explicit requirement to do so, some panelists also volunteer informa-
tion on indirect or informal ties (“this student’s mentor is a close
collaborator of mine” or “I know this applicant’s adviser very well
and trust her letter”). These disclosures are not offered systemati-
cally, perhaps because such ties are not uncommon. Given the degree
of specialization in American academia, panelists are likely to know
personally or by reputation most of the scholars whose areas closely
overlap with their own, especially if they are very active in research
circles, as are most evaluators.
Still, panelists’ awareness of the importance of limiting the “cor-
rupting impact” of personal relationships makes disclosing informal
ties common enough to be a customary rule. A historian’s reply to
my asking him to describe his favorite proposal is illustrative: “That’s
a little bit difficult because to be completely honest about it, the one
that I liked the most was by a student who quoted me at length. I
recused myself. I didn’t enter the discussion on that. She’s not liter-
ally my student. She did some independent work with me.” A sociol-
ogist, too, says she chose to remain silent when a former colleague’s
work, which she does not appreciate, was discussed.
Both these scholars were pleased when the decision the panels
reached in each case endorsed their own viewpoints. In contrast,
all of the panelists I interviewed who had written letters for appli-
cants (and who therefore excused themselves from the deliberations)
were upset when these proposals were not funded. Such decisions,
they said, made them feel awkward, as if their fellow evaluators
are giving them a vote of no confidence. They said they abstained
from asking the other panelists for an account of the deliberations,
and the other panelists did not volunteer such information. Interest-
ingly, the “slighted” parties seemed to expect such a breach of con-
fidentiality, despite their commitment to universalism, and were dis-
appointed when details were not forthcoming.
Because scholarly expertise is superposed onto the social networks
Pragmatic Fairness / 127