Listen up! Developmental differences in the impact of IDS on speech segmentation

Article (PDF Available)inCognition 160 · January 2017with 87 Reads 
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
DOI: 10.1016/j.cognition.2016.12.003
Cite this publication
While American English infants typically segment words from fluent speech by 7.5-months, studies of infants from other language backgrounds have difficulty replicating this finding. One possible explanation for this cross-linguistic difference is that the input infants from different language backgrounds receive is not as infant-directed as American English infant-directed speech (Floccia et al., 2016). Against this background, the current study investigates whether German 7.5- and 9-month-old infants segment words from fluent speech when the input is prosodically similar to American English IDS. While 9-month-olds showed successful segmentation of words from exaggerated IDS, 7.5-month-olds did not. These findings highlight (a) the beneficial impact of exaggerated IDS on infant speech segmentation, (b) cross-linguistic differences in word segmentation that are based not just on the kind of input available to children and suggest (c) developmental differences in the role of IDS as an attentional spotlight in speech segmentation.
Author Note
Correspondence regarding this article should be addressed to Melanie S. Schreiner, Georg-Elias-Müller
Institute for Psychology, Research Group Psychology of Language, Georg-August-University Göttingen,
Goßlerstr. 14, 37073 Göttingen, Germany. E-mail:
Listen up! Developmental differences in the impact of IDS on speech
Melanie S. Schreiner and Nivedita Mani
Research Group Psychology of Language, Georg-August-University ttingen,
Goßlerstr.14, 37073 Germany
While American English infants typically segment words from fluent speech
by 7.5-months, studies of infants from other language backgrounds have
difficulty replicating this finding. One possible explanation for this cross-
linguistic difference is that the input infants from different language
backgrounds receive is not as infant-directed as American English infant-
directed speech (Floccia et al., 2016). Against this background, the current
study investigates whether German 7.5- and 9-month-old infants segment
words from fluent speech when the input is prosodically similar to American
English IDS. While 9-month-olds showed successful segmentation of words
from exaggerated IDS, 7.5-month-olds did not. These findings highlight a) the
beneficial impact of exaggerated IDS on infant speech segmentation, b)
cross-linguistic differences in word segmentation that are based not just on
the kind of input available to children and suggest c) developmental
differences in the role of IDS as an attentional spotlight in speech
Infant language acquisition, speech perception, word segmentation, exaggerated infant-
directed speech
German infants can segment exaggerated infant-directed speech
Exaggerated IDS as a tool to guide attention and facilitate language learning
Developmental differences in the role of IDS as an attentional spotlight in speech
One of the critical aspects of acquiring a language is the ability to segment the
fluent speech stream into its constituent units, i.e., words. In first language acquisition,
this ability seems to be in place by approximately 7.5-months, at least for American
English infants (Jusczyk & Aslin, 1995), with some studies showing even earlier
evidence of segmentation (e.g., Bortfeld, Morgan, Golinkoff, & Rathburn, 2005).
However, it has proved difficult for studies examining infants learning other native
languages to replicate such findings at the same ages. For instance, one recent study
finds that German 9-month-olds familiarized (in the laboratory) with words embedded in
fluent speech, do not differentiate these familiarized from unfamiliar control words
(Schreiner, Altvater-Mackensen, & Mani, 2016). Studies with Dutch (Kooijman, Hagoort,
& Cutler, 2005) and French infants (Nazzi, Mersad, Sundara, Iakimova, & Polka, 2014)
find similar inconsistencies with the pattern of results reported with American English
infants. Thus, French 8-month-olds familiarized with words in isolation seem unable to
recognize the same words in fluent speech, while German 9-month-olds perform
successfully in this task so long as the words tested are highly frequent function words
(Höhle & Weissenborn, 2003). In contrast, French 8-month-olds do recognize words in
isolation when previously familiarized with the same words in fluent speech. Thus, there
appears to be considerable variation in the circumstances under which infants
successfully segment words from fluent speech across languages.
Why do we find such differences? While there are likely to be considerable cross-
cultural phenomena that may underlie such behavioral differences, we focus here on
one possible explanation for the differences found across language cultures, namely, the
differences in the kind of speech presented to infants in the studies, and in their native
language, at large. Importantly, the speech presented to infants in the Jusczyk and Aslin
(1995) study, and indeed, in most studies on speech segmentation, was in the infant-
directed speech register (hereafter, IDS), the speech register typically used in
communication with young infants. It differs from speech used in normal communication
between adults, i.e., adult-directed speech (hereafter, ADS): Speech addressed to
infants is slower, higher in pitch, with longer pauses between words, and greater pitch
variation within utterances (Kuhl et al., 1997).
