Seite 204 [204]
MIRJAN DAMASKA
likely that criminal cases will be decided by overwhelming statistical evidence,
although due to the magic of numbers statistics is the most serious candidate for
replacing ordinary reasoning in adjudication. This seems to follow from differences
between testimonial (anecdotal) and statistical evidence. Witnesses testify about
concrete facts they perceived concerning relevant events, while statistical experts
are ignorant of these facts and merely inform the court about the frequency
of relevant events (quod plerumque fit). And while the probability of error in
witness testimony remains unexpressed, it is clearly indicated in the statement of
statistical experts. If the latter claim that the probability of the defendant’s guilt
is 99%, for example, the possibility still remains that the defendant falls in the
remaining statistical 1% niche. He can claim that he belongs in this niche and can
object that he should not be sacrificed on the altar of statistically optimal decision¬
making. Since judges strive in criminal cases to deliver individualized justice,
it is hard for them to disregard this argument. It is hard for them to disregard
it, even though statistical reasoning would in a large number of cases yield the
greater number of accurate results. This shows that the opposition to decision¬
making on statistical evidence stems less from epistemic reasons than from moral,
political and cultural background considerations. It has in fact been persuasively
argued that ordinary cognition represents the ethical and political predicate of
justice as presently conceived, and that judicial surrender of ordinary to expert
reasoning that requires sailing over deep mathematical waters would undermine
the legitimacy of criminal proceedings.® All things considered, then, it seems
unlikely that law will obligate judges to disregard their native reasoning processes
and decide cases on the basis of expert findings and opinions. The imperialism of
science is not likely to extend so far.
What one should expect in the not too distant future, however, is the increased
acceptance of negative proof rules based on scientific insight, and especially
acceptance of their presently disliked corroboration kind. This is because legal rules
of this kind, although based on scientific insights and technological achievements,
preserve the space for judges to assess the value of evidence by using ordinary
cognitive processes. Judges are not compelled by these rules to establish the
defendant’s guilt contrary to their beliefs, but are prohibited from convicting him
or her without the imprimatur of science.’ An example of such a rule has already
been given: a legal provision can easily be imagined requiring that the testimony
® Jonathan Cohen, Freedom of Poof, in William Twining (ed.), Facts and Law, 16 Archives
for Philosophy of Law and Social Philosophy, Hudson, NY, Steiner Books, 1983, 1-6.
° Rules of this genre illustrate Karoly Bärd’s contention that some norms serve neither
concerns about fact-finding accuracy, nor the protection of substantive values (such as
privacy or human dignity). Their purpose is instead to reduce the danger of wrongful
convictions. See Bärd, Fairness in Criminal Proceedings, 55.
+ 202 +