Pronominal Anaphors Research Articles

Speech Pauses and Pronominal Anaphors

This paper addresses the usefulness of speech pauses for determining whether third person neuter gender singular pronouns refer to individual or abstract entities in Danish spoken language. The annotations of dyadic map task dialogues and spontaneous first encounters are analyzed and used in machine learning experiments act to automatically identify the anaphoric functions of pronouns and the type of abstract reference. The analysis of the data shows that abstract reference is more often performed by marked (stressed or demonstrative pronouns) than by unmarked personal pronouns in Danish speech as in English, and therefore previous studies of abstract reference in the former language are corrected. The data also show that silent and filled pauses precede significantly more often third person singular neuter gender pronouns when they refer to abstract entities than when they refer to individual entities. Since abstract entities are not the most salient ones and referring to them is cognitively more hard than referring to individual entities, pauses signal this complex processes. This is in line with perception studies, which connect pauses with the expression of abstract or complex concepts. We also found that unmarked pronouns referring to an entity type usually referred to by a marked pronoun are significantly more often preceded by a speech pause than marked pronouns with the same referent type. This indicates that speech pauses can also signal that the referent of a pronoun of a certain type is not the most expected one. Finally, language models were produced from the annotated map task and first encounter dialogues in order to train machine learning experiments to predict the function of third person neuter gender singular pronouns as a first step toward the identification of the anaphoric antecedents. The language models from the map task dialogues were also used for training classifiers to determine the referent type (speech act, event, fact or proposition) of abstract anaphors. In all cases, the best results were obtained by a multilayer perceptron with an F1-score between 0.52 and 0.67 for the three-class function prediction task and of 0.73 for the referential type prediction.

Efficient Learning Approach for Pronominal Anaphora and Ellipsis Identification and Resolution in Arabic Texts

The anaphors and ellipses resolution is a challenging task for most of NLP applications. However, few works have addressed this problem in Arabic. In fact the richness of the Arabic language and the scarcity of labeled data make the task more difficult. This paper offers a generic approach that deals with the identification and the resolution of both pronominal anaphors and ellipses in Arabic texts. To identify referential anaphors and ellipses positions, we propose a self-training SVM method based on a set of pattern-based and linguistic-based criteria. For the resolution step, we propose a novel hybrid method combining a reinforcement learning method with Word Embedding models. The reinforcement learning method uses an adapted version of Q-learning algorithm to finds the optimal combination of features. It exploits a set of morphological and syntactical features. The Word Embedding based method uses word representation models to check the semantic validity of the candidates. The evaluation of the identification approach gives a precision of 99.23% for pronominal anaphors and 94.33% for ellipses. The resolution approach reaches a precision of 80% for pronominal anaphors and 97.5% for ellipses. Our proposed approaches can be easily adapted to English and French to provide multilingual approaches.

A Comparative Study of Linguistic and Computational Features Based on a Machine Learning for Arabic Anaphora Resolution

Anaphora resolution is one of the problems in natural language processing. It is the process of disambiguating the antecedent of a referring expression from the set of entities in a discourse. The correct interpretation of pronouns plays an important role in the construction of meaning. Thus, the resolution of pronominal anaphors remains a very important task for many natural language processing applications. Additionally, it plays an increasingly significant role in computational linguistics. However, a significant amount of work on anaphora resolution is focused on English; anaphora resolution for other languages, including Arabic, is still limited. In this paper, we present a new set of computational and linguistic features to resolve Arabic anaphors using a machine learning approach. In this paper, an in-depth study was conducted on a set of computational and linguistic features to exploit their effectiveness and investigate their effect on anaphora resolution. The aim was to efficiently integrate different feature sets and classification algorithms to synthesize a more accurate classification procedure. Four well-known machine learning algorithms k-nearest neighbor, maximum entropy, decision tree and meta-classifier, were employed as base-classifiers for each of the feature sets. A wide range of comparative experiments on Quran datasets was conducted, the discussion presented, and conclusions were drawn. The experimental results show that our approach gives satisfactory results.

