CHAPTER TwENTY ONE A BIOLOGICAL APPROACH TO A MODEL OF AESTHETIC ExPERIENCE OSHIN VARTANIAN AND MARcos NADAL Recently, Leder and colleagues (2004) introduced an information-processing model to account for aesthetic experience. This model breaks the computation of the aesthetic response into five stages, associating each stage with a particular process of interest. In this paper we review results from recent neuroimaging studies of visual aesthetics to determine the extent to which they support this model. In addition, we derive specific hypotheses from the model that remain to be tested at a biological level. We argue that because all the cognitive and emotional processes that comprise the model are instantiated in the brain, one should in principle be able to test this model using biological methods. We conclude that the model is a promising framework within which to conduct such work on aesthetics. There is now general agreement that the aesthetic experience is the outcome of a complex interplay of cognitive and affective processes. Recently, Leder and colleagues introduced an information-processing model to account for the interaction of various component processes in the computation of aesthetic experience (Leder el al., 2004, 2005). Their model of aesthetic experience was described at the psychological level, and unlike other models (e.g., Chatterjee, 2003) was not designed to account for the biological underpinnings of aesthetic experience per se. Nevertheless, the model of aesthetic experience has certain characteristics that make it amenable to neuroscientific investigation. First, it breaks the computation of the aesthetic response into various stages, associating each stage with a particular process of interest. Because neuroscientists have studied those processes in contexts other than aesthetics, rudimentary cortical maps of their neural correlates have begun to emerge. This feature allows one to test hypotheses about whether any particular process of interest iSOlated within this model will map onto plausible cortical structures. Second, and critically, there are built-in temporal constraints in the structure of the model. In other words, information flows in specified ways through the system, and this orderly 430 Chapter Twenty One A Biological Approach to a Model of Aesthetic Experience 431 flow has certain temporal characteristics associated with it. This feature allows one to test the temporal dynamics of information flow using time-course and functional connectivity analyses. The aim of this chapter is a~ follow.s. First, we will review some of the key features of the mo~el of aesthetIc expenence that are particularly relevant to our arguments. We wIll not present a detailed account of the model as these exist elsewhere,(Leder el al., 2004, 2005). Second, we will compare this model to Chatterjee s (2003) model of visual aesthetics, developed specifically to address the neur~phys.IOIogy of aesthetic experience. Third, we will review some work on the biOlogIcal bases of the aesthetic experience that speak to some of the predlcllons and hypotheses derived from the model of aesthetic experience. Essenllally, we beheve that biOlogical approaches have the potential to inform us about the vahdlty of thIS model, and that predictions derived from the model can 10 tum be tested at a biological level. Although in this paper we will focus on neuroimaging studies only, the arguments are also relevant to neuropsychological approaches involving patient populations (e.g., Chatterjee, 2004). Fmally, we wIll assess the current status of the model of aesthetic experience based on the available biological data, and will outline specific hypotheses that can be used to test the so-called joints in the system. A Model of Aesthetic Experience Here we pre~ent a stripped down version of Leder el al. 's (2004, 2005) model of aesthettc expenence. The model of aesthetic experience is comprised of five mformatlOn-processmg stages that are connected in sequence, as well as through. several feedback loops (see Fig. 21-1). Information flow is unldrrectiOnal lo some parts of the model and bidirectional in others such that certain phases involve bottom-up as well as top-<lown processing. fu addition, there IS an affecllve evaluation stream (Continuous Affective Evaluation) that nms p",:"lIel to thIS sequenllal stream and receives its output. The input into the system IS the artwork Itself, which for the purpose of this paper will be limited to vls~~l1 s~lmuh, speCifically paintings. Then, at each stage, a particular operation. .IS perf~nned on the artwork, therefore extracting various characlensllcs from It. The ftrst stage involves perceptual analyses. At this stage features such as compleXJty or symmetry are distilled. For example, there is mU~h research. demonstrating that people prefer more 10 less symmetrical deSIgn. Accordmg to the model of aesthetic experience, this information is processed rather early in the stream. This stage is not under the influence of top- down processes and IS