The Rise of Semiotics and the Liberal Arts: Reading Martianus Capella's "The Marriage of

Philology and Mercury"

Author(s): Han-Liang Chang
Source: Mnemosyne, Fourth Series, Vol. 51, Fasc. 5 (Oct., 1998), pp. 538-553
Published by: BRILL
Stable URL: http://www.jstor.org/stable/4432880 .
Accessed: 30/01/2011 22:14

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .
http://www.jstor.org/action/showPublisher?publisherCode=bap. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

BRILL is collaborating with JSTOR to digitize, preserve and extend access to Mnemosyne.

http://www.jstor.org
 THE RISE OF SEMIOTICS AND THE LIBERAL
ARTS: READING MARTIANUS CAPELLA'S THE
MARRIAGE OF PHILOLOGY AND MERCURY*)

BY

HAN-LIANG CHANG

Current interest in mediaeval trivium and the revival of rhetoric

have led researchers to re-examine the
implications liberal of the
arts, as outlined by Martianus Capella (fl. 410-39) and Boethius (ca.
480-524), in relation to contemporary thinking of signs. In his pro-
posals for a history of semiotics1), Umberto Eco identifies three cat-
egories of writers who can be said to have contributed to such a his-
tory. His checklist runs the
gamut from
explicit of signs,theories
such as those of Augustine and Boethius, through repressed theories
abstractable from the writings of the Church Fathers who discuss
language in general, and finally to the so-called encyclopedic semi-
otic practices, such as early Christian symbology.
Meanwhile, the revival of rhetoric, especially by American neo-
pragmatists over the last two decades, has evoked once again a time-
honoured debate between rhetoric and logic in the West, which can
be traced back to Plato and Aristode. One is reminded, among other
things, of the distinction between two critical models of demonstra-
tion and persuasion which Stanley Fish makes in the tradition of the
Sophists2). The debate has been otherwise interpreted as that be-

) The author wishes to thank the National Science Council of Taiwan for sup-
porting his research project on the history of semiotics and The University of
Manchester's Department of English and American Studies for an Honorary
Research Fellowship in 1997, during which time the paper was written and pre-
sented in the Postgraduate Seminar on 22 April 1997. He gratefully acknowledges
the learned comments of the journal's readers regarding the Latin quotations and
a reference to the ideas of Ilsebraut Hadot in her Arts lib?rauxet phibsophie dans la
pens?eantique(Paris 1984), where she argues that the system of the seven arts has its
roots in Neoplatonism, as it appears in Augustinus de Ordine.
1) Umberto Eco, Proposahfor a H?tory of Semiotics,in: Tasso Barb? (ed.), Semwtics
Unfolding:Proceedingsof the Second Congressof the InternationalAssociationfor SemwticsI
(Berlin 1983), 75-89.
2) Stanley Fish, Is Therea Text in This Class? The Authorityof Interpretive
Communities
(Cambridge, Mass. 1980).

Koninklijke Brill NV, Leiden, 1998 Mnemosyne,Vol. LI, Fase. 5

 THE RISE OF SEMIOTICS AND THE LIBERAL ARTS 539

tween the two traditionally hostile disciplines of philosophy and

rhetoric.
This opposition gets a linguistic turn in Paul de Man's celebrated
essay Semiobgy and Rhetoric where the author casts
light on the new
issue by situating it in the mediaeval liberal arts education system
and by introducing the role of the as-yet-non-existent discipline of
semiotics. According to de Man, among the three language-focused
disciplines that constitute the trivium, logic is closer than grammar
and rhetoric to the quadrivium that accounts for the mediaeval
scholar's knowledge of the world. He does not bother to dwell on
the tension between logic and rhetoric, but moves to discuss the dis-
crepancy between grammar and rhetoric, grammar and logic, which
French semiologists such as Todorov and Genette have failed to no-
tice, as revealed by their conflation of the trivium in such titles as
Grammar of the Decameron and Figures III 3).
De Man seems to be suggesting a new spectrum of relationships
among the seven arts, based on their truth-claims or their transitiv-
ity to the world under investigation, and that rhetoric, with the sig-
nified truth bracketed, standing at one extreme of the spectrum, is
less 'transitive5 than any other discipline. Notwithstanding the seven
arts' intransitivity, de Man's attempt at reconstructing semiology in
the trivium has opened up a line of inquiry into the disciplinary ori-
gin and history of semiotics.
As is well-known now, the liberal education of seven arts, though
traceable to classical antiquity, is formulated in the late Roman peri-
od and the early Middle Ages, in particular, by Martianus Capella

3) Paul de Man, Allegoriesof Reading:Figurai Languagein Rousseau,Nietzsche, Rilke,

and Proust (New Haven 1979). De Man has himself conflated two kinds of rhetoric,
i.e., as figurative language and as persuasion, in order to accommodate his metahis-
tory of modern critical discourse since French 'semiology'. In this metahistory the
first kind of rhetoric is equivalent to the semantic (and syntactic) universe of the lit-
erary text, or the first semiotic order of signification; and the second kind to prag-
matics, or the second semiotic order of communication. Thus the paradigm shift in
literary linguistics around the early 1970s can be accounted for by the classical ten-
sion between rhetoric and logic. Meanwhile, one could argue that the relationships
among the three units of the trivium have never remained stable and therefore call
for radical historicization. Then this would be the problem of Charles Sanders
Peirce too (see below). The same can be said of rhetoric. Compare its modern vari-
ations in Kenneth Burke and Wayne Booth, and you will see its wide semantic
range.
540 HAN-LIANG CHANG

in his De Nupt?is Phihlogiae et Mercurii 4) and Boethius in his commen-

taries on or adaptations of Aristotle, Nicomachus, Porphyry, Euclid,
Ptolemy, and Cicero. Although Boethius coins only the term 'qua-
drivium' in his De Insntutione Arithmetica, and the term 'trivium' is only
a mediaeval derivation, it is essentially his structure of the seven arts
that is to be followed throughout the Middle Ages5).
Boethius makes the distinction between 'trivium' and quadrivium
already on the basis of the former's linguistic instrumentality as or-
ganika and the latter's coverage of areas of speculative knowledge. A
question can be raised here: Does the trivium's instrumentality serve
the purpose of the quadrivium? Or, is there transitivity between the
two, i.e., three paths leading to four paths? The answer seems to be
both negative and positive. If one examines the quadrivium, one
finds all the four Pythagorean mathematical sciences?arithmetic,
geometry, astronomy, and music (the order being different in Mar-
tianus and Boethius)?have their own distinct representational sys-
tem which is non-verbally symbolic, nor does it need re-articulation
in a metalanguage, although neither Martianus nor Boethius can
afford not to use language as a modelling and interpreting system.
Unlike what our authors and their commentators believe, because of
the inconvertibility of systems, not only is there an unbridgeable gap
between the trivium and the quadrivium, but also there is another
gap between the quadrivium as abstract speculations and the ab-
solute knowledge in the incorporeal world to which they aspire. In
the enclosed logocentric system, the signifier of quadrivium points

4) Minneus Felix Martianus Capeila, De Nuptiis Phibbgiae et Mercurii(Bibliotheca

scriptorum Graecorum et Romanorum Teubneriana), rev. ed., ed. Adolf Dick
(Stuttgart 1969); De Nuptiis Phibbgiae et Mercurii,ed. James Willis (Leipzig 1983); The
Marriage of Phibbgy and Mercury,trans. William Harris Stahl and Richard Johnson
with E. L. B?rge, vol. 2 of Martianus CapeUaand the Seven LiberalArts (Records of
Civilization: Sources and Studies, 84), 2 vols, ed. William Harris Stahl (New York
and Oxford 1977). All the Latin quotations are from the Dick edition and herein-
after cited as De Miptiis; and all the English quotations are from the Stahl transla-
tion and hereinafter cited as The Marriage.Unless otherwise indicated, they are cited
by verse numbers in the text rather than separately footnoted.
5) Anicius Manlius Severinus Boethius, De InstitutioneArithmetica,ed. and trans.
Michael Masi, in: BoethianNumber Theory:A Translationof the De InstitutioneArithmetica
(Studies in Classical Antiquity, 6) (Amsterdam 1983), 69-188. For Boethius on the
quadrivium, see, for example, Alison White, Boethius in the Medieval (Quadrivium, in:
Margaret Gibson (ed.), Boethius:His Life, Thoughtand Influence(Oxford 1981), 162-
205; Henry Chadwick, Boethius:The Consolationsof Music, Logic, Theobgy,and Phibsophy
(Oxford 1981).
 THE RISE OF SEMIOTICS AND THE LIBERAL ARTS 541

nostalgically but futilely towards the ultimate signified that is the har-
monious order of divine creation.
It is no accident that Charles Sanders Peirce formulates his gen-
erad semiotic in terms of the trivium. First he explicitly equates semi-
otic to logic6); but in his characteristically cryptic argument for tri-
ads, semiotic contains three branches: pure grammar, logic proper,
and pure rhetoric because every sign (representamen) is connected
with three things?ground, object, and interpr?tant. Logic as semi-
otic (or the other way around, semiotic as logic) subsumes the trivi-
um, including logic itself. Or perhaps one should say: LOGIC as
theory of signs covers the trivium, including logic proper, which is
not LOGIC proper. This word-play is not so trivial as it first appears
if seen in the light of the fortune of rhetoric in 12th- and 13th-cen-
tury scholasticism, where dialectic dominates the trivium, and rhe-
toric is even reduced to a branch of logic7). At this point we could
surmise the Peircian semiotic's possible links with the trivium, the
author's indebtedness to scholasticism being already well-document-
ed8). We could further infer that the general semiotic Peirce proposes
is derived from the language-focused trivium, but it can be expand-
ed to deal with other representations, such as the mathematical qua-
drivium, as long as its members fulfill the conditions of ground, ob-

