The previous chapter dealt in part with neuroascesis as a technology of the self and with neurobics as a package of self-care practices allegedly based on scientific knowledge and capable of acting directly on its user’s brain. From mid–nineteenth century exercises for the double brain or the phrenological organs to the twenty-first-century brain gym, cerebral self-help has been a commercial industry. But even if the regimens it sells are widely shared, those who buy them pursue the exclusively personal goal of improving or maintaining themselves. Thus, insofar as neuroascesis and neurobics impel believers and consumers to adopt regimens for activities such as eating and exercising, they incarnate individual forms of “being brains.” In Chapter 3 we shall explore another context for the making of cerebral subjects: the neurobiologization of psychic distress. At the global and institutional level, such a context is more consequential than that of neuroascesis. Yet, contrary to the adoption of neurobics, which requires the voluntary acquisition of commercial products, the cerebralization of psychological suffering involves, as we shall see, choices and decisions by individuals and collectives, but it also affects subjectivities independently of those choices and decisions. It thus stands between, as it were, the world of neuroascesis and the more purely academic universe we shall explore here, namely that of the “disciplines of the neuro.”
The message that we are essentially our brains does not always require a deliberate commitment to regimens supposed to affect our brains directly or the unintentional and often unnoticed participation in mechanisms that shape us individually as cerebral subjects. Indeed, it may be independent from both and take the form of more or less professionalized and institutionalized programs of research and teaching in the human sciences. The actors of these programs do not necessarily implement the cerebralizing doctrine in their private lives. But what they advocate in their writings and investigative practices, though representing a minority viewpoint in their fields, has become a conspicuous element in the larger landscape of the neuro and the academic embodiment of the brainhood ideology. Moreover, by virtue of the “looping effects” the human sciences may have (Hacking 1995, 2006), it is conceivable that the disciplines of the neuro influence some aspects of subjectivities even beyond their actors’ scientific persona, that is, beyond a “cultural identity that simultaneously shapes the individual in body and mind and creates a collective with a shared and recognizable physiognomy” (Daston and Sibum 2003, 2).1
The disciplines of the neuro are varied but have several features in common. The most immediately visible for many of them is the way they are designated. Since the 1990s, intellectual and institutional projects whose names combine the prefix neuro- with the name of one of the human or social sciences have multiplied. An incomplete alphabetical list of the “disciplines of the neuro” or “neurodisciplines” (a collective we shall also sometimes call “neuroX”) may start with neuroanthropology, neuroarcheology, and neuroarthistory and finish with neurosociology and neurotheology; in between, we could place neuroeconomics, neuroeducation, neuroaesthetics, neuroethics, neurolaw, neuromarketing, neuropolitics, neuropsychoanalysis, and more. Such a list would be merely suggestive, since it would leave out older labels, such as neurophilosophy, as well as undertakings such as “affective” or “cultural neuroscience,” which lack the prefix neuro- yet are forms of the same pursuit. This chapter will focus on two of those fields, made up of the neuro approaches to aesthetics and to culture (in the anthropological sense).
To understand the neuroX it is not enough to describe their premises and promises. We have to look at what they actually do. That has been to some extent undertaken from the points of view of the anthropology, ethnography, and history of science as well as from that of the sociology of the professions and the scientific field. Our approach has instead been to range very broadly across the neuro spectrum—that is what allows some generalizations—but also to examine a few areas in sufficient detail to understand and analyze their inner “logic.”
We limited ourselves to areas that emerged during the Decade of the Brain and have therefore excluded those, such as neuropsychiatry and neurophilosophy, that were well established by the 1990s. The newer disciplines of the neuro share the basic premises and goals of the older ones. However, they constitute a major expansion of the range of application of neurobiological concepts and methods to problems that had been the traditional preserve of the humanities and the human sciences. It is the scale of the phenomenon sometimes characterized as a “neuro-turn” (e.g., Cooter 2014, Pedersen 2011), rather than particular instances, that makes it significant for the recent history of the cerebral subject. At the same time, concrete instances had to be chosen to show how that “turn” is carried through in scientific work and communication strategies.
Given the inherently political nature of the processes that shape persons’ lives, neuropolitics might have been an appropriate choice. The term seems first to have been used in 1977 by the psychologist Timothy Leary, the legendary American advocate of psychedelic drugs, to capture the belief that political problems can be traced to psychological issues based in brain chemistry. He declared: “The solutions to our predicament are neurological. We must assume responsibility for our nervous systems” (Leary 1977, 57), and he imagined recreating on that basis the relationship between individuals and the social order: “Young minds exposed to neurological freedom and the free spray of electronic information suddenly blossom like flowers in the spring” (Leary 1980, 33, 6). A less romantic notion emerges from William Connolly’s 2002 Neuropolitics, a cryptic Deleuzian manifesto that defines its subject as “the politics through which cultural life mixes into the composition of body/brain processes. And vice versa” (Connolly 2002, xiii). The predominant uses of “neuropolitics” since the 1990s correspond neither to Leary’s nor Connolly’s. Rather, in the spirit of the “neuro-turn,” the term designates a neurobiological (and predominantly neuroimaging) approach to questions of applied political science, such as people’s political attitudes or how they assess candidates and choose to vote. Similarly to neuroethics, which has been defined as the neuroscience of morality and the ethics of neuroscience, neuropolitics concerns both the neurologization of the political field and the study of the biopolitical implications of neuroscientific theories and practices. Considering such a double understanding, several ways in which the brain has been mobilized in politics can be identified (Meloni 2012 and Vander Valk 2012b offer taxonomies that partially overlap with ours).
One approach proposes to take “political affect” seriously in the framework of the “nonmechanistic materialism” it attributes to the contemporary neurosciences (Protevi 2009). Such “embodied” political theory tends to contradict the model of the “rational subject” said to prevail as the foundation for “neoliberal thought” (Lakoff 2008, 4). This brand of neuropolitics has been criticized for “honoring the neurosciences a good deal too much” and for “generously giving credit before actual research findings have emerged, and surely before alleged findings have settled into well-founded explanatory theories” (Slaby, Haueis, and Choudhury 2012, 60). A second approach emphasizes the role of empathy in political thinking and practice (Coles 2012; Olson 2008, 2013) and hopes that mirror neurons will provide an “Archimedean moral point from which to lever public discourse toward an appreciation of our true nature, which in turn might release powerful emancipatory forces” (Olson 2008).
A third approach focuses on the so-called political brain (Western 2008), mainly to anticipate voter behavior and preferences; it is an academic enterprise but also, as would be expected in an area teeming with consultants, a commercial one (Randall 2015). It too claims to show that political decisions do not rest on cost-benefit evaluations and argues (for example on the basis of studies on amygdala responses as detected by fMRI) that “the neural basis for political decisions extends across cultures” (Blank 2013, 269). A fourth approach, which largely overlaps with neuroethics, addresses the policy implications of neuroscientific research (see Blank 1999, 2013), for example in the context of the increasing social and economic costs of neurodegenerative disease in aging populations and the use of a range of interventions, from psychopharmacological enhancement to brain implants and neurogenetic procedures.
A fifth approach, which discusses the plastic brain in a political key, further illustrates the interpretive leeway the neuro provides. On the one hand, the widely covered notion of neuroplasticity is said to reflect the neoliberal emphasis on autonomy and self-responsibility; on the other, it is presented as designating a feature that allows us to “make” and “sculpt” our brains freely and “change our life” however we choose (Malabou 2008; see critiques by Pitts-Taylor 2010 and Rees 2011). Critical neuroscience can be seen as another neuropolitical approach. Assuming that “there is substantially more politics in the ‘neuro’ than there is ‘neuro’ in politics” (Slaby, Haueis, and Choudhury 2012, 64), it examines how “brain facts” are appropriated in different social domains, seeks to increase awareness of the factors that contribute to stabilize scientific worldviews and make them seem inevitable, and scrutinizes how people are “made up” according to brain categories. Finally, the neurodiversity movement is essentially a politics of the brain. Considering that what society usually classifies as psychiatric illness is a particular form of being derived from unusual brain “wiring,” the movement advocates the recognition of diagnosed individuals as different yet fully entitled members of society.
In Chapter 3 we examine the neurodiversity movement as a form of social action rooted in cerebralized understandings of subjectivity. Here, we explore and assess the potentially transformative action of the neuro by focusing on disciplines chosen because they derive from classic fields that played historically foundational roles among the human sciences: on the one hand, aesthetics and art history (by way of neuroaesthetics); on the other hand, and in order to tip the balance in the direction of social science fieldwork and non-Western societies, the disciplines of culture (by way of neuroanthropology and cultural neuroscience).
The NeuroX: An Overview
“Neuroskeptics” have taken the neuroX as the premier instance of the neurobabble, neuromythology, neurospeculation, neurotrash, neuromania, or neuromadness of the late twentieth century.2 However, if madness there is, it has logic in it, and it is precisely this logic we wish to examine. Indeed, in spite of their diversity, the disciplines of the neuro may be considered as a single constellation held together by a set of core common beliefs and foundational elements. We first outline those commonalities, derived inductively from countless articles and books across the neuroX spectrum.
First, the neurodisciplines postulate that “the mind is what the brain does.” This, however, does not function as a working assumption but as a general and established truth. It follows (implicitly) that only cerebral mechanisms are truly constitutive of the studied phenomena, while the rest, including all aspects of culture and society, only “modulates” universal neurobiological processes. Second, the disciplines of the neuro have a common goal, namely to discover neurobiological “foundations” or “substrates,” and assume that uncovering the processes that “underlie” the phenomena they study will unveil causes and thereby provide explanatory depth. The preferred tool to pursue such a goal is neuroimaging, particularly fMRI (functional magnetic resonance imaging); this is a third common element. Fourth, as a consequence of the predominant methodology, the results of neuroX research come mainly in the form of neural correlates of the investigated processes and behaviors; these correlates are said to reveal the brain structures “involved” in those processes and behaviors.
Nevertheless (fifth feature), given the correlational nature of the data and other features of fMRI, the neurodisciplines cannot explicate the meaning of the data and the role of the identified structures. This intrinsic limitation conflicts with a widespread desire for causality (a sixth common feature), which manifests itself in frequent interpretive slippages from correlations to causes. Hence (a seventh aspect) the great disparity between the apparent rigor of the methodologies on the one hand and on the other hand the range, generality, and speculative character of the programmatic statements and discussion of results.
More than the most widely shared utensil of the neurodisciplines, neuroimaging (in its well-documented overhyped mode) has been their condition of existence.3 It has enabled their emergence and sustained their development but has also been a major source of their drawbacks. We arrive here at two further shared characteristics. One (the eighth of our list) is the impression of irrelevance the neurodisciplines give, insofar as they seem incapable of providing a nontrivial answer to the question So what? with regard to their empirical findings. The other (ninth feature), which emerges from and epitomizes the others, is that the neuroX are not merely, as others have argued, irrelevant or ideologically problematic: they are basically self-defeating.
By “self-defeating” we mean that the assumptions and methods of the neuroX are inadequate to study the topics they ostensibly want to understand. One may object that they redefine their objects in ways that make them amenable to neuroscientific inquiry. This is perhaps so in some cases, but even then the objects turn out to be so transformed that they end up having little to do with what they originally were said to be or to embody. Whatever they are about, fields such as neuroaesthetics or cultural neuroscience are definitely not about the things, such as art, beauty, or cultural difference, whose neurobiological foundations they claim to be looking for. Yet, ironically, the disciplines of the neuro convey the belief (which they do not formulate explicitly) that revealing the neurobiological “substrates” of a phenomenon not only means knowing it “objectively” but also confers upon it the ontological consistency it allegedly lacks when treated with the tools of the human and social sciences.
These simultaneously epistemological and ontological persuasions bolster another common characteristic of the neuroX, namely (despite allegations to the contrary) that they place the neurobiological approach at the top of a hierarchy of forms of knowledge. The brain sciences are presented as dealing with the fundamental level, with the mechanisms that ultimately account for phenomena the human and social sciences merely describe and interpret. This takes us back to the premise that humans are what they are and do what they do because of the way their brains function. That premise justifies the methodological and interpretive preeminence of the neuro—even if it should be obvious that it is one thing to admit that we cannot create culture without our brains and another to posit that culture is a “product” of the brain and that therefore the neurobiological is the most essential and ultimate level of analysis for cultural phenomena.
From the hierarchy of approaches and modes of explanation follows a lack of attention to the productions of the humanities and social sciences—a neglect, most flagrantly reflected in the dearth of human science and social science research from the bibliographies of neuroX publications. To engage in neurodisciplinary activities (but also to study and eventually to critique them), scholars from the humanities and human sciences must become conversant with relevant neuroscientific conceptual, technical, and empirical issues; with very rare exceptions, the opposite does not happen. Such an asymmetry is understandable: Why would you study, say, philosophical aesthetics or theories of culture if those approaches are precisely what you want brain science to supersede? And why bother with historical details if you are convinced that history is fundamentally a consequence of brain processes? Even though the disciplines of the neuro pursue intellectual agendas dictated by philosophy and the human sciences, at the level of practice they neglect and even disdain them.
Finally, the neurodisciplines are usually expected to make conceptually, empirically, and methodologically positive contributions to the corresponding human sciences, and they are sometimes given the redemptive function of driving them out of their supposed crises and dead ends. (The art historian Norman Bryson [2003, 14], for example, found in the neural turn distinct advantages “over the broad family of accounts of the real that are based on the primacy of the signifier”—accounts said to have disconnected the human sciences from material reality and pushed them to the edge of irrelevance.)
The Case of Neuroethics
Neuroethics occupies a special place in this landscape. Unheard of in the 1990s, it required a few years in the early 2000s to become a recognized, autonomous, and professional discipline, with its own networks, platforms, societies, journals, academic centers, and teaching and research programs (Conrad and De Vries 2011, Hoyer 2010). Although it has been described as covering two related domains, the neuroscience of morality and the ethics of neuroscience, neuroethics is almost exclusively reserved for the discipline that examines the implications of brain science for society as well as the actual and anticipated ethical, social, political, and legal consequences of neuroscientific knowledge and its applications, including the ethics of neuroscientific research. By 2015, neuroethics publications amounted to hundreds of articles as well as (to mention only the English language) single-authored books (Levy 2007, Racine 2010) and seven multiauthored volumes, including two massive handbooks, three anthologies, and two collections of original essays (Chatterjee and Farah 2013, Clausen and Levy 2015, Farah 2010a, Giordano and Gordijn 2010, Glannon 2007, Illes 2006, Illes and Sahakian 2011). To this abundance corresponds an extremely broad range of inquiry, from the detailed empirical examination of particular cases, contexts, and circumstances to far-reaching ethical and philosophical reflections. We shall not deal in detail with this thriving, expansive, and sophisticated field but only suggest its function among the disciplines of the neuro.
To a very large extent, the existence of neuroethics as a research program with normative aspirations has thriven on the expanding application range of functional neuroimaging and therefore depends on the assumption, considered valid, that “imaging the brain provides information about the mind” (Farah 2010b, 4). Neuroethics fully shares such an assumption, which it reinforces even as it critically analyzes some of its instantiations. It may welcome resistance to “explanatory reductionism,” but at the same time it does not examine the premise that since “thought and behavior cannot be separated from their neurobiological underpinning,” the “empirical correctness of how the brain enables behavior is necessary to make sense of how we think and act” and therefore that people must be understood as behaving the way they do “because of their brains” (Glannon 2011, 191–192).
The usually implicit way neuroethics interprets separated, make sense, and because (to take the reasoning we just quoted) buttresses the features we identified as being common to the disciplines of the neuro and (intentionally or not) turns it into a protective barrier for their funding interests, constitutive beliefs, and methodological options. Whoever recommends, as the neuroethics group set up within the U.S. BRAIN Initiative, “Avoid Hype, Overstatement, and Unfounded Conclusions,” should also examine in great detail the claim that neuroscience “presents an unparalleled opportunity to gain a deeper understanding of the human brain and mind, including our cognition, behavior, memory, learning, mood, and social interactions”—which is instead taken for granted (Presidential Commission 2015, 9, 2). Such an “organic” function of neuroethics appears in its most subtle yet conspicuous aspect under the form of “proactive ethics.”4
For example, a survey among health providers and patients diagnosed with major depressive disorder found “high receptivity to brain scan for treatment tailoring and choice, for improving understanding of and coping with disease, and for mitigating the effects of stigma and self-blame” (Illes et al. 2008). The authors recognize that, as regards depression, there is no “translation” of neuroimaging research to the clinic and observe that “fMRI is still a long way from being used in individuals.” They nonetheless justify their work by the “rapid innovation” that, they claim, is taking place along the “trajectory of discovery to implementation.” The authors wish to promote “the development of responsible social and public policies in response to new diagnostic and prognostic capabilities for the benefit of patients and their families” and explain that the early identification of future challenges related to the clinical use of imaging technologies “can maximize benefit and prevent false hope, mitigate hype, and curtail their premature use and even misuse in the private sector.” However, by riding the “translation” hype (discussed below), unequivocally advertising the assumption that “fMRI promises significant benefit to the diagnostic process for major depression,” and confidently announcing the future realization of that promise without hinting at the doubts it has inspired, the study accomplishes just the opposite.
Neuroethics thus further legitimates itself. The consolidation of the discipline involved not only a vigorous research program, a strong media presence, and an extraordinarily rapid institutionalization but also claims to a “neuroethical exceptionalism” that rationalized its being separate from bioethics (see in particular Illes and Racine 2005 and the responses by Buford and Allhoff 2005, Doucet 2005, Schick 2005, Wilfond and Ravitsky 2005). The debates around such claims highlight the extent to which the stakes of neuroethics’ disciplinary autonomy are not merely professional or financial but, at least at the level of explicitly given reasons, primarily ontological. Indeed, the autonomous character of neuroethics has been justified by an appeal to neurotechnology’s capacity to modify the interaction between personal identity, responsibility, and free will more radically than genetics, and thus to transform our ideas about personhood, as well as by the “hypothesis,” which appears both “inevitable” and “omnipresent,” according to which “the mind is the brain” (Illes and Racine 2005, 12). More examples of similar assertions could be given.
Neuroethics has succeeded in being recognized as distinct from bioethics by defending its claims to be exceptional on grounds that represent ontological beliefs offered as empirical fact. Indeed, the discipline’s special status is said to derive from the “intimate connection” between brain and behavior, the “peculiar relationship between our brains and our selves,” and “the intuition that our ever increasing understanding of the brain mechanisms underlying diverse behaviors has unique and potentially dramatic implications for our perspective on ethics and social justice” (Roskies 2002, 21). Neuroethicists claim that, by virtue of those same connections and intuitions, the neurosciences will end up redefining “our sense of selfhood and brain-body relations” (Wolpe 2002, 8), radically transforming ancient philosophical questions, and giving rise to unheard-of moral and legal challenges (Illes and Racine 2005, 6). In spite of considerable criticism (some of it emanating from neuroscientists and neuroethicists) of the hype surrounding the neuro, in 2015 the Oxford Centre for Neuroethics still presented itself as follows:
Neuroscience has made enormous advances in recent years, challenging our traditional understanding of consciousness, responsibility, well-being and morality. Our newfound knowledge of the brain and the mind undermines previous beliefs about a number of areas of private and public life, including addiction and its treatment, criminal responsibility, the treatment of vegetative patients, medical decision making and the enhancement of normal human capacities. It also raises a new question: what are the moral limits of the use of such technology? Neuroethics is a new discipline, addressing these urgent issues.5
This paragraph illustrates how the self-promotion strategies that brought about and have successfully maintained neuroethics contribute to create the phenomena the discipline deals with, thus confirming the observation that neuroethics is part of the problem it supposedly seeks to address (Singh and Rose 2006, 100).
This observation, however, could be made differently: The unique ability of neuroethics has been to give itself as one of its fundamental tasks the analysis of the assumptions that legitimate its existence. For instance, an influential study of fMRI “in the public eye” (Racine, Bar-Ilan, and Illes 2005) identified in the media’s handling of neuroimaging research three main features: neurorealism, which refers to “how coverage of fMRI investigations can make a phenomenon uncritically real, objective or effective in the eyes of the public” (160); neuroessentialism, or the belief that neuroimaging has direct access to the mind and thus to what we “really” think and feel; and neuropolicy, a term to designate attempts to use fMRI to influence public policy. The method is flawless and the analysis, insightful, but the distancing from those positions is superficial.
In short, having instituted their discipline by arguing in favor of neuroexceptionalism, neuroethicists established themselves as competent for guiding both the public (including policy makers) and the scientists. To manage that identity, they must remain close to the sensibilities of both; they must be farsighted enough to capture the former and prudent enough to be trusted by the latter. They have succeeded on both counts, feeding public hopes and fears while adhering to the basic tenets of the neuro and supporting its advance. Indeed, neuroethicists have proclaimed that brain imaging will renew beliefs about human nature and will provide new and scientifically grounded responses to questions that have been “traditionally” examined by philosophy and the human sciences (Illes, Racine, and Kirschen 2006). In so doing, they have placed their field at the very heart of the neuro enterprise.
To recapitulate, in spite of the diversity of topics they deal with, the disciplines of the neuro share a complex of interconnected features:
Such a list will surely strike some readers as an unfair overgeneralization. And indeed, not every item applies equally to every single neuroX production. Moreover, speaking of a neurodisciplinary field or the neuroscientific turn risks obscuring not only the great diversity of topics but, more importantly, the variety of “goals and dynamics” proposed for articulating the brain sciences and the human sciences, from desirability of mutual influence to complementary interaction to far-reaching integration (Beaulieu 2012, 156). Nevertheless, other than the fact that our list results from examining a vast number of publications across all neurodisciplines, generalizing here is a matter of scale and perspective. In their introduction to The Neuroscientific Turn, Melissa Littlefield and Jenell Johnson (2012, 9) note that, beyond the differences among the neurodisciplines that make up the “turn,” the neuro “signifies a hypothetical location (i.e., the nervous system, brain, neuron) where we should look for answers to our deepest questions about consciousness, learning, selfhood, and so forth.” That is the core belief that brings all neuroX together and from which the features we have identified ultimately derive. There are in this regard no exceptions.
What Are Humanists Afraid Of ?
Does criticism or repudiation of that core belief and its corollaries manifest “fear” of the consequences of placing humans among animals? We don’t think so. The claim that “many from the social and human sciences [let us call them ‘humanists’] react with horror to this suggestion that our specifically human social capacities have neurobiological bases” (Rose 2013a, 15) may have rhetorical appeal, but it is by and large a misrepresentation. Of the numerous critics of the neuroscientific turn we have read, heard, and personally met, none “from the social and human sciences” have either reacted “with horror” or fear being considered biological creatures. In fact, none to our knowledge deny that humans are products of evolution and that what they are and do has neurobiological bases. Their point (and ours) is rather that neuroimaging methods and the neurobiological level of analysis are not always and ultimately the most appropriate to account for human phenomena. That is why some of the most relevant critics of the neuroscientific turn in the human sciences are brain scientists.
For example, as a basic neuroscientist with experience in the area of molecular biochemistry and biophysics, Susan M. Fitzpatrick is well prepared to examine whether functional neuroimaging methodologies offer “a constructive path for the neuroscientific turn in the humanities and social sciences” and, reciprocally, whether the typical neuroX topics are “actually well suited for study by functional brain imaging” (Fitzpatrick 2012, 180). To both questions, her answer is a clear “no.” (After fifteen years as vice president, Fitzpatrick became in January 2015 president of the James S. McDonnell Foundation, and her position has presumably informed the foundation’s decision not to fund projects “proposing to use functional imaging to identify the ‘neural correlates’ of cognitive or behavioral tasks.”)6
Now, is Fitzpatrick afraid of acknowledging the neurochemical bases of human behavior and therefore of seeing humans as biological creatures? Here too the answer is “no.” Her reasons for considering the neuroX misguided are of a different order. First, as she notes, “the findings rarely reveal novel insights about brain structure-function relationship … but are usually reinforcing psychological findings” (Fitzpatrick 2012, 182). Indeed, most neuroimaging research (in the neuroX) superficially enhances already established findings without illuminating them further. In the worse cases—and they are numerous—the neuro is merely tacked onto the discussion of mental phenomena. It may well be that Bernard of Clairvaux’s mystical visions “were likely mediated by the effects of sleep fragmentation on the prefrontal cortex,” while Teresa of Avila’s “by the inhibition of the thalamus” (Hendrix and May 2012, 116). But such speculations bring no insight into mystical experience, which is the purported object of study.
We recognize our own frustration in the impression Fitzpatrick (2012, 183) gained from reading neurodisciplinary papers, that “we learn very little of substance beyond what was already known from the cognitive psychological or behavioral studies—except for perhaps some converging evidence that behavior is accompanied by brain activity.” And even such “evidence” is pointless. Discovering that believers and nonbelievers display different patterns of brain response when they listen to prayers simply confirms that listening to prayers is not the same thing for the two groups. When “beauty” is operationalized as hedonic judgments (where subjects choose between like or dislike), then a correlation between “beauty” and the “activation” of brain areas that govern pleasure and positive reward is only to be expected. Describing the neurobiological correlates of the mental and bodily effects of meditation may have scientific interest, and “mindfulness neuroscience” will perhaps uncover their “underlying mechanisms” (Tang and Posner 2013). But to say that showing how mindfulness-based stress reduction alters gray matter density “demonstrates that changes in brain structure may underlie some of these reported improvements and that people are not just feeling better because they are spending time relaxing” (Press Release 2011) is both trivial and immodest: changes in the brain necessarily and always “underlie” the effects of practices (and that includes relaxing), yet those effects do not become any more real when they are shown to have neural correlates (see Walton 2015 for a self-contented catalog of the ways “meditation actually changes our brain”).7
Second, the primary problems with fMRI use in the neuroX are not technical and will not be solved with improved technology for the same kind of research. They are conceptual. As Fitzpatrick (2012, 183) emphasizes, “most neuroscientific-turn questions are framed assuming we already understand how neural substrates serve cognitive functions.” In our list, such framing shows up as the slippage from correlations to causes and in the vocabulary of neural substrates and underpinnings that sustains it. That vocabulary, the usual one in neurodisciplinary publications (Schleim and Rosier 2009), gives a wrong idea of what imaging may contribute. Fitzpatrick (2012, 188) remarks that the interpretation of the changes in neuronal activity that neuroimaging brings to light “is meaningless without a theory of the cognitive operations involved in performing the experimental tasks, and a well designed set of tasks to test the theory,” and she concludes: “Attempting the ‘neuro-turn’ via functional neuroimaging is most likely a wrong turn” (194).
In short, it is definitely not because humanists deny the existence of neurobiological bases of behavior or refuse to be considered biological creatures that some of them react critically to the neural turn. Is it then, as Nikolas Rose (2013a, 15) has also claimed, because they “feel that their space is being colonized, their expertise displaced”? Again, no, even though such feelings may arise in some situations. For example, in 2010 “neuro lit crit” was presented as the “next big thing in English,” as “the cutting edge of literary studies,” and as an approach that might “save humanities” (see here Chapter 4). Perhaps such claims generate among literary scholars a sense of “colonization” and “displacement.”
However, even when those terms make sense, the bottom line has less to do with fears of oppression or suppression than with points 11 and 12 of our list, hierarchy and neglect. As for hierarchy, the only feature that makes neuro lit crit “cutting edge” is the neuro label. The alliance with “science” moves literary studies closer to the top of the epistemic hierarchy—as if other methods provided no genuine insights, as if we hadn’t realized that reading literature may be worthwhile before brain-scanning technology existed, or as if assessments of cultural value could be brain based. The humanist’s discomfort derives from a clash of views about knowledge and method. Not all humanists are doggedly attached to the “ontological primacy” of the sociocultural or build their “neuro-critique” on that presumption (Fitzgerald and Callard 2014, 8–9). Even those who are not willing to give away all reciprocity or to engage in “interdisciplinary collaboration as a practice of subjugation” (Callard and Fitzgerald 2015, 96) recognize that there are differentials of power, and anyone who has admired a teacher knows that these differentials can be productive. Rather, much more basically, the unstylish humanist has concluded, on the basis of solid evidence, that the neurosciences, and neuroimaging experiments in particular, have hardly anything to offer when it comes to understanding the complex phenomena usually associated with cultural and historical research.
The neuro stance toward existence and knowledge can be characterized in terms of levels of analysis and interpretation. Studies of the “seductive allure of neuroscience explanations” showed that adding irrelevant neuroscientific information to an argument makes it more persuasive than it would be without it and that people are more likely to agree with scientific claims when they are supported by a brain image than when they are accompanied by another type of image (Skolnick Weisberg et al. 2008, McCabe and Castel 2008). After having been criticized (Farah and Hook 2013), those results have been corroborated (and the cycle is likely to continue). Superfluous neuroscience information was again found to make explanations of psychological phenomena more appealing (Fernandez-Duque et al. 2015), and evidence from neuroimaging research turns out to be perceived as more convincing than evidence from behavioral science research, especially among subjects motivated to disbelieve the evidence (Munro and Munro 2014). Perhaps most importantly, as historical and sociological studies demonstrate (e.g., Dumit 2004, Hagner 2009, Joyce 2008), the power of images is not purely a matter of individual psychological response but a social phenomenon embedded in contexts, ranging from business and technology to academia and epistemology, of which the neuro industry is itself a part.
NeuroX practitioners may, like everyone else, fall prey to the “illusion of explanatory depth” (Rozenblit and Keil 2002). This effect is particularly strong in connection with causally complex systems such as technological devices and natural phenomena. Studies of people’s posture toward their own explanatory knowledge about an object or phenomenon have found that the best predictor of overconfidence is the ratio of visible to hidden parts—in other words, people are more convinced that they know how something works when they can easily see or visualize its components. This finding is consistent with the persuasive power of neuroimages and the conviction that they demonstrate neuronal activations. Overconfidence also reveals a “level of analysis confusion”: functions are taken for mechanisms, and having knowledge at one level of explanation makes us feel that we also have it at another. We thus generate causal claims and get the impression that we grasp the mechanics of the phenomena involved. We are not saying that neuroX practitioners pretend to have full mechanistic knowledge of the processes they study but only that their threshold for acceptance of evidence and causation is such that simply affirming on the basis of correlations that a brain structure is “associated” with a phenomenon or that it is “involved” or “plays a role” in it stands in lieu of analysis or explanation.
It has been argued that functional neuroimages can support causal claims about brain function and are therefore not “merely correlational.” Indeed, neuroimaging research relies on the “reasonable expectation that areas activated concurrently with task performance are likely to be causally involved in the task,” even if that causal involvement “is no guarantee that the region implements the computation of interest in a given experiment.” Moreover, neuroimaging experiments give information “about the influence of behavior on brain activity—which is not identical but unquestionably relevant to influence flowing the other way” (Weber and Thompson-Schill 2010, 2415). Because of the correlational nature of its results, functional imaging cannot by itself establish causal effects of brain activity on behavior. Nevertheless, if a stimulus or experimental situation S consistently elicits a pattern P of brain activation, then S must somehow be the cause of P. This means, as Colin Klein (2010, 275) has argued, that the evidence of neuroimaging does not reside in the images it produces but that its findings may nonetheless “point us to where the evidence for functional hypotheses might be.”
We shall return to these matters; here we only emphasize that the neurodisciplines are not alone in their incapacity to attribute causal efficacy to the factors involved in their experiments. The problem is that their language and entire outlook suggest otherwise. By placing brain processes at the beginning of causal chains, they give those processes overall primacy and drive the metamorphosis of explicit correlation into implicit causality, of stated probabilities into implied causes. The hierarchy embodied in those methodological and epistemic moves is unwarranted, regardless of eventually “colonizing” intentions.
The second factor at the root of what has been tendentiously depicted as the humanists’ fear of oppression or suppression is encapsulated in point 12 of our list: neglect. Again, aside from individual intentions, the objective disregard of research in the humanities and social sciences is a structural effect of the epistemological and ontological positions just sketched. The problem is not “expertise displaced” but expertise neglected—and it is particularly objectionable given the neurodisciplines’ constant appeal to interdisciplinarity, collaboration, and “bidirectionality.” Of course not every project and publication manifests it in an equally grievous way, but the tendency is clear.
In the life sciences at large, there seem to be fruitful two-way movements. In one direction, a “biohumanities” model proposes to go beyond passive acceptance or mere commentary of the implications of the life sciences and produces research that feeds into those sciences and our understanding of them (Meloni 2013). In the other direction, biology “becomes social” (Meloni 2014). Evolutionary theory and molecular epigenetics blur nature/nurture boundaries and place environments and cultures center stage; the neurosciences no longer see the brain as an isolated data processor but as an organ both largely shaped by the external world and designed to create social relationships. The neuroscientific turn looks like a major protagonist of such integrative motions, since it is largely made up of projects that aim at developing shared conceptual and methodological frameworks and transcending disciplinary boundaries. Yet translating these good intentions into good deeds has proven singularly arduous.
As our cases will illustrate, the main reason for that difficulty lies not (as often claimed) in the fact that the neurodisciplines “are still young” but in their internal logic, which derives from the assumption that all manifestations of the human are best understood as products of the human brain. This assumption precludes any true bidirectionality. The human sciences are crucial insofar as they provide topics and questions: For the areas we shall examine, these might be “What is beauty?” or “What accounts for intercultural differences?” But then these topics and questions are neuro operationalized in ways that tend to ignore the scientific production of the source disciplines and to cancel the meanings they have in domains such as aesthetics or anthropology. Instead of mutual exchange, asymmetry and imbalance emerge as the key features of the neurodisciplinary setting. This is embodied not only in how the neuroX handle the concepts and the intellectual output of the human sciences but also in the concrete interactions between humanists and neuroscientists.
Formal testimony is rare, but a study documenting from the inside a transdisciplinary fMRI lie detection experiment brings to light some of the basic psychological and interpersonal features of those interactions. The first element of asymmetry is that the humanists and social scientists quickly ended up thinking “like neuroscientists, rather than like transdisciplinary deployed neuro-collaborators” (Littlefield et al. 2014, 8). This is only natural. In spite of the ideal of bidirectionality and the discourse of transdisciplinarity, the research is a neuroimaging experiment. It is therefore the humanists, not the neuroscientists, who experience “disciplinary double consciousness.” Positions and relations within “neuro-collaborations” are not symmetric or reciprocal: Even if by definition they take place “between neuroscientists and other scholars from a variety of disciplines,” they remain focused “on the central nervous system” (9). Such a focus necessarily structures the interactions and generates a hierarchy where the neuro is the dominant culture. This happens even in circumstances, like those of the experiment mentioned here, in which humanist participants felt ambivalence or had a critical attitude toward the very possibility of applying neuroimaging to lie detection (Fitzgerald et al. 2014).
