Argumentation and Advocacy, Summer 1996 v33 n1 p1(10)
Toward a theory of visual argument. (Visual Argument - Part 1) David S. Birdsell; Leo Groarke.
Full Text: COPYRIGHT 1996 American Forensic Association
These special, two issues are motivated by the conviction that argumentation theorists do not pay enough attention to the visual components of argument and persuasion. A better understanding of these components is especially important if we want to understand the role of advertising, film, television, video, multi-media, and the World Wide Web in our lives. A decision to take the visual seriously has important implications for every strand of argumentation theory, for they all emphasize a verbal paradigm which sees arguments as collections of words. Most scholars who study argumentation theory are, therefore, preoccupied with methods of analyzing arguments which emphasize verbal elements and show little or no recognition of other possibilities, or even the relationship between words and other symbolic forms. Students of argumentation emerge without the tools needed for proficiency in assessing visual modes of reasoning and persuasion. We hope that these essays will help spur the development of a more adequate theory of argument which makes room for the visual.
Though we are committed to the development of a theory of visual argument, we have chosen to begin with an article in which David Fleming details his skepticism. Visual images ("pictures") cannot, he claims, be arguments. We have begun with his paper because we want to recognize that many theorists explicitly or implicitly reject this possibility (Fleming has provided a useful bibliography), and because an answer to their objections must be the basis of a convincing account of visual argument. The rest of our issue therefore answers these objections. J. Anthony Blair attempts to meet them in a defense of the possibility and the nature of visual arguments. Cameron Shelley and Gretchen Barbatsis (appearing in the fall issue) examine cases which illuminate different kinds of visual argument, and propose conceptual distinctions necessary for dealing with different kinds of visual materials. The review essay by Lenore Langsdorf discusses an important book on images and persuasion and reflects more generally on the questions raised by contemporary attempts to understand visual persuasion.
In the present introduction we would like to add some comments on those concerns that strike us as most important when one considers the development of a theory of visual argument. The first issue which must be addressed is a prevalent prejudice that visual images are in some intrinsic way arbitrary, vague and ambiguous. This presumption encourages the view that visual images are less precise than words, and especially the written word. We think that this prejudice is a dogma that has outlived its usefulness, and that the first step toward a theory of visual argument must be a better appreciation of both the possibility of visual meaning and the limits of verbal meaning.
Visual images can, of course, be vague and ambiguous. But this alone does not distinguish them from words and sentences, which can also be vague and ambiguous. The inherent indeterminacy of language is one of the principal problems that confront us when we try to understand natural language argument. This is why historians endlessly debate the interpretation of historical documents, law courts struggle continuously with the implications of written and spoken claims, and personal animosities revolve around who said what and what was meant. The point that visual images are frequently vague and indeterminate cannot, in view of the demonstrable indeterminacy of verbal expressions, show that images are intrinsically less precise than spoken or written words (especially as we often clarify the latter with visual cues-as we may make the tone and meaning of a statement clear with a smile or a wink).
We can best illustrate the possibility of verbal meaning with some simple examples. We will begin with the following anti-smoking poster, which was produced by the U.S. Department of Health, Education and Welfare (now the U.S. Department of Health and Human Services). We must begin by noting that this poster is an amalgam of the verbal and the visual . The important point is that this does not make its visual components redundant or superfluous. Without the visual elements we could not understand the poster, for the verbal message it contains - "don't you get hooked!" - is vague and ambiguous. It does not explicitly refer to smoking or cigarettes and could as easily refer to drugs, alcohol, or anything else which is potentially addictive. We know it is a message about smoking only because it depicts a fish which is "hooked" to a cigarette. The message of the poster is straightforward. It can plausibly be rendered as "You should be wary of cigarettes because you could get hooked and - like a fish on a lure - endanger your health." This is a quaint argument by analogy. It does not match the sophistication of the visuals which crowd our television sets - and increasingly, our computer screens - but it is an argument in the standard sense: it provides a reason for a conclusion.
