An Enthusiastic Response to Allen's 12 Problems

Date01 December 1997
AuthorG.A. Vignaux,Bernard Robertson
Published date01 December 1997
DOI10.1177/1365712797001special13
Subject MatterArticle
An
enthusiastic
response
to
Allen's
12
problems
By
Bernard
Robertson
and
G.A.
Vignaux
*
New
Zealand
AlIen states
that
he identifies problems
that
have been raised in
the
literature
but
the
implications of which have
not
been adequately dealt with. Since he
does
not
refer to specific arguments it is unclear why he considers
the
exposition of these matters has
thus
far been inadequate. In any case, it is clearly
not
an effective attack on a theory to
point
out
that
the
current
exponents of
the
theory
are
not
able immediately to deal
with
all possible questions
that
might
be raised.
In this respect
the
challenges raised by Allen are interesting and useful. forcing
proponents to articulate aspects of
their
theory
better
or at all. But we
must
also be
beware
of
spending so
much
effort on fundamental (and therefore often
the
least
tractable) questions
that
we forget to move forward. to use Bayesian reasoning to solve
new problems and to improve its use as a yardstick against which to measure judges'
statements about how facts should be reasoned.
Allen's
12
points
1. The first
and
trite answer is
that
aprior probability is no different from any
other probability, it is assessed by reference to
the
available evidence. In
the
scientific
literature it is acknowledged
that
the
hardest task for any formal scheme
of
reasoning
is to tackle
the
earliest, formative stages
of
aproblem. At this
point
the
critic has two
choices. He can
either
deny
the
general validity of Bayesian reasoning (since this
problem is
not
peculiar to legal cases) or he can produce ascheme which offers a
better hope of formalising
the
earliest stages of problem definition
and
solution.
Allen specifically eschews
the
first course
and
does
not
seem to take
the
second.
Various answers to his specific question could be imagined. Where, as in
the
O.J.
Simpson case,
the
question is one of identity,
then
surely we could
start
with
aprior
odds of 1 to
the
population of
the
world
and
then
account for
the
evidence relating to
time and place. In fact.
when
it is said
that
it is difficult to assess a prior probability,
We
need to ask 'Prior to what?' since
the
jury
bring
with
them
the
common stock
of
knOWledge.
2. Analysing this evidence in Bayesian terms seems to have performed
the
valuable
Service
of
revealing
that
we do
not
know
what
it is evidence of, or why.
3. (a) This
chestnut
has been discussed extensively in
the
literature. This is
Certainly a theoretical problem
and
may occasionally be a practical problem. There is
~
Vignaux is Professor of Operations Research at Victoria University of Wellington, New
I
i~~iand.
Bernard Robertson is a consultant and editor of
The
New
Zealand
LawJournal.
---------------------------
THE
INTERNATIONAl
JOURNAL
OF
EVIDENCE
&
PROOF
335

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT