliptical expression. In varying the sentence length and stylistic
level Petronius has given tangible form to the sequence of solemn
prayer and sudden downpour.
23 Ancient evidence for this type in Friedlaender p.262; see esp. Quint. 5, ll ,21.
24 One aspect being the familiarity that breeds contempt.
25 Cf. Miiller-Ehlers p.444: 'normally ante'; G. Bendz 'Sprachliche Bemerkungen
zu Petron' Eranos 39 (1941) pp.27-55, esp. 35 referring to E. Lofstedt Peregrinatio ...

pp. 74f. and Synt. II p.304 n.l. Stefenelli p. 87 is no help. Cf. also Dell'Era p. 68. On the
'refined' character of an tea see Dell'Era p.24; on the contrast of styles ib. pp.57f.
26 We should bear in mind the note of theatrical solemnity that the word has
elsewhere in Petronius: 52,6 tandem ergo exoratus a nobis missionem dedit puero; cf.
also 140,7; there is a fine distinction in Coraci ... imperavit, but puellam quidem
exoravit: the boy is ready for anything, while the young lady demurs a little. Cf. also
Ov. met. 5,418 exorata tamen nee, ut haec, exterrita nupsi.
27 Cae/estes aquas imp/orare: Hor. epist. 2,1,135, cf. Ov. fast. 4,386; Liv. 4,30,7;
5,15,2. Its occurrence in technical writers (ThlL 3, I [1907]70,6-8) shows that it is far
from being an arbitrary poeticism.
130 PETRONIUS
With ita que dii pedes lanatos habent, quia nos religiosi non sumus the
discrepancy that has just been developed progressively is concen-
trated in a sharply pointed phrase: the gods are associated with the
not very lofty idea of feet wrapped in wool. Here the disproportion
between the words and what they refer to reaches its peak. 'The fields
languish'. Truly di multa neglecti dederunt Hesperiae mala luctuosae
(Hor. carm. 3,6,8). Here we have the theme of the sixth Roman ode-
but what a different style! The gap between subject-matter and what
is actually said is deliberately sought by Petronius in this section and
given dramatic expansion.
4. Elliptical expression and implicit meaning
j Petroni us' untranslatableness is due above all to his brevity. How can
et habet unde be rendered in English with three words? 28 The
expression occurs already in Early Latin. 29 Modo sic modo sic is also a
colloquial ellipse; the expressions aut tunc aut numquam and aut hoc
aut illud are close to ellipse, but toned down by adjacent predicates.
One phrase in our text is so brief that even an eminent scholar like
Friedlaender failed to understand it. On the passage decessit illius
pater, male he comments: 'Male does not give a satisfactory sense in
combination either with what precedes or what follows and is
corrupt or garbled'. 30 The key to a correct interpretation is the
elliptical use in e.g. Cicero A tt. 12,10 and 11 (cf. male factum in A tt.
15,1a,l): what we have here is a fixed expression in cases of
bereavementY Stolatae too is 'elliptical': the interlocutor can easily
guess the subject. A related phenomenon is the absolute use of
laborare: sed laborat (= colonia laborat civium bonorum penuriii).
Another characteristic feature is the tendency towards an implicit
mode of expression: nemo enim caelum caelum putat ('nobody thinks
heaven is to be taken seriously as heaven'). Similarly at 42,7 the word
mulier is picked up again with an added connotation: sed mulier quae
mulier ('real', 'that deserves the name') milvinum genus; cf. in Ovid
talis erit mater, si modo mater erit (her. 20,220). 32 There is another
28
On this expression ('il a de quai') cf. Stefenelli pp.88f.
29 30
Plaut. capt. 850; cf. also Ter. adelph. 122. Loc. cit. p.264.
31 See Krieg ThiL 8,2 (1937) 241,1-4. Ehlers translates correctly: 'I'm sorry'.
Marmorale, who does not punctuate, gives a less satisfactory rendering: 'sventura-
tamente mori'.
32 Cf. Hofmann Umgangssprache p.93. However one must disagree strongly with
his conflation of this with the type in funus, cui funeri used in official language.
Aposiopesis is also out of the question (wrongly argued by E.E. Burriss CPh 42 [1947]
p.245). On mulier, quae mulier cf. E. Samatov 'Una forma particolare di alliterazione
nel Satyricon di Petronio' Bollettino di Studi Latini 5 (1975) pp.27-29.
PETRONIUS 131
instance of this in our text: non mehercules patria melior dici pot est, si
homines ('real men') haberet.
5. Formulaic elements
A formula like quod ad me attinet comes from everyday speech; we
have already met it in Cato. The same is true of more or less faded
locutions with a religious origin like ita meosfruniscar, 33 me hercules, 34
me/ius loquere 35 and no doubt also the male dealt with already. Such
trite expressions contribute to the stereotyped character of colloquial
speech, which often replaces meaningful language with an assort-
ment of empty phrases. Thus the meaningless use of formulae that
are in origin religious is a good linguistic illustration of Ganymede's
jeremiad on the unbelief of the time. Petroni us' aim of exposing how
thoughtlessly people speak is drastically evidenced in a remark of
Trimalchio's: on a bowl two corpses were depicted as effectively as if
they were alive. 36 Just as the ideas have become detached from reality
through being repeated thousands of times, so the words have turned
into cliches and lost touch with the ideas. It is left to the sophisticated
writer to give a penetrating diagnosis of this absurd situation. Great
mastery of language is needed to realize such a survey in a novel. One
means of achieving it is through the banalities of the vulgar language.
The clear and dispassionate precision of Petronius' narrative style
forms an important contrast to this in the non-vulgar parts of the
work. Petronius' use of language is thus invested with symbolic
significance.
B. Structure and sentence connection

The first sentence of our text forms the end of the complaint that the
people put up with an aedile's excessive power without grumbling.
This theme is brought to a close by a vivid phrase, which Petronius
may not have taken from popular speech. 37 With quod ad me attinet
Ganymede passes to the personal consequences which he says the
high price of corn entails for him. As a good businessman he
33
Cf. e.g. ita me di ament; see Lumpe ThU 7,2,4 (1967) 526,23 - 527,18 with
bibliography.
34
On this formula see Hofmann Umgangssprache pp.29f.
35
Wrongly classified in ThLL 2,9 (1906) 2107,17. As regards the general sense, our
passage belongs to the group at 2093,71 - 2094, 16. It is well explained by Friedlaender
ad Joe. (euqn]!let).
36 37
Petron. 52,1. Cf. Aristoph. Peace 1189.
132 PETRONIUS
maintains that he is faced with imminent ruin. He looks for the cause
of the desperate situation in the indifference of the gods, which is a
consequence of men's indifference in matters of religion. At this
point Ganymede becomes grandiloquent. This would be appropriate
in a speech, where the conclusion is after all meant to appeal to the
feelings of the audience. In conversation on the other hand the
situation is different, and for this reason we have the tactful
interruption by the interlocutor. Grandiloquence is not however the
only trait that is striking about this lament for the decline of piety.
The drastic expressions ('in buckets', 'wet as mice') form a remarkable
contrast to the lofty theme which in a more solemn dress is familiar to
us from Horace and other Augustans. We see there is another side to
the successful businessman's complaint that everyone. thinks only of
his accounts and not of God. The lowly vocabulary is surely meant to
emphasize the emptiness of the topos.
The final part of Ganymede's speech is very carefully worked out,
not only in its structure, as we discovered earlier on, but also in the
way the sentences are connected. 38 The colometry is dominated by a
two-element pattern that is clear and easily recognized. In the more
ambitious style of the comparison between today and the good old
days we find a three-element pattern as well, which is accentuated by
anaphora and repetition of the same cases. Graphic and pointed
expressions also play an important role in rounding off sections. We
have seen already that the previous topic was brought to a close with
a vivid maxim-like antithesis. The same is true of the end of
Ganymede's speech: ita que dii pedes /anatos habent, quia nos religiosi
non sumus.
Petronius deliberately obscures the impact this has as a closing
statement by beginning a standard complaint about the hopeless
state of agriculture: agri iacent. However at this point Ganymede is
interrupted by Echion. He replaces grandiloquence with an everyday
philosophy of gracious contentment and passes thereafter to the
38 J. Feix Wortstellung und Satzbau inPetrons Roman (Diss. Breslau 1934) treats the
particles used in sentence connection only partially and in passing (pp.32-34). See
however J.K. Schonberger 'Zum Stil des Petronius' Glotta 31 (1951) pp.20-28, esp. 27:
'It is hard to find another Roman writer who has such a mass of sentence-connecting
particles as Petronius'. Nonetheless it is a surprise to discover that this is true even of
the 'vulgar' parts, as the more markedly popular style ofEchion's speech shows. Some
of these particles are also included in P. Soverini 'Sull' usa degli avverbi in Petronio:
avverbi intensivi e asseverativi' Atti della Accademia delle Scienze dell' Istituto di
Bologna, C!asse di Scienze Morali anna 69, rendiconti 63 (1974--1975) pp.200-255 (does
not relate directly to our text).
PETRONIUS 133
inexhaustible subject of gladiatorial contests. Here too there is a

drastic image immediately before the change of subject: tu si aliubi
fueris, dices hie porcos coctos ambulare. Nor is the train of thought in
Echion's speech at all arbitrary; the abundant use oflogical particles
and conjunctions leaves us in no doubt here.
The faded religious formula me!ius loquere immediately marks the
more gentle tenor of the speech, which at first draws its inspiration
from popular wisdom. With the words 'now this way, now that' the
farmer consoles himself for the loss of a spotted pig. This is the sort of
exemplum that Horace loves; they belong to the educational method
of his freedman father and to the everyday wisdom of the peasant
philosopher Ofellus. 39 Two maxims follow, which perhaps sound
more popular than they really are: quod hodie non est, eras erit. This is
surely a trivialization of the famous non si male nunc, et olim sic erit. 40
And cannot sic vita truditur be better understood against the
background of Horace's truditur dies die? 41 We can imagine how
Petronius' educated readers must have felt when they encountered
the aphorisms of their classic author stripped of cultivated form and
in the mouths of freedmen. Before speaking prematurely of 'Vulgar
Latin', we should recognize this ambiguity even in the sections that
seem quite ordinary. Petronius never ceases for a moment to be the
same widely-read but irreverent connoisseur ofliterary tradition, the
pose he adopts in the urbane narrative sections of the work. This is
why the 'vulgar' parts ought not to be isolated from the rest of the
work, but should be appreciated as artistic prose too. While
Ganymede's lament contained a trivial distortion of Roman aware-
ness of decline and Roman religio, we now have a rehearsal of the
theme of rural resignation that is underpinned by two platitudes
mischievously echoing Horace 42 (and which should not be uncriti-
cally regarded as 'popular').
After this introduction Echion starts to give a version of his own
which tends to look on the brighter side; it opens with a qualified
39 Here mention should also be made of the diatribe tradition and Bion of
Borysthenes (in this respect the first Russian realist); however there is much debate as
to the degree of influence, cf. R. MuthAnzeiger fur die Altertumswissenschaft, hrsg. v. d.
Humanistischen Gesellschaft 9 (1956) p.l4.
40
Hor. carm. 2, 10, 17f.
41 Hor. carm. 2,18, 15. There are also echoes of Horace at 99,1 (epist. 1,4, 13f.); 117,9
(sat. I ,3, 13; carm. 3, 1,30); 114,3 (carm. 1,3, 14f.), cf. Schonberger loc. cit. (n.38), esp.
p.25. The entire cena is influenced by Hor. sat. 2,8.
42 Of course the incuriosa infelicitas of the freedmen's language makes Horace's
curiosa felicitas (Petr. 118) stand out all the more.

134 PETRONIUS
acceptance of the previous speaker's complaint. Their home is

energetically praised as the best of all possible small towns, and it is
even admitted that there is a lack of good people. But (and here sed
introduces a fresh aspect) there is a similar shortage elsewhere. Two
further aphorisms repeat the old topos of traditional Roman
sturdiness (non debemus delicati esse; 43 perhaps Petroni us is thinking
once again of his direct opposite, Seneca44) and realistic down-to-
earthness ( ubique medius caelus est). Finally the idea is put even more
pointedly: if you've ever been anywhere else, you think it's paradise
here. The expression draws on the myth of Cockayne. 45 So once
again we have contrast and drastic imagery functioning as a
conclusion.
A new theme is introduced with et ecce; this is the forthcoming
gladiatorial contest, which is meant to give plausibility to the
optimistic assessment of the present situation. The announcement is
followed elliptically by a general characterization of the team. Then
with et we pass to the description of the organizer, who is an
important man called affectionately Titus noster by Echion (magnum
animum habet et est caldicerebrius). From this the conclusion is
drawn (in asyndeton) that at all events something special is to be
expected. With nam a retrospective explanation is now introduced: 'I
know him, you see; he does nothing by halves'. Echion then returns
asyndetically to the character of the coming entertainment; here it
should be noted how Petronius exposes the eager crowd's primitive
cruelty in the metaphor carnarium. 46
With et we have the reference to the organizer's wealth, and with it
an explanatory comment: he has come into money. This involves a
further explanation: his father has died. The ensuing expression of
sympathy provides a footnote that now stands third in line. These
baroque offshoots of an associative mode of thought differ consi-
derably from the ordered sequence of ideas at the end of the
Ganymede speech. Yet Echion too returns to the main theme with an
43
Cicero used the same kind of language in the struggle against Verres (Verr. II
4,57,126).
44 Cf. de brev. vit. 12.
45
Cf. Friedlaender p.263.
46
J.P. Sullivan (Joe. cit. pp.232-253, esp. 252) concludes that Petronius himself had
a tendency to voyeurism from the fact that in Petronius' novel third parties often
watch the sexual act. This is no less risky than inferring from the present passage that
Petronius was a sadist. What really matters is the dispassionate and critical standpoint
from which the writer analyses the debasement that in a world of mass consumption
affects every side of life, from the culinary and sexual to the linguistic.
PETRONIUS 135
astonishingly sure touch and carries it forward: even heavy expen-

diture can be borne by the organizer and it makes him famous. With
iam Echion then passes on to a further characterization of the
participants and thereby picks up the thread of his initial remarks
about the contest (45,5).
In this way Petroni us is able to create an artistic whole even from
the disconnected ideas in an everyday conversation. Again it is the
conjunctions and logical particles that achieve clarity. Et ecce and
iam are strong opening signals. Twofold use of et emphasizes the way
the passage is subdivided: firstly the organizer's fiery temperament,
and secondly his wealth. A retrospective parenthesis is clearly
identified as such by nam. In the threefold string of comments any
undesirable prominence is skilfully avoided by asyndeton and by a
gradual reduction in sentence length (relictum est i/li sestertium
~recenties, I decessit il/ius pater. I Male!).
I Thus while the content appears to present us with a chaotic
\ assortment of empty trivialities, the form is marked by the author's
\ own clever and logical arrangement. Hence our examination of
sentence connection bears out the inference drawn from analysis of
individual words and clauses, as well as from comparison with
literary parallels: the 'vulgar' parts in Petroni us' work do not provide
us with unfiltered raw material (hence it is not safe to use them
uncritically as evidence of colloquial Latin pure and simple). 47 On the
contrary they are a complex artistic creation with high literary
pretensions, for in two different ways they impart symbolic signi-
ficance to vulgar language. In its colourfulness it becomes an
expression of an intense feeling for life, which the narrator presents in
ironic refraction with a trace of nostalgie de Ia boue. On the other
hand in its triteness it becomes a symbol of the general debasement
and divorce from reality of traditional values and other important
aspects of life.
47 For the right approach cf. now also F.M. Frohlke Petron. Struktur und
Wirklichkeit. Bausteine zu einer Poetik des antiken Romans (Frankfurt/Bern 1977)

pp.lll-145, esp. p.25.
8
Tacitus (cos. 97 A.D.)
A speech of the Emperor Claudius

Original and literary recasting
The Texts:
Senatus consultum Claudianum (oratio Claudii)
de iure bonorum Gallis dando, 48 A.D. 1
... I mae rerum no[strarum] sit u.......... 1
Equidem primam omnium illam cogitationem hominum quam I

maxime primam occursuram mihi provideo, deprecor, ne I quasi nov am
istam rem introduci exhorrescatis, sed ilia I potius cogitetis, quam
multa in hac civitate novata sint, et I quidem statim ab origine urbis
nostrae in quod formas I statusque res p(ublica) nostra diducta sit. I
Quondam reges hanc tenuere urbem, nee tamen domesticis suc-
cesl soribus eam tradere contigit. Supervenere alieni et quidam exterlni,
ut Numa Romulo successerit ex Sabin is veniens, vicinus qui Idem, sed
tunc externus; ut Anc6 Marcia Priscus Tarquinius. [Is] I propter
temeratum sanguinem, quod patre Demaratho C[oJirinthio natus erat
et Tarquiniensi matre generosa sed inopi, I ut quae tali marito necesse
habuerit succumbere, cum domi relpelleretur a gerendis honoribus,
postquam Romam migravit, I regnum adeptus est. Huic quoque etfilio
nepotive eius (nam et I hoc inter auctores discrepat) insertus Servius
Tullius, si nostros I sequimur, captiva natus Ocresia, si Tuscos, Caeli
quondam Vii vennae soda/is fidelissimus omnisque eius casus comes,
post I quam varia fortuna exactus cum omnibus reliquis Caeliani I
exercitus Etrzirili excessit, montem Caelium occupavit eta duce suo I
Caelio ita appellita[vit], 2 mutat6que nomine (nam Tusce Mastarna I ei
nomen erat) ita appellatus est, ut dixi, et regnum summa cum rei I
p(ublicae) utilitate optinuit. Deinde post quam Tarquini Superb[ mores
1 Text in Inscriptiones Latinae selectae (ed. H. Dessau) vol. I (Berlin 1902) no. 212;
P. Fabia La Table Claudienne de Lyon (Lyon 1929) pp.62ff.

2 The inscription has appellitatus; this is accepted by P. Sage 'La table claudienne et
le style de l'empereur Claude. Essai de rehabilitation' REL 58 (1980) pp.274-312 (esp.

