Congrès de l’ACL 2014 CLA Conference...

L’Association canadienne de linguistique tiendra son congrès de 2014 lors duCongrès des sciences humaines à l'Université Brock, St. Catharines (ON), dusamedi 24 mai au lundi 26 mai 2014.

!e Canadian Linguistic Association will hold its 2014 conference as part of theCongress of the Humanities and Social Sciences at Brock University, St.Catharines, ON, from Saturday May 24 to Monday May 26, 2014.

Congrès de l’ACL 2014

CLA Conference 2014

Programmedernière mise à jour: 17 mai 2014 | latest update: May 17, 2014

Samedi 24 mai | Saturday, May 24 255 256 257

Acquisition (L1) Variation & changement | Variation & change Syntaxe | Syntax

9:00–9:30 Yvan Rose (MUN)

Interfaces entre domaines phonétique et phonologique dansl’acquisition de la phonologie

Kazuya Bamba (Toronto)

!e interaction between impersonal re"exives and syntactic #ange

Tomokazu Takehisa (NUPALS)

Non-selected arguments and the ethical strategy

9:30–10:00 Marina Sherkina-Lieber (Carleton)

Monolingual and bilingual #ildren’s production of Russian embeddedyes–no questions

Alena Barysevich (York)

Emergence des normes communautaires : cas de la variation lexicale

Carlos de Cuba (Calgary)

In defense of the truncation hypothesis for main clause phenomena

10:00–10:30 Anna Frolova (Toronto)

Développement de la transitivité verbale en russe L1

Philip Comeau (O!awa) & Anne-José Villeneuve (Toronto)

!e expression of future temporal reference in Picardie Fren#

Paul Poirier (Toronto)

Malay/Indonesian voice and pseudo-incorporation

10:30–10:45 PAUSE | BREAK

255 256 257

Acquisition (L2+) Pragmatique | Pragmatics Syntaxe | Syntax

10:45–11:15 Johannes Knaus & Mary Grantham O’Brien (Calgary)

Stress and morphology in second language production and processing

Johannes Heim, Hermann Keupdjio, Zoe Wai-Man Lam, AdrianaOsa-Gómez & Martina Wiltschko (UBC)

How to do things with particles

Julianne Doner (Toronto)

Dimensions of variation of the EPP

11:15–11:45 Gabrielle Klassen & María-Cristina Cuervo (Toronto)

An imperfect representation: !e preterit–imperfect contrast insyntactic theory and SLA

E. Allyn Smith (UQAM), Laia Mayol (UPF) & ElenaCastroviejo-Miró (CSIC)

Di$erences between predicates of personal taste and epistemic modalsacross languages

Daniel Milway (Toronto)

Null pronouns in English: evidence from particle verb constructions

11:45–12:15 Ma!hew Patience (Toronto)

!e L3 acquisition of Spanish rhotics by native Mandarin speakers

Naomi Francis (Toronto)

!is predicate is tasty: Predicates of personal taste, faultlessdisagreement, and the ideal judge

12:15–1:30 DÉJEUNER | LUNCH

Session des a!iches | Poster session North Hallway

1:30–3:00 Ewelina Barski (SUNY Brockport)Is there a di$erence? Interpretation in advanced vs. intermediate Hispanic heritage speakersVeranika Barysevich (Western)L’accentuation des “petits” motsPaul De Decker (MUN)!e nasal ‘ash’ system in English: how does it get so tense?Veena Dwivedi, Hope Magnus, Leslie Rowland & Kaitlin Curtiss (Brock)!e cost of a%ention in semantic processingJulie Goncharov (Toronto)Many and determiner systems in English and RussianMargaret Grant (McGill), Julie!e Thuiller (Rennes 2/LLF), Benoît Crabbé (ParisDiderot/ALPAGE) & Anne Abeillé (LLF/Paris Diderot)!e role of animacy in sentence production: Evidence from Fren#Rick Grimm (York)L’emploi variable du subjonctif et des expressions de nécessité dans le français parlé deslocuteurs restreints à Pembroke (Ontario)

Je"rey Lamontagne (O!awa)Examining ‘opposing’ processes in &ebec Fren# mid vowelsGary Libben (Brock), Kaitlin Curtiss (Brock), & Silke Weber (Calgary)Compound representation and individual experienceOlivia Marasco (Toronto)L2 Spanish speakers’ perception of secondary cues in Y/N question and statement intonationMonica McMillan (Western)Because Twi%er: Linguistic strategies for brevity in Internet speakMaida Percival (Toronto)Variation in ejectives in Harar OromoJames Porteous (Western)Redundant adjective is redundant: An analysis of Adj1 N is Adj1Malina Radu, Gabrielle Klassen, Laura Colantoni, Ma!hew Patience, Ana Teresa Pérez-Leroux, & Olga Tararova (Toronto)!e perception of intonational contours: a cross-linguistic study


255 256 257

Morphologie | Morphology Psycholinguistique | Psycholinguistics Syntaxe | Syntax

3:30–4:00 David Heap (Western), Michèle Oliviéri (Nice) & Jean-Pierre Lai(Nice)

Les sujets des verbes météorologiques dans les dialectes occitanstransitionnels du Nord

Joyce Bruhn de Garavito & Silvia Perpiñán (Western)

Subject pronouns and clitics in the Spanish interlanguage of Fren# L1speakers

Will Oxford (Manitoba)

!e rise and fall of split-ergative agreement in Algonquian

4:00–4:30 Daniel Schmidtke & Victor Kuperman (McMaster)

Mass counts in the World Wide Web: A corpus linguistic study of nouncountability across varieties of English

Laura Sabourin, Jean-Christophe Leclerc, Michele Burkholder &Christie Brien (O!awa)

Bilingual lexical organization: Is there a sensitive period?

Richard Compton (McGill)

An argument for genuine object agreement in Inuit

5:00–5:15 Musée canadien des langues | Canadian Language MuseumOuverture de l’exposition «Le français au Canada» | Opening of the “French in Canada” exhibit

Eye Corridor

5:30–7:30 Assemblée générale | Annual general meeting 215

Thistle Thistle Thistle


Thistle


Thistle

Academic South

Dimanche 25 mai | Sunday, May 25 257 258 259

Phonologie & phonétique | Phonology & phonetics Psycholinguistique | Psycholinguistics Syntaxe | Syntax

9:00–9:30 Emily Clare (Toronto)

Evidence of phonological knowledge from perceptual learning

Kaitlin Falkauskas & Victor Kuperman (McMaster)

Individual di$erences in the strategies used to read morphologicallycomplex words

Olena Kit (McMaster)

On propositional complexity of middles and its consequences forsyntax

9:30–10:00 Sara Mackenzie (MUN), Erin Olson (MIT), Meghan Clayards(McGill) & Michael Wagner (McGill)

Allophonic variation in English /l/: production, perception, andsegmentation

Kathleen Currie Hall, Claire Allen, Tess Fairburn, Kevin McMullin,Michael Fry & Masaki Noguchi (UBC)

&antifying perceived morphological relatedness

Kumiko Murasugi (Carleton)

A hierar#ical view of the ergative and antipassive in Inuktitut

10:00–10:30 Avery Ozburn (Toronto)

Statistical co-occurrence restrictions in Oromo consonants

Jianxun Liu (Victoria)

!e optionality property of Kiswahili applicatives


257 258 259

Phonologie & phonétique | Phonology & phonetics Sémantique | Semantics Syntaxe | Syntax

10:45–11:15 D. Sky Onosson (Victoria)

A prosodic account of Canadian Raising

Igor Yanovich (Tübingen)

No weak necessity

11:15–11:45 Christopher Spahr (Toronto)

Restricting non-segmental contrasts

Bronwyn Bjorkman & Elizabeth Cowper (Toronto)

Possession and necessity: from individuals to worlds

Heather Bliss (Victoria)

Partial (non-)con'gurationality in Bla(foot

11:45–12:15 Dennis Ryan Storoshenko (Calgary) & Robert Frank (Yale)

Whi# questions do you like the movements in? A semantic constrainton extraction

Clarissa Forbes (Toronto)

On the absence of nominal coordination in Gitksan


Assemblée générale du Musée canadien des langues | Annual general meeting of the Canadian Language Museum 257

Session des a!iches | Poster session North Hallway

1:30–3:00 Louise Beaulieu (Moncton) & Wladyslaw Cichocki (UNB)Les formes comme / comme que en français acadien du nord-est du Nouveau-Brunswi(Sophia Bello (Toronto)Null indirect objects in &ebec Fren#Christie Brien & Laura Sabourin (O!awa)Age of L2 acquisition has a greater in"uence on L1 metalinguistic awareness than pro'ciencyVeena Dwivedi & Kaitlin Curtiss (Brock)!e role of event knowledge in semantic interpretationBrandon J. Fry (O!awa)!e (un)interpretability of {)}: A derivational approa#Philippe Gauthier (Western)Faisons at(tension) aux voyelles hautes : le relâ#ement et l’harmonie vocalique en français deWindsor et de HearstDavid Heap, Jorge Emilio Rosés Labrada, Je" Tennant & Angélica Hernández (Western)Aspiration et e$acement du /-s/ en espagnol d’Holguín (Cuba) : variables phonologiques et leurconditionnement

Je"rey P. Hong, Todd R. Ferre!i & Deanna Hall (Laurier)Imagining events: In"uences of temporal information in verbs and perspective takingDanica MacDonald (Calgary)From a classi'er language to a mass-count language: What can historical data show us?Nesrine Mejri & E. Allyn Smith (UQAM)Étude expérimentale de l’interprétation des questions rhétoriquesNicholas Moroz (Rimouski) & François Poiré (Western)Variation prosodique dia#ronique : René Lévesque en trois tempsFrançois Poiré, Je" Tennant & Antony Cloutier (Western)Adaptation à l’accent français européen par une comédienne québécoise : Changementsacoustiques des voyelles et perception des accents queébécois et hexagonalDanielle Tomas & Kristen Don Paul (York)Spanish nominal word order in early and late bilingualism


257 258 259

Phonologie & phonétique | Phonology & phonetics Acquisition (L2) Syntaxe | Syntax

3:30–4:00 Daniel Currie Hall (Saint Mary’s)

On substance in phonology

Antonio A. González-Poot (Calgary)

Con"ict resolution in the Spanish SLA of Yucatec ejectives: L1, L2 anduniversal constraints

Gabriela Alboiu & Ruth King (York)

Emphatic ‘quite’ in Acadian Fren# as Focus operator

4:00–4:30 Liisa Duncan (Toronto)

Syn#ronic productivity of Finnish vowel harmony

Danielle Thomas & Katie Tetzlo" (York)

Language-internal context and VOT in bilingualism: A view fromSpanish and English

María Luisa Rivero & Vesela Simeonova (O!awa)

An evidential modal in Bulgarian: !e inferential future

4:30–5:00 Laura Colantoni (Toronto), Alana Johns (Toronto), Michelle Yuen(MIT) & Marta Ortega-Llebaria (Pi!sburgh)

Perception of English corrective focus by native Inuktitut speakers

Solveiga Armoskaite (Rochester) & Carrie Gillon (Arizona State)

Aspectual conditions on (in)de'niteness

5:00–7:00 Réception du président | President’s receptionCentre du Congrès | Congress Centre

( )



Thistle

Thistle


Walker Complex

Lundi 26 mai | Monday, May 26 258 259

Phonétique | Phonetics Syntaxe & sémantique | Syntax & semantics

9:30–10:00 Anthony Brohan (MIT)

Licensing Catalan laryngeal neutralization by cue

Mireille Tremblay (Montréal)

Les pronoms en français : pluralité et individuation

10:00–10:30 Svetlana Kaminskaïa (Waterloo)

Di$érences rythmiques dans les styles de parole en français ontarien

Jessica Mathie (Toronto)

Markedness in number features: Evidence from Ganggalida (Yukulta)

10:30–11:00 Blake Lewis (Calgary)

!e syntax and semantics of demonstratives: A DP external approa#


11:15–12:15 Communication plénière | Plenary talkJila Ghomeshi (Manitoba)

Who are we talking to when we talk to ‘the public’ about linguistics?Récipiendaire du Prix national d’excellence | Recipient of the National Achievement Award

243


1:15–4:50 Session conjointe avec l’Association canadienne de linguistique appliquée | Joint session with the Canadian Association of Applied Linguistics

Revitalizing Aboriginal LanguagesOrganizer: Keren Rice (Toronto)

Upper level of

Gra#on AntoneTea#ing the On*yota’a·ka language in an urban se%ingEileen AntoneLearning the On*yota’a·ka language as an adultCarrie Dyck & Amos Key, Jr.An immersion program for intermediate-level speakersMarie-Odile JunkerPu%ing information te#nologies to work for Aboriginal languages preservation and revitalizationMarguerite MacKenzieLanguage maintenance in East Cree, Naskapi and Innu: A forty-year perspectiveAlex McKay & Connor PionKiiwepiskaapiimon! Revitalizatize your language!

Thistle Thistle

Thistle

Pond Inlet

EMPHATIC ‘QUITE’ IN ACADIAN FRENCH AS FOCUS OPERATOR

Gabriela Alboiu (York University) and Ruth King (York University)

Introduction. The Baie Sainte-Marie, Nova Scotia variety of Acadian French (henceforth, AF)

has borrowed the English degree modifier quite in constructions such as (1):

(1) a. Vous aviez fait une quite de visite. (King 2013: 102)

‘You had had quite a visit’

b. C’était une quite de Carole.

‘Carole was quite something’

Semantically, the presence of quite emphatically evaluates the lexical noun. Furthermore, the

presence of de in (1) indicates a complex nominal construction whose syntax begs clarification.

Analysis & Conclusions. King notes that the hyperbolic connotation of the data in (1) resemble

the French construction containing epithet nouns, like un espèce de cochon (“a real pig”,

literally, “a sort of pig”) and ce putain de livre (“this bloody book”, literally, “a whore of book”).

French has a number of binominal constructions of the (Det) N1 DE N2 type (on par with

Romance, more generally): (i) partitive: un verre d’eau (“a glass of water”), (ii) quantitative:

beaucoup de livres (“many books”), (iii) possessive: la soeur de Marie (“Mary’s sister”), and

(iv) qualitative: un bijou de voiture (“a jewel of a car”). Since both the AF quite construction and

the epithet N construction are semantically evaluative, they would fall under (iv), for which both

Kayne (1994) and den Dikken (2006) propose a predicate inversion structure. For Kayne, de is a

C head in a reduced relative clause, for den Dikken, it is the overt realization of a relator (i.e. a

nominal copula). However, there is no comparison between quite and visite, as there is between

bijou and voiture. Nor does quite in AF behave as a predicate: *La visite est (une) quite so a

subject-predicate analysis is difficult to maintain. Doetjes/Rooryck (2001) argue against

predicate inversion for qualitative binominals and propose (2) with N1 in Spec,EvalP instead:

(2) [EvalP [DP ce phénomène] [Eval [DP [D de [NP fille]]]]]

The structure in (2) is appealing in view of presence of an adverbial Evaluative head (à la Cinque

1999) and the likely adverbial status of quite – (3) shows that, while N2 can be modified by an

Adj in AF, quite cannot, clearly indicating that quite is not of N category:

(3) J'avons eu un (*bon) quite de bon souper.

However, while (2) correctly rules out en-cliticization for French epithet N constructions given

the D nature of de, assuming en is a pro PP (Kayne 1975), it also rules out en-cliticization with

evaluative quite in AF, which is empirically incorrect. Compare (4a)-(4b):

(4) a. *Il en est un espèce de cochon. b. Il en contait une quite d’histoire.

Rather, (4b) points to similar en extraction properties as in partitives (Kayne 1975) and

quantitatives: Il en a acheté une douzaine de pommes, and where [PP de NP] (Kayne 1975, 1994).

As with qualitative binominals (Hulk/Tellier 2000), N(2) must be a bare NP in (1). As only DP

(but not NP) needs Case (Kayne 1999), de is not a P case-assigner (pace Jones 1996). Likely, its

function is to indicate N(2) as a property/kind, on par with the IE genitive it has replaced (Ihsane

2008). In sum, we propose (5), which captures both syntactic & semantic properties of AF quite:

(5) [DP [D une[FocP quite Foc SCALAR EMPHASIS [dP [d de [NP visite/Carole]]]]]

(5) capitalizes on Giusti’s (2006) proposal that DPs have left peripheries hosting TopP and FocP

on par with clauses (Rizzi 1997), with D equivalent to Force and ‘d’ equivalent to Fin. Quite is

an operator in Spec,FocP, checking scalar/emphatic Focus (à la Krifka 2007), and ensuring a

greatest/hyperbolic alternative reading on N. In effect, on par with de/di in Romance infinitives

(Kayne 1999, Rizzi 1997), de in (5) is the C/P category of reduced nominals (NumP in Giusti

2012), forced to lexicalize whenever D and ‘d’ project separately.

References.

Cinque, G. 1999. Adverbs and Functional Heads: A Cross-Linguistic Perspective. Oxford: OUP.

Den Dikken, M. 2006. Relators and Linkers: The Syntax of Predication, Predicate Inversion,

and Copulas. MIT Press.

Doetjes, J. and J. Rooryck. 2001. Generalizing over quantitative and qualitative constructions.

Ms. Univ. of Leiden.

Giusti, G. 2006. Parallels in clausal and nominal periphery. In Phases of Interpretation, M.

Frascarelli (ed): 151-172. Berlin: Mouton de Gruyter.

Giusti, G. 2012. On Force and Case, Fin and Num. Enjoy Linguistics! Papers offered to Luigi

Rizzi on the occasion of his 60th birthday. http://www.ciscl.unisi.it/gg60/papers/giusti.pdf

Hulk, A. and M. Tellier. 2000. Mismatches: Agreement in Qualitative Constructions. Probus 12:

33-65.

Ihsane, T. 2008. The Layered DP: Form and Meaning of French Indefinites. John Benjamins.

Jones, M.1996. Foundations of French Syntax. Cambridge University Press.

Kayne, R. 1975. French Syntax. MIT Press.

Kayne, R. 1994. The Antisymmetry of Syntax. MIT Press.

Kayne, R. 1999. Prepositional complementizers as attractors. Probus 11: 39-73.

King, R. 2013. Acadian French in Time and Space. Duke University Press.

Krifka, M. 2007. Basic notions of information structure. In Interdisciplinary Studies of

Information Structure 6, eds. Caroline Fery, Gisbert Fanselow and Manfred Krifka.

Potsdam. http://www.sfb632.uni-potsdam.de/publications/A2/volumeInfo.pdf.

Rizzi, Luigi. 1997. The fine structure of the left periphery. In Elements of Grammar, ed. Liliane

Haegeman, 281-339. Dordrecht: Kluwer.

aspectual conditions on (in)definiteness

Solveiga Armoskaite Carrie Gillon

University of Rochester Arizona State University

Introduction In this paper, we investigate the behavior of Lithuanian (Baltic) bare nominal objects in different aspectual contexts. Bare nouns in articleless languages like Lithuanian are often ambiguous in their interpretation (indefinite or definite). Often, in the presence of certain aspectual morphology, the bare noun must receive a definite interpretation. However, this is not true for all verbs; with some verbs the bare noun continues to be ambiguous. We investigate the origins of this variability. The Problem Bare nominal objects may be interpreted as either definite or indefinite in cases like (1a), but may only receive a definite interpretation in cases like (1b) (cf. Piñon 2006 on Hungarian constraints for overt DPs with particular verbs): (1) a. Sara rašė laiškus. b. Sara parašė laiškus. Sara wrote letters Sara PERF-wrote letters

‘Sara was writing the letters.’ ‘Sara wrote up the letters.’ ‘Sara was writing letters.’ *‘Sara wrote up letters.’ The prefix pa- forces a definite interpretation on the bare object laiškus ‘letters’. However, in cases like (2), pa- does not have this effect, and the interpretation remains ambiguous: (2) a. Sara mėgo šunis. b. Sara pamėgo šunis. Sara liked dogs Sara INCEP-liked dogs ‘Sara liked the dogs.’ ‘Sara started to like the dogs.’ ‘Sara liked dogs.’ ‘Sara started to like dogs.’ Why doesn’t the prefix always force a definite interpretation on the bare object nominal? How can we account for the observed differences in interpretation? Proposal The interpretations of bare nominal objects in Lithuanian is conditioned by the interplay of grammatical and lexical aspect. Although both verbs rašė ‘wrote’ and mėgo ‘liked’ are both transitive, they belong to different lexical aspectual categories. Therefore the prefix has a different effect on the internal arguments.

In (1), rašė is an accomplishment verb. Such verbs have a potential endpoint (Rothstein 2004, Verkuyl 1972, 1993, among others). When pa- is affixed to rašė, it introduces an actual termination point. The verb phrase is therefore interpreted as telic, and, crucially, the internal argument must be interpreted as definite. The addition of pa- enforces a quantized/atomic interpretation on the nominal (cf. Filip 1999 for Slavic perfectives).

In (2), mėgo is a state. States do not have natural endpoints. The prefix pa- therefore cannot introduce an actual endpoint; instead, it introduces an onset of the event. Thus, the bare object nominal cannot be quantized, because only endpoints can quantize over internal arguments.

Conclusions We argue that in each case a prefix introduces a point into the event structure, but that the effect of this point depends on the lexical aspectual class of the verb. Accomplishments + prefix result in a terminal endpoint, which enforces a quantized/atomic structure; states + prefix result in an inchoative. Nominals are only quantized by an introduced endpoint, but not by an introduced onset. Thus, the interplay of the lexical and grammatical aspects conditions the variance in the definiteness interpretations in Lithuanian. reference Filip, H. 1999. Aspect, Eventuality Types, and Nominal Reference. NY: Routledge. Krifka, M. 1998, ‘The origins of telicity’. In Rothstein, S. (ed.) Events and Grammar, 197-235.

Dordrecht: Kluwer. Piñon, C. 2006. Definiteness effect verbs. In Kiss, K. (ed.) Event structure and the Left

Periphery studies in Hungarian, 75-90. Amsterdam: Springer. Rothstein, S. 2004. Structuring events. Oxford: Blackwell. Verkuyl, H. 1972. On the Compositional Nature of the Aspects. Dordrecht: Reidel. Verkuyl, H.1993. A Theory of Aspectuality: The Interaction between Temporal and Atemporal

Structure. Cambridge: Cambridge University Press.

The interaction between impersonal reflexives and syntactic change Kazuya Bamba University of Toronto

A number of synchronic analyses (Geniušiené, 1987; Kemmer, 2003 a.o.) have revealed that reflexives can be found not only in reflexive constructions, but also in middle, impersonal and intransitive statements. One of the major topics of investigation on reflexives today is a cross-linguistic variation of their appearance. I present a contrast in how impersonal statements are expressed among Romance languages, and argue that diachronic observation is essential in accounting for this difference.

The following examples demonstrate how such statements can be expressed with a reflexive clitic. In Spanish, the clitic is commonly found in the statements (1a), so is it in Portuguese, except that in some dialects an overt subject a gente can also be present (1b). This use of reflexive, however, is in general highly marked in French (1c).

(1) a. Se come pasta todos los días 3.REFL eat.3.SG.PRES pasta all.PL the.PL day.PL ‘One eats (people eat) pasta every day’ (Spanish)

This contrast illustrates the different degrees of use of reflexive clitics among the languages. The fact that these three languages also exhibit distinctive syntactic properties (e.g., word order, pro-drop, inflectional agreement etc.) clearly suggests there to be a strong correlation between the reflexive use and the language’s syntax. My study thus makes two theoretical claims. First, through examination of how the syntactic properties listed above are related to the different realisations of reflexives, I argue that any change in these properties be responsible for the divergence observed in (1). Lahousse & Lamiroy (2012), for instance, argue that modern French exhibits the most restricted word order as a result of grammaticalisation of its originally flexible word order. Adams (1987) also shows that French used to exhibit pro-drop property, just as Spanish does today, but lost it by the time of Middle French. I provide evidence of a diachronic shift in how impersonality is expressed with reflexives by examining these independent observations.

Additionally, I argue how changes of the two modes expressing impersonality, reflexive and non-reflexive, are related to the cross-linguistic differences of reflexive uses. Spanish and French can express the same propositions in (1) by having an overt subject argument, uno (‘one’) and on (‘we’) respectively. Portuguese, in contrast, cannot express the same statement without the reflexive clitic, even though the “a gente” phrase is absent in most dialects (Martins, 2005). Following up de Schepper (2007)’s discussion, I discuss how these two forms of expression are employed differently in Romance languages and demonstrate how the uses may have contributed to the divergence of interest.

In this study, by examining these two types of diachronic changes, I demonstrate how the different patterns of use of impersonal reflexives in Romance are strongly tied to historical syntactic changes. The study ultimately shows how such a theoretical approach will be useful in analysing the diverse syntactic realisations of reflexives observed cross-linguistically.

b. E era assim que se a gente vivia! and be.PST so COMP 3.REFL the people live.3.SG.IMP ‘And that’s how we used to live’ (Portuguese; Martins (2005:18))

c. Il se rencontre à Paris des gens de toutes origines he 3.REFL meet.3.SG.PRES in Paris some people of all.PL origin.PL ‘In Paris, one meets people from all over the world’ (French; Turley (1998:137))

References Adams, M. (1987). From Old French to the theory of pro-drop. Natural Language & Linguistic

Theory, 5(1), 1-32.

De Schepper, K. (2007). Reflecting the past: Mapping the development of the Indo-European

SE-form. Linguistics in the Netherlands, 211-222.

Geniušiené, E. (1987). The Typology of Reflexives. Mouton de Gruyter: Berlin.

Kemmer, S. (2003). The Middle Voice. Amsterdam: John Benjamins.

Lahousse, K. & Lamiroy, B. (2012). Word order in French, Spanish and Italian: A

grammaticalization account. Folia Linguistica, 46(2), 387-415.

Martins, A. M. (2005). Passive and impersonal se in the history of Portuguese. In C.D. Pusch., R.

Kabatek., & W. Raible. (eds.), Romance Corpus Linguistics II: Corpora and Diachronic

Linguistics (pp.441-430). Tübingen: Gunter Narr Verlag.

Turley, J.S. (1998). A prototype analysis of Spanish indeterminate reflexive constructions.

Language Sciences, 20(2), 137-162.

Is there a difference? Interpretation in Advanced vs. Intermediate Hispanic Heritage Speakers

Ewelina Barski

The College at Brockport, SUNY

Studies looking at the language of heritage speakers are interested in learning about the stability of language before the critical period and how grammar develops under reduced input conditions (Benmamoun et al. 2010). In this study the aim is to investigate the impact of reduced input on a component within the Null-Subject Parameter - the Overt Pronoun Constraint (OPC) (Montalbetti 1984) - by focusing on the mental representation of the OPC in Hispanic heritage speakers (2nd generation immigrants of Hispanic background). The goal in this study is to probe into the interpretations that bilingual heritage speakers assign to overt pronouns in the subordinate clause with quantified and wh-word antecedents. The OPC restricts the possible antecedents that an overt pronoun can have. Specifically, it states the restrictions on this pronoun when it has a quantified expression (someone, no one) or a wh-phrase (who, which) as its antecedent. As a null subject language, Spanish allows the speaker to omit the subject of the sentence. Examples (1a) and (1b) below show that unlike the null pronoun, the overt pronoun in the subordinate clause can never bind with the quantified expression or wh-phrase: The overt pronoun needs to refer to a third person within the discourse. On the other hand, if the pronoun is covert, the sentence becomes ambiguous and allows for a less restrictive interpretation. Following Montalbetti (1984), I assume that all quantifiers will be treated equally. Moreover, following a generative framework, it is assumed that that the Null Subject Parameter is set early in the grammars of these null-subject heritage languages (Chomsky, 1981; Jaeggli, 1982; Rizzi, 1982), and thus they will demonstrate understanding of the interpretative restrictions found with subordinate overt pronouns with quantified antecedents. 20 Hispanic heritage speakers participated in the experiment. Participants were asked to complete a picture matching task, which looked at a forced interpretation of the OPC and a sentence selection task, which allowed participants to provide their own interpretation of the sentence by choosing between two pictures. Both tasks tested interpretation of the implicit knowledge of the OPC with quantified antecedents. Results for the Picture-Matching task show that advanced heritage speakers understand the interpretative contrast present with overt and null pronouns within OPC contexts. However, a difference is found between advanced and intermediate heritage speakers, where the intermediate group appears to have more difficulty in the Sentence-Selection task: They do not differentiate between null and overt pronouns. Results suggest lower-proficiency participants have difficulty with the reading/comprehension component of the task, but the OPC remains in their grammars.

Examples:

Overt (1a) Nadiei cree que élj/*i va a ganar No one believes that he will win. Null (1b) Nadiei cree que Øi/j va a ganar No one believes that pro will win.

References: Benmamoun, E., S. Montrul & M. Polinsky. (2010). White Paper: Prolegomena to Heritage Linguistics. Harvard University Press. Chomsky, Noam. (1981). Lectures on Government and Binding. Dordrecht: Foris. Jaeggli, Osvaldo. (1982). Topics in Romance Syntax. Dordrecht: Foris. Montalbetti, M. (1984). After binding. PhD dissertation, MIT. Rizzi, Luigi. (1982). Issues in Italian Syntax. Dordrecht: Foris.

Emergence des normes communautaires: cas de la variation lexicale. Alena Barysevich

(York University, Glendon)

Cette communication démontre une émergence des normes communautaires distinctives au fil des générations dans le contexte de la globalisation et du nivellement du langage. Nous analysons l’expression des notions d’«automobile» et de «travail rémunéré» dans le français de la région d’Ottawa-Hull. Nous nous appuyons sur trois corpus (Ottawa-Hull, RFQ et FdO) recueillis à trois points du temps (les années 1950-1960, 1983-1984 et 2004-2006). En recourant à l’analyse sociolinguistique variationniste, nous démontrons comment l’emploi des lexemes peut être conditionné par le changement culturel à un moment précis de l’histoire. Une attention particulière est portée à la relation entre la structure des groupes sociaux et les relations integroupes au sein des communautés (Gadet 2007; Mougeon 2005).

Dans notre echantillon, la dynamique lexicale des jeunes des communautés francophones majoritaires (Vieux-Hull, Mont-Bleu et Vanier) semble converger, en partie, à celle observée parmi les jeunes des communautés unilingues (Armstrong 1998, 2001 ; Labov 1972b ; Lodge 1989), notamment on observe un rapprochement des jeunes au parler local (usage des formes vernaculaires comme marque de l’appartenance communautaire). Par contre, dans les communautés francophones minoritaires (West-End et Basse-Ville), on observe plutôt l’assimilation à l’anglais (langue de prestige), attribuable à l’insécurité linguistique des francophones dans les communautés avec une forte concentration de la population anglophone.

Nous soulignons l’importance de considérer autant les facteurs sociaux statiques que de rajouter les facteurs mettant en relief la dynamique communautaire, avec ses normes sociales et ses valeurs de prestige. Cette communication montre que dans le cas de la variation lexicale, il apparaît plus rigoureux de considérer la dynamique lexicale des groupes sociaux (et non pas des locuteurs individuels), par exemple en fonction de la classe sociale ou de l’âge des répondants. Nous avons prêté une attention très particulière à l’analyse micro-variationnelle, c’est-à-dire l’étude de chaque communauté séparément. Notre étude a également comparé la dynamique lexicale dans les communautés francophones majoritaires versus minoritaires de la région à l’étude. References: Gadet, Françoise. 2007. La variation sociale en français. 2nd ed. Paris: Ophrys. Laforest, Marty. 2002. Attitudes, préjugés et opinions sur la langue. In C. Verreault, L. Mercier, & T.

Lavoie (eds.), Le français, une langue à appri-voiser. Québec: Les Presses de l’Université Laval. 82-91.

Mougeon, Raymond. 2005. Rôles des facteurs linguistiques et extralinguistiques dans la dévernacularisation du parler des adolescents dans les commu-nautés francophones minoritaires. In A. Valdman, J. Auger, & D. Piston-Hatlen (eds.), Le français en Amérique du Nord: État présent. Québec: Presses de l’Université Laval. 261-286.

Mougeon, Raymond, Katherine Rehner, and Terry Nadasdi. 2010. La variation lexicale dans le parler des adolescents franco-ontariens. In W. Remysen & D. Vincent (eds.), Hétérogénéité et homogénéité dans les pratiques langa-gières: Mélanges offerts à Denise Deshaies. Québec: Presses de l’Uni-versité Laval. 169-206.

Poplack, Shana. 1989. The care and handling of a mega-corpus: The Ottawa-Hull French project. In R. Fasold & D. Schifrin (eds.), Language Change and Variation. Amsterdam: Benjamins. 411-451.

Rand, David and David Sankoff. 1988. GoldVarb. Logistic regression package for the Macintosh. Montreal: Université de Montreal.

Veranika Barysevich

Université Western

Départements d'études françaises

L'ACCENTUATUION DES "PETITS" MOTS.

La problématique de la nomenclature du terme de clitique en tant qu’un « cover term » pour désigner des unités prosodiquement déficientes ou en tant que primitive linguistique (Spencer 1991) a fait couler beaucoup d'encre dans la théorie linguistique (la phonologie, la morphologie et la syntaxe). Cette recherche apporte un nouvel éclairage à la problématique de la réalisation accentuelle des clitiques dans le cadre de la phonologie prosodique. Notre recherche propose une analyse comparative des propriétés acoustiques (la durée, la fréquence fondamentale et l'intensité) des mots de fonction ou 'Little words' (Leow et al. 2009) nommés souvent "clitiques" et qui ne sont pas rare accentués dans le discours spontané (Selkirk 1984). Mertens (1993) explique ce phénomène par le fait que l’accent secondaire ( l'accent qui tombe sur la syllabe initiale du mot lexical ou sur la deuxième syllabe si ce mot lexical commence par une voyelle (Mertens 2011)) est indifférent au trait [+/-clitique]. En partant de l’observation de Mertens (1993) nous testons l’hypothèse que l’accent qui tombe sur un mot grammatical dans le domaine de la réalisation du mot prosodique est différent de l’accent initial et/ou secondaire se réalisant à l’intérieur du mot lexical. Nous proposons également l’hypothèse que la structure accentuelle en français peut être reproductive, c’est-a-dire qu’elle peut reproduire/générer de nouveaux types d’accent. Pour tester cette hypothèse, nous identifions et caractérisons l'accentuation des mots de fonction dans les trois variétés du français spontané européen de trois styles de parole en provenance du corpus multi-style du français parlé C-PROM (Avanzi et al. 2010). Nous examinons les mots grammaticaux transcrits et annotés à l'aide du logiciel d'analyse acoustique PRAAT (Boersma et Weenink 2010) et segmentés en phonèmes, syllabes et mots avec le scipt EasyAlign (Goldman 2008) qui est implanté dans PRAAT. À partir du corpus de 9 locuteurs natifs du français européen provenant de la France métropolitaine, de la Belgique francophone et de la Suisse romande, nous suivons l’étude de Poiré et Kaminskaïa (2004) sur la F0 normalisée pour mesurer les valeurs normalisée en cote-z des trois variables (la durée, la fréquence fondamentale et l'intensité). Cela nous aide à mettre en question les facteurs acoustiques exploités pour produire l'accentuation sur les mots de fonction. Nous testons également l’hypothèse de Woehrling et al. (2008) sur la non-variation prosodique des noyaux vocaliques des clitiques en français spontané.

Références

Avanzi, M., A.-C. Simon, J. P. Goldman et A. Auchlin. 2010. C-PROM : Un corpus de français parlé annoté pour l’étude des proéminences. Dans Proceedings of

JEP, Mons, Belgium, mai 2010. http ://sites.google.com/site/corpusprom. Boersma, P. et D. Weenink. 2010. Praat : doing phonetics by computer (Version 5.1.24). http ://www.praat.org. Goldman, J.-P. 2008. EasyAlign: a semi-automatic phonetic alignment tool under Praat.

http://latlcui.unige.ch/phonetique. Leow, R., H. Campos et D. Lardiere. 2009. Little words: their history, phonology, syntax,

semantics, pragmatics, and acquisition. Washington, D. C.: Georgetown University Press.

Mertens, P. 1993. Accentuation, intonation et morphosyntaxe. Dans Travaux de Linguistique 6, 21-69 Mertens P., 2011. Prosodie, syntaxe, discours : autour d’une approche prédictive. Dans Actes du colloque IDP2009 (Interface Discours Prosodie), Paris, 9-11 septembre 2009. Poiré, F. et S. Kaminskaïa, 2004. La variation intonative dans deux variétés de français, XXIVe Congrès de linguistique et de philologie romane , University of Alberystwyth, England. Selkirk, E. 1984. Phonology and Syntax: The Relation between Sound and Structure. London: The MIT Press. Spencer, A. 1991. Clitics. Dans Morphological theory, 350-394. Oxford: Basil Blackwell. Woehrling, C., Ph. B. de Mareüil et M. Adda-Decker. 2008. Aspect prosodiques du français parlé en Alsace, elgique et Suisse. Dans 27es Journées d'Etude sur la Parole,Avignon.

http://latlcui.unige.ch/phonetique

Les formes comme / comme que en français acadien du nord-est du Nouveau-Brunswick : variation synchronique et variation diachronique

Louise Beaulieu & Wladyslaw Cichocki Université de Moncton & University of New Brunswick

Cette communication présente une analyse diachronique de type transversal (trend study) de la variation dans les expressions en tête des adverbiales tensées en comme dans la grammaire du français acadien. Une première analyse synchronique de cette variable (Beaulieu et Cichocki 2002) dans le français acadien du nord-est du Nouveau-Brunswick (FANENB) a montré des tendances statistiquement non significatives dans les fréquences de la variante traditionnelle comme que, en ce qui a trait à l’âge. Selon l’hypothèse du temps apparent, la variable semble stable. L’étude de suivi présentée dans cette communication a pour but de réexaminer cette conclusion à partir de données provenant de décennies antérieures. Dans les adverbiales en comme du FANENB, la forme comme, la variante standard, et comme que, la variante vernaculaire, sont utilisées en contexte formel et informel. Notons que la variante comme que n’est pas spécifique au FANENB ; on la retrouve dans d’autres variétés informelles de français parlé (voir Holder et Starets 1982 et King et Nadasdi 2006, parmi d’autres). Les données proviennent de deux corpus enregistrés dans la même communauté acadienne. Le premier a été recueilli en 1975 auprès de 22 locuteurs nés entre 1882 à 1909. Le deuxième (qui a fait l’objet de la première analyse) a été enregistré en 1990 auprès de 16 locuteurs dont les années de naissance varient de 1936 à 1968. Ces deux corpus sont stratifiés selon l’âge, le sexe et le réseau social. La présente étude, réalisée à l’aide d’analyses de régression logistique (avec Goldvarb X), examine la variation selon quatre générations de locuteurs. Les résultats montrent une augmentation intergénérationnelle de la fréquence de la variante comme que : de 23,7% dans la génération 1882-1895 à 40,3% dans la génération 1958-1968. Il s’agit donc d’un changement dans le temps, ce qui ne correspond pas à la conclusion inférée suite à l’analyse du corpus de 1990. L’analyse des deux corpus met en évidence le rôle, au niveau intra- et intergénérationnel, des facteurs réseau social et sexe sur la trajectoire de la variante traditionnelle. À l’instar de Sankoff (2005), la présente étude montre l’importance de valider, à partir de données en temps réel, les conclusions basées sur l’hypothèse du temps apparent.

Références Beaulieu, L. et W. Cichocki. 2002. Le concept de réseau social dans une communauté acadienne

rurale. Revue canadienne de linguistique 47 : 123-150. Holder, M. et M. Starets. 1982. Étude sur les formes simples et les formes composées du type

si/si que, quand/quand que/quand ce/quand ce que, etc. dans le parler acadien de Clare, Nouvelle-Écosse. Si que 5 : 117-128.

King, R. et T. Nadasdi. 2006. Another look at que deletion. Communication présentée à NWAV 35, Ohio State University.

Sankoff, G. 2005. Cross-sectional and longitudinal studies. In U. Ammon et al. (éds) Sociolinguistics. An International Handbook of the Science of Language and Society. De Gruyter : Berlin.Vol. 2 : 1003-1013.

Null indirect objects in Quebec French Sophia Bello University of Toronto The notion that a transitive verb can appear without a direct object (DO) has been widely explored throughout various experimental and theoretical studies in French (Jakubowicz et al. 1996; Hamann 2003; Cummins & Roberge (C&R) 2004, 2005; Pérez-Leroux et al. 2008). Previous analyses have construed this object to have a generic, non-referential interpretation (1) or to be classified as referential null objects (NOs) as in (2); both cases have been identified in French (cf. C&R 2004, 2005). The goal of this presentation is to expand the analyses of NOs from transitive to ditransitive constructions and observe how they can account for missing indirect objects (IOs) in Quebec French.

(1) Wild Guns est un jeu qui défoule ___. (Larjavaara 2000: 88) ‘Wild Guns is a game that destresses ___.’

(2) “Tu as lu les pages?” Il avait lu ___. (Larjavaara 2000: 43) ‘“Did you read the pages?” He had read ___.’

I adopt C&R’s (2004: p. 133) approach and assume that all nominals should be interpreted as ‘coreferential’: [lexical noun…pronoun…NO] If we take the idea that a NO must have a referent that is salient in the previous discourse, then the absence of the object clitic (whether direct or indirect) should not render the sentence to be considered ungrammatical. Note that the difference between direct and indirect objects is of a structural nature, in terms of argument position and their association with the verb (i.e., DO merges with V while IO merges with P). Thus, this presentation will focus on analyzing the structural representation of transitive and ditransitive constructions and see how this could affect child language acquisition of object clitics in Quebec French. In acquisition, Pérez-Leroux and her colleagues (2008) propose that a null object stage exist in child grammar. The notion is that a child has the option of producing a referential null object N or a clitic. Then, it is the experience, depending on the context presented, that guides the child into producing a null object. Using transitive constructions, they found that French and English-speaking children (i.e., 34.5% and 8.3%, respectively) start off by omitting DOs early on. Costa and his colleagues (2007, 2008) conducted an experiment eliciting ditransitive constructions in European Portuguese. Their results suggest a high rate of IO omissions (~52%) in 3-4 year-olds. Finally, I conducted an experiment using ditransitive verbs in obligatory contexts and found that 3-4 year-old French-speaking children omit IOs 83% of the time (3). These findings suggest that children go through a null object stage where they generalize NOs (direct or indirect) in obligatory contexts until they have acquired the adult grammar.

