TY  - JOUR
A1  - Bijeljac-Babic, Ranka
A1  - Höhle, Barbara
A1  - Nazzi, Thierry
T1  - Early Prosodic Acquisition in Bilingual Infants: The Case of the Perceptual Trochaic Bias
JF  - Frontiers in psychology
N2  - Infants start learning the prosodic properties of their native language before 12 months, as shown by the emergence of a trochaic bias in English-learning infants between 6 and 9 months (Jusczyk et al., 1993), and in German-learning infants between 4 and 6 months (Huhle et al., 2009, 2014), while French-learning infants do not show a bias at 6 months (Hohle et al., 2009). This language-specific emergence of a trochaic bias is supported by the fact that English and German are languages with trochaic predominance in their lexicons, while French is a language with phrase-final lengthening but lacking lexical stress. We explored the emergence of a trochaic bias in bilingual French/German infants, to study whether the developmental trajectory would be similar to monolingual infants and whether amount of relative exposure to the two languages has an impact on the emergence of the bias. Accordingly, we replicated Hohle et al. (2009) with 24 bilingual 6-month-olds learning French and German simultaneously. All infants had been exposed to both languages for 30 to 70% of the time from birth. Using the Head Preference Procedure, infants were presented with two lists of stimuli, one made up of several occurrences of the pseudoword /GAba/ with word-initial stress (trochaic pattern), the second one made up of several occurrences of the pseudoword /gaBA/ with word-final stress (iambic pattern). The stimuli were recorded by a native German female speaker. Results revealed that these French/German bilingual 6-month olds have a trochaic bias (as evidenced by a preference to listen to the trochaic pattern). Hence, their listening preference is comparable to that of monolingual German-learning 6-month-olds, but differs from that of monolingual French-learning 6-month-olds who did not show any preference (Noble et al., 2009). Moreover, the size of the trochaic bias in the bilingual infants was not correlated with their amount of exposure to German. The present results thus establish that the development of a trochaic bias in simultaneous bilinguals is not delayed compared to monolingual German-learning infants (Hohle et al., 2009) and is rather independent of the amount of exposure to German relative to French.
KW  - bilinguals
KW  - infants
KW  - language
KW  - prosody
KW  - lexical stress
KW  - dominance effects
Y1  - 2016
U6  - https://doi.org/10.3389/fpsyg.2016.00210
SN  - 1664-1078
VL  - 7
SP  - 1753
EP  - 1802
PB  - Frontiers Research Foundation
CY  - Lausanne
ER  - 
TY  - JOUR
A1  - Abboub, Nawal
A1  - Boll-Avetisyan, Natalie
A1  - Bhatara, Anjali
A1  - Höhle, Barbara
A1  - Nazzi, Thierry
T1  - An Exploration of Rhythmic Grouping of Speech Sequences by French- and German-Learning Infants
JF  - Frontiers in human neuroscienc
N2  - Rhythm in music and speech can be characterized by a constellation of several acoustic cues. Individually, these cues have different effects on rhythmic perception: sequences of sounds alternating in duration are perceived as short-long pairs (weak-strong/iambic pattern), whereas sequences of sounds alternating in intensity or pitch are perceived as loud-soft, or high-low pairs (strong-weak/trochaic pattern). This perceptual bias—called the Iambic-Trochaic Law (ITL)–has been claimed to be an universal property of the auditory system applying in both the music and the language domains. Recent studies have shown that language experience can modulate the effects of the ITL on rhythmic perception of both speech and non-speech sequences in adults, and of non-speech sequences in 7.5-month-old infants. The goal of the present study was to explore whether language experience also modulates infants’ grouping of speech. To do so, we presented sequences of syllables to monolingual French- and German-learning 7.5-month-olds. Using the Headturn Preference Procedure (HPP), we examined whether they were able to perceive a rhythmic structure in sequences of syllables that alternated in duration, pitch, or intensity. Our findings show that both French- and German-learning infants perceived a rhythmic structure when it was cued by duration or pitch but not intensity. Our findings also show differences in how these infants use duration and pitch cues to group syllable sequences, suggesting that pitch cues were the easier ones to use. Moreover, performance did not differ across languages, failing to reveal early language effects on rhythmic perception. These results contribute to our understanding of the origin of rhythmic perception and perceptual mechanisms shared across music and speech, which may bootstrap language acquisition.
