Sentence Position Research Articles

Metaphor processing: Referring and predicating

The general consensus emerging from decades of empirical investigation of metaphor processing is that, when appropriately contextualised, metaphorically used language is no more demanding of processing effort than literally used language. However, there is a small number of studies which contradict this position, notably Noveck, Bianco, and Castry (2001): they maintain that relevance-based pragmatic theory predicts increased cognitive costs incurred in deriving the extra effects that metaphors typically yield, and they provide experimental results that support this prediction. In our study, we first surveyed and assessed the tasks and stimulus materials of many experiments on metaphor processing from the 1970's to the present day. The most telling result was an apparent disparity between the processing of metaphorical language used predicatively versus referentially. We then ran two self-paced reading experiments to test our hypothesis that when used as a predicate, metaphorical language is no more costly than literal language, but when used referentially, it does incur extra costs, even given a preceding biasing context. In the first experiment, all metaphorical referring expressions were in subject position so occurred early in the sentence; in the second experiment, we controlled for any effect of sentence position by placing metaphorical referring expressions in object position, thus later in the sentence, similar to the predicate metaphors. In both cases, metaphorical referring incurred significantly greater costs relative to literal equivalents than did metaphorical predication, with no effect of sentence position. We end with a brief analysis of why the referential use of metaphor is special and effort-demanding.

Young Pashtun L1 speakers' laryngeal contrast and voicing variation in Korean word-initial stops production

While age proved to have a critical influence on learning L2 phonology, the reason behind it is yet controversial. (brain maturation, LOR, quality l2 input, crib or sequential bilingualism, etc.) This paper examines 6 young people in the same household using Pashtun as their L1. As the elder children arrived in Korea at ages 10, 8, 6, and 3 respectively, the younger two were born in Korea now aged 10 and 8. All family members have equal LOR(10 years except for the youngest), equal L1, and equal L1 and L2 use extent in the home. Examining the differences between VOT and f0 production in these young people will clarify some of the explanations behind the age effects. Moreover, the data reveals the unique L1 transfer phenomena. While Pashtun is a true voicing language, Korean is an aspirating language where all 3- way contrast stops have positive VOT values. This can lead the young people to produce word-initial Korean stops in intersonorant positions with a marked voicing-ratio and to produce allophones that neither exists in their L1 nor in L2 such as voiced aspirated stop sounds ( bɦ, d̪ɦ, gɦ). This paper will also examine whether these young people’s VOT and f0 productions would align with the current tonogenesis like sound change in Korean. This paper used picture naming tasks with easy words instead of reading non-words to not test their grapheme-phoneme correspondences. Each person produces 9 phonemes of word-initial Korean stops (3-way laryngeal contrasts in 3 POA) both in the position in isolation and carrier sentence.

The Emphatic Marker in Turkish: Two Readings

This paper aims to show that the anaphor kendi in Turkish has two distinct readings depending on when it’s functioning as an emphatic marker, which hasn’t been noted in the previous literature. I propose that it has two distinct readings based on their surface sentential positions and diagnostics from Ahn (2010). Moreover, I show that the event modifying a variety of the emphatic marker in Turkish differs from the same type of emphatic marker in English by showing that (i) it allows theme/patient referents, and the DP modifying variety (ii) may be able to survive without obligatory stress. In addition, I document further observations and nuances regarding the morphological marking of the two varieties of the emphatic marker in Turkish as preliminary discussion points.

At the margins of grammar

This paper focuses on the variable placement of the particle of particle verb constructions in Dutch and German. In particular, the option to place the particle in first sentence position is studied in a judgment task. We asked native speakers of Dutch to judge utterances which differ with respect to the semantic transparency of the particle and with respect to the distance between particle and verb. In spite of an overall awkwardness of constructions with a fronted particle, the factors under investigation turned out to influence their well-formedness. Our observations thus raise some questions about the acceptability and the grammaticality status of such dubious structures. When comparing our results to those found by Trotzke et al. (2015), who did a similar study for German, we find that the patterns in the two languages are similar but that there are nonetheless some important differences to note. We suggest that this discrepancy has to do with the more flexible use of the first sentence position in German.

A Blueprint for Genetic Determinism

A Blueprint for Genetic Determinism

Caregiver object labels within supported and coordinated joint engagement during interaction with toddlers at elevated and typical likelihood of autism.

Early in development, caregivers' object labelling contributes to children's word learning. Language development is a bi-directional process, and differences in joint engagement (JE) and language among children with developmental disabilities such as autism spectrum disorder (ASD) may provide caregivers varying contexts and opportunities to provide object labels. However, potential variation in caregivers' production of object labels and its relation to language development remain relatively unexplored among toddlers with ASD. This study characterized the structural and functional features of object labels produced by parents of children with typical (TL) or elevated likelihood (EL) of ASD during naturalistic toy play. We examined features of object labels within two JE contexts, supported and coordinated JE, which are differentiated by a child's use of eye contact, as well as their relations with concurrent and future child language skills. The present study included 55 (TL = 12, EL = 43) children who completed a naturalistic parent-child interaction in the home at 18 months of age. Children's expressive and receptive language was assessed at 18, 24 and 36 months. At 36 months, EL children were assessed for ASD and classified as either EL-No Diagnosis, EL-Language Delay or EL-ASD. Videos of interactions were divided into discrete engagement states, including supported and coordinated JE. All parent speech was transcribed and coded to capture structural (types, tokens, mean length of utterance (MLU), sentence position) and functional (follow-in comments, directives, lead-in labels) features of object labels as well as parent prompts for the child to produce a label. Parents of toddlers across outcome groups labelled objects at similar rates within each engagement state. However, parents of EL-ASD children provided the lowest rates of prompts for labels in supported JE and the highest rate of labels as the final word of an utterance (sentence-final position) in coordinated JE. Additionally, parent prompts in supported JE were related to concurrent child expressive language. Labels in sentence-final position were positively related to later language outcome when delivered in supported JE but were associated with poorer language outcomes when delivered in coordinated JE. Subtle differences in parent object labels across outcome groups demonstrate the role that child language and social engagement can play in influencing parent input and the cascading impact of this input on language development. What is already known on this subject Variations in caregiver object labelling can impact child language development. However, child characteristics such as language ability also actively shape the input caregivers provide, demonstrating the bi-directionality of language development. What this paper adds to existing knowledge The present study demonstrates that characteristics of the engagement context in which a label is delivered may be important for understanding how object labelling relates to child language acquisition and whether this relation varies for children who face challenges in language learning. What are the potential or actual clinical implications of this work? As child differences in social engagement emerge, parents may be more attuned to moments their children are engaging with eye contact. Caregiver-mediated interventions might consider strategies that guide caregivers in recognizing engagement without eye contact as a similarly meaningful opportunity for learning and encourage the use of rich input within these moments.

Binary Particle Swarm Optimization with an improved genetic algorithm to solve multi-document text summarization problem of Hindi documents

Automatic text summarization plays a vital role in text retrieval. However, very little interest, as well as attention to Hindi text summarization and the wide use of hybridization of an evolutionary and swarm-based method for solving optimization problems, motivates us to develop a technique for achieving better results. We have used a combination of the Title feature, Sentence length, Sentence position, Numerical Data, Thematic word, Proper Noun and Term frequency & Inverse Sentence Frequency to find the results. The proposed BPSO-IGA comprising Binary Particle Swarm Optimization (BPSO) and an improved genetic algorithm finds a well-formatted summary from multiple documents. The method for calculating the score of the sentence is proposed initially. A combination of BPSO and IGA finds the features’ optimal scores and then is used to generate the final summary. The proposed method is compared with six well-known techniques, SummaRuNNer, ChunkRank, TGraph, UniRank, FF-SE, and MDS-ACS, on FIRE 2011 Hindi dataset. To strengthen the hypothesis that all seven features are required in the summary generation, the summary is also generated using only ONE and a combination of more than one feature. It is found that the best summary is generated when all seven features are considered. The BPSO-IGA performs best among all the methods for precision, recall, f-measure, cohesion, and non-redundancy. This research aims to design a multi-document extractive text summarizer for Hindi documents using a swarm intelligence and evolutionary algorithm-based hybrid approach.

Length, position, and functions of inter-clausal Chinese–English code-switching in a bilingual novel

AbstractBased on a written Chinese–English code-switching corpus compiled from a Singaporean bilingual novel, this paper reports on the distribution of instances of inter-clausal code-switching in terms of their length, the distribution of the position of instances of inter-clausal code-switching in mixed sentences, and the pragmatic functions of inter-clausal code-switching of different lengths in different sentence positions. The study yields four results. First, the distribution in terms of length fits Zipf’s Law well, and bilingual speakers attempt to avoid lengthy inter-clausal code-switching in their communication. Second, the majority of inter-clausal code-switching appears in sentence boundary positions. Third, the major pragmatic functions of shorter or longer instances of inter-clausal code-switching are turn-opening, turn-closing, explicitness, and accuracy. Finally, instances of inter-clausal code-switching with medium clause lengths have two main functions, which are the conveying of irony and emphasis.

