The place of grammaticalization theory vis-à-vis alternative frameworks, and especially that of Diachronic Construction Grammar, has more recently been a subject of debate. One of the topics figuring in this debate concerns the role played by explanation in accounting for grammatical change. It turns out that the two frameworks rely on contrasting views on what constitutes an explanation. Parameters favored in Diachronic Construction Grammar are most of all analogy, reanalysis, and frequency of use. Students of grammaticalization, in contrast, tend to prefer to search for explanations by looking at the functional motivations underlying discourse processing. The goal of the paper is to compare these different positions and to attempt a critical analysis of them with a view to finding out whether the two can be reconciled with one another. Unlike most of the work that has so far been carried out on both grammaticalization and constructional change, the paper has a focus on the initial phase of changes in discourse processing.

This paper investigates the evolution of the present perfect and have to in English to assess whether these forms may be considered emerging inferentials. Combining quantitative and qualitative analyses in diachrony and synchrony, it argues that the two forms may show incipient signs of the grammaticalization of inferential evidentiality. The present perfect is already fully grammaticalized in its morphology, and highly frequent. Inferential evidentiality is not part of its meaning, but a slightly rising implicature. The semi-auxiliary have to is less grammaticalized morphologically, but possesses a distinct function that is traditionally called ‘epistemic’, but is rather ‘logical inferential’. This function is still infrequent, but has been rising steadily in the past decades. Whether these changes will lead to a fully-developed evidential paradigm is unpredictable, but this study confirms that evidentiality is a necessary notion to describe English grammar and its evolution accurately.

Languages with multi-term evidential systems vary in how people speak about themselves and their own actions. The aim of this paper is to understand the differences between evidentials used with the first person (oneself) and with other persons. We focus on two groups of languages with large evidential systems - Tariana and Tukano from the Vaupes River Basin linguistic area in Amazonia, and Common Tibetan and Central Ladakhi, two Tibetic languages of the Tibetan Plateau and the Himalayas. We analyse similarities and contrasts in the expression of information source and access to information through grammatical evidentials and egophoric systems in these two groups of languages. Our analysis reveals their intricacies and the special status of speech act participants.

In some work based on the framework of Diachronic Construction Grammar, a branch of Construction Grammar (e.g. Croft 2001 ; Goldberg 2006 ), it has been argued that phenomena traditionally classified as instances of grammaticalization should not be considered as belonging to a distinct field of its own within historical linguistics. Rather, it is proposed that such phenomena are more appropriately taken care of within Diachronic Construction Grammar. The present paper takes issue with this stance, arguing that grammatical change as it is the subject of grammaticalization theory has a unique place within the field of diachronic linguistics. The paper focuses on a study by Gildea & Barðdal (2023) , which represents the most pronounced position on this issue within Diachronic Construction Grammar.

This article re-examines the literature on the evidential uses of French tenses, and evaluates what distinguishes French from languages that are said to possess fully grammaticalized evidential systems. Based on corpus analyses, semantic testing, and crosslinguistic comparisons, this study argues that the French pass & eacute; compos & eacute; and imparfait do not carry any inherent evidential meaning, unlike the futur and conditionnel. The evidential interpretations of the former two tenses are simply conveyed by the context, while those of the latter two are indeed due to their intrinsic semantic make-up. We conclude that although French encodes evidentiality with verbal inflections only infrequently, it is no different from languages usually cited to illustrate advanced evidential paradigms from a formal and semantic standpoint. Cet article a pour but de r & eacute;& eacute;tudier la litt & eacute;rature sur les usages & eacute;videntiels des temps en fran & ccedil;ais et d'examiner ce qui distingue le fran & ccedil;ais des langues qui sont g & eacute;n & eacute;ralement consid & eacute;r & eacute;es comme dot & eacute;es de syst & egrave;mes & eacute;videntiels enti & egrave;rement grammaticalis & eacute;s. Sur la base d'analyses de corpus, de tests s & eacute;mantiques et de comparaisons interlinguistiques, nous soutenons que le pass & eacute; compos & eacute; et l'imparfait fran & ccedil;ais n'ont pas de sens & eacute;videntiel inh & eacute;rent, contrairement au futur et au conditionnel. Les interpr & eacute;tations & eacute;videntielles des deux premiers temps sont simplement induites par le contexte, tandis que celles des deux derniers sont en effet dues & agrave; leur composition s & eacute;mantique intrins & egrave;que. Nous concluons que, bien que le fran & ccedil;ais n'encode que rarement l'& eacute;videntialit & eacute; au moyen de flexions verbales, il n'est pas diff & eacute;rent, d'un point de vue formel et s & eacute;mantique, des langues g & eacute;n & eacute;ralement cit & eacute;es pour illustrer les paradigmes & eacute;videntiels

This paper introduces the main notions that are addressed in this special issue, namely evidentiality, modality, and grammaticalization. It defines each notion and briefly synthesizes the literature. It also presents some of the controversies which surround the ideas that prevail in these research fields. Crosslinguistic examples illustrate the main evidential and modal categories, and clarify why the two domains are both distinct and related. The paper then sketches the main pathways of grammaticalization of modal and evidential markers as they have been documented in typological work. Finally, it introduces the contributions to this special issue, highlights the new insights, and discusses what remains to be investigated on the links between evidentiality, modality, and grammaticalization.

This paper seeks to answer three questions: (1) What is the difference between grammatical and lexical indications of information source? (2) What qualifies an element for grammaticalization as an evidential? (3) How can we identify grammatical evidentials and instances of evidential grammaticalization? The answers proposed are as follows: (1) The difference between grammatical and lexical indications of information source is a difference between indications conventionalized as discourse secondary and indications conventionalized as potentially discourse primary. (2) A candidate for grammaticalization as an evidential must (i) have propositional scope, (ii) belong in the conceptual domain of information source, (iii) be frequent enough to pass the threshold for conventionalization, and (iv) be discourse secondary, but not by convention. (3) Grammatical evidentials and instances of evidential grammaticalization can be identified based on focusablity, addressability and modifiability.

The paper examines to what extent expressing information sources with a fully grammaticalized evidential system or with lexical means may have a different pragmatic impact in conversation. A contrastive corpus of Tibetan (TSC) and English (CSC/LAC) allows me to investigate several dimensions: frequency, semantic schematicity, optionality, economy, and information hierarchy. The data from the contrastive corpus indicate that Tibetan evidentials are on average 6.7 times more frequent than English evidentials. Tibetan evidentials are also functionally more schematic, since English evidentials usually contain more semantic features. This compels English speakers to be more specific when referring to their information sources. Grammatical evidential systems may be less optional than lexical systems, but the degree of optionality differs depending on the linguistic level. The morpho-syntactic obligatoriness of many Tibetan evidentials may indeed be compensated by the availability of evidentially neutral forms in the same paradigm. However, evidentiality seems to be pragmatically less optional in Tibetan than in English. Tibetan evidentials are also more economical on a morpho-phonetic, syntactic, and cognitive level. Finally, all these parameters influence the informational status of evidentials. Because they are more frequent, schematic, obligatory, and economical, Tibetan evidentials appear more backgrounded.

This article argues that the data on English evidential markers indicate that evidentiality is not a grammatical category which is only applicable to some languages, but can be considered a universal semantic function that tends to trigger the grammaticalization of lexical forms. This study investigates five criteria that can help us locate linguistic forms on the lexicon–grammar continuum: reduction, desemanticization, backgrounding, decategorialization and paradigmatization. Although it has not reached the full maturity of so-called ‘evidential languages’, English provides evidence of the grammaticalization of evidentiality. Evidentiality is thus a relevant notion for English which substantially impacts the organization of the lexicon of the language, and the evolution of its grammar.

The Polyfunctionality of 'give'The verb 'give' may be one of the most frequent and basic verbs in the world's languages, but it captures a complex situational frame (Newman, 1996).Its polyfunctional quality seems to be universally attested, as the verb root is frequently redeployed into other syntactic categories, and commonly appears 1 We would like to thank Amanda Edmonds, Xénia De Heering, Françoise Robin, Camille Simon, and Debra Ziegeler for their valuable comments, which allowed us to deepen our understanding of this paper's topic.

AbstractEvidentiality, i.e. the linguistic encoding of the mode of access to information (direct perception, inference, hearsay), despite not being fully grammaticalized in English and French, is expressed through a variety of means. This paper seeks to determine how a relatively non-salient concept in the source and target languages can be acquired by L2 learners. Using an oral elicited narrative task, we determine what markers of direct perception and inference are commonly used by native speakers of French (n = 10) and English (n = 10) and L2 learners of those two languages (at three levels of proficiency,n = 10 per group), and at which level they emerge. Our results point to a much more frequent use of inferential markers than direct perception markers, to slightly different patterns of evidential marking in French and in English, and to a late emergence of evidential markers in the speech of learners, who display sensitivity to their discursive functions, with types and tokens increasing as a function of proficiency level.

An abstract is not available for this content so a preview has been provided. Please use the Get access link above for information on how to access this content.

The history of either as a clause-final, right-periphery marker has seen little intensive research, apart from a few isolated studies such as Rullmann (2002) and Gast (2013). This is surprising, given the recent interest in parenthetical discourse items and the controversies surrounding their development (grammaticalization vs. pragmaticalization, and other debates). In the present study, it is first questioned whether right-periphery either (RP-either) could be categorized as a bona fide example of a discourse marker, and second, how a hypothesis emerged that 18th and 19th century prescriptivism motivated its sudden shift to become a post-negation, clause-final item, replacing the now non-standard, right-periphery neither (e.g. Jespersen 1917, Fitzmaurice and Smith 2012). The present study builds on the previous accounts, suggesting that the use of either as a clause-final additive focus marker had grammaticalized from a resumptive quantifier, post-posed in apposition and gradually renovating the former functions of clause-final neither in strong negative polarity contexts by a process of grammaticalization following co-optation (Heine 2013). The social stigmatization of right-periphery neither (RP-neither) as an example of negative concord at the time must therefore have been due simply to its resulting association with recessive, dialectal or non-standard usage, as RP-either rapidly increased its earlier range of functions to take over those of the ousted RP-neither in strong negative polarity contexts during the 19th century.

This paper aims to examine the behaviour of the equivalents of ‘give’ in Lhasa Tibetan in order to confirm, qualify or invalidate the universal tendencies that previous cross-linguistic research has unveiled (Newman 1996, Ed., 1997). We will first explore the semantic relations between the various forms that can express ‘give’ in Tibetan: SPRAD, BTANG, GNANG and PHUL, on the basis of previous lexicographic and descriptive research on Lhasa Tibetan, as well as a corpus of spoken Lhasa Tibetan (TSC). We will see that the most basic term (SPRAD) has not developed much beyond its literal meaning, whereas the hypernymic BTANG is used as a light verb whose constructions can be divided into several categories of meaning. GNANG is the honorific form of SPRAD, and PHUL is its humilific form. While SPRAD is not used as a light verb, its honorific and humilific counterparts are very productive light verbs. To explain this phenomenon, we will explore the honorific domain, and its systematisation in Lhasa Tibetan (Hajime, 1975; Rdorje et al., 1993; DeLancey, 1998; Tournadre & Sangda Dorje, 1998; Dorje & Lhazom, 2002). We will see that the humilific plane is not the symmetrical opposite of the honorific plane. We will also explore the productivity of GNANG and PHUL, which can be explained by the fact that giving is one of the most basic interpersonal actions of the human behavioural repertoire. It therefore establishes a link between two humans, which is essential in order for the honorific and humilific notions to emerge.