Thanks to Regina Blass for her interesting comments. I'm reordering some of
the illustrations she gives for further analysis. She writes, "I would like
to challenge both JL and Christoph's claim that no polysemy is involved in
1. old (books & maps).
vs.
2. (old books) and maps.
examples.
"Cruse shows that items like 'book', 'map', 'CD' have FACETS. They have
[various] readings in certain contexts in which ambiguity tests are
positive. [E.g.] "book" has the reading "1. tome" and "2. text" and, as
such, a sentence [containing it] can convey different truth conditions
depending on whether [one facet or the other is] meant:
3. Did you buy an an old book?
[may get as answers]
4. Yes, I bought a novel by Charles Dickens.
5. No, the book was printed in 1999.
[I was wondering about e-books. A typical case of the text interpretant
only! "Buy? Are you crazy. What's the use of a full internet subscription
for then! I read it from Gutenberg Project!" :) JLS]
The same could apply to 'map' [where]
6. Did you buy an old map?
where a map [of pre-1974 England is meant! I like to see Rutland back
there! JLS].
[which could get as answers]
7. [Yes, I bought a map with Rutland and Huntingdonshire in it!
Lo-ve-ly! And Manchester is, as it should, in Lancashire!
8. [Well] the map was printed in 1999 [but printed by THIS ENGLAND,
that conservative publication if ever there was one, sticks to the
"Save the Old Shyres" Campaign, and so, forget about seeing
Humberside, Cleveland, and Greater London! And Yorkshire is the
greatest shire, as it always was!]
Regina writes,
"According to this test 'book' and 'map' are polysemous. However, 'book'
and 'map' can also convey both facets AT THE SAME TIME and therefore they
are not intuitively polysemous. They can also fail certain ambiguity tests.
You can say:
9. John bought an old book. So did Mary.
which could be interpreted as
10. John and Mary bought old tomes with old texts.
However, it would probably be possible to also convey that
11. John bought an old tome with a fairly new novel
and Mary an old tome with an old novel.
So the ambiguity test fails here.
[I fail to see the meaning of the first conjunct of (11), i.e.
12. John bought an old tome with a fairly new novel.
I guess the idea is that what is polysemous here is "old". i.e. an old
novel is, by anglophone standards, "OLDER" than an old tome. E.g. "Waves"
by Virginia Woolf may be said to be a "fairly new novel", but if John
bought the first edition by Hogarth Press, the tome would be "old" (or
"dated") enough - and a collector's item, at that! In general, with
Regina's examples, I was struck (not being an anglophone) that to me it
wasn't clear whether the ambiguity applied to "old" rather than to the
modified noun! JLS]
Regina writes: "To come back to your [syntactic] scope example. The facet
readings add further complication to your interpretations. Your 'old books
and maps' example will probably allow both facet readings to occur.
13. John bought old books & maps.
can be interpreted as
14. John bought old tomes & "new" maps
(content and physical make-up - or either of them)
However, [disambiguating the scope]
15. John bought old books & OLD maps.
would probably be more restricted to either a 'tome' or 'text' reading for
both.
[i.e. I see. Following, Cruse's terminology, the same "facet" would be
chosen, ceteris paribus, qua generalised conversational explicature!]
"So, since we are in fact dealing with polysemous facets we have to ask how
to approach their disambiguation in pragmatics. By the principle of
relevance the [addressee] will interpret the utterance
in the first available context consistent with the communicative principle
of relevance. If the contextual information leads to a "tome"
interpretation, the [addressee] will access encyclopaedic entries connected
to 'tome' and enrich [and disambiguate] the explicature accordingly. If it
leads to the
'text' interpretation he will enrich the explicature accordingly."
Regina adds, "I do not find Grice's answer to those problems very helpful.
After all he does not give us a *means* by which to 'avoid ambiguity', RT
does."
Well, thank you. Indeed, I like Cruse's idea of a facet and the whole idea
of interpreting the facet as modifying the truth condition(s) of a claim.
Now, as to how satisfactory Grice's original answer is, the issue is, I
find, tricky, and I would not wish to get this into a historiographical
polemic!
Regina writes, "Grice does not give us a "means" by which to avoid
ambiguity".
This is, to me, pretty ambiguous!
As I understand Grice, "avoid ambiguity" is one of the strategies of the
conversational decalogue (if we count them as being nine plus the one for
"the" being the tenth, i.e. Studies in the Way of Words, p.26 AND p.273).
As such, Grice, it seems to me, is not really WILLING to give us the
*means* (as Regina puts it) by which to avoid ambiguity. The maxim is
expressed in natural English and does not require much of an interpretation
(i.e. it does not go, "avoid Empson's seven types of ambiguity!").
As is obvious with the other maxims (including "be relevant"!), Grice seems
to be more concerned with FLOUTS to the maxim rather than the following of
it (i.e. what to avoid ambiguity would really amount to).
From Grice's two examples of FLOUTS of the "avoid ambiguity" strategy, one
can get an idea of what kind of ambiguity he was having in mind. His first
example being
16. Never seek to tell thy love,
Love that never told can be
since she [Nature] pricked thee
Out for woman's pleasure
mine be they love,
& thy love's use their treasure
W. Blake
and the second
17. Peccavi (I've Scinde),
Wrote Lord Ellen so proud.
More briefly Dalhousie
Wrote Vovi (I've Oude).
Punch, Oxford Dict. Quot.
In 16, the ambiguity resides in line 2, which can read as 18 or 19
18. Love that never can be told.
19. Love that never told can be ( = exist).
Is the utterer involving "polysemy" there? It seems to that the ambiguity
is, like in "old books & maps", of the scope, syntactic, type (which
admittedly, may yield different truth conditions). You don't seem to need
to bring in Cruse's polysemous facets or alternate readings for specific
lexical items. Just two different surface syntactic parsings (which, will
yield, admittedly, again different truth-conditions). Ie. Grice's examples
seem to aimed at illustrating the interface between syntax and pragmatics.
I.e. how alternate semantic parsings give raise to different truth
conditions (and pragmatic interpretations) without need to fulfill an
intermediate "full" lexico-semantic level. The two semantic parsings being
20. Love that never can be told by anyone.
i.e., a kind of the love that dare not speak its name: a love that IS (or
"exists"), but, for some reason, cannot be told. The example is,
admittedly, tricky since it involves the negative adverbial "never", and so
in a sense, it resembles
21. The King of France was never bald.
Does 21 presuppose (implicate, mean) that there was a king of France?
Relying on "never", one can claim that 21 does NOT presuppose (implicate,
mean) that
31. There is a king of France.
other than via a cancellable conversational implicature. Similarly, (20)
may not presuppose that such love exists. I myself opt for this latter
alternative, since I note that Blake is placing such love as object of
one's search, "NEVER SEEK TO TELL THY LOVE". Cfr.
32. Never seek the abominable Snowman.
as implicating - on occasion
33. Because your search is doomed to fail,
there not being such a man.
But I'm rambling... Now, the second "ambiguity" flout example, the
"Peccavi" couplet is so contrived, I find, and it doesn't really even
concern English, which is my topic of research! Until someone proves to me
that a similar kind of military example can be built for English, I think
Grice was being too much of a classicist there! (Recall his lectures were
given at Harvard which is reputedly the one offering the best classical
education in USA!).
=================
These above are, then, some observations regarding "avoid ambiguity". In
their essay in Werth, Sperber & Wilson argue that "avoid ambiguity" reduces
to "be relevant" - and I will not deal with this reduction here: even if
something reduces to something it doesn't mean we must stop talking about
it. E.g. Grice thinks "meaning" reduces to "intention", but most of his
essays have "meaning" in the title, not intention!
===================
Now, the other main aspect where ambiguity features large in Grice's
programme - and which I expressed in the header of this thread - is in his
"Modified Occam's Razor" which he introduces in "Further Notes" (Studies,
p.47). My focus has been the connectives, and so I will deal with that, in
particular with the first connective,
34. John bought the book & read it.
vs.
