Workshop on Verbal Elasticity @ UAB October 4, 2011 Preposition Stranding: Its Parametric Variation and Acquisition* Koji Sugisaki Mie University 1. (1) (2) (3) (4) (5) Introduction [email protected]‐u.ac.jp Model of Language Acquisition: CHILD Input Data of L → UG → Grammar/I‐language of L Two Modes of Inquiry into UG (Chomsky 2007:4): a. Top‐down Approach = Principles & Parameters How much must be attributed to UG to account for language acquisition? b. Bottom‐up Approach = Minimalist Program How little can be attributed to UG while still accounting for the variety of languages attained, relying on third factor [i.e. language‐independent] principles? The Strong Minimalist Thesis (Chomsky 2010:52): Interfaces + Merge = Language The Locus of Cross‐linguistic Variation: a. Principles & Parameters: Each of the subsystems of [grammar] is based on principles with certain possibilities of parametric variation (Chomsky 1981:6). b. Minimalist Program: Parameterization and diversity too would be mostly – maybe entirely – restricted to externalization (Chomsky 2010:60). Boeckx (2010:2): “ … if one takes minimalism and biolinguistics seriously, one should abandon the [substantive] notion of Parameter, the more so given its diminishing empirical validity …” I would like to thank Cedric Boeckx and William Snyder for their valuable comments on a number of aspects of this study, and for their constant encouragement and support. ‐ 1 ‐ *
(6) 2. (7) (8) 2.1. (9) (10) (11) Goals: a. b. c. Overview cross‐linguistic generalizations about preposition stranding (P‐stranding) proposed in the theoretical literature. Evaluate some of these cross‐linguistic generalizations with the data from child language. Discuss the question of whether the “surviving” cross‐linguistic variation of P‐stranding can be located in the “externalization” component (i.e. PF component). Parametric Variation in P‐stranding English: P‐stranding possible a. Who was Peter talking with t ? b. ?? With whom was Peter talking t ? Spanish: a. * Quién who b. Con with [Odd, in spoken English] P‐stranding impossible / Pied‐piping obligatory hablaba Pedro con t ? was‐talking Peter with quién hablaba Pedro t ? who(m) was‐talking Peter Stowell (1981) Cross‐linguistic Generalization: P‐stranding is possible only in those languages that permit transitive verb‐particle construction (especially the one with the order V‐Particle‐NP). Transitive Verb‐Particle construction: a. English: Mary lifted up the box. b. Spanish: María levantó (*arriba) la caja. Cross‐linguistic Survey (Sugisaki & Snyder 2002): Language P‐stranding? Verb Particles?
Yes Icelandic Yes Norwegian Yes Yes Swedish Yes Yes Danish Yes Yes English Yes Yes Dutch (limited) Yes Frisian (limited) Yes Afrikaans (limited) Yes ‐ 2 ‐ 2.2. (12) (13) (14) (15) German No Yes Greek No No French No No No Italian No Spanish No No Bulgarian No No Russian No No Serbo‐Croatian No No Basque No No Kayne (1981) Cross‐linguistic Generalizations: a. Prepositional Complementizer (PC) construction is possible only in those
languages that permit P‐stranding. b. Double Object (more accurately, Double Accusative) construction is possible only in those languages that permit P‐stranding. PC construction: a. English: b. French: * John wants (for) Mary to leave. Jean veut (de) Marie partir. Double Accusative construction a. English: John gave Mary a book. b. French: * Jean a donné Marie un livre. Cross‐linguistic Survey: P‐stranding? 2.3. (16) (17) (18) (19) (20) Icelandic French YES YES NO PC construction? YES NO NO Double Accusatives? YES NO NO Maling and Zaenen (1985) Cross‐linguistic Generalization: P‐stranding with A‐movement (prepositional passives or pseudopassives) is possible only in those languages that allow P‐stranding with Aʹ‐movement. English: a. What did they talk about t ? b. This problem was already accounted for t. ‐ 3 ‐ snakket med? talked with ble ledd av. was laughed at Swedish (Vikner 1995:246): a. Vem har Peter who has Peter b. … att Peter that Peter talat med? talked with skrattades åt. was‐laughed at Cross‐linguistic Survey: English
Norwegian Swedish Icelandic
Danish Spanish P‐stranding with Aʹ‐mvt?
