| Popis: |
Research investigating the timing of verb selection during production has found evidence for both early (Lindsley, 1975; Kempen & Huijbers 1983) and just-in-time verb planning (Schriefers et al. 1998; Allum & Wheeldon, 2007). Early verb planning, also known as the verb guidance hypothesis (Momma & Ferreira, 2019) advocates that the verb lemma is needed for grammatically encoding verbal arguments and hence the verb is retrieved prior to the formulation of these arguments. Just-in-time verb planning also known as the conceptual guidance hypothesis advocates that arguments can be grammatically encoded verb independently based on the conceptual message of the sentence. Thus, the verb does not need to be necessarily retrieved early thereby reducing the cost of maintaining the verb in working memory until its eventual articulation. More recently, it has also been found that verb selection can vary based on the typological properties of the language (Hwang & Kaiser, 2014): early retrieval for SVO languages and just-in-time for SOV languages or the type of semantic and syntactic dependency between the verb and its arguments (Momma et al., 2016, 2018; Momma & Ferreira, 2019): verb retrieval prior to the internal argument and not the external argument. The current study aims to investigate another source of variability in the timing of verb retrieval: the necessity of the verb for case assignment. When case assignment to a sentential argument depends on the verb, then it might be crucial to retrieve the verb lemma in advance. However, when case assignment can proceed relatively independently then such an early retrieval might not be needed. The possibility of such a differential verb retrieval strategy will be investigated via the phenomenon of split ergativity in Hindi. In Hindi, the ergative marker -ne is assigned to subjects of transitive verbs only in the perfective aspect, and the nominative case marker, which is phonologically null appears elsewhere. Compare (1) which requires an ergative marker on the subject because the verb is in the perfective aspect with (2) where the verb is in progressive aspect and the subject is assigned a nominative case marker. (1) daktarni-ne nurse-ko bulaya (The doctor has called the nurse) (2) daktarni nurse ko bulaa rahi hai (The doctor is calling the nurse) If ergative case assignment to the subject relied on perfectivity alone, then early verb retrieval may not be needed, conceptual knowledge of an event that has been completed might suffice. However, perfectivity alone does not dictate ergative case assignment in Hindi. There is a range of transitive predicates that do not take up the ergative marker even in the perfective aspect and these cannot be categorized into a class sharing some common features (e.g. agentivity, kinesis, volition). Sentences (3) and (4) are in the perfective aspect, yet they fail to take up the ergative marker: (3) kisaan phal laaya (the farmer brought fruits); (4) kisaan vakeel-se milaa (the farmer met the lawyer). Thus, a conceptual route might not be a failsafe measure for assigning ergative case to subjects in Hindi and it seems that lexical access to the verb might be crucial. In contrast to ergative case, nominative case assignment to subjects is more regular and advanced verb selection may not be needed. Based on such a rationale the research question is whether case induced syntactic dependencies influence the timing of verb retrieval in production or not. The question will be investigated using the extended Picture Word Interference Paradigm (Momma & Ferreira, 2019; Schriefers et al. 1998) ). Participants will describe images of transitive events using simple sentences. Superimposed upon the images will be words that are either semantically related or unrelated to the target verb. Observing a semantic interference effect at speech onset (longer latencies in the related distractor condition as compared to unrelated distractor condition) shall be indicative of early verb retrieval. |
