The assignment of grammatical gender in German is a notoriously problematic phenomenon due to the apparent opacity of the gender assignment system (e.g. Comrie 1999: 461). Various models of German gender assignment have been proposed (e.g. Spitz 1965, Köpcke 1982, Corbett 1991, Wegener 1995), but none of these is able to account for all of the German data. This thesis investigates a relatively under-explored, recent approach to German gender assignment in the form of Optimal Gender Assignment Theory (OGAT), proposed by Rice (2006). Using the framework of Optimality Theory, OGAT claims that the form and meaning of a noun are of equal importance with respect to its gender. This is formally represented by the crucial equal ranking of all gender assignment constraints in a block of gender features, which is in turn ranked above a default markedness hierarchy *NEUTER » *FEMININE » *MASCULINE, which is based on category size. A key weakness of OGAT is that it does not specify what constitutes a valid gender features constraint. This means that, in theory, any constraint can be proposed ad hoc to ensure that an OGAT analysis yields the correct result. In order to prevent any constraints based on 'postfactum rationalisations' (Comrie 1999: 461) from being included in the investigation, the gender features constraints which have been proposed in the literature for German are assessed according to six criteria suggested by Enger (2009), which seek to determine whether there is independent evidence for a gender features constraint. Using an independently-verified constraint set, OGAT is then tested on a sample of 592 nouns systematically selected from the Duden Rechtschreibung. The results indicate that OGAT is relatively successful in its predictions when compared to other approaches but that it cannot account fully for the sample data. Accordingly, a revised version of the theory is proposed (OGAT II), which involves the ranking of certain gender features constraints. It is found that OGAT II is able to account for the genders of around 95% of nouns in the sample. A number of specific aspects of OGAT II are then tested by means of an experiment in which native German speakers are required to assign genders to 26 pseudo-nouns. The results suggest that OGAT II comes the closest of the systems discussed in the literature to modelling how native speakers assign gender in German.

Welsh grammatical gender exhibits several unusual properties. This paper argues that these properties are necessarily connected. The argument is based on a series of corpus investigations using techniques from statistical natural language processing, specifically distinguishing properties that exhibit significant statistical patterns from those which can be used to make useable predictions. Specifically, it’s shown that the grammatical properties of Welsh gender are such that its unusual statistical properties follow.

Research on the acquisition of grammatical gender has shown that for many languages, children gain an early command of gender. However, often in these languages gender marking is quite overt and provides a clear one-to-one correspondence between a marker and the gender encoded. In Welsh, gender marking is more complex. Gender is marked by mutations, a set of morphophonological changes that affect the initial consonants of words, and the mapping between mutation and gender is quite opaque. Two mutation types are used in part to mark feminine gender: both feminine nouns modified by the definite article and adjectives following feminine nouns undergo Soft Mutation, and the feminine gender of the possessive adjective ei is marked by Aspirate Mutation on the modified noun. The four studied in this thesis examined children's productive command of gender as expressed in the mutation of nouns modified by the definite article, of adjectives modifying nouns, and of nouns modified by the homonymic feminine and masculine possessive adjective. Mutation in non-gendered contexts was also examined. Subjects were 4- to 9 1I2-year-old children from North Wales. First, a seminaturalistic study was conducted to obtain knowledge about children's ability with gender marking. A Cloze procedure was also used to elicit children's production of masculine and feminine forms, both real words and nonsense forms, in a variety of linguistic contexts. Some of these contexts provided cues to gender status, some did not. The data obtained indicated that the acquisition of the Welsh gender system is a drawn-out process, and children have not mastered the system even by 9 112years of age. In addition, children become proficient in marking feminine nouns modified by the definite article and adjectives modifying feminine nouns before they do so on nouns modified by feminine ei. Results suggest that when a language has a complex gender system that is marked by opaque morpho-phonological processes the course of development is protracted and variable.

