Auditory CognitionWe are very happy to announce that PhD student Antje Strauß got her paper
Alpha and theta brain oscillations index dissociable processes in spoken word recognition
accepted at NeuroImage. Congratulations! Find her paper here.
We are happy that our paper
A Sparse Neural Code for Some Speech Sounds but Not for Others
is scheduled for publication in PLOS ONE on July 16th, 2012.
This is also our first paper in collaboration with Alexandra Bendixen from the University of Leipzig.
The research reported in this article provides an extension of the predictive coding framework onto speech sounds and assumes that auditory processing uses predictions that are not only derived from ongoing contextual updates, but also from long-term memory representations — neural codes — of speech sounds. Using the German minimal pair [lats]/[laks] (bib/salmon) in a passive-oddball design, we find the expected Mismatch Negativity (MMN) asymmetry that is compatible with a predictive coding framework, but also with linguistic underspecification theory.
[Update]Paper is available here.
We are happy to announce that our paper “Asymmetries in the processing of vowel height” will be appearing in the Journal of Speech, Language, & Hearing Research, authored by Philip Monahan, William Idsardi and Mathias Scharinger. A short summary is given below:
[Update]Purpose: Speech perception can be described as the transformation of continuous acoustic information into discrete memory representations. Therefore, research on neural representations of speech sounds is particularly important for a better understanding of this transformation. Speech perception models make specific assumptions regarding the representation of mid vowels (e.g., [
]) that are articulated with a neutral position in regard to height. One hypothesis is that their representation is less specific than the representation of vowels with a more specific position (e.g., [æ]).
Method: In a magnetoencephalography study, we tested the underspecification of mid vowel in American English. Using a mismatch negativity (MMN) paradigm, mid and low lax vowels ([
Results: Larger MMNs occurred when the mid vowel [
Conclusion: We take our results to reflect an abstract long-term representation of vowels that do not include redundant specifications at very early stages of processing the speech signal. Moreover, the dipole locations indicate extraction of distinctive features and their mapping onto representationally faithful cortical locations (i.e., a feature map).
The paper is available here.
Phonetic cues instantaneously mapped onto dialectal categories appear to be extracted at early moments in auditory speech perception, as we try to show in our paper
You had me at “Hello”: Rapid extraction of dialect information from spoken words
to appear in NeuroImage (Mathias Scharinger, Philip Monahan, William Idsardi).
In a modified passive oddball design, we compare the Mismatch Negativity (MMN) to deviants in one American English dialect (Standard American English or African-American Vernacular English) to the standards of the respective other dialect. In a control condition, deviants within the same dialects have the same averaged acoustic distance to their standards than the cross-dialectal averaged acoustic distance. Standards and deviants were always spoken exemplars of ‘Hello’ in both dialects (ca. 500 ms). MMN effects are significant in the cross-dialectal condition only, implying that a pure acoustic standard-deviant distance is not sufficient to elicit substantial mismatch effects. We interpret these findings, together with N1m source localization data, as evidence for a rapid extraction of dialect information via salient acoustic-phonetic cues. From the location and orientation of the N1m source activity, we can infer that dialect switches from standards to deviants engage areas in superior temporal sulcus/gyrus.
We are glad to announce that our paper (Mathias Scharinger, Samantha Poe, & William Idsardi) on cortical representations of Turkish vowels is in press in Journal of Cognitive Neuroscience. In this paper, we extend previous methods of obtaining centers of cortical activity evoked by vowel exemplars (e.g. Obleser et al., 2003, on German) and provide an N1m ECD (Equivalent Current Dipole) map of the entire vowel space of Turkish. Intriguingly, ECD locations mapped nearly perfect to locations in F2/F1 space, although our model comparison suggested that inclusion of discrete feature based predictors for both locations as well as collocations of vowels in auditory cortex improves the model fits substantially. We discuss the findings on the background of neural coding schemes for speech-related auditory categories.
UPDATE: Paper is available here.
A new paper is about to appear in Neuroimage on the interaction of syntactic complexity and acoustic degradation.
It is written by myself, PhD student Lars Meyer, and Angela Friederici. In a way, the paper brings together one of Angela’s main research questions (which brain circuits mediate the processing of syntax?) with a long-standing interest of mine, that is, how do adverse listening situations affect the comprehension of speech.
The paper is entitled
Dynamic assignment of neural resources in auditory comprehension of complex sentences
The paper first establishes that acoustic variants of increasingly complex sentences essentially behave like written versions of these sentences.
The data then neatly show that processing challenging (but legal) syntax under various levels of degradation has a very different effect on the neural circuits involved than profiting from semantics: While the latter has been shown previously to involve more widespread, heteromodal brain areas, the double demand of increasingly complex syntax and an increasingly degraded speech signal (from which the complex syntax has to be parsed) elicit an “upstream” shift of activation back to less abstract processing areas in the superior temporal and prefrontal/frontal cortex.
We tentatively have termed this process “upstream delegation”. We have also tried and established a slightly unusual method to do justice to the fMRI activation data: We have included all z‑scores gathered along certain spatial dimensions, irrespective of whether they were sub- or suprathreshold, and have treated them as distributions. Check it out and let us know what you think.
This went online just a day before Christmas:
Neuromagnetic evidence for a featural distinction of English consonants: Sensor- and source-space data
by Mathias Scharinger, Jennifer Merickel, Joshua Riley, and William Idsardi
http://dx.doi.org/10.1016/j.bandl.2010.11.002
We wanted to look at featural (categorical) place of articulation distinctions in English consonants, and selected labial and coronal fricatives and glides for an MMN study. In this study, we looked at sensor- and source-space effects of labial deviants preceded by coronal standards and coronal deviants preceded by labial standards, across the two manners of articulation, i.e. fricatives and glides. Note that there are rather dramatic acoustic differences between these manners of articulation: uncorrelated noise through narrow constriction vs. vowel-like sound with typical resonance frequencies. We found consistent place-of-articulation effects, independent of manner of articulation: labial deviants produced larger MMN, contra a directional hypothesis of underspecification, and dipole source locations followed the Obleser-gradient in that labials elicited N1m dipoles anterior to dipoles of coronals in auditory cortex.
Late 2010 was particularly good to us:
Multiple brain signatures of integration in the comprehension of degraded speech
by Jonas Obleser and Sonja Kotz, in NeuroImage.
The final pdf will hopefully be available online very soon. Meanwhile the figure below captures our main results:
