Art and technology have been converging rapidly in the past few years; an important example of this convergence is the alliance of neuroscience with aesthetics, which has produced the new field of neuroaesthetics.
The Neural Imagination examines this alliance. Neuroscience has demonstrated its relevance for aesthetics in several ways. First, it has identified specific neural accompaniments to aesthetic activities of both artists and audiences. One of the means by which such "localizations" are established is through the study of artists (such as Ravel) who have brain lesions. Nowadays, of course, this work is greatly facilitated by the use of brain imaging techniques. The changes in an artist's work after a stroke, for instance, help to localize the constituent elements of artistic production at the neurological level. Thus an injury to a particular area in the right hemisphere might interfere with a musician's appreciation of melodic contour without affecting other aspects of his/her musical abilities, and one might conclude that the injured area has an important role in "processing" melodies. As for audiences, Jonah Lehrer (2007, pp. 141-142) proposes that the 1913 riot at the first performance of Stravinsky's "Rite of Spring" was caused by the effect of unfamiliar sounds on certain neurons which, when overstimulated, precipitate a flood of dopamine that can, in turn, produce symptoms resembling those of schizophrenia. (The example of individuals such as Boris Pavlovich Nikonov, in whom music induced seizures, is perhaps more persuasive [cf. Avanzini, 2003]. I understand that Jock Murray of Halifax, Nova Scotia is studying musicogenic epilepsy.)
A second way in which neuroscience has become relevant for the arts is by providing the content of certain works. So, for instance, the behaviors associated with neurological syndromes such as Tourette's or Asperger's, or the effects of commissurotomy (the division of the two halves of the brain), are now regularly the subject of novels and films. An even more direct contribution of neurology to the arts comes in the form of "readouts" from the brain: the patterns of electrical activity that carry the codes through which we experience our environment are transcribed directly into sound or into visual displays and are then used as the basis of musical compositions or projected into the plastic arts.
On the other hand, the area of motivation—that is, why we want to have aesthetic experiences—has remained relatively opaque to neurological investigation. It is true that some of the events in the brain accompanying the aesthetic response have been localized. Semir Zeki (1999) argues that painters who specialize in certain techniques stimulate particular electrochemical processes that give us satisfaction. Other researchers have identified groups of neurons that seem to respond to beauty as such, or, at least, to whatever the subjects of those experiments consider beautiful. For the most part, though, the aesthetic as a global experience has remained hard to explain, and the motivation for that experience is not easily distinguished from other forms of pleasure. What makes the problem still more difficult, from the point of view of localization, is the fact that some of the most important elements of the aesthetic experience do not have anything to do with beauty or pleasure, but may be primarily strenuous, thought-provoking, or even painful. It then becomes harder to know what we are trying to localize.
Some of the above subjects are mentioned in other works on the psychology of art. What is particularly new in The Neural Imagination is the comparison of a neuroscientific approach to the arts with a humanistic approach. If we are to understand the contribution of neuroscience to aesthetics fully, such a comparison is essential. In addition to examining the general problems outlined above, then, this book juxtaposes three essays on the arts that use a neuroscientific approach (Chapters 2, 3, and 4) with a second part in which the differences between a humanistic and a neurological approach are emphasized. Each of the essays in the first group deals with one large area: the visual arts (Chapter 2); literature (Chapter 3); and music (Chapter 4). In these chapters, several hypotheses are advanced.
In the chapter on the visual arts (Chapter 2), I begin with a historical exploration of possible relations between nineteenth-century neuro-psychological theories of form and the rise of cubism. I then go on to modern neurological studies of the visual system in relation to art, ending with the thought that the so-called mirror neurons may have a role in the persistence of representationalism.
In the literature chapter (Chapter 3), I take up the question of dream language and its dissociation from other elements of dreams, the problem of aphasia, the relation between thought and expression, and the bearing of these issues on the language of literature. Ending with the topic of metaphor, I offer the conjecture that the experience of metaphor may be associated with what is called, in psychology, "contrast enhancement" and "the appeal of the rare"—possibly, even, in chemical terms, with an increase in oxytocin levels.
In the music chapter (Chapter 4), I begin with a historical study of the musical dream from Wagner onward. I then go on to explore the implications of this phenomenon for the relations of music, language, and imagery. Finally, I argue that, in song, language is subordinated to musical effects because the language function is relatively weak in the brain's right hemisphere, which tends to be dominant for music.
As I have said, Part 2 of the book, "The Imagination, Plain," emphasizes differences between the humanistic and the neuroscientific approach to the imagination.
For the topic of vision, I work with the observation that forms suggesting closure are generally preferred to open forms, and I compare the impressionistic defense of closed forms by a French nineteenth-century psychologist (G. Calame) with Ramachandran's neurologically informed discussion of this preference, as well as with some recent work by Bar and Meta.
For the area of literature, I explore Patrick Fermor's travelogues as well as what one learns from reading Fermor: namely, that language is not a set of defining rules but a handbook of suggestions. This is not the kind of conclusion that one would reach by using neuroscientific methods alone. Here I emphasize the difference between the kinds of generalizations drawn by literary critics and those drawn by scientists from the study of their materials.
For music, I attempt a new reading of Schubert's song "The Trout," but in this case I conclude that the neurological description of the word-music relation that I offer in Chapter 4 actually supports my impressionistic reading.
Chapter 6, which begins with the "Values" section, uses a variety of references, from Rembrandt to Yiddish poetry, in an attempt to define the tragic moment in art. The purpose of this section is to set up a clear antithesis to any utilitarian conception of art, on which scientific approaches to the arts tend to be predicated. It is intended to establish a clear-cut boundary between what is appropriate for neuroscientific investigation and what answers to a different principle.
In the last sections of Chapter 6, I confront some of the difficult philosophical problems that I encountered during the course of the book and try to reconcile the conflicting positions that emerged during my inquiry. I deal with the question of whether neural events cause or merely accompany mental events and with the uncertain neural status of hypothetical entities (such as "essence," "intentionality," or "imagination" itself).