sllmulus driven. The second stage involves implicit memory IUtegratton, where the perceptual information is related to past expenence. For example, we know that people prefer colors that are more prototypical (Martindale & Moore, 1988). However, we also know that what is deemed prototypical depends in part on personal expenence. Essenllally, at thIS stage people compare what they see to what they know, and this affects their responses to it. This stage is presumed to be under the indirect influence of top- down processes. The third stage mvolves exphClt clasSIficatiOn, and this IS where expertise comes into play. At this point, the person analyzes content information, and also explicit information about the style of the artwork. There is much evidence demonstrating that expertise affects the way in which artworks are processed (Hekkert & van Wieringen, 1990), and this is one of the stages where the difference between experts and novices would be apparent. In some ways the fmal two stages of the model are the most interesting not only because they tap higher-level cognition, but also because.they probably exert the most influence on aesthetic experience. The penultunate stage 15 referred to as cognitive mastering, the moment at which interpretation or meaning is imposed on the artwork. Thus, having already distilled its perceptual properties and placed it within self-referential (implicit memory integrallon) and explicit (explicit classification) contexts, we make sense of what It IS that we see. Of course, what one observes is also influenced by experllse 10 the VIsual arts so that different cues become more or less important in giving meaning to the' artwork (parsons, 1987). In the final stage referred to as evaluation we appraise the meaning or interpretation that was placed on the. artwork durmg mastering. This evaluative stage generates two outputs: aesthetIc judgment and aesthetic emotion, which are the endpoints of the aesthellc expenence. If cognitive mastering is successful and the subject has successfully mterpreted the artwork it will be evaluated as either a good or a poor work of art. Those aestheti~ judgments will be accompanied by po~itive and negative aesthetic emotions respectively. On the other hand, Ifcogmtlve mastenng IS unsuccessful, the artwork will likely be evaluated as a poor work of art, accompanIed by negative aesthetic emotion. . . In summary, the model of aesthetic experience has fiv~ cognitive stages that are interwoven by an affective component, although the lOfluence of affect on the computational process. varies along the stream. The model of aesthetIc experience presents a hypothetical route for the generatIon of the two most common dependent variables in aesthetic research: AesthetIC judgment and aesthetic emotion. Aesthetic emotion can be seen as the fmal affectIve byproduct of successfuJ mastering, whereas aesthetic judgment can tap eIther the cognitive outcome of the mastering stage (i.e., quahty), or Its affecllve consequences. 432 Chapter Twenty One A Biological Approach to a Model of Aesthetic Experience 433 Theoretical Links to Other Models Chattetjee's (2003) model of visual aesthetics represents a recent neuroscientific framework for investigating aesthetic experience. Chatterjee (2003) has suggested that aesthetic experiences related to visual ohjects involve three visual processing stages common to the perception of any visual stimulus, as well as an emotional response, a decision, and the modulating effect of attention. In the frrst stage early visual processes hreak the stimulus down into simple components, such as color, shape, and so on, which are extracted and analyzed in different brain areas. The second stage, intennediate vision, includes a series of operations that segregate some elements and group others, forming coherent representations. In late visual stages, included under the representational domain in this model, certain regions of the ohject are selected for further scrutiny. At this moment, memories are activated, and ohjects are recognized and associated with meanings. This visual analysis leads to emotions associated with. the aesthetic experience, and it grounds decisions about the stimulus. However, this is not a strictly linear model. In fact, it posits an important feedback flow of information via attentional processes, from higher visual and emotional levels towards early visual processing. A comparison of the models proposed hy Leder and colleagues (2004) and Chattetjee (2003) reveals similarities and differences. Both models acknowledge the importance of early and late visual processes in the generation of an emotional response and the elaboration of a decision. They also take into consideration the influence of complexity, order, grouping, and many other variables familiar to experimental aestheticians, as well as the interaction between affective and cognitive processes such as the activation of memories and the search for meaning. Additionally, hoth models suggest two different outputs: an emotional response or aesthetic emotion versus a decision or aesthetic judgment. However, at a more specific level, these models have emphasized different aspects of aesthetic experience. Chattetjee's (2003) model deals extensively with perceptual processes, hut makes little mention of higher cognitive processes, such as interpretation or classification. In contrast, Leder and colleagues (2004) suhsumed all perceptual processes in a single stage and did not explicitly consider a function for attention, instead specifying higher cognitive processes in detail, and awarding them a central role in the aesthetic experience. Fig. 21-1 shows a comhined representation of both models, illustrating their similarities and differences. There could be several reasons behind the differences between Chattetjee's and Leder and colleagues' models. First, Chattetjee's (2003) ohjective was to create a framework for neuroaesthetics that was fumly based on frodings from visual neuroscience. In his mode~ the processes involved in visual ohject recognition Elllliualioll - , 434 Chapter Twenty One A Biological Approach to a Model of Aesthetic Experience 435 constitute the starting point for visual aesthetics, so it is not surprising that they figure so prominently in his modeL This is also the reason why attention is. awarded a central role: It is known to exert top down modulation of early visual processing. On the other hand, Leder and colleagues (2004) aimed to present an information-processing model of the stages involved in the aesthetic processing of visual artistic stimuli. In this sense, the starting point of the model was their analysis of modern art (Leder ef al., 2004, p. 491). They believe that understanding plays a critical role in the aesthetic experience of modern art, in the sense that comprehending an artwork alters the way in which it is experienced, However, there is also a deeper difference between the two models, and it refers to the way in which they conceive of the aesthetic experience itself, Chatterjee (2003) believes the notion of disinterested interest adequately captures the aesthetic experience such that "the viewer experiences pleasure without obvious utilitarian consequences of this pleasure" (Chatterjee, 2003, p.55). From this perspective 'judgments about an aesthetic object might be considered outside the core aesthetic experience" (Chatterjee, 2003, p. 56). In fact, the model seems to include a decision phase only as an approach to laboratory settings, where participants are usually asked to state their preferences or make decisions about a certain aspect of the stimulus. Conversely, Leder and colleagues (2004) believe aesthetic experiences arise when "exposure to art provides the perceiver with a challenging situation to classify, understand and cognitively master the artwork successfully" (Leder et aI., 2004, p. 493). This successful mastering of the artwork involves, cspecially in relation to modern art, style-related processing, which results from the acquisition of expertise, In this model, the judgment of the aesthetic object is an important element; in fact, together with aesthetic emotion, it is the main output' of the model. In sum, whereas Chatterjee's (2003) proposal can be considered as a neuroscientific model of aesthetic preference for a broad range of visual objects, Leder and colleagues' (2004) proposal is an information-processing model of aesthetic judgment of visual works of art. Here we chose to focus on the model of aesthetic experience because we were particularly interested in the higher-level cognitive and emotional processes that mediate aesthetic experience, and those are treated more thoroughly in the model of aesthetic experience, Neurophysiology and the Aesthetic Experience Why is the model of aesthetic experience useful for biological approaches to the study of aesthetics? There are three reasons. First, the model incorporates cognition and emotion-broadly defined-in the computation of the aesthetic response, and recently neuroscientists have made significant inroads in dissociating the neural pathways belonging to those modes of information processing. What we know from neuroscience can be used to test predictions from the model. Second, at a more micro level, many of the component processes that characterize each of the five stages, namely perceptual (visual) analyses, implicit memory integration, explicit classification, cognitive mastering, and evaluation have been studied extensively by neuroscientists as well, and at least at a rather gross level we know a little about their neural correlates. This makes it possible to test more specific hypotheses about the differential engagement of each of these processes in the computation of the aesthetic response, Finally, the structure of the model of aesthetic experience places temporal constraints on the process. For example, by definition, one cannot engage in cognitive mastering unless one has carried out a perceptual analysis first. This can be a valuable tool in neuroimaging because one can conduct time-course