YES YES YES YES YES NO P‐stranding with A‐mvt? YES YES YES NO NO NO 2.4. (21) (22) Law (1998, 2006) Cross‐linguistic Generalization: Pied‐piping of prepositions (P‐pied‐piping) is obligatory in those languages that have suppletive forms of prepositions and determiners (P+D suppletive forms). English Norwegian (Vikner 1995:246): a. Hvem har Per who has Per b. … at Petter that Peter (23) P+D suppletive forms: a. French: Jean a parlé du sujet le plus Jean have talked about‐the subject the most ‘Jean talked about the most difficult subject.’ b. Italian: Gianni ha parlato del sogetto più Gianni have talked about‐the subject most ‘Gianni talked about the most difficult subject.’ difficile. difficult difficile. difficult Cross‐linguistic Survey: French Italian Portuguese
German P+D Suppletive forms? YES YES YES YES NO NO Obligatory P‐pied‐piping?
YES YES YES YES NO NO ‐ 4 ‐ English Icelandic 2.5. (24) (25) (26) (27) Merchant (2001) Cross‐linguistic Generalization (Merchant 2001:92):1 A language L will allow preposition stranding under sluicing iff L allows preposition stranding under regular wh‐movement. English: a. Peter was talking with someone, but I don’t know (with) who. b. Who was Peter talking with? German: a. Anna Anna aber but b. * Wem who hat has ich I hat has mit with weiβ know sie she jemandem gesprochen, someone spoken nicht *(mit) wem. not with who mit gesprochen? with spoken 2.6. (28) (29) Cross‐linguistic Survey: Language P‐stranding? P‐stranding under sluicing? English YES YES Frisian YES YES Swedish YES YES Norwegian YES YES Danish YES YES Icelandic YES YES German NO NO Greek NO NO Yiddish NO NO Czech NO NO Russian NO NO Slovene NO NO Polish NO NO Bulgarian NO NO Serbo‐Croatian NO NO Persian NO NO 1. For (potential) problems to this generalization, see Almeida & Yoshida (2007), Sato (in press), Stejepanović (2008), and Szczegelniak (2008). For an alternative analysis of (some of) these counterexamples, see Rodrigues, Nevins, and Vicente (2009). ‐ 5 ‐ (30) Hebrew NO Moroccan Arabic
NO NO Basque NO NO Hasegawa (2007) Cross‐linguistic Generalization: Swiping is possible only in those languages that allow P‐stranding. Swiping (Merchant 2002): (sluiced wh‐word inversion with prepositions in Northern Germanic) a. English: Peter went to the movies, but I don’t know who with. b. Danish: Per er gået I biografen, men jeg ved ikke Per is gone to cinema but I know not hven med. who with ‘Per went to the movies but I don’t know who with.’ c. Norwegian: % Per gikk på kino, men jeg veit ikke Per went to cinema but I know not hvem med. who with ‘Per went to the movies but I don’t know who with.’ Cross‐linguistic Survey: 2.7. (31) NO English Danish Icelandic Spanish French P‐stranding? YES YES YES NO NO Swiping? YES YES NO NO NO Truswell (2009) Cross‐linguistic Generalization: A’‐extraction from Bare Present Participial Adjuncts (BPPA) is possible only in those languages which allow pseudopassivization. ‐ 6 ‐ (32) (33) English A’‐extraction out of Bare Present Participial Adjuncts (an adjunct headed by a present participle): a. English: What did John arrive [whistling t ]? b. Norwegian: Hvilken sang kom han [ plystrende på t ]? which song came he whistling on ‘Which song did he arrive whistling?’ c. Swedish: Vilken sång kom han in i rummet which song came he in in room.the [ visslande på t ] ? whistling on ‘Which song did he come into the room whistling?’ (37) Cross‐linguistic Survey: (38) (39) (40) P‐stranding with Aʹ‐mvt? P‐stranding with A‐mvt? Aʹ‐extraction out of BPPA? YES YES YES Norwegian YES YES YES Swedish YES YES YES Icelandic YES NO NO Danish YES NO NO Faroese YES NO NO Dutch NO NO NO Greek 3. (34) (35) (36) NO NO NO 2. Acquisitional Evaluation [1]: P‐stranding and Verb Particles2 Cross‐linguistic Generalization (Stowell 1981): P‐stranding is possible only in those languages that permit transitive verb‐particle construction (especially the one with the order V‐Particle‐NP). In acquisitional terms, the grammatical knowledge required for V‐Particle‐NP construction is a proper subset of that required for P‐stranding. Prediction for Child English: The age of acquisition for V‐Particle‐NP construction should always be less than or equal to the age of acquisition for P‐stranding. (No child should acquire P‐stranding significantly earlier than V‐Particle‐NP construction.) This section is based on Sugisaki & Snyder (2002). ‐ 7 ‐ (41) Corpora Analyzed: Child Collected by Age Span Abe Kuczaj (1976) 2;04 – 5;00 # Child Utterances 4,214 Adam Brown (1973) 2;03 – 4;10 9,253 Allison
