Search

There is much evidence that individuals with aphasia (IWA) are impaired in tense or aspectual inflections and the production of non-canonical sentences like passives. On the Cartographic Approach (CA), passives are represented at the lower node in the syntactic tree, with tense or aspectual elements at the higher node. This hierarchical structure can be used to account for the selective impairments in aphasic production. The present study reports the empirical findings from Chinese agrammatism using the theory of cartography. It is found that the performance of IWA with different degrees of severity is associated with different nodes (CP, TP, and vP) in the syntactic tree. Notably, the Chinese-speaking IWA are severely impaired in vP-related elements. Finally, the present study highlights the significance of adopting a cartographic approach to aphasia studies.

Low proficiency ESL learners often encounter a lot of learning problems, particularly due to their lack of vocabulary, which is evident in their incorrect usage of words. This study was an attempt to aid 85 low proficiency ESL learners to easily comprehend the forms and functions of 80 target words, selected from the 2000 level General Word List, using the syntactic tree structure approach. Prior to the intervention, a vocabulary test based on 80 words, which were later introduced to the students in their course book, was administered to the students. The same test was given again to the students upon the completion of the third lesson. To determine the possible significant influence of the intervention, the pre-test and post-test data were analysed using the paired-samples t-test. The analysis yielded a significant result (t (84) = 7.31, p < .05 (two-tailed)). When the students’ mean test scores of the two tests were compared, a statistically significant increase of 7.5 points was observed, implying the success of the intervention. The calculated effect size was also large (ŋ=.39). However, since no control group was introduced in the study, caution should be taken in making any strong claims on a causal relationship.

We propose a method of simulation-based natural language generation that accounts for both building a correct syntactic structure and reflecting the given situational information as input for the generated sentence. We employ the Monte Carlo tree search for this nontrivial search problem in simulation, using context-free grammar rules as search operators. We evaluated numerous generation results from two aspects: the appropriateness of sentence contents for the given input information and the sequence of words in a generated sentence. Furthermore, in order to realize an efficient search in simulation, we introduced procedures to unfold syntactic structures from words strongly related to the given situational information, and increased the probability of selecting those related words. Through a numbers of experiments, we confirmed that our method can effectively generate a sentence with various words and phrasings.

It is argued that the notion of classical entailment faces two problems, the second argument projection problem and the P -to- Q problem, which arise because classical entailment is not designed to handle partial functions. It is shown that while the second argument projection problem can be solved either by flattening the syntactic tree or with naive multi-valued logics, the P -to- Q problem cannot. Both problems are solved by introducing a new notion of entailment that is defined in terms of Strawson entailment (in the sense of von Fintel 1999, 2001). BibTeX info

This paper describes our system to predict enhanced dependencies for Universal Dependencies (UD) treebanks, which ranked 2nd in the Shared Task on Enhanced Dependency Parsing with an average ELAS of 82.60%. Our system uses a hybrid two-step approach. First, we use a graph-based parser to extract a basic syntactic dependency tree. Then, we use a set of linguistic rules which generate the enhanced dependencies for the syntactic tree. The application of these rules is optimized using a classifier which predicts their suitability in the given context. A key advantage of this approach is its language independence, as rules rely solely on dependency trees and UPOS tags which are shared across all languages.

This thesis aims at a modular method for parsing unrestricted English text into a Systemic Functional constituency structure using fragments of Systemic Functional Grammar (SFG) and dependency parse trees.

The question arises, how to represent non-continuous syntactic n-grams without resorting to their graphic form? Recall that continuous syntactic n-grams are simply sequences of words (obtained by following paths in a syntactic tree), but the case of the non-continuous syntactic n-gram is rather different.

In this paper, we present a Korean syntactic analysis system which can generate all possible syntactic trees in a given sentence. Therefore, the number of syntactic trees by this syntactic analysis system can be increased exponentially. To solve this problem, we suggest a segmentation method and maximum connected unit in a segmentation. Maximum connected unit is a combined unit which contains all morphemes in a segmentation. According to the input sentence, it is possible one or more maximum connected unit in a segmentation. We extract 516 sentences to experiment randomly from the text book of Korean middle school. We could reduce about 28% of the number of syntactic trees.

This paper is concerned with automatic generation of all possible questions from a topic of interest. Specifically, we consider that each topic is associated with a body of texts containing useful information about the topic. Then, questions are generated by exploiting the named entity information and the predicate argument structures of the sentences present in the body of texts. The importance of the generated questions is measured using Latent Dirichlet Allocation by identifying the subtopics (which are closely related to the original topic) in the given body of texts and applying the Extended String Subsequence Kernel to calculate their similarity with the questions. We also propose the use of syntactic tree kernels for the automatic judgment of the syntactic correctness of the questions. The questions are ranked by considering both their importance (in the context of the given body of texts) and syntactic correctness. To the best of our knowledge, no previous study has accomplished this task in our setting. A series of experiments demonstrate that the proposed topic-to-question generation approach can significantly outperform the state-of-the-art results.