The use of IDS in studies with infants is well-grounded: Not only do infants show
a preference for IDS from birth onwards (Cooper, Abraham, Berman, & Staska, 1997;
Werker, Pegg, & McLeod, 1994) but they also seem to be better in extracting words from
fluent IDS compared to ADS (Singh, Nestor, Parikh, & Yull, 2009; Thiessen, Hill, &
Saffran, 2005). Furthermore, IDS appears to facilitate word learning (Graf-Estes &
Hurley, 2013; Song, Demuth, & Morgan, 2010), and its use in communication with
infants can predict vocabulary growth (Shneidman, Arroyo, Levine, & Goldin-Meadow,
2013; Weisleder & Fernald, 2013). However, it is important to note that most of this
research has been conducted with American English infants using American English
There is considerable variation in the prosodic characteristics of IDS across
languages, with different studies finding that American English IDS is the most modified
compared to ADS amongst the languages tested (Cooper, et al., 1997; Fernald et al.,
1989, Shute & Wheldall, 1989). Against this background, is it possible that the above-
mentioned studies with infants of other languages (e.g., French, Dutch, German) fail to
replicate the pattern of segmentation reported in American English infants due to the
characteristics of IDS in the different languages? Or to put it differently, given that infants
show improved segmentation of fluent speech from IDS relative to ADS (Singh et al.,
2009; Thiessen et al., 2005) and that American English IDS is more exaggerated relative
to IDS in other languages (Cooper et al., 1997, Ferguson, 1964), would we find similar
segmentation abilities in infants learning other languages if the speech input presented
to them is as exaggerated as American English IDS?
One recent study testing speech segmentation in British English infants offers
considerable support for this possibility (Floccia et al., 2016): Only one of 13
experiments found successful word segmentation, and only when the stimuli were
presented to 10.5-month-old infants in exaggerated IDS. This suggests that the different
styles of IDS used to address infants of different dialects and different languages
critically impacts their performance in segmentation tasks
. Nevertheless, this study
finds successful segmentation in infants three months later than similar findings have
been reported with American English infants. The possibility remains, therefore, that
Note that the lack of segmentation abilities in 9-month-old British English tested with American English
IDS suggests that exaggeration might not be sufficient but that the native accent is required to succeed in
segmenting speech.
infants of other languages, e.g., German, may not be able to segment words at this
younger age even given more exaggerated IDS.
Examining this possibility is critical for the following reason. On the one hand,
were infants learning other languages, e.g., German, able to segment words from fluent
speech at 7.5-months given exaggerated IDS, this would suggest that the differences
between the studies reported to-date with infants learning other languages and
American English infants come down to the input presented. In other words, infants from
different language backgrounds would be able to segment words from fluent speech at
the same age as American English infants as long as the input is adequately
exaggerated and engaging. While this might have consequences for lexical development
in infants hearing such less engaging input on a regular basis, this would at least
suggest that there is no long-term cognitive impact of hearing such less exaggerated
IDS on day-to-day language processing. Conversely, were we to find that infants
learning German are unable to segment words at 7.5-months, even given exaggerated
input, this would suggest that merely exaggerated input is inadequate to drive successful
segmentation, at least in German infants. This would further imply that there may be
other cross-cultural (including cross-linguistic) differences between infants from different
language backgrounds that induce more long-term differences in the language behavior
of these infants. Against this background, the current study sets out to explore German
7.5- and 9-month-olds’ segmentation abilities given exaggerated IDS resembling that
heard by American English infants.
Twenty-two 7.5-month-old, and 22 9-month-old monolingual German infants
participated in the study (Appendix.A).
Material and Design
Four passages with one of four phonotactically legal German monosyllabic
pseudowords, Jopp [ˈjɔp], Riel [ri:l], Mauf [mauf], and Lenn [lɛn], were recorded in an
exaggerated speech register resembling American English IDS (Table 1; APPENDIX B).
The same female speaker recorded five different isolated tokens of each pseudoword
which were repeated three times to form lists of 15 tokens. Stimuli were selected for
their acoustic properties to match those of American English IDS (Figure 1).
Table 1 Mean, minimum, and maximum fundamental frequency in Hz and mean duration in s for the
passages and isolated tokens of the study. Standard deviations are provided in brackets. For the
recordings of the exaggerated German IDS stimuli, a female native speaker of German imagined
herself as speaking to a child. In addition, she was asked to produce the passages and isolated
tokens in a slower and more exaggerated way than she typically would.
mean F0
min F0
max F0
mean duration
299.33 (22.06)
149.13 (36.76)
440.61 (31.53)
35.12 (3.24)
isolated tokens
322.63 (80.01)
266.65 (86.33)
377.02 (96.67)
22.77 (0.33)
Figure 1 Mean, minimum and maximum fundamental frequency for typical German IDS, typical English IDS
(taken from Fernald et al., 1989), and the exaggerated German IDS stimuli used in the current study.
A trained experimenter controlled the experiment from the adjacent room using
the stimulus-presenting software Look (Meints & Woodford, 2008). During each trial,
infants were presented with a blinking checkerboard on screen whilst simultaneously
being presented with an auditory stimulus. Using silent video images of the infant, the
experimenter initiated a trial when the infant looked towards the screen and continued to
indicate throughout the remainder of the trial whether the infant was looking towards the
screen or away by pressing a corresponding button on the keyboard. The auditory and
visual stimulus continued to play either until the trial was complete or until the infant
looked away for more than 2 s (see Mani & Paetzold, in press, for an identical
procedure). The experimenter was blind to the experimental condition as no information
mean min max
Frequency (Hz)
exaggerated German IDS
typical German IDS
typical American IDS
on the stimuli being presented was provided by the computer and the stimuli played in
the adjacent booth were masked by music.
Familiarization Phase. Infants listened to alternating blocks of two passages in
exaggerated IDS. Passages were either repeated for a total of 12 times or until the child
had accumulated 100 s of listening time for both passages.
Test Phase. Infants were presented with isolated tokens of the words they had
heard embedded in passages during the familiarization phase and control words they
had never heard before. Each infant received three trials of isolated tokens of either the
two familiarized, or the two control words, i.e., totalling 12 trials. Trial order within test
blocks was randomized.
Test Phase. A repeated-measures ANOVA with the within-subject factor
familiarity (familiarized vs. control word) and the between-subject factor age (7.5 vs. 9
months) revealed a significant interaction of familiarity and age (F(1, 42)=4.11, p=0.049,
ƞp2=0.09) and a significant main effect of age (F(1, 42)=4.70, p=0.036, ƞp2=0.10). There
was no significant main effect of familiarity (F(1, 42)=1.13, p=0.293, ƞp2=0.03). Hence,
we ran planned contrasts within each age-group to further examine infants’
segmentation abilities. For the 7.5-month-olds, there were no significant differences
between listening times to familiarized and control items (t(43)=-0.93, p=0.357, d=-0.14).
However, 9-month-olds listened significantly longer to the familiarized relative to the
control words (t(43)=2.99, p=0.005, d=0.45) indicating successful word segmentation
(Figure 2; APPENDIX C). Thus, our results suggest that German infants at 9-months
benefit from exaggerated speech in segmenting the speech stream, whereas 7.5-month-
olds did not show a similar benefit.
Figure 2 Difference scores for the mean listening times of the familiarized and the novel control words for
the 7.5- and 9-month-old infants.
Familiarization Phase. Comparing infants’ mean listening times to the
familiarization trials (Table 2), an independent-samples t-test revealed a significant
difference between 7.5- and 9-month-olds (t(42)=3.67, p=0.001, d=1.11). Thus while
7.5-month-olds listened longer to the familiarization trials relative to the 9-month-olds, it
appears that they looked away less than the 9-month-olds, thereby initiating fewer trials
during the familiarization phase.
Table 2 Infants‘ mean listening times (s) for the familiarization phase. Standard deviations are provided in
age group
mean listening time
mean number of trials
26.85 (8.48)
4.68 (2.17)
18.41 (6.67)
6.77 (2.72)
Previous studies on infants’ speech segmentation report that infants from
language backgrounds other than American English do not seem able to segment words
from fluent speech to the same degree as American English infants (e.g., British English:
Floccia et al., 2016; Dutch: Junge, Cutler, & Hagoort, 2014; however, note Spanish and
Catalan infants already show segmentation abilities around 6-months of age: Bosch,
Figueras, Teixidó, & Ramon-Casas, 2013). For instance, German infants are able to
successfully segment words from fluent speech only under certain conditions, e.g., when
familiarized with isolated tokens of highly frequent function words (9 months: Höhle &
Weissenborn, 2003), presented with accentuated words (10 months: Braun, Pohl, &
Zahner, 2014), previously familiarized with similar-sounding words (Altvater-Mackensen
& Mani, 2013), or tested with words previously familiarized at home (Schreiner et al.,
2016). Similarly, British infants showed segmentation of words from fluent speech only
when presented with exaggerated IDS, similar to American English IDS (Floccia et al.,
2016), but again, only at 10.5-months. In contrast, American English infants succeed in
this task already at 7.5-months without any additional cues (Jusczyk & Aslin, 1995).
Against this background, we examined whether more pronounced IDS also facilitates
word segmentation in younger German-learning infants.
The main finding of the study was that 9-month-old infants listened longer to the
familiarized words relative to the control words suggesting that infants indeed
recognized these words after an exaggerated IDS familiarization phase. Seven-and-a-
half-month-old infants did not listen longer to familiarized words, even when familiarized
in exaggerated IDS.
On the one hand, the findings with the 9-month-olds contrast previous studies
with German infants (Schreiner et al., 2016), for instance, with 9-month-olds only
listening longer to familiarized relative to control words when familiarized with these
words embedded in stories over a six-week period at home but not when presented with
a brief 100s familiarization phase. The stimuli in the Schreiner et al. (2016) study were,
however, in standard German IDS and not the exaggerated IDS presented to infants in
the current study. Thus, it is likely, that the difference in the findings can be attributed to
the speech register presented to infants across the two studies. This echoes findings
from British 10.5-month-olds (Floccia et al., 2016) while highlighting that even at a
younger age, exaggerated IDS positively impacts speech segmentation.
The results of the current study, taken together with the results reported by
Floccia et al. (2016) point to at least one potential factor underlying the cross-
linguistic/dialectal differences in speech segmentation in infants from different language
backgrounds and highlight again, the importance of IDS in early language development.
Our findings reveal differences in the ability to segment words from fluent speech
at 7.5- and 9-months as infants in the younger group failed to show significant
differences in listening times to familiarized and control tokens. This finding has
important implications for our understanding of the cross-linguistic differences in early
speech segmentation. Firstly, this suggests that at the same age at which American
English infants successfully segment words from fluent speech German infants fail to
show evidence of segmentation despite being provided with exaggerated speech input.
This places some limitations on the conclusions drawn by Floccia et al. (2016) and the
results with 9-month-olds in the current study as to the facilitatory impact of exaggerated
IDS on speech segmentation. Thus, it does not appear that presenting exaggerated IDS
alone induces successful segmentation in younger infants. What, then, might explain the
differences in performance between German and American English infants?
One possibility for the difference between the 7.5-month-olds and the 9-month-
olds in the current study is the difference in looking times during the familiarization
phase. 7.5-month-olds listened longer to the familiarization trials initiating fewer look-
aways than the 9-month-olds. Hence, 7.5-month-olds might not have learned the
relationship between their look-aways and stimulus presentation. It might, therefore, be
that the absence of a difference between listening times to familiarized and control
words at 7.5-months of age is due to their not performing as required in the task.
However, we note, that even 7-month-old German infants successfully discriminate
between familiarized and control words in this task given additional familiarization input
(Altvater-Mackensen & Mani, 2013). Thus, while we cannot exclude the possibility that
the 7.5-month-olds in the current task were not, in general, performing as expected, it is
unlikely that the lack of a significant difference in listening times to familiarized and
control words is solely due to this factor.
A second, more tantalizing, possibility is that the difference may lie in the
language backgrounds of the two groups of infants, including very likely, the speech
register used to address infants in the two languages. Might the absence of evidence for
segmentation in 7.5-month-olds be indicative of more long-lasting differences between
infants from the two language backgrounds that cannot be nullified by merely presenting
infants with more exaggerated speech input, as at 9-months of age? Here, we include
not just the differences in the kind of IDS presented to infants from the two language
backgrounds but also the degree of lexical and morphosyntactic complexity in the two
languages, as well as cultural differences in parent-child interactions. At the very least,
the difference between the 7.5- and 9-month-olds suggests that merely the presentation
of more exaggerated input does not induce successful segmentation in German infants
across development. This raises the question whether the findings of Floccia et al.
(2016) could be replicated with younger British infants, e.g., at 7.5-months, and the
extent to which exaggerated IDS induces successful segmentation in British infants
across development.
It is, however, important to note that we in no way imply that German infants
are unable to segment words at the same age as American English infants. We note that
the effect size reported for the 7.5-month-old age group is close to the meta-analytic
effect size by Bergmann and Cristia (2016). Indeed, as we can see from Figure 2, a
proportion of the 7.5-month-olds show a similar pattern as 9-month-olds, with increased
listening times to familiarized relative to novel control words. Furthermore, previous
results from our lab suggest that even younger German infants are able to segment
words from fluent speech provided they have additional cues. Taken together, this
suggests that our findings can only be taken to conclude that German infants may
require different kinds of/increased exposure to speech relative to American English
infants to show successful segmentation in this task. Against this background, ongoing
studies in our lab are currently employing an ERP-task to examine speech segmentation
in younger infants in greater detail.
The results of the current study speak to the role of IDS as an attentional spotlight
in speech processing (Kuhl, 2007; Zangl & Mills, 2007). In Altvater-Mackensen and Mani
(2013), the ability to segment similar-sounding words from fluent speech was interpreted
in terms of word-form familiarity bootstrapping segmentation. The similarity of the to-be-
segmented words to the previously familiarized words captures infants’ attention in the
otherwise unfamiliar speech stream and drives segmentation. In Schreiner et al. (2016),
recognition of familiarized words correlated significantly with infants’ attention to the
stories in ADS highlighting again the importance of attraction to speech in order to learn.
Similarly, our finding that IDS influences at least 9-month-olds segmentation of
speech can be interpreted as the exaggerated speech input facilitating segmentation by
capturing infants’ attention to a greater extent than other less exaggerated input.
IDS may therefore function as an ostensive cue that alerts the infant to a
referential communication that is directed towards her (Saint-Georges et al., 2013), even
during sleep at least in neonates (Saito, Aoyama, Kondo, Fukomoto, & Konishi, 2007).