The final section of Chapter 6, "John Keats to the Rescue," provides a more optimistic perspective. Keats had been a medical student and was aware of the most advanced neurological ideas of his day, but he did not think of neuroscience as a threat to the arts. His poem "Ode to Psyche" is a credo and a manifesto which demonstrates that a reconciliation between neuroscience and the arts is feasible after all. Keats imagines Psyche—once the last but now, by his act, the first among the immortals—as the very goddess of neuroscience. The temple that he will erect for her will not stand among pines, but will be woven of dendrites, "the wreath'd trellis of a working brain," on which the ever-changing buds of imagination will grow. Keats celebrates neuroplasticity: "branched thoughts, new grown with pleasant pain." Scientific thought is welcomed, not feared, by poetry. The only condition for this reconciliation is that it take place within inspiration itself. The primacy of the imagination is the precondition for the appropriate pursuit of neuroscience, of the science of mind; it is only from within the aesthetic that the two can collaborate successfully. Keats' poem, which enshrines Psyche, the soul, as the goddess of neuroscience, is the ideal expression of such a collaboration.
1. A Beauty Spot
Much of the neurological investigation into the arts has been concentrated on localization, that is, on the relation of the musical experience of timbre or rhythm, for instance, to a particular set of neurons in a particular region of the brain. There is also the more general question: whether the "beauty response" itself can be localized. If it can, might it then be amenable to electrical inducement? After all, if rats will pedal themselves practically to death by pressing on the "pleasure" bar, and if a subject could be made to find absolutely everything around her intensely funny by having a small area in her medial ventral prefrontal cortex electrically stimulated (Fried et al., 1998; Shammi and Stuss, 1999; and Goel and Dolan, 2001; for the eliciting of a hemi-smile accompanied by euphoria, see Springer et al., 2006), why couldn't one be made to experience beauty automatically, by similar means? Jamie Ward is cited as claiming that beauty is an innate, hardwired response (Garfield, 2006). In fact, Kawabata and Zeki (2004) claim to have located a "beauty spot," a response to beauty in the orbito-frontal cortex, and Vartanian and Goel (2004) have attempted something similar. There is also an article by Cela-Conde et al. (2004), in which the experimenters claim to have found an area in the left dorsolateral prefrontal cortex that shows increased activity when subjects are exposed to pictures that they judge to be beautiful. More recently, Limb and Braun (2008) have done fMRI studies of musicians doing jazz improvisations.
I was much inclined to dismiss all this news as so much cyberphrenology: finally, what does it tell us about beauty to know that a certain area of our brain shows greater activity when we find something beautiful? After all, in principle, you should be able to localize anything to which you can give a name. At the same time, the case of the artificial inducement of humor, a function almost as subjective and elusive as beauty, left me feeling uneasy.
All this was the last thing on my mind when I went for a walk with my wife last February in the wintry wastes of the so-called Alabama Swamp, about twenty miles outside Buffalo. I was in a bad mood, and I was not enjoying the walk particularly; everything seemed pretty drab. In fact, the swamp is a dreary place, flat and featureless, not exactly what you would think of as a "beauty spot." When, after several hours of walking, we were close to our car again, I started to lag behind. I think, uninteresting as the swamp was, I wasn't eager to get back to the city. There was a small fenced field on my left. Suddenly I noticed a few black teasels high on a bush, then a small tuft of purple-sepia grass reaching out across the snow, and at the back of the little field a mass of yellow reeds like a snowman made out of corn husks. Each of these details was as vivid as if an ophthalmologist had slipped exactly the right lens into the heavy apparatus on the bridge of my nose. I called out to my wife, "I'm being assaulted by inscapes!"
"Inscape" is the word that the poet Gerard Manley Hopkins (1959, p. 127) uses to describe the moment when we catch the essence of what we are perceiving. It is the moment when we do the "quiddity" of our percepts justice. A typical inscape from the Hopkins journals reads,
there was one light raft of beech which the wind footed and strained on, ruffling the leaves which came out in their triplets threaded round with a bright brim like an edge of white ice, the sun sitting at one end of the branch in a pash of soap-sud-coloured gummy bimbeams rowing over the leaves but sometimes flaring out so as to let a blue crust or platter from quite the quick of the orb sail in the eye. (p. 233)
On the drive back to Buffalo, three or four times, in flashing glimpses, I saw other details from the fields on my right with that perfection and intensity, almost arbitrarily, surrounded by meaningless space. It was as though something in me chose to slip the inscape lens into place a few last times before my giving up and going back to the city.
I continue to meditate on the implications of this excursion.
2. Purpose of the Neurological Investigation of the Arts
a. Benefits and Limitations
I cannot pretend to shake off entirely the uneasy thoughts aroused by this episode, even as I address the more practical question: What is the purpose of exploring the arts by neurological methods? Perhaps a hint of suspicion haunts the question itself, in the form in which I have phrased it. But what could be more desirable and more innocent than a scientific inquiry, with newly available techniques, into a major area of human activity? The humanists' fear seems to be that the very objectivity of the inquiry, and its proven capacity for correlating the details of psychology with physiological events, will end up subverting the arts, reducing them to ordinary functions without any special privilege. This chapter will begin to explore both the possible advantages and some implied threats of a neurological approach to the arts. I will first mention a few of the obvious benefits in introducing neuroscience to the arts, before allowing these plausible purposes to shade off into objections.
Neuroscience has made tremendous progress in optics in recent years. In consequence, some art schools now find it necessary to offer course modules in neurology, having concluded that it is important to teach students about discoveries concerning vision that may enrich their techniques or otherwise inform their work. There is an annual symposium on art and neuroscience at the Italian Academy in New York. The architectural theorist John Eberhard (2007) has published a book titled Architecture and the Brain: A New Knowledge Base from Neuroscience, which serves as an introduction to neuroscience for architecture students. Such developments may be comparable to the introduction of treatises on perspective during the Renaissance, which may have contributed to the massive changes in representational style that we now take for granted. Alexander Calder, for instance, was perhaps intuitively aware that adding color to a mobile reduces our awareness of its motion (because we respond more quickly to color than to motion; Zeki, 1999, p. 66), but such an intuitive awareness is quite different from the knowledge of students who have been taught the color/motion principle as a fact and have assimilated it as a standard part of their artistic repertoires. As for music, I have myself, in one of the chapters of this book, attempted a neurological explanation of the ease with which music overrides text in a song setting; whether my explanation, if correct, could contribute to the technique of a composer is another question. On the other hand, David Huron (2006, p. x) claims that from now on composers will indeed have to understand the brain's functions in order themselves to function successfully as composers: "in the absence of knowledge, our only recourse is to follow our intuitions. But intuition is not the foundation for artistic freedom or creative innovation." (One wonders how Chopin or Beethoven managed.)