(3) Question: Qu’est-ce que Marc fait pour que Julie puisse manger ses céréales? ‘What does Mark do so that Julie can eat her cereal?’ Child: elle Ø donne du lait. (C17, 3;09) She Ø-DAT gives some milk ‘She gives Ø some milk.’

This presentation provides structural and experimental evidence on null objects in Quebec French. Such a study is fundamental to understanding the role of the verb, the status of the verb’s internal arguments, and ultimately, what causes children’s grammar to diverge from the adult grammar.

References Costa, J. and Lobo, M. (2007). Clitic omission, null objects or both in the acquisition of European Portuguese? In S. Baauw, F. Drijkoningen and M. Pinto (Eds.), Romance Languages and Linguistic Theory 2005. Amsterdam/Philadephia: John Benjamins, pp. 59-71. Costa, J., Lobo, M., Carmona, J. and Silva, C. (2008). Clitic Omission in European Portuguese: Correlation with Null Objects? In A. Gavarró and M. João Freitas (Eds.), Language Acquisition and Development: Proceedings of GALA 2007. Newcastle: Cambridge Scholars Press, pp. 133-143. Cummins, S. and Roberge, Y. (2004). Null Objects in French and English. In J. Auger, C. Clements, & B. Vance (Eds.), Contemporary Approaches to Romance Linguistics. Amsterdam: John Benjamins, pp. 121-138. Cummins, S. and Roberge, Y. (2005). A modular account of null objects in French. Syntax 8(1): 44-64. Hamann, C. (2003). Phenomena in French normal and impaired language acquisition and their implications for hypotheses on language development. Probus 15: 91-122. Jakubowicz, C., Müller, N., Kang, O., Riemer, B., & Rigaut, C. (1996). On the acquisition of the pronominal system in French and German. In A. Stringfellow, D. Cahana-Amitay, E. Huges, & A. Zukowski (Eds.), Proceedings of the 20th Annual Boston University Conference on Language Development. Sommerville, MA: Cascadilla Press 1, pp. 331- 342. Larjavaara, Meri. (2000). Présence ou absence de l’objet. Limites du possible en français contemporain. Helsinki: Academia Scientiarum Fennica. Pérez-Leroux, A. T., Pirvulescu, M., and Roberge, Y. (2008). Null Objects in Child Language: Syntax and the Lexicon. Lingua 118: 370-398.

Possession and necessity: from individuals to worldsBronwyn M. Bjorkman and Elizabeth Cowper, University of Toronto

This paper investigates the use of possessive morphosyntax for modal necessity, as in (1). Pos-sessive modality (PM) occurs both in languages with a verb have (English, German, Span-ish, Catalan), and in those expressing possession with be (Hindi, Bhatt 1997; Russian, Jung2011). We claim that PM constructions arise because both possession and necessity express anINCL(usion) relation between two arguments of the same semantic type: possession expressesINCL between two ⟨e⟩-type arguments, while necessity expresses INCL between sets of worlds.This relation arises in two distinct structures: possessive have is syntactically transitive, whilemodal have conceals one argument within the modal head.

(1) a. That cyclist has a helmet. (poss’n) b. Cyclists have to obey traffic laws. (nec.)The semantics of clausal possession are not well understood, but one aspect is the part-

whole or INCL relation (Aikhenvald, 2013), expressed in the syntax by a transitive head relatingtwo nominal arguments (Boneh and Sichel, 2010; Harley, 1995; Levinson, 2011; Ritter andRosen, 1997). Just as possession in (2) involves inclusion between individuals, the formalsemantics of necessity involve inclusion between sets of worlds. Since Kratzer (1981, 1986),modal constructions are taken to include a modal operator (∀ or ∃), which composes first witha modal base (a set of accessible worlds), and then with a proposition (also a set of worlds).With a universal modal operator, the proposition is true in all accessible worlds—i.e. the setof worlds corresponding to the modal base is included in the set of worlds corresponding tothe proposition. Extending have to modal necessity requires only reanalysis of an interpretablefeature INCL, broadening the arguments it relates from individuals to sets of worlds.

(2) a. the tree with branches b. coffee with milkIf both possession and necessity are semantically transitive, however, why is only posses-

sion syntactically so? Syntactic transitivity has been argued to be the defining property ofpossessive have (Hoekstra, 1984; Cowper, 1989), but modals, including modal have (to) areintransitive, with raising syntax (Bhatt, 1997, a.o.). Semantic work often assumes complexstructure within modal heads (allowing composition under sisterhood between a modal oper-ator and modal base, e.g.). Syntactic Merge, however, cannot create head-internal structure:a first-merged argument is by definition a syntactic complement. We resolve the mismatchby proposing instead that the head-internal structure of modals consists of two interpretablefeatures, encoding modal force and modal base. Function Application can apply not onlyto structures created by Merge, but also to heads bearing more than one semantically inter-pretable feature: the semantic transitivity of modals arises due to their featural, rather thanstructural, complexity. The morphology can realize either of these features (or both): whileEnglish modals primarily express modal force, Matthewson et al. (2006, et seq.) show thatmodal systems can also primarily express the modal base.

The advantage of this proposal, in contrast to previous approaches to PM (e.g. Bhatt 1997and Bybee and Pagliuca 1985, who treat PM as expressing the possession or existence of anobligation), is that it directly explains why necessity, but not possibility, is expressed by posses-sive morphosyntax. For Bhatt, PM expressions assert the existence of an obligation, expressedby a silent necessity operator, making it mysterious why there is no corresponding silent possi-bility operator. For us, the universal force PM constructions follows from the inclusion seman-tics of possession.

PM constructions thus shed light not only on the semantics of possession but also on thecompositional syntax of modal operators. Our account supports the idea that inclusion is at leastpart of the semantics of possession, and also explains possible mismatches between syntacticand semantic transitivity.

1

ReferencesAikhenvald, Alexandra Y. 2013. Possession and ownership: A cross-linguistic perspective. In Possession

and ownership: A cross-linguistic typology, ed. Alexandra Y. Aikhenvald and R. M. W. Dixon, 1–64.Oxford University Press.

Bhatt, Rajesh. 1997. Obligation and possession. In The proceedings of the UPenn/MIT workshop onargument structure and aspect, number 32 in MIT Working Papers in Linguistics. MIT Departmentof Linguistics and Philosophy.

Bybee, Joan L., and William Pagliuca. 1985. Cross-linguistic comparison and the development of gram-matical meaning. In Historical semantics, historical word-formation, ed. Jacek Fisiak, 59–83. Berlin:Mouton.

Cowper, Elizabeth. 1989. Thematic underspecification: the case of have. Toronto Working Papers inLinguistics 10:85–94.

Halle, Morris, and Alec Marantz. 1993. Distributed morphology and the pieces of inflection. In Theview from Building 20: Essays in linguistics in honor of Sylvain Bromberger, ed. Kenneth Hale andSamuel Jay Keyser, 111–176. Cambridge, MA: MIT Press.

Hoekstra, Teun. 1984. Transitivity: Grammatical relations in government-binding theory. Dordrecht:Foris Publications.

Jung, Hakyung. 2011. The syntax of the BE-possessive: Parametric variation and surface diversities.Amsterdam: John Benjamins.

Kratzer, Angelika. 1981. Partition and revision: The semantics of counterfactuals. Journal of Philo-sophical Logic 10:201–216.

Kratzer, Angelika. 1986. Conditionals. In Papers from the parasession on pragmatics and grammaticaltheory: Twenty-second regioual meeting of the Chicago Linguistics Society, volume 2, 1–15. Chicago:University of Chicago.

Levinson, Lisa. 2011. Possessive WITH in Germanic: HAVE and the role of P. Syntax 14:355–393.

Matthewson, Lisa, Hotze Rullmann, and Henry Davis. 2006. Modality in St’at’imcets. In Studies in Sal-ishan, ed. Shannon T. Bischoff, Lynika Butler, Peter Norquest, and Daniel Siddiqi, volume 7 of MITWorking Papers in Linguistics on Endangered and Less Familiar Languages, 93–112. Cambridge,MA: MIT Department of Linguistics and Philosophy.

Boneh, Nora, and Ivy Sichel. 2010. Deconstructing possession. Natural Language & Linguistic Theory28:1–40.

Harley, Heidi. 1995. Subjects, events and licensing. Doctoral Dissertation, Massachusetts Institute ofTechnology.

Ritter, Elizabeth, and Sara T. Rosen. 1997. The function of have. Lingua 101:295–321.

2

Heather Bliss University of Victoria

Partial (Non-)Configurationality in Blackfoot Introduction Algonquian languages are often described as non-configurational, yet there is variation in the underlying source(s) of non-configurationality. Some languages are argued to pattern as Pronominal Argument (PA) languages, in which argument expressions (AEs) are adjoined to the clause and bind pronominal arguments (Reinholtz & Russell 1995, Reinholtz 1999 for Swampy Cree; Brittain 2001 for Western Naskapi; Junker 1994, 2004 for East Cree). However, the PA analysis is rejected for other languages (Bruening 2001, LeSourd 2006 for Passamaquoddy; Christianson 2002 for Odawa; Hamilton 2012 for Mi’qmaq). Main Claims Blackfoot is a partial PA language, exhibiting a split conditioned by obviation (a reference-tracking system for 3rd persons). In particular, proximate AEs are generated as clause-external adjuncts, but obviative AEs are generated in argument positions inside the clause. Clitics and Agreement Proximate AEs exhibit canonically adjunct-like behaviour, consistent with a PA analysis. For example, they can be freely moved or omitted (1). Obviative AEs, on the other hand, must be resumed by an enclitic –áyi if moved to a preverbal position or omitted (2). (1) a. A’páwaawahkaawa anna Piohkomiaaki. A’p-a-waawahkaa-wa ann-wa ipi-ohkomi-aakii

around-IMPF-walk.AI-PROX DEM-PROX far-sound-woman “Far Sounding Woman is walking around.”

b. (Anna Piohkomiaaki) a’páwaawahkaawa. “{Far Sounding Woman/ she} is walking around.”

(2) a. Áókatakiyini anni ónssts. a-okataki-yini ann-yi w-insst-yi IMPF-bead.AI-OBV DEM-OBV 3-sister-OBV “His sister does beadwork.” b. (Anni ónssts) áókatakiyin*(áyi).

“{His sister /she} does beadwork.” Under the PA analysis, agreement affixes either occupy argument positions (Jelinek 1984) or absorb case (Baker 1991, 1996). Either way, the prediction is that agreement affixes and clitics should not co-occur. This is borne out for proximate but not obviative arguments. C-Command If proximate but not obviative AEs are clause-external (3), then we predict that proximate AEs should asymmetrically c-command obviative ones, regardless of grammatical function. This is borne out: whether an obviative AE is construed as an object (4) or subject (5), it can be bound by a (null) proximate AE. (Conversely, proximate AEs cannot be bound.) (3) [CP DPPROX [CP … DPOBV… ]] (4) Ikáóhkanawáákomiimmiiyaa oksísts.

ik-a-ohkana-waakomii-mm-yii-yi-aawa w-iksist-yi

INTNS-IMPF-all-love-TA-3:4-3PL-3PL.PRN 3-mother-OBV

“Everybodyi loves hisi/j mother.”

(5) Otáóhkanawáákomiimokyaa oksísts.

ot-a-ohkana-waakomii-mm-ok-yi-aawa w-iksist-yi

3-IMPF-all-love-TA-INV-3PL-3PL.PRN 3-mother-OBV

“Hisi/j mother loves everybodyi.”

Implications Blackfoot exhibits a split system, in which proximate but not obviative AEs are adjuncts that bind pronominal arguments. That such a split exists provides evidence against the view of non-configurationality as a macro-parameter (Chomsky 1981, Hale 1983, Baker 1996) and supports a finer-grained approach to the typology of (non-)configurationality.

Heather Bliss University of Victoria

References Baker, Mark C. 1991. On some subject/object non-asymmetries in Mohawk. Natural Language

& Linguistic Theory 9(4): 537-576. Baker, Mark C. 1996. The polysynthesis parameter. New York: Oxford University Press. Brittain, Julie. 2001. The morphosyntax of the Algonquian conjunct verb: A minimalist approach.

New York: Garland. Bruening, Benjamin. 2001. Syntax at the edge: Cross-clausal phenomena and the syntax of

Passamaquoddy. PhD dissertation, MIT. Chomsky, Noam. 1981. Lectures on government and binding. Dordrecht: Foris. Christianson, Keil T. 2002. Sentence processing in a “nonconfigurational” language. PhD

dissertation, Michigan State University Hale, Ken. 1983. Warlpiri and the grammar of non-configurational languages. Natural Language

& Linguistic Theory 1(1):5-47. Hamilton, Michael David. 2012. (Non-)Configurationality in M’igmaq. Paper presented at the

44th Algonquian Conference. University of Chicago, October 25-28, 2012. Junker, Marie-Odile. 1994. La syntaxe du quantifieur universel en algonquin. Canadian Journal

of Linguistics 39(1): 1-13. Junker, Marie-Odile. 2004. Focus, obviation, and word order in East Cree. Lingua 114(3): 345-

365. LeSourd, Philip S. 2006. Problems for the pronominal argument hypothesis in Maliseet-

Passamaquoddy. Language 82(3): 486-514. Reinholtz, Charlotte. 1999. On the characterization of discontinuous constituents: Evidence from

Swampy Cree. International Journal of American Linguistics 65(2): 201-227. Reinholtz, Charlotte and Kevin Russell. 1995. Quantified NPs in pronominal argument

languages: Evidence from Swampy Cree. In Proceedings of the North East Linguistic Society 25, ed. Jill Beckman, 389-402. Amherst, Mass: GLSA.

Age of L2 acquisition has a greater influence on L1 metalinguistic awareness than proficiency.

Brien, C. and Sabourin, L.

University of Ottawa

The age at which a speaker acquires a second language (L2) has been found to influence homonym

processing in the first language (L1). In a combined ERP and cross-modal lexical decision task,

monolinguals were found to reveal a context-by-frequency-interaction which slowed their processing of

target words that were appropriately-related to the subordinate reading of the priming homonym,

supporting the Reordered Access Model1. The results of the bilingual groups did not, even though the

task was carried out in English, the L1 of all participants. The diverging performances of the bilinguals

from the monolinguals were apparent in behavioural responses as well as in the amplitude, scalp

distribution, and latency of ERPs. Bilingual effects were found that varied by age of L2 acquisition (AoA)

suggesting that AoA influences processing in the L12. Specifically, the later bilingual groups exhibited a

marked divergence from the monolingual group which was correlated with AoA. The later bilinguals

revealed greater priming effects (p<.001) and ERP modulations compared to the simultaneous bilinguals

and the monolinguals, suggesting a heightened metalinguistic awareness due to the L2 influencing

homonym processing in the L13,2.

The current study intends to discount proficiency as a factor influencing these particular results.

Participants were grouped according to proficiency levels using French cloze tasks4 and were compared

to the above-reported AoA-grouped results. As anticipated, participants with the highest proficiency

scores correlate with the earliest AoA. However, the remaining participants revealed a continuum of

proficiency scores which did not correlate with the heightened metalinguistic awareness that was found

by AoA. These results are anticipated to support these previous findings and suggest that AoA, rather

than proficiency, has the greatest influence on a second language influencing the first in regards to a

speaker’s metalinguistic awareness of lexical relationships in homonym processing.

1. Duffy, S.A., Morris, R.K., & Rayner, K. (1988). Lexical ambiguity and fixation times in reading. Journal of

Memory and Language, 27, 429-446.

2. Dussias, P.E., & Sagarra, N. (2007). The effect of exposure on syntactic parsing in Spanish- English

bilinguals. Bilingualism: Language and Cognition, 10, 101-116.

3. Cook, V. (2003). Effects of the Second Language on the first. Clevedon: Multilingual Matters.

4. Tremblay, A. (2011). Proficiency assessment standards in second language acquisition research:

“Clozing” the gap. Studies in Second Language Acquisition, 33, 339-372.

Licensing Catalan Laryngeal Neutralization by CueAnthony Brohan - MITCatalan has a pattern of voicing neutralization which has been analyzed under a licensing-by-prosody framework as coda neutralization [4]. Word-internally and across word-boundaries, stops neutralize before nasals (hipnosi [bn]) and [tl] clusters are neutralized(atleta [dl]). [s] is neutralized before nasals (esnob [zn]) and before laterals (legislar [zl],deslletar [zL]. This pattern of neutralization is problematic under a basic licensing bycue approach [2], which holds that contrast follows from cue availability, and that allpre-sonorant cues are equal. Catalan licenses contrast before sonorants in tautosyllabicsequences (a.kla ∼ a.gla) , but not in heterosyllabic TR sequences (ab.na).This paper elaborates on the licensing by cue approach for the licensing of stops contrast inCatalan. The pattern of neutralization is a result of stop voicing cues being impoverishedwhen the release of a stop is obscured by a following sonorant. Acoustic inspection ofCatalan neutralized TN sequences indicates that stops are nasally released. Furthermore,Catalan has an prevalent process of realizing TN and [tl] clusters as geminates (ritme [dm]∼ [mm], atleta [dl] ∼ [ll]). I take this to be a reflex of more general coarticulatory constraintin Catalan. In Catalan TR clusters, velum lowering and C2 gestures are timed earlierthan in other languages, which provides for nasal releases of stops in non-geminatedstop-nasal sequences.To test the perceptibility of a stop voicing contrast is less perceptible with an obscuredburst, a perceptual experiment (currently being piloted) was conducted to determineperceptibility stop-sonorant clusters with obscured and clear releases in Russian. InRussian, homorganic [tn] and [tl] are nasally and laterally released, heterorganic [kn] and[kl] have clear release. Aggregated d′ measures from a 2AFC identification task in noisewith 5 native Russian speakers show decreased discrimination of voicing in clusters withobscured releases.

Contrast t ∼ d tn ∼ dn tl ∼ dl kn ∼ gn kl ∼ gld′ 1.64 1.18 0.93 1.49 1.75

This approach holds Catalan neutralization as a product of a pervasive process of coartic-ulation masking release bursts, yielding a poorer environment for cue realization. Exten-sions of this proposal to sandhi voicing and behavior of TN clusters cross-linguisticallywill be discussed. The pattern of fricative neutralization is accounted for separately asa gemination contrast [3]. Data from West Flemish [2] suggests a separate licensingmechanism to account for fricative neutralization, as fricatives but not stops neutralizebefore sonorants across word boundaries (dat men [tm], ‘that person’, zes noten [zn] ‘sixnuts’).References[1] Steriade, Donca (1999). Phonetics in phonology: the case of laryngeal neutralization. In Matthew Gordon (eds.)Papers in Phonology 3. (UCLA Working Papers in Linguistics 2) Los Angeles Department of Linguistics,UCLA, 25-146) [2] Strycharczuk, Patrycja, and Ellen Simon. Obstruent voicing before sonorants. The case ofWest-Flemish. Natural Language & Linguistic Theory (2013): 1-26. [3] Kawahara, Shiegeto (2012) Amplitudechanges facilitate categorization and discrimination of length contrasts. IEICE Technical Report - The Institute OfElectronics, Information and Communication Engineers - 112: 67-72 [4] Wheeler, Max (2005). Phonology ofCatalan. Oxford University Press

Subject pronouns and clitics in the Spanish interlanguage of French L1 speakers Joyce Bruhn de Garavito & Silvia Perpiñán

The University of Western Ontario

There has been a recent debate regarding the initial state in L3 acquisition and the conditions for syntactic transfer in this type of multilingualism. According to the Cumulative Enhancement Model (Flynn et at., 2004), all previous linguistic knowledge (L1 + L2) may affect the attainment of the L3; on the other hand, Bardel and Falk (2007) proposed that it is the L2 that plays a key role in L3 acquisition. On the other hand, Rothman (2011) believes that syntactic transfer is selective and that typological proximity plays a crucial role in L3 transfer: independently from the order of acquisition, it will be the typological closer language the source of transfer. In this study, we investigate further this issue by examining the initial state of L3 Spanish learners in French native speakers, with English as their L2.

Certain grammatical categories such as personal pronouns appear to be similar in nature but may belong to different classes, a crucial distinction that entails important syntactic differences (Cardinatelli & Starke, 1999). Spanish is uncontroversially a null subject language. When subject pronouns are present, they are strong pronouns; that is, they may be separated from the verb, they may appear alone, they may be focused and they may be coordinated (1). They behave in a similar fashion to English pronouns. French is considered by many a non-null subject language because subject pronouns are obligatory. Following Roberge’s (1990) seminal work, we assume that these pronouns are clitics, that is, they cannot be separated from the verb, they cannot be focused, nor can they appear alone. In other words, they behave more in line with morphological agreement features than strong pronouns.

The question that arises is whether third language learners at the initial stages of acquisition will be able to distinguish the different nature of subject pronouns in Spanish and French, and whether they will resort to English, a typologically different language but more proximate than French with respect to subject pronouns when acquiring Spanish. In particular, we wonder whether they will be able to recognize in sentences such as (1-3) that Spanish pronouns have different syntactic behaviour than their French counterparts, and that the default pronoun in Spanish is in nominative case, unlike in English or French.

(1) Élnom y yonom estudiamos español por la mañana. He and I study-we Spanish in the morning. (2) *Míobl, yonom como el helado de fresa, no élnom. Me, I eat the ice-cream of strawberry, not him. (3) *Eres tiobl quien canta bien, no Jaime. are you who sings well, not Jaime

We tested 20 native speakers of French and 20 native speakers of English learning Spanish as their L3 in their third week of exposure to Spanish. Participants completed a written and oral Acceptability Judgment Task with coordinated pronouns (1), pronouns in contrastive focus (2), cleft-sentences (3) and other similar constructions in which French pronouns would act differently from Spanish pronouns. They also completed an oral task in which they had to answer with one word who the actor of several house duties was. This task elicited responses with pronouns, such as (4) “Who has to clean the dishes?, and the expected response was: “Yo”.

Results indicated that learners at early stages of exposition to Spanish as L3 had significant problems restructuring their grammars, presenting pronouns in the wrong case (i.e.: they produced ‘mí’, and sometimes ‘ti’ for second person to answer to questions such as (4). At the

same time, they had indeterminate knowledge about the possibility of coordination of pronouns, or their appearance in isolation, showing that their source of transfer can be both, the L1 and the L2, as the Cumulative Enhancement Model (Flynn et al., 2004) would predict. References: Bardel, C., & Falk, Y. (2007). The role of the second language in third language acquisition: The

case of Germanic syntax. Second Language Research, 23(4), 459–484. Cardinaletti, A., & Starke, M. (1999). The typology of structural deficiency: A case study of the

three classes of pronouns. In H. C. V. Riemsdijk (Ed.), Clitics in the Languages of Europe (pp. 145–233). Walter de Gruyter.

Flynn, S., Foley, C., & Vinnitskaya, I. (2004). The Cumulative-Enhancement Model for Language Acquisition: Comparing Adults’ and Children’s Patterns of Development in First, Second and Third Language Acquisition of Relative Clauses. International Journal of Multilingualism, 1(1), 3–16.

Roberge, Y. (1990). The syntactic recoverability of null arguments. Kingston, Ont.: McGill-Queen’s University Press.

Rothman, J. (2011). L3 syntactic transfer selectivity and typological determinacy: The typological primacy model. Second Language Research, 27(1), 107–127.

Evidence of phonological knowledge from perceptual learningEmily Clare, University of Toronto

In a groundbreaking study of perceptual learning, Norris et al. (2003) found that listeners shift the boundary betweentwo phonemes along a perceptual continuum following brief exposure to speech exhibiting that pattern. Further stud-ies, such as Kraljic and Samuel (2006) and Nielsen (2011), have tested whether participants would extend their knowl-edge about the continuum they were exposed to (e.g. [d]-[t]) on the basis of natural classes; listeners indeed shifted theboundary of a related continuum (e.g. [b]-[p]) that was absent from exposure. Although it is tempting to attribute thisto speakers’ knowledge of phonological features, it is not clear that this is the case. The difference between [d] and [t]acoustically is the same as the difference between [b] and [p], which creates a confound; participants could have beengeneralising across acoustic traits rather than accessing phonological categories. In this paper, I present results from anew method designed to circumvent this issue, which show that natural classes can aid perceptual learning.

Experiment 1 replicated past studies, where participants shifted their category boundary after listening to speechincluding an ambiguous sound [?] which was consistently disambiguated via lexical knowledge. For example, anambiguous sound [?] between [d] and [t] in the word dash ([?]ash) would be categorised as D because tash is not aword. Hearing [?] as [d] caused listeners’ VOT boundary to increase (from 30ms to 40ms, for example), as evidencedby their subsequent categorisation of that same [?] in nonsense syllables as D. The rate of D responses by group isshown in Figure 1 (left). A mixed-effects logistic regression model indicated that the main effect of group is significant,showing that participants are adjusting their VOT boundary based on exposure, as has been shown before.

Experiment 2 introduced second-order learning (along the lines of Onishi et al., 2002), whereby the direction of theVOT boundary shift was dependent on the neighbouring vowel. For example, one group of listeners heard [?] in placeof [t] before [i] and [?] in place of [d] before [u]. At test, these participants more often identified the ambiguous [?]token as D before [i], but as T before [u], following their exposure. This shows that listeners are able to make perceptualadjustments which are conditioned by a particular adjacent vowel. The difference between D response rates in the twoenvironments by group is shown in Figure 1 (centre). A mixed-effects logistic regression model indicated that theinteraction between vowel and group is significant; thus participants succeeded at learning this complex pattern.

Experiment 3 examined whether new participants could learn a shift that was dependent on groupings of vowels ratherthan individual vowels. For one group of participants, the conditioning vowels were grouped by the natural classesformed by [±high] ([i, I] vs. [e, E]); for the other group of participants, the conditioning vowels were grouped by anunattested feature, which split those same four vowels into ([i, E] vs. [I, e]). The difference between D response rates inthe two environments by group is shown in Figure 1 (right). Only the group whose conditioning environment formeda natural class showed a difference in D response rates between the two environments, meaning only they learned thepattern. This suggests that the natural class facilitated learning, which is the first evidence of this kind.

Figure 1: (Left) Results of Experiment 1, current N=28, total N=32; (Centre) Experiment 2, current N=30, total N=32; (Right)Experiment 3, current N=11, total N=32.

These findings show that phonological features are not only psychologically meaningful for representing contrastsand patterns, but also for the constant subconscious accommodation to complex phonetic patterns. The method itself,second-order perceptual learning, constitutes an additional contribution to this area of study: it supplies a mecha-nism for future work that can directly compare the salience of different phonological features in learners’ minds andtherefore straightforwardly test a theory’s predictions about the structure of speakers’ phonology.

References

Kraljic, T. and Samuel, A. G. (2006). Generalization in perceptual learning for speech. Psychonomic Bulletin &Review, 13(2):262–268.

Nielsen, K. (2011). Specificity and abstractness of VOT imitation. Journal of Phonetics, 39:132–142.

Norris, D., McQueen, J. M., and Cutler, A. (2003). Perceptual learning in speech. Cognitive Psychology, 47(2):204–238.

Onishi, K. H., Chambers, K. E., and Fisher, C. (2002). Learning phonotactic constraints from brief auditory experience.Cognition, 83:B13–B23.

Perception of English corrective focus by native Inuktitut speakers

Laura Colantoni1, Alana Johns1, Michelle Yuan2 & Marta Ortega-Llebaria3

1University of Toronto, 2MIT, 3University of Pittsburgh

What happens when native speakers of a language with very limited use of intonation acquire an intonational language? We seek to answer this question by analyzing the perception of English corrective focus (e.g. Is Bobby the dog? No, TOBY is the dog) by native speakers of Inuktitut, a language in which the use of intonation is restricted to phrasing (Fortescue 1984; Shokeir 2009); i.e. pitch movements appear only at the end of utterances. As a result, the contrastive meaning expressed by sentence prominence in English is expressed by means of morphology in Inuktitut; the use of polysinthesis is a characteristic of Eskimo-Aleut languages (Johns 2010). Therefore, it is expected that Inuktitut speakers will find it difficult to associate pitch movements with prominence, because their L1 meaning-form association will direct their attention to morphology. However, given that form-meaning mappings are not available in low-pass filter stimuli, we expect to find better identification scores in a task that targets auditory processing. To test these predictions, we compared the performance of a group of 22 speakers of Inuktitut against 13 English controls. All the Inuktitut speakers were first exposed to English in elementary school and use English in their daily lives but not to the same extent. Most speakers (N=16) rated themselves as Advanced or Near Native, while the remaining 6 speakers rated themselves as Intermediate in some but not all the language skills. The test included two force-choice identification experiments plus two production tasks (not reported here). Experiment 1 was designed to test the use of acoustic cues, and consisted of non-speech like stimuli (i.e. low-pass filtered utterances of original English sentences with focus on the subject, verb or object). Participants heard an English sentence followed by three low-pass filtered utterances and were asked to choose the contour that matched the sentence more closely. Experiment 2 was a contextualized task designed to test intonation-meaning mappings. Participants heard the story “Frog, where are you?” by M. Meyer, and they were asked to chose the most appropriate answer to a question out of three possible answers. In both cases, there were 15 target stimuli (5 each with focus on the subject, verb or object) plus distractors. Results showed that native and L2 learners behaved consistently across task but diverged in their overall performance. As predicted both groups performed better in Experiment 1, which focused on acoustic cues (Mean correct responses: English 4.89 vs. Inuktitut 1.98) than in Experiment 2 (English 4.69 vs. Inuktitut 1.48), where the answers were related to the comprehension of a story. The results of our experimental group also support our predictions based on L1 transfer. Interestingly, and in spite of the low number of correct responses, there was a relatively wide-range of variability across participants (Experiment 1: 1.3-3.5; Experiment 2: 0.3-2.3) and, to a lesser extent, across focus conditions (with the lowest number of errors when the object was focalized and the highest when the verb was on focus). Thus, although results are expected based on patterns of cross-linguistic influence, and are consistent with recent findings suggesting that L1 prosodic transfer affects auditory as well as non-auditory perception (e.g. Ortega Llebaria & Colantoni 2013), they should be taken with caution considering the variability in the linguistic experience of each of the speakers (Dorais 2010; Johns 2010) and their relatively lack of familiarity with more metalinguistic tasks.

Works Cited: Dorais, L-J. (2010). The Language of the Inuit. Montreal & Kingston: McGill-Queen’s

University Press. Fortescue, M. (1984). West Greelandic. New York: Routledge. Johns, A. (2010). Eskimo-Aleut languages. Language and Linguistics Compass 4.1041-1055. Ortega-Llebaria, M. & L. Colantoni. (2013). The L2 acquisition of English intonation: Relations

between form-meaning associations, access to meaning and L1 transfer. Studies in Second Language Acquisition 36 (2).

Shokeir, V. (2009). Intonation in Inuktitut. Ms. University of Toronto.

The expression of future temporal reference in Picardie French

Philip Comeau Anne-José Villeneuve (University of Ottawa) (University of Toronto)

The expression of future temporal reference (FTR) has been widely studied across Romance languages (e.g. Poplack & Malvar 2007; Aaron 2010), including spoken French varieties. For centuries, French grammarians (Maupas 1607, Antonini 1753) described the choice between two main variants—the periphrastic future (ça va être cette année ‘it’s going to be this year’) and the inflected future (ça sera au mois d'octobre ‘it’ll in October’)—as influenced by temporal distance; specifically, periphrasis was argued to express a proximate future (i.e. le futur proche). However, results from different varieties of French display surprising heterogeneity with respect to this linguistic variable. For instance, studies of Laurentian (Poplack & Turpin 1999, Wagner & Sankoff 2011) and Continental French (Roberts 2012) have challenged grammarians’ descriptions by showing that the temporal distance constraint is weak or inoperative. In these varieties, the strongest predictor of variant choice is sentential polarity: negative clauses strongly favour the inflected future. In contrast, conservative varieties of Acadian (King & Nadasdi 2003) and Martinique French (Roberts 2013) show a lack of polarity effect: in these varieties, temporal distance is the strongest constraint on variant choice, a finding that supports grammarians’ description. Thus, varieties of spoken French appear to be divided between two types of systems with respect to future temporal reference. To contribute to these lines of research, we examine FTR in a recent corpus of 24 French interviews collected in a rural area of northwestern France where Picard, a Gallo-Romance language in which the inflected future is strongly preferred, still enjoys a relative vitality. By analyzing this contact variety, we also tap into the role that Picard may have played on the development of FTR variation in the area. To determine the factors that condition variant choice, we analyze spoken French data from Picard–French bilinguals and French monolinguals and consider a number of social (sex, age, class, bilingualism status) and linguistic (temporal distance, sentential polarity, subject type, etc.) factors proposed in the literature. Our results show that the inflected future is used at a rate of 38% in Picardie French, much higher than in Laurentian varieties. Bilingualism status was also found to play no role on this variable. Instead, socioeconomic class was the only social statistically significant factor: upper-class speakers had higher rates of the inflected future variant (54%, N=66) than middle- and lower- class speakers (33%, N=125 and 30%, N=76). Linguistic conditioning also differed not only from Laurentian but also from other Continental French varieties: preliminary results from multivariate analyses reveal that polarity, despite its strong effect in most varieties of French studied to date, does not constrain variant choice in Picardie French. Instead, temporal distance highly constrains variant choice, with proximate events (within the hour or sooner) strongly favouring periphrasis. This finding, while a contrast to other studies, mirrors closely the patterns reported for Acadian and Martinique French.

Our study contributes to our understanding of this variable in French by showing that varieties of Continental French, like their Canadian counterparts, can fall along either type of systems with respect to future temporal reference.

References Aaron, J. E. (2010). Pushing the envelope: Looking beyond the variable context in Iberian

Spanish futures. Language Variation and Change, 22.1: 1-36. Antonini, A. (1753). Principes de la grammaire françoise, pratique et raisonnée. Paris:

Duchesne. King, R. and Nadasdi, T. (2003). Back to the future in Acadian French. Journal of French

Language Studies, 13.3: 323-337. Maupas, C. (1607). Grammaire et syntaxe françoise. Paris: Adrian Bacot. Poplack, S and Malvar, E. (2007). Elucidating the transition period in linguistic change: The

expression of the future in Brazilian Portuguese. Probus, 19.1: 121-169. Poplack, S. and Turpin, D. (1999). Does the futur have a future in (Canadian) French? Probus,

11.1: 133-164. Roberts, N. S. (2012). Future Temporal Reference in Hexagonal French. University of

Pennsylvania Working Papers in Linguistics, 18.2: Article 12. Roberts, N. S. (2013). The influence of linguistic factors on the expression of futurity in

Martinique French. Newcastle Working Papers in Linguistics, 19.1: 138-151. Wagner, S. E. and Sankoff, G. (2011). Age grading in the Montréal French inflected future.

Language Variation and Change, 23.3: 275-313.

An argument for genuine object agreement in InuitRichard Compton, McGill University

Claim: I argue that ϕ-indexing morphology in Inuit includes genuine cases of object agreementexponence, contra recent work (see below) that has called into question the existence of objectagreement cross-linguistically and recast apparent instances thereof as pronominal clitics (and thusclitic doubling when an object is present). While tense-variance—proposed by Nevins (2011) as adiagnostic for differentiating agreement from clitics—is inadequate to diagnose the status of Inuitϕ-indexing morphology, mood-variance can instead serve to distinguish real agreement.Background: Recent work by Preminger (2009), Woolford (2010), Arregi & Nevins (2008),Nevins (2011), and Kramer (to appear) has recast a number of apparent cases of agreement asactually being clitics. Kramer suggests that further instances of “purported object agreement”(p.30) cross-linguistically may in fact also be clitics and Nevins (2011) proposes “an analysis ofall cases of object agreement as pronominal clitics in languages with agreement with both sub-ject and object” (p.967). For Inuit, such analyses would mean that the subject/object ϕ-indexingelements in (1)-(2) would consist (at least in part) of object clitics (Dorais 1988; Lowe 1985).(1) arna-up

woman-ERG.SG

niri-ja-Naeat-DECL.TR-3SG.3SG

aapuapple

‘The woman is eating the apple.’

(2) taku-ja-gitsee-DECL.TR-1SG.2SG‘I see you (sg.).’

While Nevins (p.959) argues that “morphophonolgical clitichood and morphosyntactic clitichoodare orthogonal” and that phonological criteria should not be used to establish syntactic clitichood,he proposes that pronominal clitics can be distinguished from agreement using the criterion oftense-invariance (along with Person-Case Constraints and Omnivorous number). If pronominalclitics belong to the category D, as argued by Nevins and a number of other works cited above (orperhaps pro-ϕ heads; Dechaine & Wiltschko 2002), we do not expect them to be sensitive to tense.Conversely, genuine agreement can be conditioned by tense (e.g., she walks vs. she walked).

However, the structure of Inuit verbal complexes is such that tense markers are separated fromϕ-indexing morphemes by mood and often elements such as the perfective marker and negation.Despite the lack tense-variance, which Nevins identifies as crucial to identifying genuine agree-ment, I argue that all this ϕ-indexing morphology is agreement—object agreement included.Evidence from mood: While Inuit ϕ-indexing morphemes are invariant with respect to tense, weinstead observe that they are variant with respect to mood. For example, in Eastern Inuktitut wefind distinct agreement morphology for 2SG.1SG in the indicative and interrogative moods:(3) taku-va-rma

see-INDIC.TR-2SG.1SG‘You (sg.) see me.’

(4) taku-viNasee-INTERR.TR.2SG.1SG‘Do you (sg.) see me?’

While some ϕ-indexing morphemes are stable across moods, others exhibit distinct forms betweenthe participial, declarative, interrogative and conditional, and conjunctive moods (not shown). Justas there is no principled reason why pronominal clitics of category D should vary with tense, it isalso unexpected that they should vary with mood. Conversely, agreement often conditions mood.Discussion: While Chomsky (2004) proposes that ϕ-features are inherited from C in languageslike English, the C head itself is the locus of agreement in Inuit. This explains (i) the positionof agreement, (ii) its form being conditioned by mood, and (iii) the existence of portmanteaumood/agreement (not shown). This paper expands Nevins’ (2011) criteria for distinguishing agree-ment from clitics, adding mood-variance to the set of properties exhibited by genuine agreement.

1

His diagnostic of variance is crucial, but what kind of variance depends on the locus of agreement(Dechaine & Wiltschko, to appear; Oxford, to appear, on object agreement in Algonquian.)References:Arregi, Karlos & Andrew Nevins (2008). Agreement & Clitic Restrictions in Basque. In Roberta

D’Alessandro, Susann Fischer and Gunnar Hrafn Hrafnbjargarson (eds.) Agreement Re-strictions, 49–86. New York: Mouton de Gruyter.

Dechaine, Rose-Marie & Martina Wiltschko (2002). Decomposing Pronouns. LI 33.3:409–442Dechaine, Rose-Marie & Martina Wiltschko (to appear). Micro-variation in Agreement, Clause-

typing and Finiteness: Comparative Evidence from Blackfoot and Plains Cree. The Pro-ceedings of the 42nd Algonquian Conference.

Dorais, Louis-Jacques (1988). Tukilik: An Inuktitut Grammar for All. Inuit Studies, OccasionalPaper, no. 2. Qebec: Association Inuksiutit Katimajit and Groupe d’etudes inuit et circum-polaires (GETIC).

Johns, Alana (to appear). Anaphoric arguments in Unangax and Eastern Canadian Inuktitut.Johns, Alana (2013). Ergativity lives: Eastern Canadian Inuktitut and *clitic doubling. Presented

at the annual meeting of the Canadian Linguistics Association, University of Victoria.Lowe, Ronald (1988). Kangiryuarmiut uqauhingita ilihautdjutikhangit—Basic Kangiryuarmiut

Eskimo Grammar. Inuvik, NWT: Committee for Original Peoples Entitlement.Lowe, Ronald (1985). Siglit Inuvialuit Uqausiita Ilisarviksait—Basic Siglit Inuvialuit Eskimo

Grammar. Inuvik, NWT: Committee for Original Peoples Entitlement.Merchant, Jason (2011). Aleut Case Matters. In Etsuyo Yuasa, Tista Bagchi and Katharine Beals

(eds.) Pragmatics and Autolexical Grammar. In honor of Jerry Sadock, pp. 193–210. NewYork: John Benjamins.

Nevins, Andrew (2011). Multiple agree with clitics: person complementarity vs. omnivorousnumber. NLLT 29.4:939–971

Oxford, Will (to appear). Multiple Instances of Agreement in the Clausal Spine: Evidence fromAlgonquian. Proceedings of WCCFL 31.

Pittman, Christine (2009). Complex verb formation revisited: Restructuring in Inuktitut and Nuu-chah-nulth. In M.-A. Mahieu and N. Tersis (Eds.), Variations on Polysynthesis, pp. 135–147. Amsterdam: John Benjamins.

Preminger, Omer (2009). Breaking Agreements: Distinguishing Agreement and Clitic Doublingby Their Failures. LI 40.4.

Woolford, Ellen. 2010. Active-Stative Agreement in Choctaw and Lakota. ReVEL 8:6–46. (Re-vista Virtual de Estudos da Linguagem) Special Issue No.4 Optimality Theoretic Syntax,edited by Gabriel de vila Othero and Srgio de Moura Menuzzi.

2

In defense of the truncation hypothesis for main clause phenomena Carlos de Cuba

University of Calgary

Introduction: In order to account for the fact that main clause phenomena (MCP) like topicalization are available in peripheral adverbial clauses (PACs) but not central adverbial clauses (CACs), Haegeman (2006a,b) proposes that the CP layer of CACs is structurally deficient, while PACs are fully articulated. Haegeman (2006b) extends the truncation analysis to finite that clauses, claiming that the CP layer of factive complements is also truncated, providing an explanation for why factive complements resist MCP. However, in subsequent work Haegeman (2012:189-193) argues against the truncation hypothesis, providing a series of what she sees as problems with this type of analysis. In this paper I make two main claims: (a) Haegeman’s (2006a,b) truncation account is problematic; (b) Haegeman’s (2012) arguments against truncation are specific to her (2006a,b) view of the left periphery, and they do not create serious problems for other truncation analyses (such as de Cuba 2007, de Cuba & Ürögdi 2009). Problems with Haegeman’s (2006a,b) account: Although she takes a cartographic view, Haegeman’s truncation account assumes a left periphery, shown for English in (1a-c), that deviates substantially from a more standard Rizzian (1997 et seq.) implementation (1d). (1) a. Root clauses: Mod Top Foc Top Force Mod Fin

b. Embedded MCP clauses/PACs: Sub Mod Top Foc Top Force Mod Fin c. Embedded factive clauses/CACs: Sub Mod Fin d. Rizzi (2004): Force Top Int Top Foc Mod Top Fin

Note the innovations that Force is below a number of projections and that Sub is the highest projection in both adverbial clauses and that clauses. For the latter innovation Haegeman appeals to Bhatt & Yoon (1992), who differentiate a pure subordinator position from a position encoding illocutionary force. For adverbial clauses this position hosts the subordinating conjunction, which can then select a PAC with Force or a CAC without Force. Haegeman then extends the analysis to embedded that clauses, treating that as a subordinating conjunction. However, this is a curious move given that Bhatt & Yoon (see also Szabolcsi 1994) specifically analyze English as a language that conflates subordinators and force in one position, as opposed to say Yiddish, Korean or Hungarian. If this is correct, then an analysis like (1b-c), which crucially depends on the independence of Sub and Force, is untenable for English that clauses. Also problematic is the fact that that does not seem to pattern syntactically with subordinating conjunctions (before, when, because, etc.) externally (2), or in extraction (3). (2) a. I closed the door before/when/because/*that John was yelling at Mary. b. I said *before/*when/*because/that John was yelling at Mary. (3) a. *Who did you close the door before/when/because John was yelling at? b. Who did you say that John was yelling at? An existing alternative: Haegeman (2012:261) claims that her arguments against truncation in adverbial clauses hold for truncation accounts of that clauses. However, the majority of the arguments are against the innovations in (1a-c). I argue specifically that the truncation account of de Cuba & Ürögdi 2009 (truncated referential [CP], fully articulated non-referential [cP [CP]) stands up to the challenges posed by Haegeman (additionally, de Cuba & Ürögdi do not appeal to the problematic concept “assertion”). Thus, the truncation hypothesis remains as a challenger to the event operator intervention account of MCP (Haegeman & Ürögdi 2010, Haegeman 2012).

References: Bhatt, Rakesh, and James Yoon. 1992. On the composition of COMP and parameters of V2. In

Proceedings of WCCFL, Vol. 10, pp. 41-52. de Cuba, Carlos. 2007. On (Non)Factivity, Clausal Complementation and the CP-Field. Doctoral

Dissertation. Stony Brook University. de Cuba, Carlos, and Barbara Ürögdi. 2009. Eliminating factivity from syntax: Sentential

complements in Hungarian. In: Approaches to Hungarian: Vol. 11. Marcel Den Dikken and Robert Vago (eds.). John Benjamins. 29-63.

Haegeman, Liliane. 2006a. Argument fronting in English, Romance CLLD and the left Periphery. In: Zanuttini, R., Campos, H. Herburger, E., Portner , P. (eds.), Negation, Tense and Clausal Architecture: Cross-linguistic Investigations. Georgetown University Press, 27-52.

Haegeman, Liliane. 2006b. Conditionals, factives and the left periphery. Lingua 116:1651-1669. Haegeman, Liliane. 2012. Adverbial Clauses, Main Clause Phenomena, and the Composition of

the Left Periphery: The Cartography of Syntactic Structures, Volume 8. Oxford: Oxford University Press.

Haegeman, Liliane, and Barbara Ürögdi. 2010. Referential CPs and DPs: An operator movement account. Theoretical linguistics, 36(2-3), 111-152.

Rizzi, Luigi. 1997. The Fine Structure of the Left Periphery. In: Elements of Grammar: A Handbook in Generative Syntax. Liliane Haegeman (ed.). Kluwer: Dordrecht. 281–337.

Rizzi, Luigi. 2004. Locality and left periphery. In Belletti, A., (Ed.), Structures and Beyond. The Cartography of Syntactic Structures, Vol. 3. Oxford University Press, Oxford. 223-251.

Szabolcsi, Anna. 1994. The noun phrase. Syntax and semantics, 27, 179-274.

The Nasal 'Ash' System in English: how does it get so tense?Dr. Paul De Decker

Memorial University of Newfoundland

This paper investigates the production of the low front vowel /æ/ which undergoes “tensing” and “raising” (Ferguson, 1972; Labov, 1989) when followed by a nasal consonant in some dialects of English (Labov, Ash and Boberg 2006). Speakers in their early twenties, and life long residents of the province of Newfoundland read a word list which included five tokens of both hand and hat. All speakers exhibited significantly lower F1 and higher F2 values for the vowel in the word hand compared to hat. As a way to explain the acoustic differences found in nasal systems, we test the model outlined by Krakow et al. (1988), that such lowering and raising is a result of co-articulation between the vowel and the following nasal consonant. We predict that:

1. If nasal co-articulation is responsible for tensing, then nasality, measured by A1-P0 (Chen 1997, Chen et al. 2007) should be higher in the nasal environment compared oral ones (i.e. elsewhere).

2. (a) Nasality and (b) its concomitant effects on F1 and F2 should be weakest early in the vowel and strongest immediately preceding the nasal consonant. This follows from Cohn's (1993) formulation of co-articulatory nasalization as a phonetically gradient rule as opposed to a categorical phonological one.

All acoustic analyses were conducted using Praat (Boersma and Weenink 2013) at two temporal locations (20% and 80%) into each vowel token. Two-tailed t-tests revealed significantly lower A1-P0 values (more nasality) in the nasal environment for Speaker 1 (t (18) = 5.5011, p < 0.0001), but not for Speaker 2 (t (18) = 1.9266, p = 0.07) or Speaker 3, (t (18) = 0.2687, p = 0.79). This presents a challenge to Hypothesis (1) above suggesting that tensing might not directly related to nasalization.

To test Hypothesis (2), three two-tailed t-tests were run for each speaker to examine the effect of nasality over the course of /æ/. No statistically significant differences were found in A1-P0 across the duration of the vowel for Speaker 1 (t (8) =1.008, p = 0.34). However, significant differences were found for F1 (t (8) = 2.8361, p = 0.02, and for F2 (t (8) = 16.6312, p < 0.0001), though not in the direction expected if affected by nasalization. Likewise, no significant differences were found in A1-P0over the duration of /æ/ for Speaker 2 (t (8) = 0.399, p = 0.70). While a significant difference was found for F2 of Speaker 2 (t (8)= 3.1884, p = 0.01), and not in the direction expected, no effect was found for F1 (t (8) = 0.5862, p = 0.57). Finally, for Speaker 3 neither A1-P0 (t (8) = 1.0342, p = 0.33), F1 (t (8) = 0.0224, p = 0.98) nor F2 (t (8)= 1.1659, p = 0.0.28) were significantly different over the course of the vowel. These results suggest that the level of nasality is as high at 20% into the vowel as it is at 80% and that /æ/ was no higher or more fronted 80% into the vowel than at 20%.Taken together, these results run against hypothesis (2).

Two critical findings are discussed. First, some speakers may not have higher levels of nasality for /æ/ in the nasal environment. Therefore, nasalization is not likely responsible for tensing in the speech of these speakers (contra Hypothesis 1). This is consistent with observations of other tensing systems (De Decker and Nycz 2012) where more fronted and raised lingual gestures were found in the nasal environment. Second, speakers who do exhibit higher levels of nasality for /æ/ in the nasal environment, show that co-articulatory nasalization does not apply in a gradient manner (contra Hypothesis 2). Rather, nasalization is applied categorically, affecting the acoustics of the whole vowel. The overall significance of this study reveals two types of nasal systems mediated through either a) lingual specification or b) nasalization. In both sub-systems /æ/-tensing is understood as a categorical, phonological phenomenon, not one driven by phonetic implementation.

References

Boersma, P., Weenink, D. 2013. Praat: Doing phonetics by computer. Retrieved from http://www.praat.org/.

Chen, M. 1997. Acoustic correlates of English and French nasalized vowels. The Journal of the Acoustical Society of America 102(4), 2360–2370.

Chen, N. F., J. S., Stevens, K.N. 2007. Vowel nasalization in American English: acoustic variability due to phonetic context. Proceedings of 16th International Congress of Phonetic Sciences, 905–908.

Cohn, A. 1993. Nasalisation in English: phonology or phonetics. Phonology 10(1), 43–81.

De Decker P., Nycz, J. 2012. Are tense [æ]s really tense? The mapping between articulation and acoustics. Lingua 122(7), 810–821.

Ferguson, C. 1972. ‘Short a’ in Philadelphia English, in: Smith, E. (Ed.), Studies in Linguistics in Honor of George L. Trager. Mouton.

Krakow, R.A., Beddor, P.S., Goldstein, L., Fowler, C.A. 1988. Coarticulatory influences on the perceived height of nasal vowels. The Journal of the Acoustical Society of America 83, 1146–1158.

Labov, W. 1989. The exact description of the speech community: Short a in Philadelphia, in: Fasold, R., Schiffrin, D. (Eds.), Language Change and Variation. John Benjamins, 1–57.

Labov, W., Ash, S., Boberg, C. 2006. The Atlas of North American English: Phonetics, Phonology, and Sound Change: A Multimedia Reference Tool. Walter de Gruyter.

Dimensions of Variation of the EPP Julianne Doner, University of Toronto

Two different dimensions of cross-linguistic variation for the EPP have been proposed. First, Alexiadou and Anagnastopoulou (1998) propose a contrast between languages which check the EPP with a phrase, and those which check it with a head. Second, analyses such as Davies and Dubinsky (2001) and Massam and Smallwood (1997) propose a contrast between a verbal and a nominal EPP. I argue that these two dimensions of variation cross-classify to create a total of four EPP types, as shown in the table in (1) (see also Doner 2012). (1) The EPP checked by… A Nominal Element A Verbal Element A Phrasal Element DP-EPP (e.g., English) vP-EPP (e.g., Niuean) A Head Element Dº-EPP (e.g., Greek) Vº-EPP (e.g., Irish) At this point, the notion of the EPP covers such a large number of distinct processes that it becomes less clear that they are all the results of the same requirement. However, I show that the EPP varies along these dimensions within a single language, giving evidence that they are, in fact, equivalent. I show this for Arabic, which alternates between a Dº-EPP and a DP-EPP, and for Afrikaans, which alternates between DP-EPP and vP-EPP. Modern Standard Arabic exhibits two different word orders, each associated with a different mechanism of EPP-checking. The VSO word order, which has only partial subject-verb agreement, has a DP-EPP, as shown by the insertion of the expletive in spec,Infl in (2). Note that Aoun et al. (2010) argue that the verb raises above Infl in the VSO word order.

(2) kaana hunaaka Taalib-un fii l-ħadiiqati (Aoun et al. 2010: 70) was.3M.SG there student-NOM in the-garden ‘There was a student in the garden.’ In contrast, the SVO order (3) has rich agreement and allows null subjects. Here, as wth pro-drop languages (Alexiadou and Anagnostopoulou 1998), the EPP is checked by a Dº on the verb.

(3) a. l-muʕallim-uun ʔakal-uu (Aoun et al. 2010: 76) the-teacher-M.PL.NOM ate-3M.PL

‘The teachers ate.’ b. ya-drus-uun (Aoun et al. 2010: 59) 3-study-M.PL ‘They study.’

While Arabic alternates between a DP-EPP and a Dº-EPP, modern spoken Afrikaans shows an alternation on the other dimension, between a vP-EPP and a DP-EPP. Biberauer (2010) argues that the word order difference shown below occurs because in some cases (4a), the entire vP raises to spec,Infl to check the EPP, while in other cases (4b), only the subject pronoun raises.

(4) a. Ek weet [CP dat [InflP [vP sy dikwels Chopin gespeel] het <vP>]]. I know that she often Chopin played has b. Ek weet [CP dat [InflP [DP sy] het [vP <DP> dikwels Chopin gespeel]]]. I know that she has often Chopin played ‘I know that she often played Chopin.’ (Biberauer 2010: 171) Thus, the typology in (1) is supported empirically by synchronic intra-linguistic variation in Arabic, which varies as to the size of the element which checks the EPP, and in Afrikaans, which varies as to the kind of element which checks the EPP. This typology allows us to maintain that the EPP is cross-linguistic, that the EPP is obligatorily checked in some form in every instance of Infl, and that it must always be satisfied by a local relationship.

Selected References Alexiadou, Artemis and Elena Anagnostopoulou. 1998. Parametrizing AGR: Word order, V-

movement and EPP-checking. NLLT 16: 491-539. Aoun, Joseph, Elabbas Benmamoun, and Lina Choueiri. 2010. The Syntax of Arabic. Cambridge:

Cambridge University Press. Barbosa, Pilar. 2011. Pro-drop and theories of pro in the Minimalist Program. Language and

Linguistics Compass 5(8): 551-587. Biberauer, Theresa. 2010. Semi null-subject languages, expletives, and expletive pro

reconsidered. In Theresa Biberauer, Anders Holmberg, Ian Roberts, and Michelle Sheehan, eds. Parametric Variation: Null Subjects in Minimalist Theory. Cambridge: Cambridge University Press. 153-199.

Davies, William D. and Stanley Dubinsky. 2001. Functional architecture and the distribution of subject properties. In William D. Davies and Stanley Dubinsky, eds. Objects and Other Subjects: Grammatical Functions, Functional Categories, and Configurationality. Netherlands: Kluwer Academic Publishers. 247-279.

Doner, Julianne. 2012. A Typology of EPP-Checking Mechanisms. MA Forum Paper, University of Toronto.

Massam, Diane and Carolyn Smallwood. 1997. Essential features of predication in English and Niuean. In Kiyomi Kusumoto, ed. Proceedings of NELS 27. Graduate Linguistics Student Union, University of Massachusetts, Amherst. 263-272.

Synchronic Productivity of Finnish Vowel Harmony Liisa Duncan University of Toronto

Finnish has a well-studied palatal harmony system whereby front and back vowels cannot co-occur in non-compound words. Suffixes alternate according to the stem’s harmonic class. This paper examines the phonetics of stem and suffix harmony, showing it is not synchronically productive for all speakers.

To determine whether harmony is synchronically productive, a phonetic experiment examining the stem and suffix harmony of harmonic and disharmonic loanwords was performed. The experiment included a language game involving the transposition of initial CV sequences of adjacent words, providing a nonce test for both types of harmony. While 29% of disharmonic loans were harmonized in the normal reading, re-harmonization varied in the game from 10-43%, depending on the original word type and the harmonic class of the switched vowel and the remaining word portion. Though some harmonization did occur, these results run counter to claims by Campbell (1980) and Harrikari (2000) that re-harmonization is automatic and fully productive in such games. The lack of productive harmonization of stems in the nonce setting of the game indicates that stem harmony is no longer fully productive, at least for some speakers.

Many studies addressing suffixal harmony utilize orthographic data, assuming it is representative of the spoken realizations. However, phonetic studies indicate that this is not necessarily the case. Välimaa-Blum (1999) found that 5-19% of her loan tokens were affixed with a central vowel rather than the expected harmonic vowel; this vowel is not distinctly realized orthographically. My experiments on loans found that some speakers, especially females, produced almost exclusively front suffixes, which did not conform with their written forms.

The unusual suffixation of loans, though of interest, does not necessarily indicate that suffixal harmony is in a state of decline as these words could form a separate stratum or be lexical exceptions. However, acoustic studies of native words indicate that suffixal harmony in these words may also be less than fully productive, at least for some speakers or dialects. For certain speakers/dialects, the low vowels may be only barely distinct. Kuronen’s (2000) examination of Tampere Finnish vowels shows that low vowels were situated extremely close. Iivonen & Harnud (2005: 65) state that “it is a well-known phenomenon that an auditory confusion of /æ/ and /a/ is possible in the region.” Eerola & Savela’s (2012) work in south-west Finland also shows a similar small difference between these vowels. In Mahonen’s (2011) study of suffix vowels of Helsinki speakers, 20% of her speakers showed overlap of front and back suffixes. In my game data, speakers produced a significant number of neutral suffixes and more front harmonic suffixes.

There thus appears to be evidence that stem and suffixal harmony may not be synchronically productive for all speakers/dialects. If there has been a weakening of the system, what might cause such a shift? Many languages which have lost harmony have been influenced by massive borrowing of disharmonic loans or have succumbed to the influence of internal pressures on the system including vowel mergers and shifts or a change of the [±back] distinction from vowels to consonants (Comrie 1981). Though Finnish does not have these pressures, it has other internal pressures. Most inflectional suffixes contain low vowels which, of the harmonic vowels, are closest in the vowel space. Sentence and word intonation typically employ falling intonation (Suomi et al 2008). Glottalization, breathy voice, and devoicing are all common on final vowels, which are normally unstressed. This conspiracy of phonetic factors may result in lessened prominence of the final vowels, which are often the very harmonic vowels under discussion. Together, these internal pressures may have conspired to reduce the perceptibility of the harmonic suffix vowels, leading to the weakening of harmony.

References

Campbell, L. 1980. The Psychological and Sociological Reality of Finnish Vowel Harmony. In R. Vago (ed.) Issues in Vowel Harmony. Amsterdam: john Benjamins. pp. 245-266.

Comrie, B. 1981. The Languages of the Soviet Union. London: Cambridge University Press.

Eerola, O. and J. Savela 2012. Production of Short and Long Finnish Vowels with and without Noise Masking. Linguistica Uralica. 48/3. pp. 200-2008.

Harrikari, H. 2000. Segmental length in Finnish - studies within a constraint-based approach. Doctoral dissertation, University of Helsinki.

Iivonen, A. and H. Harnud 2005. Acoustical Comparison of the Monophthong Systems in Finnish, Mongolican, and Udmurt. Journal of the International Phonetic Association. 35/1. pp. 59-71.

Mahonen, K. 2011. The Productivity of Finnish Vowel Harmony: An analysis of disharmonic words. Doctoral Dissertation, The University at Buffalo, State University of New York.

Suomi K., J. Toivanen, and R. Ylitalo 2008. Finnish Sound Structure: Phonetics, phonology, phonotactics, and prosody. Oulu: University of Oulu Press.

Välimaa-Blum, R. 1999. A Feature Geometric Description of Finnish Vowel Harmony Covering both Loans and Native Words. Lingua 108. pp. 247-268.

The role of event knowledge in semantic interpretation

Veena Dwivedi & Kaitlin Curtiss

In order to investigate the role that heuristic vs. algorithmic mechanisms play in language comprehension, we conducted a picture verification study with word triplets that described a conceptual script, such as KID CLIMB TREE (cf., Chwilla & Kolk, 2005). These words formed the N1VN2 lexical skeleton from quantifier scope ambiguous sentences such as Every kid climbed a tree. Such sentences are semantically ambiguous; either one or several trees were climbed on inverse vs. surface scope interpretations, respectively. This ambiguity is attributed to the algorithmic computation of quantifier scope, where it has been shown that participants prefer the plural interpretation (Kurtzman & MacDonald, 1993; Raffray & Pickering, 2010). However, perhaps the interpretation is not derived via computation but instead from heuristic knowledge regarding events (Kim & Sikos, 2011; Kuperberg, 2007) and the likely number of participants in events. This hypothesis emerges from a previous norming study (Dwivedi et al., 2010), discussed in our recent publication (Dwivedi, 2013), where 32 participants judged tree in Every kid climbed a tree as plural 100% of the time but diamond in Every jeweler appraised a diamond was judged as plural at 60%. Given that these sentences are exactly alike in terms of syntactic structure, differences in interpretation must be due to the individual lexical items, and their contribution to a schema (Schank & Ableson, 1977). In the current word triplet study, 45 subjects were instructed to interpret word chunks such as KID CLIMB TREE as telegrams, and select a picture via button press regarding the number associated with TREE. The hypothesis was that responses to word triplets could serve as a predictor for judgments of full sentences, that is, those displaying quantifier scope ambiguity. In the present work, a binary logistic regression analysis was conducted to test the likelihood of a plural judgment in the norming study using the word triplet responses as a predictor. Results confirmed that the triplet judgments were reliable predictors of the sentence judgments. Implications for the on-line computation of meaning are discussed.

References

Dwivedi, V.D. (2013). Interpreting quantifier scope ambiguity: Evidence of heuristic first, algorithmic second processing. PLoS ONE 8(11): e81461. doi:10.1371/journal.pone.0081461

Dwivedi, V.D., Phillips, N.A., Einagel, S. & Baum, S.R. (2010). The neural underpinnings of semantic ambiguity and anaphora. Brain Research, 1311, 93-109.

Chwilla, D. J. & Kolk, H. H. J. (2005). Accessing world knowledge: Evidence from N400 and reaction time priming. Cognitive Brain Research, 25, 289-606.

Kim, A., & Sikos, L. (2011). Conflict and surrender during sentence processing: An ERP study of syntax-semantics interaction. Brain and Language, 118, 15-22.

Kuperberg, G. R. (2007). Neural mechanisms of language comprehension: Challenges to syntax. Brain Research, 1146, 23-49.

Kurtzman, H. S. & MacDonald, M. C. (1993). Resolution of quantifier scope ambiguities. Cognition, 48, 243-279.

Raffray, C. N. & Pickering, M. J. (2010). How do people construct logical form during language comprehension? Psychological Science. DOI: 10.1177/0956797610375446.

Schank, R. C. & Abelson, R. P.(1977). Scripts, Plans, Goals and Understanding. Hillsdale, NJ: Lawrence Erlbaum Associates.

The cost of attention in semantic processing

Veena Dwivedi, Hope Magnue, Leslie Rowland & Kaitlin Curtiss

In a recent paper (Dwivedi, 2013), 3 self-paced reading experiments examined the processing of quantifier scope ambiguous (QSA) sentences such as Every kid climbed a tree. A new model of language comprehension was proposed, where it was argued that the parser operates in a serial manner but grammatical algorithms are not primary (in contrast to claims by Frazier & Clifton, 1996). Instead, the parser operates in a heuristic first, algorithmic second manner. It is assumed that heuristic processing is a shallow form of perception, whereas algorithmic processing requires higher levels of attention. As such, it is hypothesized that factors contributing to allocation of attention would affect heuristic vs. algorithmic processes. The present experiment used the paradigm as developed in Kurtzman & MacDonald (1993), with a slight change modification to Control unambiguous contexts in order to ensure their referential nature (Kaplan,1978)

The table above summarizes which condition should be dispreferred on algorithmic grounds: the inverse scope interpretation of QSA sentences. This pattern of results was found (Exp2 in Dwivedi, 2013)) in the question response accuracy of participants who had just read QSA context sentences followed by singular vs. plural continuation sentences. That is, even after having just read the singular continuation sentence The tree was in the playground, when asked How many trees were climbed? ONE SEVERAL, participants performed at chance levels. However, RTs for continuation sentences did not pattern as expected on algorithmic grounds, instead, these showed patterns consistent with lexical-pragmatic biases. This is consistent with Ferreira et al.’s claim that the parser operates using ‘good enough’ processing strategies. Deep processing of stimuli is not predicted in real-time language comprehension.

The present work seeks to test this claim by investigating whether the addition of an appropriate linguistic ‘pre- context’ could modulate attention. Perhaps participants do not process QSA sentences deeply because the strong quantifier every does not have a context set over which it can be interpreted. At least 30 participants will be run by April; 10 have been run so far. Results reveal a 10% drop in accuracy rates, surprisingly, for the Control conditions, whereas accuracy rates did not change for the Ambiguous conditions. This could be due to the fact that whereas the Ambiguous pre-context sentence merely sets the scene for the upcoming QSA context, The kids spotted the park on the long walk. Every kid climbed a tree. The tree(s) was/were… for Control conditions, the pre-contexts had a synonym for the referential direct object NP, such as The kids spotted the oak(s) on the long walk. Every kid climbed that/those tree(s). The tree(s) was/were… Thus, when participants had to answer How many trees were climbed? they would have undergone what Garrod & Sanford (1994) call the ‘resolution’ phase of co-reference, which is integration of the NP trees in the question. In the Ambiguous condition, no extra integration for the meaning of trees is required at the question. Thus, at this early phase of data collection, it

AMBIGUOUS CONTEXT UNAMBIGUOUS CONTROL PLURAL CONTINUATION Every kid climbed a tree.

The trees were in the playground. (surface scope) Every kid climbed those trees

The trees were in the playground.

SINGULAR

CONTINUATION Every kid climbed a tree. ?? The tree was in the playground. (inverse scope condition)

Every kid climbed that tree. The tree was in the playground.

seems that there is a cost in terms of accurately responding to the question How many trees were climbed?; where the cost does not have to do with scope computation but instead has to do with the attentional resources required to interpret trees in the question itself.

References Dwivedi, V.D. (2013). Interpreting quantifier scope ambiguity: Evidence of heuristic first,

algorithmic second processing. PLoS ONE 8(11): e81461. doi:10.1371/journal.pone.0081461

Ferreira, F. (2003). The misinterpretation of non-canonical sentences. Cognitive Psychology, 47, 164-203.

Ferreira, F., & Patson, N. D. (2007). The “good enough” approach to language comprehension. Language and Linguistics Compass, 1, 71–83.

Frazier, L, & Clifton, C., Jr. (1996). Construal. Cambridge, MA: MIT Press.

Garrod, S., Sanford, A., (1994). Resolving sentences in a discourse context: how discourse representation affects language understanding. In: Gernsbacher, M. (Ed.), Handbook of Psycholinguistics. Academic Press, pp. 675–698.

Kaplan, D. (1978). Dthat. In P. Cole (Ed.), Pragmatics: Syntax and semantics (pp. 221-243).

New York: Academic Press. Kurtzman H. S., & MacDonald, M. C. (1993). Resolution of quantifier scope ambiguities.

Cognition, 48, 243-279. MacDonald, M. C., Pearlmutter, N. J., & Seidenberg, M. S. (1994). The lexical nature of

syntactic ambiguity resolution. Psychological Review, 101, 676-703.

Individual differences in the strategies used to read morphologically complex words Kaitlin Falkauskas & Victor Kuperman (McMaster University)

Background When reading morphologically complex words, there is a growing consensus that individuals can simultaneously access the lexical form of the whole word (catcher) and the morphemes of that word (catch, -er). The lexical properties of words, including word frequency, and other sources of information affect the balance between the use of these processing routes (Kuperman, Bertram, & Baayen, 2010). For example, words that are shorter and more frequent are more likely to be recognized as whole words rather than via their morphemes. It is unclear, however, how individual differences in reading skill and differences in the strength of individuals' lexical representations for complex words and their component morphemes may affect the extent to which these routes are used. Following the Lexical Quality Hypothesis (Perfetti, 2007), we hypothesized that individuals with different levels of reading skill or experience will vary in the strengths of their lexical representations for complex words and morphemes, and thus will make differential use of the processing routes. We take an effect of whole word frequency as indication that the whole word route is used, and effects of the stem frequency as an indication that the word is being decomposed into its morphemes. To explore this hypothesis we conducted an eye-tracking study in which participants read sentences, such as (1), containing derived words with the suffixes: -er, -ness, -ment and -ation. (1) The only catcher had heartburn so the game was postponed. Method Twenty-eight undergraduate students read sentences containing the derived words while their eye-movements were monitored. The target words varied in whole word frequency and stem frequency. Participants also completed a battery of tests including tests of verbal and cognitive skills (Wechsler, 1999), phonological awareness (Torgesen, Wagner & Rashotte, 1999), exposure to print (Acheson, Wells, & MacDonald, 2008), and word segmentation. We fitted linear mixed effect regression models to the eye-movement measures, with critical interactions of the test scores and either whole word frequency or stem frequencies. Results Individuals with both higher and lower scores on the reading experience test read higher-frequency words faster, with less-proficient individuals being facilitated more by higher whole-word frequencies. At the same time, only individuals less proficient in reading benefitted from higher-frequency stems when reading derived words, while the best-performing individuals were not affected by the words' morphological structures. Discussion The results suggest that individuals with higher scores on a number of skill measures are more likely to recognize words as whole lexical items, rather than decomposing the words into their morphemes. Although these individuals are likely better at segmenting words into morphemes, they do not appear to be using this as a word processing strategy. This suggests that individual differences in the component skills of reading are a major factor in shifting the balance between the processing routes used.

References Acheson, D. J., Wells, J. B., & MacDonald, M. C. (2008). New and updated tests of print

exposure and reading abilities in college students. Behavior Research Methods, 40, 278-289. Kuperman, V., Bertram, R., & Baayen, R. (2010). Processing trade-offs in the reading of Dutch

derived words. Journal of Memory and Language, 62, 83-97. Perfetti, C. A. (2007). Reading ability: Lexical quality to comprehension. Scientific Studies of

Reading, 11, 357–383. Torgesen, J., Wagner, R. & Rashotte, C. (1999). Test of Word Reading Efficiency (TOWRE).

Austin, TX: Pro-ED Inc. Wechsler, D. (1999). Wechsler abbreviated intelligence scale. San Antonio: The Psychological

Corporation.

On the absence of nominal coordination in Gitksan Clarissa Forbes; University of Toronto

The syntactic structure of coordination has been the subject of discussion for many years. Under the generally accepted analysis (Munn 1993, Kayne 1994), a coordinating element serves as the head of a functional projection &P which takes conjuncts in its complement and specifier positions. In this paper, I show that Gitksan (Tsimshianic, northern interior BC) definitively lacks an element which conjoins nominals in such a fashion, and may also lack an element with this structure to coordinate clauses. This claim lies in contrast to Johannessen's (1998) proposal that the functional structure of coordination is crosslinguistically available.

Gitksan's apparent nominal coordinator, an, has previously been described as having a number of properties unexpected for a traditional coordinator (Rigsby 1986, Tarpent 1987). The first conjunct may be a pronominal affix; in addition, the rest of the "conjunction" may dislocate from the first conjunct and appear clause-finally. Both these properties are exemplified in (1).

(1) Jap-i-'y=hl ts'el ky'oots 'niin make-CTRL=1SG.II=CN half.dried.sal yesterday and=CN 2SG.III 'You and I made half-dried salmon yesterday.'

Livingston's (1989) analysis of this morpheme (in mutually-intelligible the traditional coordinate structure, utilizing pro; however, I present additional data that suggests such an analysis in Gitksan would be at best uneconomical, if not entirely untenable. As noted by Davis & Brown (2011), the first conjunct of a set linked by an can be A'-extracted. This directly violates the Coordinate Structure Constraint (Ross 1967), which prevents the extraction of full conjuncts and is known to hold crosslinguistically of coordinate constructions. In light of this evidence, I analyze an as a comitative oblique which adjoins to either full nominals or clauses.

Johannessen (1998) argues that the structure of coordination is universal. When considering languages with no overt coordinator, where nominal elements are simply juxtaposed, a null coordinator is assumed. However, in Gitksan, it is ungrammatical to link two nominals only by juxtaposing them; nominals are linked via the an-oblique or they are not linked at all. It is consequently impossible to posit an additional null element which projects the &P structure.

Where clausal coordination is concerned, the distinct morphology of pronouns in clauses which appear before versus after the clausal coordinator ii presents a similar obstacle to straightforward analysis of this element as a coordinator rather than a subordinating oblique.

(2) a) Di-dal -'y=hl sim'oogit ii na=di- =hl naks-t TR-speak-1SG.II=CN chief and 1SG.I=TR-speak=CN spouse-3SG.II 'I talked to the chief and (then) I talked to his/her spouse.'

b) *… - -'y=hl naks-t

To conclude, this paper shows that a standard coordination structure is incompatible with the facts of nominal "coordination" in Gitksan. I present an alternative analysis where the morpheme adjoins to either the level of the DP or the clause; such a structure is useful for comparison with similar comitative constructions in other languages (e.g. Slavic, Dył 1988). Finally, while further work remains to be done on the structure of coordination in the clausal domain, Gitksan so far appears to be a counterexample to the claim that true coordinate phrases, containing two conjuncts linked by a single functional head, are universal linguistic structures.

References Davis, H. & J. Brown. 2011. "On A'-dependencies in Gitksan." In Proceedings of the 46th

International Conference on Salish and Neighbouring Languages. 43-80.

Dył , S. 1988. "Qu -comitative coordination in Polish." Linguistics 26: 383-414.

Johannessen, J. 1998. Coordination. New York: Oxford University Press.

Kayne, R. 1994. The antisymmetry of syntax. MIT Press.

Livingston, E. 1989. "Conjoined arguments in Nisgha." In Proceedings of the 24th International Conference on Salish and Neighbouring Languages.

Munn, A. 1993. Topics in the syntax and semantics of coordinate structures. PhD dissertation, University of Maryland.

Rigsby, B. 1986. Gitxsan grammar. Ms. University of Queensland.

Ross, J. 1967. Constraints on variables in syntax. PhD dissertation, MIT.

Tarpent, M.-L. 1987. A grammar of the Nisgha language. PhD dissertation, University of Victoria.

Naomi Francis University of Toronto

This predicate is tasty: Predicates of personal taste, faultless disagreement, and the ideal judge

This investigation examines the behaviour of predicates of personal taste (PoPTs) – expressions such as fun and tasty – in English. Their truth depends on whose tastes are under discussion. They also exhibit faultless disagreement, illustrated in (1); native speakers routinely judge that Bill and Sue are disagreeing, but that neither of their utterances is false. This is in contrast to disagreement over non-subjective predicates, as in (2).

(1) Bill: This cake is tasty! (2) Bill: It’s raining outside. Sue: No it’s not! This cake isn’t tasty at all. Sue: No it isn’t! It’s snowing.

This investigation seeks to answer the following research question: What are the semantics of PoPTs, and how can we account for their behaviour in discourse? I propose a new analysis of these predicates; along the way, I also discuss the question of whether faultless disagreement truly exists.

Two analyses of PoPTs are dominant in the literature. Lasersohn’s (2005) indexical approach holds that they are evaluated relative to the tastes of a judge, formalized as a parameter on evaluation parallel to world (w) and time (t) indices. Thus, in (1) Bill and Sue disagree because they assert contradictory contents, but their utterances can both be true because they are evaluated relative to different judges. Stephenson’s (2007) null argument analysis, in contrast, holds that PoPTs are two-place predicates which can take the judge as a covert pronominal argument in the syntax. However, a null argument analysis makes problematic predictions; if PoPTs involve covert pronominal arguments in the syntax, we might reasonably expect these arguments to behave like overt pronouns. However, this is not the case; given a rich enough context, multiple overt pronouns in a sentence can have different referents (3), but multiple PoPTs cannot have different judges (4).

(3) Bill: He1 likes it, but he2 doesn’t. (4) Bill: ??It’s tasty, but it’s not tasty. Intended: Fido likes it, but Spot doesn’t. Intended: It’s tasty for Fido, but it’s not tasty for Spot.

While Lasersohn’s (2005) treatment of the judge as a parameter on interpretation instead of an argument in the syntax avoids the above problem, dialogues such as (1) do not behave like other cases where contradictory contents are evaluated relative to different contextual parameters.

(5) A: Sherlock Holmes lives at 221B Baker Street. w = the world of the novels B: No he doesn’t! He doesn’t even exist. w = the actual world A: Well, I just meant in the books. (Bouchard 2012: 3)

Whereas the dialogue in (5) feels like a simple misunderstanding, Bill and Sue seem to be having a real disagreement in (1), suggesting that parameters on interpretation may not be the locus of faultlessness. I propose an alternative analysis where PoPTs are evaluated relative to an ideal judge which is selected by an ordering source similar to Kratzer’s (1991) modal ordering source; possible judges are ranked according to a contextually determined ideal set of dispositional traits. This analysis holds that, when conversation partners disagree about PoPTs, they do not disagree about whether something is tasty for anyone in the discourse context; instead, they disagree about whether it would be tasty for the ideal judge. Thus, I argue that Bill and Sue are making objective and contradictory claims in (1) as to whether the cake would be tasty for the ideal human judge; the intuition of faultlessness comes from the fact that, as humans, they are equally entitled to make claims about what a human might find tasty. This proposal offers an alternative to the current analyses of PoPTs and makes testable predictions about their behaviour in context. It will therefore have consequences for our understanding of both these subjective predicates and the intuition of faultless disagreement.

Naomi Francis University of Toronto

References

Bouchard, D. (2012). The partial factivity of opinion verbs. Proceedings of Sinn und Bedeutung 17, 133-148. Egan, A., Hawthorne, J., & Weatherson, B. (2005). Epistemic modals in context. Contextualism in philosophy, 131–168. Kaplan, D. (1989). Demonstratives. In J. Almog, J. Perry, & H. Wettstein (Eds.), Themes from Kaplan. Oxford: Oxford University Press. Kratzer, A. (1991). Modality. In A. von Stechow & D. Wunderlich (Eds.), Semantik. Ein internationals Handbuch der zeitgenössischen Forschung (pp. 639–650). Berlin: de Gruyter. Kratzer, A. (2009). Modals & Context Dependency. Colloquium, Harvard University, February 20, 2009. Lasersohn, P. (2005). Context-dependence, disagreement, and predicates of personal taste. Linguistics and Philosophy 28, 643-686. Link, G. (1995). Generic information and dependent generics. In G. Carlson & F. Pelletier (Eds.) The generic book (pp. 358-382). Chicago: University of Chicago Press. Stephenson, T. (2007). Judge dependence, epistemic modals, and predicates of personal taste. Linguistics and Philosophy 30, 487-525. Stojanovic, I. (2007). Talking about taste: Disagreement, implicit arguments, and relative truth. Linguistics and Philosophy 30, 691–706.

Anna Frolova

Développement de la transitivité verbale en russe L1

Cette étude explore le développement de la transitivité verbale en russe L1 dans le cadre de la grammaire générative avec une attention particulière aux omissions illicites de l’objet direct, et ce en comparaison avec le développement du français et de l’anglais L1. Les omissions optionnelles de l’objet direct dans différentes langues sont normalement étudiées dans des contextes référentiels et associées à l’acquisition du clitique accusatif (Wexler 1998 ; Hamann 2003, entre autres). Pourtant, une période d’omissions est également attestée dans les langues sans clitique, comme l’anglais (Pérez-Leroux et coll. 2008). En russe, l’objet direct non référentiel est obligatoire dans certains contextes perfectifs. Un objet lexical est requis dans ces contextes téliques, contenant une projection d’aspect (AspQ) qui permet au verbe et à l’objet de s’accorder en quantité (Borer 2005). En revanche, l’objet est optionnel dans les contextes imperfectifs qui sont atéliques. Nous avons examiné l’emploi de l’objet direct en russe L1 à partir d’une étude de production induite d’enfants monolingues russes âgés de 3 à 5 ans afin de mieux comprendre la nature des omissions illicites en L1. L’expérience a été basée sur la méthodologie de Pérez-Leroux et coll. (2008) et ciblait l’emploi de l’objet dans les contextes optionnellement ou fortement transitifs. Les participants devaient répondre aux questions de l’expérimentateur après avoir écouté des histoires illustrées. Les résultats de notre étude expérimentale, qui s’inscrivent dans le cadre théorique du développement langagier continu (dans sa version faible), révèlent que, premièrement, les enfants russes de 3 à 5 ans omettent optionnellement l’objet dans le contexte perfectif fortement transitif, tandis que les adultes y emploient seulement l’objet lexical. La comparaison des groupes selon le nombre d’objets nuls à l’aide du test Mann-Whitney a démontré des résultats significatifs entre les adultes et les enfants de 3 ans (U = 15 ; p = 0,034 ; r = 0,5), les adultes et les enfants de 4 ans (U = 15 ; p = 0,006 ; r = 0,56), les enfants de 4 et de 5 ans (U = 85 ; p = 0,033 ; r = 0,34). Une diminution des omissions entre 4 et 5 ans indique un développement graduel. Bien que la différence entre le groupe de 5 ans et les adultes ne soit plus statistiquement significative, on constate encore 16 % d’omissions illicites dans la production des enfants de 5 ans. Deuxièmement, ces omissions semblent être indépendantes du développement de l’aspect parce que les enfants utilisent de manière adulte la morphologie verbale, les verbes perfectifs sont préfixés et correctement conjugués. Cependant, les enfants de tous les âges ont occasionnellement employé les verbes à l’imperfectif dans le contexte perfectif (dans 15 - 20 % des cas pour tous les groupes d’âge). La comparaison des groupes selon l’emploi de l’imperfectif dans le contexte perfectif au moyen du test Mann-Whitney a établi une différence significative entre les adultes et les enfants de 3 ans (U = 12 ; p = 0,037 ; r = 0,58) ; les adultes et les enfants de 4 ans (U = 21 ; p = 0,027 ; r = 0,48) ; les adultes et les enfants de 5 ans (U = 18 ; p = 0,027, r = 0,53) ; il n’y a pas de différence entre les groupes dʼenfants. Ce comportement des enfants correspond aux résultats des recherches antérieures et peut être attribué au développement des compétences discursives (Van Hout 2005, Kazanina et Phillips 2007). L’emploi incorrect de l’imperfectif ne semble pas influencer la production de l’objet nul dans le contexte perfectif. En plus, les tests statistiques montrent une différence significative dans le taux d’omissions entre les enfants qui utilisent l’aspect correctement et les adultes (Mann-Whitney : U = 12 ; p = 0,025 ; r = 0,56). En nous basant sur ce résultat, nous pouvons conclure qu’au moins une partie des objets nuls illicites dans la production L1 n’est pas liée au développement aspectuel. Nous proposons que les omissions

illicites de l’objet direct en russe L1 sont causées par le développement de la quantification nominale.

Bibliographie

Borer, H. 2005. Structuring sense. Oxford : Oxford University Press.

Hamann, C. 2003. Phenomena in French normal and impaired language acquisition and their implications for hypotheses on language development. Probus 15, 91-122.

Kazanina, N. et C. Phillips. 2007. A developmental perspective on the imperfective paradox. Cognition 105 : 65-102.

Pérez-Leroux, A.T., M. Pirvulescu et Y. Roberge. 2008. Null Objects in Child Language : Syntax and the Lexicon. Lingua 118 : 370-398.

Van Hout, A. 2005. Imperfect imperfectives : on the acquisition of aspect in Polish. Dans Aspectual Inquiries, sous la direction de P. Kempchinsky et R. Slabakova, 317-344.

Wexler, K. 1998. Very early parameter setting and the unique checking constraint : a new explanation of the optional infinitive stage. Lingua 106 : 23-79.

The (un)interpretability of {α}: a derivational approach

Brandon J. FryUniversity of Ottawa

This paper proposes that the interpretability of a syntactic object (SO) of the form {α} depends on itsderivational origin. If this object is derived by Self Merge, it is uninterpretable because it delivers ambiguousinstructions to the interfaces. If it is derived by Transfer, it is interpretable because ambiguous instructionsare not generated. This lends support to a derivational approach to syntax.

