marks, traces, "traits," contours, "orli," and "splendores":...

Marks Traces Traits Contours Orli and Splendores NonsemioticElements in Pictures

James Elkins

Critical Inquiry Vol 21 No 4 (Summer 1995) pp 822-860

Stable URL

httplinksjstororgsicisici=0093-18962819952229213A43C8223AMT22C22A3E20CO3B2-3

Critical Inquiry is currently published by The University of Chicago Press

Your use of the JSTOR archive indicates your acceptance of JSTORs Terms and Conditions of Use available athttpwwwjstororgabouttermshtml JSTORs Terms and Conditions of Use provides in part that unless you have obtainedprior permission you may not download an entire issue of a journal or multiple copies of articles and you may use content inthe JSTOR archive only for your personal non-commercial use

Please contact the publisher regarding any further use of this work Publisher contact information may be obtained athttpwwwjstororgjournalsucpresshtml

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

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academicjournals and scholarly literature from around the world The Archive is supported by libraries scholarly societies publishersand foundations It is an initiative of JSTOR a not-for-profit organization with a mission to help the scholarly community takeadvantage of advances in technology For more information regarding JSTOR please contact supportjstororg

httpwwwjstororgWed Oct 24 182211 2007

Marks Traces Traits Contours Orli and Splendores Nonsemiotic Elements in Pictures

James Elkins

Art history lacks a persuasive account of the nature of graphic marks and that limits what can be said about pictures If a sign as Charles Sanders Peirce said is something which stands to somebody for something in some respect or capacityn-a formula as vague as it is compact-then every mark in a picture is also a sign every brushstroke pencil line smudge and erasure must function as a sign and have meaning In practice that would spell trouble for accounts of pictures that take sign to mean the forms that are made out of the marks-such as in the typical examples figures scenes and narratives For the most part-and with important exceptions in all periods and subjects-art history has concen- trated on the larger-scale forms or on the large-scale properties held by groups of marks such as facture or handling That emphasis has left more exacting questions about graphic marks to the domain of practical criticism so that the people who are said to understand marks best are other artists and the most incisive critiques are taken to be other paintings instead of texts

Although it might seem that semiotic art history would address that inequality-perhaps by reinstating something like Peirces wider sense of signs-it is closely wedded to the distinction between meaningless marks and meaningful signs and for that reason semiotic accounts tend to gloss over marks in favor of the scenes they compose In Mieke Bals view the

1 Charles Sanders Peirce Elements of Logzc in The Collected Papers of Charles Sanders Pezrce ed Charles Hartshorne and Paul Weiss 6 vols (1931 Cambridge Mass 1974) 2135

CnItcal l n q u t ~21 (Summe1 1995)

O 1995 by The Unive~sity of Chicago 0093-189619512104-0003f0100 All rights reserved

Critical Inquiry Summer 1995 823

field of signs and meanings in pictures can be divided into three classes signs properly so called marks that are subsemiotic and those that are suprasemiotic or holistic Subsemiotic marks are those that are part of what make us interpret the work even though we do not usually give them meaning in them~elves~ They include stylistic variation light and dark composition or more technical aspects like brushstrokes paint thickness and lines Suprasemiotic signs on the other hand are holistic aspects of works including entire pictures considered as signs (RR p 400 n 16)This schema excludes painted or drawn marks from visual semiotics by denying them the status of signs and by demoting them to the domain of the technical Both judgments can be read as strategies that allow the wider project of semiotic art history to get under way and later I will argue that semiotic art history sometimes depends on suppressing the semiotic nature of marks in order to proceed with readings that hinge on narrative

These two decisions about marks and signs are at the root of competing versions of how pictures mean and of what happens in pictures below or apart from the level of figures and narratives In the first account (the one I am representing as Peirces though I only mean to say it is compatible with what he writes about signs in general) semiotics compre- hends the entirety of marks whether they are the discrete systematic morphemes of writing or the slurred brushstrokes that make up a painting In the second account which I will be ascribing by synecdoche to Mieke Bal and Norman Bryson it would not make sense to call the morass of painted marks signs for them semiotics in painting properly begins with larger units Let me call Bal and Brysons account semiotic and Peirces following his own spelling semeiotic In semiotics graphic marks somehow build to make signs but are not signs themselves They are technical irrelevant or irrecoverable and in the strict sense meaningless In semeiotics on the other hand any mark could function as a sign though signs for scenes and figures might be different in kind from signs that are nameless brushstrokes

2 Mieke Bal Reading Rembrandt Beyond the Word-Image Opposition (Cambridge 1991) p 400 n 16 hereafter abbreviated R R There are many varieties of this kind of claim and some try to recover some linguistic status for subsemiotic elements Norman Bryson talks of asemantic elements mutilated language and infrasentences (Norman Bryson Word and Image French Painting oftheilncien Rigzme [Cambridge 19811 pp 13 28)

James Elkins is associate professor in the Department of Art History Theory and Criticism at the School of the Art Institute of Chicago His recent work includes The Poetics ofPerspective (1994) and essays on art criticism the concept of style the use of schemata in visual art Nelson Good- man the place of the unconscious and ambiguity in pictorial meanings Another book The Object Stares Back is forthcoming

824 James Elkins Nonsemiotic Elements In Pzc t~r~e~

In what follows I am going to argue against semiotics and propose ways to think about subsemiotic marks by attending to their syntactic properties Semiotics I think has several deleterious effects on the ways we understand pictures Despite its claims to be neutral between linguistic and other sign systems semiotics slights the meaning of marks bringing visual narratives unpleasantly close to written ones (so that without illustrations in the texts it would sometimes be difficult to tell if a semiotic account were referring to a painting or another text) In the end semiotics shrinks the notion of what a picture is assimilating pictures to texts and overlooking their painted strangeness Semiotics makes pictures too easy I want pictures to be harder to look at and harder to describe so that we cannot get as quickly from the slurry of niarks to orderly historical meanings

Pictures would be somewhat difficult to write about if their subsemiotic marks were ultimately beyond the reach of linguistic analogies as many people take them to be but pictures would be even more difficult and far more analytically engaging if marks were neither hopelessly beyond the reach of analysis nor entirely assimilable into the systems of semeiosis Both of those possibilities are reductive in their own ways positing that marks are nonsemiotic gives up a large part of what pictures are and claiming marks are signs more or less avoids reading pictures as such by making them into other kinds of objects I do not think pictorial marks can be discussed on the model of written marks but neither are they inarticulable inchoate mutterings forever divided from the signs of language IYhat I will be saying about previous theories can be put as two interlocked claims first about semiotics that its own logic of the sign prohibits it from assuming that marks are meaningless subsemiotic elements and then about semeiotics that it need not-as Peirce often did- assume that everything about the taxonomy of signs can be analyzed But most of this essay is occupied wit11 what marks are rather than with what they are not I will be presenting a series of modes of graphic marks that have connections to art history art practice and linguistics-but no exclusive allegiances to any of them The idea is to make a start in describing graphic marks by showing some ways that they can exist between linguistic signs and painterly babble

What Is Visr~alSemiotics

To begin it is reasonable to ask how semiotics can make sense outside of linguistic structures and analogies Is there any such thing as visual semiotics apart from the application of linguistic models to nonlinguistic forms Bal and Bryson come upon these issues in their essay Semiotics and Art History in the course of reviewing Saussures relevance to a

Critical Inquiq Summer 1995 825

semiotically informed art historyVt may seem they remark that Saus- sures emphasis on the significant units of language makes his doctrine inapplicable to visual art where there may not be units of any kind

We might try to say that below a certain threshold perhaps roughly corresponding to phonetics in language there are marks that con- tribute to but which do not yet produce signification-individual brushstrokes or lines or dots or pixels and that above that threshold these as yet nonsemantic marks emerge as productive of meaning But can we say that marks below the threshold are units Or above the threshold Particularly in the Saussurean tradition the positing of meaning-bearing units-signifiers-seems essential But a painting is a continuous surface with marks that blend together inextrica- bly If no minimal units for images can be found then a visual semiotics deriving from Saussure must be an impossible endeavour we cannot establish where the signifier actually is$

Their response is that the problem is ill conceived because not only pictures but language itself lacks those static signifiers

The objection is understandable but it may be misplaced The problem of a mismatch between words and images can in fact lead us in a rather different direction toward the question whether the individual word actually is languages prima materia The quest is a reflection of a philologically derived linguistics that posits meaning as occurring at the level of the word or the sentence but does not consider the larger aggregates the bonding together of words and sentences in social practice as discourse At this level signs are not discrete but dense individual signs become molar consolidated fundamentally inseparable To think of semiosis as process and as movement is to conceive the sign not as a thing but as an event the issue being not to delimit and isolate the one sign from other signs but to trace the possible emergence of the sign in a concrete situation as an event in the world [S p 19415

It is odd to defend the use of a linguistic theory that depends on fundamental units as Saussures does by arguing that languages primary units are not fundamental units however this way of answering the question runs into serious logical problems when it implies that it is ac- ceptable that pictures do not possess discrete units because language does

3 Bal and Bryson repeat and amplify their position in the Encyclopedia of Aesthetics (forthcoming)

4 Bal and Bryson Semiotics and Art History Art Bulletin 73 (June 1991) 194 hereafter abbreviated S

5 A footnote to the word dense reads With apology to Nelson Goodman for this reversal of his term density (S p 194 n 96)

826 James Elkins Nonsemiotic Elements in Pictures

not possess discrete units at higher levels of organization Does the fact that pictures do not have significant units below a certain level mean that they can be better compared with language which lacks discrete units above a certain level

Bals Reading Rembrandt repeats the defense of semiotics against those who would claim that signs in visual works are dense The distinction between oppositional language and dense pictures Bal says

is deceptively self-evident and can be deconstructed only by re-versing it and arguing that to some extent verbal texts are dense- the sign of the effect of the real cannot be distinguished from the work as a whole on which it sheds a specific meaning-and that visual texts are discrete which sometimes and in some respects they are [RR p 401 n 161

It is the same faulty argument if it seems inappropriate to use semiotics on dense images then it is not less inappropriate just because texts are also dense Both linguistics and textual semiotics depend on the existence of minimal units of meaning whether they are morphemes phonemes or entire propositions and no matter how much attention we choose to pay to larger structures atomic units remain essential for the sense of the enterprise Even if they werent essential-even if it were possible to imagine a linguistics or a semiotics independent of the oppositional discrete character of linguistic signs-that still wouldnt be an adequate re- ply to the claim that the density of pictures is different from the density of texts It is an evasion to claim that these densities are different and it contradicts the tripartite division of signs into subsemiotic semiotic and suprasemiotic because if suprasemiotic holistic aspects of artworks are not to be considered as signs how can they be the only important or legitimate units of meaning in texts

