Thesis (DPhil)--Stellenbosch University, 2012. / ENGLISH ABSTRACT: In this dissertation I enquire into the status, scope and limits of human knowledge, given the fact that our perceptual and cognitive faculties are the product of evolution by natural selection. I argue that the commonsense representations these faculties provide us with yield a particular, species-specific scope on the world that does not ‘correspond’ in any straightforward way to the external world. We are, however, not bound by these commonsense representations. This particular, species-specific view of the world can be transgressed. Nevertheless, our transgressing representations remain confined to the conceptual space defined by the combinatorial possibilities of the various representational tools we possess. Furthermore, the way in which we fit representations to the external world is by means of our biologically determined epistemic orientation. Based on the fact that we are endowed with a particular set of perceptual and cognitive resources and are guided by a particular epistemic orientation, I conclude that we have a particular cognitive relation to the world. Therefore, an accurate representation for us is a particular fit (our epistemic orientation) with particular means (our perceptual and cognitive resources). / AFRIKAANSE OPSOMMING: Hierdie tesis handel oor die aard, omvang en limiete van kennis, gegewe dat ons perseptuele en kognitiewe vermoëns die resultaat van evolusie deur middel van natuurlike seleksie is. Eerstens, word daar geargumenteer dat die algemene voorstellings wat hierdie vermoëns aan ons bied ‘n partikuliere, spesie-spesifieke siening van die wêreld aan ons gee, wat nie op ‘n eenvoudige manier korrespondeer aan die werklikheid nie. Ons is egter nie gebonde aan hierdie voorstellings nie. Hierdie partikuliere, spesie-spesifieke siening van die wêreld kan oorskry word. Ons is egter wel beperk tot die konseptuele ruimte wat gedefinieër word deur die kombinatoriese moontlikhede van die voorstellingsmiddele tot ons beskikking. Verder word die manier waarop ons hierdie voorstellings aan die wêreld laat pas deur ons biologies gedetermineerde epistemiese oriëntasie bepaal. Dus, gegewe dat ons ‘n spesifieke stel perseptuele en kognitiewe vermoëns het en deur ‘n spesifieke kognitiewe epistemiese oriëntasie gelei word, staan ons in ‘n spesifieke kognitiewe verhouding tot die wêreld. ‘n Akkurate voorstelling (m.a.w. kennis vir ons) is om spesifieke vermoëns (perseptuele en kognitiewe vermoëns) op ‘n spesifieke manier (epsitemiese oriëntasie) aan die wêreld te laat pas.

Studies have shown that linguistic and perceptual constraints are important for speech processing. First, rule-like structures are more easily learned over vowels than over consonants. Second, sequences varying in pitch and duration are grouped following the Iambic – Trochaic Law (ITL). In this research, I investigated the origins of these linguistic and perceptual constraints. My aim was to test if vowels’ acoustic saliency was the reason why they are the preferred target for abstract computations, and to explore the extent to which the principles of the ITL come from evolutionary heritage or language experience. Results show that rats learn rules over consonants and vowels with the same ease, so saliency is insufficient to explain the asymmetries observed in humans. This also shows that animals share with humans the trochaic principle of the ITL, but they lack the iambic-grouping bias, which might rely on language experience. / Diversos estudios han encontrado que limitaciones perceptuales y de aprendizaje intervienen en el procesamiento del lenguaje. Primero, que el aprendizaje de reglas se realiza mejor sobre las vocales. Segundo, que secuencias alternando en frecuencia y duración son agrupadas siguiendo la Ley Yámbico-Trocaico (LYT). En esta investigación busqué esclarecer el origen de estas limitaciones lingüísticas y perceptuales. Mi objetivo fue estudiar si la preferencia por las vocales se debe a su prominencia acústica e investigar hasta qué punto la LYT es producto de la herencia evolutiva o de la experiencia lingüística. Los resultados muestran que las ratas computan reglas sobre vocales y consonantes, por lo que las asimetrías funcionales observadas en humanos no se explican por la saliencia acústica de las vocales. También sugieren que animales y humanos comparten el principio trocaico de la LYT, pero no el yámbico, el cual podría emerger tras años de experiencia con el lenguaje nativo.