Elsabé Taljard

Part-of-Speech Tagging of Northern Sotho: Disambiguating Polysemous Function Words

A major obstacle to part-of-speech (=POS) tagging of Northern Sotho (Bantu, S 32) are ambiguous function words. Many are highly polysemous and very frequent in texts, and their local context is not always distinctive. With certain taggers, this issue leads to comparatively poor results (between 88 and 92% accuracy), especially when sizeable tagsets (over 100 tags) are used. We use the RF-tagger (Schmid and Laws, 2008), which is particularly designed for the annotation of fine-grained tagsets (e.g. including agreement information), and we restructure the 141 tags of the tagset proposed by Taljard et al. (2008) in a way to fit the RF tagger. This leads to over 94% accuracy. Error analysis in addition shows which types of phenomena cause trouble in the POS-tagging of Northern Sotho.

Locative trigrams in Northern Sotho, preceded by analyses of formative bigrams*

Abstract In Northern Sotho one of the strategies to express locality makes use of locative particle groups, being complements preceded by any of the so-called locative particles ka, kua, mo, ga, or go. Current linguistic descriptions shy away from those cases where sequences of such particles are employed. In this article these sequences are termed “locative n-grams” and are studied for the first time. It will be shown that, synchronically, just a handful of locative trigrams and bigrams do actually occur in a relatively large corpus. An in-depth study of the examples allows taking stock of the existing structures, provides data regarding the distribution of all the n-grams, and hints at the semantic content as well as the restrictions posed on the nature of the complements. In order to get clarity on the latter two aspects, a diachronic approach is often pursued. As a by-product, the study of the higher-order n-grams also brings hitherto overlooked features of the unigrams to light. The main research question that drove this investigation was thus to find out whether or not higher-order locative n-grams exist in Northern Sotho. As the answer was found to be positive, the major objective became to describe the found structures minutely by drawing on corpus data.

The use of LSP dictionaries in secondary schools - a South African case study

This article reports on the results of a broad evaluation of the efficacy of the Multilingual Explanatory Science Dictionary and Multilingual Explanatory Math Dictionary in a multilingual educational environment. The aim of the investigation is to ascertain (a) whether the target users possess the necessary dictionary using skills to make use of the dictionaries effectively, and (b) whether the benefit of exposure to definitions of terms in the home language is significant in the decoding of the meaning of science and mathematical terms. Data were collected by means of two questionnaires that were completed by members of the intended target user group. Participants in the study revealed themselves inexperienced and untrained dictionary users with rudimentary dictionary using skills. They were able to perform simple look-up procedures but performed badly in cases where a more sophisticated approach is called for.

Corpus-based linguistic investigation for the South African Bantu languages: a Northern Sotho case study

The advent of electronic corpora has revolutionized linguistic investigation internationally and is starting to make inroads into the study of the South African Bantu languages. In this article, it is argued that the move from traditional language studies to corpus-based investigation would be a natural one for South African Bantuists, since it proffers a familiar methodology in a new guise. To illustrate the value and potential of corpus-based investigation, a case study was done on Northern Sotho adjectives, highlighting certain semantic aspects which have hitherto not been noted by grammarians, and illustrating the historical relationship between adjectives and enumeratives.

Corpus-based language teaching : an African language perspective

Abstract Studies on corpus-based language teaching are notably absent within the South African educational context; more so with regard to the teaching of African languages. This article explores the possibilities offered by the availability of an electronic corpus to enhance language teaching, and more specifically, the teaching of Northern Sotho as a second additional language at first year university level to first-time learners of the language. Particular attention is paid to corpus-based selection and sequencing of learning material, an activity that has hitherto depended on anecdotal evidence and the intuition of the language teacher. A critical evaluation of existing pedagogical material for Northern Sotho reveals that although excellent sources of reference, these works are inadequate for the purpose of teaching Northern Sotho to first-time learners. It is indicated that information gleaned from a corpus provides the language teacher with guidance on both micro and macro level with regard to selection and sequencing of learning content.

Designing a verb guesser for part of speech tagging in Northern Sotho

The aim of this article is to describe the design and implementation of a verb guesser that will enhance the results of statistical part of speech (POS) tagging of verbs in Northern Sotho. It will be illustrated that verb stems in Northern Sotho can successfully be recognised by examining their suffixes and combinations of suffixes. Two approaches to verbal derivation analysis will be utilised, namely morphological analysis and corpus querying of suffixes and combinations of suffixes.

Collocations and grammatical patterns in a multilingual online term bank

This article considers the importance of including various types of collocations in a terminological database, with the aim of making this information available to the user via the user interface. We refer specifically to the inclusion of empirical and phraseological collocations, and information on grammatical patterning. Attention is also given to provision of information on semantic prosody and semantic preferences — aspects which have been rather neglected in South African terminological databanks and language for special purposes (LSP) dictionaries. Various strategies for the effective semi-automatic extraction of collocational data from specialized corpora are explored. Possibilities regarding access to and presentation of collocational information to the user are briefly considered. It is argued that users should have direct access to collocational information, and that collocations should not only be accessible via the lemmatic address of the term appearing as part of the collocation. The research is done within the context of the establishment of an Open Access Resource Term Bank, which is developed as a pedagogical tool to support students whose language of learning and teaching is not the L1.

The Semantics of Eating in Afrikaans and Northern Sotho: Cross-linguistic Variation in Metaphor

The abundant and systematic presence of metaphor in language has in particular been explored by departing from the embodied nature of many metaphors. In the current research we investigate the manner in which the concept EATING in two nonrelated languages, namely Afrikaans (a Germanic language) and Northern Sotho (a Bantu language) gives rise to metaphorical expressions in these two languages. The two notions of cultural model and metaphor form the cornerstones of our research. The basic question guiding our research is whether the metaphorical mappings originating from the same source domain (EATING) onto various target domains are the same in the two languages and secondly, whether there is any evidence that differences—if any—are culturally motivated. Our study is corpus-based. Lexical items belonging to the source domain of eating were used as search nodes in our corpus search. Our analysis indicates that the metaphorical source-domain–target-domain mappings in the two languages show a large amount of overlap. As far as the metaphors that we identified are concerned, remarkable similarities and very few—and these not significant—differences were found.

Designing a noun guesser for part of speech tagging in Northern Sotho

In this article, we describe an element of a suite of computational tools for assigning word-class tags (as a preparation for part of speech (POS) tagging) to word forms in unrestricted Northern Sotho texts. POS-tagging is a step towards a linguistic analysis of the texts, which in turn allows for advanced data extraction. The tool component that is described, identifies (and classifies) noun forms. Several types of linguistic knowledge are used to recognize nouns that are not contained in the noun lexicon of the system. These include the relationship between singular and plural noun prefixes, knowledge about noun derivation, and data about the co-occurrence of the candidate with concords, pronouns and adjectives in a local context. Our implementation is a symbolic, voting-based process: together, all tests determine whether a candidate is a noun; accuracy on unseen test data is around 92%.

Cattle and cattle colour terminology in South Africa

A comparison between cattle colour terminology found in the Khoikhoi languages and that found in South African Bantu languages brings noticeable similarities to light, leading to the assumption that both the cattle and the terms used to describe their characteristic features were acquired by the South Eastern Bantu-speakers through contact with Khoikhoi pastoralists. In this article we investigate the possibility that these terminological similarities are the result of mutual relatedness to a single progenitor set of terms, rather than of borrowing that took place in a restricted geographical area. To this end, three hypotheses are proposed, respectively termed the East African link-hypothesis, the Northern Botswana link-hypothesis and the Sudan link-hypothesis. We argue that the latter, pointing to a common Nilotic proto-source, represents the most plausible explanation for the origin of cattle colour terminology shared by amongst others, the South African Bantu and Khoikhoi languages.