stepgrandparents and relational ambiguity

[Victor]

Laurence Horn wrote:
> At 9:39 PM -0500 1/22/09, Victor wrote:
>
>> I am rather concerned about a dictionary being cited as proof of lack of
>> semantic ambiguity...
>>
>> I would say exactly the opposite. "Murder" is *very* ambiguous in legal
>> context, leading some states to redefine crimes related to a loss of
>> life without using the word "murder" or, occasionally, without referring
>> to "manslaughter"....
> Maybe this is a purely terminological dispute.  Like Arnold, I find
> it useful to distinguish between true ambiguity and lack of
> specification/underspecification.
Yes and yes. I guess, I got lost in the details and forgot to put in the
punchline. Yes, I do believe that "murder" is *legally* ambiguous
despite a compact dictionary definition. No, I don't believe there is
any real ambiguity involved in lexical terms other than simple
underspecification.

What I am trying to do is actually distinguish between homonymy and
homophony on one hand and relational ambiguity on the other. I suppose,
the difference is ontological and perhaps qualitative. I don't see
"bat", "bank" and "bear" as ambiguous--they can easily be contextually
determined and, as AZ points out, they don't mix. This is not the same
for what I have been referring to as ambiguous kinship terms. I
mentioned three different meanings of "uncle" which can't be determined
either from context or by modifiers--to find out which kind of "uncle"
is being used, a complete explanation needs to be given. The same
applies to "stepgrandmother"--nothing in the general context and no
modifiers can help determine which of the two possible meanings are
implied. This is not the same, however, for "brother", "sister",
"sibling", "cousin", etc. These "units" are interchangeable and, if
there is a difference in meaning, i.e., gender in "sibling" or "cousin",
it can be narrowed down contextually (e.g., a respective pronoun
elsewhere in the passage). So there is a qualitative difference between
these three cases.

So, 1) I don't count underdeterminacy or underspecification as
ambiguity, 2) homonymy and homophony are not structural/relational
ambiguities, unless the lexemes actually have closely related meanings
(e.g., "cheek"). Aside from the "b-" examples already discussed, I would
say that "arm" is not ambiguous, even though it refers to a body part, a
class of parts on several classes of furniture (arm of a chair or a
couch), a robotic or a mechanical device (the robotic arm on the Space
Station), a part of a lever or a balance (the "arm" that opens the latch
on an escape hatch), etc. These are radically different meanings that
share some semantic components (e.g., similar function, similar
[physical, non-linguistic] morphology). The distinction I am making is
more philosophical (ontological) than lexical. The two sets of
stepgrandparents are really the same sort of objects, yet, they are
distinct (by the generation of adoption). In contrast, "stepgrandparent"
is also underspecified with respect to the genealogical line and gender,
but this is not the ambiguity of which I speak.

>    There is no compelling evidence
> from crossed-reading tests or other sources that compels us to regard
> "kill" as ambiguous between the instrument of the killing, the
> motivation or intention of the killer, etc.  Yes, someone can be
> killed by a falling rock or an assassin, but there's no reason to
> consider "kill" as ambiguous for that reason, any more than because
> the killing could take place in New Haven, New York, or the New
> Hebrides.  This is simply not a lexical ambiguity or homonymy or
> homophony of the kind we have in "bank" vs. "bank", "bear" vs.
> "bear", "bat" vs. "bat", etc.  But if you find it more convenient to
> apply "ambiguity" as a cover term for both kinds of lexical
> relations--the "uncle" kind (that we're calling underspecification)
> and the "bat" kind (that we're calling ambiguity or homonymy)--that's
> all there is to it.
>

There is no structural distinction between different modes of "kill".
The issue is specificity. The distinctions for "banks", "bats" and
"bears" is ontological--they are different sorts of objects. Neither
case is an ambiguity. Homonymy is not "ambiguity" as I have been using it.

    VS-)

------------------------------------------------------------
The American Dialect Society - http://www.americandialect.org