Projects and Studies

Current Projects

Language interaction between parents and their children

Director: Prof. Dr. Bettina Braun
Researchers: Katharina Zahner, Moritz Jakob

What the study is about...

Children acquire their native language without any effort, in an apparently implicit way. The linguistic input a child receives from its environment is essential in this process. When adults talk to infants, they often intuitively use a manner of speaking which increases the attention of children. This so-called infant-directed speech (IDS) – in comparison to adult-directed speech (ADS) – is generally characterized by simpler and shorter sentences, slower speech rate, hyperarticulated vowels, higher and more variable pitch, and more variability in the voice quality. The increased attention towards IDS, as compared to ADS, suggests that children prefer infant-directed speech, as the ManyBabies Study in which our BabyspeechLab also participated has recently confirmed (ManyBabies Consortium 2020, see completed projects).
The use of IDS has been one of the focal points in language acquisition research for years. In our current study, we investigate whether and how today’s generation of parents differs in terms of their usage of IDS from the results of previous studies, i.e. whether the use of IDS has changed over the years. Additionally, we are interested if speakers of different varieties of German (e.g. Standard and Alemannic) have different habits when it comes to the interaction with their children (and the use of dialect).

How we investigate this…

For this study, we are particularly interested in natural interactions between parents and their children. These often differ heavily from interactions in a laboratory setting. To obtain such recordings, we ask families to use one of our mobile audio-devices to record everyday situations (e.g. playing, changing diapers, reading a story, etc.) in a home environment.
We are looking for infants between the ages of 0 and 3 years. Please get in touch if you are interested.


ManyBabies Consortium (2020). Quantifying sources of variability in infancy research using the infant-directed speech preference. Advances in Methods and Practices in Psychological Science, 3(1), 24-52.

Cooper, R. P., & Aslin, R. N. (1990). Preference for infant-directed speech in the first month after birth. Child Development, 61, 1584-1595.

Completed Projects

Discrimination of Portuguese Intonation Patterns

Director: Prof. Dr. Bettina Braun
Researchers: Katharina Zahner, Nathalie Czeke, Jasmin Rimpler

What the study is about...

„We support the research of the BabySpeechLab today.“
„Do we support the research of the BabySpeechLab today?“

In many languages, e.g., German and English, statements and questions differ in terms of their syntactic structure (word order, use of “do” etc.). Moreover, speakers in these languages also use the speech melody (intonation) to mark the difference: In statements, our speech melody usually falls at the end of the sentence whereas it typically rises at the end of yes/no-questions. While German and English use both syntax and intonation, there are languages which only mark the difference between statements and questions using intonation only, e.g., Portuguese and Basque. In the Babylab in Lisbon, infant researchers found that Portuguese babies are able to discriminate Portuguese rising and falling intonation contours at the age of 5-6 months (Frota et al. 2014). From other studies we know that Basque babies show similar abilities at that age (Sundara et al. 2015). English babies, by contrast, have difficulty discriminating Portuguese rising and falling intonation. In our study, we aim to find out whether German babies at the age of 5-8 months are able to discriminate statements and questions using the cue of intonation only. To this end, we replicate the study by Frota et al. (2014), testing German babies.

How we investigate this...

Twenty German babies at the age of 5-6 months and 20 German babies at the age of 8-9 months took part in our perception study. The method we used is called Switch Procedure. The child sits on a parent’s lap and listens to sequences of sound. While listening, the child looks at a multi-coloured pattern on a screen. In the so-called habituation phase, the child listens to one type of stimuli, i.e., words with either rising or falling intonation. After a while, the child loses interest in the sounds played and the attention to the screen decreases (habituation). This is when the second phase, the test phase, starts. In the test phase, two sound sequences are played, one sequence with the same intonation as during habituation (same trial) and one with different intonation (switch trial). We measure the duration of looking times to the screen for the two intonation types. Babies usually look longer to new than to familiar stimuli because new things are more interesting and exciting. Infants are thus expected to look longer to the screen when they hear the switch trial as compared to the same trial in the test phase when discrimination (falling vs. rising intonation) is successful.

First results...

Our first results showed that German babies had difficulty in discriminating the Portuguese falling vs. rising intonation contours in both age groups (Czeke et al. 2019). They behaved differently from Portuguese and Basque infants, but similarly to English infants. It hence seems that the way a language marks the difference between statements and questions influences the sensitivity infants display to this contrast. Portuguese and Basque infants who grow up with languages where the difference is marked by intonation only show early discrimination abilities. German and English infants, on the other hand, grow up with languages where the difference between questions and statements is marked by intonation and syntax and might hence be less sensitive to the intonation contrast. We plan to test German infants in a simplified procedure (see Sundara et al. 2015 for English infants) to assess their discrimination abilities further.


