Marcus Smith wrote:

> Tom Wier wrote:
>
> >> I agree, but I didn't invent the terminology!  :-)  That's the
> >> terminology I've seen in every description of Romanian I've read.
> >
> >Well, then.  A thoughtful linguist would change it.
>
> And divorce himself from the previous literature.  I don't think so.  Rarely
> happens, [....] IMHO changes should only be made when the current terminology
> leads to confusion.

But terminology like 'unergative' verbs do in fact lead to confusion, and
Dixon has remarked as such in that case (see his _Ergativity_, pg. 235 or so).
No, on the contrary, inappropriate terminology by definition leads to confusion,
because it uses a word in such a way as to imply characteristics and qualities
into the object under discussion that aren't there:  the illocutionary effect is
wrong.

The reality is that you *always* have to set out what terminology you are
using and what you mean when you intend to analyze a grammatical system;
you can't just inherit whatever the rest of the field uses, if you really want to
understand what others are saying.  Otherwise, it leads to the very confusions
that you spoke of:  it's difficult to tell bad terminology from bad thinking.

> An inappropriate label for a case is not such a situation.  It is, after all,
> only a label, and nothing important hangs on it.

Nothing important hangs on it? Respectfully, I disagree. There are two
extremes to the use of terminology.  On the one hand, there are those
who act like that the words have some inherent, external meaning other than
that which we imbue it with.  Then there is the opposite extreme:  that any
ol' word will do, since they're all arbitrary anyways. This latter argument
is more meaningful, except that it too misses the point: yes, they are arbitrary,
but it doesn't follow that conceptual consistency (*not* terminological
consistency) is also arbitrary.   Presumably, any empiricist tries to
describe different phenomena both as independent entities and as relative
to one another.  So, what I'm basically arguing for, in other words, is a
terminology that, whether because of its morphology or because of its
semantics, better captures the generalizations within and between those
phenomena.  "Unergative" and "categories" fail on both of those counts,
respectively.

======================================
Tom Wier   |     "Cogito ergo sum, sed credo ergo ero."
======================================