The System of

CO ICS

Thierry Groensteen
Tronsloted by Bart Beoty ond Nick Nguyen
 CONTE NT$

www.uprt"&~State.ms.u.s

Tbe l:niversiry Press ofMississippi is a member of rhe A~sociation of Americm

University Presses.

Originally published in 1999 by Presses Universitain.>s de FraElCe as -~Y#hne rk Ia

bande dessinie
Copyright© 2007 l'resses Universitain."S de France Foreward vii
Translation and !Oreword copyright © J..007 by University Press of Mississippi
All rights re.~erved
lntroduction I
.\l[:mufactun."<i in the Unired Stares of Amcrica

Print-on-Demand Edition
Chapter One. The Spatio·Topical System 24
@)
Library ofCongrcss Gtaloging-in-Publication Data (~apter Two. Restrained Arthrology: The Sequence 103
Groensrecn, Thk:rry.
Chapter Three. General Arthrology: The Network 144
[Sysn!me de la bande dessinc!e. English]
The- system of comics I Thierry Groensreen ; tr.mslared by Ban Beaty and Nick
Nguyen. -rsr ed.
(onclusion 159
p.cm.
Indude.~ bibliogrnphical references and index. Notes 165
ISBK-13: 97S-1-578o6-925-5 (clorh : alk. paper)
ISBN-xo: k578o6-925-4 (doch: alk. paper) 1. Comic books, srrips, erc.- Index 179
History and criridsm. 2. Semiotics. I. Tide.
PN67t4.G76q 2007
74'·5'69-dcn

British Libr,uy Caraloging-in-PubJication Dam avaibbJe

 FOREWORD

Thierry Groensreen's The System of Comics (Systeme de Ia bande dessinte, Presses

Universitaires de France, 1999) contains a ground-breaking analysis of ehe 6pera-
rion of the language of comics, offering ehe most imporrant semiotic analysis of
ehe medium published to date. A rigorously argued work, The System of Comics
functions as its own best introduction. Our foreword, therefore, will serve only
to lay a basic foundation for what is to follow, and to offer some direction for
readers coming to this work wichout the author's deep knowled.ge of comics,
particularly of ehe Franco-Belgian school.
Questions of cernies form have received relatively litcle attention in English-
language scholarship, which has tended to view the medium through histori-
cal, sociological, aesthetic (literary), and thematic lenses. Notahle exceptions to
thesedominant approaches include Will Eisner's Comics and SequentialArt (1985)
and Scott McCloud's Understanding Comics: TheInvisible Art (1993), two books
authored by practicing cartoonisrs. Both of these works have offered a significant
contribution to the dialogue about ehe comics form, suggesting new avenues for
investigacion and providing a tool box of terminology that continues to be used
to this day. Nonetheless, it is fair to say that both of these contributions have
been criticized for rheir Iack of theoretical sophisrication. Moreover, each work
exists sui generis, removed from ehe scholarly traditionswich which it might best
intersect.
 viii Foreword
One of the great strengrhs of Groensteen's book is the fact that it is deeply
-r
integrated into the dominant schools of visual analysis, where it makes an impor-
tant and unique conrriburion. Originally published in rhe "Semioric Forms" col-
lecrion, The System of Comics forcefully brings rhe medium of comics into rhe
field of semiorics, or the study of signs and sign systems. Generally, semiotics
involves the production of signs; communication through signs; rhe systematic
structuring of signs into codes; the social function of signs; and, ultimately, the
meaning of signs. In short, semiotics asks not simply what signs mean but how
rhey mean.
While semiotics has traditionally been applied across the humanities in rhe
study of language, culrure, and the ans, rhe application of semiotic thoughr to
the field of comics has been relarively rare. This oversight stems, perhaps, from
the low culrural value that has historically been assigned to comics, which has
rendered it an unattractive object of study. Yet, as Groensteen demonstrates, rhis
blind spot has litde to do with the specific formal qualities of comics themselves.
lndeed, as a language rhat is composed of image sequences and, often, the inre-
grarion of text, cernies would appear to offer a wide range of possible insights,
inro the spatial and temporal operations of the image.
lt is the elaboration of rhese insighrs which grace rhe pages that follow. J;ly
approaching comics primarily as a language, Groensreen reveals enrirely new
avenues for scholarly invesrigation. Beginning with an analysis of the numet-
ous attempts to define comics as a particular medium or mode of expfession,
Groensreen finds fault with each and every proposed definirion, countering them
all with his own definition of the form. Founded on the norion of"iconic solidar-
iry," his own book-length definirion reveals, through minutely detailed analysis
of case stud.ies, that comics are ~ preponderancly visuallanguage in which text
plays a subordinate (rhough fat ftom superfluous) role.
Throughour The System ofComics, Groensteen introduces key concepts for rhe
study of comics form. The first of rhese is the spatio-topical system, in which rhe
importance ofspace and place in thecomics systern is established. Here Greensteen
demoostrares that meaning is constructed first and foremost in comics by the
specific placemenr of parreis upon the page. Processes of breakdown and page
Iayout are shown to be ceritral to the production of reading, wich aesthetic effects
generared by the panel, the gutter, the frame, and the margin proving central ro
the operative logic of comics as a system that communicates meaning. The second
key concept introduced by Groensteen isthat of arthrology, a neologism from the
Greek arthon (articularion) which deals with rhe study of the relarions berween
Foreward ix
panels, wherhet linear ("restricted arthrology") or distant ("general arthrology").
Ir is wirhin these explications of arthrology that Greensteen raises the idea of
braiding wirhin comics: rhe way parreis (more specifically, rhe images in the pan-
els) can be linked in series (continuous or discontinuous) through non-narrative
correspondences, be it iconic or other means. Whether the relation between the
panels is linear rhrough a sequence or distant wirhin a network, Groensteen's
approach moves beyond the descriptive to provide important and useful tools for
analyzing rhe specific formal functioning of comics as a system that speaks by and
rhrough images.
Iftherewill be a Iimitation regarding The System ofComics for an English-reading
audience it will necessarily stem from a Iack of familiarity. Where Greensteen
takes the time and space to outline detailed readings ofindividual works or pages,
as is ehe case wich works by Tardi, Baudoin, Cuvelier, Yslaire, Mufi.oz, Geerts,
and many orhers, readers will find rhat his merhod can provide fascinating and
illuminating revelations. However, readers for whom the preceding names are
unfamiliar may, unforrunately, find that some of Groensteen's nuance will slide
by rheir attention. To call E. P. Jacobs a "wordy" cartoonist is one thing, but
for readers who have not been raised on a diet of Blake and Mortimer albums,
rhe specificiry of rhis off-hand· commenr may weil be lost. Groensreen's refer-
ences throughout The System o/C:omics are remarkably hererogeneous, ranging
from avant-garde comics srylists to cherished creators of children's comics, from
artists associated with superheroes to those firmly rooted in the Franco-Belgian
adventure tradirions. The breadth of Groensteen's understanding of the comics
medium and its rich history points to the greatest sttengrhs of this book, which
ulrimarely challenge readers to keep pace.
Of course, matehing rhe author's pace may be simpler said than clone.
Groensteen has sperrt a liferime studying comics, and he has accomplished more
in rhis field rhan most scholars could dream of. The formet director of the comics
museuro in Angouleme, France, Greensteen has edited two of the most impor-
tant magazines dedicared to comics that have appeared anywhere in the world:
Les Cahiers de Ia bande dessinie and 9e Art. Furrher, he is the author of more than
a dozen books on comics, including works on Alix, Tardi, Herge, and manga.
He has edited an even greater nurober of books, ranging from essay collections
ro art caralogues. Furrher, as a publisher ofEdirions de !'An 02, he has fadlirared
the release of some of the most important comics art currently being published.
This is ro say nothing ofhis own seenarios for published albums and hisextensive
writing on subjects other than comics.
x Foreword

Thierry Groensreen is nor only the most prolilic scholar on rhe subject of com-
ics, he is indisputably one of the best. The System of Comics is his chefd'oeuvre,
his masterpiece, finally available to readers in this English edition. Wehave Iinie
doubt that this work will once again inspire new investigations into the field of
comics, raise new quesrions, incite newdebares, and open new doors for approach-
ing rhis lirrle-undersrood an form rhar we know as comics. The System of COMICS
Bart Beaty and Nick Nguyen
March 2006
 INTRODUCTION

lnvemor of "srories in etchings" at ehe end of ehe 182os, the Genevan Rodolphe
Töpfler (1799-1846) initiated the rheoriurion of this new form of storytelling.
For ehe reader at ehe end of ehe twenriech century, ehe first "defense and illusrra-
tion" of comics,I his Essai de physiognomonie (1845), opens stimulating perspectives
for a refiection on an an which, in the interverring period, has contributed in a
decisive manner to the shaping of the modern imagination, rhereby confirming
ehe intuitions of ehe genial precursor.
Since ehis initial thunderclap, it is rarely noted ehat pracrice has become
divorced from eheory. The works that have contributed to ehe understanding of
the comics phenomenon are extremely limited in number, and ehe relative Iegiti-
mation of the "ninth an" in France has not actually led to their multiplication.
Myopie scholarship, nostalgia, and idolatry have structured ehe discourses around
comics for abour ehree decades. All too often ehe history of the medium takes the
form of an egalitarian chronicle where masterpieces and less glorious works are
treated as equivalents, while, at the same time, rhe artists who "sell" are continu-
ally ehe object of fetishistic celebrations in which critical analysis has little place.

TOWARD A NEW SEMIOLOGY OF COMICS

As rare as rhey have been, the milestones of think.ing about comics neverthe-
less demoostrate an evolution in the approach to the subject. Pierre Fresnault-
Deruelle-who, wiehin ehe French universities, was, for a long time, alone in his
interest--distinguishes four successive layers in the critical discourse:

the archeological age of ehe 196os, where nostalgic authors exhumed readings
from eheir childhoods (Lacassin 1971)
 2 Introdurnon

the sociohistorical and philosophic age of the 197os, where the crirics esrablished
the texts in rheir variants, reconstituted the relationships, etc. (Le Gal!o 96 ;
1 7
I lntroduction 3

elements rhat have a proper meaning," ro use Christian Metz's phrase.• I hold
Kunzle 1973) that this merhod cannot bring forth that which is truly specific about the lan-
gnage of comics. .
the structuralist age (Fresnault-Deruelle 1972, 1977; Gubern 1972) The second idea is that comics are essentially a mixture of text and Images, a
specific combination of linguistic and visual codes, a meeting place berween ~o
rhe semiotic and psychoanalytic age (Rey 1978; Apostolides 1984; Tisseron 98 ,
1987)' 1 5 "subjects of expression" (in rhe sense of the Iinguist Louis Hjelmslev). Agairrst
rhis conception, I intend ro demonstrate the primacy of rhe image and, there-

~ subscribe grosso modo to

fore, rhe necessity ro accord a theoretical precedence to that which, provisionally,
this periodi>ation, but it is still neoessary to qualifY
I designate under the generic term of"visual codes."
It. Of the four tendencies, none has been torally abandoned; rhey continue ro
I will begin by explaining myself on these rwo points.
coexis: rather like divergent, or parallel, roads affered to rhe invesrigator, not
exclus1ve from others (in panicular, rhematic criricism and genre studies: humor,.
fantasy, wesrern, etc.). 'Whar interests me more is thar Pierre Fresnaulr-Deruelle
has marked the recenr emergence of a "fifth stratum," thar of a "neo-semioric
The useless dispule about signifying units
criticism where the accent will be placed on the poetic dimension of comics. "3 Ir
For certain researchers, all drawing-ru1d, singularly, the often wülfully schematic
seems to me that this precisely recogni2es the ambition of this book.
linework of traditional comics--can be broken down into discreet units that
Comics will be considered here as a language, thar is ro say, not as a his-
can rhen be idenrified-points, line sections, spots-as equivalents (according
torical, sociological, or economic phenomena, which it is also, bur as an origi-
ro a precise system of homology or of analogy) to those oflexemes, morphem":',
nal ensembie of productive mechanisms of meaning. This langnage will not be
and phonemes in naturallanguages. Guy Gauthier, for example, defended th!s
passed through rhe sieve of a grand constituted theory, such as strucrural analy-
option in 1976: "We postulate therefore that, in every image, it is possible to
sis or narrative semiotics. Taking inro accounr the given objecr, rhe perspecrive
isolate lines or groups of lines, spots or groups of spots, and to locate, for each
that I propose can no doubr be described as semiologic (or semiotic) in rhe
signifier rhus determined, a precise signified, itself corresponding to a patt of the
broadest sense of thar term. However, as there will be hardly any discussion of
global signified. "' The same author insisted: "The discrete units gene:ared m the
the sign in these pages-for reasons that will become clear in a moment-I situ-
drawing style of Peanuts can be compared to the units of the first artlculanon of
ate myself, in regard to semiology, on the fringes of its disciplinary orrhodoxy.
language, the image can be compared to one or more syntagms" (p. 126).
I will not forego a few short detours through the realms of rhe semantic and rhe
According ro other researchers, the pertinent units are more highly elaborated
aesthetic, turning to my adva~tage everyrhing that can conrribute inrelligibil-
and correspond ro rhe illusrrated message or to the figures--objects, characters,
ity to the medium. That is the reason that the term "neo-semiotic" appears to
body parts. In an essay entitled "Comics lesen," Ulrich Krafft distinguished four
me completely adequate to qualifY the point of view that The System of Comics
demands. kinds of patterns, respectively: character in the foreground, object in the fOre-
ground, character in the background, object in the background. Then he broke
Reading the researchers who have preceded me and, above all, of the vulgare
up the "character" into smaller and smaller signs (Anzeichen), thus categonzm~
spread _by the media and by insrruction manuals, has convinced me rhat a rheory
Donald Duck as the head wirhin the body, the eye wirhin the head, and the pup!l
of COmics must definitively renounce two current ideas that, even though inspired
wirhin the eye. 6
for the most part by the semiotic approaches produced up to this point, appear ro
Following the terminology proposed by the Groupe Mu in their Traiti du signe
me tobe obstacles to real comprehension of the object. The firstwidespread idea is
thar the study ~f c~mics, like ~hat ofevery other semiotic system, must pass rhrough
visual (terminology that I raketobe essential), the elementary units distinguished
by Krafft correspond to "sub-entities" of iconic signifiers, while those discussed
a decomposlt10n mto constitutive elementary units: the "smallest commutable
by Ganthier are of an inferior Standard, that of "marks. "'
 4 lntroduction
As we know, the simultaneous existence of similar units wirhin an image is
conrroversial. If the Groupe Mu gives credence to this thesis in supplying a gen-
r Introdurnon 5
eral and sysremaric descriprion, no doubt the most convincing ro rhis day, orher
eminent researchers have pleaded for the recognition of a semanricism specific
to the image, which makes the economy of srable unirs analogous to those of
language.' This was already Emile Benveniste's poinr of view:
The signifying relations of ortistic "longuage" are revealed w~hin ocomposition. Art is never more here
thon a porticular work of art, where the ortist freely establishes oppositions and values which he com-
mands with total sovereignty, hoving ne~er an "answer" to wait upon, nor a contradiction ta eliminate,
but only a vision ta express.... The signilicance of art never retums ta a convention identically received
between partners. Each time it must discover the terms, limitless in number and unpredictable in nature,
so as Ia be reinvented for each worlc; in short, they are inapt ta be fixed in an institution'
The image provides the example of a semioric sysrem devoid of signs, or at least
not relianr on a finished sysrem of signs. Ir is in rhis sense rhat Benveniste main-
rained that "none of the plasric arrs considered in rheir entirery can reproduce
[the] model [of language]," the language in which he needs ro resign hirnself to
see "rhe only model of a sysrem rhar can be semioric at the sarne time in irs formal
srructure and in irs functioning. "10
Although I adhere wichout reservarion to Benveniste's affirmation, I am not
trying to demonstrate the well-founded. I do not assume that the qu~tion of
existence or nonexistence of visual signs is central in ehe analysis of ehe language
of comics. I especially want to establish that ehe most important codes concern
!arger units, which are already highly elaborated. In this case, these codes govern
th~ a~iculation, in time and space, of the units that we call "panels"; they obey
cntena that are just as much visual as narrative--er, more precisely, discursive.
These two orders of preoccupation sometimes superimpose themselves ro the
point of indistincrion.
Enteringinside the frame, in order to dissect the image by counting ehe iconic
or plastic elements rhat compose the image, then studying the methods of articu-
lation for these elemenrs, supposes a profusion of concepts bur does not Iead to
any significanrly advanced theory. By this I mean that we touch upon only the
most general mechanisms, none of which is particularly weil suited to shed light
on comics. I am convinced that we will not arrive at a coherent and thought-
fu! description of the language of comics by approaching them on this Ievel of
derail and incorporating a progressive enlargement. On the contrary, we need
to approach from on high, from the Ievel of grand articu!ations. (I don't use the
rerm "articulation" in the specifi.c meaning that it has in linguistics but in ehe
sensethat it highlights the fact that every operation consists to "organize the col-
lection of,units functioning at ehe same level."n)
In the concluding pages of his essay on Les dessaus de Ia peinture, Hubett
Damisch writes: "There, where semiology is vainly exhausted updating the
'minimal units' that would allow it to deal with painting as a 'system of signs,'
painting demonstraees, in its very texture, that ehe problern demands to be taken
upside down, at the Ievel of relations berween the terms, to the Ievel, not of the
ropes, but of ehe knots." 1z On ehe surface, this position is very close to mine bue
perhaps marred by a certain ambiguity: When describing the "relations bern:een
ehe eerms" it is important to know wich precision that which ehe terms bmd.
In so doing, ehe theory of painting will make the economy of a micro-semiocic
approach much more diflicult than the theory of comics. The reason for th~s
methodical inequaliry is simple. The image in painting is unique and global; tt
cannot arouse delicate apprehension except at the price of decomposition (this
was endlessly demonstrated by Alain Jauberr in Palettes, his remarkable television
series about painting).'' On the contrary, the comics panel is fragmentary and
caught in a system of proliferaeion; it never makes up ehe totality of ehe utterari.ce
but can and must be understood as a component in a larger apparatus.
Perhaps one objects that the fact of esrablishing the image as a base unit does
not exempt an examination of ehe inferior elemenes that conseitute it. It is true
that these rwo approaches are not exclusive and that they can even complement
each other. The Groupe Mu speaks of "this constant oscillation of the theory
berween ehe micro- and macro-semiotic, ehe first exhausted in the search for
minimal srable units, ehe second challenging the existence of these in the name of
originality each time renewed of complex utterances. "14 Iris not importane forme
co challenge, alongside Benvenisre, ehe exiseence of these units. It is only a ques-
tion ofknowing whae, from ehe micro- or ehe macro-semiotic, is most useful for
the elaboration of a complete model of the language of comics. I repeat: For the
particular subject chat is comics, the operativity of ehe micro-semiotic is revealed
to be, in practice, extremely weak.
Guy Gauthier is elsewhere obliged to admit this. For one thing, he writes that,
despite irs "apparenr complication," the image "can always be reduced, some-
times, ie is crue, thanks to a work outofproportion with the results obtained," on ehe
other hand, in explaining his method allows at most the arrival of "the descrip-
tion of a code, or rather to a sub-code, since it characterizes a single areist while
 6 Introduttion
be!ng accessible to millions of readers. "I5 Despite his pretensions to scienrificiry,
this method, when it distinguishes as many codes as there are arrists, rerurns ro
r
. lntrodudion 7
stylistic analysis and not to rhe semiology of comics as such.
If the image is rhe base unir of the comics language ir can be seen ro confirm
rhar rhe live "types of dererminarions" rhar characterize rhe "visual signs" accord-
mg to Groupe Mu (the global properties, superordination, coordination, subordi-
narion, and preordination)'' all perfecrly apply ro this unit, and in a much clearer
manner than ro units of rhe inferior rank, such as, for example, rhe character.
Ir does not appear to me useflll to fetishize a priori certain codes that are more
specific to cernies rhan oth.ers. This point merirs a brief clarification. Christian
Merz has insisred in several places rhar cinemarographic language resulrs from rhe
combination of specific codes and nonspecific codes.I7 In comics rhe codes rhat
are truly specific ro rhe form are perhaps less numerous than rhey are for film (if
rhey even exisr). Thus, rhe spario-ropical code, which organizes rhe co-presence
of panels wirhin space (and which I will esrablish larer as a rheoretical founda-
tion), equally governs rhe frarning relarions of phoro-novels. Funher, this relared
medium has also adopted rhespeech balloon as a method ofinserringwriring into
rhe hearr of rhe image. At rhe end of the day, whar makes comics a language thar
cannot be confUsed with any other is, on the one hand, the simultaneaus mobiliza-
tion of the entirety of codes (visual and discursive) rhar constirure ir, and, ar rhe
same time, the facr rhar none of these codes probably belongs purely ro ir, conse-
~uen~y speci~ng themselves when they apply to panicular "subjecrs of expres-
SIOn, wh1ch 1s the drawing. Their "efficiency"' 8 finds irself norably singularized.
?'mies are ther~fore an original combination of a (or two, with writing)
subjecr(s) of expresswn, and of a collecrion of codes. This is rhe reason thar ir can
only be described in rhe terms of a system. From then on, the problern posed ro
rhe analysr is not which code io privilege; ir is ro find an access road ro rhe inre-
rior of the system that permits exploration in its totality so as to find coherence.
Pur an~ther way, rhe objecrive musr be to define the sufficiently encompassing
caregones for rhe majority, or rhe rorality, of linguisric processes and rhe observ-
able tropes in rhe field thar can be explained by rhese conceprs. In elaboraring the
conceprs of spario-ropia, anhrology, and braiding, all rhree of which draw upon
the macro-semiotic, I am obligated to realize this program.
If, at certain analyrical moments, we move to the interior of the panel in order
to concern ourselves with certain component elements, we will always do so wich
reference to the codes that, at a more elevated. level of interrogarion, derermine
these components. To give a simple example, one can see that a dose-up has no
value in irself, but as long as it is opposed to a general outline or is a part in a pro-
gression observable only if one rakes into considerarion the syntagm for'::ed by:
number of consecurive panels. In addirion, this !arge framework can also rhyme
with anorher !arge framework, and the two images thus bound are able ro occupy
opposing or symmetrical places on the page. The colors, andin a general way any
units of an iconic or plastic nature, are simultaneously informed by the neigh-
boring images and somerimes by rhe distant images. In short, rhe codes weave
themselves inside a cernies image in a specific fashion, which places the Image
in a narrative chain where the links are spread across space, in a Situation of co-
presence. The Quebecois Yves Lacroix summed up the specificity of rhe medium
vety ably in speaking of "the soul of comics, irs fundamental immobility, simul-
taneity and panopticism compels its units, orherwise known as rhe serial srarus." 19
Acting in its production or reading, the cernies image is not that of paint-
ing. The meaning of this work will be to disengage and ro analyze rhar which,
berween the unique fixed image (picture or illustration) and the animared image,
is common to fixed sequential images.
Adominantly visual narrative species
''A thousand-year-old-logocentric tradirion has rrained us to conceive a relation-
ship of rhe suzerainty of rhe verb ro rhe image," Michel Thevoz jusrifiably remi~ds
us.2o This tradition has, in realiry, produced rwo important consequences, wh1ch
arenot always as sufficienrly distinguished from each orher as rhey could be. One
belongs ro general semiology, the orher to narrarology:
The Iangue has been raken as rhe model of alllanguage.
Fictionallirerarure is considered almosr everywhere and by nearly all as rhe model
of all narrative forms. (This second consequence is panly a logical corollary
of the firsr.)
Though ir is hisrorically based, rhis last conception is nonerheless theoretically
unrenable. The fact thar wrirren Iirerarure (irself preceded, and at one time accom-
panied, by orallirerarure) preceded by several millennia the quasi-simulr":"eous
advent of cinema and modern cernies confers on it no mono poly on the pnvilege
of rights, merely a de facto anteriority. In other words, it is no Ionger possible to
 8 lntroduction
confuse narrative and literature, exposed as we have been to a range of med.ia rhat
have, more or less, recourse to the structures of the story.
The narrative genre, with all of its caregories (intrigue, diegesis, situarions,
themes, dramatic conflicrs, characters, etc.), exists in itsdf and can be analyzed
as such, as a system of thought, as a manner of appropriating the world, or as
an immemorial activity of the human soul. Ir cuts across different serniotic sys-
tems ~d can incarnate itself indifferencly in each of them (or rarher: differendy,
but Without renouncing -its particular technical nature, which is nothing other
than the art of storyrelling). I agree here, wirh Paul Ricoeur," rhat there exists a
narrative genre and several narrative species: novel, film, stage play, but also com-
ics, the photo-novel, and-why not?-also ballet and opera, without prejudging
rhose thar will be born tomorrow from rechnological progress (since comics and
cinema owe their lare birrh-relarive to Iiterature-co rechnological evolurions,
~t is to. say, for comics, the invenrion oflithography).u Naturally, every narra-
tiVe spec1es proposes to the public another exposirory model of storyrelling and
is indined to its particular competencies. Thus, as Ricoeur writes: "No mimetic
arr had gone as far in ehe representarion of thought, feelings, and discourse as has
the novel. "'' For its part, film has orher assets, and comics have theirs also, a fact
that is demonstrated by its continuing populariry after a cenrury and a half of
existence, despite the competition of cinema and of all the new images born of
whar Regis Debray calls the "videosphere."
!öpffer saw in the texr and in ehe image two equal components of comies,
wh1ch he defined from their mixed character. This point of view, which was sup-
portable at the time, is no Ionger today. Indeed, those who recognize in the verbal
an equal Status, in the economy of comics, to the image, begin from rhe principle
that writing is the vehicle of storyrelling in general. Yet the multiplicity of narra-
tive forms has rendered this posrulare obsolete.
To suppose that cernies are essentially the site of a confrontation between rhe
verbal and the iconic is, in my opinion, a theoretical counter-truth rhat Ieads ro
an impasse."' Need I be specific? Ifi plead fur the recognition of the image as pre-
eminent in Status, it is not for the reason that, except on rare occasions, in cernies
ir occupies a more imporrant space rhan that which is reserved for writing. Its
predominance·within the system atraches to what is essential ro the production
of rhe meaning that is made through ir.
Some will surely meet this assenion wirh skepricism. Since Lessing, western
thoughr has in effect clung to these two categories, "the story" and "the image,"
taken as antinomic, beginning from the distinction between space and time. The
lntroduction 9
cinematographic imagewas a time-image; it did not arouse the same theoretical
embarrassmenr as the comics image. Of the rwo great forms of storyrelling wirh
images, it is undoubredly comics that pose the most q~estions to the Iitera?'
and plastic arts. Now, the apparent irreducibiliry of the Image and the Story ~s
dialectically resolved rhrough the play of successive images and through rhe1r
coexistence, through their diegetic connections, and through their panoptic dis-
play, in which we have recognized the foundarion of rhe medium.~ we ~n see,
it is through this collaboration berween the arthrology and rhe spano-topia that
the sequential image is seen to be plainly narrative, wichout necessarily needing
any verbal help. . . .
The 196os and 1970s, it is true, witnessed "a massive transfer of lmgutsttc
notions into rhe domain of the analysis of visual arts: we rhus speak frequent!y of
pictorial utterances, of filmic syntagms, etc.," this application expanded li~guistic
concepts rhat relied on the idea that "all representation [can] not be anythmg but
coded and rhat all contemplation of a figurative representation [is] a reading."z.5
Yet this idea conrinues ro be opposed by theorists who defend a more restricrive
(dogmatic?) conception of ehe notion of narration and who refuse to extend it
ro the visual arts. Jean-Marie Schaeffer is one of the most convincing advocates
of this ]inguistic orthodoxy. I am rempted primarily to oppose to this refusal the
fact that it is manifescly counter-intuitive, that it goes against common experi-
ence: indeed, for the viewer of a film or the reader of a comic there is no doubt
that one is being rold a story! One also recalls rhat the generative process of all
0 [ these works usually begins through the creation of a scenario. But Schaeffer
argues precisely rhat "narration is not given in or by ehe im~ (whereas in _r~e
case of a verbal structure, it is given in and by rhe connecnon of phrases): It IS
at once upstream from ehe work (as a narrative program) and downstream (as a
reconstruction on the part of the spectator)." 26
There is a certain kind of sophism in this position, which admits narration
upstream and downstrearn but refuses to recognize it acting in ehe work itself!
And one must ask by what miraculous. cognitive alchemy the reader or specta-
tor can reconstruct a story if that story wasn't already contained in the work to
which he is exposed. Schaeffer's response is that the spectator extrapolares a story
"beginning from what it (rhe image) represents thanks to what is shown.":' He
continues, "to teil a story in ehe first meaning of the term does not auromattcally
imply that rhere is a narrative in the technical sense of term, that is to _say an
enunciative act assumed by the narrator." Thus, he defines the two essential fea-
tures rhat characterize narration as an enunciative act: "The specifi.city of logical
 10 lntroduclion
links that carry out the reciprocal inregration of elemenrary propositions rhanks
to the links of consecution (a and rhen band rhen c and then ...) and of causaliry
(. · · c because b because a)" and the facr rhat "narrative assertions must refer ro a
speaker; ir follows that all narration implies a narrator. "tS
. This demonstracion, once again, is symptomaric of rhe linguisric hegemony
29
m general semiotics and, therefore, of rhe mo-frequent mechanical applicarion
of dogmas of literary narratology to every orher form of storyrelling. Linguistics
always reduces the category of "story" to the aurhoriry of "narration" and does
not reco~nize rhe presence of a narraror except insofar as certain markersbelang
to a pamculat ~erballanguage. Consequenrly, it can only discredit image-based
srones as narranve forms; rhe verdicr is reached before rhe rrial begins.
Instead of concluding rhar "it is appropriate ro restriet rhe application of rhe
technical norion of 'narration' to rhe verbal domain, "30 for my parr I think: r) thar
irisurgent to revise the technical notion rhat has ceased ro be operative because it
is in flagrant Contradiction with the experience of the modern reader-viewer; and
2) that it is no less necessary to invent specific concepts to report on the extra-
linguis~i~ "l~~callinks," rhat "carry out rhe reciprocal integration of elementary
proposmons m stories in images.
In Schaeffer's argument, however, there is a point that may be retained and
whic_h is applicable to comics: This is the insistence on the active cooperation
~rov1ded by the reader. Comics is a genre founded on reticence. Not only do the
sdenr and im.mobile images Iack rhe illusionist power of the lilmic image, but
thetr connecnons, far from producing a continuiry that mirnies realiry, offer the
reader a srory that is full of holes, which appear as gaps in the meaning. jf this
double reticence recalls a "reconstruction on the part of the spectator," the story
"ro be ':'c~nstructed" is no less set in rhe images, driven by the complex play of
sequenna!Ity. Moreover, if one believes Frans:ois Dagognet, it is the role of arr
in general to manufacture "the surreal with the elliptic. "31 Every comics reader
kn~ws that, from rhe insranr where he is projected inro the liction (rhe diegetic
umverse), he forgets, up to a certain point, the fragmented character and discon-
tinuity of the enunciarion. Allow me to recycle something that I wrote elsewhere
on the particular illusionism of the narrative arr of comics:
The ponels retum nothing but the frogmeniS of the implied world in which the story unfolds, but this world
is supposed to be continuous ond homogenous, everything tronspirtng os if the reoder, hoving entered into
the world, will never ogoin leove the imoge to which he hos been ollered occess. The crossing of fromes
becomes olorgely unconscious ond mechonicol operotion, mosked by on investment (obsorption) in the
lnttoduclion 11
virtuol world postuloted by the story. The diegesis, this fontostic virtuol imoge, which comprises oll of the
ponels, tronscends them, ond is where the reoder con reside.lf. occording to Pierre Sterckx's term, Icon
build 0 nest [nidifiml in oponel, it is becouse, in retuming, eoch imoge comes to represent metonymtcolly
the totolity of this world .... the multiplicity ond spreod of these imoges, the ubiquity of the chorocters,
mokes comics truly open to ocons~tent world, os Ipersuode myse~ oll the more eosi~ thot Icon live there
thm ... 1do not ceose, in reoding, to enter within ond to exit.31
To sum up, the srory is possibly full of holes, bur it projects me inro a world rhat
is portrayed as consistent, and it is the continuity attributed to the fi.ctional world
that allows me to efforrlessly fill in rhe gaps of the narration.
A nurober 0 f comics Iovers have said very similar things. Thus Pierre Fresnault-
Deruelle: "the fascination that comics can carry out on the reader rests, among
other elements, on its capadry to make us imagine everything other than what
is acrually shown to us: there is a rusrling of voiceless signs (just like there is a
swarming of the motionless) behind rhese docilely aligned frames" ;" or Federico
Fellini: "Comics, more than film, benefirs from the collaboration of the readers:
one teils them a story that they teil to rhemselves; with theit parricular rhyrhm
and imagination, in moving forward and backward." 34 • •
In comics, as I have said before, narration passes first and prmctpally (save
for exceptions) by way of rhe images. Andre Gaudreault recalls that "for Plato
mimesis is not, contrary to what is roo frequencly suggested, a category opposed
ro diegesis but weil and truly one of the forms of it. "35 In truth, in an image-based
story, as in film or comics, each element, whether it is visual, linguistic, or aural,
.parricipates fully in the narration. Christian Merz had the upper hand and one
can't say it more clearly: "in a narrative film, everything becomes narrauve, even
the grain of rhe film or rhe rone of voice."'6 The subjects rhat follow apply equally
to comics as to cinema:
The tenninology wos principol~ fixed in reference to finguislic norrotion, in porticulor to novels. There, the
narrolive codings ore superimposed to oftrst stuge of mojor odjustmeniS, those of longuoge; it is becouse
of them thm we speok of enunciotion, since the term is linguistic. By repercussion, il need be one con
reserve "norrotion" to the lower Ievei. But the norrotive film is based on nothing, it doesn't pile up on
on equivolent of olonguoge; it is itse~. or rother it mokes everything thot it will be on the order of "lmr
guoge." At the some time thlll enunciotion is mode norrotive, norrotion tokes chorge of oll enunciotion''
1 want to contribute to clarifying in these pages the notion of stories in images,
beginning with the parricular case of comics, which I will posrulate from here
 12 lntrodudion
lntrodudion 13
on .as a predominantfy visua! n · fo
Ri -. d fi arratzve rm. It seems to me in any case that Paul Bur one also meets definitions of cernies that are Ionger and more articulared,
stocoe;::, e. ned the proper perspec~ive when he separated the aurhority of the
better conforming to the definition of a definition: "An enunciarion of attributes
ry m tts diverse concrete mantfesrarions, and siruared each on I f
th ·· t"cal r Th. . a pane o that distinguish something, that belongs in particular to the exclusion ofall others"
ieseo~~ I dequa I?'. Is pnncipled petition opens the field to compararive srud-
(Littre). These differing definitions are rerained as pertinent for the number and
an to eepenmg the semioric sysrems in their respective singularities.
the idendry of rheir attribures. Researchers have not failed ro butt heads on rhis
point, as one can see by looking at some clarifying examples.
THE IMPOSSIBLE DEFINITION3.a The work ofDavid Kunzle, The Ear!y ComicStrip, launched a series intended
to cover the entire history of comics. This first book examines the pre-Töpfferian
The definitions of comics that can be fu d . d" . . period, ftom 1450 to 1825, grouping not only anonymous popular imagery but
and also in rh . . . un m ICtlonanes and encyclopedias, also painting and engraving cycles by arrisrs such as Callot, Rubens, Greuze, and
is easy to de moredsphectahzed hterarure, are, as a general rule, unsatisfactory. Ir
un erstan r e reasons. Hogarth, to name but a few. Kunzle formulates "four conditions" under which
These definitions are of two sorrs The fi ft . these stories in images can be considered proro-comics or, if one prefers, assimi-
ess ·ar . h . rsr, o en conctse, participares in an
lated a posteriori:
enti. Ist ap~roac and Iooks to lock up some synthetic form of the "essence"
of comJcs. This enterprise is no doubr doomed to fail "f "d
i
1,: far from verifYing the Ion ass d u~e I one const ers that, Iwould propose odefinition in which o"comic slrip" of ony penod, in ony country, fulfills the following
rilism . g ume poveny of expressiOn and inrrinsic infan-
' comtcs rest on a group f di . conditions: 1I There must be o sequence of seporote imoges; 21 There must be o preponderonce of
the representation and the o coor: nanng mechanisms that participate in imoge over lext; 31 The medium in which the strip oppeo5 ond for which it wos originolly intended must
language, and that these mechanisms govern in th .
~ove:enrs numerous and disparate paramerers, of which the dynamic inte:~r be reproductive, thot is, in printed form, omoss medium; 4I The sequence must teil ostory which is both
non t es on extremely varied forms from one comic to anorher Wh .- morol ond topicol''
succ h 1 f · arever 1rs
esses on ~ e p ane o art, one musr recognize rhar any comic: .
Bill Blackbeard, another, and no less eminent, American researcher, is vio-
is necessarily (constitutionally) a sophisticated strucrure lendy opposed to rhis view. Challenging, and not withour some bad faith, each
of the conditions proposed by Kunzle, Blackbeard formulared rhe following
only actualizes certain potentialities of the medium, to the derrim f h
that are reduced or excluded ent o ot ers definirion:

Asenolly published, episodic, operrended dromotic nonolive or series of linked onecdotes obout recurrent
Consequendy, searching for rhe essence of comics is to be assured ffi d"
shorra b fu · o n mg not a idenlified chorocters, !old in successive drowings regulorly enclosing bollooned diologue or ils equivolent
ge ur a pro Sion of responses. In the brilliant essay by Alain Rey e ,· ded
Les S>pectres de Ia ba de. " h ds n 1 ond generolly minimol nonolive text41
ics is in "th ~ ' one t us rea on page I02 thar "rhe essential" of com-
e organized space that chears berween the two d" . f h
fi d th . 1mens10ns o r e These two definitions are, to my undersranding, borh unacceptable. They are
ormat an e percepnve suggesrion of the world". pag th " h
between the textual and figural values creates the ~n:':;·co:~~"·e exchange equally normative and self-inreresred, each made ro measure in order to Support

~~~: no; on th~ ~edium characterizes above all "a creative bar;!:~:::~
an narrat~VIty, not between image and texr, this last assumes nothin
but the mosr superficial aspecr of the sto " Th . g
. ry. ese are many different and fertile
s~ggesnon~ and no doubr it would not be diflicult to find in rhis single book half
a ozen ot er analogous formulas that suggest some part of the trurh.
an arbiuary slice of hisrory. For example, rhe third of Kunzle's condirions only
serves ro jusrify rhe facr rhat he chose rhe invention of printing as a starring
point for The Early Comic Strip. While Blackbeard's definition, which defends
the thesis of rhe American origin for comics, applies only to prinred comics and
is destined to dismiss rhe entire field of comics that predates the appearance of
the Yellow Kid in 1896.

I4 Introdurnon
In France, let us recall that Antoine Roux proposed a definition in six points in
La Bande dessinie peut etre !ducative (Ed. de l'Ecole 1970), a definition backhand-
edly swept aside (and, here again, in part unjustly) by Yves Fremion in L:4BC de
la BD, where one reads: "In ten years, none of these criteria, although a priori
serious, has with.srood history. "4z
The difficulry of producing a valid definition of comics, a definition that per-
mits discrimination in that which it is not but which excludes none of its his-
torical manifestations, induding its marginal Or experimental visionaries (I arn
thinking, for exarnple, of the works of Jean Teule and of Manin Vaughn-James,
where the reception can seem to be problematic), was indicated by Pierre
Couperie in 1972:
Comics would be ostory (but it is not necessorily ostory .. .) constiMed by hondmode imoges from one
or severol ortisls \rt must eliminote cinemo ond the photo-novel), fixed imoges On difference from onimo·
tion), multiple (controry to the corroon), (Dld juxtoposed (in difference from illustrotion ond engroved
novels .. .) . Butthis definilion opplies equolly weil to Trojon's Column ond the Boyeux topestry. ~
And Coupetie adds that neither the framing of images, nor the use of the bal-
loon, nor the mode of disrribution are determining criteria.
So great is the diversiry ofwhat has been claimed as comics, orwhat is claimed
today under diverse latirudes, that it has become almost impossible to retain any
definitive criteria that is universally held to be true. I want to demonstrate this for
two of the pertinent traits ofren erected as doctrinal demenrs:
rhe insertion, in rhe image, of verbal enunciarions
rhe permanence, wirhin the pa.nels, of at least one identifiable character (a crite-
rion notably insisted upon by Blackbeard)
Although used overwhelmingly, these elements must be seen to be contingent
characteristics, suffering a number of exceptions. lt fullows that they can only
produce reductive definitions.
Here, first of all, are some authors who have produced "mute" comics, that is to
say, devoid of verbal eminciations, without dialogue or the narrational text (cap-
tions). Coming from Germany, this parricular narrative form was widespread at
the end of the nineteenth century with the pantomines of Caran d'Ache, K-Hito,
or A B. Frost, to mention not a single French, Spanish, or American ar,tist. One
later finds works "without words" in every category ofcomics: the daily comic strip
lnttoduction I5
and/or independent pages (Adamson by Oscar Jacobsson [19201; The Little King by
Otto Soglow [19311; Vaterund Sohn bye. o. plauen [1934]; HenrybyCar!Anderson
[19341: Globi by J. K. Schiefe and R. Lips [1934]; Professeur Nimbus by Andre
Daix [1934]; M. Subito by Roben Velter [1935]; Max l'explorateur by Guy Bara
[19551: etc.); complete stories published in the illustrated press (here the exarnples
abound, recall only, among the successes, AlM! il est vivant by Raymond Po"ivet
[1964]; Sanguine by Philippe Caza [1976]; many episodes of Ken Parker by Milazw
and Berardi [mid-198os1: Magie Glasses by Keko [1986]; or again the skerches
of the German artist Sperzel, such as those that can be found in recent years in
U-Comix and Kowalskz); finally, in books, from Milt Gross (He DoneHer Wrong
[1930]) to Thierry Robin (La Teigne [1998]), passing through Moebius (Arzach
[1975]), Crepax (La Lanterne Magique [1979]), AnaJuan (Requiem, with Gordillo
[1985]), Avril and Petit-Roulet (Soirs de Paris [1989]), Hendrik Dargathen (Space
Dog [1993]), Albeno Breccia (Dracula, Dracul, Vlad?, bah . .. [1993]), Fabio (L'Oeil
du Chat [1995]), Lewis Trondheim (La Mouche [1995]), Anna Sommer (Remue-
menage [1996]), and Peter Kuper (TheSystem [1997]), and this Iist has no pretence
to completeness. 44
The permanence--and the present viraliry--of this tradition does not prevent
some researchers from assening that "what distinguishes a comic from a cycle of
frescoes is the facr that the wrinen words areessential to the undersranding of the
story."" An amusing derail-and indicative of his blindness-the author next
produced, in suppon of this observation, a Krazy Kat page in which the texts
were masked, without seeming to notice that, unfortunately for him, the narra-
tion, developed in eleven images, remained perfectly intelligible despite the verbal
amputation!
As for the presence of a. recurrent character, there are diverse ways to bypass
this. I will note six:
The first is radical: it is sufficient that no human being is depicted in the
I.
story; in rhis case, these works have the unique motor of a metamorphosis of
a place or of a population of objects. Examples: The Cage by Martin Vaughn-
James (1975), Interieurs by Regis Franc (1979), A Short History ofAmerica by
Roben Crumb (1979).
The second case can be considered as an attenuation of the first. Although the
2.
recurrent character is not shown, his presence is suggested "in absentia" by
the use of a verbal narration in the first person, andlor a focus of perception
assumed by the images (a practice in cinema that is known by the expression
 Introduttion 17
16 Introdurnon

works by the Bazooka Group illustrate this tendency, as weil as the five pages
"subjective camera"). Andre Juillard's contribution to the collecrive anthol-
by Crumb enticled City of the Future (1967). The first chapter of C'etait Ia
ogy Le Violon et l'archer (1990) illustrares this second case. One mighr also
guerre des tranchies by Jacques Tardi (published in f)l suivre} no. 50 [March
remember the fumous page by McCay, in Dreams ofa Rarebit Fiend, where
1982]) is not very far removed from this; its polyphonic structure attests to
rhe prora~onisr assists in his burial ar the bottom of his coffin. (The series
ehe collective nature of an ouclook (ehe absurdity of war) that is not suitable
was pubhshed fi-om 1904 to 19n, then resrarred in 1913; ehe precise dare
to personalization, and which is under pains to reduce it.
of rh~ parti~ular page is not, to my knowledge, menrioned in any edirion.)
A ne1ghbonng case is one where rhe characrer is simply hdd permanencly
Thus, two dogmatic criteria, retained for the most part in current definitions
off-screen--one can-hear him speak withour seeing him-as in Calma chicha
of comics, must be dismissed. The difficulty encountered here is not particular to
(1985), a short story by the Spanish artist Marti.
comics. It arises in almest identical terms for the most part, if not completely, in
3· :rhere is also, while present in the image, rhe characrer rhar is not physically forms of modern art, like the cinema, and for forms where the evolution over the
~dennfiable, because rhe elemenrs rhar form his identity (and, in rhe firsr course of a century has smashed the traditional definition (novel, painting, music)
msrance, his fuce) are sysrematically evaded. The book Capets' bazaar by into pieces. For example, Roger Odin shows clearly that it is almost impossible
F:an<;oiS Mutterer and Marnne Van (1983) rises to rhis challenge. A slightly to express a definition of cinema that also applies to animared films and to all
differentexample would be Un jlip coca by Edmond Baudoin (1984), where the forms of experimental or "widened" cinema. The aporia that the semiotician
rhe fearures of rhe heroine are not revealed ro rhe reader except in ehe last
necessarily unblocks is thus described:
rh_ree pages of rhe book. (She is, until rhat point, depicred fi-om rhe back or
With her face covered by her hair.) By what nght da we exdude from cinemo these produdions when their authors present them explicitly
The "stabiliry" of rhe character can also be given a pounding by inces-
as 'films"? The fuct thot these produdions do not enter into our definition of the "cinemo; is thot o
sufficient juslificonon for this exdusion? lf not, must we revise our definition of cinemo in a more genero~
4·
sant mutarions of rhe corporeal envelope or by the graphic treatment rhar
IS rese~ed for him. An experimental book such as john et Betty by Didier izable manner in order to integrote these counter-exomples? But if so, where do westop this generolizo-
Eber~n~ (1985) proposed an approximation of this pracrice. Rerie Perillon
tion: otthe obsence ofthefilm? Atthe obsence ofthe screen? Atthe absence of the projector? Won't we
46
used It m a humorous mode in depicring the "head of directory enquiries of ornve ot a sort of definition thot teils us nothing obout its object?
Ter~ · · · o~e of these unsrable dass B14 muranrs, which constantly change
Roger Odin suggests that it is necessary to surpass this immanent approach
rhe1r heads (Btenvenue aux terriens [1982], p. 25).
to cinema in order to take into account its social uses. No Ionger considering the
5· The character as a recognizable individual dissolves when all the charac- "cinematic objecr" but the "cinematic field," he concludes (p. 57) that "cinematic
ters. resemble each other, ruining the very idea of identiry. With"m a popu- objects are definable objects, but variable objects in space and time."
Ianon such as thar of rhe Smur& • ehe phys1"cal marks of m · d"lVI"duanon
· are
extremely rare (initially reserved for Papa Smurf, Brainy Smurf, and, of
course, Smurfette). Here, the process of naming (under a form of qualified ICONIC SOLIDARITY AS A FOUNDATIONAL PRINCIPLE
ep1rher: Grouchy Smurf, Poet Smurf, Jokey Smurf, etc.) allows ehe Story to
adapt tO .rhe Stare rhat Bruno Lecigne has precisely baptized hyper-twinhood If one wishes to provide ehe basis of a reasonable definition for rhe rotality of
(hypergemellite). Cerrain stories by Francis Masse or by Florence Cestac have hisrorical manifesrations of ehe medium, and also for all of the other produc-
also come close to rhe total indifferentiation of ehe body. tions unrealized at this time but theoretically conceivable, one must recognize the
relational play of a plurality of interdependent images as ehe unique onrological
6. Moreover the cas~ of cernies where ehe "actors" renew rhemselves from panel
foundation of comics. The relationship escablished berween these images admits
to panel, each seemg his role limited to a single, unique appeara,nce. Several
 18 Introdurnon
several degrees and combines several operations, which I will distinguish later.
But their common denominarar and, therefore, the central dement of comics,
the first criteria in rhe foundational order, is iconic solidarity. I define rhis as inter-
dependent images that, participating in a series, present the double characreristic
of being separated-this specification dismisses unique endosed images wirhin
a profusion of patterns or anecdotes-and which ate plastically and semantically
over-determined by the fact of rheir coexistence in praesentia.
No doubt giving the word "comics" such an extensive meaning is not wichout
inconveniences. This is the danger noted by Pierre Couperie. From the steles,
frescoes, and the ancient Egyptian books of the dead to the predellas of medi-
eval painting, and from the Bayeux Tapestty to the polyptychs of every age, a11
the way to the pre-Colombian codex, the stations of the cross, the Emakimono
Qapanese picture scrolls), Storyboards for filmsandmodern photo-novels, there
ate probably too many of these works of art that can find refuge in this potJuck
collection. 47
Comics will encounter a problern similar to that which has long concerned the
world ofliterature. Everyone admits rhat it is not sufficient ro simply align words
in order to make a literary work, for the reason that "of all the materials that
humanity can utilize among others in the fine arts, language is perhaps the least
specific, the least dosely reserved to this end."48 Resuming a debate begun in the
time of Aristotle, Gerard Genette struggles to define the criteria of!iterarity, that
is to say the conditions by which a text can be recognized as literary. I concede in
rhe same way to the "essentialisrs" rhat ir is not sufficient to simply align images,
even interdependencly, to produce a comic. Many orher condirions can be legiri-
mately debated, which would touch in priority, initially the "nature" of these
images (their substance, rhei~ mode of production, their formal characreristics),
followed by their mode(s) of articulation, eventually even the published form
that they take, their distribution and the conditions of their reception-in short,
everything thar inscribes ehern in ehe specific process of communicarion.49 Bur ir
is improbable that unanimity will be reached on any of these conditions.
In realiry, research on rhe essence of comics is not quire on ehe same order as
thar of a definirion of lirerarity. The poim is, in rhe second case, ro separate ehe
literary discourse frorri all the other forms of discourse, starring with day-to-day
language. Literature is characterized by "a rupture wich the ordinary regime of
the language." The clearly posed question from then on is ro define "that which
makes a verbal message a work of att," according to the formulation of Roman
Jakobson recalled by Genette. For the latter, the rupture can be analyzed in terms
Introdurnon 19
offiction (in sofaras a work of ficrion develops in ehe reader an "aesthetic attirude"
and a relative "disinterest" with regard to the real world), or perhaps in terms of
diction, that is to say by the observation of formal traits that are "facts of style."
This opposition stretches to coincide with the division of the field of literaeure
into "two great types: on the one hand fiction (drarnatic or narrative), on the other
lyric poetty, more and more often designated by the term poetty all told. "''
Comics rest on a device that is not known from familiar usage. Ir is not noted
that evetything can be expressed by this means~en if the practice of comics
is, technically and financially speaking, available to everyone, as is confirmed by
ehe aptitude of those children who devote themselves to it. One cannot help but
compare ir wich other forms of creation (those, norably, rhat we have enun:-er-
ated above) that panicipate with complete rights in the domains of att or ficnon.
Since comics are not based on a parricular usage of a language, rhere is no place
to define them in terms of diction. Bur neither are they bound exdusively with
ficrional forms, since there are examples of publicity or propagandisric comics,
political and pedagogical comics, and, occasionally, comics journalism, where the
concem is ro inform or to resrify. We can also add rhat rhe proliferarion of auto-
biographical comics is a remarkable phenomenon of recent years, stemming from
Arnerica, where the works ofRobert Crumb, Art Spiegelman, and Harvey Pekar,
notably, have opened the door. This plasticity of comics, which allows them to
put in place messages of evety order and narrations other than the fictional, dem-
onstrates that befure being an art, comics areweil and truly a language.
Bur it is not necessary, at rhis srage of reßection, to push rhe concern for ehe
delimitation of ehe medium further ahead. It will be enough for us that one
cannot conceprualize comics wichout verifying rhe general rule, rhar of iconic
solidarity. The necessary, if not sufficient, condition required to speak of cernies
is that the images will.be multiple and correlated in some fashion.
This fact is empirically verified by whoever leafs through a comic book or
comics magazine. What is put on view is always a space that has been divided
up, companmentalized, a collection of juxtaposed frames, where, to eire the fine
formula of Henri Van Lier, a "mulri-framed aircrafr" sails in suspension, "in the
white nothingness of the printed page. "" A page of comics is offered at first to a
synthetic global vision, butthat cannot be satisfactory. lt demands tobe traversed,
crossed, glanced at, and analyrically deciphered. This moment-to-moment read-
ing does not take a lesser account of the totality of the panoptic field that con-
stitutes rhe page (or ehe double page), since the focal vision never ceases to be
enriched by peripheral vision.
 20 lntroduction
Irisobservable that the words for the French rerm "bandes dessim!es" (drawn
strips) irselfimplies a restrictive perception of the field rhat it is supposed ro cover.
The epithet, specifically assuming that rhe imagewill be rhe product of a draw-
ing (dessin), seems to remove a priori all recourse to the photo, to typography,
and even to painting. More seriously, the notion of the strip (bande) abusively
privileges one of the components of the medium, rhe horizontal segmene· that
sometimes constitutes a micro-story, sometimes nothing other than an ongoing
conrinuing story, or only a portion of a page. If one believes Jean-Claude Glasser,
rhe reign of this term is historically jusrified:
lt is truly in the buildings of the Agence Opern Mundi thot the expression 'bonde dessinee' wos formed
Un the 1930s], then progressive~ imposed itse~.... lt remoined to designote the doily strips ... which
exploins why it is not found in the illustroted mogozines (illustres) of the oge where the Sundoy poges
predominoted .... lt is only in the 1950s thot it ceosed to opply only to doily strips''
Bur whar was formerly nothing bur a lexical generalization has become a veri-
table impropriery. Now rhat the book [album] is, in Europe, the preponderant
vehicle for comics, ir follows rhar the page is the technical unit, market and
aesthetic reference.H
Iconic solidariry is only the necessary condition so rhat visual messages can,
in first approximation, be assimilated wirhin a comic. As a physical object, every
comic can be ·described as a collection of separate icons and interd~pendent
images. If one considers any given producrion, one quickly notices thar comics
that satisfY this minimal condition are naturally longer, but also that they do not
all obey rhe same intentions and do not mobilize rhe same mechanisms. All rheo-
retical generalizations are cog~izant of the trap of dogmatism. FarJrom wanting
to defend a school of rhoughr, an era or a Standard agairrst others, or agairr ro
prescribe any recipes, I want to force myself to note rhe diversiry of aU forms of
cernies and spare my reflecrions from any normative character.
That is why I have chosen the notion of the system, which defines an ideal, as
emblematic of rhis rellection. The comics system will be a conceprual Frame in
which all of the actualizations of the "ninth art" can find their place and be thought
of in relarion to each other, taking into account their differences and rheir com-
monalities wirhin the same medium. In rhis meaning, the norion of the system,
"an ensemble of things thar are held" (Littre), advances the fundamemal concepr
ofsolidarity.
Introduttion 21
INTRODUCING ARTHROLOGY AND THE
SPATIO-TOPIA
Ir is important now to define the exact nature of this iconic solidarity. Indeed,
comics submit the images of which they are composed to different sorts ofrela-
tions. To describe the entirety of these relations, I will use a generic term with a
very broad meaning: namely, arrhrology (from the Greek arthron: articulation)."
Every drawn image is incarnated and is displayed in a space. The fixed image,
contrary to the moving image of cinema, which Gilles Deleuze has shown is at
tb.e same time a "movement-image" and a "time-image, " 56 only exists in a single
dimension. Comics panels, situared relationally, are, necessarily, placed in relation
to space and operate on a share of space. These are the fundamental principles
of this spatial distribution that will be examined at rhe sign of the spario-topia,
a term created by gathering, while maintaining distinct, rhe concept of space
(espace) and that of place (/ieu)." The specific spaces of comics, like the word bal-
loon (bulle), rhe panel and its frame, rhe strip (the horizontal band that is the first
Ievel of arrangement for the panels), and the pagewill be successively summoned,
and their interactions analyzed.
The precedence accorded to the order of spatial and topological relations goes
agairrst most widespread opinion, which holds that, in comics, spatial organiza-
tion will be rotally pledged to rhe narrative strategies, and commanded by rhem.
The story will create or dictate, relative to its development, the number, the
dimension, and rhe disposition of panels. I believe on the contrary that, from the
insram that an author begins the comics srory thar he undertakes, he rhinks of
this story, and his work still ro be born, wirhin a given mental form with which
he must negotiate. This form is precisely the spatio-topical apparatus, one of the
keys to the system of comics, a complex of units, parameters, and functions that
it is up to us to describe. Thetaking into account of the form and the preconcep-
tion of the mode of spatial organization that will be adopted are, as I hope to
demonstrate, the preliminary conditions to every beginning, and the constraints
that never cease to inform each phase of creation. From the moment of skerching
the first panel of a comic, the author has always already raken, as for rhe behavior
of engaging with the medium, some !arge Strategieoptions (evidencly modifiable
by what follows), which concem the disrribution of spaces and rhe occupation
of places.'8 Ir will belong ro the page Iayout (mise-en-page) to specifY rhese Options
and ro provide each page with its definitive configuration.
 22 lntroduction
Bur comics is not only an art of Fragments, of scattering, of distribution;
ir is also an art of conjuncrion, of repetition, of link.ing rogerher. Wirhin rhe
spatio-ropical operarion-rhar is, wirhin rhe space rhat cernies appropriates and
develops--one can distinguish two degrees in the relarions between the images.
The elementary relarions, of the linear type, compose what we will call the
restricted arthrology. Governed by the operation of breaking down (decoupage),
they put in place the sequential syntagms, which are most often subordinated ro
the narrative ends. Iris at this Ievel that writirig rakes priority, as a complementary
fimction of narration. The other relations, translinear or distant, emerge from
generat arthrology and dedine all of the modalities of braiding (tressage). They
represent a more elaborared Ievel ofintegration between ehe narrative ßux (which
can also be called rhe narrative energy or, again, to adopr an expression from
Huben Damisch, ehe "story shurde" [navette du recit]) and ehe spatio-topical
operarion, in which ehe essential component, as Henri Van Lier has named it, is
ehe "multiframe" (multicadre).
This is not, on the one side, a comparison of spaces that will adopt the spatio-
topia, and on the other a comparison of content that comes out of arthrology.
The articulations of the comics discourse are indistinguishable from ehe content-
incamated-in-space, or, if one prefers, the spaces-invested-with-content. Thus,
the spatio-topia is a patt of arthrology, an arbitrarily detached subgroup, with
no other autonomy than that which it recognizes for irself, at a given moment,
to the heurisiic ends. Indeed, it is useful, in order to apprehend certarn Ievels of
the functionality of the comics language, to intellectually conceive of this reduc-
tion of the page as an assemblage of ftames and empty bubbles. In realiry, this
assemblage is in no way observable as such, and does not preexist, in an already
elaborated form, the final, complete version of the page.
Yet, it seems to me, the study of the system of comics must come to terms with
the spatio-topia. This precedence is not justified, I wish to add, by the chronol-
ogy of successive operations carried out in the course of a process of elaborat-
ing a comic. !t holds to the preexistence ofthat which I have called a "mental
form." A scenario destined for comics (but what follows applies a fortiori co an
improvised comic wichout a preliminary scenario-as is found, for example, in
The Airtight Garage of]erry Cornelius by Moebius) is not constructed in a purely
abstract and speculative manner. It cannot be developed except in a dialogue wich
a certain preliminary idea of the medium, ofits nature, ofits competencies and its
prescriptions. We must invent a scenario that can be incarnated in this medium
(or, sometimes, adapt itself in function to a preexisting story), to make the best
59
usage. This generaland diffuse represent;uion of comics, on which creation rests,
lntroduction 23
withholds the spatio-topical components of the system, because rhey simulrane-
ously constimte the ftamework and the base. The spatio-ropia is the point of view
that can be had on comics before thinking about any single comic, and starring
&om which it is possible tothink about a new performance of the medium.
Very quickly, when elaborating the contents, when a discourse invests the mul-
tiframe, the question of linkages and of articulations will become preponderant.
To articulate ehe frames is the process of page Iayout. Breakdown and page Iayout
areehe two fundamental operations ofarthrology, on which the braidingeventually
puts the finishing touches. The one and the other, however, help themselves
to elements that stem at fitst ftom the spatio-topia. !t is evident for the page Iay-
out, whose own role is to define a share of space, founded on rwo essential and
complementary functions of the Frame (that I will develop further along), the
separarive funcrion and the readerly function.
Thus, one can define the mode of interaction between the authoriry of the
spario-topia and the arrhrology as "dialogic" and "recursive." Ed.gar Morin, from
whom I borrowed these notions, defines them in the following way. The dialogic
is "complex associations of necessary authoriry essential for rhe exisrence of a phe-
nomenon." The phenomena of "reciprocal Feedback" can be qualified as recursive
berween rhe aurhorities rhat "are inter-regulated amongsr themselves," such as
·
rhat "the effecrs and the products are ar the same nme d prod
causaJan ""''
ucuve.
This is what I hope to establish wich respect to the complex degree of interaction
thar underlies the comics system.
From rhis perspective, the privilege &equendy accorded by critics and theo-
rericians ro certain processes assumed to be specific to comics no doubr need ro
be revised. I point, for example, in Bande dessinee et jiguration narrative (a book
which can be seen as foundarional), ro a passage asserring thar So percenr of com-
ics artists "neglect the techniques of page Iayout and breakdown that are specific
w it."'' ("It" refers to ehe language of the ninth art.) However, comics rarely
mobilize truly specific processes and techniques. On the other hand, all comics,
even those thar provide rhe simplest appearance, are parricular avatars of a sysrem
in which rhe components, and rheir inreractions, draw a complex and unpub-
lished totality. It is this system that we are now going to dissect.
I would like, at the moment of closing this introduction, t0 express my debt
ro Benoit Peerers and Thierry Smolderen, who have contribured in a decisive
fashion to the mamration of cerrain ideas formulated here. I must also thank, for
the precious commentaries that they have happily fotmulated on many of these
pages, Gilles Ciment, Pascal Lefevre, and Bernard Magne.
 The Spa~o-Tapi<al System 25

Umberto Eco to highlight how it is difficult to locate, wirhin the discourse of

comics, the stable and formalizable elementary units:

ln an icanic syntagm, such camplex cantextual cannectians intervene that it appea~ diffi:ult ta distlnguish
CHAPIER ONE the pertinent uniiS af aptianal variants amang them .... The pertinent aspeds vary: samefirnes they are
large representatians recagnizable by canvenlion; samefirnes they aresmall segments af a line, paints,
white spaces, as in the cose af a human prafile where ane paint represents the eye, a semH:irde the
eyelid. We knaw that, in analher cantext, the sametype af paint and semH:irde represent, far example,
THE SPATIO-TOPICAL SYSTEM a banana and a grape seed. 1

But the choice of the panel as a reference unit is particul3.rly necessary since

lt is ~ec~a~ that the idea and the first farm af a wark is a space, a simple place where iiS material
one is interested primarily in ehe mode of occupation of ehe specific space of

can e pace ' arranged, and nat a matenal ta be placed ar ta be arranged.
-Joseph Joubert, Carnets
(an_ ane see, indeed, that apainter _ar a paet never starts a painting ar a paem befare they have been
corrred ta see ~ ~~re ar less by therr spirit, in iiS simultaneity af principle elemeniS ... ?
comics. In its habitual configuration, the panel is presented as a portion of space
isolated by blank spaces and enclosed by a frame that insures its integriry. Thus,
whatever its contents (iconic, plastic, verbal) and ehe complexity that it eventu-
ally shows, the panel is an entiry that Ieads to general manipulations. One can
take it, for example, in order to enlarge it and create a seriegraph; one can also

-Rodolphe Topffer, Reflexionsetmenus propos d'un peintre genevoi~ chap. XXII move it.
The proofis provided when a comic, given a change in physical support (from
the daily newspaper to a book, or from an album to a pocketbook edition) is
1.0. subjected to a "reassembly": it is at that moment rhar the order of panels is com-
pletely modified. The exercise consisrs of redefining rheir respective positions. As

~u: a~temp~ ar a ::rsrematic description of the physical essence of comics will

for the images, .they are not direcdy touched, or, if they are, it is always with an
eye toward preserving the alignment of the frames, to conserve, on the newly cre-
egm r~mdt e notlOn of the multifi:ame proposed by Henri Van Lier. Although
ated page, a steady outward form. 2 The point is to make an intervention on the
a comp ete page never ceases to be a multiframe th"
h "d f . , rs rerm suggests besides &ames. Every alteration imposed on the image itself, by the fact of this interven-
t e ~ ea o a m~ltipliciry, the reduction of images to their frame, eithe; to their
tion, is of a consequential order, and can be considered as indifferent at worsr,
oucline or, especrally, to the feature that delimits it Thus it allo . .
a contenrless comic ''deansed" of it . . . ' ws us to Imagme and at best (?) as a necessary evil. When an image is re&amed, whether it is by
.h d . ' s Icomc and verbal contents, and consrructed ampuration or extension, it appears that the publishers in charge have less respect
as a fi ms e senes of supportirr fram . h .
to its sp t" . al g es-Ins ort, a comtc provisionally reduced for the internal composition (its balance, its tension, its dynamism) than for
a IO-tOplc parameters.
the coalescence of the page. The objective that is pursued is the mainrenance
of a form of geometric solidariry between the support and the panels that share
1.1-THE PREGNANCY OF THE PANEL the surface. In sum, it is notable that ehe frame dictates its law to the image.
This experiential fact reinfurces the theoretical ptivilege that must necessarily be
accorded to the panel above all other interior units.
We have alr":'dy seen why it does not seem profitable to me to a roach th
srudy of comrcs beginning from unirs smaller than the panel Hereplp . . e A!though the temporal parameters inrerest me here less rhan the sparial param-
· agam Ctte eters, I will note here that the comics panel is not the comics equivalent of the shot

26 The Spofio-Topicol System
in ehe cinematographic language. Wich regard to ehe Jengeh of time that it "rep-
resents" and condenses, its loose starus is intermediate between that of ehe shor
and that of ehe photogram, sometimes bringing together ehe one and ehe other
according to what occurs. Having imported this detail, a short detour through
ehe theory of cinema-we will make others--can be enlightening. lndeed, one
can transpose to the subject of ehe panel this remark by Chrisrian Metz: "lf the
shot is not the smallest unit of filmic signification (for a single shot may convey
several informational elements), it is at least diesmallest unit of the filmic chain. "3
And again: "One can break up a shot, one cannot reduce it." 4
lt has been said: framed, isolated by empry space (a redoubling of the frame),
and generally of small dimensions, ehe panel is easily contained by and rakes parr
in ehe sequenrial continuum. This signifies rhat at ehe perceptive and cognitive
Ievels the panel exists Ionger for the comics reader than the shot exists for a film
spectator. 'When warehing a film, "the cinema specraror does not experience ...
rhe sensation of being placed in front of a multitude of narrative utterances of
the first order that accumulate piece by piece to give birth to the second order
narrative utterance, ehe enrirecy of the filmic story. "5 The cernies reader, on ehe
contrary, experiences precisely a sensation of this rype.
These particulariries of the panel explain that it is offered., in certain cases, to
the affective investment of the reader, so that it is transformed into some sort of
fetish-as is seen, over ehe last fifteen years, in ehe vogue for seriographs based on
panel enlargements. The weil of fiction, ehe "open window on the story" (accord-
ing to a formula that has served since Alberti, but which is no less pertinenr in
the case of ehe comics image), miniature, and often wasteful in detailandin ehe
nuance of colors, ehe panel has the power to hail the reader, momentarily Frustrat-
ing ehe "passion to read" that. drives ehe images so as always to be in ehe Iead.
At certain times, at least, this power undoubtedly finds its explanation in what
Roland Barthes called ehe "obruse meaning." Beyond the informative aspect of
communication and the symbolic aspect of signification, this "third meaning"
spread.s itself to the plane of the signifier. Born from a sense of an «interrogative
reading" or of a "poetic seizure," and which clings par excellence to the "signifY-
ing accidents." Barthes specifies that "the obtuse meaning is clearly the epitome
of counter-narrative; dlsseminated, reversible, trapped in its own temporaliry, it
can establish (if followed) only an altogerher different 'script' from ehe one of
shots, sequences, and syntagms (whether technical or narrative)."' Elaborated
from a corpus borrowed from cinema (some photograms by S. M. Eisenstein),
ehe concept of ehe obtuse meaning-which must, I suggest, bring together the
The Spotio-Tapirol Sy~em 27
question of the description of the image, as weshall see it in 2.7-is not exclusive
to comics, as Barthes hirnself signals:
There ore other "orts" which combine the still (or ot leost the drowing) ond the story: these ore the photo
nov~ ond the comicstrip.l om con~nced thot these "orts," born in the lower depths of hrgh culture, pos-
sess 0 theoreficol quormcofion ond offord onew signifier (reloted to the obtuse meoning).'
The most systematic studies published up ro this point on ri:e subject ~f com-
ics generally follow an almost identical outline: they successively examme r_he
tangling of the internal relations of the panel (notably, those of ehe three majot
h -
components: t e Image,
ehe story and the ftame- but rhere are evidendy others,
• ..
since the image on its own admirs numerous parameters: reference.. compositlon,
Jighting, color, qualities of ehe line, and the writing does ehe same), then the rela-
- th th
t10ns at weave ems ves
e1 berween ehe panels • and the mode(s) of arnculauon .
of these complex units. This double-pronged approach can be fo~nd n~t~bly m
Pierre Fresnault_Deruelle; the first part ofhis book La bande dessmee, essat d analyse
sbniotique (Hachette 19?2) is broken into four chapters: "Th: Image in ,~tself
(Without Text)," "The Balloons," "Language/Drawi~gRelations, and finally_ :he
Relations Bernreen the Images"; it is confirmed also m Pierre Masson, who divides
hisLire la bande dessinte (Presses Universiraires de Lyon 1985) into two r:ts respec-
tively entided "Morphology" (on ehe "material of the image" and :he readmg ~f
a panel") and "Syntax" (on rhe page, the continuiry, and the scenar10); o~e finds lt
finally in Gase, planche, rlcit by Benolt Peeters who, respecting the prom1ses ~'f ehe
tide, suggests as ehe first chapters "The Frame Framed" (De case en :as_e) then The
Adventures of the Page," saving for the end questions of the scenanstiC ord~r.
Although 1 also begin with the panel, I want to follow a somewhat different
road. I will not be successively examining in these pages ]arger and !arger ~tter-
e finally the enrirery of the story. Instead, I Will try,
ances: t h e paneI, then the Pag • . . 1
as much as possible, not ro disassociate rhese multistage umts, but to .separate Y
analyze their different Ievels of interaction, that being ehe spatiallevel ~n the fi~st
place, and, second, the syntagrnatic Ievel of discou~, or the story (wh1ch admus
in irs rurn rwo degrees of relations: linear and rranslmear).
1.2-THE FIRST SPATIO·TOPICAL PARAMETERS
With regard to a brief digression on ehe reassembly inflicted on certain comics,
one can see that, in a ]arge measure, it is the ftame that makes the panel (ehe
 28 Thelpotio-Topicollystem
diverse funccions of ehe frame will be specified in 1.7). Ac ehe same time, ehe
page, rhis conglomeration of juxtaposed panels, is easily reduced to irs &ame-
work, which we have called ehe multiframe. The traditional schemacic represen-
tation of a comics page is nothing more rhan a grid where the comparrmenrs are
left empty, ehe "skelecon" being only ehe body of ehe evoked object. One can see
numerous examples of this in series wich reflexive characters, which show rhe an-
ist at his drawing table, exhibiting ehe so-called pages in ehe process of creation; I
am thinking notably of The Dreamer by Will Eisner and two series published in
Spirou in ehe 198os, Le plus mauvaise BD du monde and Le Gang Mazda, 8 where
pages wichout defined iconic contents can be seen.
This reduccion of ehe page to a collection of empty frames not only responds
to the imperatives of schematization. (The page being shown ar a distance, ehe
areist maintains nothing bur the minimal distinctive traits, because of ehe impos-
sibility of representing decails on so small a scale.) A convenience of drawing, ic
is also ehe most faichful representation of a general theorecicaJ· model. To draw
an ordinary mulriframe is ro consider not any particular comics page, bur com-
ics itself, to the device upon which the language is founded. These miniature
representations of comics pages are kinds of symbolic pictograms; chey give value
to their signs, they express a concept, they endose an implicit definition. Behind
their apparenc povercy, these piccograms bring us back to whac is essential abouc
comics. They plainly confirm ehe cwo fundamental incuicions thac guide me: chat
comics are Composed of interdependent images; and that these im3.ges, before
knowing any ocher kind of relation, have ehe sharing of a space as cheir first char-
acceriscic. And, remarkably, they do not say anything other chan that.
lndeed, we will see later on that this "grid" effectively incarnates comics as
a "mental form," and that ~he anist can refer to it at a very precocious stage of
creation, at a stage that has been given ehe name gridding (quadrillage). This stage
in ehe process of creation can be briefly described as ehe firsc appropriation of ehe
space that is invested in. Bur, for now, it is important to be more precise wich
regard to ehe mode of division and of ehe occupation of spaces upon which com-
ics rest, in short to describe the spatio-topical apparatus.
1t is necessary to mobilize-at least three parameters in order to precisely describe
any panel, wichout i-egard to its contents. These spatio-topical parameters are
always observable, even if ehe panel is free ftom all forms of inscription and
consists of nothing but an empty frame. The first two are geometric: they are ehe
form of ehe panel (reccangular, square, round, trapezoidal, ecc.) and its area, mea-
surable in square centimeters. This spatial dimension of ehe panel is summarized
The lpotio-Topicollystem 29
and resides in the frame. The frame is at the same time the trace and measure of
ehe space inhabited by ehe image.
The third parameter, which is the site of the panel, concerns its location on the
page and, beyond rhat, wirhin the entire work. I will come back to rhis in section
1-5- Initially, my attencion was attracced to ehe importance of ehe "site" by Jean-
Claude Raillon (in an arcicle where ehe main subject was something complecely
different).' The term emerged-admittedly, wichout ehe precise significacion chat
I atrribute to it from this point on-from a comparison between two sequences
in Herge's book, King Ottokars Sceptre (situated on pages 15 and 16 and on page
45 of ehe current edition, to which I invite ehe reader to refer), two sequences
that obey ehe "same narrative schema ... : a race, the crossing of a doorstep,
chen an unexpected discovery." Jean-Claude Raillon provides ehe following
commentary:
The observation, from the point o!view ofthe topicol poromerer, o!the ponels in queslion shows ... thot
their locotion on the poge is not compomble from one sequence to the other. The first series is distributed
overtwo poges, more prec~ely over recto ond verso of the some poge, whi~ the second offers oreadoble
denouement within the frame of the poge on which it is written.
And everything changes, of course, with regord to the nonolive suspense that orgonizes the structure
of representotion, but more certoinly, we need to be ottentive, in the ropport between the porametric
compositions that they armnge. lndeed, the relation thot allies the chomcter's movement ot the instunt
where, arriving otthe end of poge 15, he crosses overthe doorstep o( the building, und the gesture of the
render who, occomponying him, turns the poge, is remarkable. Thus one finds estoblished, between the
motenal fromework of the dmwings and the represented sequence, odouble similitude: the first associoted
with rototion, oround their respective oxes, the planes thot ore a door and osheet of paper; the second
monifests the common displocement of the chomcter und the render toword onother site. 10
This example will suffice (but I will verify it later on wich others) to testify to
ehe imporcance, for cettain panels at lease, of their "locacion on ehe page." In ehe
example of King Ottokars Sceptre, ehe panel that ends page 15 is over-determined
by a concerted coincidence between its representation and its location. 1t is com-
mon in comics that panels find themselves "automatically" reinforced by the fact
thac they occupy one of ehe places on ehe page that enjoys a natural privilege, like
ehe upper left hand corner, ehe geomerric center or ehe lower right hand corner-
and also, to a lesser degree, ehe upper right and lower left corners. Numerous
artists have assimilated this fact and made, in a more or less systematic man-
ner, key moments of the story coincide wich these initial, cenrral, and terminal
 30 The Spatio-Topical System
positions, to "rhyme" the first and last panels of a page, instimring a manner of
Iooping that we will recognize further on as an effect ofbraiding.
1.3-THE HYPERFRAME AND THE PAGE
However, rhis notion of rhe "location on rhe page" remains very approximate.
Before trying co define what must be understood by ehe site, it is necessary to
specifY a bit of ehe reference space wiehin which ehe reading is cacried out. (The
concept of the "page" is revealed, in rhis respect, to be insuffi.cient.)
Alehough often separated by ehe thin blank spaces, panels can be considered
as interdependent ftagments of a global form, something ehat is made all ehe
more clear and consistent when the exterior edges of the panels are traditionally
aligned. This form generally takes on ehe aspect of a recrangle, where the dimen-
sions are more or less homothetic to those of ehe page. The e:Xterior oucline of
this form, its perimeter, can be given the name hyperframe, in borrowing a rerm
suggested by Benoie Peeters.u Ir is possible to continue to speak of rhe drawing
board (planche) in order to designate ehe "complete" group of panels arranged on
a page (page).
The hypetftame is to ehe page what ehe frame is to ehe paneL But, in discinc-
tion to ehe panel's ftame, ehe hyperftame encloses nothing but a given homoge-
neity, and its oucline is, wich exceptions, inrermirtent. However, some authors are
pleased by reinforcing ehe hyperframe, by comaining ehe interior page wiehin a
continuous thin line, or, sometimes, by a line that is thicker than that of the pand
frames. Among others, artists such as Philippe Druillet or ehe Cosey of ehe first
Jonathan books, provided eheir pages wich an ornamental border, which reached
to elevate ehe page to ehe "dignicy" of a painting.
The notions of ehe hyperframe and ehe multiftame must not be confused.
The notion of ehe hyperftame applies itself to a single unit, which is that of ehe
page. The forms of ehe multiftame, on the oeher hand, are multiple. The strip,
ehe page, ehe double page, and ehe book are multistage multiftames, systems of
panel proliferation ehat are ·increasingly inclusive. If one wishes, it is possible co
speak of the simple ·multiframe ehat is the page, or of every unit of lesser rank
that joins several panels (the half page or ehe strip). Piling up ehe printed pages on
the recto and ehe verso, the book itself conscitutes a paged multiftame. 1t cannot
be comprehended in ehe tocality of its printed surfaces; at any place where it is
opened it can only be comemplated as a double-page spread (cf. r.6).
The Spo~o-Topicol System 31
In discinction to ehe hyperftame, ehe mulciftame does not have stable bordets,
assigned a priori. Its borders are chose of ehe entire work, whether it is an isolated
strip or a story of two hundred pages. The multiftame is ehe sum of ehe fcames
that compose a given cornic-ehat is, also, ehe sum of ehe hyperframes.
1.4-0N THE IMPORTANCE OF THE MARGIN
As we have defined it, ehe hyperframe separates ehe useable surface of ehe page
ftom its peripheral zone, or margin. (This definition accords wich ehe usual
meaning of the word ftame, since the first function of a fra~e is always to detach
a form ftom its base.) To ehe degree that it infers a cohesion between ehe different
panels that comprise it, there is a corollary that is the assimilation of the margin
as the only circumference of the page. Some emphasize that this is a reductive
definition of ehe margin, which pushes its benefits to ehe imerior of ehe page.
Indeed, ehe empty interstices ehat separate ehe panels can actually be perceived
as reticular extensions of the margin. From the hollow quadeilateral that it was,
ehese are transfurmed into a Iabyrinth. This interpreration is notably supported
by Antonio Altarriba.12 The margin, according to Altarriba's definition, is ·noth-
ing more than ehe base upon which ehe multiframe bceaks away like a form in
archipelago. In this usage, the term margin becomes synonymaus with "the part
not recovered by the base."
But ehis definition poses problems, because ehe margin, as weshall see, is not
required. to remain empty; it is not forbidden to forms of representation. This
is why I will restriet myself here to ehe moce narrow definition of ehe margin as
"part of ehe extetior support of ehe hyperframe." Thus, reduced to ehe circumfer-
ence of ehe page, ehe margin remains far from indifferent on ehe aesehetic plane,
or even ehe semantic plane. Even empcy, ehe space of ehe margin cannot be torally
neutral. And furthermore, this space is also defined by its area or, to be more pre-
cise, by its breadeh (flexible on each of ehe four sides of ehe page).
Indeed, a page is appreciated differendy according to ehe wideh of ehe margins
that surrounds it. This can be seen in a comparison of differern editions of the
same comic (for example, at Glenat, ehe first edition of Passagers du vent and ehe
re-ed.ition in the collection "CaractC:re," or, at Casterman, the current edition
and ehe deluxe edition of some of ehe Corto Maltese books, or again, at Dargaud,
ehe successive editions, accompanied by a format change, of La Quete de l'oiseau
du temps). Just as ehe interpanelblank spaces redouble ehe frame of each of ehese
 The Spotio-Topitol System 33
32 The Spotio-Topical System

pands, similarly the margin acts as a supplemental Frame with respect to the exte-
rior outline of the hyperframe (an outline partially vinual in the sense that, as we
have already noted, it genetally includes interruptions). Now-I will return to
this point-the frame of an artwork participates fully in its enunciative apparatus
and in ehe conditions of its visual reception. In autonomizing ehe work, in the
isolation of the exterior realiry, it accomplishes its closure and constirutes it as an
object of contemplation; in the case of comics, an object of reading/3
lt must also be noted that the margin is not necessarily virginal. It frequently
welcomes a title, a signarure, a page number, inscriprions in which rhe structur-
ing effect is not negligible. Most of Franquin's Idees noires are bordered, in the
upper margin, by a word game by Yvan Ddporte, and in the lower margin, by
the artist's signature, in which one finds a propensity to reproduce, in miniarure,
the principal theme of the page (cf. Idees noires, Audie, "Les albums Fluide
Glacial" 1981).
Iris easy to imagine ways to populate the margin, including through drawings,
as was formerly seen in the farnous Hauts de page by Yann and Conrad, published
in the weekly magazine Spirou beginning in 1981, and the no less famous gags by
Sergio Aragones in the pages of the monthly Mad magazine. Forthose who recall
them, these few examples suffice to demoostrate ehe diversity of relations that
marginal animation can maintain wich the page itself: this relationwas indiffer-
ent for Aragones (the gags had no necessary connection to the page that they
accompanied) was of a slightly higher order in Franquin, and was aggressive or
parodic in Yann and Conrad.
Finally, as with all representation, the margin is not pledged to whiteness. Beb
Deum and Gabrion, notably (to Iimit myself to two authors of the French expres-
sion, but one must also cite ehe British areist Dave McKean and several others),
have opted, in some of their stories, for colored margins. A book such as Rehelle
by Pierre-Yves Gabrion interests my subject all the more since it combines rwo
seldom seen principles. (I reference the first book in the series, L'Homme de Java.
I reproduce here page 26, cf. fig. r.) On the one hand, the pands are separated
Fig. 1. fro mL'Homme de Javo' 1: Rebelle (1990), by ~erre-Yves Gobrion. © Edilons Vents d'Ouest.
by black spans,4 and the entire page takes part in a hyperframe of the same
color and noticeably of the same thickness; on the other hand, the margins are,
themsdves, primed in a yellowish-brown shade. In this appararus, it can be seen
which ehe book is printed), once again becoming a color like ehe others, likely to
