Morphosyntactc producton in a head-marking languagelbutler/PsychLing...100% 0% 2% 4% 6% 8% 10% 12% NPs with RCs R C s w i t h o u t t h a t Optonal that in English relatve clauses

Morphosyntactc producton in a head-marking language

Order, agreement, and optonal morphology in Yucatec Maya

Lindsay K. Butler, U. RochesterElisabeth Norcliffe, MPI for Psycholinguistics

Jürgen Bohnemeyer, U. at Buffalo, SUNYT. Florian Jaeger, U. Rochester

1

Acknowledgments: It takes a team

• Space & participant recruitment:

– Carlos Pérez, Director of UNO

– Marta Beatriz Poot Nahuat

– Ángel Viriglio Salazar

– Michal Brody

– Betsy Kraf

• Programming of experiments:

– Andrew Watts

– Carlos Gomez Gallo (post-doc, Miami)

• Experiment and travel logistics

– Carlos Gomez Gallo (post-doc, Miami)

– Ashlee Shinn (Univ. at Buffalo)

– Katrina Furth (grad, Boston U)

• Transcription and annotation:

– Samuel Canul Yah (UNO)

– José Cano Sosaya (UNO)

Stimulus preparationYucatec recordings:

Samuel Canul Yah (UNO)

Gerónimo Can Tec (UNO)

Serapio Canul Dzib (UNO)

Video creation

Katrina Furth (grad, Boston U)

Cassandra Jacobs, Irene Minkina, Andy Wood (undergrads, Rochester)

Funding:NSF Grant BCS-0844472 to JB and TFJ

Wilmot Award, Alfred P. Sloan Fellowship to TFJ

Dissertation Improvement Grant from SBSRI, Univ. of Arizona to LKB

Mellon/ACLS dissertation completion fellowship awarded to EJN

[2]

• Psycholinguistics and the empiricist turn in the social/behavioral sciences

• Moving away from data exclusively from• College students

• who are members of the WEIRDest societies– “western, educated, industrialized, rich, and democratic”

[Henrich et al 2010]

• and speak mostly English or closely related languages

[3]

Why study the processing of “exotc” languages?

Sources of potental language-specific effects

• Variation – configurationality

– constituent order

– head-marking vs. dependent-marking

– argument ellipsis

– presence and organization of grammatical relations

– voice and alignment systems

– other functional categories

– lexical categories …

[4]

Dryer, Matthew S., 2011. Order of Object and Verb. In: Dryer, Matthew S. & Haspelmath, Martin (eds.) WASL. Munich: Max Planck Digital Library, feature 83A. Available online at http://wals.info/feature/83A.Accessed on 2011-12-20.

Short-before-long reversed for head-final languages [Hawkins 2004,2007, Yamashita and Chang 2001, Chang 2001, Choi 1997]

Roadmap

• Background– Part 1a: Challenges and methods– Part 1b: Introducing Yucatec Maya

• Example studies– Part 2: Redundancy and reduction– Part 3: Accessibility-based production– Part 4: Optional plural marking

• Part 5: Revisitng methods and conclusions

[5]

Part 1a

Challenges and methods

[6]

Challenges and methodological issues

• High uncertainty about linguistic structures of target language• Methodological and cultural issues

– Literacy– Computer literacy– Attention span– Interpretation of the task (more practice trials, more instructions)– Different norms about privacy, personal information

• Logistical issues– Maximizing output per visit – Participant recruitment

[7]

Part 1b

Yucatec Maya

[8]

Yucatec basics

JB’s field site

- Yaxley

Figure 1. Approximate geographic area where Yucatec is spoken

Our study site: UNO Valladolid

• Mayan, Yucatecan branch• 759,000 speakers age

5+ in Mexico in 2005– http://www.inegi.gob.mx

• Polysynthetic– but relatively rigid order

• Verb-initial, VOS– but lef-dislocation

pervasive in discourse• Typologically different

from English

[9]

Yucatec: Our study sites

• La Universidad de Oriente in Valladolid, Yucatán, Mexico– Sound-proof recording room– Computer-literate

participants– Familiar with testing

paradigms

• Other field sites– Valladolid surrounding

villages– Yaxley

[10]

Part 2

Redundancy and Reducton(Norcliffe 2009, Jaeger & Norcliffe, in prep)

[11]

Redundancy and grammatcal choice

• The language production system exhibits a bias to reduce the expected (contextually redundant), e.g.:

– Phonetic or phonological reduction is more common for contextually expected instances of words[e.g. AylettTurk04,06; BellETAL03,09; GahlGarnsey04; TilyETAL09]

– Morphological contraction of negation or auxiliaries is more common when the contractible element is contextually expected [FrankJaeger08; Melnick10; cf. BybeeScheibman99]

– Optional function words are likely to be omitted if the phrase they introduce is contextually expected [Jaeger10,11; LevyJaeger07; WasowETAL11]

– Optional arguments are more likely to be omitted if their meaning is more expected given the verb [Resnik96]

[12]

Theoretcal relevance

• Findings like these have been taken by some as evidence that the mechanisms underlying language production are organized to facilitate robust communicaton[Jaeger 06,10; LevyJaeger07; see also AylettTurk04; Fenk-Oczlon01; Lindblom90; vanSon&vanSanten05 and related ideas: e.g. Zipf49, Givon92]

Also offers potential account of:

• But, except for some work on phonetic reduction, all evidence comes from English.

Differential case-marking [FedzechkinaETAL11; KurumadaJaeger12]Pronominalization [Arnold98; ArnoldGriffin07; TilyPiantadosi09]Word order alternations [MauritsETAL10]Derivation of Zipf’s law [PiantadosiETAL11]

[13]

Assessing the effects of redundancy in Yucatec sentence producton

• We present a first step as to how to explore this question for a language like Yucatec.

• The phenomenon: Optional morphology in Yucatec Maya relative clauses [Bricker78, Gutiérrez-BravoMonforte09, Norcliffe09]

a. Le turista ku-t’aan-ik maya-o’DEF tourist ASP.A3-speak-INC maya-D2

b. Le turista t’aan-ik maya-o’ [ ‘Agent-Focus voice’]DEF tourist speak maya-D2“The tourist who speaks Maya”

[14]

Assessing the effects of redundancy in Yucatec sentence producton

• Hypothesis: choice of morphological form is infuenced by the expectedness of the relative clause

• Point of departure: parallels with English omission phenomena– Optional that in English object relative clauses

The cake that he baked

The cake he baked

• That omission correlates with expectedness of the relative clause [Wasow, Jaeger, Orr, 2011]

[15]

Relative Clause Rate and that Rate by Determiner

0%

25%

50%

75%

100%

0% 2% 4% 6% 8% 10% 12%

NPs with RCs

RC

s w

itho

ut th

at

Optonal that in English relatve clauses

• For pragmatic reasons, some properties of noun phrases lead to increased probability of a relative clause …

… which correlates with increased omission of that

This is the thickest book [that I had ever seen …]DT ADJ N

Expectedness of RC given DT

Om

issi

on o

f tha

t

[Wasow, Jaeger, Orr, 2011]adjusted r2=.91

[16]

• Yucatec boundary morphology is a cue to the likelihood of an upcoming RC–Definite NPs require a NP-final deictic particle (–o’)

• Therefore, absence of –o’ particle directly afer the noun is a strong cue that the NP contains post-nominal modification (including relative clauses)

ku t’aanik maya-o’

Language-specific morphosyntactc cue

[17]

Xmariae’ tu-che’ehtah le turista-o’

Absence of -o’

Maria laughed at the tourist who speaks Maya

Predicton

• The distribution of Yucatec boundary morphology increases the expectedness of relative clauses afer definite NPs, compared to indefinite NPs …

• … speakers should prefer to use reduced verb forms afer definite NPs.

[18]

Method

Spoken sentence recall

[19]

Le turista ku-t’aan-ik maya-o’

Result• Click to edit Master text styles– Second level

• Third level– Fourth level

» Fifh level

Fewer full RC verb forms (with ku) if modified NP is definite and lacks the -o’ particle.

*

Modified NP Relative clause

Full RC verb

Final particle

Conclusion

• Yucatec speakers prefer morphological reduction where RCs are highly expected

An effect of a preference for communicative robustness seems to show up in Yucatec as in English

This generalization only becomes apparent once the language-specific morpho-syntactic cue is taken into account [cf. Hawkins04,07,11]

[21]

Part 3

Accessibility(Butler, Jaeger, Bohnemeyer, Gómez Gallo, Furth, in prep)

[22]

Accessibility-based producton

• Crosslinguistically, conceptually accessible, e.g. more animate, tend to be ordered early and aligned with prominent grammatical function, e.g. subject [BockWarren85, Branigan et al. 2007, Tanaka et al. 2007]– Does this effect hold across languages , e.g. head-marking?

[from Prat-Sala & Branigan 2000][23]

“The man was hit by the swing”

“The swing hit the

scooter”

Experiment

• Video description task: Human and animal agents and undergoers

[24]

Experiment (cntd)

• Video description task: Human and inanimate “agents” and undergoers

[25]

Results

• Animacy significantly affected constituent order (human patients more likely to result in OVS) (X2 (1) = 17.1, p < 0.0001)

• Animacy of the patient, however, did not significantly affect voice choice (active vs. passive)

[26]

“The man, the truck pulled him”

“The man was chased by the dog”

Additonal results

• Universal effects of animacy on constituent order

Animacy and order in Yucatec Animacy and order in Spanish

[27]

• Variation in size of the effect and language-particulars

Part 4

Plural Marking(Butler 2011)

[28]

Language-specific morphosyntax

• The optional nominal plural marker in Yucatec, –o’ob, is right-adjoined to the DP (occupying a high position and occuring linearly late in the phrase) [Butler11]

– Predicted by the syntax of plural marking [Wiltschko08]

[29]

b. The girlsPL and the womanSG

c. The girlSG and the womenPL

d. The girlsPL and the womenPL

a. The girlSG and the womanSG

Experiment Design

• Is there experimental evidence for the DP-adjoined plural hypothesis in Yucatec Maya?

[30]

Translation task with conjoined noun phrases● N1-SG and N2-SG Verb (intransitive)● N1-SG and N2-PL Verb (intransitive)● N1-PL and N2-SG Verb (intransitve)● N1-PL and N2-PL Verb (intransitive)

The DP-adjoined plural hypothesis predicts N1-Ø N2-PL responses in Yucatec to be possible in Cond. 3

Results

• Phrase-final, DP-adjoined plural hypothesis predicts

[31]

Spanish stmulus conditons

Yucatec responses

Results

• Responses ruled out by “underspecification”

[32]

Spanish stmulus conditons

Yucatec responses

Point of departure

• Morphosyntactc priming in translaton: Plural marking is obligatory in Spanish and optional in Yucatec, thus an inherent potential for crosslinguistic priming in the task

Spanish

Stimulus: Las muchachas[PL] y las mujeres[PL] …The girls and the women

Yucatec:

Response: Le x-ch’úupal-o’ob[PL] yéetel le ko’olel-o’ob[PL]-o’The girls and the women

[33]

Translaton vs. picture descripton

• Use of plural marking in singular/one, two, and plural/many conditions compared

• TRANSLATION TASK: Singular, “Two”, Plural

• PICTURE DESCRIPTION TASK: One, Two, Many (seven)

[34]

“The baby is crying”“Two babies are crying” “The babies are crying”

Translaton vs. picture descripton results

• Click to edit Master text styles– Second level

• Third level– Fourth level

» Fifh level

[35]

Plural use in translation task Plural use in picture description task

* *

Results

• Accounted for by “underspecification” and priming

[36]

Spanish stmulus conditons

Yucatec responses

Remaining data unambiguously accounted for by DP-adjoined, phrase-

final morphosyntax

Conclusion

• Some responses only accounted for by the DP-adjoined phrase-final plural hypothesis

• Production results informing linguistic theory

[37]

Part 5

Revisitng methods & Conclusions

[38]

Methods revisited

Advantages Disadvantages

Video description tasks

Elicits unscripted speech

Limited to messages that can be unambiguously depicted

Recall tasks Especially useful for messages that are not easily depictable

Unfamiliar task

Translation tasks More familiar (in bilingual communities)

Priming from stimulus language

[39]

