Neurolinguistic and psycholinguistic approaches to studying tense, aspect, and unaccusativity

The experimental investigation of grammatical aspect was the heart and soul of this PhD project. Nevertheless, since aspect is semantically related to tense as they both convey temporal information in the sentence, we compared the processing of tense and aspect. Furthermore, we experimentally explored the interplay between aspect and unaccusativity for the linguistic literature suggests an inherent link between perfectivity and unaccusativity. In essence, the concepts of tense, aspect, and unaccusativity were described linguistically and studied experimentally in the four chapters of the thesis.

In Chapter 1, we provided a theoretical description of tense, aspect, and unaccusativity. We discussed tense and aspect realization and distribution in Bosnian/Croatian/Serbian (BCS) with cross-linguistic comparisons to English. We defined the syntax and semantics of unaccusative verbs as well as the interplay between perfective aspect and unaccusativity. In the end of the chapter, we formulated three research questions that our three experiments were designed to answer:
1. Do native speakers recognize tense and aspect violations in BCS at the critical word?
2. What are the electrophysiological correlates of aspect violations?
3. What is the relationship between aspect and unaccusativity?

In Chapter 2, we presented the set-up and the results of the behavioral self-paced reading (SPR) experiment on tense and aspect processing as well as the ERP experiment on aspect processing in BCS that attempted to answer research questions 1 and 2. In the self-paced reading experiment, we investigated the processing of tense and aspect violations in BCS. As an answer to the first research question, we showed that tense violations are not detected online at the critical word where the violation is disambiguated whilst aspect violations are detected at the point of the violation. We argued that aspectual meanings conveyed by the perfective and the imperfective verb forms are rather distinct which made aspect violations salient. For tense violations, the parser did not register the violation of a future time frame of a sentence by a past tense verb at the spot because past tense verbs can have future time reference in restricted contexts. We argued that online, the parser leaves open all possibilities and activates all the meanings of the sentence. After the sentence had been processed, in the offline grammaticality judgment task, the native speakers rejected tense violations because they selected the canonical interpretation of the past tense verb form
In the ERP experiment, we investigated the ERP correlates of aspect processing in BCS. The collected ERPs show a P600 for aspect violations. We summarized the behavioral and the ERP data on aspect processing in BCS claiming that aspect violations in BCS are salient and processed at the critical word. We also made a cross-linguistic comparison with the English time reference system. We argued for a clear difference between BCS and English aspect systems. In BCS, aspectual meanings are straightforwardly encoded by either perfective or imperfective verb form, whereas in English, one aspectual meaning can be conveyed by different forms which means that English aspect violations are not salient.

In Chapter 3, we presented the set-up and the results of the behavioral cross-modal lexical priming (CMLP) experiment that investigated the interplay between aspect and unaccusativity that was formulated in the third research question of this PhD project. The results of our CMLP experiment show the inextricable link between aspect and unaccusativity and that only subjects of perfective unaccusative verbs forms are base-generated as internal arguments. Finally, we compared the results of our experiment on aspect and unaccusative verbs in BCS to a study in English unaccusatives by Friedmann et al. (2008). We addressed cross-linguistic differences and how they affect processing. We suggested that English unaccusative verbs in their base form express perfective meaning too. Finally, we pointed out that perfectivity might be a universal feature of unaccusativity.

In Chapter 4, we reminded the reader of the current views of tense, aspect and unaccusativity in the linguistic literature. We outlined the three research questions of the PhD project and explained how our three experiments contributed to answering those research questions. We provided a comprehensive description of the findings of our three experiments and their implications. We also made an observation that is at the core of our experimental results and that is in line with the immensely important finding of Swinney (1979): When a verb form is used to convey more than one meaning, all the meanings are activated and subsequently, in the course of the sentence, the relevant interpretation is selected and other interpretations are discarded. Finally, we acknowledged the limitations of our three studies and provided suggestions for future research.