Re: evaluation of brain size/IQ refs

Fri, 4 Nov 1994 14:47:39 CST

>From Kim Hill: warning, this file is about 15 pages long. anyone
who wants a hard copy instead can send me their mailing adress.

Because of the long and detailed discussion on the HBES net
about alleged brain and life history differences across the races,
I asked to see the evidence that such differences really do exist
(and therefore demand an explanation). Although I find Rushton's
use of r and K theory logically incoherent and inconsistent with
what is known about life history evolution, I am interested in
knowing whether the patterns that he cites (but not the
explantions) can be accepted.

Rushton cited several papers on brain size and IQ differences
and I have now had a chance to carefully read them. I have
volunteered to give my assessment on the net since I had mentioned
that I thought people were being too quick to accept the fact that
such patterns exist and too anxious to 'explain' such patterns
before carefully assessing whether there is anything to explain.
Below is my assessment. Before I give a detailed critique I will
present a summary statement for those who wish to be spared the
details (and for those like monique who dont want to read more
than 1 page). The simple answer is that the patterns that Rushton
is trying to explain (racially based rather than environmentally
based differences in brain size and IQ) have NOT been adequately
established. For those of you who are familiar with multivariate
modeling, I can summarize the difficulties by noting that a
variety of factors are known or suspected to correlate with brain
size and IQ. Since these factors have not been adequately
controlled in order to isolate the effect (statistically) of race,
we cannot at this point determine whether observed racial
differences are due to racial covariation with environmental
factors (both immediate and developmental), or truly heritable
phenotypic variation that is patterned accross racial strata.
Furthermore, it is unclear whether differences in cranial volumes
across populations produce differences in IQ scores.
Several other electronic networks have also discussed this
issue in recent years including at least two in evolutionary
neurobiology. (The address to obtain access to a number of
authorities on
brain size issues is This is the
Neuroanthropology Network , founded about 4-5 years ago by Charles
Laughlin at Carleton University. This group has now merged with a
sociological one-- which puts out a Social Neuroscience Bulletin.
The address above connects you to the present editor-- TenHouten
(a sociologist at UCLA).). Since these specialists have generally
rejected Rushton's claims, it seems redundant for us to get
involved, but since our theoretical perspective is being invoked
we need to be somewhat informed on this issue. I will warn the
readers now that this response is about 15 pages long. In this
discussion I will refer to people descendent from the three major
geographical areas as Afro, Euro, and Asian. I will not get into
the details of whether such racial categories are real or useful.

BRAIN SIZE- Rushton submitted 6 references to support the
conclusion that brain size is largest among Asians, intermediate
among Euros and smallest among Afros. I will summarize the main
results of each and then discuss the overall pattern.