that the black reinforces the cohesion of the page, and that ehe ydlowish-brown,
in its difference, confirms ehe margin as a frame; but ehe major gain is no doubt combine with them wirhin the panels. .
Thus, ehe margin can, in playing wirhin diverse parameters, mf~rm :he con-
to add white to the palette of the colorist as a color in its own right. The white,
tents of the page and inflect its perception. These parameters are: ltS w!dth, the
indeed, ceases to appear as the natural color of the published book (the paper on
34 The Spatio-Topical System
drawings and ehe inscripcions chat it hosts, its color, and, finally, its degree of
auronomy, which depends on two binary factors, the closure or rhe aperture of
ehe hyperframe (concinual outline/incermittent outline) on ehe one hand, and,
on the orher, as one has seen, the idendry or the chromatic difference between the
margin and the interpanel interstices.
1.5-THE SITE
Afrer rhis consideration of the page, it is time ro revisit the panel, rhe base unit of
rhe comics system. As we have seen above (1.2), it is defined first, from rhe spatio-
ropical poinr of view, by its form and by its area. Now, under rhis double aspect,
ehe panel emers into a particular rappott wich ehe hyper&ame. Relative to. ehe
form, this rapport is of homomorphism or heteromorphism. Pur anocher way, if
it is posculaced that ehe hyper&ame isareetangle in which ehe base is ehe smallest
side (in ehe case of ehe craditional page), chere exists an important alternative: ehe
panel is itself a venical rectangle, or it assumes any other form and is opposed,
chrough chis, co ehe hyper&ame (ehe second cerm of ehe recovered alternative,
of course, of a very !arge range of possibilicies). Wich regsrd to ehe area, a pro-
portional relationship is established, a rapport chat ehe eye of ehe reader appreci-
ates with some approximation, but which the researcher can establish accurately.
Thus, a panel of8 X 23 cm will occupy, for example, close to one fifth of ehe area
of a hyperframe of 20 X 26.5cm.
From ehe topical point of view, ehe rapport chat is escablished becween ehe two
uni es is one of regionalization. The panel is a portion of ehe page and occupies,
in ehe hyperftame, a precise position. Following from chis position (central, lat-
eral, in ehe corner) and ehe general configuration of ehe page Iayout, it maincains
numerous neighboring relarions with orher contiguous panels.
The panel's spatial coordinateswiehin ehe page defines its site. The site of a
panel determines irs place in rhe reading prorocol. Indeed, it is from the respec-
tive localization of rhe different pieces of rhe mulriframe that rhe reader can
deduce ehe pachway to follow in order to pass from one panel to ehe other. At
each "step," rhe qu6ition is asked at least virtually: Where must I direct my gaze
next? Which is ehe panel chat follows, in ehe order assigned by ehe narrative? In
practice, rhe question often is not asked, because rhe response is evident righr
away. Bur one knows (and sometimes a laborious sequence of arrows regrerrably
attesrs) rhat it is not always so easy.
The Spatio-Topitol System 35
The positional coordinares of ehe panel do not scem merely from ehe parceling
0[ ehe space; chey ace also determined by a partition of time. The position of a
panel in rhe page corresponds to a particular moment in the unfolding of the story,
and also in ehe process of reading. If ehe page Iayout defines ehe spatio-topical
parameters of ehe panel (its form, its area, and its site), ic is ehe breakdown-ehe
agent of resrrained arthrology-that confers its temporal coordinares.
1.6-THE COMPOSITION OF THE DOUBLE PAGE
In principle, ehe part of ehe support (magazine or book), and ehe segment of ehe
work, that is offered to the reader's gaze is a double page. From the point of view
of perception, rhe double page constitutes a pertinent unit and merirs our atten-
tion at this rime. 15
Let us note first of all chat ehe pages on ehe left and ehe pages on ehe right
ace not equivalent wich regard to ehe utilization of sites. The panel where Tinein
races through a door, highlighced by Jean-Claude Raillon, will have its meta-
represencative'' virtue noticeably diminished if it occupies ehe lower right band
corner of a left band page, since it is incended to puncruate a right band page, and
thus ro coincide wirh the instant where rhe reader is invited ro turn the page.
However, pages situated opposite each other are dependent on a natural soli-
dariry, and predisposed co speak to each ocher. If it is possible for ehe areist to
ignore this predisposirion, there are, nonetheless, numerous ways to henefit from
it. Since, in ehe francophone marker, ehe album has supplanted ehe newspaper
as rhe dominant form of publication, authors are increasingly likely to rake inro
accounr rhe natural complexity between adjoining pages, and to conceprualize
cheir pages cwo at a time. The Iayout, ehe color, and ehe effects of interweaving
are rhe principle parameters implicated in this conception of "doubling." On this
poinc, I offer ehe cestimony ofFranc;ois Schuiten: "I always work wich an aware-
ness of ehe double page spread ... ; I pay attention to ehe general balance of ehe
pages chat will be side by side in ehe book. Some double pages that are clearer or
darker rhan others, and each time I can align the strips, I do it. "•7
The solidariry of ehe left band page and ehe right band page is never pushed
fanher chan it is in ehe case of a story cold in two pages, placed in permanent
opposirion, leaving rhe eye ro carry out a synthetic apprehension of rhe story in irs
cotality. The French areist Edmond Baudoin and ehe Spanish areist Federico Dei
Bacrio, nocably, have pulled off remarkable effects wich just such an apparacus.
 36 The Spotio-Topical System The Spotio-Topical System 37

~-11{lt;tPj :M ( (q![f/)lo
: ~ ~VI ~J/l.~- ,-~~
\ "J . ,..

' \
../
/fi"--,

RETIRI: WTitE MAIII{ {)[T_fl_ rcx.tr A'couP.

Fig.1. From Journal de Kafko (1989), by Edmond 8oudoin. © Edmond 8oudoin. F~.1. (Confinued)

Effects of symmetry, in one case, and inversion, in rhe other, are differendy moti-
vated and carried out. One can judge their results by examining figures 2 and 3,
upon which I am going to linger and comment at length.
The two pages by Baudoin (fig. 2) are an adaptation of an extract from Kafkis
JournaL'' The second reverses head to toe ehe structure of the first: ehe vertical
series of the first page (text + ehe !arger strip + text + ehe narrower strip) gives
way to rhe second page, the inverse series (rhe narrewer strip + text + the larger
strip + text). The repetition of ehe image established by the flat of ehe hand
assures rhe transition berween the two pages. however the identity of the person
who is reflected in the mirror and, correlatively, the position of the secend person,
38 The Spafia-Topical System
l A
IJRILI.A
I reading. The three panels of the first page are horizontal. !t is possible to see the
fig. 3. lo Otilfa 0985), by fedenco Dcl8arno. © Fedenco Dei Borrio.
are inverted. It is the introduction of rhis mirror theme and rhe rheme of reflecrion
(absent from ·ehe original text) that not only allows but actually generates these
diverse figures.
The two pagesby Dei Barrio, entided La Orilla ("The Shore"; fig. 3),'9 sum-
marize the life of a woman in six wordless images. The passage from the first to
the second page corresponds grosso modo to the middle of this existence. The long
diagonals that draw the su~essive positions of the people on the page respond
symmetrical!y: the descending diagonal on the left hand page is "reflected" in the
ascending diagonal of the right hand page, the pair tracing a figure V. One can
note ehe reappearance (in a miniaturized form) of these two diagonals crossed in
the footprints that are left on the sand by the aging heroine and her daughter.
The directions of these two diagonals can seem paradoxical at first, in the sense
that, in descending before re-ascending, they move backward from ehe phases of
corporeal evolution over the course of a lifetime (the body grows at first, then,
wich coming age, shrinks), and those of existence itself, that of the rise toward
maturity, followed by a decline. This apparent paradox probably has no end other
than to allow the Ietter V to appear, the first Ietter in the word vida (life), which
The Spatio-Topical System 39
redoubles and summarizes the theme of the werk. Bur it is also somewhat can-
celed by the changing of ehe panel axes, of which it is possible to risk a symbolic
idea rhat, in the first case, life is lived in the mode of contemplation (panel 2)
or expenditure (panel 3), which appears inexhaustible. The horizontality is like
infinity, like carelessness. But ehe axis reverses itse!f on ehe second page, which is
made up of three vertical panels. From then on, the gaze encounters that which
the young want ro ignore. Moving forward is now going toward rhe end, and
leaving behind imprints, which are also vestiges. If the end of life has a form, it
is no Ionger rhat of an open territory, but that of a road offering nothing more
than a single trajectoty.
I suspect rhat these examples are far from exhausting rhe connections or rhe
oppositions of all sorts that, in a comic, can be prepared between the "belle page"
(on the right hand side) and the "Fausse page" (on the left). Thus, in some books,
such as Le Bandard fou by Moebius (1974) or The Rail by Claude Renard and
Fran~ois Schuiten (1982), the actual comics pages occupy only the right hand
pages, those of the left band side are invested wich "il!ustrated" pages where the
succession creates a kind of sequential counterpoint. But this particularly book-
. ish organization of material privileges distant relationships, in absentia (ehe lirik-
ages are always made between pages that are not simultaneously offered to the
gaze), which I will speak of under the tide of general arthrology.
Yet, before progressing in the analysis of the diverse ways by which panels are
,. ~~: ~iculated, it is still necessary to refine our perception of ehe constitutive spaces
the spatio-topical system and to derail the multiple functions that fill the panel
its frarne.
.7-THE FUNCTIONS OF THE FRAME
are six important functions of the frarne, which I call the function of clo-
the separative function, the rhythmic function, the structural function, ehe
o;expressi,;e function, and the readerly function. All of these functions exert their
on the contents of the panel (a voluntari!y general expression, by which I
the totality of the figurative elementswirhin the frame) and, especial!y, on
perceptive and cognitive processes of the reader. Most of ehe functions also
up a range of formal possibi!ities, allowing the frarne to ful!y participate in
specific rhetoric of each author.
 40 The Spotio-Topicol System
1.7 .1 The Function of Closure
For its first function, the frame has to dose the panel and, also, ro confer upon
ir a particular form.
In the exercise of this function, the comics frame is opposed to the cinemato-
graphic frame. This opposition is first of all technical_ The ßexibility afforded ro
comics with regard to the form of its frames, the "elasricity" of the drawn pan-
els, highlights the rigidity of the cinematogiaphic apparatus, which is practically
condemned to equip the projected image wirh a fixed and constant fOrm (even if
other systems theoretically exist-and have existed historically).
The difference, it follows, can be qualified as onrologic, and I want, for rhe
instant, to linger a bitlanger on rhis second aspect. Ifone agrees with Guy Gauthier
rhat "to choose, for a figurative image, is not only to decide what is going ro be
visible, but also what musr be concealed, "20 one musr immediarely add rhat rhe
question of choice is posed differently to the filmmaker ( ro ehe ·camera operator,
to the framer) and to the canoonisr. In cinema, rhe frame is, from rhe moment of
shooting, u the insrrumenr of an exrraction, of a deducrion. Ir cuts up a pertinerit
zone called the "field" wirhin a profilmic continuum that overßows it, drawing a
mask around material that, not being ptinted on the film, will be absent from rhe
screen, that is to say, rhe "off-screen." The frame assigns Iimits ro rhe profusion of
the represented elements, and it elects a privileged fragment.
The frame of a comics pand does not remove anyrhing; it is cont~nted to cir-
cumscribe. Ir delimies an area affered to ehe inscription of a drawing and, if need
be, to verbal Statements.
To close rhe panel is not to stop the drawing. The graphic materiality cannot
flee or flow out; no need, tPen, to Iimit it rhrough coercive means. To close the
panel is to enclose a fragmenr of space-time belanging to the diegesis, to signifY
the coherence. (To change the frame is ofren rhe equivalent, for the reader, of
causing a displacement in space, then in time--or in these rwo dimensions at rhe
same time.)
lri concrete terms, rhe frame can be ourlined before or after rhe elaborarion of
the drawing-it must rhen enclose rhe already drawn image, crimp it-but rhe
alternative has· hardly an effect, since the mental image thar inspires rhe drawing
hand is always already framed, grosso modo. With connecrion to the first implicit
frame, sponraneously meeting and wirhaut srudy, rhe effective frame rhat is finally
inscribed on the present page is generally oflirrle difference: it is at most adjusted
to the "body of the image," in the mannet of a piece of clorhing (tailoring).
The Spotio-Topitol System 41
In resorting to rracing (that can be to ink on tracing paper, as it is norably done
by Alex Varenne, or in order to transfer rhe sketch to the original, as wirh Herge
and Jacobs), cenain anists provide themselves a supplemenrary facility to adjust
the frame to the nearest millimeter.
When it transcribes a mental image, rhe panel is first of all an image without a
body. In elaborating it on the page, the anist almost necessarily begins by creating
the frame, however approximate and provisional. All rhis occurs as if rhe frame,
having structured the space, will then favor the emergence of the icon. I later give
the name of gridding to this preliminary appropriation of space.
An artist is essentially preoccupied bywhat he wanrs to pur in his image (that is,
in his frame), not bywhat he must exclude. When even, by someeffect ofdecen-
tering or of the arbitrary cut of a pattern, or again by the writing of a voice off, he
mak.es sure that we will be led to presuppose the existence off screen of an element
that has become invisible, what is not represented has never had any physical exis-
tence wirhin the story (as in the filmic off-screen or profilmic at the moment of
- shooting):u it will remain a pure construction of the spirit, a virtualiry.
In film, the operations that exist to construct or to locate a setting, to light it,
to choose the acrors, dress rhem, supervise their sraging, in shon all of the dimen-
sions of the direction, fully panicipate in the preparation of the image, and this
preparation stans well before the actual filming. As fOr the framing (choice of rhe
lens, site-and eventual movemenr--of the camera, framing), it can be conceived
at the last minute, in the momenrs that immediarely proceed the shooring. Ir is
significant that the filmmakers least indined to improvise these filming decisions
are compelled to create elaborate storyboards, that is, to draw (or to have drawn
on their instructions) each of the shots that will be filmed. In rhe economy of
the seventh an, it is the mediation of drawing that allows, better than any other
method, to preconceive the framing. :!.3
However, there is a case where this opposition berween a c~nematographic
.. frame that rejecrs as much as elecrs and a comics frame that is contenr to hast
:. or, better, to accompany (since, from rhe initial instant of conception, the frame
the icon are interdependent and consubsrantial) is blurry. lndeed, sometimes
·--~;: :the panel is notapure translarion of the mental image, a product of the imagina-
;tion; it takes up or integrates, with or wichout modifi.cation, an earlier, generally
', photog;ra[Jhi,c, document. This document, selected in the library of the author or
.. 'drawn on the spot of the diegesis, constitutes what can be called a "prographic"
~··•.;noanerh•l The cartoonist is quite free to take only, by an operation of reframing,
peninent area that will be drawn. The initial frame of the document has no

42 The Spotio-Topical System
definitive character; irs sratus is that of a simple proposition. Undoubtedly irs
existeilce automatically confers a certain coefficient of pregnancy, an advantage in
fuct on all ehe other possible framings. But this advanrage is nothing more chan
the benefit of anteriority.
The preexisting image that inspires ehe panel can also be another panel origi-
nating from rhe same hand, belanging to rhe sarne work, or even rhe same page:
in this case, rhe variant that rhe reframing inrroduces produces an analogaus
effect to what is produced in a film, a wom Or a movement of rhe camera (travel-
ing, panorarna). Iris only when the cartoonist reframes rhat he attribures to rhe
cernies frame the extracrive function that is panicular to rhe cinemarographic
frame. Again, is ir not the same thing to frame a profilmic element and to reframe
a prographic Fragment. (The profilmic opposes icself to ehe prographic just as ehe
monument opposes irself ro the document-to use a terminology dear to Michel
Foucault.)" The distinction is taken from ehe fuct that ehe prographic is always
already framed. There exists no icon that does not camp wirhin borders, which
are always more or less arbitrary. To intervene on rhis first frame is to return ro a
fi.rst enunciative gesture; it is inevitably to produce a second degree utterance, an
image of rhe image, a cirarion.
It is necesSary to specif}r that rhis intervention does not necessarily move in
rhe sense of a reduction of rhe cired icon. To reframe is not like tailoring a suit
out of a single piece of doch. The possibility also exists to enlarge ehe frame, to
add to it one (or several) portion(s)-since rhis expansion is carried oUt on one or
several side(s)-that belong to ehe virtual off-screen of ehe previous image. The
drawing fully manifesrs, in this case, its demiurgic potential: there where there
never was anything (there where ehe document is stopped), it has ehe power to
generate a depiction that, ~though not informed by a referent, will manifest, if
the cartoonist wants it, rhe same qualiries of precision and veraciry as rhe adjacent
documented parts.
This point calls for a brief digression. An important aspect of modern comics
is the mixtute ofimaginary drawings and documented drawings, fusing rhese rwo
categories almost to ehe point where they become idencical. The dichotomy is
not pereinem except in regard to its genesis (ehe reason that I have called for it in
the midst of a ·refleciion on framing as a constitutive gesture of the image). The
usage desires that comics delete every trace of its double origin, that it conceal it
behind a homogeneaus execution: at once a producrion of the imagination and
a recycling of icons from every provenance-in eminencly variable proportions.
The Hergean line, unifier par excellence, is exemplary of this natural direction.
-- --~------------
The Spotio-Topicol Syslem 43
Largely dominant (for ehe simple reason that it reinforces the credibiliry of ehe
fiction), this tendency enteetains only cerrain exceptions, abour which I do not
believe I have to extend myself here.''
Allow me ro repeat: ehe menral image comes from ehe imagination along wich
its frame, equally a mental product, while ehe prographic document necessarily
camps wiehin the incerior of a real frame. Bur ehe materialiry of the frame does
not at all guarantee that it will be preserved; on ehe concrary, ehe real frame has
a greater chance to be altered than ehe mental frame accompanying an image
conceived ex nihilo and all of one piece.
Ir is advisable, on the other hand, to be precise: when a mental image is given
birth by a cartoonist, ehe preconceived parameters of its frame are principally
its proportians and its form (precisely rhe issue that the cinema, for its part, has
always already resolved). The dimensions of ehe frame can vacy; one knows of car-
roonists who make miniscule sketches and do not create the im.age in its real size
except at ehe moment that they carry out the transfer to ehe final page. Bur these
are ehe exception: The variations of ehe frame, berween ehe mental conception,
ehe drafr(s) and ehe final execution, are genetally of a weak amplitude.
If, in ehe mental image, ehe frame and ehe contencs are immediately inter-
dependent and consubstantial, that does not mean to say that, even if they
were thought of fiese, they so remain in ehe completed image. They can later be
modified together, in order to satisfy ehe superior exigencies of ehe page layout-
which I will speak about in ehe coming pages. Indeed, ehe frame of a panel will
not be definitively concluded wichout consideration for ehe surrounding panels.
Bound ro ehe contents rhat it encloses, ehe frame is no less attached to the frames
rhat surround it.
1.7.2 The Separative Fundion
If ehe panel is equipped wich a vircual diegetic off-screen, it also possesses a physi-
eal off-screen, which is composed of ehe bordering panels." Also, it is a condition
of reading that ehe panels are physically isolaced from each other, or cognitively
isolarable, of ehe sort that they can be read separately. In this consideration, ehe
panel frame plays an analogous role to that of punctuation marks in langnage
(here comprised of ehe elementary sign that is ehe blank white space that sepa-
rates rwo words), ehese signs that divide, wirhin a continuum, the pertinent units,
thus allowing--or fucilitating-the comptehension of the text.
44 The lpotio-Topicollystem
The dominant usage similarly rests on an insistent Separation, to the Iimit of
ehe caucology. Thac which sepacates two panels is indeed nothing less than the tri-
ple frontier constituted by the fracne of the first panel, the inter-iconic blank space
that follows, and the fracne of the second paneL Undoubtedly, fracnes are so weil
perceived as an integral pact of panels that it is necessary to distinguish the one
from the other; it is initially to the interstitial space (nocably termed, depending
on the aurhor, "intericonicspace," "inrerframes," "berween images," or "gurter")z7
rhat the reader recognizes a separative virtue: Bur it would not appear like this if
it was not bordered, on both sides, by the thin lines of the fracnes of the panels.
Gutcersare not framed themselves, but all the sacne they are calibrated with preci-
sion and, it can be said, "protected" agairrst the hegemonic pressure of the image.
However, rhe ordinary apparatus of compartmentalization in ehe "multiframe"
knows of simplified versions and can be bypassed. lt is possible to distinguish
three imponanr breaks with respecr ro rhe dominant usage:
r. Sometimes, the separation is assured by a simple outline, which simultane-
ously and indistinctly belongs to the frames of each of the two adjacent
panels. This was the case in Töpffer long ago, later in a number of American
daily Strips (from Mutt and ]ejfto Krazy Kat), and it is still the case in the
majoriry of Claire Bretecher's pages.
2. Similarly, often nothing tangible separates the different elements.of the nar-
rative sequence, nothing except rhe white of the paper, or the space that the
drawing does not occupy. In Reiser, Copi or Wolinski, the narration con-
centrates itself generally only in the characters, solitary figures developed in
an empty decor or one that is minimally suggested by a few elements. The
repetition of the sacne figure suffices to signifY the passage from one "panel"
to another (if we can be allowed to use this term).
3· In the recent works of Will Eisner, -it is common that panels are neither
framed nor separated by gutters, but interpenetrate each other in an easy
fashion. A quick examination shows, however, that most of the pages are
organized around a framed panel, where the regulac form structures the
totality of the suirounding space of the page; the elements of the decor, such
as doors and windows, are themselves strongly solicited for their structuring
effect, and frequently function as frames; finally, the contrasts between the
background blacks, whites, and greys (cross-hatched) reinforces the differen-
tiation of the images. The respect of the conventions goveming the sense of
The lpofio-Topicollystem 45
reading (from top tO bottom and from left tO right) suffices from this point
to assure the efficiency of the apparatus.
In Wolinski, as in Eisner, one can say, in borrowing a distinction proposed
by Groupe Mu, that that which is missing is the artificial "border" that desig-
nates the panel as an "organic unit"; however what remains is an "oucline'" that
undoubtedly belongs to the illustration.'8
Thus, the principle of the separation of images can never be truly denied. The
spatio-topia, Iet us not forget, is a part and a condition of arthrology: one could
not connect the visual utterances if they were not distinct. The separative func-
tion is always in the work, even if the frame, which is ordinarily its privileged
instrument, finds itself deliberately dismissed.
The function of closure and the separative function are, in truth, nothing but
the same function, successively envisaged as it exerts itself on the interior space of
the fracne and toward the exterior field.
1.7 .3 The Rhythmic Function
To cut up a text is to scan it. The "text" of comics obeys a rhythm that is imposed
on it by the succession of frames-a basic heartbeat that, as is seen in music, can
be developed, nuanced, and recovered by more elaborate rhythmic effects stressed
by other "instruments" (parameters), like those of the distribution of word bal-
loons, ehe opposition of colors, or even ehe play of the graphic forms.
A superb formula from Jean-Luc Godard defined the cinema as "the art of
making music with painting." This definition applies itself equally t0 comics; at
first because its images maintain as many affinities with painting as the shifring
images of cinema; and because comics, in displaying intervals (in the same way as
persistence of vision erases the discretization of the cinematic medium) rhythmi-
cally disrributes the tale rhat is entrusted to it. To ignore speed-its images are
immobile and no voice imprints a delivery on the dialogue--does not suggest
any less of a cadenced reading, or an operation given rhythm by the crossing of
the fracnes. !ts speech in this particularity is intermittent, elliptical, jerky. Each
new panel hastens the Story and, simultaneously, holds it back The frame is the
agent of this double maneuver of progression/retenrion.
At this time, I want to registermyself against a false temptation to which, in the
past, more than one specialist has succumbed: that of establishing an automatic
 46 The Spatio-Topi10l System
eorrespondenee between the form and rhe dimensions of the frame and rhe lengrh
of the supposed aetion rhat ir enframes. One finds, notably in Pierre Masson, sev-
eral indications along these lines: thus "a frame stretehed in height ... suggests a
Suspension of rhythm, the discovery of a high point," while the "juxtaposition of
several panels of the same format" will translare into "a rapid succession of actions
or of replies. ":t9 The same author has also written that "ehe greater ehe nurober of
pands, the stronger the impression of rapidiry." All normative propositions do
'' not do justice to ehe diversity of the expressive techniques and to ehe aesthetics
I of the authors, and it would be easy to oppose a quantiry of counter-examples
(beginning with the pages by Gabrion, Baudoin, and Dd Barrio reprodueed
above). Against the dogmatism that has sullied too many theoretical struetures,
rhe comics sysrem that is outlined here is meant to contrast a resolutely pragmatic
approach, summarizable in these terms: the fi.mction of a parameter, of a unit or
a 6gure is not prejudiced to its usage and to its signification in a given contexr
(narrative, artistic, edirorial).
1.7 .4 The Structuring Fundion
Painters, photographers, and representational theorists have abundancly eom-
mented on this point. A frame, while it structures space, is a determinant element
of the eomP.,sition of the image: it informs, during all phases of exeeution, the
drawing that is elaborated wirhin it, jusras it later inf!eets its reading. GillesDeleuze
summarized an essential dimension of rhis influence: "The Frame is therefore
inseparable from two t;ndencies: rowards saturation or roward.s rarefacrion."3o
Despite the variety of possibilities that are open to it, the eomics frame has
most ofren adopted the fo.,;, that easel painting and photography have historically
granred preeminence, namely the rectangle. Guy Gauehier sees in this canoni-
cal form "a pure produet of the western technoeratie civilization, undoubtedly
in association with the general use of perspeccive, geometric rationality and rhe
imperatives ofhandling. "• Very well; but it is appropriate to add that two faetors
more specific to cernies make this form appear natural.
First: the printed. support (book or magazine) is itself reetangular, and as a
consequenee of the hyperframe of eaeh page, the panels tend to enter into a
mimetic rapporr (ofhomology, or, in Ricardou's terms, of"autorepresentation")3l
with the imposed form. In reproducing the form of the support, the image coop-
erates with it rather than denying it or confronting it.
The Spolio-Topitol System 47
Second: more easily than a circle, a diamond, a Star, a triangle, or a trap-
ezoid, the reetangle (or its regular stand-in, the square) is able to be plaeed in a
sequenee, arranged in strips. Just as one builds walls with bricks, a multiframe
is also constructed most easily by raking on recrilinear pieces that are cut on a
right angle. Cartoonists know this weH: every reeourse to a form other than the
quadrilateral, if it is allowed to shape, by its exceptional charaeter, each eleeted
pand, presents a serious ineonvenienee in obliging the neighboring panels to be
contorred in order to make space for rhe intrusion.
Thus, the panelwill most ofren be reetangular or square. As mueh as every
other, this form, and irs particular dimensions, will induce or at least inflect
cerrain choices reuehing upon ehe composition of rhe image, and subsequencly
influenee its pereeption by the reader. The enclosed spaee of the frame is always
already struetured, and it would be for the reason that a closed space already pro-
vides a geometric center, and that this center offers a naturally privileged zone to
the representation. More generally, whether rhe image will be static or dynamic,
its centering or its "deframing," ehe distribution of presences and absences, ehe
presence of a text and irs location, and its spreading on the planes, rhe viewing
angle ehosen and the eventual hollowing out of the space by perspeetive, in short
the entirery of the formal parameters that organize the image are indexed by the
form and the dimension of the frame, as mueh as by its localization on the page,
irs site. In addirion to irs inscription in a multiframe, ehe panel's frame presenrs a
second major differenee from "the canvas on whieh the old painters colleeted and
focused their environment. "33 Indeed, it is necessarily a positioned frame.
A panel is not presented as isolared. lt participates in a series (most often
sequential, or narrative) offered ro the reader. Now, in Western culture reading
respecrs an unebanging direetion, whieh moves from the left to the right. When
the comics page respeets the dassie division of generally watertight horiwntal
("srrips"), it imposes on the panels an alignment that faeilitates the sweep
ofthe gaze.
Every comics reader knows from experience that, in practice, even when rhe
-gaze fimctions like an "irremovable beam, "34 the eye's movements on the sur-
face of the page are relatively erratie and do not respeet any precise protoeol.
'Ernpiirieal observations made by reading mechanisms unfortunately bring us
: : very lircle with regarcl to this point, and the research undertaken in Grenoble by
'' Christian Alberelli with the aid of a "eye path follower"" has not yet been pub-
:. "ISn>ea. !t is eertain that the eye does not apprehend the panel frontally, as might
generally supposed, in the way that it rakes in a painting. lt slides, wirhin the
48 The Spotio-Topicol System
hyperframe, along the surface of the plane of the page; it always arrives, and in a
justified manner, from anorher point situated wirhin the plane. Ir is in this sense
that every frame is positioned: An exit is always indicated, peinring to another
series (the following panel), which in turn solicits our attention.
The cartoonist ta.kes this natural orientation into account. Let us simply recall,
although it has often been cited, rhese remarks by Herge:
The reoder must be oble to eosily follow the norrotion. There is, notobly, on absolute rule: in our country,
one reods from left to right. ... When Ishow ochorocter who is running, he generolly goes from the left
to the right, in virtue ofthissimple rule; ond then, thot corresponds to ohobit of the eye, which follows
the movement ond which Ioccentuote: from left to right, the speed oppeo~ foster thon from right to left.
I use the other direction when ochorocter retums on his footsteps. II I olwoys moke him run from right to
left, he will hove the oir, in eoch drowing, of retuming, of chosing himself''
The dominant rule-which sometimes finds more sophisticated applications
than the very simple example chosen by Herg&-is rhat ehe dynamic of the acrion
submirs to rhe imagined movement of ehe gaze.
However, sparial organizarion and orienration are posed in slightly different
terms when the page, radically breaking wich traditional prescriprions, presents
irself as a mosaic of panels in which none are rectangular. Except perhaps in
Druillet, who has made it a formal clause, such a difference from the dominant
model is generally morivated by a will to expand the expression of a key moment
in ehe story; for example, it will aim ro make the reader feel something that
shares ~he sarne sense of disequilibrium, dread, or exultation attributed to the
characters.
Thus in certain of Guido Ctepax's diffracted Iayouts, where ehe panel's frames
do not have two parallel borders and are not square, the page has been subjected
to the empire of obliques, of points, and of apparencly arbitrary cuts. Bruno
Lecigne has perfectly analyzed ehe reasoning behind rhese "destabilizing grids,"
which coincide with erotic scenes or with violence: "The page seeks to circum-
scribe ehe Iimits of pleasure through formalization. Ir must enclose (signiJY) the
inexpressible, and thereby confer it to the realiry.... Voyeur, ehe reader is equally
constrained to interiOrize with this constant laceration of space the processes of
sadism itselE""With his unstable Iayouts, baroque frames find themselves equally
common in the work of Andreas (for example, in the album Cromwell Stone, pub-
lished by Michel Deligne in 1984 and republished by Guy Deicourt in 1993). lf
they do not always escape to a certain gratuitous nature-! spoke elsewhere of
The Spatio-Topical System 49
a "hysterization of the medium"38-they most often adopt the climaxes and the
fractures of a story in ehe faneasy genre.
Thus, it can be seen, when ehe format of the frame is differentiated from the
norm, that its structuring function tends to be confused with its expressive func-
tions (cf. 1.7.5). The image must then be accommodated to the oucline where
the irregularities almost inevitably Iead, that is, a swing of the horizon, to an
amputation of some motif, in shon, a composition that is "aberrant" in some
manner with regard to representational orthodoxy. Butthisaberration is recuper-
ated semantically as a participant in a global strategy in the service of expression.
And what is looked to express there is a situation that also has no Standard, or has
an exceptional sentiment. In this sense, the adventures of the frame in Philippe
Druillet, in spiee of appearances, follow ehe logic of appropriateness berween
the base and the form in terms of mimicry or of raising the stakes: the grandilo-
quence of the compositions, which explode the traditional frame, areehe excesses
of a "cosmic" intention. How can one suggest the eternal silence of infinite space ii
i.i
in a frame of petry dimensions? ii
In light of these extreme cases, strik.ing in proponion to their exemplarity,
one can ask oneself if ehe same law will not be always and wherever verified. Will
111',
not every organization of representative space, such as that which the frame Cre-
ates and oudines at the same time, have an expressive value? From this perspec- 'I
tive it must be posrulated that pages, using no other rypes of frames than the
canonical squares and rectangles, do not correspond to a "zero degree" of spatio-
topical expression bur, on ehe contrary, express a vision of the world founded on
the notion of order, on Cartesian logic, on rationality. Here it suffices to think
about Watchmen, the apocalyptic story by Alan Moore and Dave Gibbons, and
to that very regular Iayout, to reject this hypothesis. I will show nevertheless,
with regard to this notable example, rhat in contempotary comics, because all the
configurations of pages have been authorized, the faithfulness to a "classical" Iay-
out is generally significant (which was not the case in cenain comics magazines
of the 1950s, for example, where the classicism was imposed by ehe editors on all
· of their collaborators).
The Expressive Function
Ir is becoming clear: the frame of the comics panel can connote or index the
(\:im:>ge that it encloses. It can go so far as to instruct the reader on what must be
50 The Spotio-Topicol System The Spolio·Topicol System 51

read, or even as far as to supply a reading protocol, or even an interpretation of

the panel. lndeed, if the frame and ehe image are often unired by a relationship of
transparency or redundancy, the frame can also connote a certain form of irony
or denial.
Some time ago, Michel Rio suggested ways of rhinkingabout ehe comics ftame,
being careful to indicate: "we approach this problern in srudying (always wirhout
preoccupying ourselves wich ehe signifieds of comics) some general tendencies in
the choice of organization of ftames and of ehe Iayouts. "39 This idea has already
been picked up by Benoit Peeters, who denounced it in ehe following terms:

On this point, my pe5pective will be entirely different, to be the exoct oppostte of this seemingly
insignlficont porenthesis. The only monner of escoping from formolism thot, often rightly, wos reprooched
by the semiologists, is to toke into occount these "orgonizotionol choices" in rehman with thot which
Michel Rio colls "the signlfieds of comics," thot is to soy, to not destroy this quoskngonic coherence thot
gives eoch element its reoson for being 40

The position of Peeters is also mine. But I allow myself to add that only a
reasoned description of ehe spatio-topical appararus can make the organization of
different possible choices comprehensible wiehin ehe system, and thereby supply
a view of ehe totality of the criteria that allow the appreciation of rhe pertinence
of the choices that have been made.
Regarding the relation between ehe formal organization and rhe signified, ehe
example <hat fol!ows will supply the best demonsttation. !t involves a two-page
comic created in 1980 by ehe American author Bill Griffith. Ir bears the ticle The
Plot Thickens (cf. fig. 4)" and constitutes one of the most obvious applications
to ehe comicl of ehe methods of OuLiPo (l'Ouvroir de Iitterature potentielle)-"
lndeed, ehe formal rule that orders the number, ehe size, and ehe disposition of
frames is arbitrary. The constraint is arnong the simplest: It requires thar each
new horizontal Strip contains one more panel than the preceding strip. Eleven
strips follow each other in this way: ehe first is composed of one unique panel, ehe
last counts eleven. The width of the page is a given constant, so these panels must
necessarily be increasingly reduced in size, wich the height of ehe Strips diminish-
ing as the panels beCome narrower.
In terms of this striccly formal description, ehe exercise undertaken by Bill :ruurro,wirtg c>f ehe surfaces invested by the represenration? This opposition is redis-
Griffith appears relatively futüe. But it is judged different!y when one takes note m1rer<:d elsewhere on ehe image plane, in the sense rhat, tendentially, ehe avaüable
of ehe "signifieds." These reveal that irony and paradox command this brief ~omic. is in inverse proportion tothat which is required by the representation. The
Is not ehe idea of thickening suggested in the tit!e opposed to the progressive panel, ehe largest, shows a room furnished with dozens of rables, in which
 The Spofio-Topical Systam 53
52 The Spofio-Topicol System

In the last four strips of images on the second page, on the other hand, at
the moment where the panels attain a format equal to, then smaller than, that
of a postage stamp, rhe events become numerous and tumulruous: the protage-
nist is drugged, converted to religion, makes a short stay in prison, and finally
rediscovers his Iust for life. In sum, everything that could be called the anecdotal
developments and a more or less spectacular "direction" is sacrificed to ehe driest
enunciation, to rhe mosr minimalist graphic trearment, and the smallness of rhe
&ames does not permit anything more. Whereas, on rhe conrrary, rhe storywas
launched by a uselessly [arge panel considering the small amount of information
that it contained. ·Ii
!!·
(I well understand that, considered in isolation, and independent of the
particular dynamic of rhis srory, this panel could be coherent; it is necessary to
enlarge the field in order to note the absence of other customers and to illustrate
the isolation and immobiliry of the protagonist. On the pedagogical plane it
would be profitable to demonstrate, wich regard to this comic, that the ultimate
signification of a comics panel does not reside in itself but in the rotaliry of rela-
tions in the network that it maintains with the interdependent panels; in short,
that it borrows from general arthrology.)
In comics as in film it is common that a text establishes an ironic counterpoint
to the image. The Plot Thickens shows, in a possibly unexpecred way, that rela-
tions of the same order can be carefully handled between the image and its frame
as weiL
Another constitutive parameter of the frame, to which I have already made
allusion, is primarily concerned with the expressive function: it acts upon the
Iayout, or, if one prefers, upon the physical characteristics of the oudine. Indeed,
when these. terms change between two consecutive panels, the modification
serves, in principle, to draw attention to a rupture in the level of enunciation
regarding the status of the image, and to indicate, for example, a flashbackor the
beginning of a dream sequence. In a similar case, it is possible that it falls to
the frame to supply the "directions" (mode d'emploz) for the panel, to prescribe
ehe appropriate reading regime.
Fig. 4. ((ontinueäl

a solitary character o'ccupies a single chair. Everything works ro illustrate rhe

empciness of the room and the absence of action. The accompanying text raises
the srakes: Not conrent to signal that "the cafereria ~as empty," it specifies again
"there was nothing in the paper." The paper spread out on the rable is, indeed, free
from all information.
1.7 .6 The Readerly Function
As obvious as it seems, the last of these six functions of the frame is of no less
importance than the first. I will note that a frame is always the sign of something
1,,,
 The Spofio-Topicol System 55
54 The Spotio-Topicol System

to be read. When he "meets" a frame, the reader is taken to presuppose rhat,

wirhin the perimeter thar has been drawn, there is a content to be deciphered.
The frame is always an invitation to stop and to scrurinize.
The Groupe Mu highlighted ehis point: A border delimiting a space "assumes
an important semiotic role vis-3.-vis rhis space-indeed, it designates it as homo-
geneous. The image to come rhus receives, before it is even emitted, the status of
a unit: an isolated or enunciated sign."43 This is what Louis Marin, for his pan,
expresses in a slighrly different fashion. The frame, he argues, "reserves ehe repre-
sentation ro contemplarion" and "defines ehe place of a symbolic operarion."«
This funcrion is trivial and, in the most cases, superfluous. In theory, the panel
sufficiently manifesrs its enunciative character, its Status as a link in the discursive
chain; it attracts ehe eye so much that, even if it is unframed, ehe reader pauses
there for at least an instant. Yet, the function rhat I have called "readerly" acquires
all of its significance in ehe case where a part of ehe image on the page might
appear insignificanr, because it doesn't allow enough space for the acrion or rhe
spectacle, or merges in irs immediate environment to the point where this section
risks not being seen (or, if it is noticed, of being deliberarely "skipped" by rhe
reader, who is always driven by his eagerness to know ehe rest of ehe srory).
An example will bettet allow rhis function of rhe frame to be undersrood. Let
us refer to figure 5, which reproduces a page by ehe Spanish artist Aleix Barba,
rhe first in a srory enrirledA Wint.-r Story (drawn in 1985)." Iris an inrerior scene,
where we find---<:oncealed by rhe newspaper rhat he is in rhe process of reading-
a single characrer, in rhe midst of his familiar decor. No action, no idenrifiable
proragonist, no dialogue (bur nonerheless a nurober of linguistic enunciations:
rhe lyrics of ehe song, rhe eitles of books and records, ehe rexr of rhe newspaper).
Iris difficult to imagine an e?-try into a story that would be more sluggish or more
conremplarive than what this page construcrs as a "panorama," wirh each strip
connecting itse1f precisely to its precedent, which it prolongs.
Now, it is preciselywith an introduction such as rhis one rhat it all rakes place;
it is here rhat, ro creare in the reader a desire to keep moving, ir is necessary to
arouse his interest or pique his curiosiry. It is here thar an implicit reading con-
Fig. 5. From "A Winter Stur( {1985), by Neix Sorbe. © Neix Borbo.
tract is established, because the tide and the first panels are promises of a cerrain
mood, of a reward eXchanged for the reader's atrention. Ir is imaginable wirh
whar fuciliry and what good conscience rhe ordinary comics reader skims rhrough of ehe necessiry of deraining ehe reader and conraining his gaze, Barba has the
intelligence ro help hirnself ro ehe frames, ehar is to say, in ehis case, to arbirrarily
a page like rhis, a simple glance sufficing ro reveal ehe absence of any evenrs. To
mulriply ehern, ro produce more rhan what would ask, in consideration of the
enter into this intimate story in a hurried way is to condemn oneself to miss the
essential, to taste nothing of what gives it an original qualiry. Also, conscious image, the single function of separation.
56 The Spofio-Topicol System
Indeed, chere are only chree images here, each occupying one strip of ehe page.
Three frames would have sufliced: yet chere are nine. Thus, che reader is doubly
mobilized. At ehe regional Ievel, chat of che hyperframe, che division of each srrip
into equal chirds tends to create che illusion of a sequenriallinkage, and cherefore
a narrative process. Ac che locallevel, chat of ehe panels, che gaze concained by ehe
frame is invited ro stop and to rake account of rhe information that is suggested,
iconic as much as verbal. By vircue of rhis segmentacion, ehe reader discovers,
apparently where there was nothing to read, a Profusion of morifs and references
char, if one cakes che time to examine ehern, "communicate" a Iot of chings. (Iris
evident that this apparatus operates on the modd of ehe invitation and that it has
no coercive power. Similarly, noching is able to oblige anyone to read anyching.)
In cercain pages of ehe book Blues (Kesselring 1979), Chantal Moncellier divides
large fllll-page composirions into several frames, and certain of rhese frames
enclose nothing more chan an apparendy trivial decail, such as, for example, a
coac hanger suspended from a nail (ehe page encided Clooney, not numbered).
lf ehe page was not divided up, chis coat hanger would have only contribured,
amongst the other details that compose rhe decor, ro that which it is suitable to
call an "effect of ehe real," that which, in a ficcion, arises from elements that are
apparendy not mocivaced by che dramaturgy. Bur ehe existence of a Frame makes
ehe coat hanger appear as a privileged element of chis porcion of che scene, invit-
ing the reader, first, to register irs presence (ehe inscription of a textwirhin ehe
same frame evidently contributes to fixing a time for attention ro rhis panel),
and second, to search for some reason rhar morivates this presence. The coat
hanger can no Ionger be indifferent; it is percinent-anecdotally or symbolically
peninent---or it is unusual.
Every porcion of ehe image. isolated by a frame reaches, by chac same facc, ehe
status of a complete utterance. To dedicate a frame to an element is the same as
testifying that this el~ment constitutes a specific contribution, however slim, to
ehe Story in which ic parcicipaces. This contribution is that which asks to be read.
Ofren, its evaluation can cause a problem. Bur ehe doubt chat cakes hold of ehe
reader proves again, in its way, che readerly function of che border. This would
also apply to a frame devoid of all content, a panel in which che whiceness would
be, definicely, ~ignifica~t-
One might perhaps ask which of ehe six functions attributed to che frame are
particular to the apparatus of comics, and which also characterize the operations
of ehe Frame of any kind of icon. The response will appear simple enough. The
separative and rhythmic functions cannot exist except when two conditions are
The Spatio-Topicol System 57
joined rogecher: chey suppose a pluraliry of frames in ehe situation of co-presence
wiehin ehe same supporr; they require in addition thac ehe succession of these
frames compose, perhaps not necessarily a story, but at least an articulated dis-
course. & for ehe readerly function, it goes beyond the semiotic function inher-
ent in framing because, since ehe panel contributes to a sequential discourse, its
Frame calls for not only a concemplacion but also a reading.
As can be seen, chese three funccions follow from ehe foundarional principle
of che language of comics, chat of iconic solidaricy, and chey specifY it already.
As for ehe chree orhers, ehe funccion of closure, ehe srruccuring function, and
the expressive :function, these are exercised by frames other ehan those of a cern-
ies panel: They can be found in ehe general categories of iconic representation.
However, when these functions are exercised wiehin a multiframe, the effects
thac they produce appreciace, not absolucely, but relatively to ehe neighboring
frames; if a frame is endowed with a strucruring or expressive power, this compe-
tence exercises irself differently according to how ehe frame is similar or different
from ehe surrounding frames and, in a !arger sense, according to ehe coeflicient
of regularity ehat characterizes ehe entirety of the page Iayout. Finally, it must
be recalled, anocher variable, independent from ehe iso- or hecero-morphism of
che frames, is ehe additional decerminacion chac each Frame is subjected to by ics
own site.
1.8-AN INTERMEDIARY SPACE: THE STRIP
The cradicional model of ehe comics page (sracistically in ehe majoricy, despite
ehe number of observable breaks wich trad.icion over ehe past quarcer century)
arranges ehe panels in horiwncal rows separated by whice intersrices. In France,
chese rows--which, ic should be noced, gave rise to che term "bande dessinee"-
are commonly designated in the profession under their American name of strip. !I
The hegemony of chis imporced term is historically juscified. lndeed, if, at :II
~ !:I
ehe turn of ehe twenriech cenrury, comics were developed beginning in the color
Sunday supplements (wich major creations such as The Katzenjammer Kids,
Buster Brown, or Little Nemo in Slumberland, to eire only chose that spread chem-
