The Analogical Peacock Hypothesis: The Sexual Selection of Mind-Reading and Relational Cognition in Human Communication

Sexual Selection

Darwin (1859, 1871) was troubled by certain features and behaviors of animals, usually male, that seemed certain to hinder their survival, a stag's antlers, a lion's mane, or a peacock's tail. He suggested that differences between males and females of a species that sought to survive in the same environments were most likely due to sexual selection. The strange ornaments, armaments, and behaviors give their holder a competitive advantage allowing them to leave more offspring than those with lesser traits. This could occur as contests between members of one sex—intrasexual selection, or through selecting a mate with the best genes—intersexual selection.

Parental investment theory (Trivers, 1972) explains why females usually choose and males are normally the chosen. A fundamental disparity in the resources parents invest in offspring care means the sex investing more, normally females, maximizes reproductive success by being choosier about the genetic qualities of potential mates. Conversely the sex investing less, normally males, maximizes reproductive success by increasing mating opportunities and being less choosy about mate quality. This influences both intrasexual and intersexual selection strategies.

Intrasexual selection occurs between rivals, usually males seeking access to females, through direct competition or holding a resource that maximizes access to the opposite sex. Violent conflict is the ultimate resolution but it is usually avoided through signaling, often using extravagant armaments and behaviors that enable animals to preempt the likely outcome.

In the intersexual case, there are strong pressures to choose the highest quality mate and correctly discern quality. Consequently, an animal being chosen seeks to manipulate a choosing animal by signaling a high level of quality through fitness indicators. The costly signaling debate plays out most fervently and contentiously in this signaling with fitness indicators scenario. Fitness indicators have been suggested in many forms, the carotenoid pigmentation of many fish and birds (Olson & Owens, 1998; Peters, Denk, Delhey, & Kempenaers, 2004), the bowers built by bowerbirds (Madden, 2001), the song of the nightingale (Kipper, Mundry, Sommer, Hultsch, & Todt, 2006), the ornaments on a peacock's train (Loyau, Jalme, & Sorci, 2005), and human culture and communication (Miller, 2001a).

In sexual selection situations, handicaps can act as fitness indicating signals correlated with a signaler's genetic quality. Costly signals display the ability to support burdens despite high levels of predation (Møller, Christiansen, & Mousseau, 2011), they can indicate low rates of mutation (Iwasa, Pomiankowski, & Nee, 1991; Pomiankowski, Iwasa, & Nee, 1991), or signal developmental stability indicating low levels of parasitic perturbation (Hamilton & Zuk, 1982). Large handicaps require resources to grow and support them; they indicate the resources available to a signaler after the satisfaction of other survival related constraints. Greater burdens or levels of precision make it more difficult for lower quality cheaters to fake the signal.

Sexual Selection in Human Evolution

Sexual dimorphism is often a sign of sexual selection and human physical sexual dimorphism suggests that sexual selection may have played a role in human evolution (for recent reviews see Gangestad & Scheyd, 2005; Roberts & Little, 2008, for the intersexual case; and Puts, 2010, for the intrasexual case). Mental or cognitive sexual dimorphism in humans is more controversial but has been suggested in cognition (Baron-Cohen, 2002) and communication (Locke, 2011), although this is contested and is certainly not obviously apparent (Fine, 2010; Hyde, 2005).

Sexual strategies theory (Buss & Schmitt, 1993) and strategic pluralism theory (Gangestad & Simpson, 2000) give accounts of human mating strategies based on parental investment theory (Trivers, 1972). In brief, males seek to maximize partners in the short term but seek genetically valuable partners in the long term; females seek high quality genetic partners (using good-gene indicators) in the short-term but can offset some genetic quality against parenting skills and resource provisioning in the long term as it increases the likelihood of offspring survival (using good-parent indicators). These strategies are malleable depending on the prevailing environmental conditions, for example: in conditions favoring biparental care females should seek long-term strategy males and monogamy will be prevalent; where there is high environmental pathogen prevalence—requiring a high quality immune system—females should seek males with good-gene fitness indicators.

These theories suggest that females (and males) will assess mate quality using good-gene and good-parent indicators to differing extents. The next section reviews the mental fitness indicator literature and experimental evidence. Increased use of a mental attribute in an experiment that primes mating provides evidence that it is a mental fitness indicator. Furthermore, preference for a mental attribute in short-term mating situations suggests it is a good-gene fitness indicator; even more so for females in times of near peak fertility. Preference in long-term situations would suggest it is a good-parent fitness indicator. These distinctions are not sharp, both sexes will seek to maximize both good-genes and good-parenting in a partner.

Intelligence

Perhaps the most studied mental fitness indicator is intelligence, it is highly heritable and polygenic (Davies et al., 2011; Plomin & Spinath, 2004), and ranks highly on lists of desired mate characteristics for both sexes (Buss et al., 1990; Lippa, 2007). Intelligence also correlates with proposed physical fitness indicators and indicators of developmental stability such as symmetrical features (Banks, Batchelor, & McDaniel, 2010).

Intelligence or more specifically the psychometric construct of general intelligence g is a latent factor that consistently correlates with the results of many cognitive tasks (Jensen, 1998). A long-standing case has been made that a two-factor model is more suitable than the unitary g (Cattell, 1963; Cattell & Horn, 1978; Nisbett et al., 2012); these factors are fluid intelligence (Gf) and crystallized intelligence (Gc)—although most recent literature equates g with Gf and less attention is paid to Gc. Fluid intelligence involves abstract reasoning and captures the ability to perceive relationships independent of acquired knowledge. Crystallized intelligence contains the verbal and cultural knowledge that has been learned through exposure to education or culture. Tests that load highly onto fluid intelligence are Raven's Advanced Progressive Matrices (RAPM), a test of nonverbal relational reasoning, and Cattell's Culture Fair Intelligence Test, a nonverbal test that requires participants to perceive relationships between shapes and figures. Crystallized intelligence is typically measured using tests of vocabulary and general facts. These two components align with two of the unifying features of mental fitness indicators, mechanisms for the relational combination of concepts (Gf) and large conceptual knowledge networks (Gc).

A recent research effort has proposed a considerable overlap between working memory (Baddeley & Hitch, 1974) and fluid intelligence (Waltz, Knowlton, & Holyoak, 1999; Wilhelm & Oberauer, 2006; although see Ackerman, Beier, & Boyle, 2005). Halford, Cowan, and Andrews (2007) argue that the functionality is similar; fluid intelligence ability depends on the number of elements with active interrelationships between them that can be sustained in working memory. They use the relational complexity metric to determine the capacity limitations on both working memory (typically three to four chunks) and reasoning in fluid intelligence tasks (relations between approximately four variables). Halford, Wilson, and Phillips (2010) argue that relational knowledge is the basis of most of higher cognition. Relational reasoning is a key mechanism of the APH; intelligence functions as a mental fitness indicator primarily through the ability to rapidly find and manipulate relations between concepts. The extent of the conceptual knowledge networks is measured by crystallized intelligence and how fast and efficiently relations can be found between them is measured by fluid intelligence. Working memory provides the cognitive machinery for relational binding in these displays.

What are the costs associated with intelligence? Fluid intelligence is likely to incur developmental costs—growing a mechanism that can sustain high levels of relational reasoning over a prolonged period is difficult and intelligence seems to be susceptible to developmental instability. Production costs can arise from sustained mental effort in fluid intelligence tasks (Fairclough & Houston, 2004; Gailliot, 2008; Marcora, Staiano, & Manning, 2009). There is also the cost of brain volume which McDaniel (2005), in a meta-analysis using 37 samples across 1,530 people, estimates has a population correlation of 0.33 with intelligence. Crystallized intelligence incurs developmental costs, particularly opportunity costs associated with the gathering together and maintaining a large vocabulary, a large body of cultural knowledge, knowing how the concepts relate to each other and when they are relevant and contextually appropriate. Time lost to cultural pursuits could be usefully spent gathering food, resources, and improving physical prowess.

The evidence for a relationship between intelligence and attractiveness is equivocal and inconsistent. Intelligence can function as either a good-parent indicator by increasing provisioning abilities, or a good-genes indicator offering greater inherited intelligence in offspring. If the latter, it should be highly attractive in short-term and peak fertility situations. Haselton and Miller (2006) found that heterosexual female students desired creative intelligence in short-term partners at times of peak fertility but the methodology did not disentangle creativity from intelligence. Prokosch, Coss, Scheib, and Blozis (2009) found intelligence (WAIS-III Vocabulary subtest) indicated mate appeal for undergraduate females in both long and short-term situations but not peak fertility situations. A similar sample rated males who were perceived to be intelligent as more attractive long-term than short-term mates (Gangestad, Garver-Apgar, Simpson, & Cousins, 2007). Gangestad, Thornhill, and Garver-Apgar (2010), using 66 romantically involved heterosexual couples, directly examined the link between attractiveness and intelligence in both extrapair males and long-term partners. They found no evidence of sexual interest due to intelligence across the ovulatory cycle.

The balance of evidence is mixed but perhaps leans toward intelligence functioning as a good-parent fitness indicator rather than a good-gene fitness indicator. Another possibility is that intelligence was once a fitness indicator but now functions in a threshold role. Signals may become uninformative as discriminators of quality but be retained as thresholds that maintain female interest, beyond the threshold other signals offer greater discriminability. Li, Bailey, Kenrick, and Linsenmeier (2002) found some evidence in support of this. Using an economic budgeting methodology a broad male and female sample, interviewed at a U.S. international airport, allocated low, medium, or high resource budgets to attributes in potential mates. Low budgets led females to spend more on intelligence and resource acquisition abilities although males favored physical attractiveness and intelligence, suggesting these were necessities. When budget restrictions eased, allowing more luxury based allocation, creativity was highly favored by both males and females.

Li et al. (2002) suggest that intelligence is required as a threshold that males must pass to be considered but beyond a certain level there are diminishing returns in its power to aid discrimination. Creativity, however, was desired as a luxury and perhaps it offers greater discriminative power. Unfortunately, they only asked about long-term partner contexts and the situation may differ in short-term or peak fertility scenarios.

One argument against intelligence as a fitness indicator is that it has qualities that make it useful from a natural selection point of view that would erode its value as a costly signaling mechanism. A possible explanation is that once intelligence was established for social intelligence it became useful as a more general problem solving exaptation serving as a compensatory mechanism to offset costs (Husak & Swallow, 2011; Oufiero & Garland, 2007). This becomes even more likely if intelligence itself is not directly selected but is mediated by other signals such as creative artistic expression, and as Oscar Wilde reminds us “all art is quite useless” (Wilde, 1891). ¹

Creativity

Creativity concerns the ability to generate new ideas, insights, and solutions and plays an important part in our most valued activities: arts, science, literature, design, and engineering. A key aspect of creativity is the synthesis of existing concepts into novel and innovative combinations; an early and influential definition described it “as the forming of associative elements into new combinations” which became the basis of the Remote Associates Test (RAT) commonly used to measure creative ability (Mednick, 1962). More recently Vartanian, Martindale, and Matthews (2009) defined creativity “as the novel and useful combination of concepts previously thought to be unrelated.” They argue that creativity has two stages, first, the relatedness of concepts must be judged, then the concepts can be combined; there is little creative value in combining obviously related concepts. Faster relatedness judgments allow the quick rejection of unpromising combinations. Vartanian et al. (2009) using an all female sample found that participants who scored higher on a divergent thinking task were faster at judging the relatedness of two concepts. Creative people have a speed advantage as they can consider a greater number of conceptual relationships in a given amount of time.

Creativity in this sense involves the same three unifying features as intelligence—relational combination of concepts, a large network of concepts to search and assess, and processing speed. However, artistic creativity is perhaps the easiest indicator in which to examine the unifying feature receiver knowledge, this will be addressed in the remainder of this section.

Following Darwin (1871) who viewed visual aesthetics as a product of sexual selection through mate choice, Miller (2000, 2001b, 2001a) argues that the arts—verbal, visual, and musical—provide examples of human sexually selected signaling and that creativity functions as a mental fitness indicator. Similarly, Dutton (2009) has claimed that sexual selection forms the basis of the human “art instinct.” The art producer signals their fitness qualities to a receiver by producing an artifact that is aesthetically pleasing and difficult to produce. It is here that Miller's (2009) conspicuous precision argument is most useful. There must be costs associated with the production of art, specifically a cost differential that makes art hard-to-fake for lower quality rivals. The production of art entails costs in energy, endurance, and opportunity lost through the time taken to learn the skills, build knowledge of what appeals to an audience, and produce the art. Precision costs are displayed through levels of hand-eye coordination, diligence, and attention to detail. Dutton (2009) also notes a general preference for art produced with high levels of virtuosity—modern abstract art provides an interesting exception.

A similar costly signaling process based on a female “aesthetic taste” can be seen in the bowerbirds of Australia and New Guinea (Miller, 2001b). Male bowerbirds construct elaborate and intricate bowers designed to appeal to an aesthetic determined by the choice of the female (e.g., Madden, 2001, 2002), the Vogelkop bowerbird (Amblyornis inornatus) provides a particularly good example (Uy & Borgia, 2000). These constructions can take very extravagant forms, arranged with flowers, beetle wings, and berries among other things. Females search for the best bowers and choose the one most appealing to her aesthetic instinct and mate with the constructor. The bowers are not used for nesting; after mating females build small functional nests and expect no male parental investment. Successful males mate with many females while those unable to appeal to female aesthetic instincts do not mate. The incurred developmental, maintenance, and production costs to males are substantial. Principally there are the costs in a stable energy and nutrition supply throughout the growth period required to develop the neural mechanisms necessary for such a complex behavioral display. These mechanisms must not only encode the behavioral display but also the “aesthetic taste” of the female choosers—at some level, probably genetic, they must encode receiver knowledge.

Human art too also represents a combination of virtuosity—displays of precision costs, and receiver knowledge—displays of aesthetic taste. Costs are incurred through lost opportunities as time investment is needed to learn an audience's aesthetic sensibilities, expectations, fashions, and trends. A highly skilled artist could pay a lot of attention to creating an intricate and detailed art piece representing a farm tractor and although the virtuosity and detail may be exquisite the lack of audience knowledge will result in few plaudits and little elevation of social status—except perhaps in the cultures of some communist regimes. Receiver knowledge is an important component of high quality display and usually culture specific—this accounts for some of the difference in the appeal of modern abstract art. A culture must be learned to fully appreciate an aesthetic, whereas high levels of virtuosity are evident to cultured and uncultured alike.

Music and especially song lyrics in popular culture often reflect sexual and evolutionarily relevant themes which may serve as creative displays or perhaps prime more performative displays (Hobbs & Gallup, 2011; Saad, 2012). All forms of musical performance offer opportunities for the display of virtuosity and fine motor control that demonstrate enormous investment in time and lost opportunity (Miller, 2000). Here too receiver knowledge is important, developing the cultural knowledge of a large repertoire of performance pieces is important in displaying ability; but the highest levels of virtuosity come with fine-grained knowledge of an audience's expectations. Knowing where to slightly delay or alter a piece of music to deviate from the standard performance in a satisfying manner demonstrates a keen mind-reading ability and alignment with an audiences aesthetic sensibilities (Huron, 2006). Knowing which elements of a piece of music have a strong emotional resonance exhibits a deep receiver knowledge and empathic alignment (Scherer, 2004, 2005; Schubert, 2010).

