This paper aims to answer why the Uralic languages use, or used until intensive contacts with Indo-European languages, only non-finite subordination. It argues against regarding the evolution of finite subordination language development, showing that languages with non-finite subordination and parataxis have the same expressive power as languages with finite subordination. It claims that non-finite subordination is a concomitant of SOV word order, and the growing proportion of finite subordination in the Uralic languages from east to west, and in the history of Hungarian is a consequence of the loosening of the SOV order and the emergence of SVO. The paper examines two hypotheses about the correlations between SOV and non-finite subordination, and SVO and finite subordination, the Final-Over-Final Condition of Biberauer, Holmberg & Roberts (2014, etc.), a formal principle constraining clausal architecture, and the Minimize Domains Principle of Hawkins (2004, etc.), a functional principle of processing efficiency. The two theories make largely overlapping correct predictions for the Uralic languages, which suggests that the Final-Over-Final Condition may be the syntacticization of the condition that ensures processing efficiency in SOV and SVO languages.
{"title":"The (non-)finiteness of subordination correlates with basic word order: Evidence from Uralic","authors":"K. Kiss","doi":"10.1556/2062.2023.00647","DOIUrl":"https://doi.org/10.1556/2062.2023.00647","url":null,"abstract":"This paper aims to answer why the Uralic languages use, or used until intensive contacts with Indo-European languages, only non-finite subordination. It argues against regarding the evolution of finite subordination language development, showing that languages with non-finite subordination and parataxis have the same expressive power as languages with finite subordination. It claims that non-finite subordination is a concomitant of SOV word order, and the growing proportion of finite subordination in the Uralic languages from east to west, and in the history of Hungarian is a consequence of the loosening of the SOV order and the emergence of SVO. The paper examines two hypotheses about the correlations between SOV and non-finite subordination, and SVO and finite subordination, the Final-Over-Final Condition of Biberauer, Holmberg & Roberts (2014, etc.), a formal principle constraining clausal architecture, and the Minimize Domains Principle of Hawkins (2004, etc.), a functional principle of processing efficiency. The two theories make largely overlapping correct predictions for the Uralic languages, which suggests that the Final-Over-Final Condition may be the syntacticization of the condition that ensures processing efficiency in SOV and SVO languages.","PeriodicalId":37594,"journal":{"name":"Acta Linguistica Academica","volume":null,"pages":null},"PeriodicalIF":0.5,"publicationDate":"2023-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44974058","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"文学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Diachronic changes in phrase or clause structure are vectored rather than oscillating. A century ago, E. Sapir identified a drift towards fixed word order and another one towards the invariant word (including the levelling of the forms for subject and object marking). What is still missing is a theory that predicts such drifts. As will be argued, the theory that explains Sapir's observations and, in passing, makes the concept of Universal Grammar dispensable is the theory that grammars are targets and products of cognitive evolution. Sapir's drifts are shifts from systems based primarily on the consciously accessible declarative network to systems based on the consciously inaccessible procedural network. This also explains why the [S[VO]] clause-structure is a point of no return and why languages do not change in the reverse direction, starting from a grammar like English and eventually moving to a grammar like Sanskrit.