and functional connectivity analyses to see whether the time courses of activation corresponding to various structures occur in accordance with the predictions of the modeL Next we will review some neuroscientific evidence that can be used to assess the validity of the model. Although we will discuss studies in visual aesthetics specificaIly, we emphasize that biological data collected in studies of vision, memory, attention, and emotion can shed light on this process as ~ell. FoIlowing this review, we will highlight specific hypotheses that can be denved from this model and tested empirically to validate the model at a blOlog!cal level. Components of the Aesthetic Experience To date, five neuroimaging studies have appeared that have anc~pted to shed light on the cortical underpinnings of the aesthetic response. Four IOvolved the technique of functional magnetic resonance imaging (tMRI), and one involved magnetoencephalography (MEG). Although none of the studIes was conducted with the specific aim of testing any predIctIon denved from the model of aesthetic experience, their results nevertheless inform us about the accuracy of the model. We wiIl next review the key findings of each study, and their bearing on the model of aesthetic experience. Aesthetic Judgment Jacobsen ef al (2006) asked a fundamental question: What are the specific neural correlates that distinguish aesthetic judgment from other types of judgment? The "other" judgment in their study involved judgment of symmetry. 436 Chapter Twenty One A Biological Approach to a Model of Aesthetic Experience 437 Recall from the description of the model of aesthetic experience that judgment of symmetry occurs at the first stage involving perceptual analysis, whereas aesthetic judgment follows the fifth stage of the process-evaluation. Because subjects were presented with the same kind of stimuli in the symmetry and aesthetic judgment conditions but asked to make different types of judgments, Jacobsen el at. (2006) argued that the contrast between aesthetic and symmetry trials would reveal the brain areas that are involved in aesthetic judgment, in relation to symmetry judgment. This contrast revealed activation in several frontal, parietal, and temporal structures including the frontomedian cortex, the precuneus, the temporal pole, and the t~rnporoparietal junction. According to the model of aesthetic experience, the activation pattern that reflects aesthetic judgment must differ from the pattern that reflects symmetry judgment, and Jacobsen el aI's (2006) results confirm this hypothesis. Recall that the activation pattern that was revealed in the contrast of aesthetic versus symmetry judgment reflects multiple cognitive and affective processes that operate on the artwork following the perceptual analysis at the first stage, until an aesthetic judgment is eventually fonned. These processes include implicit memory integration, explicit classification, cognitive mastering, and evaluation. From the vantage point of the model of aesthetic experience. what is interesting about Jacobsen et al.s (2006) resulls is that several of the activated structures have been linked to the aforementioned component processes. Nevertheless, additional studies in which the design allows the comparison of successive stages of information processing are necessary for determining the contribution of each cortical structure 10 a specific component process. Affect, Cognition, and Aesthetic Experience No study to date has investigated differences in the neural correlates of aesthetic judgment versus aesthetic emotion directly. At a rudimentary level, this would involve presenting subjects with the same stimuli under two different conditions: In onc condition they would be asked to rate the stimuli on quality, thus tapping the cognitive component of aesthetic judgment exclusively (Leder el al., 2005; Vartanian & Goel, 2004a). In the other condition, they would be asked to rate the same stimuli on pleasure, thus tapping the affective component of aesthetic emotion exclusively (Leder el al., 2005; Vartanian & Goel, 2004a). However, three fMRl studies to date have tackled facets of aesthetic affect or emotion, and can thus shed light on whether variations in aesthetic emotion correspond to variations in cortical and subcortical activation. Vartanian and Goel (2oo4b) sought to determine whether aesthetic preference toward works of art is characterized by a "disinterested" or cognitive stance as presumed by some, or whether is it underwritten by an emotional response toward properties of artworks. They hypothesized that if aesthetic preference were mediated by emotion, then it should involve brain structures that have been implicated in processing emotion. On the other hand, if aesthetic preference were primarily a cognitive process, then it should involve brain structures that have been implicated in evaluation under emotionally neutral conditions. In the scanner, subjects viewed and rated paintings on aesthetic preference. Preference was defined as the degree