CLASSIFICATION OF SHORT POSSESSIVE CLITIC PRONOUN NYA IN MALAY TEXT TO SUPPORT ANAPHOR CANDIDATE DETERMINATION

Anaphor candidate determination is an important process in anaphora resolution (AR) systems. There are several types of anaphor, one of which is pronominal anaphor. Pronominal anaphor is an anaphor that involves pronouns. In some of the cases, certain pronouns can be used without referring to any situation or entity in a text, and this phenomenon is known as pleonastic. In the case of the Malay language, it usually occurs for the pronoun nya. The pleonastic that exists in every text causes a severe problem to the anaphora resolution systems. The process to determine the pleonastic nya is not the same as identifying the pleonastic ‘it’ in the English language, where the syntactic pattern could not be used because the structure of nya comes at the end of a word. As an alternative, semantic classes are used to identify the pleonastic itself and the anaphoric nya. In this paper, the automatic semantic tag was used to determine the type of nya, which at the same time could determine nya as an anaphor candidate. The new algorithms and MalayAR architecture were proposed. The results of the F-measure showed the detection of clitic nya as a separate word achieved a perfect 100% result. In comparison, the clitic nya as a pleonastic achieved 88%, clitic nya referring to humans achieved 94%, and clitic nya referring to non-humans achieved 63%. The results showed that the proposed algorithms were acceptable to solve the issue of the clitic nya as pleonastic, human referral as well as non-human referral.

Pronominal Anaphor Resolution and the Underlying Cognitive Mechanism

When people tend to refer to a previous mentioned discourse entity, they can use various types of anaphoric expressions, such as pronouns, definite noun phrases, proper names or expressions that are introduced by a lexical modifier or a comparative adjective. Among the different types of anaphoric expressions, pronoun is most common and frequently used which contains only a little information about the referent, while definite expressions contain much richer semantic information. As pronouns contain very little information on their own, they are always ambiguous. The cognitive mechanism that underlies the comprehension of pronouns, especially the ones with ambiguities, catches the interests of many researchers. Various tasks have been used to investigate this issue and the results did not go for the same direction. A combination of multiple research methods is needed for a better and deeper understanding of anaphora.

Co-varying Pronominal Anaphors: Focusing on Bound Variable Pronouns

Co-varying Pronominal Anaphors: Focusing on Bound Variable Pronouns

Who do you refer to? How young students with mild intellectual disability confront anaphoric ambiguities in texts and sentences.

Along 2 experiments we tested the anaphoric pronoun resolution abilities of readers with intellectual disability in comparison with chronological and reading age-matched groups. In Experiment 1, the anaphor test of Elosúa, Carriedo, and García-Madruga (2009) confirmed that readers with intellectual disability (ID) are slower than control readers resolving clitic anaphoric pronouns, especially when the use of morphological cues (e.g. gender) is necessary. In order to test if the poor performance could be due to low levels of metacognitive skills during reading, an inconsistency detection task combined with eye tracking was designed in Experiment 2. Participants read short texts with an anaphoric pronoun in the fifth sentence, either morphologically (gender) consistent or not with the information provided in the second sentence. The scores in the anaphor comprehension questions presented after the text confirmed that readers with ID are affected by the gender inconsistency but they are unable to explicitly report it and recover from it, as the number of re-fixations after reading the critical sentence suggests. As their answers to the explicit detection questions showed, the adults control group did not show any preference for morphosyntax or semantics in spite of being aware of the inconsistency. In sum, both groups of readers with and without ID are affected by inconsistencies, but ID readers do not have appropriate metacognitive skills to explicitly identify the source of the inconsistency and fix it.