{"title":"SVO – Attractor in the declarative-to-procedural shift in grammar evolution","authors":"H. Haider","doi":"10.1556/2062.2023.00642","DOIUrl":"https://doi.org/10.1556/2062.2023.00642","url":null,"abstract":"Diachronic changes in phrase or clause structure are vectored rather than oscillating. A century ago, E. Sapir identified a drift towards fixed word order and another one towards the invariant word (including the levelling of the forms for subject and object marking). What is still missing is a theory that predicts such drifts. As will be argued, the theory that explains Sapir's observations and, in passing, makes the concept of Universal Grammar dispensable is the theory that grammars are targets and products of cognitive evolution. Sapir's drifts are shifts from systems based primarily on the consciously accessible declarative network to systems based on the consciously inaccessible procedural network. This also explains why the [S[VO]] clause-structure is a point of no return and why languages do not change in the reverse direction, starting from a grammar like English and eventually moving to a grammar like Sanskrit.","PeriodicalId":37594,"journal":{"name":"Acta Linguistica Academica","volume":null,"pages":null},"PeriodicalIF":0.5,"publicationDate":"2023-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44330595","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"文学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper, I present a novel corpus investigation of quantified and negated objects in the Middle English and Early Modern English period, which is embedded within the wider language change scenario from linear OV to linear VO in the history of English. It will be shown that evidence for preverbal positioning of such objects is mostly limited to translated texts in Middle English in the PPCME2 corpus, and that by late Middle English, most of the hits consist of negated elements, as shown in the PCEEC corpus, which consists of native texts. The different constraints governing spell out of positive objects in Old English and Middle English are discussed and compared to the licensing of negated and quantified objects. The data provided in this paper constitute further evidence for Ingham's (2000, 2002, 2007) analysis of preposed negated objects in late ME and their correlation with the Negative Cycle, and complement previous investigations on negated and quantified objects in Middle English and Early Modern English.
{"title":"Some notes on negated and quantified objects in Middle English and Early Modern English","authors":"Chiara De Bastiani","doi":"10.1556/2062.2023.00650","DOIUrl":"https://doi.org/10.1556/2062.2023.00650","url":null,"abstract":"In this paper, I present a novel corpus investigation of quantified and negated objects in the Middle English and Early Modern English period, which is embedded within the wider language change scenario from linear OV to linear VO in the history of English. It will be shown that evidence for preverbal positioning of such objects is mostly limited to translated texts in Middle English in the PPCME2 corpus, and that by late Middle English, most of the hits consist of negated elements, as shown in the PCEEC corpus, which consists of native texts. The different constraints governing spell out of positive objects in Old English and Middle English are discussed and compared to the licensing of negated and quantified objects. The data provided in this paper constitute further evidence for Ingham's (2000, 2002, 2007) analysis of preposed negated objects in late ME and their correlation with the Negative Cycle, and complement previous investigations on negated and quantified objects in Middle English and Early Modern English.","PeriodicalId":37594,"journal":{"name":"Acta Linguistica Academica","volume":null,"pages":null},"PeriodicalIF":0.5,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43923629","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"文学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Manner/Result Complementarity (Rappaport Hovav & Levin 2010) has been argued to have consequences for argument realization: only manner verbs permit object deletion and non-selected objects. In contrast, result verbs always co-appear with their object, because they are required to express the undergoer of the change that they entail. We discuss new data involving result verbs in constructions where the undergoer of the change encoded by the result verb is not realized as the object of the predicate. We argue these data display result verbs whose root is integrated into the argument structure of the predicate in such a way that it is interpreted as specifying a co-event of the main event denoted by the predicate, whereby the result entailed by the root is not necessarily intended to hold of the direct object. This follows if verb roots do not come with a syntactically relevant specification for manner or result from the lexicon, but acquire it on the basis of their association with the syntactic structure.
{"title":"On the argument structure realization of result verbs: A syntactic approach","authors":"Josep Ausensi, Alessandro Bigolin","doi":"10.1556/2062.2023.00567","DOIUrl":"https://doi.org/10.1556/2062.2023.00567","url":null,"abstract":"Manner/Result Complementarity (Rappaport Hovav & Levin 2010) has been argued to have consequences for argument realization: only manner verbs permit object deletion and non-selected objects. In contrast, result verbs always co-appear with their object, because they are required to express the undergoer of the change that they entail. We discuss new data involving result verbs in constructions where the undergoer of the change encoded by the result verb is not realized as the object of the predicate. We argue these data display result verbs whose root is integrated into the argument structure of the predicate in such a way that it is interpreted as specifying a co-event of the main event denoted by the predicate, whereby the result entailed by the root is not necessarily intended to hold of the direct object. This follows if verb roots do not come with a syntactically relevant specification for manner or result from the lexicon, but acquire it on the basis of their association with the syntactic structure.","PeriodicalId":37594,"journal":{"name":"Acta Linguistica Academica","volume":null,"pages":null},"PeriodicalIF":0.5,"publicationDate":"2023-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"46618459","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"文学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In the evidential system of Uzbek, the speaker has different grammatical options in marking the source of information, such as -ibdi, ekan, emish, etc., although it is not compulsory to mark this category in the utterance. In addition to these established markers, new markers have developed into evidentials, and they encode specific sub-categories of evidentiality. In this study, after a brief overview of grammatical markers of evidentiality in Uzbek, the marker chog‘i is examined with a syntactic and semantic approach based on a corpus of selected texts. Its development into an inferential marker is evaluated with special attention to sources of evidentials.
{"title":"The development and functions of the inferential marker chog‘i in Uzbek","authors":"Melike Üzüm","doi":"10.1556/2062.2023.00568","DOIUrl":"https://doi.org/10.1556/2062.2023.00568","url":null,"abstract":"In the evidential system of Uzbek, the speaker has different grammatical options in marking the source of information, such as -ibdi, ekan, emish, etc., although it is not compulsory to mark this category in the utterance. In addition to these established markers, new markers have developed into evidentials, and they encode specific sub-categories of evidentiality. In this study, after a brief overview of grammatical markers of evidentiality in Uzbek, the marker chog‘i is examined with a syntactic and semantic approach based on a corpus of selected texts. Its development into an inferential marker is evaluated with special attention to sources of evidentials.","PeriodicalId":37594,"journal":{"name":"Acta Linguistica Academica","volume":null,"pages":null},"PeriodicalIF":0.5,"publicationDate":"2023-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47388637","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"文学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The Word-in-Context corpus, which forms part of the SuperGLUE benchmark dataset, focuses on a specific sense disambiguation task: it has to be decided whether two occurrences of a given target word in two different contexts convey the same meaning or not. Unfortunately, the WiC database exhibits a relatively low consistency in terms of inter-annotator agreement, which implies that the meaning discrimination task is not well defined even for humans. The present paper aims at tackling this problem through anchoring semantic information to observable surface data. For doing so, we have experimented with a graph-based distributional approach, where both sparse and dense adjectival vector representations served as input. According to our expectations the algorithm is able to anchor the semantic information to contextual data, and therefore it is able to provide clear and explicit criteria as to when the same meaning should be assigned to the occurrences. Moreover, since this method does not rely on any external knowledge base, it should be suitable for any low- or medium-resourced language.
word -in- context语料库是SuperGLUE基准数据集的一部分，它专注于一个特定的意义消歧义任务:它必须确定给定目标单词在两个不同的上下文中的两次出现是否传达相同的含义。不幸的是，WiC数据库在注释者之间的一致性方面表现出相对较低的一致性，这意味着即使对于人类来说，意义区分任务也没有很好地定义。本文旨在通过将语义信息锚定到可观测表面数据来解决这一问题。为此，我们尝试了一种基于图的分布方法，其中稀疏和密集的形容词向量表示都作为输入。根据我们的期望，该算法能够将语义信息锚定到上下文数据，因此它能够提供清晰明确的标准，以确定何时应该为出现的事件分配相同的含义。此外，由于这种方法不依赖于任何外部知识库，因此它应该适用于任何低资源或中等资源的语言。
{"title":"A proof-of-concept meaning discrimination experiment to compile a word-in-context dataset for adjectives – A graph-based distributional approach","authors":"Enikő Héja, Noémi Ligeti-Nagy","doi":"10.1556/2062.2022.00579","DOIUrl":"https://doi.org/10.1556/2062.2022.00579","url":null,"abstract":"The Word-in-Context corpus, which forms part of the SuperGLUE benchmark dataset, focuses on a specific sense disambiguation task: it has to be decided whether two occurrences of a given target word in two different contexts convey the same meaning or not. Unfortunately, the WiC database exhibits a relatively low consistency in terms of inter-annotator agreement, which implies that the meaning discrimination task is not well defined even for humans. The present paper aims at tackling this problem through anchoring semantic information to observable surface data. For doing so, we have experimented with a graph-based distributional approach, where both sparse and dense adjectival vector representations served as input. According to our expectations the algorithm is able to anchor the semantic information to contextual data, and therefore it is able to provide clear and explicit criteria as to when the same meaning should be assigned to the occurrences. Moreover, since this method does not rely on any external knowledge base, it should be suitable for any low- or medium-resourced language.","PeriodicalId":37594,"journal":{"name":"Acta Linguistica Academica","volume":null,"pages":null},"PeriodicalIF":0.5,"publicationDate":"2022-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43712821","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"文学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Transformer-based NLP models have achieved state-of-the-art results in many NLP tasks including text classification and text generation. However, the layers of these models do not output