This and countless similar examples make it difficult to sustain the kind of skepticism of those who maintain that the visual is radically indeterminate and cannot, therefore, sustain an argument. Consider Fleming's claim that a picture itself "makes no claim which can be contested, doubted, or otherwise improved upon by others. If I oppose the 'position' you articulate in a picture, you can simply deny that your picture ever articulated that, or any other, position." As common as such views are in academic discussions of the visual, they make little sense in the context of examples like the present one. Here the argument that you should be wary of cigarettes because they can hook you and endanger your health is forwarded by means of visual images, even though it is just the sort of claim that can be contested, doubted and improved upon. We too easily forget that there was a time when debates raged about the addictive qualities and the health effects of cigarettes. If someone viewing our sample poster did not "read" it as an attack on smoking (or arbitrarily denied that it "ever articulated that, or any other position"), then we are forced to the conclusion that they have radically misunderstood the visual image - to a point where we might reasonably wonder about their ability to comprehend the visual (much as we would wonder about someone's ability to understand English if they did not understand the corresponding verbal argument to be an attack on smoking).
Consider a second case which also illustrates the point that visual meaning can be in some cases neither arbitrary nor indeterminate. The following drawing is based on a 1926 editorial cartoon by S.K. Suvanto. The original cartoon was published in The Daily Worker, a socialist newspaper published in Chicago from 1924-1958 . Though we are far removed from the context which produced this cartoon, we still readily understand it, even if we ignore its title (the words in the title add nothing which is not obvious in the image itself). In the background we see the flag of the former Soviet Union - a hammer and sickle - and the silhouette of a Russian worker. The sky suggests dawn. The lattice of new buildings suggests the new industrial communist society. In the foreground we see a painter with an easel. His exaggerated obesity, his suit and his bald head are standard symbols of the capitalist. He is painting the scene in the background but what he paints bears scant resemblance to the "actual" image we see. In his canvas, the hammer and sickle in the flag - symbols of work - become a skull and cross bones. The hammer in the worker's hand becomes a bloody dagger which the worker - who has become a ruthless soldier - is plunging into a victim he grasps with his other hand.
Once again, meaning in our example is straightforward. We can discern a whole set of visual claims: Soviet communism is hard at work building a new industrial society; commentators who portray the Soviets as bent on violence and repression distort the facts; they themselves are greedy and self-interested capitalists. Taken together, these claims lead to the obvious conclusion that we should not listen to those who attack the new Soviet experiment.
Such examples leave little room for the presumption that visual meaning is necessarily arbitrary or indeterminate. The claims Suvanto makes in his cartoon are, moreover, just the sorts of claims which are open to debate, confirmation and argument. Someone who does not see his cartoon as an answer to criticisms of Soviet communism has radically misunderstood the point. Of course, one might debate specific points of interpretation (whether there is, for example, some significance in the fact that the left hand of the soldier in the painter's painting clutches his victim's throat) and one might fail to understand the visual vocabulary (a teenager might, for example, not understand the references to the Soviet Union or to capitalism). But these issues of interpretation are comparable to the issues that arise in the attempt to interpret verbal claims - the remarks of a political speaker, for example and cannot be used to show that visual claims are radically indeterminate.
What we have said about these two examples applies equally well to more sophisticated visual images. In the articles in this double issue, Blair shows how a Benetton ad can reasonably be deciphered, Barbatsis illustrates how a television camera can convey an argument and Shelley shows how drawings taken from articles on paleontology forward two different kinds of visual arguments. Fleming is right to point out that argumentation theory lacks a well developed account of the distinction between visual premises and conclusions, but this is because we have not taken seriously the possibility of visual meaning, not because visual images are - as so many commentators presume - necessarily indeterminate.
It does not follow that verbal and visual meanings are equivalent or identical. There are good reasons for questioning whether they have a similar capacity to convey relatively precise meanings. We merely observe that both can be ambiguous or cogent and that both can convey claims and arguments. The meaning of a visual claim or argument obviously depends on a complex set of relationships between a particular image/text and a given set of interpreters. The recognition that visual meaning is not necessarily arbitrary is the crucial first step that we must take in our development of a theory of visual argument.
The importance of context is the second issue that we feel must be addressed in developing such a theory. We do not expect words (at least not all words) to have solid, unassailable meanings of their own. Instead, we look to companion sentences and paragraphs to ascertain contextual meanings which may or may not be corroborated by dictionary definitions. The word "well," standing alone, could refer to my health, my skepticism, or the municipal water supply. If you read the sentence "I am well, thank you," then the context makes it clear that the first meaning is intended. Context plays a similar role when you hear someone ask me how I feel, in which case the single word "well" would be a terse but perfectly intelligible reply.