Conclusions

• Despite inherent challenges to field-based psycholinguistics, the crosslinguistic perspective provided by typologically diverse languages is essential to research on human language processing

• Language-specific effects can explain results that are otherwise counter to known effects [cf. Hawkins04,07,11]

• Quantitative production data to address structural differences informed by linguistic theory

[40]

Acknowledgments

• Space & participant recruitment:

– Carlos Pérez, director of UNO

– Marta Beatriz Poot Nahuat

– Ángel Viriglio Salazar

– Michal Brody

– Betsy Kraf

• Programming of experiments:

– Andrew Watts (U. Rochester)

– Carlos Gomez Gallo (post-doc, Miami)

• Experiment and travel logistics

– Carlos Gomez Gallo (post-doc, Miami)

– Ashlee Shinn (Univ. at Buffalo)

– Katrina Furth (grad, Boston U)

• Transcription and annotation:

– Samuel Canul Yah (UNO)

– José Cano Sosaya (UNO)

Stimulus preparationYucatec recordings:

Samuel Canul Yah (UNO)

Gerónimo Can Tec (UNO)

Serapio Canul Dzib (UNO)

Video creation

Katrina Furth (grad, Boston U)

Cassandra Jacobs, Irene Minkina, Andy Wood (undergrads, Rochester)

Funding:NSF Grant BCS-0844472 to JB and TFJ

Wilmot Award, Alfred P. Sloan Fellowship to TFJ

Dissertation Improvement Grant from SBSRI, Univ. of Arizona to LKB

Mellon/ACLS dissertation completion fellowship awarded to EJN

[41]

Selected referencesArnold, J. E. (1998). Reference Form and Discourse Patterns. Dissertation, Stanford UniversityArnold, J.E., & Griffin, Z. (2007). The Effect of Additional Characters on Choice of Referring Expression: Everyone Competes. Journal of Memory and LanguageAylett, M. P., & Turk, A. (2004). The Smooth Signal Redundancy Hypothesis: A Functional Explanation for Relationships between Redundancy, Prosodic Prominence, and Duration in Spontaneous Speech. Language and Speech, 47(1), 31-56.Bell, A., Jurafsky, D., Fosler-Lussier, E., Girand, C., Gregory, M., & Gildea, D. (2003). Effects of disfuencies, predictability, and utterance position on word form variation in Englishconversation. Journal of the Acoustical Society of America, 113(2), 1001-1024.Bell, A., Brenier, J. M., Gregory, M., Girand, C., & Jurafsky, D. (2009). Predictability effects on durations of content and function words in conversational English. Journal of Memory and Language, 60(1), 92-111.Bock, J. K. (1986). Syntactic persistence in language production. Cognitive Psychology, 18, 355- 387.Bock, J. K., & Warren, R. (1985). Conceptual accessibility and syntactic structure in sentence formulation. Cognition, 21.Branigan, H. P., Pickering, M. J., & Tanaka, M. (2008). Contributions of animacy to grammatical function assignment and word order during production. Lingua, 118 (172-189).Bresnan, J., Cueni, A., Nikitina, T., & Baayen, H. (2007). Predicting the Dative Alternation. In G. Boume, I. Kraemer & J. Zwarts (Eds.), Cognitive Foundations of Interpretation (pp. 69-94).Amsterdam: Royal Netherlands Academy of Science.Bricker, V. R. 1978. The source of the ergative split in Yucatec Maya. Journal of Mayan Linguistics, 2, pp. 83–127. Butler, L. K. 2011. The morphosyntax and processing of number marking in Yucatec Maya. Ph.D. thesis. University of Arizona.Bybee, Joan and Joanne Scheibman. 1999. The effect of usage on degrees of constituency: the reduction of don't in English. Linguistics 37-4. 575-596.Christianson, K. and F. Ferreira. 2005. Conceptual accessibility and sentence production in a free word order language. Cognition 98: 105—135.Fedzechkina, M., Jaeger, T. F. and E. Newport. Functional biases in language learning: Evidence from word order and case-marking interaction. The 33rd Annual Meeting of the Cognitive Science Society, Boston, July 2011.Fenk-Oczlon, G. 2001. Familiarity, Information Flow, and Linguistic Form. In: J. Bybee and P. Hopper (eds.) Frequency and the Emergence of Linguistic Structure, 431-448. Amsterdam/Philadelphia: John Benjamins Publishing Company.Ferreira, V. and H. Yoshita. 2003. Given-new ordering effects on the proudction of scrambled sentences in Japanese. Journal of Psycholinguistic Research 32: 6, 669-692.Frank, A., & Jaeger, T. F. 2008. Speaking Rationally: Uniform Information Density as an Optimal Strategy for Language Production The 30th Annual Meeting of the Cognitive Science Society (CogSci08) (pp. 933-938).Gahl, S., & Garnsey, S. M. (2004). Knowledge of grammar, knowledge of usage: syntactic probabilities affect pronunciation variation. Language, 80(4), 748-775.

