Evolutionary Approaches to Creativity - Scott Barry Kaufman (PDF)

P1: OJL Trim: 7in × 10in Top: 0.498in Gutter: 0.871in CUUS1027-15 cuus1027/Kaufman ISBN: 978 0 521 51366 1 May 10, 2010 20:17 CHAPTER 15 Evolutionary Approaches to Creativity Liane Gabora and Scott Barry Kaufman 1. Introduction Many species engage in acts that could be called creative (J.C. Kaufman & A.B. Kauf- man, 2004). However, human creativity is unique in that it has completely transformed the planet we live on. We build skyscrap- ers, play breathtaking cello sonatas, send ourselves into space, and even decode our own DNA. Given that the anatomy of the human brain is not so different from that of the great apes, what enables us to be so creative? Recent collaborations at the fron- tier of anthropology, archaeology, psychol- ogy, and cognitive science are culminating in speculative but increasingly sophisticated efforts to piece together an answer to this question. Examining the skeletons of our ancestors gives cues as to the anatomical constraints that hindered or enabled vari- ous kinds of creative expression. Relics of the past have much to tell us about the thoughts, beliefs, and creative abilities of the people who invented and used them. How the spectacular creativity of humans came about is the first topic addressed in this chapter. Studies at the intersection of creativity and evolution are not limited to investi- gations into the biological evolution of a highly creative species. Creative ideas them- selves might be said to evolve through cul- ture. Human creativity is distinctive because of the adaptive and open-ended manner in which change accumulates. Inventions build on previous ones in ways that enhance their utility or aesthetic appeal, or make them applicable in different situations. There is no a priori limit to how a creative idea might unfold over time. A cartoon character may inspire the name and logo for a hockey team (the Mighty Ducks), which might in turn inspire toys, cereal shapes, cigarette lighter designs, or for that matter work its way into an academic book chapter. It is this procliv- ity to take an idea and make it our own, or “put our own spin on it,” that makes creative ideas appear to evolve. The next section of this chapter investigates in what sense cre- ative ideas evolve through culture. Finally, we address the question of why creativity evolved. What forces supported the evolution of creativity? Does being cre- ative help us live longer, or attract mates? 279  P1: OJL Trim: 7in × 10in Top: 0.498in Gutter: 0.871in CUUS1027-15 cuus1027/Kaufman ISBN: 978 0 521 51366 1 May 10, 2010 20:17 280 LIANE GABORA AND SCOTT BARRY KAUFMAN Do creative projects sometimes interfere with survival and reproductive fitness; are there nonbiological factors that compel us to create? This is a third topic addressed in this chapter. 2. The Birth of Human Creativity Looking at an artifact that was fashioned thousands or millions of years ago is an awe- inspiring experience because it gives us a glimpse into the lives and worldviews of our earliest ancestors. To be sure, creative works disintegrate. The farther back in time we look for signs of creativity, the fewer creative works of that time remain with us today. But by corroborating theory and data from different fields, we are on our way toward putting together a coherent picture of how and when the creative abilities of humans arose. We begin this section by examining the archaeological evidence for the earliest indi- cations of human creativity, and the anthro- pological evidence for concurrent changes in the size and shape of the cranial cavity. We then examine various hypotheses that have been put forward to explain these data. 2.1 The Earliest Evidence of Human Creativity: Homo habilis It is generally agreed that ancestral humans started diverging from ancestral apes approximately six million years ago. The first Homo lineage, Homo habilis, appeared approximately 2.4 million years ago in the Lower Paleolithic. The earliest known human inventions, referred to as Oldowan artifacts (after Olduvai Gorge, Tanzania, where they were first found), are widely attributed to Homo habilis (Semaw et al., 1997), although it is possible that they were also used by late australopithecines (de Beaune, 2004). They were simple, mostly single faced stone tools, pointed at one end (M.D. Leakey, 1971). These tools were most likely used to split fruits and nuts (de Beaune, 2004), though some of the more recent ones have sharp edges, and are found with cut-marked bones, suggesting that they were used to sharpen wood implements and butcher small game (M.D. Leakey, 1971; Bunn & Kroll, 1986). These early tools were functional but sim- ple and unspecialized; by our standards they were not very creative. Feist (2008) refers to the minds of these early hominids as pre- representational, suggesting that hominids at this time were not capable of forming rep- resentations that deviated from their con- crete sensory perceptions; their experience was tied to the present moment. Similarly, Mithen (1996) refers to minds at this time as possessing generalized intelligence, reflect- ing his belief that domain general learning mechanisms, such as Pavlovian conditioning and implicit learning (e.g., A.S. Reber, 1993), predominated. Nevertheless, the early tools of this period mark a momentous breakthrough for our species. Today we are accustomed to seeing everywhere the outcomes of what began as a spark of insight in someone’s mind, but when the world consisted solely of naturally formed objects, the capacity to imagine something and turn it into a real- ity may well have seemed almost magical. As de Baune (2004) puts it, “the moment when a hominin . . . produced a cutting tool by using a thrusting percussion . . . marks a break between our predecessors and the specifically human” (p. 142). 