There is of course more to the process of assessing meaning and its context than examining words on a page or puzzling through sounds we hear. "Context" can involve a wide range of cultural assumptions, situational cues, time-sensitive information, and/or knowledge of a specific interlocutor. The immediate verbal context of a sentence is only one source of information interpreters use in determining the meaning of a string of words. Imagine that you overhear the following exchange:
Jonathan: Do you think the faculty will get a raise this year?
Maryann: Oh, sure. Now that we have a growing deficit, enormous new demands on our operating budget, flat revenues, and a government hostile to public education, I expect 15%!
In such circumstances, it is hard to imagine Jonathan concluding that Maryann actually means that a 15% raise is in the offing, or, more naively still, that Maryann has made a poor argument. Assuming minimal communicative competency on Mary ann's part, tone of voice alone will indicate her sarcasm. Assuming that there has never been a raise as significant as 15%, the contextually initiated will recognize that Maryann's response should not be taken at face value. The words alone do not convey these meanings, which are instead conveyed by the contextual cues.
Considered against the background of this familiar feature of verbal communication, there is no reason to assume that a visual image must conduct its contributions to argument in perfect isolation. Yet this assumption undergirds David Fleming's examination of visual argument and drives a good deal of the thinking that presupposes significant, inherent, and universal differences separating the verbal and the visual. We would never banish the consideration of contextual evidence when we consider verbal arguments, especially if we wish to understand their real-world efficacy. It would make no sense to take single words as units of argumentation unless they were clearly understandable as truncated references to more complete propositions. Why then would we assume that photographs should be examined in isolation from one another, or from verbal statements with which they are juxtaposed?
At least three kinds of context are important in the evaluation of visual arguments: immediate visual context, immediate verbal context, and visual culture. The significance of immediate visual context is most obvious in film, for it incorporates a progression of images which allows us to recognize a single frame as part of an overarching argument. Depending on the sequence of frames of which it is a part, an image of a man holding a knife may represent someone preparing to cook, a knife salesman or, more insidiously, evidence that someone is prepared to commit a murder. Sequences of images also play a role in other contexts. Instructional diagrams often use a progression of images to show viewers how to perform simple tasks. Cameron Shelley (in part two of this issue) shows that such diagrams can forward arguments.
Immediate visual contexts, however, encompass more than sequences of images. In judging such contexts we must often pay attention to visual cues beyond a single message source. Elements of the ambient visual environment can be equally influential in providing contextual cues to the interpretation of visual materials.
Immediate verbal context also provides a basis for the interpretation of visual images. A number of commentators (see Fleming in this issue) treat captions and other direct verbal references acting in concert with images as special cases, as indeed they are. It does not follow that the role of the image in a verbal-visual equation is unimportant, or secondary. Words can establish a context of meaning into which images can enter with a high degree of specificity while achieving a meaning different from the words alone. We see this in our first example, in which the words tell us that we are dealing with something which is addictive and harmful, while the visual image establishes that the topic is smoking.
Fleming explains another verbal-visual relationship in his remarks about visual evidence. But Fleming's formulation is limited by his emphasis on immediate verbal contexts which incorporate explicit claim/image interactions. The drawing based on the Suvanto cartoon invokes a much richer relationship between a larger and more general verbal context (communist narratives of the hostility of capital to the achievements of labor) and a specific visual rendition of the assertion that capitalists lie about communism. The implicit verbal backdrop that allows us to derive arguments from images is clearly different from the immediate context created by the placement of a caption beside an image.
When we incorporate conventionalized, situation-specific meanings within the process of interpreting visual arguments, we effectively extend the traditional verbal enthymeme. Suvanto's capitalist, for example, is a conventionalized image, easily recognizable as a type that could be invoked in a narrative description as readily as a visual depiction, particularly in the pages of The Daily Worker. But the imagistic recall is likely to be different from the verbal; we still need to be attentive to the way that a given image calls attention to the type. In this case, the drawing emphasizes physical characteristics, implicitly arguing against romantic images of capitalism - the beauty, glamor and power of Hollywood, for example - by emphasizing "undesirable" physical traits like corpulence, baldness and age. In other images, the dyslogies of depiction extend to demonizing qualities, such as the appearance of fangs and claws. In part two of this double issue, Shelley shows how subtle physical characteristics portrayed in a visual image can convey arguments about human evolution.
