R1.1. Misled by the aesthetic–artistic confound

Theories that neglect the distinction between the aesthetic and the artistic stem from what we propose to term the aesthetic–artistic confound, which is a theoretical assertion that identifies the aesthetic domain with the artistic domain, or at least significantly obscures their differences. Views that promote the aesthetic–artistic confound are expressed in numerous contributions to philosophical aesthetics (Beardsley Reference Beardsley1958/1981; Reference Beardsley and Curtler1983; Kant Reference Kant, Guyer and Matthews1793/2000; Stolnitz Reference Stolnitz1960) and aesthetic science (Dissanayake Reference Dissanayake1992; Jacobsen Reference Jacobsen2006; Leder et al. Reference Leder, Belke, Oeberst and Augustin2004; Locher 2012; Ramachandran Reference Ramachandran2011; Skov & Vartanian Reference Skov and Vartanian2009); see also Leder and Vartanian & Kaufman. Even one of us (Reber) could not free himself from the seductive appeal of this confound (Reber Reference Reber2008; Reference Reber, Shimamura and Palmer2012).

A philosophical view derived from the aesthetic–artistic confound is the aesthetic theory of art (Osborne Reference Osborne1981; Tolhurst Reference Tolhurst1984), which asserts that the chief function of art is to induce aesthetic experience (for a critique of this theory, see Carroll Reference Carroll1986; Reference Carroll2001).

The prevalence of this confound in psychology is illustrated by the fact that the journal of the International Association for Empirical Aesthetics is titled Empirical Studies of the Arts. Likewise, many researchers take for granted that neuroaesthetics is an adequate term to denote research on the neural bases of art appreciation, or that the term aesthetic science is acceptable to describe the scientific study of art (Shimamura Reference Shimamura2012).

Silvia was puzzled by the fact that “aesthetic science has an ambivalent relationship with art itself.” However, this “ambivalent relationship” is only puzzling if one takes for granted the seductive but misleading aesthetic–artistic confound. Likewise, Currie (Reference Currie2004), Gilmore, Hyman (Reference Hyman2006; Reference Hyman, Frigg and Hunter2010), and Noë (Reference Noë2011) have criticized aesthetic science on the contextualist ground that this science neglects the arts and their singular histories. Their criticisms echo earlier assessments (Dickie Reference Dickie1964; Munro Reference Munro1951, pp. 178–80). Such disputes would not arise if researchers could agree on how to demarcate the science of art from aesthetic science.

R1.2. Sensitive to historical conceptions of the arts

The psycho-historical research program defends a contextualist foundation for the science of art, which is currently missing in aesthetic science. In response to our critique of aesthetic science, Chatterjee claims that “artistic meaning” can “be ahistorical.” In our opinion, asserting that artistic meaning can be ahistorical is problematic because it invites or legitimizes disregarding of the art appreciator's sensitivity to art-historical contexts (see Gilmore; Hogan; Levinson). In contrast to ahistorical views, our aim is to argue that the science of art needs to take into account art-historical phenomena studied in social sciences and the humanities – such as art-historical categories and cultural learning (Richerson & Boyd Reference Richerson and Boyd2005; Sterelny Reference Sterelny2012) – to avoid the pitfalls of radical forms of anti-contextualism.

Consider the varieties of art-historical categories. Agents involved in art-historical contexts and scholars studying these contexts develop different conceptions of the arts (Shiner Reference Shiner2003; Tatarkiewicz Reference Tatarkiewicz1971). For example, techne in Greek (τέχνη, technique), ars in Latin, the concept of liberal arts, or the romantic concept of art are distinct categories. We agree with Fitch & Westphal-Fitch's and Leder's claim that these different historical categories should not be identified with the modern concept of fine arts (Shiner Reference Shiner2003). Art-historical categories like techne or liberal arts refer to distinct historical kinds (Hacking Reference Hacking, Cohen and Roth1995; Reference Hacking1999; Reference Hacking2002; Millikan Reference Millikan1999; Reference Millikan2000). Art-historical names like Yirrkala bark petitions 1963 (Museum of Australian Democracy) or Edvard Munch refer to distinct historical individuals (Danto Reference Danto and Dray1966; Strawson Reference Strawson1959). Such historical kinds and individuals are generated by singular causal processes that can often be discovered by means of historical inquiry (see R3.2; and De Smedt & De Cruz for a discussion of problematic cases). The psycho-historical program is based on the thought that one of the core tasks of a science of art is to account for the appreciator's sensitivity to such art-historical kinds and individuals. Because the psycho-historical program stresses the variety of art-historical contexts, kinds, and individuals, we disagree with Fitch & Westphal-Fitch's and Thompson & Antliff's claim that our program is irreversibly tied to Western fine arts (see also R3.2).

R2.1. Singular art histories amenable to scientific explanation

Bloom (Reference Bloom2010), Gelman et al., and Newman agree with our core hypothesis that the cognition of historical individuals and kinds is essential to art appreciation. As illustrated in Figure 1 of the target article and as argued by Newman, art appreciators are often exquisitely sensitive to the fact that works of art are causal extensions of the individual agents who produced them. Such a sensitivity to unique artistic histories is demonstrated by phenomena associated with appreciators' interest in authenticity (Newman & Bloom Reference Newman and Bloom2012) and with contagion (Newman et al. Reference Newman, Diesendruck and Bloom2011). Does art appreciators' interest in the uniqueness of art-historical agency raise a problem for the science of art?

We disagree with Chatterjee's claim that scrutinizing layered historical meanings of an individual artwork is “too fine-grained a level of analysis to be resolved by the lens of scientific experimental methods.” Humans routinely rely on the tracking of historical individuals over time to serve the identification of such individuals (Gutheil et al. Reference Gutheil, Gelman, Klein, Michos and Kelaita2008) and the scientific explanations of their behavior (consider the need to track individual organisms in ecology; see, e.g., Block et al. Reference Block, Teo, Walli, Boustany, Stokesbury, Farwell and Williams2005). Furthermore, there is scientific work on the mental mechanisms engaged in tracking and identifying historical individuals, such as visual tracking (Kahneman et al. Reference Kahneman, Treisman and Gibbs1992; Pylyshyn & Storm Reference Pylyshyn and Storm1988), multimodal tracking (Bullot Reference Bullot2009b; Bullot & Droulez Reference Bullot and Droulez2008), memory systems for self-knowledge (Conway Reference Conway1990; Reference Conway2005), face recognition for identification (Gobbini & Haxby Reference Gobbini and Haxby2007), and other mechanisms that track agents' identities (Bullot Reference Bullot2006; Bullot & Rysiew Reference Bullot and Rysiew2007; Gutheil et al. Reference Gutheil, Gelman, Klein, Michos and Kelaita2008; Rips et al. Reference Rips, Blok and Newman2006). Regarding art, the empirical research indicating appreciators' sensitivity to artistic historical individuals by Hood and Bloom (Reference Hood and Bloom2008), Newman and Bloom (Reference Newman and Bloom2012), and Gelman et al. (Gelman & Bloom Reference Gelman and Bloom2000; Gelman et al. Reference Gelman, Coley, Gottfried, Hirschfeld and Gelman1994; Gelman & Ebeling Reference Gelman and Ebeling1998; Gutheil et al. Reference Gutheil, Gelman, Klein, Michos and Kelaita2008) qualify as experimental research on the sensitivity to “layered historical meanings” (Chatterjee) of an art-historical individual. Lastly, though it rarely uses experimental methods and may often lead to historical fallacies (Fischer Reference Fischer1971), research in the historical social sciences and humanities that attempts to explain historical events can be based on rigorous evidence-based reasoning (McCullagh Reference McCullagh1984; Shafer Reference Shafer1969/1974; White Reference White1965; Wigmore Reference Wigmore1913) and source assessment (Gottschalk Reference Gottschalk1950/1969; Howell & Prevenier Reference Howell and Prevenier2001).

In contrast to Chatterjee's skepticism about a science of the sensitivity to historical individuals and unique events, we think that such sensitivity can be explained by theories that combine models of cognitive mechanisms with models of particular historical contexts. How can this integration be achieved? As indicated in Figures 1 and 2 of the target article, the psycho-historical program proposes to apprehend art-historical contexts, artists, works, and the mechanisms of appreciation as hierarchical and nearly decomposable complex systems – in the senses expounded by Bechtel (Reference Bechtel2008; Bechtel & Richardson Reference Bechtel and Richardson1993) and Simon (Reference Simon1969/1996). Our program seeks to identify some relations of hierarchical dependence (or loci of control) between these complex systems. For example, we hypothesize that the artist's work depends on both an originative art-historical context and a particular sequence of the artist's actions; or that the appreciator's artistic understanding depends both on basic exposure and the design stance.

Graham claims that our “radical” contextualist approach to complex systems is a “variant of holism” that dismisses “the viewpoint of the opposing side” understood as either universalism or reductionism. However, this interpretation misses the fact that the psycho-historical program can incorporate universalistic hypotheses and be locally reductionist. The multilevel and multicomponent structure of complex systems encourages pluralistic analyses of causal structures at different levels of organization (Mitchell Reference Mitchell2009). Such complex systems analysis can aim at identifying hierarchies of modular mechanisms whose workings are explained by means of reference to interactions between parts and subparts. This kind of view is remote from the holism criticized by Graham.

This hierarchical analysis of a nearly decomposable complex system can also be used in reply to an objection raised by Thompson & Antliff, who admit the lack of interaction between the psychology of art and art history but maintain that “it is not clear that such interaction will replace a status quo that is polarized.” According to the psycho-historical program, the psychology of art and art history often have interdependent (rather than independent) explanatory objectives because they study the same complex systems. Thus, in principle, an integrative explanation that combines psychological and historical descriptions of an artistic complex system will be preferable to explanations that are not integrated.

R2.2. Embodied or extended cognitive systems

Gibbs, Malafouris, Rollins, and Wilson offer commentaries from the standpoint of externalist and embodied theories of cognition and art (see also Brinck Reference Brinck2007; Manzotti Reference Manzotti2011; Tribble & Sutton Reference Tribble and Sutton2011). How does the psycho-historical framework relate to externalist theories of the mind (Clark Reference Clark2008; Menary Reference Menary2010; Putnam Reference Putnam1975; Wilson Reference Wilson2004) and theories of embodied cognition (Barsalou Reference Barsalou1999; Reference Barsalou2008; Gibbs Reference Gibbs2006)?

