Attention to lexical stress and early vocabulary growth in 5-month-olds at risk for autism spectrum disorder

https://doi.org/10.1016/j.jecp.2013.08.006Get rights and content

Highlights

  • We studied 5-month-olds’ ability to discriminate word stress patterns.

  • Tested infant siblings of children with ASD and typically developing children (TD).

  • Examine whether early discrimination is related to vocabulary size at 12 months.

  • Only SIBS-TD discriminate stress patterns and this relates to later vocabulary.

  • Suggests early processing of lexical stress is related to later vocabulary size.

Abstract

Typically developing infants differentiate strong–weak (trochaic) and weak–strong (iambic) stress patterns by 2 months of age. The ability to discriminate rhythmical patterns, such as lexical stress, has been argued to facilitate language development, suggesting that a difficulty in discriminating stress might affect early word learning as reflected in vocabulary size. Children with autism spectrum disorder (ASD) often have difficulty in correctly producing lexical stress, yet little is known about how they perceive it. The current study tested 5-month-old infants with typically developing older siblings (SIBS-TD) and infants with an older sibling diagnosed with ASD (SIBS-A) on their ability to differentiate the trochaic and iambic stress patterns of the word form gaba. SIBS-TD infants showed an increased interest in attention to the trochaic stress pattern, which was also positively correlated with vocabulary comprehension at 12 months of age. In contrast, SIBS-A infants attended equally to these stress patterns, although this was unrelated to later vocabulary size.

Introduction

Before producing their first words, infants are learning the sounds and patterns of their language as they build and refine their lexicons. From birth, infants are faced with the daunting task of deciphering the speech signal through their discovery of statistical regularities and properties of speech that will help them to identify and learn new words (Jusczyk, 1997). Prosody (i.e., intonation, lexical stress, and rhythm of a language) is one such property of speech to which infants are highly sensitive. Indeed, infants can discriminate their native language from those of others within days after birth based simply on the rhythmical class of those languages (Mehler et al., 1988). This attentional sensitivity to rhythm in speech has also been shown to be positively associated with later vocabulary growth (e.g., Weber, Hahne, Friedrich, & Friederici, 2005), suggesting that atypical attention to prosody during infancy may, at the very least, be associated with delays in language development.

Typically developing (TD) infants demonstrate an attentional bias for human speech over nonspeech sounds early in development (Vouloumanos and Werker, 2004, Vouloumanos and Werker, 2007), which remains specific to speech and not just any sound produced by humans such as laughing, coughing, or sneezing (Shultz & Vouloumanos, 2010). Infants also prefer infant-directed speech to adult-directed speech within the first month of life (Cooper and Aslin, 1990, Fernald, 1985). Researchers have argued that these early preferences are likely to play an important role in infants’ linguistic and socioemotional development (Droucker et al., 2013, Shultz and Vouloumanos, 2010). In a more general way, the heightened prosodic structure of infant-directed speech may assist infants in their ability to segment speech (Thiessen, Hill, & Saffran, 2005) and learn new words (Graf Estes, 2008). However, there may be even more specific links between early speech processing and language development. For instance, Vouloumanos and Curtin (accepted for publication) found that TD infants’ longer looking to speech trials at 12 months of age positively correlates with vocabulary size at 18 months. Tsao, Liu, and Kuhl (2004) found that 6-month-olds who could discriminate various phonemes (e.g., /u/ and /i/) had superior language competence when they turned 13, 16, and 24 months of age compared with 6-month-olds who could not discriminate them. Furthermore, Bernhardt, Kemp, and Werker (2007) found that infants who successfully mapped word forms to objects at 17 and 20 months of age also demonstrated stronger vocabulary production skills up to 2½ years later. Taken together, these studies suggest a link between early speech perception skills and later vocabulary competence.