Influences of grammatical and stereotypical gender during reading: eye movements in pronominal and noun phrase anaphor resolution

Two eye-tracking studies addressed the processing of grammatical and stereotypical gender cues in anaphor resolution in German. The authors investigated pronominal (er ‘he’/sie ‘she’) and noun phrase (dieser Mann ‘this man’/diese Frau ‘this woman’) anaphors in sentences containing stereotypical role nouns as antecedents (Example: Oft hatte der Elektriker gute Einfälle, regelmässig plante er/dieser Mann neue Projekte' Often had the electrician good ideas, regularly planned he/this man new projects'). Participants were native speakers of German (N=40 and N=24 in Experiments 1 and 2, respectively). Results show that influences of grammatical gender occur in early stages of processing, whereas the influences of stereotypical gender appear only in later measures. Both effects, however, strongly depend on the type of anaphor. Furthermore, the results provide evidence for asymmetries in processing feminine and masculine grammatical gender and are discussed with reference to two-stage models of anaphor resolution.

Issues on the Categorial Status of ‘Òun’ in Yorùbá

This paper examines issues on the categorial status of òun in Yoruba grammar, particularly its status as conjunction. It draws insight from binding theory to show that in Standard Yoruba and other related dialects, such as Ìlàje, òun is constantly anaphorically bound by the first of the two nominal constituents it purportedly coordinates in line with binding rule and binding principle ‘B’ within the particular phrasal context where it is often analysed as conjunction. Relying on data from South-Eastern Yoruba dialects (Oǹdó and Ào) and a Central Yoruba dialect (Òmùò-Aráròmí) where there is clear cut evidence for two òun in the lexicon, one of which is non-referential unlike the òun in Standard Yoruba, the paper shows that the use of òun as conjunction in Ào, reported in Taiwo (2005), is not exclusive as there are other dialects in the language that do the same. The paper however provide additional evidence to demonstrate that the claim that òun is a conjunction in Yoruba based on the Ào data cannot be generalized to Standard Yoruba and other related dialects (e.g., Ìlàjė) where the non-referential òun found in Ào, Òmùò-Aráròmí, and Oǹdó dialects is absent. These make argument in favour of the so-called conjunction òun in Standard Yoruba (School) grammar being a 3sg pronominal anaphor compelling.

ANAPHORS IN CASUAL AND CONTROLLED TALK GROUPS OF IRANIAN EFL LEARNERS

By relying mainly on the accessibility approach to anaphora, this article intends to analyze the types, distributions and retrieval of anaphors in two forms of spoken discourse: casual and controlled talk. For the specific purposes of the study, twenty sophomore Iranian students were randomly selected to conduct the talks. The subjects were divided into two groups of casual and controlled talk. According to the settings and adopted topics, the overall casual talk group was further divided into two groups of dorm and academic talk. In the end, it was observed that as the talk situations vary, types, frequencies, distances, retrieval qualities and thematic structure (patterning) of anaphors undergo dramatic changes too. Further analyses of the obtained data show that the number of pronominal anaphors is by far more than NP anaphors in dorm casual talk whereas in academic casual talk the number of NP anaphors exceeds that of the former talk groups. However, the distribution of anaphors in the performance of controlled talk groups has shown to be more moderate with regard to the types of anaphors used in it. Overall, the distributional patterns of various anaphoric devices in different talk situations are considered to be a function of the speakers’ evaluation of the cognitive states of the listeners/addressees. Average distances and frequencies of the different types of zero, pronominal, and NP anaphors have also been shown to undergo dramatic changes as talk situations vary.

Does the frequency of the antecedent noun affect the resolution of pronominal anaphors?: An ERP study