of liking for a painling. The results demonstrated that activation in several cortical structures that have been implicated in processing emotion or reward covaried as a function of preference ratings, including the visual cortex, the caudate nucleus, and the cingulate sulcus. What do the results of Vartanian and Goel (2004b) tell us about the model of aesthetic experience? Recall that the affective evaluation stream runs parallel to the information-processing stream, and it receives continuous input from it. This means that a subject interacting with a work of art can provide a preference rating for that artwork at any given point along the information-processing sequence, and need not have processed the artwork up to a particular stage in the sequence before a rating can be generated. Therefore, one possibility is that the results of Vartanian and Goel (2oo4b) shed light on the cortical and subcortical structures that mediate Continuous Affcctive Evaluation, and indicate the areas thal one should expect to see activated whenever subjects are asked to indicate their liking for given artworks. Another possibility is that the ratings offered by the subjects in this study reflect aesthetic emotion, which can only occur following the evaluation stage. An additional sludy in which ratings are collected at specific time points can address this issue. Kawabata and Zeki (2004) presented their subjects with paintings that they had rated as beautiful or ugly prior to viewing, and rated them again in the scanner. It is important to note that although beauty has affective and cognitive components, it draws more heavily from the latter than from the. fonner component (Leder el al., 2005). In contrast, preference also has affective and cognitive components, but it draws more heavily from the fonner than fr~~ t~e latter component (Leder el al., 2005). Their results demonstrated that acllvtly m the orbital frontal cortex was greater for stimuli classified as beautiful, and the authors argued that this activation in the orbital frontal cortex was due to the reward value of beautiful paintings. In the third fMRl study on this topic, Skov el al., (2005) presented their subjects not with paintings, but with stimuli from the Inte~ation.al Affective Picture System. The lntemational Affective Picture System IS an lDventory of pictures that have been categorized as emotionally positive, ~egative, or neutral. Subjects were asked to view and rate each stimulus as beautiful, u~ly, or neu~al in the scanner. Compared to ugly pictures, beautiful pictures activated a Wide 438 Chapter Twenty One A Biological Approach to a Model of Aesthetic Experience 439 network of areas including the occipital, parietal, and frontal lobes. However, when subjects rated pictures as beautiful despite the fact that they were emotionally negative (e.g., finding a scene that exhibits death or injury beautiful), there was activation in a somewhat different network than before, again including the occipital, temporal, and the frontal lobes, but in particular bilateral orbital frontal corlex. What do the results of Kawabata and Zeki (2004) and Skov ef al. (2005) tell us about the model of aesthetic experience? Despite methodological differences, these studies share a critical feature: Both studies attempted to isoJate those cortical structures that were activated relatively more by stimuli evaluated as beautiful. According to the model of aesthetic experience, evaluations of beauty tap aesthetic judgment. Thus, its computation can only occur following processing along all five stages of the model, culminated by evaluation. In other words, activation in the orbital frontal cortex is likely not in relation to Continuous Affective Evaluation. Rather, it is more likely that certain affective properlies of beauty are computed in the orbital frontal corlex. Activation in the orbital frontal corlex has in turn been linked to a wide array of processes, but in particular to complex reward, hedonic, and emotion interactions (Kringelbach, 2005; Kringelbach & Rolls, 2004). Using magnetoencephalography (MEG), Cela-Conde ef al. (2004) recorded brain activity while subjects judged the beauty of a series of stimuli. The greatest difference between MEG and fMRI has to do with temporal and spatial resolution. Whereas MEG has a low spatial resolution compared with fMRI, its temporal resolution is significantly greater. In addition to locating brain activity in space, this technique affords information about its temporal course. In Cela- Conde and colleagues' (2004) study, brain activity during the first second after stimulus onset was broken down into two phases, early latencies (100 400 ms) and late latencies (400-1000 ms). The results demonstrated that during late latencies activii)' in the left dorsolateral prefrontal corlex (DLPFC) was significantly greater when participants judged stimuli as beautiful as compared to the non-beautiful condition. Previous studies can help us clarify the relation between these results and the model of aesthetic experience. The literature suggests that the dorsolateral prefrontal corlex is involved in the process of decision-making based on perceptual (Heekereo ef al., 2004) and/or affective (Davidson & Irwin, 1999; Herrington ef al., 2005) information. Krawczyk (2002) provided an integrative view of the role of this area: "The left DLPFC may playa privileged role in decision making that is better constrained, has fewer options, and which may have preexisting reward characteristics that make for a more confmed set of rules for deciding" (Krawczyk, 2002, p.66I). Thus, dorsolateral prefrontal cortex activity seems to be related to conscious deliberation about different options, influenced by emotional information from orbital frontal corlex and certain limbic areas (Wallis and Miller, 2003). The fact that activity in early latencies was unrelated with beauty ratings, coupled with results from previous studies, suggest that Cela-Conde and colleagues' (2004) results reflect the neural correlates of the last two cognitive stages posited in the model of aesthetic experience. Itis during these stages that the success of cognitive mastering in producing satIsfactory understanding IS monitored. Also, it is suggested that the subjective experience of success or failure in understanding can initialize top-down information processing. In fact, monitoring other cognitive processes, as well as initiating top-down processes, has often been associated with dorsolateral prefrontal corlex activity. The task that Cela-Conde and colleagues' (2004) participants were asked to perform was a quick judgment of the image. Images were presented for only 3 seconds, .and most of the participants' response times were below 2s. Therefore, we beheve that it is plausible that as Leder and colleagues (2004, p. 503) anticipated, their judgments may have relied heavily on affect-based heuristics. These considerations reinforce the idea that the left dorsolateral prefrontal corlex might be involved in a number of processes, including the evaluation phase, the initiation of the feedback loops posited by the model of aesthetIc expenence, and the interaction between cognitive and affective states. Summary Vartanian and Goel (2004b), Kawabata and Zeki (2004), Skov ef at. (2005) and Cela-Conde ef al. (2004) were interested in determining the neural correlates of preference and beauty, two variables that have affective and cognitive components. The areas activated by Vartanian and Goel (2004b) may have highlighted those cOrlical structures that mediate Contmuous Affecllve Evaluation, or those associated with aesthetic emotion. As expected, they include the visual corlex, the caudate nucleus, and the cingulate sulcus. These areas have been shown to be activated by emotions, and in particular by sahent stimuli about which one can form an affective impression rather automatically, such as faces or pictures from the International Affective Picture System. In contrast, the studies by Kawabata and Zeki (2004) and Skov ef al. (2005) attempted to isolate those cortical structures that are actIvated more when a stimulus is evaluated as beautiful. Presumably, both studies tap aspects of aesthetic judgments. According to the model of aesthetic experience? beauty bas affective and cognitive components and it results from an evaluallon that can only occur following processing along all five stages of the model. The results indicate that evaluating a stimulus as beautiful was asSOCIated WIth mcreased activation in the orbital frontal cortex. Activation in the orbital frontal cortex has 440 Chapter Twenty One A Biological Approach to a Model of Aesthetic Experience 441 in turn been linked to a wide array of processes, but in particular tn complex hedonic and reward-emotion interactions (Kringelbach, 2005; Kringelbach & Rolls, 2004). Finally, Cela-Conde et aI's (2004) results seem to reflect cognitive processes included in later stages of the model of aesthetic experience, including cognllIve mastering and evaluation. Activity in the dorsolateral prefrontal cortex has often been associated with monitoring and initiating top <lown infonnation flow, as well as with decision-making. However, we suggest that in this particular instance these cognitive processes were also influenced by affective information received from orbital frontal cortex or subcortical structures. Testing the model of aesthetic experience at a biological level It goes without saying that the validity of the model of aesthetic experience as an accurate explanatory framework for aesthetic experience must be detennioed at the behavioral level flfSt. However, we believe that because 0/1 the processes that comprise this model, including perceptual analysis, implicit memory integration, explicit classification, cognitive mastering, and evaluation are instantiated in the brain, one can also ascertain the extent to which hypotheses derived from the model can be validated at a biological level. We have already discussed the extent to which results from a number of tMRl and MEG studies fit