Reconsidering the epistemic step: The role of speaker and listener perspectives for the processing of quantity and temporal implicatures
Final Report Abstract
In my project, I investigated the role of world knowledge, relevance and information access in deriving implicatures. The project consists of two major parts focusing on (i) scalar implicatures related to the use of quantifiers, (ii) temporal implicatures that arise in relation to conjunctive past tense sentences. In Gricean tradition, a certain level of mindreading is considered to be required for pragmatic inferencing. For instance, from the speaker’s statement “Some students passed the exam”, one can infer that the speaker has a reason for not using the stronger alternative “all”, which otherwise would be expected (Maxim of Quantity). This could be the case since she does not know about whether all students passed, or since she believes a statement with “all” to be false. Thus, inferring that not all students passed the exam (scalar implicature) requires the assumption that the speaker is informed about the status of the stronger alternative (competence assumption). Although the Gricean framework assumes that this sort of epistemic reasoning is involved in inferring implicatures, such an interaction between language understanding and our believes about the speaker’s knowledge states is not predicted by strongly modular theories that place scalar implicatures within the semantics module. In the first part of the project, I designed ERP experiments to investigate the role of competence assumption in deriving and processing scalar implicatures. We showed that in contexts in which the speaker’s competence could not be assumed, scalar implicatures were derived less frequently and less incrementally. In addition, contexts with only a partial information access to the domain led to reprocessing mechanisms whenever the full information was relevant for the sentence evaluation, which was marked with a sustained negativity effect. In the second part, I investigated temporal implicatures, that is, inferences of temporal order arising in relation to past tense event reports. For instance, “Julia finished her PhD and published her results” is truth-conditionally equivalent to “Julia published her results and finished her PhD”; yet, these two sentences suggest a different order of events, which is consistent with the order in which the events are mentioned. In Gricean pragmatics, this phenomenon is explained in relation to the Maxim of Manner requiring the interlocutors to be orderly in their communication. However, it is debated whether such inferences arise by means of Gricean-like mechanism and involve reasoning about the speaker’s believe states or other aspects of the context, or whether they arise in a default manner, due to certain constrains of the discourse structure. Given that no prior ERP studies on order effects in conjunctions are available, I decided to primary focus on the question of how violations of temporal order in conjunctions are processed and whether the relevance of the order in the context modulates the processing of reversed order conjunctive sentences. As a novel result, we show that order violations led to a P600 effect on the first event that violated the temporal order. Similarly, sentences violating temporal order expectations based on our real-life experiences showed a P600 effect relative to sentences that followed the expected order. A modulation of the N400 in response to temporal order expectations was also observed but only if the temporal order was relevant in the global context. In addition, sentences that were semantically false due to the use of a specific temporal connective led to larger N400 ERPs relative to pragmatically infelicitous sentences with “and”, supporting the view that the order effects in conjunctions arise pragmatically. The results of these studies provide a ground for further experiments that would directly focus on the interaction between the speaker knowledge state and temporal implicatures.
Publications
-
The cost of the epistemic step: Investigating scalar implicatures in full and partial information contexts. Frontiers in Psychology: Language Science (Experimental Approaches to Pragmatics)
Spychalska, M.; Reimer, L.; Schumacher, P. & Werning, M.