KW  - language acquisition
KW  - prosody
KW  - grouping
KW  - iambic-trochaic law
KW  - perceptual biases
KW  - french-learning infants
KW  - german-learning infants
Y1  - 2016
U6  - https://doi.org/10.3389/fnhum.2016.00292
SN  - 1662-5161
VL  - 10
SP  - 6707
EP  - 6712
PB  - Frontiers Research Foundation
CY  - Lausanne
ER  - 
TY  - JOUR
A1  - Boll-Avetisyan, Natalie
A1  - Bhatara, Anjali
A1  - Unger, Annika
A1  - Nazzi, Thierry
A1  - Höhle, Barbara
T1  - Effects of experience with L2 and music on rhythmic grouping by French listeners
JF  - Bilingualism : language and cognition.
KW  - rhythmic grouping
KW  - second language acquisition
KW  - prosody
KW  - musicality
KW  - Iambic
KW  - Trochaic Law
Y1  - 2016
U6  - https://doi.org/10.1017/S1366728915000425
SN  - 1366-7289
SN  - 1469-1841
VL  - 19
SP  - 971
EP  - 986
PB  - Cambridge Univ. Press
CY  - New York
ER  - 
TY  - THES
A1  - Marimon Tarter, Mireia
T1  - Word segmentation in German-learning infants and German-speaking adults
T1  - Wortsegmentierung bei deutschsprachigen Kindern und Erwachsenen
BT  - prosodic and statistical cues
BT  - prosodische und statitistische Hinweisreize
N2  - There is evidence that infants start extracting words from fluent speech around 7.5 months of age (e.g., Jusczyk & Aslin, 1995) and that they use at least two mechanisms to segment words forms from fluent speech: prosodic information (e.g., Jusczyk, Cutler & Redanz, 1993) and statistical information (e.g., Saffran, Aslin & Newport, 1996). However, how these two mechanisms interact and whether they change during development is still not fully understood. 
The main aim of the present work is to understand in what way different cues to word segmentation are exploited by infants when learning the language in their environment, as well as to explore whether this ability is related to later language skills. In Chapter 3 we pursued to determine the reliability of the method used in most of the experiments in the present thesis (the Headturn Preference Procedure), as well as to examine correlations and individual differences between infants’ performance and later language outcomes. In Chapter 4 we investigated how German-speaking adults weigh statistical and prosodic information for word segmentation. We familiarized adults with an auditory string in which statistical and prosodic information indicated different word boundaries and obtained both behavioral and pupillometry responses. Then, we conducted further experiments to understand in what way different cues to word segmentation are exploited by 9-month-old German-learning infants (Chapter 5) and by 6-month-old German-learning infants (Chapter 6). In addition, we conducted follow-up questionnaires with the infants and obtained language outcomes at later stages of development. 
Our findings from this thesis revealed that (1) German-speaking adults show a strong weight of prosodic cues, at least for the materials used in this study and that (2) German-learning infants weight these two kind of cues differently depending on age and/or language experience. We observed that, unlike English-learning infants, 6-month-old infants relied more strongly on prosodic cues. Nine-month-olds do not show any preference for either of the cues in the word segmentation task. From the present results it remains unclear whether the ability to use prosodic cues to word segmentation relates to later language vocabulary. We speculate that prosody provides infants with their first window into the specific acoustic regularities in the signal, which enables them to master the specific stress pattern of German rapidly. Our findings are a step forwards in the understanding of an early impact of the native prosody compared to statistical learning in early word segmentation.
N2  - Es gibt Beweise, dass Babys im Alter um 7.5 Monate anfangen, Wörter aus flüssiger Sprache zu extrahieren (z.B., Jusczyk & Aslin, 1995), dabei werden mindestens zwei Mechanismen angewandt, um  Wörter aus dem Sprachfluss zu segmentieren: prosodische Information (z.B. Jusczyk, Cutler & Redanz, 1993) und statistische Information (z.B., Saffran, Aslin & Newport, 1996). Aber wie diese zwei Mechanismen zusammenwirken und ob sie sich während der Entwicklung verändern, wurde noch nicht vollständig geklärt. 
Das Hauptziel dieser Dissertation ist es zu verstehen, inwiefern sich Kleinkinder die verschiedenen Hinweisreize für die Segmentierung von Wörtern erschließen, wenn sie die Sprache ihres Umfelds lernen, sowie ob diese Fähigkeit in Verbindung mit späteren Sprachfähigkeiten steht. In Kapitel 3 wurde die Zuverlässigkeit der in dieser Dissertation benutzten Methode (the Headturn Preference Procedure) eruiert, sowie die Korrelationen und individuellen Differenzen zwischen den Ergebnissen einzelner Babys und späteren Sprachfähigkeiten untersucht. Im Kapitel 4 wurde untersucht, wie deutschsprachige Erwachsene auf statistische und prosodische Informationen für Wortsegmentierung reagieren. Wir haben Erwachsene mit einer akustischen Sequenz familiarisiert, bei der die statistischen und prosodischen Informationen verschiedene Wortgrenzen anzeigen, und Verhaltens- und Pupillometriedaten erhoben. Anschließend haben wir weitere Experimente durchgeführt, um zu verstehen, wie verschiedene Hinweisreize von 9-Monate (Kapitel 5) und 6-Monate alten deutschsprachigen Säuglingen (Kapitel 6) aufgenommen werden. Außerdem haben wir nach den Versuchen mit Fragebögen Daten über die Sprachfähigkeiten in späteren Entwicklungsstadien gesammelt. 
Unsere Ergebnisse lassen darauf schließen, dass (1) deutschsprachige Erwachsene eine hohe Gewichtung der prosodischen Hinweisreize zeigen, zumindest bei dem für die Studie genutzten akustischen Material, und dass (2) bei deutschsprachigen Babys in Abhängigkeit von Alter und/oder Spracherfahrung diese zwei Hinweisreize unterschiedlich gewichtet werden. Wir stellen fest, dass sich 6-Monate alte deutschsprachige Babys stärker auf die prosodichen Hinweisreize verlassen haben als die englischsprachigen Babys. Neun Monate alte Babys haben bei der Wortsegmentierung der Audiosequenz gescheitert und haben keine Vorliebe für einen der beiden Hinweisreize gezeigt. Mit den momentanen Ergebnissen bleibt weiterhin unklar, ob die Fähigkeit, prosodische Hinweisreize für die Wortsegmentierung zu nutzen, in Verbindung mit der späteren Weiterentwicklung des Wortschatzes steht. Wir vermuten, dass Prosodie den Babys einen ersten Einblick in die spezifischen akustischen Regelmäßigkeiten des Signals verschafft, was ihnen dabei hilft, das Betonungsmuster der deutsche Sprache schnell zu erlernen. Unsere Ergebnisse sind  ein weiterer Schritt vorwärts was das Verständnis eines frühen Einflusses von muttersprachlicher Prosodie im Vergleich zu statistischem Erlernen in der frühen Wortsegmentierung betrifft.
KW  - word segmentation
KW  - prosody
KW  - statistical learning
KW  - Wortsegmentierung
KW  - Prosodie
Y1  - 2019
U6  - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:kobv:517-opus4-437400
ER  - 
TY  - THES
A1  - Kühn, Jane
T1  - Functionally-driven language change
T1  - Funktional gesteuerter Sprachwandel
BT  - prosodic focus and sentence type marking in German-Turkish bilingual yes/no questions
BT  - prosodische Fokus-und Satztypmarkierung in Deutsch-Türkischen Ja/ Nein Fragen
N2  - Since the 1960ies, Germany has been host to a large Turkish immigrant community. While migrant communities often shift to the majority language over the course of time, Turkish is a very vital minority language in Germany and bilingualism in this community is an obvious fact which has been subject to several studies. The main focus usually is on German, the second language (L2) of these speakers (e.g. Hinnenkamp 2000, Keim 2001, Auer 2003, Cindark & Aslan (2004), Kern & Selting 2006, Selting 2009, Kern 2013). Research on the Turkish spoken by Turkish bilinguals has also attracted attention although to a lesser extend mainly in the framework of so called heritage language research (cf. Polinski 2011). Bilingual Turkish has been investigated under the perspective of code-switching and codemixing (e.g. Kallmeyer & Keim 2003, Keim 2003, 2004, Keim & Cindark 2003, Hinnenkamp 2003, 2005, 2008, Dirim & Auer 2004), and with respect to changes in the morphologic, the syntactic and the orthographic system (e.g. Rehbein & Karakoç 2004, Schroeder 2007). Attention to the changes in the prosodic system of bilingual Turkish on the other side has been exceptional so far (Queen 2001, 2006).
With the present dissertation, I provide a study on contact induced linguistic changes on the prosodic level in the Turkish heritage language of adult early German-Turkish bilinguals. It describes structural changes in the L1 Turkish intonation of yes/no questions of a representative sample of bilingual Turkish speakers. All speakers share a similar sociolinguistic background. All acquired Turkish as their first language from their families and the majority language German as an early L2 at latest in the kinder garden by the age of 3.