I would read the foreclosure of sign-theory under the rubric of philologically derived linguistics as less a logical problem than a political move In order to get on with the business of reading pictures (for example with the questions of psychoanalysis gender disciplines and other social structures that occupy the bulk of Semiotics and Art His- tory) they need to finesse the nature of graphic signs5 In the end I

6 The same kinds of questions can be asked about the higher-level conglomerates of signs If there are no isolated signs then events are either sparked by groups of signs or by molar consolidated fundamentally inseparable signs If it is the former then it remains to be said what a single sign is and why it does not create meaning in isolation If it is the latter the sign-groups would be functioning as single signs and therefore they could not exist as such What is an event if it is not sparked by the presence of some sign whether or not that sign is linked to others

7 I agree with LV J T Mitchell that Bal and Brysons underdeveloped sign-theory is mostly a platform for questions of gender and power (UJ T Mitchell Picturr T h e o ~ Essays on VPrbnl and Vzsunl Repr~sentation [Chicago 19941 p 87 n 8)


dont think that Bal and Bryson want to present an account of how signs work in pictures as much as they want to show how semiotics can open questions of meaning by remaining alert to dissemination reception and the production of meaning A visual semiotics pitched at a higher level as they say could benefit from the wider field of meaning-as-event and avoid the restrictive interpretation of significant units Still the price for that liberation is very high since it entails relegating the very founda- tion of a semiotic reading-that is the sense that is to be made of marks and signs-to a matter of misplaced concern about local structures It begs questions about the way pictorial meaning happens at all

Even so Semiotics and Art History proceeds on the assumption that there is a specifically visual semiotics parallel to the linguistic model Bal and Bryson mention visual and verbal practices of the sign and they point to semiotics as a transdisciplinary theory that helps to avoid the bias of privileging language-as if the only initial problem with visual signs is making sure they are not ignored in favor of written signs (S pp 194 175) But the question of the relation between visual semiotics and linguistic semiotics is not as easy as it seems and theories that begin from linguistics have often remained within linguistics As W J T Mitch-ell puts it although Bal and Bryson insist that they are proposing a semiotic turn for art history rather than a linguistic turn they underes- timate the extent to which semiotics privileges textuallinguistic de- scriptive frameworksThe three major sources Bal and Bryson adduce concerning visual signs-Fernande Saint-Martin Koland Barthes and Nelson Goodman-can each be read as evidence that visual semiotics may not exist in the developed state that Bal and Bryson require

Saint-Martins book Semiotics of Visual Language for example is cited as a source for the semiotic theory of visual works but her categories are often rudimentary or awkward by art-historical standardsThey are too schematic to be much help in understanding visual signs in artworks and neither Bal nor Bryson make much use of her classifications Goodman also appears in their notes but principally so they can import his concept of density a concept that would if it were taken the way Goodman pres- ents it vitiate any attempt to read visual marks as signs (S pp 176 194) For those who wish to argue that painted and drawn marks are beyond the pale of logical analysis Goodman has long been a point of reference He has several ways of distinguishing graphic marks from those in writing or notations he says painted and drawn marks are syn- tactically dense so that unlike the discrete characters in an alphabet each blends into the others in seamless infinitesimal variation Among the dense symbol systems Goodman also distinguishes the replete marks

8 Ibid p 99 n 31 Mitchell quotes S p 175 9 See Fernande Saint-Martin Semiotics of Tisual Language trans Saint-Martin (Bloom-

ington Ind 1990)

828 James Elkzns Nonsemzotzc Elements zn Pzctures

common in pictures from the attenuated marks in schemata such as graphs A stock market graph for instance is attenuated because only a few properties of the lines matter-their height denotes stock prices and their horizontal position denotes the date-but it does not matter if they are printed in red or black ink or even drawn by hand Paintings are typically different according to Goodman because any change in a painted mark might change its meaning

Density and repleteness are important concepts what gets decided about graphic marks largely depends on how they (or terms like them) are understood If Goodmans claims hold true then it would probably not be helpful to say that marks can be signs in the same way as figures or painted objects are Painted and drawn marks would be left in a kind of paradoxical perdition since they would be so sensitive so attuned to nuance that they would be incapable of saying any one thing But there are also important ways in which Goodmans claim does not make sense Certainly graphic marks are often dense-though there are also many exceptions-but is any graphic mark functionally replete The idea is true enough as an exercise in classification-each tiny change in a mark cozild alter its syntactic and semantic function-but it does not correspond well to the ways that pictures are actually made or viewed The problem is that marks in pictures are not perceived as dense depending on the context some changes might appear meaningful but most would be ignored In a print of Mount Fuji by Hokusai-Goodmans example of repleteness-a seriously overinked impression might attract attention but routine shifts in ink color paper texture or printing force would not and neither would an infinity of slightly different contours in place of the one Hokusai chose When those differences are noticed they tend to be classified in terms of print states runs editions and versions but they are rarely said to have meaning I have argued elsewhere that repleteness is a fictional construct since we would be hard pressed to come up with more than a few variants of any given mark that might be expressively meaningful Many changes are invisible because they do not correspond to any known styles periods strategies or genres that we know how to read So density is a reasonable notion but it is not well related to the ways that pictures are interpreted

Barthes is another principal source for Bal and Bryson but Barthess

10 Nelson Goodman Languages o fArt A n Apflrouch to a T h r o v of Symbols (1968 India- napolis 1976) pp 136 230 230 n 2

11 Bal and Bryson say semiotic art history requires a thinking-through of the status of signs and meaning in visual art-for example of the delimitation of discrete signs in a medium that is supposed to be given over to density (S p 176) But they do not offer the thinking-through which would in fact be impossible whenever visual marks are con- strued as dense in Goodmans sense

12 See Goodman Lnngungrs of Art p 229 13 See James Elkins What Really Happens in Pictures Misreading with Nelson

Goodman Word and Image 9 (0ct-Dec 1993) 349-62


essays on signs such as The Third Meaning Research Notes on Some Eisenstein Stills and The Photographic Message are deeply problem- atic attempts to understand what a truly uncoded image might be (see S p 191)14 In Elements of Semiology Barthes says he finds it

increasingly more difficult to conceive a system of images and objects whose sign$eds can exist independently of language to perceive what a substance signifies is inevitably to fall back on the individua- tion of language there is no meaning which is not designated and the world of signifieds is none other than that of language15

Nor is Barthes the only semiotician who was troubled in this way Mitchell has noted several other attempts to resolve the issue Umberto Eco he observes was skeptical of a tendency in the 1960s in which semiotics was dominated by a dangerous verbocentic dogmatism whereby the dignity of language was only conferred on systems ruled by a double articula- tion The problem as Eco saw it is that there can be no secure way of saying what an iconic sign might be Some iconic signs are like the significant units of language and others result from the correlation of an imprecise expressive texture and convey a vast and unanalyz- able portion of content likewise some rules that govern iconic signs include rules for other kinds of signs and vice versa For Eco these confu- sions make the entire notion unworkable

One and only one conclusion seems possible at this point iconism is not a single phenomenon nor indeed a uniquely semiotic one It is a collection of phenomena bundled together under an all-purpose la- bel (just as in the Dark Ages the word plague probably covered a lot of different diseases)16

In a later essay Eco remarks that the study of iconic signs is reaching more and more complicated levels of sophistication but the examples he cites-Panofsky Gombrich Goodman Martin Krampen Christian Metz and his own work-are principally concerned with what I have been calling higher-order forms and they have little to do with the strictly pictorial nature of elementary marks

14 See Roland Barthes The Third Meaning Research Notes on Some Eisenstein Stills and The Photographic Message Image Music Text trans and ed Stephen Heath (New York 1977) pp 52-68 and 15-31

15 Barthes Elements of Semzolog3 trans Annette Lavers and Colin Smith (New York 1977) p 10

16 Umberto Eco A Theory of S~nliotics (Bloomington Ind 1976) pp 228 216 quoted in part in Mitchell Iconology Image Text Ideology (Chicago 1986) pp 56 n 3 57

17 Eco Semiotics A Discipline or an Interdisciplinary Method in Sight Sound nnd Sense ed Thomas A Sebeok (Bloomington Ind 1978) p 75 See Christian Metz Film Language A Senliotzcs of the Cinema (New York 1974) Dze Welt als Zeichen Klnssiker der moder- nen Semzotik ed Martin Krampen et al (Berlin 1981) trans under the title Classicsof Semiot- ics (New York 1987) and Krampen Signs and Symbols i n Graphic Communication

830 James Elkzns Nonsemzotic Elements zn Pictures

And there are other writers as well who have been skeptical of the idea of an independent semiotics of pictures In Structuralist Poetics Jona-than Culler renames Peirces semeiotic triad of icon index and symbol as the icon the index and the sign proper and he does so in order to exclude both iconic and indexical signs from semiotic^^ Icons differ markedly from other signs he writes because an indeterminate part of their function is due to cultural conditioning Icons cannot even be said to make sense until there is a reasonable account of the way in which a drawing of a horse represents a horse which Culler thinks is perhaps more properly the concern of a philosophical theory of representation than of a linguistically based semi~logy~ from its crite- Cullers judgment-apart ria which have to do with the way signs refer rather than their structure in the image-is in line with Barthess and Ecos visual semiotics is not a coherent enterprise because it rests on poorly formulated notions

The most persistent attempt to define the sign-theory that underlies visual semiotics is Anne Hknaults essay Semiotics in France Like Bal Bryson and Goodman H6nault takes the extreme antirealist stance claiming that semiosis is as arbitrary for visual languages as for any other and she uses that position to avoid the dead-end of Barthes semiology where the writer is forced to stop interpreting in the face of the purely visual uncoded image But she does not object to the idea that there might be some purely visual realm that is exempted from the domain of language and she faults Barthes for transforming the visual text into a linguistic text and for being satisfied with naming one by one the various items which could be seen What is needed in her opinion is to establish the semiotic status of the constituent parts of the visual plane of expression

the eye must come back to a genuine perception and avoid seizing upon the would-be figures of content which might be verbalized at once On the contrary one must forget that this photograph shows a railway or a female torso which language would call units of meaning this is the prerequisite for allowing perceptions to appear for which verbal automatism makes no p r o v i s i ~ n ~ ~