1. 1. We are not aware of works in the externalist tradition that integrate the psychological and historical approaches in the way our program does. For example, when they refer to history, advocates of semantic externalism like Kripke (Reference Kripke1980) and Putnam (Reference Putnam1981) rely on philosophical conceptions of causal and social history that engage with neither particular models of psychological mechanisms nor actual works by historians. Likewise, the works singled out by Wilson on the extended mind thesis (Clark & Chalmers Reference Clark and Chalmers1998; Wilson & Clark Reference Wilson, Clark, Aydede and Robbins2009) or Sutton's “historical cognitive science” derived from active externalism (Sutton Reference Sutton, Freeland and Corones2000; Reference Sutton, Floyd-Wilson and Sullivan2006; Reference Sutton, Knappett and Malafouris2008; Reference Sutton and Menary2010) have not examined the psycho-historical hypotheses that we propose about art appreciation.

2. 2. Several core hypotheses of the psycho-historical program seem neutral with respect to the alternatives between externalism and internalism. For example, the relevance and truth of our hypotheses on the artistic design stance and artistic understanding do not seem to directly depend on the truth of the externalist theses defended by Wilson and Malafouris.

3. 3. Malafouris claims that our framework is “internalist.” But such a claim misses the fact that the psycho-historical framework defends an historical externalism that is interpretable as “externalist” in at least two important senses specified in philosophy. First, the framework posits that modes and episodes of art appreciation are relations of epistemic tracking in the sense expounded by externalist theories in epistemology (Azzouni Reference Azzouni2004; Goldman Reference Goldman1967; Reference Goldman1999; Kornblith Reference Kornblith2001; Liebenberg Reference Liebenberg1990). On this account, a mental episode is an act of artistic appreciation because it tracks (is sensitive to) objective art-historical kinds and individuals (see sect. R1). Furthermore, our program argues that the classification of a token mental event as an episode of artistic appreciation depends on the identification of determinative relations between the mental episode (e.g., perceptual state, emotion) and historical categories and functions of arts and crafts (e.g., didjeridu, sfumato, or serialism; see sect. R1). Thus, this account is also externalist in a taxonomical sense (Wilson Reference Wilson2004, pp. 81–82).

4. 4. Certain auxiliary hypotheses of the psycho-historical program can be developed in the direction of an embodied approach as suggested by Gibbs or an approach based on perceptual strategies as proposed by Rollins. Embodied cognition (Gibbs) and the theory of perceptual strategies (Rollins Reference Rollins, Hecht, Schwartz and Atherton2003a; Reference Rollins2003b; Reference Rollins2004; Reference Rollins, Schellekens and Goldie2011) provide interesting hypotheses on the appreciator's simulation of the artist's actions during the creation of the artwork (Hirstein; Rollins). For example, some implicit processes might bypass explicit forms of the design stance if traces afforded by the artwork enable tracking of the artist's actions by means of basic exposure alone. However, such implicit processes may explain only part of the process of artistic understanding because causal information from the artwork alone is often insufficient, as Levinson and Ross rightly argue (see sect. R2.3).

5. 5. Our historical externalism is also reflected in our critique of the internalism of contemporary researchers in aesthetic science. The latter seem tempted to argue that episodes of artistic appreciation can be individualized independently of the relations of brain states to art-historical categories and contexts. In contrast to the externalist/contextualist approach, they may refer to knowledge stored in memory about the art-historical work and context (Hirstein). For example, although Leder et al.'s (Reference Leder, Belke, Oeberst and Augustin2004) proposal is one of the most advanced models in the psychological approach, this model follows the traditional internalist methodology that dominates aesthetic science, and it lacks the contextualist and externalist characteristics of the psycho-historical framework. We concede to Leder that Leder et al. (Reference Leder, Belke, Oeberst and Augustin2004) “postulated a stage of cognitive mastering, in which interpretation and assigning meaning are crucial.” However, Leder et al.'s model does not account for the determinative dependence of art-historical understanding on the appreciator's sensitivity to objective historical individuals, kinds, and contexts. Consequently, the model can account for neither the genealogy of context-specific artistic functions (Parsons & Carlson Reference Parsons and Carlson2008) nor the appreciators' sensitivity to such functions. For example, although the model could be integrated into the right part of Figure 1 (modes of art appreciation) in the target article, it circumvents the fundamental reference to the art-historical context depicted by the left part of Figure 1.

R2.3. Traces of intentions and inheritors of functions

We agree with Levinson that the concept of “causal-historical traces left in artworks” – used by Bullot (Reference Bullot2009a, pp. 96–97); see also Leyton (Reference Leyton1992), Shafer (Reference Shafer1969/1974), and Smail (Reference Smail2008) on historical traces – is adequate for analyzing each particular artwork as a causal extension of its maker and originative context (Newman). We used the concept of causal traces in prior versions of the manuscript. However, one reviewer's objections about its generality led us to use the term “carrier of causal information.”

