Antecedent (grammar)

About: Antecedent (grammar) is a research topic. Over the lifetime, 1392 publications have been published within this topic receiving 41824 citations.

How natural are conceptual anaphors

Abstract: This paper reports three experiments on the interpretation of “conceptual” anaphors. These are anaphors that do not have an explicit linguistic antecedent, but one constructed from context. For instance, if one says “I need a knife. Where do you keep them?”, them means something like “the knives that I presume you have in your house”. In the first experiment, subjects rated sentences containing conceptual anaphors, of three different types, to be as natural as ones with a “linguistically correct” antecedent (e.g. “I need an iron. Where do you keep it?”), and as more natural than ones with neither a plausible conceptual antecedent nor a plausible linguistic one. In a second (self-paced) experiment, subjects judged whether the second sentence in such pairs was a sensible continuation from the first, and the time to make these judgements was measured. We found that acceptability judgements were high, and judgement times low, in just those sentences that were rated as more natural in the first experiment. These first two experiments showed that conceptual anaphors are quite easily understood. However, they did not show that such anaphors are processed without difficulty. In the third experiment, we therefore compared conceptual anaphors (“plate … them”) with matched plural anaphors whose antecedents were explicit (“some plates … them”). The results were different for different types of anaphor: in one case (pronouns that referred to collective sets), the conceptual version followed by a plural pronoun was easier than the explicit plural version. For the other two types (references to generics and to implied multiple items), the explicit plurals were understood more rapidly than their conceptual counterparts.

Antecedents and Consequences of Envy

Abstract: The aim of this study was to examine the relationship between individual attributes and envy, and to determine how envy may impact personal response variables in the workplace. To address these issues we apply Vecchio’s theory on antecedents and consequences of envy (1995) as a theoretical framework. The present study relied on a cross-sectional measurement design. A total of 135 leaders and 772 followers employed in business organizations participated. SEM analysis shows that span of supervision serves as an important antecedent of envy, where span of supervision is significantly associated to envy via supportive leadership. Furthermore, envy seems to be indirectly and negatively related to self-esteem via distress and directly related to social loafing. The implications of these findings are discussed, and suggestions for future research are outlined.

The Role of the Antecedent in Anaphoric Relations

Abstract: In this article, I attempt to provide an account of the fact that the binding properties of subjects of NPs systematically distinguish two groups of languages In languages like Chinese, Japanese, Malayalam, subjects of NPs seem to have a sufficient “proximity” to an NP external antecedent to allow a reflexive, and yet a sufficient “distance” from it to also allow a pronoun, as in (1)

Eye-catching Anaphora

Abstract: Anaphoric elements (e.g. reflexives such as himself, and pronouns such as she and his) are among the most frequently encountered words in virtually all human languages. The most important feature of these linguistic elements is that they cannot be fully interpreted in isolation, but depend on other elements in the utterance or text instead. Hence, an important question for linguists and psycholinguists is: How does our language system construct anaphoric dependencies? A straightforward possibility is that when readers or listeners encounter an anaphoric element, this will trigger a resolution mechanism which searches the mental representation of a text and picks the most plausible referent for the anaphoric element. However, linguistic research has shown that such a single-mechanism account would be too simple: we should distinguish between syntactic dependencies (i.e. A-Chain formation between a reflexive and its antecedent), semantic dependencies (i.e. bound-variable pronouns) and discourse dependencies (i.e. coreferential pronouns). In this dissertation Arnout Koornneef discusses the findings of a series of eye-tracking and self-paced reading experiments, aimed at evaluating the linguistic ‘Primitives of Binding’ model (Reuland, 2001). In this account it is proposed that a general economy principle governs the division of labor between the different subcomponents of the anaphora resolution system: the syntactic process is thought to be more economic than the semantic process and, similarly, the semantic process is thought to be more economic than the discourse process. The results suggest that this economy hierarchy has ‘behavioral reality’ and, furthermore, that anaphora resolution occurs in a fixed sequential order, i.e. syntactic anaphoric dependencies emerge before semantic anaphoric dependencies and, likewise, semantic anaphoric dependencies emerge before discourse anaphoric dependencies. At a more general level of conception the reported behavioral results demonstrate that incorporating linguistic research into (neuro)cognitive approaches to language comprehension is often helpful – or even mandatory – since linguistic theories clearly define the object of interest (e.g. the nature of anaphoric dependencies) and, furthermore, they generate research questions that would not have emerged otherwise. The author combines insights from theoretical linguistics, psycholinguistics and neurolinguistics and this dissertation should therefore be of interest to scholars in any of these domains.

The KnowRef Coreference Corpus: Removing Gender and Number Cues for Difficult Pronominal Anaphora Resolution

Abstract: We introduce a new benchmark for coreference resolution and NLI, KnowRef, that targets common-sense understanding and world knowledge. Previous coreference resolution tasks can largely be solved by exploiting the number and gender of the antecedents, or have been handcrafted and do not reflect the diversity of naturally occurring text. We present a corpus of over 8,000 annotated text passages with ambiguous pronominal anaphora. These instances are both challenging and realistic. We show that various coreference systems, whether rule-based, feature-rich, or neural, perform significantly worse on the task than humans, who display high inter-annotator agreement. To explain this performance gap, we show empirically that state-of-the art models often fail to capture context, instead relying on the gender or number of candidate antecedents to make a decision. We then use problem-specific insights to propose a data-augmentation trick called antecedent switching to alleviate this tendency in models. Finally, we show that antecedent switching yields promising results on other tasks as well: we use it to achieve state-of-the-art results on the GAP coreference task.