5.3 Narrow Evolutionary Psychology and the Emergence of Modularity

However, there is a debate among evolutionary psychologists as to (1) the

type and number of mental modules the human mind contains, (2) the

exact time period or time periods when these mental modules were solidifi

ed in the early hominin psyche, and (3) whether these mental modules

have arisen directly through adaptation, or indirectly as an evolutionary

side effect/by-product through exaptation (as in Gould & Vrba’s, 1982,

spandrels), or even through some form of cultural evolution.

Scher & Rauscher (2003) and Wilson (2003) have drawn a distinction

between what they call narrow evolutionary psychology (NEP) and broad

evolutionary psychology (BEP). Advocates of NEP follow the groundbreaking

work of Cosmides & Tooby (1987, 1992, 1994), arguing that the mind

is like a Swiss Army knife loaded with specifi c mental tools that evolved in

our Pleistocene past to solve specifi c problems of survival, such as face

recognition, mental mapping, intuitive mechanics, intuitive biology,

kinship, language acquisition, mate selection, and detection of cheaters

(the list could be longer or shorter; cf. Palmer & Palmer, 2002; Buss, 1999;

Pinker, 1994, 1997, 2002; Gardner, 1993; Shettleworth, 2000).

In response to (1)–(3), adherents to NEP argue that (1) the mind is a host

of specialized, domain-specifi c mental modules, (2) the Pleistocene epoch

is the time period in which the basic psychological structure of the modern

human mind was solidifi ed in our genetic makeup, and (3) these modules

have arisen directly through specifi c adaptive problems that early hominins

faced.

In contrast to NEP, advocates of BEP consider alternative approaches to

Cosmides & Tooby’s Pleistocene-epoch-forming, Swiss Army knife model

of the mind and want to argue that (1) the mind probably does not contain

the myriad of specialized, domain-specifi c mental modules that the NEPers

would have us believe but relies more upon domain-general mental capacities

that have evolved to handle the various and sundry problems a human

faces (Samuels, 1998; Wheeler & Atkinson, 2001; Laland & Brown, 2002),

(2) although the Pleistocene epoch is a signifi cant time period in our evolutionary

past, it is by no means a single environment, nor is it the only

environment that has shaped the modern mind (Foley, 1995; Boyd & Silk,

1997; Daly & Wilson, 1999), and (3) this mental architecture probably has

evolved through adaptive, as well as exaptative and cultural forms of evolutionary

processes (Barrett, Dunbar, & Lycett, 2002; Laland & Brown,

2002; Otto, Christiansen, & Feldman, 1995; Buller, 2005).

All evolutionary psychologists are in agreement with the fact that

certain environmental selection forces were present in our early hominin

past and that these forces contributed to the mind’s formation. It seems,

then, that forming an accurate picture of what those selection forces were

like is integral to our understanding of the mental mechanisms that have

survived the process. At the same time, once we have an understanding

of the environmental challenges faced by our early hominins, we can get

a better picture of what our mental architecture has evolved to look

like.

Part of what I will do in this chapter is to try and adjudicate between

NEP and BEP by utilizing Mithen’s idea of cognitive fl uidity and my

account of scenario visualization that is rooted in problem solving tasks

our early hominins would have faced in their environments (also see Arp,

2006a, 2007a). After a presentation of Cosmides & Tooby’s NEP approach,

as contrasted with one BEP approach put forward by Mithen, I will develop

my account of scenario visualization further so as to get a more accurate

picture of our mental architecture and the conditions that occasioned its

evolution.

The fossil and paleogeographic evidence suggests that speciation (the

formation of new species) occurs at times of rapid, punctuated environmental

change, rather than during periods of relative stability (Gould, 1977,

2002; Eldredge, 2001; Calvin, 1998, 2001, 2004; Potts, 1996). Advocates of

NEP wager that primate evolution is no different and took place against a

rapidly changing climatic and geographic background. Global climates

have changed greatly during the past 60 my, and especially in the past

20 my. Overall, the world’s climate has become cooler and more seasonal, with less forestation, more deserts, and more ice on its surface. The key

period of climatic change that occasioned the evolution of mental modularity

was around 2.5 mya, just prior to the Pleistocene epoch, when there

was a global shift from warm and wetter to cooler and drier conditions.

The climate during the time period just prior to the Pleistocene exhibited

more unpredictability than it had in the past, “fl ip-fl opping”—a term

Calvin (1996) uses—from warm, to hot, to cool, to dry, to warm and dry,

to cool and dry, and so forth. In Africa, Europe, Asia, and North America,

given the newer environmental niches, species of animals and plants

appeared in bursts (Dawkins, 2005; Eldredge, 2001; Calvin, 1998).

