CHILD DEVELOPMENT PERSPECTIVES
Cognitive and Neural Mechanisms Underlying
Socioeconomic Gradients in Language Development:
New Answers to Old Questions
€
Ozlem Ece Demir1 and Aylin C. K€
untay2,3
1
Northwestern University, 2Kocß University, and 3Utrecht University
ABSTRACT—Although researchers have studied disparities (Bradley & Corwyn, 2002; Brooks-Gunn & Duncan, 1997;
in early language development related to socioeconomic D’Angiulli, Lipina, & Olesinska, 2012; Hackman & Farah,
status (SES), it is unclear how early and through which 2009). Differences related to SES in young children’s linguistic
mechanisms these differences emerge. As income inequal- skills are well described (e.g., Hart & Risley, 1995; Hoff, 2003),
ity continues to widen across the world, it is crucial to and are more pronounced than other cognitive domains (Hack-
examine the child-level mechanisms that mediate the man & Farah, 2009).
effects of SES on individual differences in language devel- Research on the SES-related disparities in language develop-
opment. A deeper understanding of the nature of the dif- ment does not commonly address the changes child-level mech-
ferences will allow development of more effective anisms of language learning accrue in detrimental environments
intervention techniques. In this article, we discuss work on but emphasizes parent-level mechanisms underlying the effects
child-level cognitive and neural mechanisms underlying of SES on language outcomes (Hoff, 2003). Given that early lan-
the relation between SES and early language develop- guage predicts later reading skills and academic achievement,
ment. We discuss possible factors behind individual differ- uncovering the child-level cognitive and neural mechanisms
ences in child-level mechanisms and cascading effects of underlying the relation between SES and language development
these differences. We conclude with recommendations for is crucial (NICHD Early Child Care Research Network, 2005).
research. Such mechanistic understanding of child-level mechanisms can
inform efforts to remediate the effects of low SES during the for-
KEYWORDS—language development; socioeconomic status;
mative years of language development.
neural and cognitive mechanisms
In this article, we first focus on the accumulating literature on
the child-level cognitive and neural mechanisms that underlie
Children from lower socioeconomic status (SES) fall behind their the relation between SES and language development. We exam-
middle-class peers on many aspects of cognitive development ine SES effects within the normal range (for effects of extreme
neglect on child development, see Nelson, 2007). We then dis-
cuss factors behind individual differences in child-level mecha-
€
Ozlem Ece Demir, Department of Communication Sciences &
Disorders, Northwestern University; Aylin C. K€untay, Department
nisms and longer term cascading effects of these early
of Psychology, Kocß University, Educational and Learning Sciences, differences in later development. We end by suggesting new
Utrecht University. research to improve our understanding of SES gradients under-
We thank Lauren R. Applebaum and Matthew E. Lira for com- lying children’s language development.
ments on earlier versions of the manuscript.
Correspondence concerning this article should be addressed to CHILD-LEVEL MECHANISMS THAT UNDERLIE THE
€
Ozlem Ece Demir, Roxelyn and Richard Pepper Department of EFFECTS OF SES ON LANGUAGE DEVELOPMENT
Communication Sciences & Disorders, Northwestern University,
2240 Campus Drive, Frances Searle Building, Room 2-342, Evans-
ton, IL 60208; e-mail: . By child-level mechanisms, we refer to the sociocognitive skills
and neural processes underlying language learning, especially
© 2014 The Authors
Child Development Perspectives © 2014 The Society for Research in Child Development
those that are affected by SES and that in turn predict individ-
DOI: 10.1111/cdep.12069 ual differences in children’s language outcomes. We discuss
Volume 0, Number 0, 2014, Pages 1–6
, 2 €
Ozlem Ece Demir and Aylin C. K€
untay
behavioral and neuroimaging studies examining three types of in children. Individual differences in neural structure and func-
child-level mechanisms: communicative skills, speech- and lan- tion of these regions are important because they relate to concur-
guage-processing skills, and attentional-control skills. rent and future language skills (Deniz Can, Richards, & Kuhl,
2013; Mills, Coffey-Corina, & Neville, 1997). Regarding differ-
Communicative Skills ences in brain structure, SES is positively related to gray matter
SES-related individual differences evident in communicative volume and gyrification in language-related regions, including
abilities emerge before speech and establish a base for chil- the frontal, temporal, and hippocampal regions (Hair, 2013; Han-
dren’s later language development. Children start communicat- son et al., 2013; Jednorog et al., 2012; Noble, Houston, Kan, &
ing through gestures before they use speech, and use of early Sowell, 2012; Raizada, Richards, Meltzoff, & Kuhl, 2008). The
gestures predicts later individual differences in language growth effect of SES differences on brain structure is more pronounced
(Rowe & Goldin-Meadow, 2009). SES affects children’s early for language-relevant regions, in frontal, temporal, and hippo-
gesture production. For example, in a recent study, SES related campal areas, than for other brain areas (Jednorog et al., 2012).