general predictions from the model of aesthetic experience. However, none ·of those studies was designed to test predictions derived from the model of aesthetic experience. We believe that five issues in particular can be tested effectively at a biological level, and that those results Can be used to assess the validity of the model. First, Leder el 01. (2004) argued that the context in which an object is viewed affects the way in which it is processed. For the predictions of the model of aesthetic experience to hold, the input into the system must be designated as an artwork. Essentially, this is based on the argument that the cognitive and emotional processes that are brought to bear when processing an object as an artwork will differ from the processes that will be involved when the same object is not processed as an artwork. It is possible to test this hypothesis at a biological level, and to determine whether performing identical tasks on a set of stimuli will activate different cortical structures depending on whether they are deSIgnated as artworks or not. This could help determine whether a critical assumption of the model holds true at a biological level. . Second, the model of aesthetic experience does not include top-down influences on perceptual analyses. However, several studies, such as Kaestner and Ungerleider's (2000) or Poghosyan and colleagues' (2005) have shown that attentton modulates the processing of relevant visual stimuli by enhancing neuronal responses at different levels of visual processing in the brain. Attention seems to modulate neural responses to certain locations of the visual field, whole visual objects. or specific visual features, such as color or shape. Although it has been noted that these modulatory effects are stronger in extrastriate visual areas, it seems that different features of selective attention can also affect activity in striate cortex. Furthennore, it has also been shown that the emotional valence of images can modulate activity in visual areas (Lang el 01., 1998; Schulman el 01., 1997). Hence, if future neuroimaging studies of aesthetic appreciation include strategies to control affective and attentional processes, they might be able to detennine whether these processes exert top <lown influences on early perceptual analyses. Third it is well established that expertise plays an important role in the way artworks' are processed. Numerous studies have demonstrated systematic differences in the ways experts and novices view artworks (Bekkert & van Wieringen, 1990; Nodine el 01., 1993). This hypothesis can be tested at a biological level. According to the model of aesthettcexpenen.ce, differences due to expertise become evident in the third stage. (ex!'"clt classificatton) when the person analyzes content information and expliCit mformatton about the style of the artwork. This process should draw on categonzatlon and memory, and different activation patterns should characterize those processes in experts and novices. Fourth, a critical topic in the early days of aesthetic research involved the aesthetic threshold, although interest in this topic has subSIded over the years (Jacobsen, 2006). Rather than calculating the aesthetic threshold,researchers are using presentation thresholds that are appropriate for thelT particular Issues of interest (e.g., Leder el 01., 2006). By reliance on a combination of be?avioural and neuroimaging techniques (especially MEG), the temporal dynanucs. of the aesthetic experience, as well as the engagement and disengage~ent of different stages (processes), can be investigated For ex~mple, what IS the mmlJ~al amount of time necessary for explICIt classIlkatton, and does explIcit classification in fact require more time than implicit memory ~t.egratlO~, as ~e model suggests? Are implicit memory integration and expllclt clas~Ilicatlon associated with different' patterns of cortical actIvation and lInked to characteristic time courses? Fifth what is the relationship between the two major outputs of the system, namely ;"'sthetic judgment and aesth~tic emotion? This is a p~oblem that has haunted philosophers and psychologISts at least smce the 18 cenDlfy. ThIs requires a design in which subjects are instructed to process art stlfOuII .that can be evaluated successfully under two different conditIOns: In one condltton th~y will rate them on quality, broadly speaking, and in the other conditton they WIll rate them on a measure of liking, broadly speaking. If the model of aesthetIc 442 Chapter Twenty One A Biological Approach to a Model of Aesthetic Experience 443 experience is correct, the neural correlates of these two outputs must be different. Conclusion Th~re exist. several models in experimental aesthetics that deal with the aestheltc expenence along narrow levels of analysis. In contrast the model of aesthetic expe~ence provides a general framework for aesthetic' experience at the psychologIcal le~el, and yet can be tested experimentally using biological methods at a more ffi!cro level (see also Martindale, 