35. John read the book & bought it.
Suppose the book is "The Owl &the Pussycat", so that it is possible that
36. John had time at the bookshop to go thru all the poem by Edward Lear,
i.e. read it, to mainly note that no main errata were found, and since he
loved the illustrations, he proceeded to buy it.
i.e. (36) would be truth conditionally equivalent to
37. John read the book & THEN bought it.
On the other hand, it is possible that he did not read the poem in the
bookshop, but, as usually happens, just bought it. It was a gift anyway. It
was only after buying and giving it as a gift to his nephew that he found
himself having to read the poem, too! So both (34) and (35) would be truth
condtitionally equivalent to
38. John bought the book & then read it (to his nephew).
The idea being of course, that for an entry like
39. and.
we have two polysemous facets, as Cruse would call them, viz.
40. and. 1. = "&"
"p & q" is true iff both p & q are true
2. temporal sequence. AND THEN.
This seems simple enough to me, and what Grice is saying is, follow M.O.R.
(Modified Occam's Razor) i.e.
41. Avoid multiplying senses (or polysemous facets) without necessity.
Grice does note that the status of M.O.R. is questionable (and PhD's have
been written against it!), but his whole programme seems to depend on the
idea that it is sometimes possible to know when to apply it, and
successfully too, especially when there is an alternate explication via
conversational implicature (I'm not so sure "explicatures" play a role
here) to account for the SECOND, derived, so-called alleged polysemous facet.
==========
Now, it would be nice if we could compare the maxim "avoid ambiguity" with
M.O.R., but I'm not sure that that was Grice's idea. I would need textual
criticism. It seems that with the maxim, Grice is concerned with either
syntactic alternate parsings (or contrived examples in foreign languages),
whereas with M.O.R. he is concerned with specific lexical items, and
notably the bare connectives - as he describes them (rather than whole
syntactic paradigms, or nominal phrases, etc).
=========
One example of Grice's that may be illustrative as providing a link,
though, may be
42. He is in the grip of a vice.
This, Grice, says, is "ambiguous" - although he does not use the word
(Studies, p.25) - between 43 and 44
43. About some particular male person or animal x,
at the time of utterance, x was unable to rid
himself of a certain kind of bad character.
44. About some particular male person or animal x,
at the time of utterance, some part of x's person
was caught in a certain kind of tool or instrument.
Here I would not doubt to consider a Cruse polysemous facet - and fail!
(see below):
45. vice 1. corruption of morals
from the Latin, vitium
2. screw.
from the Latin, vitis
Now, the obvious question I'd like to ask a RTheorist is whether they would
consider "vice #1" and "vice #2" to be an example of a polysemous lexical
entry or just TWO lexical entries! (Grice did not tell me. He just
implicated his answer). Perhaps he thought it is obvious that he thought
they were two lexical entries involved). The interesting thing (to me) and
from which I infer Grice's position - philosophers seldom deal with "words"
as such! - is that when discussing M.O.R. he brings up the same example. He
writes,
"Assuming for the moment that the tests
[to differentiate "mean-n" from "mean-nn"]
are roughly adequate, what I want to do now
is NOT TO EMPHASISE the DIFFERENCES BETWEEN
THESE CASES, BECAUSE THAT HAS ALREADY BEEN DONE"
And ad nauseam, too.(So much so, taht DH Mellor says Grice died of
non-natural causes!)
BUT RATHER TO LOOK AT WHAT THEY HAVE IN COMMON. IS
THIS DOUBLE USE
Note the use of USE, a favourite word with Oxford philosophers, cf. G.
Ryle's essay, "Use, Usage, and Meaning" in G. Parkinson, The Theories of
Meaning. Oxford Readings in Philosophy).
of the word "mean" just LIKE THE DOUBLE USE
note again the use of "use"
of the word "vice" to refer SOMETIMES TO
SOMETHING APPROXIMATING TO A SIN AND SOMETIMES
TO A CERTAIN SORT OF INSTRUMENT USED BY CARPENTERS.
