Consciousness, Literature and the Arts

Archive

Volume 6 Number 2, August 2005

Special Issue: Literary Universals

_______________________________________________________________

Special Issue:

Literary Universals

edited by

Patrick Colm Hogan

University of Connecticut

Introduction

 VERBAL ART AND THE HUMAN MIND:

NOTES ON A RESEARCH PROGRAM IN COGNITION AND CULTURE

 by

Patrick Colm Hogan

            Between ten and fifteen years ago, I began to write articles on literary universals. At the time, the response of humanists ranged roughly from confused to appalled. Some people seemed to be unsure whether or not I was completely sane. Others suspected a sinister political program. In any case, the near universal consensus was that there could not possibly be such a thing as a literary—or, more generally, cultural--universal.

            The response was predictable, given the theoretical tendencies of the period. But it was nonetheless strange. One might have thought that the work of Noam Chomsky and others would have established the existence of extensive cultural universals at least in language. Moreover, language is an area that has obvious relevance to literature. Indeed, Paul Kiparsky had already isolated and explained some likely literary universals, drawing on Chomskyan principles (see Kiparsky 1981 and Kiparsky 1987). Additionally, one might have expected humanists to hesitate over a blanket identification of universalism with reactionary politics when the foremost universalist of the time was also the most prominent left-wing intellectual of the time. Of course, even without the example of Chomsky and others like him, the political suspicions about universalism made no sense. As I have repeatedly noted, racists did not defend slavery by claiming that Africans and Europeans share universal human attributes. They defended slavery by asserting differences between Africans and Europeans. Anti-Semites did not rationalize the Holocaust by claiming that Jews and Aryans are the same, but by insisting that they are different. How, then, could it become received wisdom that universalism is in league with political oppression and that the assertion of difference is liberatory?

            In the following pages, I will consider the study of literary universals today, which constitutes a very different field of research than it did fifteen years ago. I also wish to locate the contributions to this issue within that field, indicating how they represent current trends and point to future possibilities. However, before going on to this, it is worth considering just what gave rise to the vehement opposition faced by universalists in the recent past, an opposition that has not entirely dissipated.

Understanding Universals: A Problem of Words

            The automatic rejection of universals seems to have derived primarily from two common misunderstandings of the term “universal.” “Universal” has a technical meaning in linguistic theory, and thus in the literary study of universals, which has been broadly modeled on the linguistic study of universals.[1] Specifically, a universal is any feature that recurs across genetically and areally distinct languages—or, in the case of literature, literary traditions--with a greater frequency than would be predicted by chance. Genetically distinct languages or traditions are languages or traditions that do not have a common origin. Areally distinct languages or traditions are languages or traditions that have not had interacted through contact.[2] Humanists have often been unaware of this technical usage and, at least until recently, tended to assume that “universal” has one of two quite different meanings.

            The first commonly supposed meaning of “universal” was, roughly, “having undiminished aesthetic or related value cross-culturally.” “Universal” in this sense is a normative term that applies to individual works of literature or art. Thus someone might say that the Mahabharata is “universal” in this sense, but not the Iliad. It should be clear that this usage is entirely unrelated to the technical sense of the term in linguistic and, by extension, literary research on universals. The latter is not normative but descriptive. Moreover, it does not apply to individual works, but to features that recur across works. For example, one likely universal is that literary traditions have standard line lengths for poetry. This does not say anything about the value of standard line lengths. For example, a poem is not inferior if it does not use the standard line length for its tradition, employing, say, hexameter rather than pentameter. However offensive the non-technical (normative and particularistic) usage of “universal” may be, it has no bearing on the technical (descriptive and general) usage. One cannot reasonably criticize the study of genetically and areally unrelated patterns because it shares a name with the celebration of individual works. That is comparable to the actions of the characters in Julius Caesar who kill a man named “Cinna” because they are angry with someone else who also happens to be named “Cinna.”

            Moreover, it is not clear that this non-technical use of “universal” is necessarily objectionable anyway. It is true that writers have tended to use the notion in a Eurocentric way. But that is not necessary. Perhaps it is the case that some works cross traditions more easily than others. For example, Kalidasa’s Sakuntala seems to have achieved success in different traditions more readily than other Sanskrit works. Indeed, David Damrosch has recently discussed the ways in which works become part of “world literature” precisely by crossing the borders of literary traditions (see Damrosch 2003). Of course, this does not—or should not--have any consequences for literary evaluation. For instance, a work that crosses borders easily may simply be more shallow—thus more readily accessible—than a work that is less enthusiastically received outside its home culture. In this way, one might reasonably argue that the study of the transcultural migration of works is valuable, but empirical, not normative. In any case, it is not clear that such study—even in its normative guise—has any intrinsic politics. And, once again, it is clear that in either guise, it is a completely different project from the study of universals in the technical sense.

            The second common misunderstanding of “universal” is also normative. In this case, however, it refers to general evaluative principles, rather than specific works. Here, then, “universal” means roughly, “an aesthetic or related norm that defines literary excellence cross-culturally.” Again, this is clearly a different usage from the technical one. A descriptively universal pattern in no way implies any normative judgment. The universality of the heroic plot does not imply that heroic plots are good, or that individual works are good simply by the fact of following a heroic prototype.