A third kind of context is supplied by visual culture, which differs from the first two categories principally in its indirect influence on the production of visual meaning. Many scholars have argued that visual culture changes significantly over time, and that developments in art, technology, philosophy, and science promote different ways of seeing over time. These scholars are for the most part careful to distinguish between the notions of "change" and "progress." They argue not that painting, or sculpture, or any other form of art has necessarily improved over time, but quite precisely that it differs, reflecting different values, conditions of production, and habits of interpretation. Cultural conventions of vision in this sense include what it means to see, or to represent seeing, as well as changes in the meaning of particular elements of visual vocabulary.
This is not the place for a comprehensive discussion of such complex ideas, but the basic concept can be illustrated readily by changes in television styles over the past thirty years. In the 1960's, television shots were considerably longer than those we find in the jumpy, quick edits typified by music videos in the 1990's. This change reflects something more than the difference between the evening news and MTV. Shot length has been reduced in almost all commercial television, and the number of shots per minute has surged. Much as cubism tried to present multiple perspectives unfolding over time and/or space on a single, two-dimensional frame, the quick-cut video editing style of the 1990's prefers several quick perspectives on a subject over the single, probing, shot that holds an image for minutes at a time. The result is a combination of visuals that decenters a unitary perspectivalism. No one camera is all-knowing and the subject is deliberately distorted with the use of negative effects or other filters that "reveal" different elements of the subject-as-source for videographic play.
Visual culture provides the broad master narratives of design which are the background for more specific visual (or for that matter, verbal) texts which perpetuate or challenge those narratives. Martin Jay's work identifies "scopic regimes" peculiar to historical periods. Students of argumentation have accepted since Aristotle the influence of acculturation in the production of verbal enthymemes. We are now arguing that the same allowances must be made for visual commonplaces as well, allowing potential visual arguments to draw on the same range of resources that we afford potential verbal arguments.
The changes in visual meaning made plain in studies of visual culture suggest a third issue which must be the basis of a satisfactory theory of visual argument. It concerns the meaning of "resemblance." In his article, David Fleming restricts his analysis to images that are created in an effort to resemble what they represent. We do not dispute the existence of a category of imagery that purports to represent reality, but we want both to problematize the notion and note that argumentation plays a key role in determining resemblance and representation (which constitute another way in which visuals are linked to argument). At issue here is a complex set of relationships having to do with representation and resemblance per se. The topic is too large to address thoroughly in this introduction, but three of its elements bear mention: the disjunction between resemblance and representation, the consequent conventionalization of representation, and the susceptibility of resemblance to visual and verbal challenge.
While most observers would say that a well-executed "realistic" portrait resembles the sitter, it may or may not adequately represent the sitter. If I sit for a portrait wearing a gorilla suit, a realistic painting, even a photograph, will resemble me (sitting in front of the artist, in a gorilla suit). But does it represent me? A caricaturist's line drawing (consider the famous profile of Alfred Hitchcock that became the lead-in to his television series) that cannot be said to resemble anyone in any detailed way may serve as a good representation of a sitter. Such examples show that while representation is a more ambiguous concept than resemblance, resemblance is itself fraught with judgment. What, exactly, should a successful visual image of a sitter "resemble?" Should it be the sitter's present attitude, the sitter's most common expression, a characteristic gesture?
These difficult questions posed by resemblance and representation have encouraged a wide reliance on conventionalized representations that are easily used in arguments. Heraldry is a conventionalized representation of a family. King Francis I of France was represented by the salamander, though he could hardly be said to have resembled one. In the sixteenth century, the Visconti family was visually represented by a serpent eating a child. While there is no "photographic" resemblance one might say that this demonstration of raw power represents (or metaphorically "resembles") the family's own. Likewise, the President of the United States is represented by his seal, which does not "resemble" him. In fact, because the seal's eagle motif is highly abstracted, appearing in a posture that no "real" eagle could attain in life, it is debatable whether the symbol even resembles a real eagle.