Bloom (1973) 1;04 – 2;10 2,192 April Higginson (1985) 1;10 – 2;11 2,321 Eve Brown (1973) 1;06 – 2;03 12,473 Naomi
Sachs (1973) 1;02 – 4;09 16,634 Nina Suppes (1973) 1;11 – 3;03 22,957 Peter Bloom (1970) 1;09 – 3;01 24,422 Sarah Brown (1973) 2;03 – 5;01 20,787 Shem Clark (1978) 2;02 – 3;02 9,178 For each child, we located the first clear uses of: a. V‐Particle‐NP construction b. a direct‐object wh‐question c. a wh‐question or a null‐operator construction with P‐stranding. Among the ten children, eight of them acquired V‐Particle‐NP constructions, direct‐object wh‐questions, and P‐stranding by the end of their corpora. Statistical method: Binomial Test a. Begin at the childʹs first direct‐object wh‐question. Count the number of clear uses of the earlier construction (either V‐Particle‐NP or P‐stranding), up to the point when the child begins using both constructions. Determine the relative frequency of the two constructions in the next four transcripts, or until the end of the child’s corpus, whichever comes first. b. Use the Binomial Test to calculate the probability of the child’s producing at least the observed number of examples of the first construction, before starting to use the second construction, simply by chance. c. Null hypothesis: Both constructions became available concurrently, and
had the same relative frequency observed in later transcripts. relative frequency of P‐stranding Eve: wh‐question V‐Prt‐NP P‐stranding = .667 1;08 1;10 2;02 9 instances of relative frequency of V‐Prt‐NP V‐Particle‐NP = .333 The probability of producing 9 instances of V‐Particle‐NP construction before P‐stranding simply by chance: (.333)^9 < .001 ‐ 8 ‐ (42) (43) 4. (44) (45) (46) (47) (48) (49) 3. Results: a. Six of the eight children acquired the V‐Particle‐NP construction significantly earlier than P‐stranding, by Binomial Test (Eve, Naomi, Nina, Peter, Sarah, Shem). b. The remaining two children acquired the V‐Particle‐NP construction and P‐stranding at approximately the same age (no significant difference, by Binomial Test). c. Crucially, no child in our study acquired P‐stranding significantly earlier than the V‐Particle‐NP construction. These findings from child English strengthen Stowell’s generalization that languages permitting P‐stranding are a proper subset of those permitting transitive verb‐particle construction. Acquisitional Evaluation [2]: P‐pied‐piping and P+D Suppletion3 Cross‐linguistic Generalization (Law 1998, 2006): Pied‐piping of prepositions (P‐pied‐piping) is obligatory in those languages that have suppletive forms of prepositions and determiners (P+D suppletive forms). Parameter of D‐to‐P incorporation (Law 1998, 2006): A language {has, does not have} D‐to‐P incorporation. Syntactic constraint on suppletion (Law 1998:227): Elements that undergo suppletive rules must form a syntactic unit X0. D‐to‐P incorporation: PP P ei DP ty ru P D t NP Given the constraint in (45), only those languages that have the positive setting of the relevant parameter are permitted to have P+D suppletive forms. Languages with the positive setting: [CP [IP … [VP … [PP P+D [DP t NP ]] about + which topic This section is based on Isobe & Sugisaki (2002). ‐ 9 ‐ ]]] (50) (51) (52) (53) Languages with the negative setting: [CP [IP … [VP … [PP P about [DP D NP ] ] which topic ]]] In acquisitional terms, the analysis by Law (1998, 2006) suggests that the only language‐particular knowledge that French‐learning children need to acquire in order to induce obligatory P‐pied‐piping is (i) the knowledge about the availability of P+D suppletive forms and (ii) the knowledge about overt wh‐movement. Prediction for Child French: A French‐learning child should exhibit P‐pied‐piping as soon as she acquires P+D suppletive forms and overt wh‐movement. Corpora Analyzed: Child Collected by Age Span # Child Utterances Grégoire
Champaud 1;09 – 2;05 3,237 Philippe Suppes et al. (1974) 2;01 – 3;03 13,739 (54) For each child, we located the first clear uses of: a. an overt wh‐movement of que ‘what’ or qui ‘who’ from a complement position b. a wh‐question with P‐pied‐piping c. a P+D suppletive form. (55) To count as a clear use, we required the P+D suppletive form to appear after verbs, nouns, or adjectives that take a PP complement, thereby eliminating the possibility that the child is using the relevant form as a pure determiner. (56) One child (Philippe) acquired direct‐object wh‐questions, wh‐questions with P‐pied‐piping, and P+D suppletive forms by the end of his corpus. (57) Statistical method: Binomial Test P+D relative frequency of & direct‐object wh‐question P‐pied‐piping P‐pied‐piping = .165 2;02 2;06 19 instances of P+D relative frequency of P+D = .835 The probability of producing 19 instances of P+D before P‐pied‐piping simply by chance: (.835)^19 = .033 ‐ 10 ‐ (58) (59) (60) 5. (61) (62) (63) 4. Results: a. The age‐discrepancy between the P+D suppletive form and P‐pied‐piping was statistically significant (p < .05, by Binomial Test). b. This indicates that Philippe acquired P‐pied‐piping significantly later than overt wh‐movement and P+D suppletion, contrary to the prediction from the parametric proposal by Law (1998, 2006). c. Our findings from the acquisition of French directly contradict Law’s view that the existence of P+D suppletive forms in a given language constitutes a sufficient condition for the obligatory pied‐piping of prepositions. Our results are compatible only with an analysis in which the availability of P+D suppletive forms is a necessary condition for P‐pied‐piping. Yet, such an analysis would be far from appealing, given that it permits adult grammars that have P+D suppletive forms but still permit P‐stranding. Evidence from child French argues against the view that the parameter of P‐stranding should relate the lack of preposition stranding to the existence of suppletive forms of prepositions and determiners. Acquisitional Evaluation [3]: P‐stranding and Swiping4 Cross‐linguistic Generalization (Hasegawa 2007): Swiping is possible only in those languages that allow P‐stranding. “P‐stranding + PP movement” analysis: (Hasegawa 2006, Kim 1997, Nakao and Yoshida 2006) (John was talking, but I don’t remember …) a. rightward movement of PP: [CP [IP he was talking [PP about what] ] b. wh‐movement + P‐stranding: [CP [IP he was talking ] [PP about what c. sluicing (IP‐deletion) in PF: [CP what [IP he was talking ] [PP ] ] ] about t ] ] Under the “P‐stranding + PP movement” approach, the grammatical knowledge required for P‐stranding constitutes a proper subset of the grammatical knowledge required for swiping. (64) (65) Prediction for Child English: English‐learning children should acquire P‐stranding with wh‐movement significantly earlier than or at around the same time as swiping. Corpora Analyzed: Child Collected by Age Span #Child Utterances 22,633 Abe Kuczaj (1976) 2;04 – 5;00 Adam Brown (1973) 2;03 – 4;10 45,555 Anne Theakston et al. (2001) 1;10 – 2;09 19,902 Aran Theakston et al. (2001) 1;11 – 2;10 17,193 Becky Theakston et al. (2001) 2;00 – 2;11 23,339 Carl Theakston et al. (2001) 1;08 – 2;08 25,084 Dominic Theakston et al. (2001) 1;10 – 2;10 21,180 Eve Brown (1973) 1;06 – 2;03 11,563 Gail Theakston et al. (2001) 1;11 – 2;11 16,973 Joel Theakston et al. (2001) 1;11 – 2;10 17,916 John Theakston et al. (2001) 1;11 – 2;10 13,390 Liz Theakston et al. (2001) 1;11 – 2;10 16,569 Naomi Sachs (1973) 1;02 – 4;09 15,960 Nicole Theakston et al. (2001) 2;00 – 3;00 16,950 Nina Suppes (1973) 1;11 – 3;03 31,505 Peter Bloom (1970) 1;09 – 3;01 26,891 Ruth Theakston et al. (2001) 1;11 – 2;11 20,419 Sarah Brown (1973) 2;03 – 5;01 37,012 Shem Clark (1978) 2;02 – 3;02 17,507 Warren Theakston et al. (2001) 1;10 – 2;09 (66) For each child, we located the first clear uses of: a. swiping b. wh‐movement involving P‐stranding. (67) Aran exhibited the first clear use of swiping at the age of 2;07. (68) a. *CHI: what in ? (Aran26a.cha) b. *CHI: who for ? (Aran27a.cha) (Aran28b.cha) c. *CHI: who from ? d. *CHI: what with ? (Aran33a.cha) This section is based on Sugisaki (2008). ‐ 11 ‐ ‐ 12 ‐ 16,651 (69) (70) P‐stranding under wh‐movement was frequently observed in Aran’s speech. The first clear use of P‐stranding appeared at the age of 2;05. Statistical method: Binomial Test P‐stranding swiping relative frequency of swiping = .40 2;05 2;07 relative frequency of P‐stranding = .60 14 instances of P‐stranding The probability of producing 14 instances of P‐stranding before swiping simply by chance: (.60)^14 = .0007 (71) The finding from child English is consistent with Hasegawa’s generalization that swiping is possible only in those languages that permit P‐stranding. The Parameter of P‐stranding: PF Approach (Tokizaki 2010, 2011) 6. 6.1. P‐stranding and Recursive Compounds (72) Cross‐linguistic Generalization (Stowell 1981): P‐stranding is possible only in those languages that permit transitive verb‐particle construction (especially the one with the order V‐Particle‐NP). (73) Snyder (2001): Transitive verb‐particle constructions are permitted only in those languages that allow recursive compounds. (74) a. English: waste disposal plan b. Italian: * rifiuti smaltimento piano (75) Cross‐linguistic Survey: Language Transitive Verb Particles? Recursive compounds? Khmer YES YES Estonian YES YES Dutch YES YES Mandarin YES YES Thai YES YES Japanese NO YES American Sign Language NO YES Egyptian Arabic NO NO Javanese NO NO Spanish NO NO Serbo‐Croatian NO NO ‐ 13 ‐ (76) 6.2. (77) (78) (79) (80) (81) Revised Cross‐linguistic Generalization: P‐stranding is possible only in those languages that allow recursive compounds. PF Approach to the Parameter of P‐stranding Tokizaki (2010, 2011): The generalization in (76) can be derived from the canonical word‐stress location in a language. The Canonical Word‐stress Location: a. English, Swedish, Norwegian … = Right‐oriented stress Stress is located on antepenultimate, penultimate, or ultimate syllable. b. French, Italian, Spanish … = Right‐edge stress Stress is located on either penultimate or ultimate syllable. Phonological Constraint on Recursive Compounds: The main stress location of compounds must correspond to the canonical word‐stress location in that language. English: a. Assign stress to the most deeply embedded element (Cinque 1993): [ plan [ disposal [ waste ] ] ] b. Movement to the specifier position: (i) [ plan [ disposal [ waste ] ] ] (ii) [ plan [ [ waste ] disposal ] ] (iii) [ [ [ waste ] disposal ] plan ] → Correspond to antepenultimate stress, and hence satisfy (79) Italian: a. Assign stress to the most deeply embedded element (Cinque 1993): [ piano [ smaltimento [ rifiuti ] ] ] b. Movement to the specifier position: (i) [ piano [ smaltimento [ rifiuti ] ] ] (ii) [ piano [ [ rifiuti ] smaltimento ] ] (iii) [ [ [ rifiuti ] smaltimento ] piano ] → Correspond to antepenultimate stress, and hence violate (79) ‐ 14 ‐ (82) Phonological Constraint on P‐stranding: a. In order for P‐stranding to be possible, the verb and the preposition must undergo “phonological word‐formation”. b. The resulting “word” must conform to the canonical word‐stress location in that language. (83) a. A verb tends to have more than one syllable, and the preposition tends to be mono‐syllabic. b. Prepositions generally do not have stress. (84) a. English: … wórking with t ( σ σ σ ) → antepenultimate stress, which satisfies (82) b. French: … traváillez avec t ( σ σ σ ) → antepenultimate stress, which violates (82) (85) a. Under what condition is ellipsis possible? b. * What condition is ellipsis possible under? (Chung et al. 1995) c. Ellipsis is possible under some condition, but we are not sure what condition. (86) Remaining problem: Cross‐linguistic Generalization (Merchant 2001:92) A language L will allow preposition stranding under sluicing iff L allows preposition stranding under regular wh‐movement. (87) German: a. Anna hat mit jemandem gesprochen, Anna has with someone spoken aber ich weiβ nicht *(mit) wem. but I know not with who b. * Wem hat sie mit gesprochen? she with spoken who has 7. Conclusion (88) Evidence from child language strengthens Stowell’s (1981) cross‐linguistic generalization that languages permitting P‐stranding are a proper subset of those permitting transitive verb‐particle construction. (89) This “parametric” variation may stem from independently motivated cross‐linguistic differences in the “externalization” (i.e. PF) component. (90) Child language is potentially a very useful tool to find out “significant” cross‐linguistic generalizations, those that should be subjected to minimalist / biolinguistic scrutiny and reformulation. References Almeida, Diogo, and Masaya Yoshida. 2007. A problem for preposition stranding generalization. Linguistic Inquiry 38, 349‐642. Bloom, Lois. 1970. Language Development: Form and Function in Emerging Grammars. Cambridge, Massachusetts: MIT Press. Bloom, Lois. 1973. One Word at a Time: The Use of Single Word Utterances before Syntax. The Hague: Mouton de Gruyter. Boeckx, Cedric. 2010. What principles and parameters got wrong. Ms., ICREA/UAB. Brown, Roger. 1973. A First Language: The Early Stages. Cambridge, Massachusetts: Harvard University Press. Chomsky, Noam. 1981. Lectures on Government and Binding. Foris: Dordrecht. Chomsky, Noam. 1995. The Minimalist Program. Cambridge, Massachusetts: MIT Press. Chomsky, Noam. 2007. Approaching UG from below. In Interfaces + Recursion = Language? Chomsky’s Minimalism and the View from Semantics, ed. Uli Sauerland and Hans‐Martin Gärtner. Berlin: Mouton de Gruyter. Chomsky, Noam. 2010. Some simple evo devo theses: How true might they be for language? In The Evolution of Human Language, ed. Richard K. Larson, Viviane Déprez, and Hiroko Yamakido, 45‐62. Cambridge: Cambridge University Press. Chung, Sandra, William Ladusaw, and James McCloskey. 1995. Sluicing and logical form. Natural Language Semantics 3, 239‐282. Cinque, Guglielmo. 1993. A null theory of phrase and compound stress. Linguistic Inquiry 24, 239‐298. Clark, Eve. 1978. Strategies for communicating. Child Development 49:953‐959. van Craenenbroeck, Jeroen. 2010. The Syntax of Ellipsis: Evidence from Dutch Dialects. New York: Oxford University Press. Hasegawa, Hiroshi. 2007. Swiping involves preposition stranding, not pied‐piping. A paper presented at GLOW 30, University of Tromso, Norway. April 12‐14. Higginson, Roy. 1985. Fixing: A Demonstration of the Childʹs Active Role in Language Acquisition. Doctoral dissertation, Washington State University, Washington. Isobe, Miwa, and Koji Sugisaki. 2002. The Acquisition of pied‐piping in French and its theoretical implications. Paper presented at Going Romance 2002, Workshop on Acquisition, Rijksuniversiteit Groningen. Kim, Jeong‐Seok. 1997. Syntactic Focus Movement and Ellipsis: A Minimalist Approach. Doctoral dissertation, University of Connecticut, Storrs. Kuczaj, Stan. 1976. ‐ing, ‐s, and ‐ed: A study of the acquisition of certain verb inflections. Doctoral dissertation, University of Minnesota. Law, Paul. 1998. A unified analysis of P‐stranding in Romance and Germanic. In Proceedings of NELS 28, eds. Pius N. Tamanji and Kiyomi Kusumoto, 219‐234. University of Massachusetts, Amherst: GLSA. Law, Paul. 2006. Preposition stranding. In The Blackwell Companion to Syntax, eds. Martin Everaert and Henk van Riemsdijk, 631‐684. Oxford: Blackwell. MacWhinney, Brian. 