276 and 279; however Sage also interprets duce suo as a solecism). Appellitatum is
proposed by J. Carcopino Points de vue sur /'imperialisme romain (Paris 1934) pp.l59-
199, esp. 169 and 184 (= 1959 2 , pp.l74-208, esp. 195); cf. L. Pareti Studi Etruschi 5
(1931) p.156 and A. Momigliano L' opera del/'imperatore Claudio (Florence 1932) p.31.
136
A SPEECH OF THE EMPEROR CLAUDIUS 137
inlvisi cfvitati nostrae esse coeperunt, qua ipsius quafiliorum ei[us], I

nempe pertaesum est mentes regni et ad consules, annuos magis I tratus,
administratio rei p(ublicae) translata est. I
Quid nunc commemorem dictaturae hoc ipso consulari impel rium
valentius repertum apud maiores nostros, quo in a[s]lperioribus bellis
. aut in civili motu difficiliore uterentur? I aut in auxilium plebis creatos
tribunos plebei? quid a consullibus ad decemviros translatum imperium,
solutoque postea I decemvirali regno ad consules rusus reditum? quid in
[plJulris distributum consulare imperium tribunosque mil[itu]m I con-
sulari imperio appellatos, qui seni et saepe octoni crearenl tur? quid
communicatos postremo cum plebe honores, non imperi I solum sed
sacerdotiorum quoque? lam si narrem bella, a quibus I coeperint
maiores nostri, et quo processerimus, vereor ne nimio I insolentior esse
videar et quaesisse iactationem gloriae prollati imperi ultra oceanum.
Sed illoc potius revertar. Civitat[em] I ... [lacuna]
... [po ]test. Sane I novo m[ore] et divus Aug[ustus av]onc[ulus m]eus
et patruus Ti. I Caesar omnem jlorem ubique coloniarum ac munici-
piorum, bolnorum scilicet virorum et locupletium, in hac curia esse
voluit. I Quid ergo? non Italicus senator provinciali potior est? lam I
vobis, cum hanc partem censurae meae adprobare coepero, quid I de ea
re sentiam, rebus ostendam. Sed ne provinciales quidem, I si modo
ornare curiam poterint, reiciendos puto.l
Ornatissima ecce colonia valentissimaque Viennensium, quam I
Iongo iam tempore senatores huic curiae confert! Ex qua cololnia inter
paucos equestris ordinis ornamentum L. Vestinum falmi!iarissime
diligo et hodieque in rebus meis de tin eo, cuius libel ri fruantur quae so
primo sacerdotiorum gradu, post modo cum I annis promoturi digni-
tatis suae incrementa; ut dirum nomen lal tronis taceam, et odi illud
palaestricum prodigium, q~od ante in dolmum consulatum intulit,
quam colonia sua solidum cfvitatis Romalnae benificium consecuta est.
Idem defratre eius possum dicere, I miserabili quidem indignissimoque
hoc casu, ut vobis uti/is I senator esse non possit.I-
Tempus est iam, Ti. Caesar Germanice, detegere te patribus
conscriptis, quo tendat oratio tua; iam enim ad extremos fines Galliae
Narlbonensis venisti 1-
Tot ecce insignes iuvenes, quot intueor, non magis sunt paenitendi I
senatores, quam paenitet Persicum, nobilissimum virum, ami Icum
meum, inter imagines maiorum suorum Allobrogici nolmen Iegere.
Quod si haec ita esse consentitis, quid ultra desideral tis, quam ut vobis
digito demonstrem, solum ipsum ultra fines I provinciae Narbonensis
138 TACITUS
iam vobis senatores mittere, quando I ex Luguduno habere nos nostri

ordinis viros non paenitet. I Timide quidem, p(atres) c(onscripti),
egressus adsuetos familiaresque vobis pro Ivinciarum terminos sum, sed
destricte iam Comatae Galliae I causa agenda est, in qua si quis hoc
intuetur, quod bello per del cem annos exercuerunt dfvom Iulium, idem
opponat centum I annorum imm6bilem fidem obsequiumque multfs
trepidis rei bus nostris plus quam expertum. Illi patri meo Druso
Germaniam I subigenti tutam quiete sua securamque a tergo pacem
praesl titerunt, et quidem cum [a] census novo tum opere et inadsuel to
Gallis ad bellum advocatus esset; quod opus quam arlduum sit nobis,
nunc cum maxime, quamvis nihil ultra, quam I ut pub/ice notae sint
facultates nostrae, exquiratur, nimis Imagno experimento cognoscimus.
I at least foresee that I shall be confronted first of all by what is usually men's
very first thought, and I ask you not to be outraged, as if it were an
innovation that was being introduced here, but rather to reflect how much
novelty there has been already in this commonwealth and how many
constitutional forms our state has adopted, right from the foundation of our
city onwards in fact.
Kings once possessed this city, and yet they were unable to pass it on to
indigenous successors. Strangers intervened, and even some foreigners; thus
Numa, Romulus' successor, came from the Sabines, a neighbour it is true,
but at that time still a foreigner; thus Ancus Marcius was followed by Priscus
Tarquinius. Because he was a bastard (his father was the Corinthian
Demaratos, and his mother an aristocratic but impoverished Tarquin; so she
had no choice but to take such a husband), he was excluded from state office
in his own country; when for this reason he then emigrated to Rome, he
attained the kingship. Similarly between him and his son or grandson (for
writers disagree on this point too) came Servius Tullius, who according to
our historians was the son of a prisoner called Ocresia, but according to the
Etruscans was once the faithful companion of Caelius (Caeles) Vivenna and
shared his every adversity. After the caprice of fortune had expelled him and
all survivors of the Caelian army from Etruria, he occupied the Caelian Hill
and gave it this name after his general Caelius. He changed his own name (for
in Etruscan he was called Mastarna), was given the name I used above, and
attained the kingship to the great benefit of the state. Then after the
behaviour of both Tarquinius Super bus and his sons had roused the hostility
of our citizens, they of course grew tired of the monarchy, and adminis-
tration of the state was transferred to annually elected consuls.
What is the point now of mentioning the introduction of the dictatorship
among our ancestors, a stronger form of government than that of consuls for
more difficult times of war and for more serious civil strife? Or the creation of
tribunes of the people to help the plebeians? The transfer of government
from the consuls to the decemvirs, and after abolition of the unlimited power
of the decemvirs, the return to a consular constitution? The division of
consular authority, and the creation of six to eight military tribunes with
consular imperium? Finally the admission of the plebs to positions of

honour, not only to government offices, but also to priesthoods? If I were to
give an account of the wars with which our forefathers began, and of how far
we have come, then I fear it could look as ifl were too presumptuous and had
sought an opportunity to boast of my extension of the empire beyond the
ocean. But I prefer to return to my subject ... [lacuna]
It was certainly in a 'new' spirit that both my great uncle, the divine
Augustus, and my uncle Tiberius Caesar wanted to see represented in this
curia the whole elite of all the colonies and municipia, naturally the good and
well-to-do. Well now! Does not an Italic senator take precedence over a
provincial one? I shall demonstrate to you my opinion on this by my actual
practice, when I start to make clear to you this part of my activity as censor.
But I think that even the provincials should not be turned down, as long as
they are the sort of people that would do credit to the curia.
Look at the respected and flourishing colony of Vienna. How long it has
been sending senators into our curia! From it comes L. Vestinus, an
ornament of the equestrian order, to whom I feel the closest attachment,
such as I do to few others, and whom I still to this day entrust with my
business; may his children, I pray, come to enjoy the first grade of the
priesthoods and later as the years go by rise to ever higher rank; however I
shall pass over in silence the baneful name of the brigand/ and I hate that
monster from the wrestling school, that brought home the rank of consul
before his colony was granted full Roman citizenship. I can say the same
about his brother, who is however to be pitied and does not deserve the
misfortune of not being allowed to serve you as senator.
It is now time, Tiberi us Caesar Germanicus, to reveal to the senators what
your speech is getting at; for you have already reached the extreme edge of
Gallia Narbonensis. We shall not be sorry about admitting to the Senate the
many admirable men I see before me, any more than my friend, the excellent
Persicus, is sorry to read the name Allobrogicus on the images of his
ancestors. But if you agree with this, what else are you asking than for me to
point with my finger for you: the ground beyond the frontiers of the province
of Narbo does indeed send senators to you already, for to our joy there are
men of our rank from Lugudunum among us. It is only with timidity,
senators, that I have overstepped the province boundaries you know so well,
but the cause of Gallia Comata must now be resolutely pressed; if in this
connection anyone takes into account the fact that for ten years the Gauls
caused trouble to the god Julius in war, then he should also set against this
their immutable loyalty for over a hundred years and their obedience that
has been more than tested in many critical situations. While my father
Drusus was subjugating Germany, their quiet behaviour gave him the cover
of a safe and carefree peace, and this was at a time when he had been called
away to the war from the business of the census which was then something
new and unaccustomed for the Gauls; we are learning at this very moment
from all too abundant experience how difficult this business is for us,
although no more is being asked than that our resources should be officially
known.
3
Valerius Asiaticus.
140 TACITUS
Tac. ann. 11,244

Maiores mei, quorum antiquissimus Clausus origine Sabina simul in
civitatem Romanam et in familias patricio rum adscitus est, hortantur
uti paribus consiliis in re publica capessenda, transferendo hue quod
usquam egregium fuerit. neque enim ignoro lulios Alba, Coruncanios
Camerio, Porcios Tusculo, et ne vetera scrutemur, Etruria Lucaniaque
et omni ltalia in senatum accitos, postremo ipsam ad Alpes promo tam,
ut non modo singuli viritim, sed terrae, gentes in nomen nostrum
coalescerent. tunc so/ida domi quies; et adversus externa jloruimus,
cum Transpadani in civitatem receptz~ cum specie deductarum per
orbem terrae legionum additis provincialium validissimis fesso imperio
subventum est. num paenitet Balbos ex Hispania nee minus insignes
viros e Gallia Narbonensi transivisse? manent posteri eo rum nee am ore
in hanc patriam nobis concedunt. quid aliud exitio Lacedaemoniis et
Atheniensibus fuit, quamquam armis pollerent, nisi quod victos pro
alienigenis arcebant? at conditor nostri Romulus tantum sapientia
valuit, ut plerosque populos eadem die hastes, dein cives habuerit.
advenae in nos regnaverunt; libertinorum filiis magistratus mandare
non, ut plerique falluntur, repens, sed priori populo facti tatum est. at
cum Senonibus pugnavimus: scilicet Vulsci et Aequi numquam ad-
versam nobis aciem instruxere. capti a Gal/is sumus: sed et Tuscis
obsides dedimus et Samnitium iugum subiimus. ac tamen, si cuncta
bella recenseas nullum breviore spatia quam adversus Gallas con-
fectum: continua inde ac fida pax. iam moribus artibus adfinitatibus
nostris mixti aurum et opes suas inferant potius quam separati habeant.
omnia, patres conscripti, quae nunc vetustissima creduntur, novafuere:
plebei magistratus post patricios, Latini post plebeios, ceterarum
Italiae gentium post Latinos. inveterascet hoc quoque et quod hodie
exemplis tuemur, inter exempla erit.
My forefathers, of whom the eldest, Clausus, came from the Sabines and was
raised to patrician rank at the same time as he obtained Roman citizenship,
urge me, on taking over state affairs, to decide in the same spirit, by
transferring here all that has anywhere achieved distinction. For I am well
aware that the Julians were called to the Senate from Alba, the Coruncanians
from Camerium, the Porcians from Tusculum and, so as not to pursue
stories from the past any further, that people were called to it from Etruria,
Lucania and the whole of Italy, and that in the end Italy itself was extended
4 Text of H. Heubner (Stuttgart 1983). Commentary by E. Koestermann, vol. III
(Heidelberg 1967). Review of the most recent scholarship in F. Romer Anzeiger fur die
Altertumswissenschaft (hrsg. v. d. Humanistischen Gesellschaft) 37 (1984) pp.154-208
and 38 (1985) pp.129-204 (for 1973-1982).
as far as the Alps, so that not only single individuals, but countries and even
races merged with our name. At that time internal peace was guaranteed; and
against outside dangers we were at the height of our power, when the
Transpadanes obtained citizenship, when all over the world the most
powerful provincials were admitted to our ranks under the pretence of
settling legions and thus help came to the exhausted empire. Or are you sorry
that the Balbi came over from Spain and that equally distinguished people
came from Gallia Narbonensis? Their descendants are still among us and feel
no less love .for our country than we do. What else destroyed the Spartans or
Athenians, for all their military achievements, save the fact that they kept the
defeated apart from themselves as aliens? Yet Romulus, the founder of our
state, possessed such great wisdom that he regarded most races on one and
the same day first as enemies and then as citizens. Among us, foreigners had
the rank of kings. Entrusting state office to sons of freedmen is no sudden
innovation, as most people wrongly believe, but an ancient custom in our
nation. 'But we did after all fight with the Senones!' So the Volsci andAequi
have never deployed their forces against us? 'The Gauls captured our city'.
But we also had to give hostages to the Etruscans and go under the yoke of
the Samnites. Yet, if one runs through every war, none was ended in less time
than the one against the Gauls; from then on, without interruption, there was
profound peace. Now that they are permeated by our character and our ways
and are connected with us by marriage, they should rather bring their gold
and treasures to us than keep them for themselves. Everything, senators, that
is now considered very old, was once new: plebeians as officials after the
patricians, Latins after the plebeians, representatives of the other races of
Italy after the Latins. This too will one day be old, and what today we defend
with precedents, will rank among the precedents.
A. Overall structure and sequence of ideas

1. Claudius
After a lacuna our text starts with the praemunitio that one should
not reject Claudius' proposal as an innovation but reflect that the
whole of Roman constitutional history has been one long process of
renewal. (The two occurrences of primam seem to indicate that this
objection, which receives such strong emphasis as 'the first', cannot
have been preceded by very much more than the main announcement
of the theme).
Next come examples. The first is that, as long as kings ruled, there
were foreigners among them (here Claudius frequently strays into
biographical details). Secondly, the Roman constitution has often
changed. The second section does not quite form a strict conceptual
whole. It is true that everything in the enumeration fits the theme of
'changes in the Roman constitution', but only part of it fits the point
142 TACITUS
at issue, 'extension of the class of people entitled to participate in

government'.
In the form of an appendix Claudius speaks of the Roman empire's
military conquests and its outward expansion; this is something that
really deserved full treatment as a separate item, because it is linked
ultimately with the need to give the incorporated peoples a share in
government. However Claudius oddly describes this approach as a
digression; for after it he explains that he is now returning to the
subject. Yet at this point he deals with the extension of Roman
citizenship to the defeated peoples, which follows naturally from the
aspect he has just mentioned. After the ensuing lacuna the further
point is made that the right of actual representation in the Senate was
granted by Augustus and Julius Caesar to the leading men in colonies
and municipia. Without encroaching on the pre-eminence of Italic
senators, Claudius then stresses that the provincials should not be
ignored.
Claudius mentions a specific community and particular indi-
viduals on whose behalf he applies for privileges or whom he abuses.
After telling himself to come to the point, Claudius finally draws
attention to the many distinguished men who deserve to be senators.
By mentioning Persicus 5 the emperor shows that he is not just an
innovator, but also a respecter of tradition. In the next sentence he
recalls that men from Lugudunum are already in the Senate.
It is true that Caesar waged war against the Gauls for ten years;
since then however they have been loyal to Rome for a hundred
years. When Drusus fought the Germans, the Gauls remained quiet,
though at that particular time the census was being carried out
among them. Claudius now knows from experience how many
problems are involved in a census even in Rome.
Vittinghoffli assumes that Tacitus' hard-headed argument to the
effect that the Gallic senators should bring their money to Rome was
immortalized in Gaul itself on the bronze tablet in Lyon: this absurd
view needs no refutation. 7 It is therefore impossible to see why he
5 Q. Fabius Maxim us Allobrogicus had at one time triumphed over the Allobroges.
Hence Paullus Fabius Persicus might well have taken the strongest exception to the
admission of Allobroges into the Senate (F. Vittinghoff 'Zur Rede des Kaisers
Claudius .. .' Hermes 82 [1954] pp.348-371, esp. 360). Koestermann, comm. p.79,
grotesquely calls Persicus an 'Allobrogian'.
6 Loc. cit. (n.5) pp.36lf.
7 Nor does Vittinghoffs assumption carry conviction (pp.364f.) that at the end
Claudius needs to mention his real purpose once again and prepare the way for the
Senate's decree. There is no proof that straight after the speech the decree itself was
postulates any continuation at all, especially as the archaeological

data offer no support for it.
2. Tacitus
The best men in the provinces should be attracted to Rome: this had
always been the case. The ancestor of the Claudians came from
Sabine territory; the same is true of many famous families. Italy itself
was extended as far as the Alps; so it was not just individuals, but
whole nations that were granted the citizenship. Examples are given.
The Ttanspadanes added new strength to the exhausted empire.
Roman patriotism is shown by descendants of Spanish citizens and
men of Gallic origin. Then comes an example to the contrary: Sparta
and Athens were ruined by their racial exclusiveness. Romulus by
contrast accepted enemies as citizens; foreigners became kings and
even in the early days sons of freedmen became officers of state. The
imaginary objection that the Senones waged war against Rome is
refuted by the observation that so did Vulsci and Aequi. The second
objection is that the Gauls conquered Rome; it is refuted by the fact
that Rome was also humiliated by the Etruscans and Samnites.
Moreover the Gallic war was brought to an end with particular
speed, and since then the Gauls have been unwaveringly loyal. They
are already Romanized and should make the benefits of their wealth
available also in Rome. In the course of Roman history the ius
honorum was extended to an ever wider class of people. The present
step too is only part of a broader development.
3. Comparison
a) What does Tacitus omit?
Tacitus' speech is considerably shorter than the original. He passes
over the complicated discourse on the period of the kings (one of
Claudius' fields of study 8), the sketch of Roman constitutional
history, the reference to Claudius' wars and successes in Britain, the
written on a supposed third tablet. A far more natural assumption is that the decree
was mentioned in the lost opening of the tablet and explicitly identified with the
emperor's words as subsequently quoted. If we assume that the general heading
extended in somewhat larger letters over the whole double tablet, it is clear that the
lacunae at the start and in the middle of the speech do not need to be very big. Hence
the conclusions we have drawn from the content are also in line with the external data.
A brief senatorial decree at the start of the tablet is also posited by E. Schoen bauer Jura
6 (1955) p.l69. On the interpretation of the last part of the inscription cf. now Sage
pp.305-312.
8 Inter alia Claudius wrote twenty books of Etruscan history in Greek (Suet. Claud.
42,5).
144 TACITUS
allusions to measures of Augustus and Tiberius, 9 and the mentions of

