Categories
Auditory Neuroscience Auditory Perception Degraded Acoustics Executive Functions Papers Publications

Lis­ten­ing: The strat­e­gy mat­ters [Update]

In press on Neuropsychologia

Thal­a­m­ic and pari­etal brain mor­phol­o­gy pre­dicts audi­to­ry cat­e­go­ry learning

figure_mat

Cat­e­go­riz­ing sounds is vital for adap­tive human behav­ior. Accord­ing­ly, chang­ing lis­ten­ing sit­u­a­tions (exter­nal noise, but also periph­er­al hear­ing loss in aging) require lis­ten­ers to flex­i­bly adjust their cat­e­go­riza­tion strate­gies, e.g., switch amongst avail­able acoustic cues. How­ev­er, lis­ten­ers dif­fer con­sid­er­ably in these adap­tive capa­bil­i­ties. For this rea­son, we employed vox­el-based mor­phom­e­try (VBM) in our study (Neu­ropsy­cholo­gia, In press), in order to assess the degree to which indi­vid­ual brain mor­phol­o­gy is pre­dic­tive of such adap­tive lis­ten­ing behavior.

Ref­er­ences

  • Scharinger M1, Hen­ry MJ2, Erb J2, Mey­er L3, Obleser J2. Thal­a­m­ic and pari­etal brain mor­phol­o­gy pre­dicts audi­to­ry cat­e­go­ry learn­ing. Neu­ropsy­cholo­gia. 2014 Jan;53:75–83. PMID: 24035788. [Open with Read]
Categories
Papers Perception Publications

New Paper in press — Scharinger, Hen­ry, Obleser, in Mem­o­ry & Cognition

For nor­mal-hear­ing humans, cat­e­go­riz­ing com­plex acoustic stim­uli is a seem­ing­ly effort­less process, even if one has nev­er heard the par­tic­u­lar sounds before. Nev­er­the­less, pri­or expe­ri­ence with spe­cif­ic cor­re­la­tions between acoustic stim­u­lus prop­er­ties affects the cat­e­go­riza­tion in a ben­e­fi­cial way, as we show in our paper:

Pri­or expe­ri­ence with neg­a­tive spec­tral cor­re­la­tions pro­motes infor­ma­tion inte­gra­tion dur­ing audi­to­ry cat­e­go­ry learning

(by Math­ias Scharinger, Mol­ly Hen­ry, and Jonas Obleser).

The arti­cle is in press at Mem­o­ry & Cog­ni­tion (avail­able online). Our main find­ing is that stim­uli dif­fer­ing in the loca­tion of two spec­tral peaks were bet­ter cat­e­go­rized if there was a neg­a­tive cor­re­la­tion between the two spec­tral peaks than if there was a pos­i­tive cor­re­la­tion. Since neg­a­tive spec­tral cor­re­la­tions char­ac­ter­ize pho­net­ic speech prop­er­ties, our find­ings sug­gest that short-term audi­to­ry cat­e­go­ry learn­ing is influ­enced by long-term rep­re­sen­ta­tions of abstract acoustic-pho­net­ic prop­er­ties (here: spec­tral correlations).

Ref­er­ences

  • Scharinger M, Hen­ry MJ, Obleser J. Pri­or expe­ri­ence with neg­a­tive spec­tral cor­re­la­tions pro­motes infor­ma­tion inte­gra­tion dur­ing audi­to­ry cat­e­go­ry learn­ing. Mem Cog­nit. 2013 Jul;41(5):752–68. PMID: 23354998. [Open with Read]
Categories
EEG / MEG Evoked Activity Linguistics Papers Perception Place of Articulation Features Publications Speech

New paper in press — Scharinger et al., PLOS ONE [Update]

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.

Ref­er­ences

  • Scharinger M, Ben­dix­en A, Tru­jil­lo-Bar­reto NJ, Obleser J. A sparse neur­al code for some speech sounds but not for oth­ers. PLoS One. 2012;7(7):e40953. PMID: 22815876. [Open with Read]
Categories
Auditory Perception EEG / MEG Events Evoked Activity Posters Publications Speech

Poster Pre­sen­ta­tions at SFN

There will be two poster pre­sen­ta­tions at SFN in Wash­ing­ton, DC., on the top­ic of audi­to­ry pre­dic­tions in speech per­cep­tion. The first poster, authored by Alexan­dra Ben­dix­en, Math­ias Scharinger, and Jonas Obleser, sum­ma­rizes as follows:

Speech sig­nals are often com­pro­mised by dis­rup­tions orig­i­nat­ing from exter­nal (e.g., mask­ing noise) or inter­nal (e.g., slug­gish artic­u­la­tion) sources. Speech com­pre­hen­sion thus entails detect­ing and replac­ing miss­ing infor­ma­tion based on pre­dic­tive and restora­tive mech­a­nisms. The nature of the under­ly­ing neur­al mech­a­nisms is not yet well under­stood. In the present study, we inves­ti­gat­ed the detec­tion of miss­ing infor­ma­tion by occa­sion­al­ly omit­ting the final con­so­nants of the Ger­man words “Lachs” (salmon) or “Latz” (bib), result­ing in the syl­la­ble “La” (no seman­tic mean­ing). In three dif­fer­ent con­di­tions, stim­u­lus pre­sen­ta­tion was set up so that sub­jects expect­ed only the word “Lachs” (con­di­tion 1), only the word “Latz” (con­di­tion 2), or the words “Lachs” or “Latz” with equal prob­a­bil­i­ty (con­di­tion 3). Thus essen­tial­ly, the final seg­ment was pre­dictable in con­di­tions 1 and 2, but unpre­dictable in con­di­tion 3. Stim­uli were pre­sent­ed out­side the focus of atten­tion while sub­jects were watch­ing a silent video. Brain respons­es were mea­sured with mul­ti-chan­nel elec­troen­cephalo­gram (EEG) record­ings. In all con­di­tions, an omis­sion response was elicit­ed from 125 to 165 ms after the expect­ed onset of the final seg­ment. The omis­sion response shared char­ac­ter­is­tics of the omis­sion mis­match neg­a­tiv­i­ty (MMN) with gen­er­a­tors in audi­to­ry cor­ti­cal areas. Crit­i­cal­ly, the omis­sion response was enhanced in ampli­tude in the two pre­dictable con­di­tions (1, 2) com­pared to the unpre­dictable con­di­tion (3). Vio­lat­ing a strong pre­dic­tion thus elicit­ed a more pro­nounced omis­sion response. Con­sis­tent with a pre­dic­tive cod­ing account, the pro­cess­ing of miss­ing lin­guis­tic infor­ma­tion appears to be mod­u­lat­ed by pre­dic­tive context.

The sec­ond poster looks at sim­i­lar mate­r­i­al, but con­trasts coro­nal [t] with dor­sal [k], yield­ing inter­est­ing asym­me­tries in MMN responses:

Research in audi­to­ry neu­ro­science has lead to a bet­ter under­stand­ing of the neur­al bases of speech per­cep­tion, but the rep­re­sen­ta­tion­al nature of speech sounds with­in words is still a mat­ter of debate. Elec­tro­phys­i­o­log­i­cal research on sin­gle speech sounds pro­vid­ed evi­dence for abstract rep­re­sen­ta­tion­al units that com­prise infor­ma­tion about both acoustic struc­ture and artic­u­la­tor con­fig­u­ra­tion (Phillips et al., 2000), there­by refer­ring to phono­log­i­cal cat­e­gories. Here, we test the pro­cess­ing of word-final con­so­nants dif­fer­ing in their place of artic­u­la­tion (coro­nal [ts] vs. dor­sal [ks]) and acoustic struc­ture, as seen in the time-vary­ing for­mant (res­o­nance) fre­quen­cies. The respec­tive con­so­nants dis­tin­guish between the Ger­man nouns Latz (bib) and Lachs (salmon), record­ed from a female native speak­er. Ini­tial con­so­nant-vow­el sequences were aver­aged across the two nouns in order to avoid coar­tic­u­la­to­ry cues before the release of the con­so­nants. Latz and Lachs served as stan­dard and deviant in a pas­sive odd­ball par­a­digm, while the EEG from 20 par­tic­i­pants was record­ed. The change from stan­dard [ts] to deviant [ks] and vice ver­sa was accom­pa­nied by a dis­cernible Mis­match Neg­a­tiv­i­ty (MMN) response (Näätä­nen et al., 2007). This response showed an intrigu­ing asym­me­try, as seen in a main effect con­di­tion (deviant Latz vs. deviant Lachs, F(1,1920) = 291.84, p < 0.001) of an omnibus mixed-effect mod­el. Cru­cial­ly, the MMN for the deviant Latz was on aver­age more neg­a­tive than the MMN for the deviant Lachs from 135 to 185 ms post deviance onset (p < 0.001). We inter­pret these find­ings as reflect­ing a dif­fer­ence in phono­log­i­cal speci­fici­ty: Fol­low­ing Eulitz and Lahiri, 2004, we assume coro­nal seg­ments ([ts]) to have less spe­cif­ic (‘fea­t­u­ral­ly under­spec­i­fied’) rep­re­sen­ta­tions than dor­sal seg­ments ([ks]). While in stan­dard posi­tion, Lachs acti­vat­ed a mem­o­ry trace with a more spe­cif­ic final con­so­nant for which the deviant pro­vid­ed a stronger mis­match than vice ver­sa, i.e. when Latz acti­vat­ed a mem­o­ry trace with a less spe­cif­ic final con­so­nant. Our results sup­port a mod­el of speech per­cep­tion where sen­so­ry infor­ma­tion is processed in terms of dis­crete units inde­pen­dent of high­er lex­i­cal prop­er­ties, as the asym­me­try can­not be explained by dif­fer­ences in lex­i­cal sur­face fre­quen­cies between Latz and Lachs (both log-fre­quen­cies of 0.69). We can also rule out a fre­quen­cy effect on the seg­men­tal lev­el. Thus, it appears that speech per­cep­tion involves a lev­el of pro­cess­ing where indi­vid­ual seg­men­tal rep­re­sen­ta­tions with­in words are evaluated.

