DSCN4364

AftTiGLES

184 EvoMJonary Anthropołogy

CO

CL

a

p

s.

O

c

m

9

s

10

& 3* J

h

'Goniła

1119

►o

Hylobatcs

0.1 1 10 Neocortex Ratio

Fłgure 3 Mean group size ptotted against neocoftex rabo for indhńdual genera. shown separatek for proswnian. simian, and hominoid prima tes. Prosi mian group size data. from Dunbar and Joffe,** inctude species for which neocortex ratio is estimated from total brain votume Anthropoad data are from Dunbar.** Simians: 1. h/tiopithecus; Z Papio; 3. Macaca: 4, Procoiobus,; 5, Saimrt 6. Eiytfwooebus 7, Cercopkhecus; 8, Lagothm: 9. Cebus; 10. Atetes: 11. Cercocebus12, NasaSs 13. CaUcebus: 14. Alouatta; 15, CaMmico; 16, CebueSa; 17. Saguinus; 18 A otul 19. Wiece. 20, CaScebus. Prosimians: a. Lemur: b. Warece.c. Fułemurd. Propithe-cus, e. Incti f. Mkmcebus; g, Ga&go; h. Hapalemun i, Avahfcj, Perodictus.

size of the cortical processing machin ery increases, at least relative to the opportunity cost of taking cortical neu rons away from other cognitiye pro-cesses.

It seems equaliy unlikely that the problem lies with a pure memory constraint, though memory capacity obvi ously musi impose some kind of upper limit on the number of relationships that an animal can have. There are three reasons for this clalm. First, In humans at least, memory for faces is an order of magnltude larger than the predicted cognitive group size: Humans are said to be able to altach names to around 2,000 faces but have a cognitive group size of only about

100

Simians

Apes

U

Prosimians

f

1

REFINING THE RELATIONSHIP

The social brain hypothesis implies that constraints on group size arise from the information-processing ca-pacity of the primate brain, and that the neooortex plays a major role in this. However. even this proposal is open to several imerpretations as to how the relationship is mediated. At least five possibilities can be usefully considered. The constraint on group size coułd be a result of the ability to recognize and interpret visual signals for identifying either irtdividuals or their behavior; limitacions on memory for faces; the ability to remember who has a relationship with whom (e g., all dyadic relałionships within the group as a whole); the ability to manipulate Information about a set of relation-ships; and the capacity to process emo-tional information, particularly with respect to recognizing and acting on cuęs to other anłmais’ emotional States. These are not all necessarily mutually exclusive. bul they do identlfy differ-ent points in the cognitive mechanism that might be the crucial Information' processing bottleneck.

Although visual mechanisms are likely to be import ant for social inter-action, and may well have been the

b

c

d

13 rj s

i C.

j CJ initial kick for the evolution of large brains in primates.¹⁰ it seems intrinsi-cally unlikely that the ultimate constraint lies in the mechanisms of the visual system itself.²⁸ Although there is a correlation between the relative size of the vtsual cortex and group size in anthropoid primates, the fit is much poorer, and the slope significantly shal-lower than that between the nonvisual neocortex and group size (z² = 0.31 vs i² = 0.61, respectively) (Fig. 4). Partial correlation analysis indicates that only the correlation for the nonvisual relationship remains significant when the other component is held constant²⁸ (though this is not true for prosimians²⁵). A morę important point is that the volume of the lateral geniculate nucieus. a major subcortlcal way sta-tion in visual processing, does not correlate with group size at all, indicat-ing that pattem recognitlon per se is unlikely to be the issue.²⁸ It may be of some significance that the absolute size of the vlsual cortex seems to reach an asymptotlc value in the great ape clade, whereas the nonvisual neocor-tex continues to increase in size. One interpretation of this is that visual processing does not necessarily con-tinue to improye indefinitely as the

The social brain hypothesis implies that constraints on group size arise from the information-processing capacity of the primate brain, and that the neocortex plays a major role in this. However, even this proposal is open to several interpretations as to how the relationship is mediated.

150. Second, there is no intrinsic rea-son to suppose that memory per se is the issue. The social brain hypothesis is about the ability to manipulate information, not sirnply to remember it. Third, and perhaps most significantly. memories appear to be stored niamly in the teniporal lobes,⁴² whereas re-cent PET scan studies implicate the prefrontal neocortex, notabły Brod-man area 8, as the area for social skills and, specifically, theory of inind.⁴³ Frith⁴⁴ has suggested that memories and representations for objects or events may lnvolve interactions between several levels of the neocortex depending on the kinds of operations