Epstein (2007) notes that Transfer creates objects which are not SOs according to Chomsky (1995). Basedon his definition of Merge as Merge(α, β) = {α, β}, Chomsky defines SOs as (i) lexical items (LIs) or (ii) ofthe form {α, β} where α, β are SOs. Transfer is triggered by phase heads (PHs) and removes the complementsof PHs (PHCs) from the workspace, relaying these PHCs to the interfaces. However, the product of Transferis not an SO. Consider the derivation in (1), where α is a PH.

(1) a. Merge(γ, β) = {γ, β}b. Merge(α, {γ, β}) = {α, {γ, β}}c. Transfer({α, {γ, β}}) = {α}

The object created by Transfer in (1-c) is not an SO because it is neither (i) an LI, nor (ii) of the form {α, β}.This generation of a non-SO should cancel the derivation. The problem, then, is that Transfer creates objectswhich should result in the cancellation of the derivation. Let us call this Epstein’s Problem.

The claim here is that there is an unprincipled condition on the application of Merge and that once thisstipulation is lifted, Epstein’s Problem dissolves. The argument is as follows.

It has been claimed that Merge should apply freely (i.e. without constraints). However, there is a(implicit) distinctness condition in the standard definition of Merge such that α 6= β. This unprincipledcondition should be eliminated. Once this is accepted, the definition of SO must be amended such that SOsmay be (i) LIs or (ii) of the form {α, β} where α, β are SOs and where α = β or α 6= β.

By eliminating the distinctness condition, objects of the form {α} are SOs, since SOs may be of the form{α, β} where α = β, and since {α, α} = {α}. Therefore, objects created by Transfer are now SOs. Epstein’sProblem, thus, dissolves.

The elimination of the distinctness condition also results in the possibility of Self Merge (Adger 2013).Thus, objects of the form {α} derived by Merge(α, α) are generable by the narrow syntax. However, they donot survive at the interfaces. Since α undergoes Merge with α in the narrow syntax, the SO {α} instructs theinterfaces to order α before α. This is an ambiguous instruction, which the interfaces reject (Boeckx 2008).

In summary, this paper hypothesizes that objects of the form {α} may be derived by Transfer or by SelfMerge, but that only those derived by Transfer are interpretable at the interfaces.

References

Adger, David. 2013. A syntax of substance. Cambridge: MIT Press.

Boeckx, Cedric. 2008. Bare syntax . Oxford: Oxford University Press.

Chomsky, Noam. 1995. The minimalist program. Cambridge, MA: MIT Press.

Epstein, Samuel David. 2007. On i(nternalist) functional explanation in minimalism. Linguistic Analysis33:20–53.

1

Faisons at(tension) aux voyelles hautes : le relâchement et l’harmonie vocalique en français de Windsor et de Hearst La situation linguistique du français laurentien (dorénavant FL) en Ontario est compliquée par plusieurs facteurs sociaux, géographiques et langagiers, dont la question de l’influence de l’anglais ainsi que le statut majoritaire ou minoritaire des locuteurs au sein de la communauté figurent parmi les plus saillantes voies de recherches (ex. Mougeon, 1997). Cette œuvre a pour but la description et la comparaison de deux dialectes du FL en Ontario, à savoir celui de Windsor et de Hearst, dans l’intention d’élaborer une analyse du relâchement des voyelles hautes en FL en Ontario, phénomène encore mal connu. Ce projet analyse les données de deux corpus du projet phonologie du français contemporain (PFC) avec le logiciel Praat pour pouvoir être dans la mesure de déterminer avec précision les caractéristiques du relâchement des voyelles hautes.

Cette communication présente d’abord les deux dialectes du FL. Nous soutenons que le relâchement des voyelles hautes manifeste des propriétés intéressantes dans la mesure où les deux dialectes ne démontrent des tendances de relâchement unique. Pour ce faire, nous faisons une première distinction entre le français de l’Ontario et celui du Québec. Avec quelques exceptions, le FL sortant de Montréal est souvent pris pour la forme standard (ex. Dumas, 1981, 1987; Poliquin, 2007). Bien qu’il existe de la variation dans les divers dialectes parlés au Québec, elle est encore plus marquée entre ceux-ci et ceux en Ontario. Également, les deux dialectes mis en évidence dans le présent papier font preuve d’une deuxième distinction marquée. À première vue, le FL à Windsor et à Hearst existe dans deux contextes nettement différents. Hearst compte 86,9% de sa population qui est francophone comme langue primaire, comparé à 2,6% à Windsor; également, 69% de la population de Hearst sont capable de mener une conversation en anglais et français, comparé à 8,3% à Windsor (Recensement 2011, Statistique Canada). Il est donc fort probable que le relâchement des voyelles se manifeste de façon différente dans ces deux milieux.

Selon l’analyse standard (Dumas, 1987; Ostiguy & Tousignant, 1993; Poliquin, 2007; Walker, 1984), les voyelles hautes en FL, à savoir [i y u I Y U], sont caractérisées par une opposition non-contrastive de tension. En général, ces voyelles sont réalisées de façon tendue sauf dans les contextes suivants : une voyelle haute est relâchée quand elle se trouve (A) en position finale de mot au sein d’une syllabe fermée; et (B) en position non-finale de mot si la syllabe est ouverte mais aussi suivie d’une syllabe comprenant une voyelle haute relâchée. Ce dernier contexte est souvent nommé un processus d’harmonie vocalique. À titre d’exemple, observons les données suivantes :

(1) si [i] *[sI] (syllabe ouverte : relâchement non-permis) (2) pipe [I] *[pip] (syllabe fermée : relâchement obligatoire) (3) pilule [i…Y] ~ [I…Y] (syllabe non-final : relâchement facultatif) L’analyse standard propose que le relâchement en (2) est obligatoire tandis que celui en (3) n’est que facultatif. Par contre, les résultats démontrent que les locuteurs de Hearst et de Windsor ont tendance à ne pas relâcher la voyelle haute en (2). Ceci va à l’encontre de l’analyse standard. Il semble aussi que l’harmonisation des voyelles (3) manifeste des caractéristiques uniques à Windsor, en raison de l’influence de l’anglais, mais non à Hearst.

Bibliographie Dumas, D. (1981). Structure de la diphtongaison québécoise. The Canadian Journal of Linguistics, 26(1), 1–61. Dumas, D. (1987). Nos façons de parler. Sillery, Québec: Presses de l’Université du Québec. Mougeon, R. (1997). La recherche sociolinguistique sur le français du Canada. In De la polyphonie à la symphonie. Méthodes, théories, et faits

de la recherche pluridisciplinaire sur le français du Canada (pp. 183–208). Leipzig: Leipziger Universitäts Verlag. Ostiguy, L., & Tousignant, C. (1993). usages. Montréal: uérin universitaire. Poliquin, G. C. (2007). Canadian French vowel harmony. Thesis dissertation. Harvard University. Walker, D. (1984). The pronunciation of Canadian French. Ottawa, Canada: University of Ottawa Press.

Many and determiner systems in English and Russian

Julie Goncharov, University of Toronto

The ambiguity of quantity expressions in English, such as many, is traditionally identified asthe distinction between cardinal and proportional determiners, e.g. Milsark 1977, Barwise andCooper 1981. Although many can appear in canonical adjectival and quantificational positions,the ambiguity holds only for quantificational many. The adjectival many in English is unam-biguously cardinal (Partee 1989: 9). This creates an asymmetric picture, see (1-a). Russianmany is also ambiguous and has two morphological forms - an adverbial uninflected mnog-o‘many-ADV’ and an agreeing adjectival mnog-ie ‘many-AGR’, see Krasikova 2011. The Rus-sian system is also asymmetric, but it is strikingly different from English, see (1-b). Theseobservations raise a number of questions. The present paper answers two of them: 1) Why dothe gaps in (1-a) and (1-b) exist in the first place? 2) Why do these particular gaps exist?

(1) a. Quantificational Adjectivalprop. manycard. many the many

b. Adverbial Adjectivalprop. mnogiecard. (ind.) mnogo

(group) mnogoProposal In this paper, I argue that the differences between the Russian and English many

are attributed to the differences between the determiner systems of these languages. Building onthe proposals in Kyriakaki 2011 and Pereltsvaig 2006, I propose to decompose the D-head intothree heads: i that spells out the uniqueness, Fam that carries the information about familiarityand Ind that provides an ability to have individual referents, see (2):

(2) [iP i [FamP (AP) [Fam0 Fam [IndP Indf [QP Q [NumP (AP) [NumP Num [NP Nf ]]]]]]]]Formally, Ind has a set of unvalued f -features that are valued by the closest nominal in itsc-command domain and then are used to value the features of the main predicate. The pres-ence/absence of Ind is responsible for the ambiguity of Russian mnogo. This proposal capturesnaturally the correlation between the individual/group reading of a nominal and the main pred-icate agreement/its absence, see Pereltsvaig 2006. This proposal also makes correct predictionswith respect to pseudo-partitive constructions, such as a bunch of flowers, in which the Q-headis occupied by an interfering nominal that values the features of Ind triggering singular agree-ment on the main predicate. In addition, this proposal allows room for parametric variation:if Ind merges below the measure nominal in Q or the two nominals are equidistant, we expectto find the agreement with the low nominal or with either. The latter situation is attested inGreek, see Alexiadou et al. 2007. I propose to analyze mnogie as a definite adjective mergingin the specifier of FamP, as in Aljovic 2002. This treatment of mnogie explains its obligatoryadjectival inflection, the absence of genitive of quantification on the following nominal, theagreement of the main predicate (via Ind) and the definiteness effects observed with mnogie.

Asymmetry explained: Russian Suppose that Russian once had two adjectival manys:definite (a long-form which we see today as mnogie) and indefinite (a short-form without AGRmerged as an adjunct to NumP). When Russian lost short-form adjectives in the attributiveposition, the cardinal adjectival many was lost; its role now is performed by mnogo with Ind(to some extent). Russian does not have a proportional adverbial many simply because it doesnot have (and has never had) means to add definiteness apart from as an inflection on adjectives.

Asymmetry explained: English Suppose that English spells out iP and FamP as the, as inKyriakaki 2011. The adjectival many will merge low in NumP and combine with NP intersec-tively; then combined with the, it would be interpreted as unique, definite and cardinal. English

does not have definite adjectives that can combine with a noun to acquire familiarity readingwith the exclusion of uniqueness, thus English cannot have a proportional adjectival many.

References Alexiadou, Artemis, Liliane M. V. Haegeman, and Melita Stavrou. 2007. NounPhrase in the Generative Perspective. Berlin: Mouton de Gruyter; Aljovic, Nadira. 2002.Long adjectival inflection and specificity in Serbo-Croatian. In Recherches linguistiques deVincennes, volume 31, 27-42; Barwise, Jon, and Robin Cooper. 1981. Generalized quantifiersand natural language. Linguistics and Philosophy 4:159-219; Krasikova, Sveta. 2011. On pro-portional and cardinal ‘many’. Generative Grammar in Geneva 7:93-114; Kyriakaki, Maria.2011. DETs in the Functional Syntax of Greek Nominals. Doctoral Dissertation, University ofToronto; Milsark, Gary. 1977. Toward an explanation of certain peculiarities of the existentialconstruction in english. Linguistic Analysis 3:1-29; Partee, Barbara H. 1989. Many Quanti-fiers. ESCOL 89, 383-402; Pereltsvaig, Asya. 2006. Small Nominals. Natural Language andLinguistic Theory 24: 433-500.

Conflict resolution in the Spanish SLA of Yucatec ejectives: L1, L2 and universal constraints

Antonio A. González-Poot This study focuses on the Spanish second language (L2) acquisition of plain-ejective plosive (p-p’, t-t’, k-k’) and affricate (ts-ts’, tʃ-tʃ’) Yucatec contrasts, an interesting learning situation given the inactive status of constricted glottis ([CG]) – the contrasting feature – in the learners’ native (L1) grammar. It has been claimed (Brown 2000) that second language learners (L2ers) filter incoming perceptual input through their featural system: If a feature is inactive in the L1, it will preclude acquisition of new contrasts. It is shown here, however, that Spanish L2ers are able to accurately perceive target [CG]-based contrasts in Yucatec.

A review of the L2 perceptual literature strongly suggests that the first check that the interlanguage grammar makes when in the process of acquiring a new L2 contrast is whether or not the contrasting feature is active or inactive in the learner’s L1 grammar. If active, as predicted by Brown, acquisition of the new contrast will be given a head start. However, evidence from both L1 and L2 acquisition suggests that not all active target features are acquired at the same time: some features are acquired earlier than others. It is claimed here that this differential rate of acquisition is dependent upon 1) the intrinsic perceptual saliency of the feature, 2) the place/manner specifications of its segmental host, and 3) the syllabic position of this host. As shown by the result of the Spanish L2 acquisition of Yucatec ejectives, new L2 contrasts based on a feature inactive in the learner’s L1 grammar are still likely to be acquired, but they will also be subject to these 3 criteria.

These predicted differences in the perceptual acquisition of L2 contrasts are formalized here within an OT framework (Prince & Smolensky 2004 [1993]) via a set of perceptually and prosodically motivated constraints, introduced in the grammar in the form of both prominence hierarchies (e.g. IDENT[AirStreamMechanism] >> IDENT[Manner] >> IDENT[Place]), and faithfulness-based harmonic scales (Howe & Pulleyblank 2004), such as IDENT[CG]/ONSStop[Periph] >> IDENT[CG]/ONSStop[Cor]. It is claimed here that the interaction of harmonically aligned faithfulness constraints with structural restrictions (e.g. *CG) accounts for the attested perceptual patterns among Spanish L2ers. An OT analysis also allows illustrating the conflict of L1, L2 and universal constraints in the gradual processing of perceptual L2 input.

The analysis of the Spanish L1-Yucatec L2 learning situation constitutes a first within the field of studies in SLA. By providing new L1-L2 data, this study significantly contributes to a theory that greatly relies on cross-linguistic surveys in order to test the validity of hypotheses about Universal Grammar. References Brown, Cynthia A. (2000). The interrelation between speech perception and phonological acquisition: from infant to adult. In John Archibald (Ed.) Second Language Acquisition and Linguistic Theory (pp. 4-63). Malden, MA: Blackwell. Howe, Darin & Douglas Pulleyblank (2004). Harmonic Scales as faithfulness. Canadian Journal of Linguistics, 49, 1-49. Prince, Alan & Paul Smolensky (2004). Optimality Theory: Constraint Interaction in Generative Grammar. Malden, MA: Blackwell.

The role of animacy in sentence production: Evidence from French Margaret Grant1, Juliette Thuilier2,3, Benoît Crabbé4,5 and Anne Abeillé3,4 1McGill University, 2Univ. Rennes 2, 3LLF, 4Univ. Paris Diderot, 5ALPAGE

We present evidence from a sentence recall paradigm in French bearing on the role of conceptual information, specifically animacy, in sentence production. Previous studies have found that rates of inversion errors in production indicated a preference to produce more conceptually accessible DPs (on a scale such as that of Keenan and Comrie, 1977) earlier in a sentence. However, there is a debate as to the stage of sentence production at which accessibility plays a role. Models of sentence production have proposed separate stages for the assignment of lexical items to grammatical roles and for the linear ordering of constituents (see Bock and Levelt 1994). Previous studies have found evidence for an accessibility effect on grammatical assignment only (Bock and Warren 1985, Tanaka 2011) or an effect on both levels (see Branigan et al., 2007 for a review). We studied the recall of three sentence types: transitives with voice alternations [1], sentences with coordinated DPs [2], and ditransitives [3]. Modern French does not have an equivalent to the English dative alternation, however word order variation in ditransitives is attested; postverbal arguments of ditransitive verbs may appear in NP-‐PP order or PP-‐NP order, although NP-‐PP order is considered to be canonical (see Thuilier 2012 for corpus evidence). Preliminary results show that for transitives (N=30), speakers were more likely to produce inversions from passive to active when doing so would avoid having an inanimate argument in subject position (39% of otherwise correct recalls) than when both arguments were animate (11.4%, p < .01 in a mixed-‐effects logistic regression analysis). Active-‐to-‐passive inversions were very rare (only 2 instances) and were therefore excluded from analysis. No significant differences were found for inversions in coordinations (N=32) based on animacy. For ditransitives (N=33), the results did not show a preference for ordering animate arguments before inanimates. On the contrary, there was an interaction (p < .01) such that ditransitives with inanimate themes had the most inversions from PP-‐NP to NP-‐PP order (22%) and the lowest percentage of inversions from NP-‐PP to PP-‐NP order (4.5%), with animate argument in between (15% and 12%, respectively). Our results for transitives and coordinations support a role for animacy in grammatical role assignment but not for simple linear order. The ditransitive result, however, is not predicted by conceptual accessibility. Because of the free recall nature of the task, subjects were able to vary factors such as length and definiteness that were controlled in the presented items. We are in the process of coding the productions for these factors to gain insight into the nature of this surprising effect. Example sentence (animacy alternatives shown in {}, recall prompt, translation) 1 Au bout de la ruelle, {le voleur/le revolver} a été trouvé par le policier.

‘At the end of the alley, the thief/revolver was found by the policeman.’ 2 Ce jeune homme a toujours fui les traîtres et {les lâches/les échecs}

‘This young man has always avoided traitors and cowards/failures.’ Ce jeune homme a toujours fui {les lâches/les échecs} et les traîtres

‘This young man has always avoided cowards/failures and traitors.’ 3 Le chef de projet a confié {un agent commercial/un nouveau budget} à un décorateur. Le chef de projet a confié à un décorateur {un agent commercial/un nouveau budget}.

‘The project manager assigned a commercial agent/new budget to a decorator.’

The role of animacy in sentence production: Evidence from French Margaret Grant1, Juliette Thuilier2,3, Benoît Crabbé4,5 and Anne Abeillé3,4 1McGill University, 2Univ. Rennes 2, 3LLF, 4Univ. Paris Diderot, 5ALPAGE

References: Bock, J. K. and Warren, R. K. 1985. Conceptual accessibility and syntactic structure in

sentence formulation. Cognition 21:47–67.

Bock, J.K. and Levelt, W. 1994. Language Production: Grammatical Encoding. In the Handbook of Psycholinguistics, M.A. Gernsbacher (ed.). Academic Press.

Branigan, H. P., Pickering, M. J., & Tanaka, M. 2008. Contributions of animacy to grammatical function assignment and word order during production. Lingua, 118(2), 172-‐189.

Keenan, E. L. and Comrie, B. 1977. Noun phrase accessibility and universal grammar. Linguistic Inquiry 8(1): 63-‐99.

Tanaka, M. N., Branigan, H. P., McLean, J. F., and Pickering, M. J. 2011. Con-‐ ceptual

influences on word order and voice in sentence production: Evidence from Japanese. Jounal of Memory and Language, 65:318–330.

Thuilier, J. 2012. Contraintes préférentielles et ordre des mots en français. Thèse de doctorat en sciences du langage, Université Paris Diderot (Paris 7), Paris.

L’EMPLOI VARIABLE DU SUBJONCTIF ET DES EXPRESSIONS DE NÉCESSITÉ DANS LE FRANÇAIS PARLÉ DES LOCUTEURS RESTREINTS À PEMBROKE (ONTARIO)

Rick Grimm, Université York

En français, diverses constructions verbales régissent le mode subjonctif. Les études ayant examiné l’emploi du subjonctif en français parlé au Canada, et qui reposent sur des données tirées de corpus sociolinguistiques, ont trouvé qu’au moins deux tiers des occurrences du subjonctif suivent un seul verbe, soit le verbe impersonnel falloir + que (Comeau 2011, Grimm 2012, Poplack 1990, Poplack et al. 2013). Non seulement ce verbe est-il le plus abondamment utilisé dans la matrice, mais il amène aussi le subjonctif systématiquement ou presque dans la subordonnée. Donc, parmi tous les verbes susceptibles d’entraîner le subjonctif, c’est le verbe de nécessité falloir + que qui en fournit la majorité des occurrences analysables. La présente étude examine l’emploi variable du mode subjonctif, avec un accent particulier sur falloir + que, dans le français parlé à Pembroke (Ontario), localité où les francophones ne constituent qu’une faible minorité (6% des 13 000 habitants). Les données analysées proviennent d’un corpus d’entrevues sociolinguistiques (177 000 mots) menées auprès de 31 adolescents francophones inscrits dans une école de langue française. Il s’agit de locuteurs ‘restreints’ qui déclarent favoriser le français (versus l’anglais) comme langue de communication de 4% à 44% du temps. Bon nombre d’entre eux se servent peu ou jamais du français au foyer, ce qui fait de l’école l’un des seuls milieux où ils sont exposés à cette langue (voir Mougeon & Beniak 1991). Selon nos analyses quantitatives, à la fois le mode subjonctif et les constructions verbales pouvant le régir sont assez rares à Pembroke : il existe 42 cas du mode dans seulement 68 contextes possibles. Nous constatons que la maigre présence du subjonctif est en grande partie attribuable à un emploi marginal de falloir + que (N=19). Ce résultat a suscité une question importante : si ces locuteurs évitent falloir + que, quelle(s) expression(s) privilégient-ils pour exprimer la nécessité ? En pareil contexte, falloir + que (N=37, 14%) est plus ou moins évité au profit de structures suivies d’un complément infinitif, surtout devoir (N=141, 53%), et dans une moindre mesure falloir à titre de verbe personnel (N=34, 13%) (p. ex., je faut conduire), falloir + infinitif (N=16, 6%), et avoir à, être obligé de et avoir besoin de (N=39, 14%). La distribution des variantes à Pembroke diffère de manière marquée de celle calculée pour les communautés dont la population francophone est plus forte : falloir + que (58%–72%) représente l’expression de choix par rapport à devoir (1%–9%), qui conserve son statut de variante formelle (Lealess 2005, Thibault 1991). Il semble que l’emploi prépondérant de devoir dans le parler semi-informel des locuteurs à Pembroke puisse s’expliquer par l’input auquel ces derniers ont accès. Il ressort d’une analyse du discours de 13 enseignants (15 heures d’interactions enregistrées) que dans la salle de classe, c’est devoir qui prévaut (36,5%). Qui plus est, le logiciel Rbrul, lequel permet de traiter la restriction linguistique en tant que variable continue (vs. une catégorie fixe), révèle que le choix des variantes est corrélé avec le niveau relatif de restriction dans l’emploi du français. Par exemple, plus un locuteur se sert du français, plus il utilise falloir + que; inversement, moins un locuteur se sert du français, plus il utilise devoir. En somme, notre étude démontre, d’une part, comment le comportement d’une variable sociolinguistique peut influer sur celui d’une autre et, d’autre part, qu’il est possible de déceler des corrélations statistiquement significatives à

l’intérieur d’un seul groupe de restriction linguistique lorsque l’effet de ce facteur social est mesuré à l’aide d’un continuum plus étendu. OEUVRES CITÉES Comeau, P. (2011). A window on the past, a move toward the future: sociolinguistic and formal

perspectives on variation in Acadian French. Thèse de doctorat inédite, Université York, Toronto.

Grimm, D. R. (2012). L’emploi variable du mode subjonctif dans le français parlé à Hawkesbury (Ontario). Communication livrée au colloque Les français d’ici, Université de Sherbrooke, juin.

Lealess, A. (2005). En français, il faut qu’on parle bien: Assessing native-like proficiency in L2 French. Mémoire de maîtrise inédit, Université d’Ottawa.

Mougeon, R., & Beniak, É. (1991). Linguistic consequences of language contact and restriction: The case of French in Ontario. Oxford: Clarendon Press.

Poplack, S. (1990). Prescription, intuition et usage: le subjonctif français et la variabilité inhérente. Langage et société, 54, 5–33.

Poplack, S., Lealess, A., & Dion, N. (2013). The evolving grammar of the French subjunctive. Probus, 25(1), 139–195.

Thibault, P. (1991). Semantic overlaps of French modal expressions. Language Variation and Change, 3, 191–222.

On substance in phonology Daniel Currie Hall, Saint Mary’s UniversityWhile ‘substance-ee’ phonology owes its name to Hale & Reiss (2000), the idea echoes Fudge’s

(1967: 26) proposal that phonologists “ought to burn their phonetic boats and turn to a genuinelyabstract amework.” One motivation for this is the desire to avoid the redundant formal encoding ofphysiological facts: Hale & Reiss (2000) argue that no insight is to be gained by positing phoneticallymotivated universal markedness constraints, and Mielke (2008) argues against attributing to UG aninventory of features that could be derived om the properties of the human vocal and auditory ap-paratus. However, theories in which phonetic substance is altogether banished om phonology canend up looking surprisingly similar to phonetically based theories in the explanations they posit forphonological patterns that are phonetically ‘natural.’ If phonology is oblivious to phonetic content,then the fact that many phonological patterns are natural must be attributed to phonetics itself. In thecase of substance-ee theories, the influence of phonetics can exert itself only through acquisition anddiachrony, rather than through phonetically based synchronic rules or constraints, but if phonologyis “a genuinely abstract amework,” much of its explanatory burden must be shied to phonetics.Is anything lost in this transfer? This paper argues that something is lost; that it can be regained

through the moderate use of phonetic substance in phonology; and that the banishment of substancehas been based in part on unwarranted assumptions about the rigidity of phonological representations.Mielke’s case for emergent features draws support om the existence of phonological patterns involv-

ing unnatural classes of sounds. If phonology is “a genuinely abstract amework,” it offers little reasonfor skepticism about such patterns. They may arise diachronically through uncommon combinationsof phonetically natural changes, but the synchronic learner can easily represent them. However, Hall(2010) and Godey (2012) show that several of the ‘unnatural’ patterns reported by Mielke are subjectto reanalysis either as natural or as combinations of natural patterns with independent motivation. Forexample, what Mielke treats as deletion of nasals before the unnatural class of nasals and icatives (tothe exclusion of obstruent stops) in Bukusu, Hall (2010) analyzes as independently motivated patternsof nasal effacement before icatives, nasal place assimilation, and a systematic ban on geminates. Thereis, then, at least a methodological case for pursuing a theory that forces one to look for naturalness.The proponents of substance-ee approaches are correct in observing that the phonetic properties

of phonemes do not dictate their phonological behaviour. But there is a way to curtail the role ofsubstance without eliminating it altogether. Contrastive specification based on a cross-linguisticallyvariable hierarchy of features, as proposed by Dresher (2009), offers a principled explanation for the factthat phonemes with a particular phonetic property are sometimes ignored by phonological processesthat refer to the corresponding feature. Consider an example om Mackenzie (2013). A three-waycontrast among implosives and voiced and voiceless plosives may be encoded by either of two hierar-chical orderings of [voice] and [constricted glottis]. If [c.g.] takes wider scope, it distinguishes theimplosives, and [voice] is relevant only for the plosives; if [voice] takes wider scope, it distinguishes thevoiceless plosives, and [c.g.] is relevant only for voiced stops. As Mackenzie shows, both possibilitiesare attested. Ngizim [voice] harmony requires agreement between plosives, but ignores implosives.Hausa [c.g.] harmony requires agreement between (homorganic) voiced stops, but ignores the voice-less plosives. Under this view, the task of the learner in acquiring phonological representations is toset up a system of features that is just sufficient to differentiate the phonemic inventory and that allowsfor the encoding of observed patterns. If the features themselves must be phonetically interpretable,then the learner’s job is simplified, and the analyst’s hypothesis space is constrained. Representationsare substantive enough to make ‘natural’ patterns the norm, but also abstract enough to account forthe fact that phonetics does not determine phonological destiny.

References

Dresher, B. Elan. 200⒐ The contrastive hierarchy in phonology. Cambridge: Cambridge UniversityPress.

Fudge, Erik C. 196⒎ The nature of phonological primes. Journal of Linguistics ⒊ 1–3⒍Godey, Ross. 20⒓ “Crazy” classes are not crazy: An argument against Mielke 200⒏ Ms., Universityof Toronto.

Hale, Mark & Charles Reiss. 2000. ‘Substance abuse’ and ‘dysfunctionalism’: Current trends inphonology. Linguistic Inquiry 31⑴. 157–16⒐

Hall, Daniel Currie. 20⒑ Probing the unnatural. In Jacqueline van Kampen & Rick Nouwen (eds.),Linguistics in the Netherlands 2010, 71–8⒊ Amsterdam: John Beǌamins.

Mackenzie, Sara. 20⒔ Laryngeal co-occurrence restrictions in Aymara: Contrastive representationsand constraint interaction. Phonology 30⑵. 297–34⒌

Mielke, Jeff. 200⒏ The emergence of distinctive features. Oxford: Oxford University Press.

Quantifying Perceived Morphological Relatedness Kathleen Currie Hall, Claire Allen, Tess Fairburn, Kevin McMullin, Michael Fry, Masaki Noguchi

A key tool in a phonologist’s toolkit is the use of morphophonological alternations, i.e., cases where a single morpheme has multiple phonological representations, varying by context. Most predictable phonological processes are known through their application in alternations. Whether a single morpheme is undergoing alternation is in many cases fairly trivial; e.g., inflections of a single verb provide easily interpretable data. But there are other cases in which determining whether two words are morphologically related is not so easy: does the morpheme face occur in face-lift? facial? surface? Etymologically speaking, the answer may be yes, but this does not necessarily reflect the morphological awareness of speakers. And yet, the recognition of such relatedness is crucial to the identification of alternations and thus to phonological processes. Much of the morphological literature has focused on processing models (e.g., Stockall & Marantz 2006) and/or mostly dealt with inflection (e.g., Ackerman et al. 2009). In this paper, we present a quantification that does not attempt to isolate individual morphemes but rather gives an overview of perceived word similarity. An AXB discrimination task was used to elicit similarity judgments from English speakers. 180 individual key words (base morphemes, e.g, press) were matched with each of nine different words of the following types (given with examples): (1) inflection (pressed), (2) relatively transparent derivation (pressure), (3) relatively opaque derivation (expressway), (4) primary semantic sense (push), (5) secondary semantic sense (media), (6) rhyme (mess), (7) cohort (preppy), (8-9) unrelated words (sofa, table). For any given key word, two of these nine words were presented in orthographically; participants indicated which was “more similar” to the key. No instructions on what defines similarity were given. All combinations of all pairs were presented, randomly, with a single participant seeing only two samples of any given key word in 360 total trials. Results here are based on 26 pilot participants. For each type of comparison word, the percentage of the time that that type was chosen when it was an option is given in the table below. Morphologically related forms are indeed interpreted as being most similar to the key, followed by semantically and then phonetically related forms. The behavioural (AXB) results were also correlated with similarity scores for spelling and pronunciation (cf. Khorsi 2012) and semantic relatedness (cf. Han et al. 2013), as shown below; R2 values for the three are 0.31, 0.22, 0.30, respectively. Given, however, that morphological relatedness should really reflect a combination of sound and meaning relatedness, a logistic mixed-effects regression model of the behavioural results was developed. The model has a classification accuracy of 75% and a McFadden’s R2 value of 0.49. It crucially relies on spelling, pronunciation, and semantics as factors, suggesting that all three are relevant for predicting perceived morphological relatedness. This model can thus be used to independently approximate the degree to which any two words are morphologically related in the minds of speakers, which in turn can be used to determine the extent to which phonological processes that depend on alternations between morphologically related words are in fact psychologically viable.

0.00

0.25

0.50

0.75

1.00

−50 −25 0 25Spelling Relatedness Score

AXB

Rel

ated

ness

Sco

re

0.00

0.25

0.50

0.75

1.00

−60 −40 −20 0 20Transcription Relatedness Score

AXB

Rel

ated

ness

Sco

re

0.00

0.25

0.50

0.75

1.00

0.00 0.25 0.50 0.75 1.00Semantic Relatedness Score

AXB

Rel

ated

ness

Sco

re

Relation Percent Inflected 0.83 Derived1 0.74 Derived2 0.65 Meaning1 0.60 Meaning2 0.51 Rhyme 0.48 Cohort 0.38 Unrelated 0.17

References: Ackerman, F., Blevins, J. P., & Malouf, R. (2009). Parts and wholes: Patterns of relatedness in

complex morphological systems and why they matter. In J. P. Blevins & J. Blevins (Eds.), Analogy in grammar: Form and acquisition (pp. 54-82). Oxford: Oxford University Press.

Han, L., Kashyap, A., Finin, T., Mayfield, J., & Weese, J. (2013). UMBC_EBIQUITY-CORE: Semantic Textual Similarity Systems Proceedings of the Second Joint Conference on Lexical and Computational Semantics.

Khorsi, A. (2012). On Morphological Relatedness. Natural Language Engineering, 1-19. Stockall, L., & Marantz, A. (2006). A single route, full decomposition model of morphological

complexity. The Mental Lexicon, 1(1), 85-123.

Les sujets des verbes météorologiques dans les dialectes occitans transitionnels du Nord David Heap (UWO), Michèle Oliviéri et Jean-Pierre Lai (Université de Nice)

Alors que la plupart des langues romanes sont des systèmes « à sujet nul », certains dialectes ont développé des paradigmes complets ou partiels de clitiques nominatifs. Entre les grammaires avec un paradigme complet de clitiques sujets (français standard) et le cas non-marqué des parlers qui n’en ont pas (espagnol, italien, la plupart des dialectes occitans), se trouve un continuum de variétés transitionnelles dont les paradigmes présentent entre un et cinq clitiques nominatifs. Dans ces langues « partiellement à sujet nul », les verbes météorologiques requièrent typiquement soit l'absence de pronom sujet, soit la présence d'un clitique sujet explétif qui est le plus souvent formellement identique à celui de la troisième personne du masculin singulier (comme en français standard ou en rhéto-roman). Cependant, les dialectes occitans du nord de l'aire (les NODs), qui se trouvent dans une zone de transition dans le centre de la France, rompent avec ce choix binaire : ils présentent des pronoms sujets avec des verbes météorologiques, qui sont de la troisième personne du singulier mais avec des formes complètement différentes (kɔ ou ka, comme dans (1ab)) des pronoms référentiels du masculin singulier (1c). En outre, les pronoms sujets dans les NODs font preuve d’autres particularités. Dans les dialectes de l’Italie du Nord (NIDs, mieux documentés), les verbes météorologiques sont parmi les derniers à présenter des sujets réalisés, qui apparaissent après les autres pronoms de la troisième personne (P3, cf. Oliviéri 2011). Des enquêtes récentes en Corrèze, en Creuse et en Dordogne four-nissent de nouvelles données qui montrent que dans cette aire, le sujet des verbes météorologiques peut parfois être le premier et seul clitique du paradigme, comme dans (2). Dans le sud de la région considérée (en Corrèze notamment), ce ko apparaît de manière variable dans les constructions météorologiques, et avec moins de fréquence dans certains contextes (Kaiser et al. 2013). Plus au nord (en Creuse), ce phénomène s'étend à d’autres constructions impersonnelles, mais toujours avec une forme distincte des tous les autres clitiques nominatifs, comme on le voit en (3), où la seule forme verbale sans pronom sujet réalisé est le verbe impersonnel équivalent de « falloir ». Ce même élément apparaît aussi dans d’autres contextes où il n'est manifestement pas clitique, comparable au ça du français, comme sujet (4a) ou objet (4b). La structure précise des différents paradigmes qui apparaissent dans ces NODs est d’une importance considérable pour la typologie morpho-syntaxique. Bien que l'apparition de ces éléments ne soit pas contrainte au point de ne suivre qu'un seul chemin linéaire (de 0 à 6), le nombre de paradigmes attestés est beaucoup plus réduit que ce qu'on pourrait prédire si la variation entre les personnes grammaticales était aléatoire. La forme de ces paradigmes partiels correspond moins à l’ordre dans lequel apparaissent les pronoms qu'à l’ordre d’apparition des contrastes entre les pronoms (Heap & Oliviéri 2013). Les différentes progressions observées dans l'émergence des clitiques nominatifs des NODs fournissent des indices sur la structure interne des contrastes entre les traits morphologiques, d’après une analyse sous-spécifiée inspirée de Harley & Ritter (2002). Notre modèle de l’introduction progressive des contrastes suppose des traits monovalents, organisés hiérarchiquement pour refléter des dépendances implicationnelles. Si dans les NIDs (Poletto 2000, Tortora 2002, Manzini & Savoia 2005), le clitique de P2 apparaît en premier, dans les NODs, cela peut également être ceux de P3 ou de P1. Cela indique que le premier contraste implique le trait Participant (soit P2 ou P1) qui s’oppose au trait Individu (P3 et P6), mais cela n'explique pas les faits observés pour les constructions météorologiques des NODs. Cependant, dans un système qui permet la sous-spécification variable, il peut y avoir une option encore moins spécifiée : le clitique « nu », qui n’est ni Participant ni Individu. Nous proposons que cette configuration soit celle des sujets météorologiques comme ko, précisément parce qu’elle est dénuée de tout trait morphologique et peut donc représenter des éléments non référentiels. Cette perspective nous permet alors de préciser le statut de ces éléments, que l'on peut considérer comme des quasi-arguments (cf. aussi Kaiser et al. 2013), dans le sens de Chomsky (1981).

Exemples

(1) a. [ka pl'øw], [kɔ pl'øj] b. [ka gr'ɛlø], [kɔ gʁ'ɛlə] c. [ɔ e fatiɡˈa] ‘il pleut’ ‘il grêle’ ‘il est fatigué’

(2) Paradigmes verbaux en Corrèze Treignac ˈvɔw ˈva 'vaj aˈnã aˈna ˈvɔ ka ˈpløw Le Vedeix ˈvɔw ˈva 'vaj aˈnɛ aˈna ˈvɔ kɔ ˈpløj

‘aller’ P1 P2 P3 P4 P5 P6 ‘il pleut’ P3 (3) Émergence progressive de ko Corrèze Creuse Treignac Faux-Mazurat Royère-de-Vassivière ka pløu ka plo kɔ mɔʎɔ ‘il pleut’ faj fʁɛ ka fe frɛ kɔ fej fʀe ‘il fait froid’ fɛ nɥɛ ka fɛ nœʲ kɔ faj nɛ ‘il fait nuit’ mø ʃɑblə me sẽblᵊ kɔ me semblɔ ‘il me semble’ sɔu fo fo ‘il faut’ (4) Le Vedeix a. [kɔ vɔ la pɛnø] ‘ça vaut la peine’, [kɔ to plɛi] ‘ça te plaît?’, [kɔ ʃɛ bu] ‘ça sent bon’

b. [j a di kɔ] ‘tu lui as dit’, [ʃe a vi kɔ] ‘si vous aviez vu ça’, [pyʁki fa kɔ] ‘pourquoi tu fais ça?’ Références

Chomsky, N. (1981), Lectures on Government and Binding. The Pisa Lectures, Dordrecht, Foris. Harley, H. & E. Ritter (2002), "Person and number in pronouns: A feature-geometric analysis",

Language 78, 482–526. Heap, D. (2002), "Split subject pronoun paradigms : Feature geometry and underspecification" In T.

Satterfield et al. (eds.) Current Issues in Romance Languages : Selected Papers from the 29th Linguistic Symposium on Romance Languages (LSRL), Ann Arbor, 8-11 April 1999, Amsterdam-Philadelphia, John Benjamins, 129-144.

Heap, D. & M. Oliviéri (2013), "On the emergence of nominative clitics in Romance dialects", paper presented at Workshop European Dialect Syntax VII, Konstanz, 13-15 juin 2013.

Kaiser, G.A., Oliviéri, M. & Palasis, K. (2013). "Impersonal constructions in northern Occitan", In X.A. Álvarez Perez, E. Carrilho, and C. Magro (eds.) Current Approaches to Limits and Areas in Dialectology. Cambridge: Cambridge Scholars Press/CSP, 345-366.

Manzini, M.R. & L.M. Savoia (2005), I dialetti italiani e romanci: morfosintassi generativa, Alessandria, Edizioni dell'Orso.

Oliviéri, M. (2011), "Typology or Reconstruction: the Benefits of Dialectology for Diachronic Analysis" In J. Berns et al. (eds.) Romance Languages and Linguistic Theory 2009. Selected papers from 'Going Romance' Nice 2009, Amsterdam, John Benjamins, 239-253.

Poletto, C. (2000), The Higher Functional Field : Evidence from Northern Italian Dialects, New York & Oxford, Oxford University Press.

Tortora, C.M. (ed.) (2002), The Syntax of Italian Dialects, Oxford, Oxford University Press.

Aspiration et effacement du /-s/ en espagnol d’Holguín (Cuba) : variables phonologiques et leur conditionnement.

David Heap, Jorge Emilio Rosés Labrada, Jeff Tennant, Angélica Hernández (UWO)

En espagnol, la fricative sifflante sourde /s/ peut apparaître dans deux positions syllabiques différentes : en attaque et en coda (Hualde 2005:74). En position de coda, ce phonème /s/ peut être interne au mot (position médiane) ou bien en finale absolue (frescos “frais” masc. pl.) et il subit ou bien l’aspiration (/s/ -> [h]/__$) ou bien l’effacement (/s/ -> [Ø]/__$) (Lipski 1999:198) dans plusieurs variétés de l’espagnol. Comme l’observe Hualde (2005:28), tous les locuteurs des variétés antillaises de l’espagnol “aspirate preconsonantal and word-final /s/ to a certain extent [...], [therefore] this is a highly variable process.” Cette étude se concentre sur l’aspiration et l’effacement de /-s/ dans un corpus de l’espagnol cubain de la ville et la province d’Holguín, dans un cadre variationniste labovien (Labov 1972). L’importance de cette étude réside en partie en la nature du corpus. D’autres études de l’aspiration et de l’effacement de /-s/ en espagnol cubain (Terrel 1977, 1979 ; Lynch 2009 ; entre autres) se basent sur la langue d’émigrants cubains établis à Miami. Notre corpus, en contraste, inclut uniquement des locuteurs et des locutrices cubains—spécifiquement d’Holguín) —qui ont vécu toute leur vie dans le pays et qui représentent donc plus précisément le parler actuel de cette région de Cuba. Notre étude se base sur des échantillons de trois styles (liste de mots, passage de lecture, conversation) pour 10 locuteurs. Ces passages sont d’abord transcrits orthographiquement par le biais du logiciel Praat afin de faciliter le codage ainsi que l’analyse acoustique de cette variable (s). Nous nous concentrons dans cette étude sur les variables phonologiques qui conditionnent ces deux phenomènes, à savoir la position syllabique, les pauses, l’accent tonique t la longueur du mot, ainsi que les traits phonologiques des segnments qui suivent : par exemple [± vocalique], [± continuant] et [± voix]. Notre analyse démontre que la variable dépendante (s) est conditionnée par sa position dans le mot, par la longueur du mot, et par l’accent. De façon significative, on voit aussi que le trait [±voix] du segment suivant joue aussi un rôle important : les consonnes voisées défavorisent le maintien du /-s/. Le style et le niveau d’éducation des locuteurs semble aussi représenter des variables significatives.

