... The natural language processing module uses a state of the art statistical English parser, Universal Dependency parserstanza and the semantic representation of the sentences is a simplified form of a discourse representation structure [19,20]. ...
... Then, an algorithm is applied to build the discourse representation structure. This algorithm is based on the Discourse Representation Theory [19] but it only takes into account some discourse phenomena, such as variables are always existentially quantified, conditional are not considered, and events and time are not represented. In the future, the algorithm defined can be extended or replaced by a new tool such as a semantic parser. ...
... Figure 3 shows the universal dependency parse of the utterance example. The dependency tags are used to construct the utterance partial semantic representation, a simplified discourse representation structure [19], where: ...
An ontology-based task-oriented dialogue is presented as an interface for a user developing applications in the OutSystems environment, using natural language instructions. The dialogue system represents the domain of knowledge, the instructions, as an OWL2 ontology that is consulted and updated with the interpretation of the user’s utterance. The Instructions Ontology, is defined to express the concepts and instructions related to the OutSystems environment in the creation, parameterization, or updating, of software components. The user utterance interpretations are the instructions that will be executed in the OutSystems environment, according to the instructions ontology populated with the corresponding utterance interpretation and expressing the user’s intentions. The utterances are processed by applying a Universal Dependency Parser and then interpreted to obtain a partial semantic representation. Pragmatic interpretation computes a set of possible interpretations by matching the partial representation with the ontology classes, properties, instances and data properties values, such as names. The dialogue manager uses soft constraints to choose the set of best interpretations. The development of an ontology-based task-oriented dialogue system, which establishes the natural language interface of an intelligent application, to help unskilled users build applications to display and manipulate database contents. A set of preliminary experimental cases with promising results. The domain-dependent instruction ontology, which models specific discourse actions, instructions, and the domain of knowledge, allows defining a Natural Language module that is domain-independent that uses a partial discourse representation structure of the user’s utterance to determine the matching terms in the ontology and to obtain a set of possible semantic representations of the user’s utterance.
... It was pitch dark around him (s1), because the Venetian blinds were closed (s2). (Partee 1984:254-255) Partee (1984), Hinrichs (1986) and Kamp & Reyle (1993) notice that the effect of Reference time movement is tightly linked to the telicity of predicates. Whereas Partee (1984) and Hinrichs (1984) consider that telic events (which they refer to as 'events') introduce their own R, they find that atelic events (which they refer to as 'states') reuse the previously introduced R, as is shown in (41), a discourse representation of (40). ...
... The clear distinction between 'events' (telic events) and 'states' (atelic events) is also made in Kamp & Reyle (1993). Like Partee (1984) and Hinrichs (1986), 'events' are included in, and 'states' overlap with the 'temporal location' t, a notion that partly corresponds to R. Kamp & Reyle (1993) break R up into three different notions: Besides t, there is also the so-called 'reference point' Rpt which accounts for the narrative progression, and the 'temporal progressive point' Tpt necessary for the analysis of the Past Perfect. ...
... The clear distinction between 'events' (telic events) and 'states' (atelic events) is also made in Kamp & Reyle (1993). Like Partee (1984) and Hinrichs (1986), 'events' are included in, and 'states' overlap with the 'temporal location' t, a notion that partly corresponds to R. Kamp & Reyle (1993) break R up into three different notions: Besides t, there is also the so-called 'reference point' Rpt which accounts for the narrative progression, and the 'temporal progressive point' Tpt necessary for the analysis of the Past Perfect. ...
In this dissertation, I investigate the semantics of tense and aspect in Xhosa, including their interaction with information structure.
... Sentence (2a), in contrast to (1a), implies that each girl was wearing more than one hat, and thus differs sharply in its truth conditions from (2b) (cf. De Mey 1981;Zweig 2008Zweig , 2009Kamp and Reyle 1993;Champollion 2010b for similar observations and discussion). ...
... If dependent plurals have the same underlying semantics as singular indefinites, the interpretation in (5a) should be correct. Kamp and Reyle (1993) demonstrate that this is in fact not the case. The following is a slightly modified version of the example they discuss: ...
... Subsequently, accounts along these lines have been developed by Kamp and Reyle (1993) and Spector (2003). 8 Distributive approaches are able to account for the co-distributive interpretation characteristic of dependent plurals. ...
The paper focuses on the semantics of distributivity, grammatical number, and cardinality predicates (numerals and modifiers like several). I argue that constructions involving so-called ‘dependent plurals’, i.e. plurals lacking cardinality predicates occurring in the scope of certain quantificational items such as all and most (e.g. All the girls were wearing hats), pose a challenge to familiar semantic frameworks that distinguish between two sources of multiplicity: mereological plurality and distributive quantification. I argue that dependent plural readings should be analysed as distinct both from cumulative readings and distributive readings, in the classical sense. I demonstrate how this can be accomplished in a semantic framework where expressions are evaluated relative to sets of assignments, or plural info states (van den Berg, in Stokhof and Torenvliet (eds) Proceedings of the 7th Amsterdam Colloquium, ILLC, University of Amsterdam, Amsterdam, 1990, in Dekker and Stokhof (eds) Proceedings of the 9th Amsterdam Colloquium, ILLC, University of Amsterdam, Amsterdam, 1994, Some aspects of the Internal Structure of Discourse. The Dynamics of Nominal Anaphora. PhD thesis, University of Amsterdam, 1996). The specific formal implementation is based on a modified version of Brasoveanu’s (Structured nominal and modal reference. PhD thesis, Rutgers, The State University of New Jersey, 2007, Linguist Philos 31(2):129–209. https://doi.org/10.1007/s10988-008-9035-0, 2008) Plural Compositional DRT. In this framework we are able to distinguish between two types of distributivity: weak distributivity across the assignments in a single plural info state and strong distributivity across multiple info states. I argue that both of these types of distributivity play a role in the semantics of natural language, accounting for the contrasting properties of ‘singular quantifiers’, such as each and every, and ‘plural quantifiers’, such as all and most. The contrasting properties of bare plurals and plurals involving cardinality modifiers are analysed in terms of the distinction between state-level and assignment-level (mereological) plurality.
... La structure des phrases y est contrainte avec des règles de construction et d'interprétation, tout en restant très proche de l'anglais courant. Associé à un moteur d'analyse, l'APE 10 , il permet de générer des représentations sémantiques sous la forme de structures de représentation du discours (DRS 11 ) [26] et d'expressions logiques du premier ordre. Par extension, les DRS peuvent ensuite être traduites dans diverses représentations telles que OWL 12 [27], SWRL 13 [28] ou RuleML [29]. ...
... L'UCCA [47] et la couche TL du corpus PTL [58,59] présentent des struc-tures à plusieurs niveaux pour l'annotation des distinctions sémantiques, semblables aux arbres de dépendance. D'autres représentations plus anciennes, telles que les DRT [42,26], constituent également une option envisageable, au moins sur le plan théorique. ...
... La théorie de la représentation du discours de Kamp (DRT 1 , [42,26]) fournit un premier cadre d'exploration du sens selon une approche sémantique formelle. Elle inclut un niveau de représentations mentales abstraites, appelé structure de représentation du discours (DRS 2 ). ...
La spécification de systèmes techniques est une tâche complexe et source d'erreurs. D'un point de vue méthodologique, les caractéristiques attendues doivent être rigoureusement spécifiées. En pratique, des cahiers des charges regroupent les propriétés voulues sous la forme d'une liste de conditions à vérifier, nommées exigences. L'enjeu de cet ouvrage est la construction d'un processus d'analyse pour une application sur des documents textuels, rédigés dans une langue naturelle telle que l'anglais. La mise en œuvre visée est une chaîne de traitement, automatisée de bout en bout, et intégrant des facultés d'interprétation et de raisonnement sur les données traitées. Précisément, nous proposons d'étudier comment relier des énoncés de langue anglaise à des modèles formels exploitables dans un cadre théorique adapté. Premièrement, le principe de transduction sémantique est avancé pour extraire et formaliser des énoncés du langage naturel. Dans un second temps, les propriétés algébriques de modèles de spécification sont étudiées pour définir une théorie permettant de vérifier la consistance des exigences d'un cahier des charges.