Within neuroscience, the “translational imperative” (Harrington and Hauskeller 2014), which demands that research be applicable in the form of products and therapies, is in practice driven by a “promise of porosity,” by the expectation that, one day in the distant future, laboratory work will lead to clinical interventions (Brosnan and Michael 2014); against rampant hype in the area, John Ioannidis (2015, 39) has called for failures and negative results to be recognized “probably as the most useful outcomes that translational research efforts can offer.” But nobody pursues that as a goal. Neuro collaborations proclaim a commitment to inter- or transdisciplinarity, and when failure arises, they urge more voguish “entanglements.” Such commitment, mandatory for a research proposal to have chances of being funded, has (in the neurodisciplines) been by and large intellectually unsuccessful. In practice, it is mainly rhetoric and promises, and in the end, only the neuro remains.
A Word on Neuroimaging
Contrary to what any responsible neuroscientist could explain, in 2007 an official document of the American Psychological Association, the world’s largest professional society in the field, affirmed that fMRI “isn’t quite a mind reader, but it comes close,” declared that it “produces movies starring the brain,” and added that psychologists and others “aren’t using fMRI just to see what lights up in people’s brains as they perform different mental tasks” but also “to help answer classic questions within psychology” (APA 2007). This is merely one example in thousands of the hype and the misleading publicity that have sustained the growth in the range of application of fMRI. We do not mean to downplay the extraordinary achievement of having developed such a tool nor ignore its potential or deny the existence of many scientifically significant results. Any honest specialist would nonetheless have to concur with Bruce Rosen, one of the world’s leading fMRI experts, that over twenty years after its invention, the technology has not “changed the world” in the ways suggested in the APA document and assumed and proclaimed in many other contexts (Rosen and Savoy 2012).8
fMRI has left a modest mark on presurgical planning, treatment evaluation, and clinical assessment; it has been widely adopted as a tool for research into mental illness but plays no diagnostic role and has not influenced psychiatric care. Although it has transformed cognitive neuroscience, where it is the most frequently used technology, the gargantuan increase in the quantity of neuroimaging papers is not universally seen as having been accompanied by a proportional contribution to the understanding of the mind; while, as we shall see, it has been extensively applied to issues related to personhood, society, and culture, its most significant impact in that connection has been to drive the emergence and justify the existence of the neurodisciplines we shall discuss here. Its major role and effects have been cultural, sociological, and economic. Neuroimages have become contemporary fetishes whose power derives from the belief that fMRI offers “windows into the mind” (the metaphor is widespread). The technology has nourished the growth of various commercial and academic industries; it has dramatically contributed to “neurologize” research into the human (and the criteria by which it is evaluated) and has been occasionally misused for worthy purposes (as when it was wielded as evidence that mental illness is organic, decisively advancing the adoption of the U.S. Mental Health Parity Act of 2008).
As far as the neurodisciplines are concerned, much more than a tool, fMRI has been a condition of existence. The main empirical form of the neuroX consists of applications of fMRI to a range of humanistic and social science issues, and insofar as the availability of this form of neuroimaging has been the engine for the rise and growth of the neurodisciplines, the neuroscientific turn has been technology driven. Nevertheless, the inherent flimsiness of the neuroX does not derive from their dependence on a particular technology but from their premises, regardless of the particular imaging methods they might favor at any point. This distinction is not always clear.
One can take measures to face the “crisis of confidence” in the replicability and reliability of published neuroscientific findings, try to diminish the impact of excessively small samples and the mind-boggling proportion of false-positive findings (Eklund, Nichols, and Knutsson 2016; Miller 2016), and in general call for vigilance with respect to the uses of neuroimaging and the interpretation of its results (Boekel et al. 2015; Boekel, Forstmann, and Wagenmakers 2016; Button et al. 2013; Rachul and Zarzeczny 2012; Whelan and Garavan 2014).9 Reality checks, however, are not incompatible with adhering to the neuroX project and trying to solve “from the inside” problems such as the abundance of false positives, the cover-up of negative results, or many investigations’ lack of validity and statistical power (see for example the “nine ideas for a better neuroscience” by Bareither, Hasler, and Strasser 2015).
These shortcomings can be overcome, but it is not because of them that the neurodisciplines are, as we said above, self-defeating. Nikos Logothetis (2008, 876–877), a specialist in the neural mechanisms of perception, director at the Max Planck Institute for Biological Cybernetics, and one of the most authoritative neuroimaging experts worldwide, has pointed out that the limitations of fMRI “are not related to physics or poor engineering, and are unlikely to be resolved by increasing the sophistication and power of the scanners; they are instead due to the circuitry and functional organization of the brain, as well as to inappropriate experimental protocols that ignore this organization.” In some areas, fMRI can be certainly used “for gaining insights into brain function and formulating interesting and eventually testable hypotheses” (877). However, as Logothetis’s observations about brain organization and experimental design make clear, it cannot be an adequate tool for investigating the phenomena the neurodisciplines claim to study.
While the availability of functional neuroimaging has driven the neuroX, their existence also depends on the principle (mentioned above in connection with neuroethics) that “imaging the brain provides information about the mind” (Farah 2010b, 4). Historically, such a principle has been often associated with the belief that brain images allow some form of “mind reading.” The boom of optical instrumentation that began in the mid–nineteenth century gave plausibility to brain/mind-reading fictions. With the development of electroencephalography in the early 1930s, some researchers became convinced that the recorded waves would offer direct insights into mental life (Borck 2005). Five decades later, after a period of relative “iconophobia” and the spread of brain-as-computer models, the rise of digital brain imaging technologies has given mind-reading hopes a new lease of life (Hagner 2009).
As far as fMRI is concerned, the assertion that “imaging the brain provides information about the mind” implies that correlations between brain activity and the realization of a particular task during scanning capture a relationship between mental states and brain states that somehow illuminates the former. How does this work? During the control and experimental conditions, fMRI registers a hemodynamic signal known as BOLD (“blood-oxygen level-dependent”), whose intensity depends on local glucose metabolism. The precise relationship between neuronal activity and the BOLD signal is still under investigation, but it is widely assumed that the signals correlate with neuronal or local synaptic activity. It is in any case a surrogate that does not straightforwardly “reflect” that activity, and it is not clear “how it is coupled to the underlying neurophysiology, and how this coupling varies across the brain, across tasks and across individuals” (Singh 2012, 1121). The main output of neuroX studies consists of BOLD signals that correlate at a statistical level judged significant with responses to the tasks subjects carry out inside a scanner.
Through complex statistical procedures, the signals are ultimately transmuted into the familiar images, displaying the brain areas colloquially said to become “active” when subjects perform the experimental task. Even when they look like realistic reproductions of an entire brain (a gross external view, either flat or three-dimensional) or of traditional brain sections (sagittal, coronal, or transverse), fMRI images neither apply optical principles nor reproduce anything: they are “belief-opaque in a way photography is not” (Roskies 2007, 871). Their appearance results from decisions about how to generate, process, and represent numerical data recorded from a functioning brain; this data could be given the form of graphs or curves rather than vibrantly colored brains.
FIGURE 2.1 Illustration of subtractive technique. Data collected during a control condition are subtracted from data collected during performance of task of interest. Fig. 1 from Roskies (2010). Courtesy Adina Roskies and Marcus E. Raichle.
The areas designated as being active are obtained by “subtraction” (Figure 2.1). Schematically: BOLD signals generated during task T1 (the control condition) indicate activity in brain areas B1 and B2; these are assumed to be “associated with” the neurobiological processes that “underlie” T1. BOLD signals generated by performing experimental task T2 demonstrate increased or decreased activity in B1 and B2 and/or “activation” of new areas. The experimental result emerges from subtracting the data of T2 and T1 (or “contrasting” the conditions). Even though the otherwise or newly “activated” areas thus identified represent correlations, they are said to “underlie” or “underpin” the functions involved in the task, in the sense of constituting their anatomical and physiological “substrates” or “bases” (Schleim and Rosier 2009).
Although by the early twenty-first century the range of subtractive techniques had become broader than this “classic” paradigm, they involve the same principles and assumptions and give rise to similar criticisms (Roskies 2010, 638). Debates around these have been highly technical and theoretical (see for example Coltheart 2006; Hardcastle and Stewart 2002; Roskies 2009, 2010; Van Orden and Papp 1997). A particularly relevant point for our discussion concerns the possibility of attributing function or significance to the results. Neuroimaging has often assumed that the brain is modular and that localization therefore provides information about mental functions. Various sources of evidence, including neuroimaging itself when it reveals replicable patterns and reliable activations, demonstrate that the brain indeed has a certain degree of functional specialization. At the same time, while not equipotential, the brain is highly plastic, interconnected, and integrated. It is characterized by “causal density”: any task is likely to have effects across the brain, and “there is a causal path between changes in any explanatory variable and any other variable” (Klein 2010, 269). Moreover, different neural states may realize the same mental state, and different mental states possess the same neural correlates (a feature akin to what philosophers call “multiple realizability”; Bickle 2013). And given that B1 and B2 are also active in T2, neuroimaging cannot discriminate the function of the various activations nor provide the basis for interpreting modified or new activation beyond noting an “association” between them and the experimental task. That is why it cannot answer the So what? question.
In defense of the subtraction method, it has been judiciously noted that neuroimaging results, hardly meaningful in themselves, are not or should not be considered piecemeal and in isolation (Roskies 2010, Rugg and Thompson-Schill 2013). Rather, they need to be interpreted by means of “triangulation,” with reference to other results of the same subjects, other neuroimaging experiments, and information of different sorts, such as that obtained (for example in psychology or neurophysiology) via other methods. A multimodal approach is indispensable. It has so far played a negligible role in the neurodisciplines but should in principle be beneficial. However, would going beyond the verbs (“underpin” and so forth) commonly used to describe the role of structures and circuits with regard to the function or task of interest give the neurodisciplines the power to say something relevant about their topics?
Results obtained until now suggest that it is unlikely and that research will reinforce the problem of multiple realizability and deepen the challenge of interpreting neural correlates. For example, each person has a “connectivity profile” that distinguishes him or her “regardless of how the brain is engaged during imaging” (Finn et al. 2015). Similarly, contrary to legend, while there are sex differences in the brain, largely attributable to hormone exposure during perinatal life (McCarthy 2015), brains don’t come in male and female forms; rather, most brains are unique “mosaics” of features, “some more common in females compared with males, some more common in males compared with females, and some common in both females and males” (Joel et al. 2015). Thus, while “your brain scans are quintessentially you” (Finn 2015), you are not quintessentially your brain scan.
Discussions about the uses and abuses of neuroimaging, especially in neurodisciplinary contexts, have not abated. In 2014, the Hastings Center, a major bioethics institution, devoted its Report to assessing functional neuroimaging. In it, Martha Farah, a cognitive neuroscientist and a leading figure of neuroethics, systematically probed the critiques and concluded that while each had a “kernel of truth,” each could also be rebutted. Yes, Farah noticed, a BOLD signal is not a direct measurement of brain activity, and we do not know to which aspects of neural activity it corresponds. But the relationship between the two is strong enough to make fMRI a useful tool. Images, like graphs or maps, are indeed manufactured, but they are not fabricated. Yes, fMRI is about localization, but it is not for localization’s sake, and most neuroimaging is not motivated by it. (That may be an accurate generalization across its many fields of application, but the neurodisciplines seem to be an exception.) Subtraction? Yes, neuroimaging initially assumed context-independent modules, such that “a cognitive process A will have the same neural instantiation whether it is accompanied by cognitive process B, C, D, or E” (Farah 2014, S23). But other approaches, such as the analysis of functional connectivity, should allow fMRI to transcend such limitation.
Neuroimaging research has been criticized for its use of “reverse inference,” that is to say for attempts to infer the presence of specific mental processes from detected “activations” (Poldrack 2008): If there is activation in area B1, then mental process M1 must be taking place. If there were a one-to-one correspondence between brain regions B and such processes M, one could infer which Ms are taking place by identifying active Bs. However, since (as mentioned above) psychological processes involve several brain regions, and since a single brain region is usually involved in multiple processes, such inference is hardly possible. Farah (2014, S24) argued that, with appropriate precautions and information derived from manipulating processes and observing brain activation, reverse inference may be legitimate. If performance of a task T involving mental process M consistently “activates” brain region B, then activation of B may indicate that M is taking place. Although such inferences are not the goal of neurodisciplinary research, they may serve, as Poldrack (2008, 224) pointed out, to “drive subsequent behavioral or neuroimaging studies, rather than as a direct means to interpreting neuroimaging results.”
Farah is right to emphasize that (as also mentioned above) many of the problems with fMRI (reverse inference, highly artificial experimental environments, small samples, low reliability, weak validity, and inflated statistical significance) are shared by other areas of scientific inquiry and that there is a difference between specifically criticizing particular applications or studies and “wholesale criticism” of neuroimaging. The latter concerns the method itself and casts doubt “on the conclusions of any research carried out with imaging, no matter how well designed and carefully executed” (Farah 2014, S28). Such an indiscriminate position would certainly be unfair, and there is no doubt that, since the neuroX wave began in the 1990s, enormous progress has been done in meeting methodological and technical challenges. Yet the congenital shortcoming of the neuroX does not reside in the quality of the design and execution of the experiments but in their adequacy to the objects they claim to be studying. We illustrate this below.
It is revealing of the self-assurance that prevails in the neurodisciplines that such a fundamental issue goes unexamined. In an article of 1999 entitled “If Neuroimaging Is the Answer, What Is the Question?” the cognitive psychologist Stephen Kosslyn doubted that mental processes can ever be better understood by observing which neural sites are “activated” when subjects perform a task. Rather, he argued, one should start with questions that inspire experimental tasks in ways that take advantage of the strengths of neuroimaging techniques. Although Kosslyn (1999) was reflecting on work done mainly since the early 1990s, his question remains as topical as ever. Over fifteen years later, and after considerable hype and groping, the neurodisciplines show evident signs of methodological and theoretical refinement. At the same time, we must still ask whether the assumptions and approaches of the neuroX are suitable for the goals, questions, and objects they define for themselves. The answer, as our case studies suggest, is that they are not.
The Neurodisciplines of Culture
Most neurodisciplines aim at capturing the commonality that underlies the heterogeneity of behaviors and experience—in other words, universal neurobiological processes, which are “modulated” by contextual factors. In contrast, the neurodisciplines of culture, such as neuroanthropology and cultural neuroscience, focus less on commonality than on difference, on that which gives cultures their specificity and on how culture is “inscribed” in the brain. Like all disciplines of the neuro, they attempt to draw their explanations from knowledge about the brain. But they are particularly careful to emphasize the brain’s “enculturation” and the interactions of culture and brain. They thus offer an opportunity to examine how the notion of culture operates within a framework built to study transcultural neurobiological processes.
The editors of The Encultured Brain, a book that presents itself as “an introduction to neuroanthropology,” state that the discipline’s project is “to examine different neural systems empirically, understand how neural capacities develop, and document which biological and environmental factors shape their realization” (Downey and Lende 2012, 24). Such a project has been considered as part of an “exciting shift” toward a more “integrative” biological anthropology, insofar as it demonstrates “that anthropology can provide for neuroscience contextual examples of how enculturation can help explain differences in brain functioning, while neuroscience offers anthropology direct evidence of neuroplasticity’s role in social and cultural dynamics” (MacKinnon 2014, 357). The Encultured Brain claims to break with earlier notions of culture:
For a long time, anthropologists have focused on culture as a system of symbolic associations, public signs, or shared meaning. But from the perspective of the nervous system, patterns of variation among different groups also include significant unconscious, non-symbolic traits, such as patterns of behavior, automatized response, skills, and perceptual biases. This neuroanthropological framing opens more space for considering why all types of cognition may not operate in identical fashion, and how non-cognitive forms of neural enculturation may influence thought and action. (Downey and Lende 2012, 37)
In other words, culture is not only about shared representations but also about “shared conditionings of the nervous system.” This may seem self-evident, since it is unlikely that there can be shared patterns of behavior, either symbolic or automatic, in the absence of some shared brain processes. Yet for the authors we just quoted, it is the “implications” of that principle that appear “obvious.”
For them, “the predominant reasons that culture becomes embodied … is that neuroanatomy inherently makes experience material” (37). The trivial observation that “without material change in the brain, learning, memory, maturation, and even trauma could not happen” leads them to the apparently significant statement that “cultural concepts and meanings become neurological anatomy” (37). All of this, as the authors say, is obvious. The questions are whether, or in what sense, examining changes in the brain significantly adds to an understanding of culture beyond reiterating that neurobiological processes are involved, and how the notion of culture operates within the conceptual and methodological framework of the neuro.
These issues can be explored in at least two ways. On the one hand, with respect to research itself, we may ask: How do the neurodisciplines of culture translate their emphasis on brain-culture “bidirectionality” into concrete investigative strategies, and what are their empirical results? On the other hand, these disciplines can be examined in regard to their implicit values and epistemologies. In spite of their insistence on the two-way processes that embody the brain in culture and culture in the brain, they usually assert the ontological primacy of the brain and see the human groups that constitute cultures as a “community of brains” (Domínguez Duque 2015, 292). Such a premise turns culture, however defined, into an external factor that “shapes,” “influences,” and “impacts on” neural activity, function, and processes. The spontaneous use of such action verbs is emblematic of how the neurodisciplines of culture approach their object (Gutchess and Goh 2013; other references are given below). What is the consequence for those disciplines and for the very concept of culture?
Neurologizing Culture
Like other neurodisciplines, those that concern culture passed in a few years from being an informal group of scholars with common interests to having their own name and Wikipedia article, professional institutions, journals, societies, colloquia, educational events, blogs and websites, programs, and graduate students. Special issues of journals not specifically devoted to them highlight synergies between those new disciplines and more established fields. Thus, cultural neuroscience has been the subject of special issues of Psychological Inquiry (2013), Social Cognitive and Affective Neuroscience (2010), the Asian Journal of Social Psychology (2010), and Progress in Brain Research (2009). The Handbook of Social Neuroscience offers an overview (Chiao 2011), and the collective volume Cultural and Neural Frames of Cognition and Communication (Han and Pöppel 2011) includes several contributions from the discipline. As for neuroanthropology, calls for such an endeavor first emerged in the late 1970s; the word was used in the early 1990s and had entered reference works of anthropology by the middle of the decade (Marcus 1997, Downey 2012a). In 2012, the same year The Encultured Brain appeared, the journals Anthropological Theory and Annals of Anthropological Practice devoted special issues to neuroanthropology.
In turn, the term “cultural neuroscience” seems to have first appeared in print in 2007, in a chapter for the Handbook of Cultural Psychology. It was then defined as “an area of research that investigates cultural variation in psychological, neural, and genomic processes as a means of articulating the interrelationship of these processes and their emergent properties” (Chiao and Ambady 2007, 238). Cultural neuroscientists of course acknowledge that social factors offer more than “minimal interest” for understanding brain and behavioral processes (Zhou and Cacioppo 2010). At the same time, considering that the sociocultural level of analysis is by itself insufficient, they highlight the interdisciplinarity of their endeavor and the bidirectionality of the processes they investigate and speak of “biocultural co-constructivism” and “multiple” or “reciprocal determinism” (Zhou and Cacioppo 2010). Cultural neuroscientists hold that values, practices, and beliefs both “shape and are shaped by the mind, brain and genes” and that the study of “cultural variation in mental, neural and genomic processes” therefore constitutes a means of “articulating the bidirectional relationship of these processes and their emergent properties” (Chiao and Cheon 2012, 288; Chiao et al. 2013; Kim and Sasaki 2014).
While the notion that complex behavior “results from the dynamic interaction of genes and cultural environment” is not new, cultural neuroscience is supposed to represent “a novel empirical approach to demonstrating bidirectional interactions between culture and biology by integrating theory and methods from cultural psychology, neuroscience and neurogenetics” (Chiao and Cheon 2012, 289). It seeks “to explain a given mental phenomenon in terms of a synergistic product of mental, neural and genetic events” (289) and claims to have “potential implications” not only for psychiatry, business, and technology but also for global public policy issues in health, globalization, immigration, and interethnic ideology (Chiao 2009b, Denkhaus and Bös 2012). At the research level, cultural neuroscientists are motivated by two “still unanswered” questions: How do cultural traits “shape” neurobiology and behavior? And how do neurobiological mechanisms “facilitate the emergence and transmission of cultural traits”? (Chiao et al. 2010, 356).
Neither neuroanthropology nor cultural neuroscience represents the first attempt at broaching culture with neuroscientific tools. Since the early 1990s, cognitive neuroscience has incorporated the study of interpersonal and social behavior, and “social neuroscience” emerged toward the end of that decade (see Cacioppo and Berntson 1992 for an early use of the term); by 2005, a handbook of “key readings” had already been published (Cacioppo and Berntson 2005). The field derives from findings in cross-cultural psychology that show how social cognition and behavior depend on sociocultural context, and it combines neuroimaging, cognitive science, and social psychology to investigate the neural “representation” of social interaction and the neural “substrates” of social processes (Han and Northoff 2008, Zhou and Cacioppo 2010). The journals Social Neuroscience and Social Cognitive and Affective Neuroscience were launched in 2006, a Social and Affective Neuroscience Society “committed to research investigating the neural basis of social and affective processes” was established in 2008, and it was followed in 2010 by a Society for Social Neuroscience, intended to support “the interdisciplinary academic field devoted to understanding how biological systems implement social processes and behavior, and how these social structures and processes impact the brain and biology.”10 The journal Culture and Brain was founded in 2013, with a focus on “cultural differences in neural activity” and “the mutual constitution of culture and the brain” (Han 2013).
The social, affective, and cultural neurosciences largely overlap with one another as well as with neuroanthropology and transcultural neuroimaging (Domínguez Duque et al. 2009, 2010; Han and Northoff 2008; Lende and Downey 2012a); labels such as “sociocultural neuroscience” are forged to underline interconnections (Wajman et al. 2015). At the same time, these emerging disciplines are engaged in dynamics of differentiation. In particular, neuroanthropologists have emphasized the differences between their approach and that of cultural neuroscience (Domínguez Duque 2012, Lende and Downey 2012a). While, in their view, cultural neuroscience wishes above all to offer brain-level explanations, neuroanthropology aims to combine such explanations with an ethnographic perspective. It is thus “in a better position to move back and forth between the neural, the phenomenal and the cultural domains” (Domínguez Duque 2012, 22) and to test neuroscientific hypotheses “against the reality of what people actually do, say, and experience” (Downey and Lende 2012, 42). Ethnography, in short, should be able to provide “empirical access” to the ways social and cultural processes shape brain function, meaning and behavior (42). Some neuroanthropologists have expressed concern about cultural biases in research and called for an increased awareness of the historical, social, and political circumstances under which experiments are conducted (Domínguez Duque et al. 2010); others see the interface between anthropology and the neurosciences as a way of doing anthropology experimentally (Roepstorff and Frith 2012).
In short, there is a cluster of neuro disciplines aimed at understanding how the brain “mediates” social interactions and culture and produces emotion and cognition. The question is how and whether these questions and programmatic statements translate into research results capable of going beyond generalities such as “cultural practices adapt to neural constraints, and the brain adapts to cultural practice” (Ambady and Bharucha 2009, 342), which simply reiterate the field’s premise.
Causes, Correlations, Plasticity
Insofar as neuroimaging is supposed to show “how ‘deep’ culture can go into the human brain” (Kitayama and Park 2010, 124), it has been the method of choice for directly studying the “encultured” brain. However, to the extent that neuroanthropology draws its main concepts and questions from cultural anthropology, it emphasizes fieldwork as its empirical basis and is as a consequence not inclined to use neuroimaging, which requires an experimental setting. That is why most neuroanthropological studies limit themselves to citing brain research and juxtaposing it to other kinds of materials, drawn directly from the study of cultural settings and situations, under the assumption that such a juxtaposition demonstrates the impact of those situations on the brain or the “interplay” of culture, brains, and experience (see for example The Encultured Brain [Lende and Downey 2012a] or the special issue “Neuroanthropology and Its Applications,” Annals of Anthropological Practice 36, 2012). In short, “neuroanthropology” is mainly the name of a potential framework with seemingly little impact on concrete anthropological work. A study on the anthropology of opioid maintenance treatments for addiction may redescribe itself as “neuroanthropology” and its topic as “the neuroeconomics and neuroracial politics of opioid pharmaceuticals” (Hansen and Skinner 2012). Similarly, it is simply by attaching the label “neurocognitive” to the skills involved that ethnographies of rugby or capoeira become instances of neuroanthropology (Downey 2012b, 2012c). The renaming operation is purely cosmetic and constitutes good marketing, but it brings no methodological, empirical, or conceptual gains.
In contrast to neuroanthropology, cultural neuroscience uses neuroimaging so systematically that it is often described as “cultural neuroimaging.” This is not to say that neuroanthropology would benefit from turning to neuroimaging but that imaging methods have so far been the chief way of going empirically beyond merely juxtaposing the neurobiological and the cultural. The question is whether they satisfy the stated purpose of illuminating culture.
The difference between neuroanthropology and cultural neuroscience with regard to neuroimaging is consistent with their conceptual and disciplinary roots in cultural anthropology and cultural psychology respectively. Cultural psychology is indeed cultural neuroscience’s “parent discipline” (Denkhaus and Bös 2012)—but in a manner that involves little more than replacing the “mind” of the psy by the “brain” of the neuro. Indeed, the University of Chicago anthropologist Richard Shweder (1991, 72) defined cultural psychology as the study of “the way cultural traditions and social practices regulate, express, and transform the human psyche, resulting less in psychic unity for humankind than in ethnic divergences in mind, self, and emotion.” If we here substitute human brain for human psyche and neural unity for psychic unity and then add “brain” to the loci of ethnic divergence, we obtain an accurate depiction of cultural neuroscience.
This field assumes that “understanding cultural and genetic influences on brain function likely holds the key to articulating better psychological theory” (Chiao 2009b, 290). The quest for “influences” is reinforced by the premise that “human behavior results from neural activity” and by the further inference that behavioral variation among cultures “likely emerges from cultural variation in neural mechanisms underlying these behaviors” (Chiao 2009b, 290, our emphasis; see also Chiao and Cheon 2012, 289). Though here left vague through the use of “likely,” the reasoning presupposes a direction and hierarchy of causes, from genetics and the brain toward mind and culture. Neuroimaging and genomic methods for “mapping” neural processes and genes “to” neural, mental, and cultural processes produce correlations, but these are presented in a causal perspective reinforced by the belief that cultural traits constitute evolutionary adaptations (Chiao and Blizinsky 2010).
The tension between correlational results and causal claims, as well as the existence of an implicit epistemic hierarchy, undermine cultural neuroscience’s calls for synergy and bidirectionality. We later examine the relevant research, but let us first take as an example the assertion that cultural values, practices, and beliefs “impact human behavior” or that the “cultural dimension” of individualism-collectivism (a favorite of cultural psychology) “affect[s] a wide variety of human mental processes at a behavioral level” and “modulate[s] neural and electrophysiological responses” (Chiao 2009b, 291, 295). Such statements embody a circular reasoning. On the one hand, a cultural “dimension” includes by definition mental and behavioral processes, and these necessarily correlate with some feature of brain functioning. On the other hand, the cultural dimension is itself defined, at least in part, on the basis of the mental and behavioral processes it is supposed to “affect.”
Culture “influences” brain functioning, “modulates” neural mechanisms, “shapes” neural systems (291). Thus, people living in culture X may develop “distinct neural mechanisms.” Yet these mechanisms may “underlie” behaviors identical to those observable in culture Y, where they correlate with other neural processes (290). Cultural neuroscience has explored these effects in connection with emotion (providing “evidence that culture influences how people infer emotional states”; 296), interpersonal perception (showing that individuals from egalitarian vs. hierarchical cultures display greater mesolimbic activity to dominant vs. facial cues; Freeman et al. 2009), and social cognition (demonstrating that cultural values rather than ethnic affiliation “modulate neural response during self-evaluation”; Chiao 2009b, 297). The discipline studies a wide range of psychological process, from visual and semantic processing (Goh et al. 2010, Gutchess et al. 2010) to fear (Chiao et al. 2008), empathy (Cheon et al. 2011), and self-representation (Kitayama and Park 2010; Mrazek, Harada, and Chiao 2014). The basic claim is always the same: culture shapes activity in some part of the brain, which in turn guides behavior.
Cultural neuroscientists consider their findings supported by the existence of neuroplasticity (the brain’s ability to change as a result of experience) and its main theoretical consequence: to challenge the belief that brain functions have fixed localizations and that the brain is malleable only within strictly limited critical periods. Celebrated, as we saw in Chapter 1, as a revolutionary finding and immediately taken up by a large spectrum of interested individuals from brain fitness dealers to philosophers and psychiatrists, political scientists, and rehabilitation specialists, neuroplasticity also seems to confirm that cultural differences at the neural level reside in patterns of connectivity. Sustained engagement in cultural tasks, understood as repeated participation in routinized behaviors, results in different patterns of brain activation and functional and structural modifications (see Hanawaka et al. 2003, for Japanese abacus experts, or Maguire et al. 2000, for London taxi drivers). Brain plasticity can therefore be depicted as the feature that enables the interaction of brain and culture at the three interrelated levels of explicit values, conventions, and routines; socially shared scripts for action; and individual idiosyncrasy. It thus explains intercultural neural differences as a consequence of practice and experience.
Investigative Practices
Cultural neuroscience has followed two strategies. One, “culture mapping,” involves “determining which cognitive or neural processes vary across cultures without determining whether the differences are learned or innate” (Ambady and Bharucha 2009, 342). For example, while performing numerical tasks, native English speakers showed more activation in brain areas “associated” with language processing; native Chinese speakers showed more activation in an area “associated” with visual-spatial processing (Tang et al. 2006). The finding is hypothetically attributed to exposure to different visual patterns. Greater premotor activity in the Chinese “could be due” to the visuospatial nature of their language, whereas activation of language areas in English speakers suggests that “the retrieval of mathematical facts may be mediated by phonological processing” (Ambady and Bharucha 2009, 342–343). The second strategy, “source analysis,” attempts to determine “the source or causes” of cultural mappings, including genetic commonality or difference, cultural learning “mediated by brain plasticity,” and the degree of similarity between cultural environments. This strategy has been less pursued than mapping, “but new technologies promise to advance” it “quickly” (343, 344); by 2016, it seems to have remained programmatic.
In turn, neuroanthropologists see themselves as uniquely situated to explore brain-culture bidirectionality, and (as we mentioned) they take a critical stance vis-à-vis cultural neuroscience. Yet they too focus on how culture “influences” or “changes” brain function and structure and how brain areas “respond to regularities in the cultural stream of experience” (Domínguez Duque et al. 2009, 43). They too hail as “extraordinary” the fact that culture “affects” not only brain function but also brain structure (60; see also Domínguez Duque 2012, 22). Indeed, as neuroanthropologists explain, the prefrontal cortex “stands first to be modified or constituted by cultural experience as it is the structure that lays culture’s foundations” (Domínguez Duque et al. 2009, 60, 61, our emphasis). The notion that the prefrontal cortex is constituted by culture while also being what ultimately generates it goes beyond describing the reciprocal interaction, which obtains at all levels, of body and world. It highlights the foundational asymmetry of the neurodisciplines of culture. The assertion that culture, as a complex of activities including forms of learning, “modifies” the brain is certainly substantiated by empirical observation. In contrast, except in its most diluted interpretation, the claim that the prefrontal cortex “lays” the foundations of culture formulates an ontological assumption. And this assumption translates into the way research is performed.
Let us take a frequently quoted article in the field, published in 2009 in Human Brain Mapping and entitled “Neural Basis of Individualistic and Collectivistic Views of the Self.” Its goal was to understand how individualism and collectivism “modulate neural representations underlying social cognition” (Chiao et al. 2009, 2813). According to earlier studies, people who support individualistic values think of themselves and others as independent and as having stable personal traits, whereas those who endorse collectivistic ideals see people as interconnected and describe themselves as immersed in a social context. The authors drew on the notion of self-construal style (SCS), which has been used to differentiate Western and East Asian views of the self; they do not refer to research questioning the capacity of self-construal to reflect individual-level cultural orientation or mediate and explain cross-cultural differences (Levine et al. 2003).
Based on previous work suggesting that activity in the medial prefrontal cortex (MPFC) “reflect[s] the neural basis of self-knowledge” (Chiao et al. 2009, 2814; Kelley et al. 2002), the authors hypothesized that individualists would show a greater response for general self-descriptions, and collectivists, for contextual self-descriptions, in the anterior rostral portion of the MPFC. Twenty-four right-handed university students were recruited for the study, twelve native Japanese from Nagoya and twelve “Caucasian-Americans” from Chicago. They were shown seventy-two stimuli, in Japanese or English respectively: twenty-four general self-descriptions, twenty-four contextual self-descriptions, and twenty-four self-descriptions in italicized or roman font. In comparison with collectivists, individualists exhibited during self-judgments greater activation in the bilateral thalamus, right putamen, bilateral cuneus, right insula, bilateral cerebellum, and right superior frontal gyrus; in turn, collectivists showed during the same task greater activation in the left middle temporal gyrus. Self-construal style, as established after scanning, interacted in a statistically significant manner with “type of self judgment in neural response within the anterior rostral region of MPFC,” the bilateral parahippocampal gyri, right middle temporal gyrus, and left superior occipital gyrus (Chiao et al. 2009, 2817).
The results seemed to demonstrate that “self-relevant processing within MPFC varies as a function of [self-construal style].” People who endorse individualistic values show greater MPFC activation during general self-descriptions, while those who endorse collectivist values display greater MPFC activation during contextual self-descriptions. In both cases, increased MPFC activity “reflects the role SCS plays in how knowledge about the self is formed, and possibly also stored and retrieved.” The researchers concluded that “knowledge self-representations of one’s self … are culturally specific at the neural level.” They also speculated that greater activity within the right superior frontal gyrus may “reflect evidence of enhanced self-relevant processing in individualists relative to collectivists” and called for more research to elucidate how cultural values “affect” neural processing (Chiao et al. 2009, 2819). A meta-analysis of research in the area published between 2003 and 2014 confirmed that “East Asian cultures are associated with increased neural activity in the brain regions related to inference of others’ mind and emotion regulation whereas Western cultures are associated with enhanced neural activity in the brain areas related to self-relevance encoding and emotional responses during social cognitive/affective processes” (Han and Ma 2014, 293).
