[Lingtyp] 1992 LSA presidential address

Sat Dec 21 18:42:36 UTC 2019

Follow-up to my note of the 19th about the handout for my 1992 LSA Presidential Address (delivered in January 1993 at the Biltmore in Los Angeles): Luigi Talamo at the Univ. of the Saarland -- from this list -- has managed to find and use a copy of the software that the handout was written in and produce a .pdf of the things, formatted just as in the original. A text version of that:
[A posting on my blog will eventually appear.]

                                                                                    LSA – 1/9/93
                                                                                    Arnold M. Zwicky
                                                                                    Ohio State & Stanford

Mapping the Ordinary into the Rare
Basic/Derived Reasoning in Theory Construction

(1) The ‘elsewhere’, general, or ordinary case is treated as basic, and the constrained, special, or rare case as derived from this, via a mapping rule.

(2.1) English syntax: The Auxiliary + Subject + Complement (SAI) type of finite clause (Had they realized the problem,...; Had they realized the problem?), is analyzed as derived from the Subject + VP Predicate (SVP) type of finite clause (main clauses like They had realized the problem, conditional clauses like If they had realized the problem,..., relative clauses like anyone who had realized the problem, complement clauses like I knew that they had realized the problem and I wondered when they had realized the problem, etc.).

(2.2) German morphology: The past participle form without prefix ge-, used for stems with unstressed initial syllables (verletz-t ‘injured’, ertrunk-en ‘drowned’, posaun-t ‘trumpeted’), is analyzed as derived by morpheme deletion from the form with the prefix, used for the bulk of verb stems in German, which are initially stressed (ge-hol-t ‘fetched’, ge-trag-en ‘carried’).

(2.3) (American) English phonology: The glottal stop allophone of /t/, which occurs between a stressed syllabic and (unstressed) syllabic n (cotton, button, and mitten), is analyzed as derived from a basic voicless alveolar stop allophone (top, stop, pot, spittoon).

(3)  Assumption that significant generalizations about linguistic form are stated with respect to the basic (rather than the derived) alternants.

(4)  The system that is (statistically) modal, the ordinary case, is treated as universally basic; any system, within a particular language, that is unusual, an instance of the rare case, is treated as derived from this, via a mapping rule.

(5.1) Halle’s (1990) ‘abstract morphemes’ in cases of cumulative and zero exponence in inflectional morphology.

(5.1.1) Grammatical-category content without any real phonetic shape: Plural for English sheep

(5.1.2) Phonetic material that cannot be divided into parts in a principled way but codes two or more grammatical categories: –e of Latin Voc Sg ami:ce for ami:c-us ‘friend’

(5.2) Assumption that all morphology (derivational as well as inflectional) is concatenative, combinatorial.

(5.3) Sadock’s (1991) (covert) assumption that clitics with the semantics of clausal conjunctions have the syntax of such conjunctions. ‘Let us assume that this lexeme has just the syntactic properties of an ordinary conjunction in Latin, but that it also is subject to an independent morphological requirement that it attaches as a suffix to a fully inflected word.’ (63)

(5.4) Assumption that all clause organization, even in VSO and OSV languages, involves the combination of a Subject and a (VP) Predicate.

(6)  One-to-one association of meaning and form: ambiguity (I like Pat better than Chris) and paraphrase (move more quickly than did the penguins, move more quickly than the penguins did).

(7)  Prediction of determined properties of expressions: subcategorization, government, grammatical agreement, anaphoric agreement, etc. (She sees penguins).

(8)  Relations of presupposition between rules: generalization A presupposes that the conditions in generalizations A1 through An have been satisfied (SAI presupposes that conditions on VPs consisting of Auxiliary Verb + Complement have been satisfied and that conditions on clauses consisting of a Subject + VP Predicate have been satisfied).

(9)  Virtual time: the rules for the presupposed requirements apply ‘before’ the rules for the presupposing requirement.

(10) Determining that an integer m is a perfect square, using the fact that (n+1)2 = n2+2n+1 = n2+O(n+1), where O(i) is the ith odd integer:

(10.1) Start with 1 and add the other odd integers in succession (3, 5, 7,...); if the result at any stage is m, then m is a perfect square, but if the result is an integer greater than m, then m is not.

(10.2) Start with m and subtract the odd integers in succession (1, 3, 5,...); if the result at any stage is 0, then m is a perfect square, but if the result is a negative number, then m is not.

(11) Checking an expression (Are penguins flying?) for wellformedness according to a grammar

12.1) Virtual time again: Default requirements are imposed first, then the overriding requirements override these, and so change the properties of an expression that is being analyzed.

(12.2) Alternative: default/override relationship holds between rules, or (equivalently) between the constructions that rules describe. If we’re checking the wellformedness of a clause, e.g. Can penguins fly?, we ask if the construction that the clause exemplifies – Yes-No Question, here – is one that stipulates SAI. If, as in this case, it is, the requirements of SAI must be satisfied; if it is not, then the requirements of the default rule, SVP, must be satisfied.

(13.1) Resulting theory is more restrictive – since every language now has the same system (or at least has one of a smaller number of systems).

(13.2) The resulting theory involves fewer ontological commitments, so that Occam’s razor is on its side even if the set of possible languages is not constrained.

(14.1) Open the door to the full complexity of derivational frameworks for linguistic description.

(14.2) Such analyses are gratuitous.

(15) Marsupial mammals are the special case vis-a-vis placental mammals.

(16.1) Laqueur (1990): ‘Language constrained the seeing of opposites and sustained the male body as the canonical human form. And, conversely, the fact that one saw only one sex made even words for female parts ultimately refer to male organs. There was in an important sense, no female reproductive anatomy, and hence modern terms that refer to it – vagina, uterus, vulva, labia, Fallopian tubes, clitoris – cannot quite find their Renaissance equivalents.’ (96)

‘...neither were there two sexes juxtaposed in various proportions: there was but one sex whose more perfect exemplars were easily deemed males at birth and whose decidedly less perfect ones were labeled female.’ (124)

(16.2) Boswell (1989): Type A, in which ‘all humans are polymorphously sexual’; Type B, which posit a small number of fixed types (e.g., heterosexual, homosexual, bisexual); Type C, which ‘consider one type of sexual response normal (or “natural” or “moral” or all three) and all other variants abnormal (“unnatural,” “immoral”).’ (23)

‘Type C theories are essentially normative rather than epistemological, but borrow from both sides of the universals question in assuming, by and large, that people are born into the normal category but become members of a deviant grouping by an act of the will, although some Type C adherents regard “deviants” as inculpably belonging to an “abnormal” category through mental or physical illness or defect.’ (23f.)

(17) Rather than taking the ordinary cases as a sort of ground state, theorists should insofar as possible devise their frameworks so as to embrace equally the full range of relevant phenomena.

(18) relative frequency and extratheoretical considerations (e.g., ease of processing, production, or learning)

References

Boswell, John. 1989. Revolutions, universals, and sexual categories. Martin Duberman, Martha Vicinus, and George Chauncey, Jr. (eds.), Hidden from history: Reclaiming the gay and lesbian past (New York: New American Library), 17-36.

Goldsmith, John A. 1976. Autosegmental phonology. Ph.D. dissertation, MIT.

Halle, Morris. 1990. An approach to morphology. NELS 20.1.150-84.

Janda, Richard & María Sandoval. 1984. “Elsewhere” in morphology. Bloomington: IULC.

Laqueur, Thomas. 1990. Making sex: Body and gender from the Greeks to Freud. Cambridge MA: Harvard Univ. Press.

Lieber, Rochelle. 1987. An integrated theory of autosegmental processes. Albany: SUNY Press.

McCarthy, John J. 1979. Formal problems in Semitic phonology and morphology. Ph.D. dissertation, MIT.

McCarthy, John J. 1981. A prosodic theory of nonconcatenative morphology. LingI 12.3.373-418.

Pullum, Geoffrey K. & Arnold M. Zwicky. 1991. Condition duplication, paradigm homonymy, and transconstructional constraints. BLS 17.252-66.

Pullum, Geoffrey K. & Arnold M. Zwicky. 1992. A misconceived approach to morphology. WCCFL 10.387-98.

Sadock, Jerrold M. 1991. Autolexical syntax: A theory of parallel grammatical relations. Chicago: Univ. of Chicago Press.

Zwicky, Arnold M. 1986. The general case: Basic form versus default form. BLS 12.305-14.

Zwicky, Arnold M. 1989. What’s become of derivations? Defaults and invocations. BLS 15.303-20.