Let us be clear: Huron is not saying that the composer can use information from cognitive science as a suggestion for possibilities of expression (as chaos theory or fractals have been used in other media, i.e., as metaphors to stimulate the creative mill). Huron apparently thinks that composers need this information in order to understand what they are already doing, so that they may then be able to do it better or differently. Interestingly enough, he ends without offering a single hypothetical example of this imagined process, or what its results might look like.
Data from neuroscience can, of course, be used as materials in any medium. Composers such as Horatio Radulescu (2003) and Diego Minciacchi (2003) (not to mention the pop music group "Heaven 17"), attempt to translate brain data directly into music, by using a special notation. For literature, Lisa Zunshine (2006) shows how dialogue follows the rules of cognitive science in imputing thoughts to an interlocutor, but it is not clear that the formalization of this understanding would help one to write a dialogue. Still, neuroscience has provided an abundance of subject matter for any number of novels and films, from Rain Man on.
In practical terms, then, some knowledge of neuroscience may be of greater value to the visual artist than to the musician or the author, though the materials of neuroscience can furnish subject matter to both, as some artists have used DNA patterns or microbial cultures in their work. (For more on this subject, see Anker and Nelkin, 2004, and Markoff, 2005. It is amusing to see an ad in the April 2006 issue of Discover offering DNA variety items for sale.) One of the most promising areas for a beneficial dialogue between neuroscience and the arts is dance. I have been introduced to the work of I. G. Hagendoorn by Daniel Glaser of the Wellcome Institute, and, although dance is a subject about which I do not feel qualified to write, there do seem to be great possibilities in the projects sponsored by Hagendoorn and his group and in Steven Brown's (2008) experiments. But even where no specific use is made of neurological information, no one, and certainly no artist, can help being influenced by the popularization of neuroscientific discoveries and ideas during the last few decades, beginning, perhaps, with Wilder Penfield's stimulating the exposed brain during surgery and producing realistic hallucinations (Penfield and Rasmussen, 1950, p. 36 and passim).
There have been numerous observations of what happens to artists after suffering a stroke. One of the most famous cases is that of Ravel, who lost some, but not all, of his musical capacities in consequence of damage to his brain (Mithen, 2005, pp. 53-55; see also Chapter 4, below). Such selective losses of function reveal the existence of what one might call subfunctions; this situation is most familiar in the case of language, where a stroke may deprive one of only the ability to speak, or to understand, or to read, and so forth. The same is true of music, as the area, areas, or networks affected in the brain may serve only one or a few aspects of musical function. What has also been observed is that artists may change their styles after suffering a stroke even if their ability to produce has not been reduced. A well-known example is that of the composer Vissarion Shebalin, again described by Steven Mithen (p. 33). Some physicians specialize in cases of this kind (B. Carey, 2005; Lythgoe et al., 2005; and W. W. Seeley et al., 2008; see also the useful volumes edited by Bogousslavsky and Boller, 2005, and Bogousslavsky and Hennerici, 2007, as well as Chapter 2, Section 2, below).
In view of such observations, it becomes difficult to deny that physiology is destiny; even under the best of circumstances, as Alfred de Vigny ("La Flûte," 1948, p. 201) said, "Des organes mauvais servent l'intelligence" [literally, "Bad organs serve our intelligence"] though it is hard to know what perfect organs would be. For all our attempts to carve out a privileged position for the arts vis-à-vis neurology, when someone acts strangely in a nonaesthetic context, we do not hesitate to assign a physical cause to that person's behavior (see Chapter VI, below). Why should we regard the aesthetic in any other way? If you have a stroke or a brain tumor, it will make you act differently. No one will deny that fact.
Nevertheless, perhaps only as a rear-guard action, I will raise some of the problems associated with such a physiological determinism. For one thing, the expected changes in creative style after someone has suffered brain damage do not always occur (see the article on Visconti and Fellini by Dieguez et al., 2007). (Apparently Balthus' style did not change despite the numerous strokes that the artist suffered.) What is more important, even determined efforts to use the lesion approach to triangulate the source of creative thinking have not yielded satisfactory results. One might assume that the sudden emergence of apparently creative activity after a clearly identifiable injury would enable researchers to locate the origin of the newly expressed abilities. Nevertheless, in their important article, "De novo Artistic Behaviour Following Brain Injury," Pollak et al. (2007) are at pains to dissociate themselves from any confident conclusions concerning the localization of creativity or the combination of neurochemical circumstances that may make creativity possible. As Raymond Tallis (2008, p. 15) remarks, the claim that "the future of aesthetics is in neurology" seems, to say the least, premature.
More broadly speaking, when an artist has a stroke, what changes? Before the stroke, the artist had to work with whatever neurological equipment was at his/her disposal. After the stroke, the artist still has to work with those "organes mauvais," only somewhat more "mauvais" than before. In other words, there is still the assumption of a central "intelligence" or, at least, consciousness, which does what it can with the equipment available to it. I do not see how any number of theories of "distributed processing" or "connectionism" can alter this fact (see Port and Van Gelder, 1995, and Lloyd, 2004).
Again, there is no doubt that neurology can explain certain artistic effects, especially in the plastic arts (see Chapter 2, below). These tend to be of the nature of optical illusions: for instance, Isia Leviant's "Enigma" (Figure 3a), in which concentric circles seem to spin. What neurology cannot do, though, as I will point out again in Chapter 2, is tell us whether "Enigma" is a good or a bad picture, and why. Is it because of its conceptual value, its showing that a physically stationary image is not in fact a fixed image for the brain? Is it because of the particular color scheme that Leviant uses? Is it because of the way the various lines and circles interact? In other words, there are so many elements in and parameters of the picture besides the fact that the circles spin that we are faced in the end with the old, familiar situation in aesthetics, in which a subjective judgment must be arrived at and justified on the basis of a multitude of variables. At that point, neurology has been left behind.