Claudius' personal friends and the application for privileges on their
behalf.
Tacitus deliberately restricts himself to the immediate question at
issue and keeps it more sharply in mind than the emperor, who thinks
more in terms of the generc;tl development of the Roman constitution
and also adduces more details. To a degree these two elements in
Claudius are interdependent. Since he proposes admission to priestly
office for certain persons, he is also justified in not restricting himself
to political magistracies in his historical survey. By contrast the
limitation in Tacitus is characteristic: plebei magistratus post patricios,
Latini post plebeios, exterarum ltaliae gentium post Latinos. Tacitus
thus omits what has only limited historical reference and does not
belong to the subject. On the other hand there is no trace of Claudius'
humour.
b) What does Tacitus add?
Since Claudius' original speech is not preserved in its entirety, there
cannot be complete certainty on this point. 10 Given the state of the
texts, we can make the following observations.
Firstly, Tacitus brings out more clearly the principle behind
Claudius' action: ut non modo singuli viritim, sed terrae, gentes in
nomen nostrum coalescerent. He thus encapsulates the basic idea, 11
which in Claudius is there in the background while he is speaking of
the Romans' wars and the expansion of the empire even beyond the
Ocean.
Secondly, Tacitus also emphasizes the reliability of the Gauls, 12
but he goes beyond this and stresses the truly Roman patriotism of
those senators that come from the provinces. Among these he
significantly gives prominence to the Spaniards, who in his day were
9 However deductarum ... legionum presumably refers principally to colonies
founded by Augustus.
10 Vittinghoff Joe. cit. emphasizes the fragmentary state of the original speech and
the possibility of an intermediate (historical) source for Tacitus. For direct use of the
acta Senatus cf. F.A. Marx 'Untersuchungen zu Tacitus' Annales' Hermes 60 (1925)
pp. 74-93, esp. 82-90. Direct use of the original speech is also assumed by Carcopino
(1934 p.168; 1959 p.183).
11 The same basic concept is already found in Sallust epist. 2,5,7. According to
Vretska C. Sallustius Crispus, Invektive und Episteln, hrsg., ubers. und komm.
(Heidelberg 1961) vol. II pp.l15f., Sallust is thinking here of the enfranchisement of
the very Transpadanes mentioned in our Tacitus text as already a historical precedent.
12 Both authors therefore pass over in silence the revolt of Sacrovir under Tiberius
(Fabia p.6).
especially important. 13
Thirdly, Tacitus draws attention to what is typical about the case
by means of a historical example illustrating the opposite: Athens
and Sparta brought political ruin on themselves because of their
inability to overcome the prejudices of blood. This is certainly a
striking example, but one that is more likely in a rhetorically trained
man of letters than in a politician. Here Tacitus seems exceptionally
to show a wider historical range than Claudius; however his purpose
is to give the basic idea of the speech all the more prominence
through an example to the contrary. He thus offers an interpretation
of the emperor's action in setting it against a different type of
behaviour. This is in essence an interpretation concerned with morals
and with patterns of history, which does not deal primarily with the
historical event but presents 'anthropological' generalizations. In
Claudius the frame of reference is the whole development of the
Roman constitution; in Tacitus it is the potentialities of human
behaviour. 14
Fourthly, in contrast to Athens and Sparta Tacitus then introduces
Romulus, who made full citizens out of people that on the same day
had been enemies. This accords, both overall and in detail, with his
technique of bringing together opposites.
Fifthly, the Gauls were not just enemies of Rome; they actually
conquered it. Tacitus makes a stronger appeal to the imagination.
The refutation is hardly less vivid (iugum subiimus). This is evidence
for his method of striking illustration in the form of 'mental
pictures'. 15
Sixthly, the historian uses two arguments that would have been out
of place on an inscription published in Gaul: the military strength
and the finances of the province should now benefit Rome. In line
13
The emperor Trajan was a Spaniard. The subject of 'new citizens in the Senate'
also had importance for Tacitus in view of his own career and that of his admired
father-in-law Agricola. E. Paratore Tacito (Rome 19622) pp.498ff. is of course right to
stress that subjectively an extension of the ius honorum need not necessarily have been
welcomed byTacitus.
14
K. Wellesley 'Can You Trust Tacitus?' Greece and Rome 2nd ser. I (1954) pp.14-
33 fails to see this when he says: 'a statesmanlike utterance is turned into a school
exercise' (p.31). Value judgments like this depend on whether an observer is more
interested in 'how it really was' or 'how it always will be given human nature' or in
both. Ultimately they say more about the observer than about the thing observed.
15
Cf. H. Hommel Die Bildkunst des Tacitus Wiirzburger Studien 9 (Stuttgart 1936)
pp.116-148, esp. 119 n.12: Tacitus' 'mind's eye' style (after P. Friedlander in Norden
Einl. in die Altertumswissenschaft3 I 4 [1923] p. 81 ); now in Symbola I (Hildesheim/New
York 1976) pp.365-392, esp. 367 n.l2.
146 TACITUS
with the practice of speeches in classical historians, no further

significance is attached to such short-term factors.
Seventhly, Tacitus asserts the foreign origin of the most important
patrician houses, including the Claudians; however this point may
also have occurred in the lost opening of the original speech. What
the historian has undoubtedly added, as we saw, is the mention of the
new Spanish senators: if the purpose is to supply evidence for the
speech's objective, Tacitus himself can use examples freely. On the
other hand it is significant that as a senator he passes over Claudius'
characteristic reference to his 'great-uncle' Augustus and his 'uncle'
Tiberius; he thus eliminates the dynastic frame of reference, but not
the traditional Roman clan-consciousness in the appeal to the
maiores. This is an important shift of emphasis.
Eighthly, Tacitus objectifies the immediate event by setting it
within the historical process (inveterascet hoc quoque). He also
stresses the exemplary character of the step for the future. Both these
points are in line with his tendency to think in terms of the
'exemplum'. It is however true that this general aspect does not
appear until the end of the speech.
The writer's approach has thus freed the speech from its limited
historical reference and in Tacitean fashion invested it with para-
digmatic significance as an instance of universal human nature.
c) Arrangement
Claudius states at the beginning that the proposed step is new.
Tacitus saves this idea to the end and allows it to suggest itself
through the series of examples. At the end he skilfully combines it
with a short recapitulation of the extension of the ius honorum.
Tacitus begins very abruptly with an example and it is not until the
end of the first sentence in a kind of supplementary ablatival gerund
that he introduces the basic concept. He thus allows the general point
to develop out of the particular in a matter of detail as well as overall.
His procedure here is of course very far from being indiscriminate;
this is shown by the structure. In the original speech the emperor
strings the historical events together chronologically without always
distinguishing the different aspects clearly from each other. In
Tacitus on the other hand the grouping of the exempla is determined
by their role within the argument: foreigners became kings in Rome.
Placed impressively in the middle of the speech, the sentence has a
more powerful effect than the many names and details at the opening
of the emperor's real speech. However Claudius' wide-ranging
historical panorama is replaced in Tacitus by just a functional

excerpt at the end. 16
Suetonius said of Claudius' style: scripsit magis inepte quam
ineleganter. 17 Here inepte means neither 'crack-brained' nor 'clumsy',
but denotes a certain 'inappropriateness', the opposite of aptum.
Whereas Tacitus adapts every detail of his thought to subject, type of
situation, and personal status, Claudius likes to indulge in digressions
and sometimes shows his humour without too much regard for
dignitas.
So much then for ineptia of content in Claudius and for Tacitus'
feeling for aptum. 18 The following section will show how far the
emperor may after aH have found the tone that suited his situation; it
will also clarify what we are to make of Claudius' stylistic elegantia
and what the contrasting characteristics are that occur in Tacitus.
B. Language and style

1. Claudius
A characterization of Claudius' speech in terms oflanguage and style
has to move between two extremes. 19
a) Conventional style
In the vocabulary of the speech one is struck by certain political
catchwords of the imperial era, for example quies and pax (2,36f.) as
important civic values, and boni et locupletes as a qualification for
admission to the Senate (2,5). Another characteristic feature are the
stereotyped superlatives ( elatives) used to designate the persons and
16
On the apparent widening of the historical perspective in Tacitus see above
pp.l45 and 146.
17
Suet. Claud. 41. On Claudius' elegantia cf. also Tac. ann. 13,3,6. A subtle
interpretation of Tacitus' estimate is given by Sage p.297: Claudius possessed
elegantia, but only when he meditata dissereret. The Lyon inscription preserves (in
' Sage's view) an improvised speech (on this cf. below p.l52).
18
When N.P. Miller RhM99 (1956) p.314 interprets the 'antiquarian arguments' of
the speech in Tacitus as an attempt to characterize Claudius as a pedant, she fails to see
that here Tacitus is replacing antiquarianism with anthropology, and is thus 'Tacitean'
in the choice of arguments as well. Miller herself recognizes that the style of the speech
is Tacitean. On the different criteria with regard to aptum in language cf. H. La us berg
Handbuch der literarischen Rhetorik (Munich 1960) pp.507-511.
19
An excellent survey of all 'particularites grammaticales et stylistiques de Ia Table
claudienne' is given by Sage pp.276-285.
148 TACITUS
communities on which the emperor bestows his favour: ornatissima 20

ecce colonia valentissimaque (2,10); L. Vestinumfamiliarissime diligo 21
(2, 12); de fratre ... miserabili quidem indignissimoque hoc casu (2, 17f. );
nobilissimum virum (2,25); to the same class belong terms that are
only apparent expressions of feeling: ins ignes iuvenes (2,24); ami cum
meum (2,25). Abuse of a political opponent as 'brigand' and 'monster
from the wrestling school' is also conventional (2,15). Here it is
wrong to detect the unbridled temperament of the tyrant;2 2 rather we
should realize that language like this, which is also familiar from
Cicero, is not an outburst of passion but a cliche.
However political speeches are also characterized by a certain
caution and restraint. One example of this is the choice of verbs. The
subjective puto (2,9) and the imploring quaeso (2,13) are formulae of
modesty; such empty phrases are also part of the stock-in-trade of
speeches and motions in the Senate. 23 A similar tendency is found in
vereor ne ... videar (1,39f.) and timide quidem (2,30). Caution is also
indicated by the indirect form of polemic and debate: si quis hoc
intuetur ... idem opponat (2,32-34). On the other hand the repeated use
of non paenitet in the sense of 'we are glad' is a piece of diplomatic
understatement.
The tone ofthe speech has a decisive effect also on syntax and style.
This explains elements that are only apparently('pleonastic', 1e.g.
equidem primam omnium illam cogitationem hominum~quam maxime
primam occursuram mihi provideo: 'that I shall be confronted first by
what men always think first of all'. Here the repetition has a
generalizing function that weakens the objection from the start.
Duplications are also part of the traditional Curial style: 24 formas
statusque (1,7); coloniarum ac municipiorum (2,3); bonorum ... et
locupletium (2,4); ornatissima ... valentissimaque (2,9);familiarissime
diligo et hodieque in rebus meis detineo (2,llf.); miserabili quidem
20 In our text the use of ornare and related expressions is entirely stereotyped: cf.
also ornare curiam 2,9; equestris ordinis ornamentum 2, 11. The term ornamentum also
occurs Sen. ad Polyb. 14,3 (on the 'typically Claudian' style of this speech cf. H.
Dahlmann Hermes 71 [1936] pp.374f.; D.M. Last and R.M. Ogilvie 'Claudius and
Livy' Latomus 17 [1958] pp.476-487, esp. 485).
21 Diligo is fairly rare in Tacitus, cf. R. Syme Tacitus (Oxford 1958) p.345.
22 Vittinghoff p.362: 'The furious attack on Valerius Asiaticus, without any basis in
fact, is sufficient· indication of the character of an impulsive man who lacks self-
control.'
23 Cf. e.g. the use of arbitror in Cato's speech for the Rhodians (above p.l8).
24 Here it is not a question of synonyms, but of the stylistic co-ordination of related
concepts. On the frequent occurrence of duplications in col. 2,3f. cf. Sage p.284.
Claudius may actually have intended the sequence of heavy genitive plurals.
indignissimoque hoc casu (2, 18); fidem obsequiumque (2,34).

Certain forms of traductio (ploke) also belong to the traditional
style of the Curia. We are reminded of the ancient type diem, quo die
when Claudius says: non magis sunt paenitendi senatores, quam
paenitet (2,23f.; cf. also 2,29 non paenitet).
b) Choice expression and historiographical style
There is more to the speech than just the conventional traits of the
Curial style. Thus where duplication is used we find a tendency to
invest it with fresh brilliance by means of additional devices. The
collocation fidem obsequiumque has a more vivid impact because the
marginal position of the elements modifying each word gives rise to a
chiasmus. The traductio mentioned above achieves pointedness
thanks to antithesis. Synonymous adjectives often have other words
inserted between them (above 2,4 and 2,9); alternatively the second
element is expanded (above 2,11f.; 2,18).
While most features of the Curial style are found in the second
column and in the opening lines of the first, the main part of this first
column is rather different. With quondam reges hanc tenuere urbem
Claudius enters the sphere of historiography in regard both to
content and style. 25 We can compare the opening of Tacitus' Annals:
urbem Romam a principia reges habuere. The same 'distinguished'-
sounding termination -ere occurs in Claudius, where it is also
retained in the following sentence (supervenere 1,9). 26 The explana-
tory parentheses with nam (1, 16 and 22) are likewise typical of the
historiographical style. The intensive appellito 27 also fits the archaic
25
It is no surprise to find in our text many phrases from Livy, Claudius' admired
teacher (this is fully treated by Last and Ogilvie Joe. cit.; although they definitely go too
far in classing expressions like prim am omnium or qua- qua among the reminiscences,
this still leaves such characteristic echoes as supervenere; discrepat inter; invisi civitati;
trepidis rebus; securam ... pacem praestiterunt; bello per ... annos exercuerunt and
various others). For a comparison with the speech of Canuleius (Liv. 4,3f.) see A.
Momigliano (cited above p.l36 n.2) p.38 (with older lit.). The emperor presents
himself as a tribune of the people.
26
On the other hand the 'normal' ending coeperunt is used with great sensitivity in a
non-narrative section concerned with appraisal (1,25).
27 In the archaizer Gellius the verb occurs often with a frequentative sense. On one
occasion in Tacitus it is also used intensively; oddly enough this too is in connection
with the mons Caelius (ann. 4,65). The semantically striking use of this very rare verb in
the same material context suggests that it occurred in an early historian, on whom
Tacitus and Claudius are ultimately dependent. It is unlikely that here too Tacitus is
drawing on Claudius, since the matter presented in Tacitus is different. The passage
from Claudius' speech is strangely absent from the Thesaurus, although it is the oldest
direct evidence for the word. However Claudius cannot have been the first to use the
word in the sense of appellare (i.e. as an intensive, not a frequentative), for Paul. Fest.
!50 TACITUS
colouring of the historical part of the speech.

The way important concepts are assigned to less important parts of
the sentence at times indicates a striving after asymmetry. Participial
constructions 28 that can show considerable refinement play a
particular role: these are still relatively rare in Cicero. However this
practice, which is also to be found in Tacitus, is not put into effect by
Claudius at all strictly; what gives Claudius' speech its distinctive
character is neither the studied asymmetry nor the preciosity, but a
very objective style that is characterized by Suetonius without
particular commendation as 'non ineleganter'. 29
c) Unconstrained style
Other elements seem to point to an oral quality about the style that
on occasion almost suggests improvisation. We can cite cases of
repetitiveness like ornare- ornatissima- ornamentum (close together
at 2,9-11 and without any recognizable stylistic purpose, especially as
there is no close conceptual link between the sentences); cf. also ita
appellitavit- ita appellatus est (1,22f.).
At times it seems possible to identify traces of oral improvisation 30
in the sentence structure as well. We should note how in the following
sentence the construction develops gradually and undergoes a
certain shift ( 1,5-7) ... cogitetis, quam multa in hdc civitate novata sint,
et I quidem statim aborigine urbis nostrae in quodformas statusque res
p(ublica) nostra diducta sit. Initially et quidem adds a supplementary
adverbial phrase which could mark the end of the sentence (urbis
nostrae); however this addition develops into an independent new
27M.(= 24 L.) presumably goes back via Festus to Verrius Flaccus (Augustan period),
who in turn must have taken the word, which he thought needed explaining, from the
common early tradition that is used by Claudius and Tacitus. There is an interesting
piece of evidence to support this. In Paulus, Muller reads: appellitavisse : <saepe>
appellasse. The added word is a modern supplement and is out of place in Claudius and
Tacitus and the tradition they depend on. It should be rejected, since frequentative
appellitare = saepe appellare was well-known and needed no explanation for a Roman;
on the other hand appellitare = appellare (for giving names) was unusual, and Verrius
recorded it.
28 E.g. the future participle promoturi; this is presumably not dependent on quae so
with ellipse of sint (on the sequence of tenses in final object-clauses see Kuhner-
Stegmann II p.l80), but is a loosely attached predicative participle. Cf. also patri ...
subigenti, even though this second construction is known to Cicero.
29
The pleasure Claudius takes in the 'right' term will become even clearer in
comparison with Tacitus.
30
The formpoterint, for which there are occasional parallels (Sommer Hdb. p.531:
erint [auct. Her. 1,10; 2,10] is 'scarcely old'; aderint in inscriptions from the 1st cent.
A.D.), may be due to the workman who inscribed the tablet. Cf. Sage p.277: 'Si Claude
a reellement dit poterint, i! a commis un quasi-barbarisme.'
statement with its own subject and predicate.