A study of changes in bilingual varieties requires a previous cross-linguistic comparison of both of the involved languages in language contact in order to draw conclusions on the contact-induced language change in delimitation to language-internal development.
While German is one of the best investigated languages with respect to its prosodic system, research on Turkish intonational phonology is not as progressed. To this effect, the analysis of bilingual Turkish, as elicited for the present dissertation, is preceded by an experimental study on monolingual Turkish. In this regard an additional experiment with 11 monolingual university students of non-linguistic subjects was conducted at the Ege University in Izmir in 2013. On these grounds the present dissertation additionally contributes new insights with respect to Turkish intonational phonology and typology. The results of the contrastive analysis of German and Turkish bring to light that the prosodic systems of both languages differ with respect to the use of prosodic cues in the marking of information structure (IS) and sentence type. Whereas German distinguishes in the prosodic marking between explicit categories for focus and givenness, Turkish uses only one prosodic cue to mark IS. Furthermore it is shown that Turkish in contrast to German does not use a prosodic correlate to mark yes/no questions, but a morphological question marker. 
To elicit Turkish yes/no questions in a bilingual context which differ with respect to their information structure in a further step the methodology of Xu (1999) to elicit in-situ focus on different constituents was adapted in the experimental study. A data set of 400 Turkish yes/no questions of 20 bilingual Turkish speakers was compiled at the Zentrum für Allgemeine Sprachwissenschaft (ZAS) in Berlin and at the University of Potsdam in 2013. The prosodic structure of the yes/no questions was phonologically and phonetically analyzed with respect to changes in the f0 contour according to IS modifications and the use of prosodic cues to indicate sentence type.
The results of the analyses contribute surprising observations to the research of bilingual prosody. Studies on bilingual language change and language acquisition have repeatedly shown that the use of prosodic features that are considered as marked by means of lower and implicational use across and within a language cause difficulties in language contact and second language acquisition. Especially, they are not expected to pass from one language to another through language contact. However, this structurally determined expectation on language development is refuted by the results of the present study. Functionally related prosody, such as the cues to indicate IS, are transferred from German L2 to the Turkish L1 of German-Turkish bilingual speakers. This astonishing observation provides the base for an approach to language change centered on functional motivation. Based on Matras’ (2007, 2010) assumption of functionality in language change, Paradis’ (1993, 2004, 2008) approach of Language Activation and the Subsystem Theory and the Theory of Language as a Dynamic System (Heredina & Jessner 2002), it will be shown that prosodic features which are absent in one of the languages of bilingual speech communities are transferred from the respective language to the other when they contribute to the contextualization of a pragmatic concept which is not expressed by other linguistic means in the target language. To this effect language interaction is based on language activation and inhibition mechanisms dealing with differences in the implicit pragmatic knowledge between bilinguals and monolinguals. The motivator for this process of language change is the contextualization of the message itself and not the structure of the respective feature on the surface. It is shown that structural consideration may influence language change but that bilingual language change does not depend on structural restrictions nor does the structure cause a change. The conclusions drawn on the basis of empirical facts can especially contribute to a better understanding of the processes of bilingual language development as it combines methodologies and theoretical aspects of different linguistic subfields.
N2  - Seit den 1960er Jahren und der damit einhergehenden Arbeitsmigration ist Deutschland eines der europäischen Länder mit der größten türkischstämmigen Einwanderergesellschaft, die sich auf etwa 2.8 Millionen Personen beläuft (Woellert et al. 2009). Obwohl Sprachminderheiten im Laufe der Generationen gewöhnlich zur Mehrheitssprache wechseln, ist das Türkische in Deutschland zweifelsohne eine lebendige Sprache. Die Zweisprachigkeit der türkischen Gemeinschaft in Deutschland ist eine nicht von der Hand zu weisende Tatsache und liegt im Fokus sprachwissenschaftlicher Forschung. Hierbei liegt das Hauptaugenmerk auf der Betrachtung der sprachlichen Merkmale und Veränderungen in der Zweitsprache (L2) Deutsch (e.g. Hinnenkamp 2000, Keim 2001, Auer 2003, Cindark & Aslan 2004, Kern & Selting 2006, Selting 2009, Kern 2013). Die Erstsprache (L1) Türkisch findet vorrangig Beachtung in der sogenannten heritage language Forschung. Obwohl der Begriff keiner klaren Definition oder Abgrenzung bezüglich ähnlicher Konzepte, wie der Minderheitensprache unterliegt, bezieht sich heritage language im Allgemeinen auf die Erstsprache bilingualer Sprecher in der Diaspora (vgl. Polinski 2011). Das Türkische in Deutschland wurde bislang vor allem in Bezug auf Sprechstil und den einhergehenden Mechanismen von code-switching und code-mixing (z.B. Kallmeyer & Keim 2003, Keim 2003, 2004, Keim & Cindark 2003, Hinnenkamp 2003, 2005, 2008, Dirim & Auer 2004), aber auch in Bezug auf grammatikalische und orthographische Veränderungen untersucht (e.g. Rehbein & Karacoç 2004, Schroeder 2007). In Bezug auf prosodische Merkmale von bilingualem Türkisch beschreibt Queen (2001, 2006) eine Vermischung (fusion) steigender Intonationskonturen in Konversationen bilingualer Deutsch- Türkischer Kinder. Backus, Jorgenson & Pfaff (2010) bieten einen Überblick zu bisheriger Forschung zum Türkischen in der Diaspora.
