Feedback on paper of Jones et al.

FLAWS

	The paper makes no explicit claims, nor does it provide evidence in the
form of proofs or experiments. There are implicit claims, but these have to
teased out by careful reading. There is no explicit evaluation of these implicit
claims, although supporting evidence, of a kind, them can also be found by
careful reading.

	More generally, this is a good example of the use of conceptual analysis
to propose a framework, in this case a logic-based one. Conceptual analysis is a
methodology borrowed from philosophy and consists of careful analysis of a
concept in order to tease out its structure and properties. Usually, there is no
extrinsic assessment of the correctness of the analysis and its conclusions ---
you must assess the analysis by its intrinsic quality, i.e. how thorough and
convincing you found it to be. Research of this kind is, therefore, very
difficult to assess. 


HYPOTHESES

	There are a number of implicit claims that can be extracted from the
paper by careful reading. 

1. The normative positions N1-N15 exhaust the possibilities. The evidence for
this is that they were constructed using the 'generative method', which is, it
is claimed, known to be complete. You can think of this as an implicit theorem,
but the proof is not given here; it is assumed to be implied by the earlier work
of Jones and Sergot 1992.

2. Knowing that N1-N15 exhaust the possibilities is useful when it comes to
formalising certain kinds of legal or quasi-legal situations. The evidence for
this is experimental -- a number of case studies are analysed using this
formalism. Note that there is no 'gold standard' by which we can assess whether
the analyses give the 'right' answer. The correctness of the analysis is,
instead, to be assessed by how convincing it is to the reader. This analysis
usually takes the form of rejecting all but one of the Ni, but the accuracy of
the remaining Ni is not usually investigated in any depth. One could, in
principle, appeal at this point to some independent authority, e.g., a judge or
lawyer, to confirm that this Ni was indeed the right one. This was not done.



COMMON MISTAKES

* Many of you quote the first sentence of the abstract, that the contribution is
an "application of the formal-logical theory of normative positions to the
characterisation of normative informational positions". I thought this was
confusing and misleading. The technique being applied here was what he later
calls "the generative method" to exhaustively generate all normative
informational positions expressed in a modal logic (see hypothesis 1 above). I
gave more credit if you managed to extract some clarity out of this misleading
jargon, rather than just quoting it verbatim.

* Several people spotted hypothesis 2 above, but no one mentioned hypothesis
1. It is stated most explicitly in the first sentence of section 5.

* Several people identified "exploratory investigation to suggest a hypothesis"
as a contribution, but (a) there were no exploratory experiments reported and
(b) it wasn't clear what hypothesis emerged from the research that wasn't
implicit from the start, e.g., 1 and 2 above.