Such studies of the neural “bases” of individualism and collectivism are typical of the neurodisciplines in at least two ways. First, they illustrate a characteristic slippage between the establishment of statistical correlations (here, with culture as predictor) and the identification of anatomo-functional “bases” or “underpinnings.” Second, the outcomes that could matter are predictable without neuroscience or neuroimaging. The authors of “Neural Basis of Individualistic and Collectivistic Views of the Self” point to “an intriguing aspect” of their findings, namely that participants’ cultural values (individualism or collectivism) rather than cultural affiliation (being white American or native Japanese) “modulated” neural response during self-judgments (Chiao et al. 2009, 2819). But in the Western and East Asian contexts from which the study drew its subjects, people adjust to various environmental demands, so that culture, as defined by ethnic or national affiliation, cannot be expected always to match individual behavior. Its findings are therefore far from “intriguing.” The main thing a study such as the one we just summarized does is to convey the assumption that culture is based on the brain and the belief that a phenomenon becomes more real or objective by virtue of having a neural correlate. Unless these assumptions are made, there is no need for neuroscience to apprehend the “dynamic nature of cultural values across individuals and cultural groups” (2819).
Cultural neuroscientists may retort that they have not simply corroborated results from the social sciences but added something essential by showing “how such dynamic cultural values shape neural representations” (2819). However, in the same way that they cannot demonstrate the neural “bases” of culturally contingent values or attitudes, they cannot show how particular values or attitudes shape the brain. To be sure, “cultural values, beliefs, and practices must be important for social brain functioning” (2819). This, however, is so by definition. First, because anything brained organisms do is related to brain function. Second, because given that “social brain” refers to the brain regions involved in understanding others (Blakemore 2008) and that social cognition is, in humans at least, inseparable from culturally determined ways of interacting with others, culture is necessarily “important” for the social brain. The contrary would in both cases constitute a sensational finding, if not a contradictio in adjecto.
Cultural Diversity as “Neurodiversity”
On the one hand, with respect to their significance for understanding culture, neuroimaging experiments recover at the end what they put in at the beginning, namely the notion that culture has “neural bases.” On the other hand, the rhetoric of wonder, in which findings are always “intriguing” or “extraordinary,” betrays the persistence of a dualistic attitude. Celebrating the discovery that “culture” somehow “modifies” brain function implies imagining at least two dualities: brain and person, culture and individual. Yet, as has been pointed out from inside the discipline, “it should not be surprising per se that there exists a neural difference underlying a psychological difference”—in fact, the existence of such a difference is “an axiomatic assumption” of cultural neuroscience, not an “empirical question” (Freeman 2013, 26).
The cultural neuroscientists whose study we just sketched reported on the “influence of cultural values on neural responses within MPFC during self judgments, despite the absence of differences at the behavioral level,” and concluded that their results “reveal an advantage of examining cultural values such as SCS at the neural level” (Chiao et al. 2009, 2819). The “advantage” resides in the potential to discover cultural affiliation in the absence of overt behavior. Now, such inscription of cultural values “at the neural level” could mean two things. One is that culture, including beliefs, norms, and meanings, is somehow embodied in individuals, specifically in their brains, prereflexively shaping their actions (Choudhury and Slaby 2012a, Gallagher and Zahavi 2008, Noë 2009). Another is that the neural level displays a truth about humans as cultural beings that is not knowable by examining social and cultural practices. Although the programmatic statements of cultural neuroscience seem to favor the former interpretation, cultural neuroscientific practice materializes the latter.
A frequently cited study on the “neural basis of cultural influence on self representation” provides another illustration of such a perspective (Zhu et al. 2007; see also the replications: Ng et al. 2010, Ray et al. 2010). The authors used fMRI to analyze brain activity of Western and Chinese subjects while they judged personal-trait adjectives regarding self, the mother, or a public person. As others in the field, they started with the observation that whereas North Americans and Europeans tend to view the self as independent, autonomous, and separate from others, East Asians emphasize interdependence and interconnectedness. The experimental design was standard: thirteen Chinese and thirteen Western college students were scanned while asked to judge if an adjective was adequate to describe the self, the mother, and other. They were also asked (as a neutral control condition) to judge the font of the words.
The findings were said to provide evidence of a neural distinction between self and intimate persons for Westerners but not for Chinese. Thus, in Chinese subjects, “mother-judgments” generated enhanced MPFC activity (compared with “other-judgments” and the neutral condition), such that “the representation of Chinese mother” could not be distinguished “from the representation of their selves, in terms of the MPFC activity.” This result was said to indicate that whereas the Chinese “use” MPFC to “represent” both mother and the self, MPFC activity in Western subjects corresponds to a “representation” of the individual self alone (Zhu et al. 2007, 1314). The data seemed to the authors significant for both anthropology and neuroscience because it suggested “that culture influences the functional neuroanatomy of self-representation” and provided evidence of an “interplay of biology and culture in shaping the mind and the brain” (1315).
The study apparently mediated between a social constructivism that downplays the role of biology in cultural and social processes and practices and a naturalistic reductionism according to which interpersonal and cultural relations arise in the brain. However, unless one holds one of the two positions and thereby engages in a form of dualism, it is hard to justify costly experiments to arrive at statements such as “culture influences the functional neuroanatomy of self-representation” or “habitual cognitive processes are accompanied by detectible [sic] parallel neural processes” (1315, 1314). The paradox is that a significant Cartesian bias persists behind the explicit emphasis on brain-culture reciprocal interactions.
As the authors explained, social psychology demonstrated behavioral and cognitive differences between the Western and the East Asian self. But since it did not tell “whether cultures influence the relevant neural mechanisms,” it remained necessary to look for neuroimaging evidence that Western and Chinese selves effectively differ “at a neural level” (1313, 1315). Two poles were thereby joined: culture both “affects the psychological structure of self ” and “shapes the functional anatomy of self-representation” (1310). The trouble with such a claim is twofold. On the one hand, correlations do not reveal relations that can be captured by verbs such as “affect” and “shape.” On the other hand, the use of those verbs manifests a peculiarly abstract and mechanical view of culture. Contrary to the way they are here conceptualized, notions, attitudes, and practices connected to the self are integral parts of culture; they are among the key features that contribute to enact it, not something that a mysterious agent called “culture” shapes from the outside.
Insofar as cultural diversity is conceptualized essentially as a form of neurodiversity, the experimental setups and results of cultural neuroscience may become part of identity politics (Roepstorff 2011, 40). At the same time, by positing the existence of difference between selves “at a neural level,” cultural neuroscience contributes to the downplaying of diversity within the group. In both scenarios (interethnic difference and intragroup identity), the brain is endowed with ontological primacy: The mind is what the brain does, and culture is included in the process. One of the main concerns here is not that cultural neuroscience seems to suggest that universal values do not exist (Begley 2010) but that it naturalizes cultural stereotypes in the laboratory (Choudhury 2010, Choudhury and Kirmayer 2009). There have been calls for a more nuanced consideration of socioecological factors (Cheon et al. 2013), but so far they have not been systematically implemented in experimental work, and cultural neuroscience still has to draw consequences from the complex intellectual and political histories of sampling categories such as the usual “Caucasian-American” (see Painter 2010 for an overview). Indeed, as critics have pointed out, in practice cultural neuroscience tends to classify subjects on the basis of outer appearance at the expense of behavior or sociological or cultural dimensions, has “an understanding of ‘culture’ and ‘race’ which still appeals to biology, blood and ancestry,” and may therefore seem to reinforce “Western dominance in a postcolonial situation” (Martinez Mateo et al. 2012, 160; 2013, 3). Regardless of whether cultural neuroscience is really so politically incorrect, the notion of “culture” it assumes indeed functions as a proxy of “race” (Heinz et al. 2014).
From Culture to Brain
One could object that individualism/collectivism and self-representation are particularly problematic subjects or that we confined ourselves to investigations that explicitly claim to be about “neural basis” (for a synthesis, see Zhu and Han 2008). After all, research has been carried out on topics such as perceptual processing (Kitayama et al. 2003), attentional modulation (Hedden et al. 2008), language (Tan et al. 2005; Lei, Akama, and Murphy 2014), music (Nan et al. 2008), number representation and mental calculation (Tang et al. 2006), emotional processes (Chiao et al. 2008), mental attribution (Tang et al. 2006), and self-representation and self-awareness (Han and Northoff 2008); other topics, such as the default mode network; regulation and inhibition of feelings, thoughts, and actions; prejudice and dehumanization; and fundamental warmth and competence judgments (Ames and Fiske 2010), have been identified as up-and-coming research areas. Moreover, important integrative effort has been made concerning the neuroscience of intergroup and intercultural relations (Cikara and Van Bavel 2014, Warnick and Landis 2015).
The studies we chose are nonetheless representative. Let us take another one that has been justly described as “foundational”11 for the field: Shihui Han and Georg Northoff’s 2008 overview of the area, implications, and future directions of transcultural neuroimaging in relation to “culture-sensitive neural substrates of human cognition.” The authors position themselves clearly from the start: “A fascinating mystery facing human beings is how the brain gives rise to the mind.” Transcultural neuroimaging emerges as a way to deal with that mystery and is considered promising insofar as it “can bridge the gap between neuroscientific investigations of supposedly culture-invariant neural mechanisms and psychological evidence of culture-sensitive cognition” (Han and Northoff 2008, 646). Once again, then, the mind, and culture as its ultimate collective product, are essentially what the brain does.
The authors aptly ask whether cultural experiences modulate or determine preexisting patterns of neural activity. This is a crucial question, one common to all attempts at bridging brain and culture. But is it is relevant for understanding culture? As the authors themselves point out, even if the same brain region is “recruited” by different groups for the same task, “two cultures might have different meanings for the concepts involved in a task” (652). The significant level of analysis must therefore be that of meanings and practices.
Han and Northoff realize that the notion of culture involves complexities that cannot be studied via the usual experimental designs. They recognize, for example, that there is no such thing as a homogenous “Western” or “East Asian” culture. Research practices, however, are less nuanced. It has been remarked that cultural psychology may give the impression that “there is a very small number of cultural identities (North American vs. East or Southeast Asian) that vary principally on the dimensions of individualism-collectivism or independent-interdependent self-construal” (Cohen 2009, 194). The same applies to cultural neurosciences, whose methods and experimental designs inevitably homogenize and factorize culture. More importantly, cultural neuroscience does not take culture as its object of study but as an independent variable on which a dependent one, such as the individualist-collectivist position, rests.
We have already noted that some contributors to the neurodisciplines of culture think of their object in a subtler way. Anthropologists have suggested an experimental approach that would take into account both the anthropology of experimentation and research subjects’ lived experiences (Roepstorff and Frith 2012, Roepstorff and Vogeley 2009). Together with a number of colleagues, Juan F. Domínguez Duque, “the first Ph.D. in Neuroanthropology,”12 has criticized cultural neuroscience’s “primarily psychological” concept of culture, understood as a set of variables affecting the brain. Such an approach ends up setting aside “the actual social processes by which cultural knowledge is constituted” (Domínguez Duque et al. 2010, 143, 144).
For example, individualism and collectivism cannot be reduced to a simple variable and may be integrated in the same person to handle pragmatically different perspectives on the same situation. Domínguez Duque and colleagues would like to reduce the projection of the investigator’s cultural values and beliefs onto the groups analyzed and to situate the circumstances under which the experiments take place. For them, therefore, neuroanthropology is a sort of self-reflexive radicalization of cultural neuroscience, one in which “research and analysis techniques from cultural (and more broadly, social) neuroscience are integrated into and embedded in ethnographic research” (Domínguez Duque 2012, 25). In a similar vein, Suparna Choudhury (2010) proposes to approach cultural neuroscience from the angle of “critical neuroscience.” To do that, she suggests attending to the conceptualization of culture in the design and interpretation of experiments, taking into account the historical contexts of the phenomenon under scrutiny, considering the meanings that experimental categories may have in different cultures, and identifying how cultural biases and beliefs may influence the design and results of experiments (see also Choudhury, Nagel, and Slaby 2009; Choudhury and Slaby 2012b). It is hoped that such suggestions will inspire fruitful synergies between cultural neuroscience, neuroanthropology, and critical neuroscience (Lende and Downey 2012b, 411).
As for the concept of culture itself, an alternative to the psychologism of cultural neuroscience emphasizes that culture is socially created and transmitted and should be understood as “shared structures of meaning” through which people interact with one another (Domínguez Duque et al. 2010, 139; Domínguez Duque 2012). Such criticism of the notion of culture implicitly used by the “first generation” of cultural neuroscientists, as well as the emphasis on the contested and evolving nature of the concept, are accompanied by proposals to incorporate an anthropological understanding of culture into experimental settings. Those laudable goals, however, are not specific to the neuro in “neuroanthropology” or “neuroethnography.” Rather, they can be achieved by complementing various qualitative and quantitative methods with critical theory and reflexive ethnography and by “historically, socially and politically contextualizing the circumstances under which enquiry takes place” (Domínguez Duque et al. 2010, 144).
In a similar perspective, the German scholars Ruth Denkhaus and Mathias Bös point out that most criticisms of cultural neuroscience have already been raised with regard to its mother discipline, cultural psychology. They propose to replace the “entity conception of culture” underlying the homogenizing and essentializing tendencies of cultural neuroscience with a notion of culture as “patterns of representations, actions and artifacts that are distributed or spread by social interaction” (Denkhaus and Bös 2012, 445). Reference to “actions and artifacts” implies that culture is not in people’s heads but is simultaneously in the individuals, their brains and minds, and the world they inhabit (450). The authors thus criticize the assumption that culture is “stored in people’s brains” (Ames and Fiskes 2010, 72). Han et al. (2013) have also underlined the constitutive, rather than merely modulatory, role of context. While in the modulatory context-dependence model, neuronal and cultural influences interact but remain separate and independent, the notion of constitutive context-dependence implies no clear-cut separation between the biological domain of the brain and the social domain of culture. In this model, brains are “biosocial” and culture is “sociobiological” (353). Some cultural neuroscientists have proposed to redefine culture as that which is manifest in “the direct dependence of the brain’s neural activity” on context (Northoff 2013b, 95), and others would like to integrate factors such as socioeconomic status, unemployment rate, residential mobility, or population density in their definition of cultural influences as a way to address within-nation variation (Ng, Morris, and Oishi 2013).
Such views, however, do not alter the basic assumption that neuroscience provides “the most fundamental perspective yet available” on how people appropriate culture (Domínguez Duque et al. 2010, 140). That is why the declarations of intention about the co-construction of brain and culture have not had a major impact on how experimental and fieldwork are conducted, nor have they prevented neuroanthropologists from claiming that “the shared webs of signification that make up culture are primarily the product of the activity of the PFC [prefrontal cortex]” (Domínguez Duque et al. 2009, 60). Such a statement gives expression to the foundational creed of the field’s epistemology and thought style.
A 2015 overview asserts that, from a functional connectivity analysis showing that neural connections between MPFC and bilateral temporoparietal function (said to be “implicated” in perspective taking) “were much stronger for Chinese than for Danes during the judgment of social attributes of the self,” it can be concluded
that the Chinese self is constituted by a more integrated, or holistic, representation of both direct and indirect appraisals. In comparison, the Western self appears more one-dimensional in the sense that it is defined largely on the basis of the first-person perspective alone. (Kitayama and Huff 2015, 6)
Such an inference, however, is fallacious. This is not because it implies the questionable existence of a homogeneous, perfectly self-consistent Western or Chinese self. In fact, the study itself reports that in Asian American individuals, who have multiple cultural identities, brain response patterns depend on which “cultural frame” is made salient (10), but it does not establish that this cannot happen in allegedly monocultural persons. The inference is fallacious because the nature of a self cannot be inferred or even hypothesized from the existence of certain “neural connections.” However, such an inference exhibits the ultimate implicit goal of much neurodisciplinary research: to diagnose and classify on the basis of brain data, thus saving the trouble of engaging in apparently messier and less objective human science research. (We shall see in Chapter 3 that a similar ambition characterizes some areas of psychiatric neuroimaging.)
Culture?
What, then, other or beyond a “product” of prefrontal cortex activity, is culture for the neurodisciplines of culture? The concept was notoriously capacious by the time that the anthropologists Alfred Kroeber and Clyde Kluckhohn (1952) enumerated over 150 definitions, and it has remained so (Shweder 2001). In Primitive Culture, Edward Tylor (1871, 1) defined “Culture or Civilization” as “that complex whole which includes knowledge, belief, art, morals, law, custom, and any other capabilities and habits acquired by man as a member of society.” Since then, many others have followed more or less his lead, seeing in culture “the complex of values, customs, beliefs and practices which constitute the way of life of a specific group” (Eagleton 2000, 34). Different emphases are also to be found, with a range and overlap of meanings, as illustrated in Raymond Williams’s (1985, 91) observation that “in archaeology and in cultural anthropology the reference to culture or a culture is primarily to material production, while in history and cultural studies the reference is primarily to signifying or symbolic systems.”
What exactly goes into “culture” has been long discussed. Kroeber and Kluckhohn identified different types of definitions (descriptive, historical, normative, psychological, structural, and “genetic” in the sense of developmental) and came up with a lengthy list of the conceptual elements that went into them, from acts and activities to feelings, language, and traditions. They themselves proposed a very encompassing definition but recognized that there is no way to proceed normatively. Nonetheless, two things emerge. One is that students of culture tend to characterize their object as “the organization of human experience and action by symbolic means” (Sahlins 2000, 158). The other is that those organizations and means are not static and do not form systematic and homogenous totalities. Early twentieth-century anthropologists sometimes did regard culture in that way, producing what Marshall Sahlins (159) critically called “anthropology-cultures.” In that framework, it was always possible somehow to dig out the authentic native who reflected the culture. Indeed, as James Clifford (1988, 338) noted, the very idea of culture “carries with it an expectation of roots, of a stable, territorialized existence” (see also the neuroanthropological point of view in Roepstorff, Niewöhner, and Beck 2010).
Such bounded, functionally integrated, coherent units operating as a consistent totality probably never existed, and if they did, they certainly no longer do in the context of what has been famously characterized as “locally lived lives in a globally interconnected world” (Gupta and Ferguson 1992, 11). Two points about this context are especially relevant here. One is the possibility of internal contradiction. The debate around Margaret Mead’s 1928 Coming of Age in Samoa is in this respect illuminating. Mead offered the image of a harmonious society with a liberal attitude towards sexuality. Her book had a huge social impact and became the bible of an entire generation. Then in 1983, Derek Freeman published Margaret Mead and Samoa: The Making and Unmaking of an Anthropological Myth, where he argued (on bases that were later questioned) that Mead was misled by native informants and ignored evidence contrary to her depiction of Samoan life.
Details aside, a major lesson of this controversy is that Samoan culture contains paradoxes and contradictions, which are, as Nancy Scheper-Hughes (1984, 90) put it, “culturally structured but never actually resolved.” Aggressive and harmonious behavioral poles and values may function in the same individuals and groups depending on circumstances. Mead therefore captured a Samoan truth, not the Samoan truth. Anthropologists have given up the idea “that everything in a society must adhere to a single configuration or pattern” and no longer “think of ‘culture’ as a single integrated reality” (90, 91). Yet whenever cultural neuroscience draws on tools such as the Self-Construal Scale, it assumes exactly such a view, according to which any one factor (being “independent” or “interdependent”) must correlate with some basic principle or attitude considered defining of the culture (such as individualism or collectivism). Like neuroaesthetics trying to establish the neural correlates of beauty but incapable of taking into account the fact that the same stimulus may be judged both ugly and beautiful, cultural neuroscience can only identify the supposed neural correlates of isolated factors, under the postulate that these correlates represent the cerebral embodiment of culture. The definition of “culture” as “factors that affect the biological and psychological processes that shape beliefs and norms shared by groups of individuals” illustrates precisely that (Hyde et al. 2015, 76).
The second consequence of lives being “locally lived” in an “interconnected world” is that cultural difference is not a basic given that correlates with being or belonging to some form of “people” (Western, Asian) but is rather “a product of a shared historical process that differentiates the world as it connects it” (Gupta and Ferguson 1992, 16). Cultural neuroscientific practice assumes separate and discrete cultures, which it juxtaposes in its experimental designs. It thereby participates in the processes whereby differences are constructed. This is in itself unproblematic and, perhaps, an inevitable artifact of studying culture. The problem and the challenge lie deeper, and they apply to all the neurodisciplines of culture: their assumption that culture is essentially, both ontologically and causally, a byproduct of the brain does not equip them well to deal with cultural phenomena—while at the same time it gives them a powerful tool for shaping culture itself.
Varieties of Neuroaesthetic Inquiry
As both aesthetics and neuroscience, neuroaesthetics is pursued in a variety of contexts and from different methodological perspectives; such diversity, as that of other neuro fields, is expected to generate and enable the testing of “foundational hypotheses” across domains (Skov and Vartanian 2009b, 1). The official story of how neuroaesthetics came about follows a familiar pattern: first, “the possibility of equating an aspect of mind to brain processes is introduced and hotly debated”; models and hypotheses are then advanced, followed by experimental work; the area receives a decisive impulse from the advent of noninvasive neuroimaging techniques; and now, “just as the rise of the experimental study of consciousness and moral psychology has increased neuroscience’s relevance to these fields of enquiry,” a similar trend is foreseen for aesthetics (2–3). Insofar as mechanisms involved in aesthetic judgment also intervene in the case of objects that are not considered artworks, the relevant neural processes are likely to be common to different functions, and that is why, it is said, neuroaesthetics should be seen “as a basic part of the larger neuroscience program” (5).
The word itself seems to have been coined in the late 1990s by Semir Zeki, a specialist of the primate visual brain at University College London. A decade later, neuroaesthetics was depicted as coming of age “as a field of study in its own right,” devoted to exploring the “neural foundations of aesthetic appreciation” (Cela-Conde et al. 2011, Chatterjee 2010, Nadal and Pearce 2011). The new discipline might be undergoing “growing pains” (Chatterjee 2012), but that has not prevented it from being outstandingly productive.13 Moreover, in 2013 a major institutional step was accomplished with the establishment of a Max Planck Institute for Empirical Aesthetics in Frankfurt. While this institute does not identify itself with “neuroaesthetics,” it includes a neuroscience department, and its self-presentation as an interdisciplinary enterprise that “focuses primarily on the foundations of aesthetic evaluation, perception and experience” largely matches the discipline’s profile.14
Neuroaesthetics has been characterized as studying “the neural bases for the contemplation and creation of a work of art” (Nalbantian 2008, 357) in the visual arts, literature, music, dance, theater, and film, but it is actually both narrower and broader. Narrower because, as we shall see, it tends to concentrate on issues of appreciation, which it operationalizes as subjects’ liking or disliking stimuli or judging them for their “aesthetic” qualities; broader because it should be made to encompass two areas of study that place today’s brains—those of the scanned subjects—in long-term temporal perspectives.
On the one hand, there is evolutionary aesthetics, which generally considers art, and more specifically aesthetic preferences, as adaptations that evolved to enhance reproductive success (see the concise discussion in Davies 2009). More precisely, the “exhibitory form of art” is said to have roots in “courtship rituals by animals where genetic qualities of the displaying animal are assessed by the potential mate” (Zaidel 2013, 229; see also Schaeffer’s [2010] interpretation, based on evolutionary costly signaling theory, or Menninghaus’s [2008] claim that the evolutionary perspective confirms Kant’s theory of aesthetic judgment). Discussing these conjectures would lead us too far away from our topic. Suffice it here to say that the main relevance of evolutionary aesthetics for neuroaesthetics lies in the belief that the alleged adaptations have become inscribed in neuronal function (such that, for example, they determine humans’ supposedly innate preference for savanna-like landscapes; Falk and Balling 2010). Neuroaesthetics, in fact, has been said to belong in a wider field of “bioaesthetics” aimed at understanding representation, emotion, and creativity from a neuroscientific and evolutionary point of view (Fitch, von Graevenitz and Nicolas 2009; Skov 2006).
On the other hand, there is “neuroarthistory” (Onians 2008a). The book that launched this barbarous term was advertised as “a fascinating account” of “one of the newest and most exciting fields in the human sciences,” and its author was praised by his university for “cracking the real Da Vinci code.”15 He did that thanks to the concept of neuroplasticity, which allegedly opens the way to explaining why art looks the way it does. This, it is claimed, will for the first time in art history allow access to artists’ minds. For example: the reason why the animals painted in the Chauvet-Pont-d’Arc Cave look so “strikingly naturalistic” (a questionable assertion) is that paleolithic humans observed them intently; the neural networks of a contemporary Australian bark painter were particularly attuned to parallel lines because as a child he admired his father’s expertise in using fibers to make fish traps and was later exposed to parallel lines through op art; a similar reasoning explains why Florentine Renaissance artists made more use of lines, while Venetians, more use of color. In such a perspective, the answer to the question “What might neuroarthistory add to the discussion of Du-champ?” is that the objects Duchamp used in his readymades
had become so familiar as valuable devices that viewers would necessarily have enjoyed seeing them. They might not have placed them in the category of art, but the response they evoked was one at the centre of artistic experience, an unconscious pleasure, a pleasure enhanced by the additional references associated with title and text and context. It was such a neurally-based response that Duchamp unconsciously exploited. (Onians 2008b)
The word “nothing” is a more straightforward way of answering that question, and, as a reviewer noted, “it is disturbing that these often ludicrously tendentious ideas … are being advanced not by some mad autodidact on a park bench but by a serious academic” (Tallis 2008b, 19).
While neuroaesthetics has some methodological roots in the neurobiology of perception, it should be properly considered as part of a broader neuroscience of art, which could itself be situated within “empirical aesthetics.” Since its beginnings in the nineteenth century, empirical aesthetics has used mainly the methods of experimental psychology to restore aisthesis—perception and sensation—at the heart of aesthetic experience, and has devoted itself mostly “to the discovery of a finite set of universal laws that govern people’s interactions with objects” (Vartanian 2014, 8; see also the other chapters in Tinio and Smith 2014, part I; as well as Lichtenstein, Maigné, and Pierre 2013 for the French context). Although these disciplines have in principle a range of concerns that encompasses human responses to non–art objects too, questions related to art predominate in their actual productions, and they see their most important cultural significance in the answers they can provide to questions such as “What is art?” “What is beauty?” “How does aesthetic judgment work?” and, in the case of neuroaesthetics, “How do they all emerge from the brain?”
The neuroaesthetic mode of inquiry into these questions, as well as the answers provided, presuppose that art is essentially an extension of the brain’s cognitive and adaptive functions. Both reflect a “truth” Zeki (2002, 54) declared to be “axiomatic,” namely “that all human activity is dictated by the organization and laws of the brain; that, therefore, there can be no real theory of art and aesthetics unless neurobiologically based.” That is indeed the assumption of the entire field. From it follows neuroaesthetics’ most typical hermeneutic device, which consists of considering artists as neuroscientists qui s’ignorent, as individuals whose main task is to explore the brain with the tools of their trade. What accounts for the impact of Michelangelo’s Pietà? The fact that the sculptor “instinctively understood the common visual and emotional organization and workings of the brain,” which in turn “allowed him to exploit our common visual organization and arouse shared experiences beyond the reach of words” (Zeki n.d.; specifically on the widespread idea of the artist as neuroscientist, see also Cavanagh 2005, Ramachandran and Hirstein 1999, Zeki 2000).
In all fairness, the notion of the artist as neuroscientist also appears more metaphorically in connection with artistic artifices known to be consistent with the physiology of vision. For example, shadows must be darker than their immediate surroundings, but they can have the wrong color or shape yet function as recognizable shadows; given some basic cues, transparent material can be recognized in spite of gross deviations from the optics of refraction, and almost any reflection (as in mirrors) will be accepted as such; figures need not be complete in order to be meaningfully understood. These observations, however, were made without using neuroimaging and initially without reference to neuroscientific results at all. Most importantly, they were made in contexts where the main goal was not to explain art but to use artworks as stimuli for studying the visual system. It has, for example, been argued that finding statistical regularities (in amplitude spectrum and fractal dimension, among others) in paintings’ spatial structure may “grant insight into the range of spatial patterns that humans find compelling” and that art, both representational and nonrepresentational, replicates the “basic statistics” of the world (Graham and Field 2007, 149, 150). The goal here is to investigate the visual system through the analysis of artworks; the shared and divergent statistics of art and natural scenes “may provide new tools for uncovering the coding strategies of the visual system” (162).
The same approach is at work when the neurophysiologist Margaret Livingstone examines the reasons for the elusive quality of the Gioconda’s smile (it is more apparent in the low-spatial-frequency range and therefore more visible as a smile when one is not looking at her mouth) or claims that Rembrandt and many other artists were stereoblind, a condition that might have been an asset for flattening three-dimensional scenes onto a canvas (Livingstone 2000, Livingstone and Conway 2004). The accuracy and relevance of such interpretations have been questioned. The point, however, is that their object is not primarily art but “the biology of seeing”—how certain effects of artworks depend on the neurobiology of luminance, stereoscopy, shading, and other processes—as well as how art may enter into research about the visual brain.
This is very different from, say, suggesting that impressionism is “effective” because of, or thanks to, the amygdala, a bilateral brain structure involved in emotional learning and memory consolidation. Research has shown that the amygdala responds more strongly to a blurry version of faces expressing fear than to the same stimuli depicted in sharp detail (Vuilleumier et al. 2003). But does this mean that the impact of impressionist works results from their connecting “more directly to emotional centres than to conscious image-recognition areas” (Cavanagh 2005, 305)? In one case, artworks are used as stimuli to explore the functioning of the visual system; in the other, information about the functioning of a brain structure is extrapolated to explain in the simplest possible terms a complex historical and experiential phenomenon.
It is one thing to acknowledge that experience of artworks begins with the automatic analysis of perceptual stimuli; it is another to affirm that contextual, educational, biographical, and other nonbiological factors are nonessential because they only “modulate” neurophysiological processes. Such an assertion would simply get rid of art and aesthetics. A German neuroanatomist, for example, has proposed a “universal model of aesthetic perception based on the sensory coding of natural stimuli,” on the assumption that cultural, historical, and social factors are variable and that “being variable [they] cannot be relevant in the search for a universal theory of aesthetic experience” (Redies 2007, 100). The model boils down to the idea that artworks induce a particular “resonance” based on the evolutionary adaptation of the visual system to natural scenes and that sensory stimuli are “more or less aesthetic” depending on the degree of “neural resonance” they induce (106).16 Even when “aesthetic” is, as here, unduly employed as a synonym of “likeable,” those “variable” factors remain constitutive of human’s relation to art.
As the definitions and conjectures we just sketched illustrate, neuroaesthetics oscillates when it comes to specify its object. It seems to distinguish implicitly between the aesthetic relation and the artistic relation or experience.17 Partly following Kant, the French literary theorist Gérard Genette finds at the bottom of the aesthetic relation a “disinterested” attitude in which we pay attention to and judge an object independently of considerations about usefulness, focusing on its appearance rather than on its function. We can engage in this kind of relation with any object (including natural ones), but any object we approach aesthetically can also be treated in other ways (as a commodity, as a piece of decoration, as an instrument of power). From a biological and evolutionary standpoint, the aesthetic relation is seen as part of the natural behavioral repertory of the human species. In the artistic relation we recognize in addition that objects are imbued with an aesthetic intention, but of course, such objects too can also be handled in nonaesthetic ways and, inversely, numerous nonartistic objects have become art (see for example Fraenkel 2007, on the case of prehistoric cave drawings). As Genette (1999, 11) puts it, “it is not the object that makes the relation aesthetic, but the relation that makes the object aesthetic.”
In practice, neuroaesthetics focuses on responses to artworks, with the response generally operationalized as hedonic judgments concerning preference or degrees of liking. It thus systematically reduces the aesthetic relation to those judgments, which it uses as a proxy for the response to art. Its neglect of the specificity of the artistic relation is embodied in its very methodology, which assumes that subjects lying in a scanner can engage in an aesthetic and specifically artistic relation to the small-size copy of an artwork displayed on a screen for a few seconds and for an explicitly scientific purpose. Although some neuroaestheticians worry about the prodigious lack of ecological validity of studies in their field (Chatterjee 2010, Lacey et al. 2011, Nadal and Pearce 2011), the adjective “aesthetic” continues to be identified with “hedonic” and to be applied indiscriminately to perception, assessment, preference, and experience.
Beauty
As mentioned, neuroaesthetics is varied, ranging from investigations into the neuronal processing of lines, color, or kinetic contours to research into the neural correlates of aesthetic judgment. We shall first look at a fundamental concern of aesthetics, neuro and otherwise: beauty, with a focus on the contributions of the neuroaesthetics pioneer Semir Zeki, long the most visible figure of the field.
Zeki is a prolific author and blogger, fond of somersaulting over the abyss that separates single-cell measurements from the biggest names of the Western canon. He is of course perfectly aware that there is such a thing as culture and that, whatever its definition, it does not exist exclusively as encephalic mass. He recognizes, for example, that the correspondence between Mondrian’s or the constructivists’ preference for straight lines and the existence of brain cells that selectively respond to straight lines of different orientations and the fact that kinetic art seems “admirably suited for stimulating the cells in V5 and anticipated artistically the physiological properties of motion-selective cells” (Zeki 2001, 51) do not mean that the aesthetic feelings aroused by a Malevich or a Calder are attributable exclusively to the individual activity of certain neurons. Rather, the correlation between the existence of certain types of artwork and certain types of neurons shows that an artwork’s “constituent elements” are “a powerful stimulus for these cells” and that a brain deprived of the appropriate neurons “will not be able to appreciate” the art in question (Zeki 1998, 14). In other words, constructivist or kinetic art “would give no aesthetic feeling in the absence of those cells” (Zeki 2000, 100). Details aside, this apparent insight boils down to saying that one cannot enjoy painting without a visual system or music without an auditory one (or their functional equivalents).18 It therefore adds nothing to aesthetics. And since those perceptual systems are involved in nonaesthetic functions, an analysis such as Zeki’s “tells us nothing about Picasso and Cézanne that does not apply equally to Häagen Dazs and McDonalds” (Hyman 2006). For all their superficial references to Plato, Kant, or Schopenhauer, speculations à la Zeki (for example about the origins of Dante’s, Michelangelo’s, and Wagner’s art in their having “formed an ideal of love in their brains” [Zeki 2002, 62]) leave out so much of relevance that their contribution to aesthetics amounts to nil (Ione 2003). What, however, do neuroaestheticians do in the lab?
