Investigating Cognitive and Functional Biases for Learning Acoustic Categories

Sabine van der Ham

Research output: ThesisPhD Thesis

Abstract

Do cognitive adaptations related to language exist? A broad
spectrum of theoretical accounts exists that aim to explain
human language abilities and how these abilities emerged or
evolved. However, many of the properties of language can be
explained by the effects of cultural evolution: the constant
interaction between language learners results in a learnable
(i.e. structured and compressible) language for its speakers.
That does not mean that nothing could have evolved for
language: after all, cultural evolution intrinsically depends on
the cognitive abilities of its learners. I propose that we should
investigate 'continuously variable cognitive traits that are used
in language'. These cognitive abilities can optimize under
pressure from the cultural environment, and are therefore
suitable candidates for empirical investigation into cognitive
adaptations related to language.
In the current thesis I aim to uncover two potential cognitive
adaptations for learning language-like sound categories, by
investigating whether learning and producing such categories
show indications of optimization related to speech.
In my experiments, human participants are trained on
artificial sound categories. I manipulate the distribution of the
sounds in the training set to examine whether learners
produce even more extreme categories than what they have
observed (Experiment 1), as well as the type of signal that
they have to learn (Experiment 2) to examine whether
learning and production is affected by signal and modality
type. I also investigate the role of functional pressure in
learning, producing and communicating signals (Experiment 3).
Although the experiments are unable to provide unequivocal
evidence for language related cognitive adaptations, the
current line of investigation is able to provide a relevant
contribution to the discussion of whether and how cognitive
adaptations can or should be investigated.
Original languageEnglish
QualificationDoctor of Sciences
Awarding Institution
  • Vrije Universiteit Brussel
Supervisors/Advisors
  • De Boer, Bart, Supervisor
Award date7 Sep 2021
Place of PublicationBrussels
Publication statusUnpublished - 2021

Keywords

  • Evolution of Speech

Fingerprint

Dive into the research topics of 'Investigating Cognitive and Functional Biases for Learning Acoustic Categories'. Together they form a unique fingerprint.

Cite this