2.2 The Adaptive Larger-Brained Homo erectus Homo habilis persisted from approximately 2.4 to 1.5 million years ago. Approximately 1.8 million years ago, Homo erectus appeared, followed by Homo ergaster, archaic Homo sapiens, and Homo neanderthalensis. The size of the Homo erectus brain was approxi- mately 1,000 cc, about 25% larger than that of Homo habilis, and 75% of the cranial capac- ity of modern humans (Aiello, 1996; Ruff, Trinkaus, & Holliday, 1997; Lewin, 1999). Homo erectus exhibit many indications of enhanced ability to creatively adapt to the environment to meet the demands of sur- vival, including sophisticated, task-specific  P1: OJL Trim: 7in × 10in Top: 0.498in Gutter: 0.871in CUUS1027-15 cuus1027/Kaufman ISBN: 978 0 521 51366 1 May 10, 2010 20:17 EVOLUTIONARY APPROACHES TO CREATIVITY 281 stone hand axes, complex stable seasonal habitats, and long-distance hunting strate- gies involving large game. By 1.6 million years ago, Homo erectus had dispersed as far as Southeast Asia, indicating the ability to adjust lifestyle to vastly different climates (Ant´on & Swisher, 2004; Cachel & Harris, 1995; Swisher et al., 1994; Walker & Leakey, 1993). In Africa, West Asia, and Europe, by 1.4 million years ago Homo erectus developed the Aschulean hand axe (Asfaw et al., 1992), a do-it-all tool that may even have had some function as a social status symbol (Kohn & Mithen, 1999). These symmetrical biface stone tools probably required several stages of production, bifacial knapping, and con- siderable skill and spatial ability to achieve their final form. Though the anatomical capacity for lan- guage was present by this time (Wynn, 1998), verbal communication is thought to have been limited to (at best) presyntacti- cal protolanguage (Dunbar, 1996). Thought during this time period was most likely only first order; the capacity for thinking about thinking (i.e., metacognition) had not yet developed. 2.3 Possible Explanations for the Onset of Human Creativity It has been suggested that these early archae- ological finds do not reflect any underly- ing biological change but were simply a response to climactic change (Richerson & Boyd, 2000). However, given the signifi- cant increase in cranial capacity, it seems parsimonious to posit that this dramatic encephalization allowed a more sophisti- cated mode of cognitive functioning and is thus at least partly responsible for the appearance of cultural artifacts (and the beginnings of an archaeological record). There are multiple versions of the hypothesis that the onset of the archaeo- logical record reflects an underlying cogni- tive transition. One suggestion is that the appearance of archaeological novelty is due to the onset of the capacity to imitate (Dugatkin, 2001), or the onset of a theory of mind – the capacity to reason about the men- tal states of others (Premack & Woodruff, 1978). However, other species possess a theory of mind (Heyes, 1998) and imitate (Byrne & Russon, 1998; Darwin, 1871), yet they do not compare to hominids with respect to creativity. Moreover, although these hypotheses may explain how new ideas, once in place, spread from one indi- vidual to another, they are inadequate as an explanation of the enhanced capacity for coming up with new ideas in the first place. Yet another proposal is that Homo under- went a transition at this time from an episodic mode of cognitive functioning to a mimetic mode (Donald, 1991). The episodic mind of Homo habilis was sensitive to the sig- nificance of episodes, and it could encode them in memory and coordinate appropriate responses. But it could not voluntarily access them independent of cues. The enlarged cranial capacity of Homo erectus enabled it to acquire a mimetic form of cognition, characterized by possession of what Donald (1991) refers to as a “self-triggered recall and rehearsal loop,” or SRRL. The SRRL enabled hominids to voluntarily access memories independent of cues and thereby act out events that occurred in the past, or that could occur in the future (indeed the term mimetic is derived from the word “mime”). Thus not only could the mimetic mind temporarily escape the here and now, but through gesture it could bring about a sim- ilar escape in other minds. The SRRL also enabled hominids to engage in a stream of thought, such that attention is directed away from the external world toward one’s inter- nal model of it, and one thought or idea evokes another, revised version of it, which evokes yet another, and so forth recursively. Finally, the SRRL enabled the capacity to evaluate and improve motor acts through repetition or rehearsal, and adapt them to new situations, resulting in more refined artifacts and survival tactics. It seems reasonable that a larger brain might be more likely to engage in self- triggered recall and rehearsal, but Donald’s scenario becomes even more plausible when considered in light of the structure and dynamics of associative memory (Gabora,  P1: OJL Trim: 7in × 10in Top: 0.498in Gutter: 0.871in CUUS1027-15 cuus1027/Kaufman ISBN: 978 0 521 51366 1 May 10, 2010 20:17 282 LIANE GABORA AND SCOTT BARRY KAUFMAN 2003, 2010). We know that neurons are sen- sitive to microfeatures – primitive attributes of a stimulus, such as a sound of a particu- lar pitch, or a line of a particular orientation (Churchland & Sejnowski, 1992; Smolensky, 1988). Episodes etched in memory are dis- tributed across a bundle or cell assembly of these neurons, and likewise, each neu- ron participates in the encoding of many episodes. Finally, memory is content address- able such that similar stimuli activate and get encoded in overlapping distributions of neu- rons. It seems reasonable that brain enlarge- ment entails a transition from a more coarse- grained to a more fine-grained memory, such that episodes are encoded in more detail. This means there are more ways in which distributions can overlap, and thus more routes by which one can evoke another, pro- viding an anatomical basis for self-triggered recall and rehearsal, and the forging of cre- ative connections. The enhanced ability to make connections would in turn have paved the way for a more integrated internal model of the world, or worldview. 3. Over A Million Years of Creative Stasis The hand axe persisted as the almost exclu- sive tool of choice for over a million years, spreading by 500,000 years ago into Europe, where was it used until about 200,000 years ago. Indeed during this period not only is there almost no change in tool design, but little evidence of creative insight of any kind, with the exception of the first solid evidence for controlled use of fire some 800,000 years ago in the Levant (Goren-Inbar et al., 2004). 