Our findings with the 9-month-olds support the idea that prosody is an important
contributor to early language processing that assists infants’ development of
segmentation abilities (Morgan, 1996). We note that these findings are similar to those
reported with British English infants (Floccia et al., 2016), albeit at a younger age. Our
findings with the 7.5-month-olds, in contrast, suggest that merely exaggerated speech
may not be adequate at all ages to drive segmentation of speech, at least in infants from
German language backgrounds and highlight the need for future studies to examine the
reasons for the differences in segmentation in infants from different language
This work was founded by the German Excellence Initiative Award to Georg-
August-Universität Göttingen (Third funding line: Institutional Strategy). We also thank
parents and babies for their participation. Finally, we thank the editor and anonymous
reviewers for insightful and constructive feedback on earlier versions of this paper.
Altvater-Mackensen, N., & Mani, N. (2013). Word-form familiarity bootstraps infant speech
segmentation. Developmental Science, 16(6), 980-990.
Bergmann, C. & Cristia, A. (2016), Development of infants' segmentation of words from native
speech: a meta-analytic approach. Dev Sci, 19: 901917.
Bortfeld, H., Morgan, J. L., Golinkoff, R. M., & Rathbun, K. (2005). Mommy and me: Familiar
names help launch babies into speech steam segmentation. Psychological Science, 4,
Bosch, L., Figueras, M., Teixidó, M., & Ramon-Casas, M. (2013). Rapid gains in segmenting
fluent speech when words match the rhythmic unit: Evidence from infants acquiring
syllable-timed languages. Frontiers in Psychology, 4, 106.
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.
Cooper, R. P., Abraham, J., Berman, S., & Staska, M. (1997). The development of infants'
preference for motherese. Infant Behavior and Development, 20, 477-488.
Ferguson, C. A. (1964). Bay talk in six languages. American Anthropologist, 66, 103-114.
Fernald, A., Taeschner, T., Dunn, J., Papousek, M.,de Boysson-Bardies, B., & Fukui, I. (1989). A
cross-language study of prosodic modifications in mothers' and fathers' speech to
preverbal infants. Journal of Child Language, 16, 477-501.
Floccia, C. Keren-Portnoy, T., DePaolis, R., Duff, H., Delle Luche, C., Durrant, S., White, L.,
Goslin, J., & Vihman, M. M. (2016). British infants segment words only with exaggerated
infant-directed speech stimuli. Cognition, 148, 1-9.
Graf Estes, K., & Hurley, K. (2013). Infant-directed prosody helps infants map sounds to
meanings. Infancy, 18(5), 797-824.
Höhle, B., & Weissenborn, J. (2003). German-learning infants‘ ability to detect unstressed
closed-class elements in continuous speech. Developmental Science, 6, 122-127.
Junge, C., Cutler, A. & Hagoort, P. (2014). Successful word recognition by 10-month-olds given
continuous speech both at initial exposure and test. Infancy, 19(2), 179-193.
Jusczyk, P. W., & Aslin, R. N. (1995). Infants’ detection of the sound patterns of words in fluent
speech. Cognitive Psychology, 29, 1-23.
Kooijman, V., Hagoort, P. and Cutler, A. (2005). Electrophysiological evidence for prelinguistic
infants' word recognition in continuous speech. Cognitive Brain Research, 24(1), 109 -
Kuhl, P. K. (2007). Is speech learning ‘gated’ by the social brain? Developmental Science, 10,
Kuhl, P. K., Andruski, J. E., Chistovich, I. A., Chistovich, L. A., Kozhevnikova, E. V., Ryskina, V.
L., Stolyarova, E. I., Sundberg, U., & Lacerda, F. (1997). Cross-language analysis of
phonetic units in language addressed to infants. Science, 277, 684-686.
Mani, N., & Pätzold, W. (2016). Sixteen-month-olds segment words from adult directed speech.
Language Learning and Development 12, 499-508.
Meints, K., & Woodford, A. (2008). Lincoln Infant Lab Package 1.0: A new programme package
for IPL, Preferential Listening, Habituation and Eyetracking. [WWW document: Computer
software & manual]. Available from
Morgan, J. L. (1996). Prosody and the roots of parsing. Language and Cognitive Processes, 11,
Nazzi, T., Mersad, K., Sundara, M., Iakimova, G., & Polka, L. (2014). Early word segmentation in
infants acquiring Parisian French: task-dependent and dialect-specific aspects. Journal of
Child Language, 41, 600-633.
Saint-Georges C., Chetouani M., Cassel R., Apicella F., Mahdhaoui A., Muratori F., Laznik, M.,
& Cohen, D.. (2013). Motherese in interaction: At the cross-road of emotion and
cognition? (A systematic review). Plos One, 8(10), 1-17.