2000. The CHILDES project: Tools for analyzing talk. Mahwah: Lawrence Erlbaum. ‐ 15 ‐ ‐ 16 ‐ Mailing, Joan, and Annie Zaenen. 1985. Preposition‐stranding and oblique case. In Cornell Working Papers in Linguistics 7, 149‐161. Merhcant, Jason. 2001. The Syntax of Silence: Sluicing, Islands, and the Theory of Ellipsis. Oxford: Oxford University Press. Merchant, Jason. 2002. Swiping in Germanic. In Studies in Comparative Germanic Syntax, eds. C. Jan‐Wouter Zwart and Werner Abraham, 295‐321. Amsterdam: John Benjamins. Nakao, Chizuru, and Masaya Yoshida. 2006. ʺNot‐so‐propositionalʺ islands and their implications for swiping. Proceedings of Western Conference on Linguistics (WECOL) 2006, 322‐333. California State University, Fresno, CA. (online publication) Rodrigues, Cilene, Andrew Ira Nevins and Luis Vicente. 2009. Cleaving the interactions between sluicing and P‐stranding. In Romance Languages and Linguistic Theory 2006, eds. Danièle Torck and W. Leo Wetzels , 175–198. Amsterdam: John Benjamins. Rosen, Carol. 1976. Guess what about? Proceedings of the 6th Annual Meeting of the North Eastern Linguistic Society, 205‐211. Ross, John Robert. 1969. Guess who? Papers from the 5th Regional Meeting of the Chicago Linguistic Society, 252‐286. Sachs, Jacqueline. 1973. Talking about here and then: The emergence of displaced reference in parent‐child discourse. In Childrenʹs Language, Volume 4, ed. K. E. Nelson. Hillsdale: Lawrence Erlbaum. Sato Yosuke. In press. P‐stranding under sluicing and repair by ellipsis: Why is Indonesian (not) special? Journal of East Asian Linguistics. Snyder, William. 2007. Child Language: The Parametric Approach. New York: Oxford University Press. Stjepanović, Sandra. 2008. P‐stranding under sluicing in a non‐P‐stranding language? Linguistic Inquiry 39, 179‐190. Stowell, Timothy. 1981. Origins of Phrase Structure. Doctoral dissertation, MIT, Cambridge, Massachusetts. Stromswold, Karin. 1996. Analyzing childrenʹs spontaneous speech. In Methods for assessing childrenʹs syntax, eds. Dana McDaniel, Cecile McKee and Helen Smith Cairns, 23‐53. Cambridge, Massachusetts: MIT Press. Sugisaki, Koji. 2008. Swiping in child English. In Proceedings of the 27th West Coast Conference on Formal Linguistics Poster Session, ed. Kevin Ryan, 71‐77. UCLA Working Papers in Linguistics. Sugisaki, Koji, and William Snyder. 2002. Preposition stranding and the compounding parameter: A developmental perspective. In Proceedings of the 26th annual Boston University Conference on Language Development, eds. Barbora Skarabela, Sarah Fish and Anna H.‐J. Do, 677‐688. Somerville, Massachusetts: Cascadilla Press. Suppes, Patrick. 1973. The semantics of childrenʹs language. American Psychologist 88:103‐114. Suppes, Patrick, R. Smith, and M. Léveillé. 1974. The French syntax of a childʹs noun phrase. Archives de Psychologie 42, 207‐269. Szczegielniak, Adam. 2008. Islands in sluicing in Polish. In Proceedings of the 27th West Coast Conference on Formal Linguistics, eds. Natasha Abner and Jason Bishop. 404‐412. Somerville, MA: Cascadilla Press. ‐ 17 ‐ Theakston, Anna. L., Elena V. M. Lieven, Julian M. Pine, & Caroline F. Rowland. 2001. The role of performance limitations in the acquisition of verb‐argument structure: an alternative account. Journal of Child Language 28: 127‐152. Tokizaki, HIsao. 2010. Recursive compounds and word‐stress location. Paper presented at On Linguistic Interfaces II, University of Ulster, Belfast, Ireland. Tokizaki, Hisao. 2011. Stress location and the acquisition of morpho‐syntactic parameters. In WCCFL 28 Online Proceedings. https://sites.google.com/site/wccfl28pro/tokizaki Truswell, Robert. 2009. Preposition‐stranding, passivisation, and extraction from adjuncts in Germanic. In Linguistic Variation Yearbook 9, eds. Jeroen van Craenenbroeck and Johan Rooryck, 131‐177. ‐ 18 ‐