The word quadrivium(early manuscripts show quadruvium)first appears in the

Proemium of Book I of Boethius's De InstitutioneArithmeticawhere the author
observes: 'Among all the men of ancient authority who, following the lead of
Pythagoras, have flourished in the purer reasoning of the mind, it is clearly obvious
that hardly anyone has been able to reach the highest perfection of the disciplines
of philosophy unless the nobility of such wisdom was investigated by him in a cer-
tain four-part study, the quadrivium...'(71). Credits must also be given to Boethius's
predecessors, such as Varr?, Macrobius, Augustine, and Martianus Capella.
6) Charles Sanders Peirce, The CollectedPapers of CharlesS. Peirce,8 vols., vols. 1-
6, ed. Charles Hartshorne and Paul Weiss; vols. 7-8, ed. Arthur Burks (Cambridge,
Mass. 1980), 2.227. The citation of Peirce here follows the standardized practice by
referring to the volume number first and section number next.
7) Martin Camargo, Rhetoric,in: David Wagner (ed.), The SevenLiberalArts in the
Middle Ages (Bloomington 1983), 107-8.
8) Charles Sanders Peirce, Nominalismversus Realism,Early Nominalismand Realism,
Ockam, in: Writings of Charles S. Parce: A Chronobgual Edition, 5 vols, to date
(Bloomington 1982-1993), vol. 2, ed. Edward C. Moore et al. (1984), 144-54, 310-
36; Alan R. Perreiah, Peirce'sSemewticand ScholasticL?^:, Transactions of the Charles
S. Peirce Society 25: 1 (1989), 41-9; James Jak?b Liszka, A GeneralIntroductionto the
Semewticof Charles Sanders Parce (Bloomington and Indianapolis 1996). I have not
been able to locate the following article cited by Liszka, David Savan, Parce and the
Trivium,Cruzeiro Semiotico 8 (1988), 50-6.
542 HAN-UANG CHANG

ject, and interpr?tant. In fact, it is quite possible to recode the qua-

drivium in terms of the tripartite icon, index, and symbol9).
It seems that in both Martianus Capella and Boethius, the math-
ematical quadrivium is taken with higher regard than the 'trivial'
language Organon; but the matter is much more complicated than it
seems. A closer look will show their difference, especially regarding
the functions of language. In his De Institutione Arithmetica Boethius
proposes the quadrivium as speculative knowledge and asserts its
mastery is a prerequisite for the philosopher, ultimate goal is
whose
metaphysical contemplation. In his hierarchy, there is an ascending
order towards abstraction, from multitude (or numerical quantity),
represented by the twinning arithmetic and music, to magnitude (or
spatial quantity), represented by geometry and astronomy. A com-
prehension of both numerical and spatial
quantities helps us to as-
cend the lofty heights of philosophy.
What about the quadrivium's relationship to the Aristotelian Or-
ganon which, and whose commentaries, Boethius is to translate and
comment later? One say his interest in Peripatetic
could logic lies
primarily in its subservient
function to truth, i.e., in forming discov-
eries (topics) and judgements (analytics)10). Pythagorian mathematics
and Aristotelian logic are then the two systems which join to lay the
foundation of metaphysics. There is no question of one's modelling
the other. It should be noted that, under the general rubric of organi-
ka, rhetoric and logic are combined studies, but to Boethius logic is
clearly more important than rhetoric. In this sense, de Man's obser-
vation is accurate when he says logic is closer to the quadrivium than
the two other sister arts.
However, in De Consolatane Philosophiae, Boethius expresses his ap-

9) The Boethian quadrivium is based on a metalanguage dealing with the

mathematical concepts of multitude and magnitude, which can be adequately recast
in Peircian terms. Peirce constantly refers to the sign nature of algebra though he
seldom mentions music. However, in the growing interest in semiotics of music and
the already substantial body of secondary literature, two major trends can be iden-
tified and respectively termed Peircian and Greimasian, with the Peircian on the
rise because of the pragmatic aspect of indexicality. See Vladimir Karbusicky, The
Index Sign in Music, trans. Louise Duchesneau and Sid McLauchlan, Semiotica
66:1/3 (1987), 22-35. For a recent review and prolegomena, see, for instance, R.
Keith Sawyer, The Semwtics of Improvisation:The Pragmatics of Musical and Verbal
Performance,Semiotica 31:1/2 (1996), 269-306.
10) Jonathan Barnes, Boethius and the Study of L?gic, in: Margaret Gibson (ed.),
Boethius:His ufe, Thoughtand Influence(Oxford 1981), 73-89.
 THE RISE OF SEMIOTICS AND THE LIBERAL ARTS 543

prehension about logicians' verbal games (III. 12) and rhetoricians'

likelihood of forsaking philosophy's Ordinances' (II. 1). If we accept
de Man's rather hasty but useful equation of literature to rhetoric?
this is exactly what Martianus does in Book V?we will better appre-
ciate Boethius's valorization of philosophy at the expense of poetry?
maybe even music as performance. In the famous beginning of De
Consolatone, as if echoing Plato, the dying Boethius allows Philosophy
to scold the poetical Muses as 'tragical harlots' who 'kill the fruitful
crop of reason' (1,1); and later in Book II he hears her identify Music
as 'a little slave belonging to [her] house' (II,l)11). With Boethius,
music as performance is not accepted as part of speculative philoso-
phy, a point on which we will elaborate when analyzing Martianus's
text.
The picture is quite different in Martianus Capella. The whole
text of De Nuptiis is a highly embellished and often carnivalesque alle-
gory, within the frame of which the seven bridesmaids expound the
arts they govern. The author utilizes poetic devices and rhetorical
figures so profusely that he is even chided by Satura, the genre itself,
for tu?ngere ludiera praestas/uiliaque astriloquae praefers commenta puellae!
('fashion [ing] cheap and silly fictions') (808) without true intellectual
nourishment12). There is a fascinating moment in the course of nar-

11) Anicius Manlius Severinus Boethius,The Consolationof Phibsophy, trans. I.T.