The historical part of the speech is also marked by the same
unconstrained style that allows the idea to unfold gradually with the
sentence. The result is a very singular blend of both historiographical
solemnity and a style that adds things on in a manner close to
ordinary speech. Here one idea engenders the next as clarification.
The statement propter temeratum sanguinem is put in concrete terms
in a supplementary quad-clause; we here have a reference to the
mother's poverty, which is in turn explained more fully in a further
subordinate clause. However Claudius has not lost sight of his goal in
the meantime, for in two further subordinate clauses he now heads
resolutely for the close that contains the main idea: cum domi
relpelleretur d gerendis honoribus, postquam Romam migravit, I
regnum adeptus est. Though at first the emperor indulged his taste for
relating a few details in conversational tone from his abundant
historical knowledge, he is in no danger of losing the thread; on the
contrary he follows the seemingly chaotic diastole with a definite and
systematic systole. The same thing can also be observed in the
following sentence. First the basic idea 'between him and his son
came Servius Tullius' is expanded by an annotation to the wordji/ia:
'or his grandson'. This explanatory remark is then followed by
another involving a reference to the sources: 'for writers disagree on
this point too'. All these annotations are put, not at the end of the
sentence, but in its centre; the important final position is reserved for
the grammatical subject Servius Tullius. (It is noteworthy that here
the verb forfeits its traditional place, 31 in order to give greater
emphasis to the proper name.) This name is followed by two
observations. Firstly, according to Roman tradition Servius is the
son of a prisoner (captiva natus Ocresid: the important words are on
the outside). Secondly, in the Etruscan version he is the comrade-in-
arms of Caelius (Caeles) Vivenna. If the sentence finished here,
nothing would be missing; insertus would then be an elliptical main
verb. However Claudius now expands his review of the Etruscan
version and at the same time moves gradually towards the real goal of
the sentence: postquam ... Etruria excessit has a similar function to
cum domi relpelleretur d gerendis honoribus in the previous sentence.
It is true that before he attains the kingship we also have a report of
his occupation of the Caelian Hill and of how both were given new
names. This involves use of parenthesis and rather long-winded
31
It is true that with predicative participles this is often the case.
152 TACITUS
repetition and periphrasis. On the whole however the sentence we

have just examined reveals the same principle of construction as the
previous one, only in extended form: beginning, annotations, main
point, annotations, gradual return in the temporal clause, annota-
tions, main point. In the following sentence we again notice a
tendency for the annotation to move into the centre (qua ipsius qua
filiorum eius). Again the crucial political change comes at the end. 32
Thus the section considered so far does indeed exhibit elements of
a looseness of structure that is close to oral expression. However this
is no reason to regard Claudius' style as associative and erratic, or
even as symptomatic of a weak personality. It would after all be
strange if from the middle of the sentence onwards Claudius'
personality consistently changed back to being a strong one, aiming
resolutely for the thematically relevant end.
This approach is therefore deliberately chosen and has two
components: the first is artistic and the second practical and
psychological. The artistic component is the attempt to achieve
greater colour and movement by breaking up the surface. This is the
same approach that leaves the walls and columns of the contem-
porary Porta Maggiore in Rome in a state -of half-finished rusti-
cation33 and thereby makes the light that shines in softer. The
psychological component is barely separable from the artistic one.
There is so much less rigidity to a style of argument which does not
aim straight for its goal without regard for the audience, but takes
them first on a pleasant trip along the seeming byways of early
Roman history. The cosy oral manner and the scholarly absent-
mindedness of 'dear old Emperor Claudius' are also a device he uses
to influence the audience diplomatically.
2. Tacitus
Our discussion follows the text. Each heading indicates a stylistic
device that is especially prominent in the sentence under consi-
deration. However mention is also made of the other recurrent
stylistic features so as to present a complete picture of the text.
a) Main idea in subordinate position
The very first sentence contains an unusual feature. The main idea
does not in fact occur in the main clause, but in an ablatival addition:
32 A detailed analysis of the two periods just discussed is to be found in Sage pp.281-
284 and 296, with reference to sentence structure and improvisation.

33 L. Curtius-A. Nawrath Das antike Rom 4 , ed. E. Nash (Vienna and Munich 1963)
pp.209f.
A SPEECH OF THE EMPEROR CLAUDIUS !53
transferendo 34 hue quod usquam egregium fuerit. The effect of this

stylistic device is not to devalue or modify the idea, for this is given
sufficient weight by the final position itself. However the ablatival
form adds an impersonal and self-contained element that belongs to
the monumentality of Tacitus' style. ..· ~
The second sentence shows twofold ellipse of essclhis again is a
typical piece of compression. At the same timeTacitti~ demonstrates
his feeling for climax. We twice have a sequence of increasing
magnitude: then in non modo ... sed a further climax is achieved. This
sentence too has an ending that is important in its content and form.
The main idea of the first sentence is developed and surpassed, as
transplanting the best foreigners to Rome is replaced by the more
comprehensive idea of general Romanization. The important con-
cluding idea again comes in a subordinate position. This will often be
the case in the following sentences. In Tacitus the 'exception' almost
becomes the rule.
b) Two concepts superimposed
The form in which the process of Romanization finds expression is I
typically Tacitean: in nomen nostrum coalescerent. The peculiar \
appeal of this collocation is due to the fact that preposition and prefix
are not identical: in signifies a 'growing into' the Roman name,
whereas co- on the other hand indicates a 'growing together'. The
translator or commentator is obliged to undo the 'yoke' with which
the writer has combined two not wholly congruent ideas; this is a
difficulty constantly facing anyone who is concerned with Tacitus'
style. A similar phenomenon recurs in our text later on (see p.155
under 'f).
In virtually no other writer is zeugm~ so marked a pattern of
thought as in Tacitus. 35 The slight clash here between preposition
and prefix will also have been felt by a Roman. The phrase owes its
particular density to the intentional creation of two different levels.
We could say that Tacitus places one musical note on top of another,
instead of letting us hear them one by one.
34 On 'ablative of the gerund with accusative' cf. A. Draeger Ueber Syntax und Stil
des Tacitus (Leipzig 1882 3) p.81, with further examples; on additions to the sentence
see especially F. Klingner 'Beobachtungen iiber Sprache und Stil des Tacitus am
Anfang des 13. Annalenbuches' in: V. Posch! (ed.) Tacitus Wege der Forschung 97
(Darmstadt 1986 2) pp.557-574; K. Seitz Studien zur Stilentwicklung und zur Satz-
struktur innerhalb der Annalen des Tacitus (Diss. Marburg 1958); for a specific
treatment of the abl. abs. as the additional element cf. R. Enghofer Der Ablativus
abso!utus bei Tacitus (Diss. Wiirzburg 1961) pp.39-46 and 130-138.
35 On zeugma in Tacitus cf. Draeger p.107. For related expressions cf. pp.105f.
!54 TACITUS
c) Asymmetry
The third sentence again contains two ellipses. Here again the main
idea, coming as it does at the end, is not in the main clause. However
the most typical feature of the sentence in question is its asymmetry.
So/ida domi quies is expressed impersonally. The personal adversus
externa jloruimus is co-ordinated with this construction by means of
simple et. The constructions of the following cum-clauses are in
reverse sequence: the first is personal, the second impersonal.
Owing to its elliptical character the fourth sentence is very nearly
ambiguous. That num paenitet means the same as num vos paenitet
has to be guessed from the context; otherwise we have to consult
Claudius' original speech to make sure. 36 The idea of naturalization
and of moving to Rome for good again comes emphatically at the end
of the sentence, and again it is not a main verb that expresses it.
d) Compactness
The constructive half of the speech is rounded off by a sentence
whose corner-stones are two finite verbs (manent ... concedunt) and
whose centre stresses the Roman patriotism of the new citizens
(amore is in the middle of the sentence).
e) Overstatement 37 and contrast
With quid aliud Tacitus now puts forward an example to the contrary
that carries immediate conviction. The main idea, which this time is
the rejection of foreign assimilation, is once again moved to the end,
and again stands in a subordinate clause. The contrasting positive
example of Romulus likewise puts the emphasis on the end: this time
it is in a consecutive clause. The following sentence, the shortest so
far, overstates the general idea: in Rome even the kingship was open
to foreigners. And then comes another extreme case: sons of
freedmen obtained state office. Tacitus had already paved the way
for this kind of overstatement earlier, when he made the trans-
formation from enemy to citizen take place within a single day. Cases
like these are meant to serve as a contrast to set off the innocuousness
of the present step. Now for the first time there is an idea at the end of
the sentence that differs from what has gone before: the extension of
the ius honorum accords with ancient custom. The importance of this
fact is underlined by the choice of vocabulary: factitatum est is a
36
Col. 2,23; 24; 29.
37
On the subject in general cf. B.-R. Voss Der pointierte Stil des Tacitus (Munster
2
1980 ).
frequentative. 38 To this we should add the striking rhythm (double

trochee).
f) Conciseness
An imaginary objection follows: the Senones were our enemies. This
is countered by the argument that so were the Volsci and Aequi. Then
comes a second objection: the Gauls conquered our city. The
counter-argument here is that the Etruscans and Samnites have
already humiliated us as well.
In these four sentences the style is very concise: the facts are simply
set one against the other. Initially at serves to indicate the logical
relationship; in the fourth sentence sed does this. The third sentence
is put in asyndetically; the second contains an ironic scilicet.
This is followed by two arguments in the Gauls' favour. Quickly
subjugated, they have been staunchly loyal to the Romans ever since.
Besides, they should now invest their wealth also in Rome. Here
again typically _'}'<tcitean ellipses occur, such as we have often
encountered throughout ourtexf(iialtum ... confectum: continua inde
... pax). In addition we have a collocation that is close to zeugma:
moribus artibus adfinitatibus nostris mixti.
At the end the perspective broadens. All that is now old was once
new. The ius honorum was constantly extended to cover a wider class
of people. Today's decision will also be among the precedents in
future. The final section has been given greater animation than those
preceding it. This is due in part to the technique of introducing an
objection and rebutting it with counter-arguments, as well as to the
exceptional conciseness of the wording.
g) Position of the main verb and sentence connection
In Tacitus the position of the main verb is more flexible than in
Claudius' speech. We find not only the traditional final position, but
also the 'logical' centre position (hortantur uti ... ) and the animated
initial position: this last occurs in an interrogative clause (num
paenitet ... ) as well as in a statement clause (manent posteri eo rum;
similarly inveterascet hoc quoque). 39 The significant part of the
predicate stands emphatically at the beginning in the following
hyperbaton: capti a Callis sumus.
If the significant verb comes to the front, the result is a lively
38
This verb is often used to refer to ancient customs. Tacitus employs it with
somewhat greater variety than Cicero (see ThlL 6,1 [1913] 139f.).
39
In both cases the inverted word order is supported by chiasmus (marginal
position of the verbs).
156 TACITUS
staccato. In fact the sentences are often not connected to each other
by logical particles, but juxtaposed asyndetically as in poetry. 40 The
only sentence-connecting particles that occur rather more frequently
are adversative. 41 This too is characteristic of Tacitus' concise style
with its love of contrast and pointed expression.
h) Irony
It is typical that irony should play a role in Tacitus: scilicet Vulsci et
Aequi numquam adversam nobis aciem instruxere. We shall come back
to this when we compare Claudius.
3. Stylistic comparison
We have mentioned already that comparison of content involves a
number of uncertainties; when it comes to language and style we can
however be more definite. Claudius' language is not uniform. On one
side there is the normal language of senatorial speeches with its
conventionality in praise as well as blame and a diction that is
restrained and circumspect to the point of imprecision; in contrast to
this we have elements of the historiographical style and certain
participial constructions that appear elegant and even affected.
However these differences are not arbitrary: they are the result of
accommodation to the particular subject-matter and situation.
In Tacitus linguistic usage is more homogeneous. This is clear
from details like the ending -ere (3rd person, perf. ind.). Tacitus uses
it not only in referring to historical facts (instruxere) but, unlike
Claudius, in the actual discussion as well: omnia ... , quae nunc
vetustissima creduntw. nova fuere. Likewise the Sallustian fessus
(significantly avoided altogether by Caesar) stands in a clause of
appraisal. This participle normally refers to persons and its associ-
ation with imperium produces a poetic effect. 42
The process of Romanization is twice given eloquent expression in
the form of zeugma. A voidance of conventional phrases is every-
where discernible (Claudius on the other hand consistently uses the
technical terms of political life, 43 e.g. 1,34f. distributum consulare
40
On Tacitus and Vergil (with reference to content) cf. W. Edelmaier Tacitus und
die Gegner Roms (Diss. Heidelberg 1964) pp.134-139. On the stylistic aspect cf. Syme
pp.357f. Passages are assembled in Draeger pp.l27-129. In our text regnare in is
unusual (by analogy with dominan}.
41 At in an imaginary objection is rhetorical, cf. Draeger p.122.
42 Cf. ThLL 6,3 (1916) 611,22ff.; 612,3ff.
43
Dignitatis incrementa also occurs in Val. Max. 4, 7,5 and is attested in inscriptions
(ThLL 7,1,7 [1941] 1047,40ff. [Bulhart]). The phrase is less artificial than it first
appears. Claudius positively delights in the 'right' expression.
imperium tribunosque militum consulari imperio appellatos 44 ). The

coloniae and municipia (Claudius 2,3) are also missing here in
Tacitus; instead he expresses the point in an enigmatic paraphrase:
specie deductarum per orbem terrae legionum (there is a hint here of
deductarum co!oniarum, but no use of the technical term). 45 Tacitus
observes similar restraint in his use of elatives, insofar as conventional
formulae of commendation are involved (cf. just the positive ins ignes
viros). 46
In Claudius' style we kept noting cases of pleonasm; in Tacitus it
was cases of ellipse. Before making inferences from this as to the
personalities of the two writers, we need to take into account the
social context. Claudius is speaking in front of the Senate, which he
has to convince tactfully. A certain long-windedness is often a
requirement of diplomacy and can be part of the oratorical style. 47
Tacitus on the other hand has no need to adjust to an audience that
is often indifferent and only half attentive. He can expect a receptive
readership that is able to look forward eagerly to all the sophisti-
cations of his style and to enjoy them in peace. 48 In practice a speech
like this one of Tacitus would have sounded much the same as
Brutus' speech in Shakespeare's Julius Caesar: clever and concen-
trated, but somewhat aloof and too subtle to be effective.
Ultimately therefore the differences in literary genre here are
determined by the differing audiences. If Tacitus' stylistic approach
to the speeches in his work does differ somewhat from his treatment
44
The 'tribuni militum consulari potestate' were a permanent institution of the
early Republic (Liv. 4,6,8 and often); on the consular tribunate cf. T. Mommsen
Romisches Staatsrecht II, I (Leipzig !887 3) pp.l8!-192. Tacitus does not mention the
latus clavus here, although it is the real point (cf. 11,23,1; on this cf. Syme p.344).
45
A tendency to avoid technical terms can be observed throughout Tacitus (cf.
Syme pp.343f.; F. Kuntz Die Sprache des Tacitus [Diss. Heidelberg 1962] p.l48;
Norden I p.331; Koestermann on II ,23, 1). On adoption and avoidance ofterminology
from Claudius' speech cf. J.N. Adams 'The Vocabulary of the Speeches in Tacitus'
Historical Works' BICS 20 (1973) pp.l24-144(esp. 127f.): 'Tacitushas borrowed a few
words or expressions deliberately, but only if they fit easily into his own style' etc.; ib.
pp.l38f. (n.) a bibliography on the style of Tacitus' speeches.
46
Claudius also uses this form (2,23 ins ignes iuvenes), but with him it is not the only
one (see above).
47
Pliny (epist. I ,20) tries to convince Tacitus that in some sense a speech's quality
increases with its length. This is of course a one-sided view, but in principle it is based
on a correct psychological observation: the audience want to get the feeling that the
orator is interested in them and deserves their trust.
48
Quintilian has in mind the difference of audience when he praises Sallust's style
but at the same time describes it as unsuitable for practical oratory (10, 1,32).
158 TACITUS
of the historical narrative itself, 49 our discussion has shown that the
I
.;
gap separating this speech from one that could actually be delivered
is still far greater than any such inbuilt divergence. Tacitus' style in
the imaginary speech under consideration is marked by brachylogy,
ellipse, zeugma and the search for elevated language; the same
features are also present elsewhere.
Claudius' original speech is altogether more communicative.
Irony is a dangerous weapon in the orator's hand: since it arouses
hostility, it can very easily turn against the speaker. Tacitus can use it
without danger, as we have seen. Claudius on the other hand fills his
speech with humour 50 and thereby wins the senators' good-will. He
dwells happily on the situation of the impoverished noblewoman
who has no choice but to marry a foreigner. He praises the loyalty of
the Gauls, who did not revolt during the German War although the
census was being carried out just then, and he points out that this is
still a problem even today and in Rome. The way he touches on his
conquest of Britain is also far from straightforward, being varied by a
streak of innocent self-irony 51 and given a slightly humorous tone.
Thus in real life majesty and humour go together better than in a
work of history.
We have noted a whole series of linguistic devices whereby
Claudius endeavours to take the edge off his remarks. We see the
opposite in Tacitus: eadem die hastes, dein cives; advenae in nos
regnaverunt. The trend towards epigrammatic overstatement can be
observed in almost every sentence. The reverse tendency is exhibited
in Claudius' 'supplementary' style (his supplements depart from the
subject, whereas in Tacitus 52 they are strictly relevant).
To what extent can we move beyond generic boundaries to the
writers' stylistic individuality? We know that Claudius too was a
historian, though not a follower of Sallust but a student of Livy. 53
49 In Tacitus inveterascere occurs only in the present speech, though it is familiar to
Caesar and Cicero. Words that Tacitus uses only in speeches are collected by Syme II
pp.719f.
50 This is not the only place where Claudius shows a sense of humour, cf. V.M.
Scramuzza The Emperor Claudius (Cambridge 1940) p.lll; C.G. Bruns Fontes Juris
Romani (Tiibingen 1909 7) pp.l99f. (no. 53).
51 Self-irony is much closer to humour than to irony proper, see A. Haury L'ironie et
!'humour chez Ciceron (Leiden 1955). Cf. also the present writer 'Ovids Humor' Der
altsprachliche Unterricht 6,2(1963) pp.47-72; now in: M. v. Albrecht and E. Zinn (edd.)
OvidWege der Forschung 92 (Darmstadt 1968) pp.405-437, esp. 408.
52 On 'supplements' in Tacitus cf. Klingner (cited above p.l53 n.34).
53 The fact that Tacitus also in part follows Livy is stressed by Syme II pp. 733f.;
however he also recognizes that Sallust's influence is more important. Kuntz talks
The diametrical opposition to Livy's serene lactea ubertas 54 is an

indication that the style of the Tacitus passage is not merely due to
generic tradition, but to the independent creativity of a highly
disciplined newcomer to the Senate. If on the other hand we recall the
sober and straightforward diction of normal senatorial decrees and
imperial edicts, we can detect a personal note in the scholarly long-
windedness and the relaxed humour of the aristocratic and unpre-
tentious emperor.
Despite occasional preciosity the language of Claudius' speech is
on the whole conventional; however the train of thought is Claudius'
own. With Tacitus the reverse is in a sense the case. As regards the\,
thought, he looks for what is typical; but as to the wording he\
underlines the distinctiveness of the historiographical style by\
looking for what is unusual. Here avoidance of normality and\
repetition of unfamiliar structures in fact produce a new norm of
linguistic detachment. 55
Hence the occurrence of what is factually ineptum and stylistically
proprium in Claudius and of the factually aptum and stylistically
improprium in Tacitus is more than just a mechanical consequence of
the difference of genre. In its particular configuration it is charac-
teristic of the distinctive quality and stylistic approach of both
authors. 56
about 'historiographical language' throughout. Yet the real issue starts where he
stops: every historian has his own different style.
54
Quint. 10, 1,32, cf. also 10, 1,101: adfectus quidem praecipue eos, qui sunt dulciores,
ut parcissime dicam, nemo historicorum commendavit magis.
55
For example in the way the main concepts are transferred to subordinate parts of
the sentence.
56
Tacitus' deliberate elaboration of his language and style is shown by the major
studies of his stylistic development, of which we need mention only those by E.
Wi:ilfflin (Philologus 25 [1867] pp.92ff.) and E. Li:ifstedt (most recently in: V. Posch!
[ed.] Tacitus [Darmstadt 1986 2] pp.70-84). For literary assessments of Claudius
(besides the works already cited) the following deserve particular mention: E. Huzar
'Claudius- The Erudite Emperor' ANRWII 32, I pp.611-650 (on the Lyon tablet 627-
632; for an estimate of the emperor as orator cf. also esp. 635); M.T. Griffin 'The
Lyons Tablet and Tacitean Hindsight' CQ 32 (1982) pp.404-418.
9
The Younger Pliny (cos. 100 A.D.)
A writer's success in the hunt 1