Behavioral studies investigating the influence of the relative word frequency of antecedent nouns on the processing of anaphoric pronouns have yielded contradictory results. While some researchers found no effect of an antecedent's frequency of occurrence on coreference resolution [J. Simner, R. Smyth, Phonological activation in anaphoric lexical access (ALA), Brain Lang. 68 (1999) 40–45], others report shorter reading times for pronouns referring to low compared to high frequency nouns [R.G.P. van Gompel, A. Majid, Antecedent frequency effects during the processing of pronouns, Cognition 90 (2004) 255–264]. Using event-related potentials, our study aimed to further investigate the issue. Participants were presented with sentence pairs, of which the first contained either a high frequency, a middle frequency or a low frequency noun. The second sentence contained a pronoun which referred back to the noun in the first sentence. ERP waves were determined, time-locked to both the nouns and the anaphoric pronouns. We observed a graded N400 effect for antecedents of the three frequency classes with amplitudes reversely related to the word's lexical frequency. Coreferential pronouns elicited a P300, with amplitudes dependent on the noun's relative frequency of occurrence, i.e. the lower the antecedent's word frequency, the higher was the amplitude of the P300. This amplitude effect at the pronoun is interpreted in terms of the allocation of attentional resources to salient discourse entities.

Null Subjects in European and Brazilian Portuguese

The goals of this paper are twofold: a) to provide a structural account of the effects of the informal ‘Avoid Pronoun Principle’, proposed in Chomsky (1981: 65) for the Null Subject Languages (NSLs), and b) to compare, in European and Brazilian Portuguese (EP and BP), the distribution of the third person pronouns in its full and null forms, to check whether in written corpora BP incorporates signs of the ongoing loss of the null subject, largely attested in its contemporary spoken language. The strong theoretical claim is that in the Romance non-NSLs the pre-verbal subject is sitting in Spec of IP, while in the Romance NSLs it is Clitic Left-Dislocated (or is extracted by A-bar movement if it belongs to a restricted set of non-referential quantified expressions). The paper provides quantitative evidence that BP is losing the properties associated with the Null Subject Parameter. In its qualitative analysis, it shows that the contrasts between EP and BP are easily accounted for if the two derivations are assumed and if the null subjects in the two varieties are considered to be of a different nature: a pronoun in EP and a pronominal anaphor in BP.

Syntactic Principles and Limits in Language Usage with Reference to Pronominal Anaphors in Second Language Acquisition

Language has been argued as communication tool to use and/or to teach how to use it in terms of its function, form, distribution, meaning in the field...

Referent ontology and centering in discourse

This article describes three production experiments in which the role of the ontology of referents is investigated with respect to establishing and maintaining local coherence in discourse. The experiments elaborate on the basic claims put forward by Centering Theory. Subjects were asked to complete sentences or fill in the most adequate anaphoric device in two- or three-sentenc e sequences in which concrete and abstract referents were manipulated along the lines of variables which in centering theory are claimed to determine the type of center transition and hence the accessibility of referents. The distribution of different types of anaphoric expressions indicated that—all other things being equal—abstract referents require more 'referential force' than concrete referents, which is claimed to be due to their being less accessible by their very nature. Furthermore, demonstrative nominal and demonstrative pronominal anaphors appeared to function quite differently in expressing differences in transition stages of discourse referents.

The Time Course of Anaphor Resolution: Effects of Implicit Verb Causality and Gender

This paper examines the time course of the use of implicit verb causality and gender agreement information in determining the antecedent of a pronominal anaphor. A crossmodal probe paradigm measured the activation of candidate antecedents at four time points during comprehension. Sentences contained two participants, an implicit causality verb, and a subsequent "because"-clause (e.g., Gary amazed Ellen time after time because he was so talented.). In Experiment 1, a general advantage for the first-mentioned participant was nullified immediately upon encountering the pronoun if the implicit causality of the verb favored the second participant. Experiment 2 showed that this effect was due to anaphor resolution processes and not priming from the verb. Experiment 3, which deleted the causal conjunction "because" and placed the pronoun in a separate sentence, continued to find an effect of implicit causality. In addition, the first-mentioned advantage was nullified at the pronoun for sentences without a gender contrast. The overall results are consistent with a model of anaphor resolution in which resolution processes are initiated immediately upon encountering the pronoun, but cannot be completed until information becomes available from other processes.

Comprehending Conceptual Anaphors in Spanish.

