Composition as Conditioning: Collective Aesthetics and the Mere Exposure Effect
Alexandra Kleeman (UC, Berkeley)
Of course it is beautiful but first all beauty in it is denied and then all the beauty of it is accepted…Automatically with the acceptance of the time sense comes the recognition of the beauty and once the beauty is accepted the beauty never fails any one.
— Gertrude Stein, “Composition as Explanation”
As an interdependent and fundamentally interactive dynamic system, literature is not merely reflective of the social but is in fact a force in its creation. The literary canon is both the concrete product and the active producer of collective aesthetics: each generation makes changes to the canon of their time by adding new works, removing works that have come to be considered irrelevant or inappropriate, and reassessing the canon in relation to new “contingencies of value.” These alterations reverberate in the reading practices of subsequent generations, shaping their aesthetic criteria and affecting their own modifications to the corpus of the literary tradition.
Myriad different mechanisms of canon alteration can be identified within literary history, including the recuperation of “forgotten” works by academic scholars and the redemption of “unappreciated” works by advocacy on the part of artists and critics. The inclusion of Petronius’ Satyricon and the dramatic and fictional works of Samuel Beckett within our present-day literary canon can be traced to the efforts of individual artists who raised the cultural value of these texts by alluding to them or actively promoting their consumption among members of their own audience (Kermode 2004: 77). But while these individual-driven narratives of canon changemake forcompelling literary histories of individual texts, large-scale alterations in the canon over an extended timeline arebetter describedby collective sociohistorical dynamics than by anecdote. The influence of a collective aesthetic may be visible in the isolated text, but is more often grounded in a subjective readerlyaffect that escapes conscious analysis. Modified by factors both external to the text, such as publication frequency, and internal (repetition, style, grammatical structure) the reader becomes a locus for the convergence of repetition and resonance, a site for the synthesis of the text’s formal properties and material context.
New methodologies of literary historiography have been developed by theorists such as Franco Moretti, who argues that “a field this large cannot be understood by stitching together separate bits of knowledge about individual cases, because it isn’t a sum of individual cases: it’s a collective system, that should be grasped as such, as a whole” (Moretti, 4). In the first chapter of his book Graphs, Maps, Trees, Moretti performs an extensive analysis of the annual output of British novels over a two-hundred year span, sorted according to forty-four major novelistic genres (e.g. the Utopian novel, the Epistolary novel, the Courtship novel). Moretti finds that new novelistic genres tend to emerge and disappear in clusters, with five or six new types of novels emerging at roughly the same time, remaining static for about twenty-five years, and then being replaced by five or six new forms simultaneously at roughly the end of that time span. Lacking a clear explanation for this phenomenon, Moretti suggests that these destabilizations could correspond, roughly, to periods of generational stability—to shared mental climate and cultural values (perhaps among the middle-aged economically and artistically powerful members of society) (Moretti, 21).
This unexpected finding, derived from the extended temporal analysis of the rise and fall of genres, indicates points to a larger influence for collective social influences on literary production and reception than has traditionally been supposed. Though the rhetoric of canonization prefers to portray literary merit as a concrete property discovered within particular texts, new approaches argue that literary merit is a construct of social and psychological processes that often escape conscious awareness. Evaluative analysis is oblivious to its own blind spots, unable to recognize and identify influences that operate on either a larger-scale collective level, or at the level of subconscious, automatic psychological mechanisms. Compensating for these blind spots calls for an interdisciplinary approach combining the methodology of literary studies with that of sociology and psychology, in an attempt to discover uncover general social or psychological effects dynamics that could help to explain phenomena such as the acceptance of avant-garde literary texts into the canon.