C. P/inius Cornelio Tacita suo s.
Ridebis, et licet rideas. ego ille, quem nosti, apros tres et quidem
pulcherrimos cepi. 'ipse?' inquis. ipse; non tamen ut omnino ab inertia
mea et quiete discederem. ad retia sedebam; erat in proximo non
venabulum aut lancea, sed stilus et pugillares; meditabar aliquid
enotabamque, ut, si manus vacuas, plenas tam en ceras reportarem. non
est quod contemnas hoc studendi genus; mirum est ut animus agitatione
motuque corporis excitetur: iam undique silvae et solitudo ipsumque
illud silentium quod venationi datur magna cogitationis incitamenta
sunt. proinde cum venabere, licebit auctore me ut panarium et
lagunculam sic etiam pugillares feras: experieris non Dianam magis
montibus quam Minervam inerrare. vale.
Pliny sends greetings to his dear Cornelius Tacitus.
You will laugh- and you're quite free to. Your friend Pliny has caught
three boars, and beauties at that. 'Yourself?' you ask. Myself! Though
without giving up my quiet and easy life completely. I sat by the nets. 2
Nearby was neither hunting-spear nor lance, but pen and writing-tablets. 3 I
thought something over and wrote it down, so as to come home, if empty-
handed, at least with full tablets. There is no reason for you to scorn this kind
of intellectual activity. It is marvellous how the mind is stimulated by
vigorous movement of the body. The very woods and the solitude round
about and just the silence that hunting requires provide a powerful incentive
to reflection. 4 So when you hunt in future, you can follow my lead and take
not only bread-bag and water-bottle, but writing materials as well. You will
1
Plin. epist. 1,6. Text of M. Schuster and R. Hanslik (Leipzig 1958 3); cf. now also
R.A.B. Mynors (Oxford 1963). Commentary by A.N. Sherwin-White (Oxford 1966).
Index by X. Jacques and J. van Ooteghem (Brussels 1965; Acad. Royale de Belgique,
Cl. de Lettres, Mem. 8°, 2e ser. 58,3).
2
On hunting nets see J. Aymard Les chasses romaines (Paris 1951) pp.207-218.
3
Sherwin-White ad Joe. has in mind either ordinary wax tablets or pugillares
membranei. But would the stilus mentioned here fit the latter?
4
On cogitatio cf. 9,36,2. On the idea cf. 1,9,6; 9, 10,2; and Sherwin-White (Tac. dial.
9,6; Quint. inst. 10,3,22f.).
160
1HE YOUNGER PLINY 161
discover that Minerva's brain-waves are as common in the forest as Diana's

successes in the hunt. 5 Farewell.
Pliny's very first word places his letter in the category of the light-
hearted epistle. The elements of an epistolographical theory 6 that
recognizes such a subdivision have been found in Cicero: however
with such light material even more care than usual must be taken not
to exaggerate the importance of abstract theories. Let us therefore
start by saying simply that Pliny wants to write a light-hearted,
artistic letter. The first sentence is charmingly phrased. The main
verb in the future already anticipates the letter's effect, even before its
content is known. The indicative statement 'you will laugh' is
followed by words of approval: 'And you are allowed to laugh'-
rather like the conductor's friendly nod, when an instrument starts
up correctly without him. The next sentence is framed by subject and
predicate. The fact that in this case the subject is actually expressed is
in itself remarkable: cepi on its own would have been sufficient for a
simple statement. However we are given more than just use of
emphatic ego: there is further expansion in the addition of ille quem
5
If at 9, 10,1 a similar idea is ascribed to Tacitus, it does not of course follow that
one of the letters was written by Tacitus (as argued by L. HerrmannLato~us 14 [1955]
pp.343ff.). A compelling case for Pliny's authorship is made by E. Lefevre 'Plinius-
Studien II, Diana und Minerva, Die heiden Jagdbilette an Tacitus (1,6; 9,10)'
Gy,mnasium 85 (1978) pp.37-47.
6
Cf. Cic.fam. 2,4,1; E. FraenkelJRS 47 (1957) p.69; H. Koskenniemi 'Cicero iiber
die Briefarten' Arctos (Commentationes in honorem E. Linkomies) (Helsinki 1954)
pp.97-102. On the connection of this idea with the concept of the 'literary Jetter' cf. A.-
M. Guillemin Pline et Ia vie litteraire de son temps (Paris 1929) p.133; on the relation to
the poetic epistle pp. 134-146. In the case of a writer who works self-consciously, theory
may influence practice; see the fundamental study of H. Peter Der Brief in der
romischen Literatur Abh. der Kgl. Sachs. Ges. der Wiss., phil.-hist. Kl. 20 (1903) (on
Pliny esp. ch. 5); he connects Pliny's type ofletter rather one-sidedly with the rhetorical
digression (a poetic analogue would be Statius' Silvae); Guillemin's reference to
Martial and the 'epigram' character of many of Pliny's letters is also important. On
Pliny's own epistolographical theory cf. Sherwin-White pp.3-l I. On his attention to
style see 7,9,8 Vola epistulam diligentius scribas. nam saepe in oratione quoque non
historica modo sed prope poetica descriptionum necessitas incidit, et pressus sermo
purusque ex epistulis petitur. Guillemin is correct in her general observation on the
blend of poetic elements and plainness in Pliny's letters, even though a number of
individual parallels seem forced (Joe. cit. ch. 3). 'Pliny's simplicity is decidedly studied'
(Sherwin-White p.3). Our analysis will confirm the high level of self-consciousness in
his style. Prose rhythm too receives careful attention; we find cretic +trochee; double
cretic; double trochee; cretic; also the type' esse videatur' (retia sedebam ). On the genus
iocosum, to which the present Jetter is to be assigned, see K. Thraede Grundzuge
griechisch-romischer Brieftopik Zetemata 48 (Munich 1970) pp.27-34. On Pliny in
general ib. 74-77 (not on our Jetter).
162 THE YOUNGER PLINY
nosti. Stress is thus laid not only on the fact that Pliny was
particularly successful in the hunt, but that it was the Pliny familiar
to Tacitus - usually more accustomed to victory with words than
with weapons. The word order is expressive even in detail: the boars
are placed directly beside Pliny, and that there were three of them is
emphasized in catalogue style by putting the numeral behind its noun
in analytic sequence: 'boars: three'. Only then are we told that these
are 'beauties'. 7 The flanking position of subject and predicate reflects
the capture in a way that is no less attractive for being perhaps
unintentional. Again the reaction of the addressee is taken into
account: 'ipse?' inquis. This searching question can also be given an
affirmative answer: ipse. We thus have three words and at the same
time three sentences. No better example could be found of the
stylistic principle of brevity as befitting the epistolary mode generally
and in particular the recipient Tacitus. Moreover the rapid change of
person illustrates the classical concept of letter-writing as a con-
versation with those absent. 8 The final and rather striking answer
requires qualification: non tamen ut omnino ab inertia mea et quiete
discederem. So Pliny has not been untrue to his normal character. On
one hand this puts Tacitus' mind at rest; on the other this explanation
makes the puzzle even greater. Before we continue, a brief comment
is required on the way Pliny characterizes himself by means of inertia
et quiete. These words appear to give a negative impression. Pliny
may however be referring to his intellectual inclinations with the
same understatement that the business-man in Petronius uses in
talking of his 'few rags' and his 'shacks'. The language is not without
self-irony; it is considerably less provocative than when an elegist like
Ovid (to take just one) refers to himself as il/e ego nequitiae Naso
poeta meae. 9 Since the days of Sallust and Cicero the writer's self-
confidence regarding the value of his way of life has increased
substantially; apologies or provocativeness can be replaced by gentle
humour.
The next section supplies the solution to the puzzle: ad retia
sedebam. So in this regard the savant has not changed his habits of
life after all. Nor do the attributes that he mentions fit the present
situation, but his real character: erat in proximo non venabulum aut
lancea, sed stilus et pugillares. The antithesis between hunting-gear
7
Pulcher is a technical term for animals that are without blemish, e.g. Naev. fr. 3,3
Morel. Posch (cited below n.l3) gives a different interpretation.
8
Cf. e.g. Cic.fam. 1,7,1; 15,14,3; Att. 6,1,24; 7,15,1; 12,39,2; 13,18.
9
Am. 2,1,2.
THE YOUNGER PLINY 163
and writing-materials is significant. The change of perspective

should be noted. Initially the surprise was due to having the writer
presented as a hunter; but now conversely the 'hunter' turns out to be
a writer at work: meditabar a liquid enotabamque, ut si manus vacuas,
plenas tam en ceras reportarem. Thus for a moment writing once again
appears in the light of ancient Roman tradition as 8E(m;po~ rcA.oli~,
an alternative in case external success is denied. Yet in the course of
the letter what seems to be a secondary activity becomes increasingly
the main subject. This is already stressed by Pliny in the following
sentence: non est, quod contemnas hoc studendi genus. The initial
position of the negative lends weight to the idea, and the inclusion of
the addressee creates a personal note. Hunting is not an end in itself,
but only a frame for a studendi genus. These are the words of a man of
letters who sees everything in the light of his intellectual interests: this
is virtually the opposite of the usual picture of a Roman, based on a
narrow misconception of what Roman character is capable of. 10
mirum est, ut animus agitatione motuque corporis excitetur. However
an important feature of Roman thought cannot conceal itself even
here. The peace and solitude of the forest are not an occasion to
indulge in romantic fancies; it is true that they are seen in relation to
intellectual activity, but the standpoint is utilitarian.
The recipient is now drawn into these reflections in a tone of
friendship. With proinde, which is also used elsewhere to point the
moral, he is advised to imitate Pliny. Towards the end an impressive
vividness is achieved by the specific mention of the equipment (the
traditional packed meal, but also the writing-tablet that henceforth
has to be taken along as well). The climax of the letter is the
embodiment of its two underlying themes (hunting and intellectual
activity) in mythological figures.
In the sustained juxtaposition of the two contrasting elements that
make up the letter, Pliny achieves some very varied effects. One
important device is the word-order itself: ego il!e quem nosti stands
close beside apros. The reverse sequence occurs on a larger scale: on
one hand apros tres ... cepi, on the other non tam en ut omnino ab
inertia mea et quiete discederem. The contrast of equipment is pointed
up sharply: non venabulum aut lancea, sed stilus et pugillares; the same
happens with the physical and intellectual gain: si manus vacuas,
10
Pliny is incidentally less rigorous than Sallust, who rejects hunting as servile
officium ( Catil. 4, 1). However in his own day Pliny no longer needs to lay special stress
on the superiority of intellectual pursuits -rather on the 'utility' of hunting even for
the man of letters.
plenas tamen ceras. Here variety is produced by the chiasmus, but

also by the cross between conditional and concessive modes. For all
the conceptual similarity, the stylistic devices are continually varied
down to the end of the letter: the banal non solum, sed etiam is
skilfully avoided twice: ut panarium et lagunculam sic etiam pugillares;
and again in a different way: non Dianam magis ... quam Minervam.
However in the third quarter of the text the two aspects are not set
one against the other in the form of a climax, as in the passages
considered so far; instead they are shown in their interconnection
with each other: mirum est ut animus agitatione motuque corporis
excitetur. It should be noted how the word agitatione, that belongs to
corporis, is put next to animus, and conversely how excitetur, which
refers to animus, stands directly beside corporis: we have here a
linguistic reflection of the interaction of mind and body. In the next
sentence there is a linear symmetry in syntax, content and sound.
This is marked by an antithetical correspondence between venationi
and cogitationis in the centre and silentium and incitamenta as an
outer frame.
The tendency to work with contrasts is one special case of a more
general preference for two-element structures, even when the
concepts are similar. We find this stylistic principle applied in almost
every sentence: inertia ... et quiete; venabulum aut lancea; stilus et
pugillares; meditabar ... enotabamque; agitatione motuque; silvae et
solitudo; panarium et lagunculam; cf. also tres et quidem pulcherrimos.
Two main features of our letter are two-element structures where
the elements either form a contrast or are similar: these two types
have different functions. The contrasting type generally serves to
emphasize in climactic form one of the two elements: the dramatic
development comes from the alternate predominance of either of
them or their temporary equality. The exclusive occurrence of one of
the two elements produces in the reader the unconscious expectation
of the other; 'anticipatory incompleteness' is thus the prerequisite for
tension and climax. A different purpose is served where the two
elements have similar meaning. Here a uniform concept unfolds
under two complementary aspects: writing materials (stilus et
pugillares), intellectual activity (meditabar ... enotabamque), pro-
visions (panarium et lagunculam). Pliny has an eye for small,
expressive details. 11 With nouns the process leads to greater vividness
and concreteness, and with verbs to a more exact, step-by-step
11
Cf. also Guillemin p.l53.
THE YOUNGER PLINY 165
description of an event, in which chronological distinctions are

possible. Both have to do with the rhetorical requirement of
f:vapyEta; in relation to events, the technique has been encountered
in the speech of Cicero analysed above. 12
The contrasting use of two elements has led us to see Pliny's letter
as in some sense a piece of music with two themes that in different
ways are both played off against each other and combined with each
other: in the third quarter a climax 13 is produced by interpenetration
and overlapping.
The use of two similar elements indicates how Pliny gives concrete
shape to his ideas and provides his reader with vivid impressions.
Here we see a rhetorically trained imagination at work, which also
shows in the fact that the letter, for all its shortness, appears to
consist of prologue (ridebis ... ), narratio (ad retia sedebam ... ),
argumentatio (non est quod contemnas ... ) and peroration (proinde
cum venabere ... ). Though really self-evident, this observation dispels
any remaining doubts as to the literary character of the epistle: more
important however is the fact that the addressee is actively involved
at beginning, middle and end, and that each section of the letter is
rounded off with an epigrammatic point. 14 One particularly attractive
feature of the letter is that the same words and ideas keep recurring in
a new light: ridebis is an objective statement, et licet rideas gives a
subjective assent to this. Ipse? is the recipient's supposed question,
and ipse is again the writer's reply. Pliny's account of his intellectual
activity in the forest is reflected in Tacitus' mind without verbal
repetition but in conceptual summary as hoc studendi genus.
12 See above pp.41-44. Hence we should not stress Plinian 'breadth' alone, without
also taking account of the £vupy£ta this achieves.

13 This is the only place where we have one instance of a three-element structure; it
too is a climactic device: iam silvae et so!itudo ipsumque ... silentium ... On the tricolon
(Cicero - Vergil - the orators - Pliny) cf. Guillemin pp.l50f. Posch identifies (with
extreme boldness) the three elements with the three 'boars': S. Posch 'Eine Eberjagd
mit Gansefiisschen (Zu Plin. epist. 1,6)' in: P. Handel and W. Meid(edd.)Festschrift
fur R. Muth (Innsbruck 1983) pp.375-383.
14 I: ipse; non tamen ut omnino ab inertia mea et quiete discederem.
II: ut, si manus vacuas, plenas tamen ceras reportarem.