This paper examines the mechanisms involved in the assignment of an antecedent to an anaphoric element. In general, pronouns must match their antecedents at least with respect to number and gender. Sensitivity to such constraints has been shown in several experiments. But Gernsbacher (1991) has also shown that people have no difficulty comprehending a plural pronoun with an antecedent that is grammatically singular but conceptually plural. In the first three experiments, we tested whether such a "conceptual effect" was preserved with zero anaphors in Spanish. (The typical omission of pronouns in subject position in Spanish.) Verbs in a second clause were marked with plural or singular endings. Plural verbs were rated more natural than singular verbs when they followed three types of singular but conceptually plural antecedents (Experiment 1). Clauses containing plural verbs were read faster when they followed one type of singular but conceptually plural antecedents, i.e. collective sets (Experiments 2 and 3). In fact, clauses containing plural verbs were read equally fast when they followed literally singular collective sets or explicitly group nouns. Using pronominal anaphors, these reading time effects were replicated and extended to sentences that contained generic types as antecedents (Experiment 4). The results are discussed in terms of the use of information during the comprehension of anaphors.

Dimensioni della linguistica Ed. by Maria-Elisabeth Conte, Anna Giacalone Ramat, and Paolo Ramat (review)

BOOK NOTICES Dimension! della lingüistica. Ed. by Maria-Elisabeth Conte, Anna Giacalone Ramat, and Paolo Ramat. (Materiali Linguistici, Université di Pavia, 1.) Milano: Franco Angelí, 1990. Pp. 239. L. 28.000. This volume is the first of an interdisciplinary series entitled Materiali Linguistici, a collection of the current linguistic research of several scholars at the University of Pavia. This first collection is rather heterogeneous, but the articles can be grouped into the three general categories of historical, typological, and textual linguistics. Part 1 is concerned with historical linguistics. Anna Giacolone Ramat, in 'Clitici latini e romanzi ' (11-30), explains the syntactic changes observed in the evolution of Romance clitics from Latin as the variable application of Wackernagel 's Law regarding the distribution of clitics after stressed words. Silvia Luraghi, in 'Osservazioni sulla Legge di Wackernagel e la posizione del verbo nelle lingue indoeuropee' (31-60), examines some exceptions to Wackernagel 's Law in the position of pronominal clitics of Ancient Greek and Hittite. Gianguido Manzelli's 'Il plurale cappatico in armeno e ungherese' (61-1 18) is a re-evaluation ofthe hypothesis of a genealogical connection between Armenian and Hungarian in the origin of Armenian plural -k\ Manzelli proposes an alternative derivation for -k' from Proto-IndoEuropean *-s". Two papers in typological linguistics follow in Part 2. The article by Giuliano Bernini, 'Per una tipología delle repliche brevi' (1 19-50), is a study of yes-no questions and their replies in many Indo-European languages and Hungarian; Bernini's goal is to explore certain crosslinguistic typological parameters. Paolo Ramat, in 'Nessuno, Ulisse e Polifemo' (151-68), gives a preliminary typology based on the functional distribution of negative quantifiers (e.g. Italian nessuno 'nobody') and negators (e.g. Italian non 'not') in European languages. The remaining three articles in the collection cover several topics in pragmatics and textual linguistics. Claudia Caffi, in 'Modulazione, mitigazione, litote' (169-200), analyzes the degree of illocutionary force in conversational modulation, mitigation, and minimization. In 'Pronomi anaforici non-coreferenziali' (20116 ), Maria-Elisabeth Conte presents a ternary classification of noncoreferential pronominal anaphors on the basis of anaphorantecedent relations. Finally, Michèle Prandi's 'Una figura testuale del silenzio: La reticenza' (217-39) describes the formal and metacommunicative functions of silence in its various forms (e.g. ellipsis and reticence) and undertakes to establish a typology according to their role in literary texts. The editors have successfully brought together a collection of articles from various subfields of linguistic research that will appeal to the scholar who values a broad approach to the study of language. It is to be hoped that future volumes in the Materiali Linguistici series will bring to light more of the splendid work done by linguists at the University of Pavia. [Gerardo A. Lorenzino, City University of New York, Graduate Center.] Current research in natural language generation. Ed. by Robert Dale, Chris Mellish, and Michael Zock. (Cognitive Science Series, 4.) London: Academic Press, 1990. Pp. ix, 356. $45.00. This volume contains selected papers from the Second European NLG Workshop, held in Edinburgh in April 1989, together with some additional texts. In the words of the editors, its goal is to provide 'a snapshot of the state-ofthe -art in a number of important research themes in NLG' (3). A first section on text planning comprises four papers. Eduard Hovy's 'Unresolved issues in paragraph planning' (17-45) uses a paragraph structure planner based on Rhetorical Structure Theory as a basis for discussing problems in planning, including the theory and representation of coherence relations. Donia Scott & Clarisse de Souza's 'Getting the message across in RST-based text generation' (47-73) draws on psycholinguistic research to propose a heuristic-based model of style predicated on ease of cognitive processing. Alison Cawsey, 222 ...

