David S. Webster (1999) How Does Brain Size Matter?. Psycoloquy: 10(030) Brain Expertise (5)

Volume: 10 (next, prev) Issue: 030 (next, prev) Article: 5 (next prev first) Alternate versions: ASCII Summary
PSYCOLOQUY (ISSN 1055-0143) is sponsored by the American Psychological Association (APA).
Psycoloquy 10(030): How Does Brain Size Matter?

Commentary on Skoyles on Brain-Expertise

David S. Webster
Department of Psychology
University of Durham
Science Laboratories
South Road
Durham DH1 3LE

Ken Richardson
Centre for Human Development and Learning
School of Education
Open University
Milton Keynes

D.S.Webster@durham.ac.uk & D.S. Webster@btinternet.com K.Richardson@open.ac.uk


In this commentary on Skoyles we argue that the distinction between expertise and IQ is not of the right kind to explain the brain-size paradox. The evidence suggests that the capacity for expertise has (a) a longer history than Skoyles would allow; (b) is qualitatively related to increased brain corticalization and restructuring with the emergence of H. sapiens; (c) is developed against new demands made by a changing ecology as a consequence of human global expansion, and (d) with H. sapiens sapiens (i.e., Cro-Magnon and after), attains both qualitative and quantitative independence from gross brain size.


brain size, brain imaging, evolution, expertise, hemispherectomy, Homo erectus, individual differences, intelligence, IQ, language, microcephaly, MRI volumetrics, psychometrics.
1. Skoyles (1999) asks why modern humans have larger brains than members of our ancestral species, since a propensity for large brains may cause our species problems; in particular, making child birth more hazardous than it need be. The answer Skoyles offers is that large brains offer a pay off in the form of an increased capacity for expertise. In contradistinction to expertise, Skoyles offers a variety of evidence that, taken together, suggests that IQ is independent of brain size. Skoyles therefore contends that, (a) "The capacity for expertise correlates absolutely with brain size: to have more capacity for expertise, a brain must be bigger." (par. 3 i. emphasis added); and, (b) "Expertise plays a critical role in the survival of simple hunter-gatherers" (par. 3 ii).

2. Skoyles poses his question (why large brains?) in light of the unargued assumption that the near 50% increase in brain size between H. erectus and H. sapiens was driven by cognitive demands. He poses another question, namely, "What aspect of early cognition drove this increase in size?" (par. 1).

3. What is most problematic in Skoyles's thesis is the ambiguity with regard to what is taken as primitive expertise. The examples of expertise that Skoyles turns to are, with one exception, very modern and non-survival related at the level of the species (gamblers, chess players), and so it is difficult to relate them to the hominid situation 1.6 million years ago (mya). Perhaps more pertinent but still very problematic is Skoyles's use of the hunting skills of !Kung as an example of primitive expertise. The !Kung are a fully modern people living under the extreme pressure of modernisation around them, constrained to marginal lands and facing probably catastrophic climatic change. To use the !Kung in this way is questionable at the very least.

4. Another problem with Skoyles's thesis is the unnecessary crudity of using gross measures such as brain weight or volume as a comparative tool. A more pertinent measure for cross-species comparison is the ratio of corticalization to body weight expressed as the Encephalization Quotient (EQ). Moreover, the enlargement of the hominid brain had two distinct phases, with the first phase split into two sub-phases. The first sub-phase is the Australopithecus to H. habilis transition (roughly 3.5 million to 2 million years ago), and shows EQ rising from 3 to 4.5, with a concomitant increase in both brain and body size. Thus corticalization increases absolutely against a general increase in brain size: but crucially, brain size in hominids is a close correlate of body size (cf. Walker & Shipman 1996). The second sub-phase is the H. habilis to H. erectus transition (2 million to 1.6 million years ago), with Homo erectus attaining an EQ of about 4.8 but with an 46% increase in brain size from 650cc to 950cc, together with a substantial increase in body size. Thus the Nariokotome boy (H. erectus) was as tall as modern humans; ranging between five feet two inches to six feet one inch. The difference in the two sub-phases is that in the first sub-phase, corticalization increases absolutely against increasing brain size, whereas in the second sub-phase, the increase in corticalization is cancelled out by erectus having a much larger body.

5. The second full phase in brain development comes with the H. erectus to H. sapiens transition. Here there is a 31% average increase in brain size for H. erectus (950cc-1371cc), and a dramatic increase in EQ from an average of 4.8 to 7 for H. sapiens (range 5 -10.3); all without the attendant increase in average body size. The brain, however, underwent a change of form with cortical folding and internal restructuring; there was an increase in the amount of association cortex, and repositioning and reduction of other areas such as the striate cortex (cf. Armstrong et al, 1993). A further point is worth noting: in the primate line, large-bodied Gorillas have relatively smaller brains and lower EQ than that of the much smaller pygmy chimpanzees. The H. habilis to H. erectus transition marks a point of strong selection for increasing relative brain size, since, the EQ of erectus holds steady against a leap in body size. On the primate pattern the EQ of erectus would be expected to decrease from that of habilis (cf. Walker and Shipman, 1996).

6. Thus over a period of 2 million years the hominid brain increased in size in close accordance with an increase in body size. But more importantly, the brain increased in its degree of corticalization, and underwent a restructuring. Behind the change in body size, and, just as importantly, body shape, were new demands regarding thermoregulation. Coppens (1994) points out that the hominid/chimpanzee split (dated to 3.5 mya) has at least one of its origins in the tectonic shifts that created the Rift Valley and a line of peaks forming the western rim of the valley (dated to 8 mya). This contributed to a new drier climate on the eastern side of the valley, in which humid forest was replaced with open savannah. With this much drier climate came new demands for thermoregulation,to be met by the loss of body hair, increases in body size and changes in body shape (Nariokotome boy was very slim and tall).

7. Skoyles, in supporting his more-brain/more-expertise thesis, draws our attention to an important -- perhaps a key -- event in hominid evolution; that of secondary altriciality (this term is not used by Skoyles). Secondary altriciality involves the truncation of the gestation period and the birthing of "foetal" infants with the extension of foetal growth rates in to the first year of life. With secondary altriciality, hominid brains could undergo a significant increase in size (hence Skoyles's interest in it), but most importantly, this particular developmental pattern places nurturing practices at the centre of hominid survival strategies, since the hominid infant is completely helpless at birth and requires many years of close support and attention. The demands of child rearing may be seen to have had a profound effect on the total conduct of the early hominid groups, and formed the basis of human social relations.

8. Cameron (1993: 404) suggests (based on a cladistic analysis of Pliocene and Protochimpanzee behavioural morphotypes) that ancestral hominid behavioural morphotypes had the following form, ."..males patrolling alone; strong bonding among all classes; dependant upon concentrated food patches; capture of prey dependant on complex organisation; usually removes food from its source location and most frequently taken back to home base, where it is processed; and nests are almost always shared and used over extended period of times." This behavioural morphotype dates back 5 million years, but already evinces a fair degree of division of labour and a degree of group cohesion that would allow secondary altriciality to emerge. Secondary altriciality, moreover, is dated to the advent of H. habilis some 2 mya (cf. Walker and Shipman, 1996), and is intensified by the selection for tall and slim bodies in H. erectus. Secondary altriciality also clearly worked to differentiate the ancestral behavioural morphotype; intensifying behaviours relating to social actions and nurture, while loosening others such as dependence on concentrated food patches (herein lies the whole historical trajectory and diversity of human culture).

9. The preliminary picture which emerges is one of climatic change initially selecting for tall, slim and hairless hominids with increased corticalization, through successful group strategies and close bonding between individuals.

10. Skoyles is, to an extent, correct in linking secondary altriciality, larger brains and human expertise. The issue, however, is not the fact (if it is a fact) that "individuals with the brains that had the greatest capacity to acquire expertise would survive more successfully" (par. 37), but rather, that those larger corticalized and restructured brains are at the service of hominid groups. Hominid groups, structured and driven by the demands of secondary altriciality into socially differentiated but yet cohesive units, could maintain the solidarity and coherence required in the face of rapidly changing resources and environment. Here we are thinking about the adversity arising from the human species spreading throughout the world (e.g., "Out of Africa" 1 & 2). That is, social cohesion with increasing divisions of labour afforded the active and rapid adaptation to novel environments, and reflexively allowed for further human expansions.

11. As a model for archaic human expertise Skoyles offers the example of the !Kung expertise in hunting. Skoyles's use of this example reveals a certain blindness to the social and institutional dimensions of expertise. However, generalising, and granting Skoyles's homology between the !Kung and archaic humans, the development of perceptual learning -- the ability gained within a social group to make further differentiations within a structured perceptual array -- seems a key feature. Here, it may be noted that the first people to be named as "experts" were witnesses in French courts of law (19th century), who were called to identify forgeries of hand-written scripts; a good example of perceptual learning mobilised through a social role (i.e., a witness) within an institutionalised setting.

12. As Lee (1979, 1984: 47) cited by Skoyles (par., 36), explains, "the !Kung hunter can deduce many kinds of information about an animal...the species...is identified by the shape of hoof print and by dung or scat...and any 12 year old can accurately reproduce in the sand the prints of a dozen species. The size or age of an animal correlates directly with the size of its print. An old or infirm animal may be distinguished by a halting gait or uneven stride length. Evidence of crippling is eagerly sought and is discerned when one hoof print is deeper that others."

13. Understanding an animal's daily habits along with the type of infill in the hoof print helps to determine how long it has been since it passed. Thus, what the !Kung rely upon is the nesting of covariant information in the form of the variable values of size, depth, infill of hoof print, zig-zaging hoof prints, and the relative concentration of hoof prints. Elsewhere (e.g., Richardson & Webster, 1996a), we describe this type of nesting of variable values as hyperstructural, resulting in the enhancement of predictive judgements in complex environments. To take another of Skoyles's examples of expertise, that of expert gamblers (par. 35), the significance of the findings from racetrack betting experts is that values of up to eleven variables were not combined additively, but integrated across as many as seven levels of interaction. This means that the covariation between two variables (e.g., the probability of winning and jockey weight) is conditioned by the values of another variable (e.g., the state of the course), and this conditioning is itself conditioned by values of yet another variable (and so on, up to seven variables "deep").

14. With regard to implicit learning, studies have shown that verbal descriptions of relations acquired are confounded, because those relations are deeper than linear bivariate correlations. Thus Reber (1989: 219) notes that implicit learning produces a knowledge base that is "abstract and representative of the structure of the environment," and is induced from "the complex covariations among events that characterize the environment."

15. Again, studies of complex cognition in simulated "real-life" situations, such as factory-production, have shown how subjects acquire multivariate, interactive, relations of considerable depth (for review see Funke 1991). And Sylvia Scribner's study of dairy workers (in Scribner, 1997) showed how they were able to deal with demanding situations in which direct associations became conditioned by novel variables. The game of chess, as Grand Masters Bronstein and Smolyan (1982: 24) understand it, should "be a combination or a manoeuvre, a trap or a study-like ending, the complex logic of a plan or the geometrical harmony of co-ordination -- all of these are distinctive aesthetic "invariants." What grand masters bring to the game of chess is quality of play in terms of "correctness," "difficulty," "vivacity," "richness" and "logical unity." What an expert chess player "sees" from a glance at the board is not amorphous information chunks, but dynamic configurations engendering a variety of possible moves (Chase & Simon, 1973).

16. Piaget, of course, repeatedly stressed the importance of representing "deeper" coordinations as distinct from direct ("perceptual") relations as the basis of cognitive development. In one study (Richardson 1992), in which eight and eleven year olds were asked to predict the speed of bikes from information about rider-size, bike-size and quality of road-surface, it was shown that what the older children seemed to appreciate, relative to the younger ones, was the interaction effects (in which, for example, the covariation between speed and bike size is conditioned by rider size, which in turn may be conditioned by road surface). In another study (Richardson & Webster 1996a) it was shown how children's and other people's recovery of an object "image" from very sparse point light stimuli -- some consisting of only four points -- was strongly related to the "depth" of information in the stimuli, as revealed by log-linear analysis of interactions and mutual information measures. Finally, a study of contextual effects on standard analogical reasoning tasks (Richardson & Webster 1996b)revealed the importance of understanding the deeper coordinations between elements and their features.

17. The root of expertise (we might as well call it "intelligence" - see Richardson, 1999) we take to be a utilization of the deeper covariation or coordinate structure in a domain, such as to predict the consequences of action upon it. This involves the utilizing ever-deeper hyperstructural information as our species has evolved and has been able to move into increasingly complex environments (in humans, social ones), and explains how bigger corticalized and restructured brains are brought into play.

18. We further suggest that even IQ reflects in-depth knowledge and reasoning (even simply understanding instructions for taking the test requires that), but only of those forms and formats specific to a narrow culture. This is most obviously the case with the verbal items which comprise the content of most IQ tests, questions such as "Who was Genghis Khan?," and "What is the boiling point of water?." These are obviously associated with school learning (the main criterion of item selection). But the same arguments apply to the non-verbal items which are often considered to be closer to a "pure" measure of "g" (the hypothetical intelligence "power").

19. Take for example, the Ravens Matrices. It is usually presented as a "culture-fair" test, because it is free of verbal content. Yet it isn't difficult to show that Ravens items are also based on in-depth knowledge, albeit knowledge culturally-situated. First of all, the items are presented as black and white figures, flat on the paper, with the "pick-up" of information arranged from top-left to bottom right. These are culture-specific tools for handling information, more prominent in some groups than others.

20. The same applies to the "deeper" relations embedded in the items themselves. These require the induction of the logical "rule" implicit in the arrangement and transformation of the visual patterns, which is such that covariation across columns is embedded in (conditioned by) the covariation down rows. These rules are emphatically culture-loaded, in the sense that they reflect further information-handling "tools" for storing and extracting information from text, for example, from tables of figures, accounts, or timetables, all of which are more prominent in some cultures and subcultures than others.

21. In other words, items like the Ravens are also based on deep covariation relations, but these are such as to make them more, not less, culturally-steeped than any other kind of "intelligence" testing item: "the most systematically acculturated tests," as Keating and Maclean put it. So it is not the case that IQ tests do not require "in-depth" knowledge and reasoning, but that they require forms that are highly culturally specific.

22. In conclusion, we do not think that Skoyles's distinction between IQ and expertise is of the right kind to explain the brain size-intelligence paradox. The evidence suggests that the capacity for expertise -- understood as a capacity for perceptual learning and the development of hyperstructures within an intensely social context -- has (a) a longer history than Skoyles would allow; (b) is qualitatively related to increased brain corticalization and restructuring with the emergence of H. sapiens; (c) is developed against new demands made by a changing ecology as a consequence of human global expansion, and (d) with H. sapiens sapiens (i.e., Cro-Magnon and after), attains both qualitative and quantitative independence from gross brain size. We suspect that this last point will need to be resolved at the level of other variables: for example, within the context of functionally equivalent connectivity and architecture there may be wide variation in neuron-size and density, glial-size and density, myelination, and vascularization. Assessing such possibilities will, however, require methods other than gross scanning and volumetric analysis.


Armstrong, E., Zilles, K., & Schleicher, A. (1993). "Cortical folding and the evolution of the human brain." Journal of Human Evolution 25, 387-392.

Bronstein, D., & Smolyan, G. (1982). Chess in the Eighties. London: Pergamon Press.

Cameron, D. W. (1993). "The Pliocene Hominid and Protochimpanzee Behavioral Morphotype." Journal of Anthropological Archaeology 12, 384-414.

Chase, W.G. & Simon, H.A. (1973). "Perception in chess." Cognitive Psychology, 4, 55-81.

Coppens, Y. (1994) "East Side Story: The Origins of Humankind." Scientific American Vol 270 (5), May 1994.

Funke, J. (1991). "Solving complex problems: exploration and comtrol of complex systems." In R.J. Sternberg & P.A. Frensch (eds.) Complex problem solving: principles and mechamisms. Hove: Erlbaum.

Gibson, J. J., & Gibson, E. J. (1955). "Perceptual Learning: Differentiation of Enrichment." Psychological Review 62, (1), 32-41.

Hayden, B. (1993). "The cultural capacity of Neanderthals: a review and re-evaluation." Journal of Human Evolution. 24, 113-146.

Keating D. P. and MacLean, D. J. (1987) "Cognitive processing, cognitive ability, and development: a reconsideration." In P. A. Vernon (ed.), Speed of Information Processing and Intelligence. New York. Ablex.

Larkin, J.H. (1983). "The role of problem-representation in physics." In D.Gentner & A.L. Stevens (eds.), Mental Models. Hillsdale, N.J: Erlbaum.

Lee, R. B. (1979). The Dobe !Kung. New York: Holt, Rinehart and Winston.

Reber, A.S. (1989). "Implicit learning and tacit knowledge." Journal of Experimental Psychology: General, 118, 219-235.

Richardson, K. (forthcoming). "Hyperstructures in brain and cognition." Psycoloquy.

Richardson, K. (1999) "The making of intelligence." London, Weidenfeld & Nicolson.

Richardson, K. (1992) "Covariation analysis of knowledge representation: some developmental studies." Journal of Experimental Child Psychology, 53: 129-150.

Richardson, K. & Webster, D.S. (1996a) "Recognition of objects from point-light stimuli: evidence for covariation hierarchies in conceptual representation." British Journal of Psychology, 87: 567-591

Richardson, K. & Webster, D.S. (1996b). "Analogical reasoning and the nature of context." Britsh Journal of Educational Psychology," 65, 23-32.

Scribner, S. (1997b) "Knowledge at work." In E. Tobach, L.M.W. Martin, R.J. Falmagne, A.S. Scribner & M.B. Parlee (eds.) Mind & social practice: Selected writings of Sylvia Scribner. Cambridge: Cambridge University Press.

Skoyles, J.R. (1999). Human evolution expanded brains to increase expertise capacity, not IQ. PSYCOLOQUY 10(002). ftp://ftp.princeton.edu/pub/harnad/Psycoloquy/1999.volume.10/ psyc.99.10.002.brain-expertise.1.skoyles http://www.cogsci.soton.ac.uk/cgi/psyc/newpsy?10.002

Walker, A., & Shipman, P. (1996). The Wisdom of the Bones. In Search of Human Origins. London: Phoenix.

Volume: 10 (next, prev) Issue: 030 (next, prev) Article: 5 (next prev first) Alternate versions: ASCII Summary