III: ... illud silentium, quod venationi datur, magna cogitationis incitamenta sunt.
IV: experieris non Dianam magis montibus quam Minervam inerrare.
Guillemin loc. cit. p.l50 speaks of 'veritables epigrammes en prose'. Cf. ib. for
remarks on the difference in the structure of sentences and periods between Pliny and
Cicero; here. however there should be a stricter comparison between each author's
letters, and between the Panegyricus and the speeches. Guillemin is right to stress the
clear arrangement in Pliny (e.g. p.l49).
The pursuit of variation is also highly significant. The verb venari

is avoided over a long period (cepi; ad retia sedebam; venabulum;
venationi). It is not employed until the end, and then not with
reference to Pliny but to Tacitus. The constant interchange has
developed into a complete reversal of roles. The key event happened
to Pliny but is stated in connection with Tacitus. The original
experience is transposed to the mind of the addressee in heightened
mythological form. Is this then a conversation between people far
apart? There is surely more to it than that. It is a 'communication>~ 5
in the fullest sense of the word, skilful and at the same time
unassuming and humorous.
15
The question 'imaginary or real letter' is out of place in a civilized society (R.
Hanslik Anzeiger fur die Altertumswissenschaft [hrsg. v. d. Humanistischen Gesell-
schaft] 17 [1964] pp.6f. with bibliogr.).
10
Apuleius (b. c. 125 A.D.)
An abortive bid for salvation 1

Sic adfectus atque in solitudinem relegatus angulo stabuli concesseram,
dumque de insolentia collegarum meorum mecum cogito atque in
alterum diem auxilio rosario Lucius denuo futurus equi perfidi
vindictam meditor, respicio pilae mediae, quae stabuli trabes sustinebat,
in ipso jere meditullio Eponae deae simulacrum residens aediculae,
quod accurate corollis roseis equidem recentibus fuerat ornatum.
denique adgnito salutari praesidia pronus spei, quantum extensis
prioribus pedibus adniti poteram, insurgo valide et cervice prolixa
nimiumque porrectis labiis, quanta maxime nisu poteram, corollas
adpetebam. quod me pessima scilicet sorte conantem servulus meus, cui
semper equi cura mandata fuerat, repente conspiciens indignatus
exsurgit et: 'quo usque tandem', inquit, 'cantherium patiemur istum
paulo ante cibariis iumentorum, nunc etiam simulacris deorum in-
festum? quin iam ego istum sacrilegum debilem claudumque reddam'; et
statim telum aliquod quaeritans temere fascem lignorum positum
offendit rimatusque frondosum fustem cunctis vastiorem non prius
miserum me tundere desiit, quam sonitu vehementi et largo strepitu
percussis ianuis, trepido etiam rumore viciniae conclamatis latronibus
profugit territus.
Being thus treated and sent into lonely banishment, I had withdrawn into the
corner of the stable, and while I reflect on the impudence of my colleagues
and think how on the following day with the aid of roses I should be back to
being Lucius and should avenge-myself on the faithless horse, I notice on the
middle pillar, that carried the rafters of the stable, almost exactly in its
centre, an image of the goddess Epona enthroned in a shrine and lovingly
adorned with little wreaths of still perfectly fresh roses. At last with the
1 Apul. met. 3,27. Text ofR. Helm (Leipzig 1955 4 ). Index by W.A. Oldfather, H.V.
Canter, B.E. Perry (Middletown, Conn. 1934). Commentaries by R.T. van der Paardt
L. Apuleius Madaurensis, The Metamorphoses, A commentary on book III with text and
introduction (Amsterdam 1971); J.C. Fredouille Apulee, Metamorphoseon fiber XI
Coil. Eras me (Paris 1975); J. Gwyn Griffiths Apuleius of Madauros, The Isis-Book, ed.
with an introduction, translation and commentary (Leiden 1975).
168 APULEIUS
means of deliverance before me and full of eager hope, I reared up

powerfully, as high as I could reach with outstretched front feet, extended my
neck, turned my lips out a very long way, and tried to reach the wreaths with
all my strength. In this attempt, which was sadly doomed to failure, I was
suddenly spotted by my servant-boy, who had always had to look after my
horse; he jumped up in anger and said: 'How much longer must we put up
with this beast, that earlier on attacked the draught-animals' food, and now
even attacks the images of the gods? Why don't I beat this impious creature
rightaway, till it can't stand up?' No sooner said than done. As he looked for
a weapon his eye fell on a bundle of wood that happened to be lying there; he
picked out the very thickest cane with branches, 2 and did not leave off
thrashing my poor self, till the door was broken down with a loud din and
violent crash, and the neighbours shouted anxiously 'Robbers, robbers!',
and he took fright and fled.
1. Narrative structure
Lucius has been transformed into a donkey, and he has to endure
being driven from the manger by a fellow-donkey and his own
faithful horse. At the start of our section he is sulking in the corner.
The pluperfect (concesseram) marks this statement as the back-
ground for what follows. The thoughts on the wrong suffered and the
plan for revenge are similarly not regarded as the main action, but are
treated as an element of suspense and put in a temporal clause
coming before the main clause, which then produces the big surprise:
respicio. The donkey notices the rose-decked image of the equine
goddess Epona. Within the main clause too the sequence of
impressions produces an increase of tension. First of all we see before
us the pillar, whose position and function are explained, then the
image of the goddess, and not until the end the wreaths of flowers
that can bring salvation. There has just been a brief reference to the
liberating effect of roses in Lucius' deliberations: preparation and
suspense are thus equally balanced. Apuleius then presupposes
knowleqge of the chance of rescue in a preliminary participial phrase,
stresses hope as the driving force in an apposition, and then in the
main clause (insurgo) uses very vivid ablatival amplifications to
present the grotesque contortions of the donkey in its yearning for
the flowers. The sentence ends in the imperfect: thus the attempt has
not yet succeeded. The suspense is prolonged by a contemporaneous
participle: quod me ... conantem. A glimpse of the unfortunate sequel
is inserted in ablatival form, as with the introduction of the rose
motive referred to earlier. Only then is the new subject mentioned
2 Cf. Pseudo-Lucian Lucius 29.

APULEIUS 169
(viz. the slave); it is presented in a subordinate clause as in the

description of the pillar. After the suspense of these preliminary
observations the events come in rapid succession: repente conspiciens
indignatus exsurgit et 'quousque tandem' inquit. The dramatic effect is
heightened by placing the inquit after the first two words of the
speech it introduces: this isolates the lofty quotation from Cicero.
After the speech, which we shall consider separately, the final
sentence covers a tremendous amount of action. The rhythm
'participle/finite verb' is repeated three times in it, first in two main
clauses, and then in a subordinate clause. Finding a stick is spread
over one and a half sentences with great psychological subtlety
(looking for a weapon - discovery of the pile of wood- rummaging
out the strongest stick). The middle of the sentence is occupied
by the thrashing. The end of the affair is artfully transferred to a
subordinate clause (non prius .. . quam ... profugit): this is a device
often encountered in mature narrative style, and particularly in
combination with donee. 3 However Apuleius adds one further
refinement: the surprising element that moves the action forward
(viz. the break-in by the robbers) is reduced to an ablatival
construction within the subordinate clause. In this way the reader is
kept in breathless suspense.
The rhythm of Apuleian narrative is determined by the way in
which the tension gradually built up in subordinate clauses and in the
forward positioning of participles is succeeded by rapid resolution in
the main clause. The striking transfer of the final incident to a
subordinate clause forms a deliberately sought climax. Particular
attention should be paid to the method of giving advance notice of
themes and events which subsequently acquire importance: this
involves skilful integration into the narrative by means of ablatival
constructions. Also noteworthy is the technique whereby the tension
achieved at the end of one sentence is prolonged at the start of the
next by a present participle.
The picture of Apuleian narrative style that has emerged so far is
barely more than an outline. In what follows we propose to give it
depth and life through an analysis of standpoint and methods.
2. A donkey's standpoint
Throughout the entire text the viewpoint of Lucius the donkey is
3
On donee cf. J.-P. Chausserie-Lapree 'Les structures et les techniques de
!'expression narrative chez les historiens latins' REL 41 (1963) pp.281-296, esp. 295f.
170 APULEIUS
maintained. From the stable floor the reader shares his view of the
post and the image of Epona on it. There is a systematic transfer of
social, moral and legal concepts from his human past to his new
surroundings. The other animals are called collegae; they have
'relegated' the hero from the manger, and the hitherto faithful horse,
just described as vector meus probissimus, 4 has turned out to be an
equus perfidus. Lucius plans revenge; vindicta in this sense is a poetic
and Tacitean word. Its application to a horse is astonishing. The
language becomes even grander at the sight of the roses that will
bring deliverance: we shall deal with this later. A contrast is provided
by the realistic account that follows immediately of the futile
contortions of the clumsy donkey. The ablative absolutes give the
description an element of technical precision which may recall the
famous account in Cornelius Nepos 5 of Chabrias' battle-stance as
immortalized in art or the descriptions of strategic manoeuvres and
devices in Caesar. In terms of content the repetition quantum ... adniti
poteram '-quanta maxime nisu poteram underlines the effort involved,
and in formal terms it may also indicate the awkwardness of the
donkey. We thus have a tremendous contrast between the poetic
account of a state of mind and the exact description of an animal
standing on its hind legs. Here too the standpoint of first-person
narrative is retained, and the 'self-presentation' is carried through
with grotesque consistency.
The paradox of a continuity of consciousness despite altered shape
is worked out in the encounter between master and servant: servulus
meus refers to the earlier self; quod me .. . conant em in the same
sentence to the new self. This underlines the comic element of the
switch in the exercise of power. The slave apostrophizes his master as
a jaded nag and gives him a thrashing: this is a burlesque variant of a
situation that is given tragic form by Ovid in the story of Actaeon. 6
What is perhaps most striking is the retention of the chosen
4
Met. 3,26. On the difficulties of a donkey's standpoint in regard to narrative
technique cf. R.T. van der Paardt 'Various aspects ofnarrativetechniquein Apuleius'
Metamorphoses' in: B.L. Hijmans Jr. and R.T. van der Paardt (edd.) Aspects of
Apuleius' Golden Ass (Groningen 1978) pp.75-94, esp. 76-79.
5 Nep. Chabr. I ,2f.: fugatis iam ab eo conducticiis catervis reliquam phalangem loco
vetuit cedere obnixoque genu scuta, proiecta hasta impetum excipere hostium docuit ...
hoc usque eo tota Graeciafama celebratum est, ut ilia statu Chabrias sibi statuamfieri
voluerit. On this description cf. G .E. Lessing Laokoon ch. 27-28.
6
Heujamulosfugit ipse suos (Ov. met. 3,229). Apuleius uses the myth of Actaeon at
2,4; there is an obvious analogy to the fate of Lucius. H. Riefstahl Der Roman des
Apuleius Frankfurter Studien 15 (Frankfurt/Main 1938) pp.67-69 does not however
deal with the slave's role which we examined above.
APULEIUS 171
standpoint at the end of our section: in the donkey's view the robbers'
raid is not a major event to begin with, but only of importance insofar
as it makes the slave take to his heels, and so frees the animal from its
tormentor.
3. Clarity of expression
The adjective rosarius, which as a rule is used in a concrete sense/
retains its 'technical' character in this passage too: however the
context removes it from the sphere of concrete reality. This
combination of technical language and mysterious content produces
the special atmosphere of magic. Auxilium too, which originally was
a military term, 8 adds a down-to-earth distinctness to the style. The
language is as precise as the idea is vague. The objective term 'roses' is
avoided. The abstract phrasing signifies something operating through
roses or something rose-like, in which the aid is manifested. By
transcending objectivity the expression directs attention beyond
itself to the mystical salvation in book eleven.
The combination adgnito sa/utari praesidia points in the same
direction. The preciosity that marks the reference to Lucius' feelings
at the sight of the roses has a distancing effect. While the individual
words all occur in Cicero, the collocations with the resultant religious
colouring are nonetheless peculiar to Apuleius. Praesidium is in
origin a military technical term like auxilium. In the present work
Apuleius generally uses it in a wholly concrete sense. 9 It occurs once
in a prayer-like context: a girl fleeing from robbers prays to the gods
for help and addresses Lucius the donkey as 'praesidium meae
Iibert a tis meaeque sa/utis.' At 2,18,18 the dagger is grandly described
as sa/utis meae praesidia. A technical word like this, which is used
almost in the sense of 'means', acquires a magical and religious tone
in the passage we began with, thanks to the combination with the
adjective sa/utare. Specific evidence can be adduced to support this.
In book elevert Isis, the Queen of Heaven, says: iam tibi providentia
mea illucescit dies salutaris ( 11 ,5,20). 10
The meaning of the cryptic praesidium sa/utare will be made clear
by the scene in the last book where the priest gives the donkey a
garland of roses and in this way brings about his deliverance. We
should note the bold use of sa/us here: fat a sa/utemque ips am me am
7 Cf. Forcellini s.v.

8
See Walde-Hofmann s.v.
9
See the Index Apuleianus s.v.
10 Cf. II ,22,5 deae potentis benignitas salutaris; II ,25,6 salutarem porrigas dexteram.
172 APULEIUS
gerens sacerdos adpropinquat ... mihi coronam. 11 Here the gramma-

tical construction suggests that the priest brings both sa/us and
corona. This parallelism, which is close to an identification, goes
beyond the language of the text we have been discussing and makes
clear what lies behind its abstract vagueness.
Thirdly, we need to bring out the perspicuity of the phrase pronus
spei. From a stylistic point of view the adjective pronus, which
significantly does not occur in Cicero's speeches and letters, is far
from being an everyday word. 12 On a semantic level it indicates the
way animals lean forward in contrast to man's upright bearing; 13
hence it often denotes an animal's unbridled nature. 14 Syntactically
speaking, the collocation pronus spei is a mannered type of con-
struction with roots in Early Latin that underwent Greek influence
and gained currency chiefly in poetry. 15 Thus Lucius is 'unbridled in
his hope'. 16 The description of the donkey's up reared body with its
outstretched neck and protruding lips is simply a grotesque portrait
oflust. An instructive contrast is provided by the donkey's deliberate,
'human' restraint at the time of the real deliverance in the last book:
Nee tamen gaudio subitario commotus inclementi me cursu proripui,
verens scilicet, ne repentino quadripedis impetu religionis quietus
turbaretur ordo, sed placida ac prorsus humano gradu cunctabundus
paulatim obliquato corpore, sane divinitus decedente populo, sensim
inrepo (11, 12). While the text we have been discussing emphasizes
unbridled behaviour and animal shape, the 'human' character of the
gait in the actual scene of deliverance prepares the way for the
retransformationY Only when the garland is given to him, is he
allowed to swallow it. 18
11
Met. 11,12.
12 Cicero uses it in four places in the philosophical works; two of them are in the
'historical' second book of rep. (2,41,68; 2,26,47). The word is familiar to Sallust and
Tacitus. In poets too it is not rare.
13 Ov. met. 1,84f. pronaque cum spectent animalia cetera terram, os homini sublime
dedit.
14 Thus in the present novel the eyes of the uncomplicated Fotis, who is really to
blame for Lucius' transformation into a donkey, are described as prona libidine
marcidos (3, 14).
15 The evidence is collected in Kuhner-Stegmann pp.435-451, esp. 445. On its
interpretation cf. Hofmann-Szantyr pp. 77-81, esp. 79c and 75~.

16 Exactly similar is 8,11 inrepit cubiculum pronus spei.
17 Technically speaking, we are dealing here with a mode of presentation often
employed by Ovid; cf. the present writer's 'Die Verwandlung bei E.T.A. Hoffmann
und bei Ovid' Antike und Abendland 10 (1961) pp.161-180, esp. 162ff.
18 Avido ore susceptam cupidus promissi devoravi (11,13). Lucius has already
restrained himself once before, not indeed out of pure self-knowledge, but because he
APULEIUS 173
In this same scene the transformation of spes too is indicated by the

choice of language. Pronus spei is replaced by cupidus promissi. The
two incidents are separated by the revelation which has led up to the
change and which through a promise has turned mere hope into firm
confidence. Significantly the divine epiphany was preceded by
Lucius' feeling that his sufferings would soon be over: jato scilicet
iam meis tot tantisque cladibus satiato et spem salutis, licet tardam,
subministrante (11,1). 19
Even a detail of this scene like the image of the goddess Epona has
wider significance and points to the Isis of the final book. Whereas in
the earlier passage the donkey tries to deprive the goddess forcibly of
the roses dedicated to her, at the start of book eleven he prays
sincerely, whereupon Isis graciously appears to him in a dream: this
is an important counterpart to the 'sacrilege' for which Lucius is so
severely punished by his own slave.
These two scenes of abortive and mystical deliverance are absent in
Pseudo-Lucian; hence the interrelation between them, which extends
to points of language, is particularly characteristic of the way
Apuleius writes. 20 At the same time both the content and language of
this scene are clear arguments in favour of the intimate connection of
the ending with the rest of the work. 21
4. Highlights
Vocabulary and style are vivid and colourful: this is the case not just
felt safer from the robbers in his donkey-shape (3,29, 16). The instinct of self-
preservation plays the same role when he forgoes the unsuitable and poisonous laurel-
roses (4,2). The roses play a double role in Apuleius: the mystical role is preceded by
the erotic (2, 16,2 in the bedroom; cf. Hor. carm. I ,5, 1; in the story of Amor and Psyche
6,11,3; cf. roseus of bodily charms 2,17,5; 2,8,13; of Venus' feet in the story of Amor
and Psyche 4,31,12; Fotis is an earthly Venus: 3,22,13). A symbolic anticipation can be
seen in the opening words of the third book that describes the transformation
(Aurora's rosy arm); cf. R. Heinze Virgils epische Technik (Leipzig 1914 3) pp.366-370.
We have a mystical reflection in Psyche's rosy blood (5,23,6; Lucius the donkey is of
course an Amor manque, cf. 3,22, 13) and in the roses of the heavenly wedding (6,24, 7).
Parallels between books III and XI are drawn by J. TatumApuleius and the Golden Ass
(Ithaca and London 1979) p.43. Cf. in general A. Wlosok 'Zur Einheit der
Metamorphosen des Apuleius' Philologus 113 (1969) pp.68-84.
19 A further step is the expectation of being initiated (11,21).
20 A. Lesky ('Apuleius von Madaura und Lukios von Patrai' Hermes 76 [1941]
pp.43-74) gives a cautious treatment ofthe question of sources in line with the general
assumption that Apuleius and Pseudo-Lucian depend on a common source. As far as
the present writer can see, he does not deal with our passage.
21 Undue stress is placed on the separateness of the ending by R. Helm in the
introduction to Apuleius Metamorphosen (lateinisch und deutsch) (Berlin and Darm-

stadt 1956) p.6.
174 APULEIUS
in elevated contexts, but also, and perhaps even more so, in the
thrashing scene. Quaeritans, which occurs in Early and Late Latin, is
more expressive than quaerens, offendit more graphic than vidit,
rimatusque more descriptive than quaerens, vastiorem infinitely more
menacing than maiorem, and finally tundere is a very strong term
owing to its connotation of intensity and professional thoroughness.
Rhetorical brilliance marks the clear and generally bipartite
sentence structure we have already described: this is further accen-
tuated by rhythmical devices, as the clausulae show. 22 The play on
reminiscences of Cicero in the slave's speech is meant to tease the
reader: we shall now look at this speech.
Quousque tandem is followed, after a dramatic pause, not by the
name Catiline, but by a term of abuse starting with the same letters:
cantherius. The connection between the previous and the present
scene is supplied by a rhymed isocolon characteristic of Apuleius
(cibariis iumentorum ... simulacris deorum)Y The adjective infestus
maintains the stylistic level of the Catilinarians. 24 The reproach of
sacrilege recalls the equally famous Verrines; cf. II 1, 18,4 7 illine tu
temp/a ... tam sane to, tam religioso manus impias ac sacrilegas afferre
conatus es? 25 There is a certain similarity to Cicero's Verrines in the
way Apuleius follows the indictment de re frumentaria 26 with the
charge de signis (on images of the gods). This contrast puts a playful
emphasis on the significance of the present scene, which is new and
with the theme of deliverance marks an improvement on the 'fodder
motif of the Greek original. 27 The end consists of a brutal threat. By
a method often encountered in Tacitus, Apuleius gives distinction to
an everyday collocation of words by changing their order: while
claudus normally goes in front of debilis, 28 Apuleius puts the two-
22
On clausula rhythm in Apuleius seeM. BernhardDer Stil des ApuleiusTiib. Beitr.
2 (Stuttgart 1927) pp.249-255.
23 According to F. Polheim Lateinische Reimprosa (Berlin 1925) p.206, Apuleius is
'more partial to rhyme than any other writer of antiquity'. On isocolon in Apuleius see
Bernhard pp. 87ff., 289ff., 307ff., 345ff. We should note the heightening of tone due to
the generalizing plural; similarly we can say 'There are ladies present', even where only
one lady is concerned.
24 Cf. Cic. Catil. 1,5,11: quod hanc tam taetram, tam horribilem tamque infestam rei
publicae pestem totiens iam effugimus.