(1609), rev. H.F. Stewart (The Loeb Classical Library) (London 1918), 129, 133.
12) Throughout the text, the poetic function is always foregrounded, even in the
books dealing with the seven arts. Towards the end of book II, the narrator says he
will nunc ergomythosterminatur('put aside all fable') (220), but book III opens with his
questioning the Muse who insists on using embellishment. Her argument is rather
persuasive: although the Arts 'tell that which is the truth', they have to be clothed
and cannot be given 'naked to the bridal couple' (219-222).

'atquin prioris ille

titulus monet libelli
mythos ab ore pulsos
Artesque uera fantes
uoluminum sequentum
praecepta comparare.'

'nil mentiamur' inquit

'et uestiantur Artes,
an tu gregem sororum
nudum dabis iugandis,
et sic petent Tonantis
et caelitum senatum?
aut si tacere cultum
placet, ordo quis probatur?'
544 HAN-LIANG CHANG

ration, which takes place after the three Latin arts of language have
been exposed and before the four Greek arts of mathematics are
introduced. The narrator Martianus, though well versed in lyrics13)
and mythical fabulization, should fail to recognize Philosophia (phi-
losophy) and Paedia (learning) despite hisenkyklopaedeia (all-round
education). There are several occasions on which the narrator teas-
es himself for his ignorance, including punning on his own name as
a cape lia1*).
Menippean satire and literary embellishments put aside, it would
be more interesting to compare Martianus's and Boethius's ordering
of the seven arts15). We would like to comment, in particular, on
their representations of music, because we believe the topic will
usher in the confrontation of opposing camps of semiotic thinkers of
our time, regarding the functions of language in representing other
semiotic systems.
Earlier we mentioned Boethius's relegation of performing music to
the servile division of footmen. Let us elaborate on this. With Boe-

13) Fanny LeMoine has identified fifteen different meters used by Martianus
Capella. See Fanny LeMoine, Martianus Capello:A Uterary Re-evaluation(M?nchener
Beitr?ge zur Medi?vistik und Renaissance-Forschung, 10) (M?nchen 1972), 7.
14) LeMoine (1972: 132; see ?. 13).
15) In style, De Nuptiis follows the tradition of Menippean satire attributed to
Varr? in mixing prose and poetry; where subject matter is concerned, it is a pro-
thalamion followed by convivial treatises on the seven arts, thus fully displaying the
generic trait of encyclopedic erudition. See Northrop Frye, Anatomyof Criticism:Four
Essays (Princeton 1957), 311. It would be an example dear to Frye, although in his
discussion of the encyclopedic genre Martianus's work is strangely absent. The
work, as Martianus tells his son in the frame, is 'a story which Satire [Satura]
invented in the long winter nights' (2); 'an old man's tale [senilemfabulam], a m?lange
sportively composed by Satire under lamplight...in nine books...[where she has]
heaped learned doctrines upon unlearned,...crammed sacred matters into secu-
lar;...commingled gods and the Muses, and...had uncouth figures prating in a rus-
tic fiction about the encyclopedic arts [disciplinascyclicasY(997-999).
Frye observes, "The Menippean satirist, dealing with intellectual themes and atti-
tudes, shows his exuberance in intellectual ways, by piling up an enormous mass of
erudition about his theme or in overwhelming his pedantic targets with an aval-
anche of their own jargon. A species, or rather sub-species, of the form is the kind
of encyclopedic farrago represented by Athenaeus' Deipnosophistsand Macrobius'
Saturnalia,where people sit at a banquet and pour out a vast mass of erudition on
every subject that might conceivably come up in a conversation. The display of eru-
dition had probably been associated with the Menippean tradition by Varr?, who
was enough of a polymath to make Quintilian, if not stare and gasp, at any rate
call him vir Romanorumeruditissimus'1 (ibid.).
 THE RISE OF SEMIOTICS AND THE LIBERAL ARTS 545

thius, music comes after arithmetic as representing the more intelli-

gible multitude, and as such is more basic if not lower than geome-
try and astronomy, both representing magnitude or spatial quantity.
Boethius distinguishes among three types of music: cosmic music
(musica mundana) which is natural though imperceptible to human
senses, human music (musica humana) which deals with the microcosm
of human body and its correspondence to soul, and, finally, instru-
mental music (musica Instrumentalis). He mentions the first two types
rather sketchily and is never to return to them as he promises. The
last one, which amounts to what we mean by music today, is, how-
ever, denounced by Boethius because it has recourse to the human
body only andthus fails to be speculative16). This 'militantly theo-
retical'17) position in favor of musica theorica has very little concern
with performance, or with what Roland Barthes has reinstated as
musica practica, a type not mentioned by Boethius but worth our
inquiry18). In fact, Boethius looks down upon composers and per-
formers because of the servile nature of their work. A true musician
is 'one who has gained
knowledge of making music by weighing (i.e.,
forming judgement) with the reason, not through the servitude of
work, but through the sovereignty of speculation'19). In short, the
true musician is what we know today as the music critic, who is curi-
ously denounced by Barthes for his linguistic mediation.
Boethius's concept is widely accepted throughout the Middle
Ages. A Carolingian exegete Remigius of Auxerre, who has com-
mented on both Boethius and Martianus Capella, even applies it to
glossing Martianus's allusion to Orpheus and Eurydice in Book DC
of De Nuptiis, the last book which deals with Harmony, in Martianus
Capella's order, the last and also the supreme one among the seven
arts. According to Remigius, Eurydice represents musical theory
while Orpheus performance, and the musician's loss of his wife sym-
bolizes the loss of his true art20). This is among half a dozen of scho-