A Pragmatic Function-Position Analysis of Address Terms: Tendencies in Turkish

This study, with the broader aim of describing the addressing system in Turkish, analyzes the addressing terms as inseparable parts of communication, in terms of their sentential positions, pragmatic functions and function-position mapping properties on the database of Turkish National Corpus and presents the findings within a synthesis of broader literature on different languages. With that purpose in mind, three research questions will be answered. The first one is about the pragmatic functions of address terms in Turkish in the corpus data, the second one is about the interaction of these functions with their sentential positions and third one is about the extent to which the findings of the study and the findings in the literature for other languages. A specified sub-set from the Turkish National Corpus (TNCV3) (Aksan, 2012) was used as the database of pieces of natural language. The three major findings of the study are that Turkish address terms have ten basic functions, these functions occur in four different sentential positions, there are some tendencies in function-position mappings and that although there are Turkish specific aspects, correspondences with the studies especially in English are much more. With its Turkish specific findings, the study will hopefully contribute to the analyses of the addressing systems in languages as a whole.

The Alethic Platitudes, Deflationism, and Adverbial Quantification

Abstract Alethic pluralists often claim that accommodating certain alethic platitudes motivates rejecting deflationism in favour of a pluralist inflationism about truth. Deflationists claim that the logical role of the truth predicate, viz providing something equivalent to variables for sentence-in-use positions and quantifiers governing them, is sufficient to account for the appeal to truth in the alethic platitudes. Surprisingly, however, most deflationists face an insufficiently acknowledged problem with respect to explaining how this mode of generalizing works. The standard substitutional or higher-order interpretations of sentential quantifiers and variables do not meet two desiderata that we claim any adequate account of them must satisfy. To address this issue, we review and extend A. N. Prior's adverbial understanding of sentential quantification, explain how it satisfies the desiderata, and respond to some objections. This shows that deflationists can accommodate and account for the alethic platitudes by applying this non-nominal understanding of generalizing on sentence positions.

DeepSumm: Exploiting topic models and sequence to sequence networks for extractive text summarization

In this paper, we propose DeepSumm, a novel method based on topic modeling and word embeddings for the extractive summarization of single documents. Recent summarization methods based on sequence networks fail to capture the long range semantics of the document which are encapsulated in the topic vectors of the document. In DeepSumm, our aim is to utilize the latent information in the document estimated via topic vectors and sequence networks to improve the quality and accuracy of the summarized text. Each sentence is encoded through two different recurrent neural networks based on probabilistic topic distributions and word embeddings, and then a sequence to sequence network is applied to each sentence encoding. The outputs of the encoder and the decoder in the sequence to sequence networks are combined after weighting using an attention mechanism and converted into a score through a multi-layer perceptron network. We refer to the score obtained through the topic model as Sentence Topic Score (STS) and to the score generated through word embeddings as Sentence Content Score (SCS). In addition, we propose Sentence Novelty Score (SNS) and Sentence Position Score (SPS) and perform a weighted fusion of the four scores for each sentence in the document to compute a Final Sentence Score (FSS). The proposed DeepSumm framework was evaluated on the standard DUC 2002 benchmark and CNN/DailyMail datasets. Experimentally, it was demonstrated that our method captures both the global and the local semantic information of the document and essentially outperforms existing state-of-the-art approaches for extractive text summarization with ROUGE-1, ROUGE-2, and ROUGE-L scores of 53.2, 28.7 and 49.2 on DUC 2002 and 43.3, 19.0 and 38.9 on CNN/DailyMail dataset.

Improving extractive document summarization with sentence centrality

Extractive document summarization (EDS) is usually seen as a sequence labeling task, which extracts sentences from a document one by one to form a summary. However, extracting sentences separately ignores the relationship between the sentences and documents. One solution is to use sentence position information to enhance sentence representation, but this will cause the sentence-leading bias problem, especially in news datasets. In this paper, we propose a novel sentence centrality for the EDS task to address these two problems. The sentence centrality is based on directed graphs, while reflecting the sentence-document relationship, it also reflects the sentence position information in the document. We implicitly strengthen the relevance of sentences and documents by using sentence centrality to enhance sentence representation. Notably, we replaced the sentence position information with sentence centrality to reduce sentence-leading bias without causing model performance degradation. Experiments on the CNN/Daily Mail dataset showed that EDS models with sentence centrality significantly improved compared with baseline models.