25
Cf. also Verr. II 5,72,188: quorum templis et religionibus isle ... bellum sacrilegum
semper impiumque habuit indicium.
26
Bernhard (pp.311-314) notes the imitation of Cicero only in the Florida, but not
in the Metamorphoses. It is significant that the speeches imitated in the Florida include
the Verrines, especially de re frumentaria.
27
Cf. Pseudo-Lucian Lucius 15.
28
Cf. Hofmann ThLL 5,1,1 (1910) 108,50f. and 62.
APULEIUS 175
syllable word last; the harshness is softened by the addition of -que.

5. Detachment and a higher level of communication
Compared with Petronius, the language of Apuleius is more
uniform; 29 the narrative itself is not as plain, and the dialogue is never
as markedly vulgar as in some parts of Trimalchio's dinner party.
Apuleius writes a literary language that is heavily permeated with
poetic elements. This happened occasionally in the Satyricon (for
example in some transitional passages between prose and verse
sections); however the range of styles in Petronius is wider, though
here too we find a uniform artistic purpose. Prose and poetry
permeate each other inextricably in Apuleius. The style of Petroni us
charms us with its clarity and precision, while qualifying every
expression of feeling with irony: this is true also of the urbane
sections and even of the poetic parts. By contrast the 'Golden Ass' is
pervaded by a gentle humour which uses wording either playful or
pretentious to suffuse the narrative with a fairy-land atmosphere.
For all its freedom, the novel still has a distinctive character and tone;
one might almost speak of a mask of naivety half hiding and half
disclosing a philosophic and religious world-view. Apuleius is not so
much concerned with employing a critical eye to expose falsehood
and superficiality; rather he wants to present human existence in all
its fullness from the standpoint of the miraculous and unexpected, so
as to reveal a higher plane of meaning within it. This is also the reason
for the artistic and elevated language, which in its playfulness and
preciosity imparts a mystical shimmer to all that happens. In
considering this wider framework we do not of course mean to
overlook Apuleius' tolerant and broad humanity. It would impair
the artistic effect if the metaphysical foundations of this humane
outlook were to be more conspicuous in the novel than is absolutely
necessary. 30
Notwithstanding its personal and humorous tone, the artificiality
of Apuleius' language leaves an impression of remoteness and
grandeur. For neither cosmopolitan power of observation nor an
atmosphere of philosophy and religion are able on their own to
restore spontaneity and force to the language. The result of trying to
29
Bernhard pp.255-258 rightly stressed the uniformity oftheMetamorphoses' style.
The sense of moral outrage, with which Norden Kunstprosa pp.600-605 derides
Apuleius' use of language, can only strike us as comical today. Irrespective of that, he
of course rightly understood Apuleius' connection with the second Sophistic.
3
° Cf. 11,23 Ecce tibi rettuli quae, quamvis audita, ignores tam en necesse est.
176 APULEIUS
direct attention beyond external events by making them appear

outwardly either alluring or odd is in fact a higher level of
communication. The Christian church had to proceed from a similar
linguistic position in the Latin-speaking world. Though Latin
became less and less a mother tongue spoken by the unsophisticated,
it still survived as a symbolic and impersonal means of expression:
this is due largely to the changed attitude to language that marked the
late epoch to which Apuleius already belongs.
List of Abbreviations
For Latin authors the abbreviations are those of Thesaurus Linguae

Latinae.
AJPh =American Journal of Philology (Baltimore 1880 ff.)

v. Albrecht, Die Parenthese ... = M. v. Albrecht Die Parenthese in Ovids
Metamorphosen und ihre dichterische Funktion Spudasmata, vol.
7 (Hildesheim 1964)
ANRW = Aufstieg und Niedergang der romischen Welt. Geschichte und Kultur
Roms im Spiegel der neueren Forschung, edd. H. Temporini and
W. Haase (Berlin/New York 1972 ff.)
BICS =Bulletin of the Institute of Classical Studies (London 1955 ff.)
CE = Carmina Latina Epigraphica, ed. F. Buecheler, 2 vols. (Leipzig 1895-
1897), supplemented. E. Lommatzsch (ib. 1926)
CIL =Corpus Inscriptionum Latinarum (Leipzig/Berlin 1862-1943; 1893 ff.2)
cos. = consul
CPh = Classical Philology (Chicago 1906 ff.)
CQ = Classical Quarterly (Oxford 1907 ff.)
Emout-Meillet = A. Ernout-A. Meillet Dictionnaire hymologique de Ia
langue Latine (Paris 195P)
Forcellini =Lexicon totius Latinitatis ab A. Forcellini ... lucubratum, latest ed.
(various contributors), 6 vols. (Patavii 1940)
GL = Grammatici Latini, ed. H. Keil, 7vols. and 1 supplement (Leipzig 1855-
1880)
Goethe, W.A. = Goethes Werke, hrsg. im Auftr. d. Grossherzogin Sophie von
Sachsen, Abth. I-IV, 133 vols. (in 143) (Weimar 1887-1919)
Herder, Suphan = J. G. Herder Siimtliche Werke, hrsg. von B. Suphan, 33
vols. (Berlin 1877-1913)
Hofmann, Umgangssprache = J. B. Hofmann Lateinische Umgangssprache
(Heidelberg 1936 2 ; reprint 1951)
Hofmann-Szantyr = J.B. Hofmann Lateinische Syntax und Stilistik, revised
by A. Szantyr (Munich 1965) (Hdb. d. A W II 2,2)
HRR = Historicorum Romanorum Reliquiae, ed. H. Peter, 2 vols. (Leipzig
1883-1906; vol. 12 1914)
HSCP =Harvard Studies in Classical Philology (Cambridge, Mass. 1890 ff.)
IRS= Journal of Roman Studies (London 1911 ff.)
180 ABBREVIATIONS
Kuhner-Gerth = R. Kuhner Ausfiihrliche Grammatik der griech. Sprache.

Teil 2, Satzlehre, revised by B. Gerth, 2 vols. (Hannover/
Leipzig 1898-1904)
Kuhner-Stegmann = R. Kuhner Ausfuhrliche Grammatik der lat. Sprache,
Satzlehre, 2 vols., 2nd ed. revised by C. Stegmann, 4th ed.
revised by A. Thierfelder (Darmstadt 1962)
Leeman= A.D. Leeman Orationis ratio, 2 vols. (Amsterdam 1963)
Leo =F. Leo Geschichte der romischen Literatur, vol. I (Berlin 1913; reprint
1958)
LOfstedt, Peregrinatio = E. Lofstedt Philologischer Kommentar zur Pere-
grinatio Aetheriae (Uppsala 1911; reprint 1936)
Lofstedt, Synt. = E. Lofstedt Syntactica, 2 vols. (Lund 1928-1933; vol. F
1942)
Malcovati see ORF
Meyer-Lubke, REW= W. Meyer-LubkeRomanisches Etymologisches Worter-
buch (Heidelberg 1930-19323 )
NGG = Nachrichten von der Ges. d. Wiss. zu Gottingen, phil.-hist. Kl.
(Gottingen 1894 ff.)
Nlbb =Neue Jahrbacher fiir Philologie und Piidagogik (Leipzig 1898 ff. ); Neue
Jahrbiicher f d. klass. Altertum, Gesch. u. dt. Literatur u. f
Piidagogik (ib. 1925 ff.)
Norden= E. Norden Die antike Kunstprosa vom 6. Jh. v. Chr. bis in die Zeit
der Renaissance, 2 vols. (Leipzig 1898; Leipzig/Berlin 1923 4 ;
reprint 1961)
OLD= Oxford Latin Dictionary (Oxford/London 1968 ff.)
ORF = Oratorum Romanorum Fragmenta, ed. H. Malcovati, vol. I (Turin
195SZ; 19764); Index Verborum (Turin 1979)
PhW = Philologische Wochenschrift (Leipzig 1881 ff.)
RE = Paulys Realencyclopiidie d. class. Altertumswiss., revised by G. Wissowa
(Stuttgart 1893 ff.)
REA= Revue des Etudes Anciennes (Bordeaux 1889 ff.)
REL =Revue des Etudes Latines (Paris 1923 ff.)
RhM = Rheinisches Museum fiir Philologie, Gesch. und griech. Philos.
(Frankfurt/Main n. s. 1842 ff.)
RPh =Revue de Philologie, de Litt. et d'Histoire anciennes (Paris n. s. 1877 ff.)
SC = Senatus consultum
SO= Symbolae Osloenses (Oslo 1922 ff.)
Sommer, Hdb. =F. Sommer Handbuch der lateinischen Laut- und Formen-
lehre (Heidelberg 1913 2 ; 1948 3)
ABBREVIATIONS 181
SVF= Stoicorum Veterum Fragment a, ed. H. v. Arnim, 4 vols. (Leipzig 1903-

1924; reprint 1964)
ThLL = Thesaurus Linguae Latinae (Leipzig 1900 ff.)
Walde-Hofmann =A. Walde Lat. Etymologisches Worterbuch, 3rd revised
ed. by J.B. Hofmann (Heidelberg 1938)
WSt = Wiener Studien (Vienna 1879 ff.)
INDEX OF SELECTED PASSAGES
ALCIDAMAS
de soph. 18: 7 & n. 34
APULEIUS
met. 2,4: 170 n. 6; 2,18,18: 171; 3,14: 172 n. 14; 3,27: 167ff.; 8,11: 172n. 16; 11,1: 173;
11,5,20: 171; 11,12: 172; 11,13: 172 & n. 18
CAESAR
Gall. 7,27: 59jf; 7,89,3: 81 n. 50
oral. ( ORF') fr. 29: 54ff
CATO
ad fil. (Jordan) fr. 15: 7
agr. praef: iff; I ,3: 25 n. 94
oral. (ORF') fr. 29: 30 n. J15; fr. 58: 43f; fr. 163-169: 9ff
orig. (HRR 2 ) fr. 83: 2lff
CICERO
A tt. 12, !Of.: 130
Brut. 104: 39 & n. 31; 125: 36n. 16,40 & n. 37; 210f.: 39 n. 31; 252: 57; 261:57 & n. 16, 58;
262: 65 & n. 55; 294: 24
Catil. 1,1: 174
de or. 3,214: 48, 49 & n. 92; 3,224: 48; 3,227: 48
Mur. 88f.: 49f
rep. 2,1: 2 & n. 7; 6,25f.: J02ff.
Verr. II 1,47: 174; II 5,161-163: 33ff, 37ff
CLAUDIUS IMPERATOR
oral. de iure han. Callis dando (lnscr. lat. ed. Dessau no. 212): 136ff
DEMOSTHENES
or. 28,18: 49 & n. 89
ENNIUS
trag. 231 (Ribbeck): 49ff & n. 89
EURIPIDES
Med. 502-505: 49 & n. 89
FRONTO (ed. van den Hout)
p. 117: 63; p. 132: 84
GELLIUS (A.)
3,7: 2lff; 6,3: llff.; 9,13: 86f; 10,3: 33ff; 12,2,1: 123; 13,29,4: 69; 17,18: 85
C. GRACCHUS (ORF')
fr. 17: 45; fr. 24: 45; fr. 32: 45f; fr. 43: 46; fr. 44: 46; fr. 48: 33ff; fr. 49: 38f, 47; fr. 61:
48ff
HIRTIUS
Gall. 8 praef. 4-6: 65 & n. 56
HORACE
carm. 2,10,17f.: 133; 2,18,15: 133; 3,6,8: 130
LIVY
7,9,6- 7, 10,14: 87ff.
LUCIAN
de hist. conscr. 48: 65 & n. 58
1 R1
184 INDEX OF SELECTED PASSAGES
NAEVIUS
fr. 5,1 (Morel): 70
NEPOS (CORN.)
Chabr. 1,2f.: 170 & n. 5
OVID
am. 2, 1,2; 162
her. 20,220: 130
met. 3,229: 170 & n. 6
PETRONIUS
42,7: 130; 44,14-45,6: 125ff; 52,1: 131
PLAUTUS
Persa 753-755: 60
PLINY, YOUNGER
epist. 1,6: 160ff; 1,20: 157 n. 47; 7,9,8: 161 n. 6; 9,10,1: 161 n. 5
PLUTARCH
Caes. 3: 56f; 50,3: 40 n. 40
Cato maior 12,7: 40 n. 40
Tib. Gracch. 2: 36, 40 n. 37, 44 n. 57, 48 & n. 85
QUADRIGARIUS (CLAUD.)
fr. lOb (HRR 2 ): 86JI
QUINTILIAN
inst. 8,3,48: 25 & n. 95; 10,1,7: 38f; 10,1,32: 157 n. 48, 159; 10,1,98: 124 & n. 54;
10,1,114: 64; 10,1,130f.: 123 & n. 50
RHETORICA AD HERENNIUM
3,2,3: 3 & n. 12
SALLUST
Cali!. 8,2f.: 24n. 90; 10,1: 76; 10,5f.: 68ff; 11,1:76
his!. (Maurenbrecher) fr. 4: 24 & n. 92; fr. 55,24: 76
lug. 113,3-7: 78fi
SENECA, ELDER
contr. 9,1,13: 82 & n. 55
SENECA, YOUNGER
dial. 10 (= brev. vi/.),12: 134
epist. 1,1: 112ff; 114,17: 76
SUETONIUS
Claud. 41: 147
TACITUS
ann. 11,24: 140ff
dial. 18: 36 & n. 16; 28,5f.: 64
WORD AND SUBJECT INDEX
by Wilfried Stroh
The names of the authors discussed are only included in the index if
the discussion cannot be located by means of the table of contents.
ablative absolute: as feature of official language (commoner in Caesar than

in historians) 60; in Apul., Caes. and Corn. Nepos 170; in Sail. 81;
subject of, picked up in different case (in Caes.) 61f.; word order
within (in Caes.) 61 (see also 'participial constructions')
abstract nouns: as subject of active verbs 108; psychological, in Cic. 42
action instead of abstract interpretation, in Livy 97
acutum (6~1'n:11<;), in Gracchus 44ff.
addressee: in Pliny 162ff.; in Sen. 115ff.
adjectives: emotional, in Cic. 42; forward position of 47 n. 80, 98
advance notice, in ablat. construction: in Apul. 168f.
•adversative particles, in Tac. 156
aesthetic and ethical considerations, in Livy 90
alliteration: in Livy 97; more Roman than Greek 15
anaphora: in style of hymnic predication, in Cic. 109 (see also 'repetition,
verbal')
annalistic style, different from commentarii style 59f., 63 n. 46
'annotations' in text: in Claudius 151f.
announcement of theme, at start of sections: in Cato 29f.
anticlimax, in Cato 15 (see also 'end cola, shorter')
antithesis: frequent in Gracch. 45; non ... sed, in Cic. 109 (see also 'contrast';
'non solum ... sed etiam')
appellitare 149f. & n. 27
aptum, in Tac. 147 (see also 'ineptum')
Apuleius: end of Met. connected with rest 173 (cf. table of contents)
aqua, of rain in religious language 129 & n. 27
archaic: duplication (in Cic.) 106, (in Livy) 97 (see also 'synonyms, accumu-
lation of); elements (see 'word order'); fullness, in Cato 27 (see also
'ubertas'); language (in Cato) 4, (in Quadrig.) 94; ring composition (in
Cato) 4, 8, 15, 18, (in Sal!.) 73; sequence of ideas, in Cato 11; thought
(in Cato) 3, (in Sail.) 77
archaizers, on Sen. 123
archaizing orthography, in Sail. 68
argumentation, technique of: in Cato 1lff.
Asianic eloquence 36
asymmetry: in Claudius 150; in Sal!. 74ff.; in Tac. 154 (see also 'inconcinnity')
asyndeton: in Caes. 81 n. 50; in Cato 30; in Cic. 104; in Gracch. 39; in Sail. 81,
83f. (see also 'quickening of pace'; 'sentence connection')
Atticist, Caesar as? 57f.
audience: difference between listeners and readers 157
auxilium 171
185
186 WORD AND SUBJECT INDEX
Behaghel's law: see 'end cola, shorter'; 'law of increasing members'

brevity (brevitas, cruv-ro~-tia): as stylistic virtue 40 n. 40; in Cato 115; in
Gracch. 40ff.; in Petron. 130; in Quadrig. 95; in Sail. 71f. & n. 18, 80,
84; in Sen. 116; in Tac. 155, 157; of individual sentences, in Sen. 120f.;
of letters 162 (see also 'Latin'; 'letter')
Caesar: style of, compared to Sal!. 85 (cf. table of contents)

caldicerebrius and similar compounds 127f.
Cato: influence of, on Sail. 76f.; style of, compared to Sail. 8lff. (cf. table of
contents)
cautious expression, in Claudius 148
'centripetal' style, in Sal!. 79, 81ff.
chiasmus: in Cic. 105, 109f.; in Pliny 163f.; in Sail. 74 (see also 'archaic ring
composition'; 'parallelism'; 'sequence of thought'; 'word order')
Cicero: narrative technique of, in relation to Sail. 82; on Caes. 65; on Cato
24; on Gracch. 40, 48; quotation of, in Apul. 169; reminiscences of, in
Apul. 174; style of, in relation to Sail. 85; style of, in relation to Sen.
120f. (see also 'miseratio'; cf. table of contents)
clarity of form, in rendering trivial speeches: in Petron. 135
classics, mockery of: in vulgar sections of Petron. 133
clausulae: see 'prose rhythm'
climax: in Sen. 122; through breakdown into individual phases (in Cic.) 41ff.,
50, 82, (in Pliny) 164f.
cogitatio 160 n. 4
colloquial language: in Cato 16, 23ff.; in Gracch.? 38ff.; in Petron. 126ff.,
133, 135 (see also 'oral style')
combination of words, unusual: in Cic. 107
commentarii 65; style of 56, 59f.
communicative style: in Claudius 158; in Pliny 166
compounds, Early Latin 29 & n. 114
conclusions, 'fading' 105, 121
contrast: between subject-matter and expression (in Petron.) 130; and drastic
image, to round off section (in Petron.) 133; in Pliny 164; in Tac. 154;
through change of stylistic level (in Petron.) 129 (see also 'juxta-
position ... ')
conventional language: avoided by Tac. 156; exposed by Petron. 131; in
Claudius 147ff., 156
Curial style 7 n. 35, 148f.
'decrease', rhetorical: see 'end cola, shorter'

delivery, skilful: in Caes. 58; in Gracch. 48f.
demeanour, outward: used to represent inner feelings, in Sal!. 79f.
demonstrative, emphatic: in Cato 5 (see also 'sentence connection')
Demosthenes, imitated by Gracch.? 49
'detachment', linguistic: in Tac. !59
diatribe, Stoic-Cynic: in relation to Sen. 124
dicitur, distancing 80
diminutives, in Petron. 129
donee, for dramatic turn of events 169
WORD AND SUBJECT INDEX 187
duplication: in Cato 26; in Curial style and in Claudius 148f. (see also
'archaic duplication'; 'synonyms, accumulation of)
elatives: see 'superlatives'

elegantia: in archaizers 123 n. 51; in Caes. 55, 60, 119; in Claudius 147
elevated style: in Cato 20, 24f.; in Tac. 158
ellipses: in Petron. 130; in Tac. 154f.
emotion: highlighted in Quadrig. 93; presentation of, in Cic. (as compared to
Gracch. and Cato) 42f.
emphatic positions in sentence: emphasis on beginning and end, in Cato and
Cic. 5, 8, 105; emphasis on centre, in Cic. 105; emphasis on
penultimate position, in Cic. 105 (see also 'proper names'; 'verb';
'word order')
end cola, shorter: in Cato 5, 8, 20; in Sen. 115f. (see also 'lawofincreasing
members')
'enharmonic change', in Sail. 75
Ennius, in Gracch. 49ff.
epic elements: in Livy 91f., 96f.; in vocabulary of historians 69f. (see also
'poetry')
'essay' character of Seneca's letters 114f. n. 27, 124
est with infinitive 1 n. 2
Euripides, in Gracch. 49
figura etymologica, in Cato 27

figures of speech, in Cato 19
form and content, in Cato 31f. (see also 'objective style')
Pronto, on Sail. 84
fullness: see 'archaic fullness'; 'long-windedness'; 'ubertas'
Gellius: on Cato llff.; on Gracch. 40

generic style pref., 56f.; determined by audience 157; and stylistic approach
(in Sen.) 120, (in Tac.) 159
gerund and gerundive, in Early Latin 27 & n. 106
grandeur, in Petron. 129 (see also 'oral style')
gravitas: in Gracch. 40, 44f.; in Quadrig. 95
Grecisms, in Sail. 79
hie = meus 27f. & n. 108