In the midst of all of the climate change, new food sources, and different

species emerging on the scene, Cosmides & Tooby (1994, p. 90) tell us that

“simply to survive and reproduce, our Pleistocene ancestors had to be good

at solving an enormously broad array of adaptive problems—problems that

would defeat any modern artifi cial intelligence system.” The analogy to a

computer is appropriate. Generalized computer programs equipped with

step-by-step algorithmic processing perform slowly and fail to perform the

simplest of tasks that even earthworms can perform, like negotiating a

maze. However, parallel processing computer mechanisms fare much better

in terms of learning and negotiating environments (see Cziko, 1995; French

& Sougne, 2001; Lek & Guegan, 2000; Lerman & Rudolph, 1994).

In their experiments comparing general-purpose computational mechanisms

and parallel processing computational mechanisms, Rumelhart &

McClelland (1985) have shown that the rate at which general-purpose

mechanisms process multiple pieces of disparate information is much

slower than that of parallel processing mechanisms. This is so because the

general-purpose mechanisms have to work longer and harder at cataloguing,

categorizing, and then storing the disparate pieces of information,

whereas the parallel units are composed of processors that are specialized

to recognize a particular piece of information and work simultaneously

(thus, the parallel processing) to store information (also see Roosta, 1999;

Copeland, 1993; Searle, 1992; Fodor, 2001; Churchland, 1986; McFarland

& Bosser, 1993). Further tests performed by Connell (1989), Brooks (1991),

and Franceschini, Pichon, & Blanes (1992) have shown that parallel processing

robotic mechanisms have a quicker and easier time of collecting

Coke cans from around MIT labs or navigating to some light source than

do general-purpose kinds of robotic mechanisms. In the words of Culler &

Singh (1999, p. 4), “whatever the performance of a single processor at a

given time, higher performance can, in principle, be achieved by utilizing

many such processors.” Evolutionary psychologists reason similarly that a single calculating

mechanism with the same set of rules, meant to cover a multitude of tasks,

would have processed information slowly and led to many errors. To use

an example from Buss (1999), if our early ancestors had a generalized rule

like “have sexual intercourse with any partner you can,” then, in terms of

the ultimate goal of propagating genes, such a rule would be benefi cial

with respect to nonkin but would backfi re with respect to kin. A parallel

processing, modular kind of mind would fare much better because it would

have more specialized routines designed to handle a variety of situations.

In other words, an individual module that has emerged to handle only

one kind of problem likely will be able to handle that problem swiftly and

effi ciently because it has to handle only that particular kind of problem,

and no other one.

In essence, modularity minimizes errors and allows systems to perform

optimally. We should not underestimate the importance of this kind of

reasoning on the part of NEPers. A speedy response and the minimization

of error grant the system a competitive advantage—hence, the likelihood

of such specifi ed, parallel processing mechanisms being selected for in our

early hominin mental architecture (see Arp, 2008b). This comports with

the general evolutionary principle of economy, recognized by every evolutionist

since Darwin, namely, whatever trait gives an organism a competitive

advantage most likely will be naturally selected as fi t for that organism

in relation to an environment and likely will be passed on to that organism’s

progeny.

It appears that the parallel kind of processing has been selected for with

respect to at least some of our mental architecture. As I intimated above

and in the third chapter, there is ample evidence that the visual system is

considered as a suite of coordinated physiological or brain-process modules

engaged in the parallel processing of visual information. So too, Broca’s

area and Wernicke’s area in the human brain are engaged in the parallel

processing of grammar–usage and language comprehension, respectively.

Further, the face-recognition area in the IT cortex works in parallel with

other areas of the IT cortex, and other areas of the brain, to help someone

distinguish faces from other objects.