... A 4-6. szakaszban szemléltetjük ezeket a pontokat, miután a 2. szakaszban olyan félformális ismertetést adunk a eALIS-ről, amit elégségesnek remélünk az imént említett integráció és továbbfejlesztés lényegének a megértetéséhez, a 3. szakaszban pedig összehasonlítjuk a eALIS-t és az illokúciós logikát (S&V). A 2. szakaszban tehát felvázoljuk a eALIS lényegét, részben a közvetlen tudományos előzményének tekinthető kampi DRT diskurzusreprezentációs elmélettel (Kamp-Reyle 1993) való összevetésében. Rámutatunk, hogy az, ahogyan a eALIS új matematikai definíciós hátteret ad a diskurzusreprezentációnak, természetes alapot teremt a funkcionális megközelítés integrálására. ...
... A reprezentációs technikák (és kapcsolatuk) mélyebb megértése végett tekintsünk most egy bonyolult logikai struktúrájú mondatot (3a) (NB: az egyik bírálónk tanácsát megfogadva azt javasoljuk a formalizmus finom részletei iránt kevésbé érdeklődő olvasók számára, hogy a (3) ponthoz kapcsolódó elemzést ugorják át)! Bár csak egy "ha…, akkor…" felépítésű következtetést figyelhetünk meg benne, a minden kvantordetermináns kétszeri megjelenése gyakorlatilag két további következtetést nyújt: "ha találunk egy diákot/professzort (egy értelemszerűen korlátozott halmazon belül), akkor az ilyen és ilyen…". Egy következtetési viszonynak a DRTben (Kamp-Reyle 1993) egy olyan, három dobozból álló részbenrendezett struktúra felel meg, ahol a legnagyobb doboz tartalmazza a két kisebbet, amelyeket egy '⇒' szimbólum kapcsol össze. A (3b)-ben látható dobozrendszer e háromdobozos konfiguráció háromszori rekurzív alkalmazásával áll elő. ...
A tanulmány gondolatmenete két szálon fut, ahogy a kettős címe is utal rá. Az egyik szál az illokúciós aktusok természetének, illetőleg a mondattípusokhoz fűződő kapcsolatának az áttekintéséből indul ki. Az előbbi területen Searle funkcionalistának/pszichologistának nevezhető szempontrendszerét helyeztük figyelmünk középpontjába (illokúciós cél/erő, perlokúciós hatás, világ–szó illesztési irány, kifejezett pszichológiai állapotok). Az utóbbi területet, ahol nyelvtipológiai és logikai–szemantikai szempontok mentén alakult ki a mondattipizálási és pragmatikai kategóriák rendszere, azzal jellemezhetjük, hogy a nyelvi forma szolgál a rendszerépítés kiindulópontjául. A másik szálon bemutatunk egy, a formális irányzatban gyökeredző, ℜeALIS nevű pragmaszemantikai megközelítést. A két szál ott fut össze, hogy azt állítjuk: a ℜeALIS – élethossziglani elmereprezentációvá kiteljesedő dinamikus megközelítéséből adódóan – kézenfekvően kínálja a pszichologista/funkcionalista megközelítés eredményeinek maradéktalan integrálását, miközben számot ad a funkcionalista beszédaktustípusok és a formalista mondattípusok kapcsolatrendszeréről.
... Though intimately related to it, this phase cannot be considered a portion of the nuclear structure of the situation at hand since it does not take part, by itself, in the internal structure characterising the event (cf. Moens, 1987;Moens & Steedman, 1988;Binnick, 1991;Kamp & Reyle, 1993). Since this phase is clearly stative in nature, it is called a pre-preparatory state. ...
... In fact, in all the interpretations considered, there seems to be a relationship of prospective projection between the state that overlaps a given interval of time that constitutes its temporal perspective point (cf. Kamp & Reyle, 1993) and the eventuality in its scope. ...
In contrast to most aspectual operators whose meaning remains unchanged regardless of their context of occurrence, the structureestar para + infinitive reveals a great deal of variability in this respect. Thus, in addition to the primary prospective aspect value, this construction also conveys modal meanings that can express uncertainty, counterfactuality, or volition/intention. Factors such as the tenses in which the sentence appears (in particular, the pretérito imperfeito (past imperfect) and the pretérito perfeito (terminative past tense)) and the presence of sentential negation contribute decisively to bringing about this kind of variation. However, there seems to be a shared semantic profile underlying the construction estar para + infinitive, namely the presence of a stative predication that induces a prospective projection.
... Motivated by meaning-preserving translation, the parallel meaning bank (PMB) [1] consists of sentences in English, German, Italian and Dutch languages mapped to their meaning using Discourse Representation Theory (DRT) [25]. The PMB can be used for machine learning semantic parsing. ...
... • Triple extraction [3,18,52]. • Parsing into a semantic representation, like DRS [25,35,51]. • Parsing into FOL either from text [6] or a controlled language [15]. ...
Commonsense reasoning (CSR) is the ability to reason about everyday situations. In artificial intelligence systems, such reasoning requires extensive background knowledge. We implement an integration pipeline that transforms existing large commonsense knowledge (CSK) resources into a logic representation designed to support contradictory and uncertain information. Our representation is an extension of first-order logic with default rules and numeric confidences. Using the pipeline, we integrate eight large CSK graphs to build a large knowledge base for symbolic hybrid AI. To the best of our knowledge, our contribution is the first large-scale commonsense knowledge base using default logic, available online. The CSK sources complement each other, having 5% or less overlap in knowledge between any pair of input graphs. The integrated knowledge supports taxonomy related queries. Significant gaps in commonsense knowledge remain, because most of the CSK sources lack rules that enable further reasoning beyond taxonomy.
... EA NLU uses Allen's (1994) parser for syntactic processing and construction of initial semantic representations. It uses Discourse Representation Theory (Kamp & Reyle, 1993) for dealing with tense, quotation, logical and numerical quantification, and counterfactuals. EA NLU is useful for this type of experiment because it focuses on generating rich semantic representations. ...
... The average number of choices per disambiguation is 2.9. Anaphora resolution, required to identify distinct mentions of concepts as the same entity, is handled automatically (Kamp & Reyle, 1993). Aside from the simplification process and disambiguation, the model runs autonomously from start to finish. ...
Analogy is heavily used in instructional texts. We introduce the concept of analogical dialogue acts (ADAs), which represent the roles utterances play in instructional analogies. We describe a catalog of such acts, based on ideas from structure-mapping theory. We focus on the operations that these acts lead to while understanding instructional texts, using the Structure-Mapping Engine (SME) and dynamic case construction in a computational model. We test this model on a small corpus of instructional analogies expressed in simplified English, which were understood via a semi-automatic natural language system using analogical dialogue acts. The model enabled a system to answer questions after understanding the analogies that it was not able to answer without them.
... Where a pronoun that cannot reefer to an entity introduced in the previous sentence is preceded by *. According to DRT [81], negation blocks the introduction of (some) drefs in its scope, which are otherwise introduced by indefinites. Names, unlike indefinites, project the dref they introduce outside the scope of negation. ...
... The most obvious theories to compare the present work to are Discourse Representation Theory (DRT) [80,81], Dynamic Predicate Logic (DPL) [64], Dekker's Predicate Logic with Anaphora [37] and Predicate Logic with Indices [38]. We briefly discuss these theories in turn. ...
Notre travail de thèse se situe au carrefour de plusieurs disciplines :d'une part, la logique mathématique et l'informatique théorique, d'autre part le traitement automatique du langage naturel et plus particulièrement la sémantique formelle du langage naturel. Le fil conducteur est la présence constante des méthodes logiques issues de la théorie de la preuve et par le problème philosophique qui a motivé notre thèse : quels sont les liens entre la notion de preuve et celle de signification linguistique ou logique ?Plus concrètement, nous étudions des systèmes formels dont les preuves sont vues comme des stratégies gagnantes pour des jeux à deux joueurs. Dans ces jeux, un jouer, appelé Proposant, essaye de construire une justification pour un certain énoncé tandis que l'autre, l'Opposant, essaye de construire une réfutation de cet énoncé.La thèse est composée de trois parties, chaque partie contenant deux ou trois chapitres.La première partie est propédeutique. Dans les deux chapitres qui la composent nous présentons les outils mathématiques utilisés dans notre thèse ainsi que les principes logiques et philosophiques qui ont guidés nos travaux, notamment la sémantique inférentialiste.La deuxième partie de notre thèse contient deux longs chapitres, lesquels présentent les résultats de théorie de la démonstration qui constituent le cœur de notre thèse.En particulier, dans le premier chapitre de cette partie, nous définissons précisément un système de logique dialogique pour la logique classique du premier ordre avec termes. Nous montrons que, pour une formule A, l'existence d'une stratégie gagnante pour A équivaut au fait que A est une théorème logique. Bien que des systèmes de logique dialogique pour la logique classique du premier ordre existent depuis les années 1960 il n'existait pas à ce jour de preuve convaincante publiée de ce résultat, notamment en présence de termes. Dans le deuxième chapitre de cette deuxième partie, nous présentons une sémantique dénotationnelle pour la variante constructive de la logique modale K. En particulier notre sémantique dénotationnelle est une sémantique des jeux dans laquelle les preuves de la logique modale sont interprétées par des stratégies gagnantes pour des jeux à deux jouer. Nous montrons que notre sémantique possède une propriété remarquable : elle est 'totalement adéquate' (fully complete) c'est-à-dire que toute stratégie gagnante est l'interprétation d'au moins une preuve de la logique modale.La troisième et dernière partie se compose de trois chapitres, chacun étant consacré à une application de nos travaux en théorie de la démonstration à la sémantique du langage naturel.Dans le premier chapitre, nous étudions le rapport entre les analyses syntaxiques catégorielles d'une même phrase et les représentations sémantiques logiques de la phrase analysée. Nous montrons que, lorsque certaines conditions sont respectées, la fonction qui transforme analyses syntaxiques d'une phrase en représentations sémantique logiques est injective.Dans le deuxième chapitre de cette troisième partie nous appliquons notre système de logique dialogique à la résolution au problème de la reconnaissance d'inférences en langage naturel en utilisant un analyseur syntaxique et sémantique catégoriel.Dans le dernier chapitre de cette partie, nous présentons une système formel pour la résolution d'anaphore et ellipses, problème généralement abordé par des méthodes de théorie des modèles. Nous, au contraire, présentons une solution basée sur la théorie de la démonstration, en développant un système de logique dialogique qui permet de résoudre simplement les anaphores et les ellipses.Dans la conclusion, nous faisons le bilan de notre travail de thèse et essayons de décrire les développements futurs possibles de notre recherche, tant du point de vue mathématique et logique que du point de vue des applications au langage naturel.
... Where a pronoun that cannot reefer to an entity introduced in the previous sentence is preceded by *. According to DRT [81], negation blocks the introduction of (some) drefs in its scope, which are otherwise introduced by indefinites. Names, unlike indefinites, project the dref they introduce outside the scope of negation. ...
... The most obvious theories to compare the present work to are Discourse Representation Theory (DRT) [80,81], Dynamic Predicate Logic (DPL) [64], Dekker's Predicate Logic with Anaphora [37] and Predicate Logic with Indices [38]. We briefly discuss these theories in turn. ...
This thesis is situated at the intersection of several disciplines: on the one hand,mathematical logic and theoretical computer science, on the other hand, natural languageprocessing and formal semantics of natural language. The thread tying thesetopics together is the constant use of tools and methodologies of proof theory and thephilosophical problem that motivated our thesis: what are the links between the notionof proof and that of linguistic meaning? More concretely, we study formal proofssystems. in these systems proofs are seen as winning strategies for two-player games.In the games one player, called the Proponent, tries to construct a justification for a certainstatement while the other, the Opponent, tries to refute this statement. Our thesisis composed of three parts, each part containing a maximum of three chapters.The first part is preparatory. In the two chapters that compose it we present the mathematicaltools used in our thesis as well as the philosophical question that underlieour research.The second part consists of two long chapters and presents the central proof-theoreticalresults of our thesis. In the first chapter of this part we present a dialogical logic systemfor classical first order logic. We show that, given a formula A, A is a logical theoremif and only if there is a proponent winning strategy for A. Dialogical logic systems forclassical first-order logic have existed since the 1960’s. However there is no convincingproof of this result in the literature. In the second chapter of this second part wepresent a denotational semantics for the constructive variant of the modal logic K. Ourdenotational semantics is a game semantics: the proofs of modal logic are interpretedby winning strategies for two-player games. We show that our game semantics has aremarkable property; it is ’fully complete’: every winning strategy is the interpretationof a proof of modal logic.The third and last part of our thesis consists of three chapters. Each chapter is devotedto an application of proof theory to the semantics of natural language. In thefirst chapter, we study the relationship between the categorical syntactic analyses ofa sentence and the logical representations of the sentence. We show that, when certainconditions are met, the function that transforms syntactic analyses of a sentenceinto logical representations is injective. In the second chapter of this third part, we useour dialogical logic system, together with type logical grammars, to solve textual entailmentproblems. In the last chapter of this section we present a formal system forthe resolution of anaphora and ellipsis. This problem is usually addressed by model theoreticmethods. We, on the contrary, present a solution based on proof theory. Wedevelop a dialogical logic system in which anaphora and ellipsis can be solved in asimple way.In the conclusion, we sketch possible future developments of our research. Bothfrom a mathematical and logical point of view and from the point of view of naturallanguage applications.
... Discourse Representation Theory (DRT) was introduced by Hans Kamp and others in the 1980s as a dynamic semantic theory designed to address the limitations of Montague grammar in handling natural language phenomena [2]. Montague grammar gives us formal ways to look at semantics, but it is based on a truth-conditional semantics that is static, which doesn't work well when dealing with the fact that natural language semantics change over time and depend on the context. ...
This research conducts a thematic analysis of The Everlasting Regret based on Segmented Discourse Representation Theory (SDRT), aiming to uncover its deeper meanings by exploring the text's discourse structure and rhetorical relations. Unlike traditional literary criticism approaches, this study adopts the SDRT framework, with a focus on analyzing the rhetorical relations between Compound Discourse Units (CDUs). The academic debate surrounding the theme of The Everlasting Regret is extensive, with major perspectives including allegorical, romantic, and dual-theme interpretations. Through detailed analysis, this research argues that The Everlasting Regret does not present a dual theme, but rather a single one. Moreover, this theme is not a celebration of love, but rather a subtle and nuanced form of satire. Additionally, the research offers an in-depth examination of key discourse units within the text, revealing that the rhetorical relations among them can be interpreted as Contrast, Result, and Narration. This not only highlights the complexity of discourse coherence and textual structure, but also suggests a potential intrinsic link among these three rhetorical relations.
... It relies on two assumptions: the meaning can be decomposed into atomic elements, and there exists a linear and incremental mechanism assembling these elements into abstract structures. This is the basis of many semantic formal frameworks, in particular those focusing on the syntax/semantics/discourse interface, such as the Discourse Representation Theory [Kamp and Reyle, 1993] or Combinatory Categorial Grammars [Steedman, 2000]. In a computational perspective as well, natural language understanding approaches relied for decades on the montagovian view of compositionality: a lot of work has been done in this direction in the logic programming paradigm [Colmerauer, 1982, Shieber andPereira, 1987], and more recently around Categorial Grammars [Bos et al., 2004, Moot, 2012. ...
The mechanisms of comprehension during language processing remains an open question. Classically, building the meaning of a linguistic utterance is said to be incremental, step-by-step, based on a compositional process. However, many different works have shown for a long time that non-compositional phenomena are also at work. It is therefore necessary to propose a framework bringing together both approaches. We present in this paper an approach based on Construction Grammars and completing this framework in order to account for these different mechanisms. We propose first a formal definition of this framework by completing the feature structure representation proposed in Sign-Based Construction Grammars. In a second step, we present a general representation of the meaning based on the interaction of constructions, frames and events. This framework opens the door to a processing mechanism for building the meaning based on the notion of activation evaluated in terms of similarity and unification. This new approach integrates features from distributional semantics into the constructionist framework, leading to what we call Distributional Construction Grammars.
... In comparison to the syntactic and semantic extensions of MT metrics, there have been very few previous attempts to incorporate discourse information. One example are the semantics-aware metrics of Giménez and Màrquez (2009) and Giménez et al. (2010), which used the Discourse Representation Theory (Kamp and Reyle 1993) and tree-based discourse representation structures (DRS) produced by a semantic parser. They calculated the similarity between the MT output and the references based on DRS subtree matching as defined in (Liu and Gildea 2005), using also DRS lexical overlap, and DRS morpho-syntactic overlap. ...
In this article, we explore the potential of using sentence-level discourse structure for machine translation evaluation. We first design discourse-aware similarity measures, which use all-subtree kernels to compare discourse parse trees in accordance with the Rhetorical Structure Theory (RST). Then, we show that a simple linear combination with these measures can help improve various existing machine translation evaluation metrics regarding correlation with human judgments both at the segment- and at the system-level. This suggests that discourse information is complementary to the information used by many of the existing evaluation metrics, and thus it could be taken into account when developing richer evaluation metrics, such as the WMT-14 winning combined metric DiscoTKparty. We also provide a detailed analysis of the relevance of various discourse elements and relations from the RST parse trees for machine translation evaluation. In particular we show that: (i) all aspects of the RST tree are relevant, (ii) nuclearity is more useful than relation type, and (iii) the similarity of the translation RST tree to the reference tree is positively correlated with translation quality.
... Opitz and Frank (2019) propose an LSTM based model that performs a multi-variate quality analysis of AMRs (constituting the baseline which we compared against). We believe that quality estimation approaches may also prove valuable for other meaning representation formalisms (MRs), such as, e.g., discourse representations (Kamp and Reyle, 1993;Kamp, 2008;Abzianidze et al., 2019) or universal semantic dependencies (Reisinger et al., 2015;Stengel-Eskin et al., 2020). For example, since the manual creation of MRs is a notoriously laborious task, automatic quality assessment tools could assist humans in the annotation process (e.g., by serving as a cheap annotation quality check or by filtering automatic parses in active learning). ...
... DRT was proposed by Hans Kamp (Kamp, 1981;Kamp & Reyle, 1993) and aimed at interpreting discourse through its semantics (i.e. through the kinetics of the linguistically meaningful units that follow one another), and thus offered a representational theory of discourse that postulated an intermediate level of representation between the syntax of a text and its interpretation in a model (Busquets et al., 2001). It is based on the representation of Discourse Representation Structures (DRS), making this theory a non-compositional one. ...
Shwachman-Diamond disease is a rare autosomal recessive disorder with a ubiquitous impact on the physiology, life, and autonomy of individuals, including functional and cognitive abnormalities, leading to academic and social difficulties.In general, how the nervous system of individuals with Shwachman-Diamond syndrome is affected has received relatively little attention compared to the large number of studies that have investigated the physiological and biomolecular characteristics of this clinical condition.The objectives of this doctoral work were multiple: first, to (i) characterize the impact of the SBDS mutation and Shwachman-Diamond syndrome on intellectual and cognitive functions; and (ii) to develop a differential diagnosis method to discriminate Shwachman-Diamond syndrome from other neurodevelopmental pathologies.The methodological approach adopted throughout the project was based on multidimensional analyses, first of all in the psychometric evaluation of cognitive functions on the intellectual, attentional-executive and social levels; but also through an ecological evaluation of pragmatic competences with the help of an original device developed specifically for the study; and also through an integrated interactive evaluation by algebraic analysis using the Topological and Kinetic model of Trognon (2TK), specifically developed in this doctoral work and applied to all the verbal materials observed during the thesis. The major originality of this work being the development of computational and algebraic tools, notably by machine-learning and deep-learning, and allowing to make use of all the collectable data in the field of human and social sciences and psychology in a combined way, in order to carry out clinical predictions or being able to be used as a dynamic tool of assistance to the clinical monitoring or the decision; and whose main strength is that all the tools developed have been tested beforehand on material from published archives, allowing us to verify that the tools developed allow (i) access to all the information highlighted in the original work and (ii) offer reading frameworks specific to the 2TK model and congruent with the models from which it originates; even before analyzing the first original data from this work.Our results suggest that Shwachman-Diamond syndrome primarily impacts the individual's integrative abilities, including difficulties in integrating local cues into a global context and selective difficulties in second-order theory of mind. More specifically, studies suggest that individuals with Shwachman-Diamond syndrome show a dissociated profile, with elementary cognitive functions (working memory, speed of information processing, first-order theory of mind) preserved; in contrast with integrated cognitive functions (perceptual reasoning, second-order theory of mind, pragmatic skills) impaired. They are thus congruent with data from the neuroscientific literature, which suggests that both computational centers of the central nervous system, the prefrontal cortex and the anterior cingulate cortex, are affected in Shwachman-Diamond disease.Thus, these results challenge the current view of Shwachman-Diamond patients, which considered intellectual deficit as the cognitive feature of the disease. Indeed, we have shown that only integrative skills allow the diagnosis of Shwachman-Diamond disease by automated methods.
... 3 Topic-related notions, such as 'topic' or 'givenness', are also implemented, from an information structural point of view, to the meaning description of atomic sentences (Halliday 1967, Chafe 1974, Prince 1981, Ariel 1988, Gundel et al. 1993, Lambrecht 1996. More recently, they have been used in the field of formal-discourse semantics (Kamp 1981, Heim 1982, Kamp and Reyle 1993 with a view to discussing the contribution of projection phenomena (e.g., presupposition, conventional implicature, anaphora) and their information structure (e.g., givenness) to asserted (semantic) content (Venhuizen et al. 2018). The current paper derives impetus from the literature on the functions of complex sentences, developing a focus on the impact of syntactic slots (pre-posed subordinate clause) on the overall meaning of a complex sentence. ...
The current paper aims to discuss the impact of pre-posing a subordinate clause (as in connective p, q) on the overall meaning of a conjunction. This issue has traditionally been explored from a functional perspective and, more specifically, in relation to 'topic', 'framework' or other related notions. The current work reveals that, among the inadequacies that a topic-based account presents, it essentially falls short of a context-sensitive and unified perspective. In response to these inadequacies, a relevance-theoretic approach is proposed on procedural grounds. More specifically, it is argued that a sentence-initial subordinate clause, or, more accurately, the entire context associated with it, serves as evidence for the delivery of procedure r, namely the inference that the main-clause-proposition that may follow will be relevant to the foregoing pre-posed context. From a communicative point of view, the procedural impact of sentence-initial subordinate clauses is justified as a useful rhetorical instrument that serves to manipulate the hearer's epistemic assessment of the main-clause-proposition as genuinely relevant to the context associated with the (pre-posed) subordinate-clause proposition.
... Telicity is a compositional interpretation of the temporal profile of events influenced by multiple elements in a sentence, including verbs and noun phrases-as in (1) (Bach, 1986;Champollion, 2017;De Swart, 1998;Jackendoff, 1991Jackendoff, , 1997Kamp & Reyle, 1993;Krifka, 1989;Moens & Steedman, 1988;Pustejovsky & Bouillon, 1995;Talmy, 1978;van Lambalgen & Hamm, 2005;Verkuyl, 1972Verkuyl, , 1993 but also adverbial phrases, verb particles, prefixes, prepositional phrases, and even contextual elements (Brinton, 1985;Filip, 1993; van Hout, 1996;Jackendoff, 1997;Moens & Steedman, 1987;Pustejovsky, 1991Pustejovsky, , 1995. Sometimes, the addition of linguistic material can change the canonical interpretation of a sentence (Brennan & Pylkkänen, 2008;Jackendoff, 1997;Moens & Steedman, 1987;Piñango, Zurif, & Jackendoff, 1999;Pustejovsky, 1991Pustejovsky, , 1995Todorova, Straub, Badecker, & Frank, 2000): while (1a) is a telic sentence, the interpretation shifts to an atelic one when a durative adverbial is introduced as in (3a). ...
What is the relationship between language and event cognition? Past work has suggested that linguistic/aspectual distinctions encoding the internal temporal profile of events map onto nonlinguistic event representations. Here, we use a novel visual detection task to directly test the hypothesis that processing telic versus atelic sentences (e.g., "Ebony folded a napkin in 10 seconds" vs. "Ebony did some folding for 10 seconds") can influence whether the very same visual event is processed as containing distinct temporal stages including a well-defined endpoint or lacking such structure, respectively. In two experiments, we show that processing (a)telicity in language shifts how people later construe the temporal structure of identical visual stimuli. We conclude that event construals are malleable representations that can align with the linguistic framing of events.
... The problem arises while two references from the two different sign systems and both refer the same sense, the nature of their identity is unknown. Because, even by considering discourse representation theory and Kamp and Reyle's (1993) ideas that they believed; people keep some information in their memories and the new information can restore as synonym; and if someone don't understand a sign according to default semantics by Jaszczolt (2005) she/he doesn't know these signals are synonym; we still need a hypothesis that finds reasonable answers for some questions. First, how it is possible to reach image or though called concept by consider the reference of the outside world for verbal units such as 'of', 'on', 'for', etc. or grammatical units such as 'on behalf' or 'because'? ...
Purpose: This paper is trying to show that the meaning of the sign has become a kind of mystery and has not yet been analyzed in a convincing way. Theoretical Framework: In this article, a model will be mentioned that can turn this mystery into a problem and then propose a solution for this problem. This article has theoretical aspect, and proceeds with a hypothesis that can maintain its efficiency for all kinds of signs. Result and Discussion: In this article, the author tried to explained three hypotheses. First, all the sentences that form in linguistic context will perceive based on sentence or sentences from background knowledge combined with sentences from situational context and the message. Second, the concept of unites in sentence shapes by proposition. Third, the assumption introduced in this article expand in all kinds of signs, whether symbol, icon or index. In another word, the concept of sign, from whatever kinds, forms by some sentences. Value/Originality: Each sign, from whatever kind, must signify something other than itself. This signification process links the form of a sign to its meaning and involved us in a problem that has always been taken for granted. What is taken for granted in the meantime is the ‘meaning’ of the sign, which has never been explicitly investigated.
... We only 26 Or alternatively, a single underspecified covert aspect is interpreted as a habitual or as a perfective, with eventive predicates. 27 A common explanation for the incompatibility between present tense and perfective viewpoint aspect is that the speech time is grammatically represented as being of a very short duration, too short to accommodate perfective aspect (Bennett & Partee, 1978;Kamp & Reyle, 1993;Smith, 1997:110, Wyngaerd, 2005. Because present tense contributes the meaning that the reference time equals the speech time, a present reference time is too short of an interval to include the event time. ...
Paraguayan Guarani does not overtly mark tense in its inflectional system. Prior accounts of languages without obligatory morphological tense have posited a phonologically covert lexical tense, or have introduced tense semantics via a rule, in the post-syntactic interpretative component. We offer a more radical approach: Paraguayan Guarani does not have tense at the level of lexical or logical semantics. We propose that evaluation time shift, a mechanism independently attested in the narrative present in languages with tense, is more widely used in Paraguayan Guarani for encoding temporal meaning. The broader consequence of our proposal is that tense is not a linguistic universal.
... In Montague's theory, syntax of language is modeled via a grammar. Syntactic rules are then associated with semantic rules that deliver the interpretation of a sentence from the interpretations of its parts (Kamp & Reyle, 1993). The Montagovian approach is a three-stage process (the upcoming description is somewhat crude but not incorrect). ...
... As noted earlier, DRS(Kamp and Reyle, 1993) represents these semantic forms in an equivalent way. ...
VoxML is a modeling language used to map natural language expressions into real-time visualizations using commonsense semantic knowledge of objects and events. Its utility has been demonstrated in embodied simulation environments and in agent-object interactions in situated multimodal human-agent collaboration and communication. It introduces the notion of object affordance (both Gibsonian and Telic) from HRI and robotics, as well as the concept of habitat (an object's context of use) for interactions between a rational agent and an object. This paper aims to specify VoxML as an annotation language in general abstract terms. It then shows how it works on annotating linguistic data that express visually perceptible human-object interactions. The annotation structures thus generated will be interpreted against the enriched minimal model created by VoxML as a modeling language while supporting the modeling purposes of VoxML linguistically.
... In classical dynamic semantics (Kamp, 1981;Heim, 1982;Groenendijk and Stokhof, 1991), indefinites are the primary means of introducing discourse referents. Subsequent developments in the 1990s (Van den Berg, 1996;Chierchia, 1995;Kamp and Reyle, 1993;Kanazawa, 1993Kanazawa, , 1994 introduced two important ideas: selective generalised quantifiers and quantificational subordination. Let us discuss them in turn. ...
This paper explores the idea that scalar implicatures are computed with respect to discourse referents. Given the general consensus that a proper account of pronominal anaphora in natural language requires discourse referents separately from the truth-conditional meaning, it is naturally expected that the anaphoric information that discourse referents carry play a role in the computation of scalar implicatures, but the literature has so far mostly exclusively focused on the truth-conditional dimension of meaning. This paper offers a formal theory of scalar implicatures with discourse referents couched in dynamic semantics, and demonstrates its usefulness through a case study on the plurality inferences of plural nouns in English.
... Hence each conceptually viable relation within minimal recursion semantics is governed by predicate-argument relations, whereas in the current context a meaning relation has to merely fulfill the requirements of standing in an antisymmetric relation with or without any regard for syntactically governed compositions. A similar difference holds with respect to Discourse Representation Theory [7] as well. ...
This paper proposes a fresh formulation of conceptually grounded meaning relations by way of construction of certain well-defined relations over the lexicon of a natural language. These relations are constrained by the logical structures of linguistic meanings across sentence and discourse contexts. One of the biggest advantages of such meaning relations is that they are not defined over, or do not ride on, the syntactic structure of a given language. Nor do they turn on compositional relations for the computation of meaning values. This helps in the formulation of meaning relations to be defined on the symbolic elements of a lexicon on the one hand, and to be extracted from the surface structure of linguistic constructions on the other. This has consequences not merely for the nature of lexical meaning but also for the construction of a kind of (shal-low) semantic networks that can be used for semantic processing in natural language understanding or machine translation systems that are driven by a kind of shallow processing of linguistic meanings. Thus, this paper aims to show the usefulness of a kind of conceptually based characterization of linguistic meaning for its relevance to computational language processing.
... Furthermore, if this conclusion is correct then it points toward the need to dissociate the rules of narrative sequencing from grammatical aspect. Specifically, the pattern we observe in Examples (6) and (10) suggests that in English these rules are tied to the sortal distinction between stative and non-stative eventualities, with states requiring overlap and non-states favoring a sequential narrative interpretation (Kamp and Reyle, 1993). Progressive verbs pattern with statives which suggests that the function of the Progressive -ing form may be to turn non-stative predicates into (derived) statives (Vlach, 1981;Hallman, 2010;Ramchand, 2018). ...
The study employed a combination of a picture selection task and Visual World eye-tracking to investigate the processing of grammatical aspect (perfective vs. imperfective) in three languages: Russian, Spanish and English. In order to probe into the cognitive representations triggered by the aspectual forms we contrasted visual representations of different temporal portions of telic events—a snapshot of the process stage (ongoing event) and a snapshot of the immediate aftermath of the event/the result state (completed event). In all three languages, the gaze patterns and offline responses revealed a strong preference for representations of ongoing events in the imperfective condition. This confirms that the imperfective forms in all the three languages draw attention to the in-progress portion of a telic event. In the perfective condition, however, we found robust differences. Russian uses verbal prefixes to mark perfective aspect, and our results suggest that perfective telic verbs in Russian strongly highlight the result state of an event. In Spanish, the perfective past tense form (Preterite) also highlights event completion, but to a lesser extent than in Russian—in line with its less restrictive semantics in not requiring an inherent boundary. In contrast to Russian and Spanish, English speakers did not show a preference for representations of completed events in the perfective (Simple Past) condition. This suggests that the English Simple Past form does not encode a preferential cognitive salience for either the activity portion of an event or its result state, and lends support to the analysis of the English Simple Past as a non-aspectual tense form.
... The Explanation Agent Natural Language Understanding System (EA NLU; Tomai & Forbus, 2009) uses a syntactic parser (Allen, 1994) and lexical information from COMLEX (Grishem et al. 1993) for syntactic processing. It also uses lexical and semantic representations from ResearchCyc 1 , extended with an implementation of Discourse Representation Theory (Kamp & Reyle, 1993) that uses Cyc microtheories to handle contexts. ...
The naturalness of qualitative reasoning suggests that qualitative representations might be an important component of the semantics of natural language. Prior work showed that frame-based representations of qualitative process theory constructs could indeed be extracted from natural language texts. That technique relied on the parser recognizing specific syntactic constructions, which had limited coverage. This paper describes a new approach, using narrative function to represent the higher-order relationships between the constituents of a sentence and between sentences in a discourse. We outline how narrative function combined with query-driven abduction enables the same kinds of information to be extracted from natural language texts. Moreover, we also show how the same technique can be used to extract type-level qualitative representations from text, and used to improve performance in playing a strategy game.
... The Segmented Discourse Representation Theory (SDRT) (Asher and Lascarides, 2003) is based on the Discourse Representation Theory (Kamp and Reyle, 1993), with discourse relations added, and discourse structure is represented with directed acyclic graphs (DAGs). Elementary discourse units may be combined recursively to form a complex discourse unit (CDU), which can be linked with another EDU or CDU (Asher et al., 2017). ...
... There has been a growing interest in semantics as a suitable research area since the beginning of the 21st century. Researchers become more interested in the complexity of language, particularly how meanings are constructed in the language (Kamp & Reyle, 2013;Bickerton & Bickerton, 2016). Semantics has been extensively applied in the study of the human mind, its processes of thoughts, conceptualization, and cognition (Reyes, Rosso & Buscaldi, 2012). ...
APA Citation: Harara, N. M. (2021). The comprehension of semantic interpretation of metaphor in the expository writing of EFL learners at Albalqa Abstract The present study endeavors to investigate the learning of the comprehension of semantic interpretation of metaphor in the expository writing by EFL learners at Albalqa Applied University, Jordan. Moreover, the study attempts to see the role of mother tongue influences in the learning and comprehension of semantic interpretation of metaphor in expository writing by Jordanian EFL learners. Thereafter, the collected data on the semantic interpretation of metaphor are analyzed by using SPSS (22.0), One-Way ANOVA, and Mann-Whitney U test. The results indicate that: (1) the learners have good learning and comprehension of semantic interpretation of metaphor in the expository writing by EFL learners at Albalqa Applied University (mean = 3.25). (2) There is no significant difference between the male and female learners in their learning and comprehension of semantic interpretation of metaphor in the expository writing by EFL learners (P. = .057). (3) There is an agreement on using learning and comprehension of semantic interpretation of metaphor in the expository writing (M = 2.96). Finally, it was found that learners' mother tongue (i.e., Arabic) has a negative effect in the course of learning and comprehension of semantic interpretation of metaphor in the expository writing by the Jordanian EFL learners at Albalqa Applied University as most of the errors were attributed to learners' mother tongue.
... The Segmented Discourse Representation Theory (SDRT) (Asher and Lascarides, 2003) is based on the Discourse Representation Theory (Kamp and Reyle, 1993), with discourse relations added, and discourse structure is represented with directed acyclic graphs (DAGs). Elementary discourse units may be combined recursively to form a complex discourse unit (CDU), which can be linked with another EDU or CDU (Asher et al., 2017). ...
Discourse information is difficult to represent and annotate. Among the major frameworks for annotating discourse information, RST, PDTB and SDRT are widely discussed and used, each having its own theoretical foundation and focus. Corpora annotated under different frameworks vary considerably. To make better use of the existing discourse corpora and achieve the possible synergy of different frameworks, it is worthwhile to investigate the systematic relations between different frameworks and devise methods of unifying the frameworks. Although the issue of framework unification has been a topic of discussion for a long time, there is currently no comprehensive approach which considers unifying both discourse structure and discourse relations and evaluates the unified framework intrinsically and extrinsically. We plan to use automatic means for the unification task and evaluate the result with structural complexity and downstream tasks. We will also explore the application of the unified framework in multi-task learning and graphical models.
... It maps from English words to concepts in NextKB (Forbus and Hinrichs 2017) and builds a semantic interpretation of the input using frame semantics extended from FrameNet (Fillmore, Wooters, and Baker 2001). Discourse Representation Theory (Kamp & Reyle, 1993) is used to handle contexts needed for modals and counterfactuals, as well as logical and numerical quantification. ...
To interact with humans, artificial intelligence (AI) systems must understand our social world. Within this world norms play an important role in motivating and guiding agents. However, very few computational theories for learning social norms have been proposed. There also exists a long history of debate on the distinction between what is normal (is) and what is normative (ought). Many have argued that being capable of learning both concepts and recognizing the difference is necessary for all social agents. This paper introduces and demonstrates a computational approach to learning norms from natural language text that accounts for both what is normal and what is normative. It provides a foundation for everyday people to train AI systems about social norms.
... Section 5 concludes. Bos and Markert (2005) utilize a CCG parser (Bos et al. 2004) to represent the text and hypothesis in discourse representation structures (DRSs, Kamp and Reyle 1993) that encapsulate information on argument structure, polarity, etc. The DRSs of the text and hypothesis are then translated into formulae in first order logic, and a theorem prover is used in order to search whether there is a logical proof from the text formula to the hypothesis formula. ...
We introduce a new formal semantic model for annotating textual entailments that describes restrictive, intersective, and appositive modification. The model contains a formally defined interpreted lexicon, which specifies the inventory of symbols and the supported semantic operators, and an informally defined annotation scheme that instructs annotators in which way to bind words and constructions from a given pair of premise and hypothesis to the interpreted lexicon. We explore the applicability of the proposed model to the Recognizing Textual Entailment (RTE) 1–4 corpora and describe a first-stage annotation scheme on which we based the manual annotation work. The constructions we annotated were found to occur in 80.65% of the entailments in RTE 1–4 and were annotated with cross-annotator agreement of 68% on average. The annotated parts of the RTE corpora are publicly available for further research.
... When more third options could be generated, the children were less likely to accept an invalid conclusion as certain and hence fell into the false dilemma fallacy. Our findings thus further support the idea that reasoning with incompatibilities is influenced by the same semantic retrieval processes related to human conditional reasoning (e.g., De Neys et al., 2002;De Neys, Schaeken, & d'Ydewalle, 2003a, 2003bKamp & Reyle, 1993). Furthermore, this study expands on the evidence provided by Brisson et al. (2018) by observing those same background knowledge effects with elementary school children. ...
The false dilemma or dichotomy is a logical fallacy that occurs when interlocuters accept the premises in an incompatibility statement as being jointly exhaustive (i.e., leaving no third option), whereas that is in fact not the case. Brisson et al. [Memory & Cognition (2018), Vol. 46, pp. 657–670] investigated this fallacy in an adult sample and discovered a content effect that influenced participants’ performance. The current study aimed to elaborate on these findings by establishing whether similar patterns could be observed with children. A number of age-appropriate incompatibility premises were constructed. For every item, four different inferential problems were presented (Affirm First, Affirm Second, Deny First, and Deny Second) with three potential answers to choose from (X, not X, or uncertainty regarding X). A sample of 192 volunteer children, with ages ranging from 8 to 13 years, was collected. Statistical analysis showed no significant effect for participants’ age but did reveal main effects for premise validity and the amount of available “third options” (possibilities outside of the presented dichotomy). These results are a clear replication of the general effects on adults found by Brisson et al. Affirm inferences were also easy for children, Deny inferences were difficult (even more so than for adults), and content had a profound effect on participants’ performance. Whenever more third options could be generated, children were less likely to fall into the false dilemma fallacy. Our findings thus further support the idea that reasoning with incompatibilities is influenced by the same semantic retrieval processes that have been previously related to human conditional reasoning.
... Rhetorical Structure Thoery (RST) [122], a theory of text organization designed for discourse analysis and text generation, has been applied to natural language generation, parsing, summarization, argument evaluation, machine translation, and essay scoring [181]. Segmented Discourse Representation Theory (SDRT) [109], an expansion of Discourse Representation Theory [90], is a dynamic semantic theory of discourse interpretation that uses rhetorical relations to model the semantics/pragmatics interface. In SDRT, discourse segments are linked with rhetorical relations reflecting different characteristics of textual coherence, such as temporal order and communicative intentions. ...
Semantics, the study of meaning, is central to research in Natural Language Processing (NLP) and many other fields connected to Artificial Intelligence. Nevertheless, how semantics is understood in NLP ranges from traditional, formal linguistic definitions based on logic and the principle of compositionality to more applied notions based on grounding meaning in real-world objects and real-time interaction. “Semantic” methods may additionally strive for meaningful representation of language that integrates broader aspects of human cognition and embodied experience, calling into question how adequate a representation of meaning based on linguistic signal alone is for current research agendas. We review the state of computational semantics in NLP and investigate how different lines of inquiry reflect distinct understandings of semantics and prioritize different layers of linguistic meaning. In conclusion, we identify several important goals of the field and describe how current research addresses them.
Situation theory is a mathematical theory of meaning introduced by Jon Barwise and John Perry. It has evoked great theoretical and practical interest and motivated the framework of a few `computational' systems. PROSIT is the pioneering work in this direction. Unfortunately, there is a lack of real-life applications on these systems and this study is a preliminary attempt to remedy this deficiency. Here, we examine how much PROSIT reflects situation-theoretic concepts and solve a group of epistemic puzzles using the constructs provided by this programming language.
The new field of machine ethics is concerned with giving machines ethical principles, or a procedure for discovering a way to resolve the ethical dilemmas they might encounter, enabling them to function in an ethically responsible manner through their own ethical decision making. Developing ethics for machines, in contrast to developing ethics for human beings who use machines, is by its nature an interdisciplinary endeavor. The essays in this volume represent the first steps by philosophers and artificial intelligence researchers toward explaining why it is necessary to add an ethical dimension to machines that function autonomously, what is required in order to add this dimension, philosophical and practical challenges to the machine ethics project, various approaches that could be considered in attempting to add an ethical dimension to machines, work that has been done to date in implementing these approaches, and visions of the future of machine ethics research.
This thesis presents an empirically oriented investigation into Mandarin-Chinese control and complementation. It provides an in-depth systematic classification of Chinese matrix predicates based on their control and complementation properties. The classification draws on a range of diagnostics, including cross-linguistically valid ones and language-specific ones. Besides incorporating corpus data, the study has adopted sentence-acceptability experiments followed by mixed-effects statistical analyses to test a subset of linguistic generalisations. The obtained empirical patterns are modelled within the parallel constraint-based architecture of Lexical-Functional Grammar, which is augmented by Glue Semantics and Partial Compositional Discourse Representation Theory to address issues at syntax-semantics-discourse interfaces. The empirical chapters (Chapter 2 and Chapter 3) focus on the control and complementation properties of the following lexical-semantic verb classes: aspectual, attitudinal, comitative, commissive, communication, desiderative, directive/permissive, factive, implicative, interrogative, and situational. Chapter 2 focuses on developing five types of linguistic diagnostics: (i) complementation diagnostics, which reveal phrasal structures, grammatical relations, and predicate-argument relations; (ii) coreferentiality diagnostics, which distinguish various control properties (e.g., obligatory, non-obligatory, exhaustive, partial, split, implicit, unconstrained); (iii) clausehood diagnostics; (iv) diagnostics for identifying the distribution of future modals (hui and yao) and aspectual markers (-le and -guo) in the complement clause, which has implications for a finiteness distinction for the complement clause; (v) diagnostics for revealing correlational relationships between control properties and two displacement phenomena -- inner topicalisation and focus fronting. Chapter 3 applies these five types of diagnostics systematically to every verb in the data set, resulting in detailed empirical classifications. The chapter also addresses issues related to finiteness, which is approached as a relative notion of semantic dependency in Chinese. Sentence-acceptability experiments, followed by cumulative link mixed-effects statistical analyses, have been adopted to test the generalisations related to inner topicalisation and focus fronting. The theoretical chapters (Chapter 4 and Chapter 5) focus on formal-language modelling. Chapter 4 addresses control relations at syntax-semantics-discourse interfaces by integrating Lexical-Functional Grammar with Glue Semantics and Partial Compositional Discourse Representation Theory. The discussion leads to a refined typology of model-theoretic control mechanisms, adding novel sub-types to the existing literature. Furthermore, the chapter addresses modelling complexities related to different classes of equi, partial-control variants, and (partial) copy control. Chapter 5 provides a formally explicit analysis for inner topicalisation and focus fronting, capturing their correlation with control and complementation properties. It also contributes to the modelling of restructuring effects and shows how restructuring constraints interact with other constraints in the formal grammar.
Space and time representation in language is important in linguistics and cognitive science research, as well as artificial intelligence applications like conversational robots and navigation systems. This book is the first for linguists and computer scientists that shows how to do model-theoretic semantics for temporal or spatial information in natural language, based on annotation structures. The book covers the entire cycle of developing a specification for annotation and the implementation of the model over the appropriate corpus for linguistic annotation. Its representation language is a type-theoretic, first-order logic in shallow semantics. Each interpretation model is delimited by a set of definitions of logical predicates used in semantic representations (e.g., past) or measuring expressions (e.g., counts or k). The counting function is then defined as a set and its cardinality, involving a universal quantification in a model. This definition then delineates a set of admissible models for interpretation.
Una Stojnić's Context and Coherence: The Logic and Grammar of Prominence offers a series of interesting criticisms of the classical dynamic paradigm in natural language semantics and offers a sophisticated alternative outlook, one that does recognize a dynamic, context change inducing dimension of meaning but at the same preserves the idea that (declarative) utterances express propositions in context. The purpose of this note is to set the record straight: existing dynamic analyses of modals and conditionals compare favorably with Stojnić's dynamic propositionalism.
This research is aimed at studying the role played by discourse structure in text comprehension. In a series of three experiments, volunteered subjects are asked for to reorder a set of sentences, randomly presented, to make a narrative text composed of five sentences –the first experiment-, an expository text composed of five sentences –the second experiment-, or an expository text composed of nine sentences –the third experiment- coherent. Each subject is presented with a specific and unique set of scrambled sentences. Results from the 1st experiment show that referential relationships among entities and/or events contribute to determine text coherence. In a text, sentences organized according to local coherence contain lexical and non lexical cues that allow subjects to link every pair of sentences to create a particular structure; sentences organized according to global coherence contain lexical and non lexical cues that allow subjects to relate every sentence with a common topic. Under these conditions a different strategy is applied by subjects in the second and the third experiments. Results from the 2nd and 3rd experiments reveal that human subjects are willing to exploit information from the way text sentence order is being constrained but in a different way for five-sentence texts and for nine-sentence texts, in five-sentence texts through local coherence and concrete semantic relations, and in nine-sentence texts through global coherence and abstrat semantic relations. This difference arises from the fact that contrary to five-sentence texts, nine-sentence texts represent a heavier cognitive processing load for working memory. Referential relations play an identical role in all the experiments. The results are discussed according to Latent Semantic Analysis, Rethorical Structure Theory and Centering Theory models. No theoretical model can account for all the experimental findings, but, discarding the role played by reference, they are more compatible with a Centering Theory model, since this model is the only one which make specific predictions about the role played by sentence order.
In this paper, we argue that the very convincing performance of recent deep-neural-model-based NLP applications has demonstrated that the distributionalist approach to language description has proven to be more successful than the earlier subtle rule-based models created by the generative school. The now ubiquitous neural models can naturally handle ambiguity and achieve human-like linguistic performance with most of their training consisting only of noisy raw linguistic data without any multimodal grounding or external supervision refuting Chomsky's argument that some generic neural architecture cannot arrive at the linguistic performance exhibited by humans given the limited input available to children. In addition, we demonstrate in experiments with Hungarian as the target language that the shared internal representations in multilingually trained versions of these models make them able to transfer specific linguistic skills, including structured annotation skills, from one language to another remarkably efficiently.
Sign Language is fundamental communication language of deaf people. Endeavors to develop sign language generation systems can make the life of these people smooth and effortless. Despite the significance of sign language generation systems, there is a paucity of a Systematic Literature Review. This is the foremost recognizable scholastic literature review of sign language generation systems. It presents a scholastic database of literature between the duration of 1998–2020 and suggests a classification criterion to systematize the research studies. 414 research studies were recognized and reviewed for their direct pertinence to sign language generation systems. 162 research studies were subsequently chosen, examined and classified. Each of the 162 chosen research papers was categorized based on thirty sign languages and was further comparatively analyzed based on seven comparison parameters (input form, translation technologies, application domain, use of parsers/grammars, manual/non-manual features, accuracy and output form). It is evident from our research findings that the major research on sign language generation has been carried out using data-driven approaches in absence of proper grammar rules and generating only manual signs. This research study may provide amateur researchers a roadmap towards future research directions and facilitate the compilation of information in the field of sign language generation.
This article addresses the question as to whether time is a primitive concept or rather is composed of conceptually more basic building blocks. After a brief analysis of tense-time mismatches with examples from English, Polish, Thai and Swahili, I present a hypothesis that time is conceptualized in terms of degrees of epistemic modality. Expressions with future, present and past reference are ordered on scales of epistemic commitment. I demonstrate that the theory of Default Semantics has no difficulty with representing tense-time mismatches in that it reflects the fact that information about temporality is conveyed via a variety of processes, some of them pertaining not to the processing of the lexicon or grammar but even to pragmatic inference. The theory also gives support to the thesis of time as modal detachment.
The study explores the scope of adverbial relations expressed by the Old English (OE) forms an(an), anlice and butan functioning as restrictive focusing adverbs that find their equivalents in Present-Day English only. While most investigations address the statistical representation of the aforementioned adverbs in Old English (see among others Rissanen, 1985; Nevalainen, 1991; Brinton, 2017), the current research examines the correlation of the OE prototypes of focusing adverbs with information structure and regularities of sentence word order patterns. The investigation is based on Discourse Representation Theory (Kamp&Reile, 2013; Geurts et al., 2020), which allows keeping the file of discourse referents (Андрушенко 2021), i.e., given-new information, as well as various types of Topics and Foci (Krifka&Musan, 2012). The sentence components are tagged in on T-CODEX annotation scheme (Petrova, 2009). According the data from OED (2021) the adverbial form of butan is used in the restrictive function after a negative clause, e.g., Naefde ic nǽnne hiht on óðerne nǽfre búton on ðé (AElfc. Gr. Z. 270, 12). Regarding the Modern English only and its direct Old English equivalents an(an) and anlice there is no unanimity among linguists as far as the origin is concerned. Despite the existence of OE adjective anlic/aenlic "one, single, solitary" and the adverb anlice/aenlice "singularly, uniquely", these forms according to Rissanen (1985) are not regarded a primary source of ME free adverb only formation in further centuries. The examples from Borsworth Toller's Anglo-Saxon Dictionary (2021) indicate that the adjectival forms were used in the sense of "beautiful or elegant". E.g., He haefde án swíðe ǽnlíc wíf Bt. 35, 6; Fox 166, 30 (he had a very excellent wife). Another channel for Modern English only emergence is related to the numeral an, used adjectivally in Old English (Brinton, 2017), viz., Ic bydde ðé, ðaet ðú lǽ te me sprecan áne feáwa worda (Nicod. 11; Thw. 5, 40) I pray thee, that thou let me speak only [once for all] few words. Based on the abovementioned, the numerical and adjectival usages of an(an), as well as adverbial form of butan have been taken into account from the perspective of sentence structure. Therefore, the following patterns have been singled out: nales þat an þat SOV, where an(an) functions as a numeral that creates a separate sentence projection with OV-order following it; SOV butan → O anan (with the last element functioning as a post-modifying adjective); SOV butan → O anre (with the form being of dual adjectival and numerical character); SVO buton ← ane; SX(V) buton → X, SO ← anum (anre)V(v); O ← anan (SVX) (with anan being used as an adjective), etc. The results of the investigation show that in its numerical function an(an) creates a separate projection preceding the word it modifies in the following sentence, while in its adjectival usage the post-positive adjacent placement is more typical.
In this paper, we extend Abstract Meaning Representation (AMR) in order to represent situated multimodal dialogue, with a focus on the modality of gesture. AMR is a general-purpose meaning representation that has become popular for its transparent structure, its ease of annotation and available corpora, and its overall expressiveness. While AMR was designed to represent meaning in language as text or speech, gesture accompanying speech conveys a number of novel communicative dimensions, including situational reference, spatial locations, manner, attitude, orientation, backchanneling, and others. In this paper, we explore how to combine multimodal elements into a single representation for alignment and grounded meaning, using gesture as a case study. As a platform for multimodal situated dialogue annotation, we believe that Gesture AMR has several attractive properties. It is adequately expressive at both utterance and dialogue levels, while easily accommodating the structures inherent in gestural expressions. Further, the native reentrancy facilitates both the linking between modalities and the eventual situational grounding to contextual bindings.
Studies focused on conversations involving individuals with pathologies such as schizophrenia have revealed the importance of underspecification. Consequently, semantic interpretation and context play a more crucial role. In this article, I present a framework for analyzing schizophrenic discourse that principally relies on a novel pragmatic context, one that takes its inspiration from analyses commonly applied to fictional discourse. This perspective sheds new light on schizophrenia and, more generally, on conversation context management.
We propose a novel solution to anaphora and ellipsis resolution using multi-sorted first order logic. Our theory is proof-theoretic, employing methods from the study of dialogical logic. The first order propositions are extracted from reduced lambda terms, which are themselves derived from Lambek Categorial Grammar proofs.
In this paper, we present a formal approach to the compositional processing of questions and answers presented in the corpus SLAM-Schizophrenia and Language: Analysis and Modeling (Amblard et al. in Traitement Automatique des Langues, 55(3), 91115, 2015). In particular, we want to address issues surrounding dialogue lexicality by beginning with definitions as formalized in the framework of Düsseldorf Frame Semantics presented in Kallmeyer and Osswald (Journal of Language Modelling, 1(2), 267330, 2014). We then introduce a view of dialogue that emerges from compositions of negotiation phases that may be studied as separate elements while remaining linked by a common dialogue context (shared among all dialogue participants). Finally, we produce an analysis of English and French interrogative words toward an operationalization of our model for real-life data.
In this paper we present Uniform Meaning Representation (UMR), a meaning representation designed to annotate the semantic content of a text. UMR is primarily based on Abstract Meaning Representation (AMR), an annotation framework initially designed for English, but also draws from other meaning representations. UMR extends AMR to other languages, particularly morphologically complex, low-resource languages. UMR also adds features to AMR that are critical to semantic interpretation and enhances AMR by proposing a companion document-level representation that captures linguistic phenomena such as coreference as well as temporal and modal dependencies that potentially go beyond sentence boundaries.
ResearchGate has not been able to resolve any references for this publication.