2001). We believe that the model of aestheti~ expcrience is a promising model for biological investigations of aesthetic expenence. References Cela-Conde, C. 1., Marty, G., Maetsu, F., Ortiz, T., Munar, E., Fernandez, A., Roca, M., Rossello, J:, & Quesney, F. (2004). Activatinn of the prefrontal cortex m the human VIsual aesthetic perception. Proceedings of the National AcademyofSdences USA, 101,6321-6325. ChatteIjee, A. (2003). Prospects for a cognitive neuroscience of visual aesthetics. Bulletin of Psychology and the Arts, 4, 55-60. (2004). The neuropsychology of visual artistic production. Neuropsychologia, 42, 1568-1583. Davidson, R.J., & Irwin, W. (1999). The functional neuroanatomy of emotion and affecltve style. Trends in Cognitive Sciences, 3, 11-21. Heekeren, H. R., Marrett, S., Bandettini, P. A, &. Ungerleider, L. G. (2004). A general mechamsm for perceptual decision-making in the human brain. Nature, 431, 859-862. Hekkert, P., & v~ Wietingen? P.C.W. (1996). The impact of level of expertise on. ~e evaluatIon of ongmal and altered versions of post-impressionistic pamltngs. Acta Psychologica. 94, 117-131. Hertington, J. D., Mohanty, A, Koven, N. S., Fisher, J. E., Stewart, J. L., Bamch, M. T., Webb, A. G., Miller, G. A., & Heller, W. (2005). Emotion- modulated performance and activity in left dorsolateral prefrontal cortex. Emotion, 5, 200-207. Jacobsen, T. (2006). Bridging the arts and sciences: A framework of the psychology of aesthetics. Leonardo, 39, 155-162. Jacobsen, T., Schubotz, R. I., HOfel, L., & Cramon, D.V. (2006). Brain correlates of aesthetic judgment of beauty. Neuroimage, 29, 276-285. Kaestner, S., & Ungerleider, L. G. (2000). Mechanisms of visual attention in the human cortex. AnnualReview of Neuroscience, 23, 315-341. Kawabata, H., & Zeki, S. (2004). Neural correlates of beauty. Journal of Neurophysiology, 91, 1699-1705. Krawczyk, D. C. (2002). Contributions of the prefrontal cortex to the neural basis of human decision making. Neuroscience and Biobehavioral Reviews 26,63H;64. Kringelbach, M. L. (2005). The human orbitofrontal cortex: Linking reward to hedonic experience. Na/ure Reviews Neuroscience. 6, 691-702. Kringelbacb, M., & Rolls, E. (2004). The functional neuroanatomy of the human orbitofrontal cortex: Evidence from neuroimaging and neurophysiology. Progress in Neurobiology, 72,341-372. Lang, P. J., Bradley, M. M., Fitzsimmons, J. R., Cuthbert, B. N., Scott, J. D., Moulder, B., & Nangia, V. (1998). Emotional arousal and activation of the visual cortex, and fMRl analysis. Psychophysiology, 35, 199-210. Leder, H., Augustin, D., & Belke, B. (2005). Art and cognition: Consequences for experimental aesthetics. Bulletin of Psychology and the Arts, 5, 11-20. Leder, H., Belke, B., Oebers!, A., & Augustin, D. (2004). A model of aesthetic appreciation and aesthetic judgments. British Journal of Psychology, 95, 489-508. Leder, H., Carbon, C c., & Ripsas, A-L. (2006). Entitling art: Influence of title information on understanding and appreciation of paintings. Acta Psychologica, 121, 176-198. Martindale, C. (2001). How does the brain compute aesthetic experience? The General Psychologist, 36,25-35. Martindale, C., & Moore, K. (1988). Priming, prototypicality, and preference. Journal of Experimental Psychology: Human Perception and Performance, 14,661-670. Nodine, C. F., Locher, P. J., & Krupinski, E. E. (1993). The role of formal art training on perception and aesthetic judgment of art compositions. Leonardo, 26, 219-227. Parsons, M. J. (1987). How we understand art: A cognitive developmental account of aesthetic experience. Cambridge, UK: Cambridge University Press. Poghosyan, V., Shibata, T., & Ioanoides, A. A (2005). Effects of attention and arousal on early responses in striate cortex. European Journal of Neuroscience, 22, 225-234. Schu1man, G. L., Corbelta, M., Buckner, R. L., Raichele, M. E., Fiez, 1. A, Miezin, F. E., & Petersen, S. E. (1997). Top-<!own modulation of early sensory cortex. Cerebral Cortex, 7, 193-206. Skov, M., Christensen, M. S., Rowe, J. B., & Paulson, O. B. (2005, June). Specific activations underlie aesthetic judgment of affective pictures. Poster 444 Chapter Twenty One presented at the 11 'h annual meeting of Human Brain Mapping, Toronto, ON, Canada. Vartanian, 0., & Gael, V. (2004a). Emotion pathways in the brain mediate aesthetic preference. Bulletin 0/Psychology and the Arts, 5,37-42. Vartanian, 0., & Gael, V. (2004b). Neuroanatomical correlates of aesthetic preference for paintings. NeuroReport, 15,893-897. Wallis, J. D., & Miller, E. K. (2003). Neuronal activity in primate dorsolateral and orbital prefrontal cortex during performance of a reward preference task. European Journal o/Neuroscience, 18,2069-2081. . CHAPTER TwENTY ONE A BIOLOGICAL APPROACH TO A MODEL OF AESTHETIC ExPERIENCE OSHIN VARTANIAN AND MARcos NADAL Recently, Leder and colleagues (2004). speaking. If the model of aesthetIc 442 Chapter Twenty One A Biological Approach to a Model of Aesthetic Experience 443 experience is correct, the neural correlates of these