Numerous studies on infants, children, and adults seem to confi rm the

fact that people have innate mental modules seemingly designated for

specifi c tasks. For example, Chomsky (1964) has argued convincingly that

there must be some innate capacity for language, since young children

from any culture can pick up language easily, as well as being able to learn

any language (see also Jackendoff, 1987, 1992, 1994). This makes sense from the NEP perspective, since Pleistocene hominins formed social groups

and eventually communicated with one another through language during

that time (Aiello, 1996; Aiello & Dunbar, 1993). Spelke (1991) has demonstrated

that children as young as two years old have an apparent innate

understanding of physical properties of objects like solidity, gravity, and

inertia (see also Pinker, 1994). This also makes sense from the NEP perspective,

since Pleistocene hominins constructed and handled a variety of tools

in a variety of ways during that time. Palmer & Palmer (2002) demonstrate

how people have mental modules attuned to certain fears, detection of

cheaters, empathizing, and spatial reasoning (cf. Adolphs, Tranel, Damasio,

& Damasio, 1995; Nesse & Abelson, 1995). Gardner’s (1993) list of multiple

intelligences of the mind is accepted by so many psychologists and educators

that it forms the basis for the curricula of many primary and middle

schools (Gardner, 1999). Pinker’s (2002) list includes an intuitive knowledge

of physics, biology, engineering, psychology, spatial sense, number

sense, probability, economics, logic, and language. Kandel et al. (2000, p.

412) tell us simply, “A newborn’s mind is not blank.” According to NEPers,

all of these capacities—language, intuitive physics, automatic knee-jerk

responses to snakes and spiders, detection of cheaters, and so forth—most

likely were solidifi ed in our species’ psyche during the Pleistocene epoch.

According to advocates of NEP, the adaptive problems in the Pleistocene

environments occasioned the emergence of psychological modules designed

to handle the various and sundry problems of such environments. Several

basic components of our present-day psyches were solidifi ed back then

and, in the words of Cosmides & Tooby (1987, p. 34), “the complex architecture

of the human psyche can be expected to have assumed approximately

modern form during the Pleistocene . . . and to have undergone

only minor modifi cation since then.” We must remember that the claim

NEPers make regarding the solidifi cation of our mental architecture occurs

over a period of many years, since the Pleistocene epoch spans approximately

2 mya to 10,000 ya. There are more than 65,000 generations of one

family line of one population alone that lived during that time! Thus, given

this vast amount of time, it is not wholly implausible and, in fact, it is

very possible that our human psyche was formed during this time period.

In other words, given what we know about laws of probabilities in relation

to genetic variability and natural selection, prima facie the hypothesis is

not that outlandish.

Evolutionary psychologists speak of these modules as domains of specifi

city. What this means is that a module handles only one kind of adaptive

problem to the exclusion of others. Modules are encapsulated in this sense and do not share information with one another (Fodor, 1983, 1985). For

example, my cheater-detection module evolved under a certain set of circumstances

and has no direct connection to my fear-of-snakes module,

which evolved under a different set of circumstances. Like the various

kinds of tools in a Swiss Army knife, the various mental modules are supposed

to solve the various problems that arise in circumstances; however,

they do so to the exclusion of each other. The scissors of the Swiss Army

knife are not directly functionally related to the Phillip’s-head screwdriver,

which is not directly functionally related to the toothpick, and so forth.

This kind of encapsulation works best for environments where the

responses need to be quick and routine—such developments enabled these

organisms to respond effi ciently and effectively in their environments.

This being the case, the modules could perform quite well as long as the

environments remained relatively unchanged and typical. In fact, most of

this modular processing in mammals occurs at the unconscious level. It is

arguable that since this processing occurs at the unconscious level and,

further, since information can become memorized, mammals quickly are

able to respond to the pressures associated with fi ghting, fl eeing, eating,

mating, and so forth (Fodor, 1983, 1985; Hermer & Spelke, 1996; Cosmides

& Tooby, 1992; Shettleworth, 2000).

However, there seems to be a fundamental limitation in the NEPers’

reasoning, especially if the environment in which the domain-specifi c

module has been selected is supposed to have remained fairly stable. Cosmides

& Tooby (1994) note that these domain-specifi c modules have

evolved “for solving long-enduring [my italics] adaptive problems,” and

Hirschfeld & Gelman (1994, p. 21) characterize a module as a “stable

response to a set of recurring [my italics] and complex problems faced by

the organism.” Now Daly & Wilson (1999), Foley (1995), and Boyd & Silk

(1997) have shown that the Pleistocene did not consist of a single hunter–

gatherer type of environment but was actually a constellation of environments

that presented a host of challenges to the early hominin mind. Thus,

the fi rst problem for advocates of NEP has to do with the possibility of the

environment in which a particular module evolved being stable enough for

the module to have evolved. In other words, Daly et al.’s criticism of NEP

is that the environments in which our Pleistocene ancestors lived were too

varied and too erratic for the Swiss Army knife blades to be solidifi ed in

our genetic makeup.

