Frameworks vs. theories

[SHALOM LAPPIN]

> I also wanted
> to add that, at least in cases like this, `frameworks' don't make predictions.
> Particular theories, some of which might be formulated in more than one
> framework, make predictions. Getting clear on this is important for any
> discussion that tries to cross framework boundaries...
>
> All best,
> Ivan

I think that one has to be careful here in assuming a clear distinction
between frameworks and theories. The defining assumptions of a framework
are, in general, simply more abstract claims located closer to the
center of the a general theory than other assertions, that can be more
easily revised without inducing major changes in the theory. Do we take
Newtonian mechanics and relativity to be alternative frameworks
or competing theories with more general application to the physical
universe than hypotheses about a specific phenomenon formulated in
either framework/theory? Kuhn's notion of paradigm change
aside, it is hard to see how one can resist the latter
view. Carl's comparison of theoretical frameworks in linguistics to
programming languages seems a bit problematic here. These languages
can be shown to be formally equivalent at least to
the extent that for any program that one can write in one language a
counterpart program performing the same basic set of computations can be
formulated in the other languages. But it is not obviously the case that
any theoretical analysis that one formulates in one grammatical
framework can be stated in an alterative framework without revising some
of the latter's basic assumptions. So, for example, a classical movement
account of unbounded dependency is commited to treating all such
dependencies in the same way, while a typed feature structure account
naturally allows for independent patterning of distinct varieties of
this relation. On the movement view extraction and wh-scope should obey
basically the same constraints, while on an HPSG approach these two
types of relation are, prima facie, expected to exhibit some degree of
independence, given that they are handled by distinct features (SLASH
and REL). It is, of course, possible, to modify these theories to
accomodate either uniformity or independence of these two instances of
UD relations. But either move represents a weakening of the theory in
which it is made. To the extent that evidence can be supported to
motivate one of these conclusions, there is an interesting empirical
difference between the frameworks. Given that frameworks are simply
general theories of grammar, this is not suprising. It seems to me that
we should be concerned to find the best (simplest, formally most
elegant, most empirically motivated, etc.) theory rather than to
work either for the convergence of alternative frameworks or for the
sharpening of the distinctions between them. To achieve this objective,
it is necessary to be aware of work done in alternative
frameworks and to constantly evaluate it comparatively. Regards.
                           Shalom