            On the other hand, there is a difference between this and the first non-technical sense of “universal.” Since this usage applies to general principles, it may be partially connected with the technical usage. Specifically, it is quite possible that one could isolate features that tend to give rise to aesthetic appreciation cross-culturally. In other words, it is possible, in principle, to isolate universals of aesthetic evaluation. However, these universals are not themselves normative. They remain descriptive. In other words, an empirical study of cross-cultural patterns in aesthetic preferences does not establish those preferences as valid. Indeed, in some cases, it is actually impossible for a universal aesthetic preference to have universal validity as a norm. This is because some general preferences necessarily lead to particularizations that are mutually exclusive. Specifically, some preferences are “indexical,” which is to say, they refer to culturally or individually particular features of the evaluator (e.g., the reader of a literary work). Since these features vary from person to person, universal preferences of this sort cannot yield any truly universal norm. Put differently, no such universal can be used for establishing a cross-cultural aesthetic hierarchy. For example, one likely aesthetic universal is that, other things being equal, readers will prefer works that concern members of their own in-groups (thus people with whom they share consequential and salient identity categories). Since different people belong to different in-groups, one cannot use this preference as a globally valid norm for ranking works in terms of aesthetic excellence. One set of works will be given the highest ranking in accordance with my in-groups. But another set will be given the highest ranking in accordance with your in-groups.[3]

            This is not to say that the empirical study of universals could not contribute to a normative study of universals. It probably could. The point is that empirical study of universals does not by itself yield norms, even in cases where the study focuses on aesthetic preferences. Indeed, it could operate in precisely the opposite way. For example, one might view the universality of in-group bias as aesthetically deleterious and argue for self-conscious bias in favor of out-groups in order to counter-balance the initial, universal in-group bias.

            Here too, then, there is a very serious misunderstanding of universalism, for the study of universals in the technical sense is radically different from the assertion of general evaluative norms.

            At the same time, the political condemnation of general norms, common in humanistic circles until recently, itself appears invalid. Undoubtedly, the invocation of universal norms has been used toward oppressive ends. On the other hand, it has been used toward liberatory ends as well. This is just what one would expect. Everyone wishes to claim that his or her actions are justified. Once some norm appears to be widely accepted as a justification for action, everyone will claim that his or her action fits the norm. It does not follow from this that there are no justifications for anything. George Bush and Tony Blair invoke universal norms in defending Bush’s policies. Amnesty International invokes universal norms in criticizing those policies. Does Bush’s invocation of universal norms invalidate everyone else’s invocation of universal norms, including that of Amnesty International? Of course, it is the case that the people with the most power will be the ones whose invocation of universals is most widely disseminated. Thus Bush’s invocation of universal norms is more generally familiar in the U.S. than that of Amnesty International. But it hardly follows from this that Amnesty International should stop invoking universal norms. Of course, one may or may not believe that there are any normative universals—ethical, aesthetic, or whatever. Moreover, one may or may not accept the validity of particular norms that someone asserts are universal. As Kwame Appiah has emphasized, the problem with, say, colonialist claims of universality is not that they are claims of universality. The problem is that they are wrong (Appiah 1992, 58). But none of this affects the fact that the assertion of normative universality as such does not in any way entail oppressive policies. Again, it just as readily entails policies that diminish oppression. The result depends on the specific norms themselves and on their application in particular cases.

            In sum, objections to the study of universals that were standard in the past seem to have been based on misunderstandings of the term “universal.” They do not in fact apply to the study of universals in the technical sense. Moreover, they do not appear to be valid even in their own terms.

Tacit Universalism: How We Misunderstand Our Own Beliefs

            In recent years, I have on occasion been faced with a different sort of objection to a research program on literary or cultural universals, an objection that does not derive from a simple misunderstanding of the term “universal.” Specifically, I am sometimes told that there cannot be cultural universals as different societies believe there are different universals. This objection is problematic at two levels. First, it is almost certainly untrue that people believe in entirely different universals. (Obviously, they believe in some different universals.) For example, I suspect that most people in most societies believe that every society tells stories. Indeed, judging from theories of poetics, which are found in all major literary traditions, it seems that a wide range of features are cross-culturally believed to be universal features of verbal art (on these different traditions of poetic theory and some cross-cultural patterns that they manifest, see Hogan and Pandit 2004).[4] Second, even if different cultures did accept entirely different universals, it would not follow that there are no universals. That is simply a non sequitur. For example, suppose it turns out that there are no patterns to folk medical beliefs cross-culturally. That hardly means that there are no medical patterns cross-culturally.

            Despite these rather obvious problems, objections of this sort indicate that there has been considerable progress in the general reception of the study of universals. A decade ago, the commonplace was that the whole idea of universals was European. Now, humanists—including humanists who reject any cultural universality--seem to recognize that a wide range of societies, perhaps all societies, believe in universals. On the other hand, as this suggests, there is something very strange about this objection. I have just mentioned that people who reject cultural universals may accept that all societies believe in universals. But if someone accepts that all societies believe in universals, then he or she accepts at least one universal—that all societies believe in universals. It does not matter whether the specific beliefs about particular universals overlap. The existence of this one, general, shared belief would still constitute a universal. In a very real sense, then, people who make this objection simply do not believe what they are saying (i.e., that there are no cultural universals). Indeed, they are not alone in this. No anti-universalist really believes that nothing is common across cultures. This is not to say that they are lying. They are not lying. Thy simply do not realize that they believe in cultural universals.

            It may seem bizarre—and not a little aggressive—to claim that anti-universalists do not know what they believe. But, far from being strange, it is almost banal. We are all in the same boat with respect to knowing our own beliefs. Of course, we do generally know what we believe. However, we do not know it directly. We have to infer what we believe. In consequence, we—universalists and anti-universalists alike--are often mistaken. In Martha Nussbaum’s words, “as Aristotle stresses (and as Socrates showed before him), most people, when asked to generalize, make claims that are false to the complexity and the content of their actual beliefs. They need to learn what they really think” (Nussbaum 1986, 10-11). The point is empirically well-established (see, for example, Nisbett and Ross 1980, 226-27 on inference in self-explanation). Beliefs about cultural  universals are no exception. In recent years, I have found humanists increasingly open to the idea of cross-cultural patterns. However, they almost invariably qualify comments about universals with such phrases as “if there are any.” But, again, everyone who says this believes that there are patterns that recur across cultures. For example, suppose someone thinks he or she believes that there are no cross-cultural patterns in cuisine. The person still believes that people in other cultures eat, that they care about eating, that they care about having enough to eat, that they do things with plants and animals in order to get enough to eat, that they eat things that give them sufficient nutrients to survive, and so forth. No one is a cultural relativist on everything. But if one is not a cultural relativist on everything, then one necessarily believes in some universals. Thus the only issue should be what the universals are, not whether or not there are any.

            There seem to be two significant reasons for the common failure to recognize that, in fact, we all believe in universals.[5] First, we tacitly concern ourselves with the wrong level of particularity. In effect, we tend to assume unreflectively that a universal must be specific enough to guide behavior. In the case of literature, we regularly ignore patterns that are too general to orient the production or interpretation of a particular work. “Having a standard line length” cannot, in practice, guide the writing or analysis of a poem in a particular tradition. “Having iambic pentameter as the standard meter” can guide both. Suppose I want to write a poem in the Persian tradition. I cannot write the poem simply by knowing that Persian literature has genres. I need to know details about a specific form, say the ghazal. The same point holds for English literature and the sonnet. Similarly, I cannot evaluate the relation of a given poem to Persian tradition if I know only that the tradition has genres. I need to know particulars, so that I can judge whether it is violating conventions, combining them in an unusual manner, or simply repeating clichés. Put differently, our evaluation of information is almost always related to potential action. Our actions with respect to literature involve a degree of specificity that is often below that of our tacit beliefs about universals. In practice, we do not self-consciously think of “having a standard line length” or “having genres” as features of any tradition since they cannot guide action. We do self-consciously think of “having iambic pentameter as a standard line length” or “having sonnets” as a feature of a given tradition. Most often, we would self-consciously think of a property as a literary universal only if it has this degree of practically consequential specificity (e.g., if all traditions had iambic pentameter as the standard line length).

            The same point holds about other cultural universals. We believe that all societies have language. But that is not a belief that is functional for us. To communicate with someone, we need to know their particular language. Similarly, to feed someone from another culture or to eat food from another culture, we need to know things that are much more specific than that the people care about eating or eat things that give them nutrients. This is not to say that there are no universals at a more specific level. There are. But the universals we tacitly believe in without any research tend to be “non-operational” (i.e., they cannot serve to guide practical activity).

            On the other hand, guiding practical activity is a relative concern. To see this, we only need to imagine encountering a species of space aliens who do not care about eating (perhaps they survive on water and sunshine). Knowing that all people care about eating does have practical consequences. So does recognizing that all societies have language or that different traditions have genres and standard poetic line lengths. Thus functionality alone does not explain even these cases. This leads us to the second, and more fundamental reason that we do not recognize our belief in universals. Our brains are simply designed in such a way as to habituate to sameness and to focus our attention on differences (see, for example, Frijda 1986, 318-20 on habituation and 272-73, 318, and 386 on attention to novelty or difference from expectation). In driving down a suburban street, we do not take notice of the fact that the first house has a lawn, the second house has a lawn, the third house has a lawn, and so forth. However, we do notice that one house has only a dirt lot, while another has a pond and a rock garden. The same point holds for food, language, and stories. Tacitly believing in a particular universal is cognitively related to ignoring that universal. Our implicit assumption of its universality is precisely what keeps us from noticing it.

The Beginning of a Research Program

            Once again, however, the mere fact that people say such things as “what universals there are, if any,” suggests a sea change in attitudes toward the study of universals. The topic of literary and related universals is no longer dismissed as beneath discussion, even by those who dispute the existence of such universals. Rather, discussions of the topic have begun to appear with surprising frequency in books, articles, and conferences.[6]

            The essays in this special issue, which would itself have been unimaginable ten years ago, are prime examples of this change in attitude. Moreover, they are not simply instances of universalist study that has come to be tolerated in a more open-minded academy. They are part of a growing body of work that is beginning to point beyond mere acceptance of a research topic toward an active professional research program.

            The following essays span different literary traditions and periods. They consider different aspects of literature and formulate different kinds of possible universals. Most importantly, none of the authors claims to have the last word on the topic. Each contribution is a sort of field report on the state of research now. In that way, they are perhaps less important for their positive claims than for the future research toward which they implicitly direct us.

            More exactly, in the first essay, I take up imagery. It is somewhat surprising that imagery of any sort would be a cross-culturally recurring feature of literary works. Since it has no direct function in narrative, it would seem to be an unlikely candidate for universality. Yet we find it across a wide range of literary and oral traditions, probably all such traditions. After considering this very general universal, I turn to particular types and uses of imagery, isolating universal patterns in these as well. Finally, I consider some culture-specific image patterns, arguing that these too are part of broader, universal patterns. Put differently, cultural specificity in image patterns is not really a matter of cultural uniqueness or extreme cultural difference. Rather, it is a matter of variation on universal themes.

            Vito Evola’s contribution considers universal patterns in the reception and interpretation of religious texts. First, he points to recurring features in the context for reading sacred texts. Second, he isolates a common assumption that the text involves direct address to the reader. Finally, he considers the view of devotees in different religious traditions that the sacred text has direct application to their own lives—indeed, that it is absolutely important and absolutely relevant to their lives. Expanding slightly on Evola’s ideas, we might say that most interpretation of pragmatic messages is highly selective. When given driving instructions, we do not pay attention to the precise word choice and the multiple resonances of the phrasing. It does not matter if we are told “turn right at the petrol station,” “go west on Washington,” or “take the right turn at George’s Place.” Rather, we select what information is relevant to our specific, practical goals. When interpreting poetry, in contrast, we tend to take everything as relevant. It matters whether the work says “turn right at George’s Place” or mentions “Washington” and “petrol.” But there is not necessarily any direct, pragmatic significance to this. Evola’s argument is, in effect, that the interpretation of sacred texts takes up the pragmatism of ordinary interpretation, but absolutizes it by making all aspects of the text relevant, as in literary interpretation.