We need only refl ect on the kind of circumstances one experiences

through the course of the proverbial bad day to see that one is bombarded

constantly with novel pieces of information. Throughout this bad day, one fi nds oneself in a repertoire of novel mini-environments, from waking up

to discover your toilet has overfl owed, to locating a bus schedule and then

riding a bus to work because you discovered your car broke down, to

searching for some twine in your cubicle at work so that you can hold your

pants up because in the rush out the door you forgot to wear a belt, and

so forth. All of this was atypical of your normal day.

In fact, as Crick & Koch (1990) and Kosslyn & Koenig (1995) note, novel

scenes are being presented to the visual system regularly. Everyday objects

and circumstances in our visual fi eld constantly are obscured, occluded, or

observed from different angles: “There are an almost infi nite number of

possible, different objects that we are capable of seeing. . . . The combinatorial

possibilities for representing so many objects at all different values of

depth, motion, color, orientation and spatial location are simply too staggering”

(Crick & Koch, 1990, p. 268). Today, I made a turn onto Main

Street and looked down the block to see Mr. Jones’ oak tree bathed in

sunlight; yesterday, I made the same turn in the pouring rain but had to

squint through the rain-soaked glass of my car to see the same tree darkened

by the storm. Also, as I move around my cat teasing her with a ball

of string, her body contorts into a variety of positions, and I experience

her shape anew with every different angle.

Even if we tempered Daly et al.’s claims regarding the multitude of environments

faced by our ancestors and grant that the Pleistocene consisted

of a more unitary Stone Age, hunter–gatherer, life-out-on-the-savanna kind

of existence—like the one Cosmides & Tooby would have us believe—then

we have the further problem of the possibility of some stable and routine

module being able to handle the unstable and nonroutine events occurring

in some environment. When routine perceptual and knowledge structures

fail, or when atypical environments present themselves, it is then that we

need to be innovative in dealing with this novelty. If mental modules are

encapsulated and are designed to perform certain routine functions, how

can this modularity account for novel circumstances? The problem for NEP

can be phrased in the form of a disjunction: either (1) the environment

was not stable enough to occasion the emergence of domain-specifi c

modules, as is part of the thrust of Daly et al.’s criticism, or (2) the environment

was stable, allowing for domain-specifi c mental modules to

emerge, but then the environment changed, making it such that the

modules specifi ed for the old environment would no longer be helpful in

the new environment.

Now, imagine the Pleistocene epoch. The climate shift in Africa from

jungle life to desert/savanna life forced early hominins to come out of the trees and survive in totally new environments. Given a fortuitous genetic

code, some hominins readapted to new African landscapes, and some

migrated to new places like Europe and Asia, but most died out. This environmental

shift had a dramatic effect on modularity, since now the specifi c

content of the information from the environment in a particular module

was no longer relevant. The information that was formerly suited for life in a

certain environment could no longer be relied upon in the new environmental

niches. Appeal to modularity alone would have led to certain death and

extinction of many mammalian species. In fact, countless thousands of

mammalian species did become extinct, as fossil data indicate (Novacek,

2002; Dingus, 1990). Elsewhere, I have called mental disruptions of this

nature cognitive dissonance (see Arp, 2004a, 2004b).

The successful progression from typical kinds of environments to other

atypical kinds of environments would have required some other kind of

mental capacity to emerge in our hominin ancestors that creatively could

handle the new environments. Mere mental associations, or trial-and-error

kinds of mental activities, would not be enough, since the environments

in which these hominins found themselves were wholly new, and there

would have been no precedent by or through which one could form

mental associations utilizing past information. Mental associations deal

with the familiar. What is one to do when encountering the wholly unfamiliar?

Although important, modules have their limitations, since they do

all of their associative work in routine environments. What happens if an

environment radically changes, making the information that a particular

module characteristically selects in a familiar environment no longer relevant

in a wholly new environment? A radical readaptation and readjustment

would be needed, one that transcends the limitations of the

routine.

Recall that nonroutine creative problem solving involves fi nding a solution

to a problem that has not been solved previously. The invention of a

new tool would be an example of nonroutine creative problem solving

because the inventor did not possess a way to solve the problem already.

This totally new environment would require that we be creative or innovative

in order to survive. But how is it that we can be creative? The signifi cant

question becomes, then, this: How is it that humans evolved the ability to

engage in forms of nonroutine creative problem solving, especially given

either that the Pleistocene environment in which early hominins existed

was really a constellation of ever-changing environments (Daly et al.’s criticism)

or that any one environment was fi lled with a myriad of nonroutine problems

that seem only to be able to be handled creatively?

