234 Wybo Houkes
This dichotomy is supplemented with a hypothesis about the underlying processes and
with an explicit explanatory goal. First, functional attributes are understood as fi tness-
conferring adaptations, subject to natural selection. Second, styles are assumed to be
“selectively neutral” (Dunnell 1978: 199) results of cultural transmission; the idea is that
toolmakers transmitted stylistic features to one another, occasionally varying on the theme,
producing old-fashioned or avant-garde arrowheads, and so forth.^12 Thus a combination
of two mechanisms, natural selection and cultural transmission, is used to reconstruct
archaeological traditions or lineages, and to explain them.^13
The primary goal of EA is, however, not classifi catory but explanatory. Specifi cally the
application of evolutionary theory is supposed to circumvent the need for intentionalist or
mentalist explanations of the archaeological record. In EA, artifact traditions are con-
structed and explained, not by reconstructing the intentions or mentality of the producers
of ancient artifacts, but by directly constructing artifact lineages and identifying styles and
cultural transmission. The reason is quite simple: “Individuals do make decisions, but
evidence for these decisions cannot be recovered by archaeologists” (Flannery 1967: 122);
and more recently, “Although we endorse the notion that new variants are intentionally
created at least some of the time... we have yet to determine how such intentions are to
be identifi ed analytically in the archaeological record” (Lyman and O’Brien 2000: 41).^14
Thus researchers in EA seek to combine evolutionary concepts and models with a com-
mitment to designer intentions: the former are regarded as potential explanatory replace-
ments of the latter.
13.3.2 Current State of Development
Very roughly, this is the conceptual framework of EA. To fulfi ll their promise to avoid
intentions in explaining the archaeological record, advocates of EA need to demonstrate
that it is actually possible to construct a lineage of phylogenetically related artifacts and
to explain this lineage as a product of cultural transmission. This is far from easy: we
know from personal experience, ethnographical studies, and the historical record that there
are many different processes that may result in similarities and differences among artifacts
of consecutive generations. Thus an underdetermination problem arises. Suppose that we
fi nd two slightly different projectile points, A and B, in adjacent layers of some excavation
site, and that dating methods show that they are approximately one generation apart in
age.^15 It makes sense to place these artifacts in a lineage and to presume that B was pro-
duced by person P, who (directly or indirectly) learned his trade from Q, the producer of
A. This may explain the similarity between A and B. For their differences, however, a
large variety of explanations is available: P might have come up with a functionally
equivalent stylistic variation, he might accidentally have produced an unfaithful copy, he
might have used a slightly different production process, or he might have adapted the
arrowhead to a perceived change in environments. Alternatively, B might have been robbed