The evidence that humans view creativity as a fitness indicator is less equivocal than the evidence for intelligence. Nettle and Clegg (2006) found a positive correlation between creative activity and mating success in a sample that partly targeted male and female artists and poets—serious and professional art producers have more sexual partners than hobby and nonproducing people. As mentioned, Haselton and Miller (2006) found creative intelligence was more desirable than wealth in vignettes describing both short-term and long-term partners, but correlations between female fertility and preference for creativity occurred only for short-term mates. The phrase creative intelligence conflates creativity and intelligence but the focus of the most popular vignette for short-term mating was a financially poor artist suggesting a more classically creative interpretation.

Prokosch et al. (2009) also examined creativity as separate from intelligence and found that creativity had a stronger effect than verbal intelligence, independently predicting mate appeal in both long- and short-term relationships, but no effect was found due to level of fertility.

Salient mating opportunities can motivate creativity. Griskevicius, Cialdini, and Kenrick (2006) primed psychology undergraduates with photographs of desirable potential mates leading males to write more creative stories for both short- and long-term mates, although females showed no increase. However, when priming emphasized a “committed” long-term partner, creativity (RAT) increased in both females and males.

Sexual selection of creativity has been proposed as an explanation of the persistence and heritability of schizophrenia. Nettle and Clegg (2006), in their artist and poet sample, found two components of schizophrenia—unusual experiences and impulsive nonconformity—positively correlated with mating success and unusual experiences was mediated by creative activity.

This section introduced receiver knowledge in creativity and again highlighted the relational combination of concepts, large conceptual knowledge networks, and processing speed. Contextualization is mostly involved in creativity that involves components of improvisation, as occurs often in creative linguistic communication.

Language

Dunbar, Marriott, and Duncan (1997) listened to 45 human conversations in three samples. People mostly exchanged social information, however, there were sex differences. Males talked more about intellectual and work topics and even more so in the presence of females; males were also concerned with self-promotion and display “in what has all the characteristics of a mating lek.” Female conversations were oriented toward the service and maintenance of relationships within their social networks.

These differences suggest there may be a role for sexual selection in language, and this has been proposed in some detail by Darwin (1871), Miller (2001a), and Locke (2011) among others, with different emphasis placed on intrasexual and intersexual selection. Locke (2008, 2011) emphasizes male intrasexual competition and distinct linguistic styles in the two sexes in a similar vein to the differences noted by Dunbar et al. (1997). Locke (2011) views much of male language use as verbal “dueling” in which males compete with rivals to increase their social status or to impress females. The focus of within-sex male conversation involves boasting, teasing, ritualized insults, jokes designed to elevate their social status, or in mixed sex situations to impress females with their eloquence and storytelling prowess. Many of these contests become ritualized as in the “flyting” of medieval Nordic and British cultures, or the “playing the dozens” and “sounding” of U.S. inner city Black adolescents and young adults—precursors to the modern “freestyle rap battle” (Liu et al., 2012). The prize for better displays of “verbal plumage” is increased status and consequent access to more females.

In Locke's (2011) view, within-sex female conversation places little emphasis on personal display and greater emphasis on forming and maintaining affiliative social bonds. This typically involves intimacy, empathy, and conveys self and other disclosing information. These female “duets” cement social relationships, and Locke speculates that the pressures of human birth and altricial neonates led to close female bonding; verbal grooming ensures help is available when it is required. Locke states that no specialized intersexual linguistic style exists resulting in communicative difficulties between the sexes. He pays no special attention to courtship language, and provides no explanation for why the divergence stops at the level it does and diverges no further.

Male social status is a key factor in the evolution of language for Burling (1986, 2005) who argues that a simple pidgin would suffice if the function of language is to exchange propositional information concerning hunting or subsistence. There is no need for the baroque extravagances of linguistic style found in most languages. Somehow complexity in language is being selected and he implicates the relationship between social group leadership and reproduction. In Burling's (1986) view, better linguistic skills result in elevated status and leadership positions, and leaders have more surviving offspring; this alone ensures the propagation of genes favoring complex linguistic ability. However, this view says little about the form that language takes.

Social status is also a driver for Dessalles (1998) who invokes linguistic relevance as a tool for gaining status. Drawing on Relevance Theory (Sperber & Wilson, 1986), Dessalles notes that a sender can create a relevant communication by providing a receiver with contextually appropriate and useful information related to a problem. This seems like an altruistic act but for Dessalles it is better viewed as a trade for social status from the receiver or social group. Providing contextually appropriate and relevant information is important to the APH, however, it is not a trade for social status, it is a display of mind-reading ability that may result in elevated status if it is well executed. Being aware of what is relevant to a receiver displays receiver knowledge and including context makes a signal hard to fake.

Relevance theory states that utterances in human communications only provide a small part of the information a receiver needs to understand the sender's meaning; a receiver has to make contextual assumptions to correctly interpret the sender's meaning. For example, the meaning of the utterance “John is a soldier” is underdetermined by itself, but context would make it readily apparent if it was meant to be understood literally or metaphorically (Sperber & Wilson, 2002). Relevance theory also highlights the “ostensive–inferential” nature of communication, any communicative act provides two levels of information, at an obvious level there is the content of the message, but there is also the information that someone is trying to communicate. Ostensive acts—pointing, direct eye-gaze, an utterance—signal this second level of an intention to inform. These ostensive acts let a receiver know that a communication is probably worth putting in the cognitive effort required to understand its meaning; in short, if someone is trying to communicate then it is probably worthwhile listening to them. The value of the communication is its cognitive effect, useful information that improves a receiver's representational knowledge. Ostensive acts aid communication by alerting the receiver to the fact that the sender thinks the information is optimally relevant, the receiver can then infer the meaning by using the path of least effort to understanding. An enormous amount of receiver knowledge is required to do this. Knowledge that the information is useful enough to the receiver to be worth the effort and that the receiver will be able to interpret it given evidence from the utterance and current context. In the APH a sender can display mind-reading abilities by crafting optimally relevant utterances making them hard to fake by incorporating as much context as possible.

Franks and Rigby (2005) examined deception in mate selection using the theoretical framework of ostensive–inferential communications in relevance theory. They suggest a further level of information exists in sexual selection situations that informs the receiver that the communication is a display. This signals the receiver to put more cognitive effort into a more difficult interpretation to assess the sender's ability and ensure there is no deception. A hearer is often not aware of the information at higher levels than the propositional content of the message. They present experimental evidence that in mating primed situations a sender expects that a receiver will expend the extra cognitive effort; we will return to these experiments later.

Whereas Burling (1986, 2005), Dessalles (1998), and Locke (2011) emphasize intrasexual components and the importance of status in human communication, Miller's (2001a) account emphasizes language use in human courtship. Rivals and social status can be barriers that determine access to females but in most human courtship scenarios females ultimately decide who is successful. Miller notes that “much of human courtship is verbal courtship” and mate choice is based around a long and intricate interview process. In these courtship interchanges life stories: are exchanged; grandiloquent low frequency vocabulary is engaged; literary and poetic constraints are added and; sadly for the romantics among us, the costs of linguistic extravagance means that male language use becomes limited and more functional once a relationship is initiated. This period of direct face to face verbal flirtation is often accompanied by public displays of linguistic abilities especially in mixed company that can be viewed as indirect or covert courtship. Miller also proposes a female to male linguistic courtship strategy; this uses courtship language to avoid a partner leaving in the later stages of a relationship as commitment-phobic males become choosier. This longer-term strategy keeps a relationship conversationally interesting throughout its duration.

The experimental evidence for language use in mate primed situations is more limited. Rosen and López (2009) examined courtship language played to 30 heterosexual females across the ovulatory cycle using a dichotic listening task. Distracting courtship language played in one ear caused more task errors during fertile than nonfertile phases. S. Miller and Maner (2011) found that undergraduate males smelling a t-shirt recently worn by a female near peak fertility generated more sexual words in a word stem completion task than males in equivalent low fertility or unworn conditions. These studies provide evidence that mating primes influence low-level linguistic mechanisms in both males and females.

Miller (2001a) invoked a similar vocabulary size argument to Burling (1986), we have many more words than we need for simple information exchange—4,000 words account for 98% of conversation yet the average adult has a 60,000 word vocabulary (see also Nation, 2006). Rosenberg and Tunney (2008) examined this in a psychology undergraduate sample. They found that mating primed males used more low frequency words than females or nonprimed males. This suggests that a large network of concepts may be playing a direct selection role rather than just a supportive role in providing disparate concepts to link.

In APH terms, there are developmental costs in growing the neural apparatus required to engage in linguistic interaction, and opportunity costs of acquiring vocabulary and receiver knowledge. Three unifying features appear in this section; large conceptual knowledge networks are important and contextualization and its interaction with receiver knowledge can be displayed by using relevant communications. Although important for language, little mention of conceptual combination and processing speed has been made in this section. However, humor creates linguistic situations in which the fast relational combination of concepts becomes crucial.

Humor

Humor is a human universal—meaning it is likely to have arisen early in human evolution through some form of selection pressure. Laughter, humor's closely associated social signal, is one of the first vocalizations made by humans infants and at about 4–6 months of age they start to participate in humorous interactions (Kaufman, Kozbelt, Bromley, & Miller, 2007). The speed with which humor is processed and the fast onset of laughter also suggest that it involves a low level but sophisticated cognitive mechanism (Juckel et al., 2011; Marinkovic et al., 2010). Theories of humor have suggested superiority, surprise, social bonding, and resolving incongruities are at its core. It is a common mate preference (Buss & Barnes, 1986; Lippa, 2007; Provine, 2000) and a likely mental fitness indicator in both intrasexual and intersexual selection cases.

In the intrasexual case, Locke (2011) stresses the importance of humor and humorous insults in the verbal duels between males. Greengross and Miller (2008) found 32 male college students self-reported more use of other-deprecating humor than a sample of 64 females. They noted that use of this type of humor declined with age (range was 18–30) but a larger sample is needed to clarify this effect.

Progovac and Locke (2009) point out that success in “duels” between males is a result of good humor, fluency, and timing. They highlight a conceptual combination phenomena with an important processing speed component. Males often deliver ritualized humorous insults to other males and gain greater expressive power using two word compounds—specifically exocentric verb–noun compounds—rather than single word insults. Progovac and Locke (2009) suggest that this humorous name-calling allowed males to vie for status and dominance by demonstrating verbal skills and quick wittedness without the need to resort to outright conflict—a classic intrasexual selection competition. The insults often involved vulgar belittling components made socially acceptable by humor and ritual, some slightly tamer examples are, kill-joy or turn-coat in English, and vrti-guz “spin-butt” (restless person, fidget) in Serbian. Following Jackendoff (1999), they claim that many of these kinds of word compounds exist as syntactic fossils in current language. They invoke Chomsky's (1995) Minimalist Program calling this a “proto-Merge” phenomenon, and speculate that such contests may provide a route to a proto-syntax.

In the case of intersexual selection of humor, the issue of mutual mate choice comes into sharpest focus: Does humor appeal equally to males and females? If mate choice is balanced between sexes we would expect similar patterns of display and assessment in females and males. This question of mutual mate choice has been addressed most fully with respect to humor, although it is not yet fully resolved. A meta-analysis that included 12 samples with a measure of humor concluded that there were no notable differences in preference between the sexes (Feingold, 1992). More recently, in a vignette study McGee and Shevlin (2009) found that undergraduate students of both sexes rated “a good sense of humor” as more attractive in a long-term partner than “an average” or “no sense of humor.”

Other studies found important differences between the sexes. Bressler and Balshine (2006) had 210 male and female undergraduate psychology students rate “autobiographical statements” purported to be from undergraduate students and found that only females evaluating males found humor desirable. However, they were unsure about how the participants interpreted the phrase “good sense of humor.” Grammer and Eibl-Eibesfeldt (1990) found differences in laughter when it occurred in mixed sex and same-sex groups of student aged participants, both males and females laugh more in same-sex groups and less in mixed sex groups, but in mixed sex groups females laughed more often and even more so in the presence of a male independently rated as attractive. They found that the amount a female laughs in a conversation is predictive of both female and male interest in dating. Bressler, Martin, and Balshine (2006) directly tested the hypothesis that there are sex differences in preference for humor production and receptivity. Male and female undergraduate students both valued humor and receptivity to their own humor in a partner, but females placed greater value in humor than males and only females valued humor production. When participants were forced to choose between production and receptivity for different types of relationship females had clear preferences for humor production and males for humor receptivity in long-term relationships, but short-term results were more equivocal. Despite this issue there seems to be evidence that males and females value humor in different ways.

Kaufman, Kozbelt, Bromley, and Miller (2007) make the important point that we should distinguish between humor production, comprehension, and appreciation. Humor production may be more important in mating effort whereas humor comprehension and appreciation are mechanisms for mate choice. Correlations exist between these three components except notably production and appreciation. They suggest this is because high quality individuals are choosier and do not need to be receptive to lower quality humor. In this view, humor production and comprehension are fitness indicators, and appreciation serves as a mate choice threshold.

Humor has an interesting relationship with intelligence. Lundy, Tan, and Cunningham (1998) found that in undergraduate students self-deprecating humor enhanced desirability of a long-term relationship with physically attractive people, but it decreased perceptions of intellect. Similarly, Bressler and Balshine (2006) found humorous individuals were viewed as more socially adept but less intelligent and less honest and trustworthy. The humorous “autobiographical statements” used in the Bressler and Balshine (2006) study were, in their words, of a “sophomoric nature,” which does not necessarily imply intelligence but might imply a social prowess in aligning oneself with a peer group, and culture specific receiver knowledge. A desire to appear intelligent may be context dependent and outweighed by opportunities for status elevation within one's proximal cultural grouping.

Two recent studies directly examined the link between humor production and intelligence. Howrigan and MacDonald (2008) gave humor production tasks to 185 college-age students. Participants with higher intelligence (RAPM) produced humor rated as funnier and males scored higher than females in humor production tasks. Greengross and Miller (2011) measured humor production by ratings of captions added to cartoons using a sample of 400 university students. Similarly, they found that intelligence (vocabulary score and RAPM) predicted humor production ability, and that males had higher average humor production ability than females. They also looked at the relationship between humor production and mating success; higher scoring humor producers reported having more short-term, uncommitted sex. Finally, they created a structural equation model and found that verbal humor strongly mediated intelligence on mating success concluding that “intelligence may be sexually attractive mainly insofar as it is manifest through verbal humor.” This effect held for both males and females suggesting some male mate choice regarding humor ability.

The relationship between creativity and humor parallels that of intelligence; there is very little relationship between humor appreciation and creativity (Kaufman et al., 2007). However, Rouff (1975), in a sample of 108 psychology undergraduate students, reports a strong positive correlation between humor comprehension and creativity (RAT) that remained after intelligence had been partialed out; she notes “these results strongly suggest that comprehension of humor and creative thinking are related and that they have a common basis in the ability to link disparities.”

The studies addressed in this review of mental fitness indicators have many issues. They often depend on participant interpretation of terms such as “short-term” and “date” which may invoke evaluations based to differing degrees on physical or mental characteristics. Often fertility is categorized as high or low depending on the stage in the ovulatory cycle; Fischer et al. (2011) have recommended more fine grained sampling regimes. The participant samples mostly use undergraduate students and, although this age group is of specific interest regarding mental fitness indicators, broader cross-cultural sampling is desirable (Henrich, Heine, & Norenzayan, 2010). Studies often rely on self-report which can be vulnerable to dishonest replies, especially concerning mating history. Vignette studies or other non-natural assessment situations are often used; more natural courtship situations would enable stronger conclusions.

Methodological issues notwithstanding, there seems to be a body of evidence that the candidate mental fitness indicators are involved in human mating situations. The unifying features of relational combination of concepts, large conceptual knowledge networks, processing speed, contextualization, and receiver knowledge also have a more prominent role when courtship and mating are salient. A major issue outstanding in this research is the lack of a mechanism to explain how mental fitness indicators function. The APH proposes a mechanism based around these five unifying features.