In 2004, Zeki and Hideaki Kawabata, a professor at Keio University in Japan, published in the Journal of Neurophysiology an article entitled “Neural Correlates of Beauty.” This brings us right to the heart of the neuroaesthetic enterprise: The topic is beauty, the methodology is brain imaging, and the main product is a correlation. The purpose of the study was to find brain areas that would respond specifically to one category of painting (for example, the portrait) as well as brain areas that would be “consistently active” across subjects when they perceive a painting they judge beautiful or ugly. This strategy, the authors explained, allowed them to circumvent the question of how what an individual regards as ugly or beautiful is conditioned by culture, education, and inclination (Kawabata and Zeki 2004, 1699). As it turns out, since these factors cannot be switched off, the authors do not really circumvent them but just ignore them—and with them, most of what makes something be “art” or “aesthetic,” “beautiful” or “ugly.”
Four categories of painting were retained: abstract, still life, landscape, and portrait. Each of the ten twenty- to thirty-one-year-old subjects (“5 females”) was shown three hundred paintings on a computer monitor and was instructed to give each a score for ugly (1–4), neutral (5–6), or beautiful (7–10). Of these, 192 were selected (those with scores 1–2, 5–6, and 9–10), and they were randomly shown to each subject, who again had to rate them while inside the fMRI scanner. The procedure generated a two-factor 3 × 4 event-related design: one factor was three different response conditions (beautiful, neutral, ugly); another, the four genres of painting.
The study revealed a functional specialization of the visual brain, particularly for faces and landscapes, regardless of aesthetic judgment. A still life produced the greatest change in the visual cortex V3 area, and landscapes in the parahippocampal place area. Second main result: diverse contrasts (that is, subtractions in the sense of fMRI methodology) revealed activity in the medial orbitofrontal cortex, the anterior cingulate gyrus, the parietal cortex, and the motor cortex. With responses to the neutral stimulus acting as a baseline, the authors also found that, regardless of painting categories, signals increased for beautiful judgment at the orbitofrontal cortex and at the motor cortex for ugly judgment. Moreover, aesthetic judgment correlated with brain structures “known to be engaged during the perception of rewarding stimuli” (1702)—something to be expected since, at least in the lab, beauty is associated with pleasure, and reward is pleasurable by definition. But no separate area was “specifically engaged” when stimuli were judged to be ugly. The conclusion was that, since both beautiful and ugly stimuli “modulate activity in the same cortical areas … it is the modulation of activity within those areas that correlates with the judgment of a stimulus as being beautiful or not” (1704). Such a result, the authors emphasized, does not tell “what constitutes beauty in neural terms” but rather addresses the “Kantian” questions of “what are the conditions implied by the existence of the phenomenon of beauty (or its absence) … and what are the presuppositions that give validity to our aesthetic judgment. In aesthetics,” they declared, “the answer to both questions must be an activation of the brain’s reward system with a certain intensity” (1704).
This conclusion is consistent with the postulate that “the almost infinite creative variability” of art “arises out of common neurobiological processes” (Zeki 2001, 51). Does such commonality illuminate artistic processes beyond the fact that art making and aesthetic response share mechanisms with making and responding in other domains? Do these processes differ in artists and nonartists? The late Robert Solso, the author of Cognition and the Visual Arts and The Psychology of Art and the Evolution of the Conscious Brain, conducted MRI scans of a portrait artist while making thirty-second drawings of faces inside the scanner. When the results were compared to those of a psychology graduate student with no formal art training, they revealed decreased activity in the artist’s fusiform face brain area (FFA), frequently associated with face identification, and increased activity in his right middle frontal area, usually associated with “more complex associations and manipulations of visual forms” (Solso 2000, 82). The lower FFA level of activation indicated that the artist “may be more efficient” than the novice at processing facial features, and the involvement of the right frontal part of the brain suggested that he engaged in a “ ‘higher order’ interpretation” and an “abstracted representation” of the perceived face (83; also Solso 2001).
What have we learned? A visual artist can be expected to be more efficient than a novice at processing visual stimuli, and morphological and physiological brain changes always accompany doing and learning. The hippocampus of taxi drivers grows larger as the drivers spend more time in the job (Maguire et al., 2000), meditation has been shown to increase thickness in brain regions associated with attention and processing sensory input (Lazar et al. 2005), pianists’ brains display increased gray matter density and white matter integrity (Han et al. 2009)—and so on for the rest of us who are neither cabbies nor Buddhist monks nor pianists. Of course empirical details are important, and they may be useful to detect, diagnose, and treat any number of conditions attributable to a brain lesion. Art, however, is not in the brain as a lesion is in the brain, and the commonality neuroaestheticians claim to discover does not illuminate the differentiated aesthetic reactions individuals can have vis-à-vis a Mondrian and a Malevich: you can find the one beautiful and the other ugly even though perceiving both necessitates the firing of neurons specialized in straight lines. The fact that perceiving a Mondrian and a garden fence activates the same cells reveals nothing specifically about hedonic preference and even less about the aesthetic relation.
So does neuroaesthetics say anything about beauty that would be significant for aesthetics? Kawabata and Zeki refrained from theorizing beauty or venturing normative definitions. But almost a decade later, Zeki and another Japanese collaborator finally concluded that beauty is “some quality in bodies that correlates with activity in the mOFC [medial orbitofrontal cortex] by the intervention of the senses” and that such definition could give rise to standards of judgment: “A painting by Francis Bacon may be executed in a painterly style and have great artistic merit but may not qualify as beautiful to a subject, because the experience of viewing it does not correlate with activity in his or her mOFC” (Ishizu and Zeki 2011, 8–9, 7; for more on Bacon along the same lines, see Zeki and Ishizu 2013).
Whether Bacon or Beauty, these studies miss their putative objects. They postulate dichotomy as the basic structure of judgment, thus excluding the possibility that you might admire and enjoy Rubens’s or Bacon’s composition, palette, and brushwork but be displeased by the looks of their characters—not to say anything of the aesthetics of ugliness (Eco 2007) or of the fact that one can find an object or person simultaneously superb and hideous. Yet from the neuroaesthetician’s point of view, the problem with the study by Kawabata and Zeki was that it did not deal with “aesthetic judgment per se” (Jacobsen et al. 2006, 276). That is why researchers at the Institute of Psychology of the University of Leipzig and the Max Planck Institute of Human Cognitive and Brain Sciences, also in Leipzig, proposed to identify “the neural correlates of genuine aesthetic judgments of beauty” (281) through the use of geometrical shapes rather than paintings. Subjects were asked to describe each of four stimuli as beautiful or not and as symmetric or not. Since earlier studies found that aesthetic judgment is often ruled by symmetry, it was expected that “differences between the brain correlates of aesthetic judgment and symmetry judgment should be solely due to differences of judgment processes per se” (276–277). The use of abstract black-and-white patterns was intended to factor out elements, such as attitudes and memory, said to interfere with aesthetic judgment.
As anticipated, symmetry was the most important predictor of judgments; symmetric and regular pictures were generally found to be more beautiful than the others. Some brain areas were differentially activated. Aesthetic judgments elicited activation in the right frontomedial cortex; symmetry judgments, bilateral activation in the dorsal premotor cortex, the superior parietal lobule, the intraparietal sulcus, the left ventral premotor cortex, left fusiform gyrus, and the visual cortex. While symmetry had no significant influence on BOLD signal changes, beautiful judgments led to higher signal changes than not-beautiful judgments in the brain areas “specifically engaged” in aesthetic and symmetry judgments. Moreover, aesthetic judgments “recruited” areas that partly overlapped with brain networks involved in social and moral judgments. In conclusion, judgments of beauty were found to “trigger” activation in a brain network “that generally underlies evaluative judgments, and hence share neural substrate with, e.g., social and moral judgments,” and brain activations during aesthetic judgment can therefore not “be reduced to an assessment of symmetry but are actually due to a particular mode of judgment” (284).
Again, it is difficult to detect here a contribution to aesthetics. Neither a certain preference for symmetry nor the fact that beauty is irreducible to symmetry ranks as a novel insight. In addition, to the extent that aesthetic judgment is judgment, it must involve brain areas engaged in judgment. The ecological validity of the experimental tasks is yet another issue: shouldn’t the “per se” in “aesthetic judgment per se” include the memory, attitudes, and other factors the experiments were designed to exclude (but naturally couldn’t)? And does the experiment succeed in this? Isn’t someone familiar with and liking or disliking Malevich (or op, kinetic, constructivist, or geometric art in general) likely to react to a geometric pattern differently from someone who isn’t? And wouldn’t the judgment be different depending on whether the pattern measures 200 × 200 cm and hangs in an art gallery with a famous signature on one corner or is ten times smaller and displayed for two seconds on a screen inside a loud and cramped scanner? Surely the frictionless plane is an important epistemic device. But neuroaesthetics goes in this respect well beyond the Galilean conceptual model, and it is not enough to be asked to decide in a few seconds whether something is beautiful for the response to be germane to aesthetics.
FIGURE 2.2 Example of canonical and modified stimuli. The original image (Doryphoros by Polykleitos) is shown at the center of the figure. This sculpture obeys canonical proportions (golden ratio = 1:1.618). Two modified versions of the same sculpture are presented on its left and right sides. The left image was modified by creating a short legs:long trunk relation (ratio = 1:0.74), the right image by creating the opposite relation pattern (ratio = 1:0.36). All images were used in behavioral testing. The central image (judged-as-beautiful on 100%) and left image (judged-as-ugly on 64%) were employed in the fMRI study. From Di Dio, Macaluso, and Rizzolatti (2007). Courtesy Dr. C. Di Dio.
Other neuroaestheticians may retort that the shortcomings of Jacobsen and colleagues is that they stayed too close to Kawabata and Zeki and hence too far away from aesthetic judgment “per se.” An attempt at getting closer, carried out in Italy, examined “brain response” to images of classical and Renaissance sculptures manipulated so as to alter the golden ratio of the originals (Figure 2.2).
The purpose of the study was to find out whether there is “an objective, biological basis for the experience of beauty” or whether such experience is “entirely subjective” (Di Dio, Macaluso, and Rizzolatti 2007, 1). The question can be translated into whether “objective parameters intrinsic to works of art are able to elicit a specific neural pattern underlying the sense of beauty in the observer” (6). The authors hypothesized a positive answer, namely that humans are endowed “with species-specific mechanisms that resonate in response to certain parameters present in works of art” (6). The German team had chosen symmetry; the Italians preferred the golden ratio. Both symmetry and the golden ratio have for centuries figured prominently in empirical and philosophical inquiries into art and aesthetic experience; the fact that the debate about their role and status still goes on suggests the extent to which they do not represent merely factual matters but are dense condensations of values and worldviews. More down to earth, neither the Germans nor the Italians explained in which sense the “objective parameters” they claimed to investigate would be “intrinsic” to artworks, since they are absent in some artworks and present in non-art artifacts as well as in nature.
The Italian team used fifteen sets of three images, each including fifteen originals and fifteen modified pictures (seven with long trunks and short legs, eight with short trunks and long legs); twenty sculptures represented male bodies and ten, female bodies. The stimuli, thirty for each of six separate fMRI runs, were presented, in a randomized order and for two seconds each, in three conditions: observation, in which subjects were asked “to observe the sculptures as if they were in a museum,” aesthetic judgment (subjects were asked if they liked the image), and proportion judgment (whether they found the image proportionate).
Two types of analysis were carried out. One contrasted brain activations in reaction to canonical vs. modified stimuli. This was supposed to uncover “the neural responses to objective beauty parameters”; the hypothesis was that the golden ratio would elicit enhanced activity in areas mediating pleasure and that signal increase would be particularly strong during the observation condition, “where brain response to the artworks was not interfered with by additional cognitive requests” of judgment (2). (Asking the subjects to look at the stimulus as if in a museum was obviously not considered a significant request, cognitively or otherwise.) A second analysis, aimed at brain responses “related to the overt subjective appreciation of the stimuli,” contrasted brain activations obtained during the presentations of judged-as-beautiful against judged-as-ugly images.
The behavioral results showed that canonical images were evaluated more positively and modified ones more negatively. In MRI analysis, the results of viewing canonical and modified conditions were first taken together and then contrasted with the rest across all three conditions (observation, aesthetic judgment, and proportion judgment). This contrast revealed “activations” in several areas. Especially significant for the authors was signal enhancement in the insula during the observation condition. The insula is one of the most beloved brain structures of the neuro industry. It is involved in motor control and homeostasis, as well as in interoception and visceral states associated with emotional experience, in addition to self-awareness and sense of agency. It seems to play such a crucial role in combining information about bodily states into higher-order cognitive and emotional processes that, as a New York Times science writer put it: “The bottom line is that mind and body are integrated in the insula” (Blakeslee 2007).
Di Dio and her colleagues attributed the weaker insula activation effect in the aesthetic and proportion conditions to the explicit request to judge, which might have “diverted the volunteers’ attention resources towards a specific cognitive demand, thus lessening the natural neural response within the insula” (Di Dio, Macaluso, and Rizzolatti 2007, 4). From this it seemed to follow that “the positive emotional feeling elicited in the viewer by the canonical images was determined by a preferential coding of these images, relative to the modified ones, by various cortical areas and by a concurrent, joint activation of the anterior insula” (6). Thus, since the golden ratio “determined brain activations different to those where this parameter was violated,” the question about the existence of “objective beauty” was answered positively (6). The authors acknowledge that it would be “too reductive” to imagine that the sense of beauty “occurs because” of insula activation; joint activation of many areas and circuits is needed. In sum, could artworks “ever become a permanent patrimony of humankind without a resonance induced by some biologically inherent parameters?” (8). The answer is obviously “no.”
However, like any catalog of humanity’s essential attributes or components (from variations on the Western canon to enumerations of features characterizing human nature), the list of those parameters is negotiable, debatable, and likely to reflect particular interests and specific circumstances. Moreover, the fact that our brains may be predisposed one way or another toward certain qualities, such as proportion, symmetry, or scale invariance, does not help us better understand aesthetic experience as aesthetic and as experience. Common perceptual mechanisms are surely involved in looking at Andy Warhol’s 1964 sculpture Brillo Box in a gallery and at James Harvey’s original Brillo boxes in a supermarket. Harvey’s representative complained about Warhol’s use of his client’s design, yet he despondently admitted, “What’s one man’s box, may be another man’s art” (Gaddy 2007). While neuroaesthetics does not aim to analyze such a situation, it should at least have the means of taking it into account. All it has done so far, and all it can do with its methods and conceptual frameworks, is to correlate known facts (we react emotionally to artworks) or some narrowly operationalized aspects of the artistic relation (especially appreciation) with the activity of various brain areas, which are then said to “be involved,” “associated with,” “underlie,” “contribute to,” or somehow or other “reflect” those facts or aspects. Such are “the insights from neuroimaging” for understanding the “experience of art” (Nadal 2013).
Empathy
Although beauty has been a central topic in aesthetics, the neuroaesthetics of beauty has brought us to the point of wondering if, in spite of its name, the new discipline is about aesthetics at all. With David Freedberg, we enter a different world—one that promises a more sophisticated treatment of art as well as smarter ways of linking neuroscientific knowledge and the aesthetic relation. Indeed, Freedberg, formerly the Pierre Matisse Professor of the History of Art at Columbia University, and since 2015 the director of the Warburg Institute, had, before turning his attention to the brain, published widely on Dutch, Flemish, French, and Italian seventeenth-century art (including painting, drawing, and print), iconoclasm, the intersection of art and science, and, to a lesser extent, contemporary art. Freedberg’s interest in the neurosciences relates directly to historical events he explored in his seminal The Power of Images.
In that book, Freedberg (1989) wished, as Ernst Gombrich (1990) noted in a sharp review, “to lead the response to art back to our elementary reactions.” The author’s turn to neuroscience fulfills that original desire and explains the power of images on the basis of the neurobiological mechanisms of automatic empathetic reactions. The concern for the universality of response and its transcultural psychobiological roots drives Freedberg to consider that power as an immanent property and to look in images themselves for the principle of their efficacy, instead of considering them as configurations that, in order to have power beyond perceptual primitives, must encompass the spectator and historical and cultural forces (Prévost 2003). Christian devotion, for example, can be inspired not only by the more or less realistic depiction of religious scenes and characters but also by abstractions such as pictures of the Sacred Heart or the monogram of Christ. The metaphoric or metonymic sign may have a cognitive, affective, and existential power equal to that of the relevant related object, and that is why, as Gombrich (1990) put it, “there is no line one could ever draw between sacred images, words, or signs.” In all cases, it is context that makes those objects sacred.