Beals, Smith and Dodd 1984. This paper examines brain size
(generally measured as endocranial volume in cc) across 122 ethnic
groups around the world and some fossil hominids. The sample size
of total individuals is unknown (but large) because most of the
data come from Hrdlicka's 1924-42 catalogue of human crania and
some data are found only in population summary form. Single
populations are represented by as few as 4 individuals or as many
as ??(I was unable to check the source of some data, especially
european and asian data that do not come from Hrdlicka's
catalogue). Many individuals are of unknown age, and a
reasonable number are of unknown (but guessed at) sex. Also,
single individual skulls cannot be matched with body size, etc in
this data set, so population means for all variables are used
throughout. In essence there are 122 data points, each
represented a population mean ( (male volume + female volume)/2 )
for relevant variables. The paper's main conclusion is that a
good deal of variation in brain size among modern humans (and
hominids) can be explained as adaptation to climatological
factors. For example, the authors show that 83% of the variance
in cranial volume across 6 major geographic regions (North
America, Asia, Europe, S.America, Oceania, Africa) is
statistically explained by what percent of the samples from each
region live in areas that experience frost (P.306). They also
show that in a world combined sample, the mean cranial volume from
temperate or cold climates is 1385 cc (both sexes) whereas from
hot or warm climates it is 1297 cc (table 5). They further show
that cranial volume correlates well with other body features know
to be affected by climate (eg. body size, surface area, nasal
index, etc) but that cranial volume correlates MORE STRONGLY with
climate than do other morphological features (table 6). They also
show that the relation between climate and cranial volume is
stronger in the Old World than in the New World (as would be
expected given evolutionary time depth). They interpret data
within the context of a model of adaptation in which certain head
shapes and sizes are favored in cold climates for thermoregulatory
purposes. More specifically rounded heads (and flatter faces) are
believed to be adaptive in cold climates whereas longer heads are
supposedly adapted to hotter climates. Because of geometrical
considerations rounder heads end up slightly larger according to
the authors. Short robust bodies with large heads (eg. the
neanderthals and current arctic peoples) are suggested to be
particularly indicative of cold climate. Importantly, the same
trends found in the worldwide sample are found within different
geographical regions (table 7).
The authors in this paper also point out that brain size in
mammals varies as a function of body size by a constant allometric
ratio, and that in their sample body weight could explain 39% of
the variance in cranial volume across groups (p.305). This paper
contains the geographical means widely cited by Rushton, namely
that the mean cranial volume for 26 Asian societies was 1380 cc,
the mean volume for 10 European societies was 1362, and the mean
for 10 African societies was 1276. Notably, the African sample
contained 5 groups that are characterized by exceptionally small
body size (2 pygmy groups and 2 bushman groups). Indeed the Akka
pygmies (representing 1/10 of the whole African sample) had the
smallest cranial volume ever found in extant humans (1085 cc).
Tremendous variance is found within "racial groups". For example
the Aleut Eskimos have a mean cranial volume of 1518cc whereas the
Choctaw (who are the same size) have a mean cranial volume of
1292cc. The authors believe that such observations strongly
support their climatological model (more on this below). The
authors (perhaps anticipating Rushton's later useage of their
data) make the following statement: "The implication is that any
effort to attribute racial or cognitive significance to brain size
is probably meaningless unless climate is controlled. For
example, the endocranial volumes of Europeans and Africans differ
little from what one would expect given the difference in their
respective winters."

Ricklan and Tobias 1986. This paper reports a mean endocranial
volume for 50 male and 50 female Zulu of South Africa. The sample
came from cadavers which the authors point out come mainly from
the lower socioeconomic strata of society. Male cranial volume
was 1373.3cc and female volume was 1251.2. This gives a combined
sex mean of 1312.3cc. The main point of the paper is to emphasize
the low level of sexual dimorphism in this population. The
authors review a good deal of material on cranial volume in
African populations and report means for 1) American Negros; 2)
other African Negros; and 3) South African Negros. Their data
(p.289) show largest cranial volumes for South Africans Negros
(1310cc mixed sex mean), followed by other African Negros (1285 cc
mixed sex mean), and then American Negros (1282cc mixed sex mean).
The authors also present some data that suggest a negative secular
trend in cranial capacity among the Zulus during the past 100
years (ie. volumes may have been decreasing in time). It is known
that Zulus have shown a negative secular trend in stature during
the past century and also that stature correlates with cranial
volume. Thus a negative secular trend in cranial volume might be
expected. Cranial volume measurements of the Zulu in the late
19th century show males with a mean of 1450cc (or a 6% larger
volume) at that time.

Shibata 1936 This paper presents results from autopsies on 153
adult Koreans done in the 1930s. Brain weights (rather than
cranial volumes) are reported and compared similar data from
studies done in Japan. Mean brain weights are 1369.5g for males
and 1276.9g for females. The sex combined mean is thus 1323g.
Table 1 shows that these weights are very similar to weights
reported in a large number of Japaneese autopsies. The paper also
shows that brain weight peaks in early adulthood and declines
through the lifespan (or else there is a secular trend in the
data) and that taller individuals have larger brains (tables 2-
4). Finally brains are RELATIVELY larger in young and smaller
individuals (an allometric effect).

Takahashi and Suzuki 1961. This paper reports brain weights for
470 adult Japanese. Mean weight for males is 1397 g and for
females is 1229 g for a sex combined mean of 1313g. The paper
also shows that brain weight decreases with age in adulthood,
especially after age 50. Finally it reports regional differences
in brain weight in Japan. The regional mixed sex means vary from
1299g to 1339g and some differences are statistically significant.
The authors call for more research into regional variation in
brain size in Japan.

Ho et al 1980 part I. This paper reports brain weights from
autopsies on 1261 Americans in Cleveland, Ohio (in the 1970s).
Races are categorized as White or Black in the standard American
fashion (ie. by self report or phenotypic inspection). The data
show that brain weight peaks at age 25 and declines slowly (2 g
per year) until age 80 at which point it declines very rapidly.