3.1 A Second Increase in Brain Size Between 600,000 and 150,000 years ago there was a second spurt in brain enlargement (Aiello 1996; Ruff et al., 1997). But although anatomically modern humans had arrived, behavioral modernity had not. It would make our story simple if the increase in brain size coincided with the burst of creativity in the Middle/Upper Paleolithic (Bickerton, 1990; Mithen, 1998), to be discussed in the next section. It does correspond with the revolutionary advancement of the Levallois flake, which came into prominence approxi- mately 250,000 years ago in the Neanderthal line. But although one sees in the artifacts of this time the germ of modern-day repre- sentational thought, it is clear that cognitive processes are still primarily first order – tied to concrete sensory experience – rather than second order – derivative, or abstract. R. Leakey (1984) writes of anatomically mod- ern human populations in the Middle East with little in the way of culture, and con- cludes that “the link between anatomy and behavior therefore seems to break” (p. 95). It may be that this second spurt in brain size exerted an impact on expressions of cre- ativity that leave little trace in the archae- ological record, such as finding ways of manipulating competitors for purposes of survival and reproduction (Baron-Cohen, 1995; Byre & Whiten, 1988; Humphrey, 1976; Whiten, 1991; Whiten & Byrne, 1997; Wil- son, Near, & Miller, 1996; Dunbar, 1996). However it is possible that what we see in the archaeological record really does reflect what was happening at the time, i.e. that there really was a rift between anatomi- cal and behavioral modernity. We e will return to this mystery after examining how the spectacular creativity of modern humans came about. 4. The Creative Minds of Modern Humans The European archaeological record indi- cates that a truly unparalleled cultural transition occurred between 60,000 and 30,000 years ago at the onset of the Upper Paleolithic (Bar-Yosef et al., 1986; Klein, 1989a; Mellars, 1973, 1989a, 1989b; Soffer, 1994; Stringer & Gamble, 1993). Considering it “evidence of the modern human mind at work,” Richard Leakey (1984:93–94) describes the Upper Pale- olithic as “unlike previous eras, when sta- sis dominated, . . . [with] change being mea- sured in millennia rather than hundreds of  P1: OJL Trim: 7in × 10in Top: 0.498in Gutter: 0.871in CUUS1027-15 cuus1027/Kaufman ISBN: 978 0 521 51366 1 May 10, 2010 20:17 EVOLUTIONARY APPROACHES TO CREATIVITY 283 millennia.” Similarly, Mithen (1996) refers to the Upper Paleolithic as the “big bang” of human culture, exhibiting more inno- vation than in the previous six million years of human evolution. At this time that we see the more-or-less simultaneous appearance of traits considered diagnostic of behavioral modernity. It marks the begin- ning of a more organized, strategic, season- specific style of hunting involving specific animals at specific sites, elaborate burial sites indicative of ritual and religion, evidence of dance, magic, and totemism, the colo- nization of Australia, and replacement of Levallois tool technology by blade cores in the Near East. In Europe, complex hearths and many forms of art appeared, including naturalistic cave paintings of animals, deco- rated tools and pottery, bone and antler tools with engraved designs, ivory statues of ani- mals and sea shells, and personal decoration such as beads, pendants, and perforated ani- mal teeth, many of which may have been used to indicate social status (White, 1989a, 1989b). Indeed, White (1982:176) also writes of a “total restructuring of social relations.” What is perhaps most impressive about this period is not the novelty of any particu- lar artifact but that the overall pattern of cultural change is cumulative; more recent artifacts resemble older ones but have modifi- cations that enhance their appearance or fu- nctionality. This appears to be uniquely hu- man (Donald, 1998) and it has been referred to as the ratchet effect (Tomasello, 1999). Despite a lack of any overall increase in cranial capacity, there was a signifi- cant increase in the size of the prefrontal cortex – and particularly the orbitofrontal region (Deacon, 1997; Dunbar, 1993; Jeri- son, 1973; Krasnegor, Lyon, & Goldman- Rakic, 1997; Rumbaugh, 1997), and it was likely a time of major neural reorganiza- tion (Klein, 1999; Henshilwood, d’Errico, Vanhaeren, van Niekerk, & Jacobs, 2004; Pinker, 2002). These brain changes may have given rise to what Feist (2008) refers to as “meta-representational thought,” or the abil- ity to reflect on representations and think about thinking. Along similar lines, Den- nett (1976) suggests that an important tran- sition in the evolution of Homo sapiens is from first-order intentionality to second- order intentionality. A first-order intentional system has beliefs and desires but cannot reflect on those beliefs and desires, whereas second-order intentional system has beliefs and desires about the beliefs and desires of themselves and others. Whether this period was a genuine rev- olution culminating in behavioral moder- nity is hotly debated because claims to this effect are based on the European Paleolithic record, and largely exclude the African record (McBrearty & Brooks, 2000; Hen- shilwood & Marean, 2003). Indeed, most of the artifacts associated with a rapid transition to behavioral modernity between 40,000 and 50,000 years ago in Europe are found in the African Middle Stone Age tens of thousands of years earlier. These include blades and microliths, bone tools, specialized hunting, long-distance trade, art and decoration (McBrearty & Brooks, 2000), the Berekhat Ram figurine from Israel (d’Errico & Nowell, 2000), and an anthropomorphic figurine of quartzite from the Middle Ascheulian (ca. 400 ka) site of Tan-tan in Morocco (Bednarik, 2003). Moreover, gradualist models of the evolu- tion of behavioral modernity well before the Upper Paleolithic find some support in archaeological data (Bahn, 1991; Har- rold, 1992; Henshilwood & Marean, 2003; White, 1993; White et al., 2003). If mod- ern human behaviors were indeed gradu- ally assembled as early as 250,000 to 300,000 years ago, as McBrearty and Brooks (2000) argue, it pushes the transition into align- ment with the most recent spurt in human brain enlargement. However, the traditional and probably currently dominant view is that behaviorally modern humans appeared in Africa approximately 50,000 years ago, and spread throughout in Europe, replac- ing others who had not achieved behavioral modernity, including the Neanderthals (e.g., Ambrose, 1998; Gamble, 1994; Klein, 2003; Stringer & Gamble, 1993). From this point onward, anatomically and behaviorally modern Homo sapiens were the only living hominids.  P1: OJL Trim: 7in × 10in Top: 0.498in Gutter: 0.871in CUUS1027-15 cuus1027/Kaufman ISBN: 978 0 521 51366 1 May 10, 2010 20:17 284 LIANE GABORA AND SCOTT BARRY KAUFMAN 4.1 Cognitive Explanations Whether one believes the change happened gradually or suddenly, it is accepted that the Middle/Upper Paleolithic was a period of unprecedented creativity. How and why did it occur? What kind of cognitive processes were involved? We now review the most popular hypotheses for what kind of biolog- ically evolved cognitive advantages gave rise to behavioral modernity at this time. 4.1.1 ADVENT OF SYNTACTIC LANGUAGE It has been suggested that humans under- went at this time a transition from a predom- inantly gestural to a vocal form of commu- nication (Corballis, 2002). Although owing to the ambiguity of the archaeological evi- dence we may never know exactly when language began (Bednarik, 1992:30; David- son & Noble, 1989), most scholars agree that although earlier Homo and even Nean- derthals may have been capable of primitive protolanguage, the grammatical and syntac- tic aspects of language emerged near the beginning of the Upper Paleolithic (Aiello & Dunbar, 1993; Bickerton, 1990, 1996; Dun- bar, 1993, 1996; Tomasello, 1999). Carstairs- McCarthy (1999) presents a modified version of this proposal, suggesting that although some form of syntax was present in the earliest languages, most of the later elab- oration, including recursive embedding of syntactic structure, emerged in the Upper Paleolithic. Syntax enabled language to become general purpose and put to use in a variety of situations. It enhanced not just the ability to communicate with others, spread ideas from one individual to the next, and collaborate on creative projects (thereby speeding up cultural innovation), but also the ability to think things through precisely for oneself and manipulate ideas in a con- trolled, deliberate fashion (Reboul, 2007). 4.1.2 SYMBOLIC REASONING Another suggestion is that the creativity of the Middle/Upper Paleolithic was due to the emergence of an ability to internally rep- resent complex, abstract, internally coher- ent systems of meaning, including sym- bols and the causal relationships amongst them (Deacon, 1997). According to Dea- con, we shifted from iconic representation, in which the representation physically resem- bles what it ‘stands for’, to indexical represen- tation, in which the representation implies or ‘points’ to the thing it stands for, to sym- bolic representation, in which there is no sim- ilarity or implied relationship between the representation and what it stands for. Dea- con claims that the onset of symbol use colored our existence by making us view objects and people in terms of the roles they could play in stories, and the point or mean- ing they could potentially have, or partici- pate in. 4.1.3 COGNITIVE FLUIDITY It is undoubtedly the case that symbolic representation plays a fundamental role in the mental life of modern humans. Oth- ers however believe that the transition from iconic to indexical to symbolic representa- tion was a secondary consequence of onset of the intuitive, divergent, associative pro- cesses by which we unearth relationships of correlation (or roughly, similarity), such as through the discovery of analogies. Faucon- nier & Turner (2002) propose that the excep- tional creativity of the Middle/Upper Pale- olithic was due to the onset of the capacity to blend concepts, which facilitated anal- ogy formation and the weaving of expe- riences into stories and parables. A simi- lar explanation is put forward by Mithen (1996), drawing on the evolutionary psy- chologist’s notion of massive modularity (Buss, 1999/2004; Cosmides & Tooby, 1992; Dunbar et al., 1994; Rozin, 1976; for an extensive critique see Buller, 2005). Mithen suggests that the creativity of the modern mind arose through the onset of cognitive fluidity, resulting in the connecting of what were previously encapsulated (functionally isolated) brain modules devoted to natu- ral history, technology, social processes, and language. This he claims gave us the ability to map, explore, and transform conceptual spaces, referring to Boden’s (1990) definition of a conceptual space as a “style of think- ing – in music, sculpture, choreography,  P1: OJL Trim: 7in × 10in Top: 0.498in Gutter: 0.871in CUUS1027-15 cuus1027/Kaufman ISBN: 978 0 521 51366 1 May 10, 2010 20:17 EVOLUTIONARY APPROACHES TO CREATIVITY 285 chemistry, etc.” Sperber (1994) proposes that the connecting of modules involved a special module, the “module of meta- representation” or MMR, which contains “concepts of concepts,” and enabled cross- domain thinking, and particularly analogies and metaphors. Note that the notion of modules amounts to an explicit high-level compartmentaliza- tion of the brain for different tasks. How- ever, this kind of division of labor – and the ensuing creativity – would emerge unavoid- ably as the brain got larger without explicit high-level compartmentalization, owing to the sparse, distributed, content-addressable manner in which neurons encode informa- tion (Gabora, 2003; 2010). As noted ear- lier, neurons are tuned to respond to dif- ferent microfeatures, and there is a sys- tematic relationship between the content of a stimulus and the distributed set of neurons that respond to it, such that neu- rons that respond to similar microfea- tures tend to be near one another. Thus, as the brain got larger, the number of neurons increased, and the brain accord- ingly responded to more microfeatures, so items could be encoded in more detail. Neighboring neurons tended to respond to microfeatures that were more similar, and distant neurons tended to respond to micro- features that were more different. Therefore there were more ways in which distributed representations could overlap and creative connections could be made. Thus a weak modularity of sorts emerges naturally at the neuronal level without any explicit high- level compartmentalization going on, and it need not necessarily correspond to how humans carve up the world, that is, to cat- egories such as natural history, technology, and so forth. Moreover, explicit connecting of modules is not necessary for creative con- nections to be made; all that is necessary is that the relevant domains be simultaneously accessible. 4.1.4 CONTEXTUAL FOCUS The above proposals for what kind of cognitive change could have led to the Upper Paleolithic transition stress different aspects of cognitive modernity. Acknowl- edging a possible seed of truth in each of them, we begin to converge toward a common (if more complex) view. Concep- tual blending is characteristic of divergent or associative thought, which tends to be automatic, intuitive, and diffuse. This is quite different from convergent or analyti- cal thought, which tends to be logical, and controlled. It is widely believed that the modern mind engages in both (Arieti, 1976; Ashby & Ell, 2002; Freud, 1949; Guilford, 1950; James, 1890/1950; Johnson-Laird, 1983; Kris, 1952; Neisser, 1963; Piaget, 1926; Rips, 2001; Sloman, 1996; Stanovich & West, 2000; Werner, 1948; Wundt, 1896). This is some- times referred to as the dual-process the- ory of human cognition (Chaiken & Trope, 1999; Evans & Frankish, 2009) and it is con- sistent with current theories of creative cog- nition (Finke, Ward, & Smith, 1992; Gab- ora, 2002, 2010; Smith, Ward, & Finke, 1995; Ward, Smith, & Finke, 1999). Divergent pro- cesses are hypothesized to occur during idea generation, whereas convergent processes predominate during the refinement, imple- mentation, and testing of an idea. It has been proposed that the Paleolithic tran- sition reflects a mutation to the gene(s) involved in the fine-tuning of the biochem- ical mechanisms underlying the capacity to shift between these modes, depending on the situation, by varying the specificity of the activated cognitive receptive field (Gab- ora, 2003, 2010; for similar ideas see Howard- Jones & Murray, 2003; Martindale, 1995). This is referred to as contextual focus1 because it requires the ability to focus or defocus attention in response to the context or sit- uation one is in. Defocused attention, by diffusely activating a broad region of mem- ory, is conducive to divergent thought; it enables obscure (but potentially relevant) aspects of the situation to come into play. Focused attention is conducive to conver- gent thought; memory activation is con- strained enough to hone in and perform 1 In neural net terms, contextual focus amounts to the capacity to spontaneously and subconsciously vary the shape of the activation function, flat for divergent thought and spiky for analytical.  P1: OJL Trim: 7in × 10in Top: 0.498in Gutter: 0.871in CUUS1027-15 cuus1027/Kaufman ISBN: 978 0 521 51366 1 May 10, 2010 20:17 286 LIANE GABORA AND SCOTT BARRY KAUFMAN logical mental operations on the most clearly relevant aspects. Thus in an analytic mode of thought the concept giant might only acti- vate the notion of large size, whereas in an associative mode the giants of fairytales might come to mind. Once it was possible to shrink or expand the field of attention, and thereby tailor one’s mode of thought to the demands of the current situation, tasks requiring either convergent thought (e.g., mathematical derivation), divergent thought (e.g., poetry), or both (e.g., techno- logical invention) could be carried out more effectively. When the individual is fixated or stuck, and progress is not forthcoming, defo- cusing attention enables the individual to enter a more divergent mode of thought, and working memory expands to include periph- erally related elements of the situation. This continues until a potential solution is glimpsed, at which point attention becomes more focused and thought becomes more convergent, as befits the fine-tuning and manifestation of the creative work. Thus the onset of contextual focus would have enabled the hominid to adapt ideas to new contexts or combine them in new ways through divergent thought, and to fine-tune these strange new combinations through convergent thought. In this way the fruits of one mode of thought provide the ingre- dients for the other, culminating in a more fine-grained internal model of the world. 4.1.1 SHIFTING BETWEEN IMPLICIT AND EXPLICIT THOUGHT In a similar vein, it has been proposed that cognitive fluidity enabled hominids to move not just ‘horizontally’ between domains (as Mithen [1996] suggests), but also ‘verti- cally’ between implicit and explicit modes of thought (Feist, 2008). Implicit and explicit cognition map roughly onto divergent and convergent modes of thought. While explicit cognition is equated with advanced abilities such as planning, reasoning, and hypothesis- guided deduction, implicit cognition is asso- ciated with the ability to automatically detect complex regularities, contingencies, and covariances in our environment (e.g., A.S. Reber, 1993). Implicit cognition plays a significant role in structuring our per- ceptions and behavior (Berry & Broadbent, 1988; Cleeremans & Jim´enez, 2002; Hassin, Uleman, & Bargh, 2005; Lewicki & Hill, 1987; Lewicki, Czyzewska, & Hoffman, 1987; S.B. Kaufman, 2007; McGeorge & Burton, 1990; A.S. Reber, 1967, 1993; P.J. Reber & Kotovsky, 1997; Squire & Frambach, 1990). It is thought to be useful for making broad associations and arriving at creative ideas, and believed to be a fundamental aspect of our humanness (Bowers, Farvolden, & Mer- migis, 1995; S.B. Kaufman, 2008, in press). It may be that the fruits of associative or implicit processes come to awareness only once they have been honed into a form that is sufficiently well-defined that we can mentally operate on them, or on symbolic representations of them. Then the execu- tive functions associated with explicit cogni- tion use this information to produce thought and behavior that is more complex than could have resulted from either associa- tive/implicit or analytic/explicit processes alone. A contributing factor to the emer- gence of the ability to shift between these modes of thought may have been the expan- sion of the prefrontal cortex, and the asso- ciated executive functions and enhanced working memory2 capacity that came with the expansion. Enhanced working memory allowed humans more control over their focus of attention so as to maintain task goals in the presence of interference. Indeed, indi- vidual differences in working memory are strongly related to fluid intelligence among modern humans (Conway, Jarrold, Kane, Miyake, & Towse, 2007; Engle, Tuholski, Laughlin, & Conway, 1999; Kane, Hambrick, & Conway, 2005; S.B. Kaufman, DeYoung, Gray, Brown, & Mackintosh, 2009). 4.2 The Multi-Layered Mind and a Return to the Lag between Anatomical and Behavioral Modernity Several researchers emphasize that the mod- ern human mind consists of various “kinds of 2 Working memory is the ability to maintain, update, and manipulate information in an active state.  P1: OJL Trim: 7in × 10in Top: 0.498in Gutter: 0.871in CUUS1027-15 cuus1027/Kaufman ISBN: 978 0 521 51366 1 May 10, 2010 20:17 EVOLUTIONARY APPROACHES TO CREATIVITY 287 minds” layered on top of one another (A.S. Reber, 1989, 1993; A.S. Reber & Allen, 2000; Dennett, 1995, 1996). According to these accounts, these multiple minds are continu- ously operative, giving rise to many internal and external conflicts amongst members of our species, as well as contributing to our most distinctly human intellectual and cre- ative accomplishments. Arthur Reber pro- poses that implicit cognition is evolution- arily older than explicit cognition. It may stem from the oldest parts of the brain and in its crudest form be involved in the execu- tion of behavior patterns that are prewired, reflexive, and tied to survival-related goals, while our later-evolving explicit capacities for reflection and deliberate reasoning may allow us to override strictly survival-related goals (Stanovich, 2005). Alternatively, it may be that more brain tissue simultaneously allowed for not just the onset of explicit reasoning but also a qualitatively different kind of implicit pro- cessing, one in which the detection of gra- dations of similarity paves the way for cog- nitive flexibility. Let us return briefly to the question of why the burst of creativ- ity in the Upper Paleolithic occurred well after the second rapid increase in brain size approximately 500,000 years ago. A larger brain provided more room for episodes to be encoded, and particularly more association cortex for connections between episodes to be made, but it is not necessarily the case that this increased brain mass could straightaway be optimally navigated. There is no reason to expect that information from different domains (whether strongly mod- ular or weakly modular) would immedi- ately be compatible enough to coexist in a stream of thought, as in the production of a metaphor. It is reasonable that it took time for the anatomically modern brain to fine-tune how its components “talk” to each other such that different items could be blended together or recursively revised and recoded in a coordinated manner (Gabora, 2003). Only then could the full potential of the large brain be realized. Thus the bot- tleneck may not have been sufficient brain size but sufficient sophistication in the use of the memory already available, through con- textual focus, or shifting between implicit and explicit thought. It is worth noting that other periods of revolutionary innovation, such as the Holocene transition to agricul- ture and the modern Industrial Revolution, occurred long after the biological changes that made them cognitively possible. 4.3 “Recent” Creative Breakthroughs Of course the story of how human creativ- ity evolved does not end with the arrival of anatomical and behavioral modernity. The end of the ice age approximately 10,000 to 12,000 years ago witnessed the begin- nings of agriculture and the invention of the wheel. Written languages developed around 5,000 to 6,000 years ago, and approximately 4,000 years ago astronomy and mathematics appear on the scene. We see the expression of philosophical ideas around 2,500 years ago, invention of the printing press 1,000 years ago, and the modern scientific method about 500 years ago. And the past 100 years have yielded a technological explosion that has completely altered the daily routines of humans (as well as other species), the con- sequences of which remain to be seen. 5. Creativity and Cultural Evolution We have examined how the capacity for cre- ativity evolved over millions of years. In this section we explore the possibility that cre- ative ideas themselves evolve through cul- ture, in the sense that they exhibit “descent with modification,” or incremental adapta- tion to the constraints of their environment. (A related idea is that the creative process not at the cultural level but within the mind of one individual is Darwinian; this is dis- cussed in Chapter 9, this volume.) 5.1 Creative Cultural Change as a Darwinian Process It has been proposed that the process by which creative ideas change over time as they pass from person to person can  P1: OJL Trim: 7in × 10in Top: 0.498in Gutter: 0.871in CUUS1027-15 cuus1027/Kaufman ISBN: 978 0 521 51366 1 May 10, 2010 20:17 288 LIANE GABORA AND SCOTT BARRY KAUFMAN be described in Darwinian terms (Aunger, 2000; Blackmore, 1999; Boyd & Rich- erson, 1985; Cavalli-Sforza & Feldman, 1981; Dawkins, 1975; Durham, 1991). This approach is sometimes referred to as “dual inheritance theory,” the idea being that we inherit cultural as well as biological infor- mation, and the units of cultural informa- tion are sometimes referred to as “memes.” The rationale is clear; since natural selec- tion is useful for explaining the astonishing creativity of nature, perhaps it is also use- ful for explaining the astonishing creativity of human culture. There are many paral- lels between the two. Clearly, new