gms | German Medical Science

21. Wissenschaftliche Jahrestagung der DGPP

Deutsche Gesellschaft für Phoniatrie und Pädaudiologie

10. bis 12.09.2004, Freiburg/Breisgau

Segregation of syllables during the first year of life

Segmentierung sprachlicher Information während der Entwicklung

Vortrag

  • author presenting/speaker Stefanie Kruck - Charité-University Medicine, Department of Audiology and Phoniatrics, Campus Benjamin Franklin, Berlin, Germany
  • author Karsten Nubel - Charité-University Medicine, Department of Audiology and Phoniatrics, Campus Benjamin Franklin, Berlin, Germany
  • author Nadja Hahnemann - Charité-University Medicine, Department of Audiology and Phoniatrics, Campus Benjamin Franklin, Berlin, Germany
  • author Antje Berkau - Charité-University Medicine, Department of Audiology and Phoniatrics, Campus Benjamin Franklin, Berlin, Germany
  • author Manfred Gross - Charité-University Medicine, Department of Audiology and Phoniatrics, Campus Benjamin Franklin, Berlin, Germany

Deutsche Gesellschaft für Phoniatrie und Pädaudiologie. 21. Wissenschaftliche Jahrestagung der DGPP. Freiburg/Breisgau, 10.-12.09.2004. Düsseldorf, Köln: German Medical Science; 2004. Doc04dgppV38

Die elektronische Version dieses Artikels ist vollständig und ist verfügbar unter: http://www.egms.de/de/meetings/dgpp2004/04dgpp64.shtml

Veröffentlicht: 9. September 2004

© 2004 Kruck et al.
Dieser Artikel ist ein Open Access-Artikel und steht unter den Creative Commons Lizenzbedingungen (http://creativecommons.org/licenses/by-nc-nd/3.0/deed.de). Er darf vervielf&aauml;ltigt, verbreitet und &oauml;ffentlich zug&aauml;nglich gemacht werden, vorausgesetzt dass Autor und Quelle genannt werden.


Abstract

The following study measured the processing of different phonemes in dependence on different gap durations between two consonant-vowel items. Event related potentials (ERPs) were recorded to a phoneme change occurring during the iterative presentation of /da-da/, occasionally interrupted by a stimulus /da-ba/. The pre-attentive detection of the change is measured by the difference between the ERPs of the frequently presented standard stimulus and the rarely presented deviant stimulus.

The experiment was conducted during the first year of life to describe the maturation. The results showed that the phoneme change including a short gap can only be detected in the later maturational stadium, whereas a long gap is detected already in the age of four weeks. The greater mask effect in younger infants can be seen to be induced by a slower neural processing in early infancy.

Zusammenfassung

Der kontinuierliche Sprachschall setzt sich aus vielfältigen Variationen akustischer Parameter zusammen. Gemessen an der Vielzahl akustischer Merkmale ist nicht genau klar, woran sich das auditorische System orientiert, um eine Segmentierung und Identifizierung einzelner Wörter zu bewerkstelligen. Die Verarbeitung von akustischen Merkmalen der Sprache kann mit Hilfe von Ereignis Korrelierten Potenzialen (EKP) sichtbar gemacht werden.

In der folgenden Studie wurde die Verarbeitung von Phonemkategorien in Abhängigkeit von unterschiedlich langen Silbengrenzen (50ms und 150 ms) im Alter von 4 Wochen, fünf Monaten und 13 Monaten getestet. Um die Detektion einer Phonemänderung zu messen wurde der Reiz /da-da/ iterativ präsentiert. Die so angelegte neuronale Gedächtnisspur soll durch eine Abweichung der Phonemkategorie /da-ba/ unterbrochen werden. Die prä-attentive Detektion zeigt sich erwiesenermaßen in der Differenz der EKP beider Reize (Näätänen, R. 1992).

Die Ergebnisse zeigen, dass eine kurze Silbengrenze erst im Alter von 13 Monaten identifiziert wird. Die langsame Verarbeitung der ersten Silbe scheint hier die Verarbeitung der zweiten Silbe zu überlagern. Eine lange Silbengrenze vereinfacht die Verarbeitung und kann schon im Alter von 4 Wochen genutzt werden, um Silben zu segmentieren. Darüberhinaus zeigen die morphologischen Veränderungen anhand von Amplitude und Latenz eine deutliche Entwicklung des auditorischen Systems innerhalb des ersten Lebensjahres.


Text

Introduction

The processing of relevant environmental sounds becomes increasingly specialized during early infancy. During language aquisition babys have to learn to segregate single speech units within the complex speech stream and learn to identify relevant information [1]. Fast acoustic changes and short interruptions between the phonemes are helping to segregate these single speech units from each other. Psychoacoustic experiments showed that the gap duration which is required for the separation of two sounds is strongly age dependent. Infants of three months of age need longer gap durations than older infants of 12 months of age [2]. The underlying neuronal maturation and specialization to relevant sounds is known to be correlated with changes in the cognitive components of the event related potentials (ERPs). Among other factors, the onset latencies of ERPs as response to short auditory stimuli decrease in the course of infant development [3]. The aim was to find out how the ability to discriminate consonant-vowel sequences with different gap durations develops in the first year of life. Therefore, we examined the infants' ability to process syllables as separate units. Beside the general change in morphology of ERP components during the maturation we predicted to find correlates to the maturation of auditory system for the detection of gaps with different duration.

Method

All experiments were part of a longitudinal study about healthy and normal hearing infants conducted within the German language acquisition study (GLaD, www.glad-study.de). Parents gave their written consent according to institutional informed consent procedures and accepted to participate with their children in a multiple set of examinations. The study was conducted in infants during their first year of life (in the age of 4 weeks, 6 months and 13 months).The processing of different phonemes was measured in dependence on different gap durations between two consonant-vowel items. ERPs were recorded to a phoneme change occurring during the iterative presentation of /da-da/, occasionally interrupted by a stimulus /da-ba/. The pre-attentive detection of the change is measured by the difference between the ERPs of the frequently presented standard stimulus and the rarely presented deviant stimulus.

The presented stimuli consisted of two consonant-vowel items (CVs) /da/ and /ba/ naturally produced by a male German speaker. The CVs were matched in peak amplitude and stimulus length (100 ms). The stimulus /da-ba/ was presented as the deviant with a probability of 0.2 interspersed with the standard stimulus /da-da/. The gap between the CVs was set either to 50 ms for the short gap condition or 150 ms for the long gap condition. The interstimulus interval was set to constant 750 ms. An electrode-cap (Easy Cap, Falk Minow) was used to fix 11 electrodes according to the international 10-20 electrode system.

The EEG was amplified with PORTI-32/MREFA (Twente Medical Systems) with a sampling rate of 250 Hz. After eliminating artefacts and filtering the EEG-signal, epochs of 850 ms from stimulus onset were averaged for standards and deviants in both conditions. The mismatch response was defined as the response to the average of standards subtracted from the response to the average of deviants. Using the maximal difference of the grand averages statistical analyses of the mean amplitude were carried out for all age groups in a time window around the maximal peak.

Results

As demonstrated at electrode F3 in Figure 1 [Fig. 1], only infants of 13 months showed a mismatch response (MMR) of positive amplitude for the short gap condition in the time window of 200-300 ms after change onset (F1.60 = 11.095, p = 0.001). The mismatch response for the long gap condition occurred again 200-300 after change onset and was significant in all age groups (4 weeks: F1.101 = 5.458, p < 0.05; 5 months: F1.102 = 5.710, p < 0.05; 13 months: F1.62 = 19.438, p < 0.001).

Discussion

In the present study, infants of the age of 13 months showed a positive mismatch response for both gap conditions, which indicates the detection of the change onset within the bi-syllabic stimulus. The positive amplitude is often found for infants and is defined as mismatch response which can be correlated to the adult mismatch negativity (MMN) in its functionality [4]. In contrast to the adult MMNs the infants' MMR is instable in amplitude and shows a positive deflection for slight acoustic changes, which is also discussed to overlay the negative wave of adult MMN [5]. Yet, it is not clear which of the infant ERPs are correlated to the adult ERPs, although there is some evidence for the explanation for ERPs of school age children up to adults ERPs [6]. Younger infants showed this MMR of positive amplitude only for the long gap condition. The lacking MMR to the short gap condition can be explained by a longer processing time for the first syllable (mask effect). The auditory stream segregation which is important for the detection of sound patterns is highly dependent on the maturation of especially the central auditory system. Although changes in ERP may not be associated with clearly identifiable structural alterations [6] indicators of the electrophysiological methods can help to develop age dependent hearing tests, which can use the basic skills for language acquisition. The results showed that a short gap can only be detected in the later maturational stadium, whereas a long gap is detected already in the age of 4 weeks. The greater mask effect in younger infants is suggested to be induced by a slower neural processing in early infancy.

Acknowledgements

The authors thank Christina Rügen and Jördis Haselow for recording the ERP data. This study was supported by the German Research Foundation (DFG; 381) and Max Planck Institute of Cognitive Neuroscience in Leipzig.

The data characterising the developmental state of our subjects were kindly provided by Volker Hesse, head of the paediatric clinic of the Krankenhaus Lichtenberg, Berlin. He and his team took care for the somatic and neurological data of the children and he provided resources and manpower for recruiting subjects. Finally, we want to thank him for hosting our labs in the Lindenhof hospital.


References

1.
Jusczyk P, Aslin R. Infants' detection of the sound patterns of words in fluent speech. Cognitive Psychol. 1995; 29, 1-23
2.
Werner LA, Marean GH, Halpin CF, Spetner NB, Gillenwater JM. Infant auditory acuity: gap detection. Child Dev. 1992; 63 (2), 260-272
3.
Kushnerenko E, Ceponiene R, Balan P, Fellman V, Huotilaine M, Naatane R. Maturation of the auditory event-related potentials during the first year of life. Neuroreport. 2002; 21;13(1):47-51
4.
Maurer U, Bucher K, Brem S, Brandeis D. Development of the automatic mismatch response: from frontal positivity in kindergarten children to the mismatch negativity. Clin Neurophysiol. 2003; (114): 808-817
5.
Morr ML, Shafer V, Kreuzer J, Kurtzberg, D. Maturation of mismatch negativity in typically developing infants and preschool children. 2002; Ear Hear, 23(2):118-36
6.
Kurztberg D, Vaughan HGJ, Courchesne E, Friedman D, Harter MR, Putnam LE. Developmemtal aspects of Event-related Potentials. Ann NY Acad Sci. 1984; 425, 300-18