selves out over ehe ful! page--che Sunday page--in which che forcnat was akin to
a poster), from before 1910, ehe interior pages of ehe American press integrated
several horizontal strips in black and whice, ehe daily strips, each day of ehe week
(from Monday chrough Sacurday).
58 The 5pofio-Topicol System
Europe also knew ehese rwo formulas: ehe one (linear) of the Strip, and the
oeher{tabular), of ehe page. However, since ehe principal publishing format of ehe
comics on ehe Old Continent was not the daily press, but specialized magazines
(formerly designated as illustrt!s [ülustrated magazines]), it is natural ehat ehe page
was immediarely imposed as ehe unit of reference.
The apparatus rhat is mosr familiar to us thus presents the appearance of a dou-
ble fitting-togeeher. The strips unite the panels; the page, in its turn, unites the
strips. In rhe midst of rhe page, it seems rhat the strip, devoid of autonomy, has no
Status oeher ehan that of an intermediary unit. Thus, while ehe panel and the page
are two closed and structured spaces thar the eye is pleased ro survey and whose
contents lend themselves easily to a toralization, ehe strip, irself, appears like a tran-
sie zone, insufficiendy homogeneaus or isolared to be able to claim a true identiry.
In terms of ehe graphic oucline, and from ehe point of view of the narrarion, it is
incorrect to affirm it as a pertinent unit. The strip is nonetheless a space in which
ehe spatio-topia can and must admit specific functions and usages into evidence.
Architectural meraphors are frequendy used to describe its place in ehe general
economy of ehe page. Indeed, the page resembles a house ehat has several stories
(at least two, more often rhree, perhaps four, sometimes more-up to eighr, as we
have seen, in the example from Bill Griffith). Among other works, the operring
page of the surprising book Carpets' bazaar--a fucade in which the ehree Ievels
presenr diseinet ornamental particularities-suggests a literal illustration of this
analogy.'6
At ehe Ievel of breakdown, if it is only necessary to section an action ehat is
to be represented, a discourse to come, the succession of panels are presented
theoretically as a simple end-to-end formation, a strictly linear series. The mental
form under which ehe body is thought of in ehe work in progresswill be ehat of
ehe ribbon-or of the roll of film. For a period of time in which ehe materialiry
of the medium is not confronted, ehe projected work willliken irself to ehe hori-
zontal thread of successive panels, unwinding ehe ehread of a narration in images.
Comics actually took this imaginary form when, well before its invention by
Töpffer, the book had yet to replace the roll, the codex succeeded by the volume.
The Bayeux Tapestry, just like the Japanese Emakimono that were its contempo-
raries, or even (although wound in a spiral), the rwo hundred-meter frieze ehat
decorates Trajan's Column, is sufficient evidenceY
Bur this imaginary ribbon must be placed into ehe mold that forms the pub-
lishing format, into the page. This page Iayout has all the appearance of a violent
surgical ihtervention, of an aggression: rhe point is to segment the "ribbon," to
The Spatia-Topicol Sy5tem 59
section it, since, unless it is folded up on itselflike an accordion (and ehus made
illegible), ehe "ribbon" cannot be adopted into ehe template imposed by the pub-
lication. Wiehin the page, the segmenrs will be placed under each oeher--and
ehey wül be recognized them as strips.
Ir appears then that ehe strip, conceived as it is, does not natural~)! constitute
an integrated plastico-narrative entity. Often, it is nothing more than the rela-
tively random product of a fragmentation imposed by ehe publishing format.
When ehe artist "bumps into" the right lateral edge of the page (more precisely to
irs useable surface, a reduction from the margin), it "falls into line." !fit is notthe
object of an increased investment, if the artist has not made it the receptacle of
an additional narrative or aesthetic determination, the strip has little more perti-
nence than a line of text forwhich it is ehe randomness of photocomposition, and
not the instructions of the writer, that decides wich which word it commences
and where it will find its end. lrs unique function is of ehe readerly order: indeed,
ehe strip indicates, wiehin ehe compartmenralized space of the multiframe, a read-
ing route,.a vectorized ttajectory, which Pierre Ftesnault-Deruelle righdy argues
is, in some way, "cleared from ehe proliferating mass of images. " 48
The impossibiliry, for ehe cartoonist, of preserving in the sequence of images
ehe "ideal" form of the ribbon, clearly appears when the breakdown mirnies a
cinematographic displacement (a movement of ehe apparatus), such as the pan-
orama, or a form of lateral traveling. We have seen an example with the page by
Aleix Barba (cf. fig. 5), to which it would be interesting to compate the short
story by Regis Franc entided Interieurs. The horizontal sweeping is frustrated by
ehe changes to the strip; ehe cartoonist is subjected to ehe obligation of artanging
explicit links in order to permit the reader to heal these fractures.
Of course, the assimilation of ehe strip of images to ehe line of text is only
approximate. The area of the strip (which, inside the invariable hyperframe, is
determined by irs height) and, correlatively, ehe reduced number of strips ehat it
allows on the page, permits each to detach itself enough in ordertobe offered indi-
vidually to the evaluation of ehe gaze (whereas in ehe page of writing, ehe line of
, ·" text, which is much more indistinct, is only extracted from the typographic mass
with difficulry). lt is precisely because the strip has a body ehat it is atrached to an
aesthetic stake: which oucline, which presence is it given? The alternative is posed
in simple terms. Eicher ehe page is divided into randomly sectioned fragments,
in which case ehe form imposes irs law on the auehor, eventually working against
-him; or the strips can be cut and articulated according to an architectural per-
spective, in which case the material constraints can contribute to rhe generation
60 The lpofio-Topicollystem
of an artistic success. Benolt Peeters has shown how, in the work of Winsor
McCay, ehe strip plainly plays its role in ehe apparatus of enunciation chac is
scaged by ehe page. Bur it remains to be added that ehe facher of Little Nemo, lictle
followed on this road, remains "one of the rare aurhors to have gotten an aestheric
and narrative part out of this intermediary unit. "49
Aestherics and narrative: indeed, they are a function of rwo different criteria,
one temporal, ehe ocher spatial, which must appreciate ehe integration of ehe
panels in ehe scrip, and ehe strips in ehe page: As regards ehe temporal dimension,
taking account of the strip as a pertinent intermediary unit will allow us to be
precise wich what I referred to in ehe preceding section as ehe rhychm of ehe read-
ing scanned by ehe frames.
To abstract onself from a frame in order to "dive" inro what follows is an
operation that is accomplished in a fracrion of a secend and which rakes the
form of an unconscious automatism. lt remains that rhis displacement of atten-
cion involves ehe crossing of a void (ehe between-images) and thac, rhychmically
speaking, all ehe voids probably don't have ehe same value. One smoochly glides
along ehe panels thac, caking parc in ehe same strip, follow each other along ehe
horizontal axis, while a leap is required in order co pass ro ehe following strip. The
linear course of ehe reading is briefly subjected to a breaking. This last is clearly
more sensitive than ehe equivalent of ehe passage from line to line in ehe read-
ing of ehe page of text. On ehe one hand, because ehe reader is returned farcher
behind, by ehe fact chat ehe format of pages in a comic book is greater chan that
of an ordinary book; on ehe other hand, and especially, because of ehe height of
the strip: vertically, the leap that is accomplished represents a quarter, a third, or
even half of ehe page.
Also one can estimate that the "gutter" between the ultimate panel of one strip
and ehe first panel of ehe strip situated underneath-it is suitable to us tosaveehe
term gutter in order to designate that which is more exactly a route, wich eventual
"fly overs" of ehe drawn parts--is an interval greater than ehe simple gutter that
separates two contiguous panels. In the unfolding cadence of the strip, it inscribes
a leaner scansion. However, a more important gutter represents the passage from
one page to ehe following page. (The "value" ofthisdifferent gutter is also relative
to whether the pages· are situated side by side or printed recro-verso.)
In ehe musical vocabulary, ehe different values of silence--counted in beats-
have received names; an analogy can be risked between the gutters of cernies and,
respeccively, ehe sigh (that is, for us, ehe between-images), ehe semi-pause (ehe
hetween-scrips), ehe pause (ehe between-pages).
The lpotio-Topicallysrem 61
So, when the more or less regular Iayout observes a canonical division into sep-
arate strips, the reading of a comic obeys a natural rhythm, a breathing aroused
by its discrete apparatus of enunciation, which, discontinuous, is laid out in strips
and rabular. Pursuing ehe comparison in the musical sphere, it can be said that
the strip passes for a measure-but an irregular measure, given that the duration
of the panels is not constant.
When the layout is chaotic, this breathing becomes affected, anarchic, or even
disappears wirhin a phenomenon that accompanies the reading. If the page even-
tually wins an expressiviry, it can be thought that, correlatively, something is
lost ro ehe quasi-hypnotic power of fascination exercised by ehe drawn fiction.
In fact, two effects are produced together that go in rhe same direction. The
transformation of the layout into an ostentatious performance (instead of an
apparencly neutral apparatus, wich a cendency toward transparency) diverts ehe
formal parameters to ehe profit of a part of ehe attencion thac, otherwise, would
be entirely devoted to ehe narrative contents; and at ehe same time it is rid of the
reader's captivicy ro ehe rhychm, on which ehe comics most often recognized as
classics naively rest.
When ehe Iayout favors ehe regulariry of a rhythmic reading, ehe auchor can
look elsewhere and simply leave this effect to act; or he can make use of it to
organize ehe srory. A nurober of authors generally hold to ehe page, using it as
a narrative unit, rnaking ehe change of a page coincide wich a change of place,
of time, or of action. Perhaps less numerous are those that, wirhin a page, com-
pose mini-sequences calibrated wich regard ro ehe natural unit chat is ehe strip.
Yet a McCay, as Benoi:t Peeters points out, "in a manner at once simple and
marvelously efficient, is helped by ehe changing of ehe strip in order to care-
fully prepare ehe ellipses."'' As it occurs, ehe efficiency cakes on ehe suitabiliry
established becween ehe accelerations of ehe story and ehe rhythmic punccuation
imposed by the medium. The narration and its form share ehe same periodiza-
tion. For example, the strip appears as an adequate unit so that the secring up,
ehe development, and ehe resolution of a gag follow each ocher immediacely and
form a sequence ehat is a homogeneous compound. Also the great humorists-
Herge first of all-have often rnade cheir gags coincide wich ehe dimensions of
ehe strip.
A particular layout characterizes Tardi's m.agnificent work, which is entided
C'itait Ia guerre des tranchies." Each page is divided into three stri ps of ehe same
height and composed of a single bandeau (an image stretched over ehe entire
size of the hyperframe). I want to eire here some lines of commentary that were
62 The Spotio-Topicol System
provided by Jacques Samson, who confirms, by rhe study of rhis example, the
general principles that I have tried to bring out:
Jhe rigor, the systemoticity ond the little used choroder of this construction of poges provides evidence
of oformal double constroint opplied to the imoge: the invoriobility of the fiome und the nortilivision of
the strips.... Jhe usoge, hobituol in comics, of the vorioble fiome stretches, compresses, ond dilotes
the rhythm of the reoding, while the constroint of the invoriobility of the fiome hos on inve~e effect; it
bolonces the progression by imposing origorous ond imperturboble metric-here, the tercet-in which
repetition produces osort of spell.52
The other dimension concerned with rhe strip's mode of investment is space.
Th~ compartmenralization of rhe page threatens ro allow rhe gaze to wander and
to scatter. Wirhin the multiframe the strip not only suggests a direction to the
reading, but, according to ehe degree of visual cohesion rhat it demonstrates,
appears like a pregnant visual rone, an aesthetically profito.ble slice.
Ir is easy to see how the management of time determines that of space: An
ellipsis leads almost necessarily to a modification of rhe image (rhe changing of
the place, ehe entering of a new character into the scene, the passage from day-
time lighting to nighttime lighting, etc.) and this modification allows the oppo-
sition of rwo consecutive strips wich regard to their graphic content, assuring a
minimal identity to each. But it is no less conceivable rhat rhe visual unity of the
strip (the coalescence of panels that constitute it) can be obtained by other means
and does not necessarily suppose that it will be preceded and followed by narra-
tive ellipses. Iris fortunate, because it is hard to imagine how one can (excepr in
experimenral cases) raise to the Status of a principal that rhe breakdown would
consist in reducing each scene to the dimensions of a strip.
If the inrerventions on the temporal unfolding of rhe story concern the enun-
ciated, then the visual idendry of the strip is indeed more often assured by rhe
processes that rhemselves rouch on rhe enunciation. At this level, two types of
pararn.eters are concerned: those that address the general architecture of rhe page,
and those that participate in the graphic rreatment and the "direction" of the
enunciated.
To begin, I will mention (without taking into account rhe effects inferred by
the face to face of the left- and right-hand pages) four parameters rhat very much
belang to rhe first group.
r. Calculated by the toral height of rhe page, the propottion occupied by rhe
strip, in an absolute sense and relative to the proportians of the orher strips
The Spotio-Topirol System 63
on the same page. Asrrip standsout better ifit is larger, or ifits height differs
from that of its neighboring strips.
2. The rhickening of the blank horironral interstices rhat separate rhe strips, in
the absolute but also relative to that of rhe vertical gutters that separate the
panels. In Bourgeon, for example, the autonomy of the strip is reinforced by
rhe fact that the "berween-strips" are !arger rhan the "between-pages."
The location assigned to the word balloons. Sysrematically placed in rhe
3-
upper part of rhe panels, they reinforce the frontier already esrablished by
the intersritial gutter and contribute to separating the strips. A different dis-
position, if it is anarchic, will scramble the apparatus of rhe layour, or, if
it is concerted, will substitute another logic. This point will be more fully
discussed in 1.9.
4· The numbet of panels that make up the strip, in the absolute terms and
relative to the quanriry of panels rhat are induded in the neighboring strips.
Except when they are particularly high or excessively flattened, strips that
conrain rwo equal panels in dimension are noticeably less detached than
the orhers. From the instant where rhe format of the panels moves closer
to square, rhe effect of the strip is reduced, as is the narrative dynamic. The
rwo panels, which balance each orher, ask to be contemplated separately;
they srretch roward an aesrhetic of the picture. (The illustrative dimension
so characteristic of Hal Foster's Prince Valiant was norably reinforced ar the
beginning of the 1950s, when rhis cartoonist adopted, as a systematic rule,
placing at rhe center of each ofhis pages two !arge panelsinan almost square
format.) On the other hand, rhe autonomy of a srrip is reinforced when, as
in Clftait Ia guerre des tranchtes, it is composed of a single panel-bandeau
srretched across rhe entire width of the page, or when, on the contrary, it
hosts four or more panels. In this last case, the fragmentary character of the
panels is highlighted. Condemned by their narrowness to a relative incom-
pleteness, rhey cannot bur con-figure. Their meaning is generally to be found
in the linkages (the synragm); it is at this point that the strip provides an
insrance of interpretation.
Other susceptible parameters reinforcing the visual unity of the strip, too
numerous to be enumerated, more directly concern ehe narrative rhetoric. They
have in common the reinforeerneut of the redund.ance or the complementarity
64 The Spotio-Topicol System
of the iconic contents. lndeed, all the processes rhat, implemented wirhin a strip,
have ehe effect of highlighting eieher ehe permanence of ehe same motif in some
conjoined panels, or ehe complementarity of the scenes presented side by side (a
unirary decor broken into several images, ehe lateral movement of a character,
ehe decomposirion of a movement or a visual gag, etc.), conrribute to rhe syn-
tagmaric cohesion of ehe strip and, therefore, to its affirmation as a plasric unity
and pertinent narrative.
Cerrain carcoonists have played chis natural collusion becween panels of ehe
same strip like virruosos. "Thus, in Segar, it is space that finds itself abolished, ehe
characters passing instanrly from one panel to another, even if the places repre-
senred in each of ehernarenot contiguous. It suffi.ces chat ehe panels are."n This
example, provided by Jean-Claude Glasser, precisely demonstrates ehe possibility,
for ehe surface of inscriprion, of subsriruting ehe diegetic space; rhis shorr-circuit
berween two spaces (ehe one, continuous and of rwo dimensions; ehe orher, scat-
rered in three-dimensional fragments that are supposed to be rton-contiguous)
is the principle of numerous reflexive sequences, where cernies are arnused to
denounce their particular codes.
The allusion to Elzie C Segar, ehe creator of Popeye, allows me to raise a par-
ticular difficulty of ehe multiframe, which concerns ehe direction of a chase scene.
Perhaps one recalls ehe memorable pursuit on the deck of a boat that is one of rhe
highlights of Rodolphe Töpffer's Histoire deM. Cryptogame. Benoie Peeters has
compared ehe two known versions of chis sequence: one, originally dr3.wn by ehe
Genevan artist, and ehe other prepared by Cham for ehe engraving of ehe stocy
in L'IIlustration in 1845- Peeters remarks that Cham was not careful wich cegard to
the inversion that ehe engraving would force upon ehe images, ruining "certain of
the most effective discoveri~ ofhis model,"H such as this famous pursuit. In this
species, ehe fortunate will find that "ehe three characters [that is, M. Ccyptogame,
Elvire, and the priest] perfectly adopt ehe movement of reading, moving from left
to right wich a nice ensemble effect that allows ehern, in some way, to step over
the Frontiers between the frames." While in ehe case of ehe unhappy Cham, "ehe
dynamic of each of ehe actors finds itself frustrated and conrradicted by ehe tra-
jectoty of the gaze. The chasers have become ehe chased, and they must, in each
panel, retrace cheir steps."
Now, this readingcan be reversed to ehe benefit ofCham, because ehe Töpfferian
version of ehe chase rests on a paradox that Peeters has not, it appears, perceived:
Elvire occupies in ehe strip (which, in Töpffer, is always equivalent to ehe page) a
position ahead of Ccyptogame, while she is supposed to be running behind him.
The Spotio-Topitol System 65
Similarly, ehe priest, who must be ehe third in line, appears to occupy ehe head
of ehe trio. Töppfer had an intuitive mastety to ehe spatio-topical appatatus, and
singularly of ehe lateral dynamic particular to ehe vectorized strip by ehe move-
ment of ehe gaze; but here this logic, that of ehe published form in its physicality,
contradicts ehe logic of ehe action and of ehe diegetic topogcaphy.
Segar leaned on ehe physical space in order co institute an aberrant diegetic
space from ehe logical point of view, but one that is perceptively acceptable.
Töpffer found hirnself faced wich a dilemma that did not provide any good solu-
tion-in Cham, ehe characters chase each other in ehe right order but indeed
not in the right direction--except to adopt a unique panel-bandeau that simul-
caneously represents ehe three characters, ehe physical space coinciding wich ehe
diegetic space, but in this hypothesis, it is ehe rhythm provided by ehe captions
that would have to be sacrificed.
Having noted this paradox, which perhaps consticutes an aporia of ehe mul-
tiframe, let us return to another good example of ehe concerted use of ehe strip,
on page 52' (cf. fig. 6) of Teule and Vautrin's Bloody Mary (Glenat 1983), which
includes chree strips, each composed of three isomorphic images. Each strip man-
ifests a remarkable cohesion: ehe first by ehe fact that it opens and closes wich
ehe same diagonal, ehe following two by ehe repetition, in ehe three consecutive
images, of ehe same element. (The !arge scar on ehe face ofN'Doula, then ehe ver-
tical "upsurge" of ehe group ofbuildings-that in this case, metaphorically evoke
another type of erection.) At ehe same time that he has directed this partition of
the page into chree contrasring zones, each supplied wich a particular coherence,
Jean Teule, concerned about narrative fluidicy, has carefully established explicit
links becween ehe consecutive strips. Thus, N'Doula's face appears partly in ehe
last panel of ehe first strip--ehe link is also effected by ehe adoption of an identical
skin color, which subsequendy deteriorates in order to progressively merge wich
that of ehe sky-while ehe HLM make their appearance in ehe ultimate panel of
ehe second strip, already beginning ehe tri-panel that follows, which conserves, by
its reciprocity, ehe memoty ofits predecessor, through ehe face ofN'Doula oblit-
erated by ehe new visual element which will fotce him to fade away.
(I benefit on chis occasion to be also able to highlight, in this example, ehe
intelligent use made of ehe site in two places of ehe page, which are ehe first panel
and ehe central paneL From ehe upper left corner, ehe entrance of N'Doula into
Mary's apartment coincides wich his entrance-and that of the reader-into ehe
page, which is a left-hand page. As for ehe central panel, its situation predis-
poses it to operate on ehe synthesis becween ehe terms that are contradiccory in
66 The Spotio-Topical System The Spotio-Topicol System 67

principle: suffering and love, as the caption suggests, but also the black race and
white race, since in this panel N'Doula has a mixed complexion.)
Except when it is over-determined by processes such as those used by Teule,
the strip, it is necessary ro note, is relativized. in a certain way by the choice of a
regular Iayout. Indeed, when ehe interpanel gutters are arranged as venical con-
rinuations of each other, thanks ro the isomorphism of rhe panels, other routes
are open to ehe surface of ehe page than ehe horizontal of ehe strip: a vertical
route, or even, in the style of a checker board, a diagonal route.
Ir does not enter into my intentions to plead for a systemaric reinforcement
of ehe Strip. I sought nothing other than placing in ehe light its potentialities,
and ehe diverse usages to which it lends itself. !t belongs to each author to decide,
in the fllnction of a narrative srrategy and a global aesthetic, if and at what
moment rhe strip, this intermediary space, roo often ignored in irs funcrional
specificicy, deserves tobe reinforced.

1.9-AN ADDITIONAL SPACE: THE WORD BALLOON

The description of ehe spatio-topical apparatus is still incomplete. Indeed, ehe

form, ehe number, and ehe location of ehe word balloons (bulles), in sum, ehe
necwork that they create wiehin ehe hyperframe, also regulate ehe management of
space, and contribute in a determining Fashion to directing the gaze of the reader.
Although it is unusual to consider ehern in this way, ehe word balloon panicipates
in the constitutive spaces of the comic-as do the frames enclosing a narrative
text, which I designate, following others, ehe caption (recitative). !t is in this con-
text that I am interested in ehern, generally wichout regard for their panicular
status and the nature of the Statements that they accommodate.

1.9.1 The Balloon in the Panel

Like ehe panel, ehe balloon is, as a general rule, a closed, compact space. This
outline could not be given ehe generic narne of a frame (etymologically forged
from quadro, square) without some diffi.culty, in the sense that, as its very name
Fig. 6. From 8/oody M01y (1983), by Jenn Teufe und Jeon Vuul!in. © Edi~uns Glenut.
suggests, it generally stretches toward an elliptical form. (Although quadrangular
balloons are common-I will return to this-and, of course, captions are habitu-
ally presented in this form.) The fact remains, as weshall soon discover, that, at
68 The Spotio-Topicol System
the place of its interior space and vis-a-vis its exterior environment, ehe circum-
ference of this "word sack" exerrs ehe majority of the functions identilied above
as characteristics of the panel's frame.
Again, as wich the panel, it appears on occasion that the text is not to be
crimped, and the words are welcomed wiehin ehe representational space wirbout
the iconic element and ehe lield of writing being explicitly disassociated. Claire
Bretecher, Jules Feiffer, or the Fred of Petit Cirque worked this way, amongst
others. Jan Baetens has very justly remarked that in this case, ehe writing is placed
"in the desemantized zone of ehe image: for Iack of duly frarned captions, the
verbal information emerges in a place where ehe image appears empty. Thus, all
conflict of precedence 6nds itself evacuated. ... The scission of ehegenrein writ-
ing and image is not contested. "55
A natural hierarchy exists between the panel and ehe balloon. First of all,
because if the existence of the one conditions that of ehe other, ehe opposite is
not necessarily true. A number of panels (in variable proporrion atcording to ehe
authors) do not include balloons. On ehe other band, balloons are never presented
alone, because ehe balloon is an emission that is supposed to be resonant, and
every emission presupposes a source, or a place of origin. The panel is that place,
or it contains it. This rule is true whatever the exact nature of the source: whether
a depicted speaker or an invisible speaker, it is always situable with respect to a
diegesis of wh_ich at least a &agment is represented, or was previously.
The balloon cannot be postulated wich out, correlacively, postulating ehe paneL
This aflirmation signilies precisely this: a balloon that occupies, in the hyper-
frame, an unframed and empty position, a balloon that is detached, isolated,
wirhin an empty space, suflices to attest that there is well and truly a panel there,
and chat despite appearances the discourse of ehe page is not interrupted. Why?
Because the balloon itself is at the sarne time information (an outline invested
with a known symbolic function) and a carrier of information (the words or ehe
graphic elements that it contains) and since, from this fact, it is identified, in this
particular case, with ehe panel itself.
The case thac has just been described is relatively rare. On ehe other band, it
is more common that a single caption occupies the entirety of a frame. And this
current practice, in the work of authors as dissimilar as Godib and Mufi.oz, dem-
onstrates that the panel is not necessarily _mixed in nature, since, if certain panels
include drawings wichout text, others enclose only text.
In a hierarchy of spaces, ehe balloon is thus subordinated to ehe panel because
ehe panel can proceed wichout the balloon while the balloon necessarily implies
The Spotio-Topicol System 69
ehe paneL But this link of Subordination awakens a secend aspect: In ehe panel,
ehe balloon counts fur only a patt; it does not consist of the totality, but merely
of a closed subgroup, and ehe remairring part is, in principle, dedicated to ehe
drawing.
To this degree of generality-that is, in considering neither that which is
represented by ehe drawing nor what is said in the balloon-one can already
distinguish four different Ievels in ehe relation that is established between ehe
balloon and the host panel, between ehe part and ehe entirety. Initially we think
in cerms of depth. We also speak of a relacionship that opposes two forms, then, of
a relation between two areas, otherwise called a rapport of proportions; and the
final criterion is the positioning of the balloon wirhin the panel.S 6
We are going to explore these four Ievels one after the other. In. the course of
this investigation it will be necessary to alter the coordinates, in order to consider
ehe balloon, not only in its relation to the frame of the panel, but wirhin the
hyperfrarne of ehe page. It will not be ehe isolated balloon that interests us, but
ehe totality'.of balloons comprised in ehe page; we will discover that, occupying
relative positions, they compose a network in which the general con.6guration is
a determining factor in ehe prococol of reading.
The notion of the "ratio of depth" between the balloon and the panel is a
double understanding. At first, it designates an opposition between the "textual
zone" and "image zone." lndeed, the image, to the degree that it relies on the
petspectival code and practices the staging of the planes, creates ehe illusion of
three-dimensionality. The texc, on the other band, frees itself from this mimetic
transcendence, respecting and con.6rming the bi-dimensional materiality of the
writing surface. "When the panel is cut into two zones, one asserts a flatness that
is betrayed by the other in ehe production of the illusion of depth. In this sense,
it is Iegitimare to assert that ehe cohabitation of the drawing and the balloon gen-
erates a tension, since the three-dimensional space constructed by the cartoonist
is contradiered by the presence within it of this piece that is added, a stranger to
the representative illusion.
Obviously, the humor cernie ofren plays wich the procedures of a reflexive
character that rests on the exchange between the writing surface and the diegetic
space, ehe plastic (material) dimension and that ideal iconic dimension of the
drawing are manifested together or one after the other, but always in a contradic-
tory fashion. In doing chis, the humor exploits, prolongs, and explicitly renders
a tension that, from the moment that there are balloons, is always already in
thework
70 The Spotio-Topitol System
When comics are in color _the balloons are most often presented as white
{although, from Jacobs tO Sienkiewicz, a certain number of aurhors have colored
them). The whiteness of rhe balloon, which is that of the paper, anests to its
indifference to the illusionist convenrions that govem ehe image.
The ratio of deprh is, simulraneously, of a slighdy different order. The bal-
loon, as was notably observed by Pierre Fresnault-Deruelle, "is at first this white
presence rhat neutralizes rhe decor." 57 Indeed, the balloon, even when it tends to
be discrete {norably in occupying what Jan Baetens calls a "desemantized wne"),
subtracts from our view a portion of the image, even when this part is supposed
tObe virginal {such as a sky wirbout clouds in a black and white comic).
It is not important whether the cartoonist did or did not invest rhis partat the
moment that hecomposed his image. When rhe balloon is added to {or introduced
into) analready constructed imageehe cartoonist effectively suppresses-rhrough
erasure or by collage--a portion ofthat image. If, on rhe contrary, the balloon is
drawn right away, and then taken into account from ehe initial Conception of ehe
panel, it can happen rhat the penciling provisionally "overflows" into rhe space
reserved for the text, because eirher rhe cartoonist needed t0 skerch all the parts
of a pattern in order to tesr its construction, or, in all simplicity, his gesture drives
it. Bur, even though ehe same "textual zone" would not have at any time sheltered
even the smallest aspect of rhe drawing it does not remain less rhan the balloon,
because it is besieged by rhe image, always producing an ejfect ofconcealment. And
rhis effect is what carries us, withour considering the genesis of rhe pane}.58
In reality, the effect of concealment is a direct consequence of rhe tension
between rwo- and rhree-dimensionality evoked above. In rhe first pages of his
famous essay on Perspective as Symbolic Form, Erwin PanofSky provides rhe means
to camprehend the articul~tion berween the two phenomena that occupy us, or
rather to understand rhat it is necessarily rwo aspects of rhe same phenomena,
when he writes:
For us, the pe~pective, in the lullest sense of the term, is the obility to represent severol objects within the
port of the spoce in which they ore found, of osort so os the notion of the material support of the picture
finds itse~ completely dnven out by the notion of the transparent plane, so !hat we believe, our goze
trove~es in order to plunge into an exterior imoginory spoce !hat would contoin oll the objeds in opporent
string ond which will not be limited but on~ cut by the borders of the picture.59
If rhe frame of the panel cuts the "exterior imaginary space" that the drawing
provides access to, allowing rhe reader to mentally posrulate an off-screen, that
The Spoffo-Topi<ol Sy5tem 71
is an extension of this space beyond rhe Iimits assigned by rhe frame, the reader
is, a fortiori, held to posrulate rhat which I call an "out-of-sight" behind rhe zone
hidden by rhe balloon. Virtually unlimited in its reach, rhe exterior imaginary
space cannot comprise emptiness. The balloon does not designate a natural cavity
in the space that is depicted; it inscribes a zone of opacity wirhin the "transparent
plane" that we identifY as the paneL (Writing of painting, Hubert Damisch elo-
quendy evoked "the univocal connection that wants the figure, by its outline, to
remove itself from rhe background, or pur it another way, rhat the background,
literally, is removed from untler rhe figures_")6o
This mode of apprehending rhe panel is constructed rhrough a faith in per-
spective and of orher analogical codes that, creating the image in the resemblance
of the real, invite us to project ourselves into what Panofsky calls rhe imaginary
space, and which, when acting in a narrative genre, has come to be known as the
diegetic universe. lf rhe imagewas understood, not in irs reference but in its strict
materialiry, as a group of lines or signs inscribed on a support plane, the solurion
to continuity between the drawing and the balloon would reflect on norhing
more rhan on the nature of signs {analogic here, rhere digital) and rhere would be
no place to imply that such signs conceal or recover any others.
Ourperceprion of the figurative image beingwhat it is-aconsenring illusiOn-
it can be said rhat the balloon produces an effect of concealment. Spontaneously,
we have a rendency ro Formulare rhis effecr in terms of a recovery, as if rhe bal-
loon effecrively superimposes irself, like a mask (as in rhe dots rhat are sometimes
used to cover the most explicit parts of pornographic images), to a previously
complete and homogeneaus image. The vocabulary used a moment ago, when I
simared rhe out-of-sight "behind the zone masked by rhe balloon," reflects this
propensity.
(Let us note rhat, in the absolute, rhe balloon could also be seen as a hole, a cav-
ity in which we rediscover the plane of the printed page, located "under" the plane
ofrhe image.lt is as ifthe panelwas glued onto the page, and the balloon was acut
rilade in this panel. This version is as coherent as the first, but it does not accord
with our perception of rhe image as opening onto a three dimensional space. One
can make a hole in a plane or hollow out a solid; how would one empty, wirhin
a given perimeter, a space endowed with deprh, and also of an a priori unlimired
deprh-whatever rhe elements limiting the view offered by this space?)
Despite rhe commodity of metaphors borrowed from the lexicon of earthwork,
-rhere is no need to represent rhe balloon and rhe drawing on two different planes.
These are areas that mainrain sparial relarions of conriguiry or inclusion, on one
72 The Spotio-Topicol Sy5tem
part, and another of a shared Iayout (the oucline of the balloon) that divides the
same surface, that of the paneL The image zone and the textual zone ate like two
complementary pieces of a puzzle. The space reserved for the text is a space taken
from the drawing, but is situated on the same plane. So that one can consider the
enclave that is the balloon (especially when it doesn't touch any of the exterior
borders of the panel) as an "inrerior frame" of ehe image, that is, ehe form that,
wirhin it, assigns it a frontier and delimits the field of its visibiliry. The comics
image has the particularity to be stopped o'n two fronts: by its exterior frame
which separates it from the off-screen, and by its interior·frame which bars it from
the out-of-sight. As I will show shordy, this norion of the interior frame is justified
all the better when the six functions of the frarne that I identified are exerted,
almost identically, by the balloon-considered like an outline of a specific space.
As I have said, the balloon and the panelalso maimain relations of form and of
dimension. Reduced to its outline, the balloon is first of all a formthat is detached
wirhin another form, that which is imposed on rhe panel in its frame. A rapport
ofhomo- or hetero-morphism can exist between these two forms; either they Iook
alike, or they produce an effect of contrast, of dissonance. Let us postulace, since·
it is the norm, a reetangular or square paneL With regard to this norm it is already
possible to distinguish two !arge families of balloons: those where the line is also
quadrangular, and all the others that have an elliptical or irregular line. More pre-
cise criteria would evidently allow us to refine this dichotomy in order ro generate
a complex typology. Thus, it is possible to make note of the "notched" character
of the Hergean speech balloons, in the form of arrows or appendixes that poim to
ehe speaker, etc. To do this would be to enter into ehe analysis of particular writ-
ings by different artists. And the essentialline of division would be blurred.
Indeed, the right-angled speech balloon manifests its distinction by at least
two panicular properties: Ir economizes space and it is more discrete rhan ehe
diversely shaped balloons. The space savings that it enables holds primarily due
to the fact that its outline can more closely enclose the body of the text. (An oval
balloon, on the contrary, generally presems white pockets, the bulges on each side
of the text, providing it with an interior "margin. ")
The economy of space that the right-angled Speech balloon also has another
reason: !t is almost automatically calibrated at the same size as the paneL This
conformation is observed, for example, in the work of Herge and of Franc;ois
Bourgeon. When the panel is enlarged, the balloon (except, of course, when it
houses only a few words, or a simple interjection) is dilated in the same propor-
tion and continues to occupy all the available width. In the inverse, elliptical
The Spotio-Topicol Sy5tem 73
balloons tend to conserve their proponions as a constant, which Ieads cartoonists
to more willingly decenter them with respect to the panel's frame, or to regional-
ize them in some way. Thus, the right-angled balloon wins its place because it
very ofren occupies the two upper corners of the panel, which are the zones of
weakest profitabiliry for the composition of the image. (Bourgeon, when the text
is abundant, also allows the speech balloons to leave the panel in order to join,
side by side, the gutter. Forthis author, in whom pages with fifteen panelsarenot
uncommon, the space must be used in a most economical and rational fashion;
it is the requirement of such a dense breakdown.)
The same reason is why quadrangular speech balloons occupying the upper
part of the panel are relatively discrete: They carefully provide the drawing with a
homogeneous area and a regular form. The text hangs over the image, but it does
not overrun it; the icon does not appeat to be penetrated by this strange body.
Edgar P. Jacobs, a "wordy'' cartoonist when hewanted tobe, could not use ellip-
tical balloons. An adventure comic cannot easily accommodate such a tremendous
quantity of text, so it was necessary for him to opt for a more discrete and more
economical, in a word, a more sober, system ofballoons. It is particularly neces-
saty for Jacobs, and this sobriety is more than suitable for the rigor and the spirit
of seriousness that impregnates his stories with a scientific pretext. The problern
allows radically different solutions for a humorcartoonist such as Edika. Here the
verbal prolixity is translated by a proliferation ofballoons with irregular lines. The
string ofballoons that surround the characters sometimes oblige them to wriggle
free, participating in an exuberant styleandin a comedy of exaggeration.
!t is apparent that, if the right-angled speech balloon possesses some cerrain
objective vinues, it is not intrinsically "better" than the others. It is necessary,
in order to appreciate the pertinence of the retained option, to consult with the
general aesthetic sttategy of the author, which implies a congruence berween
elements of a diverse nature. To speak only of formal parameters (although the
subject and the general pitch equally matter): What conception governs the Iay-
out? Is the text rare or abundant? ls it a page in color or black and white? In
the first case, is the balloon white or colored? What does the oudine Iook like?
(In Baudoin, for example, the outline of the balloon and the frame of the panel
benefit from the same irregular line, as rough and sensual as the drawn images,
which has the effect oficonicizing these elements of the code.) Finally, what form
does the calligraphy take? What kind of alphabet is the lettering made of?6' There
are many elements that one can only truly appreciate in the play of their determi-
nations and reciprocal conditionings.
 74 The Spotio-Topicol System
!t is possible to conjoincly use several sorts of balloons, in order to play wich
their differences. The formal opposirions translate as oppositions of status; they
will belong to the enunciacive instances or to different "voices." In the work
of many cartoonisrs conforming to the dominant usage (for example in Tardi,
Mufioz, and Giraud), che reetangle is reserved for ehe captions; it denotes ehe
intervention of the narrator. For dialogue, on ehe other hand, the balloon is made
oval (in Mufioz and Giraud), or of a fanrastic line (in Tardi).
These formal discriminations are pushed further in certain Anglo-Saxon car-
roonists, such as Bill Sienkiewicz. In his Stray Toasters (Epic Comics 1988), for
i
I example, voices are idencified not only by rhe shape of ehe balloons, but also by a
chromatic code. Thus, four or fi.ve different voices can interject all while remairr-
ing identifi.able, each characterized by a color. In an earlier production by the same
artist, The Shadow (DC 1987), ehe balloons spoken by ehe hero beneficed &om a
discinccive line, ehe black ring was doubled, on ehe lefc, by a sorc of red shadow.
On ehe other hand, in Big Numbers (Mad Love 1990), not onlyare allehe balloons
whice and drawn wich an idencicalline, but cheir form is totally uniform: They
inscribe them.selves all in a perfect circle (somecimes ampurated by the frame of
che panel). A formal garne established becween rhe series of panels-all equal-
sized rectangles, with twelve on each page-and the series of circles that never vary
in number or in diameter. Bur ehe writer of Big Numbers, Alan Moore, had opred
in Watchmen (DC 1986) for a singularized codificacion of ehe balloons spoken by
ehe character ofDr. Manhattan, as well as for ehe caprians accribuced ro·Rorschach
or to Max Shea.
Scemming from chesarne logic, rhe saw-toothed oucline is cradicionally reserved
for voices mediared by a device (radio, relevision, telephone), and ehe special scal-
lop (in ehe form of ehe scring of small bubbles) also craditionally denoces ehe
"incerior voice" of ehe thinker. All chese resources are incended co guaranree ehe
intelligibility of the enunciative situation, that is to allow the reader to know who
expresses what and by which method.
In this disserration on ehe form of ehe balloon I have already made some
allusion to their dimensions, relative to those of ehe paneL I do not believe chat
there is a lot to say here, as ·a general rule, about the quantitative rapport that
escablishes icselfbecween ehe part of ehe cexc and chac of ehe image. This rapporr
varies, of course, in considerable proportions, from "all images" (in mute comics)
to a combination where the text occupies most of the surface. The hypothesis of a
comic composed of nothing but text, although it clashes with common sense, has
been attempted, notably by Bosc, Olivier Ka (in Fluide glacial}, Topor (in Strip),
The Spotio-Topical System 75
Jim and Gaston (for a book wich Vencs d'Ouesc encicled On iteint Ia furniere . ..
on se dit tout), and also by Lewis Trondheim, in ehe case of OuBaPo.
If a precise measure of ehe bulk of ehe rexr is useful, ic would appear ro be as
a way to prop up an analysis of the contents. lt would not be necessary to com-
pare the total surface of the "textual zone" with that of the "image wne," but to
compare between several textual segments, in order to establish the respective
imporrance of ehe voices. I arn happy here to borrow ehe study that was under-
raken by Alain Chante, beginning wich Jacques Marcin's book La GrandeMenace
(Lombard/Dargaud 1954). Ir sought to derermine by what means comics repre-
sent the notion of power. The aurhor proposed to "determine the characteristics
of the superior, wirhin professional hierarchies as wirhin the presented societal
group. " 6:t Arnongst the distinctive attributes of rhe superior, he teils us, appear
style of dress, ehe righr to address someone informally, the use of ehe impera-
tive, ehe righc co be angty, and ehe righc co cake ehe initiative. Bur Alain Chance
reached the condusion that "speech appears to be the most pertinent element,
that which can be scudied in a quantitative fashion." For each characrer, he has
then counted ehe number of balloons spoken in ehe book and their toral surface.
This surface is chen compared to ehe number of panels in which ehe charac-
cer appears. The coefficienc chus obtained defines what Alain Chance narnes che
"potential for speech rhac is available to each character, his propensity ro begin to
speak." This type of research mighc incerest ehe sociologiscs, che hiscorians, and,
from a different perspective, the narratologists. Foreign, in his aim, to the preoc-
cupations that are mine here, it neverrheless uses observations of the spatio-topi-
cal order, and it nonetheless confirrns for cernies the fundamental importance of
ehe discribucion of space.
The final paramecer clrar remains to be addressed is by no means ehe least,
since ic concerns ehe vety locacion of rhe balloon. Far removed from indifference,
ehe position chac is occupied by ehe balloon is always relative to three different
elemencs: ehe character who is speaking (ehe speaker [locuteur]), ehe &ame of ehe
panel, and, finally, ehe neiglrboring balloons (whecher they are siruaced in ehe
same panel or in a contiguous panel). 6 ;
No balloons exist that do not refer, and cannot be attributed, to a known
or supposed speaker. The relationship becween rhe speaker and ehe enunciacion
.that is uttered is so scrong thac one can speak of a sorr of funccional binomial.
This bipolar strucrure is one of ehe feeund schemes chat organizes ehe reading
of comics. This is why our perception of ehe depicced scene and, singularly, our
knowledge of ehe dialogue, are barely modified according to ehe posicion chat ehe
76 The Spolio-Topicol System
balloon occupies, beneach, co ehe left, co ehe righc, or above the speaker. (The
generarion of rhis position more precisely begins from ehe face, which is the true
place of enunciation, and the physiognomic expression constitutes the principal
instance of interpretation-in the rwo senses of the term, theatrical and herme-
neutic.) Indeed, in ehe binomial of ehe balloon-character, in principle the chacac-
ter is perceived first. As a drawn element and as a generally privileged element in
ehe composition of the image, its perception is quasi-instantaneous; the presence
of the character is the salient information rhat the reader regisrers at the same
instant in which his attention is directed by the paneL Even ifit is not knowingly
looked at, ehe character is seen instantaneously. (This is, as we shall see later on,
one of the principle aims of narrative drawing.) The text, which requires a linear,
thus progressive, deciphering, is only read in ehe second place. This order recalls
a perceptive automatism; it is largely independent of ehe respective positions of
ehe balloon and of the speaker.
Of course, rhis is not to exdude the idea that the text can arouse, afcer the ini-
rial reading, a second Iook at rhe chacacter. We can speak of an informed gaze, one
that is not turned toward seeking a presence, but toward a more detailed percep-
tion of ehe consriruent parts and rhe attribures of ehe character. Indeed, in many
cases only ehe text enables ehe reduction of ehe intrinsic polysemy of a gesrure, of
an attirude, of an expression. In other terms, if the characrer interprets rhe rexr,
ir is no less true rhar rhe rexr, in return, inrerprets ehe characrer. Thus, between
ehe two rerms of the binomial is carried out a reciprocal adjusrmenr, a filrering
of ehe diverse possible readings. This give and rake is essential to ehe production
ofmeaning.
If ehe direction rhat indicates the balloon wich respecr ro ehe character is mar-
ginally significative, ehe ~er ehar ir is close or relarively distant to him is more
important. As a general rule, carroonists try hard to place ehe balloon in rhe
immediate proximity to ehe speaker, thus favoring rhe mutual derermination that
I have described. When a balloon is placed far from the speaker (which implies a
panel supplied wich ehe necessary height orwidth), one supposes, because chis gap
goes agairrst ehe natural tendency, that it proceeds from a particular desire, from
rhe search for an effect. In addirion to ehe specific effect produced (which will
be, for example, of the order ofbraiding), one often observes in such a case that
ehe remarkable distance between ehe balloon and rhe speaker creares ehe feeling
that a silence has intervened. Ir is as if ehe balloon carries no more than ehe echo
of speech already spoken, and that ehe character has already rerumed to silence.
From this impression (I am only dcawing here on my parricular experience as a
The Spolio-Topicol System 77
reader), it can be inferred rhat ehe image ordinarily convens ehe simultaneity of
time to a proximity in space. Indeed, when rhis proximity is respected, rhere is
also a parricular time given to ehe binomial balloon-character, since, if several
speakers cohabit wirhin ehe same panel and respond to each ocher, ehe starr of
their speech cannor be synchronous; each lives "at rhe momenr of his balloon."
Sometimes ehe character is invisible, situated off-screen, or hidden by an ele-
ment that has been turned into a screen (a wall, vegetation, mist, erc.), out of
view. Sometimes he is visible but too distant for his distinctive traits to be rec-
ognizable-"drowned" in a crowd, for example. The balloon is chen indicative
of his presence, and is his appendage, pointing roward him like an arrow, exerts
a truly signaling function. The balloon can have no ochet reason to exist than
to catch ehe gaze and to manifest, by ricochet, ehe presence of a character in ehe
· midst of ehe frame. Introduced in ehe panel to ehis single end, it continues to
exist more as a symbolic space-a coded sign-than as ehe receptade of verbal
contents rhat are in rhis case incidental; it is its own particular signified.