TD infants are also highly sensitive to the prosodic properties of speech such as the intonation, lexical stress, and rhythm of a language. Lexical stress, which was the main focus of the current study, refers to prominence of a syllable within a word that is characterized by duration, intensity, and pitch (Hayes, 1995). In English, the stress pattern of most words is trochaic (strong–weak [Sw] as in DOCtor, with stress denoted by uppercase letters, often found in nouns) and fewer words are iambic (weak–strong [wS] as in reCORD, often found in verbs). Indeed, for English listeners, the likelihood of encountering a content word with a strong syllable at the beginning is three to one (Cutler & Carter, 1987). English-learning infants are highly attuned to trochaic and iambic stress and can discriminate between these patterns as early as 2 months of age (Jusczyk & Thompson, 1978). Furthermore, 9-month-old English-learning infants reliably demonstrate a preference for trochaic stress (Echols et al., 1997, Jusczyk et al., 1993). By 12 months of age, infants can map labels to objects when those labels differ solely in stress (initial [Sww] vs. medial [wSw] stress; Curtin, 2009), demonstrating a potential recognition that words minimally differing in their stress can also have different meanings. At 16 months of age, infants will map an iambic pattern to path actions but will not map forms with trochaic stress (Curtin, Campbell, & Hufnagle, 2012), suggesting that they have a rudimentary understanding of which stress pattern accompanies actions versus objects. Given the prominence of trochaic words in English, it is perhaps not surprising that English-learning infants use this information to detect and learn new words in the speech signal.

In prior research, investigators have explored the linguistic purpose of this perceptual sensitivity to stress in speech during infancy. For instance, Herold, Hohle, Walch, Weber, and Obladen (2008) tested whether preterm infants at the corrected age of 4 to 6 months could discriminate between trochaic and iambic stress because this population tends to experience delays in language acquisition. They found that whereas TD infants at 4 to 6 months attended longer to trochaic stress, preterm infants did not differentially attend to either. Similarly, Weber and colleagues (2005) investigated whether infants’ impaired perception of word stress at 5 months of age could be used as a marker for specific language impairment (SLI). Indeed, they found that the infants who did not successfully discriminate between the stress patterns at 5 months also had word production scores in a range suggestive of a future SLI diagnosis at both 12 and 24 months. Taken together, these studies suggest that attention to lexical stress in speech early on in development could serve as a marker for future language difficulties as early as 12 months of age. Here, we examined whether infants at risk for autism spectrum disorder (ASD) demonstrate atypical preferences for their native language stress pattern at 4 to 6 months of age and whether this affects early vocabulary growth at 12 months.

ASD has long been associated with impairments in speech production and language development overall (American Psychiatric Association, 1994, Paul et al., 2005), with prosodic production being the most widely affected aspect (Peppé, Cleland, Gibbon, O’Hare, & Martínez-Castilla, 2011). It is important to note that the majority of children diagnosed with ASD do not receive that diagnosis until they are between 2 and 4 years of age (Bryson et al., 2008, De Giacomo and Fombonne, 1998, Rogers and DiLalla, 1990), so research investigating speech perception during infancy in this population is almost nonexistent. In an effort to examine the early years of ASD, investigators have turned to studying infant siblings of children diagnosed with the disorder because approximately 19% of these infants will also be diagnosed by 3 years of age (Ozonoff et al., 2011). Constantino, Zhang, Frazier, Abbacchi, and Law (2010) found that 20% of those infant siblings who are not diagnosed with ASD are reported by their parents to have a history of a language delay, with approximately half of them exhibiting autistic qualities of speech. This elevated rate of impairment presents a unique opportunity to study the early development of ASD and/or language disorders (Elsabbagh & Johnson, 2010). Furthermore, research of this type allows for the examination of the hypothesized importance of attention to prosody during infancy and its relationship to language acquisition.

The range of language impairments in individuals with ASD is quite heterogeneous, ranging from mutism and little functional communication to relatively well-developed skills (Wilkinson, 1998). Despite this heterogeneity, most verbal individuals with ASD tend to produce atypical prosody when speaking, ranging from monotone, to sing-song-like, to robotic, to over-exaggerated speech (Järvinen-Pasley, Peppé, King-Smith, & Heaton, 2008). Shriberg and colleagues (2001) examined prosodic production in a group of adolescents with ASD and found that they tended to use stress, loudness, pitch, and resonance in their speech inappropriately. Although it is well known that individuals with ASD struggle with producing prosody appropriately, very little is known about their perception of it (Järvinen-Pasley et al., 2008, Järvinen-Pasley et al., 2008, McCann and Peppé, 2003, Peppé et al., 2011). That said, researchers have found that there are certain aspects of prosodic perception, such as pitch, that have in some cases remained intact in ASD and even superior to that of controls (e.g., Heaton, Davis, & Happé, 2008); however, this appears to hold true more so for intellectually higher functioning groups of individuals with ASD with stronger language skills (e.g., Järvinen-Pasley and Heaton, 2007, Heaton et al., 2008). Furthermore, this ability to detect changes in pitch appears to be limited to short words, whereas the ability to understand phrases with varied prosodic contours remains considerably limited in ASD (Järvinen-Pasley, Pasley, & Heaton, 2008). Research examining speech/prosodic perception in ASD has focused primarily on individuals who are already diagnosed with the disorder and who are often higher functioning in terms of their language and intellectual functioning. Much of this research has also concentrated more on the perception of global aspects of speech (e.g., the sound of a mother’s voice) versus nonspeech sounds (e.g., white noise) (e.g., Klin, 1991) and not on specific properties of the speech signal, including prosody.

