On the learnability of implicit arguments

Author(s):  
Victoria Mateu ◽  
Nina Hyams
Keyword(s):  
2015 ◽  
Vol 41 (4) ◽  
pp. 625-664 ◽  
Author(s):  
Michael Roth ◽  
Anette Frank

In this article, we investigate aspects of sentential meaning that are not expressed in local predicate–argument structures. In particular, we examine instances of semantic arguments that are only inferable from discourse context. The goal of this work is to automatically acquire and process such instances, which we also refer to as implicit arguments, to improve computational models of language. As contributions towards this goal, we establish an effective framework for the difficult task of inducing implicit arguments and their antecedents in discourse and empirically demonstrate the importance of modeling this phenomenon in discourse-level tasks. Our framework builds upon a novel projection approach that allows for the accurate detection of implicit arguments by aligning and comparing predicate–argument structures across pairs of comparable texts. As part of this framework, we develop a graph-based model for predicate alignment that significantly outperforms previous approaches. Based on such alignments, we show that implicit argument instances can be automatically induced and applied to improve a current model of linking implicit arguments in discourse. We further validate that decisions on argument realization, although being a subtle phenomenon most of the time, can considerably affect the perceived coherence of a text. Our experiments reveal that previous models of coherence are not able to predict this impact. Consequently, we develop a novel coherence model, which learns to accurately predict argument realization based on automatically aligned pairs of implicit and explicit arguments.


2014 ◽  
Vol 3 (2) ◽  
pp. 125 ◽  
Author(s):  
Andres Saab

<p>I show that core implicit subjects in Spanish (i.e., the ones that occur with analytical passives, impersonal <em>se</em>, and causatives) can be derived from a theory under which absence of <em>Merge</em> in external subject position is a possible syntactic output. Core implicit arguments then have no syntactic representation (<em>pace</em> Landau 2010). Absence of <em>Merge</em> can make to arise two different scenarios: (i) a conflict at the interfaces, which requires the implementation of some repair strategy, (ii) no conflict at the interfaces; i.e., a legitimate object at the interfaces. The first scenario is illustrated with reference to the so-called impersonal <em>se</em> in Spanish, and the second one with reference to analytical passives. The proposed system is able to capture a set of very intricate facts that does not have a satisfactory solution hitherto. Crucially, this particular view on implicit arguments, together with a purely syntactic theory of argument structure, derives the full distribution of impersonals and reflexives in <em>hacer</em> ‘to make’ causative contexts. Finally, it is shown that the arbitrary readings that the two scenarios above described display have a different source: whereas impersonal <em>se</em> requires (costly) default computation at the interface, arbitrary interpretations in analytical passives are calculated at the <em>v</em>P level.   </p>


2018 ◽  
Vol 15 (2) ◽  
pp. 291-313 ◽  
Author(s):  
Fabrizio Macagno

Abstract This paper advances an approach to presupposition rooted in the concept of commitment, a dialectical notion weaker than truth and belief. It investigates ancient medieval dialectical theories and develops the insights thereof for analyzing how presuppositions are evaluated and why a proposition is presupposed. In particular, at a pragmatic level, presuppositions are reconstructed as the conclusions of implicit arguments from presumptive reasoning, grounded on presumptions of different type and nature. A false (or rather unaccepted) presupposition can be thus represented as the outcome of a conflict of presumptions – the ones used by the speaker and the ones commonly accepted or backed by evidence. From an interpretative perspective, this defaulted presumptive reasoning can be explained by comparing the available presumptions and repaired by replacing the weaker and unacceptable ones.


Author(s):  
Olivia Custer

Derrida reacted to Foucault’s History of Madness by insisting that the very project of that work, namely to “make madness speak,” was as mad as it was necessary. Sharing the commitments which gave sense and force to the gesture of the book, Derrida was drawn to identify the difficulties Foucault’s project could not, on his analysis, avoid. An untenable reading of Descartes, Derrida argued, was the symptom of the conflict between project and accomplishment; unsustainable by other standards, it held up the book’s architecture and claims. Thirty years later, Derrida professed not to want to return to the quarrel, declaring that those interested should refer to the archive already constituted. Exploring what Custer takes to be a delayed reactivation of the exchange, she considers why Derrida’s posthumous publications – notably his seminar on the Death Penalty and The Animal that therefore I am – should be added to this archive. Some implicit arguments in these works sound eerily the same as those with which Derrida had taken issue in his debate with Foucault. Custer suggests that this does not mark the possibility of a reconciliation but does offer a productive place from which to revisit the problem for which “Descartes” came to stand as a symbol. Today, when the project of giving voice to those who have been silenced has become a cliché of political activism and reading practices, it is useful to reactivate Derrida’s warnings about the mad nature of Foucault’s project.


2013 ◽  
Vol 23 (5) ◽  
pp. 552-593 ◽  
Author(s):  
EDWIN BRADY

AbstractMany components of a dependently typed programming language are by now well understood, for example, the underlying type theory, type checking, unification and evaluation. How to combine these components into a realistic and usable high-level language is, however, folklore, discovered anew by successive language implementors. In this paper, I describe the implementation ofIdris, a new dependently typed functional programming language.Idrisis intended to be ageneral-purposeprogramming language and as such provides high-level concepts such as implicit syntax, type classes anddonotation. I describe the high-level language and the underlying type theory, and present a tactic-based method forelaboratingconcrete high-level syntax with implicit arguments and type classes into a fully explicit type theory. Furthermore, I show how this method facilitates the implementation of new high-level language constructs.


Sign in / Sign up

Export Citation Format

Share Document