Références

Hualde, José Ignacio. 2005. The sounds of Spanish. New York: Cambridge University Press. Labov, William. 1972. Sociolinguistic patterns. Philadelphia: University of Pennsylvania Press.

Lipski, John M. 1999. The many faces of Spanish /s/-weakening: (re)alignment and ambisyllabicity. In Advances in Hispanic linguistics. Papers from the 2nd Hispanic Linguistics Symposium, Vol I., ed. Javier Gutiérrez-Rexach and Fernando Martínez-Gil, 198-213. Somerville, Mass.: Cascadilla Press. Lynch, Andrew. 2009. A sociolinguistic analysis of final /s/ in Miami Cuban Spanish. Language Sciences 31(6): 766-790. Terrell, Tracy D. 1977. La aspiración y elisión en el español cubano: implicaciones para una teoría fonológica dialectal. In Estudios sobre el español hablado en las principales ciudades de América, ed. Juan M. Lope Blanch, 39-48. México: Universidad Nacional Autónoma de México.

Terrell, Tracy D. 1979. Final /s/ in Cuban Spanish. Hispania 62(4): 599-612.

How to Do Things with Particles Johannes Heim, Hermann Keupdjio, Zoe Wai-Man Lam, Adriana Osa-Gómez and Martina Wiltschko

University of British Columbia

According to classic definitions of speech acts, (1) S asserts p if i) S believes p and ii) S wants A to believe p (Bach & Harnish 1979). This view forms the basis for current assumptions regarding the syntax of speech acts. For

example Speas & Tenny (2003) propose a complex speech act phrase (SAP; see also Ross 1970).

(2) [SAP S [gives [SAP p [to A]]]].

According to this view, a declarative clause maps onto the assertion speech act. If this was the

case, we would expect that declaratives are only felicitous if the two conditions above are met.

This is however not the case. i) A declarative may be uttered by a speaker who does not believe p,

namely in the form of a declarative with rising intonation (Gunlogson 2003). ii) A declarative

may be uttered even if S knows perfectly well that A already knows p. This takes the form of a

declarative with surprise intonation. (3) You have a new dog. This demonstrates that intonation may serve as a speech act modifier. In this talk we explore the

formal and functional properties of speech act modification. In particular, we explore two types

of speech act modifiers: intonation and sentence peripheral particles such as those in (4): (4) Du host an neichn Hund goi? Goi Du host an neichn Hund ? Prt you have a new dog The existence of speech act modifiers such as intonation and sentence peripheral particles, we

argue, requires a more complex and nuanced understanding of speech acts. And indeed,

according to current semantic analyses it is essential to take into consideration the fact that S

doesn’t necessarily aim to change A’s Set of Beliefs (SoB) but instead propositions are tabled

and are required to be accepted by A before they are considered to be in the common ground

(CG; Stalnaker 2002). We propose that speech act modifiers are explicit linguistic means to

encode this tabling function. In particular, we explore speech act modifiers in two non-tone

languages (German and Spanish) as well as in two tonal languages (Medumba and Cantonese).

We observe that in tonal languages, a particle encodes the orientation of the speech act, whereas

in non-tonal languages the intonational contours perform this function. Based on the interaction

of particles with intonational contours, we argue that intonational contours are best analysed as

intonational morphemes/tunes (Truckenbrodt 2012, Pierrehumbert & Hirschberg 1990). We further propose that complex speech acts such as in (3-4) are indicative for another

layer above SAP which introduces the S’s representation of the A’s SoB. We refer to this layer

as GroundP. (5) [GroundP SoB(A) [Ground SAP]]

Evidence for the syntactic representation of A-orientation comes from the properties of speech

act modifiers. They may attach to the different layers in (5) resulting in systematic semantic

differences. Moreover in German dialects they agree with A depending on whether S addresses

A informally (6a) or formally (6b). In some dialects we also find number agreement (6c). (6) a. Es regnt goi? b. Es regnt goi-ns? c. % Es regnt, goi-ts

It rains prt.2inf It rains prt-2form It rains prt-2pl Finally, we show that both the linear order of speech act modifiers as well as their scopal

properties lend support to the structure in (5).

References Bach, K., & Harnish, R. M. (1979). Linguistic communication and speech acts (Vol. 4, pp. 624-

648). Cambridge, MA: MIT press. Gunlogson, C. (2013). True to form: Rising and falling declaratives as questions in English.

Routledge. Pierrehumbert, J. (1990). The Meaning of Intonational Contours in the Interpretation of

Discourse Janet Pierrehumbert and Julia Hirschberg. Intentions in communications, 271.

Ross, John R. (1970). On Declarative Sentences. In Readings in English Transformational

Grammar, ed. R. A. Jacobs and Peter S. Rosenbaum. Waltham, Mass.: Ginn.

Speas, M. & Tenny, C. (2003). Configurational properties of point of view roles. Asymmetry in

grammar, 1, 315-343. Stalnaker, R. (2002). Common ground. Linguistics and Philosophy, 25(5), 701-721. Truckenbrodt, H. (2012). The interface of semantics with phonology and morphology. In

Maienborn, von Heusinger and Portner (eds.) Semantics. De Gruyter, 2039-2069.

Imagining Events: Influences of Temporal Information in Verbs and Perspective Taking Jeffrey P. Hong Todd R. Ferretti Deanna Hall Wilfrid Laurier University

Recent psycholinguistic research shows that grammatical aspect (GA; imperfective, perfective) interacts with lexical aspect (LA; activities, accomplishments) to influence language comprehension difficulty (Yap et al., 2009; Becker et al., 2013). The purpose of the present research is to examine how these variables interact to influence the ease in which people imagine events described in simple phrases. A second goal is to examine how taking a first person perspective (as if performing the event) or third person perspective (looking at self performing event) influences the ability to imagine events. This research also employed event-related brain potential methodology to examine slow cortical brain potentials. These brain potentials are known to be sensitive to the ease of integrating text into developing situation models (King & Kutas, 1995), and to the ease of imagining events (Conway et al., 2003).

In the first experiment, participants read and imagined sentences that contained either accomplishments (build) or activities (act) that were grammatically marked as ongoing or completed (I was acting/I acted). Our slow potential results show that participants had less difficulty imagining events when the temporal properties of the two forms of verb aspect matched (imperfective activities, perfective accomplishments) versus mismatched (perfective activities, imperfective accomplishments). Furthermore, participants reported they imagined the events more often from a first person perspective for activities than accomplishments. GA had the strongest influence on imagined accomplishments; the first-person perspective was used more often for perfective than imperfective accomplishments.

In the second experiment, participants were told to take either a first or third person perspective when imagining events described in imperfective activity phrases (e.g., I was exercising). The slow potential findings demonstrated that SCP amplitudes were more negative for third-person than for first-person event representation, in left-frontal regions. This difference is taken to indicate that greater cognitive effort is required for the generation of imagined events from the third-person perspective, as compared to the first-person perspective.

This research provides novel neurocognitive and behavioural insight into how event representation is influenced by temporal information associated with verbs and the perspective from which an event is represented.

References

Becker, R. B., Ferretti, T. R., & Madden-Lombardi, C. J. (2013). Grammatical aspect, lexical

aspect, and event duration constrain the availability of events in narratives. Cognition, 129(2), 212–220.

Conway, M. A., Pleydell-Pearce, C. W., Whitecross, S. E., & Sharpe, H. (2003). Neurophysiological correlates of memory for experienced and imagined events. Neuropsychologia, 41(3), 334–340.

King, J. W., & Kutas, M. (1995). Who Did What and When? Using Word- and Clause-Level ERPs to Monitor Working Memory Usage in Reading. Journal of Cognitive Neuroscience, 7(3), 376–395.

Yap, F. H., Chu, P. C. K., Yiu, E. S. M., Wong, S. F., Kwan, S. W. M., Matthews, S., et al. (2009). Aspectual asymmetries in the mental representation of events: Role of lexical and grammatical aspect. Memory & Cognition, 37(5), 587–595.

Différences rythmiques dans les styles de parole en français ontarien

Svetlana Kaminskaïa

Université de Waterloo Les mesures rythmiques (Ramus et al. 1999, Dellwo 2006, White and Mattys 2007, Low et al. 2000, Grabe and Low 2002) appliquées à des langues différentes ont permis d’explorer largement la variation typologique et régionale. La présente analyse s’adresse à l’effet du style sur le rythme prosodique en français parlé en Ontario en situation minoritaire (corpus Windsor, PFC, Durand et al. 2002). Les lectures du texte sont comparées aux extraits de la parole spontanée produits par les mêmes locuteurs.

Notre analyse antérieure des lectures du texte a démontré que malgré la situation du contact linguistique, l’ensemble des données se caractérise par une rythmicité syllabique propre à la langue française et que la convergence avec le rythme accentuel anglais n’a pas vraiment lieu. Cependant, dans ce type de production les locuteurs tendent vers une prononciation plus normative (Simon 2003, Hambye 2008), alors que la parole spontanée démontre une prosodie plus régionale (Carton 1984). Cela permet de supposer que dans les entrevues libres le rythme peut ressortir affecté par le contact avec l’anglais, donc plus accentuel.

La parole spontanée et la lecture du texte se caractérisent normalement par des différences de vitesse d’articulation. Dans le contexte du contact linguistique où la lecture en français n’est pratiquée que dans le contexte d’école, on suppose une production plus rapide en parole spontanée. Le débit plus rapide amène typiquement à une moindre variabilité des intervalles et donc une rythmicité plus syllabique, ce qui entre en conflit avec l’hypothèse formulée tout à l’heure. Ici, nous explorons donc l’effet du style sur le débit et sur le rythme en français en situation minoritaire.

Pour l’analyse, nous appliquons les méthodes suivantes de l’analyse métrique : le modèle CCI (« Control and Compensation Index », Bertinetto et Bertini 2008) sensible aux changements du débit, ainsi que les méthodes normalisant le débit qui se sont démontrées les plus performantes dans les études antérieures (VarcoV, nPVI-V et %V). Les résultats démontrent que tous les locuteurs ont le débit significativement plus rapide en parole spontanée que lors de la lecture, et que malgré cela la rythmicité dans ce sous-corpus se révèle plus accentuelle, car les valeurs VarcoV, nPVI-V et CCI démontrent plus de variabilité des intervalles. En même temps, les résultats pour %V suggèrent le contraire.

Cette analyse permet de conclure que pour une langue en contact les observations antérieures sur le rapport entre le changement du débit et la rythmicité ne s’appliquent pas et que les caractéristiques de la prononciation locale doivent être prises en considération, aussi bien que la situation du contact linguistique.

References Bertinetto Pier Marco & Chiara Bertini 2008. On modeling the rhythm of natural languages.

Proceedings of the 4th International Conference on Speech Prosody, Campinas 2008, 427‐430.

Bertinetto Pier Marco & Chiara Bertini 2010. Towards a unified predictive model of natural language rhythm. In Russo Michela (cur.), Prosodic Universals. Comparative studies in rhythmic modeling and rhythm typology. Napoli, Aracne: 43-77.

Durand Jacques, Bernard Laks & Chantal Lyche 2002. Synopsis du projet PFC, la Phonologie du Français Contemporain : Usages, Variétés et Structure. Bulletin PFC #1: Protocole, conventions et directions d’analyse. 5-7.

Grabe Esther & Ee Ling Low. 2002. Durational variability in speech and the rhythm class hypothesis. In Warner Natasha & Carlos Gussenhoven (eds.) Papers in Laboratory Phonology 7. Berlin: Mouton de Gruyter. 515-546.

Hambye, Philippe 2008. Convergences et divergence: quelques observations sur la standardisation du français en Belgique. Standardisation et déstandardisation: le français et l’espagnol au XXe siècle. Éd. Jürgen Erfurt et Gabriele Budach. Frankfurt: Peter Lang. 35–62.

Low Ee Ling, Esther Grabe & Francis Nolan 2000. Quantitative characterizations of speech rhythm: Syllable-timing in Singapore English. Language and Speech 43(4). 377-401.

Pike Kenneth 1945. The Intonation of American English. Ann Arbor: University of Michigan Press.

Ramus Franck, Marina Nespor & Jacques Mehler 1999. Correlates of linguistic rhythm in the speech signal. Cognition 73(3). 265-292.

Simon, Anne-Catherine 2003. Analyse de la variation prosodique du français dans les données conversationnelles: propositions théoriques et méthodologiques. Bulletin Phonologie du Français Contemporain 3, 99–13.

White Laurence & Sven L. Mattys 2007. Calibrating rhythm: First and second language studies. Journal of Phonetics 35. 501-522.

On propositional complexity of middles and its consequences for syntax Olena Kit (McMaster University)

Though linguistic properties of middle constructions, as in (1), have been extensively studied, the syntactic structure remains debated. Recent work argues that middles are structurally identical to passives (Stroik, 1992; Hoekstra and Roberts, 1993; Medová, 2009), unergatives (Fagan, 1992; Ackema and Schoorlemmer, 1994), or reflexives (Massam, 1992; Steinbach, 2002). In this talk, I provide evidence that middle constructions are unlike passives, unergatives, and reflexives in that they form propositionally complex structures. I argue that even though the structure of middles is transitive-like, the structural complexity is not a result of two arguments being merged in the structure. Instead, the two-proposition structure arises as a result of the internal argument undergoing an operator triggered A’-movement, followed by A-movement. (1) a. The book reads easily. b. The bread cuts easily. Puzzle: Bale (2007) observes that there is an interesting correlation between the propositional complexity of a VP and its internal structure. Specifically, non-stative, transitive verbs allow again to introduce presuppositions that do not involve the verb’s subject, whereas intransitive verbs and stative, transitive verbs must include the subject. We can use this observation as a diagnostic for the structure of middles. An example of this paradigm is in (2): (2) Context: John drove the car quickly, but then he sold it. Assuming Mary buys the car for

its rapid acceleration, it is safe to say that… a. Mary drives the car quickly again. c. The car drives quickly again. b. #Again Mary drives the car quickly. d. Again the car drives quickly.

Middle sentences, (1c) and (1d), do not pattern with intransitive or stative, transitive verbs, instead they share propositional properties with non-stative transitive verbs. This finding suggests middles are propositionally complex. In other words, the understood subject is not a functional argument of the verb and the understood object and verb form one propositional level. This evidence is further supported by reconstruction diagnostics, namely the Coordinate Structure Constraint (Fox 2000) and the Epistemic Containment Principle (von Fintel and Iatridou, 2003), where middles and active transitives pattern alike. Consider the examples in (3): (3) a. John must boil the noodles and a chef stir the sauce. Active b. *Eggs must have been cracked and a vegetable have been chopped. Passive c. Cakes must cut easily and flour mix without difficulty. Middle Since middles seem to be transitive-like in their propositional properties, the question that arises is: What causes this behaviour? As the following examples in (4) show, this structurally complex behaviour cannot result from having a covert external argument in the structure as there is a ban on agent-oriented adverbs (4a) and agentive by-phrases (4b) in middles constructions. (4) a. *The bread cuts carefully. b. *The bread cuts easily by the chef. Therefore, middle constructions are propositionally complex structures without an agent present in syntax. This raises the question of whether an alternative approach can account for the syntactic behaviour of middles. Proposal: I propose that the internal argument is extracted using a two-step derivation, similar to Collins’s (2005) smuggling strategy. Specifically, an operator triggers the internal argument to move out of its functional projection (A’-movement). The operator introduces a generic interpretation with modal properties, which are necessary for the dispositional reading (Lekakou, 2005). After the initial step, the internal argument moves to the subject position (A-movement).

References

Ackema, P., & Schoorlemmer, M. (1994). The middle construction and the syntax-semantics interface. Lingua, 93(1), 59-90.

Bale, A. C. (2007). Quantifiers and verb phrases: An exploration of propositional complexity. Natural Language & Linguistic Theory, 25(3), 447-483.

Collins, C. (2005). A smuggling approach to the passive in English. Syntax, 8(2), 81-120. Fagan, S. M. (1992). The syntax and semantics of middle constructions: A study with special

reference to German (Vol. 60). Cambridge University Press. Fox, D. (2000). Economy and semantic interpretation (Vol. 35). The MIT Press. Hoekstra, T., & Roberts, I. (1993). Middle constructions in Dutch and English. Knowledge and

language, 2, 183-220. Massam, D. (1992). Null objects and non-thematic subjects. Journal of Linguistics, 28(1), 115-

137. Medová, L. (2009). Reflexive Clitics in the Slavic and Romance Languages. A Comparative View

from an Antipassive Perspective. ProQuest LLC. 789 East Eisenhower Parkway, PO Box 1346, Ann Arbor, MI 48106.

Lekakou, M. (2005). In the middle, somewhat elevated (Doctoral dissertation, University College London).

Steinbach, M. (1999). Argument structure and reflexivity: Middle constructions in German. In Proceedings of TLS Conference on Argument Structure.

Von Fintel, K., & Iatridou, S. (2003). Epistemic containment. Linguistic Inquiry, 34(2), 173-198.

An Imperfect Representation: the Preterit-Imperfect contrast in Syntactic Theory and SLA

Gabrielle Klassen & María Cristina Cuervo (University of Toronto)

This study examines the contrast between the two simple tenses with past reference in Spanish, the preterit and the imperfect. We employ an experimental design in second language acquisition with a twofold aim. First, we use L2 data to test Cowper’s (2005) analysis of the contrast in terms of the features of inflection. Second, we discuss our findings alongside previous L2 findings in terms of feature (under)specification rather than (missing) syntactic heads, and avoiding lexical aspect as a confounding factor. Arguing against analyses of the preterit-imperfect contrast in terms of (viewpoint) aspect, Cowper (2005) captures the complex distribution of the Spanish preterit and imperfect via one simple feature: [Entirety], dependant on the feature [Precedence] which distinguishes past reference. The Spanish preterit consists of the features [Precedence] and [Entirety]. That is, the preterit denotes that all the moments associated with the event precede the temporal anchor (1a). The imperfect, in contrast, only encodes [Precedence] and is therefore ambiguous (non-specified) with respect to whether the eventuality (the state of being in Montreal) holds at the moment of speech (1b). The feature [Entirety] is absent from the English tense system, which explains the ambiguity in this respect of the simple past (1c).

(1) a.Bill estuvo en Montreal ayer. (preterit) b. Bill estaba en Montreal ayer. (imperfect) c. Bill was in Montreal yesterday.

Recent studies have found a pattern in the morphosyntactic production of second language learners: L2 learners tend to overuse the least specified form as a “default” form in contexts consistent with the more specified form (see Bruhn de Garavito & White 2002; McCarthy 2008 for gender morphology; McCarthy 2006 for tense morphology). McCarthy (2006) captures this in terms of her Underspecification Theory. We combine this hypothesis with Cowper’s featural analysis, and so it is predicted that the imperfect would be the default form, as it is not specified for [Entirety], and would be therefore overused by L1 English L2 Spanish speakers. The preterit, in contrast, should only appear in appropriate contexts once the feature [Entirety] has been acquired. This prediction differs crucially from previous studies based on the Lexical Aspect Hypothesis (LAH) or some form thereof (e.g. Andersen 1986; Salaberry 2003, 2011; Slabakova & Montrul 2003, 2007 a.o.), which analyse acquisition in terms of verbal telicity (lexical aspect). These studies, however, fail to tease apart the acquisition of lexical aspect and the acquisition of the past tense morphology and its function, independent of the lexical aspect, and their results are more easily explained by our proposal, providing independent support for Cowper. Our study tested these predictions on two tasks performed by 10 intermediate and 10 advanced L1 English L2 Spanish speakers, as well as 5 native Spanish control speakers. The first task was a truth-value judgement task in which participants responded to written stimuli, answering questions that probed the comprehension of the feature [Entirety]. The second task was a narrative elicitation task based on a picture story. Preliminary results of the first task show that the L2 participants (advanced) have acquired the feature [Entirety]. The narrative elicitation task, in contrast, showed some deviance from the native speakers: L2 learners overextend the use of the imperfect to a strictly preterit context (9.8% errors), but not vice versa (0.006% errors). These results are consistent with an analysis of the imperfect as the less specified form (contra Salaberry 1999), providing initial support for Cowper’s featural analysis. Further analysis of the data allows us to discuss alternative analyses of the contrast, mostly based on aspect (but see Bello 1847; Rojo & Veiga 1999), and the analyses of L2 data from

the literature based on those aspectual approaches which rely on the association of the verb with telicity. References Andersen, R. (1986). El desarrollo de la morfología verbal en el español como segundo idioma

[The development of verbal morphology in Spanish as a second language]. In J. Meisel (Ed.), Adquisición del lenguaje—Aquisic¸a˜o da linguagem (pp. 115–137). Frankfurt: Klaus-Dieter Vervuert Verlag.

Bello, A. (1847). Gramática de la lengua castellana destinada al uso de los Americanos. Bruhn de Garavito, J., & White, L. (2002). L2 acquisition of Spanish DPs: The status of grammatical features. In A. T. Pérez-Leroux, & J. M. Liceras (Eds.), The Acquisition of Spanish Morphosyntax: The L1/L2 Connection (pp. 151-176). Dordrecht, The Netherlandsadrecht: Kluwer.

Cowper, Elizabeth A. (2005). The Geometry of Interpretable Features: Infl in English and Spanish. Language, 81(1), 10-46.

McCarthy, C. (2006). What’s Missing from Missing Inflection?: Features in L2 Spanish. McGill Working Papers in Linguistics, Volume, 20(2), 21-37.

McCarthy, C. (2008). Morphological variability in the comprehension of agreement: an argument for representation over computation. Second Language Research, 24(4), 459- 486.

Rojo, G., & Veiga, A. (1999). El tiempo verbal. Los tiempos simples. In I. Bosque & V. Demonte (Eds.), Gramatica descriptiva de la lengua española (pp. 2867-2934). Madrid, Spain: Espasa.

Salaberry, R. (1999). The development of past tense verbal morphology in classroom L2 Spanish. Applied Linguistics, 20, 151–178.

Salaberry (2002). Tense and Aspect in the Selection of Spanish Past Tense Verbal Morphology. In R. Salaberry & Y. Shirai (Eds.), The L2 Acquisition of Tense-Aspect Morphology (pp. 397- 416).

Amsterdam: John Benjamins. Salaberry, R. (2003). Tense Aspect in Verbal Morphology. Hispania, 86(3), 559-573. Salaberry, R. (2011). Assessing the effect of lexical aspect and grounding on the acquisition of

L2 Spanish past tense morphology among L1 English speakers. Bilingualism: Language and Cognition, 14(2), 184–202.

Salaberry (2013). Contrasting Preterite and Imperfect use among advanced L2 learners: Judgments of iterated eventualities in Spanish. IRAL, 51, 243-270.

Slabakova, R., & Montrul, S. (2003). Genericity and aspect in L2 acquisition. Language Acquisition, 11(3), 165–196.

Slabakova, R., & Montrul, S. (2007). L2 acquisition at the grammar–discourse interface: Aspectual shifts in L2 Spanish. In J. M. Liceras, H. Zobl, & H. Goodluck, (Eds.), Formal features in second language acquisition (pp. 452–483). Mahwah, NJ: Lawrence Erlbaum

Stress and Morphology in Second Language Production and Processing Johannes Knaus & Mary Grantham O’Brien

Department of Linguistics, Languages and Cultures, University of Calgary [email protected] [email protected] !

Second language (L2) learners often produce incorrectly stressed words, which can have a nega-tive impact on the comprehensibility of utterances (e.g., Trofimovich & Isaacs, 2012). However, the source such errors is still unknown: Is it merely an issue of performance or are the underlying representations of the words non-nativelike? Morphologically complex words provide an inter-esting test case for this question, as suffixes can influence stress placement in a principled way. We examine how English-German L2 learners process licit and illicit word stress patterns in morphologically complex words in an ERP (event related potential) study. In a further production study, we determine how they produce similar words with predictable lexical stress. Finally, we gain further insight into participants’ metalinguistic awareness of lexical stress assignment rules through the use of a think-aloud protocol. Taken together, the data enable us to determine the role of explicit awareness in the processing and production of L2 German word stress. !In the ERP study, 22 intermediate (B1 and B2) level English-German L2 learners heard a series of trisyllabic German words from one of the following conditions over loudspeakers:

a. morphologically complex German words with a predictable stress pattern based on the word’s suffix (e.g., ˈ Heiter+keit, Univers+iˈtät, Demonsˈtr+ant); or

b. morphologically complex German words of Latin origin (neoclassical word-formation) with English cognates in which the cognate differs in stress assignment (e.g., Eleˈf+ant, Mineˈr+al).

Half of the words were correctly stressed, and the remainder incorrectly stressed. The partici-pants’ task was to evaluate whether the word they heard was stressed correctly or not, as has been done in previous studies (e.g., Domahs et al., 2008). Responses to the explicit task along with electrophysiological (i.e., EEG) responses were measured to determine whether there are differ-ences in the processing of correctly vs. incorrectly stressed words. !For the production portion of the study, a comparable group of 24 English-German L2 learners produced similar German words as in the processing task, but the words in production study con-tained both three and four syllables, and an additional condition was also added: words with a final syllable containing schwa. After they produced the words, participants completed a think-aloud protocol along the lines of Osbourne (2003). That is, they listened to their own productions and provided information about the rule they followed when they assigned stress to the words. !Preliminary findings indicate that participants are relatively unaware of morphological and phonological regularities in lexical stress assignment in German and that they tend to rely upon a single across-the-board rule for lexical stress assignment when making their decisions. !References Domahs, U., Wiese, R., Bornkessel-Schlesewsky, I., & Schlesewsky, M. (2008). The processing

of German word stress: Evidence for the prosodic hierarchy. Phonology, 25,1–36. Osbourne, A. G. (2003). Pronunciation strategies of advanced ESOL learners. International Re-

view of Applied Linguistics in Language Teaching 41, 131–141. Trofimovich, P., & Isaacs, T. (2012). Disentangling accent from comprehensibility. Bilingualism:

Language and Cognition, 15(4), 905–916.

Examining ‘opposing’ processes in Quebec French mid vowels Jeffrey Lamontagne, University of Ottawa

INTRODUCTION

The mid vowels in French are reportedly subject to the loi de position, a long-proclaimed tendency to be pronounced as mid-high when in open final syllables and as mid-low when in closed final syllables, thereby losing their contrast (Walker, 1984). This study examines this claim in Quebec French, which is notably cited for its apparent counterexample to the process: an “opposite” alternation causing [ɛ] in open syllables to lower to [æ] rather than to raise to [e], as would be predicted by the loi de position (Morin, 1988). Although the dialect figures prominently in the debate as a result of this apparent exception, empirical and quantitative analysis of the two processes have not be conducted in tandem and are sparse individually.

METHOD

This study exploits the Laurentian sub-corpus (Côté, 2012) of the Phonologie du français contemporain corpus (Durand et al., 2002, 2009) to examine data from twenty-three speakers, chosen such that there were two speakers for each combination of age class (young or old), gender (male or female) and region (Chicoutimi, eastern Quebec and western Quebec) except for young males from eastern Quebec due to data availability. The recordings were aligned using Milne’s (2011) forced aligner and then the final-syllable mid vowels were extracted using a Praat script, which notably measures the first formant and stress correlates (pitch and intensity), in addition to coding for the surrounding segments and syllable type. As this pilot study uses GoldVarb (Sankoff, Tagliamonte and Smith, 2005), the vowels’ distributions were used to generate discrete allophone categories for the statistical analyses.

RESULTS

In total, 18 340 tokens were analysed. Regarding contexts refused by the loi de position, younger speakers produced mid-low vowels in open syllables 11% of the time (718 of 6244 tokens), compared to older speakers’ 16% of the time (1220 of 7531 tokens), and the younger speakers produced mid-high vowels in closed syllables 48% of the time (922 of 1939 tokens) as compared to older speakers’ 43% (1171 of 2716 tokens). Stress correlates were found to be more significant than syllable structure (more significant for younger speakers for raising and lowering, but not longer significant for mid-low variant selection) in the case of underlying mid-low vowels. For underlying mid-high vowels, syllable structure was not selected as significant, but the stress correlates showed an important contribution that increases in apparent time.

DISCUSSION

This pilot study suggests that the loi de position may be gaining influence in Quebec French. However, the significance of stress correlates has greater implications. This pilot study firstly lends credence to claims of a developing weight-sensitive stress pattern in Quebec French, whereas Standard French is traditionally analysed as having strictly word-final (or purely prosodic) stress (Armstrong, 1999). Secondly, it also suggests that the loi de position may more importantly describe a conspiracy against mid-low vowels in open syllables and against mid-high vowels in closed syllables not finally, but in stressed syllables.

REFERENCES

Armstrong, Susan Dawn. 1999. Stress and Weight in Québec French. Master’s thesis, University of Calgary: Calgary, Canada.

Côté, M.-H.2012. “Le projet PFC et la géophonologie du français laurentien”. Presented at the Les Français d'Ici colloquium 2012, Sherbrooke.

Durand, J., B. Laks & C. Lyche 2002. “La phonologie du français contemporain: usages, variétés et structure”. In C. Pusch & W. Raible, réd. Romanistische Korpuslinguistik- Korpora und gesprochene Sprache/Romance Corpus Linguistics - Corpora and Spoken Language. Tübingen: Gunter Narr Verlag, 93-106.

Durand, J., B. Laks & C. Lyche 2009. “Le projet PFC: une source de données primaires structures”. In J. Durand, B. Laks & C. Lyche, réd. Phonologie, variation et accents du français. Paris : Hermès, 19-61.

Milne, Peter. 2011. The acoustic properties of /R/ allophones and their distribution in a corpus of Canadian French. University of Ottawa comprehensive exam.

Morin, Yves-Charles. 1988. “La loi de position?”. Revue québécoise de linguistique 17 (1): 237-242.

Sankoff, David, Sali Tagliamonte, & Eric Smith. 2005. Goldvarb X: A variable rule application for Macintosh and Windows. Department of Linguistics, University of Toronto.

Walker, Douglas C. 1984. The Pronunciation of Laurentian French. Ottawa, Canada : University of Ottawa Press.

The Syntax and Semantics of Demonstratives: A DP External ApproachIssue: Standard syntactic and semantic theories preclude the notion that Dems, (e.g. this and that) and Det(erminers) (e.g. the and a) can co-occur in the same noun phrase, since they are traditionally classified as the same category (Wiltschko 2009). However, Classical Greek data is problematic for such a theory, as the data in (1) shows that Dems and Dets do in fact co-occur. (1) a. ekeinos ho anthropos that the man 'that man' (Morwood 2001: 145)Where are Dem(onstratives) located in syntactic and semantic structures and what information and features do they express?Syntax: Using Classical Greek data, I show that Dems are neither Dets (contra Wiltschko 2009), nor adjectives (contra Leu 2008). The primary evidence to support this claim is due to the behaviour of Dems in definiteness spreading constructions (Alexiadou et al. 2007). Giusti (1994), Rosen (2003), Guardiano (2012), and Roberts (2011) claim that Dems are DP internal (below Dº) and that they are located in a low position above nP. I suggest that Dems are in their own phrase projection above the DP, which I base on empirical evidence from Greek, Latin, Michif, Irish, and Basque. I show that this alternative solution accounts for all the data, without movement violations (Chomsky 1995), which is found in DP internal accounts. Also, I show that this alternative solution is optimal, since it can handle all the data in the literature, including Dem extraction data, where a Dem can appear in a final position with the rest of the DP extracted into a higher position such as in Latin (Iovino 2011), or as shown in the Basque surface order data in (2). (2) lau sagar eder hauek four apple beautiful these ‘these four beautiful apples‘ (Artiagoitia 2013: 74)Semantics: I further show that co-occurrence is problematic for semantic reasons as well, since Dets are standardly analyzed as the type <<e,t>e> (Heim & Kratzer 1998). If Dems and Dets are understood to have the same semantic function, one would not be able to take the other as an argument. I suggest that Dems would need to be of the type <e,e>, which takes an obligatory Det, even if it is unpronounced. I base this on the presence/absence of features, such as definiteness, since Dems can, but do not need to be definite, as shown in (3), where the Dem is comparable to the indefinite article in English. (3) a. so this guy bumped into me at the pub and spilt my drink! b. so a/*the guy bumped into me at the pub...'Features: Additionally, Dems appear to show temporal features in Languages like Blackfoot (Algonquian) and Aleut (Bergsland 1997). Certain Dems exhibit a situational time (time relative to the main verb) reading (Elbourne 2008), where only the change in Dem alters the tense interpretation of the entire clause. For instance, in Blackfoot “that man kicks the rock” is interpreted as past time, while “this man kicks the rock” would be interpreted as present time. Furthermore, by keeping the Dems external to the DP, temporal features can easily be read by the verb to provide situational time interpretations.

References: Alexiadou, Artemis, Liliane MV Haegeman, & Melita Stavrou. 2007. Noun phrase in the generative perspective. Berlin/New York: Mouton de Gruyter.

Blake Lewis

Artiagoitia, Xabier. 2013. Some arguments for complement-head order in Basque DPs. Anuario del Seminario de Filología Vasca Julio de Urquijo. 71-92.Bergsland, Knut. 1997. Aleut Grammar: Unangam Tunuganaan Achixaasix. A Descriptive Reference Grammar of the Alteutian, Pribilof, and Commander Islands Aleut Language. Alaska Native Language Center Research Paper Number 10.Chomsky, Noam. 1995. The Minimalist Program. MIT Press: Cambridge, Massachusetts.Elbourne, Paul. 2008. Demonstratives as individual concepts. Linguistics and Philosophy 31: 409-466.Giusti, Giuliana. 1994. Enclitic articles and double definiteness: A comparative analysis of nominal structure in Romance and Germanic. The Linguistic Review 11.3-4: 241-255.Guardiano, Cristina. 2012. Demonstratives, word order and the DP between syntax and semantics: Crosslinguistic remarks. Studies in Greek Linguistics 32: 100-115.Heim, Irene, & Kratzer, Angelika. 1998. Semantics in generative grammar. Blackwell Publishing. Malden: MA.Iovino, Rossella. 2011. Word Order in Latin Nominal Expression: The Syntax of Demonstratives. Formal Linguistics and the Teaching of Latin. Cambridge Scholars Publishing: Newcastle. 51-64.Leu, Thomas. 2008. The internal syntax of determiners. PhD dissertation. New York University: New York.Morwood, James. 2001. Oxford Grammar of Classical Greek. Oxford University Press: New York.Roberts, Ian. 2011. FOFC in DP: Universal 20 and the nature of demonstratives. Manuscript, University of Cambridge.Rosen, Nicole. 2003. Demonstrative Position in Michif. The Canadian Journal of Linguistics. 48(1/2). 39-69.Wiltschko, Martina. 2009. What’s in D and how did it get there? In J Ghomeshi, I. Paul, M. Wiltschko, eds. Determiners: universals and variation. Amsterdam: John Benjamins. 25-66.

Blake Lewis

Compound representation and individual experience

Gary Libben1, Kaitlin Curtiss1 & Silke Weber2 1Brock University

2University of Calgary In recent years, new psycholinguistic research has advanced the understanding of the conditions under which the meanings of compound constituents are activated in online lexical processing (e.g., Marelli & Luzzatti, 2012). At the same time, there has been considerable evidence that patterns of individual experience drive the development of meaning relations within compound words (Libben, Westbury & Jarema, 2012). This paper presents new evidence regarding the effect of such experience and the role of individual variation in compound representation and processing. The results of two experiments are presented. Experiment 1 employed a combined lexical recognition-production experiment using the P3 paradigm (Libben, Weber, & Miwa, 2012). The paradigm employs Primed Progressive Demasking, reading aloud, and typed word production to yield dependent variables for reading latency and accuracy, inter-letter typing times, and whole-word production time. In Experiment 2, the P3 paradigm was employed in dyadic word recognition and production. In this experimental paradigm, two participants are tested at the same time. One member of the dyad sees progressively demasked stimuli and says them aloud. The other participant types the compound stimuli (as one would in a classic dictation task). In total, 98 participants were tested. Experiment 1 yielded evidence that individual participant characteristics including education, linguistic background and metalinguistic knowledge affected performance on both tasks. The results of Experiment 2 showed than persons tested in pairs showed higher levels of accuracy and lower latencies that those tested alone. Moreover, there were effects of the differences between the two participants of a dyad in terms of their individual characteristics. These effects are captured using Visual Participant Profiles, a new data visualization technique. This technique may enable us to capture multivariate participant data and their relations to patterns of experimental results.

References Libben, G., Westbury, C., & Jarema, G. (2012). Embracing complexity. In G. Libben, G. Jarema, & C. Westbury (Eds.) Methodological and Analytic Frontiers in Lexical Research. Amsterdam: John Benjamins (pp. 1-12). Libben, G., Weber, S., & Miwa, K. (2012). P3: A technique for the study of perception, production, and participant properties. The Mental Lexicon, 7:2, 237-248. Marelli, M., & Luzzatti, C.G. (2012). Frequency Effects in the Processing of Italian Nominal Compounds: Modulation of Headedness and Semantic Transparency. Journal of Memory and Language, 66, 644-664.

1

The optionality property of Kiswahili applicatives

Jianxun Liu

University of Victoria

This paper reports the optionality properties of Kiswahili applicatives observed in fieldwork. Adjunct thematic roles can be introduced in prepositional phrases (PP) or realized as applicatives in Kiswahili (Ngonyani, 1998; Port, 1981). Ngonyani (1998) observes that in a clause containing both a direct object (DO) and an instrumental in post-verbal positions, the instrumental cannot be realized as applicative. Ngonyani attributes this to the existence of the post-verbal DO and generalizes this with the restriction on the post-verbal realization of both DO and applied instrumental. Based on our fieldwork, we argue that whether or not an instrumental can be realized as applicative is not affected by the existence/absence of the post-verbal DO, but is related to its position in the clause, pre-verbal or post-verbal. Our argument is evidenced by the following data.

First, in some intransitives, the instrumental still cannot be realized as applicative in a post-verbal position. In (1) the instrumental kisu ‘knife’ is obligatorily introduced in a PP, though this clause contains no DO. (1). [m.tʍu huju' a- li- khath (*-i) -a *(kua) kisu] man this 3sg. Past chop Appl FV by knife

The man chopped with a knife. Second, for some clauses when the DO is fronted to a pre-verbal position, the instrumental

cannot be realized as an applicative in a post-verbal position notwithstanding (2). (2). [hi ' ni ɟ iama ambajɔ m.ʧuhuju a- li- khath (*-i) -a *(kua) kisu] this is meet that man 3sg. Past cut Appl FV with knife

This is the meat that the man chopped with a knife. Third, when appearing in a pre-verbal position, instrumental can, or even must, be realized

as an applicative (3). (3) [hikhi ni khisu amɓaʧɔ mʧu huju a- li- khath *(-i) -a anijama]

this is knife which man this 3sg. Past cut Appl FV meat This is the knife with which the man chopped the meat. This pre-verbal tendency of applicatives also holds for other thematic roles including goals

and locatives. Goals in Kiswahili, for example, when appearing in a post-verbal position, can be realized in either applicatives or PPs. However, only the applicative form is grammatical when the goal is extracted to a pre-verbal position (4). (4) [huju ndiɔ mbwa (*kwa) ambajɛ paka a- li- m- kimbi *(-li) -a] this be dog with which cat 3sg. past OM run Appl FV

This is the dog toward which the cat ran. In summary this study finds that in Kiswahili the applicative construction is preferred or

even obligatory when an adjunct thematic role appears in a pre-verbal position. To account for these word order facts in Kiswahili within the current minimalist framework can be an interesting topic in future studies.

2

References Baker, Mark. (1988). Theta theory and the syntax of applicatives in Chichewa. Natural

Language and Linguistic Theory 6: 353-389. Gerdts, Donna B., and Lindsay Whaley. (1991). Two types of oblique applicatives in

Kinyarwanda. In Proceedings of the Western Conference on Linguistics 4, ed. by K. Hunt, T. Perry, and V. Samiian, 138-151. California State University, Fresno, CA.

Jeong, Youngmi. (2006). The landscape of applicatives. Doctoral dissertation, University of

Maryland. Kimenyi, Alexandre. (1980). A Relational Grammar of Kinyarwanda. University of California

Press, Berkeley. Ngonyani, Deo. (1998). Properties of applied objects in Kiswahili and Kindendeules. Studies

in African Linguistics 27: 67-97. Peterson, David. (2007). Applicative constructions. Oxford University Press, New York. Port, Robert. (1981). The applied suffix in Swahili. Studies in African Linguistics 12: 71-15.

From a classifier language to a mass-count language: What can historical data show us? Danica MacDonald, University of Calgary

This paper investigates both the historical development and the modern-day uses of the

Korean morpheme -tul. Korean, like other Eastern Asian languages, is considered to be a classifier language. A predominant property of classifier languages is that they lack plural-marking (Allen 1980, Chierchia 1998); however, Korean poses an interesting problem for this claim since it appears to have an optional plural-marker: -tul (Kang 1994, Baek 2002, Kim 2005). Some researchers propose that it is not a plural-marker at all, but rather a marker of information structure marking distributivity (Park 2008) or focus (Song 1975). Korean -tul has been studied extensively; however, there is little consensus as to its distribution or function. My research takes a new approach to the analysis of -tul by examining the historical development of this morpheme. The goal of this paper is to shed light on the modern-day uses of -tul through its past. The general question this paper addresses is whether the development of -tul is consistent with the properties of a classifier language and whether, based on that development, Korean still qualifies as a classifier language.

Park (2010) proposes that Modern Korean -tul has grammaticalized from an autonomous noun to an inflectional morpheme and then, subsequently, develops into an agreement marker. I observed a different developmental path for Korean -tul. Like Park (2010), I suggest that Middle Korean -tul underwent grammaticalization. However, I also suggest that -tul was originally used to mark focus on nouns that should be interpreted as plural, rather than functioning as a uniquely plural-marking morpheme. Under this approach, -tul would not have been optional, but would have only been used in certain specific contexts. I also propose that if –tul functions like an plural-marker today, it is not due to the development from Middle Korean, rather, it is, at least partially, due to language contact. My study is based on a corpus analysis which investigates the historical use and development of -tul. The study comprised 125 newspaper articles which covered, approximately, a 100-year period (1924 – 2011). I specifically looked for data relevant to the distribution of -tul, the number of instances of -tul in the article, the type of nouns which -tul attached to, as well as cases where a plural interpretation was clear, but -tul was not used.