Gertrude Stein’s essay “Composition as Explanation” describes collective influences upon aesthetic evaluation, performing an overt analysis of temporal and social factors contributing to the perception of beauty in a work of art even as it reenacts these processes in the mind of the reader. Her text speaks of the effect of familiarity on the perceived beauty of an object and of the process of canonization, wherein a work long perceived as unappealing suddenly achieves a beauty in the eyes of its audience that, once achieved, cannot be easily undone. Stein notes that public approval of a work of art is closely linked to temporal factors, stating that “for a very long time almost everybody refuses and then almost without a pause everybody accepts” (Stein, 1926). At the same time, Stein demonstrates for the reader how the acquisition of this acceptance occurs, using formal devices that initially obstruct the reader’s access to meaning, but become comprehensible and even beautiful as the act of reading progresses. In this respect, Stein’s experimental writing is convergent with empirical investigations of the Mere Exposure Effect (MEE), a psychological phenomenon linked to implicit learning and conditioning, in which simple exposure to symbols and structures enhances their attractiveness on subsequent viewings. A close reading of Stein’s text in conjunction with recent experimental work on the relationship between mere exposure and the artistic canon can help to explain why new forms of art take time to be welcomed into the preexisting canon, how cultural factors affect our experience of beauty, and how basic psychological processes involved in ordinary aspects of human cognition influence our aesthetic evaluation of art.
Mere Exposure and Affective Judgment
Research in the cognitive sciences has been markedly reluctant to address the aesthetic and affective dimensions of language, focusing instead on quantifiable aspects of specific task domains, such as lexical activation and syntactic processing. Psychological investigations of aesthetics are hindered by the view that aesthetic impulses are peripheral to cognition proper, as well as by the notion that art is created through individual acts of irrational inspiration unrelated to everyday behavior. Researchers dealing with aesthetic effects also encounter the difficulty of replicating a complex naturalistic activity such as the appreciation of art within a controlled laboratory setting. As a result, the scientific research on aesthetics is underdeveloped, centered mostly on the demonstration of empirical differences in the brain activation of individuals looking at different types of artistic stimuli, or comparing the behavior of experts and nonexperts in passive viewing of artistic objects (Zeki 1994, Ramachandran 2005). Such studies legitimize their investigations by linking their efforts to well-established branches of cognitive science such as the visual perception of form and motion, leaving the phenomenal experience of being positively or negatively affected by aesthetic stimuli more or less untouched. The mere exposure effect is a rare instance of an empirically-researched psychological effect that deals with the temporally-extended relationship For this reason, I focus on one of the few lines of research that is both centrally located with respect to the issue of aesthetic judgment and well-substantiated by a long history of experimental investigation: the mere exposure effect.
The “mere exposure effect” refers to a situation in which multiple exposures to a stimuli increases subject preference for that particular stimuli, and was first discovered during a series of experiments conducted by Zajonc in 1968. Using Chinese ideograms and nonsense words as stimuli, Zajonc showed that subjects’ ratings for “goodness of meaning” are positively correlated with the frequency with which they have been exposed to the stimuli, demonstrating a strong link between passive exposure and affect. When presented with a series of nonsense words (dilikli, for example) at frequencies ranging from 0 to 25 exposures, subjects rated the stimuli in positive correlation with how many times they had been seen. In other words, merely viewing an item under a condition that “just makes the given stimulus accessible to the individual’s perception” increased its positive affect in subsequent rating tasks (Zajonc, 1968).
In the two decades that followed Zajonc’s initial discovery, over two hundred subsequent studies of mere exposure effects were conducted, examining the relationship upon effect size of parameters such as stimulus recognition, duration of exposure, stimulus complexity, number of exposures, and delay between exposures. Additional studies dealt with individual and group differences in performance on mere exposure paradigms, exploring the performance of different age groups (children and adults) on mere exposure tasks, as well as the effect of factors such as attitude and personality type on task performance (Bornstein, 1988). Recent experiments have expanded the scope of the project to include neuroimaging research (Elliott and Dean, 1998), and studies relating affect to aspects of processing dynamics such as perceptual fluency (Winkielstein and Caccioppo, 2001).
