Gail Mauner

Arguments for adjuncts

It is commonly assumed across the language sciences that some semantic participant information is lexically encoded in the representation of verbs and some is not. In this paper, we propose that semantic obligatoriness and verb class specificity are criteria which influence whether semantic information is lexically encoded. We present a comprehensive survey of the English verbal lexicon, a sentence continuation study, and an on-line sentence processing study which confirm that both factors play a role in the lexical encoding of participant information.

What with? The Anatomy of a (Proto)-Role

This paper describes a comprehensive survey of English verbs that semantically allow or require an Instrument role. It sheds light on the nature of Instrument roles and instrumentality by examining the distribution in semantic space of those verbs. We show first that verbs that semantically require instruments are typically semantically more complex than predicted by current theories of the structural complexity of verb meanings. We also show that verbs that require or allow instruments constrain the end states of situations they describe more than they constrain the agents initial activity. Our survey further suggests that the causal role played by the instrument is more varied than suggested by previous studies and requires the introduction of a new subtype of causal relation, which we dub helping. Finally, our survey demonstrates that verbs that semantically require an instrument cluster together more closely in semantic space and constrain the instruments (causal) role and properties more than verbs that merely allow the presence of an instrument.

Semantic similarity, predictability, and models of sentence processing

The effects of word predictability and shared semantic similarity between a target word and other words that could have taken its place in a sentence on language comprehension are investigated using data from a reading time study, a sentence completion study, and linear mixed-effects regression modeling. We find that processing is facilitated if the different possible words that could occur in a given context are semantically similar to each other, meaning that processing is affected not only by the nature of the words that do occur, but also the relationships between the words that do occur and those that could have occurred. We discuss possible causes of the semantic similarity effect and point to possible limitations of using probability as a model of cognitive effort.

A-definites and the Discourse Status of Implicit Arguments

This paper focuses on the semantics of implicit arguments and compares it with that of explicit indefinites with which they can be truth-conditionally paraphrased. It is shown that once the discourse-potential of expressions is taken into account, the semantics of implicit arguments differs from their indefinite explicit counterparts. They are shown to be semantically identical to a particular kind of non-quantificational NP (a-definites) which are characterized by their inability to serve as antecedents for future reference. A model of this behavior of implicit arguments, it is argued, follows naturally from the underlying assumption of Discourse Representation Theory that semantic representations must include two kinds of information, a set of available discourse markers and a set of predicative conditions. Because implicit arguments satisfy a predicates argument positions without introducing discourse markers into the Discourse Representation Structure of a sentence, they cannot serve as the antecedent of definite pronouns. When they do enter into anaphoric relations it is not through discourse markers equality clauses, but instead is the result of either lexical identification of variables (via semantic detransitivization or meaning postulates) or of an accommodation process which involves bridging and/or factoring interferences.

When is schematic participant information encoded? Evidence from eye-monitoring

Abstract Two eye-monitoring studies examined when unexpressed schematic participant information specified by verbs is used during sentence processing. Experiment 1 compared the processing of sentences with passive and intransitive verbs hypothesized to introduce or not introduce, respectively, an agent when their main clauses were preceded by either agent-dependent rationale clauses or adverbial clause controls. While there were no differences in the processing of passive clauses following rationale and control clauses, intransitive verb clauses elicited anomaly effects following agent-dependent rationale clauses. To determine whether the source of this immediately available schematic participant information is lexically specified or instead derived solely from conceptual sources associated with verbs, Experiment 2 compared the processing of clauses with passive and middle verbs following rationale clauses (e.g., To raise money for the charity , the vase was/had sold quickly …). Although both passive and middle verb forms denote situations that logically require an agent, middle verbs, which by hypothesis do not lexically specify an agent, elicited longer processing times than passive verbs in measures of early processing. These results demonstrate that participants access and interpret lexically encoded schematic participant information in the process of recognizing a verb.

Lexical Encoding of Event Participant Information

The typical understanding of a short passive sentence like The ship was sunk is that it was sunk by someone. We investigated whether unexpressed agent information is introduced via semantic argument information associated with the lexical representations of verbs, or instead via conceptually derived inferences. We demonstrate, in self-paced reading and eye-monitoring studies, that implicit agents are derived from lexical rather than conceptual sources and that verb argument structure information is accessed as soon as a verb is recognized.

Class Specificity and the Lexical Encoding of Participant Information

It is commonly assumed across the language sciences that some semantic participant information is lexically encoded and some is not. In this article, we propose that semantic obligatoriness and verb class specificity are criteria which influence whether semantic information is lexically encoded. We present a comprehensive survey of the English verbal lexicon and two continuation studies which confirm that both factors play a role in the lexical encoding of participant information.

The Role of Specificity in the Lexical Encoding of Participants.

In addition to information about phonology, morphology and syntax, lexical entries contain semantic information about participants (e.g., Agent). However, the traditional criteria for determining how much participant information is lexically encoded have proved unreliable. We have proposed two semantic criteria (obligatoriness and selectivity) that jointly identify the participants that are lexically encoded in verbs. We tested whether one of these criteria, semantic selectivity, makes psychologically real distinctions between participant information that is lexically encoded and participant information that is not. We examined how readers integrated syntactically optional WH-constituents in filler-gap sentences when the participant information conveyed by the WH-filler was specific to a restricted class of verbs (i.e., source locations) and when it was not (i.e., event locations). Our results provide support for the role of specificity in the lexical encoding of participant information of syntactically optional constituents.

When Are Implicit Agents Encoded? Evidence from Cross-Modal Naming☆

A cross-modal naming task was used to investigate when readers access and use semantic argument information to integrate a verb into the representation of a sentence. Previous research has shown that readers include implicit agents as part of their understanding of short passive sentences like The door was shut but not in intransitive sentences like The door shut. We demonstrate that implicit agents are accessed immediately upon recognizing a passive verb. Additionally, our results suggest that cross-modal naming is sensitive to some types of lexically encoded semantic information.

The Representation and Processing of Participant Role Information

Various attempts to characterize the representation of verb argument structure and participant role information are discussed and evaluated. It is concluded that the representation of verb argument structure and participant role information must include both situation-based prototypical role-filler information, as suggested by McRae et al.s (Language and Cognitive Processes: Special Issue on Lexical Representations in Sentence Processing, 12, 137–176, 1997) verb-specific concept view, and more abstract, semantic role and role property information, as suggested by Koenig et al. (Cognition, 89, 67–103, 2003) lexical encoding hypothesis. Two questions about processing of these two aspects of participant role information are then posed: (1) Are both of aspects of participant role information automatically activated? and (2) How do these aspects of participant role information influence online sentence comprehension? Based on evidence from Su (When is semantic priming automatic? Instrument and location participant role priming as a case study, Unpublished doctoral dissertation 2013), it is concluded that situation-based prototypical role-filler information is not automatically activated by verbs. Evidence regarding the automatic activation of more abstract semantic role property information from sentence processing studies is less direct and not conclusive. Three ways in which these two aspects of participant role information might influence online language comprehension are then discussed. In particular, the role of participant role information in guiding syntactic attachment decisions, integrating constituents such as WH fillers into sentence representations, and anticipating the mention of participant role fillers is examined. It is concluded that both aspects of participant role information may influence all three processes, but when and how they are used during online comprehension may vary as a function of task demands, the comprehender’s goals, and referential contexts.