and vary through such simple, in de pen dent associations, as we shall see
in chapters 4 and 5.
Imagine each of the thousand legs of a millipede contributing to
movement in de pen dently of what every other leg is doing, or with what
is happening in the environment. Th ey would be useless without the
evolution of some sort of coordinating function. Yet behavioral ge ne-
ticists’ computations suggest that most of our individual diff erences in
mental functions are due to hundreds or thousands of genes having
in de pen dent eff ects in that sense.
Alternatively, of course, there might be something badly wrong with
their computations. In their article in the journal Social Science Research
in 2015, John Daw and colleagues argue that “ these conclusions are
erroneous due to large violations of the additivity assumption under-
lying behavioral ge ne tics methods— that sources of ge ne tic and...
environmental variance are in de pen dent and non- interactive.”^8
Th is should hardly be surprising. Just as coordination guides func-
tion, and millipedes would not have evolved without it, so genes do not
act alone but have evolved, and are utilized, through elaborate coordina-
tion functions. Th e eff ect of one or a few wayward legs behaving in de-
pen dently would not simply produce marginal increments or decrements
of per for mance, resulting in the imagined bell- shaped curve. It would
disrupt the whole system and result in an un co or di nated shambles.
In chapters 4 and 5, I show that the evolution of complex functions
involved the utilization of genes, and much else, in coordinated dynami-
cal systems. When responding to ever- changing environments, there can
be no such thing as gene in de pen dence for complex traits. Th e product of
any gene automatically becomes the environment of other genes. Th e
wider environment gets changed in the pro cess, to recruit further genes,
and so on. Some products of genes even limit their own production. In
that way, genes even create their own environments. So variation is
created from such interactions. As in static photos of any team or
cooperative group, the team members may look in de pen dent, but their
real- life activity is the result of intense interactions among them.
Th e prob lem is that the statistical models employed cannot cope with
such interactions. Behavioral ge ne ticists may talk about interactions, but
only in the shallowest form (see fi gure 2.3). Instead, the interactive
sources of variance are simply interpreted as additive eff ects and called
This content downloaded from 139.184.14.159 on Tue, 17 Oct 2017 13:51:26 UTChttp://www.ebook3000.com