Harvard University: Indo-European Workshop
Postponed
Bert Vaux
Harvard University"There was and there wasn't a Balkan Sprachbund" All are welcome, and refreshments will be served.
For more information, contact ptaylor@fas.harvard.edu
MIT Phonology Circle
Andrea Calabrese Title TBA
For more information, contact czoll@mit.edu.
MIT Linguistics Colloquium
Donca Steriade
UCLA"Context-dependency similarity and the Romanian semi-rime"
For more information, contact bissell@mit.edu
"English dialects today"
Prof. Bert Vaux of Harvard University
Abstract: Existing studies of English dialects are based almost entirely on materials collected from older speakers in the 1930s through the 1960s; moreover, these studies generally focused on the identification of obscure and archaic vocabulary items. For this reason it is easy to find discussions of whiffletrees, stone boats, ground hackies, and a rich variety of plow parts and types of broom, but it is impossible to find information about how regular people actually speak today. In this talk I discuss the sorts of linguistic tricks one can use to identify the geographical provenance of a random person who seems to speak "standard" American English. We'll consider roly polies, sunshowers, soft drinks, pins and pens, and the Three Maries, and we'll also touch on issues of crypto- and schizo-prestige, problems of defining "standard" English, and the future of non-standard dialects such as Boston, New York, Southern, and Midwestern.
Harvard Bookstore event
Noam Chomsky
Massachusetts Institute of Technlogy"NEW HORIZONS IN
THE STUDY OF LANGUAGE AND MIND"with an introduction by Steven Pinker
Harvard University: Ford Foundation GSAS Workshop on Syntax
Balkiz Ozturk "Genitive Phrases in Turkish" (1) a. Ben-im araba-m
I-gen car-1p sg
'My car'
b. Ben-im araba
I-gen car
'My car'
In this study, the two types of genitive constructions in Turkish, given in (1), will be compared structurally. The movement and modification possibilities of each form will be discussed.
As for the genitive constructions with agreement, as in the one in (1a), it will be claimed that the genitive marked possessor DPs must have thematic relations with the head noun. Based on the modification, quantification possibilities and the movement observed in genitive constructions, it will be claimed that they act as full-fledged DPs.
As for the forms in (1b), on the other hand, it will be claimed that the genitive nouns in agreementless genitive phrases are not thematically related to the head noun, but they are used to refer to topics, which were introduced to the hearer in an earlier discourse. It will be shown that only genitive pronouns and proper names can be used, and genitive marked common names are not allowed in these constructions. It will also be discussed that it is not possible to modify the genitive nouns and no movement is observed in the case of these forms. It will be shown that the genitive nouns given in (1b) compete with the quantifiers positionally. Hence, based on the modification and movement possibilities, it will be claimed that unlike the full-fledged DP possessors in (1a), the genitive marked forms in (1b) are determiners, which are used to refer to specific-referential entities.
Colloquium
Professor Gregory Ward
Northwestern University"On the Semantics and Pragmatics
of the Identifier 'so'"
For further information, contact Prof. B. Fraser.
Author appearance
Professor Steven Pinker
MITSteven Pinker will discuss and sign his recent book:
Words and Rules: The Ingredients of Language.This event is free and open to the public. It is sponsored by the Boston University chapter of the Phi Delta Kappa Honor Society.
MIT Phonology Circle
Mahasen Hasan Abu Mansour
Umm Al-Qura University, Makkah, Saudi Arabia and MIT"An OT account of Arabic Hypocoristics"
For more information, contact czoll@mit.edu.
Harvard University: Indo-European Workshop
Haraldur Bernhardsson "Verner's Law in Gothic" All are welcome, and refreshments will be served.
For more information, contact ptaylor@fas.harvard.edu
UMass, Amherst - Colloquium
Edward Gibson, MIT "Linguistic Complexity: Locality of syntactic dependency"
For more information, see http://www.umass.edu/linguist/colloquia.html
MIT LING-LUNCH
Thursday, November 30, 2000 at 12:00, MIT E39-335 (conference room)Christer Platzack
Lund University"The Computational System as a Minimal Feature Driven Device and the Tripartite TP/VP-Hypothesis of the Universal Clause" We suggest a restricted version of the computational system, based on the feature driven device outlined in Chomsky (1999) and developed into an efficient tool for describing basic syntactic phenomena by Pesetsky & Torrego (2000). In our paper we will take a third step proclaiming that the main part of the sentence is built up as a tripartite system of adjacent TP/VP-pairs, with a single CP on top as a stop site. We will claim that the syntactic computation in the TP/VP-part is driven by two types of features, tense-features and phi-features, whereas additional features are active at CP. The features assumed (except EPP) all have a semantic value, and the general description thus falls within the Relativized Extreme Functionalism of Pesetsky & Torrego (2000).
General outline: after having presented an outline of the universal clause, the feature inventory and the operations that apply to features, we illustrate with deriving active and passive in English, Swedish and Icelandic, closely related but still different languages. Concentrating mainly on the variation related to CP in these languages, we suggest a formal account of embedding that we apply to wh-clauses and that-clauses, deriving the variation at hand. In the final part we succeed in deriving the well known syntactic differences between Ice. and Mainland Scandinavian (Vikner 1995, Holmberg & Platzack 1995), like ±V-to-I, ±stylistic fronting, ±expletive inversion, ±transitive expletives, ±oblique subjects, ±that-trace, effect, ±complementizer omission. This part also contains an account of the role morphological case is playing in Ice., including its relation to agreement and word order.
Theoretical assumptions. We standardly assume that the clause is built up by merging roots with categorial values like V, N, A or P to produce a binary branching tree with the properties of a bare phrase structure (Chomsky 1995); the Economy Condition and the Minimal Link Condition (MLC) apply as well. We only make use of two functional categories, T and C. Further ingredients include: (a) The features tau and phi, with either the value interpretable, +tau and +phi, or uninterpretable, ¬tau and ¬phi, (b) The operations on features to delete ¬F when matched with +F: Agree is the double bond relation where [+F, ¬G] in the probe match [¬F, +G] in the g oal (Chomsky 1999), Modify the single bond relation with [¬F] in probe matching [+F] in goal; (c) EPP: grammatical when associated with [¬F] in probe, semantic when associated with [+F] in probe, phonological (Chomsky 1999, Holmberg 2000).
The Universal Clause: [CP XP C [TP XP T [VP XP V [TP XP T [vP XP v [TP XP T VP]]]]]]. The highest TP is similar to Rizzi's FinP, the middle TP similar to Chomsky's TP, and lowest TP similar to AspP/AgroP. Internal arguments are merged in the last VP, the external argument in vP; no argument is merged in highest VP ( the locus of auxiliaries). Argument DPs, as well as active agr. (tensed verb agr. in Ice., participle agr. in Sw. and Ice, cf. Alexiadou & Anagnostopoulou 1998) have the features [¬tau,+phi]. Each T has the features [+tau, ¬phi], where [¬phi] is associated with EPP. In Eng., Ice. and Sw. the middle T also has an EPP feature that attracts the lowest T. Consequently, the main verb is always in the middle Tƒ. Due to active agr. in Ice., the finite verb is always attracted to the highest T. Verb second is an EPP-feature associated with [¬tau] in C, available only in main clauses, since embedding is an agree relation between the matrix verb and C, using [¬tau] and the feature Force in C, with the consequence that [¬tau] in C is available for attracting the verb only in main clauses. C may also host a [¬Wh] feature.
Deriving the simple transitive clause (below CP): When a transitive verb is merged in Vƒ, there must be a DP, with features [¬tau,+phi] in the complement of V. When T is merged to VP, it introduces the feature bundle [+tau, ¬phi], (EPP assoc. with [¬phi]). DP is attracted to Spec-TP, both uninterpr. features are deleted. At the next step, v is merged to TP, introducing the subject with its features [¬tau, +phi]. In Ice. in addition agr. with features [¬tau, +phi] is merged in v, and [+phi] in subj. is identified with [+phi] in agr. The middle T is merged to vP, introducing the features [+tau,¬phi]. Being closer to middle T than the object, the subject in Eng/Sw, agreement in Ice, is attracted to Spec- TP, deleting the uninterpr. features; the lower T is also attracted (see above), and the main verb is spelled out in the middle T. In the absence of any auxiliary, the next step in the derivation is simply the addition of an empty VP and the highest T, introducing the feature bundle [+tau, ¬phi]. Once again, the subject is the closest element with [+phi] in Eng./Sw., and it is attracted to Spec-TP, deleting all the uninterpr. features in T. In Ice. it is the verb with agr. that is closest, and is therefore moved to highest T. Up to this level Sw. equals Eng., both differs from Ice. In the Ice. clause there is still an uninterpr. feature around, [¬tau] in the subject in Spec-vP; this is deleted by nominative case.
Periphrastic and morphological passive.Sw. has two passives, with aux+participle, and with s attached to main verb. The participle has agreement, therefore [¬tau, +phi]. The s-passive is derived as the simple transitive clause: no external arg. in vP, therefore the [+phi] of the only DP (internal arg.) must be used with all Ts to delete [¬phi]. The periphrastic passive is different, since the [+phi] of the participle is used to delete [¬phi] in the lower two TPs; when the auxiliary is inserted in highest V, the participle is no longer available, and the DP argument is attracted to highest Spec- TP to delete [¬phi]. This correctly predicts that only s-passive is possible with perception verbs taking an ECM small clause, since the SC is another phase, hence the ECM subject is too far away when needed to delete [¬phi] in highest TP. In Danish, where the participle does not agree, the ECM subj. is attracted to lowest TP, and therefore it is available at the highest TP. Ice. (4) is out since a necessary relation between agr. in v and the nominative in the SC cannot be established over the phase border:
The different status of the participle in Sw. and Eng., in combination with the different feature values of the expletives (det/there), explain the following differences between Sw. and Eng.
(1) Han sågs [SC springa från dem kl. 8 (Sw)
`he saw.PASS run from them at 8 o'clock'
(2) *Han blev [SC sedd springa från dem kl 8 (Sw)
`he was seen run from them at 8 o'clock'
(3) Han blev set [SC løbe fra dem klokken otte. (Da)
`he was seen run from them at 8 o'clock'
(4) *Hann var sedh [SC koma. (Ice)
`he was seen come'
Icelandic case and agreement: compared with Eng. and Sw., the Ice. subject has more options, explained here as an effect of the DP not being forced to a particular position to delete features.
(5) *There was danced the whole night.
OKDet blev dan sat hela natten.
(6) *There were written three books last year.
OKDet blev skrivet tre böcker förra året.
That the verb agrees in pers. and num. with Nom subject, but only in num. with Nom object is due to c-command and MLC: when the Nom c-commands agr. (in vP) we get full agreement, when agr. c-commands Nom in VP or in lowest TP we get partial agreement. When agr.-Nom. does not meet MLC, ungrammaticality results, see (8), where D interferes:
(7)
Thadh
mundu
thá
sennilega
einhverjir bátar
ekki
verdha
seldir
á
uppbodhinu.
(TEC)
expl
would
then
probably
some b oats.N
not
be
sold
at
auction-the
(8)
*Thadh
voru
ekki
gefnar
stráknum
gjafir.
expl
were
not
given
the boys.D
gifts
For more information, contact karlos@MIT.EDU.
Human Development Program Colloquium
Professor Catherine Snow
Harvard University"Learning to talk in preschool: Developing Language to Promote Literacy "
All are welcome. Refreshments.Graduate Program in Human Development
Department of Psychology
Boston University
64 Cummington Street
Boston 02215For more information, see http://www.bu.edu/psych/graduate/hd/humtalks.html
MIT Phonology Circle
Michael Wagner
MIT"Prosodic Neutralization vs. Paradigm Uniformity
in Laryngeal Neutralization"The interaction with morphology can be explained by taking the phonological shape of suffixes into account. There are two classes of suffixes in German. Level I affixes behave as if they formed one string of segments with the roots they attach to prior to syllabification and stress assignment. A second class of suffixes does not interact with stress assignment and can only attach outside of Level I affixation (Level II suffixes). Among this group, a suffix forms a syllabification domain of its own if it starts with a consonant and is at least bimoraic. A cyclic definition of the prosodic word (Selkirk 1995) and subcategorization of suffixes for prosodic constituents (Inkelas 1992) allow for deriving the relevant domain effects in a principled manner.
Based on typological considerations, I will furthermore argue that cluster agreement has to be explained independently of final neutralization. Again, the syllabification domain plays a crucial role: The direction of assimilation within the domain differs from the direction of assimilation spanning domain boundaries. Phonetic measurements reveal that the alleged difference between Dutch and German with respect to cluster agreement (Lombardi 1991, 1999) is an artifact of a misinterpretation of the phonetic data.
I will compare this analysis to Steriade's (1997) 'Licensing by Cue' account. The two analyses make different predictions for a set of data were Paradigm Uniformity as defined in this framework fails to predict neutralization. The experimental data support the following claims:
- The environments of neutralization cannot be characterized phonetically- No notion of Paradigm Uniformity in addition to phonetic considerations can appropriately account for the data.
- The cues used to mark a contrast are not a good predictor of phonological patterns or of phonological markedness.
For more information, contact czoll@mit.edu.
MIT Linguistics Colloquium
Joseph Stermberger
University of Minnesota"Sprang, sproing, range, and gallop: An output-based approach to hypersimilar word families in speech production and acquisition" I will present experimental evidence from inflectional morphology to argue that the generalizations are based on OUTPUTS, and not on inputs. I will present information about hypersimilar families involving phonological nondefaults such as [Labial], leading to the same conclusions. I will discuss a model of learning and processing to account for the facts, and will argue that defaults (which are less commonly associated with hypersimilar word families) can be learned and processed in the same way as nondefaults.
For more information, contact bissell@mit.edu
MIT LING-LUNCH
Thursday, November 16, 2000 at 12:00, MIT E39-335 (conference room)Peter Hallman
University of North Texas"The Morphosyntax of Passivization" A close look at the prosodic template on which passive participles are formed in Lebanese Arabic reveals that the template is bimorphemic. Part of the template is a passivizing morpheme and the other part of the template is an adjective-deriving morpheme. The adjective-deriving morpheme occurs in adjectives without the passivizing morpheme, but with a particular restriction: it only forms adjectives from non-agentive predicates (hence its occurrence in passive participles).
Extending these findings to English makes sense of a peculiarity of the participial suffix -en. It forms participles of unaccusative verbs without valency reduction, as in `the fallen snow'. This makes sense if, like its Arabic counterpart, -en is an adjectivizing morpheme that only attaches to non-agentive verbs (hence its occurrence also in passive participles). But then the actual passivizing morphology must be non-overt in English (if -en isn't passivizing), and this seems to be the case. A survey of constructions that display valency reduction other than the participial ones shows no morphological reflex of passivization (e.g. in nominals: `the destruction of the city by the Romans').
But why do -en and its templatic Arabic counterpart only appear in non-agentive contexts? To answer this question, recent proposals are examined to the effect that verbs are not syntactically atomic. Agentive verbs are composed of an agentive, or causative, part that denotes a change of state, and an unaccusative part that denotes a resultant state. The proposal I will present to explain the behavior of -en is (1) that -en attaches to the unaccusative part of the verb to the exclusion of the agentive part; this explains its association with unaccusativity: it falls out from its particular syntactic selectional frame, and (2) when -en attaches to the unaccusative part of the verb, it monopolizes the syntactic slot in which accusative case is normally assigned, triggering object promotion to the nominative case position (i.e., passivization).
This take on the morphological composition of passivization makes sense of peculiarities of the distribution of -en and sheds light on a parallel between English and Arabic derivational morphology that was not previously noticed. It also lends credence to the hypothesis that verbs are syntactically complex, since this hypothesis explains the restriction to non-agentive verbs that -en and its Arabic counterpart display. This in turn means that syntax does not stop at the word level, and suggests that morphological phenomena arise in general from the interaction of syntactic and phonological forces.
For more information, contact karlos@MIT.EDU.
UMass, Amherst - Colloquium
William Labov
University of Pennsylvania"The continental divide of American English and its consequences for theories of language change"
For more information, see http://www.umass.edu/linguist/colloquia.html
MIT LING-LUNCH
Thursday, November 9, 2000 at 12:00, MIT E39-335 (conference room)Sungeun Cho,
Massachusetts Institute of Technology
"A Form of Case Agreement in Korean:
possessor agreement constructions"This paper focuses on Case agreement phenomenon in the framework of the Minimalist Program. For our purpose, Case agreement will be defined as some systematic covariance between two separate elements in the domain of a single verb. In Korean, two or more noun phrases can bear the same Case. Possessor nominals can be marked with the same Case as their possessee nominal, as shown (1).
(1) Mary-ka John-ul tali-lul cha-ess-ta
Mary-NOM John-ACC leg-ACC kick-PAST-DECL
`Mary kicked John's leg.'
Although the verb cha `kick' is a mono-transitive verb, both the possessor John and the possessee tali `leg' are marked accusative, like normal objects.
Interestingly Case agreement between the possessor and the possessee is not always shown. In (2a), the possessor John is not marked accusative Case. Instead, it is marked with genitive Case as shown in (2b).
(2) a. *Mary-ka John-ul cha-lul cha-ess-ta
Mary-NOM John-ACC car-ACC kick-PAST-DECL
`Mary kicked John's car.'
b. Mary-ka John-uy cha-lul cha-ess-ta
Mary-NOM John-GEN car-ACC kick-PAST-DECL
`Mary kicked John's car.'
This phenomenon raises crucial questions: In what situation are possessors marked with accusative Case like objects? How can mono-transitive verbs check accusative Case on nominals twice? Regarding the former, I propose that the entailment between the verb and the possessor is the crucial factor. Regarding the latter, the leading idea is that Case-checking is always bijective and always results in erasure. The phenomenon of multiple agreement arises through the action of a distinguished class of postposition-like functional elements (Ks) that have the capacity to copy Case-features from an independent Case-bearing category (V, T, etc.). Agreeing elements therefore always check their Case feature through a K head. This analysis extends idea by Aronoff (1988) on agreement as copying and by Larson (1997) on adjectival agreement as feature-copying. And as I show, it intersects with a number of very interesting topics in syntax and semantics including the structure of possessives, passivization, affectedness and entailment.
For more information, contact karlos@MIT.EDU.
MIT Phonology Circle
Ken Hale
MIT"The Phonological Aspect of the Navajo Conjunct Verb System"
For more information, contact czoll@mit.edu.
MIT Linguistics Colloquium
Juvénal Ndayiragije
The University of Western Ontario"The Ergativity Parameter"
For more information, contact bissell@mit.edu
BU CAS/CNS Noon Colloquium
Professor Peter Jusczyk
"Infants' segmentation and recognition of words"
To reach Jusczyk's home page and list of publications see http://www.psy.jhu.edu/~jusczyk. Relevant papers to download for the upcoming talk are: the 1999 Trends in Cognitive Science Paper, the paper with Hohne & Bauman that appeared in Perception & Psychophysics, and the paper with Mattys that is "in press" at Cognition (i.e. the one on the use of phonotactic cues). These papers will also be available in the CAS/CNS library in the 2000/2001 Seminar Box. Jusczyk will be presenting new findings that build on these earlier results. For information on the CAS/CNS Fall Seminar Series see http://www.cns.bu.edu/Colloquia.
MIT LING-LUNCH
Thursday, November 2, 2000 at 12:00, MIT E39-335 (conference room)Derek Bickerton
University of Hawai'i"When c-command fails: principles of priority and finality"
For more information, contact karlos@MIT.EDU.
Human Development Program Colloquium
Professor Margret Hagen
Boston University"Psychological Syndromes and Courtroom Testimony: Child Sexual Abuse Accommodation Syndrome, Battered Woman Syndrome, and Rape Trauma Syndrome"
All are welcome. Refreshments.Graduate Program in Human Development
Department of Psychology
Boston University
64 Cummington Street
Boston 02215For more information, see http://www.bu.edu/psych/graduate/hd/humtalks.html
MIT Linguistics Colloquium
Larry Horn
Yale University"Free Choice, Free Relatives, and the Quodlibetic Indefinite" polarity (NPI-) any and free-choice (FC) any. For the univocal wide-scope universalists, any is always a wide-scope operator triggered variously by negation, modality, genericity, etc. This approach is elegant but empirically flawed in failing to account for a wide range of existential/indefinite properties of any. On the countervailing ambiguist view, NPI-any is existential and FC-any universal. While this line appears to be empirically more successful than univocal universalism, it is conceptually unsatisfactory in positing a dubious homonymy, especially in the light of diagnostics (e.g. at all, whatsoever) that co-occur with both any's but with no other operators of either universal or existential persuasion.
The third position, to be defended in this presentation, develops a unified theory of NPI-any and FC-any as specialized indefinites. We will revisit the familiar diagnostics, including some that have been taken to support universalist analyses of non-NPI any, and we will introduce new data that provide evidence for the treatment of free-choice expressions (including English any and its analogues in Greek, Romance, Hindi, and Korean) as "quodlibetic" indefinites lacking inherent quantificational force and ranging over arbitrarily selected items typically situated at scalar end-points.
This line on free-choice determiners and quantifiers will be extended to the treatment of wh-ever free relatives on their 'don't care' (rather than 'don't know') readings. While recent proposals have taken these free relatives to constitute either definite NPs or universal quantifiers, I will argue that whatever and its close relatives are best analyzed as free choice indefinites in the manner of the corresponding any expressions.
For more information, contact bissell@mit.edu
MIT Phonology Circle
Joan Mascaro "On the representation of contrasting rhotics" As previously announced 20-10- and 27-10-theories will be compared and conclusions will be drawn (we hope).
For more information, contact czoll@mit.edu.
MIT LING-LUNCH
Thursday, October 26, 2000 at 12:00, MIT E39-335 (conference room)Denis Bouchard
UQAM and Paris 8"Determinerless Noun Phrases in French and English" It is well known that English allows quite easily the absence of a determiner in common NPs, plural (1) or singular (2), whereas this is generally impossible in French (3-4).
(1) a Beavers build dams everywhere.
(1) b I hate beavers.(2) a Lion tastes awful.
(2) b I ate lion yesterday.(3) a *Castors construisent barrages.
(3) a' Les castors construisent des barrages.(3) b *Je déteste castors.
(3) b' Je déteste les castors.(4) a Le/ *Ø lion, c'est délicieux.
(4) b J'ai mangé du/*Ø lion hier.The main hypothesis I make is that these contrasts originate in the different ways these languages have chosen to express semantic number in a nominal expression. There are numerous independent indications that number in French is borne by the Det, whereas in English it is borne by the N. This variation will be shown to have its source in properties of the CI and AP systems, hence to have external motivation. The analysis rests on a fundamental condition that a nominal expression must satisfy to properly interface with the CI system. Simplifying, we can say that a sentence describes an event involving actants (in a broad sense covering arguments and adjuncts). The typical role of a nominal expression is to provide information to identify an actant. There are a few equally efficient modes to identify an actant in the CI system, so that the grammaticalization of any number of these is available for a language. I will show how number marking is one means to grammaticalize one of these modes, and that a few equally efficient modes of coding number in the AP system give rise to additional choices: English and French made different choices.
I will discuss several sets of data that follow from the main hypothesis. Since English codes number on the N, a NP without a Det can still refer minimally. On the other hand, since Number is expressed on the Det in French, a determinerless nominal gets an intensional interpretation. This peculiar interpretation explains the restricted distribution of determinerless NPs in French: they are nonreferential, but with a few interesting exceptions.
For more information, contact karlos@MIT.EDU.
Boston University undergraduate Linguistics Association
first event of this academic year ! A special screening of 
François Truffaut's
MIT Linguistics Colloquium
Howard Lasnik
University of Connecticut"Salvation by Deletion"
For more information, contact bissell@mit.edu
UMass, Amherst - Colloquium
Roger Schwarzchild
Rutgers University"Parallels in the semantics of tense and scalar predicates"
For more information, see http://www.umass.edu/linguist/colloquia.html
MIT Phonology Circle
Jim Harris, assisted by a Cuban pop vocal ensemble "The distribution of rhotics in Spanish" Abstract: It is well known that orthographic <r> and <rr> in pairs such as cero-cerro 'zero'-'hill' represent a phonological contrast in Spanish: <r> and <rr> spell segments typically realized as a flap and a trill, respectively. The flap-trill contrast is found only intervocalically; in all other positions in which either segment can occur, either (a) one or the other is obligatorily disallowed or (b) the two are in free (stylistic) variation. Thus the question arises: One underlying segment or two? If one, the intervocalic contrast presents a challenge; if two, lack of contrast elsewhere must be explained. The traditional structuralist two-phoneme view of Trager (1939) and others persists in D'Introno et al. (1995) and other current work. However, at least since Saporta & Contreras (1962), generative accounts have posited a single underlying r-segment for Spanish and Catalan (plus Portuguese, which presents analogous data). I summarize old and new arguments for the standard generative analysis, and respond to some objections to it that have been raised. Next week, Joan Mascaro presents an alternative view (abstract to follow), the two views will be compared, and conclusions will be drawn (we hope).
D'Introno, Francesco, Enrique del Teso & Rosemary Weston. 1995. Fonetica y fonologÌa actual del espanol. Madrid: C·tedra.
Saporta, Sol & Heles Contreras. 1962. A Phonological Grammar of Spanish. Seattle: University of Washington Press.
Trager, George L. 1939. The phonemes of Castilian Spanish. Travaux du Cercle Linguistique de Prague 8:217-222.
For more information, contact czoll@mit.edu.
MIT LING-LUNCH
Thursday, October 19, 2000 at 12:00, MIT E39-335 (conference room)Susana Bejar
University of Toronto"Locality, Cyclicity and Markedness in Georgian Agreement" The inventory of inflectional pieces in the agreement system is given in (1). In general, it can be said that Georgian manifests both subject and object agreement for both person and number. However subject and object features compete for the same positions of exponence in a manner that cannot be straightforwardly accounted for in purely morphological terms. Thus it is not always the case that full subject and object agreement are manifest, and a systematic account of what governs this is required. Previous accounts have appealed to principles of vocabulary insertion to explain this phenomena (Halle and Marantz 1993, Carmack 1997). I propose that in fact the principles which govern the choice of subject and object agreement are more straightforwardly syntactic.
(1) v- 1 subject
m- 1 object
gv- 1pl object
g- 2 object
-t plural
In general it can be said that the verb agrees with person of the object (2) and the number of the subject (3), but when the subject is unmarked for number (singular) there can be number agreement with the object (4).
(2) m-xedav 'You see me'
1-see
(3) m-xedav-t 'You(pl) see me'
1-see-PL
(4) gv-xedav 'You(sg) see us'
1.PL-see
If the derivational history of the clause is ignored, then the shape of this paradigm seems arbitrary. However there is a derivational acount from which the pattern in (2)-(4) follows. The account rests on two proposals: (i) the Markedness Theorem; (ii) the splitting of uninterpretable phi-features between functional categories so that T is not the sole locus of Agree in Georgian. Specifically, phi-features are distributed so that light v has person and T has number. Agree is cyclic, such that first light v Probes the closest person feature, and then T Probes the closest number feature. Given these assumptions, it follows from locality that the person features of the object -and not the subject - are indexed, while the number features of the subject - and not the object - are indexed. If the person of the object is unmarked (i.e. 3rd person) then the Probe cannot be satisfied and default agreement is manifested. If the number feature of the subject is unmarked then the subject cannot sastisfy the probe which may extend past it and be matched by the number feature on the next nearest argument in its domain.
The agreement morphology is essentially reduced to the manifestation of the probe on v and the probe on T, thus there is no subject or object agreement per se. Additional locality effects give further evidence for this analysis. (i) As expected given locality, an indirect object (merged higher than the direct object) blocks agreement with the direct object. (ii) Intransitive constructions only manifest the agreement morphology associated with T, either because there is no light v (unacussative) or there is no goal in the domain of the light verb (unergative) or the v is defective (passive). (iii) Most importantly, the so-called 'inverse' agreement pattern which is manifest in dative subject constructions can also be reduced to a locality effect.
The analysis reveals an underlying simplicity to Georgian verbal morphology which has eluded previous accounts. Perhaps more importantly, it demonstrates that markedness effects - which have long been relegated to abstract hierarchies of one sort or another - can be derived from core principles, and can be shown to be both a source of order and a source of complexity. However, the Markedness Theorem poses a problem for the framework with respect to case assignment. If the Markedness Theorem is correct, then it is not clear how, for example, a singular subject would get nominative case, or a 3rd person object accusative. Constructions with nominative objects pose similar problems for the formulation of probe, and the postulated relationship between case and agreement in Chomsky (1998, 1999) and Schutze 1997. Furthermore the Georgian data suggests that visibility - in the sense of there being an unvalued uninterpretable case feature on a DP - is not relevant to the probe. An accusative object whose case needs have been satisfied by v can nevertheless enter into agreement with T. Thus we can distinguish between symmetric associate dependencies, where both the uninterpretable feature of the target and that of the goal are valued, versus asymmetric associate dependencies where only the uninterpretable feature of the target is valued. This in turn raises the possibility that one might find the opposite asymmetry, such that only the uninterpretable feature of the goal was valued. This would account for the ability of a probe to assign case to a DP with unmarked phi-features. However many questions arise as to the conditions under which asymmetric associate dependencies may arise, and what consequences they hold for the framework.
For more information, contact karlos@MIT.EDU.
MIT Phonology Circle
Hubert Truckenbrodt
Rutgers University and MIT"On the Representation of Register in Intonation" BACKGROUND. H tones (and L tones) can systematically vary in phonetic height. Many authors have followed the suggestion of Clements 1979 that the phonetic height of tones is determined relative to an abstract register. Downstep is systematic lowering of the register, and the reset sets the register back to the initial height.
CLAIM. While downstep and reset are abstract phenomenon that can only indirectly be inferred from the phonetic height of H (and L) tones, I present three arguments that the abstract reset is itself represented in phonology as a boundary tone. I claim that the reset is represented as [h]%, a feature [h] for high register with a diacritic %, marking it as a boundary tone. The arguments draw on my experimental research on Southern German intonation.
ARGUMENTS. The first argument is that the reset, like other boundary tones, occurs in the vicinity of prosodic boundaries, where, crucially, it may either precede or follow the prosodic boundary. Thus, in English, Japanese, and Yoruba, the reset occurs at the beginning of a new prosodic domain, *following* a prosodic boundary. The present analysis treats this as an initial [h]% boundary tone. In Southern German, the reset *precedes* a medial prosodic boundary, which the present analysis treats as a final [h]% boundary tone. It will be seen that this unexpected reset phenomenon is not compatible with a number of suggestions about the representation of downstep and reset in the literature.
The second argument is that the reset may be minimally displaced from the edge to which it is assigned, in response to pressure from phonological constraints. This will be seen in a phonological and phonetic analysis of four dialectally different patterns of the pre-boundary reset in Southern German. Minimal displacement from the edge is a property that the reset will be seen to share with other boundary tones.
The third argument is that the reset can show pragmatic boundary tone 'meanings': The Southern German reset shares the meaning of announcing continuation with English and German H% boundary tones.
RELEVANCE. The result that the reset is represented as a [h]% boundary tone is important for theories about the phonology and phonetics of intonational systems more generally: It excludes some existing theories about the role of register in intonation (Pierrehumbert and her colleagues, for whom register is confined to the phonetic component; Clements and Ladd, who organize register in a hierarchical representation). It supports the conclusion reached by Hyman, Inkelas, Snider, and others, that register in intonation is autosegmentally represented. Within that theory, it provides a new perspective according to which register in intonation interacts with prosodic structure in ways that parallel the interaction of H and L tones with prosodic structure.
For more information, contact czoll@mit.edu.
MIT Phonology Circle
Charles Reiss
Concordia University"Quantification and identity references in phonology:
A case study in Universal Grammar"
For more information, contact czoll@mit.edu.
UMass, Amherst - Colloquium
Jane Grimshaw
Rutgers UniversityTitle TBA
For more information, see http://www.umass.edu/linguist/colloquia.html
MIT LING-LUNCH
Thursday, October 5, 2000 at 12:00, MIT E39-335 (conference room)Norvin Richards
MIT"Very local A-bar movement in a top-down derivation" (1) a. There seems to be a man in the room
b. *There seems a man to be in the room
c. There was heard [a rumor that a man is in the room]
d. [A rumor that there is a man in the room] was heard
The account I developed of the facts in (1) made a surprising prediction about A-bar movement: that under certain circumstances, such movement should be very local, unable to skip an A-position in which it could in principle land. In this talk I will try to show that this prediction is, under certain circumstances, correct, and yields an account of a number of recalcitrant facts.
For more information, contact karlos@MIT.EDU.
MIT - First annual Harry Bradford Stanton Lecture
Lila Gleitman
University of Pennsylvania"Does Our Language Affect the Way We Think?"
Refreshments will be served.
MIT LING-LUNCH
Thursday, September 28, 2000 at 12:00, MIT E39-335 (conference room)Cristina Cuervo,
Massachusetts Institute of Technology"Dative experiencers and the lack of intervention effects " In transparent languages such as English (Chomsky, 1998 and others) and French (Boeckx, 1998), everything proceeds as in the configuration without the experiencer: the dative experiencer does not intervene (it is not a potential target for movement), permitting raising (1). In opaque languages, such as Icelandic and Italian (Boeckx, 1998), the experiencer has a blocking effect, and so raising of the embedded subject is ungrammatical (2).
(1) Jon(1) seems to Mary [ t(1) to be the best ]
(2) * Ólafur(1) hafði virst þeim [ t(1) vera gáfaður ]
Olafur(Nom) has seemed them(Dat) to be intelligent
According to Chomsky (1998), ungrammaticality in Icelandic is due to locality. Boeckx accounts for the asymmetry in terms of datives being inherently case-marked (English) versus structurally case-marked (Icelandic). Spanish has been claimed to pattern with Icelandic in that the experiencer blocks raising (Torrego 1996; McGinnis 1998; Boeckx 1998; Anagnostopoulou to appear) (3).
(3) * Emilio le parece a Valeria ser el mejor
Emilio(Nom) Cl(Dat) seems to Valeria(Dat) to be the best
I will present data from Spanish that these analyzes cannot account for. I will show that the proposed split (transparent-inherent case versus opaque-structural case) does not adequately characterize the asymmetries observed in these languages, and that intervention effects are not the source of the observed ungrammaticality.
`Seem'+experiencer constructions are better understood when considered as one more instance of unaccusatives where a dative argument is merged higher than another argument that gets nominative case. Several structures are analyzed in English and Spanish: psych verbs, inchoatives, ditransitives and passives. The main contrast is between cases where the dative raises to subject position and cases where the lower object raises across the dative.
The weight of the contrast is placed on the characteristics of the attracting head, rather than on the case of the NP (that is, whether dative is inherent or structural case in a certain language), or on differences in mapping of arguments onto syntactic structure. In particular, I claim that some functional heads (their EPP feature) impose restrictions on the Case characteristics of the NP they attract (the NP must be `active'). Others attract the closest argument (irrespectively of the case characteristics of the NP). English T and Spanish v are of the first type; Spanish T is of the latter type.
This variation has an important consequence for `seem' constructions. In languages where the dative raises to subject position, (Icelandic and Spanish) a propositional predicate like `seem'+experiencer has an `experiencer subject', and it is expected it will not take a raising clause (IP) as a complement. The difference is that while Spanish shows the `expected' control infinitival CP complements, Icelandic tolerates the more rare raising/ECM clauses.
(4) þeim hafði virst Ólafur vera gáfaður
Them(Dat) has seemed Olafur(Nom) to be intelligent
(5) Me parece [ PRO haber visto un lindo gatito ]
Cl(Dat1stSg) seems to have seen a cute little cat
For more information, contact karlos@MIT.EDU.
MIT Phonology Circle
Zhiqiang Li
MIT"A Prosodic Account of Neutral Tone in Mandarin Chinese"
For more information, contact czoll@mit.edu.
UMass, Amherst - Colloquium
Cheryl Zoll
MIT"Markedness-driven tone mapping"
For more information, see http://www.umass.edu/linguist/colloquia.html
MIT LING-LUNCH
Thursday, September 21, 2000 at 12:10, MIT E39-335 (conference room)David Adger
" Strategies for wh: derivation and representation
in Scottish Gaelic Questions and Relatives"(1) Thuirt thu gun robh e tinn
Said you C was he ill
"You said he was ill"
(2) Co a thuirt thu a bha tinn
Who C said you C was ill
"Who did you see was ill"
The important fact to focus on here is the alternation between the complementiser `gun' in (1) and `a' in (2).
We can accommodate this into current thinking by proposing that the phasehood of CP is voided because optional addition of an EPP feature to C brings the potential extractee into a position where it is accessible to its ultimate attractor. This optional EPP feature is morphologically signalled by the use of `a'.
Attractive though this story is, we argue in this paper that it is not quite right. There is strong empirical evidence that the relation between the complementiser and the goal position is not a movement relation in Scottish Gaelic, but is rather an agree relation with a null pronominal. This evidence comes from definiteness agreement on prepositions, reconstruction phenomena and CED effects, and leads us to propose that long wh-extraction actually involves a chaining of Agree relationships which is `sealed off' by a copular construction. Schematically, for an example like (1), we have (3):
(2)[Co] Cop [RelClause C[+Rel] ... C[+Rel] ... pro[+Rel]]
A number of interesting theoretical questions arise, not least the function of the [Rel] feature. We argue for a system where this feature is unvalued on C unless C Agrees with pro[+Rel]. If it does, then the CP is interpreted as a predicate, with a lambda-abstracted variable of the same type as pro. The function of the copula is to allow predication between this Relative and the Wh-subject of the copula. The long distance effect arises because of an interpretative clash between the selectional specifications of the higher verb (in (1), the verb `say') which selects for a proposition, and the semantics of the relative (which is a predicate). Although [Rel] on C is valued, it is not interpretable yet, hence it is still active and a possible goal for the higher C[Rel]. The dependency is passed up the structure until it can be terminated by a verb of the right selectional specification (i.e. the copula).
Theoretically, the argument points to a system where LF requirements are interweaved with syntactic requirements, at least at certain points (phases) in the derivation.
For more information, contact karlos@MIT.EDU.
MIT Phonology Circle
Bert Vaux
Harvard UniversityCoronals and epenthesis available at: http://ruccs.rutgers.edu/roa.html
For more information, contact czoll@mit.edu.
MIT LING-LUNCH
Thursday, September 14, 2000 at 12:10, MIT E39-335 (conference room)Cheryl Zoll
MIT"Markedness and Tone Mapping"
For more information, contact karlos@MIT.EDU.
MIT Phonology Circle
Karlos Arregui-Urbina The semantics of reduplication
For more information, contact czoll@mit.edu.
HUMIT
A joint MIT and Harvard Student Conference on Language Research
To be held at MIT and Harvard on August 30-31, 2000
MIT Phonology Circle
Bert Vaux
Harvard University"Disharmony and derived transparency in Uyghur Vowel Harmony"
For more information, contact czoll@mit.edu.
MIT Linguistics Colloquium
Jaye Padgett
UC Santa Cruz"The Role of Contrast in Russian (Historical) Phonology"
For more information, contact asrackow@mit.edu.
MIT LING-LUNCH
Thursday, May 11, 2000 at 12:10, MIT E39-335 (conference room)CÈdric Boeckx
University of Connecticut"Interpretation by phase: some evidence" In this talk I provide further evidence for a cyclic (phase-based) model of syntax by looking at data ranging from Quantifier Scope to the interpretation of multiple questions and vehicle change. I argue that the phenomenon of Frozen Scope (esp. Bruening/Sauerland-style data) can be reconciled with Fox's theory of QR once the importance of phases is taken into account. Likewise the type of pair-list/single-pair asymmetries between wh-fronting and wh-in-situ languages recently brought to light by Boskovic and others is shown to provide evidence for a phase-based model. Finally, I show that a specific interpretation of what a phase is might shed some light on the phenomenon of, and dispense with the mechanism of, vehicle change..
For more information, contact karlos@MIT.EDU.
MIT Phonology Circle
Benjamin Bruening, Elissa Flagg, and Vivian Lin "Asymmetries in the Processing of Tone Information in Chinese and English Speakers"
For more information, contact czoll@mit.edu.
MIT Linguistics Colloquium
Angelika Kratzer
UMass Amherst"Telicity and the Semantics of Objective Case"
For more information, contact asrackow@mit.edu.
MIT LING-LUNCH
Thursday, May 4, 2000 at 12:10, MIT E39-335 (conference room)Antonella Sorace
University of Edinburgh"Gradience in auxiliary selection and the syntax-semantics mapping" CHANGE-OF-LOCATION Core unaccusative, selects 'be'
(least variation)
CHANGE-OF-STATE
CONTINUATION-OF-STATE
EXISTENCE-OF-STATE
UNCONTROLLED PROCESS
CONTROLLED PROCESSES (MOTIONAL)
CONTROLLED PROCESS (NON MOTIONAL) Core unergative, selects 'have'
(least variation)
Core verbs at the unaccusative extreme denote telic change; core verbs at the unergative extreme denote agentive activity in which the subject is unaffected; verbs between the extremes have progressively less specified event structures. It is with these verbs that most cases of 'unaccusative mismatches' or variable mappings are found. Core verbs are those on which native grammaticality judgments are maximally consistent, and are acquired early by both first and second language learners; degree of inconsistency and delay in acquisition are a function of the position of a verb along the hierarchy. The generalization seems to be that as soon as one moves from the core one finds massive, but predictable, indeterminacy in the syntax-semantics mapping with intransitive verbs.
I first argue against a purely semantic explanation of the facts, and against restricting the hierarchy to auxiliary selection (as opposed to split intransitivity in general). I then discuss the implications of this hierarchy for the Unaccusative Hypothesis, and particularly for theories of the syntax-semantics interface as they relate to unaccusativity. I show that neither existing projectionist accounts (e.g. Levin and Rappaport Hovav 1995) nor constructionist accounts (e.g. Borer 1994; McClure 1995) can completely accommodate these facts, although the latter can better handle mapping ambiguities.
For more information, contact karlos@MIT.EDU.
MIT RLE Speech Communication Group Seminar
John Kingston
Linguistics Department, University of MassachusettsDISCRIMINATION PREDICTS IDENTIFICATION
NOT VICE VERSAReference:
Macmillan, N. A., Kingston, J., Thorburn, R., Walsh Dickey, L., & Bartels, C. (1999). Integrality of nasalization and F1. II. Basic sensitivity and phonetic labeling measure distinct sensory and decision-rule interactions, J. Acoust. Soc. Am., 106, 2913-2932.
For more information, contact czoll@MIT.EDU
Harvard University
Graduate Student Society Talk
Adolfo Ausin
Department of Linguistics, University of Connecticut"Multiple Spell-Out and the locality conditions on A-movement"
You can obtain directions and other information at: http://www.fas.harvard.edu/~lingpub/resources/publtalk.htm. For more information, contact szczeg@fas.harvard.edu
presents...
A special showing of the NOVA documentary about GENIE
with an introduction by Prof. Jean Berko Gleason
For more information, contact Kate Thompson.
MIT Phonology Circle
Degif Petros Banksira
MIT"The Light-Heavy allomorphy of Chaha object suffixes" Each object suffix in Chaha has two phonologically related allomorphs called Light and Heavy. In most cases, the Light ones have the underlying or ungeminated phonemes whereas their Heavy counterparts have the strengthened or geminated allophones. The phonological content of the phoneme immediately preceding an object suffix was considered to be the conditioning factor for this allomorphy. In this talk I will present a detailed examination of the alternations and the conditioning factors and I will discuss its implications for Semitic verb morphology. In line with Banksira (1999), I will argue that verb conjugations in Chaha, as well as in Semitic generally, have the structure in (1), where Q is a fused [Aspect] and [Subject Clitic]. The linear order of Q varies depending on whether it is [+Perfect] or [-Perfect].
(1) [[[[stem] Q] AGR-S] AGR-O]
The Light-Heavy allomorphy affects Agr-O; Agr-O is Light when preceded by a phonetically null Agr-S and it is Heavy when preceded by a phonetically realized Agr-S. The quality of preceding phonemes found in constituents different from Agr-S (i.e. the stem-final phoneme when both Q and Agr-S are f, and the final phoneme of Q when Agr-S is f) does not matter. In other words, the phonemes of Agr-S, which warrant Heavy Agr-O, are followed by a Light Agr-O if they belong to a constituent different from Agr-S. The Light-Heavy allomorphy requires distinguishing Q from Agr-S, as in (1), and constitutes an argument for the presence of two independent functional heads between stem and Agr-O in both the perfective and imperfective aspects. If the presence of Q and Agr-S in both aspects is justified, the notion of Fission, used in Noyer (1997) and Halle (1997) for comparable facts, can be dispensed with from the account of Semitic imperfective conjugation.
For more information, contact czoll@mit.edu.
Harvard University
Graduate Student Society Talk
Claire Bowern & Gulsat Aygen-Tosun
Department of Linguistics, Harvard University"Titan's Tensed Prepositions "
You can obtain directions and other information at: http://www.fas.harvard.edu/~lingpub/resources/publtalk.htm. For more information, contact szczeg@fas.harvard.edu
MIT LING-LUNCH
Thursday, April 20, 2000 at 12:10, MIT E39-335 (conference room)Ora Matushansky "WHY AM I A LINGUIST IF I ONLY SEEM GOOD?
Adjectival movement in nominal predicates"1. Introduction:
In this paper, I will offer an explanation for the special properties of
an English construction in which seem is followed by an (indefinite) DP. Examples (1)-(4) show that the post-verbal DP in this construction must be modified and that there are semantic restrictions on the nature of the modifying adjective: it must be overt, as in (1), non-intersective (Larson 1995), as in (2), and scalar, as in (3). Examples (4) show that certain adverbs also license a DP-complement:
1 a. * Paganini seems a violinist.
b. Paganini seems a great violinist.
2 a. * Paganini seems an Italian violinist.
b. Paganini seems a skillful violinist.
3 a. * Paganini seems a complete violinist.
b. Paganini seems a talented violinist.
4 a. * Paganini seems a dark-haired violinist.
b. # Paganini seems a fairly dark-haired violinist.
Examples (5) show that these restrictions are not found when the complement is an AP:
5 a. Paganini seems intelligent.
b. Paganini seems Italian.
c. Paganini seems dark-haired.
2. Presupposition:
the double entailment in (6) shows that the adjective behaves as if the main verb combined solely with the AP, to the exclusion of the DP remnant:
6 a. Paganini seems a great violinist.
b. Paganini doesn't seem a great violinist.
=> Paganini is a violinist.
3. Proposal:
The modifier phrase seems to be selected by the verb. The fact that it is missing from the presupposition shows that it behaves like an argument of seem. However, since the AP is buried in the DP, compositionality ensures that a direct ?-relation between it and seem is unlikely.
In adjective stacking constructions, it is always the topmost adjective that is excluded from the presupposition (7a, b). On the other hand, when a licensing adjective, such as expensive in (7b), is not the outer one, the DP complement becomes ungrammatical (cf. (7c), where the licensing adjective is embedded under a non-licensing one).
7 a. This seems a bad expensive car.
Presupposition: This is a bad car.
b. This seems an expensive bad car.
Presupposition: This is a bad car.
c. * This seems a big expensive car.
The island-like effects in (8b, c) show that adjective licensing is
possible over short distance only.
8 a. She seems a doctor of extraordinary ability.
b. * She seems a person who has extraordinary ability.
c. * She seems a doctor of a friend of mine of extraordinary
ability.
The selectional restrictions on the modifier and locality effects indicate that movement takes place. To accommodate the Spec hosting the (covertly) moved AP, we assume that the relation between seem and its complement is mediated by a CP.
4. Motivation:
Adjective fronting constructions, such as (9), show that adjectival movement must be postulated for independent reasons:
9 a. The more reliable a doctor one is, the better one is paid.
b. How reliable a doctor was W. Somerset Maugham?
c. No matter how reliable a doctor W. Somerset Maugham was,
he is better known as a writer.
d. W. Somerset Maugham was not that reliable a doctor.
There also exist cases with similar interpretation and choice of adjectives, where the relation between the adjective and the DP remnant is mediated by an overt complementizer.
10a. W. Somerset Maugham seems very reliable as a doctor.
b. W. Somerset Maugham seems very reliable for a doctor.
5. Trigger:
The modifiers licensing a DP complement to seem are scalar, i.e. they all contain a degree variable d, whose value is contextually determined (John is tall means [John's height = d]). We suggest that the AP is moved to [Spec, CP] to check the [degree] feature of the CP and specify the conditions under which the [degree] feature is visible to the syntax. The licensing effect of degree adverbs in (4) will also be explained.
Finally, we will give arguments against a purely semantic analysis of modifier selection and specify the conditions, under which a non-modified DP can appear as the complement of seem.
6. Conclusion:
The selectional properties of the DP-complement are derived from the fact that it is contained in a CP, inside which adjective movement is necessary for convergence. The [+degree] CP thus strongly resembles the [+wh] CP.
7. Bibliography:
Larson, Richard K. (1995) Olga is a Beautiful Dancer. Paper presented at the LSA meeting, New Orleans, LA.
For more information, contact karlos@MIT.EDU.
Harvard University
Graduate Student Society Talk
Kylie Skewes
Slavic Department, Harvard University"Small Clauses in Russian, Ukrainian and Icelandic:
A Brainstorming Session"(1) Roni cut the bread wet.
(2) *Roni cut the bread white.
Russian and Ukrainian depictive Small Clauses, however, may have both Stage-level and Individual-level (adjectives denoting more permanent properties of individuals) properties; these properties depend on whether the predicative adjective agrees with its NP- controller in Case or not, e.g., Russian (Nominative subject-control):
(3) Stepan (NOM), mesjac provaljavshis' v bol'nice, vernulsja
zdorovyj (NOM) /zdorovym (INSTR)
Steven, having disappeared for a month in hospital,
returned healthy.
(example from Timberlake 1986, tested on my native informants).
The predicative adjective with nominative Case in this example has an I-level interpretation, while the adjective with the instrumental Case has a S-level interpretation. Russian (accusative object-control):
(4)Ja(NOM) voz'mu ego(ACC) zhivogo (ACC).
I'll take him alive.
(5) (Vzjat' ego(ACC)) zhivym(INSTR) ili m'ortvym(INSTR).
(Take him) alive or dead.
(Russian equivalent of 'Wanted Dead or Alive')
Ukrainian (accusative object-control):
(6) Borys znajshov Sashu(ACC) holoho(ACC)/ holym(INSTR)
Boris found Sasha naked.
We see the same contrasts in meaning between agreement and the instrumental Case in these examples with object control-- agreement imparts a I-level interpretation, the instrumental Case a S-level interpretation. These oppositions disappear in constructions with long-distance control (ie., the instrumental is obligatory once there is a non-finite verb or an overt complementizer), and in constructions in which the predicative adjective is controlled by an NP with lexical Case.
Unlike Russian and Ukrainian, Icelandic predicative adjectives agree in case with their controlling NP in both simple clauses and in constructions with long-distance control. Interestingly, in non-finite constructions the predicative adjective agrees with the Case that an overt NP-controller would have received in a finite construction (note: in the following example the verb "to drift" takes an accusative argument in Icelandic):
(7) Hana(ACC) rak a land eina(ACC).
She drifted ashore alone.
(8) Hun vonast til [PRO ad reka a land eina ]
She(NOM) hopes to drift ashore alone(ACC).
(data from Andrews 1990)
To muddy the waters even more, Russian (and Ukrainian) have a construction that resembles the Icelandic long-distance control constructions--the so-called "second datives" (Comrie 1974) 'odin' and 'sam' 'alone'. Unlike predicative adjectives, 'odin' and 'sam' always agree in case with their NP-controller in simple clauses. In certain clauses with long-distance control, however, the Small Clause either agrees in case with its controller, or occurs in the dative case (the Case of subjects of non-finite verbs in Russian):
(9)Ja(NOM) poprosyla Ivana(ACC) [PRO prijti odnogo(ACC) /
odnomu(DAT)]
I asked Ivan to come alone.
For some speakers these two constructions have a different semantic interpretation, for others only one or the other is possible. These constructions differ in both structure and interpretation from all other predicative adjective structures in Russian, and resemble the Icelandic state of affairs more closely.
The Russian, Ukrainian and Icelandic data suggest that Theta-theory, Case Theory and the semantic interpretation of Small Clauses are closely connected. The exact nature of this connection will be open for a brainstorming session.
You can obtain directions and other information at: http://www.fas.harvard.edu/~lingpub/resources/publtalk.htm. For more information, contact szczeg@fas.harvard.edu
Harvard University
Linguistics Talk
Stephen Anderson
Yale University"A-Morphous Account of Tagalog Second Position Clitics"
For more information, contact violette@fas.harvard.edu.
Joan Maling
Brandeis University
" The Innocent Dative:
Morphological case, grammatical functions, and thematic roles"
(1) a. Der Stoff w‰scht sich gut.
the fabric(N) washes REFL well
'the fabric launders well'
b. *Er hilft sich schwer.
he(N) helps REFL difficultly.
Intended: 'he's difficult to help'
However, the fact that verbs which take goal arguments fail to undergo MF even when the goal is marked accusative suggests that the relevant restriction is thematic and has nothing to do with the dative case. Even more telling is the fact that lexical case does not block MF in Icelandic, a language in which (unlike German) verbal arguments bearing the theta role of theme may be marked dative or genitive. The independence of morphological case from both grammatical function and theta role in Icelandic allows us to tease apart these grammatical notions. I argue that lexical rules need to refer to the content of theta-roles and not just their relative position in a verb's theta-grid. I show that the widespread tendency to treat Theme as the default theta-role on direct objects has led to false claims about the nature of restrictions on various rules, e.g, nominalization in English (Roeper 1993) and depictive predicates in English (Hudson 1992, Baker 1997). Lastly I discuss the implications for Bakers's Uniformity of Theta Assignment Hypothesis (UTAH), which requires that verbal arguments bearing similar thematic roles be expressed in similar initial structural positions both within and across languages (Baker 1997:104-5).
For more information, contact linguist@bu.edu.
MIT LING-LUNCH
Thursday, April 13, 2000 at 12:10, MIT E39-335 (conference room)John Frampton,
Northeastern University"Agreement is Feature Sharing" (This is joint work with Sam Gutmann.)
I will begin by discussing some significant problems with the feature checking mechanism presented in Chomksy's "Derivation by Phase." In order to overcome these problems, a new view of agreement is proposed under which agreement between heads consists of shared features. This, in turn, leads to a new and much simpler conception of chains, independent of movement.
In contrast to much recent minimalist work, which attempts to establish the architecture of the system and the mechanics of feature checking (or valuation) in such a way that the properties of NP-chains are a natural consequence, we return to a GB conception in which chain well-formedness plays a crucial role in the syntax. Chains which are not well-formed are uninterpretable. This has many consequences. Worthy of note are an account of the interaction of the case-assigning properties of preverbs with their selectional properites which Burzio's Generalization describes, the absence of object expletives, and the licensing of quirky case NP-chains in Icelandic.
The framework has no comparison of derivation, backtracking, or lookahead. It has no numerations. There is simply the alternate application of Select and Agree. Since there is no comparison of derivations, uninterpretability simply leads to a failed derivation. Overgenerate and then filter is incompatible with the point of view that Gutmann and I have advocated that the derivational computation should be efficient. It therefore falls on local design features (selection in particular) to ensure that efficient computation can be maintained and derivations are built in such a way that they do not produce uninterpretable chains and output filtering can be avoided.
For more information, contact karlos@MIT.EDU.
Harvard University
Graduate Student Society Talk
Javier MartÌn-Gonz·lez
Department of Linguistics, Harvard University"Sentential 'no' " in Spanish negative sentences is incompatible with preverbal n- phrases (Laka 1990), its occurrence being obligatory otherwise (Bosque 1980, 1994, Laka 1990, Zanuttini 1991):
(1) a. *(No) vino nadie.
not came nobody
'Nobody came.'
b. Nadie (*no) vino.
nobody not came
'Nobody came.'
However, sentential 'no' seems to be optional when the preverbal n- phrase is a Topic in a Clitic-Left Dislocated construction:
(2)A ninguna de estas personas, (no) las vi en la fiesta.
PREP none of these people not them saw-I in the party
'I didn't see any of these people at the party.'
(Lit. 'None of these people, I didn't see them at the party.')
Furthermore, there are two cases in which sentential 'no' is obligatory in spite of the fact that the sentence contains a preverbal n-phrase. One such case takes place when the Topic n-phrase appears to have been displaced from an ambedded wh-clause:
(3)A ninguna de estas personas, dime por quÈ *(no) las
PREP none of these people tell-me why not them
invitaste.
invited-you
'Tell me why you didn't invite any of these people.'
(Lit. 'None of the se people, tell me why you didn't invite them.')
The second case is instanced by sentences where the Topic n- phrase occurs between a non-wh complementizer and its doubled counterpart (i.e. 'que' +n-phrase +'que'):
(4) Me dijeron que, a ninguno de ellos, que *(no) los
me told-they that PREP none of them that not them
invitaste.
invited-you
'They told me that you didn't invite any of them.'
(Lit. 'They told me that, none of them, that you didn't invite them.')
Except for those cases like (3) involving embedded wh-clauses, analysed in SuÒer 1992, no proposals have been made for cases like (2) and (4).
In the present proposal, Move and Merge (Chomsky 1995) are adopted as the two distinct derivational processes that account for the preverbal position of n-phrases. Secondly, it is argued that Topic negative phrases, unlike non-Topic negative phrases- contain a clitic (overt or covert) as head and the negative phrase as Specifier (along the lines of Uriagereka 1995). It is also argued that Topic n-phrases may occur preverbally due to application of either Move or Merge, whereas non-Topic n-phrases -including non- arguments (Escobar Alvarez 1995)- can only appear in preverbal position due to application of Move.
In this paper it is also claimed that NegP in Spanish contains an uninterpretable feature that must be checked off before Spell-Out. Checking may be achieved by having overt 'no' as head of NegP, or by having an n-phrase move to Spec,NegP. This correctly predicts that sentences without preverbal n-phrases (e.g. 1a) will require an overt sentential 'no'. Furthermore, considering that preverbal non-Topics must have been moved to that position, it also makes the correct predictions about the obligatory absence of sentential 'no' in sentences with preverbal non-Topic n-phrases (e.g. 1b). Likewise, considering that Topics may occur preverbally due to Move or Merge, the optionality of 'no' in examples containing preverbal Topic n-phrases in examples like (2) is successfully predicted as well.
Finally, an attempt is made to account for the obligatory presence of 'no' in cases like those exemplified in (3) and (4). First of all, Rizzi's (1997) enriched left-periphery structure is adopted, together with his claim that fronted wh-phrases sit in Spec,FocP. Then, it is shown that the doubled form of the non-wh complementizer occupies Foc0 and it is claimed that an assertive Operator fills the Spec,FocP position in those cases. Furthermore, drawing from similarities between negative and interrogative sentences (Klima 1964, Haegeman 1995, Rizzi 1990), it is argued that the Focus head can also check the negative feature. Thus, any movement past the checking domain of Focus0 will constitute a violation of "shortest move" (especially under Fergusson's (1996) Shortest Move Requirement) and will result in a nonconvergent derivation. Therefore, the only possibility for a n-phrase to occur in some position to the left of FocP is by application of Merge, which makes the presence of 'no' obligatory since the uninterpretable negative feature of NegP must be checked. This makes the correct predictions about the obligatoriness of sentential 'no' in cases of preverbal n-phrases to the left of wh-complementizers (ex.3), and between non-wh complementizers and their doubled form (ex.4).
Thus, by making use of claims and postulates that are independently motivated, non-language and non-construction specific, the proposed analysis offers an elegant account for all the "new" cases (ex. 2-4), plus the familiar instances of (non- )occurrence of sentential 'no' in Spanish negative sentences (ex.1).
You can obtain directions and other information at: http://www.fas.harvard.edu/~lingpub/resources/publtalk.htm. For more information, contact szczeg@fas.harvard.edu
MIT RLE Speech Communication Group Seminar
Grzegorz Dogil
University of Stuttgart (Experimental Phonetics)
McGill University (visiting - Speech Communication; Motor Control Lab Linguistics)THE SPEAKING BRAIN
For more information, contact czoll@MIT.EDU
MIT Phonology Circle
David Embick and Morris Halle "On Syllabotonic Verse in Latin"
For more information, contact czoll@mit.edu.
Marco Haverkort,
Boston University & University of Groningen"Modularity and the Neurology of Language" In this paper I will discuss a number of experiments looking at the comprehension of sentences of different degrees of complexity, making use of positron emission tomography (PET) and functional magnetic, which cast serious doubt on the correctness of the generally accepted picture. The following issues will be discussed:
* The functions of the two traditionally distinguished language areas are not as clear-cut as is usually assumed, i.e. Broca's area underlying syntax and Wernicke's area underlying semantics. Broca's area, for instance, becomes activated when subjects read word lists as well as when they read complex sentences.
* There is evidence that there are more than just these two areas in the left hemisphere involved in language comprehension. The anterior temporal lobe, for instance, is more highly activated when subjects read sentences than when they read word lists, indicating that it supports sentence comprehension.
* It will be shown that areas in the right hemisphere, especially the right frontal lobe, and also the cerebellum are involved in language processing in addition to the traditionally distinguished left hemisphere areas. The right frontal lobe, for instance, is activated when subjects have to read sentences which contain a lexically ambiguous word that is disambiguated to the non-preferred interpretation later in the sentence.
* Language processing shares components with several non-linguistic cognitive systems, necessitating a revision of the notion of modularity. A comparison of subject and object relative clauses, for instance, shows that the visual cortex becomes activated in the latter, but not the former; the same area has been shown to be active in visual attention tasks. This suggests that the language module is not completely informationally encapsulated: some information internal to the language 'module' is available to affect processes in other cognitive 'modules'.
In conclusion, the function of a number of traditional language areas in the brain has to be re-evaluated; moreover, a number of additional areas, both in the left and the right hemisphere have been shown to be active in language understanding. As a result of these findings, the 'language module' must be conceived of as a much more complex network of functionally specialized areas than has traditionally been assumed in linguistics; some of these areas may interact with other cognitively specialized areas/networks.
For more information, contact linguist@bu.edu.
MIT Linguistics Colloquium
Mark Baker
Rutgers University"On the Universality of the Lexical Category Distinctions"
For more information, contact asrackow@mit.edu.
MIT LING-LUNCH
Thursday, April 6, 2000 at 12:10, MIT E39-335 (conference room)Arhonto Terzi
Technological Educational Institute of Patras"Clitic Misplacement in Early and SLI Language" We collected data from early and SLI Cypriot Greek and we offer a syntactic account of the phenomenon based on theoretical claims in Martins (1994), Uriagereka (1995) and Terzi (1999) regarding finite enclisis in this type of languages. We consider (finite) V-cl order to be the result of overt Verb movement to F/M, a functional head higher than T in the clausal structure. When the features of M are checked via some other mechanism (such as Merge or Move of an inflectional particle) no need for V-to-M movement arises and the result is proclisis. We claim that the extensive enclisis manifested by early and SLI Cypriot Greek is due to extensive V-to-M movement, a consequence of the fact that children have not grasped the properties of the particles that may check the features of M in their language. Empirical support for this claim is provided by the frequent omission of the subjunctive marker and the use of the inappropriate Negative marker. We believe that these errors argue for incomplete knowledge in terms of phrasal vs. head status of the relevant particles rather than in terms of their feature strength. This belief is supported empirically (Wexler p.c., concerning early Swiss German and English) and conceptually if one is willing to accept that (children know that) the language faculty does not allow operations such as right adjunction (Kayne 1994). Finally our findings do not argue for two different developmental patterns, each one involved in early and SLI Cypriot Greek respectively.
For more information, contact karlos@MIT.EDU.
Harvard University
Linguistics Talk
Maya Arad
Harvard University"Roots Across Templates in Hebrew "
For more information, contact violette@fas.harvard.edu.
Harvard University
Linguistics Talk
Cheryl Zoll
MIT"Reduplication as Morphological Doubling"
For more information, contact violette@fas.harvard.edu.
|
|
 |
is starting up again in Spring 2000 ! |
|
Tuesday, April 4 at 7:30 PM "The Neurology of Language: CAS 220 Click here for more information. |
||
MIT Phonology Circle
Caro Struijke
University of Maryland"Feature Movement and Segment-based Feature Correspondence" In the original conception of Correspondence Theory (McCarthy and Prince 1995) it is assumed that features are attributes of segments, and only segments can enter in correspondence relations. Featural faithfulness is enforced by Identity constraints which require corresponding segments to be identical with respect to feature specifications. This conception of featural faithfulness cannot account for feature movement because features cannot be preserved independent of segments. Feature preservation must be due to a faithfulness constraint, yet faithfulness constraints are violated because corresponding segments are not identical.
I will argue in this talk that a segment-based approach of featural correspondence can account for feature movement if we assume that Faithfulness constraints regulate recoverability of input material from the output, rather than identity between the two (Struijke 1998, 2000). Given an input segment and a single identical output correspondent, both identity and recoverability are achieved. However, in multiple correspondence, recoverability does not imply identity. That is, if one output correspondent is identical to an input segment but a second output correspondent is not, recoverability is served, but identity is not achieved. I argue that featural faithfulness constraints are satisfied in such a situation, because the input feature is present in (or recoverable from) the output.
In Cuzco Quechua /tala/ -> [tala] the input liquid is in correspondence with both output consonants. Faithfulness on the laryngeal feature is achieved, because one correspondent of the liquid carries the feature.
For more information, contact czoll@mit.edu.
Kristine Strand
Brandeis UniversityMetrical Phonology - From Theory to Therapy
Abstract: Metrical analysis in phonology has been shown to account for syllable omissions in the speech of young typically developing children and children with phonological delays better than more traditional accounts. In this talk, aspects of metrical phonology will be described as an example of the useful integration of linguistic analysis into clinical contexts.
For more information, contact linguist@bu.edu.
MIT Linguistics Colloquium
Larry Hyman
UC Berkeley"Cyclic identity and non-identity effects in Bantu"
For more information, contact asrackow@mit.edu.
MIT
The Biannual GASLA (Generative Approaches to Second Language Acquisition) ConferenceThe Theme : The Bilingual* Brain
(i.e: * = Multilingual)
For more information, go to http://web.mit.edu/fll/www/news/Conf.html.
MIT LING-LUNCH
Thursday, March 30, 2000 at 12:10, MIT E39-335 (conference room)William J. Idsardi (University of Delaware)
and Sun-Hoi Kim (MIT)"The phonology-syntax interface in North Kyungsang Korean"
For more information, contact karlos@MIT.EDU.
MIT Phonology Circle
Paul DeLacy
UMass Amherst"The Effects of Interpretation on Form" I aim to show that the proposal deals with a certain kind of "duplication problem": Suppose a structure S is never produced by the grammar. There are (at least) two ways to account for this fact: (1) there is an inviolable syntactic/phonological constraint *S, or (2) S is uninterpretable. Whether (1) or (2) is responsible for S's ill-formedness, it would be grossly redundant to propose that S is _both_ uninterpretable and there is a constraint *S as well. (Other reasons apart from redundancy for why this is an undesirable situation will also be discussed).
The problem is that such situations do seem to arise. For example, crossed association lines in autosegmental phonology have been argued to be uninterpretable (Sagey 1988 and others). However, there are reasons to think that crossed lines are also banned by a phonological constraint:
1. Economy: In certain circumstances a crossed-line structure is the most economical/harmonic structure possible. Without an active prohibition against crossed lines, the grammar should always produce them, predicting that certain common phonological operations are impossible (e.g. metathesis).
2. Triggering and Blocking: Crossed-lines trigger and block phonological processes; the grammar actively avoids crossed- lines by either blocking an otherwise general process that would produce crossed-lines, or by triggering a repair process. Without an actual constraint against crossed-lines, though, the grammar should not react to their presence at all.
In summary, it seems necessary to ban the same structure in two different components - once in the Phonology, and again in the Interpretive component. Such a situation is similar to the Duplication Problem in serial theories, where the same structure is banned in both the lexicon and in the phonological component (see SPE p.382).
I will argue that such a duplication is undesirable and unnecessary: with the new conception of Syntax/Phonology-Interpretation interaction proposed above, there is no need to duplicate the effects of the Interpretative component in syntactic/phonological constraints.
Attention will also be given to related cases (e.g. Phonology: strict locality of feature spreading, full featural specification, Syntax: binary branching trees, theta theory).
The implications of this proposal for several issues of current interest will also be discussed, including:
1. Levels: Whether the grammar can/must be serial (e.g. SPE vs OT).
2. Freedom of Structure-Building: How free the construction of syntactic/phonological structures are - whether they are very limited (e.g. X-bar theory) or laissez-faire (e.g. OT^Rs Freedom of Analysis).
3. Evolution: How the proposal relates to the origin of the language faculty - whether the language faculty is an evolutionary adaptation or exaptation, and how adaptive the Syntax/Phonology is.
For more information, contact czoll@mit.edu or delacy@linguist.umass.edu.
Harvard University
Graduate Student Society Talk
Juan Uriagereka
Department of Linguistics, University of Maryland"Reprojection: An Argument for Derivations "
You can obtain directions and other information at: http://www.fas.harvard.edu/~lingpub/resources/publtalk.htm. For more information, contact szczeg@fas.harvard.edu
Harvard University
Graduate Student Society Talk
Ju-Eun Lee
Department of Linguistics, Harvard University"A Note on Verb Movement, Scrambling, and Scope Interaction in Korean Negation "
You can obtain directions and other information at: http://www.fas.harvard.edu/~lingpub/resources/publtalk.htm. For more information, contact szczeg@fas.harvard.edu
MIT LING-LUNCH
Thursday, March 16, 2000 at 12:10, MIT E39-335 (conference room)Benjamin Bruening
& Andrea RackowskiNorvin Richards "Configurationality and Object Shift in Algonquian" "Some Notes on Conjunct and Independent Orders in Wampanoag" Abstract: In this talk we examine an unusual pattern of agreement in Eastern Algonquian languages, members of a family that has been described as non-configurational and polysynthetic (e.g., Russell and Reinholtz 1997). The details of the agreement pattern cannot be captured by standard approaches to this type of language (e.g., the Pronominal Argument Hypothesis of Jelinek 1984 and the Polysynthesis Parameter of Baker 1996); configurationality and movement are implicated instead. This finding has wide-ranging implications for approaches to polysynthesis and (non)configurationality.
The Eastern Algonquian languages Wampanoag (a.k.a. Massachusett, extinct), Western Abenaki, and Delaware retain a distinction reconstructed by Goddard (1967) for Proto-Algonquian, in the form of an agreement morpheme glossed as "NonSpec" or "SpecInan" below:
Wampanoag
1. Nu-ssoh-=F4-m-un(=F4n) J8nesogn-ag
1-send.out-DIR-NonSpec-1PL juryman-PL
"We sent out jurymen." (Goddard and Bragdon 1988, 17:14)
2. Koshkuhtaukquainnin shanuh wut-ahtauw-un=E2-(a)sh mitcheme.
K. these.Inan 3-have-SpecInan-Inan.PL forever
"Koshkuhtaukquainnin has these forever." (Goddard and Bragdon 1988, 74:6-7)
As shown in the contrast between (1) and (2), the choice between -m- and -un=E2- marks a contrast in the specificity and animacy of a non-proximate third person argument. In (1), the non-specific plural "jurymen" triggers the appearance of the -m- affix. In (2), by contrast, the demonstrative "these" (underlined) triggers the appearance of the -un=E2- allomorph, indicating a specific inanimate.
Of interest is the behavior of this morpheme when there is more than one non-proximate third person argument: the morpheme agrees with the lowest specific argument of the clause. When a second object is added, as in a ditransitive or in an adverbial construction called the subordinative, agreement is with the lowest argument if it is specific: if the direct object is non-specific, the verb will agree with the indirect object; but if the direct object is specific, -un=E2- will appear regardless of the specificity of the indirect object. Furthermore, in a construction known as the inverse, a non-proximate third-person argument appears as the subject of the verb; if there is no lower specific third person argument, the verb will then agree with the subject, but if there is a lower argument, it will again take over agreement. Thus specificity agreement cannot be explained in terms of grammatical roles or argument structure, but only in terms of simple structural dominance. Even more importantly, specificity is only distinguished for overt noun phrases, implying hierarchical relations among them. As in the Germanic languages, pronouns are always specific. No approach to polysynthetic languages that embraces the Pronominal Argument Hypothesis could explain the pattern we see.
Word order appears to be sensitive to specificity as well: specific arguments appear preverbally more than 50% of the time, but nonspecifics almost never do. This is similar to their behavior in German, for example, where specific objects, including pronouns, obligatorily shift to the left of the VP. If the correct account of Germanic object shift involves movement, we should expect movement of noun phrases to be the correct analysis of the Algonquian pattern as well.
Abstract: Like a number of other Algonquian languages, Wampanoag distinguishes between so-called "Conjunct" and "Independent" order verbs:
(1) a. nu-naw-uq-wunon-ak [Independent]
1 see INV 1PL PL
'They see us (excl.)'
(1) b. naw-uqeey-ak-eek [Conjunct]
see INV 1PL PL
'They see us (excl.)'
The morphological differences between the forms mainly have to do with agreement; one salient difference is that Independent verbs may have an agreement prefix agreeing with the "central" argument (the argument which is highest on an animacy hierarchy), while Conjunct verbs lack such a prefix. I will discuss the syntactic conditions on the distribution of these forms, which appear in relative clauses, some but not all kinds of wh-questions, and certain kinds of adjunct clauses. I will defend the claim that Conjunct order verbs are lower in the functional structure than their Independent order counterparts; some instances of the use of the Conjunct, I will argue, correspond to the comparatively well-studied phenomenon of Anti-agreement.
For more information, contact karlos@MIT.EDU.
Refreshments will be served.
Dr. Lori Markson
MIT"How special is word learning"
MIT Phonology Circle
Jennifer Smith
UMass Amherst"Prominence, Augmentation, and Neutralization" In positional neutralization, 'strong' positions license contrasts that are neutralized in 'weak' positions in the same language (Trubetzkoy 1939); for example, stressed syllables often license more vowel contrasts than unstressed syllables. OT analyses of positional neutralization can be divided into two major approaches: Positional Faithfulness (Selkirk 1994; Beckman 1995, 1998; Casali 1996) and Positional Markedness (Lombardi 1995; Steriade 1997; Zoll 1996, 1997, 1998). Positional Faithfulness uses high-ranking faithfulness constraints specific to strong positions (F-str), protecting strong positions from neutralization. Some versions of Positional Markedness use high-ranking markedness constraints specific to weak positions (M-wk), forcing neutralization in weak positions only.
Crucially, any approach to positional neutralization must exclude markedness constraints that are specific to strong positions (M-str). Such constraints would allow neutralization to occur in strong positions only, wrongly predicting (e.g.) languages with more vowel contrasts in unstressed syllables than in stressed syllables.
However, there do exist phonological processes that specifically target strong positions: augmentation phenomena. Examples are given in (1).
(1) Augmentation effects:
(a) Stressed syllables become heavy (Chierchia 1981)
(b) Roots obey Word Minimality (McCarthy & Prince 1986)
(c) Heads of feet become heavy (Iambic Lengthening; Hayes 1995)
Augmentation processes affect only strong positions, so they must be driven by constraints of the form M-str. Here lies the paradox: including M-str constraints in the grammar seems to make the wrong predictions for positional neutralization, but these constraints are needed in order to account for augmentation.
I argue that closer consideration of the role of prominence in phonology allows us to retain a restrictive theory of positional neutralization without incorrectly excluding augmentation. The M-str constraints that are attested serve a specific purpose: they act to increase the perceptual prominence of the strong positions with which they are associated. I propose that this is no coincidence, but is in fact a restriction imposed by the grammar. Namely, M-str constraints can exist only if they do enhance the prominence of their strong positions. Since positional neutralization processes involve featural neutralization and segmental deletion, they do not enhance prominence. Therefore, there can be no M-str constraints of this sort.
With this universal restriction on the form of M-str constraints, both positional neutralization and augmentation can be accounted for within the same theoretical framework.
For more information, contact czoll@mit.edu.
MIT LING-LUNCH
Thursday, March 9, 2000 at 12:10, MIT E39-335 (conference room)Susanne Gahl,
Harvard University"On the role of lexical biases in aphasic sentence comprehension" Many studies have shown convincingly that failure to utilize lexical information can lead to abnormalities in sentence processing. The present study argues that processing abnormalities can also ensue when the parser is relying too heavily on lexical information: in some listeners, comprehension may rely on lexical biases at the expense of syntactic parsing.
The hypothesis to be tested in this experiment was that aphasics with different clinical types of aphasia were differentially sensitive to verb biases. A plausibility judgement task was used to probe for sensitivity to lexical biases. Subjects were asked to judge the plausibility of transitive and intransitive sentences. In one condition, the sentence type (transitive or intransitive) matched the lexical bias (transitive or intransitive) of the main verb. In a second condition, the transitivity of the sentence was the opposite of the verb's bias. Our results indicate that some aphasic patients, as well as nonaphasic controls, make significantly more errors and show longer reaction times on sentences in which sentence structure and lexical bias do not match.
For more information, contact karlos@MIT.EDU.
Harvard University
Graduate Student Society Talk
Ken Nakatani
Department of Linguistics, Harvard"The processing complexity of nested structures in Japanese"
You can obtain directions at: http://www.fas.harvard.edu/~lingpub/resources/publtalk.htm. For more information, contact szczeg@fas.harvard.edu
(i) Does verb focus differ from argument focus in locality (in overt movement constructions) or with respect to focus projection (with or without overt movement)? Does it create a head-chain or a phrasal dependency? Is there categorial coercion e.g. nominalization?
(ii) In movement constructions, how is verb focus related to the C-system?
(iii) Does verb focus affect aspectuality or Aktionsart, e.g. is there stative-to-inchoative coercion or other eventive entailments such as factive-ness? Are there transitivity effects, or restrictions with complex predicates?
(iv) How does verb focus interact with sentential operators like yes/no questions, negation, tense, adverbs and modality?
Discussion will seek to establish comparative data or research questionnaires revealing micro-parametric differences in the above properties, across Niger-Congo or more widely. Participation is therefore open with the provision that participants either speak their respective object langauges or that they attend together with fluent-speaker collaborators (who may not necessarily be linguists themselves). The form of publication, if any, resulting from the workshop will be decided subsequently by participants.
Victor Manfredi
manfredi@bu.edu617-441-2534
MIT Phonology Circle
Miklos Torkenczy
MIT"Defective stems in Hungarian: a case of absolute phonological ungrammaticality"
For more information, contact czoll@mit.edu.
Date: Mon, 28 Feb 2000 17:29:31 -0500
To: Carol Neidle <carol@louis-xiv.bu.edu>
From: hutch@bu.edu (John P. Hutchison)
Subject: ACAL2000 ProgramHere is the conference program for your information (as it stands right now):
Thursday March 2nd - George Sherman Union, 775 Commonwealth Ave.
10 a.m.-Registration: 2nd floor lobby (through noon Saturday)
Book displays: Ziskind Lounge (through 5 p.m. Saturday)
Session 1c: Literacy 1
11:40-M.A. Mohamed (UTL), Indigenous medium as a tool in education: the case of Swahili
12:20-T. Moyo (U Zululand), The language-in-education policy and language politics in MalawiSession 2a: Morphoyntax
2 p.m.-S. Obeng, E. Yankey (Indiana), Negation in Nzema
2:40-J. Mugane (Ohio), Split categories in Gikuyu
3:20-M.J.C. Echeruo (Syracuse), Igbo gbaa verbsSession 2c: Literacy 2
2 p.m.-I. Diallo (Ouagadougou), J. Hutchison (Boston), L'impact du français sur l'enseignement primaire au Burkina Faso
2:40-K. Kone (SUNY-Cortland), Do proverbs constitute Africa's contribution to philosophy?
3:20-O. Alidou (Cleveland State U), The development of literaciesPlenary session
4:15-Business meeting
Friday March 3rd - George Sherman Union, 775 Commonwealth Ave.
Session 3a: Checking theory
9:30-E. Bokamba (Illinois), Head movement, agreement and Case-marking in Bantu languages
10:10-J.L. Jake (Midlands), Evidence for movement in finite control structures in Fur
10:50-J. Ndayiragije (W. Ontario), Deriving syntactic ergativitySession 3b: Tone-bearing units
9:30-M. Mous (Leiden), K. Traore (Ouagadougou), Contour tones and downstep in Seme (Kru, Burkina Faso)
10:10-J. Ntihirageza (Chicago), Quantity-sensitivity in Kirundi
10:50-M. Paster (OSU), The TBU in G„Session 3c: Grammaticalizaton 1
9:30-M. Sepota (U of the North), The Northern Sotho preposition: a lexical or non-lexical category?
10:10-M. Meeuwis (Antwerp), The Lingala verb *kolinga* 'to want' at an early stage of grammaticalization
10:50-A. Bell (Cornell), Bipartite negation in Afrikaans: a Khoesan interference feature?Session 3d: Language death
9:30-H. Fleming (Boston), Four endangered languages of unusual saliency
10:10-Z. Leyew (Kˆln), Kemantney, a language on the verge of extinction: a sketch of its grammar
10:50-B. Connell (Oxford), Phonetic/phonological variation and language deathSession 4a: Kwa syntax 1
11:40-M.C. Baker (Rutgers), A. Kawu (Rutgers), O.T. Stewart (UBC), A comparative analysis of serial verb constructions in Kwa languages
12:20-G. Masagbor (UQAM), Asymmetry in Yekhee double definitness construction1 p.m.-U.P. Ihionu (Maryland), Transitivity and multiple predicate constructions
Session 4b: OT 1
11:40-A. Akinlabi (Rutgers), Asymmetries in reduplicative and nonreduplicative defaults
12:20-C. Zoll (MIT), Output constraints on tone mapping
1 p.m.-C.R.C. Sheedy (UBC), Grammatical tones in Edo: an Optimality-Theoretical accountSession 4c: Sociolinguistics 1
11:40-E. Yankey (Indiana), A study of address forms in Nzema: a sociolinguistic and pragmatic approach
12:20-K.E. Essizewa (NYU), Forms of address in Kabiye: a case study of borrowing kinship terms of addressSession 4d: Planning
11:40-N.E. Phaswana (MSU), 11=1+1=2: is it true for South Africa's 11 official-language policy?
12:20-L.L. Muaka (Col. de Mexico), The effects of language attitudes on the development of the national language in Kenya
1 p.m.-A.S. Bobda (Illinois/Yaounde 1), Towards an English pronunciation atlas for AfricaSession 5a: Kwa syntax 2
2:40-E.O. Aboh (GenËve), Focus constructions across Kwa
3:20-O. Oyelaran (WSSU), V. Manfredi (Boston), Weak focus-movement islands in Yoruba
4 p.m.-R.-M. DÈchaine (UBC), Layered Comp in Yoruba
4:40-O. Ajiboye, R.-M. DÈchaine, O.T. Stewart (UBC), The syntax of nominalization: evidence from Edo and YorubaSession 5b: OT 2
2:40-L.J. Downing (UBC), Bukusu reduplication
3:20-D. Brassil (UCSD), Kirundi reduplicated adjectives
4 p.m.-M. Cahill (SIL), Tone polarity in Konni: an Optimality-Theoretic analysis
4:40-B. McCall (Cleveland State), Phonological ordering of elements in the Luo verbSession 5c: Grammaticalization 2; Sociolinguistics 2
2:40-B.B. Mbom (CUNY), Towards a typological appraisal of Basaa gender
3:20-E. Ochola (South Carolina), Evidence for a composite grammatical frame from Luo/English codeswitching
4 p.m.-S. Obeng (Indiana), Some pragmatic properties of Ghanaian political discourse: a pragmalinguistic analysis of intertextuality, evidentiality and contrast establishment
4:40-F. Ngom (Illinois), Religious and linguistic behaviors as means of anticolonial resistance in Subsaharan Africa: the case of the Murids in SenegalSession 5d: Area studies
2:40-N. Cyffer (Wien), Common linguistic features in languages of the wider Lake Chad area
3:20-G. Hudson (MSU), Alphabeticness, uniqueness and alleged deficiency of the Ethiopic writing system
4 p.m.-M. Garba (Temple), The Ancient Egyptian sign for lateral /L/: *L-t 'time, moment'
4:40-D. Duke (UTArlington/SIL), The Bayaka Pygmies and their 'masters': factors effecting language shiftPlenary session
5:30-Ngugi wa Thiong'o (NYU), "After Asmara: the future of African languages and literatures"
7 p.m.-Public reception
Saturday March 4th - George Sherman Union, 775 Commonwealth Ave.
Session 6a: Verb focus
9 a.m.-C. Stanley-Thorne (Moncton), "The coming that they came^J": verb topicalization and verb and sentence focus in Tikar
9:40-N. Hall (UMass Amherst), Factive relative clauses in BafutSession 6b: Loan phonology
9:00-M. Bamba (Penn), Deletion of the medial consonant and nasal stability in Mawukakan
9:40-A. Kawu (Rutgers/Ilorin), Faithfulness and markedness in Benue-Congo loan phonologySession 6c: Afroasiatic syntax
9 a.m.-G. Halefom, J. Lumsden (UQAM), Unusual agreement in Amharic
9:40-A. Benmamoun (Illinois), Agreement asymmetries without noun phrases and sentencesSession 6d: Englishes
9 a.m.-J.V. Singler (NYU), Liberian Settler English: how African is it?
9:40-M. Mutonya (Washington U), African English vowel systems: acoustic and perceptual analysisSession 7a: CP structure &c
10:30-R. Schaefer (S. Illinois), F. Egbokhare (Ibadan), Emai's classificatory 'put' verbs
11:10-D. Ngonyani (MSU), The relative marker in Bantu: a pronoun or complementizer?
11:50-B. Oduntan (Iowa), Wh-phrases and complementizers in Yoruba: evidence for a multi-layered complementizer systemSession 7b: Grammatical tone
10:30-D. Odden (Troms¯/OSU), Tachoni verbal tonology
11:10-L. Bickmore (Albany), Chilungu verbal tonology
11:50-L.M. Hyman, K.J. Olawsky (Berkeley), Dagbani verb tonologySession 7c: Afroasiatic phonology
10:30-A. Idrissi (UQAM), Phonological transfer in root-and-pattern morphology
11:10-S. Rose (UCSD), Roots, radicals and Semitic reduplication
11:50-D. Petros-Banksira (MIT), Interactions of gemination and reduplicationSession 7d: The (mis)education of the Creole speaker
10:30-M. DeGraff (MIT), Creole morphology and the morphology of an ideology
11:10-J.Hudicourt-Barnes (TERC), Style of argumentation in Haitian Creole
11:50-L. Hogu (Hyde Park School), "Lodyans": a tool to enable success for secondary Haitian students with limited formal schoolingSession 8a: Afro-Atlantic 1
1:30-A. Schwegler (Costa Rica/UCI), The Black ritual curse: the art of 'sounding' in the Americas
2:10-K. Bilby (Smithsonian), Reevaluating the African lexical component of the Surinamese Maroon Creoles: the Aluku case
2:50-K.K.B. Fu-Kiau (Boston, Mass.), African diasporadic languages: unspoken but alive and powerful
3:30-M. Warner-Lewis (UWI), Challenges in translating Trinidad Yoruba song textsSession 8b: Applicatives &c
1:30-P. Mabugu (Edinburgh), The applicative construction in Chishona
2:10-K. Demuth, J. Bulkowski, Alaka Holla (Brown), Object drop in Bantu languages: implications for learning the argument structure of verbs
2:50-L. Makhubu (M.I Sultan T), The effect of the applicative affix *-el* in isiZulu intransitive verbs
3:30-S. Mchombo (Berkeley), On reciprocals in Bantu and the syntax-semantics interfaceSession 8c: Laryngeal phonology
1:30-M. Russell (Illinois), Phonetic aspects of tone displacement in Zulu
2:10-M. Bradshaw (Chicago), Voicing phenomena in Suma
2:50-A. Miller-Ockhuizen (OSU), Laryngeal-pharyngeal interactions
3:30-N. Clements, S. Osu (Paris), Sonorant stops: new evidence from IkwereSession 8d: Morphology
1:30-I.A. Traore (Bamako), Songhay personal pronouns
2:10-S. Sow (Niamey), Le morpheme *'en* en fulfude: essai d'analyse d'une modalite nominale
2:50-A. Kimenyi (CSU), The verb perfective marker and consonant mutation in Kikongo
3:30-M. Gimba (UCLA), Verb reduplication in BoleSession 9a: Afro-Atlantic 2
4:20-I. Miller (Schomburg Center), From West African Ekpe to Cuban Abakua: linguistic evidence for the presence of Africa in Cuba
5 p.m.-R. Gonzalez (Boston, Mass.), Cantos de Lucumi: interpretacion
5:40-J. Mason (Yoruba Theological Archministry), Problem areas in translating Yoruba texts
6:20-D. Dawson (American Museum of Natural History), Palo Mayombe: the visual language of a Kongo religion in CubaSession 9b: Capeverdean language in education
4:20-M. da Luz Goncalves (Boston Public Schools), The new orthographic system - ALUPEK: its conception, legislation, approach, tryout: the Boston Experience
5 p.m.-G. Goncalves (Boston Public Schools), Bilingual education and the Capeverdean Program - the law, experiences, successes and program model for the Diaspora
5:40-A. Lima (Boston Public Schools), Kriolu and language impaired children
6:20-L. Caswell (Harvard), CCI: an update - foundation, activities and curriculumSession 9c: Phonology
4:20-J. Riggle (UCLA), Relational markedness in Bantu vowel harmony
5 p.m.-E. Koffi (Houghton), Unnasalized vowels in nasalizable environments in Anyi
5:40-S. Lyon (UCSC), An OT account of consonant gradation in Fula
6:20-K.S. Olson (Chicago/SIL), Languages in which the labial flap is foundSession 9d^K4:20-M. Johnson, K. Demuth, S. Canon (Brown), Morphological tagging and glossing of Bantu language corpora
5 p.m.-A. Kimenyi (CSU), Life as journey metaphor in Kinyarwanda
5:40-Y. Morimoto (Stanford), The role of animacy and associational harmony in Bantu
6:20-S. Mchombo (Berkeley), Choppin' up ChichewaPlenary session
8 p.m.-Dinner-dance, catered by Restaurant "Sodade Terra", Jorge Hernandez Cultural Center, Inquilinos Boricuas en Accion, 85 West Newton Street, Boston.
Sunday March 5th - General Classroom Building, 750 Commonwealth Ave.
Session 10a
10:00-S. Makoni (Cape Town/Michigan), The South African constitution and multilingualism: an argument for the disinvention of African languages
Session 10b: Gikuyu orthography
10 a.m.-Workshop chaired by C. Githiora and open to all ACAL participants
Session 10c: Igbo
10 a.m.-M.A.A.N. Uwalaka (Ibadan), Tense and movement in Igbo
10:40-P.A. Nwachukwu (Nsuka), Case theory, the theta-criterion and Igbo inherent complement predicatesSession 11a:
11:40-S.A. Sow (Niamey), Forces et faiblesses de l'enseignement bilingue langue officielle/langues nationales: cas du Niger
Session 11b: Gikuyu orthography 2
11:40-[Workshop continues.]
Session 11c: Anaang
11:40-U.I. Idem (Uyo), The pro-drop parameter in Anaang
12:20- I. Icheji (Uyo), Anaang syllable structureSession 12b: Gikuyu orthography
2 p.m.-[Workshop concludes.]
Click here for less information
MIT LING-LUNCH
Thursday, March 2, 2000 at 12:10, MIT E39-335 (conference room)Marie-Helene Cote
MIT"Edge effects" In this talk I argue that the syllabic approach to edge effects is inadequate and develop an alternative approach based on the perceptual salience of consonants and cast in a phonetically-oriented OT framework (cf. Steriade 1997). A consonant deletes or triggers epenthesis when the cues that permit a listener to detect its presence are diminished; vowel deletion is blocked when deletion would leave a consonant with diminished perceptual cues. This approach accounts naturally for two uncovered generalizations concerning edge effects: 1) stops are subject to edge effects more frequently than other consonants, 2) edge effects are found at all prosodic levels, from the word to the utterance (and not only at the word level, on which past research has focused almost exclusively), and they are cumulative as we go up the prosodic hierarchy.
Two case studies illustrating these generalizations will be analyzed in more detail: French and Ondarroa Basque.
For more information, contact karlos@MIT.EDU.
Refreshments will be served.
Dr. Raquel Jaakkola
MIT"The Development of Productivity in Children's Early Language"
MIT Phonology Circle
Molly Homer
Brown University"Combinatorics and Word Length " The number of contrastive forms available to a lexicon can be increased in two ways: by enlarging the segmental inventory, or by allowing longer words. Basic combinatorics are used to show that lengthening words just a little can generate more lexical forms than doubling the segmental inventory. In other words, a language can compensate for a small segmental inventory by having slightly longer words, whereas the least reduction in word length is only offset by a large increase in the segmental inventory. This prediction is supported by Nettle's (1995) study which indicates that large differences in segmental inventories (e.g. Hindi:41 segments vs. Hawaiian: 18) tend to be balanced by small differences in mean word length (Hindi:5.75 segments vs. Hawaiian: 7.08). I show that his results can be derived from combinatoric principles, and explore other predictions from combinatoric reasoning.
One such prediction is that the number of potential morphemes in a language should increase exponentially with morpheme length. However, in Spanish and English, the number of actual morphemes does not increase exponentially with length, but rather increases gradually to a mean length, and then tapers off. Several factors might contribute to the discrepancy. For example phonotactic restrictions will reduce the number of potential morphemes. However, phonotactic restrictions should eliminate roughly equal numbers of morphemes at all lengths, so should not change the overall shape of the distribution curve. On the other hand, avoiding forms which contain sequences that could be mistaken for shorter morphemes will eliminate vast numbers of potential morphemes, and the impact should be greater for longer morphemes. An examination of 4-segment long base morphemes in English suggests that morphemes which could be misanalyzed as being morphologically complex are in fact disfavored: prosodic patterns which consist of more than one syllable (CVCV, VCVC,or VCCV) are much less common (only 17 %) than monosyllabic patterns (CCVC or CVCC).
For more information, contact czoll@mit.edu.
Linguistics Lecture
open to members of the Boston University community
Paul Hagstrom
Johns Hopkins University"Predicting how optional infinitives really are" Background: Research of the last 10 years has established that in many languages, children at around age 2 go through an "optional infinitive" stage, during which a high proportion of their main clause verbs are non-finite (often an infinitive form). Ken Wexler and colleagues have achieved a great deal of predictive and explanatory success by analyzing this phenomenon as resulting from a specific deficiency in the representation of these children's clauses. In particular, recent work suggests that the nonfinite forms occur because children drop tense or agreement from their otherwise fully adult-like syntactic representations. While in this stage, children produce both adult-like and non-adult-like verbal forms, hence the label: "optional" infinitives.
To begin, I will present results gathered from three children in the OI stage acquiring French. Considering the proportions with which these children produced tense, agreement, both, and neither, our data reveals a systematic course of development through the "optional infinitive" stage. We divide the data into three (sub)stages. In the first, finite verbs all realize tense, but not agreement. In the second, finite verbs are roughly split between realizing tense and agreement. In the third, finite verbs realize both tense and agreement. Of particular importance is the move from the first to the second stage, where we see that tense starts to be produced less frequently, apparently in favor of agreement. Put another way, the second stage shows signs of competition: where only one functional category can be represented, tense and agreement compete for that role. Because agreement is now "in the game", tense "suffers" a dip from its previous rate of realization in the first stage.
I will present an analysis in terms of a small set of partially ranked constraints whose rankings change over the course of the "optional infinitives" stage. Quantitative use of partial rankings (e.g., as discussed by Anttila 1997 and Reynolds 1994 in the context of adult variation) allows an explanation not only of the existence of the different forms children produce at this stage, but also of the relative frequencies with which they produce such forms. Finally, I will connect this analysis with recent work by Wexler, showing that this analysis is not only compatible with his results but in fact with his analysis as well. The benefit of partial constraint ranking is that, in contrast with existing analyses of the OI phenomenon (which are concerned with explaining only the possibility of the different forms), this analysis correctly predicts the observed changes in frequency among finite and non-finite forms.
Cheryl Zoll
Department of Linguistics and Philosophy, MIT"Output Driven Tone Mapping"
For more information, contact linguist@bu.edu.
Harvard University
Graduate Student Society Talk
Albert Costa
Department of Psychology, Harvard"The cognate facilitation effect:
Implications for models of lexical access"
You can obtain directions and download the relevant background paper at: http://www.fas.harvard.edu/~lingpub/publtalk.html . For more information, contact szczeg@fas.harvard.edu
MIT LING-LUNCH
Thursday, February 24, 2000 at 12:10, MIT E39-335 (conference room)Liina Pylkkänen "What Applicative Heads Apply To" (1) a. High Applicatives:
The applicative head relates an additional participant to the event described by the VP, i.e. it adds a theta-role.
Examples: Kichaga benefactive, Chichewa Instrumental, Kinyarwanda benefactive
b. Low Applicatives:
The applicative head relates an additional individual to the direct object.
Examples: English benefactive, Chichewa benefactive, Kinyarwanda locative, Finnish bene/malefactive
High applicatives are possible even when the applied argument does not enter into any relation with the direct object, as is shown by the Kichaga high benefactive below:
(1) High Applicative: Kichaga
N-a-i-lyi-i-a m-ka k-elya
FOC-1s-PR-eat-AP-FV 1-wife 7-food
'He is eating food for his wife' (Bresnan and Moshi 1993)
(wife doesn't get the food)
Low applicatives, however, are not. The English translation of (1) is ungrammatical and even the grammatical example in (2b) can only have an interpretation where it is at least intended that Bill gets the letter -- it cannot simply be the case that Bill benefits from the letter-writing event:
(2) a. Jill wrote Bill a letter.
b. *He ate his wife food. (i.e. for the benefit of the wife)
In this talk I show that in addition to accounting for the ungrammaticality of (2b) in English, the proposal in (1) straigtforwardly accounts for a host of interactions between transitivity and the ability to applicativize a verb not captured by previous theories. Also, I argue that puzzling asymmetries between so-called adversity "passives" and adversity causatives in Japanese are, in fact, applicative asymmetries and that they receive a natural explanation under the present proposal.
For more information, contact karlos@MIT.EDU.
Linguistics Lecture
open to members of the Boston University community
Philippe Prévost
Laval University"Truncation and missing inflection in early L2 grammars" I examine contradictory predictions made by the two hypotheses in spontaneous production longitudinal data from eight learners of L2 French and L2 German (four children and four adults). On the TH, non-finite forms shouldn't appear in CPs like questions and subordinates (they should be restricted to root declaratives), they shouldn't be found in a position preceding negation, and they shouldn't occur together with a clitic subject. Only finite forms should be found in these contexts since the projection of Infl is involved. Such restrictions do not follow from the MIH for which production of non-finite forms is not structurally determined but stems from difficulties with the surface realization of inflection.
Results indicate clear differences between child and adult learners. Indeed, the distribution of non-finite verbs seems to be structurally determined in L2 child grammar, i.e. non-finite verbs only appear when VP is the root, which means that children project truncated structures in early L2 acquisition. Such a conclusion is not warranted in the case of adult learners. In particular, they seem to use the infinitival marker as a substitute for finite inflections. Hence, non-finite form are found under high functional projections.
MIT LING-LUNCH
Thursday, February 10, 2000 at 12:10, MIT E39-335 (conference room)Sabine Iatridou and Kai von Fintel
MITTitle TBA
For more information, contact karlos@MIT.EDU.
for students of linguistics
Light refreshments will be served.
Panel discussion: jobs in linguistics For more information, contact HeatherL@bu.edu .
MIT Phonology Circle
Sam Rosenthall Weak Roots in Arabic The distribution of glides in Arabic verbs is shown to be a consequence of constraint satisfaction as defined by Optimality Theory (Prince and Smolensky 1993). The surface form of a weak verb follows from conflict between correspondence constraints (McCarthy and Prince 1995) and constraints on glide syllabification. In brief, constraints determine whether a glide is syllabified as a syllable peak or in a syllable margin and these constraints conflict with input/output faithfulness constraints leading to glide deletion. Glides surface (either as a peak or in a margin) as a consequence of higher ranking constraints compelling surface violations of the glide syllabification constraints. As a result, glide alternations in Arabic follow from an independent theory of glide syllabification interacting with faithfulness constraints.
Input/output correspondence alone cannot account for the full range of weak verb phonology. Inter-candidate correspondence as defined by McCarthy's (1998) Sympathy is used to account for "opaque" glide vocalization, and has consequences for other aspects of the Optimality-Theoretic analysis of weak verbs. Optimality Theory provides an account of weak verbs that cannot be achieved in procedural approaches to phonology because there is no set of ordered rules that captures the phonology of both weak and strong verbs.
For more information, contact czoll@mit.edu.
Dr. David Caplan
Dept. of Neurology, MGH"Localization of Syntactic Processing"
Dr. Dawn MacLaughlin
Visiting Assistant Professor of Linguistics, Boston University"The critical period hypothesis
in first and second language acquisition"
