2018 Friday Session A 1400

Friday, November 2, 2018 | Session A, East Balcony | 2pm

Using Prosody to Compute Alternative Sets: The Case of Turkish particle dA
S. Topaloglu, M. Nakipoglu

Both foci and contrastive topics are known to give rise to alternatives in discourse (Rooth 1992, Büring 2016). Interestingly, Turkish has a unique polysemous particle, dA (da/de wrt vowel- harmony), which can signal both additive-focus(AddFoc)(1a) and contrastive topics(CT)(1b), where the two are differentiated on prosodic grounds(Göksel & Özsoy 2003). When dA follows and is associated with a focused & stressed subject(1a), the sentence presupposes that there is at least one other alternative (e.g., Elmo) that satisfies the open proposition. When dA follows and is associated with an unstressed subject(1b) and when the object is focused & stressed, two sets of alternatives (one for the contrastive-topic and one for the focus) are at issue. (1b) presupposes that there is at least one other alternative subject(e.g., Elmo) who reads something that constitutes an alternative to the focused object(e.g., a book).

This study investigated whether Turkish-speaking children can differentiate between these two uses of dA, and comprehend the dissimilar presuppositions they evoke. We created a novel task with a detective story where two puppets spying on Elmo and Kermit appeared in videos on a computer screen and exchanged information via their walkie-talkies, uttering pre-recorded test-sentences. The antecedent sentence that satisfies the existential presupposition of the test sentence was assumed to be uttered by the character at the other end of the line –but it was heard as gibberish(“the cartoon phone- call sound-effect”). This was followed by the test-sentence. Then children had to indicate which actions Kermit and Elmo performed, by selecting 2 pictures(one for Elmo, and one for Kermit). For example, as sentence(1a) suggests that both Kermit and Elmo are reading newspapers, the correct pictures are Picture-3 and Picture-4, while the correct pictures for(1b) are Picture-1 and Picture-4. After selecting the pictures, children were asked to explain their choice by referring to the walkie- talkie conversation they heard. There were 4 test-sentences with additive dA and 4 test-sentences with contrastive-topic-signalling dA.

The results showed that both First-graders and Second-graders performed significantly above chance-level in both additive-focus and contrastive-topic conditions, so they were able to use prosodic cues to discriminate between the two structures and compute different alternative sets. Preschoolers, too, performed well with contrastive-topic sentences; but in the additive-focus sentences, their performance was not significantly different from chance(see Table-1). Thus Turkish-speaking children are shown to reach an adult-like grasp of contrastive structures as signaled by dA at least starting from age 5. There is, however, no evidence for comprehension of additive-focus in the youngest age group. Preschoolers were significantly worse at interpreting additive-focus dA-sentences compared to the older groups(see Table-2 for between-group comparisons). We propose that the low performance in the additive condition can be attributable to the theoretical asymmetry between subject-focus (exemplified by additive dA-sentences) and object-focus (exemplified by contrastive-topic dA– sentences), induced by the neutral prosody of object-focus utterances(Szendrői 2017). Previous acquisition research demonstrated that such a bias for default object-focus can prompt errorful object-focus readings in the presence of exclusive focus-particles(Kim 2011) and it seems the same bias may be detected here with additive focus-particles as well.