However, there is a debate among evolutionary psychologists as to (1) the

type and number of mental modules the human mind contains, (2) the

exact time period or time periods when these mental modules were solidifi

ed in the early hominin psyche, and (3) whether these mental modules

have arisen directly through adaptation, or indirectly as an evolutionary

side effect/by-product through exaptation (as in Gould & Vrba’s, 1982,

spandrels), or even through some form of cultural evolution.

Scher & Rauscher (2003) and Wilson (2003) have drawn a distinction

between what they call narrow evolutionary psychology (NEP) and broad

evolutionary psychology (BEP). Advocates of NEP follow the groundbreaking

work of Cosmides & Tooby (1987, 1992, 1994), arguing that the mind

is like a Swiss Army knife loaded with specifi c mental tools that evolved in

our Pleistocene past to solve specifi c problems of survival, such as face

recognition, mental mapping, intuitive mechanics, intuitive biology,

kinship, language acquisition, mate selection, and detection of cheaters

(the list could be longer or shorter; cf. Palmer & Palmer, 2002; Buss, 1999;

Pinker, 1994, 1997, 2002; Gardner, 1993; Shettleworth, 2000).

In response to (1)–(3), adherents to NEP argue that (1) the mind is a host

of specialized, domain-specifi c mental modules, (2) the Pleistocene epoch

is the time period in which the basic psychological structure of the modern

human mind was solidifi ed in our genetic makeup, and (3) these modules

have arisen directly through specifi c adaptive problems that early hominins

faced.

In contrast to NEP, advocates of BEP consider alternative approaches to

Cosmides & Tooby’s Pleistocene-epoch-forming, Swiss Army knife model

of the mind and want to argue that (1) the mind probably does not contain

the myriad of specialized, domain-specifi c mental modules that the NEPers

would have us believe but relies more upon domain-general mental capacities

that have evolved to handle the various and sundry problems a human

faces (Samuels, 1998; Wheeler & Atkinson, 2001; Laland & Brown, 2002),

(2) although the Pleistocene epoch is a signifi cant time period in our evolutionary

past, it is by no means a single environment, nor is it the only

environment that has shaped the modern mind (Foley, 1995; Boyd & Silk,

1997; Daly & Wilson, 1999), and (3) this mental architecture probably has

evolved through adaptive, as well as exaptative and cultural forms of evolutionary

processes (Barrett, Dunbar, & Lycett, 2002; Laland & Brown,

2002; Otto, Christiansen, & Feldman, 1995; Buller, 2005).

All evolutionary psychologists are in agreement with the fact that

certain environmental selection forces were present in our early hominin

past and that these forces contributed to the mind’s formation. It seems,

then, that forming an accurate picture of what those selection forces were

like is integral to our understanding of the mental mechanisms that have

survived the process. At the same time, once we have an understanding

of the environmental challenges faced by our early hominins, we can get

a better picture of what our mental architecture has evolved to look

like.

Part of what I will do in this chapter is to try and adjudicate between

NEP and BEP by utilizing Mithen’s idea of cognitive fl uidity and my

account of scenario visualization that is rooted in problem solving tasks

our early hominins would have faced in their environments (also see Arp,

2006a, 2007a). After a presentation of Cosmides & Tooby’s NEP approach,

as contrasted with one BEP approach put forward by Mithen, I will develop

my account of scenario visualization further so as to get a more accurate

picture of our mental architecture and the conditions that occasioned its

evolution.

The fossil and paleogeographic evidence suggests that speciation (the

formation of new species) occurs at times of rapid, punctuated environmental

change, rather than during periods of relative stability (Gould, 1977,

2002; Eldredge, 2001; Calvin, 1998, 2001, 2004; Potts, 1996). Advocates of

NEP wager that primate evolution is no different and took place against a

rapidly changing climatic and geographic background. Global climates

have changed greatly during the past 60 my, and especially in the past

20 my. Overall, the world’s climate has become cooler and more seasonal, with less forestation, more deserts, and more ice on its surface. The key

period of climatic change that occasioned the evolution of mental modularity

was around 2.5 mya, just prior to the Pleistocene epoch, when there

was a global shift from warm and wetter to cooler and drier conditions.

The climate during the time period just prior to the Pleistocene exhibited

more unpredictability than it had in the past, “fl ip-fl opping”—a term

Calvin (1996) uses—from warm, to hot, to cool, to dry, to warm and dry,

to cool and dry, and so forth. In Africa, Europe, Asia, and North America,

given the newer environmental niches, species of animals and plants

appeared in bursts (Dawkins, 2005; Eldredge, 2001; Calvin, 1998).