ONE IS PRETTY MUCH INCLINED TO THINK IN THE LATTER
CASE TO SAY THAT THERE ARE TWO WORDS WHICH ARE
PRONOUNCED AND WRITTEN THE SAME.
But that inclination may be a vice! I recall the dialogue between Grice and
Austin reported by Grice: Grice: I don't care what the dictionary says.
Austin: And that is where you make your big mistake!).
On general grounds of economy,
cfr. LR Horn on the London School of Parsimony! (Parsimony being the term
by which Occam's razor - to shave Plato's beard - is known in philosophical
jargon).
I am incined to think that if one can avoid saying that
the word so-and-so has THIS SENSE, THAT SENSE, and
the other sense, or THIS MEANING, and ANOTHER MEANING,
if one can allow them to be variants under a single
principle, that is the desirable thing to do: don't
multiply senses beyond necessity" (Studies, p.291).
Now, from that passage it seems pretty obvious to me that Grice is
suggesting that we take "vice 1" and vice 2" as being "TWO WORDS PRONOUNCED
AND WRITTEN THE SAME". I.e. Homonyms, rather than polysemous words (like
"row" (of chairs) and "row" in the river).
But WHAT would for Grice actually *be* an example of a polysemous word!? I
can't find one single case. One case I do recall is from Adam Kilgariff (of
U.Sussex/Brighton)'s essay in "I don't believe in word senses". He notes
that an early ediiton of the O.E.D. features:
46. horse. 1. a quadruped animal
2. a representation of a horse
(This was later "improved", and the second usage eliminated!).
===========
I believe that the notion of truth condition may (AND SHOULD) help us here
but when it comes to polysemy, the issue is tricky, I find, because of
so-called EXTENDED USAGES, which I don't strictly call polysemous. This may
even apply to "and". I.e. it may well be that the primeval (seminal) use is
the truth-functional "&" and that "and that" (temporal or causal sequence)
is an extended "USE", not a different SENSE. But how and why do senses grow
then?
===========
I think another example by Grice (Studies, p.89), with which I'll end this
sketchy notes may be illustrative here. It's his discussion of grass to
mean "lawn" or "pot":
47. If I shall then be helping the grass to grow,
I shall have no time for reading.
as meaning upon the two polysemous facets (?) of "grass" as either 48 or 49:
48. If I shall then be assisting the kind of thing
of which lawns are composed to mature, I shall
have no time for reading.
49. If I shall then be assisting the marijuana to
mature, I hsll have no time for reading.
Again, though, Grice does not use the word "ambiguity". He goes the hard
way and writes,
"It would be true to say that the word "grass" means
(loosely speaking) "lawn material" and also true to
say that the word "grass" means "marijuana". Such
meaning specifications I shall call the specifications
of the timeless meanings of an incomplete utterance
type which may be a nonsentential word or phrase or
may be a nonlinguistic utterance type which is
analogous to a word or phrase" (p.89).
=======
Trouble is, I think, that, here, we cannot strictly say, as with "vice"
that it's two words pronounced and written the same. One is an extended,
figurative use of the second, viz.
50. grass. 1. lawn material
2. marijuana
(Incidentally, I think the same applies to cases like "gay" "1. merry. 2.
homosexual," or "turk: native of Turkey; unwanton child" [where the second
meaning, or item separated by the semi-colon may be thought of as a
metaphor - and thus a conversational implicature, Studies, p.34). So,
old-fashioned as I am, I'm quite against those dictionaries, like the new
C.O.E.D., that orders the alleged "polysemous facets" in terms of frequency
rather than history!
Sorry for the elongated ramblings. I got carried away. And thanks again to
Regina Blass for her food for thought.
Best,
JL
(Mr)
Buenos Aires, Argentina
- I write "Mr" but I'm no Mr Bean! You can call me "JL"! :)
This archive was generated by hypermail 2b29 : Mon Feb 26 2001 - 17:13:37 GMT