Mit der vorliegenden Dissertation wird ein Forschungsbeitrag zur Veränderung im prosodischen System des Türkischen als heritage language Deutsch-Türkisch bilingualer Sprecher mit frühem Zweitspracherwerb geleistet. Damit wird zum einen das Themengebiet des kontaktinduzierten Sprachwandels und zum anderen das der bilingualen Intonationsforschung aufgegriffen. Obwohl die Beschreibungen phonologischer Unterschiede bei bilingualen Sprechern in Bezug auf die segmentale Ebene reichhaltig sind, ist ein reges Interesse an phonologischen Phänomenen auf der suprasegmentalen Ebene in der Mehrsprachigkeitsforschung recht frisch und bezieht sich größtenteils auf die Zweitsprache und weniger auf die Veränderungen in der heritage language. Die vorliegende Arbeit beschreibt strukturelle Veränderungen in der L1 Intonation in türkischen Ja/Nein Fragen eines repräsentativen Samples bilingualer Türkischsprecher. Der Untersuchungsschwerpunkt liegt dabei auf der Realisierung funktional relationierter Prosodie. Funktional relationierte Prosodie gilt gemäß markiertheitsstrukturierter Ansätze (z.B. Eckmann 1977, Rasier & Hilligsmann 2007, Zerbian 2015) als markierter als strukturell relationierte Prosodie und damit als potentieller Auslöser von Veränderungen im bilingualen Spracherwerb und Sprachkontakt. In der Dissertation werden zwei funktionale Aspekte der Prosodie untersucht. Zum einen werden Veränderungen in der Intonationskontur in Abhängigkeit von Informationsstruktur (IS) untersucht. Genauer, wie kontrastiver in-situ Fokus in der Grundfrequenz umgesetzt wird. Zum anderen wird untersucht wie die Markierung von Ja/Nein Fragen in der f0 Kontur realisiert wird. Eine Einführung in die Thematik von Ja/nein Fragen erfolgt für das Türkische in Kapitel II. 6 und für das Deutsche in Kapitel V.6. Eine umfassende Erläuterung zu IS und seinen verschiedenen Kategorien erfolgt in Kapitel I.
KW  - bilingualism
KW  - prosody
KW  - pragmatics
KW  - Turkish
KW  - Zweisprachigkeit
KW  - Prosodie
KW  - Pragmatik
KW  - Türkisch
Y1  - 2016
U6  - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:kobv:517-opus4-422079
ER  - 
TY  - GEN
A1  - Boll-Avetisyan, Natalie
A1  - Bhatara, Anjali
A1  - Unger, Annika
A1  - Nazzi, Thierry
A1  - Höhle, Barbara
T1  - Effects of experience with L2 and music on rhythmic grouping by French listeners
T2  - Postprints der Universität Potsdam : Humanwissenschaftliche Reihe
N2  - Rhythm perception is assumed to be guided by a domain-general auditory principle, the Iambic/Trochaic Law, stating that sounds varying in intensity are grouped as strong-weak, and sounds varying in duration are grouped as weak-strong. Recently, Bhatara et al. (2013) showed that rhythmic grouping is influenced by native language experience, French listeners having weaker grouping preferences than German listeners. This study explores whether L2 knowledge and musical experience also affect rhythmic grouping. In a grouping task, French late learners of German listened to sequences of coarticulated syllables varying in either intensity or duration. Data on their language and musical experience were obtained by a questionnaire. Mixed-effect model comparisons showed influences of musical experience as well as L2 input quality and quantity on grouping preferences. These results imply that adult French listeners' sensitivity to rhythm can be enhanced through L2 and musical experience.