Robust effects of mere exposure upon affective evaluation have been found in nearly every study of the relation between exposure and affect, evidencing its widespread influence across species, genders, cultures, and age groups. The effect has been shown at very low exposure rates, appearing after even a single pre-exposure (Monahan, Murphy, and Zajonc, 2001), and has been discovered across experiments using a wide variety of stimulus types, from geometric shapes to musical phrases and abstract art. Structural mere exposure effects showing increased preference for previously-exposed structures containing novel content have been demonstrated in grammaticality judgments (Luka and Barsalou, 2005) and in artificial grammar learning paradigms, under conditions in which the subject is already familiar with the structural elements of the grammar (Zizak and Reber, 2004). Cross-cultural replications of the mere exposure effect have demonstrated that the effect can be found in cultural environments very different from those of American and European societies (Ishii 2005) and cross-species demonstrations of mere exposure effects in rats and prenatal chicks habituated to a specific tone hint at the evolutionary significance and survival value of preference for the familiar. The basic structures of most mere exposure experiments are similar enough that to spend much time reviewing the structure of individual experiments would be redundant—most vary only in terms of stimuli used and presentation of the experimental stimuli (although, as we will see later, these parameters often have a significant effect upon the duration and shape of the affect-exposure curve). The demonstration of mere exposure effects across these diverse stimuli and participant categories argue that mere exposure effects are highly influential in our assessment of the affective value of stimuli.
The effect of mere exposure upon affective judgments of neutral stimuli has startling implications for the nature of the artistic canon: if exposure does, in fact, play a significant role in the evaluation of art within a real-world setting, then artistic canons may form as the result of a type of feedback loop between society and individual. An individual encountering a particular artwork or style of art multiple times will begin to find that art more appealing, and will therefore help to increase its prevalence within society by placing reproductions of it within his home, increasing its appearance in publication, or attending museums that display the same work of art, or works of art similar in style. The efforts of the individual increase the frequency of the artwork in society, and cause an increase in its frequency of exposure for other members of society—which, in turn, increases its appeal to other individuals who begin to encounter the artwork more often. Thus, mere exposure could lead to a rapid cyclical increase in the perceived value of a work of art, establishing widely-supported canons that are based in large part upon external social factors rather than factors intrinsic to the work of art itself.
A recent experiment by James Cutting (2003) has tested these claims, asking whether the laboratory phenomenon of mere exposure can truly be generalized to a process as complex as the formation and maintenance of an artistic canon. Cutting’s experiment explores the effect of mere exposure through an ecologically-inspired design that uses actual Impressionist paintings to investigate the effect of publication frequency and mere exposure on aesthetic evaluation. Cutting selected sixty-six paintings from the collection of Gustave Caillebotte, a collector of Impressionist art who was widely believed to have impeccable taste. Sixty-six non-Caillebotte paintings were selected as controls and matched to the Caillebotte stimuli in terms of colorization, subject matter, and publication frequency—determined through a careful count of each painting’s frequency of appearance in Cornell University library collections. Subjects in three different age groups then viewed sets of slides, displayed for eight seconds, featuring one Caillebotte and one control painting, before stating which painting they preferred.
Statistical analyses of preference data showed no preference and no increase in recognizability for images from the Caillebotte collection, and instead showed a reliable correlation between publication frequency and preference and no main effect for complexity of the paintings themselves. Although subjects did show a slightly higher preference for images displayed in the Musée D’Orsay (54% versus 46%), this preference disappeared once the effects of frequency of publication for the D’Orsay’s higher-profile artworks were factored out, indicating that preference for the Musée D’Orsay’s collection is an effect of mere exposure rather than inherent quality in the selection of the art on display. While college-age subjects and older subjects showed similarly-sized effects of frequency on preference, data from child subjects showed no such effect: children’s preferences were correlated with bright color rather than frequency or recognition of the artworks. Cutting’s experimental data effectively demonstrates that preference for complex stimuli such as paintings is linked to the prevalence of artworks in society, and to mere exposure effects arising from their prevalence.
Because this correlation between frequency and preference could also be explained as an effect of the inherent quality of a painting, with individual viewers and publishers alike agreeing upon the objective merit of the artwork, Cutting followed up these experiments with another using the same paintings as experimental stimuli. In this follow-up investigation, students taking a perception course were shown PowerPoint slides of Impressionist paintings throughout the course of a semester. Low-frequency images were shown four times over the semester, while high-frequency images were shown only once. At the end of the course, students were shown pairs of slides corresponding to the stimuli used in preceding experiments, and given a questionnaire asking which painting they preferred. Subject data showed no correlation between preference and frequency, with a mere 48% of subjects preferring the more common image (compared to 57% in the earlier experiments). Elimination of the correlation between high publication frequency and individual preference through passive exposure is evidence of the powerful influence of mere exposure on aesthetic judgment. In addition, this experiment provides evidence of the robustness of the mere exposure effect: even at interexposure intervals of 10 to 80 days, the effect of four two-second long exposures was great enough to negate the sum total of subjects’ exposure to high-frequency artworks.