If the purpose of a neurological inquiry is to dispel the "mystery" of the arts in general, then larger objections arise, as I will try to show in Chapter 6. The fact is, there actually are some neuroscientists who would apparently be happy to "explain away" the arts, or who believe that they have already done so (see Kruglinski, 2008; cf. Ross et al., 2007). I will take up such challenges in more detail below. After my cursory survey of some of the uses, and some of the limitations, of the neurosciences in relation to the arts, I will reconsider a few of the issues that arise in this context in more detail.
b. Mechanical Art: Pros and Cons
In the background of any attempt to bring scientific knowledge to bear on the arts is the fear of ultimate success: that it may be possible to produce a work of art by artificial means. This procedure might consist in having a machine turn out something indistinguishable from the "real thing" or in manipulating the mind through electro-chemical techniques so that someone would automatically produce a work of art. If we can make a fruit fly flap its wings by the judicious application of a laser beam, might we not eventually be able to elicit an artistic production by analogous means? To quote Gero Miesenboeck, "If you had a hunch that a certain type of neuron might be important for a particular behavior . . . you could use this technique to see if that's the case—to find out whether activating those neurons elicits the behavior artificially" (quoted in Svoboda, 2006, p. 38). The artist as fruit fly. Or, perhaps, the artist's neurons could be induced to produce art without the artist's participation—or interference. After all, a decapitated frog can perform extraordinarily intricate maneuvers without its head; a male praying mantis, decapitated by its mate, goes on happily with its mating procedures.
At least at first thought, in the aesthetic sphere, these seem to be improbable outcomes. Even if we could determine with what precise parts of the brain Mozart's "Requiem" is associated, could we write another requiem as meaningful as Mozart's by stimulating those areas? Even if an entire work could be generated according to neurological principles, it would have little more than curiosity value. Computer poetry and mixed-media videos (see for instance, Glazier, 2002, and O'Gorman, 2006), computer-generated music, paintings based on optical illusions—all have to have a human impulse behind them in order for us to take them to heart. Even if that impulse should arise from a desire to seek out an impersonal secret in nature (Massey, 1977), it is still only through our conscious mind that that secret will reveal itself. Engaging nature at a more modest level, Jean-Jacques Rousseau ( 1966, p. 58) points out that even for a composer to make us want to hear frogs croaking, the composer must first give them a human voice, must make them sing.
Of course, this whole issue has by now expanded into its own vast field of inquiry and debate. The question of artificial art (a-art) is the subject of books from MIT Press, by Caroline Jones (2006), Mitchell Whitelaw (2004), and others. The Whitelaw book discusses many of the relevant problems in excellent, thoughtful, probing prose, while providing a survey of a-art involving primarily organic forms, materials, and concepts.
One crucial issue it avoids, though, is the necessity of a human investment, in the sense of the artist's committing some important, paid-for element of personal experience to the work. The discussion and the examples are largely formalistic: the book deals mainly with shapes and movements, interactive fields and processes (some, p. 133, involving EEG), and the theories and techniques that produce these results. There is considerable talk about the sociopolitical implications of the a-art movement but very little about the individual artist's experience and what it contributes to the artist's work: sorrow, pain, happiness, and the flavor of an individual's personality and life history, either as playing a role in the product or as recognized by the audience, do not attract much attention in this account.
Another problem, one that Whitelaw does bring up, but only toward the end of the book, might be thought to vitiate the whole a-art enterprise. It is clear that a-art tends toward an ideal horizon at which the machine will produce something that could not by any means have been foreseen by the machine artist. Whitelaw refers to the work of Adrian Thompson (p. 224), a computer scientist whose "breeder" program finally passed beyond his control and began to produce its results in inscrutable ways. If we are to assume that Thompson's results genuinely exemplify "emergence," they might offer the perfect model for an a-art that does not merely reflect, in complicated ways, the program that produced it, for it would have escaped the artist's intentionality entirely (Whitelaw, 2004, p. 226). At the same time, though, it would presumably have lost its special interest for the human observer, becoming a natural phenomenon to which one would respond only as if it were any other object of the senses. (For a sense of the struggle over issues of this sort in information theory, see Configurations, 2002.)
The third problem with the panorama of a-art as Whitelaw presents it is the aesthetic as such. The author frequently speaks with approval of particular works, but his comments seem entirely local, suggesting no particular criteria for the appreciation of a-art. The wider question of "What is art?," especially fraught when applied to this body of material, does not receive the treatment that it deserves. It is undoubtedly true that we will be facing an increasingly artificial world. It follows that we will also be facing forms of art increasingly mediated by machines. (It is worth noticing that Ede [2005, pp. 4, 187] still insists that scientists' images, whether diagrams based on DNA or produced according to mathematical formulae, are not ipso facto art, yet it is hard to deny an aesthetic dimension to the work of Karen Norberg and her group or to some of the images in Elkins, 2008.) There is a need for an aesthetics that considers the condition of postindustrial civilization as a major feature of our culture as we foresee it.
Besides the general fact that art and the machine are destined to become more closely interrelated, there is also the possibility, as I hint above (in "A Beauty Spot"), that one might eventually be able to induce what one might call a "creative state of mind" by some form of medical-mechanical intervention in neurological processes. After all, something of the sort has been demonstrated with epilepsy, some manifestations of which are similar to aesthetic, spiritual, or religious experiences. Moreover, artists of all kinds have used drugs on the assumption that the creative impulse can be enhanced by artificial means (see Boon, 2002). In the Baudelaire-Rimbaud circle, hashish was favored; of course, it was also understood that the quality of the poetry that was produced depended on who was taking the hashish. By the same token, not everyone as steeped in drugs as Marcel Proust was could be guaranteed to produce an À la recherche du temps perdu ([1913-1927] 1966-1969).
The possibility of genetic manipulation must also be considered. (On the genetics of music see Pulli et al., 2008.) An "anxiety gene" in mice has now been isolated, and it is possible to produce courageous "knockout" mice (Asher, 2005), just as, years ago, it was demonstrated that one could produce cowardly bulls by electrical suppression of the amygdala (see "timid bull"). By the way, a specific infection (Toxoplasma gondii) will make rats lose their fear of cats. Might one not be able to isolate a creativity gene? Or at least some neural mechanism that would disinhibit our spontaneous creativity? Perhaps, but it is hard to think that it could be so simple, that all the obstacles to expression could be swept aside by a pill or by a current. The very problems, disasters, and frustrations that block our libido also enter into artistic expression, the complexities and the flavors of which are provided by the fearful visitations of life as much as by its gifts. Assuredly, one could not, without some inner experience of tragedy, have written Hopkins' "Terrible Sonnets," or, for that matter, Shakespeare's King Lear. Or Bach's "Komm', süsser Tod."
Yet there are some aspects of creative expression that can have a curiously automatic character about them. I myself have had the experience, though very rarely, of getting into a "tuneful" state at the keyboard, so that I could produce passable tunes, in sequence, without any effort and without any natural limit, as though some melody-producing system had clicked into place for a short time. Had someone observed my brain during that period and stimulated the relevant areas, could the result have been repeated? Although vanishingly rare, those still-better times when music simply flows from the mind or the fingers, not so much as an action but as a state of being, might also be describable at the neurological level, as the lifting of an inhibitory process and the yield of a "productive" gene. In either case, though, the content of the expression could not be produced mechanically, since it would have to reflect all the richness and the variety, all the contradictions and the despair, of a particular individual's experience.
c. "Aesthetics": The Term
Before offering further comparisons between humanistic and neuroscientific approaches to the arts, I will attempt to sketch, in its broadest outlines, the field in which this investigation takes place: what is usually called aesthetics. For a general review of the subject, the reader may wish to consult C. Korsmeyer's Aesthetics: The Big Questions (1998). The term (which, etymologically, simply means "the study of sensation") began to develop its modern meaning—"the study of beauty or of the arts"—only during the eighteenth century. In that latter sense, it may be divided roughly into three areas: the sources of the art object, the characteristics of the art object, and the relation of the art object to its audience. This distribution (which is admittedly an old-fashioned way of organizing the material) allows in turn for an examination of the relations among these three areas, as in Charles W. Morris' "Esthetics and the Theory of Signs" ( 1971). Most of the issues that arise in critical writing, of whatever school, can be divided in this way, though, of course, the boundaries are often blurred. Even the multiple "Art and" approaches ("Art and Society," "Art and History," "Art and Psychology," "Art and Material Culture," "Art and Medicine," "Art and Linguistics," and so forth, ad infinitum), which bring other disciplines to bear on the arts, can usually be accommodated to this scheme. Strictly speaking, the neuroscientific approach to the arts should not be opposed to the aesthetic one, since, as a member of the "Art and" series, it is as much a part of aesthetics as any other approach.
All of this is manageable within the suggested taxonomy; on the other hand, I think that a satisfactory aesthetics for dealing with the new media, especially with the new mixed electronic media (see "Mechanical Art," above) and bio-art, remains to be developed. Whatever the principles of such an aesthetic turn out to be, I believe that the centrality of motion will be one of them, following the now widespread assumption in psychology that consciousness is largely concerned with phenomena involving motion (cf. Port and Van Gelder, 1995, and Thaut, 2005, p. 25). It gave me quite a shock to walk into the austere immobility of the Chinese sculpture gallery at the Metropolitan Museum of Art after spending a few days with the "Science + Art Festival 2006" in New York City. In her introduction to the anthology Sensorium: Embodied Experience, Technology, and Contemporary Art, Caroline A. Jones (2006), takes a first step toward formulating an aesthetics of the contemporary media. For the aesthetics of film, a useful compendium is still Dudley Andrews' The Major Film Theories: An Introduction (1976). There are also many books on the theory of conceptual art, among them Charles Harrison's (2001) and Arthur Danto's (2003).
Within the first of the three areas I mention above, that is, the sources of the artwork, we might group such perennial issues as inspiration, the nature of imagination, genius/creativity, and the identity (individual or collective) of the artist. Within the second—the characteristics of the art object—some recurrent topics are: What is art? (E.g., What is the art object in conceptual art, in Tinguely's self-destroying art, in performance art, in Joseph Beuys' sharing a room with a coyote, in mutilating or shooting oneself?) What are the components or structural features of a work—i.e., what are its rhetorical profile and its genre (tragedy, the sublime, etc.)? What are the strategies it deploys? What is tragedy? What is metaphor? What is the aesthetic status of natural objects (e.g., rocks, landscapes)? What are the similarities and differences among the arts? Are all the arts engaged in the same enterprise? What is the relation of form to content in art? What is the role of discord and ugliness in art (e.g., the grotesque)? But here we begin to shade over into the third category, the audience and its relation to the art object. In this area we might place such topics as: For whom does the artist create? (For himself, for others, or for an imaginary, a general, or a specific audience?) Why is there so little agreement about what is good or bad in art? Is taste in art determined by class or social group—i.e., is taste primarily a function of control and power?
There are at least two ways in which neuroscience can be brought to bear on some of the topics treated in discussions of the arts. It can look for neural "sources" for the author's act of production or the audience's act of reception. In other words, it can localize these events in the brain. Alternatively, it can take neurological patterns of behavior or response that are known from other situations and demonstrate their function in an artistic context. To pick a few random examples: an author who adopts a third-person perspective on a character's behavior activates his/her right inferior parietal cortex (localization) (Goldman, 2006, p. 213); the pleasure a metaphor gives us is associated with what psychologists call, in other situations, "the appeal of the rare" (function) (see Chapter 3, below); and Monet seascapes shimmer because of the way the artist manipulates the relation between the "what" and the "where" systems, systems that deal with color and luminance (function) (see Chapter 2, below).
The problem with these examples is that, in themselves, they have nothing to do with art. This difficulty is fairly obvious with regard to localization: it does not tell us anything about a work of art to say that the work activates, or is associated with, a certain spot in the brain. Functional explanations suffer from a similar defect, though in a less obvious way: they describe the event within the work as if it were just another event in daily life. This strategy is reduced to the ridiculous in such cases as Paul M. Matthews and Jeffrey McQuain's (2003, p. 75) neurological description of Macbeth's reaching for the imaginary dagger as though it were any act of reaching. It remains for traditional aesthetics to grapple with the meaning of the event within the work.
It is actually quite difficult to find areas in which this basic problem does not weaken or undermine neurology's contribution to aesthetics. One such space may be the genetic area, which has to do with the inception of the work of art. It may be possible to imagine setting in motion a process in the brain that would facilitate the production of works of art (see "Mechanical Art," above). The end product would not be the same as machine-generated art, which always encounters the objection that, in the end, it will be of little interest to live people.
Beyond such possibilities, though, a major value of neuroaesthetics is that it can provide us with a substantially expanded vocabulary for discussing the arts. We can talk in a clearer way about focal vs. peripheral vision and how this opposition affects the way we perceive a painting, even though our understanding of our visual processes may not change our experience of, or our response to, the painting. We can think with a finer discrimination about the varieties of memory instantiated in literature (even if it is only to note the differences among semantic, procedural, and episodic memory) if we know something about theories of memory. We can puzzle in a more hopeful state of mind over the various musical scales and modes that have been developed through the ages if we are aware of the ways in which scales are represented in the brain.
These are no mean achievements. As for the "final" aesthetic questions, those that have to do with the intrinsic nature of art and with the effects of art as art, it may well be argued that traditional aesthetics cannot boast of having had any notable successes in answering these either. The most it can claim is that it has not confused the category of the aesthetic with all the other domains of inquiry that seek to encroach upon it. But we should remind ourselves that, almost by definition, some subjects in traditional aesthetics simply do not lend themselves to investigation by neuroscience or, indeed, by any other system. The wide area of rhetoric and genre studies, for example, is difficult to speak of in terms of other disciplines. More important, it must be remembered that aesthetics is a branch of philosophy, and that there are certain issues within it, such as the definition of art, or the relation of art to truth, that can be addressed only in philosophical terms. Such issues lie largely outside the scope of neuroscience, and there is probably little point in trying to force them into its purview.
What will also be pointed out repeatedly below is that, at least until now, aesthetics has been grounded in acts of immersion in particular works, in an identification with what is happening in those works, and in an appreciation of their distinctive features. (For a brilliant defense of the quiddity of the work of art and the consequent resistance of art to scientific investigation, see Abbott, 2006.) "Humanistic" criticism of the arts tends to drift from observation to observation, rather like a dog inspecting its surroundings, or going from one fire hydrant to the next; it is likely to move from metaphor to metaphor, guided by instinct; it is not required to remain trained on an objective, though it may from time to time come up with an idea, and even a generalization. Again, neuroscience would have no reason to follow such tactics and has little to say about the unique qualities of any single work.
d. The "How" vs. the "Why"
Neurology is, then, of great value in exploring the "how" of aesthetic processes, if not necessarily the "why" or the "what for," or in helping to decide whether one work of art is of greater value than another. Again, neurology is usually more effective in analysis than in synthesis: it can help to localize the components of an aesthetic process, but it is less successful in enabling us to understand how they hang together.
For music, much progress has been made in the localization of rhythm, pitch, melody, and timbre. Some differences have been identified in the ways in which men and women process music. Isabelle Peretz (2003 and 2004) of the University of Montreal has made remarkable discoveries about the amusias (the global or selective failure to respond to music), as well as about the relation of language to music. For the plastic arts, Semir Zeki (1999) and Margaret Livingstone (2002) reveal in extraordinary detail the neurological processes by which we respond to colors, luminance, surface and depth, edges and angles, and straight lines (but not curved lines; Zeki, 1999, p. 116), and they have shown how these features function within actual pictures and other images. For literature, Ramachandran, among others, has attempted to localize the experience of metaphor (Graham, 2005); Lisa Zunshine (2006), as I mention above, examines the ways in which authors show characters imputing motives to other characters, a strategy possibly facilitated by "mirror neurons"; and Ronald Schleifer (2001) links certain aspects of poetry to Tourette's syndrome. For every area of the arts, an attempt has been made to link brain functions to specific features of the art in question. A field of physiological aesthetics, adumbrated in the eighteenth century, has now been firmly established.
Of course, as I have suggested, such inquiries also encounter methodological difficulties and may have inherent limitations. Neuroscience is best at connecting certain features of an aesthetic process with specific events in the brain, but artists and consumers of art care more about the integrated process, the whole experience, than they do about localization. What matters to them is not the science but the phenomenology of the artistic event. To borrow an analogy from the related field of ethics (Tancredi, 2005): even if we localize the neurology of altruism or observe the behavior of some brain area during an ethical dilemma, we still cannot turn over our ethical decisions to a brain function; somebody still has to decide what is right. Similarly, as I mention in Chapter 3, a nasal spray containing oxytocin may increase our willingness to trust others (Kosfeld et al., 2005), but we know that there are some people whom we had better not trust. At some point we must be able to take up a position outside our physiology, stand back, and judge for ourselves.
3. Scientific vs. Humanistic Methods
Needless to say, there is not a single "scientific method" or a single "humanistic method." (For one distinguished scientist's summary of scientific research methods, see Changeux, 2004, Chapters 7 and 8.) Both humanists and scientists, of whatever stripe, though, share the same fundamental interest: as Frederick Crews has reminded me, humanists are just as concerned with showing that their opinions are grounded in facts as scientists are. There is a common commitment to reliable information in both pursuits.
Because of this commonality, which should entail some mutual respect, methodological divergences should not lead humanists and scientists to encounter entirely irreconcilable differences when they evaluate the outcomes of their investigations, projects, etc. Still, the methodological differences are significant, and it is my purpose in this chapter to sketch some of them as they relate to this book's project. I will go over this ground again in Chapter 6; the issues are sufficiently important, though, to risk some repetition; they certainly warrant restatement and reformulation. I should also make it clear that, in my subtitle, "Aesthetic and Neuroscientific Approaches to the Arts," I use the word "aesthetic" in its broadest possible acceptance. For my purpose, it comprehends the ideas of the emotional, the subjective, the "plain," the traditional, and the humanistic, as well as such categories as structural analysis and rhetoric. As I mention above, some of these standard topics, such as the relation between art and truth, do not lend themselves to neuroscientific exploration. In "The Imagination, Neural" part of this book, I confine myself to discussing areas, such as the function of metaphor, where there is at least some hope of gaining a foothold in understanding the relation of neuroscience to aesthetic phenomena.
The differences between scientific and humanistic approaches to the arts probably appear obvious to most humanists, and are therefore not in need of reiteration or laborious demonstration. Perhaps this is not equally true for scientists. There has been a steady attempt to encroach on the humanities from the pragmatic, technological, and scientific side ever since Thomas Love Peacock, in "The Four Ages of Poetry," ( 1965) set out two centuries ago to prove that poetry was obsolete. Percy Bysshe Shelley's "Defence of Poetry" ( 1965) was, of course, the classic humanist response. But nowadays, technology threatens to subvert or subsume art at every level, and there are books such as David Edwards' Artscience: Creativity in the Post-Google Generation (2008) which declare unhesitatingly that there is no difference between the two domains. Steven Meyer (2001) argues that, since the beginning of the twentieth century, biological writing has converged with humanistic writing, and that the two have finally become identical. In a similar vein, Ira Livingston (2006, pp. 79-80) remarks, "What was news to biology [in the 1970's] had long been part of the classical legacy of structuralism to those who study language and culture." There is a certain amount of truth to this thesis, but the similarity between aesthetic inquiry and biology can be exaggerated. Of course, both may be concerned with the mind-body problem, but that does not mean that they are the same.
