those adjectives

[bresnan at csli.stanford.edu]

Response to Miriam Butt's message of 9/1/97.

With disarming simplicity and jaunty confidence Miriam describes her
initial proposal to reject a subject of predication for adjectives:

>The standard/traditional approach had been to have adjectives
>subcategorize
>for subjects:
>
>         `red<SUBJ>'  etc.
>
>We, the gang of fresh-faced young grammar writers, thought briefly
>(about 10 minutes) and decided that there was no real reason to really
>have this analysis any longer, given the larger reliance on argument
>structure concepts in recent years.  So, we nixed subjects in
>adjectives and lived happily with that for a long time.

With fresh-faced good humor she describes the consequences:

>With time, problems arose from several sources, some of the
>ideological, some of them practical.

She then enumerates some problems that arose in German grammar and
how she was able to evade them:

1. coordination of adjectives and participles.

        Die gelbe und rennende Katze -- `The yellow and running cat'

Miriam evaded this problem by eliminating the subject from the verbal
lexical form, to make it parallel with the adjective.  But, she notes,

     I still think this is a reasonable thing to do, but it didn't
     seem to make very many other people happy.

Consider an  English example:

	She seems/looks/appeared/became confident and convinced of her ideas.

Problems: the matrix verb is a subject raising predicate and passes
its subject to that of its predicative complement in f-structure (by
functional control: Bresnan 1982).  The complement subject gets distributed
across the conjuncts (by coordination -- Bresnan, Kaplan and Peterson;
Kaplan and Maxwell).  The passive participle needs its subject, and is
happy.  But if the coordinate adjective lacks a subject argument, the
result will be incoherent.  

In fact, elimination of a subject from adjectives causes problems with
raising constructions quite independently of coordination:

2. raising with adjectives:

	As the political problems mounted, Clinton became ever more
		liable to tergiservate. 
	He seems likely to waffle on this, too.
	It struck me as too unlikely to rain to take full Goretex.

Here the matrix raising verbs pass their subjects to their adjectival
XCOMPs, and the raising adjectives in turn pass their functional
subjects to their VP complements; these may be active, passive, or
require idiomatic or expletive subjects.  Eliminating the subjects
here means redoing much of English syntax.  (Certainly not out of the
question, as long as one realizes what is involved.)

Miriam good-humoredly recounts and dismisses other evidence for a
subject of predicative adjectives:

3. agreement: adjectives agree with their subjects in many languages.

She counters that "German makes a very strong argument *against* conflating
the predicative and attributive uses of adjectives" --because
predicate adjectives do not agree with their subjects in German, while
adnominal adjectives show nominal concord.

Of course, the issue of whether adnominal adjectives should be
analyzed in the same way as predicate adjectives is separate from the
question of whether (any) adjectives have subjects, which is where
Miriam starts in her message.  Crosslinguistically, adnominal concord
is often independent of subjects/predication.  In Bantu, for example,
adnominal modifiers show gender concord with the head, regardless of
whether they are quantifiers, determiners, adjectives, relative
clauses, etc.  In Australian, adnominal case concord works the same
way; it isn't restricted to the kinds of nominal modifiers that might
be thought to have subjects.  (See Rachel Nordlinger's Stanford
dissertation CONSTRUCTIVE CASE: DEPENDENT-MARKING
NONCONFIGURATIONALITY IN AUSTRALIA for recent discussion.)  In German,
too, the concordial adjectives are behaving in their agreement like
demonstratives, I believe; and thre is no independent reason to assume
that demonstratives have subjects of predication.  So this point
Miriam brings up about German is actually a roter Hering.

3. anaphora.  After recounting examples of binding patterns with
adjectives that support the presence of a subject, Miriam counters:

   However, I think it is also possible to construct examples
   in which there is no subject (impersonal constructions) and where
   the reflexive can still find an antecedent.  Consider the German:

   Hier wird sich nur gemalt, nicht fotografiert.
   here is   self only painted, not photographed
   `Here one only paints, not photographs oneself.'

   Hier wird gelernt, auf sich stolz zu sein.
   here is   learned, on self   proud to be
   `Here it is learned to be proud of oneself.'

   These sentences are examples of what the Germans call 
   "subjectless" constructions:  impersonals.  
   Here, the reflexive is good, even when there is no overt
   subject of the sentence. 
   If one admits such sentences into the debate, then they
   serve to cast doubt over the assumption whether one should
   indeed resolve anaphora only via a recourse to grammatical
   functions, or whether one should do anaphora resolution
   at the level of argument structure (in addition, or perhaps
   even instead).
   (In fact, argument structure effects on anaphora resolution
    have been argued for). 