The written text of ehe comics dialogue is not carried, colored, or embodied
by an identifiable voice. For us ro discover ehe identity of ehe speaker, it is neces-
sary that either---<:onforming to ehe examples cired above-che balloon distin-
guishes itself by a color or a parricular line, or that the speaker uses langnage in
a peculiar way, chat his "idiolect" assures his identification. Unless, of course, ehe
conrext is sufficient: a reply in a siruation of an already launched dialogue does
not insert, in principle, a problern of attribution, even if rhe speaker is provision-
ally invisible.
In ehe second place, ehe position of ehe balloon must be evaluated wirh respect
to ehe frame of ehe host paneL Let us be specific: wich regard to ehe space chat ehe
,ftame forms, but also wich regard to rhe frame itself, to its gcaphic materialiry. Ir
is ehe secend aspect rhar restrains me here, because, inirially, one cannor speak
except on a case-by-case basis, in considerarion of ehe balloon as an elemenr thar
enters inro rhe global composition of rhis particular icon.
Wich regard to the frarne proper, it allows four different disposirions of ehe
balloon:
I. The balloon is derached from ehe frarne, their respective lines not meeting
each ocher at any point. This is ehe system adopred by Herge during ehe
Second World War, and which he extended to his earlier books like Red
Rackhams Treasure, as a resulr of ehe revisions imposed on ehern by cheit pub-
lication in color. 64 Ir is only, I repeat, in this arrangement where ehe balloon
78 The Spatio-Topicol System
is enclaved in the icon, that it truly holds ehe place of an "inrerior frame."
The effecr of the concealment that attaches itself to the balloon plays here
fully, whereas ehe three other arrangements produce nothing but attenuared
versions.
2. The balloon is stuck to the frame, leaning agairrst a wall of the paneL In
the majority of these cases, it is hung on the "ceiling" of the panel (even if
it is also found occasionally on one of the other sides); this resu!ts, as noted
above, in an accentuation of rhe separation berween ehe panels simared one
above the other. The white mass of the balloon enlarges the emptiness of the
gutter. This effect is reinforced when all the panels of the same strip place the
balloons in rheir highest part and across rheir total width.
3· The balloon is leaned agairrst ehe frame, but open: ehe line of ehe frame is
interrupted on its width by the zone of contact. Everything happens then as
if the gutter has budded, forming outgrowths wirhin the panels, where the
text comes to take its place. These arrangements are habirual in ehe work
of cattoonists such as Andreas or Rosinski. No doubt, the text will then be
aligned along the frame itself, the letters that compose it stretch to form, in
a dotted line, the missing part of the outline. To a certain degree, rhe eye of
the reader imagines a complete Frame, because two of ehe functions of ehe
frame, the function of closure and ehe separarive function, are rhemselves
an aide for the reading of the page. Ir remains that placing this a~rangement
in the work conrributes perhaps ro rhe "lightening" of rhe page, bur ir does
not support the capture of the reader; this plunge into the imaginative space,
which PanofSky speaks of, never occurs more clearly than when a regularly
formed frame "inhales". the gaze and makes rhe reader forget it as a frame.
In perforaring the circumference of the panel, an open balloon indicates an
exit, and the irregular formthat it lends to the frame is susceptible to attract
an attention diverted from its principal object.
4- The balloon can finally overflow the frame, eirher encroaching on the mar-
gin, ot parrially overlapping one (or several) orher panel(s). One finds rhis
arrangement noiably in the advenrures of Blueberry drawn by Jean Giraud
andin thework of theArgentinian Horacio Altuna-the last in a version that
I will examine later. The balloon is often centered on the berween-images (of
which it constitutes a sort oflocal dilation) or even placed at the intersection
of rwo perpendicular blank spans, reducing rhe corners of three or four panels
The Spotio-Topicol System 79
at one time. Only rhe direction that is indicated by the appendix of rhe bal-
loon allows the attriburion to a cerrain panel rarher than to some other. In
adopting this arrangement, the cartoonist manifestly seeks to free his images,
ro derive the maximum benefit from all of the "desemantized zones" thar rhe
page has to offer, in order ro make more space available to the drawing. The
construction of this type of page sometimes seems a bit untidy, subjecting
the eye to an erratic roure. Bur the feeling of density in the narrative drive
and the constant dramatic tension that impregnates it, two of the principle
qualiries of the Blueberry series, are perhaps not withour relation to this anar-
chic proliferation ofballoons that maintain the reader in a state of unceasing
alerrness.
L9 .2 The Balloons in the Page
On the following page I have reproduced (fig. 7) page 14 of LeBout de la piste
(Novedi 1986). The four dispositions of the word balloons that I have described
can be found there.lt is possible to observe rhar, of rhe rwenry-nine balloons rhar
are induded on the page (in the rwelve panels), rwo exceed the exterior margin,
ren straddle rwo panels, and one-the antepenultirnate-encroaches upon three
panels at once.
Not unlike rhe emblernatic Giraud, cerrain authors, whose aesthetic could be
termed pragmatic, combine rhe four disposirions of the balloon with respect to
the frame, choosing, for each panel, that which seems the most appropriate, and
sometimes reconciling two or even threedifferent solurions wirhin the same frame.
Others, in which the art is more systematic, do not use, within a given book, and
somerimes wirhin their entire career, anything more than a single scyle of inserr-
ing the balloon in rhe frame. This opposition, which one can similarly find in the
management of other parameters, clearly de6nes, it seems to me, rwo "families'' of
authors, rwo fundamentally different approaches to rhe language of comics.
We have begun ro expand our field of investigation, since rhe very fact of
encroachment has necessitated that we consider several panels together. This
expansion is indispensable when examining the position of the balloon relative
to those that precede it and rhar follow it. The point is crucial, because the bal-
loon is perhaps the only element of the paginal apparatus on which the gaze
definitively srops (except when leafing through the comic without reading ir). lt
is a point of anchorage, an obligatoty passage. Because of this rhe reading can be
80 The Spofio-Topicol System
fig. 7. from Bluebe11y: LeBout de fu piste (1986), by Jeon G"rroud end JeooMichel Charlier. © Edilians Dargoud.
directed to a certain degree, driven by rhe network rhat connects rhe occupied
positions of rhe successive balloons across the page.
We have already noted rhis: To insert a written enunciation (notably, a
balloon-but also many of ehe depictions of postecs, of books, of graffiti, etc.)
into a panel allows the author to retain the reader's attention for an instantwirhin
The Spotio-Topicol System 81
a frame that, wichout it, would expose itself to the risk of simply being ignored
or skimmed over.
Two balloons siruared side by side, on both sides of an intericonic gutter, care~
fully create a sort of natural bridge becween the two joined panels. If the balloons
are moved apart, it is likely that the image zone situated between ehern will also
be swept away by the gaze. Evidently, a thousand detours are possible on the
trajectoty that Ieads from one balloon to the next; ehe positions of the balloons
does not so much indicate a road to be followed as ehe scages to be respected,
berween which every reader is free to wander areund in rheir own way, oheying
rhe solicitarions of other stimuli.
The dynamic tensions that enliven the image, rhe organization of its constitu-
ent parts, the rapport of color and of value that establish themselves between
elements, ehe suggeseed movements of the line, among ehe other parameeers, also
come into play when the goal is to attract or manage ehe gaze. The connections
of direction, of ehe gaze, and of the position rhat belong to the cinematographic
syntax find. certain equivalents in comics; eheir study emerges par excellence from
arthrology. !t remains that the spatial repartitioning of the balloons, which is an
operation specifi.c to comics, is a factor that weighs in a preponderant fashion on
the reading processes.
The cartoo~ist Altuna developed an original reticular apparatus that, from
one work to another, allows for certain variances. He has two constant traits that
can be seen on page 29 of the book lmaginaire (Dargaud 1988; fig. 8). The first
of these concerns the form of the balloons, which, as in Big Numbers, are invari-
ably circular. The second affects their location, which is almost always situated
on the periphecy of the panel, and as near as possible to ehe central vertical axis
of ehe page. (No balloon is sent toward ehe lateral borders of the hyperframe; on
ehe contracy, they are all brought toward ehe center.) The majority of ehe balloons
(in the example reproduced, nineteen balloons out of twenty-nine total) respect,
wich regard to ehe frame, ehe fourth type of arrangement described above: They
encroach upon the berween-images, or are even centered on them. This is why,
occasionally, Altuna seelts to expand this zone, as between the two final panels
of this page.
At first glance it appears that ehe balloons structure the space of the page.
Bound to each other, or sufficiently brought tagether to produce the impression
of a quasi-uninterrupeed chain, they vectorize the reading. The space of the bal-
loons does not imply that there is a narrator, but rather it suggests ehe possibility
of its insertion into the aggregative (molecular) structure thae ehe eotality of the
82 The Spolio-Topicol System
fig. 8. from lmagi111Jire (19881, by Horodo ~no. © Horocio Altuno.
balloons represent. Witness, in particular, the balloon situated at the bottom and
on the central axis (" WiUie doit te faire une bonne blague, n'est-ce pas, dis?"), whose
appendix has been exaggerated so that the speaker can be identified. Moving
chis balloon toward the left would have the effect of redirecting the reader's gaze,
The lpo~o-Topicollystem 83
while, on the contrary, the chain of balloons drives the reader toward the "exic"
ofthe page.
Ir is necessary to reexarnine, one by one, rhe six functions of th.e frame defined
above, in order to examine to what degree they also characterize the work of
the balloon. Thar the funccion of closure and rhe separative funcrion will again
be pettinent is already suppotted by rhe evidence. The oudine of the balloon •'i
i
delimies a homogeneous wne, a form derached from the background, and, when
several balloons are juxtaposed, or even overlap, they assure the perceptive dis-
crimination berween enunciarions that are physically near each other.
The rhythmic function is comprised of rwo aspects. lnitially, the posirioning of
the balloons in the space of the page creates a rhythm rhat is superimposed on rhe
heanbear that is produced by rhe crossing of the frames. Each text fragment rerains
some moment of our attention, introducing a briefpause in ehe movement rhat
sweeps across rhe page and the leaps that the eye affects in order to pass from one
balloon to the other give structure and rhythm to rhe reader's forward scanning.
The balloon can also, while gearing down, equip rhe text with a patticular
rhythm, with a breathing. Two phrases that succeed one another in the mourh
of the same speaker furm a unique enunciarion if rhey occupy the sarne balloon;
bur rhey are autonomized and become rwo diseinet enunciations if the atirhor
chooses, wichout so much as a change to rhe frame, to place them in rwo separate
balloons {often with a link berween ehern). Ladring the abiliry to entrust the
interpretation of his dialogue to voices, cernies at least enables this elementary
process of the divided reply, a practice rhar allows ir to inscribe a pause in the
speak.ing process. According to the enunciative context and the nature of the
given intencions, a specific nuance attaches irself to the contenrs of the second
balloon: it corrects, completes or reinforces that which was already said in the
firsr. A srring ofballoons (physically or implicidy) bound produces the effect of
an improvised discourse, as the character finds new ideas, supplemenrary argu-
menrs, or simply the suirable words.
Cerrain of the American "new comix" (such as, notably, those written by
Frank Miller) have esrablished this division of dialogue as a system. Ir would be
more accurate to speak of long monologues, which succeed one another or are
interrwined wichout truly replying to each other, characterized by a syncopated
writing, a panring, cur into very brief fragmenrs (a patt of the phrase more ofren
than a complere phrase) each of which has a dedicared balloon. The speech bal-
loon reaches to rake on the value of a punctuation mark. Ir adds something to the
period; instead of closing the Statement, it encloses it entirely.
84 The Spatio-Topical System
The connotations associated with a rext cut this way can vary from one
sequence to another. Commenting, in his Traitt de ponctuation franfaise, on ehe
style characterized by ehe profusion ofshort sentences and periods, Jacques Drillon
enumerated ehe principle effects that concern the process. Each fragment can take
"an affirmative, not ro say, peremptory character"; it can even acquire "an exclama-
tive power"; but elsewhere ehe feeling produced will be indifference, impassibilicy,
or lassirude; elsewhere agairr "ehe 'poinred' sentences have a comic effecr, or sug-
gest agitation." And also this suggestive remark: "One no Ionger knows what was
first, ehe brevity of sentences or rhe recurrence of the period, as one often ignores
what precedes ehe other, breathlessness or rhe accelerating hean beat. "6s Indeed, in
coffiics it is possible to wonder whether rhis choice of spatial occupation, moving
amid the proliferation of the balloons, precedes and orders the writing.
There is no doubt where Altuna is concerned. On page 29 of lmaginaire,
certain phrases appear to have been divided simply so as to restriet ehe balloons
to a pre-established diameter, or to occupy balloons drawn a priori. Look, for
example, at che five balloons hung one after che other that belong to ehe character
Ansdme, in the second paneL
What was previously said about the frame can be applied to ehe balloon: it
is always "the index of something-to-be-read." Bur it is not difficult to attribute
to che balloon a true readerly function, co the extent chat, contraty to ehe image
that somecimes seems innocuous and empty, a verbal enunciation is immediately
identified as a pertinent segment ofinformation for ehe intelligibility of ehe story,
wichout which che balloon would have no reason to exist.
As was noted above, it is rather with reference to the image that the balloon
provides assistance to ehe reading, concributes to ehe deciphering. !ts readerly vir-
tues merge with its descriptive power, that of attracting the eye to a barely visible
speaker in explicidy designating his position.
The structuring function is also turned toward the exterior of the balloon
(the images in which it is inscribed) rather than toward its contencs. lndeed, ehe
segmentacion of che cext imposed by the height of ehe balloon has no notable
consequence. If a text were cut in two, three or four lines, its comprehension
by che reader would not be affected, nor would ehe eventual appreciation of its
stylistic qualities. The balloon exerts no true influence on the writing of the text,
for che simple reason chat it is precisely calibrated wich reference to che number of
words that compose the enunciation. At most it can be conceded that there is an
implicit Iimit (impossible to determine wich precision) to che quantity of text for
which a balloon can be made ehe receptacle. An overlong discourse is necessarily
The Spatio-Topicol System 85
truncated into several fragments and placed in several balloons, chemselves dis-
tributed in several frames. But it is not so much che "object" of ehe balloon chat
imposes this Iimitation; it is a mechanism of the inherent regulation of the cern-
ies system considered in its totality.
lt is apparent chat che balloon contributes to strucruring the composition of
che imagewich which it sharesehe surface of che panel. !ts oudine, its dimension,
and its location irrform ehe Iayout of che image's different components-if only
to avoid masking a pertinent semantic zone. In particular, in the composition of
che image it is che respective arrangement of ehe characters and che balloons chat
is immediatdy conceived in an interdependent fashion.
I have saved the expressive function for the end, where not only is it revealed
to be as pertinent for the balloon as it is for the frame, but where it has aroused,
in humorous comics, an expansion of discoveries. I will not discuss here the entire
repertoire of chis particular rhetoric, which has already been weil commented
upon, in parcicular by Robert Benayoun's work from 1968. Insisting on the diver-
sity of enunciations that can assume che form of a balloon, Benayoun generared 'I
an empirieallist of seventy-two occurrences, amongst ehern the "censored bal-
loon," che "oneiric balloon," che "balloon ofillumination," the "papyrus balloon," I
ehe "poster balloon," che "atomic balloon," ehe "glacial balloon," and che "pierced
balloon"!66 ln exchanging its usual form for one of chese symbolic dress forms, che
balloon itsdf becomes iconic, affirming the commentary of the verbal enuncia-
tion that it encloses.
lt is important to add thac ehe position of ehe balloon wich respect to ehe
speak.er can, in certain cases, be revealed as expressive in itself. Referring to the
page &om Bout de Ia piste drawn by Giraud (fig. 7): in parreis three and four,
che character named Kdly, bending under ehe weight of a sack of stones, speaks
in balloons chat, since they are simared under him, appear to be made of Iead,
chereby reinforcing his overwhelmed body. The director of ehe Francisville peni-
tentiaty has lost his haughtiness; che location of ehe balloons allows us to see chis
decay, demonstrating that his manner of speaking is no Ionger arrogant.
1.1 0-0N THE INSET
Up to chis point che multiframe has been described as ifitwas necessarilycomposed
of juxtaposed frames, organized according to impermeable barriers. However,
che dialogue between the parreis &equendy passes through ocher configurations,
86 The Spotio-Topi<ol System
including those that find a frame welcomed wirhin one or several other frame(s).
This apparatus, which I will designate as the inset (incrustation), gives evidence of
the extreme suppleness that characterizes the management of space wirhin com-
ics. Ir opens up a large range of procedures in which the repartition of frames,
escaping from the relative automation of tabular compartmenralizarion (or,
anticipating a notion that will be defined later on, the logic of gridding), is more
direcdy dicrated by the semamic articulations of the story and fully patticipates
in the mise en scene.
There can be no question of establishing a complete catalogue of rhese proce-
dures in all their variations, as the narrative strategies and the rhetorics of authors
can be renewed and modified ad libitum. Bur it is at least possible to identifY the
largest principles that they appear to obey. And to note, from the srart, that the
inset is a figure in which the benefits are somerimes accrued to the base (inclu-
sive) panet and sometimes to the inset panel.
Benote Peeters defined the cernies panel as "an image 'in disequilibrium.'
inserted between that which precedes it and that which follows it, bur no less
berween irs desire for autonomy and its inscription in the story," and concluded
thar "comics rest, in each instant, on a rension between rhe story and rhe picture.
The story that, while including the imagewirhin a continuiry, stretches to allow
us to glide over it. And the isolated picture that allows us to fix upon ir.""
While reexamining this rich and suggestive alternative, I would say that the
inset serves the purpese of the pietute when it magnifies the backgrotind panel,
whereas it more dearly serves rhe story when its purpose is rhe contextualiza-
tion of the inset panel. In the first case, it allows itself to be reduced to a simple
superimposition; in rhe second, it puts in place a dialogic interaction between the
concerned panels.
When rhe inset is at the service of the indusive panel, the panel most often
represents a landscape, a large space, or a "background" where rhe characters
are depicted at a reduced size. This sort of panel frequently has for an a priori
function the establishment of the d&or in which the relared actionwill occur, or
to create an a posreriori disrance, in order to abstract from the action, leave rhe
protagonists, or condude a sequence. (These two functions are particularly vis-
ible in Franc;ois Bourgeon's trilogy Les Compagnons du mfpuscule, where a nurober
of chapters are opened andlor dosed wirh large panels that accommodate one or
more insets.) In such a case, rhe inset does not aim ro construct a specific link
between the panels; it is rather a consequence of the wish to offer the largest pos-
sible area to rhe indusive image, in order to magnif}r its decorative virtues (at rhe
The Spotio-Tapirol System 87
risk of occasionally adopting the aesthetics of a postcard). The other panels have
no other possible location except wirhin this enveloping image. I will give as an
example page 44 of the first volume of Cosey's book, Le Vryage en ltalie (fig. 9).
The view on rhe rock envelopes rwo syntagms, each composed of two panels; it is
like a common facrorizarion, and inscribes rhe entire page in a logic of rhe "pie-
eure" rhar corresponds ro rhe contemplative mood of rhese two characrers. The ' \1
tension between the stopped insraut that is depicted in rhe inclusive panel and in
il
the four other panels, rhe decomposition of an action rhar is inscribed in time, is
•II
·i~
notable. The result is rhar this succession of consecurive moments appears to be
summarized in rhe background panel, which, following from this fact, dernon-
srrares less a position than a synthesis of a route.
:!
Let us also note, wirh regard. to rhis example, rhar we can apply a number of
properries that we have described with regard to the balloon to the phenomenon
of rhe inser. I do not believe rhat it is necessary to insist on rhe effecr of con- .I
••. ·.·.11
Ii
~,
cealmenr rhar is produced by rhe inset panel, on the necessiry for rhe inclusive
panel to welcome the inser panel into a "desemantized wne" and, therefore, on ·.• •..
•I
the necessity to rake rhe location of rhe inset panel inro account at the moment il
where it elaborates rhe composition of the included image. All the same, Iet us ,I
highlight that, in this figure, the exterior frame of the inset panel is at the same
time rhe "interior frame" of rhe inclusive panel.
The second important option rhar can motivare an inser is the contextualiza-
tion of a panel (or a series of panels) and the highlighting of the privileged link
that it maintains with anorher semantically bound panel.
Victor Stoichita has long studied the splitting and the mounting of images
in paintings of the seventeenth and eighteenth centuries. The idea was often to
oppose a firsr plane to a background, under rhe form of an antithesis between
a sacred "text" (translated into images) and an "outsider-text" of a profane char-
acrer.68 Stoichira recalls that other "domains of representation" also experience
forms of "pictorial unfolding," particularly the theater. Similarly, in comics, the
rapport between the inclusive panel and the inset panel is ofren a tapport of
dialectic opposirion or of contrast, for example, berween a field and its reverse
shot, the gaze of a character and the scene or the place that he contemplates, an
inregrated narrator present at the action and rhe representation of his memory
(or ofhis dream, or ofhis fantasy), etc. The mostfrequent case, it appears to me,
is a dialectical relation berween the part and the whole, which places a view of the ,,
ensemble and of an element of the same scene in relation, separated and enlarged. ''il
The cartoonist introduces a framewirhin the frame in order to highlight a detail !I
.I
 88 U.. Spotio-Topicol System
Fig_ 9_ From Le VIJ}'Dfle en lmfm, volume 1 (1988), by Cosey. © Ediions Dupuis.
of his "picture," similai to the effecr of a cinematic zoom, beinging the rea.der
closer to the pertinent element (which could be, for example, the expression of a
character at the moment where he begins to speak). On occasion different details
are taken, the background panel welcoming several small inset panels, which
appear to spell out the ingredients of the depicted scene.
The Spotio-Topicol Sy5tem 89
Let us also retain another frequent modality of the inset, the establishment of
a relation of simultaneity berween two or more panels; this relationship represents
a pause in the flux of the temporal succession-the ordinacy regime of sequential
consecution between juxtaposed panels. To put it very simply, ehe inset translates
a relationship of the type meanwhile, when the traditional intericonic void is gen-
erally equivalent to a then."' On the page by Cosey, we can observe that, in ehe
case of ehe multiple insets, the temporal relations between ehe gamering of panels
concerned may be more complex. Indeed, it is not uncommon to find a succes-
sion of three or four inset panels "in accordance" in a background image; one
must then separately consider the internal temporality of this multipanel syntagm
and its relationship to the "moment" that corresponds to the including panel.
In comics, what fundamentally distinguishes the procedure of "pictorial split-
ting" from painring or the theater is, precisely, that it cannot be reduced to a
splitting, because the inclusive panellinset panel binomial is not bound to an
exclusive relationship but belongs to a multiframe, with which it continues to be
tied by multiple reciprocal determinations. The inset, in comics, is only a local
phenomenon. Wiehin a multiplicity of images characterized by different Ievels
and degrees of iconic solidarity, it is content to instimte or (more frequendy) to
highlight a_ privileged relationship becween two terms. These relationships, how-
ever, ask to be read and interpreted in taking account everything that, upstream
and downstrearn, can index or echo it.
As it constructs~ accompanies~ or highlights a semantic relationship between
two units of discourse, the inset supports arthrology, and can be described as an
arthrology auxiliacy figure of the breakdown. The inset, however, is rarely indis-
pensible to ehe establishment of this relationship, which a simple contiguity of
panelswill almost always suffice to make intelligible-ehe discontinuity and the
ellipse are, afret all, constitutive of the language of comics. Wich regard to the
breakdown, ehe inset is generally a form of taising the stakes. But, in the sense
that it introduces a remarkable gap in the spatio-topical ordecing of ehe multi-
frame, it also fully participates in the page Iayout. lt is in this regard that Philippe
Druillet, nocably, has made a significant contribution.
Alongside Cosey, Derib, Hermann and Andreas, Fran<;ois Bourgeon is one of
the French authors of whom the work is characterized by a massive and concened
· use of insets, at fitst placed in the service of the narration. But Bourgeon's Iayouts
are also characterized by the generally elevated number of panels that comprise
each page. He is certainly one of the cartoonists for whom the tension between the
picture and the Story is resolved from above, in thesensethat his aesthetic seeks to
90 The Spotio-Topicol System
reconcile "novelistic" density wich a specracular dimension, which occurs rhrough
a constant renewal of ehe page Iayout and by ehe repeated use of !arge images, of
sertings rhat abound in details. The recourse to frequent and multiple insets is
imposed as a natural solution: because of ehern, rhe decor saves irs prerogatives
wichout emptying ehe stoty of its substance. Bourgeon also manifests a great lib-
erty in ehe location of the text; ehe balloons frequencly pass through the Iimits
of ehe panel's frame and are sometimes exiled to ehe exterior of ehe frame. He
achieves this so well that we can see, in rhis aueher who possesses an acute aware~
ness of spatio-topical resources, that parreis and balloons are, equally, spaces
where ehe encroachments, enclosings, and other Iayouts compose a remarkable
mosaic that makes ehe best use of the entire area wirhin ehe hyperframe.
To shed light on ehe preceding, a question can be asked about the Iimits that
are necessary to recognize about ehe ofren-postualted (not wichout some basis)
analogy between ehe inset, as it is known in comics, and ehe figure of ehe insert in
cinema. According to Christian Metz, ehe insert is defined wirhin the "cinemato-
graphic syntagmatic" as an "autonomous segment, which contain~ but one shot."
The inserts "owe their autonomy to their Status as syntagmatic interpolations"
and are shared out in at least four categories.'' I borrow from Roger Odin ehe
summaty of this typology:
The non-diegetic Insert: oshotwithouto direct relotionshipwith the oction butwhich hos osymbolicvolue .. _;
the subjective Insert: memones, premonitions, dreoms; the disploced diegetic Insert: ounique shot of the
pursued in the midst of ogroup of shots dedicoted to the pu5ue5; ond the explicotive Insert: oclose-up
shot on the text of oIetter."
For a singular shot to be. autonomous, says Metz, it is necessary that it "pres-
ents an episode of the plot," but, to be more specific, this autonomy "is not an
independence, since each autonomous segment derives its final meaning in rela-
tion to ehe film as a whole" (p. 125).
This concession delights me, but I am not certain that it is necessarily sufficient
to the inset in comics. First of all, because several panelswill often be necessary in
order to produce an enunciation equivalent to that of a single cinematographic
shot (the equivalent· of a subjective insert, for example). Second, because ehe
autonomy of the inset panel is lesser than that of the insert in the sense that
the comics image, contrary to the filmic image, is offered to the gaze only in a
situation of coexistence with the surrounding images. The insert is framed by
two other shots according to a temporal order of succession; the inset image is
The Spotio-Topicol System 91
spatially enveloped by ehe inclusive image, that which signifies at once: a) thac the
same image borders it on all sides---ru1d not rwo different images--and b) that ehe
I
rwo panels of ehe binomiallend themselves to a simultaneous vision, as weil as
I
I
to ehe coming and going of ehe gaze, to an inrerval of attention. Finally, withouc l
having to question "comics as a whole," it is primarily the entire page, as a mul- i~
i
~
tiframe in which ehe inset represents a local phenomenon, that is ehe extremely I
·['
feeund first instance of interpretation. I
Let us finally note that, since no real internal (measurable) duration can be
assigned to a comics pand, the "scene" represented in the inset panel is not, fun-
·i
darnentally, eieher briefer or Ionger than that which is depicced in ehe inclusive
pand (where the cinematographic insert attaches itsdf generally to a connotation
of brevity: lt takes place--it is interpellated, as Metz says-berween rwo more
articulated syntagms).
1.11-0N THE PAGE LAYOUT
Amongst ehe diverse operations that assure ehe integration of ehe components of
a comic, that which particularly has ehe function of governing ehe spatio-ropi-
cal pararneters is commonly designated under ehe term "page Iayout." We now
have ehe complete catalogue of ehe questions that arewirhin its jurisdiction. We
have seen that it is not just a matter of proportional and positional relationships
berween ehe panels, but also of ehe perceptive and pereinem degree of autonomy
to ehe space of ehe strip and of the nerwork composed by ehe balloons "over" or
"within" ehe drawing of ehe frames. The page Iayout also determines ehe borders
of ehe hyperfrarne-its borders can be continuous or broken, and can also be
crossed by panels that are purposely adrift-as weil as ehe consistency ofits interior
space (depending on whether ehe panel's frames distribute it among themselves
in its totality, as is commonly the case in Franco-Belgian comics, or, as observed
in several modern Arnerican comic books and in mangas, whether they care-
fully place white spaces, areas left blank-ehe "background" blending itself more
closely wich ehe images). Finally, the Iayout can be exerted in an autonomous
manner for each page, or it can take account of the natural diptych that consti-
tutes pages destined to be printed side by side.
This is, briefly recapitulated, ehe entirety of ehe pararneters that govern this
fundamental operation. lt remains tobe understood which areehe principles and
ehe ends that guide ehe cartoonist at ehe moment where he makes ehe Iayout of

92 The Spafio-Topical System
ehe page. What does he wish to attain through ehe options that he takes wich
regacd to ehe distribution of rhe spaces?
Let us begin by highlighting chis evidence: The page Iayout does not opecate
on empty panels, but must take into account their contents. It is an instrument in
rhe service of a global artistic project, frequencly subordinated to a narrative, or, at
least, discursive, aim; if it submies a priori to some formal rule that constrains the
contents and, in a certain way, creates ehern," ehe page Iayout is genecally elabo-
rated from a semantically determined contenc; where ehe breakdown has already
assured discretization in successive enunciations known as panels. However, the
page Iayout cannot be defined as a phase that follows rhe breakdown, wich rhe
mission t<> adapt it to the spatio-topical apparatus; it is not invented under
ehe dictation of ehe breakdown, but according to a dialectic process where ehe
two instances are mutually determined..
To provisionally leave the essential question of rhis interaction in reserve,
we are, at present, not prevented from defining the three specific principles of
ehe page Iayout, or rhose that it can or must observe "independendy" from that
which rhe page will signifY, those that reduce rhe arrangement of ehe frames to
what it was at first: a spatio-topical configuration. I cannot do otherwise at this
point but adopt what I said in 1990 in my anicle "Du Je au 9e art: 'inventaire
des singularites'" ("From ehe ych to rhe 9th An: 'lnventory ofSingularities' ") 73 to
demonstrate chat ehe frames must be drawn and arranged in such a way rhat:
r. They respond to compatible options. lt is impossible, for example, to develop
(orher than by a full-page illustcation) ehe coexistence of a panel that occu-
pies ehe entire widrh of ehe page wich another panel that will be stretched
over its entire height. lt.is necessary to choose, and each choice of a frarne
restricts ehe range of possibilities for others.
2. They carefully deliver to ehe reader a route devoid of ambiguiry, suggesting a
direction to ehe reading (except in knowingly confronting ehe reader, by an
effect of narrative deconstruction, with several contradictory options).
3- They obey, eventually, a principle of global composition, more or less osten-
tatious, which submies them to an aesthetic order.
The term "eventually" marks ehe fact that, if ehe first two principles impose rheir
law on ehe entire Iayout (rhe first by material imperative, ehe second rhrough a
concern for intelligibiliry), rhe third principle is merely optional and contingent.
l
The Spofio-Topical System 93
l.l Ll The Typology of Benoit Peeters
French specialists agree rhat a decisive step in thinking about rhe Iayout of ehe
comics page was talren wirh ehe publication of Benoie Peeters' study entided
"Les aventures de Ia page" ("The advencures of ehe page")_ In this text, Peeters
distinguished four conceptions of ehe page, respectively designaced as conven-
' ~i
tional (where rhe panels are "of a strictly constant format"), decorative (where
"ehe aesthetic organization prizes every other consideration"), rhetorical (where
"rhe dimensions of ehe panel submit to ehe action rhat is described") and, finally,
productive (where "it is ehe organization of ehe page that <ippears to dictate
rhe story")_
This extremely stimulating text raises excellent questions, but does not always
ask them in ehe most appropriate terms. In practice, the identification of the four
suggested categories encounters sevecal difficulties.
lndeed, if a number of pages obey a Iayout that precisely responds to one of
rhe suggesred categories (so rhat ehe page by Teule reproduced above [fig. 6] is
conventional, while the pages by Cuvelier and Muiioz that I will comment on
in ehe next chapter are both rhetorical), others, in a non-negligible quanticy,
respond at once to sevecal ofPeeters's definitional categories. Several of ehe pages
rhat we have examined up to this point exemplifY this ambivalence. Let us talre,
first of all, chose by Baudoin (fig. 2). They can be termed rhetorical, in ehe sense
that ehe panels which enclose a "medium close shot" or a "close-up" are, on the
whole, smaller than those where ehe characters are framed under ehe knee. But do
they not also produce a decorative effect, by virtue of the mirror effect analyzed
previously, which malres ehe second page a double inverse of ehe first? An analo-
gaus doubt applies to ehe two pages by Del Barrio (fig. 3). Considered separately,
they are each conventional, since each is composed of three identical frames.
Neverrheless, putting ehern side by side produces an undeniable decocative effect
if one prefers, a pictorial effect, where the attention is immediately drawn
hy rhe swing of ehe horiwncal axis and ehe vertical axis. Finally, if one recalls
rhe signifieds that we have deciphered in chese pages (how, in ehe course of life,
this story provides us to read, ehe accent is moved from a quasi-unlimited
c\horiron roward an inescapable end), it is necessary to admit that rhis Iayout also
rhetorical qualities.
The page by Aleix Barba (fig. 5), built according to a conventional nine-panel
also presents a productive aspect, because of the arbitrary division of each
into three panels, aiming to slow down the reading by separately framing ehe

94 The Spotio-Topirol System
portions of the field offered to the view. As for that of Cosey (fig. 9), it appears to
me to be both rhetorical and decorative at the same time.
The frequency of similar cases, where several categories are confirmed wirhin
ehe same Iayout without contradiction and wichout the possibility of separation,
obliges us to conclude an aporia in the nonetheless seductive typology of Benolt
Peeters.
He must, it seems to me, address hirnself to three critiques. 1) Educared by
his experience as a creator-Peeters is not orfly a theoretician, but also a script-
writer-his theory doesn't sufficiencly take into account the perceptive and cog-
nitive schemas that are broughr to the work on ehe pan of the receiver; 2) Despite
appearances, it does not hold the initially stated methodological promise: that
of not separating the "choice of organization of the frames and the planes," on
the one band, and tbe "signifieds of comics," on the other band, in order to
"not destroy the quasi-organic coherence that gives each element its reason for
being";" 3) Tbe definition of the page Iayout to wbicb Peeters refers is at the same
time too vague and too resrrictive; it remains completely silent on important
aspects of tbe spatio-topical system, sucb as the consistency of the page (in the
sense defined above), tbe degree of autonomy of the strips,75 the spatial reparti-
tion of tbe balloons, and tbe process of the inset.
lf the Iayouts rhat are given to comics are, in their parricular configurations,
quasi-unlimited in number, a fundamental division remains: eieher the frame
varies &om one panel to the other (in form and dimension), or the &ame does
not vary at alL Ir is not possible to imagine a third term in this binary alter-
native. This results in an opposition of the "conventional" Iayout against all
of the others, and a zero degree Iayout will therefore begin to be seen in this
"convention."
Iris noteworthy, however, that the kind oflayout that is statistically dominant
undoubtedly corresponds to wbat Peeters terms tbe rbetorical Iayout. Indeed,
it is this model, applying nothing other than this resource in an often intuitive
and almost automatic manner, that distinguishes comics &om cinema and which
allows it in eacb instant to redefine tbe forrnat of its &ame. The rhetoricallayout
is more common because it is the most convenient and the supplest form, and
because it is entirely ·at tbe service of the srory, it is perfectly adequate to the
narrative project that most comics authors follow. Jacques Samson pertinencly
summarized this adequacy: "Tbe elasticity of tbe representation (variable &ame)
maximally 'diegetizes' the figurative space and erases all impressions of spatial or
temporal discontinuity. "76 I do not believe that it was by accident that Rodolpbe
l I
The Spolio-Topicol System 95
Töpffer, in whom an intuitive comprehension of rhe medium was attained to an
astounding degree, bad specifically opted for an apparatus oftbis type.
Ir has not been sufficiently highlighted that the rbetorical Iayout, because it
offers no resistance to the cartoonist (except the obligation to organize the com-
patibiliry of the &ames, to find the best compromise between local solutions, tbe
constraints of the support and the whole configuration), can lead to a type of
automatism and, finally, can prove tobe too easy. With a bit of practice in comics,
one quickly learns to develop an intuition for tbe &ame that appears appropriate
to each image: a narrow panel for a single character on foot, a large panel for a
group ofcharacters, a landscape, an action scene, etc. These spontaneaus solutions
give rise to a certain banaliry, and it can sometimes be morefeeund to opt for a less
anticipated solution, to thwart tbe "partial predictibiliry" oftbe form dear to Gestalt
theory; for example, the idea of drawing a panel-bandeau for a single cbaracter,
to create what appears as waste of space or a fault offraming, can Ii berate the space,
creating an interesting respiration from the plastic or rhythmic point of view.
The rbetoricallayout is not only the most common; of all the possible Iayouts,
it is also that wbicb is least remarkable, tbat wbicb tbe reader immediately accepts
as natural (even if this discretion of the paginal apparatus of the whole can be
locally disrupted by a remarkable effect, like the suppression of a &ame or an
inset, for example).
Tbe page Iayouts that more or less cleanly depart from the rbetorical model
are perceived by the reader as one of the particular apparatuses, and assimilated
specifically to remarkable deviationswith respect to the implicit normthat consti-
tutes the rhetorical conception. In most cases, this perception is notably applied
to the Iayout that Benoit Peeters calls "conventional" and that, because it is not
so banal, I prefer t0 qualifY as the regular Iayout. Indeed, when the page obeys
a uniform compartmentalization the ortbogonality of the grid is a trait that is
remarkable and to which a strong productive effect is often attached, because it
allows the page to activate and to adopt certain potentialities of the multi&ame
"colored and coloring, comparator and substitute . .. , mutational and permuta-
tional," to use Henri Van Lier's terms.n The reader-I am thinking here of the
competent reader, someone wbo is sufficiencly familiar with the language of com-
ics--notices that tbe work obeying this Iayout "is deprived of a certain nurober of
real elements placed at their disposal,"78 in thesensethat it re&ains &om playing
with the potential elasticiry of the &ame; and, knowing that this renunciation is
voluntary, he can also usually detect, at first glance, tbe strategy or the benefit
that, on any other Iayout, inspires this choice.
96 The Spotio-Topicol System
1.11.2 ADefense and Illustration of the Regular Layout
With regard to the rwo principles chat musr necessarily govern the Iayout of a
cernies page, the choice of a regular Iayout immediately eliminates all difficulties.
The frames are more than compatible because they are identical and no doubt is
introduced to the reader with regard to the order in which the panels are linked.
lf the regulat Iayout was chosen only for this convenience, it might reflect a
cerrain laziness on the parr of the cartoonisr,' an abandonment to a mechanical
situation. In fact, this Iayout has not found great favor amongst ehe theorisrs. In
the first version of his study (1983), Beno!t Peeters spoke of a "mutilated instru-
ment," emphasizing that some comicssubmit "to an imaginary constraint, inher-
ited from another discipline" 79---cinema, of course, where rhe size of screen stays
constant. This notion of mutilation did not reappear when rhe textwas rewritten
in 1991, where Peeters developed rhe advantages that he wanted to attribute ro
this layout in certciin cases.
The most interesting uses of this principle ore those thot ... push this constoncy of the frome in order
to ochieve o kind of "stotic shot" unwinding on the poge. The humorist cortoonists like Schulz, Feiffer,
Bretecher, Wolinski or Copi hove provided remorkoble exomples of sequences where the slightest
modificotion of the regulority of the whole tokes on oconsideroble volue, fovoring otrue entronce into
the drowing.
Even oreolist odventure cartoonist like Hugo Prott is often served by this type of opporotus . ·.. , which
concentrotes the ottention of the reoder on some minimal modificotions in the oction ond the ottitudes.
(p. 38)
In the secend edition of Case, planche, ricit in 1998, Peeters appeared even
better disposed to the consid~ration of the regular Iayout. But those who, in the
interim, have adopted his categories have not demonstrated a great enthusiasm
for this kind oflayout; it is in this concessionary mode that Baetens and Lefevre
admit that "the mechanical allure of the conventionallayout does not necessarily
exclude subtle and ingenious solutions." I believe that this understates the mat-
ter, and .that it is time to rehabilitate the regular orthogonal grid as a model that
is at once rigorous and which offers a number of interesting possibilities to the
breakdown and to the braiding. The argument advanced by Peecers is extremely
streng; the fact of forcing the frame to observe the greatest neutraliry allows it
to elevate other parameters whose visibiliry and, therefore, efficacity would be
considerably reduced if the Iayout were given over to useless fantasias. lt can be
The Spntio-Topical System 97
seen, for exarnple, in the play of physiognomy in Bretecher: it is evident that the
differential value of every shrug or frown of the eyebrow is accentuated when no
other element is changed.
Moreover, far from being the exclusive prerogative of creators of sketchily
drawn satirical comies, the regular Iayout has ofren characterized ambitious and
innovative works, from PoUy and Her Pals by Cliff Sterrett to Fires by Mattotti or
even Watchmen by Moore and Gibbons. With regard to Watchmen, Baetens and
Lefevre have analyzed the different checkerboard parterns that drive the intersec-
tion of rwo alternating and differentially colored scenes.So The rectilinearity and
the perpendiculatity of the interpanel gutters creates an effect of a "syntagmatic
nerwork structured at the base of alignments." I borrow these words from a com-
menrary provided by Groupe Mu on a picture by Vasarely, a commentary that
applies here word for word:
Eoch posttion [eoch poneij is ot the intersection of ohorizontol ox~, of overticol oxis ond of ot least two
diagonal o~es. There is then, for eoch of them, four possible syntogms. (four ot least, for besides, each ;;
position con enter into more complex figures thot will moke remorkoble positions: comers, center, high r:.
il
ond low poles, etc.)" !i
i!
A regular page Iayout also inspires certain braiding effects because it allows for the
;
mostsimple and striking way of organizing things from the point of view of per-
ception, and because it strengthens the bonds bervveen predetermined locations
(cf. 3-3). Finally, it possesses the ultimate virtue ofhandling the possibiliry of sud-
den and spectacular ruptures from the initially given norm. In a book in which
all the other pages ate regular, a page that is suddenly distinguished by a special
configuration carries an extremely streng impact (the example of the double-page
spread situated at the center of the fifth chapter of Watchmen comes to mind).
When all the panels are distinguished by different formats-as is suggested by the
rhetorical option-it is more difficult to have one really stand out.
1.11.3 New Propositions
If one wanrs to describe and to analyze the Iayout of a cernies page, I think that
one must begin by responding to two questions:
1. Is it regular or irregular?
2. Is it discrete or ostentatious? i
I
98 The Spotio-Topical S)"tem
One cannot reduce the terms of the :first alternative ro rhose of the second, for
the strict regularity of the frame is not the only factorthat determines the layout's
degree of visibility. I have said that the regular Iayout has been perceived, in the
most cases, as a variationwich respect to the norm, which is rhetorical. This per-
ception permits exceptions-I am thinking about cerrain pages of Gaston or of
Achille Talon, for example--and above all admits different degrees. When all the
frames that compose a page observe a consrant format this Iayout is all the more
rernarkable when: I) the number and/ or the form of the panels reaches to stray
from the norm (toward a great number of panels, as in Joe Matt, or vety length-
ened panels, as in certain pages by Andreas or by Regis Franc, which give the page
a "gaudy" Iook because it is so atypical); 2) the benefits derived from this regular-
ity of frames, at the plastic or narrative Ievel, is itself remarkable. The intertwined
chromatic series and the checkerboard effects of cerrain pages of Watchmen high-
light the principle that governs the Iayout of the whole work but which, in other
pages, remains more discrete. 8:!. Also, the "static shors" to which Peerers makes allu-
sion raise the stakes with regard to the regularity of the frame, since it is the panel
itself that is repeated in its (quasi-)totality, bringing a serial aspect to the page.
If the regular Iayout is ofren striking, conversely, certain Iayouts that Peeters
puts in the "decorative" category are excessively discrete. Because the symmetry
that governs a number of E. P. Jacobs's pages is not truly perceived, unless one
teduces the page to a schematic diagram made of empty panels, as Peeters does. In
the album, thedensity of the images, the relative homogeneity of their· formats,
the profusion of the text, and the work of the coloring screen the perception of
this particular ordering of the panels; the Iayout is, because of this discretion,
simply perceived as irregular.
The possibility for the pag~ Iayout to become ostentatious, that is to artract
attention to some remarkable quality of itself, notably corresponds to the third
optional principle, previously noted: that is, to obey a panicular aesthetic
imperative.
The combination, two by two, of characteristics designated here as fundamen-
tal, allows the definition of four categories of Iayouts:
I. regular and discrete
2. regular and ostentatious
3· irregular and discrete (which corresponds to the dassie "rhetorical" Iayout)
4· irregular and ostentatious
The Spofio-Topitol System 99
When one encounrers a Iayout deemed ostentatious, it is necessarily oppor-
tune to interrogate it, in a second Ievel of analysis, about the motivations that the
cartoonist has obeyed in the elaboration of the page. In order to proceed in this
evaluarion, one must necessarily compare the Iayout of the page to irs iconic and
narrative contents, or even, in cerrain cases, to the artistic project that animates
the whole of the work. Only this confrontation permits one to decide whether