Hirtius, on Caes. 65
historical objectivity, in Caes.? 66f.
historiography: style of, in Claudius 149f. (see also 'commentarii, style of);
theory of 65
homoioteleuton, in Cato 15 (see also 'rhyme')
Horace vulgarized, in Petron. 133
humour: in Apul. 175; in Claudius 158; in Pliny 162; and irony (or self-irony)
158
hymnic style: see 'anaphora'
hyperbaton: in Caes. 62; in Cato 27 (see also 'separation')
hypercorrectness, in Petron. 127
Ignatius Loyola: follows Seneca 123

imperfect: aids visualization, in Cic. 41f.; breaks up the narrative, in Cic.
41f.; creates suspence, in Apul. 168 (see also 'perfect, historic')
implicit meaning, in Petron. 130f.
inconcinnity: in Sail. 74, 84; in Tac. stronger than in Sal!. 84 n. 65 (see also
'asymmetry')
inconspicupusness of vocabulary, as stylistic device: in Cato 6
ineptum 147 (see also 'aptum')
infinitive: historic, in Sal!. 71f. & n. 18, 75; and indicative, to present insight
and action respectively 100
inscription-like style 31f.
inversion, in Gracch. 47
irony, in Tac. 156, 158 (see also 'humour')
juxtaposition of opposites: in Petron. 129; in Pliny 163; in SaiL 74 &n. 28 (see

also 'contrast')
Kffi!ltKOc; XUPUKTllP 40 n. 35
language: and gesture 27f. & n. 108; late antiquity's attitude to, in Apul.
171ff.
Latin: architectonic trend of 46; brevity of 40 n. 40; history of particular
words in pref; 'poverty' of6, 37; suitability of, for philosophy Ill &n.
22
Latinitas, in Gracch. 38ff.
law of increasing members: in Caes. 55; in Sail. 73, 81; in Sen. 114f.; reversed
in Cato 31f. (see also 'end cola, shorter')
letter: as conversation with those absent 162 & n. 8; theory of 161 & n. 6
Livy: influence of, on Claudius 149 n. 25 (see table of contents)
long-windedness: as means of oratorical persuasion, and required by
diplomacy 157 & n. 47 (see also 'archaic fullness'; 'ubertas')
main idea (or main action) in subordinate position: in Quadrig. 99; in Tac.
152f. (see also 'subordinate clause')
male, in cases of bereavement 130 & n. 31
many-sidedness, of Cic.'s style Ill & n. 21
metaphors: from law, finance etc. (in Sen.) 119f.; in Petron. 128f.; pectus,
lingua (in SaiL) 70
miseratio, treatment of: in Cato and Cic. 42 & n. 47, 44; in Cic. and Gracch.
49ff.
modesty, formulae of (in political speeches): in Claudius 148
mortales 68ff., 96
mundities 40
mythological figures, in Pliny 163, 166
narrative: standpoint, change of (in Cato) 30 (see also 'perspective');

structure, a,s 'large-scale syntax' pref (on individual authors see table
of contents; see also esp. 'historic present')
negative: initial position of, in Pliny 163
neuters, changed to masculines in vulgar language 127

non prius ... quam, for dramatic turn of events 169
non solum ... sed etiam, varied in Pliny 164
nouns: conversion to, produces detachment 97; 'noun style', in Cic. 110
numerals, put behind their nouns 162
objective style: in Caes. 59ff.; in Cato 6f., 31f.

observation: less important than typology, in Livy 92
oral style: in Cato 6f., 25; in Claudius 150f.; in Quadrig. 94; solemn (in Cato)
16, 59f.
oratio obliqua, in Caes. 60
otium, of writer 162
parallelism: in Caes. 55; in Cato 4; in Cic. 109f.; in Sail. 72f., 74 (see also
'chiasmus'; 'symmetry')
parataxis: and hypotaxis, in Cato 32; and periods, in Cato 25 (see also
'sentence connection')
parenthesis: explanatory, in historiographical style 149; to create deliberate
pause, in Sal!. 81; to create hyperbaton, in Quadrig. 94
participle: and finite verb (sequence in Apul.) 169; prolongs tension, in Apul.
168f.
participial constructions: in Caes. 62f.; in Cic. 104f., 111; in Cic. and
Claudius 150; in Cic. and Sal!. 82; in Gracch. 39; in Livy and Quadrig.
101; literary function of 61ff., 82, 84 (see also 'ablative absolute')·
perfect: historic, dry compared to imperfect 42; in -ere instead of -erunt 149,
156 (see also 'present, historic')
perspective: in first-person narrative (in Apul.) 169ff.; in Sal!. 79f., 82;
progressive narrowing of (in Cato) 28 (see also 'narrative standpoint')
plainness, as means to credibility in orator 52
pleonasm, apparent: in Claudius 148
rcAoK'Il: see 'traductio'
plural, generalizing 174 n. 23
Plutarch, on Gracch. 36f., 48
poetry: quotations of, in schools of rhetoric 49
poetic elements: in Apul. 175; in Cato 30 n. 115; in Quadrig. 94; in Sal!. 70,77
(see also 'epic elements'; 'Horace')
point, epigrammatic: at end of sections, in Petron. 132; at end of sentences, in
Sen. 121; with structural function, in Pliny 165
'pointed' style: in Augustan orators and in Sen. 121 & n. 45; in Gracch. 45; in
Tac. 158
pose, in Caes. 67
position, outer: emphasizing most important words, in Caes. 62
praesidium 171
prefix and preposition, different in Tac. 153
present, historic: and historic perfect (in Cato) 30 & n. 115, (in Gracch.) 41,
(in Livy) lOOf., (in Sal!.) 83; use of, in Sal!. 80
pronus 172
proper names, position of in sentence: in Claudius 151; in Gracch. 47
proprietas verborum, in Claudius 156 n. 43
prose rhythm: in Apul. 174 & n. 22; in Caes. 56, 63; in Gracch. 45 n. 69; in
Pliny 161 n. 6; in Sal!. 83; in Sen. 114, 120 n. 44; in Tac. 155 (see also
'word order')
proverbial expressions: in Petron. 128f.; in Sen. 120
quickening of pace, stylistic: in Caes. 81 n. 50; in Sail. 75, 80f., 83f.

Quintilian: against excessive variatio through synonyms 119; on Caes. 64; on
Sen. 123f. & n. 50
rationality, of style: in Cic. 46 n. 75; in Gracch. 44ff., 51

repetition, verbal: developing thought 16; functions of 5f., 15ff.; in Caes. 57,
60; in Cato 5ff., 15ff.; in Cic. 108f.; in Claudius 148, 150; in
commentarii style 59f.; intensificatory 5, 42; non-intensificatory 5f.; of
colourless words 16; 'oral' 6f.; rhetorical 108; unrhetorical 38; with
change in function of words 108 (see also 'anaphora'; 'Quintilian';
'variatio ... ')
retardation: in Cato 30; in Cic. 42; in Livy 91; in Quadrig. 99; in Sal!. 73, 79;
and preparation (in Apul.) 168
rhetoric: devices of, in Quadrig. 94f.; Greek, in Cato? 3, 11ff., 19f.;
Hellenistic, in Roman orators 51; importance of, for structuring
thought pref.; in the service of philosophic guidance, in Sen. 122f.;
structural pattern of, in a letter of Pliny 165
rhyme, in Apul. 174 & n. 23
ring composition, in Cato 30 (see also 'archaic ring composition')
rosarius, in Apul. 171
roses, symbolism of 172 & n. 18
sacral language 3, 7, 26
sa/us- salutaris, in Apul. 171f.
schools, Greek: tradition of, in Gracch. 49
Seneca, Elder: on Sal!. 82
sentence:
sentence connection: archaic 99; in Cic. 104; in Petron. 131ff.; in
Quadrig. 93, 99; in Sal!. and Cato 83f.; in Sen. 120; through asyndeton
(in Cic.) 41, 110, (in Livy) 98, (in Quadrig.) 99, (in Sen.) 120, (in Tac.)
155f. (see also 'asyndeton'); through demonstratives 39 (see also
'traductio')
sentence length: gradual reduction of, in Petron. 135; sequence oflong
and short sentence, in Caes. 63f.
sentence structure: both self-contained and setting off emphatic parts,
in Quadrig. 95 (see also 'ablative absol.'; 'asymmetry'; 'conclusions,
"fading"'; 'emphatic positions .. .'; 'law of incr. members'; 'main
idea .. .'; 'parallelism'; 'parenthesis'; 'participial constructions'; 'sub-
ordinate clause'; 'symmetry'; 'verb'; 'word order')
separation, 'natural': in Cato 17f.
sequence of thought: in Cato 3; in Sal!. 72f., 74; in Sen. 118 (see also 'archaic
ring composition'; 'archaic sequence of ideas')
sic habere 109
simplicity, artful: in Caes. 59ff.; in Cic. 106f.; in Pliny 161 n. 6
slogans, political: avoided by Sail. 77 n. 41, 83

spiritualizing: of facts of social psychology, in Sen. 120; of meanings of
words, in Cic. 107f., 111; of old-fashioned Roman concepts, in Sail. 77
n. 40
state, Roman concept of commitment to: influenced by Cato 28 n. 111
stereotyped character of colloquial speech, in Petron. 131
structural relationships, larger: absent in Quadrig. 100
style:
development of: in Caes. 67 & n. 67; in Cic. 58, 109f. &n. 17 andn. 19;
in Sail. 79 n. 47, 84f.; in Tac. 159 & n. 56
differences of: between beginning and main section of speech, inCa to
14; between speeches and historical writing, in Cato 23ff.; between
speeches and narrative, in Tac. 157f.; present in Cato 2ff. (see also
'generic style')
generic, period and individual styles pref.
subigere 71
subordinate clause: containing crucial development in narrative, in Apul.
169 (see also 'main idea ... ')
Suetonius, on Claudius' style 147
superlatives: capped by positives, in Sal!. 73, 75; (elatives) used sparingly, in
Tac. 157; stereotyped, in Claudius 147f.
'supplementary' style: of Claudius 151; of Claudius and Tac. compared 158
symmetry, in Caes. and in Cic.'s early speeches 58 (see also 'asymmetry';
'parallelism')
synonyms, accumulation of: in Cato 7f., 14, 26; in Gracch. 44 n. 56 (see also
'archaic duplication'; 'repetition, verbal'; 'variatio')
tautology: for sake of period, in Gracch. 45f.

technical terms: used by Claudius and avoided by Tac. 156f. & n. 45
tenses, use of: see 'imperfect'; 'infinitive'; 'narrative structure'; 'perfect';
'present, historic'
thought: associative style of, in Petron. 134; reader stimulated to thought by
stylistic means (in Sail.) 70f., 75, (in Sen.) 122
'three-dimensional' effect, in Livy's narrative style 101
'thrust', at beginning of sentence: in Cato 17
Tiro, critical of Cato llff.
'toning down', linguistic: in Livy 96
traductio (rcA.oKrj): in Cato 27; in Curial style and in Claudius 149; in Livy 97;
in Quadrig. 95
tricolon 165 & n. 13
two-element structures: in Gracch. 44; in Pliny 164f.
ubertas, stylistic: in Cato 7, 44; in Gracch.? 37 (see also 'archaic fullness';

'long-windedness'; 'pleonasm')
utilitarianism, Roman 163
utilitas and honestum, as rhetorical categories 3
variatio: pursuit of (in Pliny) 164, 166, (in Sail.) 75; through synonyms (in
Caes.) 57, (in Cic.) 108, (in Quadrig.) 93, (in Sen.) 119 &n. 38 (see also
'Quintilian'; 'synonyms, accumulation of; 'vocabulary, choice of)

verb(s): (main verb), position of in sentence (in Caes.) 61, (in Cato) 17, 44, (in
Cic.) 42, 105, (in Claudius) 151, (in Tac.) 155f. (see also 'emphatic
positions .. .'; 'word order'); picked up by participle (see 'traductio');
use of, in Cic. 110
verbal reference, of ancient thought 122f.
verruca 25f.
vocabulary, choice of: in Cic. 105; in Petron. 127f.; in Sall. 68ff.; in Sen. 123
(see also 'variatio through synonyms')
vulgarisms, in Petron. 126f.
winning of converts, style meant for: in Sen. 121

word order: additive, as archaic feature 17; analytic 106f., 162; change of,
lending distinction to language (in Apul.) 174f., (in Tac.) 174; in
Gracch. 47; in Livy 98; inverted for sake of rhythm, in Caes. 56;
inverted, supported by chiasmus 155 n. 39; self-enclosed 62 (see also
'ablative absol.'; 'adjective'; 'emphatic positions .. .'; 'negative'; 'nu-
merals'; 'position, outer...'; 'proper names'; 'sentence structure';
'verb')
work, intellectual 162
zeugma: as pattern of thought, in Tac. 153, 155

Masters of Roman Prose (Anthologie Stylistique Commentaires (Michael - Von - Albrecht) PDF

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Masters of Roman Prose (Anthologie Stylistique Commentaires (Michael - Von - Albrecht) PDF

Hochgeladen von

Copyright:

Verfügbare Formate

ARCA

Classical and Medieval Texts, Papers and Monographs

General Editors: Francis Cairns, Robin Seager, Frederick Williams

MICHAEL VON ALBRECHT

Translated by NEIL ADKIN

Die Meister Romische Prosa first published by

English translation first published by

All rights reserved. No part of this publication may be repro-

British Library Cataloguing in Publication Data

Albrecht, Michael von

Printed in Great Britain by

1 The beginnings of literary prose: M. Porcius Cato

2 Two great orators: C. Gracchus (154-121 B.C.)

3 Caesar (100-44 B.C.)

4 Sallust (b. 86 B.C.)

5 Sullan and Augustan historiography: Claudius

6 Two philosophical texts

7 Petronius (d. 66 A.D.)

9 The Younger Pliny (cos. 100 A.D.)

List of abbreviations 177

history, or to present the development of Latin prose style in its

Altertumswiss. I (Hildesheim/New York 1976).

Michael von Albrecht

Landwirtschaft (Berlin 1963); F. Klingner Romische Geisteswelt (Munich 1956 3) p.56

and a good husbandman'. Whoever was thus praised received in their

A different interpretation is given in Thielscher Joe. cit. (after Friedrich: institutum as

10 'Zum Stil Catos in De re rustica' WSt 43 (1922/23) pp.l68-172. Kappelmacher

leaves it open whether Cato's arrangement here is deliberate or not (p.l71).

corresponding parts in the manner of old-Italic sacral language,

11 Orationis ratio I (Amsterdam !963) pp.2!-24. esp. 22f.

that the form of the praefatio is actually determined by Hellenistic

meticulousness in crucial areas. Cato devotes attention to the

\ is its ultimate purpose.

32 On the stylistic value of existimare see B. Axelson Unpoetische Worter (Lund

device clearly indicates that the introduction to De agricultura is

II. Speech in the Senate for the Rhodians 36

36 Text in Malcovati Oratorum Romanorum Fragmenta (Turin 1976 4 ) fr. 163-171.

omnia plura habere volumus, et id nobis impoene est.

criticized Cato's speech. 44 Tiro's criticisms are worth mentioning,

to be the weakest part of the case into a strong-point.

particular reference to neologisms resulting from synonym accumulation). Early

Cato fragment 23 Malcovati 1 (= 21 Malcovati 4) egoque iam pridem cognovi atque

At the beginning of the speech there is also a predominance of

docent; transvorsum trudere.

Cato' AJPh 8 (1887) pp.292-328. Et predominates in De agricultura (Elmer pp.307f.).

clauses, 56 while et stands between connected words: 169 mihi et liberis

separation that is involved but the 'natural' kind, such as is admitted

Perhaps the question of Greek influences on Cato has been

III. A Roman Leonidas 80

perhibuit 83 il!oque facto, quod illos milites subduxit, exercitum ceterum

- 83 -Quicherat's praehibuit is preferable.

of is and of -que, which co-ordinate all kinds of secondary ideas. The

republican historians (W.O. Lebek Verba Prisca Hypomnemata 25 [Gottingen 1970]

widespread use in the Origines of the conjunction atque, 93 which in

'Per un riesame degli arcaismi Catoniani (rileggendo il Till)' Quaderni dell'Istituto di

which lends forcefulness to the account. 108

the name is not mentioned, the tribune becomes a representative of

Cato's influence than is commonly realized.

in a delightful construction: in expectando sunt. Here therefore the art

at the beginning, where in regard to content the surprising turn of

at que in foro media nudari ac deligari et virgas expediri iubet. clamabat

garment that he could change as he pleased'. 10 He demonstrated

only indirect knowledge of Gracchus' speeches. 20 Besides, the whole