Cutting’s use of Impressionist art as experimental stimuli is especially interesting in the context of the history of the Impressionist movement and the incorporation of fringe art into the artistic mainstream. The Impressionist movement in painting provides a classic example of an artwork or artistic movement’s transformation from refusal to acceptance: at its inception, Impressionism was generally considered aesthetically offensive to realist sensibilities, and it was only after several years of gradual exposure to Impressionist work that the movement was almost unanimously embraced. When Gustave Caillebotte died, his entire collection of Impressionist artworks was offered to the Musée D’Orsay, which reluctantly took possession of approximately 38 of the 70 or so paintings offered (many of which were thought to be “unsellable,” and were bought by Caillebotte in order to support painters that he had befriended). Over time, the Caillebotte collection housed at the Musée D’Orsay gained notoriety, and is now thought to reflect Caillebotte’s superb taste in art. What Cutting’s experiments indicate, however, is that their placement in a prominent art museum contributes as much to their prominent placement in the artistic canon as any internal quality of the art. The process of canon formation is thus revealed to be a self-perpetuating, self-organizing dynamic system, heavily influenced by basic psychological processes linked to feedback and implicit learning.
Mere exposure as explanation
A review of the psychological literature on mere exposure, and on Cutting’s (2003) experiment in particular, shows that mere exposure is related to the formation and maintenance of an artistic canon. Mere exposure provides a possible explanation for why “everybody refuses and then almost without a pause everybody accepts” (Stein, 1926)—collective preference for art, on a macro scale, is influenced by cultural factors such as prevalence and preexisting position in the canon. In this way, the psychological tendencies to prefer pre-encountered stimuli functions to maintain pre-existing valuations, or even to further their entrenchment in the aesthetic economy. At the same time, mere exposure as a form of implicit learning can subvert existing structures of aesthetic valuation, given a reader engaged with a text that actively works within itself to re-expose the reader to its novel formal qualities. A reader encountering Gertrude Stein’s “Composition as Explanation” for the first time is likely to find himself slowed by its difficult syntactic form and recursive, reiterative phrasing. As the reading progresses, however, repeated engagement with Steinian syntax models modes of re-exposure that characterize more canonical literary works, and thus reenacts its own progression from refusal to acceptance.
Stein begins her essay with the difficult phrase “there is singularly nothing that makes a difference a difference in beginning and in the middle and in ending except that each generation has something different at which they are all looking.” This phrase situates the reader’s initial engagement with the text in a vertiginous space, marking a discontinuity between the style of the essay and those formal qualities that characterize the pre-existing conventions of composition. Barbara Herrnstein Smith observes that aesthetic judgments never engage “raw” objects: “Not only are the objects we encounter always to some extent pre-interpreted and pre-classified for us by our particular cultures and languages; they are also pre-evaluated, bearing the marks and signs of their prior valuing and evaluations by our fellow creatures. Indeed, pre-classification is itself a form of pre-evaluation, for the labels or category names under which we encounter objects not only, as I suggested earlier, foreground certain of their possible functions, but also operate as signs—in effect, as culturally certified endorsements—of their more or less effective performance of those functions” (Smith, 43). Stein presents a grammatical object foreign to the pre-exposed patterns of conventional language use, an object that does not participate in an existing exposural feedback loop. The formation of a new circuit begins at a distance from preexisting feedback-loops of exposure and preference, grounded in a language whose classicality and comprehensibility is as yet unassimilated.