Saito., Y., Aoyama, S., Kondo, T., Fukomoto, R., & Konishi N. (2007). Frontal cerebral blood
flow change associated with infant-directed speech. Archives of disease in childhood:
Fetal and neonatal edition, 92, 113-116.
Schreiner, M. S., Altvater-Mackensen, N., & Mani, N. (2016). Early word segmentation in
naturalistic environments: Limited effects of speech register. Infancy, 21(5), 625-747.
Singh, L., Nestor, S., Parikh, C., & Yull, A. (2009). Influences of infant-directed speech on
early word recognition. Infancy, 14(6), 654-666.
Shneidman, L. A., Arroyo, M. E., Levine, S, & Goldin-Meadow, S. (2013). What counts as
effective input for word learning? Journal of Child Language, 40(3), 672-86.
Shute, B., & Wheldall, K. (1989). Pitch alterations in British motherese: Some preliminary
acoustic data. Journal of Child Language, 16, 503512.
Song, J. Y., Demuth, K., & Morgan, J. (2010). Effects of the acoustic properties of infant-directed
speech on infant word recognition. Journal of the Acoustical Society of America, 128,
Thiessen, E. D., Hill, E. A., & Saffran, J. R. (2005). Infant-directed speech facilitates word
segmentation. Infancy, 7(1), 53-71.
Weisleder, A. & Fernald, A. (2013). Talking to children matters: Early language experience
strengthens processing and builds vocabulary. Psychological Science, 24(11), 2143-
Werker, J. F., Pegg, J. E., & McLeod, P. J. (1994). A cross-language investigation of infant
preference for infant-directed communication. Infant Behavior and Development, 17, 323-
Zangl, R., & Mills, D.L. (2007). Brain activity to infant versus adult directed speech in 6- and 13-
month olds. Infancy, 11, 31-62.
APPENDIX A. Participants of the study.
age (days)
range (days)
additional infants2
7.5-month-old group
9-month-old group
APPENDIX B. Passages of the familiarization phase.
1 Das Jopp schmeckt sehr lecker.
2 Ein Jopp schmilzt in der Sonne.
3 Von dem köstlichen Jopp gibt es viele Sorten.
4 Das rote Jopp riecht nach Erdbeere.
5 Am Imbiss kauft Tom Gabi ein großes Jopp.
6 Mit Sahne wird das Jopp cremig.
1 Der Riel dient als Schutz.
2 Ein Riel ist ein eckiges Holz.
3 Wenn Mona den Riel bewegt, wird es lustig.
4 Ein Stück Riel schützt den Tisch.
5 Auf dem Boden liegt der schöne braune Riel.
6 Im Laden kann man Riel kaufen.
1 Der Mauf liest eine Geschichte.
2 Ein Mauf steigt die Treppe hinab.
3 Sobald er den Mauf hört, freut er sich.
4 Der tolle Mauf ist sehr mutig.
5 Im Dunkeln leuchten die grünen Augen des Mauf.
6 Dort draußen wohnt der Mauf allein.
1 Das Lenn ist eine Pflanze.
2 Ein Lenn hat eine große Blüte.
3 Da Pia das Lenn vergaß, ist es eingegangen.
4 Das pinke Lenn hat keine Blätter.
5 In der Erde stecken die Zwiebeln des Lenn.
6 Der Topf passt dem Lenn gut.
An additional 13 infants were tested but had to be removed from the analysis for different reasons
(unsteadiness (n=4), inability to finish the experiment (n=3), disturbance through toys (n=2), bilingualism
(n=2), and technical problems (n=1), looking times during the test phase more than two SDs away from
mean (n=1)).
APPENDIX C. Mean listening times in s for the familiarized and the control words for the
7.5- and 9-month-old age group. Standard deviations are provided in brackets.
age group
familiarized word
control word
12.02 (4.89)
12.41 (5.43)
10.00 (3.70)
8.74 (4.00)
  • Preprint
    Across cultures, infants are typically addressed using a special speech register, called infant-directed speech (IDS). Infants appear to benefit from being addressed in this register, although there seem to be some cross-linguistic differences in their learning from IDS. One possible explanation for these differences is that children from different backgrounds are addressed in a less exaggerated register than standard American English IDS. Against this background, we examined whether German 7.5-month-olds are able to segment words from exaggerated IDS and standard German IDS. Furthermore, in order to evaluate the potential long-term consequences of continued exposure to less or more exaggerated IDS, we also examined the impact of individual differences in maternal IDS on early speech segmentation. We found that 7.5-month-old infants were able to segment words from exaggerated, but not from standard German IDS. A potential explanation for this result comes from our finding that infants whose mothers used more exaggerated IDS were able to also segment words from standard German IDS. Taken together, the current study a) underlines the importance of IDS in early language acquisition, b) demonstrates successful exaggerated IDS segmentation abilities, and c) presents a possible explanation for cross-linguistic differences in infants’ segmentation abilities reported in the literature.
  • Article
    Word segmentation plays a crucial role in language acquisition, particularly for word learning and syntax development, and possibly predicts later language abilities. Previous studies have suggested that this ability develops differently across languages, possibly affected by the languages' rhythmic properties (Rhythmic Segmentation Hypothesis) and target word location in the prosodic structure (Edge Hypothesis). The present study investigates early word segmentation in a language, European Portuguese, that exhibits both stress-and syllable-timed properties, as well as strong cues to both higher-level prosodic boundaries and the word level. Infants aged 4-10 months old were tested with target words located in utterance-medial and utterance-final positions. Evidence for word segmentation was found early in development but only for utterance-edge located target words, suggesting the more salient prosodic cues play a crucial role. There was some evidence for segmentation in utterance-medial position by 10 months, demonstrating that this ability is not yet fully developed, possibly due to mixed rhythmic properties.
  • Preprint
    Familiarity with a word-form has been identified as one source of information infants use in detecting and segmenting individual words from continuous speech (Altvater-Mackensen & Mani, 2013). Here we extend this finding to examine whether prior familiarity with the phonological form and meaning of a word (based on natural language exposure) influences segmentation of words from fluent speech. The current study tested infants’ segmentation of pseudo-words that sound similar to words infants were likely to already be familiar with, given their natural exposure to language at home. A word-comprehension task confirmed that infants were not only familiar with the chosen word-forms but also their meaning. Results of the segmentation task suggest that infants are able to use their knowledge of previously familiar words to segment similar-sounding words from the speech stream. Thus, word-form similarity based segmentation is a powerful mechanism that can drive infants’ rapid vocabulary growth. In addition, the study confirms that 7-month-old infants already comprehend simple frequent words and are able to distinguish them from subtle mispronunciations of these words.
  • Article
    Most children can produce a few words by the end of their first year and rapidly acquire almost 30 times as many words in the following year. Although this general pattern remains the same for children learning different languages, the words individual children know are considerably different. In this article, we consider the possibility that children are an important source of variability in early vocabulary acquisition in the context of curiosity‐driven approaches to language learning. In particular, we review research that supports two interrelated claims: that what children know and what children are interested in interact in shaping what children learn. We suggest that this, as well as the possibility that children are motivated intrinsically to learn language, sets the stage for early vocabulary learning.
  • Article
    One of the first challenges facing the young language learner is the task of segmenting words from a natural language speech stream, without prior knowledge of how these words sound. Studies with younger children find that children find it easier to segment words from fluent speech when the words are presented in infant-directed speech, i.e., the kind of speech typically directed toward infants, compared to adult-directed speech. The current study examines whether infants continue to display similar differences in their segmentation of infant- and adult-directed speech later in development. We show that 16-month-old infants successfully segment words from a natural language speech stream presented in the adult-directed register and recognize these words later when presented in isolation. Furthermore, there were no differences in infants’ ability to segment words from infant- and adult-directed speech at this age, although infants’ success at segmenting words from adult-directed speech correlated with their vocabulary size.
  • Article
    Full-text available
    We examined 7.5-month-old infants' ability to segment words from infant- and adult-directed speech (IDS and ADS). In particular, we extended the standard design of most segmentation studies by including a phase where infants were repeatedly exposed to target word recordings at their own home (extended exposure) in addition to a laboratory-based familiarization. This enabled us to examine infants' segmentation of words from speech input in their naturalistic environment, extending current findings to learning outside the laboratory. Results of a modified preferential-listening task show that infants listened longer to isolated tokens of familiarized words from home relative to novel control words regardless of register. However, infants showed no recognition of words exposed to during purely laboratory-based familiarization. This indicates that infants succeed in retaining words in long-term memory following extended exposure and recognizing them later on with considerable flexibility. In addition, infants segmented words from both IDS and ADS, suggesting limited effects of speech register on learning from extended exposure in naturalistic environments. Moreover, there was a significant correlation between segmentation success and infants' attention to ADS, but not to IDS, during the extended exposure phase. This finding speaks to current language acquisition models assuming that infants' individual attention to language stimuli drives successful learning.
  • Article
    Infants start learning words, the building blocks of language, at least by 6 months. To do so, they must be able to extract the phonological form of words from running speech. A rich literature has investigated this process, termed word segmentation. We addressed the fundamental question of how infants of different ages segment words from their native language using a meta-analytic approach. Based on previous popular theoretical and experimental work, we expected infants to display familiarity preferences early on, with a switch to novelty preferences as infants become more proficient at processing and segmenting native speech. We also considered the possibility that this switch may occur at different points in time as a function of infants' native language and took into account the impact of various task- and stimulus-related factors that might affect difficulty. The combined results from 168 experiments reporting on data gathered from 3774 infants revealed a persistent familiarity preference across all ages. There was no significant effect of additional factors, including native language and experiment design. Further analyses revealed no sign of selective data collection or reporting. We conclude that models of infant information processing that are frequently cited in this domain may not, in fact, apply in the case of segmenting words from native speech.