UGDAS: Unsupervised graph-network based denoiser for abstractive summarization in biomedical domain.

Abstractive summarization models can generate summary auto-regressively, but the quality is often impacted by the noise in the text. Learning cross-sentence relations is a crucial step in this task and the graph-based network is more effective to capture the sentence relationship. Moreover, knowledge is very important to distinguish the noise of the text in special domain. A novel model structure called UGDAS is proposed in this paper, which combines a sentence-level denoiser based on an unsupervised graph-network and an auto-regressive generator. It utilizes domain knowledge and sentence position information to denoise the original text and further improve the quality of generated summaries. We use the recently-introduced dataset CORD-19 (COVID-19 Open Research Dataset) on text summarization task, which contains large-scale data on coronaviruses. The experimental results show that our model achieves the SOTA (state-of-the-art) result on CORD-19 dataset and outperforms the related baseline models on the PubMed Abstract dataset.

Sentential position of V[sbnd]N combination modulates the rhythmic pattern effect during Chinese sentence reading: Evidence from eye movements

Though previous research has examined how implicit meter can facilitate the processing of stress-timed languages, syllable-timed languages, such as Chinese, remain under studied. Past research has shown that among verb-noun combinations in Chinese, the processing of [2 + 2] (two disyllabic words) combination rhythmic pattern is easier than that of [2 + 1] (a disyllabic word and a monosyllabic word) pattern, though it is unclear whether this effect is modulated by the sentential position of the verb-noun combination. The present study uses eye-tracking to examine the influence of position on rhythmic pattern during silent reading. In Experiment 1, participants read sentences with [2 + 1] versus [2 + 2] VN phrases embedded in different sentential positions. Results show that the fixation duration of [2 + 1] VN phrases is significantly longer than that of [2 + 2] and that the fixation duration of VN phrases is shorter at the sentence-middle position than it is at the sentence-final position, suggesting that the rhythmic pattern effect at the sentence-middle position exhibits a reduced magnitude compared to the sentence-final position. In Experiment 2, participants read sentences with either mono- or disyllabic words after the VN phrases to further explore whether the reduction of the rhythmic pattern effect is related to the number of succeeding syllables. Results show that while the fixation duration of the [2 + 1] VN pattern is significantly longer than that of the [2 + 2] pattern, there is no significant difference between the monosyllabic versus the disyllabic conditions, nor is there a significant interaction between rhythmic pattern and syllable length post VN phrases, thus ruling out the rhythmic effect from succeeding context. Together, these patterns suggest that the reduction of the rhythmic pattern effect is caused by position rather than number of syllables after phrases.

Capacity limits in sentence comprehension: Evidence from dual-task judgements and event-related potentials

There is an ongoing controversy over whether readers can access the meaning of multiple words, simultaneously. To date, different experimental methods have generated seemingly contradictory evidence in support of serial or parallel processing accounts. For example, dual-task studies suggest that readers can process a maximum of one word at a time (White, Palmer & Boynton, 2018), while ERP studies have demonstrated neural priming effects that are more consistent with parallel activation (Wen, Snell & Grainger, 2019). To help reconcile these views, I measured neural responses and behavioral accuracy in a dual-task sentence comprehension paradigm. Participants saw masked sentences and two-word phrases and had to judge whether or not they were grammatical. Grammatically correct sentences (This girl is neat) produced smaller N400 responses compared to scrambled sentences (Those girl is fled): an N400 sentence superiority effect. Critically, participants' grammaticality judgements on the same trials showed striking capacity limitations, with dual-task deficits closely matching the predictions of a serial, all-or-none processing account. Together, these findings suggest that the N400 sentence superiority effect is fully compatible with serial word recognition, and that readers are unable to process multiple sentence positions simultaneously.

Perception of prosodic cues to correct mistakes, in listeners with normal hearing and with cochlear implants

Prosody is used to mark important information in speech yet is not an integral part of speech recognition testing among people with hearing difficulty. While a listener may correctly perceive words spoken, they may not perceive meaningful emphasis on a certain word, which could be critically important in a variety of everyday situations. This study introduces a new paradigm for assessing perception of prosodic cues in listeners with normal hearing and with cochlear implants, who have notorious difficulty perceiving pitch. Stimuli consisted of spoken sentences where one word (in various sentence positions) was emphasized in a manner that indicated that a prior statement was incorrect in a specific way. Participants used a visual analog scale to mark the timing and degree of emphasis aligned with the target words. Here, the perceptual data are linked with acoustic measures of voice pitch contour, intensity, duration, and vocal quality to characterize how contrastive stress cues are recovered by listeners with and without hearing impairment.