This language is vertiginous in the extreme, using terminology that is at times childlike, at others borrowed from academic discourse. Commonplace nouns possessing a generality that makes them at once abstract and highly concrete play key roles in many sentences: “So then naturally it was natural that one thing an enormously long thing was not everything an enormously short thing was also not everything nor was it all of it a continuous present thing nor was it always and always beginning again” (Stein, 1926). These “things” and “everythings” link recursively back to pronouns given earlier in the sentence or in earlier sentences, pronouns with referents that appear to shift as the process of reading continues and more of the composition is revealed. Even as the reference of pronouns and general terms appears unfixed, relatively precise terms are rendered unstable by their placement within the same highly recursive, repetitious grammatical structures. “And so there was the natural phenomena that was war, which had been, before war came, several generations behind the contemporary composition, because it became war and so completely needed to be contemporary became completely contemporary and so created the completed recognition of the contemporary composition.” “Contemporary,” a word that generally possesses localizable meaning, comes unmoored within the context of a Steinian syntax. Modified in reference by each new instantiation, “contemporary” becomes multiple, necessitating a distinctly different style of reading that defers evaluation indefinitely.
Stein’s distinctive syntactic structures scaffold a language that is unfamiliar, and even unpleasant to read. The lengthy sentences found in “Composition as Explanation” may contain multiple restructurings of subject-object relations as well as conversational interjections and the explicit contradiction of statements made earlier in the sentence. Highly repetitive syntax and word use make it possible for words to function as subject, objective, or modifier, while the length of the sentences in which they are embedded increases the difficulty of holding onto any particular semantic orientation. As consistency is visible even in the differences among and between sentences, the reader repeatedly encounters similarly convoluted patterns, a sequence of stimuli united by their distributed rather than localized syntax: “And so now one finds oneself interesting oneself in an equilibration, that of course means words as well as things and distribution as well as between themselves between the words and themselves and the things and themselves, a distribution as distribution. This makes what follows what follows and now there is every reason why there should be an arrangement made. Distribution is interesting and equilibration is interesting when a continuous present and a beginning again and again and using everything and everything alike and everything naturally simply different has been done” (Stein, 1926). Although the principles of Steinian syntax cannot be abstracted from their particular instantiation and are inextricable from the context in which they operate, paragraphs like this show a pattern of rhythm, recursion, and reassembly that give the sentences consistency, and therefore have the effect of re-exposing the reader to a grammar that was originally unfamiliar and disruptive.
It is through this internal mechanism of exposure and re-exposure that “Composition as Explanation” performs its own progression from refusal to acceptance.
Stein’s distinctive style, juxtaposing highly repetitious syntax and lengthy sentences with a curious semantic opacity, is present throughout the essay and feels increasingly natural as reading progresses. By the essay’s conclusion, her style of writing no longer alarms: the reader has reached his or her own acceptance of this new form of art, replicating in miniature the progression of art from refusal to acceptance. Just as experimental subjects learning begin to find “ungrammatical” sentences more acceptable and appreciably after multiple exposures, the reader has gained the ability to read what was previously illegible. Formerly opaque sentences become accessible to analysis and appreciation, joining a collective composition comprised of culturally legible units. Ultimately, Stein’s text demonstrates that “that beauty is beauty even when it is irritating and stimulating, not only when it is accepted and classic” (Stein, 1926).
Stein’s essay thus has multiple affinities with mere exposure experiments, especially those concerned with exposure effects for grammatical structures and patterns. By repeatedly exposing the reader to syntactic formations that reflect Stein’s writing style and approach to narration, Stein creates a mere exposure effect in the mind of her readers, making her unconventional style of syntax more fluent, more appealing. As the reader monitors his own reactions to the essay throughout the course of reading, he notices that sentences no longer feel as “difficult,” and that he is able to spend more resources attending to the content of the sentences, or to other aesthetic factors related to their form, such as rhythm. As a result, the same structures begin to reveal a previously unacknowledged or inaccessible beauty. Whether this effect is related to perceptual fluency created in the implicit learning of Stein’s grammar, or to a conditioned positive response to stimuli that have been shown to have no negative effects, the end result is the same: the inherent beauty of a refused object is exposed, its refusal turns to acceptance. Stein’s model of literary experimentation takes the text as a mechanism for uncovering the psychology of the reader, even as it reconstitutes the reader as a differently configured mind. The performative effects of literary experiments thus parallel those of empirical investigation.
Gertrude Stein’s “Composition as Explanation” thus provides an example of how mere exposure and cultural prevalence affects the psychology of individuals, while interrogating its effects upon the formation of an artistic canon. In its modification of the affective response of individuals during the process of reading, we see one possible mechanism for moving a work from refusal to acceptance. At the same time, these reading practices and formal textual structures are embedded in a larger social context, in a linguistic environment that works to influence and configure our own individual preferences. Many studies have shown that overexposure to stimuli result in a reversal of the mere exposure effect in which overexposed stimuli become more dislikeable with each additional exposure: it is possible that new movements rise to prominence when the older movements have begun their affective decline. As a society, we may replicate the affect-exposure curves of individuals, leading to a distributed collective preference for something new that allows new forms of art to take hold.
In conjunction with empirical research concerning the exposure-affect curve in mere exposure experiments, Stein’s essay offers insight into the ways in which individual psychologies and social structures are mutually interpenetrating, into the collectivity that inhabits the individual reader and the subjective affects that embody collective judgments about matters of aesthetic and social value. If the composition of our time is, as Stein says, “the thing seen by every one living in the living they are doing…at the time they are living in the composition of the time in which they are living,” then to see the mutual influence of individual and collective in the literary text is to see a new way in which the present might be composed.
Works Cited
Berlyne, D.E. (1970) Novelty, complexity, and hedonic value. Perception and Psychophysics, 8, 279-286.
Bornstein, R. F. (1988) Exposure and affect: Overview and meta-analysis of research, 1968-1987. Psychological Bulletin, 106(2), 265-289.
Cutting, J. (2003) Gustave Caillebotte, French Impressionism, and mere exposure. Psychonomic Bulletin and Review, 10(2), 319-343.
Cutting, J. (in press) “Mere exposure, reproduction, and the impressionist canon.” In: A. Bryzyski, ed. Partisan Canons, Durham: Duke University Press.
Elliott, R. and R.J. Dolan (1998) Neural response during preference and memory judgments for subliminally presented stimuli: a functional neuroimaging study. The Journal of Neuroscience, 18(12), 4697-4704.
Ishii, K. (2005) Does mere exposure enhance positive evaluation, independent of stimulus recognition? A replication study in Japan and the USA. Japanese Psychological Research, 47(4), 280-285.
Kermode, F. (2004) Pleasure and Change: The Aesthetics of Canon. Oxford: Oxford UP.
Luka, B.J. and L.W. Barsalou (2005) Structural facilitation: Mere exposure effects for grammatical acceptability as evidence for syntactic priming in comprehension. Journal of Memory and Language, 52, 436-459.
Monahan, J.L., S.T. Murphy, and R.B. Zajonc (2000) Subliminal mere exposure: specific, general, and diffuse effects. Psychological Science, 11(6), 462-466.
Moretti, F. (2005) Graphs, Maps, Trees: Abstract Models for a Literary History. London: Verso.
Ramachandran, V.S. and W. Hirstein (1999) The science of art: a neurological theory of aesthetic experience. Journal of Consciousness Studies, 6, 15-51.
Smith, Barbara Herrnstein (1988) Contingencies of Value: Alternative Perspectives for Critical Theory. Cambridge: Harvard University Press.
Stein, G. (1926) “Composition as Explanation.” In: What are Masterpieces, London: Hogarth Press.
Winkielman, P. and J.T. Cacioppo (2001) Mind at ease puts a smile on the face: psychophysiological evidence that processing facilitation elicits positive affect. Journal of Personality and Social Psychology, 81(6), 989-1000.
Zajonc, R.B. (1968) Attitudinal effects of mere exposure. Journal of Personality and Social Psychology Monographs, 9(2), 1-27.
Zajonc, R.B. (2001) Mere exposure: a gateway to the subliminal. Current Directions in Psychological Science, 10(6), 224-228.
Zeki, S. and M. Lamb (1994) The neurology of kinetic art. Brain, 117, 607-636.
Zizak, D.M. and A.S. Reber (2004) Implicit preferences: the role(s) of familiarity in the structural mere exposure effect. Consciousness and Cognition, 13, 336-362.
Previous Article | Next Article