to children’s production of gestures but not spoken vocabulary Furthermore, SES effects on the volume of gray matter in frontal
size, at 14 months (Rowe & Goldin-Meadow, 2009). Moreover, and temporal regions increase with age (Hanson et al., 2013;
children’s use of gestures predicted children’s spoken vocabu- Noble et al., 2012). In a study that used electroencephalogram
lary size at 54 months and mediated the relation between SES (EEG), 6- to 9-month-olds with lower SES had reduced baseline
and vocabulary outcomes at 54 months. Other early communica- resting brain activity in frontal areas compared with their higher-
tion behaviors such as vocalizations, eye-gaze, and joint atten- SES peers (Tomalski et al., 2013). Overall, the limited number
tion are also influenced by parents’ SES. In a study on the of neuroimaging studies suggests that SES-related differences in
trajectory of children’s communication-initiating behaviors language-related structures might emerge early in life, possibly
between 6 and 40 months, parents’ SES predicted the growth of before behavioral differences appear, and differences might
children’s initiating behaviors (Landry, Smith, Miller-Loncar, & increase with age.
Swank, 1997). The frequency of children’s initiating behaviors Few studies have examined SES-related differences in neural
at 18 months predicted faster growth of later language skills. processes during language tasks in children. In one, 3- to 8-
year-olds’ event-related potential (ERP) responses to semantic
Speech- and Language-Processing Skills and syntactic violations in spoken sentences were related to
Infants’ ability in processing speech sounds provides the basis SES (Pakulak, Sanders, Paulsen, & Neville, 2005). Both groups
for later developing language abilities. Foundational speech and of children showed a negative ERP component for semantic vio-
language-processing skills include the ability to differentiate lations (N400), but its amplitude and distribution varied accord-
phonetic contrasts of the native language, lexical processing effi- ing to SES. Both groups of children displayed a late positivity
ciency, and the ability to repeat words and nonwords. These (P600) for syntactic violations, but only children with higher
early skills predict later vocabulary and syntactic development SES displayed an early negativity, suggesting that the effects of
(Gathercole & Baddeley, 1990; Hurtado, Marchman, & Fernald, SES might be more pronounced for syntactic processing than for
2008; Junge, Cutler, & Hagoort, 2010; Roy & Chiat, 2013). SES semantic processing. This early negativity reflects immediate
may affect early-developing speech- and language-processing use of incoming language input to guide online syntactic parsing
skills. In one study, SES-related differences were seen in and thus greater syntactic proficiency (Pakulak & Neville,
18- and 24-month-olds’ lexical processing efficiency during 2010). A similar relation between larger amplitude negativity to
real-time language comprehension (Fernald, Marchman, & syntactic violations and childhood SES has been seen in adults
Weisleder, 2013). Lexical processing efficiency was measured (Pakulak & Neville, 2010). In a study that used functional mag-
by how quickly children turned toward the picture of a target netic resonance imaging, SES correlated with greater hemi-
object, as opposed to an alternative, when they heard the spheric specialization of the left inferior frontal gyrus during a
object’s name. Children with higher SES looked at the named rhyming task for 5-year-olds, even after controlling for behav-
target sooner and were more accurate. In another study, 3- to ioral measures of language ability (Raizada et al., 2008). These
5-year-olds with lower SES performed less accurately than their studies suggest that children with higher SES might engage lan-
higher SES peers on tasks in which they repeated words and guage-related regions of the brain more robustly and efficiently
nonwords (Roy & Chiat, 2013). during language tasks.
A small but growing body of research uses neuroimaging to
explore how SES affects the neural mechanisms underlying Attentional-Control Skills
speech and language processing. These studies reveal SES- In addition to domain-specific speech and language-processing
related differences in brain structures and neural processes rele- skills, domain-general attentional-control skills provide a basis
vant to speech and language processing, especially in frontal for language development. Language learning requires selective
(specifically inferior frontal and prefrontal areas) and temporal attention to particular verbal stimuli for further processing while
(specifically superior temporal and hippocampal regions) cortices filtering out distracting environmental stimuli (NICHD Early
Child Development Perspectives, Volume 0, Number 0, 2014, Pages 1–6
Cognitive and Neural Mechanisms Underlying
Socioeconomic Gradients in Language Development:
New Answers to Old Questions
€
Ozlem Ece Demir1 and Aylin C. K€
untay2,3
1
Northwestern University, 2Kocß University, and 3Utrecht University
ABSTRACT—Although researchers have studied disparities (Bradley & Corwyn, 2002; Brooks-Gunn & Duncan, 1997;
in early language development related to socioeconomic D’Angiulli, Lipina, & Olesinska, 2012; Hackman & Farah,
status (SES), it is unclear how early and through which 2009). Differences related to SES in young children’s linguistic
mechanisms these differences emerge. As income inequal- skills are well described (e.g., Hart & Risley, 1995; Hoff, 2003),
ity continues to widen across the world, it is crucial to and are more pronounced than other cognitive domains (Hack-
examine the child-level mechanisms that mediate the man & Farah, 2009).