Czeke, N., Zahner, K., Rimpler, J., Braun, B. & Frota, S. (2019). German infants do not to discriminate Portuguese rising vs. falling contours. Poster presented at the 4th Workshop on Infant Language Development (WILD), Potsdam, Germany. Poster, Abstract

Frota, S., Butler, J. and Vigário, M. (2014), Infants' Perception of Intonation: Is It a Statement or a Question?. Infancy, 19: 194-213. doi:10.1111/infa.12037

Sundara, M., Molnar, M., and Frota, S. (2015). The perception of boundary tones in infancy. Procedings of the 18th International Congress of Phonetic Sciences, Glasgow, UK.

Word stress, grammatical gender and verb placement in bilingual language acquisition

Director: Prof. Janet Grijzenhout, Prof. Claudia Diehl
PhD Student: Monika Lindauer

What the study is about...

The majority of the world’s population grows up with two or more languages. In Germany, as well, multilingualism is more frequent than one would expect: Children of immigrants typically acquire the language of their parents, i.e. their heritage language, as well as German, i.e. the surrounding language of the society.
In our study, we investigate the acquisition of German by children between 3 and 10 years who grow up with Italian or Turkish as their heritage language. We aim to find out whether and how the acquisition of German differs between bilingual and monolingual children. Furthermore, we analyse which of the following factors play a role for possible differences in language acquisition.

  • the influence of the heritage language on German
  • the age of onset of acquisition of German
  • language exposure from the early years until today in family, day-care, school and leisure activities
  • the socio-economic status of the family

In order to compare the linguistic development of the children, we observe three phenomena of German in different linguistic domains:

  • in rhythm/ accentuation: the duration of the stressed syllable in German word stress
  • in the lexicon and grammar: the gender of the definite article in German (der/ die/ das)
  • in syntax: the position of the finite verb in German main and subordinate clauses

How we investigate this...

We conduct two production experiments: a picture naming game and a sentence completion game. In these games, the children are prompted to produce certain sentence structures and words. That way, we are able to analyse the above-mentioned phenomena. In a parental questionnaire we collect information about the language environment, the language use and the socio-economic background of the children.

Benefits of the study...

The study is part of the EU-project AThEME (Advancing the European Multilingual Experience). The results will contribute to a broader knowledge about multilingualism. For instance, they shed more light on the question in which domains and in what way language acquisition can be promoted more specifically or on the other hand in which domains this may not be necessary. That way, we can counteract prejudices about multilingualism and enlighten the advantages of multilingualism. By the analysis of the children’s linguistic environment we will find out more about the influence of their language exposure. This may reveal some answers to the frequent question: “Which language should I use with my children at home?”

Many Babies Project – Preference for child-directed speech

Director: ManyBabies Consortium
Researchers in the BSL Konstanz: Prof. Dr. Bettina Braun, Dr. Katharina Zahner, Nathalie Czeke

What the study is about...

When adults talk to infants, they typically use a certain kind of speech, called infant-directed speech (IDS). In comparison to speech between adults (ADS), IDS is generally characterized by simpler and shorter sentences, slower speech rate, hyperarticulated vowels, higher and more variable pitch, and more variability in the voice quality. Previous research found that infants show a preference for IDS over ADS (Cooper & Aslin 1990, among many others), i.e., babies display more attention to IDS than to ADS, when given the choice.
The ManyBabies (MB) Project, directed by the ManyBabies Consortium, is a multi-lab approach promoting reproducibility in infant research by replicating influential experiments in developmental psychology. The first project of the MB initiative was to replicate the findings on the preference for IDS. To this end, in a worldwide collaboration project, 67 babylabs all over the world used the same methodology and materials to systematically investigate whether the results concerning the preference of IDS can be confirmed on a broader data basis. This extensive data collection allows us to determine how stable individual results from previous language acquisition studies with smaller sample sizes are. Moreover, we get more detailed insights into the influence of cultural, methodological and psychological factors on the preference for IDS.

How we investigate this in the ManyBabies Project...

The method we used in our lab within this collaborative project is called Head Turn Preference Procedure. It is based on children's tendency to direct their attention longer to things they find interesting than to things that don't interest them. In this paradigm, the child sits on a parent's lap and listens to speech stimuli, either recorded in IDS or ADS. The stimuli are presented from the right or the left side – accompanied by red blinking lights. We measure how long infants orient to the side light and the stimuli. In total, 2329 babies between 3 and 15 months in 67 babylabs of 16 different countries were tested with this or two other methods (central fixation and eye tracking). About 150 researchers were involved in this large-scale collaborative study. All labs used the same stimuli spoken in American English. In Konstanz, we tested 16 babies between 6 and 9 months.


Overall, the results confirmed that babies prefer listening to IDS over ADS. The preference for IDS is most pronounced when the IDS is in the child’s native language (i.e., for American English infants). However, even if the children hear IDS in a foreign language, they prefer it over ADS. With increasing age, the preference for IDS also increases. Among the three different methods used in different labs, the preference for IDS was strongest in the Head Turn Preference Procedure, which was also used in our lab.

Benefits of the study...

The large number of participants in different countries worldwide makes this project unique. It constitutes a basis for consecutive studies in providing the materials and data for future research. This will guarantee high comparability across studies. In addition, the results shed more light on how the preference for IDS is related to the input infants receive (language and duration of input). As such, the ManyBabies project is a great contribution to language acquisition research.


ManyBabies Consortium. 2020. Quantifying Sources of Variability in Infancy Research Using the Infant-Directed-Speech Preference. Advances in Methods and Practices in Psychological Science 3(1). 24-52.

Cooper, R. P., & Aslin, R. N. (1990). Preference for infant-directed speech in the first month after birth. Child Development, 61, 1584-1595.

The role of intonation in language acquisition

Director: Prof. Dr. Bettina Braun
Researchers: Muna Schönhuber (née Pohl), Katharina Zahner

What the study is about...

Language acquisition begins before birth. Already in the last trimester the fetus begins to perceive certain features of its mother tongue. The individual sounds are not yet recognizable in the vibrations that reach the ears, but prosodic features like intonation, rhythm and stress are. This explains why newborns are receptive to prosodic information. Then, starting at birth, the infant is exposed to the phonetic sounds of its native language. For German and Swiss German babies these sounds are much more important than intonation or rhythm. This observation poses the question, at what point do they learn to give more attention to the sounds than to rhythm and intonation.

In this project we ask, first, how German and Swiss German infants from 6 to 10 months old process the intonation of their mother tongue and that of a foreign language, and how strongly they weight this information in comparison to information about speech (phonetic) sounds. In the second part we investigate the extent to which the intonation contour helps children to identify individual words within the continuous speech signal.

How we investigate this...

The method we use for both experiments is called the Head Turn Preference Procedure. It is based on children's tendency to direct their attention longer to things they find interesting than to things that don't interest them. The child sits on a parent's lap and hears various speech recordings presented from the right or the left side. Researchers then analyze the length of attention to the various stimuli.


Overall the results showed that infants interpret a stressed syllable as stressed only if the syllable is high-pitched. In this condition, infants are able to recognize trochees in fluent speech. When the stressed syllable is low-pitched, infants fail to recognize the syllable as stressed (Zahner et al., 2016, Zahner & Braun 2018, Zahner, 2019). This pattern of results might be explained a) by the salience of high-pitched stressed syllables and b) by the frequent occurrence of high-pitched stressed syllables in German infant-directed speech (Zahner et al. 2016, KIDS Corpus.)


Zahner, K. (2019). Pitch accent type affects stress perception in German: Evidence from infant and adult speech processing, unpublished PhD Thesis. U of Konstanz

Zahner, K. & Braun, B. (2018). F0 peaks are a necessary condition for German infants’ perception of stress in metrical segmentation. Proc. of the 17th Speech Science and Technology Conference (SST 2018), Sydney, Australia, 73-76.

Zahner, K., Schönhuber, M., & Braun, B. (2016). The limits of metrical segmentation: intonation modulates infants' extraction of embedded trochees. Journal of Child Language, 43(6), 1338-1364. DOI: 10.1017/S0305000915000744.

Zahner, K., Schönhuber, M., Grijzenhout, J. & Braun, B. (2016). Konstanz prosodically annotated infant-directed speech corpus (KIDS Corpus). Proceedings of the 8th International Conference on Speech Prosody. Boston, USA. (KIDS corpus online)

Zahner, K., Pohl M. & Braun B. (2016). Pitch accent distribution in German infant directed speech. Proceedings of Interspeech 2015. Dresden, Germany.

Braun, B., Pohl M. & Zahner K. (2014). Speech segmentation is modulated by peak alignment: Evidence from German 10-month-olds. Proceedings of the 7th International Conference on Speech Prosody. Dublin, Ireland.