In the midst of all of the climate change, new food sources, and different

species emerging on the scene, Cosmides & Tooby (1994, p. 90) tell us that

“simply to survive and reproduce, our Pleistocene ancestors had to be good

at solving an enormously broad array of adaptive problems—problems that

would defeat any modern artifi cial intelligence system.” The analogy to a

computer is appropriate. Generalized computer programs equipped with

step-by-step algorithmic processing perform slowly and fail to perform the

simplest of tasks that even earthworms can perform, like negotiating a

maze. However, parallel processing computer mechanisms fare much better

in terms of learning and negotiating environments (see Cziko, 1995; French

& Sougne, 2001; Lek & Guegan, 2000; Lerman & Rudolph, 1994).

In their experiments comparing general-purpose computational mechanisms

and parallel processing computational mechanisms, Rumelhart &

McClelland (1985) have shown that the rate at which general-purpose

mechanisms process multiple pieces of disparate information is much

slower than that of parallel processing mechanisms. This is so because the

general-purpose mechanisms have to work longer and harder at cataloguing,

categorizing, and then storing the disparate pieces of information,

whereas the parallel units are composed of processors that are specialized

to recognize a particular piece of information and work simultaneously

(thus, the parallel processing) to store information (also see Roosta, 1999;

Copeland, 1993; Searle, 1992; Fodor, 2001; Churchland, 1986; McFarland

& Bosser, 1993). Further tests performed by Connell (1989), Brooks (1991),

and Franceschini, Pichon, & Blanes (1992) have shown that parallel processing

robotic mechanisms have a quicker and easier time of collecting

Coke cans from around MIT labs or navigating to some light source than

do general-purpose kinds of robotic mechanisms. In the words of Culler &

Singh (1999, p. 4), “whatever the performance of a single processor at a

given time, higher performance can, in principle, be achieved by utilizing

many such processors.” Evolutionary psychologists reason similarly that a single calculating

mechanism with the same set of rules, meant to cover a multitude of tasks,

would have processed information slowly and led to many errors. To use

an example from Buss (1999), if our early ancestors had a generalized rule

like “have sexual intercourse with any partner you can,” then, in terms of

the ultimate goal of propagating genes, such a rule would be benefi cial

with respect to nonkin but would backfi re with respect to kin. A parallel

processing, modular kind of mind would fare much better because it would

have more specialized routines designed to handle a variety of situations.

In other words, an individual module that has emerged to handle only

one kind of problem likely will be able to handle that problem swiftly and

effi ciently because it has to handle only that particular kind of problem,

and no other one.

In essence, modularity minimizes errors and allows systems to perform

optimally. We should not underestimate the importance of this kind of

reasoning on the part of NEPers. A speedy response and the minimization

of error grant the system a competitive advantage—hence, the likelihood

of such specifi ed, parallel processing mechanisms being selected for in our

early hominin mental architecture (see Arp, 2008b). This comports with

the general evolutionary principle of economy, recognized by every evolutionist

since Darwin, namely, whatever trait gives an organism a competitive

advantage most likely will be naturally selected as fi t for that organism

in relation to an environment and likely will be passed on to that organism’s

progeny.

It appears that the parallel kind of processing has been selected for with

respect to at least some of our mental architecture. As I intimated above

and in the third chapter, there is ample evidence that the visual system is

considered as a suite of coordinated physiological or brain-process modules

engaged in the parallel processing of visual information. So too, Broca’s

area and Wernicke’s area in the human brain are engaged in the parallel

processing of grammar–usage and language comprehension, respectively.

Further, the face-recognition area in the IT cortex works in parallel with

other areas of the IT cortex, and other areas of the brain, to help someone

distinguish faces from other objects.

