Sprog Og Lateralisering I Hjernen

Transcript

Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital SPROG OG LATERALISERING I HJERNEN Andreas Højlund Nielsen Folkeuniversitetet, Odense, 12. maj 2015 Interacting Minds Centre Aarhus University au [email protected] AARHUS UNIVERSITET Dept. of Linguistics Aarhus University HJERNEFORSKER - ER DU SÅ LÆGE? Født 1983 Bachelor i lingvistik 2008 Cand.mag. i kognitiv semiotik 2011 (Næsten) ph.d. i kognitiv neurovidenskab (27. maj) 2015 Postdoc ved Aarhus Universitet (Parkinson, DBS, sprog) 2015Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE AFTENENS PROGRAM Andreas snakker (ca. 45 min) Pause (ca. 15 min) Andreas snakker (ca. 35 min) Spørgsmål (5-10 min) Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE HJERNENS GÅDE #1 Hvor sidder sproget i hjernen? Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE AFASI NÅR SPROGET I HJERNEN GÅR I STYKKER Fra græsk: ‘uden tale’ (a- = ‘uden’, phasis = ‘tale’, aphatos = ‘målløs’) dvs. ‘nedsat evne til at bruge (eller forstå) sproget’ Oftest pga. stroke (blodprop eller blødning i hjernen) hjernesvulst traume (slag mod hovedet) visse typer demens Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE PAUL BROCA (1824-1880) Patient Leborgne - også kendt som “Tan” Brocas post-mortemundersøgelser af “Tan” førte til beskrivelsen af ‘Brocas afasi’ Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE EKSEMPEL: BROCAS AFASI Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE AFASI EFTER STROKE (BLODPROP ELLER BLØDNING I HJERNEN) Aphasia in Acute Stroke: Incidence, Mild Moderate Severe and Recovery Determinants, Aphasia p Value Table I . Basic Patient Characteristics in Reiation to Aphasia Severity No N (incidence) Age (yr) (SD) Sex, male (9%) Aphasia Aphasia Aphasia 551 Pedersen, (62.5%) MA,"101 (11.5%) 56 MD," (6.4%) (19.6%) MD," Palle Mgjller Henrik Stig Jgjrgensen, Hirofumi173 Nakayama, 73.1Hans ( 1 1 . 5Otto ) 76.5 (9.5) (9.5) 77.1 (9.4) <0.0001 Raaschou, MD,T and Tom75.8 Skyhgj Olsen, MD, PhDX 48% 29% 45% 0.04 48% Handedness, right (%) 93% 92% 98% 94% NS Knowledge of the frequency and remission of aphasia is essential for the rehabilitation of stroke patients and proSide of stroke lesion, left (96) 37% 93% 89% 87 % <0.00001 vides insight into the brain organization of language. We studied prospectively and consecutively an unselected and Mortality ($%) community-based sample 18% <0.00001 10%of 881 patients 10% with acute stroke. Assessment of aphasia47% was done at admission, weekly Prior stroke (%) 26% score of the Scandinavian 36% 0.0004 20% during the hospital stay, and at a 6-month26% follow-up using the aphasia Stroke Scale. ThirtyComorbidity (%) 21% at the time of14% 25%18% had aphasia. 27% NS eight percent had aphasia admission; at discharge Sex was not a determinant of aphasia no (12.1) sex difference 41.8 in the(9.7) anterior-posterior of lesions was found. The remission SSS o n admission (SD)in stroke, and43.9 33.5distribution (11.6) <0.0001 15.5 (11.2) 95% was reached those with initial mild aphasia, curve was(SD) steep: Stationary language function S S S excluding language 29.2 (10.4) 31.2 in (9.5) 28.0within (10.6)2 weeks in 15.0 (10.7) <0.0001 within 6 weeks in those with moderate, and within 10 weeks in those with severe aphasia. A valid prognosis of aphasia BI on admission (SD) 61.6 (38.9) 63.6 (37.0) (38.4) 16.1of(30.3) <0.0001 could be made within 1 to 4 weeks after the stroke depending44.3 on the initial severity aphasia. Initial severity of aphasia wasScale; the only clinicallylanguage relevant=predictor of aphasia outcome. Sex, and handedness, side BI of stroke lesion were SSS = Scandinavian Stroke SSS excludirig SSS on admission excluding aphasia orientationand scores; = Barthei index; not independent outcome predictors, and the influence of age was minimal. SD = standard deviation; NS = not significant. Pedersen PM, J@rgensenHS, Nakayama H, Raaschou HO, Olsen TS. Aphasia in acute stroke: incidence, determinants, and recovery. Ann Neurol 1995;38:659-666 Center of Functionally Integrative Neuroscience Aphasia is a common symptom in stroke. It is considAarhus University/ Aarhus University Hospital ered a major disability by sufferers and relatives, and knowledge of prognosis and the time course of remission is important for the planning of rehabilitation and for informing patient and family. The remission of aphasia is known to take place mainly within the first Interacting Minds Centre Aarhus University the stroke. Eighty-eight percent of all stroke patients are [email protected] pitalized in the Copenhagen area All patients were transDept. of[8]. Linguistics Aarhus ferred on acute admission toUniversity the same 60-bed stroke unit, where all stages of acute care, work-up, and rehabilitation AARHUS rook place. au INCLUSION AND EXCLUSION CRITERIA. A total of 1,014 pa- UNIVE ENDNU ET EKSEMPEL: BROCAS AFASI Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE A TALE OF I B Y D AV I D W O L M A N SPLIT-BRAIN-PATIENTER n the first months after her surgery, shopping for groceries was infuriating. Standing in the supermarket aisle, Vicki would look at an item on the shelf and know that she wanted to place it in her trolley — but she couldn’t. “I’d reach with my right for the thing I wanted, but the left would come in and they’d kind of fight,” she says. “Almost like repelling magnets.” Picking out food for the week was a two-, sometimes three-hour ordeal. Getting dressed posed a similar challenge: Vicki couldn’t reconcile what she wanted to put on with what her hands were doing. Sometimes she ended up wearing three outfits at once. “I’d have to dump all the clothes on the bed, catch my breath and start again.” In one crucial way, however, Vicki was better than her pre-surgery self. She was no longer racked by epileptic seizures that were so severe they had made her life close to unbearable. She once collapsed onto the bar of an old-fashioned oven, burning and scarring her back. “I really just couldn’t function,” she says. When, in 1978, her neurologist told Since 1960s, researchers been scrutinizing handful herthe about a radical but have dangerous surgery athat might help, she barely of patients who underwent a radical kind of brain surgery. hesitated. If the worst were to happen, she knew that her parents would The cohort has been a boon to neuroscience — but soon it will be gone. take care of her young daughter. “But of course I worried,” she says. “When you get your brain split, it doesn’t grow back together.” In June 1979, in a procedure that lasted nearly 10 hours, doctors created a firebreak to contain Vicki’s seizures by slicing through her corpus callosum, the bundle of neuronal fibres connecting the two sides of her brain. This drastic procedure, called a corpus callosotomy, disconnects the two sides of the neocortex, the home of language, conscious B Y D AV I D W O L M A N thought and movement control. Vicki’s supermarket predicament was B Y D AV I D W O L M A N e the 1960s, researchers have been scrutinizing a handful patients who underwent a radical kind of brain surgery. rt has been a boon to neuroscience — but soon it will be gone. A