16) See Ancius Manlius Severinus Boethius, FundarnentaL? of Music, trans. Calvin
Bower (New Haven and London 1989), 9; Manfred F. Bukofzer, SpeculativeThinking
in MedievalMusk, Speculum 12:2 (1942), 165-80; Lydia Goehr, The ImaginaryMuseum
of Musical Worh: An Essay in the Phibsophy of Musk (Oxford 1992), 133-4.
17) Chadwick (1981: 85; see n. 5).
18) Roland Barthes, Image?Musk?Text, trans. Stephen Heath (New York 1977).
19) Boethius, FundarnentaL?of Musk, 51.
20) Remigius Autissiodorensis, Remiga, AutissiodorensisCommentumin Martianum
Capelloni,ed. Cora E. Lutz, 2 vols. (Leiden 1962), 2: 310.
546 HAN-LIANG CHANG

lia in which Remigius uses Boethius on music, anachronistically, to

gloss Martianus's 'Harmony' (ibid. 334-357). With this instance, we
can now turn to Martianus Capella.
Why does Martianus put music at the end of a literary text? And
how is Harmony represented? We must look into the narrator's
invocation to Hymen in the very beginning of Book I.

Tu quern psallentem thalamis, quem matre Camena

progenitum perhibent, copula sacra deum, (1)
'Sacred principle of unity amongst the gods, on you I call;
you are said to grace weddings with your song;'

Here Hymen is addressed as copula sacra deum ('sacred principle of

unity among the gods'). Now this abstract Homeric epithet can cer-
tainly generate a long process of semiosis. For our purpose, three
major semantic areas can be identified from the word copula: (1) mar-
ital, (2) rhetorical, (3) musical. The first meaning of marital copula
is heavily invested with semantic value, especially within the context
of Roman Law, for it is marital because it is legal, or the other way
round. Here the linguistic function is purely denotative. But the
second and the
third meanings are rather formal concepts than
semantic concepts. What we have just provisionally termed rhetori-
cal copula refers to the link in both grammar (cf. 286, Copulatiuae...
coniunctiones) and logic, viz. syntax and proposition; and the musical
copula refers to the link between two musical feet, as Harmony
defines it in Book IX: etenim syzygia, id est copula, duorum pedum in unum
est ascripta conexio, qui [in] dissimiles nbi ponti esse uidentur (? syzygy or
copula is the joining of two feet that are seen to be dissimilar, into
one') (979)21). Thus, in both language and music, the term copula is
at once denotative and metalingual, perhaps more metalingual than
denotative. The same
signifier of copula, in both its phonic and
graphemic aspects, closes on three signifieds, sharing one distinctive
semantic feature of 'link'. The structural and metalingual functions
of copula enable the rhetorical and musical signifieds to serve, in
turn, as other signifiers on the higher sentential and global discur-
sive levels, and across semiotic systems.
This semiosis is in fact based on the sign function of copula. What
is a copula if it is not a sign which makes possible aliquid stat pro

21) See LeMoine (1972: 22-5; see n. 13).

 THE RISE OF SEMIOTICS AND THE LIBERAL ARTS 547

aliquo, i.e., a third sign which establishes the relation of renvoi or refer-
ral between two signs, a classical definition provided by Roman Ja-
kobson and followed by Eco and others22)? There is no exaggeration
in saying that Peirce has built his whole semiotic on that single ele-
ment of copula, who has extensively discussed the sign functions of
copula, such as equality, attribution, predication and subsumption,
not only in language, but also in logic and algebra. Earlier we said
Peirce has equated semiotic to logic. What he refers to is actually
our faculty of reasoning from a single sign to other signs through the
dynamic triangular relationship among ground, object, and inter-
pr?tant23).
Regarding the text of Martianus Capella, the question can be now
boiled down to: "How does the copula function, not only within one
system, but across several systems, such as marriage, rhetoric, and
music, or even among all the seven siblings?" The word copula, with
its variants and derivatives, like cum ita expers totius copulae censeatur, ne...
copulis interesset (40), copulam nuptialem (109), copulatiuae coniunctione (286),
recurs throughout the text, and most conspicuously in the last book
of Harmony.
Harmony is the last bridesmaid to address
the bride Philology and
the convivial Olympian company. She is charged with a more diffi-
cult task than the other six maids, and is faced with a dilemma
which they are not: Whether to perform music or to talk about
music? Or, how not to perform music when her art and only her art
entails performance? As Jupiter says, haec quippe et superum curas prae