As a historian of iconoclasm, Freedberg (1985) knows this; as a neuroaesthetician, however, he attributes the “energy” of the Buddhapada, the highly revered stone footprints that constitute early aniconic representations of the Buddha, to the activation of motor neurons in the viewer’s mirror neuron system (Freedberg 2009c). Like the Christian abstractions just mentioned, these sculptures are objects of devotion and carriers of specific powers by virtue of cultural meanings (whether apprehended at the highest theological level or at that of mere behavioral imitation, with all degrees in between). Determining if motor neurons are activated when we look at the trace of Buddha’s foot does not change that and (though scientifically interesting) does not add or subtract anything from an analysis of its power: If everyone’s motor neurons are activated when contemplating the sacred footprints but only some of us feel religious reverence (or indifference, aesthetic wonder, historical respect, or fundamentalist hatred), then those neurons don’t play the key role in how we experience the Buddhapada. Freedberg’s position reflects his rejection of the “standard social science model” that, in his view, blocks understanding of “the relationship between the cultural construction of responses and those aspects of response that pertain to our human nature” (Freedberg 2009c; see also 2007, 2008). The solution consists of shifting the investigative and interpretive balance to focus on “how culture modulates biology” and on how “neurology informs history” and to look for “biological and psychological invariances across cultures” (Freedberg 2009c; 2007, 17, 21).
For Freedberg as for others, the neural turn will renew the humanities and make them more meaningful; in aesthetics and art history, it will counter “intellectualizing views of art” (Freedberg 2009c) and make up for the “elimination of the emotional, the empathetic and the realm of non-cognitive corporeal response” said to characterize most of twentieth-century art history and criticism (Freedberg and Gallese 2007, 199). In those fields, Freedberg (2007, 23) explains, emotions “were felt to be too random, too embarrassing and too incidental to the transcendental value of art.” While the anthropological and art historical “orthodoxy” allegedly refuses to analyze emotional responses independently of their cultural and historical contexts, neuroscientific research since the 1990s has corroborated insights about empathy as an embodied emotion that were first formulated by several late nineteenth- and early twentieth-century philosophers, psychophysiologists, and art historians (Freedberg 2007, 27–29; 2009a, 87; 2009b, 70; Freedberg and Gallese 2007, 198).
Isn’t Freedberg setting up a straw man to boost his cause? He singles out the philosopher Arthur Danto (1924–2013) as an instance of the “intellectualizing” stance according to which aesthetic responses are “purely a matter of the way in which the concept of art is considered” (Gallese and Freedberg 2007). Danto, however, proposed no such thing. His point was rather that telling artworks from other things requires that they be constituted as “artistic” in virtue of theories and that it is therefore these theories that make “art” possible (Danto 1964). After all, there is no evidence that the painters of Lascaux or their contemporaries believed that they were making art, yet there now is such a thing as prehistoric art; before the German psychiatrist Hans Prinzhorn and among others the artists Paul Klee and Wassily Kandinsky in the early twentieth century, the drawings, paintings, and sculptures of the mentally ill, of children, and of indigenous peoples were not “art.” During the scandal that followed Australian aborigines’ painting over ancient “rock art,” a Ngarinyin leader declared, “Some told me recently that ‘rock art is dead.’ If ‘Art’ was dead, that would not matter to we Aborigines. We have never thought of our rock-paintings as ‘Art.’ To us they are images” (Mowaljarlai et al. 1988, 691). And these images, depicting the cloud and rain spirits known as Wandjina, must be repainted in order to perpetuate the presence of those spirits and “stimulate the energies that bring increase and renewal” in nature (692). The relationship to the images is largely determined by their function and status.
Pop art radicalized such connections in the making of the aesthetic relation. Danto described its challenge as being that of “indiscernible counterparts that may have radically distinct ontological affiliations” and asked, “Why is Brillo Box art when the Brillo cartons in the warehouse are merely soap-pad containers?” (Danto 1981, 4; Danto 1993). There might be, as he acknowledged, an aesthetic innate sense, yet responses will differ depending on how objects are classed, and the differences will be “as deep as those between bodily movements and actions, between a person and a zombie, between a divinity and an idol” (Danto 1981, 100). In short, if the aesthetic response to artworks involves processes that response to non-art things does not, then the processes involved in reacting to non-art cannot be what defines art or aesthetic response (see also Danto 1997, chap. 5). This is so much so that preclassification of artworks (for example, as fakes or originals, Leder 2001) actually affects people’s preferences. Neuroaesthetics holds the diametrically opposite assumption, namely that the ultimate grounds for our response to objects lies in object properties to which our brains automatically respond by virtue of their basic physiology. All the rest comes as a supplement.
For Freedberg, the key is to be found in mirror neurons and on the supposed neural substrates of empathy and embodiment. Together with Vittorio Gallese, the co-discoverer of mirror neurons, he has elaborated a “theory of empathetic responses to works of art that is not purely introspective, intuitive or metaphysical but has a precise and definable material basis in the brain” (Freedberg and Gallese 2007, 199). How would that work?
Mirror neurons constitute a system of visuomotor cells that fire not only when an organism performs an action but also when it observes a similar action performed by another organism, conspecific or not (Rizzolatti and Craighero 2004, Gallese 2009). They were discovered in the early 1990s in the premotor area F5 and subsequently in the inferior parietal lobule of macaque monkeys; mainly by inference from studies using electroencephalography, magnetoencephalography, and brain imaging, they are said to exist also in humans (Gallese 2007, 2008; review by Rizzolatti and Fabbri-Destro 2010). Although this has been doubted and major methodological difficulties persist, the most reliable conclusion so far is that “changes in the BOLD signal during action observation seem to be consistent with the existence of a mirror neuron system in humans, but they cannot yet furnish conclusive proof ” (Kilner and Lemon 2013, R1060; Caramazza et al. 2014), and the discussions around mirror neurons and their functions continues (for example as a Brain and Behavioral Sciences forum on the target article by Cook et al. 2014).
As soon as they were discovered, mirror neurons were hailed as providing the bases of language, “theory of mind” (the ability to attribute mental states to others), imitation, empathy (and therefore morality), art, social cognition, as well as social life and intersubjectivity in general (hence the hypothesis that mirror neuron dysfunction underlies autism). Major doubts have been formulated concerning the functions attributed to mirror neurons (e.g., Borg 2007 in connection with intentional attribution or “mindreading,” Hickok 2009 with regard to action understanding, or Jacob 2008 with respect to the representation of an agent’s prior intention; Rizzolatti and Sinigaglia 2010 reply that the brain observation-action execution circuit indeed gives the observing individual a grasp of another individual’s motor goals and intentions). In fact, the controversy—which empirical evidence alone is unlikely to settle—has not subsided and in any case does not appear to affect neuroaestheticians’ convictions.
A widespread way of understanding the role of mirror neuron systems is to say that they simulate (“mirror”) observed actions, whether performed or depicted, and that such “embodied simulation” is the foundation of our capacity unconsciously to make sense of the actions, emotions, and sensations of others. That is why such simulation may function as the basis for an approach to aesthetic response (Freedberg and Gallese 2007). Looking at an artwork incites in spectators (or rather, in their brains) the simulation of the action depicted or embodied in the work; the action can be that of the figures represented but also the artist’s creative motoric gesture. Thus, Michelangelo’s marble Prisoners activate in spectators the brain areas corresponding to the muscles that seem to be exerted in the sculpture itself. When we contemplate the singing angels of Hubert and Jan van Eyck’s Ghent Altarpiece, completed in 1432, “it is hard,” Freedberg (2009b, 67) claims, “not to want to imitate them, even to wrinkle one’s brows with the apparent difficulty of singing whatever it is they are singing.” Research has established that the observation of buccal movements enhances the motor excitability of the respective somatotopic areas in the brain, thus accounting for the imitative capacities of newborns; such findings, in Freedberg’s view (2009b, 76–78), add “scientific context” to much older commentaries about the vivid naturalism of the Eyckian angels.
Similarly, Freedberg asserts that, in front of a panel of Luca della Robbia’s cantoria in Florence Cathedral (1431–1438), which displays beautiful singing angels, “the desire for some form of emulation may well up within the observer.… Luca’s art is so remarkable that it seems to encourage its beholders, somehow or another, to participate in the movements he so vividly depicts” (72). Freedberg’s lyricism is effective, but it appeals to his experience and to a generic universal beholder rather than to empirical demonstration. Caravaggio’s Incredulity of Saint Thomas at Sanssouci in Potsdam, showing the apostle placing his finger in the resurrected Christ’s wound, gives rise to “empathy for tactile sensations” (201). And Goya’s Disasters of War provoke physical reactions in the same body parts that are depicted as mutilated in the pictures—a physical empathy that “easily transmutes into a feeling of empathy” (Freedberg and Gallese 2007, 197) and thereby opens the way for a moral response. (See also Freedberg 2008 on Rubens’s Peasant Dance.) Together with such investigations as Paul Ekman’s on the facial expressions of emotions or Peter Lang’s and others’ on the affective, facial, visceral, and behavioral reactions of looking at pictures, mirror neuron research seems to add empirical substance to Elaine Scarry’s or Susan Sontag’s classical essays about responses to the pain of others and even provides the basis for an “embodied narratology” (Gallese 2011, Wojciehowski and Gallese 2011).
These conjectures do not apply to figurative art alone. Viewing Jackson Pollock’s action paintings or Lucio Fontana’s cut canvases too provokes “embodied empathetic feelings” in response to the “visible traces of the artist’s creative gestures, such as vigorous modeling in clay or paint, fast brushwork and signs of movements of the hand” (Freedberg and Gallese 2007, 199). Freedberg and Gallese take the artists’ drippings and tears as “visible traces of goal-directed movement” (202). Yet a machine, or a chimp, or one of us could have randomly made those marks, or they could result from Mike Bidlo’s calculated effort to create a Not Pollock that looks exactly like an existing original. It is an empirical matter whether our brains respond in the same way to Pollock-looking images when “they” know and don’t know how the images were made or whether empathic simulation in front of Fontana’s slashed canvases is verified in subjects who lack familiarity with sharp objects and stretched surfaces. Of course, our visual processing systems may respond to a Bidlo as to a Pollock, just as we may like or dislike certain objects independently of whether they are labeled “art.” Our brains perhaps make us feel gestures that did or did not take place. But perception, awareness, and previous experience of context (including materials, size, position, authorship, and categorization) contribute decisively to the status of the traces supposed to give rise to empathic response, and this status in turn plays a crucial role in shaping our cognitive and emotional relations to them.
In experiments using various methods (EEG, transcranial magnetic stimulation, eye tracking, and event related potentials), Freedberg, Gallese, and associated researchers have reached results they see as empirical support for the “embodied simulation” thesis (Battaglia, Lisanby, and Freedberg 2011; Massaro et al. 2012; Sbriscia-Fioretti et al. 2013; Umiltà et al. 2012). They have shown, for example, cortical sensorimotor activation during the perception of static abstract-art images or corticomotor excitability effects that arise only during the observation of original works of art (as opposed to a copy or to a photograph showing a gesture they consider to be the same as that depicted in the original). Like most studies in neuroaesthetics, these articles combine highly technical methodologies with surprising weaknesses in such basic aspects as choice of stimuli and control situations. Putting these aside, their consequences for aesthetics are of two sorts.
On the one hand, these experiments, again like many in neuroaesthetics, imply the existence of biologically grounded criteria for evaluating artistic quality. This was obvious in the inquiries into beauty sketched above, whose aim, as has been justly observed, is “to extract rules that would lead to a practical definition of beauty, connecting features of objects and neural activity” (Conway and Rehding 2013, 1). Although Freedberg, Gallese, and their collaborators do not say it, they suggest the same goal when commenting on the correlation between looking at (the reproduction of) an original and a cerebral effect they interpret as evidence for embodied simulation: “Since observation of the photograph [of a gesture] did not significantly affect corticomotor excitability, we assume that this effect, in the case of the painting [depicting that gesture], must be a consequence of the artist’s skill in giving the illusory impression of movement” (Battaglia, Lisanby, and Freedberg 2011, 4). This opens the way for a dubious application of reverse inference: demonstrate cortical activation, and you have a proof of artistic quality.
On the other hand, precisely because of their focus on automatic unconscious or preconscious processes, the experiments we just mentioned differentiate between aesthetic experience, where embodied simulation is “an important component,” and aesthetic judgment, which, as “the explicit aesthetic rating of an object according to culturally and socially determined aesthetic canons,” represents “the most cognitive aspect of the relation established with works of art” (Massaro et al. 2012, 15). In the aesthetic relation, however, judgment and experience (even in the simplified senses given here) are not only continuous, as the quoted formulation suggests, but constitutionally interdependent.
Mirror-neuron-based embodied simulation theory thus brings us back to the question “What does neural hermeneutics contribute to the understanding of art and of the aesthetic and artistic relations?” It proposes mechanisms that may be universal as well as necessary for perceiving and responding to artworks. Indeed, we experience visceral sensations when we encounter tormented bodies in art. But we also feel them before early modern instructions for judicial torture or present-day journalistic reports and photographs of atrocities or gruesome accidents. In all cases we may be moved to disgust, compassion, and indignation; we may feel prompted to act or paralyzed with fear or despair; we may also be curious or admiring, find the object praiseworthy or contemptible; and we may experience any combination of the above, and more. In all these cases, the visceral-mirror component of our reactions (the supposed “embodied simulation”) is conceivably primary and automatic, but its relevance ends where questions of aesthetics barely begin. (Automaticity itself has also been questioned: de Vignemont and Singer 2006 suggest that empathy requires an appraisal related to the kinds of emotions involved, the relationship between empathizer and target, features of the empathizer, and situative context.)
Studies of the “empathic brain” (Keysers 2011) may identify the regions “involved” in the human capacity to understand other people’s sensations, intentions, and emotions. If anything, however, those regions’ activity can be no more than one condition of the aesthetic relation, and what the research contributes is nothing more “than a further ‘implementation’ story about our acknowledged abilities to respond to visual representations” (Davies 2014, 11). At bottom, the question is whether the activity of mirror and “canonical” neurons is constitutive of aesthetic response.19 In response to David Freedberg and Vittorio Gallese, two cognitive scientists and philosophers of perception argued that the authors’ proposal for a neural basis of empathic responses to art was “open to the charge of irrelevance to the issues of aesthetic experience and what constitutes artworks” (Casati and Pignocchi 2007). They emphasized that if witnessing real-life scenes that correspond to artistic depictions arouses “relevantly similar” neural responses, then neuronal activation “is not sufficient for aesthetic appraisal or judgments that something is an artwork.” Gallese and Freedberg (2007) replied that “mirror and canonical neurons are crucial elements in aesthetic response.” For them, aesthetic judgment requires the simulated embodiment and empathetic engagement that follow upon visual observation via mirror neuron activity; such processes “might be precognitive” and “not always” informed by cognition and culture. Moreover, insofar as “artistic skill” lies in artists’ eliciting emotional and felt motoric responses in the spectators, the “intellectualizing” (for example, Danto’s) view of aesthetic response must be mistaken.
The two sides do not even seem to understand “constitutive” and “crucial” in the same way. It is nonetheless telling that neuroaestheticians underline that the responses they study apply “in the case of lesser known—and sometimes everyday—images” (Gallese and Freedberg 2007). Indeed, making those responses and their mechanisms play the essential role in thinking about art implies that the differences, perceptual and other, between a Mondrian, the Parthenon, and a grid (or a picture and its reproduction, or a passport photo and the Veronica) are irrelevant for understanding aesthetic response. Not only does neuroaesthetics neglect materiality, and with it size, texture, color, and scale; it also does not differentiate between the picture of a hammer and the real hammer one can grasp, the depicted person and the real person with whom one could interact. It disregards the radical difference between the “affordances,” the distinct action possibilities offered by things and by their images.
Perceiving an artwork may neurobiologically prepare the spectator for actions or interactions, but these will not be necessarily realized; art thus transcends the actual enactive possibilities related to what it represents or conveys, but, by the same token, it opens up new, different ones (Gallagher 2010). There is no need to deny the existence of perceptual primitives or of bottom-up automatic processes to realize that neuroaesthetics disregards distinctions that are essentially relevant for the theory, practice, history, and reception of art and in general for the aesthetic relation. We concur with the philosopher David Davies (2014, 12) that empirical evidence of the kind provided by the new discipline might “inform” aesthetics but that “most of the significant philosophical issues cannot be resolved by appeal to this work.” For neuroaesthetics, such an objection is not even comprehensible, since its entire project assumes that aesthetics was for the most part misguided until it began to take the brain into account.
In short, aesthetic experience begins where neuroaesthetics ends; or, as the philosopher Alva Noë (2015, 361) put it, in neuroaesthetics “art isn’t explained; it is explained away.” Indeed, neuroaesthetics requires us to give up the very concept of “art,” to impoverish it to the point of considering its productions as a primitive form of brain imaging, and beauty as the automatic outcome of the arrangement of visual stimuli (Cappelletto 2009, 151, 152). More generally, the discipline moves art into an epistemic framework that excludes the notion of intentionally produced works (Fimiani 2009). Maybe this was the direction Freedberg (1989, 437) pointed to in The Power of Images when he complained that our perception is clouded by “the compulsion to establish whether an object is art or not.” There are occasions indeed when that compulsion and the discourse that surrounds it are obstacles to both feeling and understanding, yet if art is to remain a meaningful notion and the aesthetic relation a meaningful experience, cortex without context simply won’t do.