Abstracts

Present:

Disglossia in Korean Language

The most salient features of Korean language have arisen from the tension of diglossia that had prevailed for more than a millenium until the end of the 19th century. Diglossia refers to a sociolinguistic situation in which a community uses one or more languages, one of which is often considered superior to the other(s). In the past fifty years, English has replaced Chinese as the 'high' language of the new diglossia. However, the current situation is much more complex due to the conflict between globalization and local culture.

Future:

Past:

Zsuzsa Nagy, Rutgers Dept of Linguistics

Morphology Competes with Syntax: Universal but Violable Order of Functional Projections

In a recent book, Guillelmo Cinque proposed what can be called the Universal Order Hypothesis (UOH) "The ordering of functional projectional layers in a clause reflects a universal hierarchy" (Cinque 1999). Cross-linguistic data, however, do not always support this hypothesis. In this talk I address two common types of problems, the reversal and the doubling of functional projections. I show how the Optimality Theoretic interpretation -- as a universal, but violable constraint -- is a more accurate interpretation of the UOH. In two case studies morphological constraints compel violations of the UOH.

One pair of universally ordered functional projections in Cinque's system is future and past; Cinque states that the projection of 'future' is lower than the projection of 'past'. The Mirror Principle (Baker 1985) then predicts that the future morpheme is closer to the verb root than the past tense morpheme in all languages with overt morphemes for these two projections. In Temein, a language spoken in Ethiopia (Tucker and Bryan 1966), however, the reverse order is found. Closer examination of the data reveals that the two morphemes belong to different morphological categories: 'future' is an auxiliary, while 'past' is an affix. I argue that these morphological properties trigger the reversal of the projections. The combinations of two factors: that the future auxiliary stands on its own when the 'future' projection is outermost in the verbal projections and that the past tense morpheme is picked up by the lexical head rather than by another functional morpheme, make the tense-reversal candidate optimal in comparison to the other competitors.

Another case in which the UOH is violated in favor of a morphological constraint is observed in Tagalog. One of several aspectual prefixes of the language undoubtedly occurs separated from the others. In an influential paper Travis has proposed that the position of this aspectual morpheme provides direct evidence for a second, VP-internal Asp position (Travis 1991). Examining the morphology of the language once more provides a different view of the same facts. The aspectual morpheme in question is a reduplicant. Since reduplicants stay very close to their base (see (Yip 1999) and (Nelson 2001)) I argue that the reduplicant aspectual affix shifts from the canonical Aspect position closer to the base of reduplication, the verb root. The existence of this second Aspect projection in Tagalog, however, is not evidence for a universal position, but a language-specific result of constraint interaction.

William Labov, University of Pennsylvania, Dept of Linguistics

Is Linguistic Knowledge Useful Knowledge?

In more than one context, linguists have found it difficult to prove that knowledge of lin-guistic structure is useful for solving social problems connected with language. This is especially true for literacy. Linguistic knowledge has not played a significant role in the mainstream of reading research, even for those scholars who emphasize the importance of developing phonemic awareness. In the light of the persistent minority differential in reading achievement, it is even more striking that the extensive research on African American Vernacular English has not been used to attack the problem. This report will deal with the application of knowledge of African American Vernacular English [AAVE] to raising the reading levels of inner city struggling readers. A linguistic analysis of reading errors was used as the basis for an Individualized Reading Program, adapted to the concerns and interests of African American children. The program begins with a diagnostic reading that shows which types of onsets, nuclei andcdas create the greatest decoding problems for each student, and directs the student to those sections of the manual that provide the missing knowledge and training. For consonant clusters, a high correlation was found between difficulty in reading and simplification in spontaneous speech. Results so far show that phonemic awareness can not improve decoding skills for those codas that have the highest rates of simplification; more abstract training in morphophonemic awareness is required.

Bruce Tesar, Rutgers Linguistics Dept.

Overcoming Ambiguity in Language Learning

One of the central problems in human language learning is ambiguity in the overt data available to the learner. An overt form (the audible portion of an utterance) may be ambiguous between more than one full linguistic analysis, and different languages may choose different analyses. Because linguistic principles evaluate entire linguistic analyses (and not just overt forms), the learner must overcome this ambiguity in order to use the overt forms to learn the grammar of their native language.

This talk will present a learning algorithm, the Inconsistency Detection Learner, capable of learning from ambiguous data. This research uses the linguistic framework of Optimality Theory, in which languages are defined in terms of optimization over a ranked set of violable, universal constraints. The primary goal of the learner is to learn the language-specific ranking of the universal constraints. The learner overcomes ambiguity by temporarily considering more than one analysis of the ambiguous form, and then eliminating those analyses which are inconsistent with other data from the language. The simulation results demonstrate that this strategy can succeed without explosive growth in the number of grammatical hypotheses the learner has to maintain during the course of learning.

The key to the success of this algorithm is the interrestrictiveness of the data. Even though the overt portions of two words each admit multiple analyses, the learner will commonly be forced to a particular analysis of the first word in order to have a grammar capable of producing the second word. In isolation, each word is ambiguous, but collectively, the words of the language force the learner to a single grammar (and thus a single analysis of each word). When working within Optimality Theory, the interests of linguistic theory and computational language learning align: constraint interactions which are simultaneously implicated in the analysis of many words make for better linguistic theory, and also make for more efficient learning.

Nicole Nelson, Rutgers Linguistics Dept.

Exploring an asymmetry: left vs. right edges in reduplication

Dr. Judit Hajnal, Rutgers Institute for Hungarian Studies and the Rutgers University Program in Slavic and East European Languages and Literatures

The Structure and History of Hungarian

Hungarian, the official language of the Republic of Hungary, often regarded as an exotic language, is spoken natively by 10 million Hungarians in the country and by approximately 5 million people outside Hungary. A considerable number of native speakers live in the neighboring countries, due to historical reasons. With its total number of native speakers, Hungarian language takes fiftieth place in the world and the thirteenth in Europe. Its one of the few European languages which does not belong to the Indo-European language family, it is not related genetically to any of the Germanic, Romance or Slavic languages. Its cognate languages are Finnish and Estonian.

Hungarian belongs to the agglutinative languages. The morphology of Hungarian is characterized by an abundance of case suffixes and variant forms of stem. The word form szeretlek meaning "I love you", for example, conveys the meaning I (subject) szeret love (action) you (object). Word forms and expressions in Hungarian can express subtle shades of meaning; they are exact because the basic meaning of the word is well-defined.

The presentation will also point out how much Hungarian vocabulary shows evidence of all the historical events, changes and traumas the nation went through in the past. It will present the first written record of Hungarian (Funeral Oration), and the first poem written in Hungarian (Old Hungarian Mary's Lament) and it will explain how loan words of a more recent origin were borrowed from well-known European languages. It also focuses on the language reform to develop a standardized national language and a modern literary Hungarian with further movements to achieve national independence introduced.

The presentation will also call attention to the versatility of Hungarian: it makes language suitable for classical and accentual versification alike, which makes translation of Hungarian poetry extremely difficult.

February 14, 2001, 8PM

Ron Artstein, Graduate Student, Rutgers Linguistics Dept.

Conjunction weakening and morphological plurality

Conjunction weakening (Winter 1996, 1998) happens when two incompatible predicates are conjoined, as in (1). Rather than being contradictory, the sentence receives a ``weak'' reading which does not entail (3); this is in contrast to the ``strong'' reading of (2), which does entail (3).

(1) The birds are above the cloud and below the cloud.
(2) The birds are above the house and below the cloud.
(3) The birds are below the cloud.

I show that morphological plurality limits the possibility of conjunction weakening. Data from Hebrew, where predicates are often marked for number, show that predicates that are morphologically plural require a plural extension even when thus weakened. The talk will explore the consequences of this observation with respect to the representation of plurality and conjunction.

References:

Professor William C. Dowling, Rutgers English Department

How Literature Means: a Critical Discussion of The Senses of the Text

"My argument is that..."meaning" as it exists for literary theory does not depend on interpretive communities, or social conventions, or language games, or ideological systems, or anything else that is, as one could say in a more innocent age, outside the text. In the last analysis, the argument would be that meanings create interpretive communities rather than the other way around."
  --W.C. Dowling in: The Senses of the Text: Intensional Semantics and Literary Theory)

Is there a "right" and a "wrong" when it comes to interpreting literature? Does it makes sense to say that a line in a poem or text means one thing rather than another? Intuitively, we feel that there are some objective constraints on how a person may rationally interpret a work, and that for some works, there is a "true" interpretation. However, this idea is currently rejected by many approaches to literature interpretation.

In his book "Senses of the Text," Rutgers University Professor of English William C. Dowling defends the existence of determinate meaning in literature against the various arguments to the contrary. Dowling explains the program of "New Intensionalism," as devised by philosopher and linguist J.J. Katz, and how it may be used to defend a notion of determinate meaning in literature, one in which the meanings of expressions have a necessary existence independent of any act of interpretation.

Professor Dowling has agreed to meet with the Linguistics Club to discuss "Senses of the Text," determinacy of meaning in literature, and the relationship between linguistics and literature studies. The catalyst to this discussion will be three presentations by members of the linguistics club on the main ideas of "Senses of the Text." These presentations are intended to provide members who have not read Professor Dowling's book with the background needed to join in the discussion. Professor Dowling will then respond to these presentations and critique them. In this way, a general discussion of the work will develop.

All members of the linguistics club are encouraged to attend this meeting. One needs not have read Professor Dowling's book in advance to find these discussions comprehensible and interesting. It is also an exciting opportunity to discuss a work with its author, and, of course, may ignite an interest previously unrecognized.

(Above abstract written by Seth Cable)

Professor Matthew Stone, Rutgers Computer Science Department

Disfluent Speech

It is only with the utmost persistence and inspiration that an analyst can reconstruct the flow of a dialogue from a word-by-word transcript; the raw materials read like half-finished jumbles of thoughts piled atop one another, rather than orderly, complete, grammatical sentences. In fact, if it seems like you had to be there to understand it, you may be right. Maybe cognitive science can shed some light on the confusion. The evidence suggests that a transcript is just an incomplete record of the creative way participants in the conversation explored, refined and ratified their communicative intent, by using their knowledge of grammar, the environment and one another, and by flexibly adapting their commitments and responsibilities to talk. My presentation will explain some of the reasoning behind this from research over the last few years, as well as the many open questions that remain.

November 2, 2000:

Akin Akinlabi, Rutgers University, Dept. of Linguistics

The African continent offers a generous sample of the great variety of phonological systems to be found in the worlds languages, as well as some original features of its own. African phonological systems range from the relatively simple to the staggeringly complex. Those on the more complex end of the spectrum contain phonemic contrasts little known elsewhere in the world, rich patterns of morphophonemic alternation, and intricate tonal and accentual systems, all offering stimulating grounds for phonetic and phonological study. Largely in response to challenges raised by African language data, phonological theory over the last three decades has attempted to develop more adequate models in such areas as feature theory, syllabification, vowel harmony, and tonal phonology, to name just a few.

This presentation offers a synthesis of recent work on African language phonology, focusing especially on their phoneme inventories, feature contrasts, syllable structure, and tone. It should be emphasized that I will do this at a very elementary level.

October 5, 2000:

Jose Camacho, Rutgers University, Dept. of Linguistics

Expressions of Modality in Spanish

Many languages have a grammatical morpheme ("validation marker") that conveys the speaker's attitude towards the validity of the statement s/he is making. Other languages also express this information, but through different grammatical means (adjectives, adverbs, verbs, etc.) In this presentation, I will explore the syntactic properties of certain validation constructions in Latin American Spanish (LAS). Although LAS does not have grammatical morphemes specialized in validation, the words that express validation have a very specific syntactic distribution that needs to be explained: they cannot appear with constituents outside the VP (VP-external subjects), only with constituents inside the VP (Direct objects, Indirect Objects, PPs, VP-internal subjects). Additionally, there is a semantic restriction: validators can only appear with constituents interpreted indefinitely.

I will also present two hypotheses regarding this distribution: 1) validation markers follow a general strategy for mapping quantificational material, and 2) the distribution of validation markers correlates with the discourse status of constituents.

April 27, 2000: (Note: this meeting will begin at 9:15PM).

Veneeta Dayal, Rutgers University, Dept. of Linguistics

Scope Marking and the Syntax-Semantics Interface

Consider two distinct syntactic structures, which appear to have the same meaning. There are two plausible approaches one might take. One is to treat them as syntactically alike at an abstract level of syntax, thereby allowing the same interpretive procedure to apply to both. Or one might maintain the syntactic difference between the two and posit distinct but converging semantic interpretations.

In this talk I look at extraction and scope marking structures, illustrated by German (1a) and (1b) respectively, from this perspective and discuss the theoretical and empirical consequences of the two approaches.

1a. Mit wem glaubt Karl dass Maria gesprochen hat

    with whom thinks Karl that Maria spoken   has

b. Was glaubt Karl, mit wem Maria gesprochen hat

    what thinks Karl with whom Maria spoken   has

    'Who does Karl think Maria has spoken to?'

April 13, 2000:

Jim McCloskey, University of California at Santa Cruz

Working on Irish Syntax: The Graveyard Shift?

Irish is simultaneously the official language of a first-world nation-state, and an embattled minority language fighting for its life. That it has survived at all into the 2nd millennium as the spoken language of a community is a small miracle. The curious double status that the language enjoys creates both opportunities and dilemmas for the working linguist. The dilemmas are both ethical and methodological. Nevertheless, the interplay between ths language and linguistic theorizing over the past 20 or so years has arguably been beneficial for both. This talk will try to give a sense of what the difficulties and dilemmas are and also a sense of what has been accomplished.

April 6, 2000:

Matthew Stone, Rutgers Computer Science Department

Issues and Results in Corpus Linguistics

As speakers, we know a lot of conventions about how to fit the constructions of our language into the ongoing context of a conversation. Introspection about these pragmatic abilities is difficult; unlike our often crisp intuitions about syntax or semantics, our intuitions on questions like "would it be natural to say that now?" are vague and slippery (and perhaps even flat-out wrong) in many cases of interest.

A number of corpus linguistics projects are under way at Rutgers -- mainly on the more practical side of things. In fact, paid positions are open for corpus annotators!

March 23, 2000: Kent Johnson, graduate student in the Philosophy department will be speaking on distinguishing various projects in the study of language - what different people think theories of language (particularly theories of language meaning) are for. . .

March 2, 2000: A brief history of linguistics: Professor Ken Safir will be giving a talk on the history of linguistics Thursday at 9 pm.