I've found there to be a cline of understandings (see Zwicky & Sadock
1975 "Ambiguity tests and how to fail them" in the Syntax and Semantics
series out of Academic Press) associated with a lexeme, ranging from (i)
those understandings which correlate with subcategorial and syntactic
patterning in a language, down through (ii) those which correspond to
word class structuring, to (iii) those understandings which have just about
no structural language reflex at all. As a lexical semanticist, I'd restrict
the use of 'sense' to category (i).
Of course, 'lexical ambiguity' will correspond to contrasts holding
between senses (under the reading just defined).
For example, I've done detailed work on the polysemy of the English
verb 'understand' (see Wheeler 1993, "Sense and subsense: the lexical
semantics of the English verb 'understand' ", Cahiers de Lexicologie...).
There, I defined sense operationally as follows:
A sense of a verb will manifest a
(i) characteristic subcategorial pattern
(characteristic syntactic structure of the arguments
to the verb)
paired with
(ii) characteristic, distinctive syntactic distribution potential
(e.g., ability to occur in the imperative, under main
verb netation, or intransitively, etc.
A sense so delineated will then show distinct lexical substitutes.
I identified 4 senses of 'understand'. COMPREHEND (I understand the
problem/the idea/why she left, etc.), HEARSAY (I understand that your
aunt is leaving), REALIZE (I understand that the mortgage is due on the
first), AND READ(+MANNER) (I understand the phrase literally).
Ok, here's what's possibly interesting about this for the current
discussion.
Take
(1)a I understand the problem,
b. I understand Lucy,
(2) I understand that the mortgage is due on the first.
Examples (1)a-b exemplefy the COMPREHEND reading, and example
(2) the REALIZE reading. That is, (1)a-b exhibit the same subcategorial
frame (NP1 V NP2), and the same possibilities of syntactic distribution --
they both readily accept negation, are reluctant to accept imperative, and
are fine intransitively. Thus, examples (1) are examples of the
same sense.
However, we can then move to a level of finer detail and recognize that
there are differences of yes, word class (abstract v. human noun),
between (1)a and (1)b, and these may prompt differences of lexical
substitutes. Thus, (1)a is paraphrasable as "I comprehend the problem",
and (1)b as "I sympathize with Lucy."
I've named this level of contrast, SUBSENSE. I would say that
'understand' is not ambiguous across (1)a-b. These are the same
meaning of 'understand' because they show the same syntactic patterning.
Thus, the syntax treats them as the same, a significant fact. They do
show differences within the sense category correlating to wordclass of the
direct object nominal. This information is, of course, also important.
On the other hand, (2) is a distinct meaning of 'understand' - it shows a
distinct subcategorial frame [NP1 V that + S], and distinct syntactic
distributional properties: imperative = good.
What I'm suggesting is that there are different types of distinctions of
understanding in language. It seems to me that 'sense' and 'ambiguity'
have been being used to refer to the whole cline of understandings I
described above [(i) - (iii) and beyond]. My analysis of "I understand
the problem" and "I understand Lucy" suggests we need to distinguish
between these types of understandings.
Oh, terribly awkward. I'm now out of town till May 8th. If you
respond, and are patient, I'll catch up then.
rebecca wheeler
lexical semanticist