The representation of text-based and situation-based information in discourse comprehension

Three experiments were conducted to assess selected aspects of Dell, McKoon, and Ratcliff's (1983, Journal of Verbal Learning and Verbal Behavior, 22, 121–132) model of language comprehension. According to Dell et al., discourse is represented by interconnected propositions, and retrieval is determined by propositional structure. The experiments demonstrated that, first, recognition is facilitated for both the referents and the associates of pronominal anaphors; second, referents were recognized more efficiently than associates, regardless of anaphoric reference; and, third, manipulation of sentence voice (and, therefore topic salience) eliminated the difference between referents and associates, although it did not reverse it. Thus, while the basic findings on anaphoric reference are consistent with Dell et al.'s propositional account, because this model does not account for either the sustained difference between referents and associates or for the impact of sentence voice on recognition, they identify some limitations of the approach. The results highlight the importance of situation-based information in discourse comprehension and provide further general confirmation for the use of mental models in this domain. The proposal that discourse comprehension results in both text-based and situation-based representations was discussed.

Dialogue with machines

Three experiments are reported, dealing with a situation in which subjects carried on a dialogue, via a computer terminal, with what they believed to be either a computer system or another person. Analysis of the ensuing protocols concentrated on the use of anaphor and on lexical choice. Systematic stylistic variations result from placing subjects in a situation where they believe their interlocutor to be a computer system. Subjects in this condition sustain a dialogue with a restricted window of focussed content and, as a result, compared to person-to-person dialogues, utterances are short; lexical choice is kept to a negotiated minimum; and the use of pronominal anaphor is minimised, even when referring to the discourse topic. These dialogue characteristics persist over lengthy periods of interaction. Lack of evidence to support presuppositions concerning the capabilities of a defined computer source does not lead to a change in style. Similarly, attempts to manipulate features of computer output by producing more friendly surface forms did not influence subjects' behaviour. However, within the limits of the measures taken, subjects learned as much and performed as well, or better, in interactions defined as being with a computer.

Factors influencing anaphoric processing in ESL reading comprehension

ABSTRACTThe purpose of this study was to investigate the effects of some factors that influence English as Second Language (ESL) readers’ ability to resolve anaphoric expressions. Fifty full‐time ESL students randomly drawn from three different proficiency levels served as unpaid volunteers for the project. Data were elicited by means of multiple‐choice and cloze tests. The experimental design contains one between‐subjects factor (proficiency level) and three within‐subjects factors, all orthogonally manipulated (number of antecedents, case of pronominal anaphor, and propositional distance from antecedent). A four‐way mixed ANOVA for the multiple‐choice tests indicated significant effects for proficiency level, case of anaphor, and the interaction of case and number of antecedents. A similar four‐way ANOVA for the cloze tests indicated significant effects for proficiency level, number of antecedents, case of anaphor, distance from antecedent, and for the interactions of case and antecedents, antecedents and distance, and case and distance. Similarities in the findings from this study and other first and second language studies are noted. Latent trait measurement procedures were employed to produce a rank order of difficulty for the different item types examined in this study.