The shifting standards applied to resemblances make them subject to challenge on two argumentative levels. First, they may not in fact resemble (anyone who has argued with a photographer over the quality of a graduation or a wedding picture will have no trouble coming up with cases), and second, they may not represent. The kinds of arguments this implies can be conducted either visually or verbally. In this double issue they are reflected in Barbatsis' analysis of visual images in advertisements aired during the 1988 presidential campaign. One of the principal visual techniques she identifies is the deconstruction of an apparent resemblance in favor of an allegedly more accurate representation in political advertising. The point is not that the preferred alternative is or is not "genuinely" more accurate, but that through the application of visual techniques rather than verbal narrative, the question of resemblance has entered directly into the argument.
So far, we have suggested three prerequisites for a satisfactory account of visual argument: we must accept the possibility of visual meaning, we must make more of an effort to consider images in context, and we must recognize the argumentative aspects of representation and resemblance. We want to finish by more tentatively noting another issue raised by the attempt to formulate a theory of visual argument. Blair raises the issue in his article when he offers an account of visual argument which places significant limits on the visuals we can classify as arguments. In part, these limits are imposed by his distinction between argument and persuasion, suggesting that many of the visuals one might consider arguments are instances of persuasion rather than argument. Intuitively, there is something to his suggestion that such visual presentations are attempts to convince in a way that purposely circumvents argumentation and the reflection it implies. Considered from this point of view, the attempt to convince a dieter to eat a piece of cake by holding it under his or her nose is not, it seems, an argument.
Or is it? Why not take the holding of the cake in front of the dieter's nose to be a particularly forceful way of expressing the argument that "Eating this cake would be wonderful, therefore you should forget your diet and eat it"? So construed this is an argument. One might compare the ancient story that Diogenes the Cynic is said to have responded to Zeno's famous arguments against motion by walking a few steps and declaring "I refute Zeno thus." Surely this is an argument. But it is also an attempt to circumvent the reasoning and the reflection that accompanies Zeno's paradoxes.
Forbes I. Hill (1983) locates, in Aristotle's Rhetoric, support for the notion that visual appeals to desire influence our actions. As Hill puts it, "Aristotle's view of the pathe [feelings] is extremely intellectualized. To come into a state of feeling an auditor must make a complex judgment about himself in relation to external events. If he is incapable of making this judgment, he will not come into the state of feeling" (p. 47). Such a view collapses the distinction between "psychological and logical proof," by making appeals to feelings appeals to certain kinds of judgments. From this point of view, we "argue a person into a state of feeling."
At the very least it must be said that this way of extending the theory of visual argument has some intriguing consequences that are worth exploring. Most importantly, it allows for a significant expansion of the theory of argument. Without this expansion, argumentation theory has no way of dealing with a great many visual ploys that play a significant role in our argumentative lives - even though they can frequently be assessed from the point of view of argumentative criteria. Aristotle's notions of logos, ethos, and pathos can, for example, frequently be used to shed light on such circumstances, even when we have something that falls short of what we would normally count as a fully fledged argument. It is in view of this that the standard distinction between argument and persuasion needs to be reconsidered in the realm of visual argument.
Any account of visual argumentation must identify how we can a) identify the internal elements of a visual image, understand the contexts in which images are interpreted, c) establish the consistency of an interpretation of the visual, and d) chart changes in visual perspectives over time. These issues have been explored at length - albeit without a full appreciation of their relevance to argumentation studies - in the fields of art history, cognitive psychology, media studies, semiotics, and visual culture. The rich diversity of perspectives discussed in Langsdorf's review in this issue provides a good starting point for fruitful explorations of the literature this implies, but students of argument interested in the visual should obviously go beyond the single collection she discusses if they wish to engage the burgeoning scholarship in visual theory.
In the highly selective annotated bibliography that follows, we have chosen a few titles that speak very clearly to concerns that parallel the sort of broad understanding any argumentation scholar would want to bring to the examination of a verbal enthymeme. It is a literature to which we may reasonably expect to contribute ourselves. Missing from much of the analysis of visual imagery is the careful consideration of argumentation evidenced in the close readings of cases provided by Blair, Shelley, and Barbatsis. Though arriving at different conclusions about the project of visual argumentation, David Fleming exhibits much the same kind of concern by insisting that we actually find elements of something recognizable as argument before proceeding to an "argumentative" analysis of a picture, a condition that we feel the other three authors to have amply demonstrated in their essays.