            Kuipers turns to an issue that has not previously been considered in the context of universals or, as far as I am aware, cross-cultural comparative study—anthologization. Anthologies are clearly important. They are central to the establishment of literary canons, and literary canons are central to the development of literary traditions, the formation of new works in those traditions, and so forth. More generally, the advent of writing creates a particular problem for verbal art, a problem absent in oral cultures. Writing preserves stories and poems in such a way that we are repeatedly presented with the problem of selection. Writing also spatializes literary works and thus introduces issues of organization that are different from those found in oral traditions. Kuipers explores these topics, showing that editors in different traditions have repeatedly faced the same problems—and responded to them in much the same way.

            The preceding essays all take a fairly broad view of literary features. This is commonly the case in studies of literary universals. Wehrs shows that the study of universals can encompass the interpretation of individual works as well. Specifically, Wehrs sets out to isolate a recurring theme bearing on empathy. I have shown the importance of certain sorts of empathic concern across a range of literatures (see chapter four of Hogan 2003a). Wehrs deepens and particularizes our understanding of empathy by drawing on neurological research. Specifically, he points to the close interrelation between bodily experience and empathy. He then explores the literary development of this relationship across different periods in the European tradition from the rise of capitalism to the present. (The works which I had cited in connection with empathy were drawn from a range of traditions, but largely preceded this time period.) Remarkably, Wehrs is able to integrate his isolation of this pattern across historical periods with an awareness of historical particulars, thereby producing insightful interpretations of individual works. Interpretation is particularistic; it aims to show what is unusual about the work under consideration. For this reason, it is usually seen as opposed to the isolation of broad patterns. However, we can only recognize individuality when we know what the broad patterns are, thus when we know what is not individually unique--just as we can only isolate broad patterns by looking at individuals. Far from being incompatible, then, the study of universals and the study of particulars are inseparable. Wehrs’s essay illustrates the degree to which interpretation—including historically situated interpretation sensitive to changes in political economy--may benefit from an awareness of universals.

Universalism Today: Three Distinctive Features

            Though a genuine research program in literary universals is only in its initial stages, the studies of universalism that have appeared in recent years are not entirely without precedent. For example, some obvious precursors are to be found in writings by Carl Jung and his followers. They too sought patterns across unrelated literary traditions. However, the similarities between current research on literary universals and Jungian claims about archetypes more or less end with that general aim. The methods used to isolate universals are different; the precise universals isolated are different; the psychological principles used to orient research, articulate universals, and explain those universals are different.

            It is worth highlighting three distinctive features of the study of literary universals today. These features oppose current research not only to Jungianism, but to common (non-Jungian) presumptions about literary universals as well.  

            First, current research in universals recognizes the empirical nature and complexity of universals. The technical definition of universals, given above, stresses that the idea is fundamentally statistical. Universals that apply to all traditions are distinguished by the phrase “absolute universals.” When one wishes to highlight the non-absolute nature of a given universal, one refers to it as “statistical.”[7]

            This is not the only important distinction among universals. Universals may also be implicational or typological. An implicational universal pairs two features such that the presence of one feature entails the presence of the other (statistically or absolutely). This is commonly articulated by some “if . . . then . . .” phrase. For example, it may be that not all traditions have tragedies. However, if they do have tragedies, then the tragic ending commonly has certain properties (e.g., it is irreversible). Kiparsky gives a formal example of an implicational universal. He explains that alliteration “seems to be found as an obligatory formal element only in languages where the stress regularly falls on the same syllable in the word, which then must be the alliterating syllable” (Kiparsky 1981, 9).

            Typological universals in effect bring together a set of implicational universals, such that traditions of a certain type may be assumed to have a series of specific features, statistically or absolutely. For example, not all traditions have writing. Thus not all traditions have literature in the narrow sense of verbal art in writing.[8] Rather, some have oral verbal art or orature, but not written literature. In connection with this, it is clear that, in addition to universals common to orature and written literature, there are properties that are particularly likely to appear in orature and properties that are particularly likely to appear in written literature. Most obviously, techniques of composition in orature are different from those in written literature. Among other things, the former involves a set of mnemonic devices (see, for example, Rubin 1995 or Ong 1982). The differences between orature and written literature form themselves into two complex sets defining typological universals. Verbal art of the oral type has certain distinctive characteristics. Verbal art of the written type has certain distinctive characteristics.[9]

            One sub-type of implicational and typological universals is perhaps particularly significant for literary study. Contrary to the common view that universals are incompatible with historical and cultural particularity, certain implicational and typological universals may be social and historical. For example, some literary properties may be associated with mercantile economies, while others are associated with feudal economies, as Hauser suggests (see Hauser 1957, 49).

            These and other varieties of universals are important, not only in themselves, but--for our purposes more significantly--due to their place in a research program. A research program strives to turn statistical universals into absolute universals. It works to coordinate implicational universals into typological universals,[10] and then develop typological universals into absolute universals. Recognizing the variety of universals is an important part of developing a research program, because it allows for the initial isolation of minimal patterns, thereby giving such a program a legitimate and achievable starting point. It then orients researchers toward the progressive refinement and extension of those patterns.

            The essays in this issue illustrate the point nicely. The universals of imagery are likely to be absolute universals. In connection with this, I cite a range of genetically and areally unrelated traditions, including both written and oral works of verbal art. On the other hand, it is always possible to expand the range of traditions considered. Research in those traditions will almost certainly entail refinements in my hypotheses. Evola’s contribution treats a possible typological universal—necessarily so, for not all traditions have sacred texts. In developing his argument, Evola makes reference to Hinduism and Islam, along with Christianity. His article immediately suggests several tasks for future research. First, even the treatments of these religions are necessarily preliminary and point toward a more thorough examination of their various traditions of scriptural hermeneutics. Second, there is the issue of whether Evola’s claims apply to other traditions with sacred texts. In other words, researchers need to determine the precise scope of the typology. Finally, study of this topic should ultimately include the devotee’s relation to sacred utterances in oral traditions. In other words, it should eventually consider how the typological universal may be developed into an absolute universal. Kuipers’ isolation of a recurring process across editions of anthologies presents us with a possible historical implicational universal. Kuipers’ analysis also suggests the incorporation of this implicational universal into a typological universal. Like Evola, Kuipers focuses on a limited number of traditions—Greek, Japanese, and modern American. As Kuipers himself points out, the hypotheses put forth in his essay call out for further development in relation to other traditions in which anthologization has figured importantly. Moreover, they point toward still broader research treating the selective preservation and organization of works of verbal art by oral poets. In other words, here too there is an orientation of research from an implicational universal toward a typological universal and perhaps even to an absolute universal.

            Of course, the most important way in which hypotheses are refined and extended, in a research program on literary universals (or elsewhere), is not simply through an awareness of different sorts of universal. It is through more pointed theoretical formulations which themselves guide subsequent research. As I have just indicated, these essays, and Donald Wehrs’ contribution, point directly toward possibilities for further research. In the most basic cases, that research is more or less theoretically neutral (e.g., in the straightforward imperative to examine previously unconsidered traditions). However, as we try to refine our hypotheses further, as we try to move from statistical to absolute or implicational to typological to absolute universals, we necessarily take up theoretical issues. One reason for this is that theory organizes our attention to and construal of data. First, it suggests what sorts of patterns might be universal. In other words, it suggests where we might look for patterns. Second, once we recognize a pattern, there are usually countless ways in which we might articulate it. Theoretical principles guide us in selecting a specific construal. That construal is often crucial, for our precise formulation of a universal may make the difference between whether it turns out to be minimally statistical or absolute. Finally, theoretical principles allow us to explain universals—how they develop, why they are absolute, typological, or statistical, and so on.

            For example, it seems that there is a universal pattern of standard poetic line length. However, it is not easy to say just what that universal might be. One might think that it is a matter of having a certain number of words. That phrasing works for many, but not all traditions. It may seem that it is a matter of morphemes. That too works for many, but not all traditions. In terms of psychological theory, I have argued that standard line length is closely related to the capacity of human working memory (a related argument is found in Turner 1985, 61-108). In this context, the universal principle seems to be that standard line lengths should take about two seconds in sub-vocalization, in keeping with the capacity of working memory. This construal of the pattern, based on theoretical considerations, seems to conform more fully to the data. It moves us closer to—and perhaps even captures--an absolute universal. Moreover, it begins to suggest an explanation of the universal in terms of the capacity of working memory.[11]

            In principle, a wide range of theoretical principles might be invoked in the construal and explanation of literary universals. In practice, however, recent work has drawn almost entirely on cognitive science and associated fields. This is the second distinguishing characteristic I would like to isolate.

            Of course, to say that universalists today share a commitment to cognitivism is not to say that their theoretical principles are all identical. There are different currents of thought within cognitive science and thus within the study of universals. A few of these are particularly prominent in literary study. Specifically, there is a mainstream of cognitive science, well represented in textbooks in the field. This mainstream involves a fairly standard cognitive architecture. That architecture may be represented at different levels. The more abstract or psychological architecture is often referred to as “representational” or “symbolic.” It includes such mental structures as the auditory rehearsal loop and semantic memory. The more concrete or neurological structure is often referred to as “sub-symbolic.” It includes the standard brain structures, such as the caudate nucleus and the lateral prefrontal cortex. Cognitive scientists generally accept the ultimate identity of the symbolic and sub-symbolic levels. At the same time, they tend to see both as necessary for understanding human cognition. Thus we do not fully understand the lateral prefrontal cortex if we do not relate it to working memory, and we do not fully understand working memory if we do not relate it to the lateral prefrontal cortex (on the connection between working memory and the lateral prefrontal cortex, see, for example, LeDoux 277). Both sorts of mainstream architecture play a role in the study of universals.

            In addition to the standard architecture, there are particular schools of cognitive science that elaborate on or redefine that architecture in particular ways. These have different degrees of acceptance in the field as a whole. In literary study, there have been two particularly influential schools, one articulated primarily at the symbolic level and the other articulated primarily at the sub-symbolic level. The first is the cognitive linguistics of George Lakoff and his followers, including Mark Turner’s and Gilles Fauconnier’s conceptual blending theory, which developed out of cognitive linguistics. The second is the collection of neurologically based theories (e.g., the somatic marker hypothesis) developed by Antonio Damasio.

            Here, too, the following essays are nicely illustrative of current trends. I myself draw almost entirely on mainstream cognitive architecture at the symbolic level. Evola works within the Turner and Fauconnier paradigm, though he combines blending theory with reference to the sub-symbolic realm of neurology. Finally, Wehrs takes up the work of Damasio, developing an understanding of literary universals directly out of Damasio’s ideas.

            I have just been speaking about different cognitive architectures that operate in our construal and explanation of literary universals. “Explanation” here refers to accounts of just what psychological principles underlie literary universals—in this case, what cognitive structures and processes are manifest in recurring features of literary works, as well as the reception and production of those works. [12] Thus, in the example given above, the limited capacity of working memory serves as a partial explanation of just why standard line lengths tend to be about two seconds long in subvocalization time. (It is a partial explanation for one also needs to know why working memory would be relevant to line length; for a discussion of this point, see Hogan 2003a, 39-40.)

            But this is not all there is to explanation in cognitive science. Cognitive scientists commonly ask not only how a certain universal operates in terms of cognitive architecture. They also ask how it arose initially and how it is propagated. All universals of human activity—indeed all human activities, whether universal or not--must work their way through human cognitive architecture. That is why “proximate” or “primary” explanations, as we might call them, make reference to cognitive architecture. However, nothing follows from this about the precise origins or propagation of any particular activity. Most importantly, that activity need not have a biologically genetic means of transmission or an evolutionary origin simply because its operation is open to neurological explanation. In other words, “extended” or “secondary” explanation (of genesis and transmission) needs to be addressed separately from primary explanation. It cannot simply be inferred from the primary explanation.

            Consider baseball. One could not play baseball if it were incompatible with one’s neurocognitive architecture. When a batter looks at a pitch, then hits the ball, he or she is relying on a highly complex series of perceptual, motor, and other neurocognitive processes—including, for example, the spontaneous calculation of the trajectory of the ball (see van Leeuwen 1998, 272). The actual practices of baseball must be explained in part by reference to neurocognitive architecture. However, nothing follows from this about the origin or propagation of baseball. For example, the necessary embedment of baseball playing in the capacities of a biological organ—the human brain--does not entail that baseball has a biological origin or is propagated biologically. In fact, baseball is not passed on genetically, and it did not arise as a genetic mutation. It is passed on socially, and its origin is social.

            Though baseball is not a cultural universal, the same points apply to cultural universals. They must work through our brains. But, here too, nothing follows from this about their origin. While this is commonly recognized in the case of baseball, it is not widely understood in the case of universals. As a result, the default account of the origin and promulgation of any universal tends to be evolutionary, thus genetic. This leads to the third distinguishing characteristic of universalism today. Recent work on literary or other cultural universals tends to draw on evolutionary psychology for secondary explanation, often without giving any serious consideration to other options. In my personal view, the other distinguishing characteristics of current universalism are a vast improvement over earlier tendencies.[13] Moreover, I am entirely in favor of evolutionary explanations. However, I am deeply disturbed by this narrowing of the options for secondary explanation. Evolutionary accounts are not the only possible secondary explanations of cultural universals. Moreover, they are often not the best explanations for those universals.

Secondary Explanation and Literary Universals: The Future of a Research Program

            More precisely, beyond evolution and genetic determinacy, there are at least four other possible sources of cultural universals: the physical universe, bodily experience, childhood development, and group dynamics. It should be obvious that there are regularities in each of these areas that would lead to cross-cultural patterns. Yet, they are almost entirely ignored in current treatments of universalism. Bodily experience is the only partial exception to this. Writers in cognitive linguistics have discussed the importance of bodily direction for conceptualizing time (e.g., in “moving toward the future”), bodily position for metaphorizing well-being (as in feeling “up” or “down”), and so forth. Clearly, it would be redundant to posit some genetically specific account of cross-cultural patterns in the conceptualization of time or the metaphorization of well-being when those patterns may be explained far more simply by undeniable facts of bodily experience. The same point holds for patterns explained by regularities of the physical universe, childhood development, and group dynamics. Here, I believe, psychoanalytic accounts are, in general, superior to cognitive approaches—not necessarily in their particular explanations, but in their emphasis on childhood development and bodily experience. Moreover, I find Marxist and related approaches valuable in treating group dynamics as a source of universals.

            Again, this is not to say that I am in any way opposed to evolutionary and biological explanations. In fact, I am fully supportive of such accounts. In many cases, they are superior to explanations in terms of development or group dynamics. However, in many cases they are not. In treating the origin and dissemination of a universal, one should try to give the most complete and parsimonious explanation of the data. This cannot be done by deciding a priori that one approach to secondary explanation is presumptively correct. That is a religious attitude, not a scientific one.

            Consider, by way of illustration, the narrative universal of romantic tragi-comedy, discussed at length in The Mind and Its Stories. With respect to primary explanation, I argued that a prototype of romantic tragi-comedy is formed in part by experience of particular stories. Jones has a prototype of romantic tragi-comedy because, among other things, he has read or heard romantic tragi-comedies. Additionally, the narrative prototype is interrelated with a set of other cognitive contents through shared cognitive processes and structures. Most importantly, it is connected with a prototype for personal happiness—roughly, romantic union. Narratives generally concern some character striving to achieve a goal. That goal is almost always some form of happiness, most often a prototypical form. The prototype of personal happiness in romantic union defines the hero’s or heroine’s goal for romantic tragi-comedy, distinguishing this genre from the other universal genres of heroic and sacrificial tragi-comedy (which involve contextually distinct social and physical happiness prototypes). Jones develops his prototype for personal happiness from a wide range of sources. These sources include stories, but they also include observations of other people’s behavior, Jones’s own experiences of affection or attachment and sexual desire, and so forth.

            Thus, for any given individual, stories are in part responsible for the proximate generation of the personal happiness prototype, and the personal happiness prototype is in part responsible for the generation of the prototype for romantic tragi-comedy. Does this mean that the two should have the same status in an explanatory theory? I do not believe so. The personal happiness prototype does not rely on the existence of romantic tragi-comedies. However, one cannot have romantic tragi-comedies without the personal happiness prototype. Put differently, the narrative prototype incorporates the happiness prototype, but not vice-versa. Moreover, emotions seem more fundamental to human existence than formal, public narratives of the sort treated in the study of literary universals. Thus one would expect emotion prototypes to be more fundamental than story prototypes. Given all this, it seems most plausible for our secondary explanation of the narrative prototype to begin with the happiness prototype, rather than vice-versa. In other words, it makes most sense to say that romantic tragi-comedy is a narrative universal because romantic union is a universal happiness prototype, rather than saying that romantic union is a universal happiness prototype because romantic tragi-comedy is a universal narrative prototype.