Levinson argues that we adopt the misleading hypothesis that “causal history can be reliably inferred from what [we] call the causal-historical information carried by artworks.” Levinson's concern about information would be warranted if we had linked causal information to necessarily true information-driven belief, akin perhaps to Dretske's (Reference Dretske1981; Reference Dretske1994) theory of information-driven belief. Yet, this is not the case. We simply use causal information to refer to carriers of causal information qua appreciator-independent causal traces, which are also referred to as cues, indices, or marks in the literature. In fact, we agree with Levinson that causal information carried by the work is not a sufficient source for artistic understanding, and we do not assume that “retrieval is a simple matter” (Ross). Nor do we think that traces “transparently indicate” the artist's generative actions (Levinson). Artistic traces may be ascertained by means of multiple defeasible methods and sources that can be incorporated into the design stance. Relatedly, we agree with Levinson that beliefs, feelings, and explanations that result from the adoption of the artistic design stance are not immune to errors and misunderstandings (sect. 3.3.1), as implied in our connection between the design stance and inference to the best explanation (sect. 3.2).

The psycho-historical framework proposes hypotheses about the work as a causal trace of the artist's agency (sect. 2; Newman) and the genealogy of artifact functions (Parsons & Carlson Reference Parsons and Carlson2008; Preston Reference Preston1998). Davies thinks that we “identify artists' intentions as the primary data that appreciators attempt to retrieve from the art-creative context in the process of comprehending artworks.” But his point overlooks the fact that we acknowledge that works in the arts have a variety of complex unintended functions (Parsons & Carlson; Ross). Nevertheless, we concede to Davies that original intentions were mentioned too often without analysis of their complications. In that respect, Davies usefully lists seven complications faced by any appreciator who wishes to reliably attribute and interpret artistic agency. These complications refer to intentions that are (1) unconscious; (2) failed (see also Ross); (3) facilitated by social status and authority or (4) other factors in art-historical contexts; (5) categorial, as proposed by Levinson (Reference Levinson1996b, p. 188–89) and Rollins (Reference Rollins2004); (6) contradictorily assessed by actual intentionalism (Carroll Reference Carroll2000b) and hypothetical intentionalism (Levinson Reference Levinson2010); or (7) fancifully disconnected from actual historical intentions (sect. 3.1, R3.1).

Davies' typology of artistic intentions helpfully charts the rugged terrain that appreciators need to explore to track artists' conscious and unconscious agency (Davies Reference Davies1982;Reference Davies1996; 2006a). Davies' analysis can be expanded by the psycho-historical program. For example, unconscious intentions may not be directly known by the means of introspection (Carruthers Reference Carruthers2009; Wilson Reference Wilson2002). However, if causally efficient, our framework suggests that these intentions could be known indirectly if they leave causal traces in the artist's behavior and work (R2.3). For appreciators can adopt the design stance to retrieve information about these unconscious causes and keep track of the artist's action over time. Appreciation driven by the design stance and essentialist assumptions (Gelman et al.; Newman) might lead appreciators to posit unconscious drives – as in psycho-analytical interpretations of artistic creation (e.g., Breton Reference Breton1924/1988, p. 316) – that seek to explain manifest artistic behavior and unconscious agency. Given appreciators' propensity to overattribute intentionality and mentality (Bering Reference Bering2006; Heider & Simmel Reference Heider and Simmel1944), this interpretative process might lead to illusions and artistic misunderstanding (sect. 3.3; Gilmore; Levinson; Newman).

We agree with Rollins that the design stance may be “construed in terms of positing hypothetical intentions, based on beliefs about the actual historical context in which the work was produced;” see also Tullmann. From the standpoint of normative artistic understanding (Gilmore; Ross), the psycho-historical program implies that accounts of virtual artistic intentions have to be integrated with information from the art-historical context to obtain relevance and plausibility. If such information is not available (see De Smedt & De Cruz; Fitch & Westphal-Fitch), virtual etiologies or thought experiments (Gendler Reference Gendler2010) may be the only way to achieve a form of understanding that might be richer than basic exposure (sect. R3.2).

Parsons & Carlson elaborate on the difference between artistic intentions and proper functions. According to their account (Parsons & Carlson Reference Parsons and Carlson2008), the proper function of an artwork does not essentially depend on artistic intentions; it “must be analyzed in terms of artwork's causal histories.” Specifically, “an artwork has F as its proper function just in case it belongs to a type that has achieved selective success in the marketplace due to performing F” (Parsons & Carlson). In many cases, the artist never envisioned the proper function the artwork gained over time.

Parsons & Carlson's clarification is important because it strengthens artistic contextualism. However, we also focused on intentions as a means of stressing potential commonalities between Bloom's psychological and intentionalist account of artifact categorization (Bloom Reference Bloom1996a; Reference Bloom1998) and Parsons and Carlson's purely philosophical theory of proper functions. Furthermore, despite our endorsement of Parsons and Carlson's proper functions, we are interested in a more encompassing analysis of artistic functions because artworks may have ephemeral, albeit reproduced, effects that might correspond to artistic functions without qualifying as proper functions. For example, and in contrast to Fitch & Westphal-Fitch's charge of elitism, the psycho-historical approach can vindicate the appreciation and understanding of art brut (“low art” or “outsider art”) by self-taught or naïve art makers whose work has never been institutionalized (Dubuffet Reference Dubuffet1986). According to Parsons & Carlson, these works cannot be conferred proper artistic functions because they are not socially sanctioned or known as art. According to a more encompassing psycho-historical view, however, works of art brut nonetheless have artistic functions of a private type in which the self-taught art brut artist performs the functional roles of both artist and audience.