Unfortunately this promising program is only exemplified by a few essays H6nault cites three by Jean-Marie Floch and one by Felix Thiirle-

(Minneapolis 1965) and Zur psychologischen Begriindung einer Semiologie der visuellen Kommunikation In Angewandte Semiolik ed Tasso Borbe and Krampen (Vienna 1978) pp 93-1 11

18 Jonathan Culler Slructuralisl Poetics Slructurnlism Linguistics and the Study of Litern- ture (Ithaca N Y 1975) p 16

19 Ibid pp 16-17 20 Anne Henault Semiotics in France in The Semiotic Sphere ed Sebeok and Jean

Umiker-Sebeok (New York 1986) pp 172-73


mann Hknault says that one of Flochs essays on Kandinskys Composition I v proceeds like the deciphering of clay tablets written in an unknown language and writing and that it is one of the more vigilant attempts to stay close to what occurs in a specific picture But the fact that Hknault can only cite four texts may be evidence of how hard it is to resist linguistic labels conversely the great mass of more loosely semiotic art history shows how easy it is to assume that visual semiotics needs only to attach names to pictorial forms23

The moral I would like to draw from these texts by Barthes Eco Culler Goodman and Saint-Martin is that Bal and Bryson are precipitate in assuming that visual semiotics exists in parallel to linguistic semiotics Semiotics has traditionally made furtive attacks on the problem of pictures and then beat hasty retreats into more linguistic fields such as the interpretation of pictorial symbols or the analysis of sign language It bears saying that there is no such thing as a well-articulated concept of a visual sign that does not depend immediately and explicitly on linguistic models Visual semiotics as it appears in such texts as Semiotics and Art History is an account of visual narratives and not a full theory of the semiotic nature of pictures I would agree with Mitchell that the whole concept of sign drawn from linguistics seems inappropriate to iconicity in general unless what is at stake is only a license to read images more freely24 But pictures are different from texts and the more ways we can

21 See ibid pp 173 174 On page 173 HCnault incorrectly cites Jean-Marie Floch Kandinsky SCmiotique dun discours plastique non figuratif Communications 34 (1981) 135-57 She also cites Floch Simiotique poetique et discours mythique en photographie analyse dun nu de Boubat Priflublications dUrbino 95 (June 1980) and Simiotique plastique et langage publicitaire Actes Simiotiques Documents du Groupe de Recherches Simio- Linguistiques ~ c o l e des Hautes Etudes en Sczences Sociales 26 (1981) and Felix Thurlemann Trois peintures de Paul Klee essai danalyse simiotique ThSse de troisisme cycle (Paris 1979) See also Thurlemann Paul Klee Analyse slmiotique de troispeintures (Lausanne 1982)

22 HCnault Semiotics in France p 174 At the same time Flochs analysis is entirely dependent on linguistic categories Reading Algirdas Julien Greimas Essais de simiotzquepoi- lique (Paris 1971) Floch decides de substituer au discours phonemique le discours des figures lineaires et chromatiques and his distinctions rely on Greimasian linguistic categories such as phimes syntagmes and formants (Floch Kandinsky SCmiotique dun discours plastique non figuratif pp 153 148) He is furthest from my proposal about the ambiguously semiotic nature of graphic marks when he assumes that unites signifiantes are constructed from unitCs discritesn (D 150) On the other hand he em~lovs Lkvi-Strausss idea of bricolage (a kind of thinking that is at once in pictures and in concepts) to negotiate between pictorial elements and iconographic (that is narrative) meanings-a move that is parallel tb what I have in mind here even if it begins and ends in very different places see pp 150-51 See also Claude Levi-Strauss La Pensie sauvage (Paris 1962) trans Levi- Strauss under the title The Savage Mind (Chicago 1966)

23 It is not accidental that the most exemplary essays in this vein have been about geomerric pictures where the entire problem of infinitely disordered organic marks can be kept to one side See for example Yve-Alain Bois Painting as Model (Cambridge Mass 1990) and Rosalind Krauss Grilles Comnlunicalions 34 (1981) 167-76

24 Mitchell Iconolo~p 58

832 James Elkins Nonsemiotic Elements in Pictu~es

find to acknowledge that the better What is at stake here is nothing less than the pictorial nature of pictures-both their nature as pictures and what we want to count as an adequate description of a picture The way out of the impasse of visual semiotics I think is not to refine the linguistic analogy but to take a close patient look at marks and try to say what really happens in pictures

Marking against Sense

Consider this figure drawn by Jacopo da Pontormo nearly lost in a welter of stains and accidental marks (fig 1)The sheet has aged obscur- ing the headless torso under a soft pillar of darkness and the commercial service that made the photograph has carelessly stained the plate painting brilliant white coruscations in the middle of the figures back and scattering the page with smaller stars and droplets On the left a dark cloud perhaps a mold is encroaching on the image and at the lower right the dense ink of the Uffizi stamp is slowly spreading into the surrounding paper A miscellany of smaller discolorations mars the bottom of the sheet-fusty smudges light acid prints from connoisseurs fingers-and the whole image is swept by long arcs recording the darken- ing fabric of the paper itself The drawing was made with two kinds of chalk a sharpened light point for the very fine parallel lines that com- prise most of the figure and a coarser black point that roughens the shadows and folds of the figures right flank The blacker chalk-almost like a charcoal-picks up the texture of the paper as if it were a stone rubbing so that its curving lines are shot through with horizontal stria- tions and its blackness blends by imperceptible stages into the black pillar The finer lines for their part become so faint that they lose themselves in the texture of the paper so it is not possible to say whether most of the back may not be covered by a faint tracery

It is an exemplary image Intentional marks vie with mold stray ink tarnish from fingerprints and chemical seepage and there is no certainty about the nature of even the intentional marks their order in the making of the image or the places they begin or end By presenting this as an initial example I mean to emphasize that if we ignore these obfuscations and contaminations in other images we do so at the risk of not seeing what is actually present What sense would it make to begin talking about this figure-in relation say to Pontormos frescoes his hypochondria his sexuality-if the figure itself is so much a part of its decaying setting that we cannot even tell where it begins or ends Marks together with the figures and images they build are always compromised by age by accident and-most importantly-by each other and they are always partly illegible

Many images are much better preserved than this one (though many

FIG 1-Jacopo da Pontormo study of a torso Uffizi Florence No 6572E Cour- tesy Gabinetto Fotografico Soprintendenza beni Artistici e Storici di Firenze photo no 120853


are worse) and it may seem that I have chosen a deliberately unworkable example But is it the case that a painted figure-say a Bathsheba by Rembrandt-is so much more easily detached from its setting Are its component marks more clearly separable one from another Does it make more sense in the end to refer to the figure of Bathsheba than it does to the figure by Pontormo And is it ever entirely clear what makes up a figure what happens within a figure Does it make more sense to refer to Bathshebas hair her navel or the length of her fingers-to name a few signs that have been mentioned in the literature- than it would to try to see a head in Pontormos figure or to decipher the positions of its fingers Marks blur and fade into one another and even the freshest drawing will have uncertain moments where the texture of the paper confounds the sense of a mark or a group of marks converge into a dark confusion or a mark moves so lightly across the page that it is not securely visible No image is composed in any other way

What is a figure A faint webbing of paper fibers and remnants of chalk a morass of sticky oil But it does not follow that it is a leap to go from talking about marks to talking about a figure or that the difference cannot be experienced except as a property of pictures A vigilant account of pictures cannot afford to begin and end with figures or with overall properties of marking If I get an account underway with the words A figure drawn by Jacopo da Pontormo then I have committed an act of violence against the image which I could then redress by giving a fuller version of how the image works To speak only of what must exist in spite of the marks against which it struggles-only of the figure or the represented thing-is to capitulate to a concept of pictures that imagines there is a gap between marks and signs and that believes the way to come to terms with it is to omit both the gap and everything that comes before it To elide the crucial moments of darkness when the picture in all its incomprehensible nonlinguistic opacity confronts us as something illegible is to hope that pictures can deliquesce into sense

Trace Trait and the Transcendental Conditions of Drawing

An example of a text that goes in this direction even though it begins with one of the more interesting accounts of graphic marks is Jacques Derridas Memoirs of the Blind I want to spend a little while with it as a preamble to my exploration of marking because Derrida touches on several properties that are important in graphic marking before he allows his account of drawing to assimilate itself to writing and language He announces his description of the drawn mark or trace-as opposed to the wider concept of the written trace-as a matter of two paradoxes two great logics of the invisible at the origin of drawing and then a


moment later by correlation two blindnesse~~~ In the pages that follow those three choices-paradoxes logics blindnesses-vie for preemi- nence though I would suggest that logics is in many ways what is at stake Blindnesses often turn out to be illustrations and figures and paradoxes are only one form of logical conundrum Despite their scare quotes the two great logics really do unfold as logical possibilities apart from the mythological and biblical stories that surround them

The first logic is transcendental the invisible condition of the possibility of drawing of the act of drawing of the drawing of drawingj It is neither the object nor the theme of any drawing but is required in order for there to be drawing This is the arche-trace that Derrida dis- cusses in Of Grammatology the production of dfferance by pure move-ment that sets up the formal properties of disjunction rupture difference and deferral of meaning before the content itself becomes visible and before there is any question of shape boundary or area27 The arche-trace is the condition under which marks may emerge become visible have meaning and take on form Derrida evokes it in Memoirs of the Blind without repeating the analyses he had made in Of Grammatology Differance Dissemination and elsewhere in order to set the stage for the discussion of specifically graphic marking2x

The second logic is the sacrificial defined as what meets the eyes the narrative spectacle or representation of the blind By becoming the theme of the first logic it reflects and represents the invisible unrep- resentable nature of drawing (MB p 41 MA p 46)29Drawing is said to

25 Jacques Derrida Mimoires daveugle LAutoportrait et autres ruines (Paris 1990) p 46 hereafter abbreviated MA trans Pascale-Anne Brault and Michael Naas under the title Memoirs of the Blind The Self-Portrait and Other Ruins (Chicago 1993) p 41 hereafter abbreviated MB

26 This is my translation of de possibilitk de dessin le dessiner msme le dessin du dessin Brault and Naas translate this as the possibility of drawing drawing itself the drawing of drawing (MB p 41 MA p 46)

27 Derrida Of Grammatology trans Gayatri Chakravorty Spivak (Baltimore 1974) pp 61 62

28 See Derrida La Diffkrance Marges de laphilosophie (Paris 1972) pp 1-29 trans Alan Bass under the title Differance Margins of Philosophy (Chicago 1982) pp 1-27 and La Dissimination (Paris 1972) trans Barbara Johnson under the title Dissemination (Chi-cago 1982)

29 It may be unfortunate that Derrida adds the phrase of the blind Since his account here is presented as a logic of drawing itself the idea of actual blindness-conjuring the many anecdotes and examples of blindness throughout the text-seems unnecessary If there is a transcendental logic to all drawing why would it require a picture of blindness to bring it out Shouldnt the invisibility of the transcendental condition become visible in any picture with any mark But the passage and the pages that follow are suggestive enough that I prefer to pursue a reading based on the excision of the phrase of the blind and of the related references to pictures stories and memoirs of the blind In visual art the single blindness in Derridas text becomes many kinds ofevasion noniconicity invisibility and unrepresentability in visual images His choices of representations of physical blind-


take place in play between these two principles and the fuller doctrine of the drawn trace comprises three aspects or types of powerlessness for the eyen-his more precise way of putting the reflection of the meta- phorical blindness of the originary invisible condition of drawing (MB pp 44 41 MA pp 48 46) First is the aperspective of the graphic act Derridas phrase for the necessary voyage into blindness that every mark makes as it moves across the blank surface Even if drawing is as they say mimetic that is reproductive figurative representative even if the model is presently facing the artist the trait must proceed in the night It escapes the field of vision This is a way of emphasizing the abyssal difference between the thing drawn and the drawing trait which is radically and definitively foreign to the phenomenality of the day (MB pp 4 4 4 5 MA pp 48 50) But the t ~ a i talso bears the traces of vision and daylight since the person making the mark will remember images motions of the hand and the appearance of the model all of those mem- ories guiding the hand across the page as it travels into the whiteness of the blank paper or the darkness of the not-yet-existent image

The second aspect is the withdrawal [retrait] or the eclipse the dqferential inappearance of the trait (MB p 53 MA p 58) Although Derrida says this is not an aftereffect a second or secondary aspect this second aspect refers to what happens after the act of marking itself

What is to be thought now of the trait once traced That is not of its pathbreaking course not of the inaugural path of the trace but of that which remains of it A tracing an outline cannot be seen One should in fact not see it (lets not say however One must not see it) insofar as all the colored thickness that it retains tends to wear itself out [tend a sextenuer] so as to mark the single edge of a contour between the inside and outside of a figure Once this limit is reached there is nothing more to see not even black and white not even figurelform and this is the trait this is the line itself which is thus no longer what it is because from then on it never relates to itself without dividing itself just as soon the divisibility of the trait here interrupting all pure identification and forming-one will no doubt have understood it by now-our general hypothec for all thinking about drawing-inaccessible in the end at the limit and de jure [MB pp 53-54 MA p 581

The process Derrida is describing happens whenever it begins to make sense to think of the mark itself without also thinking of a surrounding picture It is a logic or a transcendental doctrine because it wavers and weakens in the presence of actual marks The idea is that once a mark is made it will begin to appear as a color more and more equal to the color

ness-which center on the last three centuries of mostly academic narrative art in the Uest- appear too strongly circumscribed to be helpful in a general theory of the drawn trace


surface it marked and when that happens the edge of the mark will begin to take on the force and form of a mark and will become a mark in turn and then that half-imaginary mark will begin to wear itself out so that it too will seem to be a color and so on without end Later I will argue that this is only half of a reciprocal pair of processes so that what happens is actually more deeply troubling for the notion of a stable relation between mark and surface At that point I will also put some historically grounded terms in play so as not to have to rely on Derridas idiosyn- cratic use of the word trait For the moment however I want only to take note of the dynamic that is being proposed things are dying away collapsing fading wearing out The mark is not being seen there is a general lack of interest in seeing and a concomitant fascination with the invisible The trait the linear limit is in no way ideal or intelligible nor is it sensible it is a kind of original unavoidable graphic blindness (MB p 55 MA p 59) In one respect that is entirely appropriate given the theme Derrida has set himself (it is after all the Memoirs of the Blind) but it is also a renunciation of any particularized constrained act of seeing and that general motion culminates in the third and last aspect

It is then that Derrida gives a name to the abdications that have been murmuring through the text the third aspect is the rhetoric of the trait By this he means more precisely the withdrawal of the rhetoric of the trait the motion that also enables writing-this is his inevitable conclusion though it has been hidden until now

For is it not the withdrawal [retrait] of the line-that which draws the line back draws it again [retire] at the very moment when the trait is drawn when it draws away [se tire]-that which grants speech And at the same time forbids separating drawing from the discursive murmur whose trembling transfixes it [MB p 56 MA p 601

Given the force of this idea and the structure of the entire argument to this point is it enough to add as he immediately does that the withdrawal does not imply that all graphic marks tend toward writing This question does not aim at restoring an authority of speech over sight of word over drawing or of legend over inscription It is rather a matter of understanding how this hegemony could have imposed itself (MB p 56 MA p 60) When the rhetoric of the trait-what Jean-Frangois Lyotard might call the figurality of the figure as opposed to its discur- sivity-loses its predominance the way is opened to speech that is to discourse of all kinds Is it possible not to read an unthematized and even a personal lack of engagement with images in this logic that takes us so swiftly from transcendental conditions to the possibility of writing I mention this because Derridas logic is an especially well-worked-out version of a common subterranean motion in literary theory and philosophy in which the drawn trace is constituted by absence and difference but also contaminated with graphic elements that can-or must in


Derridas terms-fall away The drawn trace is imagined as if from a distance or abstractly so that it appears to fade to deliquesce or retreat yielding its unstable rhetoric in favor of the irreducible properties of written traces Derridas is a repressive reading a way of silencing the drawn trace by letting it melt quietly away into writing The truth of this lies in the many accounts that depend on such an idea from the nonvis- ual approach of some literary theory (including at this point Memoirs of the Blind) to the privileging of the sign in semiotic art history and one of the values of Derridas text is that it rehearses the idea so e l ~ q u e n t l y ~ ~ But here I would like to argue the opposite and show how the drawn trace can be a stable object of inquiry something about which it is possible to say more than it withdraws it wears itself out it becomes invisible it becomes writing

Apelles Protogenes Pliny and Thirty-Two Commentators

There is an old story told by Pliny about the Greek painters Apelles and Protogenes On the surface it appears simple and to generations of readers it has also seemed simpleminded because it describes how the two great painters came to understand and respect one another by taking turns drawing straight lines But the very poverty and implausibility of Plinys account has made it useful since the commentators have been obliged-or so they imagined-to fill in the missing meaning Together their remarks and theories form an elaborate contrapuntal chorus on the question of painted and drawn marks one that can take us a great distance from Derridas abstract account Here is the story in full as it ap-pears in Plinys ATatural History

A clever incident took place between Protogenes and Apelles Protogenes lived at Rhodes and Apelles made the voyage there from a desire to make himself acquainted with Protogeness works because that artist was only known to him by reputation He went at once to Protogeness studio The artist was not there but there was a panel of considerable size on the easel prepared for painting and it was in the charge of a single old woman In answer to his inquiry she told him the artist was not at home and asked who it was she should say had visited

Say it was this person answered Apelles and taking a brush he painted an extremely fine line in color across the panel [lineam summae tenuitatis]

When Protogenes returned the old woman showed him what

30 I would read Michael Frieds article Between Realisms From Derrida to Manet as evidence of this since he has to leave Derridas transcendental logic in order to say something about particular images (Michael Fried Between Realisms From Derrida to Manet Cr~tiralInqui~p21 [Autumn 19941 4)


had happened The story goes that the artist after looking closely at the finish of the line said that the visitor had been Apelles because no one else could have made so perfect a work and he himself using another color drew a still finer line on top of the first one [in ipsa illa]and leaving the room told the attendant to show it to the visitor if he returned and add that this was the person he was searching for

And so it happened since Apelles came back and ashamed to be beaten cut the lines with another in a third color [tertio colore lineas secuit] leaving no more room for any further display of minute work Afterward Protogenes admitted he had been defeated and rushed down to the harbor to search for the visitor and he decided that the panel should be handed down to posterity just as it was to be admired as a marvel by everybody but particularly by artists

I am informed it was burnt in the first fire that occurred in Cae- sars palace on the Palatine Before then it had been much admired by us since it had nothing else on its cast surface except almost invisible lines [nihil aliud continentem quam lineas visum effugientes] so that among the outstanding works of many artists it looked like a blank space and by that very fact it attracted attention and was more val- ued than every masterpiece there31

Like other stories in Pliny this one has attracted so much attention in the two thousand years since it was written that it has almost disappeared beneath its commenta r i e~ ~~ For any number of reasons it has appeared misguided unreliable naive or incomplete as if it were itself the coarse line that has been refined by the commentators subtle arguments For these purposes I do not want to ask again how likely it is that Gaius Plin- ius Secundus writing in the first century AD in Rome about a painting he had not even seen might have given us a reliable account of an exchange between two Greek painters in Rhodes over three centuries before Nor am I concerned that because Pliny was not an artist he may have bungled the sense of the story even if it took place And above all I do not want to rehearse the ways that European commentators have taken the story as the antitype of their own concerns using it to defend or fight the issues of the day Instead I want to use it and its steadily accumulating commentary in a way that does justice to the richness of the Western conversation about the most rudimentary elements of images I will consider the major kinds of solutions that the story has provoked about the relation of line color area and the linea summae tenuitatis in order to construct a preliminary set of possibilities for the drawn or painted mark

An exceptional essay by the Rembrandt scholar Hans van de Waal

31 PlinyNatural History trans H Rackham 10 vols (1952 London 1968) 9321 323 (353681-83) trans mod J J Pollitt also translates this text see J J Pollitt The Art of Ancient Greece Sources and Documents (1965 Cambridge 1990) pp 159-60