TALE OF I 6 0her surgery, | N Ashopping T U RforEgroceries | VO n the first months2after wasL infuriating. Standing in the supermarket aisle, Vicki would look at an item on the shelf and know that she wanted to place it in her trolley — but she couldn’t. “I’d reach with my right for the thing I wanted, but the left would come in and they’d kind of fight,” she says. “Almost like repelling magnets.” Picking out food for the week was a two-, sometimes three-hour ordeal. Getting dressed posed a similar challenge: Vicki couldn’t reconcile what she wanted to put on with what her hands were doing. Sometimes she ended up wearing three outfits at once. “I’d have to dump all the clothes on the bed, catch my breath and start again.” In one crucial way, however, Vicki was better than her pre-surgery self. She was no longer racked by epileptic seizures that were so severe they had made her life close to unbearable. She once collapsed onto the bar of an old-fashioned oven, burning and scarring her back. “I really just couldn’t function,” she says. When, in 1978, her neurologist told her about a radical but dangerous surgery that might help, she barely hesitated. If the worst were to happen, she knew that her parents would take care of her young daughter. “But of course I worried,” she says. “When you get your brain split, it doesn’t grow back together.” Aarhus University/ Aarhus University Hospital In June 1979, in a procedure that lasted nearly 10 hours, doctors created a firebreak to contain Vicki’s seizures by slicing through her corpus callosum, the bundle of neuronal fibres connecting the two sides of her brain. This drastic procedure, called a corpus callosotomy, disconnects the two sides of the neocortex, the home of language, conscious thought and movement control. Vicki’s supermarket predicament was 4 the 83 | 1 5 ofM A Rthat Cbehaved H 2 in 0 some 1 2 ways as if it were two consequence a brain separate minds. © 2012 Macmillan After about a year, Vicki’s difficulties abated. “I could get things together,” she says. For the most part she was herself: slicing vegetables, tying her shoe laces, playing cards, even waterskiing. But what Vicki could never have known was that her surgery would turn her into an accidental superstar of neuroscience. She is one of fewer than a dozen ‘split-brain’ patients, whose brains and behaviours have been subject to countless hours of experiments, hundreds of scientific papers, and references in just about every psychology textbook of the past generation. And now their numbers are dwindling. Through studies of this group, neuroscientists now know that the healthy brain can look like two markedly different machines, cabled together and exchanging a torrent of data. But when the primary cable is severed, information — a word, an object, a picture — presented to one hemisphere goes unnoticed in the other. Michael Gazzaniga, a cognitive neuroscientist at the University of California, Santa Barbara, and the godfather of modern split-brain science, says that even after working with these patients for five decades, he still finds it thrilling to observe the disconnection effects first-hand. “You see a split-brain patient just doing a standard thing — you show NATURE.COM him an image and he can’t say what it is. But he To hear more about can pull that same object out of a grab-bag,” Gazsplit-brain patients, zaniga says. “Your heart just races!” visit: Work with the patients has teased out go.nature.com/knhmxk differences between the two hemispheres, the consequence of a brain that behaved in some ways as if it were two separate minds. After about a year, Vicki’s difficulties abated. “I could get things together,” she says. For the most part she was herself: slicing vegetables, tying her shoe laces, playing cards, even waterskiing. But what Vicki could never have known was that her surgery would turn her into an accidental superstar of neuroscience. She is one of fewer than a dozen ‘split-brain’ patients, whose brains and behaviours have been subject to countless hours of experiments, hundreds of scientific papers, and references in just about every psychology textbook of the past generation. And now their numbers are dwindling. Through studies of this group, neuroscientists now know that the healthy brain can look like two markedly different machines, cabled together and exchanging a torrent of data. But when the primary cable is severed, information — a word, an object, a picture — presented to one hemisphere goes unnoticed in the other. Michael Gazzaniga, a cognitive neuroscientist at the University of California, Santa Barbara, and the godfather of modern split-brain science, says that even after working with these patients for five decades, he still finds it thrilling to observe the disconnection effects first-hand. “You see a split-brain patient just doing a standard thing — you show NATURE.COM him an image and he can’t say what it is. But he To hear more about can pull that same object out of a grab-bag,” Gazsplit-brain patients, zaniga says. “Your heart just races!” visit: Work with the patients has teased out go.nature.com/knhmxk differences between the two hemispheres, Publishers Limited. All rights reserved her surgery, shopping for groceries was the consequence of a brain that behaved in some ways as if it were two the supermarket aisle, Vicki would look separate minds. nd know that she wanted to place it in her After about a year, Vicki’s difficulties abated. “I could get things n’t. “I’d reach with my right for the thing together,” she says. For the most part she was herself: slicing vegetables, ome in and they’d kind of fight,” she says. tying her shoe laces, playing cards, even waterskiing. of was Functionally Neuroscience ets.” Picking out foodCenter for the week a ButIntegrative what Vicki could never have known was that her surgery would ordeal. Getting dressed posed a similar turn her into an accidental superstar of neuroscience. She is one of oncile what she wanted to put on with fewer than a dozen ‘split-brain’ patients, whose brains and behaviours Sometimes she ended up wearing three have been subject to countless hours of experiments, hundreds of sciump all the clothes on the bed, catch my entific papers, and references in just about every psychology textbook of the past generation. And now their numbers are dwindling. er, Vicki was better than her pre-surgery Through studies of this group, neuroscientists now know that the d by epileptic seizures that were so severe healthy brain can look like two markedly different machines, cabled 2 6 0 | NAT U R E | VO L 4 8 3 | 1 5 M A RC H 2 0 1 2 © 2012 Macmillan Publishers Limited. All rights reserved Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE SPLIT-BRAIN-PATIENTER Split-brain patients have undergone surgery to cut the corpus callosum, the main bundle of neuronal fibres connecting the two sides of the brain. Input from the left field of view is processed by the right hemisphere and vice versa. Left hemisphere A word is flashed briefly to the right field of view, and the patient is asked what he saw. Visual fields Corpus callosum Right hemisphere Nothing Face Because the left hemisphere is dominant for verbal processing, the patient’s answer matches the word. Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Now a word is flashed to the left field of view, and the patient is asked what he saw. The right hemisphere cannot share information with the left, so the patient is unable to say what he saw, but he can draw it. Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE SPLIT-BRAIN-PATIENTER Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE HJERNENS GÅDE #1 Hvor sidder sproget i hjernen? SVAR: Sproget sidder primært i venstre side af hjernen Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE HJERNENS GÅDE #2 Hvorfor sidder sproget til venstre? Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE groups: mixed handers had a lower prevalence of moderate right-lateralization (4.3%) than moderate right and left handers (6.0% and 6.9% respectively). Moderate left-handers had a lower prevalence of bilateral lateralization (17.2%) than mixed and strong left-handers (26.1% and 29.8% respectively). See Fig. 1 and Table 2 for an overview of the frequency distribution. Frequencies peaked in the strong-left-handedness subgroup for all three measures. The overall frequency distribution of atypical lateralization (bilateral, moderate right- and strong right-hemispheric lateralization collapsed) also peaked in the strong left-handedness subgroup HÅNDETHED OG SPROG Brain & Language 144 (2015) 10–15 Contents lists available at ScienceDirect Brain & Language journal homepage: www.elsevier.com/locate/b&l Short Communication On the relationship between degree of hand-preference and degree of language lateralization Metten Somers a,⇑, Maartje F. Aukes a, Roel A. Ophoff a,b,c, Marco P. Boks a, Willemien Fleer a, Kees (C.) L. de Visser d, René S. Kahn a, Iris E. Sommer a a Brain Center Rudolf Magnus, Department of Psychiatry, University Medical Center Utrecht, Utrecht, The Netherlands Department of Human Genetics, David Geffen School of Medicine at UCLA, University of California, Los Angeles, CA, USA Center for Neurobehavioral Genetics, Semel Institute for Neuroscience and Human Behavior, University of California, Los Angeles, CA, USA d Department of General Practice, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands b c i n f o 1. Introduction Center of Functionally Integrative Neuroscience inspires our understanding of the genetic underpinnings of both Aarhus University/ Aarhus University Hospital Language lateralization and hand-preference can be described in terms of direction (right or left) as well as degree (strongly lateralized or more bilaterally represented (Isaacs, Barr, Nelson, & Devinsky, 2006)). It has been hypothesized that degree of handpreference mirrors degree of language lateralization, e.g. that mixed-handers have the highest prevalence of bilateral lateralization and that strong left/right-handers have the highest prevalence of strong language lateralization (Annett, 1999; Crow, Crow, Done, traits. In this study, we investigated the relation between degree of language lateralization and degree of hand-preference and tested whether hand-preference can be used as a predictor for atypical language lateralization. We enriched the data for atypically lateralized subjects, by including large families with multiple left-handers. Hand-preference was measured with the Edinburgh Handedness Inventory, language lateralization with functional Transcranial Doppler (fTCD) in a fairly large sample. nd hå re Keywords: Hand-preference Left-handedness Language lateralization Functional transcranial Doppler Asymmetry Language lateralization and hand-preference show inter-individual variation in the degree of lateralization to the left- or right, but their relation is not fully understood. Disentangling this relation could aid elucidating the mechanisms underlying these traits. The relation between degree of language lateralization and degree of hand-preference was investigated in extended pedigrees with multi-generational lefthandedness (n = 310). Language lateralization was measured with functional Transcranial Doppler, handpreference with the Edinburgh Handedness Inventory. Degree of hand-preference did not mirror degree of language lateralization. Instead, the prevalence of right-hemispheric and bilateral language lateralization rises with increasing strength of left-handedness. Degree of hand-preference does not predict degree of language lateralization, thus refuting genetic models in which one mechanism defines both hand-preference and language lateralization. Instead, our findings suggest a model in which increasing strength of left-handedness is associated with increased variation in directionality of cerebral dominance. ! 2015 Elsevier Inc. All rights reserved. nst Ve Article history: Recieved 5 September 2014 Accepted 17 March 2015 Available online 13 April 2015 a b s t r a c t Hø jre hå nd a r t i c l e Fig. 1. Degree of language lateralization vs. degree of hand-preference. The proportion of each of 5-language lateralization categories (strong left-hemispheric, Interacting Minds Centre moderate left-hemispheric, Aarhus bilateral, moderate right-hemispheric and strong rightUniversity hemispheric) is plotted against 5 categories of hand-preference (strong rightDept. of Linguistics handedness, moderate right-handedness, mixed-handedness, moderate left-handAarhus University edness, strong left-handedness). [email protected] au AARHUS UNIVE HÅNDETHED OG SPROG Downloaded from http://rstb.royalsocietypublishing.org/ on May 12, 2015 Phil. Trans. R. Soc. B (2009) 364, 881–894 doi:10.1098/rstb.2008.0235 Published online 5 December 2008 Downloaded from http://rstb.royalsocietypublishing.org/ on May 12, 2015 Review Why are some people left-handed? An Phil. Trans. R. Soc. B (2009) 364, 881–894 evolutionary perspective doi:10.1098/rstb.2008.0235 Downloaded from http://rstb.royalsocietypublishing.org/ on May 12, 2015 Published online 5 December 2008 V. Llaurens1,*, M. Raymond1 and C. Faurie1,2 Phil. Trans. R. Soc. B (2009) 364, 881–894 doi:10.1098/rstb.2008.0235 Published online 5 December 2008 1 Institut des Sciences de l’Evolution de Montpellier (UMR CNRS 5554 ), Universite´ de Montpellier II, C.C. 065, 34095 Montpellier Cedex 5, France 2 Department of Animal and PlantReview Sciences, University of Sheffield, Sheffield S10 2TN, UK Since prehistoric times, left-handed individuals have been ubiquitous in human populations, exhibiting geographical frequency variations. Evolutionary explanations have been proposed for Why are some people left-handed? An the persistence of the handedness polymorphism. Left-handedness could be favoured by negative frequency-dependent selection. Data have suggested that left-handedness, as the rare hand Interacting Minds Centre evolutionary perspective preference, could represent an important strategic advantage in fighting interactions. However, the Aarhus University fact that left-handedness occurs at a low frequency indicates that some evolutionary costs could 1,* 1 [email protected] be V. associated with left-handedness. Overall, the evolutionary dynamics of this1,2 polymorphism