Paul and colleagues (2005) completed one of the few studies that looked at how individuals with ASD produce and perceive prosody. They tested typical adolescents as well as adolescents with ASD and found significant between-group differences on tasks involving both the production and perception of word stress (pragmatic/affective or emphatic, lexical, and grammatical). They concluded that stress might be the most negatively affected part of prosody for individuals with ASD. Relatedly, Paul, Chawarska, Fowler, Cicchetti, and Volkmar (2007) investigated preferential attention to trochaic versus iambic stress in known words (e.g., GERbil, girAFFE) among 2-year-olds with ASD, TD toddlers (age and language matched to ASD sample) and 2-year-old toddlers with a nonspecific developmental delay (DD) using a head turn preference procedure. They found that ASD and DD toddlers did not demonstrate a preference for either stress pattern at 2 years of age. Interestingly, however, the 1-year-old language-matched TD infants preferred trochaic stress, whereas the 2-year-old age-matched TD infants did not demonstrate any clear preference. Not only do these results provide evidence of early differences in lexical stress preferences between TD toddlers and toddlers with ASD, but they also suggest a developmental progression in attention to this information in typical development.

Peppé and colleagues (2011) found that language comprehension tasks that required an understanding of prosody were the most difficult to do correctly regardless of language abilities in children with ASD. Specifically, these children struggled with tasks that required knowledge of verbal punctuation or phrasing, expression of feelings and affect by changing intonation or tone of voice, and intonation that changed utterance types (e.g., phrasing a sentence as a question, invitation to continue speaking). Taken together, it appears that individuals already diagnosed with ASD have difficulty in perceiving prosody not only in its more complex and social form (paralinguistic) but also in its more rudimentary form of lexical stress.

In the current study, we examined whether early preferences for trochaic stress over iambic stress are exhibited in English-learning infants between 4 and 6 months of age. We examined whether this preference depends on risk group. That is, do infant siblings of children diagnosed with ASD prefer stress in the same way as infant siblings of typically developing children? Furthermore, is a trochaic stress preference correlated with vocabulary size at 12 months of age?

Section snippets

Participants

Infants between 4 and 6 months of age from English-speaking homes were recruited and followed to 12 months. Specifically, 30 infants (15 female and 15 male) with one or more typically developing older siblings (SIBS-TD) and 26 infants (13 female and 13 male) with one or more older siblings diagnosed with ASD (SIBS-A) were recruited. SIBS-TD infants were free of any family history of ASD. The diagnostic status of the older siblings of SIBS-A infants was confirmed by parents. All of the older

Preference for word stress at 5 months of age

For the mean looking time to the trochaic test trials, the variances were equal for both SIBS-TD and SIBS-A infants, F(1, 48) = 1.17, p > .05. This was also true for the iambic test trials, F(1, 48) = 0.80, p > .05. The data were examined for outliers (i.e., infants with mean looking times 3 standard deviations above the mean), which were subsequently removed from the analyses (SIBS-A: n = 3). There were no group differences in mean looking time to the familiarization phase of the experiment (SIBS-TD: M = 

Discussion

The results of this experiment provide important insight into the early development of both TD infants and infants at an increased risk for being diagnosed with ASD. As hypothesized, 5-month-old SIBS-TD and SIBS-A infants did not attend to lexical stress in the same manner. That is, whereas SIBS-TD infants demonstrated an attentional preference for trochaic stress over iambic stress, SIBS-A infants did not differentially attend to either stress pattern. Our findings for the SIBS-TD group are

Acknowledgments

This research was supported by funding from an Alberta Centre for Child, Family, and Community Research (ACCFCR) research grant awarded to Suzanne Curtin. We thank the parents and infants who participated as well as Jennifer Campbell, Danielle Droucker, and Melanie Khu for their assistance with this research.

