NetNews Usenet Archive 1992 #27/NN_1992_27.iso/spool/sci/lang/8138

Text File | 1992-11-20 | 13.7 KB | 345 lines

Newsgroups: sci.lang
Path: sparky!uunet!zaphod.mps.ohio-state.edu!pacific.mps.ohio-state.edu!linac!uchinews!news
From: schiller@whorf(Eric Schiller)
Subject: Re: Autolexical Syntax: Interface
Message-ID: <[e-mail]>
Sender: [e-mail] (News System)
Reply-To: [e-mail]
Organization: University of Chicago Computing Organizations
References: <[e-mail]>
Date: Fri, 20 Nov 1992 22:53:37 GMT
Lines: 332

In article <[e-mail]> [e-mail] (Peter
Rossen Skadhauge) writes:
> I am working on a small project on the theory of autolexical
> syntax as described in Jerrold M. Sadock: Autolexical Syntax -
> A theory of Parallel Grammatical Representations. If somebody
> has heard about articles describing elaborations and formalizations
> of the General Matching Principle and Constructional Integrity
> condition - please let me know!
>
There's an awful lot! The most recent formulations can be seen
in a variety of papers, and there is a lot of discussion in the
forthcoming Mouton volume Autolexical Syntax, ideas and methods,
which I have more-or-less finished editing.

Among the more important works are Dave Kathman's paper on
control, and various papers in CLS 26, and 28 (not out yet),
also a paper on Yiddish in the McCawley Festschrift (1992).

A few selections:

Eilfort, William and Eric Schiller. 1990. "Pragmatics and Grammar:
Cross-Modular Relations in Autolexical Theory." CLS 26.1. 125-136.
Chicago:
Chicago Linguistic Socitey, 1990.

Faarlund, Jan Terje. 1989. "Autostructural Analysis." In Autolexical
Syntax:
Ideas and Methods, ed. Eric Schiller and Elisa Steinberg. Berlin: Mouton,
to
appear.

Graczyk, Randolph. 1991. "Incorporation and Cliticization in Crow
Morphosyntax." PhD, University of Chicago, 1991.

Kathman, David.  1992. "Control in Autolexical Syntax." In Autolexical
Syntax:
Ideas and Methods, ed. Eric Schiller & Elisa Steinberg.  Berlin: Mouton,
to
appear

Lapointe, Steven G. 1987. "Some Extensions of the Autolexical Approach to
Structural Mismatches." In Syntax and Semantics, Volume 20: Discontinuous
Constituency, ed. Geoffrey J. Huck and Almerindo E. Ojeda. 152-184. 20.
Orlando: Academic Press, 1987.

Leer, Jeff. 1991. "Schetic Categories of the Tlingit Verb." PhD,
University of
Chicago, 1991.

Sadock, Jerrold M. 1983. "The necessary overlapping of grammatical
components."
CLS 19.2 (1983): 198-221.

Sadock, Jerrold M. 1986. "An autolexical view of pronouns, anaphora, and
agreement." UCWPIL 2 (1986): 143-164.

Sadock, Jerrold M. 1985. "Autolexical Syntax: A Proposal for the Treatment
of
Noun Incorporation and Similar Phenomena." NLLT 3 (1985): 379-439.

Sadock, Jerrold M. 1992 "A Paper on Yiddish for James D. McCawley." In The
Joy
of Grammar, ed. Diane Brentari, Gary N. Larson, and Lynn A. MacLeod.
323-328.
Amsterdam: John Benjamins, 1992.

Schiller, Eric. 1989a. "Syntactic Polysemy and Underspecification in the
Lexicon." BLS 15 (1989): 278-290.

Schiller, Eric. 1989b. "The Case for Autolexical Case." In Autolexical
Syntax:
Ideas and Methods, ed. Eric Schiller and Elisa Steinberg. Berlin: Mouton,
to
appear.

Schiller, Eric. 1990. "Focus and the Discourse Dimension in Autolexical
Theory." ESCOL 7 (1990):

Schiller, Eric. 1991. "An Autolexical Account of Subordinating Serial Verb
Constructions." Ph.D., University of Chicago, 1991.

Schiller, Eric. 1992. "Infixes: Clitics at the Morphophonological Level."
CLS
28 (to appear):

Schiller, Eric & Barbara Need. 1992. "The Liberation of Minor Categories:
Such
a Nice Idea!" CLS 28 (to appear):

Schneider, Robinson. 1989. "Toward a tri-modular analysis of -ly adverbs."
In
Autolexical Syntax: Ideas and Methods, ed. Eric Schiller & Elisa
Steinberg.
Berlin: Mouton, to appear.

Smessaert, Hans. 1988. "An Autolexical Syntax Approach to Pronominal
Cliticization in West Flemish." M.A., University of Chicago, 1988.

Smessaert, Hans. 1991. "Pronominal Cliticization in West Flemish." In
Autolexical Syntax: Ideas and Methods, ed. Eric Schiller & Elisa
Steinberg.
Berlin: Mouton, to appear.

Copies of the unavailable or difficult to find papers can be
obtained from me, since I have used these and others in a course
I am teaching on Autolexical Theory. A rough draft of an e-mail
synopsis can also be sent.

We are setting up an FTP site for the ASCII stuff, but the papers
are often long, and xeroxing is not cheap. But if anyone needs
the papers urgently, I can arrange to make copies for the cost
of xeroxing and postage.

Here are some informal, state-of-the art notes (which is to say,
my course notes for Fall quarter 1992):

* Chapter 10: The Interface

** 10.1 The Interface Conditions

The Interface is where the independent representations of the various
components are compared, to make sure that any discrepancies fall
within the acceptable bounds of natural language. It is here that some
universals of human language will be formulated. It seems, for the
moment, that we only require three simple statements:

1. If two constituents are part of a larger constituent on one
dimension, they should be part of a corresponding constituent on each
other dimension, and at least one element on one dimension should be
assoociated with a head on another dimension.

2. The order of constituents should be the same on each dimension.

3. There are default relations which hold between constituent types on
various dimensions.

These are informal versions, since Autolexicalists are still pinning down
a
wording that we all feel comfortable with. The first condition has
been known as the Constructional Integrity Constraint (CIC), and the
second as the Linearity Condition (LC). Here are some more formal
statements of these principles:

Constructional Integrity Constraint:
    If an element combines with a phrase P on one dimension and with
element E on another dimension, then E must be associated with the head of
P.

Linearity Constraint: The associated elements of two dimensions must occur
in the same linear order.

Notice that we use the word 'should' in these statements. The point is
that language violates these principles quite frequently. The
interesting thing is that language never seems to violate more than
one of them at a time. This leads to the odd situation that our
principles must be lenient, allowing for violations to take place. But
(see Sadock 1991 for details), it seems to be the case that when one
principle is violated, the others are enforced with great rigor. Of
course we cosuld reword things so that each rule permits exactly one
violation per constiutent if and only if the other two rules were
obeyed, but that would lose the flavor of our framework, which is
motivated by the fact that language involves widesperad discrepant
representations, and that these discrepancies are not "abnormal".

Let us consider a typical violation of each rule.

@a) The wolf's at the door.

Here the problem is with "wolf's". Phonologically and morphologically,
the contracted auxiliary "s" is part of a single word. But in the
syntax, "wolf" and "s" are not only distinct items, but they belong to
two different constituents, the noun phrase "the wolf" and the verb
phrase "(i)s at the door". The phonological motivation for the
cliticization is hardly surprising, and auxiliary contraction is very
common in English.

While the previous example does violate our first rule, it complies with
the
second one. Now let us consider a violation of the latter rule. Inversion
of
auxiliaries in questions is a good example, but would require considerably
more discussion than we can supply here, so I will present instead our
canonical example from Latin, which is very clear. Latin has a suffix
"que" which is a conjunction. But instead of going between the conjoined
items as is the case in English, it comes at the end of the second
element:

@a)  puellae puerique 'girls and boys'

Here the linear order of the elements does not correspond to the
expected order, which would be

@b) puellae et pueri  'boys and girls'

which is another way of saying the same thing, using a syntactic
particle instead of the suffix. But all three elements belong to a single
constituent, conforming to our first rule. All enclitics (clitics that
are prefixes rather than suffixes) show the same discrepancy between
morphology (where they are prefixes) and syntax (where they would be
expected to follow the word).

The strong relationship between semantic content and syntactic cagtegory
has
been well studied (e.g., Croft, 1991). We expect that verbs will represent
actions, that adjectives will be properties, as will intransitive
verbs and nouns. But again we find a lot of exceptions (which are rarely
discussed by scholars concentrating on the expected relationships). Our
typical example is the triple "seem, likely, probably", three words with
very similar semantic functions, but the first is a verb, the second an
adjective, and the third an adverb. While we would be very happy if we
could predict the syntactic category of a word from its semantics, natural
language thwarts us. Often these discrepancies can be explained by
historical facts, but a synchronic account does not seem possible.

Exercise XX: Diagram the syntax and morphosyntax of the following
sentences and note any structural discrepancies between the dimensions.

a) Kusaiean

Twetwe-kihn-iyen yot uh eneuh kof
'Sharpening with the stone requires water'

twe:    'sharp'
kihn:    transitivizing instrumental affix
iyen:    nominalizing affix
yot:    'stone'
uh:    'the'
enenuh:    'requires'
kof    'water'

Sadock (1991:69) discusses the possible violations of the Linearity
Constraint and Constructional Integrity Constraint as applied to prefixes
and suffixes. I will expand his work here to cover additional
morphophonological instantiation mechanisms, including infixes,
circumfixes and reduplication.

In theory, each of our five types (prefix, infix, suffix, circumfix,
reduplication) could violate either one of our two conditions, or neither,
but not both of them. But in fact many of these combinations are not
attested. Some are simply logical contradictions. If a lexeme is part of a
syntactic constituent XP, it can come before it, or after it.
Morphologically, it can occupy any of the five positions. Sadock 1991
assumed that affixal lexemes must be either prefixes or suffixes, but did
not deal with the other three types, and also failed to distinguish
betwseen morphophonology and morphosyntax. Here we will be treating the
location of the affix at the level of morphophonology, and dealing with an
interface to the syntactic dimension.

..

** 10.2 Feature Checking

The checking of feature compatibility is also part of the Interface.
Unlike derivational theories with feature-passing mechanism, the
autolexical approaches involve static representations. Features are
normally matched in the various dimensions via the same sort of
correspondences that associate categories. So if a language has
morphological gender marking on nouns which triggers agreement on
verbs, it is expected that there will be some sort of semantic or
syntactic features which parallel these morphological markers. The
same holds for person, number, and other grammatical devices. But as
we know from English, this match is far from rigorous (consider simple
agreement problems involving morphologically singular words such as
'data' and 'media', as well as coordination and proximity problems).

A complete analysis of a linguistic expression requires first of all
that the various phenomena seen in the language be localized on their
proper dimensions. Such notions as transitivity, ergativity and
perfectivity have often been treated as syntactic phenomena. It is not
clear, however, that these notions play a direct role in the syntax in all
languages. They may be purely semantic factors.

Exercise: Analyze the morphosyntax, syntax, and logico-semantics
(with case features) of the following data, also considering the
proper treatment of the honorific marker.

@) Written Tibetan (Mi.la.ras.pa'i rnam thar, Saxon 1991)

The ergative marker occurs when the verb is in the perfective in
simple clauses (1,2), with agent-taking intransitive verbs in the
perfective (3) but not with a similar verb in the nonperfective (4) or
with non-agentive verbs in the perfective (5).

1.    yul    me    rnams-kyis    bzung        ste
    country    man    PL - ERG    take(PERF)    CLP
    The countrymen took (him).

2.    a.ma    nang-na        yos     rngod        kyin   yod-pa-s
    mother    house-LOC     grain    parch(IMPF) while  be-NOM-INST

    thos        nas
    hear(same)    CLP

    While roasting grains at home, mother heard (my singing).

3.    nged-rnams-kyis    Dbus.gtsang-gi    lam-du    zhugs-nas
    I-PL-ERG    Ue-Tsang-GEN    path-LOC begin(PF)-CLP

    phyin-pa-s        Gtsang        rong-gi        G.yag
    set out(same)-nom     place-name    center-GEN  place-name

    sde    zhes    bya.ba-r    sleb-s
    region    (THUS)    called-INST    reach-PERF

    We took the road to Ue and Tsang and arrived at Yakde (in the
valley of Tsangrong).

4.    nga    yang    mthu        slob-tu        'gro-ba-s
    I    also    blackmagic    teach-IMPF-LOC  go(IMP)-NOM-INST

    rang.re-rnams    bsdebs    'gro        byas-pa-s
    you(HON)-PL    together go(IMPF)    do-NOM-INST

    I, too, would like to go with you to learn black magic.

5.    nga    ni    yid-cig    ma    che-s        te
    I    TOP    be(IMPF)NEG    very great-PERF    CLP
    I was not great.

In any event, Jerry Sadock and I will be happy to answer any
specific or general questions. I hope to have a distributable form
of the e-mail synopsis ready by the end of the year.



--
***** ** *** **** ***** ** *** **** ***** ** *** **** ***** ** *** ****
Sometimes my head is in the clouds of theory, but often it is just a fog.
My feet, meantime, tend to stay more or less on the ground. I think.
***** ** *** **** ***** ** *** **** ***** ** *** **** ***** ** *** ****
Eric Schiller                              [e-mail]
Dept of Linguistics, Univ. of Chicago      [e-mail]