R3.1 Basic exposure

Tullmann asks whether basic exposure to the artwork is necessary in order to appreciate the work. Others questioned whether it is necessary for eliciting the design stance (Rollins; Ross) and subsequently causal “reverse engineering” (Thompson & Antliff) or artistic understanding (Schellekens). What about an artwork seen in the past (see also Hirstein)? Does a friend's testimony about a novel count as basic exposure?

Any exposure to information about the artwork, including poor reproductions or testimonies by friends, could count as rudimentary types of basic exposure. Could one appreciate a work as made by an agent in a particular context if we were unable to access any sorts of information about it? One could not. Thus, minimal basic exposure is a necessary condition for any mode of appreciation, and therefore for eliciting the design stance.

In many cases, however, minimal exposure would not be adequate in terms of searching for and finding the accurate causal information about the art-historical context. For example, poor reproductions, indistinct memories, or unreliable testimonies may misguide the artistic design stance, triggering searches for information that lack relevance. In contrast, veridical and rich external representations of artworks – for example, high-resolution visual, audio, and audiovisual depictions, “compliant notational systems” (Goodman 1968), or reliable testimonies (Lackey & Sosa Reference Lackey and Sosa2006) – and veridical internal representations of artworks (e.g., episodic memories; see Hirstein) will facilitate the appreciator's search for accurate causal information when adopting the design stance. Consequently, the availability of veridical representations should facilitate appreciation of the work based on artistic understanding.

Rollins, Ross, Schellekens, and Tullmann criticized the hypothesis of a strictly unidirectional causal relation linking basic exposure to design stance and the latter to artistic understanding. As discussed in the target article, the arrows in Figures 1 and 2 refer to necessary conditions, not temporal order. For example, when a reader knows that she is going to read a novel, she presumably does infer a categorial intention (Levinson 1996; Rollins Reference Rollins2004), eliciting the design stance before basic exposure occurs. Nevertheless, the search for causal information typical for the design stance can only start with basic exposure to information about the artwork, for example by reading a novel. The actual process of appreciation is best captured as a recursive process including feedback loops. A reader may anticipate that she will read a novel and prepares to adopt the design stance. When reading the novel (basic exposure), she looks for relevant causal information that fosters artistic understanding, and artistic understanding subsequently informs further reading. For the sake of simplicity, the psycho-historical framework as depicted in Figure 1 in the target article is unidirectional, refers to the artwork as artifact and its reproductions, and does not consider the reception history. A fuller psycho-historical theory would be recursive, referring to memories (Hirstein) and the mechanisms of collective and individual agency that control the reception history of the work.

R3.2 Artistic design stance

Hypotheses about the design stance and essentialism made by developmental psychologists (Bloom Reference Bloom2004; Reference Bloom2010; Gelman Reference Gelman2003) belong to the core of the psycho-historical program (see also Gelman et al.). This choice is justified by the fact that these theories – which have been neglected by research in aesthetic science – take into account both philosophical and historical issues that are central to the psycho-historical program.

De Smedt & De Cruz observed that if knowing the art-historical context were necessary for artistic understanding, much early art could not be understood because historical information about earliest artworks from the Pleistocene is missing. We agree that this issue is important. The possibility of insurmountable difficulties or errors in the understanding of some artworks is compatible with our framework (sect. 3.3.1, R2.3). In their interpretation of the design stance, De Smedt & De Cruz propose that “some of the designer's intentions can be gathered non-inferentially through direct experience with prehistoric artworks.” Their proposal suggests that causal information in the artwork itself may sometimes suffice to understand the designer's intentions. Although this suggestion is in the spirit of the psycho-historical framework, it faces the challenges raised by Davies, Levinson, Ross, and Gilmore. As Levinson and Ross pointed out, it seems unlikely one can transparently track the past from the perception of artwork traces without the support of independently justified beliefs about the art-historical context (sect. R2.3).

Similarly, Fitch & Westphal-Fitch claim that it is “often impossible to reconstruct the agent behind an artwork, or the context in which it was produced” and think that our framework “would confine the study of aesthetics to those works for which historical information is available, mainly post-eighteenth century Western ‘high art.’” We disagree because our psycho-historical program can be deployed to study folk art and art from non-Western cultures. There has been growing academic interest in the history of oral cultures (Prins Reference Prins and Burke1991), decorative arts and crafts (Craig et al. Reference Craig, Kernot and Anderson1999; Dutton Reference Dutton1993; Green Reference Green2007; Vlach Reference Vlach1990), popular music and dance (Bohlman Reference Bohlman1988; Buckland Reference Buckland2006; Connell & Gibson Reference Connell and Gibson2003), and folk tales (Ògúnjìmí & Na'allah Reference Ògúnjìmí and Na'allah2005; Rölleke Reference Rölleke and McGlathery1991; Yassif Reference Yassif1999; Zipes Reference Zipes2006) in both Western and non-Western cultures. This suggests that Fitch & Westphal-Fitch's claim that “we cannot know the maker of these works” is too strong. Furthermore, even if some of such artworks were unintelligible to a particular audience, the audience of the artist's time and culture – and not just the elites – would have had some form of understanding based on their knowledge of their originative art-historical context, such as the religious, ritual, and political functions of the work (Boyer & Wertsch Reference Boyer and Wertsch2009; Rappaport Reference Rappaport1999).

Several commentators (Fitch & Westphal-Fitch; Schellekens; Thompson & Antliff) questioned the degree to which the art-historic context, and therefore the design stance, plays a role for art appreciation. Schellekens asked: “can we really assume that all artworks require us to take contextual information into account in exactly the same way?” Fitch and Westphal-Fitch assigned a minor role to the design stance and emphasized the role of the biological roots of artistic appreciation. We do not object to the hypothesis that there are biological roots of aesthetic preferences and biases, such as preference for symmetry (Jacobsen et al. Reference Jacobsen, Schubotz, Höfel and von Cramon2006; Reber Reference Reber2002; Rhodes Reference Rhodes2006), that may explain ornamental functions. Our point is that if eliciting aesthetic preferences pertains to the functions and meanings of a work of art, the appreciator's understanding of these functions is dependent on an examination of the relevant art-historical context and kinds.

R3.3 Artistic understanding

We agree with Gilmore's claim that “understanding and evaluation need to be disentangled and their relations of dependence identified,” and that “artistic understanding is a precondition of artistic evaluation, even if the two approaches proceed simultaneously.” Commentators differ, however, in the assessment of whether the normative mode of understanding is a necessary ingredient of the psycho-historical framework.

Leder noted that the core of this problem lies in the “unnecessarily normative pretense that art is only truly appreciated in the artistic understanding mode.” Gilmore provides a direct response to Leder. We agree with Gilmore that “a normative conception is required to distinguish the appreciation of art qua art from appreciation of it from artistically irrelevant points of view. According to a normative account of appreciation, an artistic evaluation can be distinguished from a mere liking or preferring by being answerable to reasons” (Gilmore; see sect. R1). Leder's opposition to the normative mode originates from the concern that it could widen the gap between the “two cultures” by making the empirical study of the arts more difficult, if not impossible, because of the singular nature of artworks. As discussed in section R2.1, however, scholars can conduct rigorous psycho-historical research on the sensitivity to historical individuals and kinds. Empirical research based on the psycho-historical program is possible (see R4), albeit challenging (Ross).

R3.4 Mental and brain processes

In agreement with Silvia, we think that the psycho-historical program can be integrated with appraisal theories of emotion (Ellsworth & Scherer Reference Ellsworth, Scherer, Davidson, Scherer and Goldsmith2003; Lambie & Marcel Reference Lambie and Marcel2002; Lazarus Reference Lazarus1991; Silvia Reference Silvia2005a). Mechanisms enabling appraisal of the relationship of the appreciator to art-historical contexts are likely to determine the nature of the appreciator's sensitivity and affective responses to expressive contents (Robinson Reference Robinson2005) or artistic intentions (Rollins Reference Rollins2004; Silvia Reference Silvia2005c). Furthermore, the appraisal of the art-historical context may enable the experience of emotions informed by artistic understanding (sect. 3.3.2).

Likewise, to contribute to a science of art qua art, research in neuroscience needs to present models of the brain mechanisms determining the appreciator's sensitivity to the art-historical context. We agree with Takahashi & Ejima's claim that findings on “contextual information processing in the human brain” could enable “empirical experimentation” on the sensitivity to art-historical context. For example, recent hierarchical models of functional organization of the prefrontal cortex (Botvinick Reference Botvinick2008) may serve as a framework for developing models of the neural mechanisms implicated in contextual reasoning triggered by the design stance and associated with artistic understanding.

Hogan, Silvia, and Thompson & Antliff comment on problems regarding fluency and expectation. On Reber's (Reference Reber, Shimamura and Palmer2012) account, fluency, though influenced by it, differs from expectation because fluency is a phenomenal experience, whereas expectation and prediction are symbolic processes. In addition, surprising fluency is positive, not fluency per se. Rollins remarks that there “is no reason to think that false beliefs inevitably cause dysfluency.” Reber and Unkelbach (Reference Reber and Unkelbach2010) provide a Bayesian analysis of why false beliefs are more likely to cause disfluency than accurate beliefs. Transgressions (Freeman & Allen) may be another example of inducement of disfluency leading to alienation effects.

Silvia wrote that the appraisal approach to emotion “is probably more fertile than the processing fluency approach” in research about art appreciation. Although we agree with Silvia's suggestion that the appraisal approach to emotion can be integrated into the psycho-historical program, we do not view the appraisal and fluency approaches as mutually exclusive. Artists manipulate a multitude of mental and brain processes to generate artifacts and categories with art-historical functions. Such processes range from basic processes in vision (Zeki Reference Zeki1999), audition (Bullot & Égré Reference Bullot and Égré2010; Thompson Reference Thompson2008), or processing fluency (Reber Reference Reber, Shimamura and Palmer2012) to context-sensitive processes of theory-based reasoning (Murphy & Medin Reference Murphy and Medin1985) and emotions (Hogan Reference Hogan2011; Silvia Reference Silvia2005b). In regard to the making of art-historical functions, such processes complement each other.

R4.1 Dependent variables

1. 1. We agree with Gilmore's and Ross' claim that measuring liking for studying appreciation of art qua art is misguided. For the choice of liking as a dependent variable tends to neglect the connections between art-appreciative processes and art-historical categories and functions, and thus amounts to committing a far-reaching aesthetic–artistic confound; see also Gilmore (Reference Gilmore2000; Reference Gilmore2011). For example, measuring how much undergraduate students like artworks cannot directly provide clear information about the modes and mechanisms controlling appreciators' sensitivity to art-historical functions. Even asking experts in a category of art whether they like an artwork is pointless if it remains unclear how modulation of liking is controlled by processes sensitive to historical kinds such as the design stance or artistic understanding. Furthermore, many art-historical functions of artifacts, if not all of them (Carroll Reference Carroll2002; Goodman 1968), derive from pictorial or semantic content that demand an interpretation rather than stimuli that trigger pleasure or liking. Therefore, an appreciator's liking is unlikely to indicate the appreciator's sensitivity to categories and functions in an art-historical context. Assessing judgments of liking, quality, or interest without a concomitant assessment of artistic understanding is likely to be irrelevant to the study of art.

   Given the prevalent use of liking as a dependent variable, we think with the benefit of hindsight that the criteria used in our target article for identifying studies meeting the criteria of the psycho-historical framework were too lenient.

2. 2. Very few studies on the influence of semantic context (sect. R4.2) measured dependent variables that probed sensitivity to art-historical contexts, such as meaningfulness (Russell Reference Russell2003), or understanding (Leder et al. Reference Leder, Carbon and Ripsas2006).

3. 3. How can the same dependent measure become relevant or irrelevant for measuring art appreciation? Studies by Takahashi (Reference Takahashi1995) and Smith et al. (Reference Smith, Bousquet, Chang and Smith2006) illustrate this point. Both used semantic differential scales as dependent variable. Whereas ratings in Takahashi's study measured the dependence between participants' categorical appreciation and artists' drawing intentions, Smith et al. (Reference Smith, Bousquet, Chang and Smith2006) have not explained how ratings in their study were sensitive to categories from an art-historical context.

R4.2 Independent variables

Chatterjee, Leder, and Vartanian & Kaufman directed our attention to studies that they interpret as consistent with our psycho-historical program. Leder argued that our target article “omitted a large corpus of existing research” that would develop psycho-historical hypotheses. The studies can be classified into two categories: (1) inquiries that manipulate the semantic context and (2) inquiries that examine the effects of expertise (Lindell & Mueller Reference Lindell and Mueller2011).

1. 1. Most of the studies that manipulate semantic context assess the effects of titles or descriptions on liking of an artwork without connecting this judgment to the cognition of art-historical contexts (Millis Reference Millis2001; Specht Reference Specht2010; Temme Reference Temme1992). However, even if we ignore the problem of liking as a dependent variable and turn to independent variables, many studies that manipulated semantic context did not manipulate art-historical information. For example, they presented metaphorical titles (Millis Reference Millis2001) and left open the way in which titles related to the art-historical context (Belke et al. Reference Belke, Leder and Augustin2006; Franklin et al. Reference Franklin, Becklen and Doyle1993; Leder et al. Reference Leder, Carbon and Ripsas2006).

   In another manipulation of semantic context, Kirk et al. (Reference Kirk, Skov, Hulme, Christensen and Zeki2009b) presented abstract paintings with the labels gallery or computer, indicating that the paintings belonged to a reputed art museum or were generated by the experimenter with a computer program. Behavioral and brain imaging data indicated higher hedonic value for paintings labeled gallery. The study is similar to the thought experiment with Warhol's Brillo Boxes analyzed in the target article. However, this study lacks the controls required to determine that the observed effects reflect manipulation of the art-historical categories, and not, for example, effects of monetary appraisal because abstract paintings in a reputed museum presumably cost more than paintings purportedly created by the experimenter (see Plassmann et al. Reference Plassmann, O'Doherty, Shiv and Rangel2008, on effects of monetary value of wine on hedonic value). Future studies would have to ensure that art-historical categories are not confounded with other, less relevant variables.

2. 2. Because experts possess more knowledge about art-historical categories and functions than non-experts, comparing the two groups should provide a means for probing appreciators' sensitivity to art-historical contexts. Does this entail that existing studies of expertise have already implemented a psycho-historical research program? We do not think so. Apart from the fact that most expertise studies assessed hedonic measures (Hekkert & van Wieringen Reference Hekkert and van Wieringen1990; Kirk et al. Reference Kirk, Skov, Christensen and Nygaard2009a) that may be irrelevant to the art-historical context (sect. R4.1), they pose at least two methodological problems:

   First, experts may like some artworks more than others not because of relevant artistic understanding but because they know which artworks have to be liked more if one is to count as an expert and connoisseur (Bourdieu Reference Bourdieu and Nice1979/1987).