are Llaurens and C. Faurie , M. Raymond Review Center of Functionally Integrative Neuroscience University/ Aarhus University Hospital WhyAarhus are some people left-handed? An evolutionary perspective V. Llaurens1,*, M. Raymond1 and C. Faurie1,2 1 1 Dept. of Linguistics not fully understood. Here, we review the abundant literature available regarding the possible mechanisms and consequences of left-handedness. We point outUniversity that hand preference is heritable,II, ´ de Montpellier Institut des Sciences de l’Evolution de Montpellier (UMR CNRS 5554 ), Universite Aarhus and report how C.C. hand preference is influenced by Cedex genetic,5, hormonal, 065, 34095 Montpellier France developmental and cultural factors. We review the available information on potential fitness costs and benefits acting as selective 2 Department of Animal and Plant Sciences, University of Sheffield, Sheffield S10 2TN, UK forces on the proportion of left-handers. Thus, evolutionary perspectives on the persistence of this polymorphism in humans are gathered for thehave first time, highlighting the necessity for an assessment prehistoric times, left-handed individuals been ubiquitous in human populations, Institut des Sciences de l’Evolution de Montpellier (UMR CNRS 5554 ), Universite´ de Montpellier Since II, of fitness differences between right- and left-handers. exhibiting geographical frequency variations. Evolutionary explanations have been proposed for C.C. 065, 34095 Montpellier Cedex 5, France handedness; polymorphism; could human be favoured by negative 2 the persistence of the handedness Keywords: polymorphism. Left-handedness au AARHUS UNIVE HJERNENS GÅDE #2 Hvorfor sidder sproget til venstre? HÅNDETHED? Lateraliseringen er mangfoldig og kompliceret Ingen entydig eller samlet genetisk forklaring Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE PAUSE (15 MIN) Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE EEG (ELEKTROENCEFALOGRAFI) rå EEG-signaler Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE ERP (EVENT-RELATEREDE POTENTIALER) Råt EEG signal bib bib bib bib Gennemsnitligt ERP signal Tid (ms) Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE MISMATCH NEGATIVITY (MMN) [...sssssssssssdssssdsssssdssssdssssssdsss...] Oddball-paradigme s = standard-tone d = afvigende tone Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE MISMATCH NEGATIVITY (MMN) [...sssssssssssdssssdsssssdssssdssssssdsss...] Näätänen et al., Nature, 1997 Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE “LOKALE” DATA Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE MMN & LÆRING Forskel på finnere og estere Kendt fonem mere venstre-lateraliseret Näätänen et al., Nature, 1997 Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE MMN & LÆRING Finske og estiske børn Cheour, et al. Nature Neurosci, 1998 Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE sing, whereas in three subjects 31 the right hemisphere showed dominance.Columns the MMNm ECD JA Left 9 Oy, 49 in 39 5^ 8 indicate 46 16strengths in the 10left and right hemispheres 30 Vectorview magnetometer (4D Finland) lyses were performed on processing the basis of this division, i.e., in #6 between for the coded stimuli before and24 afterNeuromag the training dominance for Morse calculated as the di¡erence TK Left 17 course.The hemispheric 11 33 7 code 13 was #22 an MMNm acoustically and of magnetically shielded room (Euroshield each individual, the parameters the strengths the speech-MMNm dominant and non-dominant hemispheres. EK Left 51 44 27 56 29 P1m and MMNm 33 #29 were #4 Ltd, Finland). The MEG epochs for the standard and deviant compared between the speech-MMNm dominant PL Left 30 21 13 25 14 9 #12and -on- 5 stimuli were averaged online except for those standard dominant hemispheres. PH Right 21 24 46 57 30 49 11 19 stimuli that immediately followed deviant stimuli. Epochs The17effect of Morse on the 4magnitude, 21 IN Right 26 39 13 15 code acquisition 36 with electro-oculogram (EOG) or MEG signals 4 150 mV or loci, and was tested 15 TV Right 13 21 48 11 latency of3 the Morse code 18 processing#37 Spoken Morse-coded 3000 fT/cm,ofrespectively, were also excluded. repetition the stimulation after 3 months. When addiusing repeated-measures ANOVAs with factors session (1st, syllables syllables 2nd), component (P1m, MMNm), and hemisphere (speechThe hemispheric dominance for speech was determined by comparing the left and right hemisphere average of the MMNm ECD strengths (nAm) recorded tional analyses were required, ANOVAs with factors session P1m in the two sessions for the spoken-syllable contrast (left column). In four subjects the left hemisphere dominated the native-language speech-sound procesMMNm dominant, speech-MMNm non-dominant). The Source modeling: The stimulus contrasts in both spoken and hemisphere were performed for the components sing, whereas in three subjects the right hemisphere showed dominance.Columns 5^ 8 indicate the MMNm ECD strengths in the left and right hemispheres P1m and MMNm to the speech stimuli were also compared and Morse coded stimuli elicited typical MMNm responses. 20 20 separately. for the coded waves before and after training minus course.The hemispheric dominance Morse code in processing was calculated as the di¡erence between between thefor two sessions ANOVAs with the same factors Difference (responses to the deviants to To help stimuli to detect possible changes in the those hemispheric the MMNm strengths of the speech-MMNm dominantdipole and non-dominant tohemispheres. obtain possible effects of, for example, training or standards) were used in the equivalent current (ECD) Dipole moment [nAm] MMN & LÆRING SPROG OG MORSE-KODE dominance for the Morse stimuli, we calculated the difference in the MMNm magnitude for the Morse coded 10 10 COGNITIVE NEUROSCIENCE AND NEUROPSYCHOLOGY syllables between the speech-MMNm NEUROREPORT dominant and nonNEUROREPORT 16 8 4 Vol 14 No 13 15 September 2003 dominant hemispheres (i.e. the strength of the MMNm in Spoken Morse-coded repetition of the by stimulation after months. When addiPlastic cortical changes induced learning to 3hemisphere the speech-MMNm non-dominant subtracted 1st 2nd 1st 2nd Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. syllables syllables tional analyses were required, ANOVAs with factors session Session Session from that of the speech-MMNm dominant hemisphere) for communicate with non-speech sounds NEUROREPORT A. KUJALA ETAL. P1m and hemisphere(Table were1).performed for the components each individual MMNm 20 20 1 separately. Anu Kujala,1,2,CA Minna Huotilainen, Maria Uther,1,3 Yury Shtyrov,1,4 Simo Monto,1,2,5 Prior to the training, the mean Morse-MMNm was Spoken syllables 2,5 1,2,5 40lateralized to the hemisphere40 non-dominant for the Risto J. Ilmoniemi Ristoto N!!t!nen To and help detect possible changes in the hemispheric speech-MMNm. After the training, however, the MorseRESULTS for the Morse stimuli, we calculated the dominance Cognitive Brain Research Unit, Department of Psychology and Helsinki Brain Research Center, FIN- 00014 University of Helsinki, Finland; Department of MMNm became stronger in the speech-MMNm dominant 1,2,5 Psychology,University of Portsmouth, King Henry Building, King Henry1St,effect Portsmouth in PO12DY; Medical Research Council,Cognition and Brain Sciences for the Monto, The ofthe training on speech-sound processing: No difference MMNm magnitude Morse coded 10than non-dominant hemisphere 10 (Figs. 2 and 3). The Sprog Unit, CB2 2EF Cambridge, UK; BioMag Laboratory, Engineering Centre, Helsinki University Central Hospital, FIN00029 HUS, Finland the speech-MMNm statistically significant changes occurred in theand speechsyllables between the speech-MMNm dominant non30MMNm-magnitude difference between 30 Corresponding Author: anu.kujala@helsinki.¢ dominant and non-dominant hemispheres before and after MMNm or15 May speech-P1m magnitude, loci, or between dominant (i.e. the strength of latency the MMNm in 31 [nAm] 22 Received 15 April 2003; acceptedhemispheres 2003 the training highlights the change in the hemispheric the first and second sessions. The MMNm wasMorse-coded stronger than balance1stin each individual (Table 1):1st on average,2nd the speech-MMNm non-dominant hemisphere subtracted syllables nki, Finland; 3Department of 2nd DOI:the 10.1097/01.wnr.0000089132.26479.a5 difference Session before the training was#11 nAm and after Before learning Cognition and Brain Sciences Session the P1m (main effect ofdominant component F(1,6) ¼ 17.13, from thatresponse of the speech-MMNm hemisphere) for 00029With HUS,Morse Finland 20+ 11 nAm, indicating a complete 20 reversal of hemispheric code, an acoustic message is transmitted using speech sounds. After a training period of 3 months, the pattern po 0.01). lateralization for the MMNm to the Morse stimuli. Thus, each 1).code MMN became lateracombinations of tone patterns rather than the spectrally and individual reversed, however:(Table the mean Morse 1 2 Dipole moment [nAm] ning to nds 3 4 5 CA Morse før Morse code processing: The The effect of training on temporally complex speech sounds that constitute the spoken language. Using MEG recordings of the mismatch negativity (MMN, an index of permanent auditory cortical representations of native language speech sounds), we probed the dominant hemisphere for the developing Morse code representations in adult Morse code learners. Initially, the MMN to the Morse coded syllables was, on average, stronger in the hemisphere opposite to the one dominant for the MMN to native language lized to the hemisphere that was predominant for the speechsound MMN. This suggests that memory traces for the Morse coded acoustic language units develop within the hemisphere that already accommodates the permanent traces for natural speech sounds.These plastic changes manifest, presumably, the close associations formed between the neural representations of 26 the c 2003 tone patterns and phonemes. NeuroReport 14:1683^1687 ! Lippincott Williams & Wilkins. RESULTS Morse-MMNm and Morse-P1m responses were stronger in 37 MMNm was relatively stronger after the training, the Morse-MMNm the speech-MMNm dominant hemisphere than prior to 40in 40 the training even though only a small increase in the Morse1st 2nd 1st 2nd MMNm magnitude occurred in this hemisphere. Session Session With the training, the overall level of the neural activation related to Morse codeSpeech-MMNm processing considerably decreased; this was mainly due to a decrease 30 30in the MMNm magnitude Dominant hemisphere. The sum in the speech-MMNm non-dominant After learning No The effect of in training on speech-sound the first than the second session (main processing: effect of session statistically significant changes occurred in the speechKey words: Learning; Magnetoencephalography (MEG); Mismatch negativity MMNm; MMF; 0.05), Mismatch ¢eld); Morse code the communication; F(1,6) ¼ (MMN; 7.41, po while Morse-MMNm was stronNon-speechthe sounds; Plasticity f 3 months, pattern MMNm or speech-P1m magnitude, loci, or latency between ger than the Morse-P1m (main effect of component, Morse of the mean Morse-MMNm magnitudes of both hemide MMN became lateraNon-dominant the first sessions. The MMNm was stronger than spheres was 63 nAm before the training and 45 nAm after minant for the speechF(1,6) ¼ and 7.11,second p o 0.05). The P1m and MMNm also differed efter the training. Thus, more processing resources were required hemisphere y traces for the Morse the P1m response (main effect of component F(1,6) ¼ 17.13, in magnitude between the sessions and the hemispheres when the tone patterns were still novel to the subjects. With mismatch negativity (MMN), the brain’s automatic response 28 17 hin INTRODUCTION the hemisphere that 20 20 to change in repetitive auditory stimulation [5]. The communication can be mediated with sounds p(session othat 0.01). the training, activation unnecessary for a certain function to " hemisphere interaction, F(1,6) ¼ 22.41, p o 0.01). acesHuman for natural speech Speech-MMNm Fig. 2. The mean P1m and MMNm strengths (nAm) for the Morse coded principal generators of the MMN response, typically Speech-MMNm are acoustically completely different from speech. Morse occur presumably dropped out [25] and a ‘‘sharpening of esumably, the close assoactivated between 100 and 250 ms after change onset, dominant are code comprises combinations of acoustic (or written) dots hemisphere non-dominant hemisphere and spoken syllable contrasts in the speech-MMNm dominant and nonMoreover, the P1m and MMNm magnitudes varied between located in both auditory cortices [6,7]. Overlapping with the and dashes, which to the letters of the alphabet. tuning’’ of the neuronal representations [26] took place, representations ofcorrespond the dominant hemispheres before (¢rstof session) afteractivation (second session) bilateral MMN to changes in the acoustic features of speech For example, the letter ‘a’ is signalled by a dot followed by asessions, Fig. 3. The mean strength of MMNm ECDs (nAm) for the spoken and resulting in the decrease the total and neuronal the components, and hemispheres (session " ! c the reverse pattern (# " ) signals The ofa separate training on Morse code processing: The rt 14:1683^1687 sounds [8], MMN subcomponent isMorse elicitedcoded by the dash ( " #), whereas2003 the ‘n’.effect 1st 2nd 1st 2nd syllable contrasts portrayed with smoothed colouring on aMorse learning code. Bars indicate related to the Morse code s.e.m. processing. purely phonetic, or phonological [9,10], properties of the The duration of a dash is three times that of a dot. Silent component "hemisphere interaction, F(1,6) ¼ po 0.05), standard brain surfacewere over9.11, thestronger mean ECD location. Morse-MMNm and Morse-P1m responses in Results show that Session Session Interacting Minds Centre speech sounds of a language attained in infancy [11,12] or gaps equaling to the duration of the dash separate the The MMNm for the Morse stimuli prior to learning the learning the balance in processing is shifted to the in life [13]. This phoneticNeuroscience MMN is with typically left- Morse code, different letters and words.Center Thus, the coded is a MMNm University of message Functionally Integrative as the magnitude differed between thehemisphere. sessions code could only be aAarhus response to the changes in the acoustic the firstlater than in the second session (main effect of session speech-MMNm dominant For and each condition and hemirse sequence code communication; Speech-MMNm lateralized in right-handed individuals [3,14–17] and is of tone patterns which otherwise follows the [email protected] Aarhus University/ Aarhus University Hospital of theand tone second patterns, merely reflecting sphere, thefor mean ECD location projected on2). a triangle net represent- in features an index the of neural memory traces morphology and syntax of the language used inhemispheres, ordinary similar the first sessions (26 change vs 28 nAm, P1m magnitude did was not (Fig. F(1,6) ¼therefore 7.41,considered pwhereas o 0.05), while the Morse-MMNm was stronof Linguistics acoustic language elements [1] and theiring lateralization communication. a standard brain surface. The nearest node was assigned the mean detection based on Dept. short-term memory traces developed Dominant [18,19]. Therefore, an MMN-magnitude difference between Due to the one-to-one mapping between the phonemes respectively), whereas in the speech non-dominant hemiA further analysis on Morse-MMNm showed thatandinthethe Aarhussession University [5]: these sounds were not ECD strength (given below image) six neighboring ones a half ger than the Morse-P1m (main effect ofeachcomponent, during the experimental the hemispheres (which can be determined easily with and letters in the Finnish language, for Finnish-speaking of that value; all other nodes were set to zero. The intermediate values familiar to our subjects prior to the training and thus no Non-dominant MMNm, the MEG counterpart of the MMN) to a contrast of Morse code users, the tone patterns of the codedspeech-dominant message sphere, the Morse-MMNm magnitude dramatically dehemisphere (as indexed by the MMNm to F(1,6) ¼native-language 7.11, p ophonemes, 0.05). The P1m and MMNm also differed as well as to were that ofobtained larger by bilinear interpolation. match not only the letters but also the phonemes of the long-term memory traces for them could possibly exist. AARHUS hemisphere linguistic units [18–22], indicates, presumably, the dominant spoken language. Therefore, Morse code communication creased (from 37 astofor17 Fig. 3); this was traces reflected in the speech stimuli), the Morse-MMNm magnitude was in magnitude between the sessions and the hemispheres However, anynAm; language [3,4,11–13], permanent hemisphere for their long-term memory traces. might have access to the system of the permanent cortical n’srepresentations automatic response for the Morse coded letters are needed for the automatic When Morse coding skills are acquired, permanent session ! hemisphere interaction, F(1,6) ¼ 17.01, (recognition templates [1,2]) for the nativesignificant (session " hemisphere interaction, F(1,6) ¼ 22.41, p o 0.01). Fig. 2. The mean P1m and MMNm strengths (nAm) for the Morse coded language speech stimulation [5].sounds The[3,4]. These long-term memory representations for the units of a completely new type of au traces, like any experience-based memory traces for sounds, p o 0.01). communication system have to be formed. We aimed at mapping of the acoustic stimuli onto their linguistic and spoken syllable contrasts in the speech-MMNm dominant and non- UNIVE (Berliner, 1994; Monson, 1997; Vuust, 2000). This suggests that skilled jazz musicians may have developed a high sensitivity to subtle deviations of rhythm. To test this, we used MEG to study the strength and lateralization of pre-attentive responses of the central nervous system (MMNm) to deviations from a pattern of rhythm in expert jazz musicians and musically inept non-musicians. MMN & LÆRING Materials and methods To mimic communicative cues in improvised music, we made sequences of increasingly incongruent rhythm, using realistic may be experienced as if the music dstumblesT. Both musical expectancy and communicative valence, sIII c be characterized as a much stronger sign of incongru that was substantiated by subjects’ unanimous rating equally or more disturbing than sII. Eight inept non-musicians (6 men and 2 women expert jazz musicians (8 men and 1 woman; educ Sibelius Academy of Music, Helsinki, Finland), gav consent to participate in the study, approved by Committee of Helsinki University Central Hospital. W aptitude for rhythm with a modified version of the im used as part of entry examinations to music conse MUSIKERE www.elsevier.com/locate/ynimg NeuroImage 24 (2005) 560 – 564 To musicians, the message is in the meter Pre-attentive neuronal responses to incongruent rhythm are left-lateralized in musicians Peter Vuust,a,b,* Karen Johanne Pallesen,a,c,d Christopher Bailey,a,c Titia L. van Zuijen,e Albert Gjedde,a Andreas Roepstorff,a,f and Leif astergaarda a Centre for Functionally Integrative Neuroscience, University of Aarhus, Aarhus, Denmark P. Vuust et al. / NeuroImage Royal Academy of Music, Aarhus, Denmark BioMag Laboratory, Helsinki Brain Research Centre, Helsinki University Central Hospital, Helsinki, Finland d Neuroscience Unit, Helsinki Brain Research Center, Institute of Biomedicine/Physiology, University of Helsinki, Helsinki, Finland e Cognitive Brain Research Unit, Helsinki Brain Research Centre, Institute of Psychology, Helsinki University, Helsinki, Finland f Institute of Anthropology, Linguistics and Archaeology, University of Aarhus, Aarhus, Denmark b 24 (2005) 560–564 563 c Received 23 June 2004; revised 17 August 2004; accepted 27 August 2004 Available online 11 November 2004 Musicians exchange non-verbal cues as messages when they play together. This is particularly true in music with a sketchy outline. Jazz musicians receive and interpret the cues when performance parts from a regular pattern of rhythm, suggesting that they enjoy a highly developed sensitivity to subtle deviations of rhythm. We demonstrate that pre-attentive brain responses recorded with magnetoencephalography to rhythmic incongruence are left-lateralized in expert jazz musicians and right-lateralized in musically inept non-musicians. The left-lateralization of the pre-attentive responses suggests functional adaptation of the brain to a task of communication, which is much like that of language. D 2004 Elsevier Inc. All rights reserved. In support of neural dissociation between music and language processing, lesion studies and studies of people suffering from acquired and congenital amusia show a double dissociation between aspects of music and language processing (Ayotte et al., 2000; Liegeois-Chauvel et al., 1998; Mendez, 2001; Peretz and Coltheart, 2003; Peretz et al., 1994, 2002) as well as a greater involvement of the right hemisphere in basic music processing of especially pitch-related features (Samson et al., 2002; Tervaniemi and Hugdahl, 2003; Zatorre, 1988; Zatorre et al., 2002). Brain responses to auditory stimuli, however, are not only determined by physical properties of the stimuli, and the nature of cognitive operations involved. In many cases, the listeners’ competence and familiarity with the stimuli affect neuronal processing. This has Keywords: Pre-attentive neuronal responses; Incongruent rhythm; Musicians been shown for pre-attentive processing of language at 100–200 ms after stimulus onset as indicated by the mismatch negativity (MMNm), recorded with magnetoencephalography (MEG). Left lateralization of the MMNm occurs to deviating phonemes from Introduction subjects’ mother tongues only (Na¨a¨ta¨nen et al., 1997), and deviating Morse code syllables in Morse code trained subjects, Music is one of the many ways humans communicate with each only (Kujala et al., 2003), suggesting that left lateralization of other. Whether music and verbal language share neuronal networks sounds occur when they are perceived as meaningful. or music is processed in separate brain modules holds the key to Although music—in contrast to verbal language—rarely refers understanding music in an evolutionary perspective (Huron, 2001), to objects in the real world, musical performance involves as well as music in the context of human communication. Hence, communication. When musicians play, they exchange non-verbal this issue has been hotly debated in the growing field of neuroto sIIIasinmessages, one expert subject. Direction slice.1. (a) Note representation of the stimuli; arrow indicates the time of recording. 8th notes interval is 312.5 ms except at the 2nd bar of sIII (105 andand fromone thisinept interaction, music emergesofasthe a dipoles are projected onto the individual coronal MRFig. biological research in music.Fig. 2. (a) Location of dipoles (MMN) signs concrete form (Sawyer, 2004). Musical for calculated on the basis of dipole amplitudes. (c) Latency timeof courses of averaged magnetic evoked responses from (b) an inept non-musician and (c) an expert musician. Signals are recorded at the audito The relative amplitude (left expert: 60 nAm) represented by size of arrows. (b) communication, Asymmetry index example, musical humor (Huron, 2004), is often conveyed through the MMNs in experts and inept subjects. plotted separately for sI, sII and sIII. violation of musical expectancy. Music theory explains musical * Corresponding author. Centre for Functionally Integrative Neuroexpectancy as the motion between opposites as in harmony the science (CFIN), Norrebrogade 44, Hospital, 8000, Aarhus University, Aarhus University tension of the dominant chord resolved by the motion to the tonic. Aarhus, Denmark. language elicit MMN-response in This the left hemisphere, while phenomenon is referred to asnonan example ofstudy, musicalKujala syntax et al. (2003) showed that the amplitude of the E-mail address: [email protected] (P. Vuust). University/ Aarhus University Hospital MMNm to Morse code reversed lateralization from the hemisyllabic sounds activated the right hemisphere theauthentic midline. Thus,has been shown and violations or of the cadence to activate Available online on ScienceDirect (www.sciencedirect.com). Center of Functionally Integrative Neuroscience prior to attention, expert jazz musicians process rhythmic signs in hemisphere where also phonemes are processed in brains of highly competent users of a language. It is widely held that music and language employ separate neuronal networks (modules) especially for the spectral aspects of music whereas less is known about the temporal aspects of music (Peretz and Coltheart, 2003). At the level of the auditory cortex, the left hemisphere appears to be specialized for the processing of fine-grained temporal stimuli necessary for language comprehen- 1053-8119/$ - see front matter D 2004 Elsevier Inc. All rights reserved. the auditory cortex of the left doi:10.1016/j.neuroimage.2004.08.039 sphere opposite to the one dominant for the MMN to native language sounds to the speech hemisphere after 3 months of intensive training. This result indicates that once a rhythmic pattern conveys meaning it is pre-attentively processed by the left hemisphere. In the present study, we demonstrate that there is a correlation between musical competence and left lateralization of pre-attentive brain responses to rhythmic cues. We propose that this may be functionally linked to the left lateralization of MMN to vowels and syllables, as shown in previous language studies. Interacting Minds Centre [email protected] Dept. of Linguistics Aarhus University au AARHUS UNIVE HJERNENS GÅDE #2 Review Trends in Cognitive Sciences May 2012, Vol. 16, No. 5 Review Box 1. a journey in time and space Hvorfor sidder sproget til venstre? Nearly 50 years ago, Robert Efron [5] suggested that the processing of However, the time intervals at which the impairments emerged in both temporal information and speech may share some resources, showing studies (around 300-400 ms) are at the supra-segmental level in the that patients with aphasia were impaired at overt temporal order phonological hierarchy – that is, at the level of syllables, not phonemes. judgments, needing longer intervals between two stimuli before they A parallel approach used dichotic listening paradigms to reveal could determine which came first. Efron’s finding inspired a more hemispheric asymmetries in the processing of speech and nongeneral interest in how speech perception might be supported by nonspeech stimuli. Studdert-Kennedy and Shankweiler [62,63] demonlinguistic processes and raised the possibility that hemispheric strated that consonant–vowel combinations are processed with a specializations in speech perception could be accounted for in a right ear advantage (REA; suggesting a left hemisphere dominance). domain-general manner. It was striking that the temporal ordering Cutting [64] used sine wave analogues of consonant–vowel (CV; e.g. problems exhibited by the patients were notable from quite long inter‘ba’ or ‘di’) stimuli in a dichotic listening task and argued that there stimulus intervals (ISIs – around 400 ms), which in terms of speech was an REA for the processing of formant transitions, whether or not perception are on the order of syllables, rather than phonemes. these were in speech. It was striking, however, that there was no right Tallal and Piercy [6] investigated how children with aphasia (who ear advantage for the sine wave formant ‘syllables’ alone. In 1980, today would likely be diagnosed with specific language impairment) Schwartz and Tallal [65] showed that, whereas both ‘normal’ CV Institute of on Cognitive UniversityThese College London, Queen (with Square, WC1N 3AR, UK performed auditory Neuroscience, temporal order judgments. children, like 17 stimuli a 40London ms formant transition) and extended CV stimuli the adult patients with aphasia, needed longer ISIs than controls to be (modified to have a 80 ms transition) showed an REA, this was less able to discern which sound came first. As in Efron’s study, the children strong for the stimuli with the modified, longer profiles. Although no with ‘aphasia’ began to show problems at relatively long ISIs (around non-speech stimuli were tested, the authors concluded that this was Over past to for thealength of syllables andfor intonation procomparable 3 0 0 mthe strong evidence s). T h e s e 30 f i nyears d i n g s hemispheric h a v e b e e n w iasymmetries d e l y i n t e r p r e t in ed non-linguistic preference shorter, faster perceptionof have been within a doThese theories have been widely used as explanatory speech files. as demonstrations acoustic problems withconstrued rapid temporal sequencing. transitions in the left hemisphere. Cortical asymmetries in speech perception: what’s wrong, what’s right and what’s left? Carolyn McGettigan and Sophie K. Scott HÅNDETHED? Lateraliseringen er mangfoldig og kompliceret Ingen entydig eller samlet genetisk forklaring main-general framework, according to which preferential frameworks over the past decade. processing of speech is due to left-lateralized, non-linmodulations [11], it is unclear why spectral detail should ! No single acoustic cue underlies the way that informaguistic acoustic sensitivities. A prominent version of this The nature of speech tion is expressed in speech, and any one ‘phonetic’ be processed preferentially in the right hemisphere (e.g. argument holds that the left temporal lobe selectively It is important to consider whether these neural models is underpinned by a variety of Acoustiacoustic In some assumptions cases, ‘spectral’ is used meanof‘pitch’ contrastrapid/temporal [3,7]). information in sound. reasonable about the to nature the processes make Inpoor British English, theofphonetic difference two(Figure terms1).are fact not features. [3,12], although this is a characterization speech and there conveyed these in speech Thein approaches cally, information /apa/empirical and /aba/ support concernsfor onlya the voicing of the synonymous. been little left-hemisphere and Poeppel both clearly associate left auditory hasbetween of Zatorre or these /b/ phones; there are more than 16 serious issue for both of medial /p/ ! Finally, for cues.however, In sharp contrast, the right with aasensitivity to temporal or the rapidapproaches information. selectivity fields acoustic features which differ between the andmay Zatorre identification of the phoneme separable Poeppel be is anthe inaccurate characterization of temporal lobe is demonstrably sensitive to specific acousHowever, this plosives, We including the that length of the pre-stop vowel as the fundamental unit of perceptual information in properties. suggest acoustic accounts of tic two speech: longer in /aba/ andto thebeaspiration release of That is, the assumption access phoneme beingsensitivities speech. need informedonbythe the nature all the information in speechthat exists overto short time speech ! Not more pronounced fordomain-general /b/ [10]. is theare cardinal aspect /p/ being representations speech signal and that athan simple vs. Plosives, which frequently the of onlyspeech conof the scales. be characterized in terms of their spectral and (possibly hence that theories ! All sounds candichotomy perception, may be incorrect. studied because theyneed can to beaccount neatly domain-specific sonants left-dominant sensitivity to phenomena which and amplitude modulation profiles – in speech, for for a into a matrix of place/manner/presence or absence fitted Explaining imbalance most of the energy is carried in low spectral underlie perception. Theresounds), is considerexample,the might voicing cues, phoneme unlike many other speech do of well agreed that, for intelligibility most people,can both It is amplitude modulations. Speech be that this assumption is incorrect [13–18]. andgenerally able evidence very rapidly evolving acoustic structure. Howinclude production andspectral speechand perception, it serves speech when the amplitudeas modulations the growing evidence the and importance of preserved Indeed, fricatives, affricates, nasals,for liquids vowels are ever, arecoarse functions the leftcan hemisphere [1,2]. The language, (e.g. of listeners understand speech like onset-rime structure, are quite suprasegmental considerablystructures longer in duration than the (40 ms) often majority patients with channels aphasia have a left-hemivast to 6of broad spectral and considerable would call for a temporal account vocoded or syllables, specified by surely Poeppel. window lesion,ofand left dominance for however, speech and lansphere any sound, by itstime verywindow nature,on can existsignal, over ! As the aamplitude envelope); neither applies a wider theonly speech smoothing that been reported in intact dichotic guage ‘temporal’ a very wellrather specified term. There time, of also modulation in isolation canbrains yield using an intelligible order isofnot hundreds, than tens, of kindhas on the Aarhus University magnetic stimulation (TMS) and listening, no non-temporal information available in sound. In is Given that speech perception requires spectral percept.transcranial milliseconds. functional neuroimaging (fMRI). One influential body of turn, ‘temporal’ has been used in a variety of senses; research over the past two decades has concentrated on the Efron [5] used the concept of time to point out that Box 2. Neural oscillations and temporal primitives from a hypothesis that this left dominance might emerge phonemes need to be heard in the right order to result in Aarhus University The Asymmetric Sampling in Time hypothesis [4] predicts that the AST model, these theta effectsTallal were right-dominant. of meaningful to non-linguistic acoustic information which sensitivity speech, wheras and Piercy The [6] ongoing to complementary gamma activity proposed by the AST model is oscillatory activity in different frequency bands forms be critical for speech. In approaches strongly happens expressly considered the processing of temporal order computational primitives in the brain. Theta range (4-8 Hz) activity relatively lacking in these studies (but see [8,9]). Notably, speech influenced by hypotheses about the role of temporal inforjudgments to reflect similar processes to those necessary should be maximally sensitive to information at the rate of syllables in perception is not restricted to gamma and theta frequencies. Selective speechneuronal (Box 1), resultingoperating theoriesinproposed that mation of phonemes in the for the speech,in attention whereas populations the low gamma to identification speech is associated with lateralized alphaincoming (8-13 Hz) left("40 hemisphere show atopreference for processing the stream. More recently, ‘temporal’ has in been used speech range activity Hz) shouldwould be sensitive the rapid temporal changes [70], correlated with enhanced theta activity auditory rapid to temporal in sound. example, ofassumed regions. be central information to phonemic information. TheFor theory proposes Neural responses to degraded words applied show an to alpha to the amount of smoothing the to refer Læring og ekspertise spiller sandsynligvis en rolle Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Dept. of Linguistics au AARHUS UNIVE TAK FOR OPMÆRKSOMHEDEN Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE SPØRGSMÅL? Center of Functionally Integrative Neuroscience Aarhus University/ Aarhus University Hospital Interacting Minds Centre [email protected] Aarhus University Dept. of Linguistics Aarhus University au AARHUS UNIVE