This research revealed that in the earlier data, there were very few cases of -tul. The cases which were found were limited to use with human nouns (e.g. haksaeng ‘student’) and the use of -tul did not extend to other animate (e.g. khokkiri ‘elephant’) or inanimate (giep ‘enterprise’) nouns. In the early data, -tul did not seem to be functioning as a plural marker. Rather, -tul seemed to be functioning as a way to place emphasis or focus on the noun to which it attached. In the later texts, -tul is used more frequently and its use is extended to additionally include non-human nouns, and later concept-denoting abstract nouns. Corbett (2000) proposes an implicational ranking of semantic classes for languages that do not mark number for all count nouns: if a language marks number on nouns referring to inanimates, then it must also mark it on animate nouns. We could expect that Corbett's Animacy Hierarchy is mirrored in diachrony, such that if number is introduced as a morphological category, it will first appear on human nouns, then be extended to other animates, next to inanimates, and finally to nouns referring to abstract concepts. This appears to be supported by the Korean data.

The newspaper data challenges Korean’s status as a classifier language. It appears that Korean started out as a classifier language, not unlike other typical classifier languages, but the language appears to be undergoing a shift towards a mass-count system. Addressing the issue from a historical perspective is something not previously undertaken, but it provides a unique insight into the question of the status of Korean. My presentation will focus on the corpus data and its analysis and will discuss the implications of these findings for the modern-day uses of –tul and Korean’s status as a classifier language.

References

Allan, K. (1980). Nouns and countability. Language 56: 541-67. Baek, M.-H. (2002). Handkwuke pokswu uymi yenkwa. [A study on Korean plural senses.]

Discourse and Recognition. 9(2): 59-78. Chierchia, G. (1998). Reference to kinds across languages. Natural Language Semantics 6:339-

405. Corbett, G. (2000). Number. Cambridge: Cambridge University Press. Kang, B.-M. (1994). Plurality and other semantic aspects of common nouns in Korean. Journal

of East Asian Linguistics 3: 1-24. Kim, C.-H. (2005). The Korean plural marker tul and its implications. PhD Dissertation

(University of Delaware). Park, S. (2010). Grammaticalization of plurality marker tul. Seoul International Conference on

Linguistics. Park, S.-Y. (2008). Plural marking in classifier languages: a case study of the so-called plural

marking –tul in Korean. Toronto Working Papers in Linguistics 28, 281-295. Song, J.-J. (1997). The so-called plural copy in Korean as a marker of distribution and focus.

Journal of Pragmatics 27, 203-224.

Allophonic variation in English /l/: production, perception, and segmentation Sara Mackenzie1, Erin Olson2, Meghan Clayards3, Michael Wagner3

Memorial University1, Massachusetts Institute of Technology2, McGill University3

The distribution of allophones often depends on word, morpheme, or syllable boundaries, and thus encodes prosodic or morphological information. One such case is the distribution of dark and light /l/ in North American English. Light [l] is often claimed to appear in onsets and dark [ɫ] in rimes (e.g. Halle & Mohanan 1985). Instrumental studies have provided evidence for a more complex pattern, with the darkness of intervocalic /l/ being affected by morphological constituency (e.g. Sproat & Fujimura 1993, Lee-Kim et al. 2013). If the distribution of dark and light /l/ reliably correlates with the location of word, morpheme, and syllable boundaries, it may provide important cues to speech segmentation in perception. This talk reports on a series of experiments on the production and perception of dark and light /l/. These studies investigate the question of how speakers use allophonic cues to encode linguistic boundaries and whether listeners make use of these cues in determining the location of boundaries in perception.

In the perception study, subjects listened to two-word nonce strings. All strings had a C1VC2VC3VC4V shape in which C3 is ambiguous between the final C of word 1 and the initial C of word 2 (e.g. [desiledu]). For each string they heard, subjects were asked to choose between two, 2-word orthographic representations with the choices varying only in the location of a space indicating a word boundary (e.g. desi ledu vs. desil edu). Some strings contained a dark /l/ in C3 position and others contained a light /l/. Stimuli were produced by one of the authors reading the strings aloud without a pause. Sound files were cross-spliced; the dark [ɫ] condition consists of a C1VC2VC3 string containing a dark [ɫ] followed by a VC4V originally produced following a light [l] and vice versa.

Results from 60 participants show that presence of dark versus light /l/ played a significant role in segmentation. Light [l] was a strong cue to word-initial position, with listeners choosing the orthographic representation with /l/ in the initial position of word 2 at a rate of 92.4% when presented with a string containing light /l/. When participants heard a string containing dark [ɫ], they chose the orthographic form with /l/ in the initial position of word 2 43.8% of the time.

This perception study shows that listeners make use of the distinction between light and dark /l/ when locating word boundaries. However, because word boundaries coincide with morpheme and syllable boundaries, this result does not clarify the specific environments which condition the variation in production - or the type of boundaries which listeners’ recover in perception. We attempt to address these questions in a series of production studies which elicit productions of /l/ in a variety of morpho-syntactic contexts. Productions of /l/ were measured for F2-F1 values, an acoustic property known to play a role in the distinction between light and dark /l/ (e.g. Sproat & Fujimura 1993). We found word-initial /l/s (woo lasses) to be lighter than word-final ones (fool asses). Within words, we found morpheme-initial /l/s (free-ly) to be lighter than morpheme-final /l/s (meal-y). Contrary to Hayes (2000) and Sproat and Fujimura (1993), we found no difference in /l/ darkness between morpheme-final /l/s (kneel-ing) and morpheme-internal ones (ceiling), with /l/ being realized as dark in both environments.

Overall, our production studies found light /l/ only in morpheme-initial position and dark /l/ in morpheme-final and morpheme-internal position. This is consistent with an account in which dark /l/ is the basic variant for our speakers and /l/-lightening is a form of articulatory strengthening which occurs following prosodic boundaries with the presence of such boundaries being triggered by morphological structure. This constituent-initial /l/-lightening serves as a cue for listeners to disambiguate between multiple structural parses of a segmental string. References

Halle, M. & Mohanan K. P. (1985) Segmental phonology of modern English. Linguistic Inquiry

16. 57-116. Hayes, B. (2000) Gradient well-formedness in Optimality Theory. In Dekkers, J., van der Leeuw,

F. and van de Weijer, J. (eds.) Optimality Theory: Phonology, Syntax, and Acquisition Oxford University Press. 88-120.

Lee-Kim, S., Davidson, L., & S. Hwang. 2013. Morphological effects on the articulation of English intervocalic /l/. Laboratory Phonology 4(2).

Sproat, R. & O. Fujimura. (1993) Allophonic Variation in English /l/ and its implications for phonetic implementation. Journal of Phonetics 21. 291–311.

1

L2 Spanish speakers' perception of secondary cues in Y/N question and statement intonation

Olivia Marasco

University of Toronto

English broad focus statements and yes-no (Y/N) questions differ not only in their intonation pattern but also in their word order. In Spanish, however, the only difference between these two types of utterances is the intonation pattern. The word order and number of words is exactly the same: (1) Rompió la mesa del comedor. (2) ¿Rompió la mesa del comedor? (He broke the dining room table.) (Did he break the dining room table?) English speakers acquiring Spanish must learn that in their L2 they rely solely on intonation patterns to distinguish these two types of utterances. While Y/N questions in both languages make use of a final rising contour, other intonation cues differ, in some cases, quite significantly. The specific focus of the current project is the initial part of the utterance, particularly initial boundary tone height and pre-nuclear peak height .

The research questions guiding this project are: 1) Can L2 Spanish (L1 English) listeners detect intonational phonetic differences not present in the equivalent structures of their L1? 2) Do advanced L2 speakers possess the form-meaning mapping of these structures? Spanish signals a Y/N question from the very beginning of the utterance by making use of a higher initial boundary tone as well as a higher pre-nuclear peak with respect to its comparable statement. This height difference is not used in English to signal the presence of Y/N questions and comparable statements (English: Pierrehumbert, 1980; Bartels, 1999; Spanish: Sosa, 1999; Hualde, 2005). The predictions for this study were: 1) participants would be more successful in the discrimination of acoustic differences rather than the identification of the form-meaning mapping; 2) the discrimination task would have a categorical outcome while the form-mapping meaning will show a greater variety of answers that would distribute along a continuum.

Ten advanced L2 Spanish (L1 English) speakers took part in the current study, which consisted of discrimination and identification tasks. The discrimination task was an AX task where participants heard 3 blocks of 15 pairs of utterances that were one of the following: i) identical, ii) different in pre-nuclear peak height or iii) different in initial boundary tone height. Half of these utterances had one manipulated parameter. Both declarative and interrogative bases were used for manipulation. The identification task used the same recordings from the previous task but in this case participants were explicitly asked whether they were hearing a statement or a question. In order to avoid interference from the primary cue, the final boundary tone of each utterance was masked by a barking dog sound.

Preliminary findings of both tasks are reported here. In the discrimination task, differences of peak height resulted in a 'different' response while differences in initial boundary tone did not have the same response, even in cases where two utterances had initial boundary tone differences of over 200 Hz. In the identification task, utterances with both bases were identified as questions or as statements when their prenuclear peak height was raised or lowered respectively. This means that if a statement base had its peak raised, it would be identified as a question. The same did not occur, however, when the initial boundary tone was raised or lowered even by several hundred hertz. When it comes to intonation not all phonetic parameters carry the same weight. Although the initial boundary tone is the first thing listeners hear, the cue for utterance type

2

seems to come from the pre-nuclear peak for these L2 listeners. The expected outcome was a difference between tasks, but what emerges from the findings thus far is that listeners are successful within each task based on the particular intonational feature that is being highlighted. References Bartels, C. (1999). The Intonation of English Statements and Questions. New York: Garland

Publishing Inc. Hualde, J. I. (2005). The sounds of Spanish. Cambridge, M.A., Cambridge University Press. Pierrehumbert, J. (1980). The Phonology and Phonetics of English intonation (Doctoral

dissertation). MIT, Cambridge, MA. Sosa, J. M. (1999). La entonación del español: su estructura fónica, variabilidad y dialectología.

Madrid: Cátedra.

Markedness in number features: Evidence from Ganggalida (Yukulta) Jessica Mathie, University of Toronto This paper presents data from Ganggalida (Tangkic family, Australia) which suggests that the featural representation of plural number is more highly marked than dual in this language. This challenges the number feature geometry proposed by Harley and Ritter (2002), in which dual is more highly marked than plural. I show that the Ganggalida facts follow naturally from Cowper’s (2005) geometry. Ganggalida normally has a three-‐way singular/dual/plural distinction in its pronominal clitic system. However, in particular contexts (namely, clauses with third person non-‐singular subjects acting on second person non-‐singular objects) the contrast between dual and plural is neutralised. In neutralised contexts the normal dual clitic ‘rr’ cross-‐references both dual (1a) and plural (1b) entities. In these contexts it has a ‘non-‐singular’ meaning, analogous to the meaning of a plural in a two-‐way singular/plural number system. (1a) 3DU>2DU

ngamathu-‐yarrngga=rrawa-rra garna-‐ja wurlank-‐i girrwa gunawuna-‐ntha mother-‐two.ABS=2DU.DAT-‐DU cook.IND food-‐LOC 2DU.DAT child-‐DAT Your mothers are cooking food for you two children. (Keen 1983:206#34b)

(1b) 3PL>2PL dathin-‐da=rrawa-rr-‐ingg-‐a gurri-‐ja gilwan-‐ji that-‐ABS=2DU.DAT-‐DU-‐NPRES-‐RLS look-‐IND 2PL-‐OBL Those fellows are looking at you lot. (Keen 1983:236#155a)

This pattern is captured naturally by Cowper’s (2005) feature geometry, given in (2). In Ganggalida’s regular three-‐way system (2a) the dual vocabulary item (VI) ‘rr’ spells out [>1] and the plural VI ‘l’ spells out the dependant feature [>2]. In the neutralised system, however, only a subset of the plural features is spelled out; spellout of [>2] is blocked and [>1] is spelled out by the dual VI instead. This model accounts for the distribution of VIs and allows for unique and consistent lexical feature specifications for each VI. (2) Cowper (2005) (boxed features are spelled out)

(a) three-way system (b) neutralised system (singular) (dual) (plural) (singular) (dual) (plural) [#] [#] [#] [#] [#] [#]

>1 [>1] [>1] [>1] [>2] [>2]

VI Ø/rn rr l Ø/rn rr rr The geometry proposed by Harley and Ritter (2002), given in (3), does not effectively capture the Ganggalida system. Under Harley and Ritter’s approach the dual VI would normally spell out [group, minimal] while the plural VI spells out [group] (3a). In neutralised contexts, the dual must also be able to spell out [group] in order to cross-‐reference plural entities (3b). (3) Harley & Ritter (2002) (a) three-way system (b) neutralised system (singular) (plural) (dual) (singular) (plural) (dual) [#] [#] [#] [#] [#] [#] [minimal] [group] [group][minimal] [minimal] [group] [group][minimal] Ø/rn l rr Ø/rn rr rr This forces us to claim that the dual VI has two lexical representations, [group, minimal] and [group], and also results in there being two VIs with the identical lexical representation [group]. This account has less explanatory power due to these added specifications. Ganggalida data, therefore, provides evidence in support of a highly marked plural featural representation, and against a highly marked dual.

References Cowper, Elizabeth. 2005. A note on number. Linguistics Inquiry 36(3). 441-‐455. Harley, Heidi & Elizabeth Ritter. 2002. Person and number in pronouns: a feature-‐

geometric analysis. Language 78. 482-‐526. Keen, Sandra. 1983. Yukulta. In RMW Dixon & BJ Blake (eds.) Handbook of Australian

Languages Vol. 3 Canberra: Australian National University Press. 190-‐304.

Monica McMillan University of Western Ontario

Because Twitter: Linguistic Strategies for Brevity in Internet Speak

Twitter is a social networking site with 280 million active users every month. It provides brief and interactive communication and is used for everything from advertising and networking to sharing big news with friends. Twitter, along with other social networking sites (e.g. Tumblr, Facebook, Reddit), is playing a pivotal role in the current evolution of the English language. Its famous 140-character limit for status updates forces users to be brief, precise, and creative in order to convey their message.

The primary research question posed in this paper is: What linguistic strategies do Twitter users employ to clearly convey meaning in such a brief space? The secondary line of inquiry examines the pragmatic inferences these usages carry. This work-in-progress corpus study will examine 3 linguistic strategies that have evolved from this unique linguistic environment.

First, ‘because X.’ This construction conveys (roughly): “I’m so engrossed by this that I can’t be bothered to explain further” or “the reasoning is so obvious it doesn’t need to be elaborated.” As in 1):

1) “@OhLookItsNiamh: Can’t study because sleep but can’t sleep because study”

Second: ‘I can’t even,’ which conveys a sense of “This is so exciting/outrageous/ridiculous (etc.) that I’ve lost the ability to form proper sentences.”

2) “@niskakayla: MY MISSION CALL IS HERE!!! I can’t even. I’m unable to even. Ahhhh!!”

The third phrase in 2) “I’m unable to even” is also interesting, as it goes against the briefness constraint that normally limits tweets. Third: abbreviations, such as ‘imo’ (in my opinion), ‘lol’ (laugh out loud) and ‘idk’ (I don’t know), among others. It has been said that common Internet-speak abbreviations have evolved into pragmatic particles. Specifically, that ‘lol’ is now used to convey empathy, rather than it’s traditional meaning that the speaker finds something funny (McWhorter, 2013). An example of this usage can be seen in 3):

3) “@marquiepeyton: if it wasn’t for whit I would run away lol bc I hate everyone” 4) “@elfifaplayer: Voice > X factor. Much better signers and judges IMO” 5) “@CapedCrusaider: idk who’s Girl Scout cookies those were but I ate some.”

The presentation will examine instances of these three structures found in a corpus of Tweets; when they are used, who is using them, and what they mean in the various contexts in which they appear. Conclusions will be made regarding the intended meaning of these structures and the overall effectiveness of them. Inferences will be drawn regarding the overall effect of Internet speak and online social interactions on language evolution. References McWhorter, John. 2013. Txting is killing Language. JK!!! Electronic Document. www.ted.com/talks/john_mcwhorter_txtng_is_killing_language_jk.html. Accessed December 18, 2013.

Monica McMillan University of Western Ontario

Étude expérimentale de l’interprétation des questions rhétoriques Nesrine Mejri et E. Allyn Smith, L’Université du Québec à Montréal Dans cette communication, on examine les facteurs sémantico-pragmatiques qui rentrent en jeu dans l’interprétation des questions rhétoriques (QR) en français. Notre travail représente la première étude consacrée aux QR dans un cadre expérimental. Travaux antécédents: Les QR représentent un phénomène complexe très peu étudié, surtout en français. Elles se distinguent des questions de demande d'information (QI) dans la mesure où elles sont assertives et n'exigent pas de réponse informative (voir (1)). (1) Est-ce que l'argent fait le bonheur ? Oui/Non. (réponse informative)

C’est vrai. (réponse non-informative)

Borillo (1981) propose que les verbes d’opinion comme penser et les adverbes épistémiques (vraiment, réellement) favorisent l’interprétation rhétorique d’une question en français. Rohde (2006) considère que la présence du contenu véhiculé par la question dans le savoir partagé (SP) des locuteurs favorise son interprétation rhétorique. Des travaux connexes (Romero et Han 2004, Romero 2005, 2006) portant sur les questions biaisées (des QI où la polarité de la réponse attendue devient claire, avec l’ajout de n’est-ce pas, par exemple) sont consacrés au mot anglais really (‘vraiment’). Romero propose l'opérateur VERUM pour la dénotation du mot vraiment dont le rôle est de changer la dénotation d’une question comme ‘Est-ce que Marie viendra vraiment?’ d’une question où on demande si Marie va venir à une question où on demande si on est certain de vouloir ajouter l’information qu’elle va venir à notre SP. Par contraste, les théories avancés par vanRooy (2003) et Rohde (2006) proposent un calcul de la prévisibilité d'un ensemble de réponses pour expliquer le même phénomène. Nous voulions savoir si : (i) les facteurs proposés par Borillo et Rohde correspondent aux perceptions des locuteurs, (ii) certains facteurs biaisent l'effet rhétorique plus que d’autres, (iii) il y a un effet de renforcement qui se produit quand 2 ou 3 facteurs se combinent, et (iv) la perception des locuteurs pourrait être modélisée avec VERUM ou le calcul de la prévisibilité des réponses. Expérience : Nous avons conçu un sondage où nous avons testé les 3 facteurs suivants: vraiment, penser, et SP à travers la manipulation des 2 valeurs présence/absence de chaque facteur. Cela nous a donné 8 conditions pour chacune des 8 questions testées en contexte; il y avait donc 8 sondages écrits en design carré latin. 60 étudiants universitaire ont complété ces sondages. Leur tâche consistait à : (i) lire un contexte particulier faisant état du SP ou non de 2 locuteurs (Éric et Julie) ; (ii) lire à chaque fois un dialogue entre ces deux locuteurs comprenant une question cible comme en (2) ; et (iii) répondre à la question : « Selon vous, est-ce que Julie attend une réponse informative de Éric ? » en choisissant un numéro sur une échelle Likert de 6 points. (2) Éric : Geneviève pleure encore dans sa chambre. Julie : Est-ce que tu penses vraiment que Geneviève pleure parce qu'elle a raté son permis? Résultats : L’analyse statistique (via t-test) des résultats normalisés montre que les seuls conditions qui favorisent l’interpretation rhétorique sont (i) vraiment, (ii), vraiment + SP, et (iii) vraiment + SP + penser. En d'autres termes, contrairement aux hypothèses avancées respectivement par Borillo et Rohde, penser ne favorise pas l’interprétation rhétorique, et le SP ne le favorise pas systématiquement (seulement s'il est combiné avec vraiment). Vraiment favorise l’effet rhétorique, ce qui est renforcé particulièrement en combinaison avec le SP. Nous montrons comment ces résultats posent problème pour VERUM et pour le calcul de van Rooy et Rohde avant de proposer l’analyse suivante : si on suppose que vraiment implicite que notre loculteur sera surpris par la proposition en question (Romero 2005, Potts 2003), alors l’emploi de ce mot est incompatible avec la certitude réprensentée par le SP. Par conséquent, leur co-occurence signale l’impossibilité d’une vraie question et déclenche l’interprétation rhétorique.

Bibliographie : Borillo, A. (1981). Quelques aspects de la question rhétorique en français. Documentation et recherche en linguistique allemande, Vincennes (DRLAV) 25. 1-32. Potts, Chris. (2003). Expressive content and conventional implicature. In Proceedings of the North East Linguistic Society 33. Amherst, MA: GLSA. Rohde, H. (2006). Rhetorical questions as redundant interrogatives. San Diego Linguistic Papers 2. 134-168. Romero, M. (2005). Two approaches to biased yes/no questions. In Proceedings of the 24th West Coast Conference on Formal Linguistics. 352-360. Romero, M. (2006). Biased yes/no Questions: The Role of verum. Sprache und Datenverarbeitung, 30(1). Romero, M., & Han, C. H. (2004). On negative yes/no questions. Linguistics and Philosophy, 27(5). 609-658. Van Rooy, R. (2003). Negative polarity items in questions: Strength as relevance. Journal of semantics, 20(3), 239-273.

Null pronouns in English: evidence from particle verb constructionsDaniel Milway

University of Toronto

English is widely considered to lack the null pronoun, pro, null arguments being restricted totraces, PRO and bound variables. Evidence presented in this paper, however, demonstratesthat pro is present in certain English particle verb (PV) constructions. Furthermore, Englishnull pronouns differ from other null pronouns discussed in the literature by having bothdiscourse and syntactic requirements.

The PVs in question, demonstrated below in (1), are members of the class that show anargument structure alternation known as ground promotion (McIntyre, 2007), and have acomplementary alternation which I refer to as figure retention.

(1) a. V + Full PPAlex rinsed [Figurethe dust] out of [Groundthe coffee pot].

b. Ground PromotionAlex rinsed out [Groundthe coffee pot]. ∼ Alex rinsed [Groundthe coffee pot] out.

c. Figure RetentionAlex rinsed out [Figurethe dust]. ∼ Alex rinsed [Figurethe dust] out.

While both ground promotion and figure retention constructions are able to undergo particleshift, the characteristic syntactic property of PVs, Levin and Sells (2007) note an asymme-try between the two constructions with respect to their interpretation, namely, that figureretention PVs require contextual support for proper interpretation, while ground promotiondoes not. This paper shows that this asymmetry is due to the presence of a null argumentin figure retention which is interpreted as the contextually salient ground argument.

(2) a. Figure RetentionWhenever Jim wears his blazer, [he wipes the fingerprints off.]Interpretation: He wipes the fingerprints off his blazer.

b. Ground PromotionWhenever Sheila sees fingerprints, [she brushes her blazer off.]Interpretation: She brushes something off her blazer.Not:She brushes fingerprints off her blazer.

These null arguments do not show the syntactic dependencies that are characteristic of PRO,traces, or bound variables, and are interpreted as definite and specific like pronouns. As such,they are properly treated as null pronouns.

Beyond showing the presence of these null pronouns, this paper provides a structualanalysis for them, and places them in a typology null pronouns. The structural analysiscomes from the fact that these null pronouns are predicted by a small clause type analysis ofPVs (Svenonius, 2003). Null pronouns in figure retention are discourse-linked in the same wayas null objects are in Colloquial Brazillian Portuguese (Farrell, 1990) and French (Cumminsand Roberge, 2004). Unlike other discourse-linked null objects, English null grounds are alsolinked to a syntactic context, namely figure retention PVs. This adds strength to the claim,made by Iten et al. (2005), that semantic-pragmatic recoverability is not sufficient to explainnull complements, but that they must be syntactically licensed as well.

1

References

Cummins, Sarah, and Yves Roberge. 2004. Null objects in french and english. AmsterdamStudies in the Theory and History of Linguistic Science Series 4 258:121–138.

Farrell, Patrick. 1990. Null objects in brazilian portuguese. Natural Language & LinguisticTheory 8:325–346.

Huang, C-T James. 1984. On the distribution and reference of empty pronouns. Linguisticinquiry 531–574.

Iten, Corinne, M-O Junker, Aryn Pyke, Robert J Stainton, and Catherine Wearking. 2005.Null complements: Licensed by syntax or by semantics-pragmatics? In CLA AnnualConference Proceedings , 1–15.

Levin, Beth, and Peter Sells. 2007. Unpredicated particles. In Redpill (reality explorationand discovery: Pattern interaction in language and life), ed. Lian Hee Wee and UyechiLinda. CSLI Publications.

McIntyre, Andrew. 2007. Particle verbs and argument structure. Language and LinguisticsCompass 1:350–367.

Svenonius, Peter. 2003. Limits on p: filling in holes vs. falling in holes. In Nordlyd 31.2 , ed.Anne Dahl, Kristine Bentzen, and Peter Svenonius, 431–445. University of Tromsø.

2

Variation prosodique diachronique: René Lévesque en trois temps

Nicholas Moroz (Collège de Rimouski) et François Poiré (The University of Western Ontario)

À l’exception des travaux développés dans le cadre de la sociolinguistique contemporaine et

basés sur des corpus relativement récents, peu d’études ont été menées sur le changement

prosodique du français à partir d’enregistrements de qualité. Martin (2005), en utilisant des

enregistrements radiophoniques couvrant le XXième siècle a documenté la disparition graduelle

de l’accentuation sur la syllabe pénultième du groupe accentuel et une variation certaine dans la

réalisation des contours intonatifs exprimant la continuité. Boula de Mareüil et al (2008), en se

basant aussi sur des enregistrements radiophoniques faits depuis la seconde guerre mondiale, ont

pu préciser que le mouvement de l’accent de la pénultième vers la syllabe finale repose en fait

sur une augmentation de la différence de durée entre ces deux syllabes. Ils ont aussi observé que

la moyenne de fréquence fondamentale des journalistes a aussi diminué durant la même période,

sans toutefois préciser les conséquences fonctionnelles de cette diminution sur la forme des

contours intonatifs. Dans les deux cas, ces études portent sur le français de France. Comme le

mentionne Harrington (2006) au sujet des études phonétiques diachroniques: « what is lacking in

most of the very few studies that are available is an analysis over several decades from the same

person, which would eliminate the confounding influences due to the substantial differences

between different speakers’ anatomical vocal tracts » (Harrington 2006 : 441).

Nous proposons une étude du changement prosodique chez un locuteur unique, René Lévesque

(1922-1987), ancien journaliste à la radio puis à la télévision et politicien bien connu au Canada,

à partir de trois enregistrement datés de 1957 (animateur de l’émission Point de mire, Radio-

Canada), 1968 (discours public devant les membres du Mouvement Souveraineté-Association) et

1984 (invité à l’émission Le Point, Radio-Canada) provenant des archives de la société Radio-

Canada à Montréal. Cette recherche s’inscrit dans le cadre de la phonologie métrique-

autosegmentale (Ladd 1996) adaptée au français par Jun et Fougeron (2000). La méthodologie

reprend principalement celle de Kaminskaïa (2009) développée pour l’étude de la variation

prosodique régionale en français. L’analyse a été menée à l’aide du logiciel Praat (Boersma et

Weenink 2006).

Nos principaux résultats montrent premièrement que le nombre de groupes accentuels plus longs

exprimant la continuité (nombre de syllabe plus élevé, contour bbH) augmente pendant la

période étudiée. Deuxièmement, la valeur moyenne de la fréquence fondamentale diminue tout

au long de cette période mais cette baisse concerne principalement les contours intonatifs réalisés

dans la partie supérieure du registre du locuteur. Enfin, et contrairement aux données

européennes de Boula de Mareüil et al (2008), l’écart dans la durée des syllabes pénultièmes et

finale est stable mais la durée des deux positions syllabiques augmente de manière similaire entre

1957 et 1984.

Nous discutons ces résultats dans la perspective de la stabilisation du système prosodique des

variétés européennes et canadiennes de la langue française au cours des dernières décennies

(tendance des groupes accentuels à être plus longs, perte de diversité dans la réalisation des

contours intonatif) tout en soulignant le maintien de différences entre le Canada et l’Europe

(utilisation différente de la durée afin de marquer la frontière des groupes accentuels, utilisation

différente de l’étendue de registre dans la réalisation de l’intonation).

Boersma P. et D. Weenink. 2006. Praat: doing phonetics by computer. [Computer Program].

Institute of Phonetic Sciences, University of Amsterdam.

Boula de Mareüil, Ph., Albert Rilliard et Alexandre Allouzen. 2008. A diachronic study of

prosody through French audio archives. Proceedings of Speech Prosody 2008. Campaninas,

Brazil, pp. 539-542.

Harrington, Jonathan. 2006. An acoustic analysis of 'happy-tensing' in the Queen's Christmas

broadcasts, Journal of Phonetics, 34(4) 439-457.

Jun, S.-A. et C. Fougeron. 2000. A phonological Model of French Intonation. In A. Botinis (éd.)

Intonation. Kluwer Academic Publishers, 209-242.

Kaminskaïa, S. 2009. La variation intonative dialectale en français. Une approche phonologique.

LINCOM Studies in French Linguistics. Munich.

Ladd, R. 1996. Intonational phonology. Cambridge : Cambridge University Press.

Martin, Philippe. 2005. Quelques Changements prosodiques du français parlé de 1900 à 2000.

UFR Linguistique. Paris: Université de Paris VII.

A hierarchical view of the ergative and antipassive in Inuktitut Kumiko Murasugi, Carleton University

Much research on the syntax of Inuit languages has focused on the differences between the ergative and antipassive (AP) constructions in a two-argument sentence such as ‘Jaani saw a caribou’ (see (1)).

(1) Erg Jaani-up tuktu- Ø taku-lauq-tanga. Jaani-Erg caribou-Abs see-Past-Partic.3s.3s

AP Jaani- Ø tuktu-mik taku-lauq-tuq. Jaani-Abs caribou-Mod see-Past-Partic.3s

One view is that the differences are primarily semantic or discourse-related. It has been claimed, for example, that the AP object is indefinite (Swadesh 1946, Sadock 1980, Fortescue 1984), nonspecific (Manga 1996), takes narrow scope (Bittner 1987, Wharram 2003), introduces new information (Kalmár 1979), or is a non-topic (Berge 2011), while the object in an ergative structure has opposite properties. The other view focuses on the syntactic properties of the AP structure: the AP suffix is an incorporated noun (Bittner & Hale 1996), an aspect marker (Bittner 1987), or accusative case assigner (Spreng 2006, Johns 2006). Surprisingly, there is little consensus on what underlies the ergative/AP contrast in Inuit.

This paper presents a study that investigates which structure speakers of one Inuit dialect, Inuktitut, prefer in a context-neutral experimental setting. The study focuses on verbal agreement (double with ergative, single with AP) rather than Case, as pronouns in Inuktitut are normally deleted and thus cannot be used to distinguish between the two structures. In the Verb Preference Task Inuktitut speakers were given 28 subject-object pairs consisting of all person and number combinations, and were required to create a simple sentence using the verb takulauq ‘see.Past’ (e.g. ‘the man saw the woman’, ‘I saw the boys’, ‘you saw me’). They were free to produce either the ergative or antipassive form. The results revealed the following hierarchy of preference for the antipassive form (where, for example, 1>2 represents a 1st person subject and 2nd person object): 3>3, 3>2, 3>1, 2>1, 1>2, 2>3, 1>3.

This hierarchy strongly suggests the existence of a directional system in Inuktitut based on a 1>2>3 person hierarchy. The antipassive form is preferred when a subject lower on the hierarchy is acting on an equal or higher object (shown in bold), while the ergative form is preferred when a higher object is acting on a lower one. I present a sketch of an inverse agreement system in Inuktitut consisting of direct, inverse and equal status marking in domains other than 3>3, and obviation with 3>3 (see Hewson 1991). Obviation systems “rank third person nominal according to a complex function which includes grammatical function, inherent semantic properties, and discourse salience” (Aissen 1997:205). Given that most of the studies cited above focus on examples with 3rd person subjects and objects, it is perhaps not surprising that there is so much variation in their explanations of the ergative/AP structures. The current study, which takes into account all subject-object domains, can provide a more global perspective on the two structures.

References

Aissen, Judith. 1997. On the syntax of obviation. Language 73:705-750.

Berge, Anna. 2011. Topic and Discourse in West Greenlandic Agreement Constructions. Lincoln and London: University of Nebraska Press.

Bittner, Maria. 1987. On the semantics of the Greenlandic antipassive and related constructions. International Journal of American Linguistics 53:194-231.

Bittner, Maria and Ken Hale. 1996. Ergativity: Toward a theory of a heterogeneous class. Linguistic Inquiry 27:531-604.

Fortescue, Michael. 1984. West Greenlandic. London: Croom Helm.

Hewson, John. 1991. Person hierarchies in Algonkian and Inuktitut. Linguistics 29:861-875.

Johns, Alana. 2006. An inclination towards accusative. Linguistica Atlantica 23:127-144.

Kalmár, Ivan. 1979. The antipassive and grammatical relations in Eskimo. Ergativtiy: Towards a Theory of Grammatical Relations, ed. Frans Plank, pp. 117-43. London: Academic Press.

Sadock, Jerrold M. 1980. Noun incorporation in Greenlandic: A case of syntactic word formation. Language 56:300-319.

Spreng, Bettina. 2006. Little v in Inuktitut: Antipassive revisited. Linguistica Atlantica 23:155-190.

Manga, Louise. 1996. Specificity in Inuktitut and syntactic representation. Études/Inuit/Studies 20:61-85.

Swadesh, Morris. 1946. South Greenlandic (Eskimo). Linguistic Structures of Native America, ed. Harry Hoijer, pp. 30-54. New York: Viking Fund.

Wharram, Douglas. 2003. On the Interpretation of (Un)certain Indefinites in Inuktitut and Related Languages. Ph.D. dissertation, University of Connecticut.

A PROSODIC ACCOUNT OF CANADIAN RAISING D. SKY ONOSSON — UNIVERSITY OF VICTORIA ! Canadian Raising (CR) is a fairly well-known phenomenon in the phonological literature, which has been described primarily in featural or qualitative terms such as: “the diphthongs [ʌj] and [aj] are in complementary distribution: [ʌj] occurs before the class of voiceless consonants ([s, t, p], etc.) and [aj] occurs elsewhere. A parallel relationship holds between the vowels [aw] and [ʌw]” (Czaykowska-Higgins et al 2012, cf. Joos 1942, Chambers 1973). Phonetic research I have carried out on Manitoba speakers indicates that vowel shortening is a more significant descriptor for CR than raising: pre-voiceless diphthongs are nearly 50% shorter than elsewhere, while vowel height differs only slightly across all allophones. However, feature-based accounts do not account for vowel duration as an intrinsic part of CR. In this paper I propose a phonological analysis which interprets this durational difference as a reflection of prosodic structure, specifically as an effect of mora sharing, as determined by the sonority distinction between voiced and voiceless codas. There are three main theoretical components to my analysis: 1) the availability of [voice] as a sonority distinction; 2) the connection between sonority and prosodic structure; 3) the relationship between prosodic structure and phonetic duration. A cross-linguistic comparison of sonority research (Parker 2008) indicates that voicing quality is an available sonority distinction within several languages, justifying my proposal that it is also an available distinction within Canadian English. Research on the relationship between sonority and prosodic structure (Zec 1995) indicates that mora licencing/affiliation is related to sonority distinctions; I propose that sonority distinctions related to [voice] thus determine mora licencing/affiliation in Canadian English. Finally, a cross-linguistic study of the relationship between phonetic duration and prosodic structure (Broselow et al 1999) has found that different prosodic (moraic) structures are possible not only between languages, but also within them, and that these difference structures are associated with consistent differences in phonetic duration. My analysis relies on an existing optimality-based account of the prosodic structure of the English syllable (Hammond 1999) and proposes one additional constraint restricting independent morae to codas which meet a minimum level of sonority; specifically, mora-licencing codas must not be lower in sonority than a voiced obstruent. The only category below this level are voiceless obstruents which, in adhering to this constraint, must instead adjoin to the final mora of the preceding vowel. E.g. in ride the /d/ projects its own mora, while in right the /t/ does not meet the minimum sonority to project a mora and must adjoin to the preceding mora belonging to the vowel. The effect of vowel-coda mora-sharing in the case of right is to abbreviate the duration of the vowel in comparison with the vowel of ride, which does not share a mora with its coda. My analysis thus relates the observed difference in phonetic vowel duration to the phonological (prosodic) level as an effect of mora-sharing, in a similar manner to the findings in Broselow et al (1999), and incorporates the most significantly related features of CR found in my phonetic research, coda voicing and vowel duration, into a unified account.

CLA 2014 Abstract"1D. S. Onosson

LIST OF REFERENCES !Broselow, E., Chen, S.-I., & Huffman, M. (1997). Syllable Weight: Convergence of Phonology

and Phonetics. Phonology, 14(1), 47–82. Chambers, J. K. (1973). Canadian raising. Canadian Journal of Linguistics/Revue canadienne de

linguistique, 18(2), 113-134. Chambers, J. K. (1989). Canadian Raising: Blocking, fronting, etc. American Speech, 64(1), 75–

88. Czaykowska-Higgins, E., Dobrovolsky, M., Dyck, C., O’Grady, W., & Rose, Y. (2012).

Phonology: contrasts and patterns. In O’Grady, W. & Archibald, J. Contemporary Linguistic Analysis: An Introduction. Toronto: Pearson Canada.

Gordon, M. (2002). A Phonetically Driven Account of Syllable Weight. Language, 78(1), 51–80. Hammond, M. (1999). The Phonology of English: A Prosodic Optimality-Theoretic Approach.

Oxford: Oxford University Press. House, A. S. (1961). On Vowel Duration in English. Journal of the Acoustical Society of

America, 33(9), 1174–1178. Joos, M. (1942). A Phonological Dilemma in Canadian English. Language, 18(2), 141–144. Lunden, A. (2013). Reanalyzing final consonant extrametricality. The Journal of Comparative

Germanic Linguistics, 16(1), 1–31. Onosson, D. S. (2010). Canadian Raising In Manitoba: Acoustic Effects Of Articulatory Phasing

And Lexical Frequency. MA thesis. University of Manitoba. Parker, S. (2008). Sound level protrusions as physical correlates of sonority. Journal of

Phonetics, 36(1), 55–90. Parker, Steve (2011). Sonority. In van Oostendorp, Marc, Colin J. Ewen, Elizabeth Hume and

Keren Rice (eds.) The Blackwell Companion to Phonology. Blackwell Publishing, 2011. Blackwell Reference Online.

Thomas, E. R. (2000). Spectral differences in /ai/ offsets conditioned by voicing of the following consonant. Journal of Phonetics, 28, 1–25.

Umeda, N. (1975). Vowel duration in American English. Journal of the Acoustical Society of America, 58(2), 434–445.

Zec, D. (1995). Sonority constraints on syllable structure. Phonology, 12(1), 85–129. Zec, D. (2003). Prosodic Weight. In C. Féry & R. van de Vijver (Eds.), The Syllable in

Optimality Theory (pp. 123–143). Cambridge: Cambridge University Press.

CLA 2014 Abstract"2D. S. Onosson

The rise and fall of split-ergative agreement in Algonquian Will Oxford, University of Manitoba

The inflectional paradigm known as the “independent order” displays a split-ergative agreement pattern in some of the Algonquian languages (such as Ojibwe and Passamaquoddy) but not in others (such as Cree and Meskwaki). This presentation will describe and analyze the diachronic development of the split-ergative agreement pattern, which arose in pre-Proto Algonquian when the independent order was innovated and was retained in Ojibwe-type languages but lost in Cree-type languages. I will argue that the split-ergative pattern is a relatively shallow morphosyntactic phenomenon: the difference between Cree-type languages and Ojibwe-type languages results from a minor change in the featural content of two competing agreement probes. The difference between the Ojibwe-type pattern and the Cree-type pattern is illustrated in (1) and (2), which show a subset of the independent-order agreement morphology for forms involving 1s, 2s, and 3p arguments.1 Subject-indexing agreement is indicated by boxes. In the Ojibwe transitive forms in (1a), the subject is consistently indexed by the prefix (ni-, gi-, o-), but this is not the case in the Ojibwe transitive forms in (1b): while the 1s and 2s subjects continue to be indexed by the prefix (ni-, gi-), the 3p subject is instead indexed by the same -ag suffix that indexes the 3p object of a transitive form. This is a clear split-ergative pattern: a 3rd-person intransitive subject is indexed like a 3rd-person transitive object (by the suffix -ag) rather than like a 3rd-person transitive subject (by the prefix o-). The ergative nature of this pattern has been observed for Passamaquoddy by Bruening (2007) but is not otherwise widely discussed. The Cree forms in (2) match the Ojibwe forms in that 1s and 2s subjects are always indexed by the prefix (ni-, ki-). The 3rd-person forms differ from Ojibwe, however: a 3p subject is marked by the -ak suffix regardless of whether the verb is transitive. In Cree, then, it is not just intransitive 3rd-person subjects that are marked like objects—instead, all 3rd-person subjects are marked like canonical objects. This is not an ergative pattern, although it is interesting in its own right. I will show that the difference between the Ojibwe and Cree patterns can be explained if we posit a minor change in the features that agreement is sensitive to. I assume that the two agreement slots (prefix and suffix) reflect two distinct agreement probes. In the Ojibwe transitive 3p—3′ form (o-V-an), the two probes agree with separate arguments (3p and 3′). In the Ojibwe intransitive 3p form, where there is only a single 3rd-person argument, only the suffix appears (3p -ag). I conclude from this that when the prefix and suffix compete to agree with a single 3rd-person argument, it is the suffix probe that wins. This is the source of the split-ergative pattern. The pattern is disrupted in Cree, I propose, because the Cree suffix probe is more selective than its Ojibwe counterpart: it cannot agree with obviative DPs and thus overrides the prefix to agree with the 3p subject in a transitive 3p—3′ form, just as it does in an intransitive 3p form.

This analysis enables an elegant diachronic account of both systems. In Proto-Algonquian, suffix agreement was sensitive to definiteness, a pattern retained in Delaware (Goddard 2007). Ojibwe simply lost this sensitivity while Cree shifted to being sensitive to obviation. 1 3p forms are shown for clarity (the 3s suffix is sometimes null). The notation 3′ indicates an obviative 3rd person.