inven- tions build on existing ones, but it isn’t just the cumulative nature of human creativity that is reminiscent of biological evolution. Cumulative change is after all rather easy to come by; in the days of taping music, each time a tape was copied it became cumu- latively more scratched. The creativity of human cultures is reminiscent of biological evolution because of the adaptive and open- ended manner in which change accumulates. New inventions don’t just build on old ones, they do so in ways that meet our needs and appeal to our tastes, and as in biological evo- lution there is no limit to how any particular invention or creative work may inspire or influence other creative works. Moreover, culture generates phenomena observed in biological evolution, such as drift3 and niches4 (Bentley et al., 2004; Gabora, 1995, 1997). A theory that encompasses the two would put us on the road to uniting the social sciences with the biological sciences. 3 Drift refers to changes in the relative frequencies of variants through random sampling from a finite population. It is the reason why variation is reduced in reproductively isolated populations such as those living on a small island. Drift has been shown to occur in a culture context with respect to such things as baby names and dog breed preferences (Neiman, 1995; Madsen et al., 1999; Bentley et al., 2004). In a computer model of cultural evolution, the smaller the society of artificial agents, the lower the cultural diversity (Gabora, 1995). 4 Just as the biological evolution of rabbits created niches for species that eat them and parasitize their guts, the cultural evolution of cars created niches for seat belts and gas stations (Gabora, 1997, 1998). In order to see how Darwinian theory might be applied to the evolution of creativ- ity ideas in culture, let us examine what kind of process natural selection can describe, and how it works. The paradox faced by Dar- win and his contemporaries was the follow- ing: How does biological change accumu- late when traits acquired over an organism’s lifetime are obliterated? For example, a rat whose tail is cut off does not give birth to rats with cut-off tails; the rat lineage loses this trait. Note that this kind of continual “backtracking” to an earlier state (e.g., in the above example, the state of having a full tail) is unique to biology; if, for example, an asteroid crashes into a planet, the planet cannot revert to the state of having not had the asteroid crash into it.5 Darwin’s genius was to explain how living things adapt over time despite the fact that new modifications keep getting discarded, by looking from the level of the individual to the level of the population of interbreed- ing individuals. He realized that individuals who are better equipped to survive in their given environment tend to leave more off- spring (be “selected”). Thus, although their acquired traits are discarded, their inher- ited traits (loosely speaking, the traits they were born with) are more abundant in the next generation. Over generations this can lead to substantial change in the distribu- tion of traits across the population as a whole. Natural selection was not put forth to explain how biological novelty origi- nates. It assumes random variation of her- itable traits, and provides an explanation for population-level change in the distribution of variants. We now ask: Can natural selection simi- larly explain the process by which creative ideas evolve through culture? A first thing that can be noted is that the problem for 5 Although Darwin observed that this was the case, he did not know why. We now know that the rea- son acquired traits are not inherited in biology is that organisms replicate using a template – a self- assembly code that is both actively transcribed to produce a new individual, and passively copied to ensure that the new individual can itself reproduce.  P1: OJL Trim: 7in × 10in Top: 0.498in Gutter: 0.871in CUUS1027-15 cuus1027/Kaufman ISBN: 978 0 521 51366 1 May 10, 2010 20:17 EVOLUTIONARY APPROACHES TO CREATIVITY 289 which natural selection was put forward as a solution does not exist with respect to cul- ture (Gabora, 2008). That is, there is no sense in which the components of creative ideas cyclically accumulate and then get discarded at the interface between one generation and the next. For example, unlike the chopped off tail which does not get transmitted to offspring, once someone invented the spout on a teapot, teapots could forever after have spouts. One might ask if Darwin’s solution is nevertheless applicable; might processes outside of biology evolve through selection even if selection was originally advanced as a solution to a paradox that is unique to biology? The problem is that since acquired change can accumulate orders of magni- tude faster than inherited change, if it is not getting regularly discarded, it quickly over- whelms the population-level mechanism of change identified by Darwin. This is par- ticularly the case with respect to creative ideas since they do not originate through random processes – or even processes prone to canceling one another out – but through strategic or implicit, intuitive processes, making use of the associative structure of memory. Darwinian approaches to culture posit that the basic units of this second Darwinian process are discrete elements of culture that pass from one person to another intact except for random change akin to muta- tion that arises through copying error or biased transmission (preferential copying of high-status individuals). Copying error and biased transmission are sources of change that take place at the time an idea spreads from one individual to another, which creativity researchers tend to view as a relatively minor source of creative change compared with cognitive processes such as imagining, planning, analogizing, con- cept combination, and so forth. The reason that Darwinian theories of culture focus on sources of change that occur when an idea spreads from one individual to another is not accidental; it stems from the fact that nat- ural selection is of explanatory value only to the extent that there is negligible trans- mission of acquired