In the past few decades, the investigation of grammatical change using electronic corpora has made headway internationally. Although linguists previously believed that grammatical changes progress too slowly to observe, this method enables linguists to investigate even recent, or ongoing, changes. However, no comprehensive study of recent and ongoing grammatical changes in Afrikaans has appeared yet. Also, when comments about ongoing changes are made, it is usually based on anecdotal evidence, with a focus on English influence. In this study, the method of short-term diachronic comparable corpus linguistics is used to investigate grammatical changes in written Standard Afrikaans from 1911 to 2010. Four corpora were collected to this end, representing language use from 1911-1920, 1941-1950, 1971-1980 and 2001-2010. Additionally, quantitative grammaticography is used to take into account possible effects of prescriptive sources. Two research questions are adressed in this study: the first inquires into the nature and extent of grammatical changes in selected grammatical categories in written Standard Afrikaans from 1911 to 2010; the second wants to clarify the differences and similarities between internal and external language change, and in the light thereof establish to which extent external change, and specifically English influence, is relevant for grammatical changes in Standard Afrikaans during the past century. The theoretical framework within which language use and change is investigated in this study is cognitive linguistics, specifically emergent grammar and the exemplar model. Changes that become apparent from the data are described and explained in terms of processes of change and forces of change, and linked to the principles of cognitive linguistics. Three broad grammatical categories are investigated: temporal reference, pronouns and the genitive. Even though there is an extent of stability in each of the categories, there are also several bigger and smaller changes that give an overview of the nature of grammatical change in written Standard Afrikaans in the past century. These changes can be divided into different categories. The first type of change has to do with formalisation and colloquialisation – in broad strokes, there are signs of formalisation between the first two periods, during which the standard variety was being established, causing some features associated with formality to increase (e.g. passive constructions). However, at the end of the century there are signs of colloquialisation between the last two periods, where some formal features decrease (e.g. the formal second person pronoun u "you"), and some informal features increase (e.g. nou "now" as discourse marker). The second type of change is analogy, causing greater regularity and/or uniformity in a paradigm. For instance, obsolescent preterite forms (had "had", wis "knew") were replaced by regular forms (het/het gehad, het geweet). The last of the Dutch genitive was also replaced by the Afrikaans genitive with se "'s" and van "of". The third type of change is driven by speakers' desire to be expressive. Some of the pronouns specialise increasingly, meaning that they are used less and less for functions other than their main function, and other options are used less and less for that function. Examples of this is the third person pronoun dit "it", the shortened forms jul "you/your" and hul "they/their", and the indefinite pronouns almal "everybody", alles "everything" and elkeen "each one". A next type of change is actually a combination of different processes and forces: grammaticalisation. There are several instances of grammaticalisation: the use of gaan "go" for future reference, the use of dis "it's" rather than dit is "it is", the use of mens "human" rather than 'n mens "a human" as generic pronoun, the use of indefinite pronouns with enig- "any" like enigiets "anything", enigiemand "anybody", enigeen "anyone", and the use of the genitive particle se "'s". The last type of change is externally motivated change. Contrary to the view the Afrikaans literature in general promotes, there is only one instance of confirmed English influence in the data of this study: the increasing use of -self with reflexive pronouns, rather than the bare object form. However, there are instances of extra-linguistic influence, like standardisation that caused large scale variation reduction between the first and the second period, and the influence of feminism that can be seen in decreasing linguistic sexism, particularly with regard to generic pronouns. The conclusion in the end is that the process of internally motivated change and contact-induced change is not different – an innovation can originate from another language (overt transfer), or an internal innovation can be promoted through bi- or multilingualism (covert transfer); however, the same principles, processes and forces of change are at play, irrespective of how many languages are involved.

In the past few decades, the investigation of grammatical change using electronic corpora has made headway internationally. Although linguists previously believed that grammatical changes progress too slowly to observe, this method enables linguists to investigate even recent, or ongoing, changes. However, no comprehensive study of recent and ongoing grammatical changes in Afrikaans has appeared yet. Also, when comments about ongoing changes are made, it is usually based on anecdotal evidence, with a focus on English influence. In this study, the method of short-term diachronic comparable corpus linguistics is used to investigate grammatical changes in written Standard Afrikaans from 1911 to 2010. Four corpora were collected to this end, representing language use from 1911-1920, 1941-1950, 1971-1980 and 2001-2010. Additionally, quantitative grammaticography is used to take into account possible effects of prescriptive sources. Two research questions are adressed in this study: the first inquires into the nature and extent of grammatical changes in selected grammatical categories in written Standard Afrikaans from 1911 to 2010; the second wants to clarify the differences and similarities between internal and external language change, and in the light thereof establish to which extent external change, and specifically English influence, is relevant for grammatical changes in Standard Afrikaans during the past century. The theoretical framework within which language use and change is investigated in this study is cognitive linguistics, specifically emergent grammar and the exemplar model. Changes that become apparent from the data are described and explained in terms of processes of change and forces of change, and linked to the principles of cognitive linguistics. Three broad grammatical categories are investigated: temporal reference, pronouns and the genitive. Even though there is an extent of stability in each of the categories, there are also several bigger and smaller changes that give an overview of the nature of grammatical change in written Standard Afrikaans in the past century. These changes can be divided into different categories. The first type of change has to do with formalisation and colloquialisation – in broad strokes, there are signs of formalisation between the first two periods, during which the standard variety was being established, causing some features associated with formality to increase (e.g. passive constructions). However, at the end of the century there are signs of colloquialisation between the last two periods, where some formal features decrease (e.g. the formal second person pronoun u "you"), and some informal features increase (e.g. nou "now" as discourse marker). The second type of change is analogy, causing greater regularity and/or uniformity in a paradigm. For instance, obsolescent preterite forms (had "had", wis "knew") were replaced by regular forms (het/het gehad, het geweet). The last of the Dutch genitive was also replaced by the Afrikaans genitive with se "'s" and van "of". The third type of change is driven by speakers' desire to be expressive. Some of the pronouns specialise increasingly, meaning that they are used less and less for functions other than their main function, and other options are used less and less for that function. Examples of this is the third person pronoun dit "it", the shortened forms jul "you/your" and hul "they/their", and the indefinite pronouns almal "everybody", alles "everything" and elkeen "each one". A next type of change is actually a combination of different processes and forces: grammaticalisation. There are several instances of grammaticalisation: the use of gaan "go" for future reference, the use of dis "it's" rather than dit is "it is", the use of mens "human" rather than 'n mens "a human" as generic pronoun, the use of indefinite pronouns with enig- "any" like enigiets "anything", enigiemand "anybody", enigeen "anyone", and the use of the genitive particle se "'s". The last type of change is externally motivated change. Contrary to the view the Afrikaans literature in general promotes, there is only one instance of confirmed English influence in the data of this study: the increasing use of -self with reflexive pronouns, rather than the bare object form. However, there are instances of extra-linguistic influence, like standardisation that caused large scale variation reduction between the first and the second period, and the influence of feminism that can be seen in decreasing linguistic sexism, particularly with regard to generic pronouns. The conclusion in the end is that the process of internally motivated change and contact-induced change is not different – an innovation can originate from another language (overt transfer), or an internal innovation can be promoted through bi- or multilingualism (covert transfer); however, the same principles, processes and forces of change are at play, irrespective of how many languages are involved.

Politeness is a linguistic strategy used in order to avoid conflicts in verbal interaction. Politeness is an influential variable in sociolinguistic (cf. MEYERHOFF, 2006) to be related to language use: the pragmatic point of view, the social distance, power relations and the cost of enforcing variables are strongly involved in evaluation of linguistic strategies which are polished or not (BROWN; LEVINSON, 2011 [1987]); and sociolinguistic point of view, the sex/gender proves to be significant. Among the linguistic strategies used to express this value, we are interested in the verbal form of the future tense (FP). The use of this verb form may vary according to the value of temporal reference: past, present, future. A ware of this possibility of variation of the FP and considering that politeness can be seen on a continuum (the less polished the more polished), this research aimed to investigate the effects of pragmatic and sociolinguistic aspects of the uses FP as a function of time reference in the speech data of informants Itabaiana/SE. The general hypothesis guiding our research is that the FP alone does not encode politeness, but a set of contextual features in specific timeframes. To develop research in this perspective, we used as the sample speech corpus social network of university informants Itabaiana/SE. This sample consists of interactions conducted - the informants themselves lead to interaction - collected from a methodological model developed in our study specifically to capture the nuances of politeness, both in its pragmatic aspects as sociolinguistic. The collected data were categorized and analyzed statistically. The results were generated from three rounds statistics, with the variable rule the temporal reference of the form of FP (past, present, future) and the expression of politeness: past x present x future, past x no past, present x future. The results obtained in the first round showed that none of the controlled variables was significant in the expression of the phenomenon under study. In the second round, the program selected only two variables as significant: the verb form and the kind of discursive sequence. The results showed that the use of FP with past temporal reference is favored when the verb occurs with the assist going and the type of sequence is narrative. In the third round statistics, present x future, five variables were selected as significant: control of the interaction terms of sex/gender, verbal, linguistic parallelism, cost of enforcing and question-answer pair and comment. Among these, we highlight the results obtained with the control of the interaction terms of sex/gender which showed that men when they are in the field of the topic tend to use more verbal form of FP with present time reference. Regarding the variable cost of the levy, the results showed that the less imposing was the most recurring topic was the use of the FP with this timeframe. As for the question-answer pair variable and review, the results showed that the use of FP with present time reference was conditioned in contexts that were characterized as comment/contextualization the topic. In general, the use of theoretical and methodological collection procedures focusing on the pragmatic and sociolinguistic effects to capture the effects of politeness allowed us to demonstrate that there are significant differences regarding the use of the FP, especially regarding social distance and sex/gender. / A polidez é uma estratégia linguística utilizada com o objetivo de evitar conflitos na interação verbal. Trata-se de uma variável influente na sociolinguística (cf. MEYERHOFF, 2006) por estar relacionada à língua em uso: do ponto de vista pragmático, a distância social, as relações de poder e o custo da imposição são variáveis fortemente envolvidas na avaliação de quais estratégias linguísticas são polidas ou não (BROWN; LEVINSON, 2011 [1987]); e do ponto de vista sociolinguístico, o sexo/gênero mostra-se significativo. Dentre as estratégias linguísticas utilizadas para expressar esse valor, interessa-nos a forma verbal de futuro do pretérito (FP). O uso dessa forma verbal pode variar conforme o valor da referência temporal: passado, presente, futuro. A par dessa possibilidade de variação do FP e considerando que a polidez pode ser analisada em um continuum (do menos polido ao mais polido), objetivamos nesta pesquisa verificar os efeitos dos aspectos pragmáticos e sociolinguísticos nos usos do FP em função da referência temporal em dados de fala de informantes de Itabaiana/SE. A hipótese geral que norteia a nossa investigação é a de que o FP por si só não codifica polidez, mas sim um conjunto de traços contextuais em referências temporais específicas. Para desenvolvermos a pesquisa nessa perspectiva, utilizamos como corpus a amostra de fala Rede Social de Informantes Universitários de Itabaiana/SE. Esta amostra é composta por interações conduzidas - os próprios informantes conduzem a interação - coletada a partir de um modelo metodológico elaborado em nosso estudo especificamente para captar as nuanças de polidez, tanto em seus aspectos pragmáticos quanto sociolinguísticos. Os dados coletados foram categorizados e submetidos à análise estatística. Os resultados foram gerados a partir de três rodadas estatísticas, tendo como regra variável a referência temporal da forma de FP (passado, presente, futuro) e a expressão de polidez: passado x presente x futuro, passado x não passado, presente x futuro. Os resultados obtidos na primeira rodada evidenciaram que nenhuma das variáveis controladas foi significativa na expressão do fenômeno em estudo. Já na segunda rodada, o programa selecionou apenas duas variáveis como significativas: a forma verbal e o tipo de sequência discursiva. Os resultados evidenciaram que o uso do FP com referência temporal passada é favorecido quando a forma verbal ocorre com o auxiliar ir e o tipo de sequência é narrativo. Na terceira rodada estatística, presente x futuro, cinco variáveis foram selecionadas como significativas: o controle da interação quanto ao sexo/gênero, forma verbal, paralelismo linguístico, custo da imposição e par pergunta-resposta e comentário. Dentre estas, ressaltamos os resultados obtidos com o controle da interação quanto ao sexo/gênero, os quais demonstraram que os homens quando estão com domínio do tópico tendem a utilizar mais a forma verbal de FP com referência temporal presente. No que concerne a variável custo da imposição, os resultados evidenciaram que quanto menos impositivo era o tópico, mais recorrente foi o uso do FP com referência temporal presente. Quanto a variável par pergunta-resposta e comentário, os resultados mostraram que o uso do FP com referência temporal presente foi condicionado em contextos que se caracterizavam como comentário/contextualização do tópico. Em termos gerais, o uso de procedimentos teórico-metodológicos de coleta focalizando os efeitos pragmáticos e sociolinguísticos para captar os efeitos de polidez permitiu-nos comprovar que há diferenças significativas em relação ao uso do FP, principalmente, quanto à distância social e ao sexo/gênero.

Cette thèse propose une analyse du conditionnel en français basée sur des concepts énonciatifs et aspecto-temporels, dont le plus important pour ce temps grammatical se révèle être celui de « référentiel temporel ». Après avoir exposé le cadre théorique de J.-P. Desclés et Z. Guentchéva utilisé pour ce travail, nous présentons tout d’abord une application de ce modèle à des temps verbaux de l’indicatif (présent, imparfait et futur), puis nous abordons les problématiques du discours rapporté et des énoncés en si, que concernent à la fois les temps de l’indicatif mentionnés et le conditionnel. Nous passons ensuite à l’analyse du conditionnel lui-même, qui est sous-tendu par un invariant sémantique : ce temps grammatical pose la relation prédicative aspectualisée (le procès) dans un référentiel autre que le Référentiel Énonciatif, l’actualisation de ce procès vers le Référentiel Énonciatif se faisant alors de différentes manières selon les trois classes d’emplois du conditionnel que nous avons dégagées (« futur dans le passé », hypothèse, et désengagement). Chaque classe est étudiée en détail et divisée en sous-classes. Nous proposons dans une troisième partie une étude des équivalents en allemand du conditionnel. Cette langue utilise diverses formes verbales, relevant des modes Indikativ, Konjunktiv I et II, ainsi que de la forme en würde + infinitif, pour exprimer les valeurs du conditionnel. Pour finir, une comparaison entre cette dernière forme et le conditionnel met en évidence la pertinence du concept de référentiel temporel puisqu’il permet d’expliciter le rapprochement sémantique entre deux formes verbales morphologiquement différentes. / This thesis proposes an analysis of the French conditional based on enunciative and aspecto-temporal concepts, where the most important concept for this grammatical tense is shown to be that of “temporal reference framework”. Having expounded the theoretical framework of J.-P. Desclés and Z. Guentchéva, we will first present an application of this model to verb tenses in the indicative (present, imperfect and future), after which we will take up the problematics of reported speech and si utterances which both the indicative tenses mentioned and the conditional are concerned with. We will then move on to the analysis of the conditional itself which is underlain by a semantic invariant: this grammatical tense places the aspectualized predicative relation (the process) in a reference framework other than the Enunciative Reference Framework. The actualization of this process in another reference framework different from the Enunciative Reference Framework is realized in different ways according to the three classes of usage of the conditional which we have singled out (“future in the past”, hypothetical value, disengagement). Each class is studied in detail and divided into sub-classes. In a third part we study the equivalents of the conditional in German. This language makes use of some verb forms of Indikativ and Konjunktiv I and II moods as well as the form würde + infinitive, to express the values of the conditional. Finally, a comparison of this last form with the conditional will reveal the pertinence of the concept of temporal reference framework since it explains the semantic connection between two morphologically different verb forms.

In the Arabic-speaking world, sociopolitical upheaval, extended conflict and population displacement have triggered extensive contact between mutually intelligible varieties of the language. Notwithstanding these developments, Arabic sociolinguistic research on dialect contact settings remains limited to certain well-documented areas (e.g., Al-Wer 2020), with markedly less research targeting other locales believed to be highly propitious to convergent change, such as the long-term contact situation in Lebanon involving Lebanese and Palestinian Arabic (Fityan 1981; Hennessey 2011). Furthermore, few studies are embedded in a (comparative) variationist sociolinguistic framework (Owens 2013), and even fewer studies are articulated from a socio-historical perspective incorporating diachronic data sources with which to better understand the process of language change in Arabic (Owens 2013). Much previous research on Arabic dialects is also based on investigations of phonological variation (Al-Wer and de Jong 2018), with correspondingly less attention paid to (morpho-)syntactic variation (Choueiri 2019). The present study aims to address existing lacunae in the research literature by investigating the outcomes of dialect contact in Beirut between Palestinian Arabic (PA), the minority variety, and Lebanese Arabic (LA), the majority variety. Drawing on the framework of comparative variationist sociolinguistics (Poplack and Tagliamonte 2001) as well as research on dialect contact (Britain and Trudgill 2005), this study combines synchronic and diachronic data sources to compare three variables in LA and PA: a phonological variable, involving the word-medial raising of /a:/ to [e:] (e.g., [ka:n] alternating with [ke:n] 'he/it was'); and two morpho-syntactic variables: verbal negation and future temporal reference. The overarching aim of the research is to examine the extent to which PA shows evidence, as gauged from linguistic constraints on variant selection and variant repertoires, of becoming more structurally similar to LA in different linguistic components (Cheshire, Kerswill, and Williams 2005). The synchronic data come from 45 hours of spontaneous speech recorded in Beirut from 39 Palestinian and 27 Lebanese speakers stratified by age, sex, and level of education, generating 7,671 tokens representing the three targeted variables. A further 15,381 tokens of these three variables come from two diachronic datasets. The first is a sub-set of speech recordings from the Palestinian Oral History Archive, an online compendium of interviews with first-generation (older) Palestinians in Lebanon, recorded between the 1990s and early 2000s. The second diachronic dataset is the Lebanese Popular Theatre Corpus (LPTC), based on 34 televised plays dating from the 1960s and performed in colloquial LA. Results reveal that the [e:] variant, a stereotypical feature of LA, but not emblematic of PA spoken in Beirut (Hennessey 2011), is virtually absent from the speech of the older Palestinian generation in the synchronic and diachronic datasets, but it increases significantly in the speech of young (third-generation) Palestinian speakers, who replicate the linguistic conditioning of variant selection in LA. These results bolster the inference of contact-induced change in PA due to the influence of LA. With respect to verbal negation, the findings show that there is convergent change in terms of overall variant rates in this variable system in PA. Evidence suggests that this variable system is undergoing dialect levelling as a result of contact, with socially marked minority variants diminishing over time in the speech of educated Palestinians. The future temporal reference system, however, seems to be less amenable to contact-induced change, despite similarities in surface forms between LA and PA. Results indicate that this variable system is undergoing an internal change in PA independent of contact with LA, which is led by young, educated speakers, in line with what has been observed in PA spoken outside Lebanon (AbuAmsha 2016). Viewed in the aggregate, the results show that even though it is claimed that (morpho-) syntactic variables may be less susceptible to convergent change than phonological variables (Cheshire et al. 2005; Hinskens et al. 2005), we do not find a neat division between phonology and morpho-syntax. Word-medial imala is overtly commented on and explicitly identified by the targeted Palestinian speech community as a marker of Lebanese speech. Its iconic association with Lebanese speech patterns renders it particularly susceptible to long-term dialect accommodation for some Palestinians. Verbal negation is also subject to social evaluation, as gauged from explicit speaker meta-commentary, and socially marked exponents appear vulnerable to attrition over time. By contrast, the expression of the future temporal reference appears less socially indexical than the other variables and is not subject to normative commentary or overt correction. These differences implicate the social salience of the targeted variables as a key factor influencing their susceptibility to convergence. Situating the results in a wider perspective, the findings highlight the utility of the comparative variationist framework in elucidating the process of language change in spoken Arabic, especially in PA as spoken in Beirut, as well as in distinguishing contact-induced change from internally-motivated change. The results of this study indicate that the effects of dialect contact, and critically, the existence of contact-induced change cannot be fully understood without using a multi-faceted comparative approach incorporating horizontal and vertical comparisons. The results converge in demonstrating that an empirically accountable quantitative approach based on actual speech data is capable of transcending the limitations of alternative frameworks of analysis that have been used to investigate change in dialect contact scenarios in the Arabic-speaking world.