(2) Cree subject agreement (Wolfart 1973) a. TRANSITIVE (DIRECT) b. INTRANSITIVE 1s—3p ni- V -ak 1s ni- V 1 3p 1

2s—3p ki- V -ak 2s ki- V 2 3p 2

3p—3′ V -ak 3p V -ak 3p 3p

(1) Ojibwe subject agreement (Valentine 2001) a. TRANSITIVE (DIRECT) b. INTRANSITIVE 1s—3p ni- V -ag 1s ni- V 1 3p 1

2s—3p gi- V -ag 2s gi- V 2 3p 2

3p—3′ o- V -an 3p V -ag 3 3′ 3p

References Bruening, Benjamin. 2007. Passamaquoddy as a split ergative language and its consequences for

Marantz’s ergative case generalization. Manuscript, University of Delaware. Goddard, Ives. 2007. Reconstruction and history of the independent indicative. In Papers of the

38th Algonquian Conference, ed. by H. C. Wolfart, 207–271. Winnipeg: University of Manitoba.

Valentine, J. Randolph. 2001. Nishnaabemwin Reference Grammar. Toronto: University of Toronto Press.

Wolfart, H. C. 1973. Plains Cree: A Grammatical Study. Transactions of the American Philosophical Society, new series, volume 63, part 5. Philadelphia.

Statistical co-occurrence restrictions in Oromo consonants Avery Ozburn, University of Toronto

In this work, I examine a novel case of laryngeal co-occurrence restrictions in Oromo, a Cushitic language spoken in Ethiopia. This study reveals that Oromo has both ejective harmony and voicing harmony on a statistical level. This case of laryngeal harmony has never before been described in Oromo, nor in other Cushitic languages. Since laryngeal harmony is so rare cross-linguistically, this research adds an important new case to the literature. Moreover, due to the statistical nature of the harmony, and in particular the statistical directionality patterns, Oromo raises crucial questions about the place of non-categorical patterns in phonology.

Oromo has laryngeal contrasts in stops/affricates that vary from a two-way contrast for bilabials (/p’, b/), to a three-way contrast for post-alveolars (/tʃ, tʃ’, dʒ/) and velars (/k, k’, g/), to a four-way contrast for dentals (/t, t’, d, ɗ/) (Gamta 1989). The purpose of this research was to determine whether there are harmonic restrictions on the distribution of these consonants. Laryngeal harmony systems are rare, but common in other languages of Ethiopia, such as Chaha (Gallagher 2010). Since laryngeal harmony tends to involve only morpheme-internal restrictions (Hansson 2001, 2010; Rose and Walker 2004), it is easy for it to be missed, particularly when it is statistical. In this work, I look for these restrictions in Oromo words of shape CV(C)CV.

Specifically, my research tested whether Oromo voiceless stops (plain and ejective) are required to agree in [constricted glottis], as well as whether all Oromo stops are required to agree in voicing. Results show that these different categories of Oromo stops are not randomly distributed; chi-square statistics on laryngeal categories of first and second consonants show a very high level of significance (p<0.00001 in both cases). Observed over expected (O/E) values given below show that agreement is highly over-represented, while disagreement is under-represented. Thus, Oromo has statistical laryngeal harmony. However, it is not categorical, since the disagreeing cases do not have O/E values of 0, and there are many examples of disagreement.

Ejective harmony Voicing harmony C2 Ejective C2 Plain C2 Voiced C2 Voiceless C1 Ejective 1.53 0.35 C1 Voiced 1.46 0.57 C1 Plain 0.22 2.00 C1 Voiceless 0.38 1.58

This statistical harmony in Oromo has several major implications. In addition to a statistical tendency towards agreement, there are further statistical asymmetries in Oromo that mirror patterns that are categorical in other languages. In particular, the O/E values above show that within the disagreeing cases, there is greater under-representation when C2 is ejective for ejective harmony and when C2 is voiced for voicing harmony. For the voicing case, there is a notable trend towards significance (p<0.08). This result mirrors a categorical regressive effect in languages such as Ngizim, where sequences of voiceless stops followed by voiced stops are forbidden but the opposite order is allowed (Hansson 2001, 2010). Thus, in addition to the overall trend towards harmony, there is a statistical directionality effect that requires explanation.

With such results, I consider the larger question of what implications these effects have for phonology, which is often considered categorical in nature. If we want to understand how cross-linguistic regressive directionality arises, it is crucial to look at statistical cases like Oromo, since they suggest that such biases may go much deeper than accounts of categorical harmony would suggest. While implications of statistical consonant harmony to phonological theory have been considered previously (e.g. Brown 2008), the present research is novel in that it looks at asymmetries like directionality within a statistical pattern. Thus, by examining this novel case of laryngeal harmony, I conclude that Oromo suggests a need to understand statistical consonant co-occurrences in order to fully understand the categorical patterns usually treated by phonology.

References: Brown, Jason Camy. 2008. Theoretical aspects of Gitksan phonology. PhD dissertation,

University of British Columbia. Gallagher, Gillian. 2010. Perceptual distinctness and long-distance laryngeal restrictions.

Phonology, 27(03): 435-480. Gamta, Tilahun. 1989. Oromo-English Dictionary. AAU Printing Press. Hansson, Gunnar Ólafur. 2001. Theoretical and typological issues in consonant harmony. PhD

dissertation, University of California, Berkeley. Hansson, Gunnar Ólafur. 2010. Consonant harmony: long-distance interaction in phonology.

Berkeley, CA: University of California Press. Rose, Sharon, and Rachel Walker. 2004. A typology of consonant agreement as correspondence.

Language 80: 475-531.

The L3 acquisition of Spanish rhotics by native Mandarin speakers Matthew Patience

University of Toronto The L2 acquisition of the two Spanish rhotics (the tap /ɾ/ and the trill /r/) by native English speakers has received considerable attention in recent years (e.g., Face 2006; Olsen 2012). These studies have found that the tap is easier to acquire than the trill, which is partly due to positive transfer of the English flap, a highly similar sound (Colantoni & Steele 2008; Olsen 2012), as well as the more demanding aerodynamic constraints involved in producing the trill (Face 2006; Johnson 2008). Very few studies have investigated the acquisition of the Spanish rhotics by speakers of other languages (but see Rafat 2008 for the acquisition of Spanish /r/ by native Farsi speakers). In the present study, I examine the acquisition of the Spanish rhotics by native Mandarin speakers who speak English as an L2. Mandarin has one rhotic consonant (a voiced rhotic approximant/fricative; Duanmu 2007) which differs greatly from the Spanish rhotics (in place and manner). Mandarin also has a voiced alveolar stop and a voiced dental lateral, two segments that bear some similarity to the tap both perceptually and articulatorily (in terms of place and voicing). Such similarities could impede acquisition (Flege 1995). In contrast, native Mandarin speakers who have acquired the English flap may have an advantage producing the nearly identical Spanish tap. Based on these facts, I investigate the following two questions: (1) What are the developmental stages in the acquisition of the tap and the trill by native Mandarin speakers? (2) How does transfer from the learners’ L1 Mandarin or L2 English affect acquisition of the Spanish rhotics? 9 L1 Mandarin-L2 English-L3 Spanish speakers of beginner, intermediate and advanced Spanish proficiency were recorded performing a word repetition task in Spanish and English. English target stimuli consisted of words containing the English flap (e.g., [ˈwɑ.ɾə.ɹ] water) to determine whether or not the Mandarin speakers have acquired the flap. The Spanish target stimuli consisted of both rhotics in intervocalic position (e.g., ['ka.ɾo] caro ‘car’; ['pe.ro] perro ‘dog’). Duration, voicing, and manner of the Spanish rhotics and the English flap were measured and compared to the values of the control speakers. Preliminary results for the learners’ realization of the tap indicate their difficulty articulating the target manner, as 2 of 3 participants tended to produce a brief approximant (80% occurrence) in place of the tap (17% occurrence). Interestingly, the duration of such approximants matches that of a native-like tap and the segment is perceptually very similar to a tap. The one Mandarin speaker who produced the tap with some regularity (50% accuracy) also consistently produced an accurate English flap, while the other learners demonstrated difficulty articulating the flap. Preliminary results for the trill indicate that Mandarin speakers experience relatively more difficulty acquiring the trill. Only one participant was able to produce trills that were native-like along all three parameters. The other two participants produced a variety of non-target segments – approximants (70%), stops (7.5%), fricatives (7.5%), and taps (15%). The fact that Mandarin speakers produce a non-target yet perceptually similar segment in place of the tap suggests that learners may aim for native-like percepts as opposed to native-like articulations. It also appears that the Mandarin rhotic is more perceptually distinct than the English rhotic: English L2 Spanish learners tend to substitute their L1 rhotic for the tap and the trill until they acquire the target manner (Face 2006; Johnson 2008), yet no Mandarin-like rhotics were observed. L2 English-based transfer was also observed. Only the learner who was able to produce a native-like English flap was able produce a native-like Spanish tap and this only

occurred in post-tonic syllables, the same context in which the English flap occurs. This suggests that the presence of L2 allophones facilitate the acquisition of L3 phonemes, provided the phonetic contexts are the same.

References Colantoni, L., & Steele, J. (2008). Integrating articulatory constraints into models of second

language phonological acquisition. Applied Psycholinguistics, 29(3), 489–534. Duanmu, S. (2007). The phonology of standard Chinese. Oxford, UK: Oxford University Press. Face, T. L. (2006). Intervocalic rhotic pronunciation by adult learners of Spanish as a second

language. In C. A. Klee & T. L. Face (Eds.), Selected proceedings of the 7th Conference on the Acquisition of Spanish and Portuguese as First and Second Languages (pp. 47-58). Somerville, MA: Cascadilla Press.

Flege, J. E. (1995). Second language speech learning: Theory, findings, and problems. In W. Strange (Ed.), Speech perception and linguistic experience: Issues in cross-language research (pp. 233-277). Baltimore, MD: York Press.

Johnson, K. E. (2008). Second language acquisition of the Spanish multiple vibrant consonant. (Doctoral Dissertation). The University of Arizona, Tucson, AZ.

Olsen, M. K. (2012). The L2 acquisition of Spanish rhotics by L1 English speakers: The effect of L1 articulatory routines and phonetic context for allophonic variation. Hispania, 95(1), 65-82.

Rafat, Y. (2008). The acquisition of allophonic variation in Spanish as a second language. In S. Jones (Ed.), Proceedings of the Annual Conference of the Canadian Linguistic Association (pp. 1-15). Vancouver, BC.

Variation in Ejective Stops in Harar Oromo

Maida Percival, University of Toronto

Acoustic studies on ejectives have yielded inconsistent results, so that it is still unclear

how systematically their acoustic characteristics define them cross-linguistically. Kingston (1985) proposed a two-way typological classification system of stiff and slack ejectives based on differences in measurements of voice onset time (VOT), burst amplitude, and F0 perturbation, jitter perturbation, and rise time in intensity of the following vowel.

However, more recent studies have found this dichotomy to be problematic. Warner (1996) compared six acoustic measurements of ejectives in Ingush with 4 other languages, to find that no two languages patterned the same. Further studies have shown that even within a single language, ejectives differ in whether they would be traditionally classified as stiff or slack. Wright et al. (2002) found inter-speaker variation in the acoustic characteristics of the Witsuwit’en ejectives, suggesting that different speakers employ different strategies to differentiate ejectives from plosives. Vicenik (2010) also found variation in the acoustic characteristics of Georgian ejectives: intra-speaker, positional variation due to syntagmatic strengthening of consonants in higher prosodic positions. This brings yet another aspect into the complexities of describing ejective consonants.

Based on these findings, this paper investigates the acoustic characteristics of ejectives and their pulmonic counterparts in Harar Oromo, a Cushitic language. It hypothesizes that Oromo ejectives are distinguished from plosives by a combination of the acoustic characteristics that differentiate the two laryngeal types in other languages. It also hypothesizes that, following Wright et al. (2002) and Vicenik (2010), the characteristics will show variation across speakers and word-positions.

To test this hypothesis, acoustic analysis of word-initial and word-medial bilabial, dental, and velar ejective, aspirated, and voiced stops of eight speakers of Eastern Oromo is being done. Measurements of the consonants’ VOT, closure duration, and burst amplitude are being made as well as measurements of F0 perturbation, rise time, and jitter perturbation of the consonants’ following vowels.

Preliminary results indicated that, overall, VOT significantly differentiates ejective, aspirated, and voiced stops. Ejectives had the longest VOT, followed by aspirated stops, then voiced stops, which have negative VOT. Consistent with Vicenik (2010), positional variation was also apparent in VOT, with word-initial stops having greater VOT. Rise time results did not significantly differentiate the stops’ laryngeal type, but did show significant positional variation. The intensity of the vowels following the stops rose more slowly word-initially than word-medially. F0 perturbation was also measured and results indicated that pitch was significantly higher following ejectives than following voiced and aspirated stops, despite Oromo using tone for grammatical purposes. Results for burst amplitude and jitter perturbation are forthcoming, as are more detailed statistical analyses for each measurement for individual speakers.

The results of this study provide detailed phonetic descriptions of consonants in Harar Oromo, which did not previously exist for the language. These data also contribute to the typology of ejectives by helping us understand what sort of cross-linguistic and intra-language variation exists in the acoustic characteristics that differentiate ejective from pulmonic stops so that a more refined account of ejectives can be devised.

References Kingston, John. 1985. The phonetics and phonology of the timing of oral and glottal events.

Berkeley, CA: UC Berkeley dissertation. Maddieson, Ian. 2011. “Glottalized Consonants”. In Dryer, Matthew S. & Haspelmath, Martin

(eds.). The World Atlas of Language Structures Online. Munich: Max Planck Digital Library, chapter 7.

Nelson, Catherine. 2010. “Ejectives in Nez Perce.” Santa Barbara Papers in Linguistics: 21. Vicenik, Chad. 2010. “An Acoustic Study of Georgian Stop Consonants”. Journal of the

International Phonetic Association: 40(1), 59-92. Warner, Natasha. 1996. “Acoustic Characteristics of Ejectives in Ingush”. Proceedings of

the International Conference on Spoken Language Processing, 1525-1528. Philadelphia: Pennsylvania.

Wright, Richard, Hargus, Sharon & Davis, Katharine. 2002. “On the categorization of ejectives: data from Witsuwit'en”. Journal of the International Phonetic Association: 32(1), 43-77.

Adaptation à l’accent français européen par une comédienne québécoise : changements

acoustiques des voyelles et perception des accents québécois et hexagonal

François Poiré, Jeff Tennant & Antony Cloutier, The University of Western Ontario

Depuis au moins deux décennies, un certain nombre d’acteurs québécois poursuivent

aussi une carrière en France, en particulier au cinéma. Dans la vaste majorité des cas, et

ce contrairement à ce qui se passait antérieurement, ils ne jouent pas le Québécois de

passage en sol européen mais bien un personnage français parmi d’autres. Dans cette

étude de nature sociophonétique et perceptuelle, nous comparons les deux ‘accents’ de

l’actrice québécoise Marie-Josée Croze selon la variété de français utilisée d’un côté ou

de l’autre de l’Atlantique pour les besoins d’une production cinématographique donnée.

Nous avons choisi quatre films (deux productions canadiennes et deux européennes) et

avons extrait et retranscrit les trames sonores de ses personnages. Le travail d’analyse se

fait ensuite en deux étapes. Dans un premier temps, nous comparons la réalisation des

voyelles (structure formantique, durée et désonorisation) dans les deux variétés à l’aide

du logiciel Praat. Cette étude acoustique permet d’établir la dispersion formantique des

voyelles orales dans les deux variétés utilisées et de porter une attention particulière à

certains phénomènes phonétiques tels le relâchement des voyelles fermées en syllabes

fermées ([vIt] au lieu de [vit], vite, la réduction et la syncope totale de la même classe de

voyelles (dans les mêmes contextes qui favorisent la syncope du schwa) et le maintien de

certaines oppositions comme [a] et [ɑ] et [ɛ] et [ɛ:] (mettre et maître), typiques de

l’accent québécois et pratiquement disparues en France (Tranel, 1987, Walker 2001).

Dans un second temps, une tâche de perception construite à partir d’extraits de ces

mêmes bandes sonores est menée auprès de sujets tant québécois que français. Ces courts

extraits, variant du mot simple à la courte phrase, couvrent la totalité des systèmes

vocaliques des deux variétés et contiennent aussi les contextes demandant le plus

d’attention lorsqu’il s’agit de masquer un accent québécois (ou si l’on préfère, de

produire un accent français). L’analyse acoustique montre des valeurs de F1 réalisées

dans une bande de fréquence plus étroite dans les rôles européens, corrélat d’une aperture

moins variable tandis que F2 présente un système vocalique fortement antériorisé dans

les mêmes films, à l’exception de certaines voyelles postérieures encore plus

postériorisées. La durée de ses voyelles européennes montre aussi beaucoup moins de

variation. Les résultats du test de perception indiquent que les sujets tant européens que

québécois identifient clairement les deux accents de cette actrice. Nous discutons ces

résultats en tentant de répondre à la question suivante : Passe-t-on d’un ‘accent’ à l’autre

en éliminant des traits dialectaux, en réalisant des cibles articulatoires étrangères à notre

accent initial ou encore par un mélange des deux stratégies?

TRANEL, B., The sounds of French, Cambridge, Cambridge University Press, 1987.

WALKER, D. C., French sound structure, Calgary, University of Calgary Press, 2001.

Malay/Indonesian agent pseudo-incorporation and ditransitive clauses Paul Poirier – University of Toronto

This paper will explore the idiosyncratic voice system found in both standard Malay and Indonesian and explain it in a novel way using pseudo-incorporation (Massam 2001) and cliticization; it will also test analyses against ditransitive data, which has not been considered in previous accounts. The language exhibits an active clause (1), as well as two passive-like constructions, which I will call “passive” (2) and “object voice” (3), following Cole et al. (2008):

(1) Active: kami tidak akan mem-baca buku ini 1PL not will meN-read book this ‘We will not read this book.’ (2) Passive: buku ini tidak akan di-baca (oleh Siti) book this not will Pass-read by Siti ‘This book will not be read by Siti. (3) Object voice: buku ini tidak akan kami baca book this not will 1PL read. ‘This book will not be read by us.’ (adapted from Cole et al. 2008)

The active is problematic as the marker on the verb, ‘meN-,’ is optional. While the passive can be derived in the usual way, the object voice is unique in that the preverbal agent is restricted to pronominal forms, a property that has been largely ignored in previous accounts, which all solely use properties of v to explain the voice alternations. Both Cole et al. (2008) and Aldridge’s (2008) primary research goals are to account for A’-extraction in the language, which is limited to the surface subject, i.e. the agent in the active and the patient in other two constructions (see Chung 1976 for extensive evidence); while other arguments cannot extract, extraction of adjuncts is possible. Cole et al. (2008) adopt an agreement approach, arguing that v agrees with either the agent (‘meN-’) or the patient (a null affix) in their lexically marked Case features. Other arguments, whose Case clashes with v, cannot raise, while Caseless adjuncts (adverbs and PPs), can freely do so. Functional heads usually agree not in Case but other phi-features such as person, number, and gender; an account without a language-specific form of agreement would be preferable.

For Aldridge (2008), Indonesian exhibits an unstable mixed voice system, in transition between the ergative system found in other Austronesian languages (object voice) a nominative-accusative alignment (the active and passive); extraction asymmetries are accounted for by claiming that the active is derived from an old antipassive and that it retains some of its idiosyncratic properties, though the analysis offers no diachronic evidence for this. Additionally, neither analysis has a satisfying account for the optionality of ‘meN-’.

All analyses limited to a functional voice projection also have no explanation as to why the agent must be pronominal in the object voice. Here it will be argued that the object voice is generated by pseudo-incorporation of the agent (Massam 2001, Levin 2014); that it cannot be extracted in the object voice follows trivially from this new analysis. Following Fortin (2006), I will treat the active “prefix” ‘meN-’ as an object clitic, with the lexical patient dislocated: it is for this reason that it cannot extract. While this analysis makes some interesting predictions that interact nicely with Hopper and Thompson’s (1980) transitivity diagnostics, I will present novel ditransitive data that still challenge for all analyses of meN-’, and discuss possible adaptations to Fortin (2006). The present analysis is superior in that these additional facts are explained without attributing ad hoc properties to the functional heads responsible for voice.

Sources

Aldridge, Edith. 2008. Phase-based account of extraction in Indonesian. Lingua 118: 1440-1469. Cole, Peter; Hermon, Gabriella; Yanti. 2008. Voice in Malay/Indonesian. Lingua 118: 1500-1553. Chung, Sandra. 1979. On the subject of two passives in Indonesian. In Li, C. (ed.), Subject and

Topic. Academic Press, New York: 57-98. Fortin, Catherine. 2006. Reconciling MENG- and NP movement in Indonesian. Paper presented at

the 32nd Annual Meeting of the Berkeley Linguistics Society, University of California Berkeley.

Hopper, Paul & Thompson, Sandra. 1980. Transitivity in Grammar and Discourse. Language 56(2): 251-299.

Levin, Theodore. 2014. Pseudo-Noun Incorporation is M-Merger: Evidence from Balinese. Paper presented at the LSA Annual Meeting, Minneapolis.

Massam, Diane. 2001. Pseudo Noun Incorporation in Niuean. Natural Language and Linguistic Theory 19(1): 153-197.

Redundant Adjective is Redundant: An analysis of Adj1 N is Adj1

James Porteous

In popular English, and particularly English used online, there is a copular construction involving seemingly redundant adjectives, following the pattern of Adj1 N is Adj1, as seen in (1). These sentences include a noun phrase containing an adnominal adjective, a copula, and the same adjective in the predicate position. The goal of this paper is to discuss the semantics and uses of this construction which, to my knowledge, has received no attention in the literature.

1. Happy cat is happy.

My presentation makes two central claims about these sentences: The adnominal adjective serves to identify the NP subject; and the presence of the copular adjective allows at least three possible interpretations, which arise from the pragmatic context. (Thus, contra my title, the copular adjective is not redundant after all.) The three possible interpretations for sentence (1) are:

2. a. Simple Predication: The cat is happy. b. Intensification: The cat is very happy. c. Intrinsic-Temporal: The cat, which is normally happy, is currently happy.

It is important to note that, regardless of interpretation, these sentences occur in contexts where both the speaker and the listener are aware of the referent in question and treat it as ‘given’ knowledge, fresh in their minds. The subject NP is therefore definite in all of these interpretations, despite not having an overtly realized definite determiner. This definiteness appears to relate to the presence the adnominal adjective, and we can see its effect through its absence: ‘Cat is happy’ becomes ungrammatical and begs the question of ‘which cat?’. Similarly, a sentence with a plural noun such as (3) has a definite meaning for that noun (‘The successful rebels are successful’), whereas without the adnominal adjective the subject NP would be indefinite (‘Rebels are successful’). From this I conclude that the adnominal adjective allows us to clearly identify the NP subject.

3. Successful rebels are successful.

These sentences can take on what I call an intrinsic-temporal reading in contexts where both the speaker and the listener are fairly well-aware of the noun’s referent and its qualities. (The intrinsic quality corresponds approximately to Carlson’s (1977) individual-level predication in this reading; the temporary quality to stage-level predication.) An intensification reading comes from contexts where the referent is present or was recently present during the discourse; this interpretation is very difficult without the referent present. Such an intensification reading is similar to the reduplication intensification described by Jackendoff (2004), but with a distinct syntax. When the referent is not immediately present during the discourse, or when the adjective is binary and cannot be readily intensified (e.g. dead, unconscious), the most natural interpretation is a simple predicative interpretation.

In my presentation I will present attested examples used in context to show how the readings can be teased out of the different parameters that make up that context. To discuss the pragmatics where these sentences are used I will make reference to Grice’s (1975) four maxims and on implicature in general. In so doing I hope to shed light on the importance of pragmatics to the interpretation of copular sentences.

References

Carlson, Greg. 1977. Reference to Kinds in English. University of Massachusetts/Amherst

doctoral dissertation.

Grice, Paul (1975). "Logic and conversation". In Cole, P.; Morgan, J. Syntax and semantics. 3: Speech acts. New York: Academic Press. pp. 41–58.

Jackendoff, Ray, Jila Ghomeshi, Nicole Rosen, and Kevin Russell. 2004. “Contrastive Focus in English (The Salad-Salad Paper).” Natural Language & Linguistic Theory 22: 307–357.

The perception of intonational contours: a cross-linguistic study Malina Radu, Gabrielle Klassen, Laura Colantoni,

Matthew Patience, Ana Teresa Pérez-Leroux and Olga Tararova University of Toronto

Cross-linguistically, intonation is used to convey a wide-range of linguistic information, such as the type of sentence (interrogative vs. declarative) (Liu & Rodríguez 2012). Tonal and durational variations can be used to mark statements (falling contour), questions (rising contour) or that the speaker will continue speaking (mid-rising contour). To some extent, Spanish is similar to English in the use of intonation to differentiate between interrogatives and declaratives. In both languages, falling and rising contours at the end of constituents are mapped to similar meanings (e.g. Hualde 2002; 2005). In spite of the typological similarities between English and Spanish, however, some aspects of sentence prosody are not mastered even by advanced L2 speakers. Grabe et al (2003) found that linguistic meaning diminishes the discrimination capacity of L2 listeners, since speakers from different L1s discriminated intonation contours in non-speech stimuli equally well, while their discrimination of intonation contours of English sentences worsened; this suggests that while keeping their auditory resolution to intonation cues in non-speech tasks, adult speakers failed to relate these contours to appropriate L2 meanings.

Based on previous studies (e.g. Grabe et al. 2003), we expect native and non-native English speakers to behave similarly in their discrimination of intonational contours in non-meaning-related tasks, but to diverge in their capacity to perceive language-specific (English) form-meaning mappings. To test this hypothesis, we compared the perception of statements, and inverted and non-inverted interrogatives across two groups of speakers: L2 English-L1 Spanish vs. L1 English (N=15 per group). Participants completed three tasks (administered with Super Lab Pro) as well as a language background questionnaire. In Task 1 (forced-choice identification) participants had to indicate whether the low-pass filtered stimulus (N=30 plus distractors) was a statement, question or exclamation. This task was designed to test participants’ ability to use acoustic cues. Task 2 was identical to Task 1 but stimuli consisted of unaltered sentences. In Task 3, participants heard a scenario and three options, and subsequently had to choose one answer that best completed the scenario. Measurements included the mean accuracy rates per language group (L1 English and L1 Spanish) for each of the three tasks, as well as the reaction time (RT).

Preliminary results showed that overall, the L1 English speakers made fewer perception errors than the L1 Spanish speakers across all three tasks. As for the individual tasks, both groups of speakers had the highest accuracy for Task 2 (L1 English: .97; L1 Spanish: .67), followed by Task 1 (L1 English: .90; L1 Spanish: .58). As for Task 3 (i.e. the most contextualized task), both groups had the lowest accuracy rate (L1 English: .89 vs. L1 Spanish: .54), and the longest reaction times (participants took four times longer to answer than in the other two tasks). Confusion matrices also revealed interesting patterns: while L1 English speakers tend to confuse questions with statements, L1 Spanish speakers are equally likely to misinterpret questions as either exclamations or statements. These findings are consistent with the hypothesis that perception difficulties in L2 prosody are not primarily related to any loss of sensory capabilities but to the modulation of the listeners’ sensitivity to acoustic cues by selective attention to meaningful units, and confirm previous findings that L2 speakers are more successful in perception tasks where there is no context versus those in which a context is provided (e.g. Ortega Llebaria & Colantoni 2013). Results are also consistent with findings that L1 affects the perception of linguistic and auditory processing alike (Kuhl et al. 2008).

References Grabe, E., Rosner, B., García-Albea, J., & Zhou, X. (2003). Perception of English intonation by

English, Spanish and Chinese Listeners. Language and Speech, 46, 375-401. Hualde, J. I. (2002). Intonation in Romance: Introduction to the special issue. Probus, 14, 1-7. Hualde, J. I. (2005). The sounds of Spanish. Cambridge: Cambridge University Press. Kuhl, P. K., Conboy, B. T., Coffey-Corina, S., Padden, D., Rivera-Gaxiola, M., & Nelson, T.

(2008). Phonetic learning as a pathway to language: new data and native language magnet theory expanded (NLM-e). Philosophical Transactions of the Royal Society, 363, 979-1000.

Liu, C., & Rodriguez, A. (2012). Categorical Perception of Intonation Contrasts: Effects of Listeners’ Language Background. The Journal of the Acoustical Society of America, 131(6), 427-433.

Ortega-Llebaria, M., & Colantoni, L. (2013). The L2 acquisition of English intonation: Relations between form-meaning associations, access to meaning and L1 transfer. Studies in Second Language Acquisition, 36(2).

An evidential modal in Bulgarian: the inferential future. María Luisa Rivero and Vesela Simeonova (University of Ottawa).

In Bulgarian, invariable shte (FUT), also used in ordinary futures like English will, may have an evidential meaning. It may then signal a present inference about a present event when followed by a verb with present morphology, (1), or a present inference about a past event when followed by a present perfect: (2-3). (1) Context: Your friend asks which one among 3 singers in a photo won a TV context. You listen to a tape where each singer sings. Pointing to one singer in the photo you state: Tazi shte e pobeditelkata. This.Fem FUT be.Present.3Sg winner.Sg.Fem.the ‘This one must be the winner.‘ (2) Context: You wonder why Ivan never went to Paris. Since his mom lives there, you suppose that she often told him to visit. You state: Tja shte (da) mu e kazvala mnogo she FUT (da) him.DAT be.Present.3SG told.PP.Impf. many pati da ja poseti. times da her.ACC visit.Present.3SG 'She must have told him to visit her many times.' (3) Maria smjatashe che Petar shte (da) Maria consider.Ind.Imperfect.3Sg that Petar FUT (da) e iztarpjal mnogo prez vojnata. be.Pres.3Sg endure.PP.Perf a.lot during war.the ‘Maria thought that Peter must have endured a lot during the war.’ Inferential shte in (1-3), noted in descriptive grammars ([10], [11], [14], a.o.), is not discussed in the literature on Bulgarian evidentials ([5], [7], [13], [15], [16] a.o.). In this paper, we will examine its morphosyntax and semantics, arguing that they are of interest for current theories of modality and evidentiality, and for comparative purposes, since the properties of shte resemble at the same time those of must ([3]), the Greek future marker tha, and Romance future affixes ([4] a.o.).

We argue for three points. First, shte is an evidential with the properties of a modal, not an illocutionary marker: (a) it can be embedded under propositional attitude verbs: (3); (b) it can be embedded under question operators (not illustrated). Illocutionary evidentials lack these characteristics ([2], [8], [9] a.o.). Second, shte lacks a rigid quantificational force. It may express a high level of confidence close to certainty, and thus resemble a necessity modal: (1). In some contexts including questions, however, it associates with a level of certainty closer to possibility. Given that (evidential) modals need not be restricted to universal or existential force ([6], [12], [17] a.o.), we explore the idea that shte may be a degree modal. Third, shte takes a propositional complement with tense (including past), and aspect, which makes it resemble both Greek tha and fully inflected epistemic modals in Romance ([4)] and at the same time differ from English epistemic modals within analyses for English of the type in [1]. In sum, a well-known evidential system is a hallmark of Bulgarian ([5], [7], [10], [11], [13], [14], [15], [16], a. o.), but does not exhaust the grammar of evidentiality in this language. An interesting evidential is less known shte, a modal specialized for inferences, which {interacts with/scopes over} the complex morphological and semantic tense and aspect system of Bulgarian.

2

An evidential modal in Bulgarian: the inferential future. Cited references: [1] Condoravdi, C. 2002. Temporal interpretation of modals. The Construction of Meaning. [2] Davis, C. & al. 2007. The pragmatic values of evidential sentences. SALT 17. [3] von Fintel, K. & A. Gillies. 2010. Must … stay…. Strong! Natural Language Semantics 18. [4] Giannakidou, A. & A. Mari, 2013. The future of Greek and Italian: an epistemic analysis. Sinn und Bedeutung 17. [5] Izvorski, R. 1997. The Present Perfect as an Epistemic Modal. SALT 7. [6] Kratzer, A. 2012. Modals and Conditionals. [7] Koev, T. 2011. Evidentiality and temporal distance learning. SALT 21. [8] Matthewson, L. & al. 2007. Evidentials as epistemic modals. Linguistic Variation Yearbook 7. [9] Matthewson, L . 2010. On apparently non-modal evidentials. Empirical Issues in Syntax and Semantics 8. [10] Nitsolova, R. 2008. Bălgarska Gramatika. Sofia. [11] Pashov, P. 1989. Prakticeska balgarska gramatika, Sofia. [12] Rullmann, H. & al. 2008. Modals as Distributive Indefinites. Natural Language Semantics 16. [13] Sauerland, U. & M. Schenner. 2007. Shifting evidentials in Bulgarian. Sinn und Bedeutung 11. [14] Scatton, E. 1983. A reference Grammar of Modern Bulgarian. [15] Smirnova, A. 2012. Evidentiality in Bulgarian. Journal of Semantics. [16] Smirnova, A. 2013. The meaning of the Bulgarian and Turkish evidentials. Contrastive Linguistics 2/3. [17] Yalcin, S. 2007. Epistemic Modals. Mind 116.

Interfaces entre domaines phonétique et phonologiquedans l’acquisition de la phonologie

Yvan Rose Memorial University of Newfoundland

Dans la littérature récente sur l’acquisition de la phonologie par les enfants, on observe des positions parfois très opposées au sujet de la nature des représentations phonologiques et de leur origines. D’une part, Hale & Reiss (1998, 2008) arguent en faveur de l’innéisme des traits, et d’une acquisition de la phonologie par attrition de traits universellement disponible. Par exemple, l’absence de contraste de tension [ATR] d’un système adulte amène à l’attrition de ce trait dans le système de l’apprenant. D’autre part, Vihman et ses collègues rejettent la notion même de trait phonologique, et privilégient une description holistique du développement phonologique, encodé sous forme de gabarits phonologiques (p.ex. Vihman & Croft 2007; Menn & Vihman 2011).

Au cours de cette présentation, je resitue ce débat en prenant comme point de départ le point de vuede l’apprenant : tout enfant devant acquérir sa langue maternelle doit analyser cette langue d’un point de vue perceptuel, et ensuite découvrir les articulations nécessaires à la reproduction des catégories acoustiques identifiées au niveau perceptuel. J’adopte une approche émergentiste de l’acquisition du système phonologique (p.ex. Pierrehumbert 2003; Mielke 2008), sans toutefois rejeter les modèles de représentation formels formulés dans le cadre générativiste (Selkirk 1980; McCarthy & Prince 1986; voir aussi Fikkert 1994 et Goad & Rose 2004 dans le domaine de l’acquisition). Par exemple, au niveau segmental, les représentations phonologiques relient formellement les catégories perceptuelles et leurs catégories articulatoires correspondantes. Ces domaines analogiques (perceptuel et articulatoire) sont donc reliés via leur interface avec le domaine phonologique, lui-même composé de représentations discrètes.

J’illustre cette approche à partir de données d’acquisition disponibles sur la base PhonBank (http://childes.talkbank.org/phon/). Je discute l’acquisition de traits segmentaux, qui émergent à mesure que l’enfant arrive à maîtriser les articulations nécessaires à la reproduction de dimensions acoustiques repérées dans le signal. Par exemple, on note la présence de consonnes fricatives dans le signal acoustiqueà partir de leurs fréquences apériodiques de haute amplitude (p.ex. Ladefoged & Maddieson 1996). L’enfant doit donc identifier cette dimension acoustique, et ensuite maîtriser le mode articulatoire nécessaire à sa reproduction. Comme on peut voir en (1), Inês, une apprenante du portugais européen, a acquis ce trait de manière catégorique, pour tous les lieux d’articulation. Avant l’âge de 2;07.16, Inês produisait virtuellement toutes ses fricative cibles comme des occlusives. Au cours du mois suivant, elle arevu la production de toutes ces consonnes de manière catégorique, encore ici pour tous ses lieux d’articulation cible.

(1) a. Occlusivation (jusqu’à 2;07.16) b. Productions cibles (après 2;07.16)fiz [ˈfiʃ] → [ˈpiʃ] ficas [ˈfikɐʃ] → [ˈfikɐʃ]sim [ˈsĩ] → [ˈtĩ] sim [ˈsĩ] → [ˈsĩ]Julieta [ʒuliˈetɐ] → [dujˈletɐ] Joaquim [ʒuɐˈkĩ] → [ʒwɐˈkĩ]a rua [a ˈʁuɐ] → [a ˈɡuɐ] a rua [a ˈʁuɐ] → [aˈʁwɐ]

Passant ensuite aux patrons de production plus variables, je montre que la majorité d’entre eux proviennent d’interactions avec d’autres patrons de production, tout aussi systématiques. Par exemple, toujours dans les productions d’Inês en (1a), on observe que le patron d’occlusivation n’affecte que les positions d’attaque syllabique (et non les codas).

Je conclus par une discussion de certaines implications formelles. Par exemple, les consonnes approximantes présentent une résonance périodique; elles font donc partie d’une classe naturelle différente de celle des fricatives. Ceci est reflété dans les productions de l’enfant : le patron en (1a) n’affecte ni les liquides, ni les glides (p.ex. [ɾ, l, j, w]). Un trait formel tel que [continu] représente donc un niveau d’abstraction additionnel au sein des représentations.

RéférencesFikkert, Paula. 1994. On the Acquisition of Prosodic Structure. (HIL Dissertations in Linguistics 6). The

Hague: Holland Academic Graphics.Goad, Heather & Yvan Rose. 2004. Input Elaboration, Head Faithfulness and Evidence for Representation in

the Acquisition of Left-edge Clusters in West Germanic. In René Kager, Joe Pater & Wim Zonneveld (eds.), Constraints in Phonological Acquisition, 109–157. Cambridge: Cambridge University Press.

Hale, Mark & Charles Reiss. 1998. Formal and Empirical Arguments Concerning Phonological Acquisition. Linguistic Inquiry 29(4). 656–683.

Hale, Mark & Charles Reiss. 2008. The Phonological Enterprise. Oxford: Oxford University Press.Ladefoged, Peter & Ian Maddieson. 1996. The Sounds of the World’s Languages. Cambridge, MA: Blackwell.McCarthy, John J. & Alan S. Prince. 1986. Prosodic Morphology. In John A. Goldsmith (ed.), The Handbook

of Phonological Theory, 318–366. Oxford: Blackwell.Menn, Lise & Marilyn M. Vihman. 2011. Features in Child Phonology. In G. Nick Clements & Rachid

Ridouane (eds.), Where do Phonological Features Come From?, 261–301. Amsterdam: John Benjamins.

Mielke, Jeff. 2008. The Emergence of Distinctive Features. Oxford: Oxford University Press.Pierrehumbert, Janet B. 2003. Phonetic Diversity, Statistical Learning, and Acquisition of Phonology.

Language and Speech 46(2-3). 115–154.Selkirk, Elisabeth O. 1980. Prosodic Domains in Phonology: Sanskrit Revisited. In Mark Aronoff & Mary-

Louise Kean (eds.), Juncture: A Collection of Original Papers, 107–129. Saratoga, CA: Anma Libri.Vihman, Marilyn M. & William Croft. 2007. Phonological Development: Toward a “Radical” Templatic

Phonology. Linguistics 45(4). 683–725.

Bilingual Lexical Organization: Is there a sensitive period? Laura Sabourin, Jean-Christophe Leclerc, Michele Burkholder & Christie Brien

University of Ottawa A long-standing question in psycholinguistic investigations of bilingual language processing is whether bilinguals have a separate lexicon for each language, or whether they store lexical items from both languages in a single integrated memory system. While there is an emerging consensus for the latter, at least for bilinguals with an early age of L2 acquisition (AoA), there is also some evidence that bilinguals with a late AoA do not have an integrated lexicon (ex. Silverberg & Samuel, 2004). Such results are suggestive that there is a critical or sensitive period for lexical organization. Before such a conclusion can be drawn, however, it is necessary to ensure that other correlated factors, such as L2 proficiency and manner of L2 acquisition (MoA; i.e. naturalistic vs. instructional), are not instead responsible for this difference. In previous studies, these variables are often confounded, as it is difficult to tease them apart. In fact, no study to our knowledge has yet attempted to isolate the effect of MoA.

The goal of the current study is to investigate the question of whether or not AoA can account for differences in lexical organization for groups of bilinguals differing with respect to L2 proficiency and, in particular, MoA. We hypothesize that an early AoA is sufficient but not necessary for an integrated lexicon; a naturalistic MoA may also lead to an integrated bilingual lexicon, as it would enable L2 lexical items to form direct links to the semantic network. In order to investigate this, we are using a lexical decision task with masked priming (Forster & Davis, 1984). In this task, target words are preceded by subliminally presented prime words; in critical trials, these primes are the translation equivalent of the target. Translation priming effects (TPEs) are taken as evidence that words from both languages access an integrated lexicon (Altarriba & Basnight-Brown, 2007). Crucially, when participants are tested with L2 primes and first language (L1) targets, this TPE is sensitive to factors such as AoA and L2 proficiency (Duñabeitia et al., 2010; Dimitropoulou et al., 2010; Sabourin et al., in press).

A study conducted in our lab testing English-French bilinguals in this L2-to-L1 priming direction found that both simultaneous bilinguals (AoA = birth) and early L2 learners (AoA 3-6 years old) showed significant TPEs, while late L2 learners (AoA > 7 years old) showed none. Importantly, the early and late L2 learners were matched for L2 proficiency, suggesting that only early acquirers have an integrated lexicon, irrespective of their proficiency. The effect of MoA was not investigated here, as the majority of L2 French learners French in this region learn their L2 in instructional settings. In this same region, however, L2 English learners (with French L1) tend to learn their L2 in a more naturalistic way. This fundamental difference between these two groups of bilinguals creates an ideal situation in which to examine the role of MoA.

We will test our hypothesis by comparing the TPEs of 20 late French-English L2 learners to those of the group of late English-French L2 learners already tested. Crucially, these two groups will be matched in terms of AoA and L2 proficiency. We will use the same task with the same critical stimuli, though the languages of the primes and targets will be reversed in order to maintain the L2-to-L1 priming direction. If an early AoA is necessary and sufficient in order to have an integrated lexicon, then we expect these late French-English bilinguals to mirror the null TPEs of the late English-French bilinguals. If, however, we find priming effects for the late French-English bilinguals, then this would suggest that an early AoA is sufficient but not necessary for an integrated lexicon; that is, if the L2 is acquired with a naturalistic MoA, even late learners can have an integrated lexicon. These results will make an important contribution to the critical/sensitive period debate and will have implications for models of the bilingual lexicon.