characteristics. This is the case in biology, as we saw with the cut- off tail example; change acquired during an individual’s lifetime is not generally passed on to its offspring. As another example, you didn’t inherit your mother’s tattoo – some- thing she acquired between the time she was born and the time she transmitted genetic material to the next generation. However, few scholars accept that there is negligible transmission of acquired char- acteristics in culture. The cultural equiva- lent of the individual is the creative idea. A new “generation” begins when this idea is transmitted from person A to person B, and lasts until the idea is transmitted from person B to person C. Any changes to an idea between the time B learned it and the time B expressed it are “acquired character- istics.” If B mulls the idea over or puts it into her own terms or adapts it to her own frame- work, the process by which this idea changes cannot be explained by natural selection, because as mentioned earlier, this kind of ‘intragenerational’ change quickly drowns out the slower intergenerational mechanism of change identified by Darwin; it “swamps the phylogenetic signal.” The Darwinian perspective on culture therefore leads to a view of the human condition as “meme hosts,” passive imitators and transmitters of prepackaged units of culture, which evolve as separate lineages. To the extent that these lineages “contaminate one another” – that is, to the extent that we actively and creatively transform elements of culture in ways that reflect our own internal models of the world, altering or combining them to suit our needs, perspectives, or aesthetic sen- sibilities – natural selection cannot explain cultural change. It has been argued that due to this “lack-of-inheritance-of-acquired- characteristics” problem, not just the evo- lution of creative ideas (Gabora, 2005) and the evolution of culture (Gabora, 2004, 2008), but the evolution of early life itself (Gabora, 2006; Vetsigian et al., 2006), and even of many features of modern life (e.g. Jablonka & Lamb, 2005; Kauffman, 1993; Newman & M¨uller, 1999; Schwartz, 1999),  P1: OJL Trim: 7in × 10in Top: 0.498in Gutter: 0.871in CUUS1027-15 cuus1027/Kaufman ISBN: 978 0 521 51366 1 May 10, 2010 20:17 290 LIANE GABORA AND SCOTT BARRY KAUFMAN cannot be described by Darwin’s theory of natural selection. 5.2 A Non-Darwinian Theory of How Creative Ideas Evolve If creative ideas do not evolve through selec- tion, how do they evolve? One possibil- ity is that the evolution of creative ideas through culture is more akin to the evo- lution of the earliest biological life forms than to present-day DNA-based life (Gab- ora, 2000, 2004, 2008). Recent work sug- gests that early life emerged and replicated through a self-organized process referred to as autocatalysis, in which a set of molecules catalyze (speed up) the reactions that gen- erate other molecules in the set, until as a whole they self-replicate (Kauffman, 1993). Such a structure is said to be autopoietic, or self-regenerating, because the whole is reconstituted through the interactions of the parts (Maturana & Varela, 1980). These ear- liest precursors of life evolved not through natural selection at the level of the pop- ulation, like present-day life, but commu- nal exchange of innovation at the individ- ual level (Gabora, 2006; Vetsigian, Woese, & Goldenfeld, 2006). Since replication of these pre-DNA life forms occurred through regeneration of catalytic molecules rather than (as with present-day life) by using a genetic self-assembly code, acquired traits were inherited. In other words, their evolu- tion was, like that of culture, Lamarckian. This has led to the suggestion that world- views evolve through culture, through the same non-Darwinian process as the earli- est forms of life evolved, and creative prod- ucts such as tools and dances and architec- tural plans are external manifestations of this process; they reflect the states of the par- ticular worldviews that generate them. The idea is that like these early life forms, world- views evolve not through natural selection, but through self-organization and commu- nal exchange of innovations. One does not accumulate elements of culture transmitted from others like items on a grocery list, but hones them into a unique tapestry of under- standing, a worldview, which like these early life forms is autopoietic in that the whole emerges through interactions amongst the parts. It is self-mending in the sense that, just as injury to the body spontaneously evokes physiological changes that bring about heal- ing, events that are problematic or surprising or evoke cognitive dissonance spontaneously evokes streams of thought that attempt to solve the problem or reconcile the disso- nance (Gabora, 2008). Thus, according to this view it is not chance, mutation-like processes that propel creativity, but the self-organizing, self-mending nature of a worldview. 6. Why Did Creativity Evolve? We have discussed how human creativity evolved, and in what sense creative ideas can be said to evolve. We now address a funda- mental question: Why did human creativity evolve? 6.1 Creativity as Evolutionary Spandrels Some forms of creativity enhance survival and thus reproductive fitness. For exam- ple, the invention of weapons most likely evolved as a creative response to a need for protection from enemies and predators. For other forms of creative expression, however, such as art and music, the link to survival and reproduction is not so clear-cut. Why do we bother? One possibility is that art, music, humor, fiction, religion, and philosophy are not real adaptations, but evolutionary span- drels: side-effects of abilities that evolved for other purposes (Pinker, 1997; see also Carroll, 1995; Gabora, 2003; J.C. Kaufman, Lee, Baer, & Lee, 2007; McBrearty & Brooks, 2000). Pinker likens these forms of creativity to cheesecake and pornography – cultural inventions that stimulate our senses in novel ways but do not improve our biological fitness. The “spandrels” explanation assumes that what drives creativity is biological selection forces operating at the individual level, and there is some empirical support for this....