It is partly because of the inescapable differences between science and the humanities that I present Parts 1 and 2 of this book in a somewhat confrontational form. I have no desire to blur the boundary between the two styles of thought; in fact, one of the things I fear is the contamination of humanistic thinking by scientific discourse. H. Porter Abbott (2006, p. 713) remarks that "there are substantial practical and intellectual problems in treating scientific reductionism as a kind of gold standard for all areas of inquiry." J. Adams (2007, p. 204) adds: "The objection to the interest of the sciences in artistic production is founded on a fear that creativity will cede to just such a destructive mechanistic understanding." And, as Tim Dean has astutely pointed out (in conversation), it has even become dangerous to use the term "experimental" when speaking about the arts. Everything has to be "experimental." The term is recognized as honorific; it provides immediate validation, as the prestige of science becomes attached to any attempt at art to which one can relate that term.
In this sense, though certainly not polemic, the object of this book is almost the opposite of Steven Meyer's Irresistible Dictation: Gertrude Stein and the Correlations of Writing and Science (2001), as well as of the aforementioned book Artscience by David Edwards (2008). Edwards makes a comprehensive case for the fusion of the arts and sciences under the general rubric of creativity, in addition to appealing for the systematic institutional cultivation of interdisciplinary work between art and science. More insidious is the highly sophisticated suggestion of James Elkins (2008, pp. 99-100) that the border between intuition and calculation cannot be clearly defined (see also Adams, 2007, p. 70).
Actually, the present work is not focused directly on the comparison between the arts and the sciences, since it compares two different discourses about the arts rather than art and science as such. In this context, though, I try to show that, for all the obvious overlap between them, especially in this age of technology, the aesthetic and the scientific approaches to describing the arts represent two different kinds of thinking. I do not wish to reinvent Wilhelm Dilthey's ( 1961) wheel and argue for the irreducible distinction between the humanistic and the natural sciences, or merely to recall the "two solitudes" of C. P. Snow (1959); but I do wish to point out the difficulties in building a genuine bridge between the two areas of inquiry. Such recent commentators as Ede (2005, p. 194) and Barbara Herrnstein Smith (2005, p. 109) believe that these difficulties are insuperable. What is more, they even seem to be happy that they persist. Edelman's (2006) renewed attempt to span the gap between the cultures will not put an end to the debate.
In my previous essays, where I argued, for instance, that metaphor provides access to essence (1977), or that metaphor is an imperfect image seen through a perfect one (1999; see also Chapter 5, Section 3a, below), my purpose was to articulate some generalization derived from a personal experience; there is always some element of autobiography in a humanistic aesthetics. I had no concern whatsoever with the possibility that what I was writing might be expressed in physiological terms or, for that matter, that it might or might not ever be proved (cf. Chapter 6, Section 8, below). Similarly, rather than testability, my ideal in writing at least the second and third sections of Chapter 5 was also accuracy of or fidelity to my own experience, coupled with the capacity to draw out the implications of that experience and to communicate them. Finally, if I could not encapsulate the results of my thinking in a metaphor persuasive of its truth, I would feel that I had fallen short of my purpose. In some sense, the metaphor itself is the proof.
In contrast, when I set out to write the "scientific" essays, I was trying primarily to prove something. Since I am not a practicing scientist, I could not prove anything in my own right, but I could marshal information from accredited scientific sources to support my own hypotheses, such as my supposition that words function differently in song and in speech because song is primarily a right hemisphere function. The intent was to explain an aesthetic observation in physical terms.
To be sure, laboratory experiments may validate hypotheses founded on ordinary experience rather than on technical observations, even in those cases in which scientific confirmation could not have been foreseen (again, see Chapter 6, Section 8, below). In such instances, whereas later proof may be considered a bonus, it is not necessarily the only criterion of value for the original project. Leibniz's "petites perceptions" (1900, vol. 1, p. 80) were shown to exist by Wilder Penfield and Theodore Rasmussen (1950) two centuries later; Theodor Lipps' 1907 hypotheses concerning empathy (as well as Merleau-Ponty's notion of intersubjective awareness [Ostrow, 1990, pp. 42-43]) was confirmed when "mirror neurons" were discovered eighty or so years later; yet neither of these thinkers assumed that his ideas were futile because they lacked laboratory proof. But this situation may be changing.
To get a sense of the gulf that now separates humanistic from "hard" scientific inquiry, one need only read Daniel Dennett's (2005, p. 147) remarks on first-person reporting. Subjective reports about attributes of one's own consciousness (my "personal experiences") are indeed data for the scientist, but the contents of those reports are not. (For a dissenting view, see Hurlburt's contribution to Hurlburt and Schwitzgebel, 2007. I am indebted to Allan Hobson for the reference.) To quote Dennett, "they [the insights produced by introspection] are not data—the beliefs of subjects about them are the data" (p. 147). Only events that can be verified from a third-person perspective can have the status of scientific data. It would seem that humanists still believe, with Bishop Berkeley, that reality is founded in perception, whereas scientists now believe that it is founded in computation (cf. Lehrer, 2007, p. xii; for an interesting comment, derived from Rothacker and Blumenberg, on the element of subjectivity that cannot be eliminated from signification, see Halmi, 2007, p. 23).
At the level of vocabulary, there are immediate difficulties greater than those that our predecessors, the "natural philosophers," had to confront. A humanistic scholar might have trouble with the following statement: "fusiform and lingual gyri[,] also constitute extrastriate visual association areas" (Jeffries et al., 2003, p. 753). It is hard to know, however, what one could expect a scientist to do with phrases (of mine) such as "perfect vision" or "beauty vs. metaphor," or even with such commonplace judgments as "X was the greatest master of French prose of his time."