[42]

Selected referencesGutiérrez-Bravo, R. & J. Monforte. 2009. 'Focus, agent focus and relative clauses in Yucatec Maya', in New Perspectives on Mayan Linguistics, H. Avelino, J. Coon, & E. Norcliffe (eds.), MIT Working Papers in Linguistics.Hawkins, J. A. (2004). Efficiency and Complexity in Grammar. Oxford: Oxford University Press. Hawkins, J. A. (2007). Processing typology and why psychologists need to know about it. NewIdeas in Psychology, 25, 87–107.Jaeger, T. F. (2006). Redundancy and Syntactic Reduction in Spontaneous Speech. PhD thesis, Stanford University.Jaeger, T. F. (2010). Redundancy and reduction: Speakers manage syntactic information density. Cognitive Psychology, 61, 23-62.Jaeger, T. F., & Norcliffe, E. (2009). The cross-linguistic study of sentence production. Language and Linguistics Compass, 3, 866-887Kurumada, C. and T. F. Jaeger. 2012. Communicatively efficient language production and case-marker omission in Japanese. Paper presented at the 86th Annual Meeting of the Linguistic Society of America.Levy, R. and T. F. Jaeger. 2007. Speakers optimize information density through syntactic reduction. Proceedings of the Twentieth Annual Conference on Neural Information Processing Systems.Lindblom, B. (1990). Explaining phonetic variation: a sketch of the H&H theory. Speech production and speech modelling, 55, 403-439.Maurits, L., Perfors, A., & Navarro, D. (2010). Why are some word orders more common than others? A uniform information density account. Adv. in Neural Information Processing Systems, 23, 1585-1593.Norcliffe, E. 2009. Head-marking in usage and grammar: A study of variation and change in Yucatec Maya. PhD Thesis, Stanford University, Stanford, CA. Piantadosi, S. T., H. Tily, and E. Gibson. Word lengths are optimized for efficient communication. Proceedings of the National Academy of Sciences , 108(9):3526, 2011.Prat-Sala, M., & Branigan, H. P. 2000. Discourse constraints on syntactic processing in languageproduction: A corss-linguistic study in English and Spanish. Journal of Memory and Language,42, 168-182.Tanaka, M., Branigan, H.P., McLean, J.F., & Pickering, M.J. 2011. Conceptual infuences on word order and voice in sentence production: Evidence from Japanese. Journal of Memory and Language, 65, 318-330. Tily, H. and S. T. Piantadosi. 2009. Refer efficiently: Use less informative expressions for more predictable meanings. In Proceedings of the workshop on the production of referring expressions: Bridging the gap between computational and empirical approaches to reference .Van Son R. and Van Santen, J.P.H. (2005). "Duration and spectral balance of intervocalic consonants: A case for efficient communication", Speech Communication 47, 100-123.Wasow, T., Jaeger, T. F., & Orr, D. (2011). Lexical Variation in Relativizer Frequency. In H. Wiese& H. Simon (Eds.), Proceedings of the 2005 DGfS workshop “Expecting the unexpected:Exceptions in Grammar” (pp. 175-196). Berlin/NewYork: De Gruyter Mouton.Wiltschko, M. 2008. The syntax of plural marking. Natural Language and Linguistic Theory Zipf, G. K. 1949). Human Behavior and the Principle of Least Effort . Addison-Wesley.

[43]