Our contributions to understandings of the visual will come from our ability to flesh out theories of visual argumentation as rich and as rigorous as those we have developed for verbal argumentation. In the process of developing a theory of visual argument, we will have to emphasize the frequent lucidity of visual meaning, the importance of visual context, the argumentative complexities raised by the notions of representation and resemblance, and the questions visual persuasion poses for the standard distinction between argument and persuasion. Coupled with respect for existing interdisciplinary literature on the visual, such an emphasis promises a much better account of verbal and visual argument which can better understand the complexities of both visual images and ordinary argument as they are so often intertwined in our increasingly visual media.
SELECTED ANNOTATED BIBLIOGRAPHY
Brennan, T. and Jay, M. (Eds.). (1996). Vision in context: Historical and contemporary perspectives on sight. New York: Routledge. A highly eclectic collection of essays on vision from scholars in a wide range of disciplines. Argumentation scholars will find particularly useful the first five essays on the changed and changing roles vision at different points in history. Later essays address issues relating to vision and gender, vision and subjectivity, and visual studies and interpretation.
Bryson, N., Holly, M. and Moxey, K. (1991). Visual theory: Painting and interpretation. New York: Harper Collins. The articles in this collection range from elements of semiology to situated seeing. Every article is followed by at least one commentary, making this volume a particularly rich exploration of the issues raised. Argumentation scholars will want to pay close attention to the authors' treatment of arguments about visual materials and visual theory.
Foster, H. (1988). Vision and visuality. Seattle: Bay Press. This collection, sponsored by the Dia Art Foundation, is a brief (135 pp.) introduction to several important themes in the study of visual culture. Particularly valuable are Martin Jay's essay "Scopic Regimes of Modernity," and Norman Bryson's discussion of interpretive subjectivity in "The Gaze in the Expanded Field."
Gombrich, E. H. (1989). Art and illusion: A study in the psychology of pictorial representation. Princeton: Princeton UP. This classic work, originally published in 1960, sets out to explore the relationships among culture, perception, and forms of artistic production. Written before the development of most visual theories based on postmodernism or electronic media, Art and Illusion is a useful starting point for those who find the latter perspectives uncongenial.
Horace, B., Blakemore, C., and Weston-Smith, M. (Eds.); (1990). Images and understanding. Cambridge: Cambridge UP. This collection of essays, based on the Rank Prize Funds' International Symposium in October 1986, emphasizes cognitive approaches to visual understanding in the context of commentary from art historians and theorists.
Jay, M. (1993). Downcast eyes: The denigration of vision in 20th century French thought. Berkeley: U of California P. Here Jay explores vision as a cultural product, and particularly as the product of intellectual/artistic culture. The analyses of Bataille's and Lacan's contributions to perspectives on vision are invaluable.
Jenks, C. (Ed). (1995); Visual culture. London: Routledge. This volume contains essays on visual culture from a British cultural studies perspective. Readers interested in visual issues in electronic media will find this study particularly useful.
Melville, S. and Readings, B. (Eds.). (1995). Vision and textuality. Durham: Duke UP. This volume collects essays oriented around questions of disciplinarity in visual studies. Argumentation scholars hoping to understand academic institutionalization of vision will find several essays useful. Helpful as well is a repeated emphasis on verbal/visual issues.
Mitchell, W. J. T. (1986). Iconology: Image, text, ideology. Chicago: U of Chicago P. This is an enormously influential study of the shifting fortunes of visually and verbally based systems of meaning in western culture. Mitchell provides a lucid explanation of the stakes in preferring the visual the verbal and vice versa. His is also the best single-volume exploration of the broad sweep of intellectual history on these issues.
Panofsky, E. (1995). Meaning in the visual arts. Garden City: Doubleday Anchor. This is in part Panofsky's most accessible text on the nature of pre-iconographic, iconographic, and iconological analysis. These distinctions are very helpful for anyone attempting to "read" an image.
Sherman, C. R. (1995). Imaging Aristotle: Verbal and visual representation in fourteenth century France. Berkeley: U of California P. This richly illustrated volume examines the use of manuscript illumination in the first French translations of Aristotle. Sherman shows how the illuminations themselves conveyed important arguments about state power generally and Charles V in particular. The fact that the analysis focuses on the works of Aristotle makes this book uniquely accessible to argumentation scholars.