feels his panel was biased in favor of more humanistic social sciences,
so that a proposal with a multi-causal model was penalized. Famil-
iarity with languages also was favored, which meant that anthropol-
ogists were more likely to be funded than political scientists. Here
again, the legitimacy of the deliberation process may depend on
achieving a proper balance between “cognitive contextualization”
and consistency, while also avoiding biases.
Two of the panelists most insistent about the use of consistent cri-
teria are highly successful African-American scholars. Their own ex-
perience with biases and with being stereotyped and underestimated
by professors and colleagues due to their race is likely to have made
them particularly sensitive to the application of consistent criteria.^36
One, a senior scholar in his sixties, observes that considerations such
as whether a proposed project seems to be already completed, or
whether the candidate has already received several fellowships, were
raised as objections for some cases, but not for other equally prob-
lematic cases. And indeed, my observation of panels suggests that all
criteria do not remain equally salient from one proposal to the next;
the saliency of criteria varies in part with what the proposal evokes
in evaluators and how it primes (or prepares) them to frame their
own further thinking about the issue.^37
Gossip. For some respondents, disciplinary gossip and other extra-
neous factors have no place in deliberations. This position is illus-
trated by a younger, female African-American panelist who is ada-
mant that panelists should consider only the evidence in the dossier.
She recommends that all panel members be urged explicitly “to
make their decision on the basis of the materials before them as op-
posed to going beyond that.” She expresses her dismay over the fairly
low ranking a proposal by a prolific and well-known scholar received
prior to deliberations. She attributes this ranking to widely shared
negative views about the applicant’s reputation that she, the panelist,
144 / Pragmatic Fairness