Numerous studies on infants, children, and adults seem to confi rm the

fact that people have innate mental modules seemingly designated for

specifi c tasks. For example, Chomsky (1964) has argued convincingly that

there must be some innate capacity for language, since young children

from any culture can pick up language easily, as well as being able to learn

any language (see also Jackendoff, 1987, 1992, 1994). This makes sense from the NEP perspective, since Pleistocene hominins formed social groups

and eventually communicated with one another through language during

that time (Aiello, 1996; Aiello & Dunbar, 1993). Spelke (1991) has demonstrated

that children as young as two years old have an apparent innate

understanding of physical properties of objects like solidity, gravity, and

inertia (see also Pinker, 1994). This also makes sense from the NEP perspective,

since Pleistocene hominins constructed and handled a variety of tools

in a variety of ways during that time. Palmer & Palmer (2002) demonstrate

how people have mental modules attuned to certain fears, detection of

cheaters, empathizing, and spatial reasoning (cf. Adolphs, Tranel, Damasio,

& Damasio, 1995; Nesse & Abelson, 1995). Gardner’s (1993) list of multiple

intelligences of the mind is accepted by so many psychologists and educators

that it forms the basis for the curricula of many primary and middle

schools (Gardner, 1999). Pinker’s (2002) list includes an intuitive knowledge

of physics, biology, engineering, psychology, spatial sense, number

sense, probability, economics, logic, and language. Kandel et al. (2000, p.

412) tell us simply, “A newborn’s mind is not blank.” According to NEPers,

all of these capacities—language, intuitive physics, automatic knee-jerk

responses to snakes and spiders, detection of cheaters, and so forth—most

likely were solidifi ed in our species’ psyche during the Pleistocene epoch.

According to advocates of NEP, the adaptive problems in the Pleistocene

environments occasioned the emergence of psychological modules designed

to handle the various and sundry problems of such environments. Several

basic components of our present-day psyches were solidifi ed back then

and, in the words of Cosmides & Tooby (1987, p. 34), “the complex architecture

of the human psyche can be expected to have assumed approximately

modern form during the Pleistocene . . . and to have undergone

only minor modifi cation since then.” We must remember that the claim

NEPers make regarding the solidifi cation of our mental architecture occurs

over a period of many years, since the Pleistocene epoch spans approximately

2 mya to 10,000 ya. There are more than 65,000 generations of one

family line of one population alone that lived during that time! Thus, given

this vast amount of time, it is not wholly implausible and, in fact, it is

very possible that our human psyche was formed during this time period.

In other words, given what we know about laws of probabilities in relation

to genetic variability and natural selection, prima facie the hypothesis is

not that outlandish.

Evolutionary psychologists speak of these modules as domains of specifi

city. What this means is that a module handles only one kind of adaptive

problem to the exclusion of others. Modules are encapsulated in this sense and do not share information with one another (Fodor, 1983, 1985). For

example, my cheater-detection module evolved under a certain set of circumstances

and has no direct connection to my fear-of-snakes module,

which evolved under a different set of circumstances. Like the various

kinds of tools in a Swiss Army knife, the various mental modules are supposed

to solve the various problems that arise in circumstances; however,

they do so to the exclusion of each other. The scissors of the Swiss Army

knife are not directly functionally related to the Phillip’s-head screwdriver,

which is not directly functionally related to the toothpick, and so forth.

This kind of encapsulation works best for environments where the

responses need to be quick and routine—such developments enabled these

organisms to respond effi ciently and effectively in their environments.

This being the case, the modules could perform quite well as long as the

environments remained relatively unchanged and typical. In fact, most of

this modular processing in mammals occurs at the unconscious level. It is

arguable that since this processing occurs at the unconscious level and,

further, since information can become memorized, mammals quickly are

able to respond to the pressures associated with fi ghting, fl eeing, eating,

mating, and so forth (Fodor, 1983, 1985; Hermer & Spelke, 1996; Cosmides

& Tooby, 1992; Shettleworth, 2000).

However, there seems to be a fundamental limitation in the NEPers’

reasoning, especially if the environment in which the domain-specifi c

module has been selected is supposed to have remained fairly stable. Cosmides

& Tooby (1994) note that these domain-specifi c modules have

evolved “for solving long-enduring [my italics] adaptive problems,” and

Hirschfeld & Gelman (1994, p. 21) characterize a module as a “stable

response to a set of recurring [my italics] and complex problems faced by

the organism.” Now Daly & Wilson (1999), Foley (1995), and Boyd & Silk

(1997) have shown that the Pleistocene did not consist of a single hunter–

gatherer type of environment but was actually a constellation of environments

that presented a host of challenges to the early hominin mind. Thus,

the fi rst problem for advocates of NEP has to do with the possibility of the

environment in which a particular module evolved being stable enough for

the module to have evolved. In other words, Daly et al.’s criticism of NEP

is that the environments in which our Pleistocene ancestors lived were too

varied and too erratic for the Swiss Army knife blades to be solidifi ed in

our genetic makeup.