Sounds and syllables in different sentential intonations

Director: Prof. Dr. Janet Grijzenhout
Doctoral candidate: Anne Gwinner

The project "Sounds and syllables in different sentential intonations" is divided in two parts.

A perception experiment seeks to determine whether small children at 18, 27 and 36 months can distinguish between complex and simple word onsets. Previous studies have shown that children often have difficulty in enunciating consonant clusters. They solve the problem by dropping one of the consonants, so that, for example, *bread* turns into *bed*. Thus, one might think that these children hear no difference between *bread* and *bed*, but this is not the case. If we show them pictures corresponding to the two words and encourage them to find the bed, they look longer at the picture of a bed than at the bread. In our study we will present pairs of words to children that, like *bread*/*bed*, differ only in the initial sound, as well as words that contain small mispronunciations, like *b-yed* for *bed* or *beread* for *bread*. We want to see if German-speaking children at 18, 27 and 36 months judge the mispronunciations as equally severe and whether linguistic development can be seen with increasing age.

A production experiment investigates how nursery-age children store sentence intonation in the mental lexicon. In particular, we want to see how much morphological detail the children can reproduce. A number of experiments have shown that children prefer hearing a rhythmic pattern of alternating stressed and unstressed syllables to other rhythmic patterns. We want to see whether children omit unstressed syllables because they regard them as meaningless or because they do not fit the preferred stress pattern. Further, we want to see if all unstressed syllables are treated equally or if children distinguish among function words (articles), weak syllables in words (a-bout) and prefixes that have a grammatical function (German ge- in past participles). To this end we will be observing the word and sentence production of children growing up mono-lingual German and bilingual German-Italian. These languages allow interesting comparisons because German has intonation patterns different from those of Italian, while German allows more complex word onsets than does Italian.

For the first (perception) study with infants we will be using the Preferential Looking Paradigm. For the second (production) study with nursery children we use the Word Elicitation Task, in which particular words and sentences are elicited from children with dolls and picture books.


Gwinner, A. (2015). Early Language Acquisition and the Prosody-Morphology Interface : A Perception and Production Study with German and German-Italian Children. University of Konstanz dissertation.

Perception of phonological contrasts

Director: Prof. Janet Grijzenhout
Doctoral candidate: Muna Schönhuber (née Pohl)

The project Perception of Phonological Contrasts  studied the acquisition of contrasts in speech sounds (phonemes) in German and Swiss German, and, in particular, the plosives of both languages. By way of example: we want to find out how babies perceive small but important differences in sound, such as in the contrast *pair* / *bear*. Standard German makes a distinction between aspirated and non-aspirated plosives, while in Swiss German there is not laryngeal but rather a quantitative contrast, i. e., between long and short plosives. We investigated which phonetic features infants perceive in the speech signal and use to construct the phoneme inventory of their native language.

What concerned us particularly was the development in early phoneme perception. Currently popular theories assume that newborns and infants, up to an age of about 8 months, can distinguish all or nearly all sound contrasts -- regardless of whether these sounds are relevant for their native language or not. In the course of the first year of life the perceptual system gradually adapts to the native language, so that at 10 to 12 months only those contrasts are distinguished that carry meaning in that language.

In the BSL we tested whether German and Swiss German babies during very early acquisition (6 to 8 months) could distinguish the contrasts of their mother tongue (the laryngeal contrast in Standard German and the quantitative contrast in Swiss German), as well as the contrasts in the other language.

Further, we tested whether, in the perception of older infants (10 to 16 months), the ability to recognize the non-native contrasts had diminished. That is to say, we wanted to see whether the Swiss German infants at 10 months and older lose the ability to recognize the laryngeal contrasts that are meaningful in Standard German, and, vice versa, whether the German babies have difficulties with the Swiss German length contrasts, which are not relevant for their own language.

In all tests the Switch-Method  was employed.

The project "Perception of phonological contrasts", directed by Janet Grijzenhout  , was started in the former Sonderforschungsbereich (research group) 471, "Variation und Entwicklung im Lexikon" , and was financed originally by the DFG. Until its conclusion in 2011 it was continued outside of the SFB.


Pohl, Muna. 2011. The perception of laryngeal and length contrasts in early language acquisition. Ph.D. dissertation, University of Konstanz.

Pohl, Muna & Janet Grijzenhout. 2010. Phrase-medial bilabial stops in three West-Germanic languages. Linguistische Berichte 222. 141-167.

Pohl, Muna & Janet Grijzenhout. 2009. The perception of laryngeal and length contrasts in stops by German infants and their parents. In G.M. Socarras (ed.), Philological Explorations, 31-46. Athens: Atiner.