The time course of adaptation to distorted speech.

When confronted with unfamiliar or novel forms of speech, listeners' word recognition performance is known to improve with exposure, but data are lacking on the fine-grained time course of adaptation. The current study aims to fill this gap by investigating the time course of adaptation to several different types of distorted speech. Keyword scores as a function of sentence position in a block of 30 sentences were measured in response to eight forms of distorted speech. Listeners recognised twice as many words in the final sentence compared to the initial sentence with around half of the gain appearing in the first three sentences, followed by gradual gains over the rest of the block. Rapid adaptation was apparent for most of the eight distortion types tested with differences mainly in the gradual phase. Adaptation to sine-wave speech improved if listeners had heard other types of distortion prior to exposure, but no similar facilitation occurred for the other types of distortion. Rapid adaptation is unlikely to be due to procedural learning since listeners had been familiarised with the task and sentence format through exposure to undistorted speech. The mechanisms that underlie rapid adaptation are currently unclear.

Aspect construal in Mandarin: a usage-based constructionist perspective onLE

AbstractDespite extensive research efforts to explain the Mandarin Chinese particlele, confusion persists in the absence of a unitary theory and sufficient empirical evidence. This study provides a unitary account ofleby adopting a usage-based constructionist approach, one that liberates grammatical aspect from, and is able to accommodate, lexical aspect. We argue thatleparticipates in two distinct family resemblance constructions of aspect construal associated with two distinct sentential positions. The clause-internalleconstruction construes the closing or final boundary of an event and the clause-finalleconstruction construes the opening or initial boundary of an event. Corpus analysis showed that the two aspect constructions have distinct patterns in natural language uses that are consistent with the proposed construals. Results from elicited response data showed that native speakers paid attention to construction-level formal and semantic cues in making family resemblance judgments about tokens of the two constructions. This study has both theoretical and methodological implications for crosslinguistic research on grammatical aspect in relation to lexical aspect and for usage-based constructionist approaches to grammatical categories beyond aspect.

Web Blog Content Curation Using Fuzzy-Related Capsule Network-Based Auto Encoder

The internet content increases exponentially day-by-day leading to the pop-up of irrelevant data while searching. Thus, the vast availability of web data requires curation to enhance the results of the search in relevance to searched topics. The proposed F-CapsNet deals with the content curation of web blog data through the novel integration of fuzzy logic with a machine learning algorithm. The input content to be curated is initially pre-processed and seven major features such as sentence position, bigrams, TF-IDF, cosine similarity, sentence length, proper noun score and numeric token are extracted. Then the fuzzy rules are applied to generate the extractive summary. After the extractive curation, the output is passed to the novel capsule network based deep auto-encoder where the abstractive summary is produced. The performance measures such as precision, recall, F1-score, accuracy and specificity are computed and the results are compared with the existing state-of-the-art methods. From the simulations performed, it has been proven that the proposed method for content curation is more efficient than any other method.

ParaTTS: Learning Linguistic and Prosodic Cross-Sentence Information in Paragraph-Based TTS

Recent advancements in neural end-to-end text-to-speech (TTS) models have shown high-quality, natural synthesized speech in a conventional sentence-based TTS. However, it is still challenging to reproduce similar high quality when a whole paragraph is considered in TTS, where a large amount of contextual information needs to be considered in building a paragraph-based TTS model. To alleviate the difficulty in training, we propose to model linguistic and prosodic information by considering cross-sentence, embedded structure in training. Three sub-modules, including linguistics-aware, prosody-aware and sentence-position networks, are trained together with a modified Tacotron2. Specifically, to learn the information embedded in a paragraph and the relations among the corresponding component sentences, we utilize linguistics-aware and prosody-aware networks. The information in a paragraph is captured by encoders and the inter-sentence information in a paragraph is learned with multi-head attention mechanisms. The relative sentence position in a paragraph is explicitly exploited by a sentence-position network. Trained on a storytelling audio-book corpus (4.08 hours), recorded by a female Mandarin Chinese speaker, the proposed TTS model demonstrates that it can produce rather natural and good-quality speech paragraph-wise. The cross-sentence contextual information, such as break and prosodic variations between consecutive sentences, can be better predicted and rendered than the sentence-based model. Tested on paragraph texts, of which the lengths are similar to, longer than, or much longer than the typical paragraph length of the training data, the TTS speech produced by the new model is consistently preferred over the sentence-based model in subjective tests and confirmed in objective measures.