Abstract Semantic change proceeds via an A > A/B (> B) trajectory ( Traugott & Dasher 2002 ). While semantic extension (i.e., A > A/B) is now relatively well-understood, semantic loss (i.e., A/B (>B)) is not. This article argues that the mechanism of semantic extension is a determinant of the likelihood of loss. Specifically, loss is more likely following extension through inference-driven metonymy than following metaphorical extension. This is because metaphorically related senses tend to be contextually more distinct, which better supports polysemy sustenance. Conversely, because inference-driven extensions typically rely on bridging contexts, related senses often occur in overlapping contexts. This makes loss of the source sense more likely following extension. Evidence to support this argument is drawn from different corpus-based case studies into the semantic development of the English adjectives bright , dumb , narrow , and strange , and Dutch eng ‘frightening’.
Semantic change proceeds via an A > A/B (> B) trajectory (Traugott & Dasher 2002). While semantic extension (i.e., A > A/B) is now relatively well-understood, semantic loss (i.e., A/B (>B)) is not. This article argues that the mechanism of semantic extension is a determinant of the likelihood of loss. Specifically, loss is more likely following extension through inference-driven metonymy than following metaphorical extension. This is because metaphorically related senses tend to be contextually more distinct, which better supports polysemy sustenance. Conversely, because inference-driven extensions typically rely on bridging contexts, related senses often occur in overlapping contexts. This makes loss of the source sense more likely following extension. Evidence to support this argument is drawn from different corpus-based case studies into the semantic development of the English adjectives bright, dumb, narrow, and strange, and Dutch eng 'frightening'.
The negation marker in Hewram & icirc;, m & eacute;-, originated as a marker of imperfective aspect. A series of regular sound changes setting Hewram & icirc; apart from its closely related sisters caused the demorphologization of the original prefix. The subsequent morphologization of stress accent caused the imperfective prefix, fully retained only in the context of negation, to be reinterpreted as a marker of negation. In this article, we explore these developments in comparison with closely related varieties that have preserved every stage in the process.
This paper explores the relationship between ambiguity and reanalysis in syntactic change, with a particular focus on the interplay between structural and semantic ambiguity. The diachronic corpus study examines two cases of syntactic reanalysis. The first involves the development of the German directional adverb hinaus 'out, outward' into a verbal particle, while the second traces the evolution of the same adverb into an adposition, specifically as part of the circumposition [& uuml;ber NPACC hinaus] 'beyond, besides'. The paper argues that structural ambiguity may, but does not necessarily, align with semantic ambiguity, which has implications for potential diachronic shifts. Structural ambiguity alone does not necessarily lead to reanalysis and should not be regarded merely as a transitional stage in gradual diachronic change; rather, it may represent a relatively stable state. In this sense, structural ambiguity can serve as a potential starting point for reanalysis, though it does not inherently drive the process.
This paper contrasts typological and diachronic evidence on argument marking with theoretical proposals made in psycholinguistic research about the ways various types of audience design strategies affect production, with a focus on ambiguity avoidance in argument roles. The aim is to find convergent evidence from psycholinguistics as well as from typological and diachronic research. This evidence suggests that argument marking in functional constructions is shaped by generic audience design effects, while cross-linguistic and diachronic support for utterance-specific audience design is very scarce. Diachronically utterance-specific audience design tends to be abandoned in favor of functionally related generic audience design strategies. The paper reconciles earlier claims about ambiguity avoidance affecting argument marking and the view of ambiguity avoidance as having no effect on argument marking. It is suggested that ambiguity avoidance is a strong pressure in argument marking. I show that a high ambiguity potential of role identification of various constructions correlates positively with the degree of non-differential argument marking; and, vice versa, differential marking is more likely in constructions that provide reliable cues for role identification.
This paper hypothesises that there is a universal cognitively and semiotically grounded preference for word-form shapes that signal the morphotactic structure of the word forms they represent. By exploring three historical changes in English, we demonstrate how such a bias might plausibly constrain the actuation and implementation of sound changes. The changes we discuss are the emergence of /z/ as the underlying form in English plural, genitive, and 3rd person present suffixes; Middle English Open Syllable Lengthening; and the emergence and stabilization of irregular past tense and participle forms ending on sonorant+/t/ clusters instead of the expected ones resulting from regular /d/-suffixation. The present study shows how a preference for morphotactically indicative (and unambiguous) word-form shapes can be detected and measured and suggests that the existence of such a preference appears to be highly plausible. We conclude by pointing to more ways in which this hypothesis can be tested and potentially corroborated.
