LINGUIST List 2.631

Wed 09 Oct 1991

Disc: Double Modals; Specific/Referential

Editor for this issue: <>


  1. , Double Modals
  2. Barbara Johnstone, Re: 2.613 Queries
  3. Ron Hofmann, specific-contcrete-refential
  4. , specific/referential

Message 1: Double Modals

Date: Mon, 07 Oct 91 16:48:19 -0400
From: <>
Subject: Double Modals
With regard to David Denison's example of interrogative "might could", and
speaking as a native southerner, I simply cannot conceive of this as a possible
occurrence in Southern States English (or BVE, for that matter). I recall
being struck by other unrealistic-sounding constructions in Gurganus' novel at
the time I read it, although I cannot now remember exactly what they were.
This is not to say that declarative double modals are impossible, however; but
the only cases I can think of off-hand would necessitate some combination of
modal+subject+"oughta" -- for example, "Should we oughta eat dinner first?"
How do other southerners feel about this?
Bill Reynolds
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 2: Re: 2.613 Queries

Date: Mon, 07 Oct 91 17:06:34 CDT
From: Barbara Johnstone <>
Subject: Re: 2.613 Queries
References to double modals in Southern American speech can be
found in James B. McMillan and Michael B. Montgomery, Annotated
Bibliography of Southern American English, U of Alabama P 1989.
Also, Guy Bailey (Dept. of English, Oklahoma State U) and his
colleagues have done some more recent work on double modals in
Texas and Oklahoma. The McMillan and Montgomery bibliography
might be useful in connection with some of the other regional
syntactic features that have been discussed here recently, such
positive anymore.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 3: specific-contcrete-refential

Date: 08 Oct 91 08:31:14 EDT
From: Ron Hofmann <>
Subject: specific-contcrete-refential
TO: >INTERNET: LinguistTAMvm1.bitnet
To add to Haspelmath's 2.611 summary of specific-concrete-referential,
we can give an other, complementary interpretation of SPECIFIC.
 SPECIFIC: [presupposing existence in the 'real' world]
 (but please don't ask me what the real world is)
 "She wants to marry an Ainu speaker -- but I haven't met him yet."
 NON-SPECIFIC: [presupposing existence in some non-real world, here the
		world of her wants (desires, dreams, &c)]
 " --SAME--, but she hasn't met (him/the right one) yet."
Either account applies also to pronouns (examples above) or to definite
nominals; "(Smith's murderer/The guy who murdered Smith) must be insane"
is ambiguous on this point.
 This is for NP's. For N's, Aj's, Vb's, SPECIFIC contrasts with GENERAL
 (or possibly VAGUE) & alternates with HYPONYM.
I dislike Martin's main example for GENERIC, "Humboldtian views" (vs
"Humboldt's views"), for this form can be systematically interpreted as
[views that resemble/derive from those of Humboldt] where it is not
generic but REFERENTIAL. A more satisfying example (for me, anyhow) is
 GENERIC: "I want water" [the nature of what I want is water]
 REFERENTIAL: "I want some water" [what I want is (-Specific) water]
or better perhaps, "he bought apples" vs "he bought some apples"
Thus: GEN: "Humans have 2 legs" true by definition
 REF: (1)"All humans have 2 legs" almost certainly false.
 " (2)"Some humans have 2 legs" obviously true.
but -- if there are no humans, then (1) is true (vacuously) & a
logician's sense of (2) is false!
 This too is for NP's. 'Humdoldtian' -- along with other Aj, N, Vb, &c
 -- can be called non-referential, but then so are all words that are not
 complete NP's. For words, some people use GENERIC as I used GENERAL
 above. It seems worthwhile to distinguish carefully between the terms
 applied to NP's & the terms applied to N's.
All this within a modern received view of reference; there have been &
are other views, where this is probably nonsense.
There is much more to say. Over to you!
 ...Ron H.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 4: specific/referential

Date: 1 Oct 91 14:12
From: <>
Subject: specific/referential
I have recently read quite a few papers from various traditions that
deal with the semantics of specificity/referentiality (I am working on
a typological study of indefinite pronoun distinctions like English
some/any). The following are some generalizations that seem to hold, although
I have my own biases, of course:
SPECIFIC is used for an NP if the speaker presupposes the existence of a
referent. The typical example is something like the following:
She wants to get married to an Ainu speaker.
On the specific reading, there is an Ainu speaker (e.g. the one she did
fieldwork with and fell in love with) that she wants to marry. On the
NON-SPECIFIC reading, all she cares about is that her future husband speaks
Ainu, whoever he will be (e.g. because she wants her children to acquire
her ancestors' language, which she no longer speaks).
CONCRETE is a term that I have seen particularly in Russian-language
works, but also in Czech. It seems to be used in exactly the same way as
SPECIFIC. (A good place to look is Elena V. Paducheva's book "Vyskazyvanie i
ee otnesennost' s dejstvitel'nost'ju", Moskva 1985, which contains a very
clear discussion of basic notions of the semantics of reference.) By the way,
the term SPECIFIC is sometimes attributed to Fillmore (a 1967 paper in the
journal Glossa), although I would be interested to hear whether it was used
before. If this term was coined so recently, that explains why the Russians
have a different one.
REFERENTIAL is most often used in contrast to GENERIC. For example,
adjectival modifiers are generally NON-REFERENTIAL, whereas genitival
modifiers may be referential:
Humboldtian views vs. Humboldt's views
Similarly, dependent compound members and incorporated nouns tend to be
NON-REFERENTIAL: apple tree, bike rental, etc.
Sometimes people use REFERENTIAL in the sense of SPECIFIC (and NON-REFERENTIAL
in the sense of NON-SPECIFIC), e.g. Givon in the 1978 paper in Universals of
Human Language, Vol. 4 (ed. J. Greenberg, Stanford). There is of course a
clear similarity between the two senses, so to a large extent your usage
depends on your theory of these meanings.
Martin Haspelmath, Free University of Berlin
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue