Search

The aim of the present work was to investigate the processing of coreferential relations, focusing on their relationship with the working memory. In a reading process, it is essential that readers continuously perform mental operations that involve the working memory, such as storing, retrieving, and manipulating information. For this reason, it is understood that the processing of coreference resolution is closely related to the working memory. In this relationship between memory and anaphora resolution, the distance factor between the referent and anaphor can play a significant role as distance increases the processing costs when reading sentences. It is known that the working memory is limited; yet there is still no consensus among researchers regarding the nature of storable items. Moreover, the syntactic node, here taken as a representative of the structure of a phrase, was observed as possible storage and recoverable unit in the working memory. Thus, the objective of the study was to verify whether the syntactic distance factor, embeddings with 1 and 2 intervening syntactic nodes, is a storable unit in the working memory, investigating its influence in the processing of different anaphora resolution. The self-paced reading experimental technique was used in an experiment. Seventy participants, all university students, were tested individually and underwent a training session before the task. The results suggest that the syntactic distance is a relevant factor in the working memory with significant differences between conditions with syntactic distance concerning those with linear distance, indicating the formation of chunks. There was an increase in reading times depending on the number of intervening embeddings, and different processing costs for each type of anaphora resolution investigated . The aim of the present work was to investigate the processing of coreferential relations, focusing on their relationship with the working memory. In a reading process, it is essential that readers continuously perform mental operations that involve the working memory, such as storing, retrieving, and manipulating information. For this reason, it is understood that the processing of coreference resolution is closely related to the working memory. In this relationship between memory and anaphora resolution, the distance factor between the referent and anaphor can play a significant role as distance increases the processing costs when reading sentences. It is known that the working memory is limited; yet there is still no consensus among researchers regarding the nature of storable items. Moreover, the syntactic node, here taken as a representative of the structure of a phrase, was observed as possible storage and recoverable unit in the working memory. Thus, the objective of the study was to verify whether the syntactic distance factor, embeddings with 1 and 2 intervening syntactic nodes, is a storable unit in the working memory, investigating its influence in the processing of different anaphora resolution. The self-paced reading experimental technique was used in an experiment. Seventy participants, all university students, were tested individually and underwent a training session before the task. The results suggest that the syntactic distance is a relevant factor in the working memory with significant differences between conditions with syntactic distance concerning those with linear distance, indicating the formation of chunks. There was an increase in reading times depending on the number of intervening embeddings, and different processing costs for each type of anaphora resolution investigated .

Sentiment analysis of short texts is difficult for their simplicity and compactness. This goes a step further when it comes to the Chinese texts. Although deep learning achieved better accuracy in sentiment analysis, there is a lack of explain-ability. Thus, this paper evaluates the effectiveness of techniques for sentiment analysis of Chinese short financial texts with deep learning. For this, we built a Chinese short financial texts corpus (CSFC) and designed an ablation experiment. Beside the CFSC, we used a Chinese review collection and an English short-text repository in the experiment for comparison. There are five techniques involved. They are the Pinyin, the segmentation, the lexical analysis, the phrase structure and the attention mechanism. As results, we found that the phrase structure and the attention mechanism are two of the best. Therefore, the best model in the experiment is called a Phrase Structure and Attention-based Deep network model (PhraSAD). Moreover, to improve the classification accuracy on neutral data, we use a dual classifier strategy for 3-class problems. Experimental results showed that PhraSAD outperformed all other compared models on all experimental datasets.

Finding the structure of a sentence—the way its words hold together to convey meaning—is a fundamental step in language comprehension. Several brain regions, including the left inferior frontal gyrus, the left posterior superior temporal gyrus, and the left anterior temporal pole, are supposed to support this operation. The exact role of these areas is nonetheless still debated. In this paper we investigate the hypothesis that different brain regions could be sensitive to different kinds of syntactic computations. We compare the fit of phrase-structure and dependency structure descriptors to activity in brain areas using fMRI. Our results show a division between areas with regard to the type of structure computed, with the left anterior temporal pole and left inferior frontal gyrus favouring dependency structures and left posterior superior temporal gyrus favouring phrase structures.

Natural language understanding (NLU) is a core technique for implementing natural user interfaces. In this study, we propose a neural network architecture to learn syntax vector representation by employing the correspondence between texts and syntactic structures. For representing the syntactic structures of sentences, we used three methods: dependency trees, phrase structure trees, and part of speech tagging. A pretrained transformer is used to propose text-to-vector and syntax-to-vector projection approaches. The texts and syntactic structures are projected onto a common vector space, and the distances between the two vectors are minimized according to the correspondence property to learn the syntax representation. We conducted massive experiments to verify the effectiveness of the proposed methodology using Korean corpora, i.e., Weather, Navi, and Rest, and English corpora, i.e., the ATIS, SNIPS, Simulated Dialogue-Movie, Simulated Dialogue-Restaurant, and NLU-Evaluation datasets. Through the experiments, we concluded that our model is quite effective in capturing a syntactic representation and the learned syntax vector representations are useful.

The goal of this research is to design and implement a chatbot for querying Wordnet semantic relations. The study creates a contextual chatbot named WordnetBot, a web application that utilizes the use of technologies such as Dialogflow, React, NodeJS, Javascript, and MariaDB. The Wordnet database which leverages all other dictionaries due to its semantic relations representation was used as the data source. Phrase Structure Analysis extracts the keyword and the semantic relation from a user’s message or query. It complements the Machine Learning and AI capabilities of Dialogflow in the analysis. The researcher designed an architectural framework for the integration of the different components of WordnetBot.

A number of natural language processing tools for Urdu language processing have been developed in the past few years for word segmentation, part of speech tagging, chunking, named entity recognition and parsing. Corpora, especially treebanks, are essential data resources for language processing. This work presents the development and evaluation of an Urdu treebank, the CLE-UTB and a statistical parser. The treebank has been annotated with phrase structure annotation. Part of speech tagging has been performed semi-automatically by using an existing tagger and incorrect tags were corrected manually by annotators. The syntactic annotation has been performed in the Penn Treebank style to mark phrases. The annotation scheme also adds functional labels for grammatical roles. Currently, the treebank contains 7854 annotated sentences and 148,575 tokens. Completeness and correctness of the syntactic labels have been checked automatically after manual annotation. To ensure the annotation consistency of the resource, a grammar-based evaluation and an automatic consistency checking tool have been used to detect linguistically implausible constituents. The inter-annotator agreement is greater than 90%. We have developed a bidirectional long-short term memory (BiLSTM) based parser and a POS tagger which have been trained on the final version of the treebank. We have improved our results by training the word embeddings on a large Urdu text corpus. Our parser produced an f-score of 88.1% and the POS tagger performed with an accuracy of 96.3%.

Purpose The results of automatic machine scoring of the Index of Productive Syntax from the Computerized Language ANalysis (CLAN) tools of the Child Language Data Exchange System of TalkBank (MacWhinney, 2000) were compared to manual scoring to determine the accuracy of the machine-scored method. Method Twenty transcripts of 10 children from archival data of the Weismer Corpus from the Child Language Data Exchange System at 30 and 42 months were examined. Measures of absolute point difference and point-to-point accuracy were compared, as well as points erroneously given and missed. Two new measures for evaluating automatic scoring of the Index of Productive Syntax were introduced: Machine Item Accuracy (MIA) and Cascade Failure Rate— these measures further analyze points erroneously given and missed. Differences in total scores, subscale scores, and individual structures were also reported. Results Mean absolute point difference between machine and hand scoring was 3.65, point-to-point agreement was 72.6%, and MIA was 74.9%. There were large differences in subscales, with Noun Phrase and Verb Phrase subscales generally providing greater accuracy and agreement than Question/Negation and Sentence Structures subscales. There were significantly more erroneous than missed items in machine scoring, attributed to problems of mistagging of elements, imprecise search patterns, and other errors. Cascade failure resulted in an average of 4.65 points lost per transcript. Conclusions The CLAN program showed relatively inaccurate outcomes in comparison to manual scoring on both traditional and new measures of accuracy. Recommendations for improvement of the program include accounting for second exemplar violations and applying cascaded credit, among other suggestions. It was proposed that research on machine-scored syntax routinely report accuracy measures detailing erroneous and missed scores, including MIA, so that researchers and clinicians are aware of the limitations of a machine-scoring program. Supplemental Material https://doi.org/10.23641/asha.11984364

Past approaches to translate a phrase in a language L1 to a language L2 using a dictionary-based approach require grammar rules to restructure initial translations. This paper introduces a novel method without using any grammar rules to translate a given phrase in L1, which does not exist in the dictionary, to L2. We require at least one L1-L2 bilingual dictionary and n-gram data in L2. The average manual evaluation score of our translations is 4.29/5.00, which implies very high quality., Comment: 5 pages

A 20,000-sentence Burmese (Myanmar) treebank on news articles has been released under a CC BY-NC-SA license. Complete phrase structure annotation was developed for each sentence from the morphologically annotated data prepared in previous work of Ding et al. [1]. As the final result of the Burmese component in the Asian Language Treebank Project , this is the first large-scale, open-access treebank for the Burmese language. The annotation details and features of this treebank are presented.

Complex nominal groups are common in technical English (i.e., English for Specific Purposes, ESP) since they allow lexical items to be tightly packed into a clause which consequently leads to increased lexical density and syntactic ambiguity. In this paper, we analyse (complex) nominal groups in technical English. We propose that, in addition to context and extralinguistic knowledge (i.e. shared technical background that the ESP teacher does not necessarily possess), the structure of the nominal group — or, more precisely, the position of modifiers within the group also plays a role in resolving of syntactic ambiguity and disambiguation of meaning. Thus, modifiers standing farthest from the head have the least specifying potential and are followed by other modifiers that restrict the meaning of the entire nominal group. In this way, the participle reciprocating in steam reciprocating engine (vs.* reciprocating steam engine ) is more specific in meaning and is thus positioned closer to the head of the nominal group. Our results indicate the type of modification (i.e. linear or non-linear) lends support to the disambiguation of complex nominal groups. The paper‘s main contribution is in the field of ESP teacher education in the way that it helps ESP teachers who are not specialists in the field of (marine) engineering to process understand and successfully teach complex nominal groups.

© 2019 Language Learning Research Club, University of Michigan In the present study, we sought to advance the field of learner corpus research by tracking the development of phrasal vocabulary in essays produced at two different points in time. To this aim, we employed a large pool of second language (L2) learners (N = 175) from three proficiency levels—beginner, elementary, and intermediate—and focused on an underrepresented L2 (Italian). Employing mixed-effects models, a flexible and powerful tool for corpus data analysis, we analyzed learner combinations in terms of five different measures: phrase frequency, mutual information, lexical gravity, delta Pforward, and delta Pbackward. Our findings suggest a complex picture, in which higher proficiency and greater exposure to the L2 do not result in more idiomatic and targetlike output, and may, in fact, result in greater reliance on low frequency combinations whose constituent words are non-associated or mutually attracted.

This study investigated the cognitive processes underlying pauses at different textual locations (e.g., within/between words) and various levels of revision (e.g., below word/clause). We used stimulated recall, keystroke logging, and eye-tracking methodology in combination to examine pausing and revision behaviors. Thirty advanced Chinese L2 users of English performed a version of the IELTS Academic Writing Task 2. During the writing task, participants’ key strokes were logged, and their eye movements were recorded. Immediately after the writing task, 12 participants also took part in a stimulated recall interview. The results revealed that, when participants paused at larger textual units, they were more likely to look back in the text and engage in higher-order writing processes. In contrast, during pauses at lower textual units, they tended to view areas closer to the inscription point and engage in lower-order writing processes. Prior to making a revision, participants most frequently had viewed the text that they subsequently revised or their eye gazes had been off-screen. Revisions focused more on language- than content-related issues, but there was a smaller difference in the number of language- and content-focused stimulated recall comments when larger textual units were revised.

Phrase structure is the arrangement of words in a specific order based on the constraints of a specified language. This arrangement is based on some phrase structure rules which are according to the productions in context free grammar. The identification of the phrase structure can be done by breaking the specified natural language sentence into its constituents that may be lexical and phrasal categories. These phrase structures can be identified using parsing of the sentences which is nothing but syntactic analysis. The proposed system deals with this problem using Deep Learning strategy. Instead of using Rule Based technique, supervised learning with sequence labelling is done using IOB labelling. This is a sequence classification problem which has been trained and modeled using RNN LSTM. The proposed work has shown a considerable result and can be applied in many applications of NLP. Hashi Haris | Misha Ravi "Phrase Structure Identification and Classification of Sentences using Deep Learning" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-3 | Issue-4 , June 2019, URL: https://www.ijtsrd.com/papers/ijtsrd23841.pdf

Treebank is one of the important and useful resources in natural language processing represented in two different annotated schemas: phrase and dependency structures. There are many works that convert a phrase structure into a dependency structure and vice versa. Most of them are based that exploit the handcrafted head percolation table and argument table in predefined deterministic ways. In this article, we propose a method to convert a dependency structure into a phrase structure by enriching a trainable model of former hybrid strategy approach. By adding a classifier to the algorithm and using postprocessing modification, the quality of conversion is increased. We evaluate our method in two different languages, English and Persian, and then analyze the errors. The results of our experiments show a 46.01% reduction of error rate in English and 76.50% for Persian compared to our baseline. We build a new phrase structure treebank by converting 10,000 sentences of Persian dependency treebank into corresponding phrase structures and correcting them manually.

During comprehension, the meaning extracted from serial language input can be described by hierarchical phrase structure. Whether our brains explicitly encode hierarchical structure during processing is, however, debated. In this study we recorded Magnetoencephalography (MEG) during reading of structurally ambiguous sentences to probe neural activity for representations of underlying phrase structure. 10 human subjects were presented with simple sentences, each containing a prepositional phrase that was ambiguous with respect to its attachment site. Disambiguation was possible based on semantic information. We applied multivariate pattern analyses (MVPA) to the MEG data using linear classifiers as well as representational similarity analysis to probe various effects of phrase structure building on the neural signal. Using MVPA techniques we successfully decoded both syntactic (part-of-speech) as well as semantic information from the brain signal. Importantly, however, we did not find any patterns in the neural signal that differentiate between different hierarchical structures. Nor did we find neural traces of syntactic or semantic reactivation following disambiguating sentence material. These null findings suggest that subjects may not have processed the sentences with respect to their underlying phrase structure. We discuss methodological limits of our analysis as well as cognitive theories of “shallow processing”, i.e. in how far rich semantic information can prevent thorough syntactic analysis during processing.

This paper aims at utilizing paralinguistic information usually hidden in speech signals, such as pitch, short pause and sarcasm, to disambiguate user intention not easily distinguishable from speech recognition and natural language understanding results provided by a state-of-the-art spoken dialog system (SDS). We propose two methods to address the ambiguities in understanding name entities and sentence structures based on relevant speech cues and nuances. We also propose an approach to capturing sarcasm in speech and generating sarcasm-sensitive responses using an end-to-end neural network. An SDS prototype that directly feeds signal information into the understanding and response generation components has also been developed to support the three proposed applications. We have achieved encouraging experimental results in this initial study, demonstrating the potential of this new research direction.

Hierarchical sentence structure plays a role in word-by-word human sentence comprehension, but it remains unclear how best to characterize this structure and unknown how exactly it would be recognized in a step-by-step process model. With a view towards sharpening this picture, we model the time course of hemodynamic activity within the brain during an extended episode of naturalistic language comprehension using Combinatory Categorial Grammar (CCG). CCG has well-defined incremental parsing algorithms, surface compositional semantics, and can explain long-range dependencies as well as complicated cases of coordination. We find that CCG-derived predictors improve a regression model of fMRI time course in six language-relevant brain regions, over and above predictors derived from context-free phrase structure. Adding a special Revealing operator to CCG parsing, one designed to handle right-adjunction, improves the fit in three of these regions. This evidence for CCG from neuroimaging bolsters the more general case for mildly context-sensitive grammars in the cognitive science of language.

It is reported that grammatical information is useful for machine translation (MT) task. However, the annotation of grammatical information requires the highly human resources. Furthermore, it is not trivial to adapt grammatical information to MT since grammatical annotation usually adapts tokenization standards which might not be suitable to capture the relation of two languages, and the use of sub-word tokenization, e.g., Byte-Pair-Encoding, to alleviate out-of-vocabulary problem might not be compatible with those annotations. In this work, we propose two methods to explicitly incorporate grammatical information without supervising annotation; first, latent phrase structure is induced in an unsupervised fashion from a multi-head attention mechanism; second, the induced phrase structures in encoder and decoder are synchronized so that they are compatible with each other using constraints during training. We demonstrate that our approach produces better performance and explainability in two tasks, translation and alignment tasks without extra resources. Although we could not obtain the high quality phrase structure in constituency parsing when evaluated monolingually, we find that the induced phrase structures enhance the explainability of translation through the synchronization constraint.

WOS:000641203100008 TARAMAWOS TARAMASCOPUS This experimental study, using a pretest-treatment-posttest design, compared the effects of focused direct written corrective feedback and additional writing practice on L2 learners’ written syntactic complexity. The participants were 60 Iranian elementary EFL learners, whose L2 proficiency as well as L2 writing syntactic complexity and accuracy were controlled by administering the Oxford Quick Placement Test and a paragraph writing test. They were assigned to two groups: Focused direct corrective feedback (FDCF) and additional writing practice without feedback (No CF). The investigation included five sessions and lasted for three weeks. Every session, each participant wrote a paragraph of descriptive type in class. The experimental group received FDCF, while the control (i.e. No CF) group was provided only with additional writing practice. Lu's (2010) web-based L2 Syntactic Complexity Analyzer was utilized to calculate the five indices of syntactic complexity measures, including mean length of clause (MLC), mean length of sentence (MLS), mean length of T-unit (MLT), clauses per sentence (C/S), and verb phrases per T-unit (VP/T). The MANOVA test revealed no statistically significant difference between the two groups

This paper deliberates on the process of building the first constituency-to-dependency conversion tool of Turkish. The starting point of this work is a previous study in which 10,000 phrase structure trees were manually transformed into Turkish from the original PennTreebank corpus. Within the scope of this project, these Turkish phrase structure trees were automatically converted into UD-style dependency structures, using both a rule-based algorithm and a machine learning algorithm specific to the requirements of the Turkish language. The results of both algorithms were compared and the machine learning approach proved to be more accurate than the rule-based algorithm. The output was revised by a team of linguists. The refined versions were taken as gold standard annotations for the evaluation of the algorithms. In addition to its contribution to the UD Project with a large dataset of 10,000 Turkish dependency trees, this project also fulfills the important gap of a Turkish conversion tool, enabling the quick compilation of dependency corpora which can be used for the training of better dependency parsers.

Information-theoretic complexity metrics, such as Surprisal (Hale, 2001; Levy, 2008) and Entropy Reduction (Hale, 2003), are linking hypotheses that bridge theorized expectations about sentences and observed processing difficulty in comprehension. These expectations can be viewed as syntactic derivations constrained by a grammar. However, this expectation-based view is not limited to syntactic information alone. The present study combines structural and non-structural information in unified models of word-by-word sentence processing difficulty. Using probabilistic minimalist grammars (Stabler, 1997), we extend expectation-based models to include frequency information about noun phrase animacy. Entropy reductions derived from these grammars faithfully reflect the asymmetry between subject and object relatives (Staub, 2010; Staub, Dillon, & Clifton, 2017), as well as the effect of animacy on the measured difficulty profile (Lowder & Gordon, 2012; Traxler, Morris, & Seely, 2002). Visualizing probability distributions on the remaining alternatives at particular parser states allows us to explore new, linguistically plausible interpretations for the observed processing asymmetries, including the way that expectations about the relativized argument influence the processing of particular types of relative clauses (Wagers & Pendleton, 2016).

It is broadly acknowledged that collocation is a crucial aspect of lexical knowledge. A well-developed collocational knowledge is necessary to transfer receptive word knowledge into productive use. The aim of the present study was to investigate comparatively the effect of two glossing conditions (textual and audio) and skewed input on lexical collocation of intermediate EFL learners. To this end, 137 L1 Persian EFL students were given Oxford Placement Test (OPT), out of whom 80 were selected and assigned to four 20-participant groups: Textual Glossing Group (TGG), Audio Glossing Group (AGG), Skewed Input Group (SIG), henceforth as experimental groups, and Teacher Existing Method Group (TEMG), as control group. A piloted lexical collocation test, created by familiarity-rating checklist, was administered as a pretest and posttest to the participants to collect data. A delayed posttest was administered after a 2-week interval to assess the delayed effect of the independent variable on the dependent variable of lexical collocation. Collocations were presented in textual glossing, audio glossing, skewed, and L1 glossed conditions during ten instructional sessions. The data were analyzed using one-way ANOVA and post hoc test for both immediate and delayed posttests. The results showed that the different gloss conditions and skewed input technique had significantly different effects on learners’ L2 lexical collocation gain. It was revealed that textual glossing and skewing techniques resulted in greater improvement in lexical collocation. Some pedagogical implications were represented for language teachers, learners, and testers.

X' theory was a major milestone in the history of the development of generative grammar.1 It enabled important insights to be made into the phrase structure of human language, but it had a number of weaknesses, and has been essentially replaced in Chomskyan generativism by Bare Phrase Structure (BPS), which assumes fewer theoretical primitives than X0 theory, and also avoids several of the latter’s weaknesses. However, Bare Phrase Structure has not been widely adopted outside the Minimalist Program (MP), rather, X0 theory remains widespread. In this paper, we develop a new, fully formalized approach to phrase structure which incorporates insights and advances from BPS, but does not require the Minimalist-specific assumptions that come with BPS. We formulate our proposal within Lexical-Functional Grammar (LFG), providing an empirically and theoretically superior model for phrase structure compared with standard versions of X0 theory current in LFG.

This study aims at investigating the challenges that Yemeni translation students encounter when translating neologisms from English into Arabic. It also aims at comparing students' translation with outcomes of machine translation (MT). The authors follow the descriptive and comparative methods in conducting this study. To achieve the objective of the study, a test consisted of 24 items was designed to uncover such difficulties and was given to 55 translation students. The same test was applied on three translation applications for comparing the performance of students with MT. The findings show that translating neologisms accurately is hard for most of the Yemeni translation students and also for machine translation as well. It is difficult for them to understand English neologisms or to find the Arabic equivalents for English neologisms. The finding also indicates that only a few translation students can achieve accurate translation. Meanwhile, Translation applications produced poor translation and could not able to achieve accurate translation for most types of neologisms. As a result of the comparison, students produced acceptable translation better than MT. The difficulties are lexical related to the twelve types of neologisms based on Newmark classification (1988).

Voice and v is an investigation of the syntax of an understudied Western Austronesian language, Acehnese, with a particular interest in its implications for the theory of verb phrase structure under the framework of the Minimalist Program. Since Pylkkänen's seminal article, the idea that the functional projection of verb phrases involves two distinct layers—a higher one (that is, Voice) that is responsible for introducing the external thematic role and Case-licensing the internal argument, and a lower one (that is, v) that is responsible for introducing causative semantics and verbalizing the root—has been advanced in a series of works under the Minimalist Program. This book presents novel evidence for this hypothesis based on an in-depth analysis of Acehnese passive, object voice, and causative constructions. Building on the empirical observations from Acehnese, the book makes further explorations of the syntactic typology of passives and causatives, on which the Acehnese data shed light. It contributes not only to the description and analysis of an understudied language, but also to the cross-linguistic understanding of the different flavors of Voice and the architecture of verb phrase structures.

Corpora are well-known for the affordance to make linguistic regularities salient. Since the coinage of the term ‘data-driven learning’ (DDL) in the 1990s, much has been done to investigate the eff...

This study investigated the optimal conditions under which foreign language learning takes place using a smartphone. The authors proposed that two factors, chunking and imagery, would increase language learning. On the basis of previous findings, they formulated three hypotheses: (a) the use of images will have a positive effect on learning English sentences; (b) the use of chunks will have a positive effect on learning English sentences; and (c) the combined use of images and chunks will have a greater positive effect on learning English sentences than either feature alone. A total of 92 Korean seventh graders participated in this study. To examine the learning effect of chunking (i.e., sentence segmentation unit) and imagery (i.e., visual aid) in an experimental setting, they produced a smartphone learning application that incorporated the two methods. The authors measured learning effect with respect to lexical memory retention (i.e., word retrieval ability) and word order composition (i.e., ability to arrange words according to standard English syntax). The results show that the main effects of both chunking and imagery were significant and that the interaction effect between the two on lexical memory retention was also significant. The interaction effect was greater in the delayed effect measurement than in the immediate effect measurement. These findings suggest optimal conditions for designing a smartphone-based, self-learning application.

The present study aims to investigate the role of L1 transfer effects on L2 sentence processing strategies during the interpretation of relative clause (RC) attachment ambiguities. The main body of the study is divided into two sections. The first section describes Experiment 1, which is designed to test the resolution of RC attachment ambiguities by Turkish learners of L2 English both in Turkish and English through the use of an off-line task (i.e., paper-and-pencil comprehension tests) and compare their processing preferences to those of native English speakers. The second section presents Experiment 2, which aims to investigate the real-time processing of the RC attachment ambiguities by the same participant groups employing eye-tracking methodology. The results indicated that L1 Turkish and L1 English RC attachment preferences differed and that Turkish learners of L2 English tended to transfer their Turkish sentence processing pattern to real-time interpretation of the English RC attachment ambiguities.

Learning semantic representations of bilingual phrases is very important for statistical machine translation to overcome data sparsity and exploit semantic information. In this paper, we consider word alignments as a semantic bridge between the source and target phrases, and propose two neural networks based on the conventional recursive autocoder, which exploit word alignments to generate alignment-consistent bilingual phrase structures: One is Alignment Enhanced Recursive Autoencoder that incorporates a word-alignment-related error into the final objective function; The other is Alignment Guided Recursive Neural Network which treats word alignments as direct signals to guide phrase structure constructions. Then, we further establish the semantic correspondences between the source and target nodes of the generated bilingual phrase structures via word alignments. By jointly minimizing recursive autoencoder reconstruction errors, structural alignment consistency errors and cross-lingual reconstruction errors, our model not only generates alignment-consistent phrase structures, but also captures different levels of semantic correspondences within bilingual phrases. Experiments on the NIST Chinese-English translation task show that our model achieves significant improvements over the baseline.

Different languages have different means for structuring clauses which allow the coding of a thematically peripheral argument or adjunct as a core-object argument. The resulting constructions are known as double object or applicative constructions. The primary aim of this paper is to present a syntactic analysis of applicative constructions in Sudanese Arabic within the theoretical framework of the Minimalist Program (Chomsky, 1993, 1995) , in particular, the notion of phases, in combination with Pylkkanen's proposed phrase structure and the semantics of I-applicative (2000, 2008) . The overall endeavor is to provide answers to the central questions: how applicatives in Sudanese Arabic are derived? Whether applicatives in Sudanese Arabic are I-applicatives or E-applicatives? The derivation of such constructions in Sudanese Arabic involves the coding of new argument in the argument structure of the verb. This new argument is introduced via a preposition and has a benefactive/goal interpretation. This argument is c-commanded by any internal argument. The applied argument is placed in the complement position of the head ApplI, and can undergo neither A-movement nor wh-movement. This makes Sudanese Arabic fit into the general syntactic typology of I-type applicative languages cross-linguistically and consequently promoting contrastive linguistics. Keywords: Applied argument; applicative construction; minimalist program; spoken Sudanese Arabic; syntax

The use of ‘chunks’ is not only a common and characteristic feature of first language use, but may also be a distinguishing factor between less and more proficient second language users. The present study aimed at investigating potential correlations between chunk use and holistically rated proficiency scores and development over time in advanced Chinese learners of English as a second language (L2) over the course of 18 months of English classes. Neither holistically rated proficiency scores nor common complexity measures had shown any progress in these learners, but a few lexically based measures did. Therefore, the development of chunks in these learners was examined as chunks are mostly lexical in nature and might show more subtle progress over time. The use of chunks was examined in the first and last two texts of the learners and, in line with previous research, it was found that more proficient writers use relatively more chunks, specifically collocations. Chunk coverage, the total number of chunk words divided by the total number of words, also correlated with some holistic proficiency scores, and increased significantly over time. Average chunk length did not show any correlation with proficiency scores nor did it increase over time. Increased proficiency in written texts is thus specifically related to the use of more chunks. This study shows that development in advanced Chinese L2 learners of English appears to be subtle, which might be due to the fact that at this advanced level mainly lexical changes take place. Future research into development over time in advanced learners may thus have to zoom in on lexical measures especially. Implications of this study point to the importance of promoting the use of target language chunks to advanced learners of English.

Syntactic complexity is an important measure of second language (L2) writing proficiency (Larsen–Freeman, 1978; Lu, 2011). Large‐grained indices such as the mean length of T‐unit (MLTU) have been used with the most consistency in L2 writing studies (Ortega, 2003). Recently, indices such as MLTU have been criticized, both for the difficulty in interpretation (e.g., Norris & Ortega, 2009) and for a potentially misplaced focus on clausal subordination (e.g., Biber, Gray, & Poonpon, 2011). In this article, we attempt to address both of these criticisms by using traditional indices of syntactic complexity (e.g., MLTU), fine‐grained indices of clausal complexity, and fine‐grained indices of phrasal complexity to predict holistic scores of writing quality. In 4 studies, we used indices of each index type to predict holistic writing quality scores in independent essays on the Test of English as a Foreign Language (TOEFL). We then used all index types in a combined analysis to predict a holistic writing score. The results indicated that fine‐grained indices of phrasal complexity were better predictors of writing quality than either traditional or fine‐grained clausal indices, though a single fine‐grained index of clausal complexity contributed to the combined model. These results provide some support for Biber et al.’s (2011) claims regarding complexity and academic L2 writing proficiency. [ABSTRACT FROM AUTHOR]

Many sentences of adult English are analytic constructions, namely clauses with a matrix verb complemented by a dependent predicate that does not have an expressed syntactic subject. Examples are subject and object control, raising to subject or object, periphrastic tense, aspect and modality, copular predication and do-support. In this article the authors test a suggestion derived from Dependency Grammar that despite differences in detail, all such constructions are governed by a common principle of structure sharing which young children master when they produce such sentences. Analytic sentences and telegraphic sentences were examined in the speech of 439 young children, mean age 2;3.11 ( SD = 0;4.02). The production of different analytic constructions was significantly associated, raising the probability of each other by 32% on average. Telegraphic sentences overtly expressing the input’s covert predicate–argument relations also positively predicted the production of analytic sentences. These results suggest that children learn a general principle of sharing arguments, common to constructions with dependent predicates, making transfer and facilitation possible.

The aim of this work is to analyze word order variations of modern standard Arabic sentences within the linguistic framework of Generalized Phrase Structure Grammar (GPSG) by Gazdar, Pullum, Klein, and Sag (1985). This paper provides a direct interpretation as to the word order variations in Arabic either through a rule-to-rule formalism (including metarules in addition to other generalizations as for the rules of the grammar) or semantics. It tends not to suppose multistratal syntactic representations.

Software can facilitate English as a Foreign Language (EFL) students’ self-correction of their free-form writing by detecting errors; this article examines the proportion of errors which software can detect. A corpus of 13,644 words of written English was created, comprising 90 compositions written by Spanish-speaking students at levels A2-B2 (inclusive) of the Common European Framework. A total of 1,310 language errors were detected by the researcher. It was found that approximately 21% of these errors were spelling errors. A further 58% were characterised as either two-word phrases (45%), three-word phrases (9%), or four- and five-word phrases (4%) which are either absent from or rare in a large corpus of English which is known to be correct. The nature of software which can detect such words and phrases and bring them to students’ attention with a view to self-correction is briefly described. Of the remaining 21% of errors not detected by such software, most were found to be either errors of te...

Van Trijp (2013,2014) claims that Sign-Based Construction Grammar (SBCG) and Head-Driven Phrase Structure Grammar (HPSG) are fundamentally different from Fluid Construction Grammar (FCG). He claims that the former approaches are generative ones while the latter is a cognitive-functional one. I argue that it is not legitimate to draw these distinctions on the basis of what is done in FCG. Van Trijp claims that there are differences in the scientific model, the linguistic approach, formalization, the way constructions are seen, and in terms of processing. This paper discusses all these alleged differences. Van Trijp also claims that his cognitive-functional approach is superior in terms of completeness, explanatory adequacy, and theoretical parsimony. In order to facilitate a discussion and comparison, I introduce the reader to basic assumptions made in FCG and the analyses suggested by Van Trijp: I first deal with the representations that are used in FCG, talk about argument structure constructions, the combination operations fusion and merging that are used in FCG, I than discuss the analysis of nonlocal dependencies and show that the suggested FCG analysis is not explanatorily adequate since it is not descriptively adequate and that a full formalization of approaches with discontinuous constituents is not more parsimonious than existing HPSG analyses either. After the discussion of specific analyses, I then provide a detailed comparison of FCG and SBCG/HPSG and discuss questions like the competence/performance distinction, mathematical formalization vs. computer implementation, fuzziness and fluidity in grammars, and permissiveness of theories. I conclude that HPSG, SBCG, and FCG belong to the same family of theories and that all claims to the contrary are unjustified.

Statements containing quantity information are commonplace. Although there is literature explaining the way in which quantities themselves are conveyed in numbers or words (e.g., many, probably), there is less on the effects of different types of quantity description on the processing of surrounding text. Given that quantity information is usually conveyed to alter our understanding of a situation (e.g., to convey information about a risk), our understanding of the rest of the quantified statement is clearly important. In this article texts containing quantified statements expressed numerically versus verbally are compared in two text change experiments to assess how the entire quantified noun phrase is encoded in each case. On the basis of the results it is argued that numerical quantifiers place focus on the size of a subset, whereas verbal quantifiers are better integrated with nouns leading to more focus on the subset itself.

Collocations are word combinations essential for achieving fluency in a given language. Considerable emphasis should therefore be placed on teaching collocations as a part of vocabulary instruction in language teaching. However, there is no current consensus on how best to teach collocations, and few studies have addressed the issue. This quasi-experimental study investigated the effectiveness of online tools for learning English adjective-noun collocations compared to learning collocations via traditional activities. A quasi-experimental design with a pretest and immediate and delayed posttests was applied to 53 participants (n = 25 for the control group; n = 28 for the experimental group). The test results reveal that participants in the experimental group who learned collocations with online tools performed significantly better on both immediate and delayed posttests, demonstrating the effectiveness of these tools on learning adjective-noun collocations.