T3  - Zweitveröffentlichungen der Universität Potsdam : Humanwissenschaftliche Reihe - 450 
KW  - rhythmic grouping
KW  - second language acquisition
KW  - prosody
KW  - musicality
KW  - Iambic
KW  - Trochaic Law
Y1  - 2018
U6  - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:kobv:517-opus4-413786
IS  - 450
ER  - 
TY  - GEN
A1  - Bijeljac-Babic, Ranka
A1  - Höhle, Barbara
A1  - Nazzi, Thierry
T1  - Early prosodic acquisition in bilingual infants
BT  - the case of the perceptual trochaic bias
T2  - Postprints der Universität Potsdam : Humanwissenschaftliche Reihe
N2  - Infants start learning the prosodic properties of their native language before 12 months, as shown by the emergence of a trochaic bias in English-learning infants between 6 and 9 months (Jusczyk et al., 1993), and in German-learning infants between 4 and 6 months (Huhle et al., 2009, 2014), while French-learning infants do not show a bias at 6 months (Hohle et al., 2009). This language-specific emergence of a trochaic bias is supported by the fact that English and German are languages with trochaic predominance in their lexicons, while French is a language with phrase-final lengthening but lacking lexical stress. We explored the emergence of a trochaic bias in bilingual French/German infants, to study whether the developmental trajectory would be similar to monolingual infants and whether amount of relative exposure to the two languages has an impact on the emergence of the bias. Accordingly, we replicated Hohle et al. (2009) with 24 bilingual 6-month-olds learning French and German simultaneously. All infants had been exposed to both languages for 30 to 70% of the time from birth. Using the Head Preference Procedure, infants were presented with two lists of stimuli, one made up of several occurrences of the pseudoword /GAba/ with word-initial stress (trochaic pattern), the second one made up of several occurrences of the pseudoword /gaBA/ with word-final stress (iambic pattern). The stimuli were recorded by a native German female speaker. Results revealed that these French/German bilingual 6-month olds have a trochaic bias (as evidenced by a preference to listen to the trochaic pattern). Hence, their listening preference is comparable to that of monolingual German-learning 6-month-olds, but differs from that of monolingual French-learning 6-month-olds who did not show any preference (Noble et al., 2009). Moreover, the size of the trochaic bias in the bilingual infants was not correlated with their amount of exposure to German. The present results thus establish that the development of a trochaic bias in simultaneous bilinguals is not delayed compared to monolingual German-learning infants (Hohle et al., 2009) and is rather independent of the amount of exposure to German relative to French.
T3  - Zweitveröffentlichungen der Universität Potsdam : Humanwissenschaftliche Reihe - 435 
KW  - bilinguals
KW  - infants
KW  - language
KW  - prosody
KW  - lexical stress
KW  - dominance effects
Y1  - 2018
U6  - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:kobv:517-opus4-407225
IS  - 435
ER  - 
TY  - GEN
A1  - Abboub, Nawal
A1  - Boll-Avetisyan, Natalie
A1  - Bhatara, Anjali
A1  - Höhle, Barbara
A1  - Nazzi, Thierry
T1  - An exploration of rhythmic grouping of speech sequences by french- and german-learning infants
T2  - Postprints der Universität Potsdam : Humanwissenschaftliche Reihe
N2  - Rhythm in music and speech can be characterized by a constellation of several acoustic cues. Individually, these cues have different effects on rhythmic perception: sequences of sounds alternating in duration are perceived as short-long pairs (weak-strong/iambicpattern), whereas sequences of sounds alternating in intensity or pitch are perceived as loud-soft, or high-low pairs (strong-weak/trochaic pattern). This perceptual bias-called the lambic-Trochaic Law (ITL) has been claimed to be an universal property of the auditory system applying in both the music and the language domains. Recent studies have shown that language experience can modulate the effects of the ITL on rhythmic perception of both speech and non-speech sequences in adults, and of non-speech sequences in 7.5-month-old infants. The goal of the present study was to explore whether language experience also modulates infants' grouping of speech. To do so, we presented sequences of syllables to monolingual French- and German-learning 7.5-month-olds. Using the Headturn Preference Procedure (HPP), we examined whether they were able to perceive a rhythmic structure in sequences of syllables that alternated in duration, pitch, or intensity. Our findings show that both French- and German-learning infants perceived a rhythmic structure when it was cued by duration or pitch but not intensity. Our findings also show differences in how these infants use duration and pitch cues to group syllable sequences, suggesting that pitch cues were the easier ones to use. Moreover, performance did not differ across languages, failing to reveal early language effects on rhythmic perception. These results contribute to our understanding of the origin of rhythmic perception and perceptual mechanisms shared across music and speech, which may bootstrap language acquisition.
T3  - Zweitveröffentlichungen der Universität Potsdam : Humanwissenschaftliche Reihe - 427 
KW  - language acquisition
KW  - prosody
KW  - grouping
KW  - iambic-trochaic law
KW  - perceptual biases
KW  - french-learning infants
KW  - german-learning infants
Y1  - 2018
U6  - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:kobv:517-opus4-407201
IS  - 427
ER  - 
TY  - THES
A1  - Holzgrefe-Lang, Julia
T1  - Prosodic phrase boundary perception in adults and infants
T1  - Die Wahrnehmung prosodischer Phrasengrenzen bei Erwachsenen und Säuglingen
BT  - what the brain reveals about contextual influence and the impact of prosodic cues
N2  - Prosody is a rich source of information that heavily supports spoken language comprehension. In particular, prosodic phrase boundaries divide the continuous speech stream into chunks reflecting the semantic and syntactic structure of an utterance. This chunking or prosodic phrasing plays a critical role in both spoken language processing and language acquisition. Aiming at a better understanding of the underlying processing mechanisms and their acquisition, the present work investigates factors that influence prosodic phrase boundary perception in adults and infants. Using the event-related potential (ERP) technique, three experimental studies examined the role of prosodic context (i.e., phrase length) in German phrase boundary perception and of the main prosodic boundary cues, namely pitch change, final lengthening, and pause. With regard to the boundary cues, the dissertation focused on the questions which cues or cue combination are essential for the perception of a prosodic boundary and on whether and how this cue weighting develops during infancy. 
Using ERPs is advantageous because the technique captures the immediate impact of (linguistic) information during on-line processing. Moreover, as it can be applied independently of specific task demands or an overt response performance, it can be used with both infants and adults. ERPs are particularly suitable to study the time course and underlying mechanisms of boundary perception, because a specific ERP component, the Closure Positive Shift (CPS) is well established as neuro-physiological indicator of prosodic boundary perception in adults. 
The results of the three experimental studies first underpin that the prosodic context plays an immediate role in the processing of prosodic boundary information. Moreover, the second study reveals that adult listeners perceive a prosodic boundary also on the basis of a sub-set of the boundary cues available in the speech signal. Both ERP and simultaneously collected behavioral data (i.e., prosodic judgements) suggest that the combination of pitch change and final lengthening triggers boundary perception; however, when presented as single cues, neither pitch change nor final lengthening were sufficient. Finally, testing six- and eight-month-old infants shows that the early sensitivity for prosodic information is reflected in a brain response resembling the adult CPS. For both age groups, brain responses to prosodic boundaries cued by pitch change and final lengthening revealed a positivity that can be interpreted as a CPS-like infant ERP component. In contrast, but comparable to the adults’ response pattern, pitch change as a single cue does not provoke an infant CPS. These results show that infant phrase boundary perception is not exclusively based on pause detection and hint at an early ability to exploit subtle, relational prosodic cues in speech perception.
N2  - Die Wahrnehmung prosodischer Phrasengrenzen spielt eine zentrale Rolle sowohl im frühen Spracherwerb als auch bei der auditiven Sprachperzeption: Prosodische Grenzmarkierungen sind insbesondere relevant, da sie den Sprachstrom gliedern (so genanntes chunking), dabei die syntaktische Struktur einer Äußerung widerspiegeln und zusammenhängende Sinneinheiten erkennbar machen. Um die der Verarbeitung prosodischer Information zugrunde liegenden Mechanismen und deren Erwerb besser charakterisieren zu können, befasst die vorliegende Dissertation mit Faktoren, die die Wahrnehmung prosodischer Grenzmarkierungen bei Erwachsenen und Säuglingen beeinflussen. Mithilfe der Erhebung Ereigniskorrelierter Hirnpotentiale (EKPs) wurde untersucht, welche Rolle der prosodische Kontext (hier: die Phrasenlänge) sowie die an Phrasengrenzen auftretenden Hinweisreize (sogenannte prosodische Cues) bei der Wahrnehmung prosodischer Grenzen im Deutschen spielen. Die untersuchten prosodischen Cues umfassen das Auftreten von Tonhöhenveränderung (pitch change) und finaler Dehnung (final lengthening) sowie Pausensetzung (pause). Es wurde hierbei der Frage nachgegangen, welche Cues oder Cue-Kombinationen für die Wahrnehmung einer Phrasengrenze relevant sind und ob bzw. wie sich diese Gewichtung im Säuglingsalter entwickelt. 
EKPs sind insbesondere geeignet, da sie als on-line Methode die unmittelbare Integration prosodischer Information beim Sprachverstehen erfassen und sowohl bei Erwachsenen als auch bei Säuglingen angewendet werden können. Zudem gibt es mit dem closure positive shift (CPS) eine als Korrelat der Wahrnehmung größerer prosodischer Grenzen etablierte EKP-Komponente. 
Die Ergebnisse der drei experimentellen Studien untermauern, dass der Äußerungskontext eine unmittelbare Rolle bei der Verarbeitung von prosodischen Grenzmarkierungen spielt. Darüber hinaus konnte gezeigt werden, dass erwachsene Hörer eine prosodische Grenze auch basierend auf einem Sub-Set der möglichen, im akustischen Signal verfügbaren, prosodischen Cues wahrnehmen: Die Kombination von Tonhöhenveränderung und finaler Dehnung ist ausreichend, um die Wahrnehmung einer Phrasengrenze zu evozieren; als alleinige Cues sind jedoch weder Tonhöhenveränderung noch finale Dehnung hinreichend. Dies offenbarte sich sowohl im Ausbleiben des CPS als auch in simultan erhobenen Verhaltensdaten (Beurteilung der prosodischen Struktur). Schließlich ergab die Untersuchung sechs- und achtmonatiger Säuglinge, dass sich die frühe Sensitivität für prosodische Informationen in einem neurophysiologischen Korrelat widerspiegelt, welches dem CPS bei Erwachsenen gleicht. Die EKP-Daten zeigten, dass sowohl sechs- als auch achtmonatige Säuglinge prosodische Phrasengrenzen anhand der Kombination von Tonhöhenveränderung und finaler Dehnung wahrnehmen, wohingegen die Tonhöhenveränderung allein auch bei Säuglingen keinen CPS auslöst. Bereits in diesem frühen Alter ist somit die Pause als Grenzmarkierung nicht zwingend erforderlich.
KW  - speech perception
KW  - language acquisition
KW  - prosody
KW  - event-related potentials (ERP)
KW  - prosodic phrase boundaries
KW  - prosodic boundary cues
KW  - Closure Positive Shift (CPS)
KW  - infants
KW  - Sprachwahrnehmung
KW  - Spracherwerb
KW  - Prosodie
KW  - Ereigniskorrelierte Hirnpotentiale (EKP)
KW  - prosodische Phrasengrenzen
KW  - prosodische Grenzmarkierungen
KW  - Closure Positive Shift (CPS)
KW  - Säuglinge
Y1  - 2017
U6  - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:kobv:517-opus4-405943
ER  - 
TY  - GEN
A1  - Kügler, Frank
A1  - Genzel, Susanne
T1  - On the prosodic expression of pragmatic prominence
BT  - the case of pitch register lowering in Akan
N2  - This article presents data from three production experiments investigating the prosodic means of encoding information structure in Akan, a tone language that belongs to the Kwa branch of the Niger-Congo family, spoken in Ghana. Information structure was elicited via context questions that put target words either in wide, informational, or corrective focus, or in one of the experiments also in pre-focal or post-focal position rendering it as given. The prosodic parameters F0 and duration were measured on the target words. Duration is not consistently affected by information structure, but contrary to the prediction that High (H) and Low (L) tones are raised in ex situ (fronted) focus constructions we found a significantly lower realization of both H and L tones under corrective focus in ex situ and in situ focus constructions. Givenness does not seem to be marked prosodically. The data suggest that pragmatic prominence is expressed prosodically by means of a deviation from an unmarked prosodic structure. Results are thus contradicting the view of the effort code that predicts a positive correlation of more effort resulting in higher F0 targets.
T3  - Zweitveröffentlichungen der Universität Potsdam : Humanwissenschaftliche Reihe - 369 
KW  - Akan
KW  - effort code
KW  - information structure
KW  - prosody
KW  - register lowering
Y1  - 2017
U6  - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:kobv:517-opus4-403217
ER  -