This thesis uses an experimental approach to explore optimal gender assignment theory, an approach to gender assignment housed in Optimality Theory (Prince & Smolensky 1993/2004). Optimal gender assignment theory was proposed by Curt Rice (2006) and stipulates that grammatical gender is assigned based on a set of crucially non-ranked gender features constraints and markedness constraints. Thirty-seven participants who were bilingual in English and German received 40 sentences containing English loanwords with the definite article removed and asked to provide the appropriate gender marker and a lexical equivalency. The study found that the constraints employed and developed for optimal gender assignment theory are not applicable to English loanwords in German.

The present thesis investigates grammatical gender processing in French as a first and second language. It focuses mainly on whether non-native speakers can achieve native-like representation and processing of gender, and whether the native language (L1) influences the acquisition of the second language (L2). Theoretical linguistic models have made two contrasting assumptions concerning the ability of late bilinguals to acquire grammatical gender in their L2. While some models propose that grammatical features, such as gender, are no longer available for L2 acquisition if they are not present in L1 (Hawkins & Chan, 1997), others assume that these features are still available via the universal grammar if required in the L2 (Schwartz & Sprouse, 1996; White, 1989, 2003). These assumptions, however, are supported only by off-line studies and do not provide a comprehensive account for gender representation and processing. The present thesis uses online techniques to address these questions both in language comprehension and language production. The first chapters are devoted to comprehension processes and examined French native speakers, English-French and German-French bilinguals’ performance during the processing of correct and syntactically anomalous sentences, using ERPs and eye-movements to record behaviour. We concluded that, like native speakers, English-French bilinguals are sensitive to gender agreement violations. Thus, we argue that late bilinguals are able to acquire the gender system of their L2 even if this grammatical feature is not present in their L1. On the other hand, the performance of the German speakers we tested suggests that the presence of a competing gender system in the native language may hamper gender acquisition in L2. The influence of the native language may vary, however, according to both proficiency and how gender systems map across languages, as suggest the results we obtained with Spanish bilinguals tested in language production. In a second series of experiments, we examined determiner selection in French to further investigate gender representation and processing, but in language production. Using a picture-word interference paradigm, we compared the production of simple and complex noun phrases (NP) in French native speakers, English-French and Spanish-French bilinguals. From our results, we argue that gender representation is similar in L1 and L2, but that gender processing is less incremental in non-native speakers in that they do not compute agreement between the noun and other elements of the NP as automatically as native speakers do. The absence of interference between the two gender systems of the Spanish-French bilinguals we tested suggests that the gender systems of the two languages may be autonomous in highly proficient bilinguals. Our results suggest that highly proficient bilinguals can reach native-like representation and processing of gender in their L2 and that such is not constrained by either the age of onset of learning or the grammar of the learners L1.

abstract: Purpose: The present study examined grammatical gender use in child Spanish heritage speakers (HSs) in order to determine whether the differences observed in their grammar, when compared to Spanish monolinguals, stem from an incompletely acquired grammar, in which development stops, or from a restructuring process, in which features from the dominant and the weaker language converge to form a new grammatical system. In addition, this study evaluated whether the differences usually found in comprehension are also present in production. Finally, this study evaluates if HSs differences are the result of the input available to them. Method: One-hundred and four typically developing children, 48 HSs and 58 monolingual, were selected based on two age groups (Preschool vs. 3rd Grade). Two comprehension and three production experimental tasks were designed for the three different grammatical structures where Spanish expresses gender (determiners, adjectives, and clitic pronouns). Linear mixed-models were used to examine main effects between groups and grammatical structures. Results: Results from this study showed that HSs scored significantly lower than monolingual speakers in all tasks and structures; however, 3rd-Grade HSs had higher accuracy than PK-HSs. Error patterns were similar between monolinguals and HSs. Moreover, the commonly reported overgeneralization of the masculine form seems to decrease as HSs get older. Conclusion: These results suggest that HSs’ do not face a case of Incomplete Acquisition or Restructured Grammatical gender system, but instead follow a protracted language development in which grammatical skills continue to develop after preschool years and follow the same developmental patterns as monolingual children / Dissertation/Thesis / Doctoral Dissertation Speech and Hearing Science 2018

Past studies of grammatical gender have shown that native speakers encounter processing difficulties when encountering a form that does not agree in gender with previous words. However, the specific behavioral and neural responses to these difficulties have not been replicated across studies of the same type. This is in part due to different experimental designs and statistical analyses, but a crucial factor may be the lack of control between nouns of masculine and feminine gender in stimuli creation. Masculine and feminine gender show distinct distributional asymmetries and collapsing them into one condition diminishes the explanatory power of any study examining grammatical gender. We used reading times in a self-paced reading experiment to examine whether masculine and feminine gender violations differentially affect processing speeds. Fifty French speakers read sentences that were well-formed or contained a mismatch in gender between determiner and noun, half of which were masculine and half feminine. Following Beatty-Martínez et al. (2021), we added individual difference measures to determine how participant-specific factors modulate processing. Participants also completed a category verbal fluency task and the AX-CPT, a measure of cognitive control. They found that ERP components were modulated by these components for Spanish speakers and the modulation differed between masculine noun and feminine noun violations. We hypothesized that reading times would be similarly affected in French, a closely related language with the same gender categories. However, no conditions or interactions reached statistical significance. It is unclear whether this is due to the experimental manipulation or lack of control for participants’ language background, as we had a high number of bilingual and multilingual participants. Regardless, elements of the procedure may provide insight on how to design future experiments that lay a groundwork in understanding the most basic elements of gender processing. / Thesis / Master of Science (MSc) / Why is ‘pen’ masculine and ‘car’ feminine? Grammatical gender is a widespread feature of languages that comes naturally to native speakers and perplexes many second language learners. The assignment of gender seems to be random, but upon closer examination, patterns can be established. What do these differences mean for speakers of gendered languages? In the current study, we set out to determine how masculine and feminine grammatical gender is processed in French and how this is influenced by differences between individual speakers. Participants read French sentences that were either grammatical or contained a mismatch in gender between article and noun. Reading times were used to evaluate how speakers react when encountering an ungrammatical form with either masculine or feminine gender. Participants also completed tasks measuring response inhibition and verbal fluency to see how individuals with different cognitive and language skills react differently to unexpected forms.

No description available.

In the last few decades, there has been increased interest in the incorporation of data from bi- and multilingual individuals in linguistic theory: from second language acquisition and language attrition to heritage varieties and code-switching. This dissertation discusses a range of ways in which code-switching data can provide insight into the mechanisms that underlie linguistic structures. The data will be analysed within the framework of Minimalist Generative syntax and Distributed Morphology.The first part investigates grammatical gender assignment in code- switching between English, a language without grammatical gender, and two languages with grammatical gender: French and Belgian Dutch. These languages have comparable, but different gender systems. French has two genders: masculine and feminine, whereas Belgian Dutch adds a third: neuter. The study in this part of the dissertation compares gen- der assignment strategies in bilinguals with different profiles. In addition, the code-switching data provide evidence against the default status of neuter in Belgian Dutch.The second part focuses on word order and includes two studies: one on verb-second word order in Dutch-English code-switching and one on adverb placement in English-French and Dutch-English code- switching. The verb-second chapter identifies a lacuna in the traditional Generative analysis for verb second and uses the CS data to address this. The chapter on adverb position looks at placement of the adverb between the verb and its direct object, which is allowed in Dutch and French, but not in English. For all domains investigated, it is found that the finite verb predicts word order.Taken together, these studies demonstrate that bilingual data can shine a light on elements of the theory of grammar which remain in the shadows when only monolingual data is used. / Les dernières décennies ont vu croître l’intérêt pour l’intégration à la réflexion en linguistique théorique des données produites par des locuteurs/trices bilingues ou multilingues, que celles-ci concernent l’acquisition d’une langue seconde, l’attrition, les langues d’héritage ou l’alternance codique. Le présent travail développe plusieurs exemples où les données issues de l’alternance codique éclairent les mécanismes qui sous-tendent les structures linguistiques. Les données recueillies sont interprétées dans le cadre de la syntaxe générative minimaliste et de la morphologie distribuée (« distributed morphology »).Dans un premier temps, nous analysons l’attribution du genre grammatical dans l’alternance entre l’anglais, d’une part, et le français et le néerlandais de Belgique, de l’autre. Alors qu’il n’y a pas en anglais de genre grammatical, le français et le néerlandais de Belgique marquent ce genre, mais de façon différente :si le français distingue deux genres, masculin et féminin, le néerlandais de Belgique y adjoint un troisième, le neutre. Dans cette partie de la thèse, nous dressons le profil des stratégies d’attribution du genre auprès de deux types distincts de bilingues et nous établissons également que le neutre n’est pas le genre par défaut en néerlandais de Belgique.Dans un second temps, nous nous penchons sur l’ordre des constituants. Dans une première étude, nous examinons l’ordre des mots avec « verbe second » (V2) dans l’alternance anglais-néerlandais. Nous abordons ensuite le placement de l’adverbe dans l’alternance anglais- français et anglais-néerlandais. Le chapitre consacré à V2 identifie une lacune dans la littérature générative et tire profit des données de l’al- ternance pour y proposer une solution. Le chapitre consacré à l’adverbe s’intéresse au placement de celui-ci entre le verbe et son objet, position licite en français et néerlandais mais pas en anglais. Dans ces deux études, il apparaît que c’est la langue du verbe à la forme finie qui prédit l’ordre des constituants.L’ensemble des recherches ici réunies démontre que les données bilingues mettent en lumière des aspects de la théorie grammaticale qui restent dans l’ombre lorsque le chercheur se limite à des données monolingues. / Doctorat en Langues, lettres et traductologie / info:eu-repo/semantics/nonPublished

The creation of word embeddings is one of the key breakthroughs in natural language processing. Word embeddings allow for words to be represented semantically, opening the way to many new deep learning methods. Understanding what information is in word embeddings will help understanding the behaviour of embeddings in natural language processing tasks, but also allows for the quantitative study of the linguistic features such as grammatical gender. This thesis attempts to explore how grammatical gender is encoded in word embeddings, through analysing the performance of a neural network classifier on the classification of nouns by gender. This analysis is done in three experiments: an analysis of contextualized embeddings, an analysis of embeddings learned from modified corpora and an analysis of aligned embeddings in many languages. The contextualized word embedding model ELMo has multiple output layers with a gradual increasing presence of semantic information in the embedding. This differing presence of semantic information was used to test the classifier's reliance on semantic information. Swedish, German, Spanish and Russian embeddings were classified at all layers of a three layered ELMo model. The word representation layer without any contextualization was found to produce the best accuracy, indicating the noise introduced by the contextualization was more impactful than any potential extra semantic information. Swedish embeddings were learned from a corpus stripped of articles and a stemmed corpus. Both sets of embeddings showed an drop of about 6% in accuracy in comparison with the embeddings from a non-augmented corpus, indicating agreement plays a large role in the classification. Aligned multilingual embeddings were used to measure the accuracy of a grammatical gender classifier in 24 languages. The classifier models were applied to data of other languages to determine the similarity of the encoding of grammatical gender in these embeddings. Correcting the results with a random guessing baseline shows that transferred models can be highly accurate in certain language combinations and in some cases almost approach the accuracy of the model on its source data. A comparison between transfer accuracy and phylogenetic distance showed that the model transferability follows a pattern that resembles the phylogenetic distance.