            But how, then, do we explain the universality of the prototype for personal happiness? It is clear that emotion prototypes are somehow bound up with biologically given emotions. Does this mean that the prototype itself is innate, thus derived from evolution? That seems very unlikely. Indeed, such an account would almost certainly create more problems than it solves. For example, it seems clear that prototypes are not generally innate. Why would happiness prototypes in particular be innate and not others? What sort of genetic mutation or history of (individually adaptive) genetic mutations could produce just those contents of semantic memory? The whole idea appears implausible.

            In seeking to explain the prototype for personal happiness, then, I do not begin with innate emotion prototypes. However, I do begin with a set of innate emotion systems. Specifically, in The Mind and Its Stories, I drew on research from affective  neuroscience and elsewhere to propose a set of emotion systems that did not include romantic love at all (not even in a non-prototypical form), but that did include affection or attachment and sexual desire. The adaptive function of sexual desire is too obvious to require spelling out. (Recall that a trait has an adaptive function whenever it increases the likelihood that the organism with the trait will reproduce.) The adaptive value of attachment is only slightly less obvious. Attachment is crucial in parent/child bonding. That bonding is itself crucial for the care and thus survival of offspring. If one has lots of sex and lots of babies, but they all die from neglect, one will not pass on one’s genes. Attachment helps to prevent this from happening.

            Oatley and others have reported research showing that emotion episodes may involve the simultaneous experience of distinct emotions (see Oatley 104 and citations). In light of this, I went on to maintain that romantic love involves such a simultaneous experience of attachment and sexual desire for a single object, but a simultaneous experience that is not confined to a single emotion episode. Specifically, romantic love involves the joining of attachment and sexual desire, the close interconnection of the two such that, at least in certain contexts, the behaviors, the imaginations, the affective responses of the two emotion systems become closely interrelated. For example, affectionate and sexual forms of physical contact may come to overlap extensively, such that certain modes of sexual approach assume the manner of affectionate approach. This is in part a matter of biological congruence. For example, affection and lust both involve approach to the object. In contrast, it would be difficult to combine fear with sexual desire, since the former entails withdrawal from the object. However, this congruence simply allows for the joining of the emotions. It does not produce it. Rather, the joining results, I believe, from universals of childhood development and from group dynamics. For example, in infancy, the child’s bond with his or her caretaker—commonly, the mother—brings together attachment and physical intimacy in semantic structures, such as prototypes, and more importantly in emotional memories. (Emotional memories are memories that retain the emotional tone of the original experience. When activated, we experience the feeling again, even though we may not self-consciously recall the precipitating events [see, for example, LeDoux 1996, 182].) Sexual desire is, of course, inseparable from a parallel sort of physical intimacy. In another context, Frijda maintains that “Kissing, caressing” of “children and . . . partners in love” are “part of a biologically prepared care-giving behavioral system” (Frijda 1986, 104). Frijda does not try to account for why this care-giving system would be so important for the sexual behavior of partners in love. My contention is that a shared connection with physical intimacy (kissing and caressing, as well as frontal bodily skin contact, and so forth) fosters the interaction of the two emotion systems—first, through emotional memories; second, through imagination (guided by prototypes). This results in the “widespread occurrence cross-culturally” of affection-based behavior in romantic union (Frijda 1986, 104).

            While developmental factors are probably fundamental for this fusion, there are important social factors as well. As societies develop, those with stable systems of mating and child rearing are more likely to be successful than societies without stable systems of mating and child rearing. One way of fostering stable systems of this sort is through marriage practices (which are largely coercive—in, for example, restricting divorce). But another way is by fostering the fusion of affection and sexual desire. Sexual desire is periodic, thus variable in intensity, and somewhat unstable in its objects. However, affection tends to be enduring, more constant in intensity, and more consistent in its objects. It also has other stabilizing features, such as association with a sense of home.[14] As a result, intertwining sexual desire and affection should encourage stability in mating and child rearing, particularly when combined with marriage practices. Thus, unsurprisingly, such intertwining arises independently as an ideal in different societies. Or, rather, when it arises as an ideal among some individuals in a particular society, presumably for developmental reasons, it acquires this social function and thus spreads. For example, once some people see this as an ideal (perhaps entirely unselfconsciously), their pursuit of the ideal should to some extent stabilize their mating and child rearing practices.[15] As a result, the ideal will be sustained and extended through those child-rearing practices, both directly and as an example for others.

            The account I have just sketched (in a very simplified form) may or may not turn out to be accurate. That is a question for future research. However, it illustrates the sort of secondary explanatory account that is, I believe, most appropriate in the study of literary and other cultural universals. It incorporates innatism and evolutionary psychology, but also developmental patterns and features of group dynamics (and, to some extent, bodily experience). In short, in formulating this account I have tried to give a secondary explanation that covers a broad range of data, but that also satisfies the fundamental methodological principle of minimal interpretation, the principle that adjures us not to multiply explanatory elements unnecessarily. It seems clear that childhood development does operate to connect attachment with physical intimacy. It seems clear that groups develop in such a way as to foster stability in mating and child rearing. Insofar as these and related factors combine to yield a secondary explanation for the personal happiness prototype, the further positing of genetic mutations is redundant—and therefore poor science.[16]

            Though the following essays do not address the issue of secondary explanation, it will necessarily be an important part of any future research program on literary universals.

            In sum, the study of literary universals, and the broader study of cultural universals, are increasingly important areas of research. Though they certainly have not vanished, the past prejudices of humanists and social scientists have diminished. Interest in the study of literary universals may not (yet) have translated directly into a coordinated research program. Nonetheless, in comparison with some earlier work on cross-cultural literary patterns, the prospects for future research in the area are, I believe, unusually bright. This is due, first of all, to the general awareness of different types of universal, and of the relation among these types. That awareness helps to organize the field and to orient future research in a useful direction. Even more importantly, the thorough integration of this research into cognitive science makes it far more likely that contemporary universalists will follow rigorous research procedures and will isolate patterns of genuine significance.