Frequency and contingency are two primary statistical factors that drive the acquisition and processing of language. This study explores the role of phrasal frequency and contingency (the co-occurrence probability/statistical association of the constituent words in multiword sequences) during online processing of multiword sequences. Meanwhile, it also examines language users’ sensitivity to the two types of statistical information. Using the eye-tracking paradigm, native and advanced nonnative speakers of Chinese were instructed to read 80 disyllabic two-word Chinese adverbial sequences embedded in sentence contexts. Eye movements of the participants were recorded using both early and late measures. Mixed-effects modeling revealed that both phrasal frequency and contingency influenced the processing of the adverbial sequences; however, they were likely to function in different time windows. In addition, both native and nonnative speakers were sensitive to the phrasal frequency and contingency of the sequences, though their degrees of such sensitivity differed. Our findings suggest that adult language learners retain the statistical learning ability in second language acquisition and they may share a general statistical learning mechanism with native speakers when processing multiword sequences.

In order to build a natural language processing system first the words are placed into a structured form that leads to a syntactically correct sentence. Syntactic analysis of a sentence is performed by parsing technique. This paper explores the novel approach that how the shift reduce parsing technique is used for translating English sentences into a grammatically correct Kannada sentences by reordering of English parse tree structure, generating and implementing phrase structure grammar(PSG) for kannada sentences. Recursive Descent Parsing technique is used to generate English phrase tree structure and terminal symbols are tagged with Kannada equivalent words then Shift-Reduce Parsing technique is used to construct a Kannada sentence. Part-of-Speech (POS) tagger is used to tag Kannada words to English words. It is implemented by using supervised machine learning approach

A problem in automatically generated stories for image sequences is that they use overly generic vocabulary and phrase structure and fail to match the distributional characteristics of human-generated text. We address this problem by introducing explicit representations for objects and their relations by extracting scene graphs from the images. Utilizing an embedding of this scene graph enables our model to more explicitly reason over objects and their relations during story generation, compared to the global features from an object classifier used in previous work. We apply metrics that account for the diversity of words and phrases of generated stories as well as for reference to narratively-salient image features and show that our approach outperforms previous systems. Our experiments also indicate that our models obtain competitive results on reference-based metrics.

This study aims to establish whether the processing of different connectives (e.g., and, but) and different coherence relations (addition, contrast) can be modulated by a structural feature of the connected segments—namely, parallelism. While but is mainly used to contrast two expressions, and occurs in many different relations and has been shown to come with a processing cost. We report three self-paced reading experiments in which we manipulate whether the connected segments share a common verb phrase. Such parallel constructions frequently occur in contrastive relations, although they are typically treated as additive in comprehension research. We expect that parallelism will compensate for the cognitive complexity of contrast and for the ambiguity of and by further signaling the coherence relation. Our results indicate that parallelism speeds up processing and provides further evidence for priming in comprehension. However, parallelism interacted with connective ambiguity in an overt disambiguation task (Experiment 3) but not in a more natural reading task (Experiment 2). We argue that the processing of contrast remains shallow unless disambiguation is explicitly required.

Long Short-Term Memory recurrent neural network (LSTM) is widely used and known to capture informative long-term syntactic dependencies. However, how such information are reflected in its internal vectors for natural text has not yet been sufficiently investigated. We analyze them by learning a language model where syntactic structures are implicitly given. We empirically show that the context update vectors, i.e. outputs of internal gates, are approximately quantized to binary or ternary values to help the language model to count the depth of nesting accurately, as Suzgun et al. (2019) recently show for synthetic Dyck languages. For some dimensions in the context vector, we show that their activations are highly correlated with the depth of phrase structures, such as VP and NP. Moreover, with an $L_1$ regularization, we also found that it can accurately predict whether a word is inside a phrase structure or not from a small number of components of the context vector. Even for the case of learning from raw text, context vectors are shown to still correlate well with the phrase structures. Finally, we show that natural clusters of the functional words and the part of speeches that trigger phrases are represented in a small but principal subspace of the context-update vector of LSTM.

Infants exploit acoustic boundaries to perceptually organize phrases in speech. This prosodic parsing ability is well-attested and is a cornerstone to the development of speech perception and grammar. However, infants also receive linguistic input in child songs. This study provides evidence that infants parse songs into meaningful phrasal units and replicates previous research for speech. Six-month-old Dutch infants (n = 80) were tested in the song or speech modality in the head-turn preference procedure. First, infants were familiarized to two versions of the same word sequence: One version represented a well-formed unit, and the other contained a phrase boundary halfway through. At test, infants were presented two passages, each containing one version of the familiarized sequence. The results for speech replicated the previously observed preference for the passage containing the well-formed sequence, but only in a more fine-grained analysis. The preference for well-formed phrases was also observed in the song modality, indicating that infants recognize phrase structure in song. There were acoustic differences between stimuli of the current and previous studies, suggesting that infants are flexible in their processing of boundary cues while also providing a possible explanation for differences in effect sizes.