the Iayout obeys decorative-by which I mean that it expects an aesthetic benefit
by the single exhibition of its remarkable aspect, and does not respond to any
exrerior morivarion-rherorical or producrive ends. Thus, an extremely divided
page Iayout, one that is rowdy or dislocates the frames, will appear as rhetorical
(and not as decorative) ifit accompanies, highlighrs, or "translates" a chaotic situ-
ation, a breathless high-speed chase, an alcoholic crisis, or an illustration of the
madness of the protagonist.
The typology proposed by Beno!t Peeters followed from the initial opposition
between the story, "that, including the image in a continuity, stretches to make us
slide across it," and the picture "that, isolated, allows one to fix on it" (op. cit.).
The four conceptions of the Iayout that he offered are defined by the fact that
the story and picture are or are not independent, and the assumption that one
dominates the other.
The authority of the stoty and the picture appear to me too general to permit
a fine analysis of the play that governs the operation of the page Iayout. I never-
theless maintain that the pages in which the pieeure dominates are those that I
qualifY as ostentatious, that is, those in which the Iayout immediately imposes
on the reader's perception, while the story dominates the pages where the Iay-
out is discrete, and that rather follow the law of breakdown. To borrow another
vocabulaty, one could also distinguish between pages that are "predominancly
configurational" and pages that are "predominancly sequentiaL " 8'
In order to give an idea of the complexity of these questions, I will allow
myself to reuse the analysis that I have already produced elsewhere84 of Watchmen's
double-page spread, to which I have already made reference.'>They are pages four-
teen and fifreen of the fifrh chapter of Watchmen, enticled "Fearful Symmetty."
These are the central pages of the chapter, which itself torals twenty-eight pages;
they are printed face to face and mutually reflect one another. This mirror struc-
rure is highlighted in several ways: by the inhabirual character of the form and
of the disposition of panels in these two pages, which rupture the regular grid
that the rest of the work obeys; by the !arge Ietter V, which the two halves sym-
metrically share over the two joined pages; by the effect of the link between the
100 The Spotio-Topicol System
two cencral (almosc) conciguous pands, which host the two halves of a single
drawing; finally, by the fact that the characters are themsdves benc over a pond,
a reB.ecring surface thar relays ehe notion of division.
Thus alerted, the reader faced wich this arrangemenc can only be perplexed.
Indeed, ehe aesthetic of Watchmen is in no way decorarive or ostentatious; on
ehe contrary, everything in this work is sacrificed to ehe imperialism of an over-
written and minutely arranged story. The hypothesis of a simple srylistic effect
having been excluded, and ehe eitle of the chapcer having lenc a hand, ehe perspi-
caciry of the reader leads him to discover that the symmetry is excended in realiry,
in a discrere Fashion, to ehe whole of ehe chapter, the pages simared before and
alter the cencral double page spread are reflecced two by two, from a distance,
wich regard co the arrangement of the frames. The readerwill again find a supple-
menc in chis analysis that stems &om the proliferation, throughout the work, of
the graphic dements that also obey the principle of symmetry: the mask worn
by the character of Rarschach (and the cards bdonging to the cest of the same
name), ehe "smiley" badge, the atomic symbol for hydrogen engraved on the fore-
head of Dr. Manhaccan, the dial of ehe clock where the hands counc to midnighc,
the Owl's spaceship, the Burgers 'n'Borscht sign, the skull and crossbones on ehe
black flag of the pirate ship, and so on. Finally, this insistence on symmetrical
motifS--which compose a veritable network, providing a grid upon ehe whole of
the story--can, in the final inscance, unlock a global symbolic incerpretation of
the work. Thus; ehe symmetry becomes an abstract category that nocably addresses
the rdationship of man and woman, those of the superpowers, and a judgment of
moral equivalence between ehe criminals and the heroes, since they use compa-
rable methods-the category, in sum, that binds and allows us to think through
the majorehernes of Watchmen.
Evidendy a double-page spread that appears to be governed by a superficial
graphic or decorative effect can be revealed to obey, in its page layout, much more
profound motivarions.
In looking to rectifY that which does not appear to be sufliciendy operative in
the typology ofPeeters I have not truly substituted another-my system is more
open-so much as I have slighdy modified the approach to ehe problem. The
four options described by Peeters were given as proceeding from cancerred anis-
tic choices, but their use as analytic caregories can only highlight ehe receiver's
appreciation of the (sometimes problematic) equilibrium realized in each page
between the story-effect and the picture-effect.
The Spofio-Topirol System 101
In a more specific way, ehe analysis suggested here mobilizes in turn a rigor-
ously objective criteria: the regularity or irregularity of the frames; a criteria that
is partially dependent on ehe subjective appreciation of the receiver: the ostenra-
tious character of the layout; finally, a third criteria that, symmetrically, makes
an intervention into ehe intentionality of rhe author, and which is the motivation
rhat is susceptible to justify the oprion retained by its correlation wich the iconic
and narrative conrents.
The other correction rhat appears essential to me is to distinguish, in the
evaluation of ehe third criteria, between the globallevel of the page and the local
level of a particular panel, of a particular strip or syntagm. Since &equendy ehe
rhetorical or decorative value of the &ame (which is nothing but the particular
applications of what I have called the expressive fonction) only affects one unit or
a subgroup of ehe page, at ehe same time that the paginal apparatus, in its general
conception, is relarively neutral. The suppression of a frame, ehe transgression
of the hyper&arne, ehe widening or the blocking of an intericonic space, the
particular location of a group of balloons (in a string, in a cross, in a circle, in
archipelago), or ehe recourse to an inset are examples of relevant local phenomena
of the page layout, in which the rhetorical or decorative pertinence--but also
reflexive, rhythmic or other-musr be assessed separarely.
The rerm editing (montage) is encountered sometimes in studies on comics, in
order to designate either the layout or a sort of compromise between the Iayout
and ehe breakdown (which are not always conceptually distinguished). Before
closing this chapter, I would like to object to the use of this term, which has
been abusively borrowed from the lexicon of ehe seventh art. I will invoke no
more than two reasons, each suflicient: I) the linkage of shots in a film, which is
properly the work of editing, carries itself out in a single linear dimension: that
of time, while the panels of a comic are aniculared at once in time and in space;
2) editing is an operation that cakes place alter the filming, and it consists of an
intervention on a material that has already been elaborated; the page layout, on
the contrary, generally is invented at the same time that the drawings are real-
ized on the paper, or even before the scenario is drawn. (At most one can talk of
re-editing [remontage] when a newspaper comic is revisioned with an eye to its
publication in book form-let us think about the work accomplished by Herge
and his collaborators on a number ofTintin's adventures-or when an album is
"adapted" imo pocket book format.)
102 The Spotio-Topicol S)"'em

Jan Baetens has very accurately described this difference, necessary to the
comic and to the photo-novel:

For the comics creotor, the ininol problern consists to divide o blonk poge. For the director of the photo-
novel, the first difficulty is to moke oselection of the ovoiloble photos ond to best combine them within
the Iimits of the poge. Thot one determines its portinoning, this one determines its octivity os ofunction of CHAPTER TWO
colloge. Thus, the perspective chonges completely8 '

Ir is all ehe more regrettable to see the same author, in associarion wich Pascal
LefCvre, cede to ehe confllsion in ricling a chapter of his essay on comics: "The RESTRAINED ARTHROLOGY
editing of the story." 87
For this chapter, my conclusion will be: Iet us leave ehe editing to rhe cinema The Sequence
(and to the photo-novel) and fasten ourselves to the study of the page layout-
which the cinema cannot do.
He, who moves in the onologue, con do everything except prevent himse~ from bouncing ond jumping
from one imoge to the other.
-Henri Michaux, Forons d'endormi, forons d'eveille

2.0.

"Panel, page, story": if the theoretical presuppositions are not entirely the same,
the course proposed in these pages for a global comprehension of the comics
system is analogaus to that followed by Benoit Peeters. We have looked at the
functions of the frame and the parameters that define the panel in terms of form,
then how the Iayout configures the spatio-topia in support of a narrative and
artistic project. Ir is now necessary to analyze how ehe "narration" moves through
and contribures ro this system, as well as come to terms with how the dialogue
produces meaning between panels. For the moment we will undertake an exami-
nation of restrained arrhrology, that is to say, rhe linear sem.antic relations that
govern the breakdown.

2.1 -REGARDING THE TM RESHOLD OF NARRATIVITY

Immobileimages separated by gurrers: how do we tell a story with these things? Is

the narration in the images? Is it dispersed between each image, or does it emerge
:.J
.I
104 Restroined Arthrology: The Sequence
from being arranged end to end? Does the intericonic gutter have a symbolic func-
cion? What part does the text play in ehe production of meaning? These are just
some of rhe questions posed to rhose of us who want to theorize the operation of
rhe breakdown, in the same way as they are posed, at least intuitively, to the areist
who wishes to translate ehe story that is in his head into a sequence of images.
Film theorists have sought to define ehe threshold of narcativicy, and I will
open chis chapter wich a brief summary of their conclusions. If every theorist
recognizes the decisive role of editing in narratioh, opinions diverge as to whether
a single image can itself be considered narrative. Christian Metz responded
negatively:
[T] he photo is so incopoble of normting thot when it wishes to do so, it becomes cinemo. The photo-novel
is not odenvotive of the photo but of cinemo. An isoloted photo con not norrote onything; thot's for sure.
But why must it be by some stronge corollory thot two juxtoposed photos ore forced to normte something?
Moving from one imoge to two imoges is to move from imoge to longuoge. 1
Roger Odin has challenged this conception. Borrowing from Michel Colin
ehe hypothesis that "the reading of ehe filmic image (and more generally, ehe
image icself) is, in our culture, vectorized like written discourse, from left to
righc," Odin aflirms:
Afixed imoge con certoinly hove o normlive structure: it suffices thot the vectonzotion corresp6nds to
on octontiol structure of o conflicting type between o subject ond on onti~ubject or of o relotionol type
between osubject ond on object of desire. 2
Andre Gaudreault3 tak:es int<? his account the two "story principles" announced
by Tzvetan Todorov, 4 those of succession and transfonnation: "What could, indeed,
be considered as narrative ... are all utterances that relate actions, gestures or
events that have berween ehern a 'relationship of succession' and that develop 'a
cappon of tcansformation.'" He recalls that, for us to be able to ralk of transfor-
mation between two phorographs, for example, their reconciliation must "affirm
both resemblance and difference." Consequencly, to observe that "transformation
(in ehe sense of modification) can, to a degree, be considered as ehe single and
unique condition of narrativiry since, being by definition a process, it already
implicares succession." Gaudreault concludes that, wich respect to cinema, the
necessary condition of narrativiry resides wirhin movement. Moving im.ages will
always be "ready-made narratives," whatever rhe "degree of the structurization
Restroined Arthrology: The Sequence 105
of the action that they present." Bur they would only enact a narcativicy rhat is
"native," "spontaneous," or intrinsic (that is, "direcdy linked to subjects ofexpres-
sion"), ro which can be added a "second layer" of narrariviry, one that is extrinsic,
based on editing, and therefore on the arrangement of "narrative contents."
I have simplified the terms of these technical debates in order to retain only
rhat which is useful here, that is to say, applicable to comics. The images that
interest us are not moving images, but fixed images. According to Gaudreault,
we can only ralk about their location in terms of extrinsic narrativity: narration
is born ftom the articulation of its contents, but it cannot be found inside each
image (even in the "native" state), whereas for Odin, rhere is narration in the panel
irself, which represents a pertinent actanrial content, appropriately vectorized by
the composition of rhe image.
Truthfully, ifwe recognize its validiry, rhe notion ofvectorization surely applies
better to comics than to cinema. There are rwo reasons for this: on the one hand,
the panel is fixed in rhe sense that ics reading is not impeded by ehe internal
movements 'of the image; on the other hand, the panel participates in a multi-
framewirhin which, at the level of each strip, ehe succession ofimages is explicicly
vectorized from lefr to right. But if rhe act of reading obeys an obligatory direc-
rion (having already observed that several page Iayouts render rhe circulation of
the gaze much more complicated, or uncerrain), it is sufficient to take any ran-
dom comic to verify that, in rheir very composition, the vast majority of images
are not vectorized, whether their contents were simply deemed not vectorizable,
or whether their narrow and venical format blocks rhe vague impulses toward
lateral explorarion. Consequencly, we cannot resolve the quesrion of the panel's
internal narrativity on the basis of this particular criterion.
But the opposition berween the rwo categories of moving images and fixed
images is assuredly too crude, even though it is true rhat each permies an abun-
dance of semiotically and aesthetically differentiated images. Thus, as we will see
later, comics lean toward a work of narrative drawing, and irs images generally
presenr intrinsic qualiries rhat are not those of rhe illusrration or the picture.
The question of an intrinsic narrativity to the image heckans us less direcdy
than it preoccupies the theorists of the seventh art. The co-occurrence of panels
wirhin the mulriframe, their simtiltaneous presence under ehe eye of ehe reader,
and also the visibilicy of the intervals between these panels, rhat is to say, rhe
locations where their symbolic articulation is carried out, fUnction so that we
are naturally inclined to credit narration to the sequence-whereas wich cinema,
,,::
when it is a matter of a clear cut, the moment of passage between two shots is
"
,'
:1
106 Restroined Arthro1ogy: The Sequence
not visible: I am in one shot, tben suddenly I am in another. This tendency is
reinforced by tbe fact tbat ehe inscription in the text (caption or dialogue) in the
midst of the panel itself imposes the level oflanguage on tbe image, thus obscur-
ing rhe speculations on its eventual intrinsic narrativiry.
In privileging the sequence, we accomplish nothing more than displacing the
problem, since irisnot true, contrary to what Christian Metz has postulared, that
two juxraposed drawings {like two photos) are forced to tel! us sometbing (or tbree,
or n drawings that are assembled wiehin tbe same page). I demonstrated tbis point
at rhe Cerisy colloquium on comics in 1987, identifyingfive intra-narrative modes
wherein panels can be regrouped wirhin a multiframe, namely amalgam, survey,
variation, declension, and decomposition. Bur, wirhin these fundamental mod.es of
organization (qualified, at tbe time, by "prirnary distributive functions"), two at
Iease-variation (where the images define rhe same thematic paradigm) and declen-
sion (where an idemical motif is submirred to different srylistic treatments)-
verify the dual condirions of resemblance and difference berweert images and can
therefore be placed under the regime of transformation. For as such, the linking
of panels is not determined by any logical inference nor by any causal-deductive
order. !t follows that between two images, transformation does not automatically
ensure a relationship of narrative order. In fact, once images present a rapporr of
transformarion berween ehern, they constitute at the very minimum a series (the
minimum being only rwo images) but not necessarily a narrative sequence. 5
In the terms that tbe question of tbe tbreshold of narrativirywas p0sed, follow-
ing Metz, Odin, and Gaudreault, my provisory conclusion will be that if we do
not dismiss the hypothesis that an isolared image can be inrrinsically narrative-
rbis aspect will be taken up again later-we can, correlatively, be cerraintbat tbe
juxraposirion of two images? raken in a rapporr of rransformation, does not neces-
sarily produce narration.
To continue to borrow from film tbeory, I find my strengest support in the
analysis of tbe "movemem-image" (tbe shots) proposed by Gilles Deleuze. And
most particularly in the following passage: "On the one hand, rhe movement-
image expresses a total that changes and establishes itself between objects: It is a
process of dijferentiation . .. ,. On the orher hand, the movement-image includes
inrervals. ... It is ~ process of specification. "6 This dual characteristic of the
movement-image presents a subject that is "semiotically, aesthetically, [and] prag-
matically" formed, but non-linguistically:
lt is not an enunciation, ond these are not utteronces. lt is an utteroble. We mean that, when language
gets hold of this material (and it necessanly does so), then it gives rise to utteronces which come to
Restroined Arthrology: The Sequence 107
dominote or even reploce the imoges ond signs, ond which refere in turn to pertinent feotures of the
/angooge system. 7
Wouldn't ehe fixed irnage of comics, which by definition does not change, be
semiotically, aesthetically, and pragmatically formed (structured) as well, but in
a different manner? The question can now receive a partial response, since the
fitst half of this book has already brought to light the strucruring power of spatio-
topical parameters: the forms, dimensions, and conrours of the frame, the site of
the panel, ehe metbods of text inclusion, etc.
To my mind, tbe srarus of utterable can and must be extended to all forms of
images, but it is not enough to take note of its semantic potentiaL Indeed, the
image, as we will see shordy, is not only an utterable, it can also be a descriptable
and an interpretable. The meaning that the reader (of a comic) or the specrator
(of a film) constructs tbe reading that is executed from the image has as condi-
tions a selective description and a personal inrerpretation. This appropriation can
ultimately be converted into an utterance; it can also steer us roward an aesthetic
judgment, one rhat would consider the image for its appreciable qualicies.
Deleuze, however, did not extend to all images the enunciable qualiry: He
reserved it solely for movement-images. In chapter 7 of What Is Philoiophy?
("Percept, Affect, and Concept"), he assigns the meaning of all works of art, nota-
bly tbe tableau (painting), toward tbe single register of sensations (percepts and
affects),' that is tO say tbat he places it "outside all mediation oflanguage," a fact
that ought to srun Fran\X>is Wahl. Indeed, one can consider what would make
moving images a subject tbat language necessarily seizes, as "tbe visible, such tbat
painting expresses it"-or better: if it is the painting that expresses ir-will not
be "torn from the register of sensation." For Wahl, Deleuze strangely strikes an
impasse on tbe "discursive being of tbe rableau and its constitution by tbe phrase,"
which is precisely tbe subject that he consecrates in his work ticled Introduction au
discours du tableau.' Franc;ois Wahl's thesis is that "it is in ehe structure oflanguage
that perception is consrructed,"10 the conversion of the picture into linguiscic
propositions makes it all tbe more natural that tbe pieeure (more generally: the
visible) already obeys a specific discursive organization, its configuration consist-
ing of agame rulecl by contexrual rapports.
Again, demonstrating that meaning is inherent to the image is not something
that direccly speaks to comics, since ir is between the panels that the pertinent
contextual rapports esrablish rhemselves wirb respect to narration. Ir is more-
over at tbis level tbat we will quickly verifY that the linkage of images constructs
arricularions that are similar to those of language.
108 Restroined Arthrology: The Sequence Reslroined Arthrology: The Sequente 109

2.2-A PLURIVECTORIAL NARRATION

In order to come to terms with what a sequence of fixed images really com-
municates, and how, subsequencly, narration is accomplished, I will Start from
ehe observation of a page-which I have selected for irs Iack dialogue, in order
to reserve ehe srudy of ehe functions of ehe text for a later phase. Talce (fig. ro)
ehe secend page from ehe album Rencontres by Mufioz and Sampayo (Casterman
1984). Whatseems ro be narrated in ehis page'can be summarized in a fewwords:
The hero, Alack Sinner, is awalcened by ehe noise of a newspaper that ehe wind
has blown up against his window. The headline splashed across ehe front page
informs him of ehe death of John Lennon. The auehors use seven panels to pro-
duce the equivalent of this statement.
Two parricularities should be noted: r) The firsr two panels are sufficient to
deliver the apparent narrative content of rhe entirety of the sequence, whereas the
following images do not appear ro add anyching except an anecdoral prolongation
(Alack lights a cigarette). 2) The seven images brealc down into two series: on ehe
one hand, four panels in which the protagonist appears, and on the other, three
panels showing ehe newspaper and irs !arge headline in shots rhat are increasingly
rightened. No images simulraneously present Alack and ehe newspaper on which
we can read ehe words "John Lennon Killed."
If ehe first two images summarize ehe sequence, neither can be held to be
intrinsically narrative. It is from their juxraposition that I can deduce ·a narrative
proposition. Again, this involves no small amount of interpretation. ls Alack
really sleeping? If he was awalce, was his attention drawn ro ehe noise (no ono-
matopoeia is signaled) or only to ehe sight of ehe newspaper? Noehing allows me
to categorically respond to rJ:lese questions. As a reader. I construct meaning on
ehe basis of inferences rhat appear to be ehe most probable. There is ehe content
that each of these images shows, and there is the meaning that their confrontation
permits them to say.
Nevenheless. in the rest of the sequence, certain images, considered on their
own. are instantly translatable to linguistic Statements expressing an action; and.
fairhful to ehe wishes ofRoger Odin, rhis action purs into relation "a subject and Fig. 10. From Allrl Sinner. Renrorrtres (19841. by Josi Munoz ond Corlos Sompoyo. © Cosrennon.
an object of desire." Thus, for the fourth panel: Alack lights a cigarette. The ehird
panel irself can be translared by a sratement of rhis sort, which would be: Alack
reaches outhisarm to grab a cigarette and a lighter. Bur I only opt for this cransla- me a posteriori11 of the precise signification of his gesture, at once positively-he
tion because it is verified retroactively by ehe following paneL Considering only had to have raken out ehe cigarette and lighter for I can see him light ehe first by
ehe chird panel, I could have also imagined ehat our friend wanted to turn on his means of ehe second-and negatively-he did not light ehe larnp, since ehe room
bedside lamp. The fourrh panel, which shows him lighring his cigarerte, informs is still plunged in darkness.
ll 0 Restroined Arthrology: The Sequence
Furthermore, wasfit it already a retroactive determination that permitred me
to know that che newspaper appears to me, in the first panel, as so mething seen by
Alack-<>r better: that I see it wich him and "chrough his eyes"? By icself, che first
panel only shows me a newspaper fl.attened against a window, wichout allowing me
to wonder which hause chis window belongs to, or if che sight of the newspaper is
remarked upon by some other occupant. We can cherefore formulate this first rule,
chat che meaning of a panel can be informed and determined by che panel that
preceded it much like the one chat follows it. If there is a vectorization of reading,
there is no unid.irectional veccorization in the construction of meaning.
The best example of recroactive determination is furnished by an image chat
does not appear on this page, since it is the first panel of che following page. We
see Alack Sinner leaning over a sink, in a position that leaves no doubt that he
is vomiting. This panel provides its own meaning, not only wich regard to ehe
images chat immediately preceded it (and nocably che sixth panel of che repro-
duced page, which shows Alack coughing and shaking), but to che entirety of
ehe preceding page. In light of chis delayed revelation, I must correct my inicial
spontaneaus linguistic translation. The sequence is now reduced to ehe follow-
ing Statement: The news of che assassination of John Lennon affects Alack to che
point that it malces him sick. All che rest (Alack's awalcening, the newspaper, the
window, ehe cigarette) are reduced. to the rank of simple circumstances.
This panel~which we should consider to close ehe sequence-sheds light on,
and jusci6es a posteriori, ehe nearby image, as well as ehe very tight image of some
of the letters that compose the ticle of che newspaper (cf. panels 5 and 7). lt is
now evident that they materialize the emotional impact of this news on ehe hero.
Shocked, he feels it resonate stronger and strenger wirhin him, at the same time
that the nausea rises. The sta~us of these cwo panels is cherefore different from the
ochers: chey are not objective representations--ocherwise che wind should have
already carried the paper away; and also, Alack did not approach che window, so
rhere is no objective reasoning for ehe enlargement of the letters-but rather rhey
are graphic translations of effects. What chey express happens entirely in the head
(and stomach) of Alack Sinner.
2.3-THE PLANES OF MEANING
Let us formulate in more general terms all that we have just observed. lt seems
rhat ehe analysis of this example allows us to conclude astaging of meaning. This
Restroined Arthrology: The Sequente lll
is plainly revealed in rerms of the reader's crossing of several successive meaning-
ful planes.
Given rhat it has no existence other rhan ehe rheoretical (since the outer frame,
or ehe pbi-field, is always imposed on che perception of che comic reader), che
panel's plane should be principally considered. The image, seen by itself, out-
side of all context, is, as Deleuze rightly suggests, an utterable. I can translare or
express what I see inside che frame (the what of che monstration) in linguistic
terms. Sometimes, rhis virtual Statement will be a straight narrative (Alack lights
his cigarette), whereas other rimes, failing to perceive a dynamic inrernal relarion
to che image, I have to content myself to name the object-sign (or objecc-signs)
that ir shows. Rather than an intrinsic narrative, I will employ a more neutral
term, that of immanent significance. At this elementary stage, my job as a reader
is simply observation and identification.
The second plane is that of syntagm, limited, in occurrence, to the triad com-
posed of the panel chat is currently being read, the panel that preceded it, and che
panel chat immediacely follows it. At chis Ievel, my reading of che panel is already
forcibly different, informed before and after by ocher contents wich which I con-
struct (or verify) semantic relations, on the basis of a postulate of narrative coher-
ence. Plainly, I am now involved in interpretation. This arthrologic micro-chain
constitutes an instance of shifting interpretation: at any moment of my reading,
I will privilege ehe relations of immediate proximity and I will reconstruct chis
triad, which is carried along wich me. (Only che phenomenon of incrustation
contradiccs chis rule, by installing a privileged dialogue becween cwo terms, che
incrusted panel and the panel that accommodates it.)
The third plane of meaning is chat of the sequence. The semancic articu-
lations of the story allow me to identify and to circumscribe a story segment
of any lengch, characterized by a unity of accion and/or space. The sequence
allows itself to be converted into a synthetic Statement that, transcending the
observations and constructions of the inferior Ievel and stopping {at least provi-
sionally) the work of inferences, produces a global meaning that is explicit and
satisfYing.
General arthrology demonstrates that che panel can also be che object of dis-
tant semantic determinations, which overtake the frame of the sequence and
proceed to a necworked operation. Like all narrative works (deployed in time),
a comic is governed by the principle of diffirance (delay): its signification is con-
structed solely on che terms of ehe reader-freed afterward co che interpretation
deepened by the research of meaning that knows no definitive Iimit.
112 Restroined Arthrology: The Sequente
We must always remernher that a nurober of works thar are more traditional
and less Sophisdcared rhan Alack Sinner, obeying a narrative order rhar is srrictly
linear (of ehe causal-deductive type), never spare a retroaccive determination at
ehe Ievel of the sequence. Rather, ir is the plane of the syntagm rhat is dominant.
In comics for young children, the authors simplify their intention by forcibly
rendering each panel totally explicit and significant in itself. But the analysis of a
page taken from a given series (Jojo by Geerrs) will show that even here, cenain
effects, for example a humorous effect, do not hold unless the young reader effec-
ruates a reconciliarion berween distant panels and scenes. The choice of ehe net-
work as ehe ulrimate Ievel of interpretative pertinence is not exclusive to modern
comics, wich their fragmented writing, but is a general principle.
2.4-TO THE RESEARCH OF THE GUTTER
"!t requires less to make one see than to say, less to describe than to tel!," remarked
Claude-Fran<;oise Brunon." And to highlight, following others, the essence of
ehe story frequently passes "outside of the image ... between images." Cerrain
authors even strive to produce work in a way so that ehe reader's Iook, "deprived
by principle episodes," will be "carried out from one absence to the next," even if
"the text compensates for what the image refused to expose" on a frequent basis.
To conclude that meaning is produced by the intericonic gticrer (the
"entr'image") at least as much as is produced by the images themselves, there
is but one step, one that some people have been tempted to cross. Thus Benote
Peeters: "The true magic of comics operates berween the images, the rension rhat
binds ehern.... In Herge these are memorable 'gutters' that we must analyze,
these intervals between two panels lavished with accuracy and audacity. "''
In reality, the gutter in and of irself (that is to say, an empty space) does not
merit fetishization: When there is no gutter, only a simple line to separate rwo
contiguous images (as often seen in Bretecher, or not long ago in Töpffer), ehe
semantic relarions between rhe images are the same. While we may presuppose an
implicit frame to images rhat aten't composed with one, there is no point to postu-
late an implicit void wlien the illustrator did not make use of one.
Maybe, you will say to me, but the term "gutter" (blanc) lends irself mera-
phorically. We use it to designate "that-which-is-not-represented-but-which-the-
reader-cannot-help-but-to-infer." !t is therefore a vinual, and take note that this
virtual is not abandoned to the faneasy of each reader: it is a forced virtual, an
Restroined Arthrology: The Sequence 113
idencifiable absence. The gutter is simply the symbolic site of this absence. More
than a zone on ehe paper, it is the interior screen on which every reader projects
ehe missing image (or images).
I cenainly do not believe that the comics reader has to menrally construct
"ghost panels'' (cases fontome; the expression is from Peeters), except maybe in
extremely rare examples identified by theorists to prove a point. These examples
are a lircle roo weH-seleered to permit ehe construction of general conclusions.
By looking at it closely, we cannot help but be struck by an apparent paradox:
these famous examples never relate to segments composed of only two panels-a
necessary amplitude but sufficient to the exhibition of a gutter. A third panel is
almost always implicared,4 and this confirms that it is indeed at ehe minimum
a compound syntagm, or even a much Ionger sequence, that is at ehe major
Ievel of significance, the threshold where one can elaborate pertinent logical
inferences.
We would be mistaken to want to reduce the "silences" berween two consecu-
tive panek by assimilating the ellipse to a virtual image. On the contrary, this
silence often speaks volumes. It has nothing ro introduce, no gap to sucure. lt is
in this sensethat Henri Van Lier spoke of the "null blank" (blanc nu[) in which
the multiframe floats like a falling leaf. This blank, "the annulment of all conci-
nuity, "'' is the opposite of the "relay-gutter" (un blanc-relais).It is the Mallarmean
blank of Coup de des, ehe void of the music ofWebem and that of quantum phys-
ics. Reading a comic, I am here, then I am there, and this jump ftom one panel to
the next (an optical and mentalleap) is the equivalent of an electron that changes
orbit. In other words, an intermediate scate between the two panels does not
exist. The comics image is notaform that, subjected to a continual metamorpho-
sis, would be modified by investing successive frames (between which it would be
permissible to reconstirute ehe missing moments).lt is necessary, in contrast, that
ehe gutter (provisionally) cancels the already read panel in order to allow the next
panel to exist in irs own right, in terms of a complete and compact form.
Panels betonging to the same sequence are assuredly in debt to each other. On
the semantic plane, this iconic solidarity, in which we have recognized the very
foundation of the comics system, is programmed by the author at the break-
down stage, and, at the time of reception, postulated by the reader in the form
of hypothesizing a coherent narrative. Following this logical fallacy, all panels
inevitably intervene apropos. For the comics reader. the fact of presupposing that
there is a meaning necessarily Ieads him to search for the way that the panel that
he "reads'' is linked to the others, and how it re-reads in light of others.
:,' .I
114 Restroined Arthrology: The Sequence
Comics exist onlyas a satisfyingnarrative form under the condition that, despire
the discontinuous enunciarion and the intermittent monstration, the resultant
story forms an uninterrupted and intelligible totaliry. The "gutter" between the
two panels is therefore not the seat of a virrual image; it is the site of a semantic
articulation, a logical conversion, that of a series of utterables (the panels) in a
Statement that is unique and coherent (the story). The Alack Sinner page taught
us that rhis conversion is sometimes passed in stages. The first Statement, issued
from a dialogue between two or three juxtaposed panels--and nacurally, forged
under the control of the preceding ones-may be nothing but a provisory one
that must undergo, under a stroke of unforeseeable retroactive determination, a
correction in moving toward the adoption of a new, more inclusive Statement.
Clearly, this progressive construction of meaning is not exclusive to comics.
Rather, as Wolfgang Iser has norably demonstrated, it is analogaus ro the pro-
cess that structures the reading of a lirerary text. The "wandering viewpoint"
constitutes, he says, "the basic hermeneutic structure of reading.·" In a sequence
of sentences, new correlations frequendy "Iead not so much to the fulfillment
of expectations as to their continual modification . ... Each individual sentence
correlate prefigures a patticular horiron, bur this is immediarely ttansformed into
the background for the next correlare and must therefure necessarily be modified.
Since each sentence corrdate aims at things to come, the prefigured horizon will
offer a view "Yhich-however concrete it may be-must contain indeterminacies,
and so arouse expectations as to the manner in which these are to be fesolved. "16
The comics image, whose meaning often remains open when it is presented as
isolated (and wichout verbal anchorage), finds its truth in the sequence. Inversely,
the gutter, insignificant in itsdf, is invested with an arthrologic function that can
only be deciphered in light. of the singular images that it separates and unites.
Therefore, the intericonic gurrer can be qualified as "polysyntactic," following
what Anne-Marie Christirr has said abour "picrorial emptiness" (that which sepa-
rates the figures in the interior of an image, in the space of the picture). Anne-
Marie Christin suggests that the function of narrative is that which pictorial
emptiness assumes with the greatest difficulty:
[T]he cleor ond immediate designotion of the roles of the represented figures does not roise the spoce
thot mutuolly isolotes them from eoch other but the codes thot ore individuolly chorged, codes of dress,
gesturol codes especiolly, os shown in genre pointings, for exomple those of Greuze. lf the emptlness is
necessory to constitule ostoria between the poinred figures, os preconceived by Alberli, it is bocouse it is
Ioremost omerk of inrelligibility, the clue to oc0jlresence. 17
Restroined Arthrology: The Sequence 115
The intericonic gutter also marks the semantic solidarity of contiguous pands
above all, both working rhrough the codes of narrative and sequential drawings.
Between rhe polysemic images, the polysyntactic gurrer is the site of a reciprocal
determination, and it is in this dialectic.interaction that meaning is constructed,
not wichout the active participation of the reader.
2.5-REDUNDANCY
Another commonplace of comics theory wants to see the medium compelled
toward iconic redundancy, this being the "price to pay" for narrative continuity
to be assured. Called the "stuttering art" according co the depreciating formula of
Pierre Mason, 18 cernies are founded on a dialectic of repetition and difference,
each image linked to the preceding one by a partial repetition of its contents.
This conception proceeds from a dual inßuence. lt finds its sources in infor-
mation theory, wherein redundancy is indispensable for the transmission of new
elements that properly conscitute information, and in classical narratology (that
ofTodorov and Gaudreaulr), that conceives of narrarion as possible only between
images that are both similar and different (cf. supra 2.1).
We would cerrainly not deny that redundancy is a principle of ehe vasc major-
iryofcomics (even though there are cenain exceprions). Again we have to see rhac
it is generally a direct consequence of the story's organization around a central
figure (conventionally designated as the "hero") who, alone or flanked by side-
kicks, will always be continually at ehe hearr of the action. This narrative focaliza-
tion is translated ro the image by rhe ubiquiry of said character, who is depicted in
a !arge nurober of panels.'' In fact, the insiscenr character of the proragonist finds
itsdf in all narrative forms, comprised in the novel where it is a proper noun,
or the pronoun that rakes its place, and which is tirelessly repeaced ("!," "he,"
or "she," depending on whether the narration is effectuared in ehe first or rhird
person). It is only conspicuous when it isamatter ofa story in images, spread over
the multiframe in a situation of co-occurrence.
Among the researchers that have particularly insisced on the obligarion of
redundancy, I will menrion the German Ulrich Kraffr" and rheAmerican Richard
]. Watts. For Watts, the comics author, when he employs verbaland visual means
torender an action "conspicuously manifest," goes about "in a cerrain manner
that the reader is capable of distinguishing," not only "between new and old
information," but, equally, among the new information, berween those that are
 116 Restrained Arthralogy: The Sequence
pertinent or not.~ Nevertheless, the Manichean strand of a similar rypology of
information does not escape Warts. Righcly seeing that the production of infer-
ences is essential to th.e work of the reader, he must concede that "it is not pos-
sible to explain the totality of communication processes where comics are raised
through the application of models of coded communication," and that the art of
cernies may finally reside in the "discovery of the ultimate limit thac the reader is
susceptible of achieving in their capacity to produce inferences. "z2.
) ' We find in Michel Tardy the idea that, in' the construction of the image, the
pertinent information is rendered conspicuous by a set of techniques and proce-
dures that assure a hierarchization of motifs (spread berween the "central object"
and the "secondary objects"). Bur the author hirnself indicates the Iimits that
collide with the will to master the reception of the image, when, evoking a hypo-
thetical relevised sequence. he writes:
lt must be tnken into account that, as in the world, objects necessanly maintnin neighboring relations with
other objects, and run the risk that the spectatnr, by a voluntary act where jurisdiction slips in the factor
of images, overwhelms the perceptive hierarchies that are proposed to him_ There, in the surrounding
crowd, a feminine silhouette is more interesting than the dos!Hlp of the face af a character that is na
Ionger principaL23
As we caD: see, the theoretical use of rhe notion of redundancy is necessarily
driven toward the idea that each panel is organized around new and peninent
information that will advance the narrative. Now, there are numerous objections
against such a use. First of all, the concept of information itself is reductive; it
participates in an all too functionalist conception of the story in images. In a
drawing (and even more so ifit is a fixed image), "objects" or details can arouse an
interest, give pleasure, add a discreet touch to the work, and reach toward a form
of pertinence that is not direccly allied to the behavior of the intrigue.
Second, I will insist on the fact that the progression of the story is not constant
and linear. It is not true that each panel has as its mission progressing toward a
resolution. In particular, certain mangas are signaled by a massive use of panels
that are superfluous from a ·srriccly narrative point of view, their precise func-
tion is elsewhere: deCorative, documentary, rhyth.mic, or poetic, whatever the
case. These panels respect the general principle of co-reference, but their con-
tribution cannot be evaluaced in terms of information. More than the panel, it
is therefore the page or ehe sequence that, under this relationship, constitutes a
peninent unit.
Restroined Arthrology: The Sequence 117
In reality, there exists a multiplicity of possible correlations between contigu-
ous panels. Some effectively full back on redundancy, at the risk ofbeing reduced
to a minor element (a "secondary object"), like the hotellabels that George Perec's
Windder tried so hard to classifY:
What he would have liked would be to link each Iabel to the next, but each time in respect of samething
e~e: far example, they could have same detail in common, a mountain or a volcana, an illuminated bay,
some particular flower, the same red and gold edging, the beaming face of a groom, or the same dimen·
sions, or the same typeface, or similar slogans ('Pearl of the Ocean; "Diamond of the Coasf), or a
relationship based not on similanty but on opposition or a fragile, almost arbitrary association."
But sometimes the correlation slips totally to the redundancy of.the signifier, for
example when it is on the order of meraphor.
In addition, panels which compose a series (in the sense that they obey the
principle of partial iconic redundancy) are not necessarily disposed to follow one
after the other, but can alternate with panels parricipating in one, or even several
other, series-as in our last example (fig. ro), with the entwinement.of the first
series of three images representing the newspaper, and a second series where rhe
panels represent Alack Sinner.
In summary, redundancy is far from being an obligatory bridge berween rwo
consecutive panels ofa narrativesequence. In a comic, narrative continuity is assured
by the contiguity of images, but this side-by-side is not necessarily an end-to-end
of narrative instances structured according to a univocal and mechanicallogic of
repetition and difference. We must guard ourselves here against dogmatic conclu-
sions. Comics admit all sorts of narrative strategies, which are all equally mod-
ern and legitimate. As Richard Watts seemed to sense, the breakdown sometimes
manifests itself par excellence as the art of economizing redundancy. Inversely,
a cartoonist such as Lewis Trondheim, in his earliest works, amused hirnself by
pushing the principle oficonic redundancy to its paroxysm. ''
2.6-BREAKDOWN AND MISE EN ScENE
I will now analyze the breakdown of a sequence taken from a work that is more
classical in form than that of Mufioz and Sampayo. It is Royaume des eaux noires,
the last book in the Gorentin adventures, drawn by Paul Cuvelier from a scenatio
by Jean Van Hamme (Ed. Lombard 1974). Page 38 (fig. n) is an action scene that
118 Restroined Arthrology: The Sequenre
illustrates a pursuit. The hero, Corentin, is attempting to escape horsemen sent
our ro find him by ChaYran, his formidable adversary. He is accompanied by
his friend ZaYla and her grandfu.ther, rhe magician Narreddine, as weH as by his
habirual companions, Belzeburh rhe gorilla and Moloch the riger.
In irself, the acrion developed in rhis page is perfecrly "readable" when removed
from rhe contexr of rhe album. Only the event shown in rhe last panel is uninrel-
ligible, not only fur the readerburalso for the young proragonisrs, as is indicared
by rhe text bubble fearuring a double question mark. The aurhors have placed
rhis unusual and specracular image ar rhe end of rhe page (which, in rhe album,
occupies a page on rhe righr hand side) so rhar rhe sequence is inrerrupred on an
image where suspense coincides with stupor. The explanation is provided in ehe
following page: Narreddine has called upon his magic to conrain rheir aggressors,
erecting a "magneric wall." It is against this invisible wall that the horse and his
rider have crashed.
Let's take an interest in ehe dramaturgy of this sequence. In ehe first panel, ehe
heroes have fixed an objecrive. Their franric course in a deserr of rocks where they
cannot help but be captured finds a direcrion, and the entire sequence &om then
is driven (vecrorized) roward an uncertain finish: Will they succeed in reaching
rhe shelrer known as "Satan's Hook"? The second panel illustrates rhe menace of
rheir pursuers by revealing rheir numbers, their armament, and their determina-
rion. The chief of the horsemen exrends his arm in response ro the exrended arm
of Narreddine, borh poinring in rhe same direcrion. The third paner is the only
one ro encompass borh groups, fugirives and pursuers; this allows rhe reader
to take an exact measure of ehe situation, and to observe, in panicular, ehe dis-
rance rhar separates them. However, Moloch and Belzeburh do not appear in this
image; in fact, we do not-~en see them again in this page, nor on the following
page. Ir is not rhe magic of Narreddine rhar is the cause of rheir disappearance.
They are presumed to parricipate in rhe action righr unril rhe end, bur, having
only a supporring role, rhey are placed outside of rhe field of rhe panel so rhar, in
the moment where the d.anger increases, attention is reasserted on the principal
figures. Thar is dearly whar happens in panels 4 and 5, where only Corenrin and
ZaYla are represented.
This sequence co;,firms rhar a panel is not solely derermirred by rhe one rhar
preceded ir and rhe one rhar follows ir, bur by rhe global economy of rhe sequence,
which determines its true function. We can see that redundancy is limited here to
the characrers rhar are ar rhe hearr of the acrion, rherefore we can say rhar rhey are
objects of an iconic focalizarion (which relays rheir privileged status in rhe entirery
Restroined Arthrology: The Sequenre 119
of the srory). Nonerheless, everyrhing rhat need not be repeared is shown only
once; and this unique occurrence is like a generalfactaring for the entire sequence.
Therefore, rhe reader "conserves" all rhis useful informarion for rhe inrelligibiliry
of the narrative situation (be it, in occurrence, the presence of two savage animals,
the nurober ofhorsemen, the proximity of"Satan's Hook," the distance between
the two groups) without the successive images needing to repeat it.
The panels can be assimilared ro metonymic fragmenrs raken from a virrual
"esrablishing shor" rhar endoses rhe sequence in its roraliry, borh temporally and
sparially. The rerm breakdown should therefore be undersrood Iirerally and wirh
respect to two dimensions: that which is broken off are not only the moments
wirhin the narrative tissue (the narrative keys of the action), they arealso partial
views, selective framing zooming in on the pertinent zones, and placing certain
information outside the frame.
Roland Barrhes recalled: "[T]he mainspring of narrative is precisely the confu-
sion of consecution and consequence, what comes after being read in narrative as
whar is caused by."'6 However, modern narratology develops a mode of srrucrural
analysis that tends to "dechronologize" the narrative contents in order to "resyn-
chronize"27 them. Ir appears to me that the narrative organization belanging to
comics, that is to say, the manner in which information is distributed ove"r rhe
durarion, frequenrly overrakes rhe logic of post hoc, ergo propter hoc to employ
procedures thar are rhemselves already srrucrural.
The fundamental rule rhar strucrures rhis page from Gorentin is rhe principle
of economy, by virrue of rhe fu.cr thar the illusrraror indudes in his images only
that which musr be rhere. Without a doubt, rhis focalization on rhe perrinent
elements is never absolute, the image always extends, by nature, the message that
it contains:
[T]he monstrotion to its proper contingences ond the "porole" thot is ot issue is olwoys, more or less,
poly-informotionol. lf Isoy: "The mon is in the hollwoy," Iom not giving ony contextuol informotion on
the ospect of the hollwoy, on whot it contoins, on the topogrophic situotion of the mon in relotion to the
woll on the left, etc. As the entire imoge ... mokes me understand thot there is o mon in ohollwoy, it
olso teils me whether it is profoundly engoged, if it cantoins obiects, if the mon is leoning on the left wo II
or not.18
The "conrexrual informarion" that rhe image holds (in this example: rhe color
of the rocks, rhe presence of douds in the sky, the dust that rises in rhe parh of
the horses) may Iead ro a loss of meaning if, because of them, the arrenrion of the
 120 Restroined Arthrology: The Sequence
reader is scattered. Nevertheless, it seems from experience thar this risk is illusory,
as narrative and semantic coherence is never really in jeopardy.
What preserves it is rhe fact rhat the reader's attention is spontaneously ori-
ented toward narrative curiosity, and their attention organized by the particu-
