Kenneth Taylor’s book Reference
and the Rational Mind is a collection of essays (published in the last ten
years in journals and within conferences’ proceedings). It focuses on
reference, assuming there is a fairly stable intuition according to which some
(classes of) words are devices of singular reference. One may argue that
"run" refers to running, but verbs are usually not treated in
discussions of this kind. The lexical items he deals with are mostly devices of
singular reference and those are meant to denote single objects. He adopts the
view (by F. Recanati) according to which some terms have a meaning feature
(call it REF) " by virtue of which it indicates that the truth condition
[…] of the utterance where it occurs is singular" (p. 61). There are
some sound intuitions behind this: in some sense we talk about stuff, we
"refer to it." But there is also a major stumbling block, in my view.
While in every other field we are willing to listen to what is known
scientifically about a definite subject matter, the case of language appears to
be different: we "know what we mean" and we need no linguistics to
tell us otherwise. Massive bodies of evidence show how this is incorrect (e.g.
morphological phenomena are fairly superficial and not very useful even in
typing together languages for classification.) Most philosophers however simply
discard any evidence from linguistics on spurious grounds (mostly based on
epistemology.) Taylor is one of the philosophers of a new generation more
attuned and more cognizant of the actual contribution that linguistics gave to
our understanding in the last sixty or so years.

It is a fruitful
exercise for the reader to face one of the typical hurdles of referentialist
theories, from Plato on: empty names. Plainly we have naming
"devices" that name nothing at all (the largest prime number, Santa
Claus, etc.) What is the referentialist to do? Taylor steers a difficult course
between the idea that since referential devices contribute to propositions (the
raw materials of interpretation) nothing but
the object they refer to nothing and the obvious intuition that one says
something true when stating that there is no emperor of India now. In the
course of it Taylor takes into consideration the C principle (from Principle
and Parameters, see pp. 177 and ff.) and adduces an interesting observation
that linguists should take into account. The C principle regulates co-indexing
and it seems to run counter some of Taylor’s proposals; it predicts, e.g. that
"John shaved John" should be a borderline case of quasi-non
interpretable sentence, while "Saul shaved John" ought to work just
fine, equally for "John shaved him" where the claim is that, in the
latter, John_i has to be not him, hence that John_i is
different from he_j. This allows me to introduce,
given the short space, the main tenet of Taylor’s view. He thinks that in
interpretation we have to do with three domains of inquiry (see, in particular,
pp. 266-268): a lexical-referential, a compositional domain, and a contextual
domain. The novelty is that these three are not identical with the
traditionally conceived syntax, semantics, and pragmatics. To give but one
example, it is part of Taylor’s theory that pragmatic processes can precede and
not follow the "grasping" of the proposition "expressed",
to adopt H.P. Grice’s terminology. The problem the present writer sees in this
approach is that while many details will need filling (all too often, Taylor
slips into the "or something like that" clause) the main problem is
really a core problem of the language disciplines. While we have some idea (not
all that imprecise) of what a syntax is doing and accomplishing, the all task
of pragmatics in Taylor’s sense borders with general rationality. Does a
rational mind refer in the sense in which Taylor says it does? It is quite
possible though it remains to be shown stripping away any normative notion of
rationality.

That is why the most
promising section of Taylor’s book is his finale
of a couple of chapters that try to ground in psychological fact some of the
assumptions philosophers make (too?) easily. At the end of the book one starts
seeing (and this is probably one of the virtue of this book) why people would
bother for so long and so much in getting right the theory of the referential
features of the term "Santa Claus". If a ghost keeps haunting the
sleep of naturalistic philosophers, it is Gottlob Frege’s ghost. He left us
with the strange problem of fitting together completely different sets of
problems. On the one hand an impersonal mechanistically compositional syntax,
in however extended sense you want to take syntax, and a highly underdeveloped
semantics. Frege escaped his own dilemma building a formal artificial language
devoid of all the strangeness of natural languages. It is the present task to
bring the insights back to nature. Taylor with a mix of Frege (and Kaplan, and
Perry) and a very generous dose of Grice, tries exactly that.

© 2005 Adriano Palma

A. P. Palma, Duke University & Inst J. Nicod

Categories: Philosophical