any explicit representations for texts units larger than tokens (e.g. sentences), although such representations are required to perform text classification. Sentence encodings are usually obtained by applying a pooling technique during fine-tuning on a specific task. In this paper, a new sentence encoder is introduced. Relying on an autoencoder architecture, it was trained to learn sentence representations from the very beginning of its training. The model was trained on bilingual data with variational Bayesian inference. Sentence representations were evaluated in downstream and linguistic probing tasks. Although the newly introduced encoder generally performs worse than well-known Transformer-based encoders, the experiments show that it was able to learn to incorporate linguistic information in the sentence representations.
{"title":"BiVaSE: A bilingual variational sentence encoder with randomly initialized Transformer layers","authors":"Bence Nyéki","doi":"10.1556/2062.2022.00584","DOIUrl":"https://doi.org/10.1556/2062.2022.00584","url":null,"abstract":"Transformer-based NLP models have achieved state-of-the-art results in many NLP tasks including text classification and text generation. However, the layers of these models do not output any explicit representations for texts units larger than tokens (e.g. sentences), although such representations are required to perform text classification. Sentence encodings are usually obtained by applying a pooling technique during fine-tuning on a specific task. In this paper, a new sentence encoder is introduced. Relying on an autoencoder architecture, it was trained to learn sentence representations from the very beginning of its training. The model was trained on bilingual data with variational Bayesian inference. Sentence representations were evaluated in downstream and linguistic probing tasks. Although the newly introduced encoder generally performs worse than well-known Transformer-based encoders, the experiments show that it was able to learn to incorporate linguistic information in the sentence representations.","PeriodicalId":37594,"journal":{"name":"Acta Linguistica Academica","volume":null,"pages":null},"PeriodicalIF":0.5,"publicationDate":"2022-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"46375866","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"文学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In the scope of this research, we aim to give an overview of the currently existing solutions for machine translation and we assess their performance on the English-Hungarian language pair. Hungarian is considered to be a challenging language for machine translation because it has a highly different grammatical structure and word ordering compared to English. We probed various machine translation systems from both academic and industrial applications. One key highlight of our work is that our models (Marian NMT, BART) performed significantly better than the solutions offered by most of the market-leader multinational companies. Finally, we fine-tuned different pre-finetuned models (mT5, mBART, M2M100) for English-Hungarian translation, which achieved state-of-the-art results in our test corpora.
{"title":"Neural machine translation for Hungarian","authors":"L. Laki, Zijian Győző Yang","doi":"10.1556/2062.2022.00576","DOIUrl":"https://doi.org/10.1556/2062.2022.00576","url":null,"abstract":"In the scope of this research, we aim to give an overview of the currently existing solutions for machine translation and we assess their performance on the English-Hungarian language pair. Hungarian is considered to be a challenging language for machine translation because it has a highly different grammatical structure and word ordering compared to English. We probed various machine translation systems from both academic and industrial applications. One key highlight of our work is that our models (Marian NMT, BART) performed significantly better than the solutions offered by most of the market-leader multinational companies. Finally, we fine-tuned different pre-finetuned models (mT5, mBART, M2M100) for English-Hungarian translation, which achieved state-of-the-art results in our test corpora.","PeriodicalId":37594,"journal":{"name":"Acta Linguistica Academica","volume":null,"pages":null},"PeriodicalIF":0.5,"publicationDate":"2022-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42316871","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"文学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：