22) Roman Jakobson, A Glanceat theDevebpmentof Semwtics,Opening report at the

First International Congress of Semiotics, Milan, June 2, 1974, in: SelectedWritings,
8 vols., vol. 7, ed. Stephen Rudy (Berlin, New York, and Amsterdam 1985), 7: 199-
218; Eco, Proposalfor a History of Semiotics,77. The French term 'renvof has been ren-
dered as 'referral'. Although Jakobson casually alludes to aliquid stat pro aliquo, his
definition is rather based on linguistic opposition and equivalence. Therefore the
semiotic function of referral should be operated intra-systemically rather than inter-
systemically. A better translation, then, would be 'reciprocity' or 'double presuppo-
sition'.
23) When we say De Nuptiis is a Menippean satire, we are copulating (i.e., using
a sign to link) a subject to a predicate, a substance to a quality (or ground), a text
to a genre, in short, one sign to another, and to still another, thus ad infinitum.When
we say, as Peirce already did in November 1866, "There is no griffin" and "A grif-
fin is a winged quadruped", we are using a copula to inquire into the actuality and
possibility of being, into knowledge and fiction. Isn't this exactly what Martianus
does with his copulasacra deum? See Peirce, Wntingsof CharlesS. Parce: A Chronobgkal
Edition, vol. I, 517.
548 HAN-LIANG CHANG

cunctis potent permulcere aethera cantibus numerisque laetifieans... ('She in-

deed, above all others, will be able to soothe the cares of the gods,
gladdening the heavens with her song and rhythms...') (899).
Following Boethius's argument, if she is to perform music, then she
knows no music which she is supposed to master. If she talks about
music, then she has to use language to interpret a system which is
non-verbal. As Harmony says: in quibus artis praecepta edissertare pro-
hibitum ('[In] the star-studded spheres,...I am forbidden to discourse
on the precepts of my art') (921). Minerva, Goddess of Wisdom, has
already forestalled this when she admonished Grammar in Book III
not to usurp Music's office by talking about rhythm and meter or
she will be torn apart (326). In short, Harmony's mission impossible
is to copulate language and music.
In a sense, her task is easy, because a solution has been already
provided. Harmony, together with the six others who have already
lectured, is a handmaid whom Mercury presents to his human bride
Philology, and Mercury or Hermes is noted, among other things, for
his miraculous language competence and thus is linked to the myth-
ic origin of hermeneutics. Their marriage, as glossed by the afore-
mentioned Remigius of Auxerre, signifies the union of speech (sermo)
and reason (ratio): PhUologjta ergo ponitur in persona sapientiae et rationis,
Mercurius in similitudine facundiae et sermonis (66). How Mercury, while
symbolizing the three language arts of trivium, can present seven arts
to Philology who is supposed to be already in possession of the four
mathematical arts of quadrivium, does not need to bother us here
although it is a tantalizing topic for logical and narratological (i.e.,
in terms of time sequence) inquiries. It follows?and this is more
intriguing?that before their marital copulation, there should be no
connection whatsoever between trivium and quadrivium.
It will be interesting to see how Harmony is presented to the audi-
ence before giving her speech. There is no way to get around musi-
ca practica, albeit embedded in the semantic universe of the narrative,
and that only. First there is a performance of instrumental music
(organikon), followed by vocal music (odikon) and then recitation
(hypokritikon), none of which is speculative. Such an order of perfor-
mance, according to Harmony in her subsequent lecture, corre-
sponds to the order of curriculum known as Exangeltdcon or hermeneu-
tikon (exposition) of a certain Lasus, who has been identified as Pin-
dar's teacher in the sixth century B.C. (939).
 THE RISE OF SEMIOTICS AND THE LIBERAL ARTS 549

dextra autem quoddam gyris multiplicibus circulatum et miris ductibus

intertextum uelut clupeum gestitabat, quod quidem suis inuicem com-
plexibus modulatum ex illis fidibus circulatis omnium modorum conci-
nentiam personabat.... uerum ille orbis non chelys nee barbiton nee
tetrachordon apparebat, sed ignota rotunditas omnium melod?as tran-
scenderat organorum. denique mox ingressa atque eiusdem orbis
sonuere concentus (909)

In her right hand Harmony bore what appeared to be a shield, circu-

lar overall, with many inner circle, the whole interwoven with remark-
able configurations. The encompassing circles of this shield were
attuned to each other, and from the circular chords there poured forth
a concord of all the modes.... No lyre or lute or tetrachord appeared
on that circular shield, yet the strains coming from that strange round-
ed form surpassed those of all musical instruments. As soon as she
entered the hall, a symphony swelled from the shield.

After her appearance with instrumental

music, Harmony sings a
hymn, tune egersimon ineffabile uirgo concludens ad Iouem reuersa aliis mo-
dulis numerisque uoce etiam associata sic coepit ('Concluding this stirring
symphony, impossible to describe [by the story-teller], Harmony
turned to Jove and lending her voice to a new melody and meter,
began the following hymn...') (911). Her songs are accompanied by
the barbiton (and probably other instruments) because when the nar-
rator resumes his report, he says: talibus Harmoniae carminibus oblectati
omnes permulsique diui, nee minor quippe ex fidibus suauitas quam uocis mo-
dulamine resultabat ('Harmony's songs delighted and soothed the spir-
its of all the gods; and the strains that poured forth from her stringed
instruments were no less sweet than the melody of her voice...') (920).
On both
occasions, Martianus lets Harmony forestall and displace
Boethius's value hierarchy by performing her art.
After singing, Harmony changes her communication by giving in

completely to language:

denique qua industria comparatum quibusue assequendum ediscen-

dumque opibus uigil cura repromittat, ut in tarn dulcem eblanditam-
que mollitiem intima mentium liquescat alfectio, loue admirante dis-
quiritur. ac tune uirgo, cum artis praecepta a se expeti examinandae
eruditionis intentione conspiceret, paulum melicis temperans exhor-
tante quoque Delio Palladeque sic coepit: (920)

Hereupon a discussion ensued, to which Jupiter listened with admira-

tion, regarding the pains and labor involved in the production of that
music... Then, when Harmony perceived that those present were seek-
 550 HAN-LIANG CHANG

ing the precepts of her art by way of putting her learning to test,
refraining somewhat from songs... she thus began her discourse.

First she
takes note, in quibus artis praecepta edissertare prohibitum ('[In
heaven] I am forbidden to discourse on the precepts of my art...')
(921). What follows is a detailed exposition of the discipline she re-
presents, which concludes the wedding pageantry within the frame.
One could
say there are two roles played by Harmony, which are
indicated by two discursive registers, first as a performer, a cantor,
dramatized by the narrator Martianus, and then as a theorist, a mun-
cus, whose speech is reported. In the first case, she is the subject of
the enunciated; in the second, the subject of enunciation. Whether
enunciated or enunciating, she and her music are encoded in the
Latin language of Martianus Capella.
Harmony's experience raises the problematics of language's legit-
imacy, power and limits in representing music. This can be glimpsed
by Grammar's and Harmony's rivalry for the right to interpreting
rhythm and
tempus (326, 971). The dispute can be settled only by a
trans-systemic tertium which is a semiotics in itself, like the afore-
mentioned semiotics of copula provided that it does not reduce and
conflate all systems to symbolic logic at the expense of their textual-
ity. However, what we have here is a textual construct to be dealt
with not by semiotics of music, but by semiotics of language on
music. I shall briefly discuss the first before dwelling on the second.
Even in semiotics of music, linguistic models are most commonly
used; at any event, only language can be used in describing music,
to some musicologists no small resentment24). No doubt, this will
impose on the interpreter some limits; but given the fact that inter-
preters are homo loquens, or in Martianus' words, homo grammaticus,
whose interpretations are always model-bound, there seems very lit-
tle: one can do otherwise.
The later Barthes, after his association with the Tel Quel group in
the late sixties, has challenged the consuming tradition of Western
music criticism, where the regular mode of signification is the attri-

24) See, for example, Nicholas Ruwet, Langage,Munque, Po?sie(Paris 1972); Jean-
Jacques Nattiez, Fondamentsd'une s?miobgk de la musique(Paris 1975); Eero Tarasti,
Myth and Musk: A SemioticApproachto the Aestheticsof Myth in Music, Espeaally That of
Wagner, Sibelius and Stravinsky(Approaches to Semiotics, 51) (The Hague 1979);
Reinhard Schneider, SemiotikderMusik: Darstellungund Kritik(M?nchen 1980); Harold
S. Powers, LanguageModeh and Musical Analysis, Ethnomusicology 24 (1980), 1-60.
 THE RISE OF SEMIOTICS AND THE LIBERAL ARTS 551

bution of a rhetorical ethos to a piece of performed music (1977: 181).

That ethos is often expressed by an adjective, and the discourse based
on predication25). Barthes calls for the liquidation of this type of
music criticism because the music which one hears, which the critic
writes about, is never the music which one plays. The latter, musica
practica, depending on the way it is performed, is manual, muscular,
and bodily, and inscribed
not in the symbolic order of language.
(Compare this with Julia Kristeva's chora, "an essentially mobile and
extremely provisional articulation constituted by movements and
their ephemeral stases [and]... analogous only to vocal and kinetic
rhythm"26).) To displace the adjectival predicatory discourse of
music criticism, Barthes coins the expression 'grain of the voice' for
songs, as the 'space' where language and music encounter (1977:
181)27). Ironically, this materialistic view echos an idealistic, phono-
centric statement of Harmony: tonus est spatii magnitudo, qui ideo tonus
dictus, quia per hoc spatium ante omnes prima uox quae juerit extenditur ('A
tone is a magnitude of space. It is called a tone because the voice
was the first of all sounds to be 'stretched' over this space') (960).
In general, Barthes's critique is well taken; but one may question
whether this mortal coil is a sufficiently materialistic space as signi-
fier where language and music encounter, and whether, as a recent
study suggests, they evolved out of a proto-faculty which was pri-
marily musical but ceased to be so when language discovered its
double articulation28). Emile Benveniste has rightly pointed out the
two principles regarding the relationships between semiotic sys-
tems29). The first one is the principle of nonredundancy. Non-
redundancy because any sign is system-specific and semiotic systems

25) Note that the word ethos is Aristotelian, a quality or character attributed to
a tragedy or a work of art. And note further the act of attributing a quality to some-
thing or the act of predicating a subject is a logical procedure, thanks to the func-
tion of copula, which amounts to Peircian semiotic. On the sign function of music,
it seems there is much negotiation to do between Peirce and the Tel Quel members
like Barthes and Julia Kristeva.
26) Julia Kristeva, Revolutionin PoeticLanguage,trans. Leon S. Roudiez (New York
1984), 25-7.
27) See also Michael Cha?an, Musica Practica: The Social Practiceof WesternMusk
from GregorianChant to Postmodernism (London 1994).
28) Bryant G. Levman, The Genesisof Musk and Language,Ethnomusicology 36:2
(1992), 147-70.
29) Emile Benveniste, The Semwbgy of Language, trans. Genette Ashby and
Adelaide Russo, Semiotica, Special Supplement (1981), 5-23.
552 HAN-LIANG CHANG

are therefore noninterchangeable. "Two systems can have the same

sign in common without being, as a result, synonymous or redun-
dant" (Benveniste 1981: 12). We take this to be a response to the
Peircian copula and even the Jakobsonian referral.
The second
principle, which follows from the first, more concerns
our discussion here: The semiotic relationship between systems is
expressed as the relationship between interpreting system and inter-
preted system. There is only one semiotic system available, which
can interpret itself by itself, and interpret other systems. That system
is language. One of the reasons is: it is a system of signifying units
whereas other systems, such as music, are based on non-signifying
units. The relationship between language and the other systems in a
specific society, say, in Martianus and Boethius's world, is nonre-
versible. That is, every nonlinguistic system must use language as the
interpreting system, not the other way around. It is not a perfect
model and maybe a shaky link for the seven arts, but it is sufficient
to undermine the disciplinary hierarchy and to interfere with the
quadrivium's aspiration to transcendence.
Going back to the representation of Harmony quoted above, we
see that as in all literary texts, the instrumental music from her
multi-layered shield is doubly encoded, first in the primary signify-
ing system of language, and then, specifically, in the secondary sig-
nifying system of genre, in this case, an ekphrasis alluding perhaps
to the shields of Achilles and Aeneas. If anything, the muted music
cannot have an existence except in the semantic universe of the sec-
ond signifying system. The same is true with Harmony's song that
immediately follows. It is to be sure encoded both in lyrics and
melody, but only the semantic aspect of the former can be decoded
to inform the audience of the stories of, say, Endymion and Phoebe,
or Orpheus and Eurydice because there is no semantic value in
melody to transmit the story. If there were, it would be a different
kind, not known to us30). Finally, Harmony complies with Jupiter's
request to give her precepts which she was forbidden to do. What
are precepts if not language?
In a sense, Martianus Capella's text is both difficult and easy, but
certainly more interesting, to read than Boethius. As Harmony's re-
presentation demonstrates, the author has reinstated language's cen-

30) Powers (1980; see ?. 24).

 THE RISE OF SEMIOTICS AND THE LIBERAL ARTS 553

trai position in the seven arts by using it to model and interpret all
of them, including itself. This is true of his account of all the math-
ematical quadrivium. What we have is a singularly encoded text:
exclusively in language except the semiotic square, which is very dif-
ferent from Boethius on music and arithmetic where lots of number
bases and patterns, shapes and solids, and even matrices are insert-
ed into his language exposition, thus resulting in a multiply encod-
ed mixed text, whose function, though, is purely referential.
But on the other hand, Martianus Capella's singularly encoded
text is not so easy as it seems as far as the functions of language are
concerned. The text is in a different sense doubly encoded: first in
the primary signifying system of language and then in the secondary
system of literature. This makes his text difficult because he is said
to have mobilized almost all known generic conventions and stylis-
tic registers?whose appreciation will take the collaboration of a
Latinist's linguistic competence and a semiotician's analytical rigor.
Let us conclude. Like all the other Latin encyclopedists, Martianus
Capella has at his disposal a huge classical repertoire and therefore
cannot claim originality in his knowledge of the seven arts. One of
his sources of music is Aristides Quintilianus, probably the same as
Boethius. But here as elsewhere, his treatment is different from
both31), mainly because of the literariness of his language expression.
With reference to Eco's proposals for a world history of semiotics,
Martianus Capella should have a place in two categories, first as a
practitioner of literary semiotics, and then as a conscious thinker of
signs. He may not have an explicit theory of language, but he is
keenly aware of language's interference with other systems. Since all
historians constantly invent ancestors, might we do so by giving two
examples to see his possible influence on later semioticians. His dis-
cussion of propositional logic can be negotiated with Peirce (who
incidentally acknowledges Martianus Capella in his 1867 Dictionary
of Logic?a very rare reference indeed32), and his formalization of
Aristotelian categories of contradictories and contrarieties is the pro-
totype of Algirdas-Julien Greimas's semiotic square?a topic yet
unstudied but worthy our further inquiry.

TAIPEI, Taiwan, National Taiwan University

31) Chadwick (1981: 83-4; see ?. 5).

32) Peirce, Writingsof CharlesS. Peirce,vol. 2, 108.