Brain anatomical correlates of perceptual phonological proficiency and language learning aptitude

Abstract: The present dissertation concerns how brain tissue properties reflect proficiency in two aspects of language use: the ability to use tonal cues on word stems to predict how words will end and the aptitude for learning foreign languages. While it is known that people differ in their language abilities and that damage to brain tissue cause loss of cognitive functions, it is largely unknown if differences in language proficiencies correlate with differences in brain structure. The first two studies examine correlations between cortical morphometry, i.e. the thickness and surface area of the cortex, and the degree of dependency on word accents for processing upcoming suffixes in Swedish native speakers. Word accents in Swedish facilitate speech processing by having predictive associations to specific suffixes, (e.g. fläckaccent1+en ‘spot+singular’, fläckaccent2+ar ‘spot+plural’). This use of word accents, as phonological cues to inflectional suffixes, is relatively unique among the world’s languages. How much a speaker depends on word accents in speech processing can be measured as the difference in response time (RT) between valid and invalid word accent-suffix combinations when asked to identify the inflected form of a word. This can be thought of as a measure of perceptual phonological proficiency in native speakers. Perceptual phonological proficiency is otherwise very difficult to study, as most phonological contrasts are mandatory to properly interpret the meaning of utterances. Study I compares the cortical morphometrical correlates in the planum temporale and inferior frontal gyrus pars opercularis in relation to RT differences in tasks involving real words and pseudowords. We found that thickness of the left planum temporale correlates with perceptual phonological proficiency in lexical words but not pseudowords. This could implicate that word accents are part of full-form representations of familiar words. Moreover, for pseudowords but not lexical words, the thickness of the inferior frontal gyrus pars opercularis correlates with perceptual phonological proficiency. This association could reflect a greater importance for decompositional analysis in which word accents are part of a set of rules listeners need to rely on during processing of novel words. In study II, the investigation of the association between perceptual phonological proficiency in real words with cortical morphometry is expanded to the entire brain. Results show that cortical thickness and surface area of anterior temporal lobe areas, known constituents of a ventral sound-to-meaning language-processing stream is associated with greater perceptual phonological proficiency. This is consistent with a role for word accents in aiding putting together the meaning of or accessing a whole word representation of an inflected word form. Studies III and IV investigate the cortical morphometric associations with language learning aptitude. Findings in study III suggest that aptitude for grammatical inferencing, i.e. the ability to analytically discern the rules of a language, is associated with cortical thickness in the left inferior frontal gyrus pars triangularis. Furthermore, pitch discrimination proficiency, a skill related to language learning ability, correlates negatively with cortical thickness in the right homologue area. Moreover, study IV, using improved imaging techniques, reports on a correlation between vocabulary learning aptitude and cortical surface area in the left inferior precuneus as well as a negative correlation between diffusional axial kurtosis and phonetic memory in the left arcuate fasciculus and subsegment III of the superior longitudinal fasciculus. However, the finding correlation between cortical thickness and grammatical inferencing skill from study III was not replicated in study IV.Taken together, the present dissertation shows that differences in some language proficiencies are associated with regionally thicker or larger cortex and more coherent white matter tracts, the nature and spatial locus of which depend on the proficiency studied. The studies add to our understanding of how language proficiencies are represented in the brain’s anatomy.