Mental fitness indicators associated with morality and personality indicators have been left out of this review which has concentrated on the more cognitive mental fitness indicators. There is a strong likelihood that moral and personality factors are important and kindness ranks very highly on lists of mate preferences. We will return to these aspects after the more formal statement of the hypothesis. However, Miller (2007b) preempted much of the argument from morality and personality in saying they can function as indicators of high heritable phenotypic quality, “displaying a sophisticated, empathetic social intelligence requires the development of a complex Theory of Mind, which might be easily disrupted by a variety of mutations associated with autism, schizophrenia, mental retardation, social anxiety, and language impairments. Thus, a conspicuously expert level of empathy may function as a sort of neurogenetic warranty” (p. 101). The importance of these factors will be more evident after a more detailed specification of the hypothesis.

Mind-World Mappings

Mind-world mappings exist between objects in the environment and representations of these objects in the brain. These mappings are not in themselves part of the APH and are presumed to exist in any species capable of some form of mental representation of the world. They are, however, an important precursor to the APH mechanisms and provide the foundation upon which the uniquely human mechanisms of the APH exist. Conspecifics with a rudimentary representational system (Nieder & Miller, 2004; Povinelli, Reaux, & Frey, 2010) will have similar representations due to overlapping evolutionary needs. The similarity between the mental representations of conspecifics arises passively as a result of genetically predetermined mechanisms or through the action of learning mechanisms on similar environmental stimuli. There is no need for these representations to be closely aligned other than the need for them to perform the same ecological function in separate individuals. Nonetheless, we can expect a close passive alignment of mind-world mappings between conspecifics. This means that the set of mental representations in any two conspecific individuals can be thought of as passive analogous structures (in the sense that a meaningful mapping can be made between them). In other words, any network of representations will be aligned not because it needs to be aligned with other individuals but simply because it serves a similar function.

Between-Mind Mappings

The processes of the APH begin when there is adaptive value in a shift from a passive alignment to an active pursuit of representational alignment; such a shift is most likely to be motivated by social and communicative demands. Navigating social hierarchies is thought to be particularly important in many species of primates where social rank can determine an individual's level of reproductive success (Pawłowski, Lowen, & Dunbar, 1998; Rodriguez-Llanes, Verbeke, & Finlayson, 2009; Surbeck, Mundry, & Hohmann, 2010) and perspective taking and mind-reading abilities have been strongly implicated in the evolution of primate sociality (Dunbar, 1998; Tomasello, 1999). If these mind-reading abilities can be used to predict reproductive success or genetic quality, they become a target for sexual selection. Taking female mate choice sexual selection as a first stage, a female homonin observing a particularly socially astute male may surmise that he will be a good reproductive bet. If this proves to be the case, then genes for being socially astute and for favoring social astuteness will pass into the next generation.

According to the APH at some stage in the evolution of the homonin lineage, mind-reading of representational structures became a way of displaying social astuteness and part of a sexual selection process. Here the value of receiver knowledge is apparent. A female can assess male knowledge of her own representational structures as an indicator of the mind-reading abilities of a male sender. The female has privileged access to her own representational structures which can be used as metric against which to judge the mind-reading abilities of competing males. The best mind readers can display knowledge of the female's representations as evidence of their perspective taking and mind-reading abilities and consequently display their social astuteness. Quality is judged by how well a male can read the representations of the female and, importantly, how well the male can convey this mind-reading ability. Mind reading abilities alone can only indirectly influence reproductive success, once they have become a sexually selected attribute they must be actively displayed through communication. Within the APH, the first major purpose and primary motivating function of human communication is this display of mind-reading abilities.

As these communicative displays seek to show a male's receiver knowledge (of his female interlocutor), the male must model a female's representational structure. The preexisting passive analogous structure between representations in two individuals makes the male's own set of representations a parsimonious basis for a model of a receiver's representations (Gallup, 1998). This basis could be added to with model components that account for independent and different perspectives in other individuals. As displaying receiver knowledge becomes crucial to reproductive success, a strong competitive need arises to ensure the most complete and consistent between-mind mappings with potential mates. Effort and resources must be invested to actively acquire, align, and refine representational structures. The shared sets of mental representations in any two conspecific individuals change from being passive to actively maintained analogous structures. The second major purpose of human communication is this alignment of representations.

At this stage two of the unifying features of the mental fitness indicators interact—receiver knowledge and a large conceptual network. A female with a larger network of concepts has more discriminative power over a male's level of receiver knowledge, this creates a pressure for females to enlarge their conceptual network. Competing males must keep pace by accumulating an equally large conceptual network and ensuring alignment.

The APH proposes that much of human communication serves alignment goals, grooming between-mind mappings to ensure a complete and consistent representational parity (Dunbar, 1996, 2004). Communicative displays are less common but nevertheless form a prevalent part of the communicative system. These two communicative goals are not mutually exclusive—much alignment communication can have aspects of display and vice versa and the purpose of a communication can quickly switch between goals given changing opportunities. Exchange communications only minimally serve APH functions perhaps entering into the cost benefit assessments of the value of communications as suggested by Dessalles (1998) but they are not directly considered part of the APH.

The argument so far has concentrated on female selection of competing males in intersexual selection situations. If mutual selection occurs in humans APH processes will work intersexually in both directions. Additionally, these processes will result in representational alignment between members of the same sex. Multiple males seeking representational parity with multiple potential mates means that representations will not simply align between pairbonds but will generate a broad social consistency of representations as active alignment diffuses throughout a social grouping. Indeed, Dunbar and Shultz (2007) have suggested that it is typical of anthropoid primates to extend pairbonding cognitive skills to relationships beyond individuals who are potential reproductive partners. It is likely that between-mind mappings will rapidly extend beyond pairbond interactions to include larger social groupings. This means that APH mechanisms are likely to have involved intrasexual selection as well as intersexual selection with both display and alignment communications operating within and between sexes.

A consequence of this within group representational parity is that it increases the likelihood that the representational options for communication will include a high proportion concerning social relationships and group dynamics rather than the concrete perceptual representations of straightforward mind-world mappings (Mesoudi, Whiten, & Dunbar, 2006); this paves the way for a transition to more abstract representations (Read, 2010). We will return to this point in the Social Brain Hypothesis section. Between-mind mappings of representations are, therefore, likely to exist between all members of a social group and be actively maintained by that group.

Once embedded in a species, a sufficiently large conceptual network and representational alignment with the social grouping is a prerequisite to enter the mating game. It would be difficult to find a mate without any language or nonverbal communicative code in common. This necessity places pressures on the relatives of young members of a social group to ensure that their offspring and kin are sufficiently prepared with the requisite between-mind mappings to gain social status and a quality mate when they are mature. This ensures intergenerational continuity of the social group's representational knowledge. It also provides a motivation for pedagogy, older members of a social group must convey knowledge to younger members, provisioning them with the necessary communicative skills and knowledge to successfully compete for a mate in later life. As the cultural knowledge required to compete increases, longer developmental periods are required to aggregate the concepts and refine and align the interrelationships between them. This makes the requirement for adequate between-mind mappings a possible contributing factor to increased developmental length observed in human children.

At this between-mind mapping level there are many opportunities to display mind-reading abilities. Displays can exhibit receiver knowledge: through cooperation, knowing an individual's goals; through acts of generosity and sharing, knowing desires; through adherence to norms and expectations, knowing beliefs; through playful teasing, knowing acceptable moral boundaries. This is where mental fitness indicators related to personality and morality become important and displays of this form will be addressed in greater detail in the Nonverbal Communication section.

An important constraint on these displays arises due to the origin of the representational structures in mind-world mappings. The representations discussed to this point are bound to referents in the world: physical, environmental, and social phenomena. They, therefore, do not possess an arbitrary complexity, as may be the case in the sexual selection of birdsong (Bolhuis, Okanoya, & Scharff, 2010; Fitch, 2009). The communicative space is abstract but tied to an intersubjectively agreed reality.

Within-Mind Mappings

High-quality senders must differentiate themselves from low-quality senders by engaging in displays that are harder to fake. Additionally, as communications can serve three purposes—display, alignment, and exchange—a sender needs to ensure that a display communication is noticed as deliberate rather than coincidental. A solution to these problems arises as the level of complexity in representational structures increases; especially with symbolic representation. Discrete representational structures allow a new kind of hard-to-fake display that low-quality senders find harder to achieve and that makes it easier to determine a purposeful display.

If a set of representations is sufficiently complex and aligned between the minds of two communicators, a sender can search their own representations for subsets that have internal isomorphic structures, that is, subsets that have an analogous structure (see Figure 1). These internal analogies create opportunities for greater reliability in displaying mind-reading abilities. If the between-mind mappings are sufficiently complete and consistent between the two individuals, a sender can assume that the analogous structures in their own mind also exist in the mind of the intended receiver. Communications can then be crafted to take advantage of these within-mind mappings.

The most useful of these analogical structures for a display event are those that a sender suspects have not yet been noticed by the receiver; this parallels the Vartanian et al. (2009) definition of creativity “the novel and useful combination of concepts previously thought to be unrelated.” Display occurs with the communication of the novel connections to the receiver, A is like B. This creative link makes the receiver aware that the sender has a sufficiently consistent between-mind mapping with the receiver to isolate and display knowledge of their within-mind mappings. The communication of a creative combination of concepts is a display of mind-reading ability. The display provides evidence for developmental costs the sender has incurred in developing adequate neural apparatus to perform these mental feats; that a large portion of a lifetime has been spent acquiring cultural knowledge; and that a sender has the ability to gauge others and astutely assess the correct timing to reveal creatively linked connections.

This is the role played by the unifying feature of the relational combination of concepts in the APH. The sender displays mind-reading abilities that can anticipate an unnoticed analogy is present within the mind of the receiver. It sends a message to the effect of “I can read your mind well enough to know you will appreciate this link.” To take an example from fiction, when Cyrano compares Roxane's physical trembling to a leaf “like a leaf among the leaves, you tremble,” he is making a novel connection between a representation of a physical state in a human and a leaf fluttering in the wind. This display shows that he is able to read her mind sufficiently well that he is aware that a within-mind mapping exists in her set of representations and that she will be aware that such a connection is an apt descriptor for current events.

A creatively crafted analogical message can serve as a strong display of mind-reading abilities. However, not all displays of within-mind mappings are equal, they differ in the degree to which they convey mind-reading ability and there are additional strategies that can be used to make signals harder to fake and increase the cost differential. Reliability and attribution of purpose can be increased through displays that are fast, novel, spontaneous, bound to context, topical and relevant, and that utilize remote (semantically distant) and intricate linkage of representations. Prepreparing a communication diminishes its reliability; being quick witted conveys greater ability than being preprepared or scripted. This highlights the importance of the unifying feature of processing speed. This processing speed is essential in the manner suggested by Vartanian et al. (2009) who linked processing speed of relatedness judgments with creativity. They noted the importance of judging whether two concepts were related or unrelated, as only novel linking of as yet unrelated concepts is creative. Faster processing speed would provide a “substantial advantage in the number of potentially useful relationships that could be assessed per unit of time.” Leaving those with lower speeds unable to process, generate, and evaluate relational linkages as fast as those of higher ability.

Contextually bound communications provide another hard-to-fake indicator of mind-reading abilities, highlighting the importance of the unifying feature of contextualization. Including current context and new social events within a creatively crafted display limits the ability to preprepare or borrow from others. Cyrano gains additional cachet as Roxane is sitting among branches when he delivers his monologue. Additionally, knowing what is relevant to a receiver in any given context is itself a display of mind-reading ability and receiver knowledge. In relevance–theoretic terms crafting an utterance that is optimally relevant displays that the sender knows a receiver will be able to combine evidence from the utterance and the context to infer the meaning and extract the cognitive effects. There may be intrinsic exchange value in the cognitive effects, as Dessalles (1998) suggests, but it comes wrapped in a display package tailored to the receiver. In this view a skilled communicator can display the incurred costs of developing precise communicative neural mechanisms by crafting a communication that is tied to the current context and optimally relevant to the receiver.

Another example is useful at this point, this time of a weak display. The somewhat crass male chat-up line “Ten ton polar bear…well I needed something to break the ice,” is weak for many reasons. It is a preprepared display, it initiates conversation with an irrelevant concept, one most unlikely to be salient in the mind of a female. The irrelevant concept is then mapped to the current context through use of an idiomatic metaphor, providing a feeble humorous payoff. Bale, Morrison, and Caryl (2006) suggest that these prepackaged pick-up lines are seldom successful and perhaps only serve to indicate willingness on the part of the male—an ostensive act signaling intent to display—and “to identify sociosexually unrestricted women.”

The ability to link semantically distant concepts, especially in intricate ways, and be aware that these connections can be made in the mind of a receiver, displays knowledge of the extent of between-mind mappings. However, there is an important constraint on which concepts can be connected, pressures toward generating increasingly remote and intricate representational connections are constrained by the requirement to stay within the social group's—or at least the dyadic partner's—between-mind mappings. There is little point in creating and communicating a remote and intricately connected representational structure if it will not be recognized as such by the receiver of a communication.

In the analogical displays suggested by the APH, a receiver is made aware of an unforeseen connection in the world that allows them to assess the mind-reading and abstraction capabilities of the sender. The receiver then assesses the costs of producing the display and hence the genetic quality of the sender. The receiver can then decide whether to continue courtship, elevate or reduce the sender's social status or place increased value in an allegiance with the sender. A sender should also assess the result of their communication and recalibrate their internal models given the new information about the relative success or failure of the display.

To reiterate the hypothesis in terms of the five unifying features and the mental fitness indicators. A communicator wishes to display the quality of neural apparatus which has weathered a turbulent developmental history to arrive at a mating situation prepared to impress. The receiver is interested in these display abilities as they serve as a proxy measure of ability to deal with social complexity. Through the process of development the sender has gathered a large network of conceptual knowledge that will form the basis of the display mechanism. Possession of a large network is not sufficient; it must be aligned with the knowledge of others and the sender requires sufficient receiver knowledge to ensure this. A sender can then craft a signal that reveals novel connections within the receiver's network of concepts—a creative relational combination of concepts. The processes involved in finding and evaluating these relational combinations are similar to those of fluid intelligence, relational functions of working memory, and the evaluation of concepts in creativity. There are many ways to make signals harder to fake; increased processing speed aids the quick witted and allows faster searches for related concepts; including context and maximizing relevance minimizes the opportunity for prior preparation; greater verbal humor, fluency, and timing display linguistic prowess; and humor production encapsulates most of these display components in a single process.

Franks and Rigby (2005) conducted two experiments that provided an almost direct test of this hypothesis. They reasoned (following Gagné, 2000) that when making conceptual combinations from noun–noun pairs context-appropriate relation based combinations are easier to generate than property mapping combinations. In the first case there is a preexisting relationship between the two concepts; in the second case a property of one concept must be mapped on to the other concept. So if a participant is presented with the two nouns such as sponge table (using their example) a relation-linking interpretation such as a table that holds sponges is easier to generate than a property-mapping interpretation such as “a table that is soft and absorbs a lot of water” where a property (absorbency) is mapped from one concept on to the other. This greater difficulty makes the property-mapping solutions more creative than the relation-linking solutions. This is similar to the argument by Vartanian et al. (2009) that linking already established relationships is not creative.