In this article we analyze the translation strategies used by medieval and early modern Spanish translators when rendering the infinitive absolute (IA) in the Hebrew Bible. In a literal translation, this structure, which had no immediate equivalent in the Spanish of the time, would be rendered with the infinitive followed by a conjugated form of the same verb, as in Spanish llorar llorar & aacute;s 'to cry you will cry', meaning either 'for sure you will cry' or 'you will cry a lot'. The exhaustive analysis of the IA construction in a parallel corpus reveals the full repertoire of structures - from literal calque to creative paraphrases - used to convey the pragmatic values expressed by the Hebrew IA. In addition to characterizing each of the twelve Bible translations in the corpus in terms of literalness or creativity, we also address the question of the permeability of the calque into non-translated texts and even oral registers. Based on structural, chronological, and socio-historical factors, we conclude that this type of fronted infinitive did not evolve into a productive pattern, and that present-day Spanish topicalized infinitives (which superficially resemble the IA calque) are an independent development. Instead, the translation of the Hebrew IA into Spanish crystallized over time into tautological gerunds rather than infinitives.
The paper deals with how syntax of coordination interacts with the second position constraint. It is claimed, on the basis of Hittite data, that the cliticization scope of the enclitic coordinator can be limited to its complement with the exclusion of its (inner) specifier. This aspect of the interaction of coordination syntax and second position constraint has not been observed previously. Besides, Hittite broadens the dataset of the few languages with the "too-low" position of coordinators. Moreover, it is shown that the same type of cliticization domain is available for another extraCP clitic in Hittite - the direct speech particle. This is in stark contrast to CP bound clitics that take both the complement and the specifier of the head they target (C0) as their cliticization domain. Thus, we see a non-trivial correlation of the position the clitic is generated in (or targets) and its distribution. This compares to a similar (yet different in detail) correlation in Ancient Greek between sentential and clausal clitics.
We argue that we should maintain the distinction between ambiguity and vagueness in grammatical change, especially in usage-based linguistics, as it offers a more fine-grained perspective on types of context, change, and linguistic structure. Drawing on previous research, we present two case studies: the semantic history of be going to, and the morphosyntactic history of Mandarin modals and conditionals. These studies show that the initial relation involved in grammatical change is one of vagueness. We also contrast our usage-based perspective with that of frameworks where the clause is hierarchically organised such that the initial relation is more likely one of ambiguity.
The aim of this paper is to advance our understanding of the diachronic development of Basque prosody and related segmental processes. The discovery of regular correspondences between the distribution of aspiration in northeastern varieties and accentual patterns in the dialect of Goizueta (in western Navarre) allows us to establish correspondences between accentual classes for the entire Basque territory for the first time. As a result, we are able to make progress in the reconstruction of the Common Basque prosodic system by using the comparative method. I show that the Goizueta system shares a class of accentual exceptions with western varieties of Basque and also shares a different class of marked words with an earlier stage of the eastern accentual system reflected in the distribution of aspirated and unaspirated voiceless stops, confirming a hypothesis first formulated by L. Michelena [K. Mitxelena] over sixty years ago. I also consider specific hypotheses regarding the chronology of changes in accentuation and several segmental changes that interact with accent rules. As a contribution to the larger field of historical linguistics, this paper offers a demonstration of the importance of establishing regular correspondences (in this case involving aspirated consonants and accentual patterns) in phonological reconstruction.
It has often been argued that ambiguity plays a role in language change, but the cognitive mechanisms and constraints that might facilitate ambiguity-related change are as yet poorly understood. In this article I discuss syntactic ambiguity and its potential role in grammatical change in the contexts of real-time language comprehension and production. In order to comprehend in real-time, readers or listeners usually process syntactically ambiguous strings of words as if they were unambiguous. Syntactic change may come about when innovative syntactic analyses are computed during comprehension and later spread and are incorporated into the grammar. Although ambiguity will not normally be a problem for language producers, avoiding ambiguity in language production may be motivated by audience design considerations. Experimental evidence for speakers choosing to avoid syntactic ambiguity is mixed, however. Psycholinguistic models that propose a tight link between real-time production and comprehension offer an integrative perspective on ambiguity avoidance and its possible role in language change.
Analyzing quantitative corpus data from the 13th to the 19th centuries, this study applies a usage-based construction grammar approach and an onomasiological approach to the examination of change in argument structure in Spanish verbs of "liking", focusing on the verb gustar 'to like.' This verb went from having a nominative experiencer to having a dative one. I highlight how the semantic properties of the stimuli of liking verbs helped determine the linguistic encoding of the arguments of liking constructions. Furthermore, I argue that a key cause in the change in argument structure of gustar was the analogical extension of the dative-experiencer pattern via a multiplicity of source constructions. This study is the first to consider (a) the verb forms pagarse, ser gusto, and dar gusto and (b) onomasiological profiles for the concept of liking, regarding the change in argument structure of gustar. Results shed light on cross-linguistic variation in argument structures and extend scholarship on the issue of word order in psych-verb constructions.
In this contribution we examine current views of the relationship between innovation on the one hand, and diffusion of a change through a community on the other. We also note in the recent literature much discussion of the role of the listener vs. the role of the speaker in innovation and spread. This paper re-examines the dichotomies just mentioned - innovation vs. spread and speaker vs. listener - and finds them unjustified. We consider two types of change - sound change and grammaticalization - which are similar in that they both show directional patterns across time and across languages, and they are both considered to be gradual. Based on empirical evidence from a wide range of studies from interactional linguistics, sociolinguistics, and diachronic linguistics, we argue that innovation occurs jointly between speaker and listener in the same speech community and spreads gradually on all dimensions.
Albanian has a two-term demonstrative system, with proximal masculine ky and feminine kjo 'this' contrasting with distal masculine ai and feminine ajo 'that'. The distal pronouns double as personal pronouns of the third person. Both demonstratives arose from the composition of two deictic elements. The second element (with nominative singular masculine -i/-y, feminine -jo, and neuter -ta) is the same for both pronouns and probably continues the PIE demonstrative *se/o-, *te/o-, although some etymological details are not clear yet. The origin of the first elements k(& euml;)- resp. a- is also disputed. In this article I will first briefly sketch the morphology of the two demonstrative pronouns in Old Albanian and then proceed to a re-evaluation of the etymological scenarios that may explain the rise and grammaticalization of ky and ai.
This paper focuses on the origin of a polyfunctional morpheme ti(42) in Tunxi Hui Chinese, a little-studied Hui Sinitic language with about 140,000 speakers in Tunxi, Anhui Province of China. This versatile morpheme represents a radical syncretism of nine disparate functions, including that between the lexical verb give and the allative, locative, and temporal marking, which is seldom reported in the literature. By first-hand synchronic data and the historical comparative method, we propose that ti(42) originated as a send-type verb *diai6 'pass' in Middle Chinese, which has extended to a general ditransitive verb give in modern Tunxi Hui. During this process, it has developed into two separate grammaticalization clines, one from 'pass' to the allative, locative, and temporal markers, and the other from give to recipient, beneficiary, purpose, permissive, and passive markers. The polysemy of ti(42) sheds light on the complex processes of semantic shift from pass to give, and language-internal and contact-induced polygrammaticalization.
This paper elucidates the role of internal and external factors in the semantic evolution of Basque zu from ‘you. pl ’ to exclusively ‘you. sg ’ and in the development of Basque zuek ‘you. pl ’. Transparently, zuek involves suffixation of the demonstrative hek ‘those’ to zu . The neighboring Romance languages exhibit a parallel evolution that involves the adjectival collocation vōs alteros ‘you others’, but other similar constructions are recorded. This paper defends the premise that Basque zuek emerged through a pattern replication process in which the pivotal feature was the morphological expression of number. We argue that the grammaticalization of zuek as ‘you. pl ’ started in eastern areas due to contact with Occitan, a language in which grammaticalized ‘you. pl ’ forms appear earlier than in Castilian Spanish. We also address an innovation exclusive to Basque: a series of second plural verbal forms which exhibit a geographically spread pattern within the Basque Country similar to that of the Romance innovative pronouns in the Iberian Peninsula.
This paper describes and analyzes the renewal of the demonstrative system in Hungarian, concentrating on the competing strategies of the Middle Hungarian period. After the definite article had grammaticalized in Hungarian, the demonstrative system renewed via two different strategies: reinforcement strategy and determiner doubling. The old and new strategies are introduced in the first, more descriptive part of the study, which is followed by an empirical investigation based on digitized historical corpora. Numerical evidence can show the measure of their spread as well as the change in their ratio. A coherent model of the syntactic change is also provided through formal representations of each structure under discussion. Finally, exploring the distributional properties of individual patterns in Middle Hungarian allows for the formation of a hypothesis about the upcoming outcome of their competition.
Demonstratives are generally assumed to be universal, but their diachronic pathways of development are notoriously obscure. If they were completely resistant to change through time, they should match across related languages, apart from regular processes of sound change. Languages of the Iroquoian family all contain demonstratives, as would be expected, but these are not all cognate. Closer examination reveals that they have developed from slightly different sources, but along parallel pathways. Their development has not stopped there, however. Some have evolved further into articles but to varying extents in the different languages. The articles are evolving further into nominalizers, forming both participant and event nominalizations. These nominalizations not only serve as referring expressions but are also coming to function much like dependent clauses in complex sentences.
Medieval Insular Celtic languages possess definite articles, unlike their ancient Continental counterparts, which lack them. Since the 19th century, scholars have hypothesized that these articles in Insular Celtic languages evolved from earlier demonstratives. This grammaticalization process likely followed the path from demonstrative to definite article. However, systematic surveys of ancient Celtic data to understand this process have been lacking. This study aims to fill that gap by analyzing noun phrases in ancient Celtic languages featuring demonstratives to identify article-like uses and comparing them with medieval Celtic articles, especially in Old Irish. Findings reveal that ancient Celtic demonstratives often function similarly to articles in medieval languages, acting as pronouns and determiners that agree in gender, number, and case with nouns. In Celtiberian, so - and sto - demonstratives appear in noun phrases with deictic and anaphoric referents. In Gaulish, so - demonstratives are used for deictic referents, while sindo - is used for anaphoric ones. Demonstratives are not used with semantically defined referents. The grammaticalization process aligns with Dryer’s (2014) hierarchy and, despite limited ancient documentation, confirms the evolution from demonstrative to definite article.
