“heritability.” Th is is what has been shown in a recent study by Or Zuk
and colleagues, who called the result “phantom heritability.” Th ey
explained how “quantitative ge ne ticists have long known that ge ne tic
interactions can aff ect heritability calculations” but have paid little
attention to them. Th e phantom heritability grows steadily with increas-
ing numbers of interacting inputs, they show. In consequence, “current
estimates of missing heritability are not meaningful, because they ignore
ge ne tic interactions.”^9A. additive B. shallow interactive C. deep interactive1FIGURE 2.3
Eff ects of genes or gene products (red spheres) on a phenotype. (a) In de-
pen dent/additive model; (b) Shallow interaction with one product
modulating the eff ects of another; (c) Deep interactions involving mod-
ulations among numerous products, including (open circles) environ-
mental factors (gene- environment interactions).As we have seen, the results have misled generations of psychologists
and students, with serious consequences. For instance, even if a set of
data fi ts a statistical model, that does not make the model correct. And
the interactions automatically nullify attempts to predict phenotype
from genotype— for example, children’s potential from that of parents
or from a DNA “sample.”
Th e odd thing is that those who use such models to assess the herita-
bility of human potential tacitly know this. Th ey regularly acknowledge
that such interactions are likely, and even talk of gene- environment
interplay. At the very least, they may well acknowledge that genes do
not act of their own volition, but have to be turned on or off — thereby
implying that something else is doing the turning on and off.
This content downloaded from 139.184.14.159 on Tue, 17 Oct 2017 13:51:26 UTC