Categories
Auditory Perception Auditory Speech Processing EEG / MEG Evoked Activity Linguistics Papers Place of Articulation Features Publications Speech

New paper out in Jour­nal of Speech, Lan­guage, & Hear­ing Research [Update]

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:

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., [{varepsilon}]) 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 ([{varepsilon}]/[æ]), and high and low lax vow­els ([I]/[æ]), were opposed, and M100/N1 dipole source para­me­ters as well as MMN laten­cy and ampli­tude were examined.

Results: Larg­er MMNs occurred when the mid vow­el [{varepsilon}] was a deviant to the stan­dard [æ], a result con­sis­tent with less spe­cif­ic rep­re­sen­ta­tions for mid vow­els. MMNs of equal mag­ni­tude were elicit­ed in the high–low com­par­i­son, con­sis­tent with more spe­cif­ic rep­re­sen­ta­tions for both high and low vow­els. M100 dipole loca­tions sup­port ear­ly vow­el cat­e­go­riza­tion on the basis of lin­guis­ti­cal­ly rel­e­vant acoustic–phonetic features.

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).

[Update]

The paper is avail­able here.

Ref­er­ences

  • Scharinger M, Mon­a­han PJ, Idsar­di WJ. Asym­me­tries in the pro­cess­ing of vow­el height. J Speech Lang Hear Res. 2012 Jun;55(3):903–18. PMID: 22232394. [Open with Read]
Categories
Auditory Cortex EEG / MEG Evoked Activity Linguistics Papers Publications

New Paper out: HELLO? in press (Neu­roIm­age)

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.

Ref­er­ences

  • Scharinger M, Mon­a­han PJ, Idsar­di WJ. You had me at “Hel­lo”: Rapid extrac­tion of dialect infor­ma­tion from spo­ken words. Neu­roim­age. 2011 Jun 15;56(4):2329–38. PMID: 21511041. [Open with Read]
Categories
Auditory Cortex EEG / MEG Evoked Activity Papers Place of Articulation Features Publications

New Paper out: Com­pre­hen­sive map of a language’s vow­el space

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.

Fig­ure 1: Loca­tions of Turk­ish vow­el stim­uli in acoustic space (F1,F2, top pan­el) and N1m ECD loca­tions in cor­ti­cal space (lat­er­al-medi­al/an­te­ri­or-pos­te­ri­or/in­fe­ri­or-supe­ri­or, bot­tom panel).

UPDATE: Paper is avail­able here.

Ref­er­ences

  • Scharinger M, Idsar­di WJ, Poe S. A com­pre­hen­sive three-dimen­sion­al cor­ti­cal map of vow­el space. J Cogn Neu­rosci. 2011 Dec;23(12):3972–82. PMID: 21568638. [Open with Read]
Categories
Auditory Cortex Auditory Speech Processing EEG / MEG Evoked Activity Linguistics Papers Place of Articulation Features Publications

Paper in press: Are labi­als special?

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.

Ref­er­ences

  • Scharinger M, Mer­ick­el J, Riley J, Idsar­di WJ. Neu­ro­mag­net­ic evi­dence for a fea­t­ur­al dis­tinc­tion of Eng­lish con­so­nants: sen­sor- and source-space data. Brain Lang. 2011 Feb;116(2):71–82. PMID: 21185073. [Open with Read]