lar appetite that awakens all fiction. Paraphrasing a formula by Bernard Noel,
I would say that cernies readers do not see rhe image, but rather they see ehe
meaning that has, for each of them, the ponion of the image that the story des-
ignates to their arrention.29
Symmetrically, tbe hierarchy of information conveyed by the image is assured
by its intrinsic organization, rhar which obeys the instance of ehe mise en scene.
Borrowed from the world of live theatre, tbis concept can be meaningfully
extended to comics, since the cinema already uses it only as an extrapolation of
language. In ehe case of the ninth art, we must rigorously disringuish at least rwo
components to rhis organization of representation: the framing (mise en cadre)
and the drawing (mise en dessin). Bur ehe cheoretical benefit is sliin, more so since
the mise en frame appears to be relevant to a double authoriry: that of the break-
down and that of the page Iayout.
The mise en scene, therefore, organizes ehe different parameters of the image
(framing, choice of point of view, composition, "actions" of the characters, light-
,, ing, etc.) in accordance wich the internal dynamic of the sequence, to produce an
I
~~ aesthetic or dramatic effect, and for an immediate readability of what constitutes,
in the image, a pertinent utterable.
Let us observe the procedures of tbe mise en scene employed in the last four
panels of tbe page from Corentin. Becween tbe fall of Zaila (tbe crucial moment
in the sequence, judiciously placed in tbe center of che page) and the following
image that sees Cerentin giying her aid, the point of view has been shifred about
roo"; placed behind Za:ila, we are married to the movement ofher fall, but it is
witb Corentin tbat we hurry to her aid. Other than reinforcing participation, tbe
shifr of point of view has two other advantages: it permits the illustrator to show
tbe gesture and tbe face of Za!la from tbe most eloquent angle, and it prepares
for tbe rising of tbe cavalier in the following image. This image is framed in a
slight low angle shoc_ to render the assailant's stature even more menacing. Finally,
the composition of tbe last panel is no doubt the most remarkable: The chosen
point of view is perpendicular to the path of the horse, a choice that allows us to
appreciate ehe exact physical proximity of the heroes to their adversary, and above
all, spectacularly renders tbe brutalshock of the horseman and bis impactagairrst
Restroined Arthrology: The Sequence 121
tbe magnetic wall. The vertical drawn by the different points ofimpact is situated
in tbe center of tbe panel, which it divides exacrly in two.
The mise en scene tberefore fully participates in the breakdown. The panels,
which are tbe material body of tbe work, are composed as a function of semantic
articulations, where the breakdown has been decided and which it must manifest.
As an experiment (cf. fig. n), we could amuse ourselves by introducing, among the
panels, coordinative conjunctions or other linguistic padding that explicates the
aniculations of tbe sequence, its arthro-logic. lt appears tbat tbe meanwhile, there-
fore, but, immediate!y, unfortunate!y, suddenly captions that would be introduced
are redundant and perfecdy useless. The breakdown, norably the procedures of
mise en scl:ne, arealready structured and girded by these implicit syntactic opera-
tors, which the page Iayout itself can sometimes highlight. If it were otberwise,
the conversion of a suite of utterables (ehe panels) into a coherent Statement (the
narrative sequence) would be impossible.3°
2.7-DESCRIPTION AND INTERPRETATION
I said earlier tbat tbe image is not only an utterable, but also a descriptible and an
interprerable. The moment has come to specify in what manner description and
interpretation can restore to ehe image its true semantic richness (and ehe arising
emotional dimension), that the reduction to a linguistic Statement corresponding
to its immediate narrative "message" tends to mechanically overshadow.
Indeed, the reduction of the utterable to a Statement mobilizes, in the image,
only the elements direcrly concerned wich the narrative process, tbat is to say,
those engaged in action. (It is in ehis way that, wichout exception, ehe characters
are revcaled as more meaningful than tbe decor.) Besides, tbis conversion requires
nothing but a global and synthetic perception of the elements that it rerains. If,
ror example, I conven tbe second-to-last panel in tbe page of Gorentin (fig. II)
into a Statement such as "raising his saber, the first horseman falls upon Corentin
and Za:ila immobilized on the ground," the details relative to the attitudes of the
protagoniscs ehat are retained as pertinent (raising his saber, immobilized on ehe
ground) do not exhaust the visual information contained in the image. What
exact distance separates the assailant and his victims? How are the characters
dressed? What is the nature of the surr, the fields, tbe sky? On all of these ques-
tions, ehe image is not mute. But it only furnishes answers under ehe condicion

122 Restroined Arthrology: The lequence
fig . ll. Fmm (OJenlin: I.e Royovme des eovx noires (19741, by Poul Cuvelier ond Jeon Von Homme. © les eoitions du IDmbord.
rhat, modifying my regime of reading, I Iook at it like a descriptable, instead of
consuming it like an utterable.
We know that description is one of tbe fundamental operations of the liter-
ary rext.3I Considered one of rhe means of amplificatio, it is a form of insistence, a
Restroined Arthrology: The lequence 123
procedure ofemphasis: a texrstops on certain characters, sites orobjects to describe
ehern more or less at length, while ignoring others, establishing an implicit hierar~
chy. Whatever the lengtb and minutiae of tbe description, it represents, as Phillipe
Hamon recalled, "a relarively autonomaus 'piece', easily 'detachable,' 'withdraw~
able' in the texmal f!ux."''
We see at once that an image cannot be descriptive in the technical sense
of tbe term. If it shows the constitutive parts of a certain object, as weil as the
properties of these parts (forms, materials, colors, etc.), these details do not add
to the presentation of tbe object; tbey are tbemselves consubstantial. Indeed, it
is the distinctive feature of visual monstration to present the "parricular" rarher
than ehe "general. "33 The presence of these derails in the image is not contingent
and does not bear witness to a descriptive inrention. Wich a drawn image, in
ehe first instance, it is rhe particular style of rhe illustrator rhat determines ehe
image's degree of precision, ehe number of details rhat it receives. An editorial
cartoon by Wolinski or by Petillon remains basic or, if we wish, weakly descrip-
tive. Inversely, a caricature by Mulatier or an illustration by Ctumb present an
abundance of lines and further pushes ehe enumeration of the particulars of ehe
represented subject. Through these four examples we can see two radically differ-
ent conceptions of the "humor" drawing that are popularly known by two niodes
of opposed participation: I arn looking to grasp the idea that is expressed in tbe
drawings by Wolinski or Petillon, and my interest is exhausted as soon as I detect
the humor of it, instead of willingly lingering to contemplate the drawings by
Crumb or by Mulatier, which abandon me to the fuscinarion of detail.
This diversicy of graphical writing is found wirhin comics. But to it is added
a property whereby tbe pieeure stocy opposes tbe literary story again: tbe image's
degree of precision stays more or less equal, wharever rhe represented motif may
be (site, object, character). If the image is descriptive, it is equally so for all tbe
motifs convened by ehe story, granring to each ehe same care. This equity is strict
in tbe cinema, where tbe film does not distinguish between tbe diverse actors and
objects that make up the profilmic material, recording tbem all wirb tbe sarne
"objectivicy." It is approximate in comics, where rhe areist is free to modify the
entire regime ofhis graphical writing, detailing one motifwhile otbers remain at
tbe sketch stage. Nevertheless, tbis possibiliry is generally theoretical, and these
significarive occurrences are few in number-ir seems to methat they are mainly
found in the pages ofyoung artists such as Yvan Alagbe, Joann Sfar or Dominique
· · Goblet or in. tbe work of a baroque areist such as Bill Sienkiewicz. The rule tbat
prevails everywhere is that of the homogendry of style.
124 Restroined Arthrology: The Sequence
The pieeure story is therefore much less discriminating than the literary text.
To in·sist on a motif and ro designate ir as more important than another, the cam-
era must linger longer; bur in doing so, it produces nothing rhat is equivalent ro
a description. In comics, ir is by the frequency of appearances rhat rhis character
or that object will be privileged over any others, and it is not that its "optical
definition" will be superior to rhose characters or objects that are more episodic.
This point Ieads me to highlight the fact that once the same motif is repre-
senred several times it transports all of its attributes (its predicates) along wich it.
If we want to provide recognition to ehe descriptive propenies of the drawing,
we must therefore admit that ir is a description that is infinitely restarted, to
which we cannot assign a particular sire. Contrarily, in a text, the descriprions
are generally given "once and for all"; once described, a character, for example,
can thereafter be designated simply by his name, by a pronoun or a deictic. The
description cedes ehe place to ehe designation, where it can therefore be consid-
ered a form of extension.>-t
For all of the reasons given above, it seems to me that comics (and the visual
story in general) are not apt to produce, by itself, an equivalent of ehe opera-
tion known in the literaty domain as a description. On the other hand, each of
the panels is descriptible by the reader, in the same way that we recognize the
utterable. If necessary, it is up to the receptor to construct the description. This
construction is possible only under cenain conditions. Its extent is first of all sub-
ordinate to the quantity of information objectively conveyed by the drawing; it
is fairly obvious to me that the description of a panel by Copi will be made more
quickly than a panel by Giraud. It necessitates, above all, a change in the modali-
ties. of reading. To read a comic, in the first instance, is always to attach priority
to the chain of events or, if we prefer, to the dynamic of the story. Reading begins
by deploying the work at ~he first Ievel of coherence, that of a mechanical mean-
ing. At this stage, the image is apprehended-not exclusively, as we will see it in
an instant, but principally-in its enunciable quality. To accede to a descriptive
reading, it is not sufficient to become lost in ehe contemplation of ehe images.
The description is completed only through an attentive active reading that estab-
lishes an inventory of information contained in the image.
As we can imagine, the potentially descriptible image is rarely described in its
totaliry. We do not want the reader to undertake--to what profit?-a scholarly
exercise (an exercise that, under ehe name of ekphrasis, was part of ehe education
dispensed in ancient Greece, in Byzance, andin the Italy of the Quattrocento).
But it is undoubtedly rare that this description is not effectuated at all. The
Restroined Arthrology: The Sequence 125
degree of descriptivity of ehe effectuated reading varies from one image to another,
and ftom one reader to the next. My hypothesis isthat each readerwill find in the
image-and will retain-whichever details are significant at that instant. Now
this withdrawal35 exceeds the striccly functional Ievel of the conversion of the
im.age to statement; it constitutes ehe start of an inventory, however minimal, or
of a description.
Daniel Arasse recalls that "the !talian language differentiates that which is a
particolare and that which is a dettaglio."' 6 Any small part of a figure or an object
constirutes a particolare, ehat is to say, a detail in ehe objective sense. This small
part becomes a dettaglio ehe instant where it "becomes significant," in ehe sense
that it is separated out, chosen by ehe reader/spectator of ehe image, who finds in
this detail a particular interest or pleasure. Thus, for my part, I can never Iook at
the page of Corentin without my gaze being intercepted by at least one of the fol-
lowing "small patts": the tiger and the gorilla in the first panel, the mass of curly
hair ofZaila in the fourth panel, her lascivious abandon in the last panel, or again
the stunning position of the horse's legs in the penultimate panel. These details,
which are privileged solely for myself-to each his own-naturally inform the
global perception that I have toward the page, much as they nourish and inflect
the pleasure that I take ftom it.
But the image, and singularly the drawn image, is doubly descriptible. On the
one hand because it always shows more than is necessary to the intelligibiliry of
action-this addition is composed of what Andre Gaudreault and Fran~ois Jost
designated as "contextual information''---on the other hand, it is the product of
a singular gtaphical writing, where each line can be described in its specificiry
(technical, motot, aesthetic). Daniel Arasse makes the distinction (p. 12) between
"iconic" detail and "pictorial" detail, the first referring to a particularity of ehe
painted object, the second to the presence of pictorial matter irself. In the same
way, describing a drawn panel is an operation that is necessarily divided. A descrip-
tion is not realized ifit does not take into account, aside from ehe drawn dements,
the manner in which they are drawn. If I describe a scene illustrated by Muiioz to
someone who has never seen it, while omitting all ehe stylistic characteristics of
this areist (as if it made no difference ehat eheillustratorwas Mufioz or Herge),
I would never arrive at recreating in ftont of his eyes the image that I'm tallang
about. Admittedly, it is very difficult to completely describe an image in its two
dimensions (iconic and graphic), that is, to simultaneously do justice to the rep-
resented scene and the organized and sensible ensemble of materiallines that pro-
duce chis scene. lndeed, this ambition assimilates description to a form of criticism
126 Restroined Arthrology: The Sequence
in play. Describing. with a minimum of precision, rhe line or the graphical system
of any panicular artist presupposes competencies that are far &om being unani-
mously shared.
Ir remains that, once I stop myself at any given derail of a comic, it is not rare
that I taste it (rather confusingly) as a local and delectable graphic performance.
Thus, in rhe legs of the horse drawn by Paul Cuvelier, I admire rhe anatomical
science of the illustrator, his sense of movement, and rhe choice thar he made of
a less conventional posture, whose accuracy imposes itselfin an indisputable way.
My appreciarion of whar rhe artist has done, at rhis precise location, can conduct
me to an interrogarion of his method: did he use a documentary photograph for
rhis drawing? In brief. it is not because I am a horseman nor a friend ro horses
that I am irrteresred in rhese details; it is because I am an amateur draftsman and
admirer of Cuvelier.
Ir still remains for me to talk about interpretarion, or rather the image that is
interpretable.
To do so, there is no need to change our exarnple. It suflices ro suggest that, to
return all of the textual density to page 38 of Royaume des eaux noires, we should
incorporate a Ievel of meaning that is more global: that of the album. The fall of
Za1la (her foot stubbed against a rock) and the helping motion of Corentin evoke
an echo of anterior scenes, and more particularly the dispures that had already
been erected between our heroes. Let's rake a Iook at the album. On page 5 (panel
3), Za1la, furious, shouts to Corentin: "I hope that the rocks on the road will cause
you a thousand wounds . ., Page 17, her feminine ruse suggests a way of coaxing her
companion by suffering the pride of a young man: she simulates a fall to permit
Cerentin tosaveher and to takeher in his arm.s.
An image is interpretable in the sense that, in a sequential narrative such as
comics, it is always close to. other images, situated before and after in the course
of the story. We enter here into the domain of a general arthrology, which will
be treated in the third chapter. At this Ievel, comics should be apprehended as
a networked mode that allows each panel to hold privileged relations with any
others and at any distance.
Naturally, the image can· presenr characteristic trairs or constituent elements
that signal exterior ·referents to the considered work; interpretation-forever
unfinished-is therefore invired to take into account all the peninent determina-
rions that belang to culture, ro collective memory (socio-historic) or rhe indi-
vidual memory of the reader, to the encyclopedia, in the sense of Umberto Eco.
Perhaps we should highlight that a subsection of the encyclopedia is constituted
Restroined Arthrology: The Sequence 127
by the reader's knowledge of the illustrator of the image that is subjected to inrer-
pretation (other works, obsessional ehernes, fluctuations in style), and of cernies
in general. This knowledge is effectively determinant, behaving as an art that
pracrices a Iot of auto-reference, notably in rhe paradie mode, but also in rhe
more serious regimes of homage or of crirical rereading. For example, we do not
fully understand the masterpiece that is Watchmen if we do not have any prelimi-
nary familiariry with superheroes.
Michel Picard spoke of a poem as a "flake of meanings that all analysis flattens
in its lineariry. "" The same terms can be applied to comics. Standard readings,
which privilege, in each image, the enunciable qualiry, flatten the semantic rich-
ness of the image to profit from irs immediate narrative function. Only a descrip-
tive reading-attentive, notably, to its graphic materiality-and an interpretive
reading allows the image to deploy all of its significations and resonances.
2.8 ....THE FUNCTIONS OF THE VERBAL
The page by Mufioz and Sarnpayo thatwe exarnined was mute (with the exception
of the newspaper headline), and the one by Cuvelier and Van Harnme is hardly
ralkative. Whatever linguistic Statements that are enclosed are not indispensable
to the comprehension of an action that is, here, essentially visual. Nevertheless,
in this page we can already identify two functions of the verbal: a function of
dramatization-the exchanged comments add to the pathos of the situation-
and a realist function. This is a point rhat we generally leave aside: there is a
realiry effect rhat attaches to the verbal activiry of the characters, for the simple
reason that in life, people ralk-although it is undersrood that mosr of the time,
nothing important is said, or at least nothing essential.
These two functions are evidently not the only ones that return to the verbal
in comics. Ir is evident rhat they often panicipate in a determinant fashion to the
production of a global meaning. If I did not raise this issue earlier, it is for the
following reason: the abiliry of a language to signify is something that everyone
admits, and it seemed necessary ro clearly esrablish that a single image is also the
bearer of meaning, and more important to identify several of the semantic proce-
dures proper to sequences of images. Having defined comics from the beginning
as a predominancly visual medium, I was moved to jusrify this postulare by insist-
ing on what theory had, until now, nor thought through: morphology, syntax,
and the semantics of iconic sequences. Bur it would be denying evidence if we
128 Restrained Arthrology: The Sequence
held that the contribution of tbe verbal is negligible. Therefore, it is time now ro
interrogateehe process of raising the stakes of meaning between the two registers,
the iconic and the linguistic.
One has often ceded to the temptation of presenting cernies as a branch or a
subproduct of literature. lndeed, cernies are prinred, are sold in bookstores, and
the texts that participate in rheir discourse are given to us to read, not to listen to.
Bur, as for tbese texts, the fact rhat they share publishing forms does not allow us
to assign an identity to the function. Ifl prefer; for my own part, to speak of ver-
bal functions as opposed to written functions, it is because I think that speech in
cernies is doser to speech in tbe cinema than in the literacy text (even dialogue).
In spiee of the apparent nai:Vere of his comments, Herge touched on rhis when
he dedared, in 1942:
Iconsider my stories os films. Therefore, no norrotion, no description. Igive oll precedence to the imoge,
but my stories ore noturolly sound films ond l 00% tolkies, so speech comes out grophicolly from the
mouth of the chorocters''
Indeed, compared to a literacy stocy, ehe image translates and expresses in
visual terms all that it can: characters, decor, objects, atmospheric notations,
expressions, gestures, acrions-everything, in reality, except verbal exchanges
(and thoughrs), which the image is not able translate and can do nothing but
cite.39 Yet, among all the actions carried out by the characters, there is· one that
is specifically ehe act of spealting. The speech act inscribes itself in tbe chain
of actions and reactions that make up the story; it is an integral pan of the
framewerk of events. The caption, equivalent to the voiceover, encloses a form
of speech, tbat of tbe explicit narrator (who can be tbe principal narrator or ehe
delegated narrator, intra- or ~tra-diegetic, etc.).40
Using another code (digital, and not analogic), and reserving a contained
space, that of tbe word balloon, speech is simultaneously at tbe interior of ehe
image--"it emerges graphically from tbe mouth of tbe characters"-and diseinet
from it; This relative autonomy of verbal Statements allows them to be perceived
as links in a specific chain, parallel (or ratber interlaced) to those of tbe images.
In the same Way ·that the signification of an image is given by the sequence, it is
also ehe linking of tbe word balloons tbat should be taken into account in tbe
interpretation of verbal utterances. Theseare of tbe sott whereby arthrology man-
ages three levels of articulacion: The first two, which are homogenous, concern
rhe chain of images on one part, and ehe chain of word balloons on the other; tbe
Restrained Arthrology: The Sequence 129
third, heterogeneous, concerns the articulation of these two sequences, one the
iconic, the other the linguistic.
lndisputably, the visible presence of characters in a situation of elocution
within the image confers the status of an oral exchange to their inscribed speech
''i
in ehe word balloons. Materially, tbis speech is generally written, lettered on tbe
paper. Such is tbe constitutive and irreducible ambiguity of dialogue in com-
ics." This ambiguity opens several optionswich regard to ehe writing of dialogue.
Certain writers play ehe orality card, multiplying rhe effects of "natural speech"
(elisions, incomplete or incorrect phrases, familiar or trivial expressions, phonetic
transcription of accents given to characters, etc.). Jean-Michel Charlier is notable
here, as we found in discussing the Blueberry page (fig. 7), garnished wich inter-
jections and exclamations ("Hehe ... Tsstss ... Hee?? ... Ben ... Ah ... My
God! ... Hey ... Bon sang"). lnversely, a Jean-Claude Forest voluntarily intro-
duces into his word balloons a quantity of parentheses, hyphens, and incidents
that belong to tbe register of writing. The genius of a Goscinny and a Herge
undoubtedly lies in furging a language that is in reality vecy much written but
whose fluidity and rhytbm give it a natural appearance. We should note that no
autbor pushed ehe characterization of his characters as far as Herge by means of
linguistic mannerisms; noorher autbor has challenged tbe piefalls of familiarity."
Ir suffices to Iisten to ehe recordings on disc or to ehe dialogue of the animated
films inspired by The Adventures ofTintin to recognize at what point tbe idiolect
of ehe Thom(p)sons, for example, emerges as a natural arrifice, comical and cred-
ible on rhe printed page, but painful to hear.
If tbe statutoty ambiguity of cernies dialogue has inevitably brought me to
tbe terrain of style, I will returnwichout delay to my subject, which concerns ehe
function of text in the economy of a verbo-iconic sequence.
In a celebrated arricle, Barrhes named two functions of tbe linguistic message
in relation to ehe iconic message: anchorage and rehy. All images being polysemic,
tbe linguistic message helps to identify and to interpret ehe represented scene,
"they imply, underlying tbeir signifiers, a 'floating chain' of signifieds, tbe reader
able ro choose some and ignore otbers": 43 this is tbe function of anchorage. Wirh
respect to the other funccion, Barthes wrote:
The function of reloy is less common (ot leost os Im os the fixed imoge is concerned); it con be seen porticu-
lo~y in cortoons ond comic strips. Here text (most often osnotch of diologue) ond imoge stond in ocomple-
mentory relotionship, the words, in the some woy os the imoges, ore frogments of omore generol syntogm
ond the unity ofthe messoge is reolized ot ohigh er Ievei, thot of the story, the on<Kdote, the diegesis.~
130 Restroined Arthro1ogy: The lequence
We can reproach this anicle for not making a dear distinction between the
isolated image (which consists of a comical drawing, or the advertisement for
Panzini chac Barches analyzes) and che sequence of images (comics, phoco-novel
or cinema, of which Banhes tells us that rhe speech-relay is very important, since
it "cruly advances ehe action while laying out messages and meaning that ace
not found in ehe image"). Frequently dominant in ehe first case, che function
of anchoring che text sees its own imporcance considerably contexrualized in
the second. In comics, the image does not often need a linguistic message to be
anchored in a univocal signification. lt is not true that, wichout a verbal "crutch,"
ir will be condemned to polysemy. For what determines its signification, in ehe
first instance, and permits a reading in accordance ro ehe writer's program, is
precisely-and we have seen a sufficient amount of examples-its inscription in
an iconic sequence. The sequence itself exerts an anchoring fi.mction in relation
co each of ehe images chat compose ic, which consequendy discharges from ehe
rexr rhis responsibiliry, which ir assumes solely in the case of a unique image.
Wich chat said, we must specify which question(s) regarding che meaning of
rhe image will rhe rexr-and somerimes only the texr-be capable of answer-
ing. Lec's recurn, wich a new Iook, to che page from Corentin. In ehe first panel,
ehe phrase pronounced by Narreddine ("Hurry, lec's try co reach ehe Croc!") is
necessary for the comprehension of the situation. Wichout ir, I cannot assign a
meaning to che gesture of che old man. The arm designates someching, but chis
so mething could be a second group of enemies, ehe sky turning into :i storm, or
some ocher object chat might be heldoutside of che frame. !t is cherefore ehe text,
in chis precise case, which anchots che signification of che gescure and ehe panel.
If it is revealed here as necessary, che appeal of ehe text does not allow closure
ro an insufficiency of ehe iiD;age as representation (that is to say, like a semiotic
vehicle). !t simply recalls ehe fact chat our diflecent senses are channels of comple-
mencary information. If I find myself in ehe situacion of James Scewarr in Rear
Window, spying on my neighbors, I cannot understand a scene where rhe sound
does not reach me because it cakes place behind a closed window. Are chese peo-
ple who. are speaking wich forceful gesrures simply having an animated discussion
or are chey arguing? If it is an· argument, what is che pretext? The view alone does
not provide mtich informarion. Anyone can reconstruct this experience simply
by curning off che sound on cheir celevision. Similarly, in a "talking" comic (since
mute comics do exist), the intrigues of rhe characrers are only rendered entirely
intelligible because of che information provided by che dialogue, which is easily
ehe equivalenc, as we will verify here, to ehe soundtrack.
Restroined Arthrology: The lequence 131
To che cwo functions identified by Barches, Benoie Peetets suggests adding
a third: che function of suture, "by which che text aims co establish a bridge
becween che cwo separated images." !t is in che work of Edgar P. Jacobs chat
Peeters notes this function, which he seems to reserve to the captions that con-
scitute "che only elemencs capable of linking togecher cwo hopelessly disjoinced
panels on che visual plane."<> Let's be exact: ehe suture can also easily be confined
to dialogue. The fiese example chat comes to mind figures in che eighch page of
The Castafiore Emerald, where rhe repeat of rhe word "sprain" allows a reduction
of a formidable ellipse. "Noching broken, I hope?" Tintin asks Captain Haddock,
who has just fallen down ehe scairs. "Luckily not. Though I might easily have
sprained something!" answets Haddock, jusc before lecting out a cry of pain. In
che following panel, a doctor gives his diagnosis: "It's a bad sprain ... And you've
pulled rhe Iigaments." Between these two immediately consecutive images, we
have to imagine Tintin getting Haddock to sie down, searching for ehe doccor's
number, telephoning him, waiting for him to arrive, opening the door for him,
and assisting wich ehe medical examination. But ehe repetition of a single word
allows us to jump over these trivial episodes, as well as suturing a farnaus tear
in ehe Iigaments of che story. Nevettheless, ehe sucuring function is in ehe end a
particular case of the relay funcrion.
Symagmic cohesion, which allows che narracion to unravel wichout obscacles,
is assured by the cooperation between the iconic sequence and the linguistic
sequence. This cooperation allows ehe medium to play wich a very !arge range
of effects, be chey comic, poignant, or dramacic. !t is undoubcedly in humor
that the complementariry between these two instances sometimes Ieads to rhe
most subtle findings. In chis domain, che anchoring function is often recurned:
indeed, it is frequendy che image chat holds che keys chat permit a plausible
interpretation of the text, for example, in making explicit what an understate-
ment or a euphemism is implying, or again by denouncing, by its trivialiry, the
emphasis or ehe deliberate pomposity ofa pseudo-literary style. Elsewhere, I have
analyzed these ironic figuces in Töpffer," but chey also abound in a Gotlib or
a Goossens, who, diflecendy from che Genevan, use not only indirect style but
dialogue.
Anocher function of che text, generally confined solely to che captions, which
we can call the controllingfunction, concerns rhe management of narrative time.
The mosr convenient way that the narrator indicates ro the reader the large tem-
poral scansion of the story is to effectively resort to verbal announcements ("mean-
while," "one hour later," "that night," "the next day"). In rhis respect, comics are
132 Restrained Arthrology: The Sequence
akin co ehe phocographic sequences of Duane Michals, of which Daniele Meaux
remacked: "The chronology of ehe events that are represenced in a sequential
fashion is moscly defined by verbal information that can indicate ehe presence of
analepsis (flashbacks), prolepsis (flash-forward) and ellipsis."47
As ehe ellipsis is ehe basis of ehe discontinuous language of comics, these indi-
cations only concern ellipses of great amplitude, those that, coinciding wich a
scene change, generally implicate a jump in time and a displacement in space.
Bur, the temporal hiatus can be vouched for by strictly iconic means, as can the
spatial hiacus. A change in place is effeccively a monstration of a new decor, which
ehe reader can, in principle, identify as such; furthermore, the passage from one
scene co another is frequencly highlighced by a modification of ehe dominant
chromatic, the irrtemal coherence of each scene being generally attested by a
range of homogenaus colors that contribute to its dramaturgical and emotional
impact. When the case arises, rhis chromatic rupture often takes advantage of the
natural cantrast berween ehe two temporal sequences, opposing day for night or
ehe light of ehe dawn for ehe noonday sun (meceorological vaciations and oppo-
sirions berween interior scenes and exterior scenes also figure in the nurober of
evecyday juscificacions for changing ehe dominant look).
The image can, in almest all cases, assure the function of ehe control wichout
any linguistic contribution. Thisis proved, par excellence, in mute comics. (I am
chinking pacticulacly of Peter Kuper's The System, where ehe complex narrative,
8
developed over 104 pages, shines wich ingenuicy and ehe diversicy of transitions.)'
But works chat do not belong in this categocy can also impose this constrainc: lci
Meme, ehe "ralky" masterpiece by Forest and Tacdi, foregoes all captions, even co
indicate changes of place or temporal jumps.
My experimental reading of ehe page from Royaume des eaux noires showed
chat, by way of linguiscic padding, we can always nacne ehe semantic bridge for
each specific gap chat is represented in ehe passage from one panel CO ehe next.
This implicit text emerges sometimes on ehe surface of discourse: Ir correspond.s
precisely to governing Statements.
The last function rhat ehe verbal pole of comics exerts seems to me to be a
rhythmic fonction. Following· many others, Jan Baetens and Pascal Lefevre have
again raised ehe noti~n chat comics play wich a "temporal gap between ehe per-
ception of ehe image, which is almost global and quasi-simulraneous, and ehe
course of verbal signs, which is slower andin all cases more gradual. " 49 Bur to me,
the two Belgian exegetes are mistaken when they write rhat comics are subjected
to "the imperious necessity to minimize" this gap, as the linguistic messages risk
Restrained Arthrolagy: The Sequence 133
slowing down and confusing "the browsing of im.ages made, in a conventional
regime, for reading wich a sustained rhythm. " 50 On ehe contracy, a piece of dia-
logue is frequencly incroduced into a panel as a way co slow down ehe reading.
While there exist other examples, we have encountered an instance of rhis order
in ehe fifrh panel of ehe page from Corentin. The words of Za:ila form an improb-
able and too elaborate reply for ehe extremely urgent sicuation in which she finds
herself. Since ehe image corresponds to an instantaneous moment, ehe words:
"No, little roumi. Save yourself! Run!" prolong our participation in this emi-
nencly dramatic moment.
The presence or absence of a text, the eventual division of a verbal Statement
inco several balloons, ehe distribution of chese balloons in an equivalent or lesser
number of panels (according to whether or not they ace reunited in gcoups wiehin
ehe sacne fracne), ehe alternation of ehe dialogue and ehe capcions: so many ele-
ments contribute to imparting a rhythm to the narrative sequence-and a dura-
tion. I noted eaclier: even at ehe interior of ehe image, ehe gacne of replies (for
example, a quescion followed by an answer) can inscribe ehe passage of time chat
is macked in ehe desynchronized attitudes of ehe characters (each one living ehe
moment oftheir word balloon).
The pages by Teule and Cosey chat were commenced upon in ehe firsc pan
of this book can be invoked again here. The manner in which ehe linguistic
Statements are distributed effectively constitutes a significant contribution to the
internal rhythm of these two pages. Wehave observed, from a purely visual point
of view, how ehe incernal coherence of each ofTeule's chree stripswas established
(fig. 6). To complete this first analysis, it is advisable to add chac this compound
suuccure is also found at ehe verballevel. Direct dialogue in ehe firsc strip, cap-
tions in ehe second, off-screen dialogue in ehe third (ehe speakers being out of
view): each scrip has, under this aspect as well, its diseinet identicy. The page by
Cosey (fig. 9) is comprised of live panels, where ehe last panel is ehe only one chat
"speaks." This unique word balloon, sicuated in ehe lower righc-hand corner of
ehe page, therefore corresponds co a rupcuring of ehe silence. Furcher, ehe ques-
tion posed by young Keo is not important: It only serves to stan the conversation
wich Art, ehe second character. The true object of Keo's curiosicy appears in ehe
following page: Art's emotionallife. The word balloon that punctuates page 44
of Voyage en ltalie serves a dual function co ehe rhychm of ehe sequence. On ehe
one hand, arriving after an image that suggests a duration that is impossible to
quantifY (ehe two characters lost in ehe concemplation of ehe sea), it assigns an
end point co this temporization and brings ehe reader back co ehe present of ehe
134 Restroined Arthrology: The Sequence Restroined Arthrology: The Sequence 135

action. On ehe other hand, this speech act, at the end of an almost mute page,
opens onto a following page that is almost entirely filled with dialogue: page 45
closes, in symmetry wich the preceding page, on a mute image signifYing the
return to silence and contemplation.
We have acknowledged seven different functions of the verbal wiehin the
economy of comics: the effect of ehe real, dramatization, anchorage, relay, surure,
control, and rhythm. Among these functions, we can count three (the first two
and the last) that are turned toward the referential illusion and the production of
ehe story. The four others, in sum, are panicular cases of what we can summarize
as the informative function of the verbal. The point is, in each of the considered
cases, to demand from the linguistic Statements the completion of ehe informa-
tion delivered by the chain oficonic utterable. Ifl am unsure whether film theory
analyzes film dialogue in these same terms-the realiry is that wich rare excep-
tions, dialogue does not count amongst the privileged objecrs of srudy-it surely
seems to me that that these functions are also percinent, acting for the seventh
att as well as for the ninth art. One may object that the function of control is, on
screen, more rarely entrusted to the verbal. Bur the truth is if these large temporal
ellipses are rarely indicated by an off-screen voice (that of the narrator), they are
communicated most frequently by several words incrusted in the image, which,
in ehe perspecrive developed here, rerurns to irself.

2.9-AN EXERCISE OF TRANSLATION

To verifY and recapitulate all that we have seen about the operation of the
sequence, it appears opportul'l:e to consider a new example, and to use it as a
pretext for a short exercise. Ir is a simple matter of converting a page of cernies
into its linguistic equivalent, or if we prefer, to translate a series of utterables into
a series of utterances. The pedagogical interest of this exercise should not escape
language teachers, whom I hope to inspire. While the exercise is more or less at
the college level, I will deliberarely rake a page extracted from a comic readable
from the age of six and up (fig: 12), page 17 of the album Un ettf du tonnerre, ehe
Rg. 12. From lojo #5: Un eti du tonnerre, by Andre Geerts (1992). © Edilons Dupuis.
fifrh volume of the seri,;;]ojo by Geetts (Ed. Dupuis). The choice of a comic that
is assumed to be very simple--but otherwise of excellent qualiry-responds to The rules of ehe exercise are as follows. One afrer another, the panels will
the concern of verifYing that ehe quasi-toraliry of the questions raised up until be translated into linguistic terms, wich the resulring Statements being as con-
now does not concern this orthat particular cernie chosen for its sophistication,5 I cise as possible, always wichout omitting anything essential to the intelligibiliry
but rather the medium itself in its constitutive operations. of ehe action. The Statements will be subjected, if need be, to several necessary
136 Reslroined Arthro1ogy: The Sequence
adjuscments so chac ehe whole produees a satisfYing written page, that is to say,
a written text that reads as fluid as Geerts' page. We will then incerrogate ehe
remainder of what chis exereise produeed, to eome to termswich a11 that the page
showed, said, or suggested, and rhat our linguistic translation did not take into
account.
The page in question-so that this will not be too long, I will Iimit myself to
ehe upper half of ehe page and ehe last panel-having been separated from ehe
cotalicy of ehe album, it is neeessary to know what preeeded it, or at lease have
access to a summary. Here it is:
Momy brings Jojo ond h~ lriend Gros-louis for o vocotion ot the form belonging to her cousin Angelo.
Angelo's fomily consists of her son Odilon (o force of noture), her doughterirl"iow Emmo, ond her 1wo
little children, Thomos ond Morie. Since h~ orrivol, Gros.J.ouis hos been struck with Iove ot first sight for
the young Morie. We find our lriends on the moming ofter their orrivol.
Here now are ehe statements that I propose to account for the three panels of
ehe first strip.
Panel 1. The sun rises on rhe farm, salured by ehe roosrer's erow.
Panel2. Gros-Louis awakens wich a yawn.
Panel 3· He gets up and pours a pieeher of water into ehe basin.
Ler us stop here on our first eomment, wich regaid to ehe scrip irself, and ehe
proposed cranslation. Wich respeer to rhe srrip, we can first of all be rid of ehe
eaption that adorns ehe first panel: it is an indicacion of eoncrol, whieh is quiekly
revealed to be redundant in relation to whar ehe irnage shows us, and there-
fore unneeessaty (exeept perhaps for ehe young child chat is still insuffieiendy
used to ehe reading of images). The erow of the roosrer is a symptom of ehe
starr of che day, reused so many times (nocably in comics) that it has almost
become a metaphor. My Statement places the accent on the rise of the sun rather
than its consequence or its corollary: the crow of the rooster. The farm does
not appear in ehe i!llage. Nevertheless I mention it to contexrualize this panel:
indeed, ehe sunrise interests ehe story only insofar as it announces the awaken-
ing oecupants of ehe facrn. But I eould have easily wrirren: "The sun rises on
rhe field."
There is nothing chat indicares to us wich cerrainty if Gros-Louis wakes up
alone, if he heard the rooster's crow, or if it is ehe rays of the sun (that we see
Restroined Arthrology: The Sequence 137
penetrating into the room) that make him open his eyes. It is naturally con-
venient, in a transcription, to avoid all speculation on hypotheses that are not
proven. Speculation is not interpretation, to which we have just appealed, but
rather it is an abusive interpretation.
In ehe third panel, Gros-Louis is already standing. If I had translated it by
itself, by making a eomplete abstraetion of eontext, I would have simply writ-
ten: "Gros-Louis pours a pieeher of water into the basin." Bur we advance in
the reading of comics, as we do with a text, under the control of what precedes
us. Becween ehe seeond and third panel, I am obliged to postulate an ellipsis in
order to preserve the coherence of the story. I did not see Gros-Louis get up,
but I know that he did because afrer having seen him lying down I suddenly
discover him standing. The Statement that I propose reestablishes the continuity
of the action; it takes into account what I am sure has happened, even if I did
not see it.
Granted, I think my three Statements translate grosso modo that which all read-
ers, even when reading quiekly, would have pereeived, understood, and retained
of this inaugural Strip. It is the minimum rhat is required to continue with full
knowledge of ehe faets. Reading less than what we have transeribed would have
resulted in losing ehe thread of ehe stocy. Buc ir is clear that these images give us
infinitely more to see than what I have retained. The eonversion of utterables
to utterances-an operation that correspond.s approximatdy to what a hurried
reader would mencally accomplish, wich only one eoneern: to find ehe intel-
ligibilicy of ehe stoty in order to follow ehe episodes--<:onstitutes a formidable
reduction of the work to the sphere of the action and of the event, a mutilation
that retains only che aetantial chain, which is only interested in whac happens to
ehe protagonists.
If, given a second ehanee, I eonsidered ehe image as deseriprible, I would not
only modifY my regime of pereeption, I would aecede to a considerably !arger
inrelligence of ehe work. Deseriprion, as we have seen, is a split operation. If I
had taken ehe care ro deseribe these images by Geerts in their ieonic dimension, I
would discover all sorts of unnorieed details, for example ehe alarm clock on ehe
night table indicating chat it is seven o'cloek, or rhat Gros-Louis has to stand on
ehe tips ofhis toes co pour ehe pieeher ofwater inro the basin. Ifl eonsider ehern in
their plastic or graphic dimension, I would be sensitive to rounded forms, to ehe
tender eolors, to a11 thar rhe sofr, supple, and outmoded aspeets that Geerts's style
confers-Isn't ehe name of ehe village Avantieres?-and chat so properly evoke
a noscalgia for childhood, to all that the images signifj. !t must be recognized
138 Restroined Arthrology: The Sequence
chat the distinccion berween ehe iconic dimension and the graphic d.imension
is not always evident. So, the pietute frame hanging above tbe bed undergoes
several variations between tbe second and tbird panels. The house is a litde !arger
in its second occurrence, and ehe shape of ehe roof is different. Bur how do we
decide if tbese variations are deliberate--This would therefore be a little bonus,
free of charge, for ehe attentive reader-or if they are ascribable to a casual execu-
tion, if they are an iconic decail or a graphic accident?
Let's move on ro the second strip. It is distit1guished from the first strip by the
fact that it involves dialogue. How will the Statements that I write take this into
account?
Lec's make a first attempt that consisrs of abstracting the dialogue.
Panel4. While Gros-Louis is washing himself, Jojo, who was sleeping beside him,
takes his time to get up.
Panelj. But he finds it difficult to get hirnself up from the bed.
Panel 6. Gros-Louis combs his few hairs using gel.
The reader cannot help but be surprised to discover tbe presence of)ojo (invis-
ible up to tbat point) in tbe bed. This presence was not entirely above suspicion:
in the third panel, we can see that, even after Gros-Louis has risen, ehe mattress
remains sunken and its springs held, attesting to anotber body that is. holding it
down. Conrrast tbis ro the srate of tbe mattress in the sixth panel, where Jojo is
sirring on tbe exterior frame of the bed.
Throughout this second strip, tbere are two characters present. From a percep-
tive point of view, the reader. is therefore irrteresred in what each of ehern is doing
at every moment. Two parallel actions are engaged Qojo awakens, Gros-Louis
washes himself), berween which tbere exists no a priori hierarchy. In my verbal
translation, I did not believe tbat I needed to sysrematically mention one and
tbe other. It would be possible to do so, but only at tbe price of a formulation
that is heavy, complicated, and inelegant. As Theophile Gautier complained in
Le Capitaine Fracasse, "the artifice of tbe writer has an inferioriry to tbat of the
painter, in that he can only show objects in succession." What the image suggests
ro me with a single glance, tbe words have ro evoke piece by piece. If I do not
speak of what Gros-Louis does in the fiftb panel, it is because the fact of taking
a pot of gel &om the drawer will be implied by the fact that he uses the gel: tbe
ellipsis can tberefore be conceived. In tbe sixtb panel, it is Jojo that is left aside.
Restroined Arthrology: The Sequen<e 139
In effecr, it appears tbat he is engaged in no action otber tban speaking--and we
arrive at tbe abstraction of tbe dialogue.
Let us take note: tbe dialogue is not indispensable to tbe comprehension of
the sequence, nor in the rest of tbe page. lt will not carry over into tbe following
page. By tbemselves, the images render all of the important information of the
actanrial sphere perfi:cdy intelligible: Jojo and Gros-Louis descend to have break-
fast; the otbers arealready seated at the table; Gros-Louis sits next ro Marie; tbe
two litde boys are alarmed at tbe gargantuan appetites of their hosts; little Marie
is distinguished &om her family by only nibbling on a biscuit.
Does tbis signifY tbat the dialogue is superfluous, or even parasitical? Not
at all. We have seen tbat what attaches to the dialogue in general is an effuct
of tbe real: we see here that rhey are a condition of tbe "realism" of tbe scene. If
tbe two little boys stood up without exchanging a word, and if the entire family
ate breakfast in silence, we would find this abnormal, Strange, contrary to our
experience of tbe real world, and tbe whole page would find itself "de-realized."
Therefure, tbe dialogue is necessary here, even tbough it may be nothing more
than inconsequential chatter.
Bur, it is more tban this. Twice in the page the dialogue exercises a function of
relay and completion, in an otherwise subcle mode, of the given information in
tbe image. In tbe two cases, the cooperation berween the iconic and tbe linguistic
provides access to a more prolOund reading of the panel, an interpretive reading
that invites the memories of anterior episodes. This is demonstrated in the last
panel tbat we converted into a Statement. The verbal exchange could pass as
unnecessary. Jojo does not need to ask Gros-Louis if he used gel because he can
see it, and Gros-Louis does not need to indicate that he rook it from tbe drawer
because we saw him do it. But the raison d~etre of this exchange is elsewhere:
Jojo's shock is intended to make us understand that Gros-Louis does not usually
use gel; unusual behavior, extraordinary reason: he has fallen in love and wants
to make hirnself look good. (This attention to appearance produces its result
elsewhere, since in tbe eighth panel, Marie tells him: "Hmm! You smell good!")
The reader accedes to tbis reasoning because pages 4 and 15 of the album have
already established tbe romanric Sentiments of Gros-Louis for Marie. Therefore,
the verbal element here is important, since it is his mediation that allows an
interpretation to be constructed. !t is the sarne with respect to tbe last panel of
tbe page, which I will address below.
Previously, we had to experiment with other Statements to take account of the
three panels that compose the second strip, Statements that take into account the

140 Restroined Arthrology: The Sequence
dialogue. Several solutions are possible. The first consiscs of licerally citing ehe
speech pronounced by ehe characters wirhin ehe statement.
This would give us for ehe fourch panel:
While woshing, Gros-Louis interrogrrtes Jojo, who slept beside him:
"You're still sleeping, Jojo?"
"No," he replies, "l'm trying to get up.'
The secend solution consists of integrating ehe meaning of ehe speech to the
stacement, by summarizing or paraphrasing it. Applied to che fifth panel, this
would have us write:
But Jojo experiences some difficulty in getting out of such osoft bed.
The precision on che consistency of che bed was added to ehe proposed
statement above, in an echo of Jojo's exclamarion: "l've never seen such a soft
bed!"
Finally, rhe rhird solution-more discriminating, and therefore, .to my sen:e,
more pedagogical-consists of only making reference to dialogue 1~ a selecove
fashion, that is, only once it has delivered a complement ofinformanon that 1s of
real use co che intelligibility of che stoty. From this perspective, I will·rake a pass
on the dialorue of panels 4 and 5, conserving their inicially proposed Statements.
On che othe: hand, I will modifY che initial Statement corresponding to the sixth
panel, conforming co ehe particular interest that we have rc:cognized in the verbal
exchange chat it contains. T?e Statement becomes somethmg hke th1s:
To Jojo's surprise, Gros-lou~ uses gel to comb his hoir.
Before jumping direccly into the last image, I will make an observation con-
cerning the seventh paneL 1t is che largest panel on this page, and th~ one that
includes che most characters: nine, counting the litde black cat. Th1s type of