32 See Elkins On Monstrously Ambiguous Paintings Histo and Theory 32 (Oct 1993) 227-47


provides a passkey to the literature that has grown up around Plinys story In the course of reviewing its reception van de Waal lists more than thirty commentaries ranging from the early Renaissance to cub- ism Actually there are exactly thirty since two further commentators are mentioned in the literature but could not be identified33 The sequel to van de Waals essay is E H Gombrichs Heritage of Apelles Gombrich mentions van de Waal and numbers van de Waals own interpretation as number thirty-one and his own as thirty-two My reading therefore will be number thirty-three in a series that is likely to continue until the ways we think about graphic marks become a little more stable

The Ontology of Marks

It is useful to start by looking at several speculative interpretations those with the most affinity to the transcendental doctrine of the trace I have in mind two preliminary observations regarding the nature of marks and their surrounding surfaces Even though they have only a faint correlation with plausible scenarios of drawing and looking these two possibilities already go beyond the transcendental doctrine interrupting the slide from drawing to writing and blocking the implicit analogy between them The first deals with the question of how a mark can become a surface and the second asks how surfaces coalesce into marks

Plinys text seems often to have called to mind a kind of schematic diagram in which three vertical lines of different colors nestle inside one another each one appearing like a band of uniform width If we take this as the solution-and many authors have assumed something similar- then we face an interpretive dilemma An eighteenth-century editor of Pliny David Durand wondered about the number of lines the finished painting would have had Pliny speaks unmistakably of three lines he writes although van de Waal points out that it was a commentator Joannes Fredericus Gronovius who put those words in Plinys mouth (quoted in LST p 18)34TO Durand there are manifestlyfive lines five potentially equal bands of color side by side set against a monochrome background This difference of opinion which seems so trivial is in fact fundamental and it gives rise to two entirely different possibilities that are also competing functions of the mark

33 Hans van de Waal The Linea Summae Tenuitatis of Apelles Plinys Phrase and Its Interpreters trans John Walsh Jr ~eitschriyttfiir~sthetik und Allgemine Kunst Wissenschaft 12 no 1 (1967) 31 hereafter abbreviated LST

34 See Pliny Histoire de la peinture ancienne extraite de lHistoire naturelle de Pline liv X X X K trans D[avid] D[urand] (London 1725) pp 65-66 261-63 and C Plinii Secundi Naturalis historiae tomus primus [-tertiusf cum commntariis amp adnotationibus Hennolai Barbari Pintiani Rhenani Gelenii Dalechampii Scalzgeri Sa lm i i Is Vossii amp variorum ed Joannes Fredericus Gronovius (Leiden 1669)


Are there three lines or five If Durand is right then the results of the contest are nothing more than five bands of color laid side by side in other words the very act of making a mark would divide the existing surface (or as in this example the existing mark which is now to be considered as a color a field or surface of uniform aspect) into two parts producing two colors where there had been one-at the same time the mark would destroy itself making itself into a third color In- stead of three lines each inside the last there would be five colors- five fields of color-with no mark remaining On the other hand Durand saw that it is possible that the same configuration might remain three lines as Pliny said that is an observer in Rhodes or in Caesars palace would see the full width of the configuration as the first line and the next narrower width as the second line In that case it would be correct to say that a mark never creates new fields-new colors-by dividing the surface it marks but merely overlays the surface as if it were stamped upon it A picture would not be composed of marks and surfaces but of imbri- cated marks

Together these two options describe what I will call the ontological instability of the mark and they spawn a cacophony of compromises Per- haps the configuration on the panel was a mark (the central and final line) surrounded by four fields of color (the two lines that had been split and turned into surfaces flanking the remaining line) or two marks (the ultimate and penultimate lines) flanked by two fields of color or-in as-cending order of extravagance-three colors flanked by two lines or two lines a field of color and two more lines Although these solutions are counterintuitive-and as far as I know no commentator has even suggested them-they are in accord with the two grounding possibilities and they follow from Durands offhand remark Any of them could be explained by appealing in turn to one or the other of the options for marks and surfaces Either a mark becomes a surface-or two surfaces or many surfaces-when it is crossed by a second mark or it remains a mark all the conceivable combinations of markmaking proceed from those two options

In the first instance the mark must constantly wear away as Der- rida says weaken and become a surface itself That dissolution would be linked to the act of markmaking so that the presence of a mark would set the weakening process in motion by drawing attention to the newly created surfaces on either side of the mark (or all around it-I am simpli- fying in order to remain within the bounds of Plinys example) and by inexorably implying that the mark itself is merely another surface What began as a mark energetically cutting or dividing another mark would become a mark and two surfaces and finally a set of three surfaces It is possible to put this more elegantly without the necessity of calling on three lines of different colors by assigning deliquescence to the mark itself Marks exfoliate I would rather say by drawing attention to their


boundaries so that the boundaries become outlines in their own right when that happens the boundaries themselves can be perceived as marks turning both the original mark and the original surface into surfaces

In figure 2 a mural fragment from Delos (painted around 100 BC

and therefore older than Pliny but younger than Apelles) a ram being led to sacrifice has suddenly stopped and planted its hoofs on the ground A man leans forward apparently to give the ram a tug The archaeologist Vincent Bruno identifies this as an example of the loaded brush technique following a description Rudolf Wittkower proposed for baroque oil sketches35 In this method Bruno writes figures are loosely brushed in a milky translucent white pigment over a carefully prepared surface of dark purplish or brownish red so that forms come gently forward as if they were pools of milk on dark stone (HEp 42) Here it is easy to see marks soften into color areas and to imagine the edges of the marks as marks themselves In the mans folded tunic the space between marks strengthens this impression by creating what are unhelpfully called negative spaces that appear as dark brushstrokes Though this laying-down or melting of the mark is a universal phenomenon it should be noted in passing that marks also resist their fate by manifesting their fluid nature (or in general by underscoring the medium of the mark whether it is paint metal or empty space) The rams left leg is an excep- tionally beautiful example A single modulated area of white forms the joint and its bony continuation and it is so clearly both bone and milk or gristle and paint that it is not easy to see the leg as a color field and the edge as a mark Yet in general marks retreat into areas of color For theoretical purposes it is important to insist on the potentially infinite regression that this implies As Derrida says the undrawn boundary now perceived as a drawn contour can itself seem to be a mark and when that happens its boundaries are subject to the same exfoliation All marks must (as Derrida says incidentally stressing the force of will it may require to sense the process) dissolve into surfaces as if they were shedding skins In the end there is no such thing as a mark-there are only surfaces

But it is not enough to stop here as Derrida does since the opposite motion is also inherent in every mark the act of making a mark also turns the surface into a mark so that it is perceived not as an infinite or undifferentiated surface but as a region with definite boundaries and therefore ultimately a mark In Rembrandts study for the etching The Great Jewish Bride marks slither over each other like eels in a pot transforming the figure into a tangle of sinuous marks (fig 3) There is no exact term for this surface-becoming-mark and I will reserve for it the word field to denote a surface whose boundaries are constituent aspects

35 Vincent J Bruno Hellenistic Painting Techniques The Evidence of the Delos Fragments (Leiden 1985) p 42 hereafter abbreviated H R See Rudolf Wittkower et al Masters of the Loaded Brush Oil Sketches from Rubens to Tiepolo (New York 1967)

FIG 2-Fragment of a white monochrome frieze with a ram being led to sacrifice late secondtearly first century BC Le Palestre amp Granit Delos From Vincent J Bruno Hellen~tic Painting Techniques The Euidence of the Delos Frapmts (1985)

FIG 3-Rembrandt study for the etching The Great Jewish Bride detail c 1635 Na- tionalmuseum Stockholm From Otto Benesch The D7awings of Rembrandt A Critical and Chnmological Catalogue (1954) vol 2


of its being A single line such as Apelles drawn on a large panel will turn the surface of the panel (which was until that moment undifferentiated and effectively unbounded) into a field with internal boundaries where it meets the line and external boundaries where it meets the edges of the panel The more aware an observer becomes of the shape and extent of the field the more it resembles a determinate mark as Pliny informs us Protogenes had prepared his panel-he had smoothed it and probably colored it-so that the blank surface was already a mark Al-though in one sense the brides body is a blank surface crossed by nearly nonmimetic lines in another it is nothing but marks since every mark makes us more sharply aware of the forms it bounds turning them into fields and therefore finally into marks In effect markmaking turns surfaces into marks

These abstract possibilities show how easy it is-even within the mainly ahistorical account that Derrida proffers-to demonstrate that the drawn trace is incompatible with essential features of the written trace Marks exfoliate into fields and ultimately into surfaces and they also gather surfaces into fields and finally into marks so that visual arti- facts are nothing but marks Because both these possibilities are continuously true of graphic markmaking it is not sufficient to say that graphic marks wear away giving up their uniqueness repleteness or rhetoric and drawing near to writing I regard these arguments as decisive against accounts that portray graphic marks as somehow sliding toward or fading into written marks-and therefore as a sufficient obstacle to accounts such as Derridas or to the project of a semiotic art history insofar as it depends on an implicit parallel between the elementary and compound structures of graphic and written marks Graphic traces are unruly as subsemiotic elements might be expected to be but they are unruly in a different way from written marks and their instability does not fade when they combine into larger units In short the parallel is untenable and semiotic art history needs to look elsewhere for a theoretical model

Modes of Marking

What I have said so far has mostly to do with logical conditions that do not have immediate links to the ways pictures have been produced and viewed and for that reason I imagine that it may have been unconvincing to historians or artists But the best examples of the irreducible properties of marks come from historical practices and so I want to begin moving away from issues that are decided within philosophy and linguistics and toward those that can be framed best in art history or criticism

These meditations based on Durands remark are not entirely without historical purchase and in fact they have deep resonance with a

846 James Elkins lVonsemiotic Elements in Pictures

range of concerns of European painting between the sixteenth and the eighteenth centuries The Rubensistes and Poussinistes also cited Plinys story in support of their interpretive agendas-either emphasizing the fineness of the lines as evidence of the linear qualities associated with Nicolas Poussin or else doubting that the contest could have been about lines alone in accord with the colorist qualities associated chiefly with Peter Paul Rubens (see LST pp 18-19) Thus Claude Perrault for example emphasizes the thinness and agility of the lines arguing that the panel was a typical example of artists exhibitions of manual dexter- it^^ The trait and the ontological instability of the mark do not ade- quately paraphrase baroque academic concerns (the Rubensistes and Poussinistes grounded their exchanges in the practices of narrative painting and seldom strayed into improbable logics of marking) but the vacillation between marks and surfaces parallels the indecisive exchanges between proponents of line and color that echo in the literature from the sixteenth century onward Among the many reasons that this literature remains inconclusive is that line and color depend on one another in ways that were not always acknowledged As Jacqueline Lichtenstein has argued color was not excluded from rational inquiry or from philosophy (though that exclusion is still maintained as a historical fact) and form was not the exclusive province of certain schools (as was maintained at various times during the Italian Renaissance and in the French Acad- emy) In addition to the philosophic and historical interdependence of