We need only refl ect on the kind of circumstances one experiences

through the course of the proverbial bad day to see that one is bombarded

constantly with novel pieces of information. Throughout this bad day, one fi nds oneself in a repertoire of novel mini-environments, from waking up

to discover your toilet has overfl owed, to locating a bus schedule and then

riding a bus to work because you discovered your car broke down, to

searching for some twine in your cubicle at work so that you can hold your

pants up because in the rush out the door you forgot to wear a belt, and

so forth. All of this was atypical of your normal day.

In fact, as Crick & Koch (1990) and Kosslyn & Koenig (1995) note, novel

scenes are being presented to the visual system regularly. Everyday objects

and circumstances in our visual fi eld constantly are obscured, occluded, or

observed from different angles: “There are an almost infi nite number of

possible, different objects that we are capable of seeing. . . . The combinatorial

possibilities for representing so many objects at all different values of

depth, motion, color, orientation and spatial location are simply too staggering”

(Crick & Koch, 1990, p. 268). Today, I made a turn onto Main

Street and looked down the block to see Mr. Jones’ oak tree bathed in

sunlight; yesterday, I made the same turn in the pouring rain but had to

squint through the rain-soaked glass of my car to see the same tree darkened

by the storm. Also, as I move around my cat teasing her with a ball

of string, her body contorts into a variety of positions, and I experience

her shape anew with every different angle.

Even if we tempered Daly et al.’s claims regarding the multitude of environments

faced by our ancestors and grant that the Pleistocene consisted

of a more unitary Stone Age, hunter–gatherer, life-out-on-the-savanna kind

of existence—like the one Cosmides & Tooby would have us believe—then

we have the further problem of the possibility of some stable and routine

module being able to handle the unstable and nonroutine events occurring

in some environment. When routine perceptual and knowledge structures

fail, or when atypical environments present themselves, it is then that we

need to be innovative in dealing with this novelty. If mental modules are

encapsulated and are designed to perform certain routine functions, how

can this modularity account for novel circumstances? The problem for NEP

can be phrased in the form of a disjunction: either (1) the environment

was not stable enough to occasion the emergence of domain-specifi c

modules, as is part of the thrust of Daly et al.’s criticism, or (2) the environment

was stable, allowing for domain-specifi c mental modules to

emerge, but then the environment changed, making it such that the

modules specifi ed for the old environment would no longer be helpful in

the new environment.

Now, imagine the Pleistocene epoch. The climate shift in Africa from

jungle life to desert/savanna life forced early hominins to come out of the trees and survive in totally new environments. Given a fortuitous genetic

code, some hominins readapted to new African landscapes, and some

migrated to new places like Europe and Asia, but most died out. This environmental

shift had a dramatic effect on modularity, since now the specifi c

content of the information from the environment in a particular module

was no longer relevant. The information that was formerly suited for life in a

certain environment could no longer be relied upon in the new environmental

niches. Appeal to modularity alone would have led to certain death and

extinction of many mammalian species. In fact, countless thousands of

mammalian species did become extinct, as fossil data indicate (Novacek,

2002; Dingus, 1990). Elsewhere, I have called mental disruptions of this

nature cognitive dissonance (see Arp, 2004a, 2004b).

The successful progression from typical kinds of environments to other

atypical kinds of environments would have required some other kind of

mental capacity to emerge in our hominin ancestors that creatively could

handle the new environments. Mere mental associations, or trial-and-error

kinds of mental activities, would not be enough, since the environments

in which these hominins found themselves were wholly new, and there

would have been no precedent by or through which one could form

mental associations utilizing past information. Mental associations deal

with the familiar. What is one to do when encountering the wholly unfamiliar?

Although important, modules have their limitations, since they do

all of their associative work in routine environments. What happens if an

environment radically changes, making the information that a particular

module characteristically selects in a familiar environment no longer relevant

in a wholly new environment? A radical readaptation and readjustment

would be needed, one that transcends the limitations of the

routine.

Recall that nonroutine creative problem solving involves fi nding a solution

to a problem that has not been solved previously. The invention of a

new tool would be an example of nonroutine creative problem solving

because the inventor did not possess a way to solve the problem already.

This totally new environment would require that we be creative or innovative

in order to survive. But how is it that we can be creative? The signifi cant

question becomes, then, this: How is it that humans evolved the ability to

engage in forms of nonroutine creative problem solving, especially given

either that the Pleistocene environment in which early hominins existed

was really a constellation of ever-changing environments (Daly et al.’s criticism)

or that any one environment was fi lled with a myriad of nonroutine problems

that seem only to be able to be handled creatively?