Journal of Greek Linguistics 10 (2010) 143–147

brill.nl/jgl

Dissertation Summaries Kalliopi Katsika. Sentence processing strategies in children and adults: PP Attachment in corpora and psycholinguistic experiments. Aristotle University of Thessaloniki, 2009. [[email protected]] Keywords child language, corpus linguistics, experimental linguistics, PP-attachment, sentence processing

1 Introduction A very important goal of psycholinguistic research is to unravel the architecture and mechanisms of language comprehension. Sentence comprehension is a very rapid process which involves the immediate construction of incoming constituents into larger syntactic and semantic units. It is generally believed that sentence processing occurs incrementally, in a word-by-word fashion, and that the human sentence processor (parser) starts building constituent structures and interpretations the moment each word is heard or read (e.g. MarslenWilson 1973, 1975). The incremental nature of sentence comprehension may create brief ambiguity even to strings of spoken or written language that are unambiguous upon the completion of a sentence. Structural ambiguity thus provides a way to investigate the mechanisms that underlie the operation of the parser (Frazier 1979). This thesis examines a specific type of structural ambiguity, namely the Prepositional Phrase (PP) attachment ambiguity. Even though a considerable amount of research has been conducted on PP attachment ambiguities in English (Altmann & Steedman 1988, Britt 1994, Clifton, Speer, & Abney 1991, Rayner, Carlson, & Frazier 1983, Schütze & Gibson 1999, Taraban & McClelland 1988, to name a few), little research can be found in languages other than English (Frenck-Mestre & Pynte 1997, Konieczny, Hemforth, Scheepers, & Strube 1997, for French and German respectively) and no research on attachment preferences in V-NP-PP structures has been previously conducted in Greek. The main aim of this thesis is to investigate the extent to which native Greek speakers’ parsing decisions for temporary ambiguous prepositional phrase © Koninklijke Brill NV, Leiden, 2010

DOI 10.1163/156658410X491811

144

K. Katsika / Journal of Greek Linguistics 10 (2010) 143–147

structures are influenced by lexical, language-specific grammatical and frequency factors. 2 PP attachment ambiguity The PP attachment ambiguity involves the possible attachment of a PP either to the preceding VP or to the preceding NP, and can be illustrated in sentences such as: (1) The girl hit the boy with the apple. In verb-noun phrase-prepositional phrase (V-NP-PP) sequences such as (1), the ambiguity lies in the possibility of attaching the PP with the apple either to the preceding verb hit denoting the instrument of the action described by the verb, or to the preceding NP boy, as a modifier of the NP. In order to explore the frequency of PP attachment in naturally produced sentences of written and spoken register and to be able to make valid claims about the role that prior exposure to comparable structures in Greek may have on parsing decisions, a corpus study was conducted. Sample sentences were extracted from two types of corpora: (i) a very large corpus which includes written texts in Modern Greek (Institute for Language and Speech Processing corpus), and (ii) a manually compiled corpus of spoken Greek. The corpus data were analysed on two levels (coarse-grained vs. fine-grained analysis). The coarse-grained analysis allowed testing the predictions of the Tuning Hypothesis (Mitchell, Cuetos, Corley & Brysbaert 1995), which posits that the parser is expected to be “tuned” only to syntactic category (coarse-grained) information during on-line sentence comprehension. Alternatively, constraint-based accounts of sentence comprehension assume that parsing preferences are reflected on corpus frequencies at more fine-grained levels of analysis (e.g. Spivey-Knowlton & Sedivy 1995), at which information such as the lexical choice of preposition and definiteness are taken into account. The results of the written and spoken corpus data analysis indicated that VP attachment is the most frequent type of attachment in Greek V-NP-PP structures and that attachment frequencies vary depending on the preposition that heads the PPs: VP attachment was more frequent than NP attachment for me, se and apo-PPs whereas the opposite pattern was found for ja-PPs. In addition, the thesis explores native Greek speakers’ parsing preferences for temporary ambiguous PP structures by means of three psycholinguistic experiments (one off-line acceptability judgement and two on-line self-paced reading tasks). Two groups of adults and one group of 11- to 12-year old children participated in the study. All participants were native speakers of Greek

K. Katsika / Journal of Greek Linguistics 10 (2010) 143–147

145

and had never spent more than six months in a country other than Greece. The participants read sentences and made grammaticality decisions about items in which PPs were (pragmatically) forced towards either a VP attachment interpretation (2a) or an NP attachment one (2b). (2) a. O kipuros ekopse maxeri. (VP attachment)

to

klaði

me to/ena

The-nom gardener-nom cut-perf.sg the-acc branch-acc with the/a-acc knife-acc

‘The gardener cut the branch with the/a knife’ b. O kipuros ekopse to klaði luluði. (NP attachment)

me to/ena

The-nom gardener-nom cut-perf.sg the-acc branch-acc with the/a-acc flower-acc

‘The gardener cut the branch with the/a flower’ The primary aim of the psycholinguistic tasks was to investigate which type of PP attachment (VP vs. NP) would be read faster or rated as more grammatical by the participants. The experimental materials included four Greek prepositions (me [with], se [in/into], apo [from/by], ja [for]) in order to examine the extent to which the lexical choice of P affects native Greek speakers’ attachment preferences. The data were also manipulated in terms of the definiteness of the NP complement of P (definite vs. indefinite) so as to examine the extent to which a language-specific grammatical phenomenon which involves the postulation of multiple determiners in instances of nominal and adjectival modification (Definiteness Agreement) extends to complex object DPs with prepositions me, se, apo and ja (see Stavrou 2003, Stavrou & Tsimpli 2009). The results of the off-line and on-line tasks revealed a general VP attachment preference in both children and adults. The VP attachment preference was largely dependent on the lexical choice of preposition in the adult group whereas the insertion of different prepositions in the PP site affected children’s parsing decisions to a lesser extent. Although VP attachment sentences with apo-, ja- and me-PPs were initially read faster than NP attachment ones, there was no attachment preference for sentences with se-PPs. In addition, definite NP attachment sentences (to klaði me to luluði) were read significantly faster than indefinite ones (to klaði me ena luluði) by both children and adults in sentences with me- and se-PPs (but not with apoand ja-PPs). These results show that Greek native speakers were strongly garden-pathed by the indefiniteness of the NP complement of P, and provide further support to the claim that Definiteness Agreement extends to complex DPs (DP+PP). The fact that this effect was only found in me- and se-PPs

146

K. Katsika / Journal of Greek Linguistics 10 (2010) 143–147

indicates that the lexicality of P plays an important role in the licensing of the phenomenon (Stavrou & Tsimpli 2009). In addition, the investigation of children’s online parsing preferences revealed that children at the age of 11 to 12 employ essentially the same parsing strategies as adults but are less sensitive to lexical information than adults. These results are consistent with accounts that postulate continuity of parsing strategies (see Felser, Marinis & Clahsen 2003).

3 Conclusion This dissertation investigates the effect of frequency and the role of lexical and language specific grammatical factors on Greek native speakers’ (children and adults) processing of temporarily ambiguous PP structures. Taken together, the results of the corpus analyses and the psycholinguistic data indicate that there is a hierarchy of lexicality among prepositions me, se, apo and ja and that, although there is a degree of correspondence between frequency counts and parsing preferences (on a coarse-grained level), frequency alone cannot account for native Greek speakers’ parsing preferences.

References Altmann, Gerry, and Mark Steedman. 1988. Interaction with context during human sentence processing. Cognition 30: 191-328. Britt, M. Anne. 1994. The interaction of referential ambiguity and argument structure in the parsing of Prepositional Phrases. Journal of Memory and Language 33: 251-283. Clifton, Charles, Shari Speer and Steven Abney. 1991. Parsing arguments: phrase structure and argument structure as determinants of initial parsing decisions. Journal of Memory and Language 30: 251-271. Felser, Claudia, Theodore Marinis and Herald Clahsen. 2003. Children’s processing of ambiguous sentences: A study of relative clause attachment. Language Acquisition 11: 127-163. Frazier, Lyn. 1979. On comprehending sentences: Syntactic parsing strategies. West Bend, IN: Indiana University Linguistics Club. Frenck-Mestre, Cheryl and Joel Pynte. 1997. Syntactic ambiguity resolution while reading in second and native languages. The Quarterly Journal of Experimental Psychology A, 50 (1): 119-148. Konieczny, Lars, Barbara Hemforth, Christoph Scheepers, and Gerhard Strube. 1997. The role of lexical heads in parsing: evidence from German. Language and Cognitive Processes 12 (2/3): 307-348. Marslen-Wilson, William D. 1973. Linguistic structures and speech shadowing at very short latencies. Nature 244: 522-523. Marslen-Wilson, William D. 1975. Sentence perception as an interactive parallel process. Science 189: 226-228.

K. Katsika / Journal of Greek Linguistics 10 (2010) 143–147

147

Mitchell, Don C., Fernando Cuetos, Martin M. B. Corley, and Marc Brysbaert. 1995. Exposurebased models of human parsing: Evidence for the use of coarse-grained (nonlexical) statistical records. Journal of Psycholinguistic Research 24: 469-488. Rayner, Keith, Marcia Carlson, and Lyn Frazier. 1983. The interaction of syntax and semantics during sentence processing: eye movements in the analysis of semantically biased sentences. Journal of Verbal Learning and Verbal Behavior 22: 358-374. Schütze, Carson T. and Edward Gibson. 1999. Argumenthood and english prepositional phrase attachment. Journal of Memory and Language 40: 409-431. Spivey-Knowlton, Michael J. and Julie Sedivy. 1995. Resolving attachment ambiguities with multiple constraints. Cognition 55: 227-267. Stavrou, Melita. 2003. Semi-lexical nouns, classifiers, and the interpretation(s) of the pseudopartitive construction. In M. Coene, & Y. D’hulst (Eds.) From NP to DP: The Syntax and Semantics of Noun Phrases (Linguistik Aktuell 55). 329-353. Stavrou, Melita and Ianthi Maria Tsimpli. 2009. Definite Agreement in Complex Noun Phrases. Studies in Greek Linguistics 29: 193-206. Taraban, Roman and James McClelland. 1988. Constituent attachment and thematic role assignment in sentence processing: influences of content-based expectations. Journal of Memory and Language 27: 597-632.