Reflexives indeed have multiple functions, sometimes showing pure
pronominal binding effects and other times a valence-changing effect
on argument structures.  Quite a bit has been written about this in
LFG, starting with Grimshaw's work on French reflexives in the 1982
volume I edited and including a valuable 1985 BLS paper by Draga Zec
and Sells, Zaenen, and Zec 1987 "Reflexivization Variation: Relations
between Syntax, Semanatics, and Lexical Structure" in WORKING APPERS
IN GRAMMATICAL THEORYA ND DISCOURSE STRUCTUR (CSLI Publications),
ed. by Iida, Wechsler, and Zec [this is an important paper; it should
be on the top 33 lfg readings].  Alex Alsina's recent CSLI monograph
also contains a detailed discussion for Romance reflexives, and of
course there is Helge Lodrup's excellent paper given on Norwegian
reflexives at the recent lfg97 conference in San Diego.  In other
words, many LFG theoretical proposals and analyses exist which have
already been proposed to account for the types of data Miriam brings
up as problematic for a subject-based binding theory and its
implications for the adjectival subject hypothesis.  

4. copular constructions: verbless predications abound, and support a
subject of predication in nonverbs.

	About this, Miriam says:

    This is a fairly reasonable way to proceed, but I personally
    would rather adopt an approach in which the overt copulas do
    get to have predicative force. In those constructions without
    an overt copula, I would want to adopt something akin to
    a constructional grammar approach --- we know it has to be
    a copular construction by it's structure, so it gets a 
    subcat frame supplied *by virtue of the construction* 
    (in practical terms through an empty category).

Many of the cases of verbless predication and "small clauses" have no
syntactic place for an empty copula (e.g. examples like those in point
2. above).  Cf.:

 *He became to be ever more liable to tergiversate....
 *Consider yourself to be lucky.

And much has been written about the subtle semantic differences
between predicate complement constructions with and without the
copula. E.g. example (a) suggests direct perception; (b) does not.

a. I found him repellent.
b. I found him to be repellent.

Better examples could be dug up from the literature on this (e.g. Ann
Borkin's thesis).

Finally, Miriam points to the advantages of the subjectless adjective
analysis:

>1) F-structures are less cluttered.  

	Hmmm.  But they're prettier and more uniform.

>2) If you decide to believe the above arguments from anaphora 
>   and copula constructions (arguments A2 and A3), then you also
>   commit yourself to having all nouns (and PPs) have subjects.  
>   In practical terms this means that you must have a disjunction
>   in all your noun entries:  
>       
>           Hans<Subj>   
>   and     Hans 
>
>   That is because you will not always want your noun to come
>   with a subject ---- you actually only want it in predicative
>   contexts, but not in:  Hans sees a dog. 

Yes, and there is evidence for this distinction.  I refer
to the anonymous ftp-site where a draft of my chapter on predication
is sitting as a ps file FOR A LIMITED TIME ONLY TO INTERESTED READERS
OF THIS MESSAGE: ftp-csli.stanford.edu/linguistics/otw/ch11.ps.

>3) Having overt subjects in APs, PPs and NPs tends to lead you
>   to the question of what the semantics of things like
>
> 
>    Hans saw the Hamster in the cage 
>
>   are.  That is, "in the cage" now has a SUBJ.  Is this SUBJ
>   controlled by Hamster?  Or is it controlled by Hans??  Who
>   is in the cage?  

And an interesting question this is.  Some discussion of the analytic
range of PPs in English is given in the above ftp reference.

In her conclusion, Miriam tries to suggest that all this is a matter
of ideology and religious belief about semantics vs. syntax.  But I
don't think so...  I think it is a classic question that recurs in
semantics as well as syntax: what is the nature of predication?  But who
can tell us the answer?  The hardy field linguists with their dirty
notebooks out in the bush working on the morphosyntax of languages
into which Montague has not been translated?  Or the cool
technolinguists with their pages of bristling formalisms and
fourth-hand idealizations of quasi-linguistic data?  Or the
philosophers who remember some Latin and Greek and in linguistidcs
read only, maybe, Chomsky?

We can be grateful to Miriam for reraising these classic issues and
even reinventing the rote Hering, which makes us think.

Joan Bresnan