Main Article Content
Datagebaseerde aspekte van Afrikaanse reduplikasies
Abstract
Die doel van hierdie artikel is om aandag te skenk aan twee leemtes in Van Huyssteen (2000), te wete om sy bevindinge te verifieer en te verfyn aan die hand van ‘n korpus van geskrewe Afrikaans, asook om die fonologiese pool van die Afrikaanse reduplikasiekonstruksie beter te beskryf. Binne ‘n gebruiksgebaseerde benadering word daar aangetoon dat met behulp van korpusdata ‘n aangepaste konstruksieskema vir Afrikaanse reduplikasies gepostuleer kan word. So byvoorbeeld word twee betekeniswaardes, te wete ‘kindertaal' en ‘voltooid', uit die nuwe konstruksieskema verwyder, terwyl ‘intensiteit' en ‘voortdurend' eerder as die mees prototipiese betekeniswaardes van Afrikaanse reduplikasies beskou word. Met behulp van ‘n verskeidenheid eksperimente word op verskeie statistiese wyses aangetoon dat Lid 1 van ‘n reduplikasie verskil van Lid 2 daarvan, anders as wat tot dusver in die literatuur beweer is. Hierdie verskil is basies een van ‘n verskil in klemtoon, waar die akoestiese resultate duidelik aantoon dat Lid 1 die hoofklem dra. Verskeie vergelykende eksperimente met herhalings en samestellings word ook gedoen om die bevindinge te staaf. Hierdie artikel lewer nie net ‘n bydrae tot die Afrikaanse beskrywende taalkunde nie, maar toon ook die winste van datagebaseerde en datagedrewe taalondersoek.
This article is intended to address two under-explored aspects in Van Huyssteen (2000), namely, the verification of his findings and their refinement by means of a corpus of written Afrikaans, and also to provide a better description of the phonological pole of the Afrikaans reduplication construction. Within a usage-based approach it is shown that an adapted construction schema can be postulated for Afrikaans reduplications. So, for example, two semantic values, namely, ‘children's language' and ‘completed' are removed from the construction schema, while ‘intensity' and ‘continuous' are rather regarded as the most prototypical semantic values of Afrikaans reduplications. By means of a variety of experiments using various statistical methods, it is shown that Member 1 of a reduplication differs from Member 2 thereof, in contrast to what has been maintained in the literature to date. This difference basically lies in stress as the acoustic results clearly show that Member 1 carries the main accent. Various comparative experiments with word repetitions and compounds are also conducted to confirm the results. This article not only contributes to Afrikaans descriptive linguistics, but also shows the benefits of data-based and datadriven linguistic research.
Southern African Linguistics and Applied Language Studies 2007, 25(3): 419–439
This article is intended to address two under-explored aspects in Van Huyssteen (2000), namely, the verification of his findings and their refinement by means of a corpus of written Afrikaans, and also to provide a better description of the phonological pole of the Afrikaans reduplication construction. Within a usage-based approach it is shown that an adapted construction schema can be postulated for Afrikaans reduplications. So, for example, two semantic values, namely, ‘children's language' and ‘completed' are removed from the construction schema, while ‘intensity' and ‘continuous' are rather regarded as the most prototypical semantic values of Afrikaans reduplications. By means of a variety of experiments using various statistical methods, it is shown that Member 1 of a reduplication differs from Member 2 thereof, in contrast to what has been maintained in the literature to date. This difference basically lies in stress as the acoustic results clearly show that Member 1 carries the main accent. Various comparative experiments with word repetitions and compounds are also conducted to confirm the results. This article not only contributes to Afrikaans descriptive linguistics, but also shows the benefits of data-based and datadriven linguistic research.
Southern African Linguistics and Applied Language Studies 2007, 25(3): 419–439