Noun–noun combinations were presented to males and females in three experimental conditions. The only difference between the conditions was the nature of the experimenter who would be assessing their interpretations for “interest value.” One experimenter was a young attractive female, another was a young attractive male, and a third control condition had an older postmenopausal female experimenter. Franks and Rigby (2005) found that significantly greater creativity was displayed by males than females in both of the young experimenter conditions. They reasoned that in the young female condition the extra creativity was a display to the female experimenter and in the male condition it resulted from competition with the experimenter. A further experiment ruled out the possibility that they were also displaying in competition with the young female experimenter.

Franks and Rigby (2005) used these findings to argue that a display creates a situation where greater cognitive efforts on the part of the receiver of a communication are justified to interpret the display communication. Complementary to this position, I argue that these results provide evidence for the display of creative within-mind mappings in a mating primed situation. This methodology can possibly be used to distinguish between alignment and display communications.

As no creative mating response was primed in females these findings would seem to suggest that creative mate choice involves male display and female choice. However, there is still the possibility for female display and male mate choice at a later courtship stage where males become more choosy and reluctant to commit to a longer term relationship. As mentioned previously, Griskevicius et al. (2006) found increased creative production in females when priming emphasized a committed long-term partner; different mating conditions may cause enhanced creativity at different times in the sexes. Additionally, cultural factors are likely to exert an influence on the appropriate situations to initiate mating behavior and produce ostensive acts signaling intent to display.

This section has outlined in detail the mechanism of the APH; for a more detailed motivation and an extension of the hypothesis we now turn to the social brain hypothesis.

Nonverbal Communication

Although full analogical display requires symbolic representation, it is possible to display mind-reading abilities in nonverbal communication through joint attention, cooperation, and imitation.

Simply attending to the same thing as another individual displays knowledge of that individual's focus of attention. The development of a shared attentional frame can signal the ability to know another individual's concerns. Nonverbal ostensive acts such as eye contact, gaze-shift, pointing, eyebrow-raising, or eye-rolling can create a shared attentional communicative frame in which both alignment and display communications can take place. In an alignment communication the act may seek clarification on an “are you thinking what I'm thinking” basis, in a display situation such an act may communicate “I think you will be interested in this.”

Mind-reading abilities can be displayed in acts of cooperation by reading an individual's goal and actions toward that goal and displaying this ability by aiding them toward it. The sender's message is “I am skilled at reading minds and know your desires and how you want to achieve them,” if the sender aids the receiver this skill is made obvious. A receiver never need complete their course of actions as their mind has been sufficiently well read that their course of action is usurped by the mind reader and completed for them as a display. Cooperation in these circumstances results from a self-interested display of mind-reading ability that appears at a more superficial level to be an altruistic event.

This line of reasoning would predict greater cooperation where display motivation is higher. In a meta-analytic study, Langlois et al. (2000) showed that physical attractiveness led to significantly more favorable help-giving and cooperation behaviors. In a mainly student sample, Mulford, Orbell, Shatto, and Stockard (1998) found that perceived attractiveness predicted participants’ willingness to cooperate in a sequence of prisoner's dilemma games. Andreoni and Petrie (2008) using a sample of economics and business students, found that attractive people engendered more cooperation from others in public goods games.

Nowak, Page, and Sigmund (2000) explain a consistent finding in ultimatum games using a fairness due to reputation model. In these games one player splits a sum of money with a partner who can accept or reject the chosen sum. Participants most often share the money evenly or almost evenly, whereas game theory predicts that the lowest possible offer should be made and would be accepted (as happens with chimpanzees, Jensen, Call, & Tomasello, 2007; although see Proctor, Williamson, de Waal, & Brosnan, 2013). However, if we recast this as a display of knowledge of another individual's expectations then sharing may provide its own reward. The display combined with a desire for alignment could offset the costs and derive benefits from distributing the money more evenly.

The desire to display mind-reading ability provides a motivation for imitation over emulation. In emulation, the goal is copied and reached by any possible means; in imitation, the means of reaching the goal are copied. Cooperative emulation—where aid brings about the goal of the receiver rather than aiding actions to achieve the goal—provides a weaker display of mind-reading ability. Displaying knowledge of a receiver's goal and the method they have chosen to reach it shows greater mind-reading abilities. The APH predicts greater effort toward, and more favorable reception of cooperation where the means rather than the goal is assisted, particularly in mating or status primed situations. Direct experimental evidence for this prediction is hard to come by (although see Guéguen & Lamy, 2011; Lamy, Fischer-Lokou, & Guéguen, 2009). There is, however, an extensive literature on the affiliative nature of mimicry (Chartrand & Bargh, 1999; Gueguen, Jacob, & Martin, 2009; Lakin & Chartrand, 2003) and increasing evidence for the universality of imitation (Nielsen & Tomaselli, 2010). The chameleon effect refers to nonconscious mimicry of nonverbal behaviors in an interlocutor which often leads to increased likability and empathy. Lakin and Chartrand (2003) found that having an affiliative goal, and also failure to reach the goal, both increased the level of nonconscious mimicry among psychology students. Cooperative behavior can be viewed as displays of mind-reading abilities, particularly displays of receiver knowledge. In problem solving situations, creative solutions can display within-mind mappings by highlighting shared connections that have not yet been noticed by the receiver. Insightful cooperative solutions should increase the perceived mate value and social status of the sender.

Pickering and Garrod (2012) have argued that alignment plays an important role in human communication and that imitation in particular is important in this process of aligning of mental states (Pickering & Garrod, 2004; Pickering & Garrod, 2007). Wang and Hamilton (2012) argue for a social top-down response modulation (STORM) such that mimicry does not just happen for reasons of alignment but is carefully controlled to maximize social advantage—in this sense imitation works more like a display where it is used to influence social standing. The APH view accommodates both options; imitation can function both as an affiliative alignment process and as a display of mind-reading ability used to elevate social status.

Verbal Communication

A shift to the use of discrete symbolic representations in human communication allows more complex displays of mind reading abilities. There are many accounts of symbolic representation. Deacon (1997), a strong advocate of the importance of symbolic representation in the evolution of language, provides a concise account. He adopts Peirce's (1897) classic semiotic triad of icons, indices and symbols to emphasize a hierarchical nature in the representation of signals. Icons are associated with their referents because they are similar or resemble them. An index has a causal link with its referent, while at the top of the hierarchy and involving many indices is the symbol. In Deacon's view a symbol differs from an index because symbols are interrelated with one another. A symbol is not simply an arbitrary one to one relationship between a referent and a representation, it also has a relationship to an interconnected system of arbitrary sounds and referents that mutually reinforce each other. An index can be extinguished if the correlation between it and its referent becomes nonexistent or unreliable. A symbol, in contrast, persists because it has a relational scaffold with other symbols. Normally a symbol retains its link to an object but it also has mutually reinforcing relationships with other symbols. The set of all symbols form a complete system of reference.

A relationally interconnected system of symbolic representations like this provides an ideal substrate for within-mind mappings. However, the APH explicitly extends Deacon's mutually reinforcing scaffold of symbols to include the interrelationship with symbolic representations in the minds of other individuals—between-mind mappings. The meanings associated with symbols are not only supported by other symbols in one's own mind but constantly and mutually aligned by interaction with other individuals.

Gentner (2010) and Gentner and Christie (2010) propose that symbolic systems interact with relational cognition to form a mutual bootstrapping relationship creating a positive feedback loop that supports language learning in humans. They highlight an important function of symbolic naming, that it promotes reification. This reification of representations enables easier comparison and importantly cultivates uniform relational encoding so that there is a consistency in the manner in which symbols relate to one another across times and contexts. Importantly for the APH, symbolic encoding allows this consistency in relational encoding to exist between individuals, creating a socially pervasive relational encoding that maps between-minds and forms the basis for displays based on shared within-mind mappings.