References (59)

  • S.J. Rogers et al.

    Age of symptom onset in young children with pervasive developmental disorders

    Journal of the American Academy of Child and Adolescent Psychiatry

    (1990)
  • C. Weber et al.

    Reduced stress pattern discrimination in 5-month-olds as a marker of risk for later language impairment: Neurophysiologial evidence

    Cognitive Brain Research

    (2005)
  • American Psychiatric Association

    Diagnostic and statistical manual of mental disorders

    (1994)
  • B.M. Bernhardt et al.

    Early word–object associations and later language development

    First Language

    (2007)
  • Boersma, P., & Weenink, D. (2008). PRAAT: Doing phonetics by computer (Version 5.0.40). Retrieved from...
  • S.E. Bryson et al.

    The Autism Observation Scale for Infants: Scale development and reliability data

    Journal of Autism and Developmental Disorders

    (2008)
  • Cohen, L. B., Atkinson, D. J., & Chaput, H. H. (2004). A new program for obtaining and organizing data in infant...
  • J.N. Constantino et al.

    Sibling recurrence and the genetic epidemiology of autism

    American Journal of Psychiatry

    (2010)
  • R. Cooper et al.

    Preference for infant-directed speech in the first month after birth

    Child Development

    (1990)
  • R. Cooper et al.

    Developmental differences in infant attention to the spectral properties of infant-directed speech

    Child Development

    (1994)
  • S. Curtin

    Twelve-month-olds learn novel word–object pairings differing only in stress pattern

    Journal of Child Language

    (2009)
  • S. Curtin et al.

    Speech preference predicts autistic-like behavior in 18-month-olds at risk for autism spectrum disorders

    Journal of Autism and Developmental Disorders

    (2013)
  • A. De Giacomo et al.

    Parental recognition of developmental abnormalities in autism

    European Child & Adolescent Psychiatry

    (1998)
  • D. Droucker et al.

    Linking infant-directed speech and face preferences to language outcomes in infants at risk for autism spectrum disorder

    Journal of Speech, Language, and Hearing Research

    (2013)
  • H.M. Feldman et al.

    Concurrent and predictive validity of parent reports of child language at ages 2 and 3 years

    Child Development

    (2005)
  • H.M. Feldman et al.

    Measurement properties of the MacArthur Communicative Development Inventories at ages one and two years

    Child Development

    (2000)
  • Fenson, L., Dale, P. S., Reznick, J. S., Thal, D., Bates, E., Hartung, J. P., Pethick, S., & Reilly, J. S. (1993)....
  • Fenson, L., Dale, P. S., Reznick, J. S., Bates, E., Thal, D. J., & Pethick, S. J. (1994). Variability in early...
  • L.R. Gleitman et al.

    Language acquisition: The state of the art

  • Cited by (23)

    • Infant speech perception and cognitive skills as predictors of later vocabulary

      2021, Infant Behavior and Development
      Citation Excerpt :

      This is surprising, given findings from previous studies seem to suggest that infants can discriminate between trochaic and iambic stress patterns as early as 2 months (Jusczyk & Thompson, 1978), and sensitivity to stress patterns is related to vocabulary development (Ference & Curtin, 2013). For example, Ference and Curtin (2013) tested 5-month-old infants’ stress preference using a similar procedure, and collected vocabulary measures at 12 months. They showed that the trochaic stress preference score was positively correlated with receptive vocabulary.

    • Preference for speech in infancy differentially predicts language skills and autism-like behaviors

      2019, Journal of Experimental Child Psychology
      Citation Excerpt :

      SIBS-A may process speech differently than TD infants. As a group, SIBS-A do not prefer listening to their native language stress pattern at 5 months of age as do low-risk infants (SIBS-TD) (Ference & Curtin, 2013), and SIBS-A have difficulty in mapping words that differ in stress at 12 months (Ference & Curtin, 2015). However, 12-month-old SIBS-A who use lexical stress to map words have stronger expressive language at 24 months (Ference & Curtin, 2015).

    • Language Development

      2015, Neurobiology of Language
    • When phonology guides learning

      2018, Applied Psycholinguistics
    View all citing articles on Scopus
    View full text