36 A limitation of van de LVaals survey is that it tends to lump rather disparate figures under the headings of line or color thus Franciscus Junius Carel van Mander and Roger de Piles each end up as an example of the privileging of line For example van de Waal cites Franciscus Juniuss De pictura veterurn (1637 Rotterdam 1694) p 125 in the roster of those who privilege line but Juniuss opinion is not quite as clear as van de Waal suggests See LST pp 18-19 In the words of the editors of an English edition ofJuniuss text he seems to build on Demontiosius interpretation but (like Salmasius) he sees the contest as one of lines (Franciscus Junius The Painting of the Ancients De pictura veterurn trans Junius vol 1 of The Literature of Classical Art ed Keith Aldrich Philipp Fehl and Raina Fehl [Berkeley 19911 p 173 n 11)

37 See Claude Perrault Parallele des anciens et des rnodernes en ce qui regarde les arts et les sciences (1688 Munich 1964) Perrault speaks of a rival interpretation proposed by Monio- cosius (Ludovicus Demontiosius) which I will discuss below

Nest-ce pas la une temeritk insupportable mais afin que vous ne maccusiez pas de maltraiter un homme qui peut-estre a fait de gros livres je ne parle quaprks Monsieur de Saumaise qui en dit beaucoup davantage amp qui paroist avoir estamp plus bless6 que moy de cette insolence I1 est donc vray quil sagissoit entre Protegene amp Appelle dune adresse de main amp de voir a qui seroit un trait plus deliamp Cette sorte dadresse a longtemps tenu lieu dun grand merite parmi les Peintres IO di Giotto en est une preuve [P 1521

38 See Jacqueline Lichtenstein La Couleur eloquente RhHorique etpeinture a lige classique (Paris 1989) and Stephen Melville Color Has Not Yet Been Named Objectivity in Decon- struction in Deconstruction and the Ifisual Arts Art Media Architecture ed Peter Brunette and David TYills (Cambridge 1994) pp 33-48


form and color there is also the kind of formal dependence I have dem- onstrated here It shows in a logical way how form and color cannot be understood independently and how each corrodes the other Let me take that speculation then as the first mode of responding to Plinys story At least four more are latent in the literature that van de Waal has collected (I will be enumerating modes of response and not repeating van de Waals thirty interpretations By modes I mean formal configu- rations ways of imagining what the panel looked like There are many fewer of them than there are interpretations of the contest)

Perhaps because of his own interests in Dutch art van de Waal begins with a look at Care1 van Manders Het leven der ovde antycke doorluchtighe schilders (1617) Like many others van Mander doubts things could be as simple as Pliny says To give my own opinion freely he writes

I do not believe that these were simple straight lines or strokes as many think who are not painters but rather a contour of an arm or a leg or a profile of a face or something similar which were very delicately set down in various colours and intersecting with each other in various places which Pliny meant to indicate by cleaving (secuit) [Quoted in LST p 6]39

For van Mander it is inconceivable that painters would have admired something as obvious as three straight lines and so he imagines figures or parts of figures Van de Waals reading sheds light on van Manders assumptions about art and especially about the profession of painting and the plausible practices of the studio At the end of the seventeenth century Roger de Piles echoed the same notion without going so far as to suggest actual figures To him the word linea should be translated as design or outline

to think it was a simple line divided from another in its whole extent would be ridiculous and shock every one that has the least knowl- edge of Painting there being in that no sign of capacity to be shewn nor skill in the art Subtlety is not in the line considerd simply as a line but in the intelligence of the art which is shewn in the lines of a design40

39 See Carel van Mander Het leven der ovde antycke doorluchtighe schilders (Amsterdam 1617)

40 Roger de Piles Abregi de la vie des peintres (1699 Paris l715) p 117 trans under the title The Art of Painting (London 1744) p 78 The ellipsis in my quotation omits the following lines

What has givn rise to this interpretation is in my opinion the ill construction of the word Linea for by Linea in that place is to be understood either Design or Outlines Pliny himself make use of it in this signification elsewhere when he says ofApelles that he never let a day go over his head without designing Nullo Dies sine Lineamp which was not drawing simple lines but to accustom himself to correct designing

In the same manner we should understand the word Subtilitas not to give an idea


But what exactly could de Piles or van Mander have had in mind In van Manders case it is not unreasonable to picture the kind of sinuous contours that an early seventeenth-century painter especially one famil- iar with Italian developments would have practiced As we know from his text he admired a number of painters whose technique derived largely from sixteenth-century Italian practice artists such as Bartolo- meus Spranger and Hendrick Goltzius come to mind and so does the ultimate model in their genealogy the Michelangelo of the cartoon for The Battle of Cascina and the drawings for the Sistine Chapel ~ei l ing~ For these purposes it is not necessary to extract a single drawing from the many possibilities or even to specify the locus classicus of the type too exactly Even in a preliminary assessment it is clear what kind of drawing would produce lines intersecting with each other in various places The twining mingling and separating lines of Rembrandts drawing for The Great Jewish Bride would do perfectly well as a model though the sense of elusive perfection is better represented by one of Michelangelos densely reworked drawings such as this one of the Crucifixion where there are so many contours that they form a single dense braid on one side and slough off into thin air on the other (fig 4)

Just as Michelangelo went back over his own contours altering and correcting-but not erasing-so Apelles and Protogenes would have moved across each others drawings producing at last an unsurpassably fine contour Michelangelo often began with dull chalks sketching rapidly and schematically and then went more slowly and forcibly with a freshly sharpened chalk so that the last lines were sharper and darker than the others But if such an interplay of skillful contour[s] (as van Mander calls them) were at stake here then it would need to be said that Apelles did not win the contest by his line alone but with his line in palimpsestic combination with the lines that had previously been drawn (quoted in LST p 6)42The single winning line would not exist in its own right but together with the others forming a result that is difficult

of a very delicate line but of the exactness and fineness of the design Thus the Subtlety is not in the line [P 781

After the quotation in the body of the text de Piles concludes I confess however that the word Tenuitas which is in the same place of Pliny may create some difficulty in this explica- tion which I believe is not unanswerable for by that word the fineness and exactness of an outline may very well be understood (p 78)

41 See van Mander Dutch and Flernish Painters trans Constant Van de Wall (New York 1936) William Hogarths S undulating motion line is also compatible with these ideas but in this context it is better to trace them back toward their origins rather than forward to their dissolution in neoclassicism (Elkins Two Conceptions of the Human Form Bernard Siegfried Albinus and Andreas Vesalius Artibus et Historiae no 14 [1986] 96) See William Hogarth The Analysis of Beauty Written with a View of Fixing the Fluctuating Ideas of Taste (Lon-don 1753)

42 See van Mander Het leven der ovde antycke doorluchtige schilders

FIG 4-Mielangelo Christ on the Cws detail Windsor Castle 12761~


to capture in words On the one hand it would have been (or I should say it actually is in drawings such as Michelangelos) a single winning line whose fineness and correctness would be made apparent by the neighboring lines and on the other hand it would have been a line composed of three lines that did not combine to create an area This is an essential irreducible source of complexity in drawings of this kind properly called by one of its Renaissance names-the contorno or contour Because the lines overlap hesitate leave gaps twist under and over one another and vary continuously in speed darkness and width they rarely form an area in which two outer lines would become contours and a single inner line would appear to occupy a median interval Instead the concept of line predominates no matter where an observer looks in Michelangelos contours even where the lines momentarily diverge leaving white areas between them or where they collapse into a single jet- black mark In some places Apelles last line would have overrun one of Protogenes or one of his own but where that happened a careful eye would still have discerned a slight convolution in the line an irregular doubling or trebling of force that would give away the presence of two or three lines on top of one another In other places Apelles first line might have been so wide that it would ordinarily have evoked an area like a shade produced by drawing with the broad side of a charcoal but the presence of finer lines inside it would immediately recall the concept of line and cancel the concept of area so that both the broad stroke and its combination with the other strokes would act as a line or as the evocation of the idea of pure lines This is therefore a second way of imagining Apelles and Protogenes painting three intertwined lines that compose a contorno-a dance or a cascade of contours that work to form a single line or to submit to the most powerful and final line-without ever fully succeeding in erasing their own multiplicity It is a truly complicated structure already far beyond the reach of the simple trait that emp- ties itself into a written sign and there are others just as intricate and entirely different

A third mode arises from the possibility that Pliny might not just have been reporting correctly but that he may be exactly right right in every detail of his account of three differently colored lines each inside the last leaving no more room for any further display of minute work To be perfectly accurate-as previous commentators have generally not been-we must imagine an extremely thin initial line (otherwise Proto- genes would not have known who had visited while he was out) and not forget the large expanse of the panel all around43 Since it would not be

43 There is some weak contextual evidence that Pliny intended thin lines or only thin lines the passage in the Natural History before the anecdote about Protogenes concerns Apelles skill at knowing when to take his hand away from a painting a memorable warning that too much detail can be harmful (quoted in Pollitt The Art of Ancient Greece p 159)


possible to reproduce such a painting on a printed page (it would fall outside the resolution of even the best printing processes) the painting has to be approximated twice once as it might appear at arms length and again in detail approximately life-size (fig 5) Now if the third line were really unsurpassably fine (and if it were only unsurpassably fine with no further attributes) then it would have been painted with a single- hair brush and it would as we know from Dutch and English miniature painting have been at the limit of unaided human vision The speed of the mark would have had to be excruciatingly slow I can convey that best by relating another story a student of mine once attempted to copy a small painting eight inches wide that represents several dozen figures- Joachim Wtewaels Battle between the Gods and the Titans (c 1600) After he made a photograph transferred it to a copper plate and traced the outlines he began to paint and after seven weeks of work five hours a day two and sometimes four days a week he had completed only one leg of one figure When the scale decreases so does the speed From that I surmise that the panel if it existed exactly as Pliny describes it would have consisted of ever-so-slightly wobbly lines whose fineness would have made them nearly invisible (quasi invisibili as a nineteenth-century commentator put it)44 And because they were done in different colors we might reasonably expect them to have created a shimmering effect as the colors mixed and separated in the observers eye None of this is cap- tured by the thickened schematic offered in Gombrichs essay which in the original looks more like a sketch for the colors of a flag (fig 6) If we take Pliny seriously the painting would be a most curious object from a distance it would be a line of indeterminate color (just as some of Seurats color areas blur to a uniform unnameable color with increasing distance) and from up close it would be a dilemma for the eye one moment a single line the next three lines then a shimmering multicolored wire As in van Manders solution the marks would sometimes fuse becoming a single mark-but here there would be no breathing room between them A mark would be a scintillating incomplete fusion of other marks producing an effect of rapidly alternating stripes and intermediate opti- cal greys I would insist on this as the consequence of a strict reading of Pliny If we say otherwise-for example if we suppose that both Apelles and Protogenes lines would have to be visible at all times-then we cannot also be talking about unsurpassably fine lines

Gombrichs interpretation the thirty-second is that the contest was

If the stories are read together then the panel Apelles and Protogenes collaborated on would appear as a lesson and also a warning to painters about excess

44 Carlo Roberto Dati Vite depittori antichi (Milan 1806) p 264 Dati gives very little commentary on Plinys story but the context suggests he would have been sympathetic with van de Waals interpretation that Pliny should be taken at his word-and for that reason I enlist him here as evidence for the most literal reading

FIG 5-Reconstruaion of the line of Apelles Left-overall view of the panel Right- detail natural size Drawing by the author

FIG 6-Reconstruction of the line of Apelles From E H Gombrich The Heritage of ApeUes Studies in the Art of the Renaissance (1976)


to see who could produce the strongest illusion of depth or relief by making the single line seem to stand out or recede Protogenes line would have added a sense of light and shade to the initial line and then Apelles would have trounced Protogenes by adding the unsurpassable splendor the highlight that complements the lux or lumen (the light) and the umbra (the shade) by adding the gleaming ridge gloss or glitter that makes the line pop out in three dimensions Gombrich commissioned a painting to show the stages as he envisions them (see fig 6) In his words

If the prepared panel in Protogenes studio was grounded with the first colour say blue Apelles could have drawn his exquisite line with a second say brown Protogenes superimposing a thinner line with another colour (alio colore tenuiorem liniam in ipsa illa duxisse) might have used a third say ochre to produce the effect of illumination the third coloured line with which Apelles cut the lines leaving no further place for more subtlety (tertio colore lineas secuit nullum relin- quens amplius subtilitati locum) would have been done with the fourth pigment the trumping line of a gleaming white45

As Gombrich points out one of the advantages of his reading is that it makes new sense of Plinys phrase visum effugientes by reading effugientes as lines which appear to recede from sight as opposed to almost invisible lines The reading also has shortcomings It depends on the choice of colors since if the ground were not blue (an odd color for any imprima-tura even in a decorative panel) the initial line might not appear to stand out as well as it does in Gombrichs reconstruction and so it might not have set the contest in motion More seriously it is difficult to see how Apelles would have taken a thin initial line to be mostly about illusionistic relief Gombrich glosses the point once by calling Apelles first line thin and then by saying it was exquisite and he offers an indeterminately enlarged illustration with lines sufficiently wide to permit reproduc- tion-but wouldnt a thicker line have done better to inspire thoughts of plastic relieD4= (But of course if it had been thicker Protogenes could hardly be expected to guess who had made it) And finally if Apelles meant to make a statement about his ability as master of the splendor why not paint the initial line in relief complete with umbra and splendor

But all the descriptions of the story have their weak points It seems to me that one of the strongest descriptions is van de Waals own opinion that the contest was merely about manual skill and that all subsequent readers who have felt it necessary to add some further explanation have done so in part under the un-Greek assumption that there must be more to the comparison and judgment of painting than manual skill And it

45 E H Gombrich The Heritage of Apelles The Heritage ofApelles Studies in the Art of the Renaissance (Ithaca N Y 1976)pp 15-16

46 Ibid p 15


also needs to be said that Gombrichs account is not entirely new since it echoes a disagreement that took place early in the seventeenth century between Claude de Saumaise and the antiquarian Louis de Montjosieu (Ludovicus Demontiosius or Moniocosius as Saumaise calls him) (see LST p 16)47 Moniocosius thought that the contest did not turn on lines but in van de Waals words on extremely subtle variations within three basic gradations from light to reflection and from reflection to shadow (lux splendor umbra) (LST p 16) In Moniocosiuss view with painting-in-color one cannot talk about lines at all (certum est enim in pictura colorata nullum esse prorsus linearum usum) (quoted in LST p 16)48 Their disagreement puts us back in the field of academic debates over form and color and Moniocosiuss view is in accord with the speculative description I made that marks will deliquesce into colored fields-it is roughly a Rubensiste opinion Saumaise and Monio- cosius belong to that period and to those ways of putting these questions and I mention them here because of the intriguing appearance of the names for the three elements of chiaroscuro Moniocosius and Gombrich use the same terms but apply them differently In Gombrichs scenario each painter takes up one of the elements in turn-first Apelles paints lumen then Protogenes adds umbra and Apelles wins by completing the splendor In Moniocosiuss view the painters would have used all three elements in each of their lines He seems not to have had a specific theory about the appearance of the panel or to have decided whether the con- testants were painting contours or other elements but he is clear on his central point-which he shares with Gombrich and which distinguishes them both from authors such as Saumaise who want to retain what they take to be Plinys literal meaning-that thinness of line was not at issue

Let me imagine then one of the possible appearances of the panel according to Moniocosius it could have been a line perhaps even broader than the one Gombrich offers as a schematic demonstration but it would also have had marvelous gradations of light and shade and moments of twinkling or shining highlights If we are to retain Plinys description and say that Apelles line was inside Protogenes then Apelles line would have had to add to Protogenes illusionistic contours in such a way that the two lines would represent a single object in relief-certainly not an easy task But it would then also make sense that Apelles line would be the last word not only because of its thinness but because no further relief could plausibly be extracted Moniocosiuss hypothesis makes good sense as an artists contest and it is also a version of something that happens frequently in the studio instruction of painting-an ongoing contest to add nuanced rather than monochromatic relief

47 See Claude de Saumaise [Claudius Salmasius] Plinianae excertationes (Paris 1629) 48 See Louis de Montjosieu [Ludovicus Demontiosius] Gallus romue hospes ubi anti-

quorum monumenta explicantur (Rome 1585)


Nuanced relief within a continuous area is a stock-in-trade of painting as the artist works both lights and darks into a modulated field It can be challenging enough to create the required reflectivity and curvature- say of a leaf a porcelain bowl or an abstract field-and adding further relief to such a field can become nearly impossible Protogenes calling card might have been exactly that an illusionistic scroll or banner un- furled down the center of the panel done with exquisite attention to lumen umbra and splendor

Gombrichs account suggests a relatively simple linguistic model-a pictorial vocabulary obeying a determinate grammar The three terms become relative comparative adjectives and one term can be displaced in favor of the others a brighter splendor can turn a duller one into a mere lumen and so forth Splendor is meaningless or indeterminate by itself against a dark ground it becomes ambiguously lumen or splendor against a lighter ground it becomes umbra For its part umbra might even function as splendor if its surrounding fields are in darker shades Even lumen cannot stand on its own because in order to be perceived as light it needs to contrast with a previous mark or surface that would then become umbra There is a beautiful interpenetration between Gombrichs and Moniocosiuss theories By itself Gombrichs schema is compatible with the concept of disjunct graphic signs (at least where the schema exists in the way he posits it with three clearly separate tones or colors rather than indeterminate shades that seem to divide into three groups) and I take it as one of the best instances of a workable visual semiotics But with Moniocosiuss theory also in play-and I would say it virtually always is except in images that are mechanically produced-each region is also the site of its own subtle variations so that every lumen is also either part of composed of or equivalent to splendor and umbra There are no distinct marks where each is a composite of others or a fragment of some larger unity that would be a mark

And there are other possibilities as well if we take the contest to play off the basic lux umbra and splendor in a more sophisticated manner In describing the loaded brush technique of the Delos fragment Bruno gives an eloquent description of such a possibility

The method of showing form has nothing to do with ordinary shading as we see it for example in the Pella mosaics where there are also figures expressed mainly in white The loaded brush technique allows varying amounts of pigment to flow which the artist controls to create variations of tone within the white without actually having to delineate shadows [see fig 21 In some areas one may detect touches of grey that seem to have been added over the white but very often this is done by mixing a grey that so closely matches the transparent tones originally created by the moving brush that the added strokes of dark appear disguised we see them rather as part of the system of accidental transparencies characteristic of the


loaded-brush technique This disguising of the added shading helps to maintain the impression of spontaneity that is the very essence of the technique which relies on the avoidance of any building of im- pasto except upward into light where the white the only element of the medium simply thickens to accomplish the sense of density and weight [He p 471

The technique can mimic the triad lumen-umbra-splendor without repro- ducing it at any point so that it seems astonishing that so many subtle distinctions are suggested by means of the manipulation of a single white (He p 49) Could this have been the kind of contest that the two Greek painters shared each taking a turn with a monochromatic line Isnt it possible that the subtle shading (harmoge) that is mentioned by Monio- cosius and other authors might pertain also to this monochromatic tech- n i q ~ e ~ ~Within a single color and within the boundary of a single continuous mark it would be possible to create virtuoso effects more amazing than the routine and predictable umbra and splendor-which it is tempting to say anyone could have executed Perhaps too the contest alternated between lux umbra and splendos in both Gombrichs and Mo- niocosiuss senses and also this more fluid and less grammatical loaded brush technique Either way the marks would have been undecidably multiple and homogeneous

The commentators suggest several further modes but I will name only one other-a fifth way of imagining the panel first suggested by Leon Battista Alberti Of the three divisions of painting that he proposes in On Painting (circumscription composition and the reception of light) the first is the pivotal one Circumscription (circonscrizione)is the study of the turning of the outline (lhttorniare dellorlo) and Alberti says it is best if the outline is drawn as thin as humanly possible I say that in this circumscription one ought to take great pains to make these lines so fine that they can scarcely be seen (quoted in LST p 7)j0 Alberti alludes to Plinys story and van de Waal comments that it is clear why Plinys story occurred to Alberti at that moment and why these lines no longer held any mysteries for him (LST p 7) Alberti accepted the idea that thin lines were what was at stake although he also must have been imagining them as contours around figures (or practice for such contours) rather than straight lines Albertis orlo-as distinct from edges contorni and other concepts-should be nearly invisible because it represents that invisible datum of sight the visual ray that gives us the edges of things Alberti knows painters need to actually draw such rays but they should be wary of drawing them too coarsely This is the perennial studio problem of the existence of outlines as every drawing teacher cautions

49 Junius The Painting of the Ancients p 173 n 11 50 See Leon Battista Alberti On Painting trans John KSpencer (New Haven Conn

1966)p 68


we do not normally see outlines though we routinely draw them as if they were as visible as other objects in the world It is helpful to adopt Albertis other word circonscrizione to name the invisible outline itself which is inadequately suggested by the orlo Albertis idea of a very faint line is part of a curious complex of notions and forms the invisible concept (circonscrizione) the edge that lacks a form of its own the thin outline (orlo) and the sturdier outline or contorno The orlo is a twilight creature barely existing on paper or panel remaining attached to the abstract haunts ofperspectiva the geometric theory of vision When I think of the exfoliation of the mark the initial moment when my awareness begins to be attracted by the edge of the mark and to play with the idea that the edge of the mark may also be a mark even though nothing is drawn I would use Derridas word trait But I can imagine an orlo there and then it might congeal into the sense of a full outline surrounding the mark and eventually becoming a mark in its own right Without the orlo there is only the choice between mark and surface or mark and field with it there is no longer even a binary opposition from which to begin This is the simplest and most damaging equivocation that stands in the way of visual semiotics

The Thirty- Third Reading

Since I think van de Waals account of Plinys story is the most likely to be correct I am not adding a thirty-third theory of the panel itself (though I would submit that if we consider Pliny a good witness then it is reasonable to take him at his word and try to imagine what a truly thin line would look like) If I were to deliberately add a thirty-third interpretation it would be in favor of the loaded brush technique as against Gom- brichs more programmatic demonstration of graded tones that does not depend convincingly on the artists skills But I would rather consider the tradition as a whole as a uniquely rich compilation of responses to an open field of possibilities As van de Waal notes the particular interests of each of the interpreters (Gombrich included) come clearly forward in the sense they make of the original story and of one another Their opin- ions can be used to make another kind of point Each way of imagining the panel is a way of thinking about the most basic elements of painting and so together they are a useful source to begin culling examples of the ways marks have been thought to work

I want to emphasize that the five modes I have named here are not merely complicated cases of simpler properties of graphic traces There is no path leading back to the philosophic ground of the logic of the trace or to the reductive polarity of the semioticsemeiotic choice These modes belong to the domain of the graphic trace they are its ABCs its conditions of meaning (Still they are not the entire alphabet or even a


large portion of it I assume there are many more modes and certainly they are linked in many more ways to particular historical moments and local strategies for making and viewing images) Each mode I have named amounts to a different mode of markmaking and a distinct nonsemiotic structure

1)The ontologtcal instability ofthe mark is a double and conflicting condition On the one hand each mark exfoliates into fields and endlessly generates new marks out of its edges so that the mark itself is fugitive lost in a sea of fields on the other hand each mark coalesces its surrounding surface into fields and finally into other marks so that the surface is fugitive and hardens everywhere into a landscape of marks Unlike written signs drawn and painted marks are insecurely linked to their grounds and the same is true at the level of the figure-a fact that has to be suspended in order to get on with art-historical interpretations that treat figures as if they were signs detachable from their grounds

2) Contorno is a name for a serpentinate braiding of marks that is at once a mark in itself and many marks supporting a single mark The part does not exist without the whole and vice versa The practice of drawing contorni and imagining figures as objects bounded by successive twining marks began in the first decades of the sixteenth century and continued through the end of the eighteenth51 Today the practice is looser and no longer tied to formal academic ideals but any drawing by repetitive approximation will create contorni

3) A mark can also be a shimmering thing at the edge of analysis one instant it will seem to be solid and homogeneous and then it will resolve into parts or exhibit a determinate structure If the panel had had nearly invisible lines it would have been one example another is Seurats exper- iments I would call this Plinys theory since a literal reading of the text yields it as a consequence Since painted and drawn marks routinely op- erate at the limit of resolution of the eye barely perceptible forms are another universal mode for graphic marking Their indistinctness is not always meant to be resolved and so shimmering is something that graphic marks do-it is part of their way of accepting and declining determinate meaning

4) A mark might be undecidably a part rather than a whole or a composite rather than a single entity or even be equivalent to another mark These paradoxes that follow from the triad of lumen umbra and splendor also taunt visual semiotics by masquerading as the interplay of three disjunct signifiers Moniocosiuss theory is sensible and practical it is part of all studio instruction in painting and it is at work whenever a mark modulates so that it is ambiguously several marks or one When marks are swirled into washes or scumbled into larger areas or smudged

5 1 See David Gordon Karel The Teaching of Drawing at the French Royal Academy of Painting and Sculpture from 1760 to 1793 (PhD diss University of Chicago 1974)


into continuous gradations they lose their disjointness but not the idea of disjunction That play between hierarchical lumen umbra and splendor on the one hand and stepless change on the other is also part of the way marks carry meaning

5) And a mark might be something that must always be surrounded by a pale apparition hovering undecidably between drawn contour and mere edge Ideally the orlo the name of this ghost would never be securely visible Albertis negotiations with perspectiva brought him to this idea which remains fundamental in the concept of the figure The orlo is often present in figurative work and it is a reminder of the artificial nature of drawing (that is that edges are often drawn but not often seen) I like to think of the orlo as the ghost not only of visual rays and the geometry of vision but of written signs which all have determinate boundaries The orlo is as close as graphic marks get to the rigors of the written sign

These five modes may seem too colorful specific or intricate to be candidates for a serious description of graphic marks or a serious chal- lenge to semioticss basic assumption that visual elements are either disordered meaningless marks or proper signs But I think this is exactly the kind of extravagance and detail necessary to demonstrate and not merely claim that graphic marks are entirely different both from visual chaos and from the signs of writing or notation These apparently aber- rant overly complicated modes are the stable commonplace occurrences that compose visual images Like the splatters and moldspots in the Pon- tormo drawing or the goat made of cloudy milk droplets or the Jewish bride crawling with snaky lines or the figure of Jesus shivering uncon- trollably on the cross graphic marks are not disjoint orderly systematic linguistic signs But neither are they beyond the pale of analysis as if they were so immersed in the nonverbal world of artistic practice that history and theory could only reach them by making broad gestures in the direction of subsemiotic elements facture or handling And there are vir- tues to their vacillation because they are not consistently well behaved as signs they can occur once and never be repeated (they can respond to the uniqueness of pictures) but because they are also not entirely idiosyn- cratic they can sometimes be generalized into transcendental theories such as the one Derrida proposes (they can be approached from the side of logic and language)

There are two fundamental reasons why it makes sense to propose that graphic marks be understood as objects that are simultaneously signs and not signs First this way of looking at pictures demonstrates severe logical problems with a widely accepted paradigm of art history in that visual semiotics avoids subsemiotic marks or simplifies them into overall traits or structural oppositions while also presenting itself as a transdisciplinary account of the way that pictures signify To that extent semiotic art history is less a sign-theory than an attempt to use semiotics


to do other things-to continue or critique structuralism or to install an ideological or disciplinary critique of art history The second reason runs deeper By omitting marks or herding them into broad categories of surface gesture or handling art-historical accounts of all sorts make it possible to leap from the recalcitrant meaningless smears and blotches of a picture to the stories the picture seems to embody But once it is possible to be historically and analytically specific about graphic marks it becomes harder to justify that kind of omission and therefore harder to think and write about pictures It is that difficulty that I am after whether or not it results in a nonsemiotic position Both semiotics and semeiotics make pictures too easy semiotics by ceding the most fundamental and characteristic elements of pictures to the realm of the mere[ly] technical and semeiotics by assuming that marks are merely signs and that they could in theory be assimilated to linguistic or logical models Pic- tures are most interesting when neither alternative is possible so that a viewer is forced to attend to the ways that outlandish and partly incomprehensible marks both hinder and enable whatever story the picture seems to tell The incoherence of pictures begins here with the admission that things are very strange indeed

Marks Traces Traits Contours Orli and Splendores Nonsemiotic Elements in Pictures

James Elkins

Art history lacks a persuasive account of the nature of graphic marks and that limits what can be said about pictures If a sign as Charles Sanders Peirce said is something which stands to somebody for something in some respect or capacityn-a formula as vague as it is compact-then every mark in a picture is also a sign every brushstroke pencil line smudge and erasure must function as a sign and have meaning In practice that would spell trouble for accounts of pictures that take sign to mean the forms that are made out of the marks-such as in the typical examples figures scenes and narratives For the most part-and with important exceptions in all periods and subjects-art history has concen- trated on the larger-scale forms or on the large-scale properties held by groups of marks such as facture or handling That emphasis has left more exacting questions about graphic marks to the domain of practical criticism so that the people who are said to understand marks best are other artists and the most incisive critiques are taken to be other paintings instead of texts

Although it might seem that semiotic art history would address that inequality-perhaps by reinstating something like Peirces wider sense of signs-it is closely wedded to the distinction between meaningless marks and meaningful signs and for that reason semiotic accounts tend to gloss over marks in favor of the scenes they compose In Mieke Bals view the

1 Charles Sanders Peirce Elements of Logzc in The Collected Papers of Charles Sanders Pezrce ed Charles Hartshorne and Paul Weiss 6 vols (1931 Cambridge Mass 1974) 2135

CnItcal l n q u t ~21 (Summe1 1995)

O 1995 by The Unive~sity of Chicago 0093-189619512104-0003f0100 All rights reserved

































ington Ind 1990)









































































































Modes of Marking












































46 Ibid p 15
















1966)p 68




















































ington Ind 1990)









































































































Modes of Marking












































46 Ibid p 15
















1966)p 68




























































































































Modes of Marking












































46 Ibid p 15
















1966)p 68





























































46 Ibid p 15
















1966)p 68



































1966)p 68




















marks, traces, "traits," contours, "orli," and "splendores":...

Documents

joannes fredericus gronovius

de pictura veterurn

jacopo da pontormo

van de waal

carel van mander

leon battista alberti

linea summae tenuitatis

charles sanders peirce