Earlier work in analogical reasoning and structure mapping (Gentner, 1983; Gick & Holyoak, 1980, 1983) focused on the problem solving and inference properties of analogy but more recent views have placed analogy more centrally in human cognition (Gentner, 2010; Gentner, Holyoak, & Kokinov, 2001; Halford et al., 2010; Hofstadter, 2001; Penn et al., 2008) and also in its form as metaphor it has been argued that it is foundational to language and social cognition (Deutscher, 2005; Lakoff & Johnson, 1980; Landau, Meier, & Keefer, 2010).

All of the unifying features of mental fitness indicators are present in these analogical processes; however, two are particularly salient, the relational combination of concepts and large conceptual networks. The literature on dynamic relational binding integrates analogical processing research with working memory and fluid intelligence. The ability to represent relations independent of their arguments is particularly important, and closely related to working memory. Concepts can be bound to relations in a dynamic and flexible manner and this flexibility may aid in finding relationships. Holyoak and Hummel (2001) implicate working memory in this process making the strong claims that “relational processing may form the core of an executive component of prefrontal working memory” and “relational integration—and specifically, dynamic variable binding—may be the ‘work’ done by working memory.” This link that has also been made by other research programs (Halford, Wilson, & Phillips, 1998; Oberauer, Schulze, Wilhelm, & Süβ, 2005; Speed, 2010; Waltz et al., 1999). Additionally, limiting working memory resources and capacity has been shown to impair relational integration and analogical reasoning (Cho, Holyoak, & Cannon, 2007; Holyoak, Waltz, Tohill, & Lau, 1999; Morrison, Holyoak, & Truong, 2001). According to the APH it is these relational cognition processes that form the basis for displays of creativity, language, and humor as mental fitness indicators within social interactions.

Another research effort relevant to shared within-mind mappings and communicable analogical structure is the idea of conceptual blending and integration (Fauconnier & Turner, 2002; Fauconnier & Turner, 1998). This takes a broader, more discourse related view of analogy than the structure mapping view of Gentner (1983) and encompasses a greater sweep of phenomena. Additional to analogy, conceptual integration claims to account for category extension, metaphor, metanymy, framing, counterfactuals, and grammatical constructions. It is also implicated in creativity, humor, mathematics, and in artistic, scientific, and religious thought.

In this view “mental spaces” are representational structures containing discourse entities. These spaces are integrated to create new blended spaces through different kinds of mappings including identity, similarity, change, cause/effect, space, time, metaphoric, and analogical. The selectively projected blend of two spaces allows novel inferences by removing constraints from the input spaces and fusing some parallel elements producing “emergent structure” with properties not present in the original spaces. The processes involved are largely unconscious and only the product reaches the level of consciousness.

The nature of these blends is contentious (Gibbs, 2000) but for Fauconnier and Turner (2002) it is a fundamental cognitive operation working at all levels of abstraction, a universal process for integrating concepts to create new meaning. The utility of conceptual blending is often cited in terms of “productive changes in the conceptualizer's knowledge base and inferencing capacity” (Coulson & Oakley, 2005). The APH views these cognitive mechanisms as processes that create and allow the communication of within-mind mappings in order to display mind-reading abilities.

Discourse, narrative, and storytelling based communications offer culturally defined and strongly aligned between-mind mappings. This alignment allows more complex displays of higher order intentionality abilities and social relationship representations; often these have obvious within-mind mappings as in allegory and parody. The APH would predict that storytelling abilities should function as mental fitness indicators and there is some preliminary evidence that storytelling ability increases attractiveness (Donahue & Green, 2008). Narrative skills and storytelling abilities may also be susceptible to developmental insult (Demir, Levine, & Goldin-Meadow, 2010), perhaps highlighting a role for complex linguistic abilities as indicators of developmental stability. A parsimonious example of storytelling as a display is provided by the probably apocryphal tale of Ernest Hemingway's “shortest story ever told.” Hemingway supposedly used his story writing prowess and receiver knowledge to craft a story in just six words in order to win a bet: “For sale: baby shoes, never worn.” This creative combination of concepts is enough to invoke a narrative structure with a deep emotional resonance and we elevate the authors’ status accordingly. There is also evidence for the alignment function of storytelling; Mar and Oatley (2008) argue that fiction serves to improve Theory of Mind, the understanding of intentions, the nature of empathy, and the development and improvement of social skills.

In a somewhat reciprocal relationship there are those that suggest that an understanding of psychological (Zunshine, 2003) and Darwinian processes can aid our understanding of literature (Carroll, 1995). From an APH point of view we might expect literature to conform more strongly to a world defined by between-mind mappings rather than one that matches the stochastic realities of an objective world. Literature should have a bias toward between-mind mappings and an aligned intersubjectively agreed understanding of reality—it should, therefore, not be surprising that fact is stranger than fiction.

Both structure mapping and conceptual blending offer mechanisms for the display of mind-reading abilities using shared within-mind mappings. Fauconnier and Turner (2002) firmly place conceptual integration and blending as central mechanisms in human cognition, and increasingly analogy is viewed as a central feature of human cognition. Hofstadter (2001) places analogy at the core of human cognition, in his view we engage in analogical cognition without effort and our brains are constantly searching for patterns between abstract representations that function as scaffolds to structure our meanings. Gentner (2010) has suggested that a “mutual facilitation between relational ability and language” explains unique human cognition and how children develop linguistic and thinking abilities so rapidly. A mutual bootstrapping occurs where Gentner claims “structural alignment supports language learning, and that in turn language—especially relational language—supports structural alignment and reasoning.” Clearly the APH concurs with these views and offers a possible explanation for why this core centrality might exist.

There are important implications of these alignment and display processes for many aspects of communication consistently highlighted in the evolution of language literature. The combination of concepts provides a basis for compositionality, displays of intentionality provide the motivation for the inclusion of recursion in speech. There are also strong implications for cultural development. Creative connections are one-off displays, a creative connection that becomes common knowledge loses its utility for display—unless totally forgotten by a social group. The need to produce creative displays gives rise to a constant stream of novel representations. Those that are useful, shareable, and learnable can become conventionalized and included in mutually aligned sets of representations. These processes create the cultural ratchet of cyclical creative display and conventionalization that enlarge our conceptual networks and in turn our cultural knowledge.

Creative Arts and Performance

The key aspects of the APH converge in many of the creative and performance arts. These represent ritualized display opportunities with resultant status elevation for successful performance. Ritualization creates a strong cultural alignment with clearly defined roles for the participants. The fact that we elevate successful performers to the highest echelons of social status is clear from the often unconditional adulation they receive and the names we give to those that succeed, A-listers, cultural elite, and glitterati.

Mutually Entangled Encephalization

A final implication involves expectations for differences between the sexes that result from APH processes. The lack of sexual dimorphism in the enlarged brain volume of humans, encephalization, has been used as an argument against mate choice mental fitness indicators. Hooper and Miller (2008) have produced plausible accounts that assortative mating and mutual mate choice may produce extravagant ornamentation in monogamous sexual selection situations allowing encephalization without sexual dimorphism. However, from the APH point of view it is the constant need to be highly aligned to other individuals in order to produce quality displays that has prevented strong sexual dimorphism, both in cognition and in the neural apparatus that supports it.

In male to female display, competition among males drives them to enhance displays and females must be able to discern quality. However, overly ambitious male displays are too creative and combine concepts that fall outside between-mind mappings producing low quality displays. Females who are divergent from the cultural norms of between-mind mappings cannot discern quality displays and will choose poorly indicated fitness and lose the benefits of good-genes and more attractive male offspring. This creates a situation where encephalization pressures exist but intense negative feedback ensures that males and females are representationally entrained with one another. There is a need to increase creativity in display but this is always constrained within the envelope of between-mind mappings. The twin tensions of creative display and alignment keep male and female minds in lockstep, driving a mutually entangled encephalization that limits sexual dimorphism. This may explain the situation that although we might expect greater disparities between female and male cognition—given their often contrasting sexual strategies—the evidence for between sex differences in mental abilities remains equivocal (Hyde, 2005).