We are very happy to announce the next talk in the GK Colloquium, which will take place on Tuesday, June 4, 4 – 6 pm in SH 5.105.

Professor Elsi Kaiser (University of Southern California) will present „Head-final relative clauses and animacy effects: What corpus patterns and psycholinguistic studies can tell us“.


Animacy guides language processing in deep-reaching ways. In this talk, I explore the consequences of animacy for the production and processing of relative clauses, using corpus data and psycholinguistic studies. I will mostly focus on data from Mandarin Chinese and, if time permits, I will also present some preliminary data from Finnish. Both of these languages have relative clause structures that differ syntactically from Indo-European relative clauses in ways that can inform our understanding of how animacy influences fundamental aspects of language processing, such as argument structure.

It is well-known that crosslinguistically, animate entities tend to occur in subject position (often also in the sentence-initial position). However, much of the prior work has focused on simple matrix clauses, and less is known about embedding structures such as relative clauses. Gaining a better understanding of the effects of animacy on the processing of relative clauses can provide insights into the notion of animacy, in particular its relation to thematic and syntactic roles, and can inform debates about principles that guide on-line language processing.

Relative clauses (RCs) are a means of providing more information about a noun (e.g. the girl read the book => the girl [who lives next door] read the book). In many languages, this head noun can occur in various positions within the RC: In subject-extracted RCs, the head noun is the subject of the RC structure (e.g. the girl [who __ chased the cat]), and in object-extracted RCs, the head noun is the object (e.g. the girl [who the cat chased __ ] ).

I report a series of psycholinguistic studies and corpus analyses investigating animacy patterns in relative clauses in Mandarin Chinese, and how different animacy configurations are processed (joint work with F. Wu, E. Andersen). There is an ongoing debate concerning the processing ease of subject- and object-extracted RCs which has implications for models of sentence processing. In English, subject-extracted RCs are easier to process than object-extracted RCs, a finding that current theories correctly predict. Crucially, because Chinese RCs occur before the head noun (e.g. subject-extracted: [ _ chased cat DE] girl => the girl [who __ chased the cat], object-extracted: [cat chased __ DE ] girl => the girl [who the cat chased __] ), different theories of sentence processing yield different predictions about which RC type (subject-/object-extracted) should be easier to process. Prior experiments on Chinese have yielded mixed results.

We argue that it is crucial to take the animacy configurations of Chinese RCs into account when investigating their processing. Based on corpus data and experiments, we show that manipulating the animacy of the head and the RC-internal noun offers new insights into the debate on the ease of processing subject and object RCs, and indicates that subject-extracted RCs are easier to process than object-extracted RCs, even in Chinese. Time permitting, I will also discuss recent work (joint work with with F. Wu, S. Vasishth) that relates to the question of how processing ease relates to structural complexity vs. the frequency of a certain structure.

If time permits, I will also present preliminary data from Finnish relative-clause-type nominalized structures. Finnish has both ‘Indo-European-style’ RCs that follow the head noun (tyttö, [joka __ seurasi kissaa] ‘girl-NOM [who-NOM __ followed cat-PART]’), as well as nominalized structures that precede the noun (Kissaa seurannut tyttö [cat-PART follow-PTCP2 girl-NOM]). Thus, Finnish allows us to explore potential effects of linear position (pre-/post-head) on the role of animacy in RCs, to see how they relate to claims that animate nouns tend to linearly precede inanimate nouns due to animate nouns being more accessible.

More broadly, these results from corpus work and psycholinguistic studies highlight the relation between animacy and grammatical role patterns in relative clauses, and the role that animacy plays in on-line processing.

You are cordially invited!