With the popularity of online educational platforms, English learners can learn and practice no matter where they are and what they do. English grammar is one of the important components in learning English. To learn English grammar effectively, it requires students to practice questions containing focused grammar knowledge. In this paper, we study a novel problem of retrieving English grammar questions with similar grammatical focus. Since the grammatical focus similarity is different from textual similarity or sentence syntactic similarity, existing approaches cannot be applied directly to our problem. To address this problem, we propose a syntactic based approach for English grammar question retrieval which can retrieve related grammar questions with similar grammatical focus effectively. In the proposed syntactic based approach, we first propose a new syntactic tree, namely parse-key tree, to capture English grammar questions’ grammatical focus. Next, we propose two kernel functions, namely relaxed tree kernel and part-of-speech order kernel, to compute the similarity between two parse-key trees of the query and grammar questions in the collection. Then, the retrieved grammar questions are ranked according to the similarity between the parse-key trees. In addition, if a query is submitted together with answer choices, conceptual similarity and textual similarity are also incorporated to further improve the retrieval accuracy. The performance results have shown that our proposed approach outperforms the state-of-the-art methods based on statistical analysis and syntactic analysis. Accepted version

Background: Castilian-Spanish, Catalan, Galician, and European Portuguese are the most widely spoken languages of the Ibero-Romance group. An increasing number of authors have addressed the impact of aphasia on the morphosyntax of these varieties. However, accurate linguistic characterisations are scarce and the different sources of data have not been yet compiled.Aims: To stimulate state-of-the-art research, we provided a comprehensive summary of morphosyntactic aspects of Ibero-Romance and a review of how these are affected in non-fluent aphasia. The topics we dealt with are the use of verb argument structure and morphology, sentential negation and word order, definite articles, personal and reflexive pronouns, passives, topicalised constructions, questions, and relative clauses.Methods & Procedures: An exhaustive fieldwork and search of PubMed, Web of Science, and Medline records were performed to retrieve studies focused on morphosyntactic issues concerning the Ibero-Romance varieties. A total of 27 studies produced by 46 authors of varying background emerged. We did not review studies of category-specific deficits and aspects related to bilingual aphasia, although we assume that most speakers of Galician and Catalan are bilingual. Studies of spontaneous speech were included when no controlled experimental tasks were available.Outcomes & Results: The morphosyntactic commonalities of Ibero-Romance have been tackled from different theoretical perspectives. There exist asymmetries in findings which we explain with the use of different tasks (and task complexity) and individual differences between participants.Conclusions: Discourse-linking factors as well as deviations from the canonical pattern are recurrent answers to these asymmetries. A comprehensive theory of impairments in non-fluent aphasia integrating relevant aspects of both structural and processing accounts seems necessary.

The vulnerability of time reference, either marked by means of verbal morphology or by adverbs, has been established in an extensive array of languages in agrammatic aphasia. Recent studies ( Bastiaanse, 2008 , Lee et al., 2008 , Nanousi et al., 2006 , Martinez-Ferreiro, 2010 , Yarbay Duman and Bastiaanse, 2009 ; among others) have determined that far from being general, the observed deficit is sensitive to tense differences with a clear asymmetry between past and non-past forms. To account for these findings, Bastiaanse et al. (2011) formulated the PAst DIscourse LInking Hypothesis (PADILIH), in line with Avrutin's (2000) claim that discourse linking is impaired in Broca's aphasia. Past forms are impaired in opposition to non-past forms due to the fact that the former are discourse linked. However, this hypothesis entails further predictions: if the problem with tense is restricted to discourse linking, we expect present and future forms to be spared or, at least, damaged to the same extent. In this paper, we investigate time reference in Catalan and Spanish using a sentence completion task with pictures and a sentence-to-picture-matching task (adapted from the TART; Bastiaanse et al., 2008 ). The results confirm the predicted deficit: forms referring to the past are more impaired than forms referring to non-past, that is, tenses referring to the present and future were better preserved. However, in comprehension, asymmetries arise between present and future forms. Implications for the PADILIH are discussed.

We introduce a tree-structured attention neural network for sentences and small phrases and apply it to the problem of sentiment classification. Our model expands the current recursive models by incorporating structural information around a node of a syntactic tree using both bottom-up and top-down information propagation. Also, the model utilizes structural attention to identify the most salient representations during the construction of the syntactic tree. To our knowledge, the proposed models achieve state of the art performance on the Stanford Sentiment Treebank dataset., Submitted to EACL2017 for review

Illocutionary force and modality are grammatical phenomena that are traditionally considered to operate at a 'high level', scoping over entire clauses and represented in the upper regions of the syntactic tree or of the semantic representation. This is supported by the observation that (most) subordinate clauses do not have their own illocutionary force, but depend on the main clause in this respect. Still, English exhibits a number of constructions where an NP appears to come with its own illocution, independent of the illocutionary force of the clause of which the NP is a part. In other languages as well, NPs are known to occur with illocutionary and modality operators. In this article, several of these constructions will be brought together, and it is then considered how they can be accounted for in two closely related structural-functional syntactic theories, namely Role and Reference Grammar and Functional Discourse Grammar. It is shown that both approaches can accommodate these odd constructions, provided a number of minor adjustments are made to the technical apparatus. © 2011 Elsevier B.V. ispartof: Lingua: International Review of General Linguistics vol:122 issue:1 pages:1-23 status: published

Most previous approaches used various kinds of plain similarity features to represent the similarity of a sentence pair, and one of its limitations is its weak representation ability. This paper introduces the relational structures representation (shallow syntactic tree, dependency tree) to compute sentence similarity. Experimental results manifest that our approach achieves higher performance than that only uses plain features.

This research proposes and evaluates a linguistically motivated approach to extracting temporal structure from text. Pairs of events in a verb-clause construction were considered, where the first event is a verb and the second event is the head of a clausal argument to that verb. All pairs of events in the TimeBank that participated in verb-clause constructions were selected and annotated with the labels BEFORE, OVERLAP and AFTER. The resulting corpus of 895 event-event temporal relations was then used to train a machine learning model. Using a combination of event-level features like tense and aspect with syntax-level features like the paths through the syntactic tree, support vector machine (SVM) models were trained which could identify new temporal relations with 89.2% accuracy. High accuracy models like these are a first step towards automatic extraction of temporal structure from text.

Straipsnyje tiriama šešių būdo prieveiksmių pozicija anglų, ispanų ir lietuvių kalbose. Remiamasi generatyvinės sintaksės metodologija, pagal kurią prieveiksmis sintaksiniame medyje yra laikomas viena stabiliausių kalbos dalių, padedančių nusakyti veiksmažodžio poziciją sakinyje. Taip pat remia­masi Anne Rochette (1990) prieveiksmių klasifikacija, pagal kurią prieveiksmiai gali užimti vieną ar kelias iš šių pozicijų: pradinę (angl. initial), už veiksmažodžio grupės ribų (angl. VP-external/ Aux) arba veiksmažodžio grupės viduje (angl. VP-internal). Tekstyno duomenų tyrimas atskleidė, jog žemiausią vietą prieveiksmių hierarchijoje užimantys anglų kalbos prieveiksmiai well ir fast ne visada eina veiksmažodžio grupės viduje, bet gali eiti ir prieš veiksmažodį, tai leidžia manyti, kad jie gali būti ir aukštesnę vietą sakinio narių hierarchijoje užimantys vienetai. Taip pat buvo pastebėta, jog būdo prieveiksmiai ispanų ir lietuvių kalbose kartais užima skirtingas pozicijas sakinyje nei jų angliški atitikmenys ir todėl priklauso skirtingoms prieveiksmių klasėms. Pavyzdžiui, anglų ir ispanų kalbų prieveiksmiai priklauso IV prieveiksmių klasei, o lietuvių kalbos atitikmenų pozicija sakinyje leidžia juos priskirti I klasei. Straipsnyje pabrėžiama, kad anglų ir is­panų kalbų prieveiksmiai dažniausiai eina tik dviejose pozicijose, t. y. iš karto prieš veiksmažodžio grupę ir jos viduje. Lietuvių kalboje prieveiksmių pozicinė įvairovė yra didžiausia, kuri, manytina, yra susijusi su perteikiamos informacijos kalbančiojo požiūriu svarba.

This paper investigates the linguistic representation of tense and aspect upon a longitudinal case study of the initial stages of the acquisition process, based onspontaneous and semi-spontaneous speech data. We test the hypothesis that (A) when acquired, tense and aspect morphemes reflect only lexical aspect features, inherent to the verb, and that (B) tense and aspect features occupy distinct syntactic heads in the syntactic structure of a sentence. The results indicated that the past perfective morpheme is associated to the lexical aspect achievement, when acquired, at one-word stage, being also associated to accomplishments and activities at multiple-combination stage; the present tense and non-progressive continuous imperfective aspect morpheme is associated to the lexical aspect state, when acquired, at one-word stage; and the present tense and progressive continuous imperfective aspect morpheme is associated to the lexical aspect activity, when acquired, at 24 (twenty-four) months, during two-word stage, being also associated to accomplishments at 25 (twenty-five) months. So, hypothesis A was not refuted, because, when acquired, the perfective, non-progressive and progressive continuous imperfective morphemes reflected punctuality, state or duration features of the lexical aspect, respectively. To account for the association between the imperfective and accomplishments, at two-word stage, we propose that the semantic property telicity is underspecified during one and two-word stages, which is contrary to Li and Shirai (2000)´s proposal that this feature entails the association to the perfective. We defend that the Verkuyl (2002)’s proposal of aspectual classification, favouring aspectual composition over the ontological classification of Vendler (1967), appeared to be more adequate to data treatment. So, we propose one single feature of verb aspectuality, that, at one-word stage, indicates, in a restrictive form, state transition and, at two-word stage, begins to also indicate dynamic progress and non-state. In an analysis of the different verbal forms produced by the child, we verified the association of different morphemes to the same verb, plus syntactic representations indicating verbal movement or feature checking, as evidence of the fact that gerundive forms present grammatical aspect feature-checking and, at the multiple-combination stage, periphrastic forms (“to be” +-ing) present Aspect and Tense feature-checking. So, hypothesis B was not refuted. Thereby, we support the dissociation between Tense and Aspect, based on the existence of an AspP node, dominated by TP. The distinguishing feature of this study is its reliability upon the fact that these records represent, indeed, the initial productions of tense and aspect and it highlights the importance of an analysis not restricted to verifying the associations between lexical aspect and tense/aspect morphology, including the scrutiny of the language acquisition stages and the emergence of other syntactic elements and verbal movements, in order to validate the proposals regarding the functional layer of syntactic structure. Este trabalho investiga a representação linguística de Tempo e Aspecto a partir de umestudo de caso, de caráter longitudinal, dos estágios iniciais do processo de aquisição,com base em dados de produção espontânea e semiespontânea. Colocamos à prova as hipóteses de que (A) os morfemas de tempo e aspecto, ao emergirem, refletem apenas os traços de aspecto lexical, inerentes ao verbo, e (B) os traços de tempo e aspecto ocupam nódulos sintáticos distintos na estrutura sintática sentencial. Os resultados indicaram que o morfema de passado perfectivo se associa ao aspecto lexical culminação, ao emergir, no estágio de uma palavra, passando a se associar, também, a processo culminado e atividade no estágio de combinações múltiplas; o morfema de presente imperfectivo contínuo não progressivo, ao emergir, no estágio de uma palavra, associou-se ao aspecto lexical estado; e o morfema de presente imperfectivo contínuo progressivo, ao emergir, aos 24 (vinte e quatro) meses, durante o estágio de duas palavras, associou-se ao aspecto lexical atividade, passando a se associar, também, a processo culminado, aos 25 (vinte e cinco) meses. Logo, a hipótese A não foi refutada, pois, ao emergir, os morfemas de perfectivo, imperfectivo contínuo não progressivo e progressivo refletiam, respectivamente, os traços de pontualidade, estado ou duratividade do aspecto lexical. Para dar conta da associaçãodo imperfectivo ao processo culminado, no estágio de duas palavras, propomos que a propriedade semântica telicidade é subespecificada nos estágios iniciais, de uma e duas palavras, indo de encontro à proposta de Li e Shirai (2000), de que esse traço ocasiona associação ao perfectivo. Defendemos que a proposta de classificação aspectual de Verkuyl (2002), privilegiando a composicionalidade aspectual, em detrimento da classificação ontológica de Vendler (1967), mostrou-se mais adequada para o tratamento dos dados. Assim, propomos um único traço de aspectualidade para o verbo, que, no estágio de uma palavra, indica, de forma restrita, transição de estado, e, no estágio de duas palavras, passa a indicar, também, progresso dinâmico e não estatividade. Numa análise das diferentes formas verbais produzidas pela criança, verificamos a associação de morfemas diferentes a um mesmo verbo, além de representações sintáticas indicativas de movimento verbal ou checagem de traços, evidenciando que, no estágio de duas palavras, as formas gerundivas apresentam checagem de traços de aspecto gramatical, e, no estágio de combinações múltiplas, a forma perifrástica (“ser +-ndo”) apresenta checagem de traços de Tempo e Aspecto. Logo, a hipótese B não foi refutada. Defendemos, então, a dissociação entre Tempo e Aspecto, a partir da existência de um nódulo AspP, dominado por TP. Este estudo tem como diferencial a credibilidade de que os registros representam, de fato, as produções iniciais de Tempo e Aspecto e destaca a importância de uma análise que não se restrinja a verificar as associações entre o aspecto lexical e a morfologia de tempo e aspecto, incluindo o exame dos estágios de aquisição da linguagem e da emergência de outros elementos sintáticos e movimentos verbais, para validação das propostas acerca da camada funcional da estrutura sintática.