REFERENCES

Altarriba, J., & Basnight-Brown, D. M. (2007). Methodological considerations in performing semantic and translation priming experiments across languages. Behavior Research Methods, Instruments, & Computers, 39, 1–18.

Dimitropoulou, M., Duñabeitia, J. A., & Carreiras, M. (2011). Masked translation priming

effects with low proficient bilinguals. Memory & Cognition, 39, 260-275. Duñabeitia, J. A., Perea, M. & Carreiras, M. (2010). Masked translation priming effects with highly

proficient simultaneous bilinguals. Experimental Psychology, 57, p. 98-107.

Forster, K. I., & Davis, C. (1984). Repetition priming and frequency attenuation in lexical access. Journal of Experimental Psychology: Learning, Memory, & Cognition, 10, 680-698.

Sabourin, L., Brien, C., & Burkholder, M. (in press). The effect of age of L2 acquisition on the

organization of the bilingual lexicon: Evidence from masked priming. Bilingualism: Language and Cognition.

Silverberg, S., & Samuel, A. G. (2004). The effect of age of second language acquisition on the

representation and processing of second language words. Journal of Memory and Language, 51, 381–398.

Daniel Schmidtke and Victor Kuperman

Mass counts in the World Wide Web: A corpus linguistic study of noun countability across varieties of English

This research uses a corpus of global Internet English to explore the variation in the pluralisation of mass nouns (e.g. luggages, violences, and advices) across varieties of English. The countable use of mass nouns is a lexico-grammatical phenomenon that is hailed as a discernible proxy of the dividing line between native and non-native varieties of English (McArthur, 2002; Mesthrie and Bhatt, 2008). A recent study (Hall, Schmidtke & Vickers, 2013) confirmed this, using Google searches of the World Wide Web as a corpus linguistic tool. They revealed a significantly higher concentration of countable usage of 25 mass nouns among non-native L2 English users, compared to native L1 users of British English. In the current study we employed a less restricted data-driven methodology, with which we were able to probe the extent of ‘mass noun’ countability across an expansive list of nouns, and among a greater variety of Englishes.

We queried Davies’ (2013) 1.9 billion-word corpus of Global web-based English (GloWbE), for the raw frequencies of 17,757 singular noun lemmas and their plural counterparts. GloWbE represents 20 samples of English from 7 native-English speaking, ‘Inner Circle’ countries (e.g. Canada, Britain and Australia) and 13 ‘Outer Circle’ countries (e.g. India, Hong Kong and Tanzania). We identified nouns that occurred significantly more frequently in plural form in the Outer Circle compared to the Inner Circle, irrespective of the atomic (countable/mass) quality of each noun’s referent. Once these nouns were isolated, we explored the underlying semantic and morpho-syntactic causes of the dissimilarity in countability preferences across Inner and Outer Circle Englishes.

Firstly, we found a significant convergence of the nouns used more countably in the Outer Circle with nouns that are routinely cited in the literature as grammatically ‘mass’. This list represents a widespread countable usage of mass nouns, such as equipments, softwares and slangs, in Outer Circle Englishes. These nouns were also part of Hall et al.’s (2013) pre-defined mass noun word-list. Secondly, we observed a large number of previously unattested cases of pluralisation of mass nouns, more commonly in the Outer than in the Inner Circle. Using Latent Semantic Analysis (Landauer and Dumais, 1997), a computational method of calculating semantic distances between words, we found that the categories of nouns that exhibit countability preferences in Outer Circle English are semantically predictable. For example, plural forms of nouns that denote non-individuated concepts reliably cluster into the semantic category of ‘occupational terminology’, such as: assistances, trainings, welfares, recruitments, and taxations. In addition, the following nouns represent the semantic category of ‘written language paraphernalia’: alphabets, graphites, handwritings, mails and punctuations. Furthermore, we note an overrepresentation of abstract nouns with Latinate morphology used countably in the Outer Circle, such as acclaims, ascendants, destructions and servitudes.

Taken together, our results provide further confirmation of the heterogeneity of noun countability behaviour across the Inner Circle and Outer Circle varieties of English. We adopt the notion that the observed variance in noun countability supports the conception of English as dynamic and plurilithic entity (cf. Pennycook, 2009). Moreover, we report an original finding that suggests that the semantic properties of the nouns inform the unconscious cognitive processes involved in marking nominalizable concepts with plural morphology. Our method was blind to the theoretically grounded grammatical count/mass distinction, and, given the pattern of our results, we propose that the fixed binary distinction

of count/mass is not an essential component in a theory of English language structure. Instead, it is a phenomenon best viewed as a gradient that is semantically and regionally dependent.

References

Davies, M. (2013). Corpus of Global Web-Based English: 1.9 billion words from speakers in 20 countries. Available online at http://corpus2.byu.edu/glowbe/.

Hall, C. J., Schmidtke, D., & Vickers, J. (2013). Countability in world Englishes. World

Englishes, 32(1), 1-22.

Landauer, T. K., & Dumais, S. T. (1997). A solution to Plato's problem: The latent semantic analysis theory of acquisition, induction, and representation of knowledge. Psychological review, 104(2), 211.

McArthur, T. (2002). The Oxford guide to world English. Oxford: Oxford University Press.

Mesthrie, R., & Bhatt, R. M. (2008). World Englishes: The study of new linguistic varieties. Cambridge University Press.

Pennycook, A. (2009). Plurilithic Englishes: towards a 3D model. Global Englishes in Asian contexts: Current and future debates, 194-207.

Monolingual and bilingual children's production of Russian embedded yes-no questions Marina Sherkina-Lieber

Carleton University

The goal of this paper is to investigate typical development and the role of bilingualism in acquisition of Russian embedded yes-no questions by comparing monolingual and Russian-English bilingual children. In English, embedded yes-no questions are produced at 3;3 (Pozzan 2011), but their acquisition in Russian has not been studied. This construction involves focus-driven movement in Russian, but not in English; another such construction, multiple wh-fronting in Russian, is not fully acquired even at age 6 (Grebenyova 2012). Therefore, late acquisition of embedded yes-no questions can be expected as well.

In English, embedded yes-no questions are formed with if or whether. In Russian, the complementizer li 'whether' cliticizes on the focused element moved to Focus Phrase in the left periphery (Schwabe 2004), as in (1a). There is also a colloquial form without a complementizer, where the focused constituent may move (1b) or stay in its canonical position (1c), and where interrogative intonation is preserved (i.e. it is not truly embedded).

(1) a. Ja sprosi-l Glash-u, bystro li ona begaj-et. I ask-PAST Glasha-ACC fast LI she run-3SG

b. Ja sprosi-l Glash-u, bystro ona begajet?c. Ja sprosi-l Glash-u, ona bystro begaet?

‘I asked Glasha if/whether she runs fast.'

Russian monolingual children aged 5-8 (n=24), same age Russian-English bilingual children in Canada (n=24), and Russian adults (n=8) participated in an elicited production task, in which they had to ask the computer to ask an imaginary creature named Glasha yes-no questions. In the first part, the questions had to be embedded under imperative Sprosi Glashu 'Ask Glasha', in the second part, under declarative Ja sprosil Glashu 'I asked Glasha'.

Adults produced significantly more li forms than monolingual children; bilingual childrenproduced almost none. Both child groups favoured the no-complementizer form, but only monolingual children used it with focus movement. In addition to it, bilinguals – but not monolinguals – also used an English-like option, ungrammatical in Russian: overextending the conditional-only complementizer jesli 'if' and the conditional clause structure (without focus movement) to embedded questions, as in (2). I argue that its use is caused by cross-linguistic syntactic priming (cf. Vasilyeva, Waterfall, Gamez, Gomez, Bowers & Shimpi 2010).

(2) * Ja sprosi-l Glash-u, jesli ona bystro begajet. I ask-PAST Glasha-ACC if she fast run-3SG

I suggest that, while the li form is acquired late in general, its pre-requisite – focus

movement – is at a bigger disadvantage in bilingual children who spend most of their day speaking English. This is in line with adult heritage speakers' strong preference for basic (non-scrambled) word order in scrambling languages (Russian – Polinsky 2006; Inuktitut - Sherkina-Lieber, Perez-Leroux & Johns 2011), suggesting that focus movement acquisition is sensitive to the quantity of input and use.

References

Grebenyova, Lydia. 2012. Syntax, semantics and acquisition of multiple interrogatives: Who wants what? Amsterdam: John Benjamins.

Polinsky, Maria. 2006. Incomplete acquisition: American Russian. Journal of Slavic Linguistics 14(2): 191-262.

Pozzan, Lucia. 2011. Asking Questions in Learner English: First and Second Language Acquisition of Main and Embedded Interrogative Structures. Doctoral dissertation, City University of New York

Schwabe, Kerstin. 2004. The particle li and the left periphery of Slavic yes-no interrogatives. In The syntax and the semantics of the left periphery, eds. Horst Lohnstein and Susanne Trissler, 385-429. Berlin: de Gruyter.

Sherkina-Lieber, Marina, Ana Perez-Leroux, and Alana Johns. 2011. Grammar without speech production: The case of Labrador Inuttitut heritage receptive bilinguals. Bilingualism: Language and Cognition, 14(3), 301-317.

Vasilyeva, Marina, Heidi Waterfall, Perla B. Gámez, Ligia E. Gómez, Edmond Bowers, and Priya Shimpi. 2010. Cross-linguistic syntactic priming in bilingual children. Journal of Child Language 37: 1047-1064.

Differences between predicates of personal taste and epistemic modals across languages E. Allyn Smith, Laia Mayol, and Elena Castroviejo-Miró

This study experimentally tests whether disagreement using a direct non-acceptance particle (NAP) such as No or That’s not true is felicitous with predicates of personal taste (PPTs) and epistemic modals (EMs) across languages in which these constructions are thought to be similar (English, Catalan and Spanish). Our results provide further evidence against theories such as Stephenson 2007 that try to unify PPTs and EMs in support of theories such as Bouchard 2012 in which they are not assumed to have a common link.

Background: So-called ‘faultless disagreement’ is exemplified in (1), where the intuition is that Sam is expressing his opinion about the cake and Bob is not denying that Sam considers the cake tasty but rather is either expressing the the cake does not taste good to him (following Kölbel 2003, inter alia) or expressing that the cake should not be considered tasty more generally (Stojanovic 2007, inter alia).

(1) Mary: How’s the cake? (2) Mary: How's the cake? Sam: It’s tasty. Sam: It tastes good to me. Bob: No it isn’t, it tastes terrible! Bob: #No it isn't/doesn't, it tastes terrible!

Stephenson 2007, arguing for a modification of Lasersohn 2005, points out that when PPTs are explicitly relativised to the speaker, as in (2), it is no longer possible to directly disagree, drawing a parallel with epistemic modals such as might. The idea is that you can say that someone might be in their office, to which someone else could say No, I just saw him at the gym, but if you make your degree of knowledge explicit, for example by saying I don’t know, people can no longer respond using No because they would be denying that you are unsure rather than the content of what you are unsure of. Given the importance of this data to subsequent work in the field, we set out to test whether this pattern could be replicated cross-linguistically.

Experiments: 203 native-speaker participants across 4 conditions and 3 languages listened to 88 two-turn dialogues (majority fillers) via an internet survey. In the first turn, a statement is made that crucially contains one of the meaning types in question (those in (3) as well as regular assertions, etc.), and in the second turn, for the critical stimuli, participants heard one of four NAPs such as No or That’s not true. We also present the norming task that provided baseline judgments used as a fixed effect in the linear mixed effects models used to analyze the results.

(3) Meaning Type Example: Turn1 (statement) Example: Turn2 (response) PPT This soup tastes delicious No, it is not delicious PPT-relativised This soup tastes delicious to me No, it is not delicious Epist. Modal Matt may join us later No, he will not join us EM-relativised I don't know whether Matt will join us later No, he will not join us

Results and Discussion: Results of a Likert judgment task asking how good a response Turn2 was to Turn1 show that the distinction between relativised and non-relativised PPT and EM sentences meanings are significant (p < 0.001) for English. However, Spanish and Catalan show a different pattern for PPTs as compared to EMs:

Taste Predicates Relativised PPTs Epistemic Modals Relativised EMs English ✔ ✖ ✔ ✖ Catalan ✔ ✔ ✔ ✖ Spanish ✔ ✔ ✔ ✖

These results are problematic for theories arguing for a parallel between EMs and PPTs with respect to a judge parameter or implicit argument. We consider a variety of possible explanations for our results, from differences in what No targets cross-linguistically to a modification of Umbach in which Relativised PPTs have more in common with EMs than they do Relativised EMs.

References:

Bouchard, David-Etienne. 2012. Long-distance degree quantification and the grammar of subjectivity. Doctoral Dissertation, McGill.

von Fintel, K. (2004). Would you believe it? The king of France is back! Presuppositions and truth-value intuitions. In Marga Reimer and Anne Bezuidenhout (Eds.), Descriptions and Beyond, Oxford University Press.

Lasersohn, Peter. 2005. Context dependence, disagreement, and predicates of personal taste. Linguistics and Philosophy 28:643–686.

Kölbel, M. 2003. Faultless disagreement. Proceedings of the Aristotelian Society 104 (1):53–73. Stephenson, T. 2007. Towards a Theory of Subjective Meaning. Doctoral dissertation, MIT. Stojanovic, I. 2007. Talking about taste: disagreement, implicit arguments, and relative truth.

Linguistics and Philosophy 30:691–706. Umbach, C. Evaluative propositions and subjective judgments, ms.

Christopher Spahr University of Toronto

Restricting non-segmental contrasts

Non-segmental features can serve a contrastive role in the phonology, and various models have

been developed to account for different phenomena such as length (CV, X-slot, and moraic the-

ory), tone (autosegmental theory), and stress (metrical theory). I argue for a unified model of

non-segmental phonology in which word-level contrasts are captured by interactions between two

tiers: a CV-tier, representing segmental root nodes, and a prosodic tier, representing features such

as lexical tone and lexical stress. Pure quantity contrasts are generally represented by double root

node linking on the CV-tier, while tone or non-derivable (lexical) stress contrasts are generally rep-

resented on the prosodic tier. I suggest that such a model together with parametrisable restrictions

on associations between the two tiers can derive a range of prosodic systems, from stress, to pitch

accent, to pure tone contrasts, without predicting co-occurrences that are not attested.

This predicts that a single language may not use more than two non-segmental contrasts. That

is, lexical tone, lexical stress, and contrastive quantity may not co-occur within the same system,

contra existing conceptions in which tones, moras/CV-slots, and metrically strong positions are

represented by distinct primitives in the grammar, all of which might be expected to be exploited

within the same language. Instead, I propose that Universal Grammar provides the mechanism of

defining contrasts and the organization thereof (autosegmental association lines across two tiers),

in the vein of Dresher (2013); the phonetic correlates of a given contrastive feature are language

specific and based on phonological and phonetic behaviour. I illustrate my proposal with case

studies of several languages that show multiple interacting non-segmental contrasts.

Serbo-Croatian has a pitch accent system characterized by interactions between contrastive

quantity and tone. The language is also described as having stress, but this is predictable based

on the placement of tone (Inkelas and Zec 1988). Stress may thus be a perceptual or phonetic

phenomenon, but it has no contrastive status. The word-level phonology can thus be captured with

only two tiers: tone on the prosodic tier, and quantity on the CV tier.

Papiamentu is described as having both contrastive tone and contrastive stress. However, it

is not said to have a length contrast. I propose that tone is represented on the prosodic tier, while

stress is represented as quantity on the CV-tier; this is in line with measurements by Rivera-Castillo

and Pickering (2004), which show that duration is the main correlate of Papiamentu stress.

Finally, Estonian is said to have a three-way quantity contrast, but overlong quantity is ac-

companied falling pitch, which can serve as the sole perceptual correlate distinguishing it from

the long degree, such that the language is moving towards a kind of pitch accent system (Lehiste

2003). I show how long/overlong alternations can be captured by a two-tiered analysis in which

the elements on the prosodic tier have a prime phonetic realization of pitch, which is enhanced

with additional duration only in certain contexts.

References

Dresher, B. Elan. 2013. The arch not the stones: Universal feature theory without universal features. Paper

presented at the Conference on Features in Phonology, Morphology, Syntax and Semantics: What are

they?, CASTL, University of Tromsø.

Inkelas, Sharon, and Draga Zec. 1988. Serbo-Croatian pitch accent: The interaction of tone, stress, and

intonation. Language 64:227–248.

Lehiste, Ilse. 2003. Prosodic changes in progress: From quantity language to accent language. In Develop-

ment in prosodic systems, ed. Paula Fikkert and Haike Jacobs, 47–65. Berlin: Mouton de Gruyter.

Rivera-Castillo, Yolanda, and Lucy Pickering. 2004. Phonetic correlates of stress and tone in a mixed

system. Journal of Pidgin and Creole Languages 19:261–284.

1

Which Questions Do You Like the Movements In? A Semantic Constraint on ExtractionDennis Ryan Storoshenko - University of Calgary and Robert Frank - Yale University

The Issue At the periphery of discussions on wh-movement are extractions from a certain classof structures which have been called Quasi-NPs (Cattell, 1979, 1980):

(1) Which cari do you like [the gears/*girls [in ti]]?Crucially, the reported ungrammaticality of (1) is under a parallel interpretation of gears and girls,i.e. both should be answerable with “I like the ones in the Ford, not the Chevy.” In this paper, wefollow up on Cattell’s suggestion that the constraint involved is semantic, rather than syntactic. Weprovide further support for this by unifying this case with one discussed by Truswell (2007), andby showing that both exhibit parallel constraints on scope inversion.Covert Argumenthood Because the two cases in (1) are syntactically parallel, Cattell arguesthat they are distinguished by a semantic property: the lower NP denotes an entity of which thehigher noun is a distinguished part. We take the relevant distinction to be tied to the semanticsof the higher noun: extraction is possible when that noun establishes a (possibly contextuallylicensed) semantic relation with the extracted NP, such as identifying location: a set of gears canbe identified by the car in which it is installed. As shown by the contrast with have sentences in(2a), gears can function as a relational noun taking the car as an argument, while girls cannot.

(2) a. The car has gears/*girls. b. The class has girls/*a girl.Based on (2b), it appears that girls (unlike girl) is able to function as a relational noun with theclass. The corresponding extraction contrast is found in the following, first noted by Cattell:

(3) a. I like the girls in that class. → Which class do you like the girls in?b. I like the girl in that class. → *Which class do you like the girl in?

Extraction is thus possible when the higher noun takes the location as a covert argument.Event Identification Truswell (2007) points out another contrast in extractability which appearsto depend on the semantic structure of the embedding predicate.

(4) Whati did John arrive/*work whistling ti?Truswell proposes that extraction is possible when the event denoted by an adjunct predicate isidentified with an event argument determined by the matrix predicate. In (4), telic arrive intro-duces an event argument, with which the whistling event is identified, associated with the processleading up to a resulting state. Being atelic, work introduces no such event argument, with no eventidentification possible. We propose that this requirement of identification is identical in both cases,the quasi-NPs requiring an identification of entities, and the cases in (4) one of events.

A further reason to take the Truswell and quasi-NP cases as parallel stems from their be-haviour with respect to scope. In both cases, the possibility of inverse scope exactly parallels thepossibility of extraction, while surface scopes in (5a) are subject to pragmatic constraints differen-tiating telic and atelic predicates:

(5) a. Some linguist arrived/worked whistling each song.arrived: #∃ > ∀,∀ > ∃; worked: ∃ > ∀, *∀ > ∃

b. Some mechanic likes the gears/girls in every car.gears: ∃ > ∀,∀ > ∃; girls: ∃ > ∀, *∀ > ∃

Implications So far as we know, existing syntactic accounts of scope and extraction are not sen-sitive to the types of argument identification that we have argued are relevant. We must then eitherextend Truswell’s proposal, making both extraction and scope sensitive to semantic structure, oraugment syntactic structure to represent processes of argument identification, in such a way thatconstrains both phenomena. Though we do not have space to justify this here, we believe the an-swer lies in Synchronous Tree Adjoining Grammar (Schabes and Shieber, 1994), where syntacticand semantic representations are composed in parallel, using recursive structure building opera-tions. This system allows us to provide a parallel treatment for scope and extraction, while alsoexplaining why similarities between them break down in cases of extraction from finite clauses.

ReferencesCattell, Ray. 1979. On extractability from quasi-NPs. Linguistic Inquiry 10:168–172.

Cattell, Ray. 1980. More on quasi-NPs. Linguistic Inquiry 11:419–420.

Schabes, Yves, and Stuart M. Shieber. 1994. An alternative conception of tree-adjoining derivation.Computational Linguistics 20:91–124.

Truswell, Robert. 2007. Extraction from adjuncts and the structure of events. Lingua 117:1355–1377.

NON-SELECTED ARGUMENTS AND THE ETHICAL STRATEGY

Tomokazu Takehisa, NUPALS INTRODUCTION: This paper investigates non-selected arguments in Japanese, as in (1)–(3) below, and makes the following claims: [1] non-selected arguments such as non-volitional agents/causers and possessors are syntactically indistinguishable, introduced by Appl (Pylkkänen 2008), while (volitional) agents are introduced by Voice (Kratzer 1996); [2] due to the defective licensing property of Appl in Japanese, non-selected arguments can only appear as nominative subjects of lexical causative verbs or those of possessor passives; moreover, [3] interpretational differences between these arguments are derived post-syntactically, by means of a modified version of the Ethical Strategy (Rivero 2004), which draws inferences from the thematic information read off the syntax and the set of proto-role properties (Dowty 1991). [1]: I argue for the distinction between (volitional) agents, on the one hand, and non-volitional agents/causers and other non-agentive VP-external arguments, on the other. Support for this view comes from data involving (direct) passivization, soo suru (“do so”) replacement, and formation of potential constructions (Inoue 1976). [2]: While non-selected arguments can appear freely in languages like German where Appl assigns inherent dative case, the defective licensing property of Appl in Japanese severely restricts their distribution to the contexts where an agent argument is suppressed (in passive) or not required (in lexical causatives) so that they can be marked nominative by T. [3]: Despite the difference in the licensing property of Appl, non-selected arguments receive similar treatments with respect to thematic interpretation across languages: given that Appl merely relates an argument to the event denoted by its complement (cf. Cuervo 2003, Schäfer 2012), the argument introduced is underspecified with respect to the way it participates in the event. I propose that the post-syntactic inferential procedure derives the interpretation of such thematically underspecified arguments in consideration of the syntactic input and the set of proto-role properties, which serve to circumscribe the bounds of interpretation. Specifically, an argument introduced by Appl can participate in the event, roughly in one of the two ways: causing the event or being causally affected by it. More specifically, if Voice, which introduces an agent argument, is not present in the input, as in (1) and (2), then the argument introduced can be construed in either way: as a non-volitional agent/causer, as in (1), or as being causally affected, presumably by having a (possessive) relation to another entity affected by the event, as in (2). Moreover, if the presence of an agent is formally represented in the input, as in (3), then the thematically underspecified argument can only be construed as being causally affected. SUMMARY: The present analysis offers a unified syntactic treatment of non-selected arguments such as non-volitional agents/causers and possessors, while at the same time deriving their variable interpretations by the post-syntactic Ethical Strategy. To the extent that this analysis is successful, the interpretations of non-selected arguments are syntactically uninformative. (1) Taroo-ga ukkari koppu-o war-Ø-ta (>wat-ta)

T.-NOM inadvertently cup-ACC √break-CAUS-PST Non-volitional agent/causer ‘Taro inadvertently broke the cup.’

(2) Taroo-ga ziko-de ude-o or-Ø-ta (>ot-ta) T.-NOM accident-in arm-ACC √break-CAUS-PST Possessor (Causative) ‘Taro broke his arm in the accident.’

(3) Taroo-ga Hanako-ni/niyotte okasi-o tabe-rare-ta T.-NOM H.-DAT/by sweets-ACC √eat-PASS-PST Possessor (Passive) ‘Taro had his sweets eaten by Hanako.’

REFERENCES Cuervo, Cristina (2003) Datives at Large. PhD dissertation, MIT. Dowty, David (1991) “Thematic proto-roles and argument selection,” Language 67:547-619. Inoue, Kazuko (1976) Henkei-Bumpoo to Nihongo [Transformational Grammar and Japanese],

volume 2. Taishuukan, Tokyo.) Kratzer, Angelika (1996) “Severing the External Argument from its Verb,” In J. Orrick and L.

Zaring, eds., Phrase Structure and the Lexicon, 109-137. Kluwer Academic Publishers, Dordrecht.

Pylkkänen, Liina (2008) Introducing Arguments. MIT Press, Cambridge, MA. Rivero, Maria Luisa (2004) “Datives and the non-active voice/reflexive clitics in Balkan

languages,” In O. Miseska-Tomic, ed., Balkan Syntax and Semantics, 237-267, John Benjamin, Amsterdam.

Schäfer, Florian (2012) “Two types of external argument licensing–the case of causers,” Studia Linguistica 66:128-180.

Spanish Nominal Word Order in Early and Late Bilingualism Danielle Thomas & Kristen Don Paul, York University In observing how difficult it is for late second-language (L2) learners to fully acquire some domains of L2 grammar against a monolingual standard, some investigators have proposed a so-called ‘sensitive period’ in which the ability to acquire a second language significantly decreases due to age-related cognitive change (Meisel, 2011). Formally, the domains of grammar proposed to be most problematic imply the use of so-called ‘interface’ areas of language where subtle interpretational properties are encoded structurally (Sorace, 2005). The nominal syntax of Spanish presents many such interface challenges for learners in the areas of gender agreement and nominal word order, especially for speakers of languages like English, a non-gender and rigid word order language. While gender has been studied extensively for English speakers of L2 Spanish (White et al, 2004; Montrul et al., 2008), few studies have examined nominal word order (Judy et al, 2008). Here, we examine the interface proposal in relation to age-related models of language acquisition by examining the knowledge that Spanish monolinguals and early and late Spanish-English bilinguals have of nominal syntax.

Grammatically, English is a rigid word order language (adj-N), while Spanish permits both word orders (adj-N, N-adj). While qualitative adjectives in Spanish appear predominantly in the post-N position with a contrastive interpretation, as in (1), these same adjectives may occur in the pre-N position with a non-contrastive, “individualizing” interpretation, as in (2): (1) el director famoso, y no el desconocido (the famous director, not the unknown one); (2) el famoso director del cine italiano (the “famous-director” of Italian cinema, only one reference possible—Federico Fellini). Some have proposed that the pre-N structure of qualitative adjectives, such as “famoso”, occurs by overtly raising the adjective, which is generated in SpecNP, to the Spec of DegP in the “external shell” of the DP (Demonte, 1999). Previous work on nominal word order in bilingualism have exhibited two main results: 1) advanced speakers of L2 Spanish pattern with native speakers on their knowledge of pre-N adjectives, unlike intermediate speakers (Judy et al. 2008); and 2) early child bilinguals exhibit bi-directional effects for nominal word order in Germanic and Romance languages (Nicoladis, 2006; Rizzi et al, 2013).

The main question of the current study, therefore, was to examine if early and late bilinguals use overt movement of qualitative adjectives to the “external shell” of the Spanish DP to encode semantic distinctions that they encode covertly in English. If the “interface” problem is a “sensitive period” issue, then we expected late bilinguals to not exhibit target knowledge of pre-N adjectives in the same way as native speakers (early bilinguals and monolinguals). On the other hand, if the “interface” problem is not strictly related to maturation, then we expected early and late bilinguals to exhibit similar knowledge of this domain, in contrast to monolinguals.

We tested 50 speakers of Spanish (6 monolinguals, 9 intermediate level heritage speakers, 11 advanced level heritage speakers, 12 intermediate late L2, 12 advanced late L2) on their knowledge of nominal word order in an aural Acceptability Judgment Task that employed felicitous and non-felicitous pre-N and post-N adjectives. Proficiency was measured both non-linguistically (questionnaire) and linguistically (lexical & syntactic assessments). Results here support a non-maturational proficiency-based model of bilingualism. Intermediate heritage speakers rejected felicitous pre-N adjectives significantly more than monolinguals and advanced early and late bilinguals (p=.012, p=.008, p=.039, respectively), with the latter groups performing alike; intermediate L2 speakers appeared to not have acquired the overt raising of qualitative adjectives, given that they exhibit more random judgments of these pre-N contexts (individual results ~50%). We discuss these results in terms of explicit learning, age as “macrovariable” in contexts of bilingualism, and in terms of current models of bilingual development and processing.

References

Demonte, V. (1999). A Minimal account of Spanish adjective position and interpretation. In Franco, J., Landa, A. & Martín, J. Grammatical Analyses in Basque and Romance Linguistics, (45-75)

Judy, T., Guijarro-Fuentes, P., Rothman, J. (2008). Adult Accessibility to L2 Representational Features: Evidence from Spanish DP. Proceedings of the 2007 Second Language Research Forum, ed. Melissa Bowles et al., (1-21). Somerville, MA: Cascadilla Proceedings Project.

Meisel, J. (2011). First and Second Language Acquisition. New York: Cambridge UP.

Montrul, S., Foote, R., Perpiñán, S. Gender Agreement in Adult Second Language Learners and Spanish Heritage Speakers: The Effects of Age and Context of Acquisition. Language Learning, 58(3), 503-553.

Nicoladis, E. (2006). Cross-linguistics transfer in adjective-nouns strings by preschool bilingual children. Bilingualism: Language and Cognition, 9(1), 15-32.

Rizzi, S., Arnaus Gil, L., Repetto, V., Geveler, J., Muller, N. (2013). Adjective placement in bilingual Romance-German and Romance-Romance children. Studia Linguistica, 67(1), 123-147.

Sorace, A. (2005). Syntactic optionality at interfaces. In L. Cornips & K. Corrigan (Eds.), Syntax and variation: Reconciling the biological and the social (pp. 46-111). Amsterdam: John Benjamins.

White, L., Valenzuela, E., Kozlowska-Macgregor, M. & Leung, Y.-K.I. (2004). Gender agreement in non-native Spanish. Applied Psycholinguistics, 25, 105-133.

Language-internal context and VOT in bilingualism: A view from Spanish and English Danielle Thomas & Katie Tetzloff, York University

One of the most challenging aspects of learning a second language (L2) is acquiring a “native-like” accent. Flege’s (1995) Speech Learning Model proposes that in order for a new phonetic category to be established, L2 speakers must be able to discern some phonetic difference between the L1 and L2 sound; if the L2 sound is heard as an allophone of the L1 phoneme, a new phonetic category will not be formed, causing the L1 and L2 categories to assimilate. This may result in a compromised L1-L2 feature. In addition to language internal factors, language external factors, such as age of first exposure to bilingualism, may affect how successful a learner can be in acquiring L2 sounds in a native-like way (Oyama, 1976).

One area that has been studied quite extensively in the area of sound acquisition in a bilingual context is the perception and production of voiced and voiceless stops ([b,d,g] and [p,t,k] respectively) in languages like Spanish with relatively short voice-onset times (VOT) and languages like English with relatively longer VOTs (see Zampini, 2014). Two aspects of this research stand out: 1) early bilinguals (simultaneous/early child L2) have an advantage over late bilinguals (adult L2) in acquiring VOT values in the “monolingual-like” range (Thornburgh & Ryalls, 1998); and 2) bilinguals may exhibit variability in VOT values for a stop not only in their later-acquired language, but also in their L1 (Bullock et al, 2006). Following evidence that the VOT value of a given stop is different not only according to place of articulation of the stop, but also in terms of the following vocalic segment (e.g. vowel height, Yavas, 2007), we sought to examine if variability in VOT values for voiceless stops were uniform across all phonetic contexts (e.g. pa, pe, pi), or if bilinguals exhibit differential variation on VOT values according to the phonetic context (i.e. vowel height). Following the SLM, we predicted that both Spanish speakers of L2 English and English speakers of L2 Spanish would exhibit more variability (i.e. compromise VOT values) on low vowels than on high (front) vowels, given that low vowels are less congruent cross-linguistically, but not distant enough to establish new phonetic categories.

To date we have tested 16 bilingual speakers of Spanish and English, 9 high intermediate speakers of late L2 Spanish and 7 advanced speakers of L2 English. Further, we have tested 5 heritage speakers of high intermediate Spanish to determine if SLM is a model that can cover a full range of bilingual effects, not simply those related to late contexts of exposure. In addition to completing both linguistic and non-linguistic measures (lexical/syntactic measures and self-rating) of their relative proficiency in Spanish and English, participants read (non-cognate) words in a carrier phrase (Say ___ again; Diga ___ otra vez) with labial and velar stops in three different phonetic contexts in language-specific tasks: pa/pæ, pe/pɛ, pi, ka/kæ, ke/kɛ, ki. Results partially support the prediction made above. Overall, late L2 speakers of both Spanish and English are able to acquire the phonetic distinction of English and Spanish /p/ and /k/, and use place of articulation to constrain the VOT for this domain. Further, all groups appear to be sensitive to the use of vowel height as a means by which to constrain VOT. However, these preliminary results point to the following: i) English speakers of L2 Spanish produce noticeable and consistent compromise values on their Spanish VOT values, and in neither case (/p/ or /k/) are more accurate when the voiceless stop is followed by a high vowels (congruent cross-linguistically) as compared to mid- and low front vowels; and ii) Spanish speakers of L2 English exhibit native-like VOTs for /k/, but not for /p/. In the latter case, these speakers do exhibit more accuracy in VOT when the high (congruent cross-linguistically) vowel /i/ follows the stop than when followed by the lower vowels /e/ and /a/. Interestingly, when speaking English, heritage speakers of Spanish have noticeably longer VOT values for all variants as compared to native English speakers. We will discuss these results in terms of the methodology employed (bilingual vs. monolingual mode of communication), models of phonological development and bilingual models (age- vs. proficiency-based) of sound processing.

References Bullock, B., Toribio, J., González, V., Dalola, A. (2006). Language dominance and performance

outcomes in bilingual pronunciation. In Grantham, M, Shea, C., & Archibald, J. (eds) Proceedings of the 8th Generative Approaches to Language Acquisition Conference. (9-16). Somerville, MA: Cascadilla.

Flege, J. E. (1995). Second language speech learning: Theory, findings, and problems. In W.

Strange (Ed.), Speech perception and linguistic experience: Issues in cross-language research (pp. 233-277). Timonium, MD: York Press.

Lisker, L. & Abramson, A. (1964). A Cross-Language study of voicing in initial stops: acoustical

measurements. Word, 3, 384-422. Oyama, S. (1976). A sensitive period for the acquisition of a non-native phonological system.

Journal of Psycholinguistic Research, 5, 261-283. Thornburgh, D. F., & Ryalls, J. H. (1998). Voice onset time in Spanish-English bilinguals: Early

versus late learners of English. Journal of Communication Disorders, 31(3), 215-228. Yavas, M. (2007). Factors Influencing the VOT of English Long Lag Stops and Interlanguage

Phonology. New Sounds 2007: Proceedings of the Fifth International Symposium on the Acquisition of Second Language Speech. 492-498.

Zampini, M. (2014). Voice Onset Time in second language Spanish. In Geeslin, K. The

Handbook of Second Language Spanish Acquisition. (113-129). Malden, MA: Wiley & Sons.

Les pronoms en français : pluralité et individuation Mireille Tremblay

Université de Montréal Les études sur la distinction masse/comptable se sont limitées au domaine empirique des noms communs et une classe lexicale importante a été négligée : celle des pronoms du pluriel. Comme le paradigme s’inscrit dans la logique des noms comptables et que les pronoms étant marqués pour le nombre (singulier ou pluriel), on présuppose souvent que les pronoms du pluriel sont marqués pour l’individuation. Les pronoms du pluriel de l’anglais semblent confirmer cette analyse puisqu’ils se comportent comme des noms comptables et peuvent être individués. (1) a. two of us

b. many of us c. how many of us d. none of us

En revanche, les pronoms forts du pluriel du français semblent s’inscrire dans une logique différente, puisqu’ils permettent très difficilement les constructions en (2) ((Franckel & Paillard 2007). Les mêmes exemples deviennent toutefois parfaitement grammaticaux lorsque la préposition entre est insérée devant le pronom (3). (2) a. #deux de nous (3) a. deux d’entre nous

b. #plusieurs de nous b. plusieurs d’entre nous c. #combien de nous c. combien d’entre nous d. #personne de noou d. personne d’entre nous

L’étude du français québécois nous montre que l’utilisation de la préposition entre n’est pas la seule stratégie disponible en français permettant la partition de l’ensemble dénoté par le pronom. Dans cette variété, les pronoms forts non clitiques du paradigme du pluriel apparaissent très souvent avec le morphème post-pronominal autres (Morin 1982, Auger 1994, Blondeau 2011) et la présence de ce morphème permet d’obtenir un ensemble partitionné. (3) a. deux de nous-autres

b. plusieurs de nous-autres c. combien de nous-autres d. personne de nous-autres

Afin de rendre compte de différences distributionnelles entre l’anglais, le français de référence et le français québécois, nous proposons l’analyse suivante. Les pronoms pluriels de l’anglais et du français ont la même dénotation : les deux types de pronoms renvoient à des ensembles d’objets dénombrables. En revanche, les deux types de pronoms différent au niveau de la discrétion : alors que les pronoms de l’anglais se comportent comme des noms communs pluriels et réfèrent à des objets comptables individués, les pronoms du français se comportent comme des noms communs collectifs et renvoient à des ensembles d’objets comptables non individués. L’individuation des pronoms pluriels du français peut s’obtenir de deux façons : soit avec l’insertion de la préposition entre (en français de référence), soit avec l’ajout du suffixe –autres (en français québécois) Cette analyse appuie l’hypothèse selon laquelle la distinction comptable/masse serait grammaticale plutôt qu’ontologique. Références bibliographiques Auger, J. (1994). Pronominal clitics in Québec colloquial French : A morphological analysis. PH.D. Thesis, UPenn. Blondeau, H. (2011). Cet « autres » qui nous distingue. Tendances communautaires et parcours individuels dans le système des pronoms en français québécois. Québec, PUL. Franckel, J.-J. et D. Paillard. (2007). Grammaire des prépositions, Tome 1. Paris, Ophrys. Morin, Y.-C. (1982). De quelques [l] non étymologiques dans le français du Québec: notes sur les clitiques et la liaison. Revue québécoise de linguistique 2(2) 9-47.

No weak necessity

Igor Yanovich (Universitat Tubingen)

In English, a number of tests distinguish “strong” and “weak” deontic necessity modals(namely “strong” must and have to vs. “weak” should and ought). Two of such testsare shown in 1 and 2. In English, the distinction seems so real that there is considerableformal literature on what the underlying semantic property might be (e.g., [Copley, 2006],[Rubinstein, 2012]). In the typological literature, the distinction between strong and weaknecessity is also assumed to be universal (cf. [Bybee et al., 1994], a.o.)

(1) Strenthening test:

a. You should wash your hands. In fact, you have to.

b. ??You have to wash your hands. In fact, you should.

(2) Ashfield test:Context: there are many routes to Ashfield, with their pros and contrasTo go to Ashfield, you OKought to/??have to take Route 2.

Against the accepted view, I argue that necessity deontics do not form categories ofweak vs. strong. The argument is two-fold. First, on data from East Slavic I show thatnecessity deontics need not divide neatly into two classes. If they did, we would have seena complementary distribution of modals in “weak necessity tests”, as in English. Table 3,summarizing the results of weak necessity tests for six Russian deontics, shows nothing close.

(3)

Neg weak Neg str Stren weak Stren str Ashfield Bridge Insur-str Insur-weak Coffee

nado + + + + + + + + +nuzno + + + + + + + + +dolzna + + + + ? + + + −objazana + + + + − − + − −sleduet + + + + + + + + +stoit + + + + + − + + +

Second, when we look more closely into English data, their complexities become apparent.In the non-formal-semantic literature, the notion of strength is rarely assumed to producetwo neat categories. Instead, different authors speak of relative strength, and sometimes ofloose groupings. For example, [Kangasniemi, 1992] reports an experiment where 150 nativespeakers ranked 6 Finnish necessity deontics for strength, with the mean values varying from8.2 to 4.5 (out of 10) gradually, not bimodally.

From the formal-semantic perspective, things are not as simple either. Formally, we canprove that two modals differ in strength rather than in modal flavor iff they have differentinterpretations when their flavor is fixed. But [von Fintel and Iatridou, 2008] argue preciselythat the flavor is not exactly the same for weak and strong necessity modals: according tothem, weak necessity ones use an additional ordering source bringing in a special flavor,namely that of non-coercive, non-strict rules. But if the flavors cannot be fixed, we haveno reliable formal test. Moreover, when we do manage to fix the flavor strictly, “weak”should and ought may actually appear in contexts where no weakness of any sort is implied:e.g., when you read that “all documents should be received by June 1 for full consideration”,this is just as strict as statements with must.

So how do we explain the observed facts if the “weak necessity theory” doesn’t help? Ishow that we actually do not need such a theory, as the behavior of different modals in “weaknecessity tests” may be correctly predicted from their basic properties, most importantlythe types of modal bases and ordering sources they may have. For example, 1a is fine whenthe first modal may express the modal flavor of advice, and the second, that of deonticnecessity. In English, must and have to are not particularly common in performative advicestatements, so they are banned from the first, “weak” position in the test. In similar ways,other contexts are shown to test for basic modal properties that the theory of modality needsto employ anyway. No additional distinction between weak and strong necessity deontics isneeded.

References

[Bybee et al., 1994] Bybee, J. L., Perkins, R., and Pagliuca, W. (1994). The evolution of grammar: tense,aspect and modality in the languages of the world. University of Chicago Press, Chicago.

[Copley, 2006] Copley, B. (2006). What should should mean? CNRS, Universite Paris 8. http://copley.free.fr/copley.should.pdf.

[von Fintel and Iatridou, 2008] von Fintel, K. and Iatridou, S. (2008). How to say ought in foreign: Thecomposition of weak necessity modals. In Gueron, J. and Lecarne, J., editors, Time and modality, volume 75of Studies in Natural Language and Linguistic Theory, pages 115–141. Springer.

[Kangasniemi, 1992] Kangasniemi, H. (1992). Modal expressions in Finnish. Suomalaisen KirjallisuudenSeura, Helsinki.

[Rubinstein, 2012] Rubinstein, A. (2012). Roots of modality. PhD thesis, UMass Amherst.

http://copley.free.fr/copley.should.pdf

http://copley.free.fr/copley.should.pdf

Congrès de l’ACL 2014 CLA Conference...

Documents

Transcript of Congrès de l’ACL 2014 CLA Conference...