panel is predisposed to arouse a more descriptive readi~g. lts form~t i~ stretched
to a size chat compels a lateral sweeping, from left to nght, and th1s V!sual raute
coincides almest naturally wich an enumeration of the different motifs encoun-
tered (principally: che characters), which, by consequence, it favors. In the mea-
sure that characters reappear here that we have momentarily lost sight of and
Restroined Arthrology: The Sequence 141
who, for ehe most part, are not even familiar to us, ehe reader is more attentive to
recognizing each and noting the activity in which he is engaged. I only make this
Statement to suggest the following: if all the parreis are utterables, a few of them
make better decriprables than others.
I now arrive at ehe last panel of our page. To convert it into a statement such
as "Marie nibbles on a biscuit" would not be false, but this Statement, in its
brevicy, would not do justice to the interest that chis panel presents, and describe
its humoraus dimension. It happened that I worked on this pagewich some stu-
dents, where a cerrain number did not perceive the humor atrached to this panel
and which gives it its flavor; they needed me to explain it to chem. I will explain
it here too, not wichout noting that ehe fact of terminating ehe page predisposes
this panel to coincide wich a form of unspecified "punchline," all ehe more so
since the end of the page is also the end of the sequence. If this pagewas entided
"an awakening at the farm," the next page effectively opens a new four-page
sequence that would be called "a tour by tractor."
It is here that the dialogue, which we cannot atrribute to any precise speaker, in
taking over from the image, makes us smile: "She is delicate." In fact, the delicacy
of Marie is a recurrent theme in the album, as chis quality is associated wich her
several times. If this eherne is funny, it is primarily because Marie offers a knciwing
cantrast wich her parents, both of whom are obese; ic is also because Marie, if she
is of a delicate constirution, cercainly does not have a delicate behavior. Ic is on
page 14, when Odilon compares her to a little princess, that she curses loudly (the
first in a long series of"Vingttiju!"), which drives Angela to highlight Marie as the
"well-bred peasant." Here too, the panel revcals that the "delicate" Marie eats a
little noisily. But, by a second degree of the humor, the lettering of the onomato-
poeia issmalland discreet; contrasred wich ehe lettering of the onomatopoeia in
the ninth panel, it is proportioned to Marie's height and to her small appetite. !ts
smallness is highlighted by the !arge empty space above the child.
The place that the world of the peasant gives to delicacy is a theme that is
provided space in a number of humoraus variations in ehe album, where it con-
stirutes a leitmotifor a privileged topos. The meaning of each of the occurrences of
this theme can only emerge in full measure when the readerlinks each one to the
others. This nerworked reading is what allows ehe transcendence of ehe enuncia-
tive Ievel (where each panel is considered in its proper immediate signification)
to give way to that of interpretation.
It would serve no profit to push us further in our transcription exercise. But co
those who wish prolang it co the end of the page, feel free!
142 Restroined ArthrollllJY: The Sequence Restroined Arthrology: The Sequence 143

2. I 0-BREAKDOWN AND PAGE LAYOUT The adapration of a Story into the language of comics is evidencly not the
singular responsibiliry of the breakdown. As we have shown in the preceding
Jean-Claude Carriere likes to repeat that, in the cinema, there is no essential differ- chapter, there also has to be an intervention of an accomplice, the page Iayout,
ence berween an original scenario and a screen adaptation of a pre-existing rexr: that distributes the narrative chain in space, therefore adjusting a matter already
sequential to a spatio-topical apparatus, and assigning to each panel unit in nar-
ln reolity, everything is odoptonon in the dnemo. When Iom osked to find ostory in onovel, when l'm rative situation its form, its surface, and its setting.
osked for odive5e loct, ope5onol memory, thot Iseorch in my own memory, or in my 'imoginonon," ... Breakdowns and page Iayouts mutually inform one another. Neither of these
in oll manne~ my work will be one of odoptonon. I hove to transform th~ vogue ideo, or book, or this rwo Operations is in a position to control the other. In this regard, the artists
onecdote, into ofilm'' diverge. The page layout precedes the breakdown when it is arbitrary, that is to
say, regulated or controlled by an a priori formal constraint (when the frames are
Similarly, we must conceive the writing of a comic as an adaptation of a narra- defined without regard to the content that allots the breakdown). On the con-
tive project ro rhe particular resources and exigencies of the medium. The break- rrary, the breakdown precedes the page layout in the majority of cases, even if it
down is the firsr agent of rhis process. Ir seizes pre-exiscing narrative material has to undergo certain adjustments as the layout specifies. When the page layout
(drawn up or not, somewhat vague oralready structured), and it transfurms this is second, we have seen that it can accompany (and if need be, higblight), on the
fable or rhis dlscourse into a succession of discrete units, the panels. which are fre- rhetorical mode, the intentions that animate the break.down, or create its own
quendy associated with verbal utterances, and that are links of a narrative chain. pattition, seek tO produce a decorative effect. But in this last case, well animared
Thesepanelsare equipped with a frame (if only virtual), which designates them by a rival intention, it continues to tak.e account of the breakdown since, for Iack
as separate entities, enclosing each wirhin a meaningful fragment. of exalting it or serving it, it must in any case not interfere with narrative conti-
To each panel corresponds a situation in the flux of the story. This situation nuity. Definitively, the reading protocol required by a page is always the result of
arises from what we can call the chrono-topia (or temporal segmenration); it also an action conjugated by the breakdown and the page layout, the product of at
corresponds to a place in the structural economy of rhe sequential discourse. least rwo determinations.
The breakdown distributes information: !t attributes to them a· mode of
enunciation (iconic or linguistic}, then distills them in time by organizing their
diaehrenie cooperation and their reciprocal determinations.' 3 It ultimately com-
mands the mise en scene, that is to say the coordinated utilization of all the
parameters oficonic enunciation, in the measure where it takes part in the narra-
tion and conditions the perc~ption and interpretation of the reader.
(Whether comics are a product of a complete auteur or the work of a collabo-
ration changes nothing in this theoretical postulate. Even ifit often works on the
basis of propositions already specified by the scriptwriter, the role of the illustrator
is cerrainly dominant in what I call the mise en scene. This fact should not drive
us to disassociate the mise en scene from the break.down; on the contrary it allows
the establishinent that ·me cernies illustrator, far from being a simple illusrrator, is
fully implicated in the narration. Scriptwriter and illustrator each have their patt
in the mise en sd:ne at the point of ehe breakdown. Furthermore, in these pages,
I do not look tO concretely describe how the work is cartied out by the various
agents of creation, but to define the objective components of a language.)

CHAPTER THREE
GENERAL ARTHROLOGY
'
:I,,
The Network
~.
3. 1-THE STAGE OF GRIDDING ( QUADRII.I.AGE)
Befure cackling ehe domain of braiding, which I define as beyond breakdown, it
is advisable to say a word about rhat which comes before rhe Iayout, which I pro-
pese to give the name gridding. It is an operation (or at least a srage of reßection
rhat is not always incarnated) that intervenes very early in rhe process of elabora-
tion in comics; it defines ehe apparatus of rhe cernie prior to its actual appear~
ance. Griddlng consists of dividing the available space into a nurober of units or
compartments. While remairring in question, it operates as a primary reparrition
of the narrative material.
At ehe Ievel of rhe total space of the work (for exarnple, rhat of an album),
gridding starts at the instant where the writer divides the work into chaprers or
sequences, when he seeks t~ evaluate cheir respective lengths (in rhe number of
pages). In ehe hands of the illustrator chis can materialize in rhe elaboration of a
complete storyboard, or in the form of chumbnail sketches of each of the pages.
Gridding is cherefore an approximate equivalent of what is known as prepro-
duction in the cinema, an end-co-end Iayout of ehe shots. Ir is with this essen-
tial difference chat comics begirr where cinema ends: the nature of rhe finished
form does not allow 'ehe illustrator to produce imageswirhaut some preliminary
knowledge of cheir locacion in space and cheir location in rhe story.
Applied to the manufaccured unit rhat is ehe page, gridding corresponds to rhe
moment of taking possession of the original space. The operation, as I have said,
can remain purely mental and may not have a genuine graphic translation. It can
General Arthrology: The Network 145
also be realized in a sketch or a grid, which would have a double vircue: to over-
come the intimidating effect of the blank page, and to announce the emergence
of rhe drawing.
Thus, for Herge, rhis consisted of tracing, on a draft page, chree horizontal
lines divided into four Strips of more or less equal importance, in which a sug-
gestion of ehe breakdown could be developed. This is handed down to us by rhe
moving testimony of rhe forry-second (and last) page of sketches ftom Tintin et
l'Alph~rt (Casterman 1986) where, under a completed first strip, rhe second is
only Started and rhe last two remain untouched.
Gridding, rherefore, defines ehe first, and often crudest, configuration of rhe
multifrarne. This provisory configuration furnishes ehe aurhor wirh a framework,
a matrix. The page layout becomes an improved and corrected version of the
gridding: that is to say the version informed by the precise contenrs and by the
two other constitutive operations of arthrology, rhe breakdown and (if rhe case
arises) rhe braiding.
It semetimes happens rhat the contents obey a strict periodizacion imposed by
ehe narrative program. Thus, gridding is revealed as an essential operation, which
aims to assign to each story sequence a part rhat is determined by the support.
Dessous troublants (Futuropolis 1986), rhe album by Jeanne Puchol, provides us
with a good example. The action is concentrated in an apartment composed
of four rooms: bedroom, office, kicchen, and washroom. Each room is evoked
rhrough four characteristic objects (thus for the office: a book, an armchair, a
larnp, a pen). The number four governs the narrative breakdown, rhe sequences,
each consecrated to the exploration of one of the rooms, each counting four
pages; a door appears that effectuates the transition to ehe next room in ehe last
panel of rhe fourrh page. Finally, rhe epilogue of this cwenty-page album consiscs
of four pages, which makes us successively return througb each of the four rooms.
As simple as it is efficient, gridding-as shown here: the division of rhe scory
into five sequences (4 + 1) of ehe same length-furnishes a global matrix, wiehin
which the apparatus for rhe page Iayouts are constandy renewed.
3.2-A FIRST APPROACH TO BRAIDING
Less commonly used than ehe concepts ofbreakdown and page Iayout, ehe idea of
braiding, which I briefly imroduced in a special issue of Cinbna.Action published
 146 Generol Arthrology: The Network
in surnmer 1990/ has nevenhdess known, since its debut, some critical fonune.
The time has come to give it a little more precision.
,· !t has been often repeated in these pages that wirhin rhe paged multiframe
that constirutes a complere comic, every panel exists, potenrially if not actually,
in relation with each of the others. This rotaliry, where the physical form is gener-
ally, accotding to French editorial norms, that of an album, responds to a model
of organizacion that is not that of the strip nor that of the chain, but that of the
network.
Jan Baetens and Pascal Lefevre have jusdy noted that "far ftom presencing
irself as a chain of panels, rhe comic demands a reading capable of searching,
beyond linear relations, to the aspects or ftagments of panels susceptible to being
networked with certain aspects or fragments of other panels."' Braiding is pre-
cisely rhe operation that, from the point of creation, programs and carries out
rhis sort ofbridging. !t consists of an additional and remarkable structuracion
rhat, raking account of rhe breakdown and rhe page layout, defines a series wirhin
a sequencial frarnework.
lt is important here to clearly distinguish the norion of a sequence and that
of a series. I recall, without modificarion, the definicions that I gave at Cerisy in
198?:
Aseries ~ u succession of continuous or discontinuous imoges linked by o system of iconic, plostic or
semontic correspondences.... Asequence is o succession of imoges where the syntogmic linking is
determined by onarrolive proiect.3
(As weil as being infra-narrative, the nocion of the series is already opposed to
rhat of the suite, which designates a colleccion of disparate uncorrelated images.
Apart ftom the fact that they inicially stem ftom a mathemacical terminology,
these terms have a frequent usage in the histoty of art and aesthetics. For exam-
ple, "Suite, series, sequence" is the eitle of a page by the writer Herve Guiberr,
whose definicions barely differ from mine.4 !t is also the tide of a volume col-
lecting rhe acts of a symposium organized by the University of Poitiers and the
College International de Philosophie.)'
The series rhat give birth to braiding are always inscribed wirhin narrative
sequences, where the first sense, independent of the perception of the series, is
sufficient in itself. Theseries is inscribed like an addition that the text secretes
beyond its surface. Or, to put it in another way: Braiding is a supplementaty
Generol Arthrology: The Network 147
relation rhat is never indispensable to the conduct and intelligibility of rhe story,
which the breakdown makes its own afF.air.
3.3-FROM SITE TO PLACE
Contraty to breakdown and page Iayout, braiding deploys itself simultaneously
in two dimensions, requiring them to collaborate with each other: synchronically,
that of the co-presence of panels on the surface of the sarne page; and diachronic-
ally, that of the reading, which recognizes in each new term ofa series a recollection
or an echo ofan anterior rerm. A tension can be established between these two log-
ics, but far from ending in conflict, it resolves itselfhere in a semantic enrichment
and a densification of ehe "rext" of the comic. (The term braiding is inscribed in
the toposthat habitually associates notions of tissue or threads with the texr.)
By irs nature, a story develops in length in a linear and irreversible manner.
lnherem to all narration, this progression finds irself reinforced by rhe printed
form, the "bookish form" (mise en Iivre) of ehe comic. As Jan Baetens writes, "ehe
book irself induces an undeniable vectorization of discourse; the book consecrates
a linear, or more exactly monovectorized, reading, rhat distinguishes (and some-
times discriminates) a start and an end, an incipit and an explicit, a first and a last
of ehe cover. " 6
With respect to comics, this disposition finds itself consrandy embatded,
and in a certain measure neutralized, by the properties that we have seen in ehe
pands. The nerwork that they form is certainly an oriented network, since it is
crossed by the instance of rhe stoty, but it also exists in a dechronologized mode,
that of the collection, of the panoptical spread and of coexistence, considering
the possibility of translinear relations and plurivectoral courses.
To pur it in the vocabulary of Michel Tardy, reading in this case is the opera-
cion that puts into tension the plane ofthe process and the plane ofthe system.' The
panels of the disseminared series only form a constellation to the degree that the
reading detects and decrypts their complementarity and interdependence. lt is
the very efliciency of braiding that incites this crazed reading.
Wirhin the network, each panel is equipped with spatio-topical coordinates
by the page Iayout that constitutes it as a site. These temporal (or chrono-topical)
Coordinates are themselves conferred by ehe breakdown. Braiding overdetermines
the panel by equipping Coordinates that we can qualify as hyper-topical, indicating
their bdonging to one or several nocable series, and the place that it occupies.
148 General Arthrology: The Network
As it is articulated to several ofits likenesses by a relarion rhat comes under the
jurisdicrion ofbraiding, ehe panel is enriched with resonances rhat have an effect
of transcending the functionality of the site that it occupies, to confer the qual-
ity of the place. What is a place other than a habituated space that we can cross,
visit, invest in, a space where relations are made and unmade? If all rhe terms of
a sequence, and consequently all the unirs of the nerwork, constitute sites, ir is
rhe attachment, moreover, of these units ro one or more remarkable series, that
defines ehern as places. A place is therefore an activated and over-determined site,
a site where a series crosses (or is superimposed on) a sequence. Cerrain privileged
sites are naturally predisposed to become places, notably those that correspond
to the initial and final positions of the sroty, or the chapters that compose them.
(Thus, in Little Nemo, a serial where each weekly page was a chapter in itself,
the final panels constitute a remarkable collection of the hero's awakenings.) Bur
other places do not coincide with any privileged sites; it is the effect of braiding
8
thar brings ehern to our particular attention and consrrucrs ehern as places.
Iris now time to give some examples of these remarkable structures rhat define
a series. I will not attempt eo sketch a typology of ehe specific diverse procedures
ofbraiding here, as rheywould no doubt be impossible to enumerate. I will con-
eent myself wich demonstrating several of them, across exarnples presented in ehe
rising order of rheir amplitude, rhat is ro say, the distance separating the rerms of
ehe series.
The strip, the page, rhe double page, and rhe album are nested mulriframes,
systems of increasingly inclusive proliferation. The first three have an essential
property in common: They allow a dialogue in praesentia, a direct exchange
berween images that are in a situation of co~presence under the gaze of the reader.
While if a panel from page 5 maintains a privileged relationship with a panel
belonging to page 6 on. the reverse side, or with a panel from page 27-as a
simple example, imagine that the second panel is a reproduction of rhe first-this
relation esrablishes itself in absentia, at a distance. The correspondences handled
by braiding frequendy concern panels (or pluri-panel sequences) distant by sev-
eral pages, and that cannot be viewed simultaneously.
Let us nore rhat no panel can be inregrally repeated without modification. The
reprise of ehe same panel at rwo locations in a comic, contiguous or distant, does
not constitute a perfect duplication. The second occurrence of rhe panel is already
different from the first by the sole fact of the citation effect that is artached. The
repetieion raises the memory of the first occurrence, if it is a mauer of a rhyme
(distant repetition),9 or manifests a singular insistence, if ehe two occurrences are
General Arthrology: The Network 149
contiguous. But most important is that, being isomorphs, these panels cannot be
"isotopes"; by definition, they cannot occupy ehe same site. Even if it is not ehe
objecr of a particular qualificarion (which is assuredly not the case if there is a
rhyme), ehe site is an inalienable constitutive parameter of ehe panel.
3.4-SOME COMPACT SERIES
An example rhat is borh simple and famous is the triptych thar occurs on page
35 of Tintin in Tibet, when Tintin and his friends, .having given up trying to find
Chang, abandon the carcass of the downed plane and prepare to leave. At this
point in the stoty, rhe breakdown seeks to evoke the slow and derisory progres-
sion, in a vast snowy expanse, of protagonists reduced to ehe dimension of ants,
in three contiguous panels. Braiding pulls on patt of rhis contiguity to instimte a
continuity in ehe representation of ehe decor, which eakes on ehe aspect of a large
panorama; in rhe background, rhe circle of mounrains is prolonged in all rhree of
rhe contiguous panels. Not only does this relatively elementaty braiding operate
in praesentia, relying on panels offered as an ensemble at a glance, bur rhe entire
series draws a compact form and a linear suite. .
Somewhat tempering the structuralist euphoria of rhe epoch, Georges Mounin
observed not long ago rhat "a structure ... holds inrerest only ifwe can show that
it has a precise function in ehe work, that it is pertinent (and at which point of
view it is)."" This methodological requiremenr naturally applies for series rhat
produce braiding. At its occurrence, ehe narrative pertinence of ehe HergC::an appa~
ratus leaps out: widening rhe decor magnifies rhe immensity of the region to be
explored in order to eventually find Chang, dooming the hero's quest to failure.
In La Orilla, the rwo-page mute stoty by Frederico Dei Barrio analyzed earlier
(cf. r.6, fig. 3), braiding idenrifies irself as an effect of plastic composition. Ir is
the localization of the character in the image that is ehe agent. These successive
positions draw a descending diagonal, and, symmetrically, an ascending diagonal,
thus inscribing a giant V at the heart of the stoty.
The inaugural page of The Red Sea Sharks (fig. 13) was the subject of a brilliant
analysis by Jan Baetens, who brought to light a remarkable series, that of rhe three
Alcazars.n I will repeat ehe demonscration: the name of General Alcazar is cited
three times, in the last panel of each of the rhree strips. The first of rhese panels
shows a poster representing an actor that resembles rhe General; the third of rhese
panels marks the appearance of the actual Alcazar, but an Alcazar whose gaudy

150 General Arthrology: The Network
COKE EN STOCK
fig. 13. From Tlre~res ofTmlin: The RedSeo Shalks (1958) by Herg!. © MaulinsarrS.A.
civilian elegance,-considering rhe fact that he is a milirary man, appears to be in
disguise. The passage of the pseudo-Alcazar to the real one is effectuated through
ehe headless mannequin rhat occupies rhe intermediate panel at rhe end of rhe
second strip. The location of these three figures, always to the right ofTintin and
the Captain, and the permanence of the repeated colors in the clothing (red and
crrePn) sufficienrlv attest that rhev behave as a cancerred series.
General Arthrology: The Network 151
This series creares a compact, in the sense rhat the three panels are contiguous.
Bur, distributed along a vertical axis, rhey are linked in a translinear manner, strad-
dling other panels that are not concerned with the effect of braiding bur which
share the rest of the page. We must therefore emphasize that braiding invests these
sites as doubly privileged: first because it is an inaugural page, further because the
three panels occupy corresponding places at the different Ievels of the page.
lnstead of this remarkable series, what the reader cannot help bur notice in the
page is evidenrly the Fact that the album opens on a panel that contains the words
"THE END." The two phenomena (the paradox of this introductory inscription,
and the braiding that sets the stage for the first appearance of Alcazar) are not to
be dissociated. !t has been litcle remarked that the inscription "START" can be
found symmetrically in the last panel of the album (in a more discreet Fashion, it
is true). Bur rhe end to which the incipit points cannot be only that of the enrire
work, but also of the page itself, designated in anticipation as a privileged place.
To the cinematographic image of a horseman riding away peacefully, can we not
oppose the image of the General, who hirnself arrives in a violent manner?
3.5-DIALOGUE FROM PAGE TO PAGE
Everyone has in their memory the scene at the statt of The ShootingStar, where
Tintin arrives at ehe observatory and discovers, through the telescope, a gigantic
spider that seems to be attached to the asteroid that is approaching the Earth.
This panel occupies the right side of the second srrip of page 4· Tintin soon
understands that it was nothing more than a small spider magnified by the tefe-
scope. He can then directly contemplate ehe asteroid, wichout the interposition
of this disruptive intrusion. Thus, this second glance thtough the telescope occu-
pies the panel situated to the right of the second strip of page 5, which is printed
face to face with its precedent.
lt suffices to examine the album for verification: The recurrence of the "ball
of fire" observed by the telescope is much more striking because the two coupled
panels have the same spatio-topical coordinates wirhin their respecrive pages.
The rhyme effect is considerably reinforced, so weil that the disappearance of the
spider (elsewhere called to reappear throughout the album undet diverse forms)
has the force of an immediately perceptible visual event. Braiding, once again,
makes these narurally corresponding sites work.
As the two examples borrowed from Herge illusttate, braiding is generally
founded on the remarkable resurgence of an iconic motif (or of a plastic qualiry),
152 General Arthrology: The Network
General Arthrology: The Network 153

and it is concerned primarily wich sicuations, wich strong dramatic potential, of

appearance and of disappearance.
In adapting The Masque ofthe Red Death by Edgar Allan Poe, Albeno Breccia
sysrematized translinear rdarions in absentia between corresponding sites of con-
secutive pages. The action is siruared in the palace ofPrince Prospero, which Poe
fittingly described in rhese terms: "These windows were of stained glass whose
color varied in accordance with the prevailing hue of the decorations of rhe
chamber into which it opened."" In the fourth page of the story, Breccia shows
us the open orgies in all the rooms of the palace, while outside rhe palace walls
the plague ravages the country. Bur, at the stroke of midnight, the "Red Death"
penetrates through to ehe prince, in ehe appearance of a spectre, interrupting ehe
festivities. The spectre, Poe teils us, "made his way uninterruptedly ... through
the blue chamber to the purple, through the purple to the green, through the
green to the orange, through this again to the white, and even thence to the
violet, ere a decided movement had been made to arrest him. lt was then, how-
ever, that the Prince Prospero, maddening with rage and rhe shame of his own
cowardice, rushed hurriedly through the six chambers.... He borealoft a drawn
dagger."
This sequence is translated by Breccia into two consecutive pages, the eighth
and ninth of a twelve-page story (fig. 14). The crossing of rhe six rooms is material-
ized by the. repetition of the same character. In spiee of some minor variations in
ehe silhouette, it appears frozen in a hieraric posture and endowed Wich ubiquity.
Time and acrion seem suspended, as if the same instant found irself eternalized by
a means of diffraction. The same procedure is applied successively to the spectre,
then, in the following page, to the prince. The two characters never appear wirhin
the same image in this sequence (or even on the same page); the theme of the pur-
suit of one anorher across all the rooms in the palace seems to be elided. While it
does not accede to a direct representation, the theme of pursuit is accomplished
only by relating, two by two, the twelve implicated sites, namely rhe recognition of
six chromatically differentiated series. (The rwo pages in question have sometimes
been printed face to face, and sometimes not, depending on the edition.)

3.6-THE INNERVED NETWORK

Fig. 14. From Le masque dein mott rouge (1982), by Mberto Breccio, odopred hom the sto~ by Edgor Mlon Poe. © Mberto Bre<cio.
The spider in The ShootingStar, or the yellow badge in Watchmen, are two dassie
examples of motifs where the proliferation throughout the works, as appearances
154 Genero1 Arthro1ogy: The Network
Fig. 14. ([onffnued)
Generol Arthrology: The Network 155
at essential moments in ehe Story andlor naturally privileged sires by the book,
produces rhymes and remarkable configurarions. The texturing ofintrigue, which
is accomplished through the recurrence of these emblems, is itself accompanied
with a considerable symbolic richness.
Watchmen, the work by Alan Moore and Dave Gibbons that has already been
cited several times in these pages, counts more than three hundred pages and
makes intense use of all the procedures of braiding. Ir is notably structured by a
declension of ehe figure of a circle, used both as a recurrent geometric motif that
lends irself to plastic rhymes, and for its symbolic connotations (perfection, eter-
nal recommencement, etc.). One of the occurrences of ehe circle rhat contributes
in this way is the smiling badge, familiarly designated under the names happy foce
or smiley. The authors have contrived to pur in place two narrative loops, the first
circumscribed by the inaugural chapter, the second extending to the dimension
of ehe entire work.
Indeed, the famous badge appeats tight in the fitst panel of the first chap-
ter, and in the last panel of this same chapter, as weil as in the last panel of the
rwelfth and final chapter. A remarkable relationship is established berween
these antithetical locations, predisposed to correspond under the emblem of
circularity and through the use of style. (The relationship pur inro place by
Moore and Gibbons is much more elaborate than the !ittle bit that we have just
looked at.)
There are other examples of the proliferation of a motif, in which it obeys only
a sort of playful formalism. l'm notably thinking about the multiplidry ofblack
circles and ovals in Yann and Le Gall's album, La Lune noire (Les exploits de Y.ryo,
r. r., Glenat 1986).
Once a graphic motif spreads across the entirety of the network that com-
poses a comic, it can arouse several thematically or plastically differentiated
series. Braiding therefore becomes an essential dimension of the narrative project,
innerving the entirety of the nerwork that, finding itself placed in effervescence,
incites translinear and plurivectoral readings. We know of films structured in an
analogous fashion; for example, Orson WeHes' Macbeth is entirely organized, as
Jean-Pierre Benhorne has shown, "around the rwo motifs of the Celtic cross and
pitchforks which incessancly recur, they meddle and affront, each aflirming their
pretension to invade space."' 3 Bur, barring the use of a video or DVD. the vision
of the film is, by definition, monovectorized and irreversible; the filmic images
are fugirive, and the echo of an image already passed is without another reality,
no verification is possible, other than that of memory.
156 General Arthrology: The Network Generol Arthrology: The Network 157

3.7-THE IMPERIAL15M OF BRAIDING

Upon first reading, the seventeenth page of the first album of the series Sambre
by Yslaire and Balac (fig. 15) is surprising. !t seems that nothing happens, outside
of the apparition (again!) of a light in an otherwise dark room. The long and
complex "movement of the apparatus," beyond the small forest, Ieads us to the
familial home of the Sambres, nearby the cemetery where the futher has just
recencly been buried. If ir marks a remarkable pause in the action, it is cercainly
not a vain parenthesis. !t represents a superb case ofbraiding, which I have cho-
sen as my final example.
The Z path that the eye must accomplish to sweep over the seven panels that
compose this page is highlighted entirely by the succession of circular motiiS. The
round window seen in dose up in the first panel is visible in the secend image,
ftom much further away, just under the roof, afterward, the path hooks onto the
white stain made by the moon, rhree rimes repeated along a senestro-descendent
diagonal; to reach the bottom left side of the page, the path must return to the
window, now lit, which inevirably makes us see rhe rosette, pierced by a clover-
leaf opening that adorns the Sarnbre tomb.
In multiplying the circles and distributing ehern to eminencly concerted
spaces, the authors have not ceded to the temptation of a simple formalism.
Here, rhe series of circles bears meaning: rhey speak to us ofimminent justice and
power, through the metaphor of the eye that sees, that knows, and ·that judges.
Indeed, this window, whose narne in the vocabulary of the architecture is "bull's-
eye," belongs to the office ofHugo Sarnbre, the absent futher. As for the fumilial
mocto inscribed on the front wall of the tomb, it is worded: "Sarnbre, the light
of the moon Iooks upon you." If the moon is looking, we must deduce that it is
equipped with eyes, or is considered to be an eye itself. This eye in the sky that
watches the intrigues of the Sarnbres can only be that of the futher. In leaving
his office, he merely changed his place of observation, talting it to a higher Ievel.
His point of view is now confounded wich that of God. Fora character buried in
184-8, and that holds the name of Hugo, it is hard not to think of the celebrated
poet of the sarne narne: "The eye was in the tomb, and was looking at Cain."
In the last panel; if the office is newly lit, it is because, as the following page F~. 15. From Sam/xe #1: PIIIS ne m'est rien (19861. by v•aire et Boloc. © Eorlians Gl;nat.
attests, Sarah, the daughter of the deceased, has just moved inside. By talting
possession of this room, she intends to esrablish her moral authority over the
fumily, talting on the responsibilities of her futher. She installs herself to work as
weil, having decided to recopy and complete the unfinished manuscript left by
 158 General Arlhrology: The Nelwork
Hugo Sambre, a manuscript enticled The Wizr of the Eyes. (The preceding page
had already concluded with a close up of Sarah shooting the reader a livid stare.
Her eyes juscly occupy the same site as the bulls-eye that, now lit, reveals her
presence.)
Simply through agame of formal analogies a rather powerful semantic net-
work is put into place that willlater be revealed as rich in narrative consequences
and symbolic implications. In this page alone we see the rying together of themes
that will nourish the entirery of the work in' several volumes, in particula.r that
of the eye. I will quickly eire three other pages that, far from them being exhaus-
tive, sufficiendy attest to the repercussions of ehe series put into place, and to the
extension given to the procedure ofbraiding that, in this comic, exerts a veritable
imperialism on intrigue and the sequential breakdown.
Page 7 of Book r: Julie, rhe poacher, the heroine of this tragedy, has red eyes, a
sign ofher allegiance to a cursed race that has been predicted to cause the ruin
of the Sambres.
Page 37 of Book I: ehe moon returns in an oneiric scene, and it returns in an
explicit liaisonwich the idea of justice and punishment.
Page 46 (the last page) ofBook 2: finally, the moon a"aain, butthistime red like
rhe eyes of Julie. Pregnant with the seed of Bernard Sarnbre, Julie will deliver
a new life. The bloodied moon announces, for Julie, rhe imminence of the
revolution, which represents the hope of a new and better life for the people
ofParis.
Braiding thus manifests. into consciousness the notion that the panels of a
'
'
cernie constitute a network, and even a system. To the syntagmatic logic of the
sequence, it imposes another logic, the associative. Through the bias of a tele-
arthrology, images rhat the breakdown holds at a disrance, physically and contex-
tually independent, are suddenly revealed as communicating dosely, in debt to
one another-in the manner that Vermeer's paintings, when they are reunited,
are perceived to come in p'airs, or in threesomes. As Pierre Fresnault-Deruelle
justly nored, it is rh~e "thousand and one forms of deviation and correspondence
rhat makes of comics a text in the strengest sense of the term." 14
CONCLUSION
In approaching comics as a "system" I wanted to signif}r thar it consritutes an
organic toraliry rhar associares a complex combinarion of elemenrs, paramerers,
and multiple procedures. The definition proposed here in irs own language may
not be systematic, in rhe measure where cerrain quesrions were left ro rhe side---
norably those that concern authority and the functions of rhe scenariosuch as rhe
subjecr, program, and pre-texr, or even rhe different instances of enunciarion-
but it seems to me that it is ar least coherent. Recall rhat the principal fuundation
from which I departed was that of iconic solidarity. We now know that the three
operarions of breakdown, of page Iayout, and of braiding delimit rhe principal
modalities. These Operations, and notably the first two, are inconceivably linked
one ro the other, since the narrative dynamic of rhe story can do nothing orher
than aniculate irself on the physical occupation of the finished form.
As for the method followed, we undoubredly want to accord a pragmaric
qualificarion. With a choice of conceprual tools coming from aesthetics, narra-
tology, comparative theory or the history of the "ninth art," I gathered all that
appeared susceptible ro fertilize or to complete a reflection of inspiration rhat is
globally semiotic. Bur pragmatically, in the sense that, with regard ro creation as
weil as to receprion (reading), I am constancly forced ro open my rhinking ro all
possible and imaginable oprions. On the side of crearion, rhis signifies rhar no
procedure was considered as exclusively destined to rhe producrion of a cenain
type of effecr; even rhough rhere exisr dominant uses rhar are Iegitimare and per-
rinent to identif}r as such, all procedures are susceptible to entering a particular
appararus that would modifY or specifY its efficiency. From the Standpoint of
reading, rhe distinction berween different Ievels of receprion-the conversion
of utterables to utterances, the acrualization of rhe description, and finally rhe
interpretation-and berween several planes of meaning, far from heading to rbe
 160 (onclu~on
formalization of a protocol of normative and univocal reading, brings the conclu-
sion that meaning is, foreach reader, always tobe constructed and tobe completed,
succeeding only to provisory syntheses and necessary matks of subjectivity.
Several of the considerations developed in rhese pages certainly merit being
heard beyond the field of comics, and can even lead-the prejudiced academ-
ics are hardening themselves to suffer-to a reconsideration of the presupposed
theories of various disciplines. In particular, narratology suffers from having
developed in reference Only to literature, when its field of natural investigation
:•
is in realiry the narrative genre, and should no Ionger exclude the art of visual
stories. Specialized researchers have certainly been tempted, after a fashion, to
adapt to the seventh art concepts forged to take account of the literary story. But
it was to be enclosed in a dialogue (most often in a one-way direction), while it
should tadde narration in confronting the entirety of disciplines of the story. The
widening of research into comics (and into the phoro-novel) can only Iead to the
necessity of modifying or revitalizing certain concepts. It is otherwise a similar
statement of the inadequacy berween traditional caregories of knowledge and the
reality of messages that are increasingly exchanged in more and more diversified
forms rhat brought Regis Debray to sketch a new discipline: mediology-with,
nevenheless, ambitions that are very different from my own/
Comics is not a syncretic (total) artsuch as opera; it does not solicit ftom the
reader the sarne perceptual deployment that is demanded from the cinematic
spectator, &om which, as Francis Vanoye recalls, film "makes us · see moving
images, see written language, hear noises, music, oral speech. "z. But if it does not
mobilize senses other than sight, comics, which marries the visual and the ver-
bal, demonstrates a discontinuiry, a staggering, and the effects of nerworks, and
finally constitures a sort of image bank, appear to be situated not far &om the
rurning point berween the .civilization of the book and that of multimedia. With
this ticle, they merit much more critical attention than they receives.
We have seen that its modus legendi is very elaborate. Despite the inherent
complexity of comics, we nevertheless cannot conclude its artistic validity. I have
voluntarily le& aside, in this work, the evaluation of its legitimacy as art. This
book is not a militant work, or it is only in passing, through the bias of the
examples [hat it prOduces and comments upon. Of the rest, it is not up to aes-
thetic judgment {whatever it may be) to decide the legitimacy of a theoretical
research. As Deleuze wrote at the beginning of his monumental Cinbna, "one
cannot object by pointing to the vast proporrion of rubbish in cinematographic
production."~
(onclusion 161
I suggested that the cernies image, more than being an utterable, a descrip-
tible, and an interpretable, is also an appreciable. We must, undoubtedly, return
to this Statement for this conclusion. Having highlighted this evidence, it is clear
that the artistic qualities that we might wish to attribute to certain comics illus-
ttators are not prejudged in relation to the success of the works rhat bear his
talent. Finally, the evaluation engages the work in irs totaliry, in all of its dimen-
sions: narrative, plastic, symbolic, etc. In this respect, a better comprehension of
the comics system, if it is not immediately convertible to a criterion of jud.gment,
nevertheless allows a bettet apprehension of each work as a singular performance
of the medium.
Let us return to the image, and to its qualities as graphic performance. Since,
indeed, beyond the image-body that interests spatio-topia and the image-sign
that concerns arthrology, we should not forget the image-oeuvre, the signed
image that bears witness to a skill, a style, a vision, a potential. From the theoreti-
cal point of view, the only question of worth is that of the suitable criteria for
appreciation. In reality, it is a diflicult question, one which the specialists have
not said much about up to this point, since the cernies image is problematized by
its narrative finality and, if the case arises, its humoristic vocation. It follows thar
the criteria of judgment traditionally applied to drawing arts are here fourid to be
inadequate, Iirerally impertinent.
Undoubtedly, illustratorssuch as Rene Giffey, Paul Cuvelier, Frans:ois Schuiten,
and Andre Juillard (to Iimit myself to francophone artists) have shown that there
is not necessarily a contradiction between the exigencies of the medium and
the grand classical tradition, the secular "profession" of drawn art. Meanwhile,
other artists, just as great-let us take, for exarnple, Herge, Calvo, Franquin, and
Bretecher-have been inspired almost exdusively by caricaturists (in the largest
sense) that preceded them, each forging a remarkable style, and imposing it as
an original synthesis owing nothing but a minor debt to the academic canons of
"beauriful drawing."
The drawing that defines the proper laws of comics is that of narrative draw-
ing. It seems to me that the principle characteristics of narrative drawing are five
in number:
I. Anthropocentrism. The narrative drawing privileges the character, the agent
of the action; it successively accedes to each character the Ievel of protagonist,
in the etymological sense of "he who plays the primary role." Moreover, the
format of the panel often appears calculated to be married to the body of the
162 (ondusion
character represented in the frame, as if ehe panel constituted its natural hab-
itat, its vital space, delimiting the space of its immediate behavior-a spatial
volume, in sum, dose enough to what the deaf call the espace de signation or
what the choreographer Rodolphe von Laban proposed to name kinisphere.
2. Synecdochic simplification. The narrative drawing, as we have seen regarding
rhe page from Corentin, very ofcen evacuates that which is not necessary to
the inrelligibiliry of rhe represented siruation. If rhe sequenriality obliges it
to certain redundancies, they do, at every moment, privilege the elements
rhat have an immediately informative character, eliminating or backing up
rhe rest.
3· Tjpification. The logical consequence of the first two points, rypification
is simplification as it applies to characters. The abbreviation of a character
to several pertinent lines assures their characterization and their immediate
identifi.cation. Tintin's rufe and the cap and earring of Corto Malrese are
among rhe most celebrated emblems of rhis graphic strategy. Typification
presents a danger: that of rhe stereotype, that answers to the necessity of
fully visually expressing something through "exterior signs" (or clues--of its
richness, honesry, deceit, etc.) that are simple and immediately decodable.
Similarly, Satire and parody, which use stereotypes to critical ends, arenatural
vocations of the narrative drawing.
4· Expressivity. The "play" of characters is an essential factor to the comprehen-
sion of rhe situation by rhe reader. The discontinuity of narration in cernies
requires that the rendering of each image be all the more eloquent, suchthat
it constitutes a chosen moment, withdrawn from the supposed continuity of
action. The body (the gestural) and the fuce (the physiognomic expressions)
of characters should be as expressive as possible•-an expressivity that fre-
quently reinforces recourse to a whole scale of ideograms or of conventional
signs (such as those of the Iiede droplets around the fuce that emphasize
emotion). The codifi.cation of movement is an dement of expressivity, but
it consists, as Henri Van Lier noted, of a "movement wichout mobility," "a
cinematic wichout dynamism. "s
5· Rhetorical convergence. The narrative drawing obeys an imperative of optimal
legibiliry. Consequendy, it uses different parameters of the image (framing,
composition dynamies, color placement, etc.) in a manner that murually and
concurrently reinforces ehern to the production of a unique effect.
Conclusion 163
These criteria do not constitute a dogmatic chart that must be applied in
blocks. Bur if they find diverse degrees in their mise en oeuvre, they more or less
determine a horizon for the narrative drawing, pertaining as much to cernies as
to magazine illustration. On these different points, the narrative drawing can be
opposed totheillustrative drawing, which makes great sacrifices toward a decora-
tive tendency and calls for a more contemplacive reading.
Narurally, the respect of these criteria accommodates a !arge diversity of sryles,
from the more elaborate to the more schematic, from the most luxurious to the
most rough. As the Groupe Mu recalled, "a flower or a leaf can be the object of
a stylization that can be romantic, fantastic, modern style, puerile, mechanical,
psychedelic, erc. "6 We observe in all cases that the conditions of rhe narrative
drawing are perfectly satisfied by rhe most minimal graphic art. I'm rhinking here
in particular of the lircle "iron wire" characters of the Spanish areist Calpurnio,
which to me represent the superlative degree of simplification. Their efficiency
testifi.es to the fact rhat the drawing does not know rhe virrual, etc. If it is sche-
maric, it is neverthdess always fini.
Rodolphe Töpffer intuitively knew rhe essence of the narrative drawing. From
rhe origins of rhe ninrh art, he wrote that "the graphic line" of rhe narrator in
images is less concerned with the ideal of beauty or exactitude than with "all rhe
exigencies of expression like ... all rhose ofbrightness"; that this line "demands
enormous ellipses of accessories and detail"; and that rhe "cursive sketches" are
used for tracing, "as much as linking ro a series, often figuring like a recall ofideas,
like symbols. "7
The narrative drawing is, par excellence, that with which we can talk of a "sub-
jection of the line to ehe Iogos," to reclaim a formula by Jean-Marie Ponrevia.'
For the narrative drawing, showing and telling are one and rhe same rhing. The
narrative drawing does not return to a referent, but goes straighraway to being a
signified.
Undoubtedly, there are moments in cernies where narrative pressure is released.
Pierre Sterckx spoke of images where "ehe story marks a rruce, the narrative no
Ionger exerts its tyranny of cascading new developments, rhe hand of the illustra-
tor can marry rhe damming of the story, operring up the stroke, amplifYing the
forms and above all their intervals."9 The aesthetic evolution of cernies for the
past quarter of a cenrury has been toward the direction of liberating rhe image.
The traditional narrative drawing, from Töpffer to Franquin, and from Milton
Caniff to Mezieres, is seen to be concurrent wich writing rhat is freer, more picto-
rial, and more poetic. From Moebius to Alagbe, from Loustal to Barbier, from
164 (onclusion
Baudoin to Vanoli, comics has shown that it can accommodate ehe illustrative
drawing, and it can even completely abandon the linear drawing, at the profit
of a play with surfaces and colors, lights and intensities. With a historic shift
forward, comics have thus lived through the equivalent of what Pontevia called
"the insurrection of the gestural painting." With regard to this adventure, which
tempted ehe painter, how can the cernies author ta.ke this risk without ceasing to
be a storyteller? It is because ehe narrative contents have evolved. The narrative
ehernes par excellence (the voyage, the pursuit, ehe investigarion, ehe disguise, the
metamorphosis), which traditional comics have used and abused, if they are not
abandoned, have been at least relarivized by rhe conquesrs of new story spaces:
increasingly literary, increasingly immobile, increasingly poetic, increasingly sen-
sual, and increasingly inuospective. The comics system has definitively made a
demonstration of its plasricity.
NOT ES
INTRODUCTION
1. Translatorl note: Bandedessinleis technically "drawn strip"' and is used generically to mean
all forms of comics. Throughout this text we have substituted the generic "comics" for
"bande dessinte..,
z. "Semiotic Approaches to Figurative Narration.. in The Semiotic "Web, ed. T. A. Sebeok and
]. Umiker-Sebeok (1989; Berlin: Mouton de Gruyter, 1990). My citarions are taken from
the French manuscript supplied by the author. lt is notewonhy that the author has not
retained. the critical discourse on the ideology of comics, which inspired severai works in
the 1970s and 198os.
3· Ibid. The author analyzes, as an example of this new approach, the proceedings of the
Actes du colloque that I organized in Cerisy in August 1987, Bande dessinte, rtcit et moder-
nitt (Paris: Futuropolis, 1988).
4- Langage et cintma (Paris: Larousse, 1971; new edition Albatross, 1977), p. 155.
5· "Les Peanuts: un graphisme idiomatique," Communications, n. 24, Le Seuü, 1976, w8-139.
Citation, p. IIJ.
6. Cf. "Comics lesen," Untersuchungen zur Textualitität von Comics (Stuttgan: Klett-Cotta,
1978), pp. 15-35.
7. Cf. Groupe Mu. Traitt du signe visueL Pour une rhttorique de l'image (Paris: Le Seuil, "La
coleur des idc:!es," 1992), pp. 149-152.
8. Cf. notably Chrisrian Metz, Essais sur la signification au cinlma, t. I (Paris: Klincksieck,
"Esthc:!tique," 1968), _pp. 67-72, 87-92. Henri Van Lier arrived at the same conclu-
sions starring from a different approach to the notion of the sign which privüeged. the
"effects of the fidd." Cf. L'Animal signt (Rhode-Sainr-Genese: Albert De V!Sscher, 1980),
PP· 37-75·
9· "Semiologie de Ia Iangue," Semiotica I/2, La Haye: Mouton & Co, 1969, p. 129.
xo. ld., Semiotica III, p. 12, and 112, p. 132.
II. Roger Odin, Cintma et production de sens (Paris: Armand Colin, 1990), p. 89.
12. FCnetre jaune cadmium, ou Les dessaus de la peinture (Paris: Le Seuil, "Fiction & Cie," 1984),
p. 302.