Neural mechanisms of abstract rule changes in speech : exploring phonologic and attentional constraints
Loading...
Document Type
Document Version
Author
Director
Toro Soto, Juan Manuel
Tutor
Other authors
Publication Date
Pages
157 p.
Embargo date
Citation
Monte Ordoño, J. Neural mechanisms of abstract rule changes in speech : exploring phonologic and attentional constraints. Universitat Pompeu Fabra; 2018. handle: http://hdl.handle.net/10803/664264
This citation was generated automatically.
Citation
Doctoral program
Programa de doctorat en Biomedicina
Abstract
L’extracció de regles de la parla és crucial per a l’adquisició del llenguatge. La present dissertació estudia el mecanisme d’aprenentatge de regles explorant com el cervell detecta regularitats rellevants dins del senyal lingüístic. Per tal de donar resposta a aquesta pregunta he seguit tres línies de recerca. En primer lloc m’he centrat en la detecció de canvis, tant superficials com estructurals, de les regles lingüístiques. Mitjançant l’estudi dels potencials evocats, aquesta primera línia explora les respostes neuronals desencadenades després d’una violació estructural. En segon lloc, he estudiat els efectes de la manipulació fonètica amb la intenció de descobrir si les respostes neuronals associades a l’aprenentatge de regles varien quan aquestes s’implementen sobre les vocals o sobre les consonants. És a dir, m’he centrat en avaluar com les diverses categories fonètiques poden donar lloc a respostes neuronals diverses. En tercer lloc, he explorat la detecció de regles en un context d’aprenentatge heterogeni per tal d’observar com poden ser descobertes les regles abstractes dins d’un senyal sorollós. En conjunt, els resultats obtinguts mostren que la manipulació d’ambdós factors, tant de les pistes fonològiques com del context d’aprenentatge, modula el procés d’extracció de regles. Més específicament, aquestes manipulacions podrien alterar les fonts d’informació que es prioritzen durant el processament de la parla. Finalment, la presència d’una pista diferenciadora del senyal (com les diferències en la freqüència d’aparició de diverses regles) podria facilitar el processament de múltiples sistemes estructurals dins d’un input lingüístic.
The extraction of abstract rules from speech is paramount for language acquisition. The present dissertation explores the processing of linguistic rules by studying how our brain discovers the relevant abstract regularities in the signal. In order to tackle this question I followed three lines of research. First I focused on the detection of surface and structural changes of speech rules that I explored using an ERP approach. The objective was to understand the neural responses that are triggered after abstract rule violations in speech. Second, I studied the effects of the phoneme manipulations. The aim was to discover whether the ERP signatures linked to rule learning differ when the target regularity is implemented over consonants or over vowels. That is, I focused on exploring how different phonetic categories might trigger different neural responses to rule violations. And third, I explored the detection of rules from a heterogeneous context studying how abstract rules might be discovered over a noisy signal. Overall, the results we observed suggest that the manipulation of both the phonologic cues and the context of learning modulate the rule extraction process. More specifically, the present dissertation shows that both the task presented to the listeners and the phonemic cues present in the signal affect the selection of relevant sources of information from the speech. Even more, the experiments reported here show that the presence of a clear differentiating cue in the signal (such as the frequency unbalance across rules), might enhance the processing of different rule systems from the speech input.
The extraction of abstract rules from speech is paramount for language acquisition. The present dissertation explores the processing of linguistic rules by studying how our brain discovers the relevant abstract regularities in the signal. In order to tackle this question I followed three lines of research. First I focused on the detection of surface and structural changes of speech rules that I explored using an ERP approach. The objective was to understand the neural responses that are triggered after abstract rule violations in speech. Second, I studied the effects of the phoneme manipulations. The aim was to discover whether the ERP signatures linked to rule learning differ when the target regularity is implemented over consonants or over vowels. That is, I focused on exploring how different phonetic categories might trigger different neural responses to rule violations. And third, I explored the detection of rules from a heterogeneous context studying how abstract rules might be discovered over a noisy signal. Overall, the results we observed suggest that the manipulation of both the phonologic cues and the context of learning modulate the rule extraction process. More specifically, the present dissertation shows that both the task presented to the listeners and the phonemic cues present in the signal affect the selection of relevant sources of information from the speech. Even more, the experiments reported here show that the presence of a clear differentiating cue in the signal (such as the frequency unbalance across rules), might enhance the processing of different rule systems from the speech input.
Keywords
Aprenentatge de regles, Consonants, Vocals, Potencials evocats, N400, Rule learning, Consonants, Vowels, ERP
Subjects
80
Publisher
Universitat Pompeu Fabra