  • Conference Paper
    Full-text available
    In two headturn preference experiments, we tested whether German infants' segmentation strategies are sensitive to the position of a pitch peak relative to the stressed syllable. Specifically, we compared target words with early-peak accents (where the pitch peak is early with respect to the stressed syllable, i.e. H+L* accents) and medial-peak accents (where the pitch peak is aligned with the stressed syllable, i.e. H* accents). Such differences in accent type signal mostly pragmatic distinctions in German, such as the difference between contextually new and recoverable information. We familiarized infants with target words produced with one of the two intonation conditions; target words were embedded in sentences. We measured looking times to lists of trochaic part-words that were either embedded in target words or were novel to them. Results showed an effect of familiarity only in the medial-peak condition, suggesting that German infants at 10 months of age are very sensitive to pitch information for segmenting running speech.
  • Article
    Most words that infants hear occur within fluent speech. To compile a vocabulary, infants therefore need to segment words from speech contexts. This study is the first to investigate whether infants (here: 10-month-olds) can recognize words when both initial exposure and test presentation are in continuous speech. Electrophysiological evidence attests that this indeed occurs: An increased extended negativity (word recognition effect) appears for familiarized target words relative to control words. This response proved constant at the individual level: Only infants who showed this negativity at test had shown such a response, within six repetitions after first occurrence, during familiarization.
  • Article
    Adults typically use an exaggerated, distinctive speaking style when addressing infants. However, the effects of infant-directed (ID) speech on infants' learning is not yet well understood. This research investigates how ID speech affects how infants perform a key function in language acquisition, associating the sounds of words with their meanings. Seventeen-month-old infants were presented with two label-object pairs in a habituation-based word learning task. In Experiment 1, the labels were produced in adult-directed (AD) speech. In Experiment 2, the labels were produced in ID prosody; they had higher pitch, greater pitch variation, and longer durations than the AD labels. We found that infants failed to learn the labels in AD speech, but succeeded in learning the same labels when they were produced in ID speech. Experiment 3 investigated the role of variability in learning from ID speech. When the labels were presented in ID prosody with no variation across tokens, infants failed to learn them. Our findings indicate that ID prosody can affect how readily infants map sounds to meanings and that the variability in prosody that is characteristic of ID speech may play a key role in its effect on learning new words.
  • Article
    Full-text available
    Various aspects of motherese also known as infant-directed speech (IDS) have been studied for many years. As it is a widespread phenomenon, it is suspected to play some important roles in infant development. Therefore, our purpose was to provide an update of the evidence accumulated by reviewing all of the empirical or experimental studies that have been published since 1966 on IDS driving factors and impacts. Two databases were screened and 144 relevant studies were retained. General linguistic and prosodic characteristics of IDS were found in a variety of languages, and IDS was not restricted to mothers. IDS varied with factors associated with the caregiver (e.g., cultural, psychological and physiological) and the infant (e.g., reactivity and interactive feedback). IDS promoted infants' affect, attention and language learning. Cognitive aspects of IDS have been widely studied whereas affective ones still need to be developed. However, during interactions, the following two observations were notable: (1) IDS prosody reflects emotional charges and meets infants' preferences, and (2) mother-infant contingency and synchrony are crucial for IDS production and prolongation. Thus, IDS is part of an interactive loop that may play an important role in infants' cognitive and social development.
  • Article
    At about 7 months of age, infants listen longer to sentences containing familiar words - but not deviant pronunciations of familiar words (Jusczyk & Aslin, 1995). This finding suggests that infants are able to segment familiar words from fluent speech and that they store words in sufficient phonological detail to recognize deviations from a familiar word. This finding does not examine whether it is, nevertheless, easier for infants to segment words from sentences when these words sound similar to familiar words. Across three experiments, the present study investigates whether familiarity with a word helps infants segment similar-sounding words from fluent speech and if they are able to discriminate these similar-sounding words from other words later on. Results suggest that word-form familiarity may be a powerful tool bootstrapping further lexical acquisition.
  • Article
    Full-text available
    Infants differ substantially in their rates of language growth, and slow growth predicts later academic difficulties. In this study, we explored how the amount of speech directed to infants in Spanish-speaking families low in socioeconomic status influenced the development of children's skill in real-time language processing and vocabulary learning. All-day recordings of parent-infant interactions at home revealed striking variability among families in how much speech caregivers addressed to their child. Infants who experienced more child-directed speech became more efficient in processing familiar words in real time and had larger expressive vocabularies by the age of 24 months, although speech simply overheard by the child was unrelated to vocabulary outcomes. Mediation analyses showed that the effect of child-directed speech on expressive vocabulary was explained by infants' language-processing efficiency, which suggests that richer language experience strengthens processing skills that facilitate language growth.