            On the other hand, there are some limitations to current work on literary universals. First, there has been a somewhat narrow range of reference within cognitive science. A few schools have tended to dominate literary cognitivism, with much valuable work from the larger field of cognitive science being ignored. More importantly, there has been excessive reliance on innatism and evolutionary psychology in secondary explanation. Of course, there are vigorous counter-trends in both areas. Moreover, the general preference of humanists has been strongly in favor of historicism and social constructionism. As a result, it is unlikely that literary study will be overwhelmed by biological determinism.

            Once again, the work collected in this issue represents some of the main tendencies in current universalism. More importantly, it points toward areas of future empirical and theoretical study in an ongoing research program. Indeed, as I have already suggested, the value of these contributions may be less in the specific claims they put forward and more in the future research they suggest. Such ongoing research is, after all, the only means by which literary universals may be isolated, refined, and explained, at both the proximate level of immediate functioning and at the more distant level of origin and transmission. It is, in short,  the only way in which we can genuinely advance our understanding of the human mind and its verbal art.[17]

Works Cited

Appiah, Kwame Anthony. 1992. In My Father’s House: Africa in the Philosophy of Culture. New York: Oxford UP.

Bordwell, David. 2005. Figures Traced in Light: On Cinematic Staging. Berkeley, CA: U of California P.

Carroll, Joseph. 2004. Literary Darwinism: Evolution, Human Nature, and Literature. New York: Routledge.

Coste, Didier. 2004. “Is a Non-Global Universe Possible? What Universals in the Theory of Comparative Literature (1952-2002) Have to Say About It.” Comparative Literature Studies 41.1: 37-48.

Damrosch, David. 2003. What Is World Literature? Princeton, NJ: Princeton UP.

Frijda, Nico H. 1986. The Emotions. Cambridge: Cambridge UP and Paris: Editions de la Maison des Sciences de l’Homme.

Gottschall, Jonathan. 2005. “The Heroine with a Thousand Faces: Universal Trends in the Characterization of Female Folk Tale Protagonists.” Evolutionary Psychology 3: 85-103.

Hauser, Arnold. 1957. The Social History of Art (Volume I): Prehistoric, Ancient-Oriental, Greece and Rome; Middle Ages. Trans. Stanley Godman. New York: Vintage Books.

Herman, David, Manfred Jahn, and Marie-Laure Ryan, eds. 2005. The Routledge Encyclopedia of Narrative Theory. London: Routledge.

Hogan, Patrick Colm. 1994. “The Possibility of Aesthetics.” British Journal of Aesthetics 34.4): 337-49. (Reprinted on Cogweb, http://cogweb.ucla.edu.)

---. 1996. On Interpretation: Meaning and Inference in Law, Psychoanalysis, and Literature. Athens, GA: U of Georgia P.

---. 2003a. The Mind and Its Stories: Narrative Universals and Human Emotion. Cambridge: Cambridge UP and Paris: Editions de la Maison des Sciences de l’Homme.

---. 2003b. Cognitive Science, Literature, and the Arts: A Guide for Humanists. New York: Routledge.

Hogan, Patrick Colm and Lalita Pandit. 2004. “Ancient theories of narrative (non-Western).” In The Routledge Encyclopedia of Narrative Theory. Ed. David Herman, Manfred Jahn, and Marie-Laure Ryan. London: Routledge.

Kiparsky, Paul. 1981. “The Role of Linguistics in a Theory of Poetry.” In Essays in Modern Stylistics. Ed. Donald C. Freeman. London: Methuen.

---. 1987. “On Theory and Interpretation.” In The Linguistics of Writing: Arguments Between Language and Literature. Ed. Nigel Fabb, Derek Attridge, Alan Durant, and Colin MacCabe. New York: Methuen, 185-98.

LeDoux, Joseph. 1996. The Emotional Brain: The Mysterious Underpinnings of Emotional Life. New York: Touchstone.

Nisbett, Richard and Lee Ross. 1980. Human Inference: Strategies and Shortcomings of Social Judgment. Englewood Cliffs, N.J.: Prentice-Hall.

Nussbaum, Martha C. 1986. The Fragility of Goodness: Luck and Ethics in Greek Tragedy and Philosophy. Cambridge: Cambridge UP.

Oatley, Keith. 1992. Best Laid Schemes: The Psychology of Emotions. Cambridge: Cambridge UP and Paris: Editions de la Maison des Sciences de l’Homme.

Ong, Walter J. 1982. Orality and Literacy: The Technologizing of the Word. New York: Methuen.

Panksepp, Jaak. Affective Neuroscience: The Foundations of Human and Animal Emotions. New York: Oxford UP, 1998.

Rubin, David C. 1995. Memory in Oral Traditions: The Cognitive Psychology of Epic, Ballads, and Counting-out Rhymes. New York: Oxford UP.

Sternberg, Meir. 2003. “Universals of Narrative and Their Cognitivist Fortunes.” Poetics Today 24.2: 297-395 and 24.3: 517-638.

Sugiyama, Michelle Scalise. 2003. “Cultural Variation is Part of Human Nature: Literary Universals, Context-Sensitivity, and ‘Shakespeare in the Bush.’” Human Nature 14.4: 383-96.

Turner, Frederick. 1985. Natural Classicism: Essays on Literature and Science. New York: Paragon House.

van Leeuwen, Cees. 1998. “Perception.” In A Companion to Cognitive Science. Ed. William Bechtel and George Graham. Oxford: Blackwell, 265-81.