I have said that, in setting out to write the first two parts of this book, I began from different positions and with different strategies. These differences in points of departure for the humanistic and the scientific enterprises are usually self-evident; they go far beyond the choice of an inductive or a deductive approach. Obviously, evaluating a text is different from setting up a laboratory experiment, even a "thought experiment." Yet it cannot be denied that there are vast areas of study in the arts, such as "cultural studies," reception theory, or statistical analyses of authorship, that have at least as much claim to scientific status as do the so-called social sciences. These sub-disciplines do not provide the test case, though. For me, a crucial example would be a project that I have in mind as a topic of investigation. I have a sense (I have to say it in that way) that there is something about Patrick Fermor's use of metaphor (see Chapter 5, Section 3b, below) that surpasses our usual understanding of metaphor and reveals something about the "fuzziness" of language itself that we have not yet taken into account. The only way that I can pursue that "hunch" is by immersing myself—or trying to immerse myself—in Fermor's style and sniffing out the special things that are going on in it. Neurology would be of absolutely no use to me in such an effort. Even numerical approaches—e.g., counting words—could serve only to validate or qualify, at a later stage, what intuition alone and submission to the author's influence could ever make apparent. No scientific method could help.
Another, more general example of differences in approach might be taken from the field of memory studies, which are of equal interest to humanists and scientists. The humanist may be observing the broad difference between spontaneous and deliberate recall and its background in Bergsonian philosophy, whereas the scientist may be investigating the contribution of prions to the formation of mnemonic traces. In principle, there may be a way to reduce the first investigation to the second, but a number of obstacles, both technical and systematic, lie in the way. For one thing, the initial observations and hypotheses in scientific psychology have become less and less intuitive or available to the nonspecialist. Even when a first observation has been made at the behavioral level, in terms intelligible to the layman, there is pressure to move as quickly as possible to the physiological level. Here another problem arises: one cannot assume that bridges can be built readily between different orders of function, as each has a certain degree of independence. In "What's Not in your Genes," the biologist H. Allen Orr (2003, p. 40) remarks that some facts are irrelevant: "There are many questions in all fields of science that have little to do with underlying physical detail. . . . [W]e can learn a good deal about the mind . . . without worrying over the details of neurobiology. . . . [T]he answers to such higher-level questions do not change when the underlying mechanisms are discovered." The humanist is not likely to embrace the assumption that, say, human society can be explained by the scientist's molecular processes without the intervention of other explanatory regimes. (For an effort to grapple with the problem of what is often called "emergence" or, in other contexts, "supervenience," see Hogan, 2003, especially pp. 202-210; Modell, 2003, pp. 198-200; Whitelaw, 2004, pp. 207-210; and Adams, 2007, p. 117, 136-137.) Even Elizabeth A. Wilson (2004), who believes that a more flexible neurology could be useful in addressing social questions, acknowledges the difficulty of identifying the intervening stages and structures in such associations. When she asks, rhetorically, "Can historical, cultural, or economic events be serotonergic" (p. 28), the answer has to be sought in the midst of a complex negotiation among the subject, the social conditions, and a neurology that is itself an active player in the historical and affective circumstances it seeks to describe. In fact, Wilson tries to open up the very nature of determination (causation?) to interrogation (p. 26).
One major limitation of humanistic studies is that, although they can deal with mental phenomena and with observable behavior, they have no way of dealing with what goes on in the organism at the sub-behavioral level. This large area of inquiry is closed to humanists. As I have intimated, though, there are also vast areas that either are closed to neuroscientists or simply lie outside the domain of their competence. Take, for example, the large new field of specialization that has developed around mirror neurons, structures within the brain that respond automatically to the expressions on others' faces: if I see fear, the areas activated in my brain are connected with my own fear centers or circuits. However, these discoveries in neuropsychology raise for me major philosophical questions that overflow the domain of science as such. (On the philosophical and social aspects of mimesis, see the interesting book by Huhn, 2004.) Are all our perceptions polluted, so to speak, by an imitative impulse? What would a percept purged of imitation look like? (But even here, some might argue, the neurological study of autism could provide a clue [see Gravitz, 2006]. On the pathological condition known as "imitation behavior," see Tancredi, 2005, p. 70.) Is "perception" just another word for imitation? I find it difficult to think of ways in which a neurologist or even an experimental psychologist could address such a question comprehensively.
As we use language to subdivide experience and our thoughts about experience, we produce entities that do not lend themselves to secure definition but that we still somehow have to deal with. (For the neural status of hypothetical entities, see Chapter 6, Section 6, below.) Whether science can or should follow us into the thickets of onomastic fancy created by our own sorcerer's apprentice is up to scientists to decide. At some point, though, I think the path becomes too difficult, or not worth pursuing.
I write above that the humanist is far more likely than the neuroscientist to study a particular work exhaustively in order to identify its unique characteristics and effect. In Part 2, while comparing the humanistic and neurological approaches to the arts, I emphasize this difference repeatedly. In its most obvious sense, it is simply the difference between the participant's and the observer's point of view, what Smith (2005, p. 110), following anthropological usage, calls the "Etic" and the "Emic," respectively. Although, like the scientist, the humanist brings his/her preconceptions with him/her as a kind of template that isolates and emphasizes certain features of the subject, in the case of the humanist there is always the possibility that the encounter with the individual work, which necessarily overflows any scheme through which it is approached, will tease and draw one out beyond one's preconceptions. The artistic text is a potentially inexhaustible source of insight; one can go on describing its details forever, as if it were a dream. When one immerses oneself in a work of art, the very uniqueness of the work may lead one to, or give rise to, a generalization that only that work could have precipitated, a generalization that, in the absence of that work, would have remained forever unavailable (cf. Chapter 5, Section 3b, below). But the formulation or discovery of a generalization is not the purpose, and certainly not the ulterior purpose, of giving oneself over to the contemplation of a work of art (cf. Chapter 5, Section 3b, below). All one seeks is to do justice to the quiddity of the work; but the mind, in its habit of going beyond the particular, can spin off some broader thought that may have relevance to other situations.
To the best of my understanding, the body of fact available to, or produced by, the scientist is not in itself like an art object, which invites contemplation, identification, or appreciation for its own sake and in its own right. The same body of information may yield different theories to different scientists, or to the same scientist at different times, but one does not initially immerse oneself in it for its own sake (cf. Chapter 5, Section 3b, below).