   Second, experts may like and remember artworks better (see Kirk et al. Reference Kirk, Skov, Christensen and Nygaard2009a for a study on architects) not because they have become experts, but they may have become experts because they have liked and remembered artworks better from the outset. Experimental manipulation of historical knowledge may prove helpful to adjudicate this alternative (Kruger et al. Reference Kruger, Wirtz, Van Boven and Altermatt2004; Silvia Reference Silvia2005c; see sect. 4.1). In a study by Wiesmann and Ishai (Reference Wiesmann and Ishai2010), participants who were provided with more expert knowledge about cubism than the control group were better able to recognize the objects depicted by cubist paintings. This study meets the criteria of the psycho-historical framework because it provides the participants with art-historical knowledge and measures the recognition of objects in cubistic artworks, a dependent variable that might be more relevant to assessing sensitivity to an art-historical category than judgments of liking. By means of its manipulation of knowledge and use of non-evaluative variables, this study circumvents the problem that experts may provide responses that have to do with adherence to norms of a social class (Bourdieu Reference Bourdieu and Nice1979/1987), and that an observed outcome may be the cause instead of the effect of expertise.

3. 3. In conclusion, did we omit a large corpus of existing research on art appreciators' sensitivity to art-historical contexts in the target article? From the standpoint of a lenient criterion and a focus on independent variables alone (as we did in the target article), we concede that the target article overlooked a few studies that may meet the criteria of a psycho-historical framework (Russell Reference Russell2003; Smith et al. Reference Smith, Bousquet, Chang and Smith2006; Specht Reference Specht2010; Temme Reference Temme1992; Wiesmann & Ishai Reference Wiesmann and Ishai2010). These studies manipulate the appreciators' knowledge about the art-historical context in a way similar to the studies taken as examples in the target article (Kruger et al. Reference Kruger, Wirtz, Van Boven and Altermatt2004; Silvia Reference Silvia2005c). In contrast to Takahashi (Reference Takahashi1995), however, none of these studies manipulated the art-historical context directly.

Let us reiterate, however, that the aim of the target article was to propose an integrative research program and not to review advances in aesthetic science. The psycho-historical program entails that researchers in aesthetic science need to adopt stricter criteria for defining the science of art and overcoming the aesthetic–artistic confound (sect. R1.1; Gilmore; Ross). From the standpoint of strict criteria, where both the manipulation of the art-historical context and the dependent measure satisfy the criteria of the psycho-historical program, only the studies by Takahashi (Reference Takahashi1995), Russell (Reference Russell2003), Wiesmann and Ishai (Reference Wiesmann and Ishai2010), and Newman and Bloom (Reference Newman and Bloom2012) may qualify as psycho-historical. Therefore, regardless of whether we rely on lenient or strict criteria, we did not omit a large corpus of research.

R4.3 Misled by ahistorical universalism

The methodological commentaries by McManus and Graham illustrate the pervasiveness of the ahistorical universalism we criticize – see also Chatterjee (discussed in R2.1), Leder (addressed in R2.2), Locher (2012), and Martindale (Reference Martindale1990).

To vindicate his study a posteriori, McManus argues that Mondrian is an “anti-historical” and “anti-narrative” artist. McManus's commentary provides the kind of information about Mondrian's art-historical context that one would have expected to see discussed in his original article (McManus et al. Reference McManus, Cheema and Stoker1993; cited in sect. 4.2). McManus' outline seems to justify the thesis that Mondrian could be appreciated without any knowledge of the modernist art-historical context. However, both his thesis and his reliance on Krauss (Reference Krauss1979) can be challenged. Arguing that grids in modernist art have a “bivalent” structure and history, Krauss' (Reference Krauss1979) analysis responds to historical debates on the context of artistic modernity initiated by Greenberg (Reference Greenberg1961), Fried (Reference Fried1967/1998), and T. J. Clark (Reference Clark1973; Reference Clark and Mitchell1982; Reference Clark2001). Krauss is therefore thoroughly contextualist in her attempt to disclose the varied historical and psychological functions of grids in modernist art. Although debatable, her interpretation allows multiple interpretations of Mondrian's grids and does not endorse an aesthetic–artistic confound. In contrast, McManus' thesis that Mondrian's paintings “may encapsulate some universal principle of compositional order which can be detected by subjects” (McManus et al. Reference McManus, Cheema and Stoker1993) suggests the ahistorical view that appreciators have an innate or universal preference for specific types of organizations in grids, regardless of the art-historical context. This kind of statement implies an endorsement of the aesthetic–artistic confound and a neglect of the appreciator's sensitivity to Mondrian's modernist art-historical context.

Another example of the assumption of ahistorical universalism is found in Graham's commentary. In contrast to Gilmore, Silvia, and the psycho-historical program, Graham criticizes holistic methodologies from an ahistorical standpoint. We disagree with Graham's claim that the psycho-historical program entails methodological holism (see sect. R2.1). We think that the research on the non-randomness of Pollock's work he cited is irrelevant to the science of art because it assumes the validity of an ahistorical analysis of artworks. That said, we concede that “measurement of reduced properties of naturalistic stimuli can grant novel and unexpected insights – with respect to vision and to art” (Graham). Again, the point of our argument is that such research needs to offer models of the sensitivity to art-historical individuals, kinds, and contexts in order to contribute to a science of art qua art (see sect. R1; Gilmore; Takahashi & Ejima). This point is missing in Graham's discussion of artistic randomness.