effects of SES on individual differences in language devel- Research on the SES-related disparities in language develop-
opment. A deeper understanding of the nature of the dif- ment does not commonly address the changes child-level mech-
ferences will allow development of more effective anisms of language learning accrue in detrimental environments
intervention techniques. In this article, we discuss work on but emphasizes parent-level mechanisms underlying the effects
child-level cognitive and neural mechanisms underlying of SES on language outcomes (Hoff, 2003). Given that early lan-
the relation between SES and early language develop- guage predicts later reading skills and academic achievement,
ment. We discuss possible factors behind individual differ- uncovering the child-level cognitive and neural mechanisms
ences in child-level mechanisms and cascading effects of underlying the relation between SES and language development
these differences. We conclude with recommendations for is crucial (NICHD Early Child Care Research Network, 2005).
research. Such mechanistic understanding of child-level mechanisms can
inform efforts to remediate the effects of low SES during the for-
KEYWORDS—language development; socioeconomic status;
mative years of language development.
neural and cognitive mechanisms
In this article, we first focus on the accumulating literature on
the child-level cognitive and neural mechanisms that underlie
Children from lower socioeconomic status (SES) fall behind their the relation between SES and language development. We exam-
middle-class peers on many aspects of cognitive development ine SES effects within the normal range (for effects of extreme
neglect on child development, see Nelson, 2007). We then dis-
cuss factors behind individual differences in child-level mecha-
€
Ozlem Ece Demir, Department of Communication Sciences &
Disorders, Northwestern University; Aylin C. K€untay, Department
nisms and longer term cascading effects of these early
of Psychology, Kocß University, Educational and Learning Sciences, differences in later development. We end by suggesting new
Utrecht University. research to improve our understanding of SES gradients under-
We thank Lauren R. Applebaum and Matthew E. Lira for com- lying children’s language development.
ments on earlier versions of the manuscript.
Correspondence concerning this article should be addressed to CHILD-LEVEL MECHANISMS THAT UNDERLIE THE
€
Ozlem Ece Demir, Roxelyn and Richard Pepper Department of EFFECTS OF SES ON LANGUAGE DEVELOPMENT
Communication Sciences & Disorders, Northwestern University,
2240 Campus Drive, Frances Searle Building, Room 2-342, Evans-
ton, IL 60208; e-mail: . By child-level mechanisms, we refer to the sociocognitive skills
and neural processes underlying language learning, especially
© 2014 The Authors
Child Development Perspectives © 2014 The Society for Research in Child Development
those that are affected by SES and that in turn predict individ-
DOI: 10.1111/cdep.12069 ual differences in children’s language outcomes. We discuss
Volume 0, Number 0, 2014, Pages 1–6
, 2 €
Ozlem Ece Demir and Aylin C. K€
untay
behavioral and neuroimaging studies examining three types of in children. Individual differences in neural structure and func-
child-level mechanisms: communicative skills, speech- and lan- tion of these regions are important because they relate to concur-
guage-processing skills, and attentional-control skills. rent and future language skills (Deniz Can, Richards, & Kuhl,
2013; Mills, Coffey-Corina, & Neville, 1997). Regarding differ-
Communicative Skills ences in brain structure, SES is positively related to gray matter
SES-related individual differences evident in communicative volume and gyrification in language-related regions, including
abilities emerge before speech and establish a base for chil- the frontal, temporal, and hippocampal regions (Hair, 2013; Han-
dren’s later language development. Children start communicat- son et al., 2013; Jednorog et al., 2012; Noble, Houston, Kan, &
ing through gestures before they use speech, and use of early Sowell, 2012; Raizada, Richards, Meltzoff, & Kuhl, 2008). The
gestures predicts later individual differences in language growth effect of SES differences on brain structure is more pronounced
(Rowe & Goldin-Meadow, 2009). SES affects children’s early for language-relevant regions, in frontal, temporal, and hippo-
gesture production. For example, in a recent study, SES related campal areas, than for other brain areas (Jednorog et al., 2012).
to children’s production of gestures but not spoken vocabulary Furthermore, SES effects on the volume of gray matter in frontal
size, at 14 months (Rowe & Goldin-Meadow, 2009). Moreover, and temporal regions increase with age (Hanson et al., 2013;
children’s use of gestures predicted children’s spoken vocabu- Noble et al., 2012). In a study that used electroencephalogram
lary size at 54 months and mediated the relation between SES (EEG), 6- to 9-month-olds with lower SES had reduced baseline
and vocabulary outcomes at 54 months. Other early communica- resting brain activity in frontal areas compared with their higher-
tion behaviors such as vocalizations, eye-gaze, and joint atten- SES peers (Tomalski et al., 2013). Overall, the limited number
tion are also influenced by parents’ SES. In a study on the of neuroimaging studies suggests that SES-related differences in
trajectory of children’s communication-initiating behaviors language-related structures might emerge early in life, possibly
between 6 and 40 months, parents’ SES predicted the growth of before behavioral differences appear, and differences might
children’s initiating behaviors (Landry, Smith, Miller-Loncar, & increase with age.
Swank, 1997). The frequency of children’s initiating behaviors Few studies have examined SES-related differences in neural
at 18 months predicted faster growth of later language skills. processes during language tasks in children. In one, 3- to 8-
year-olds’ event-related potential (ERP) responses to semantic
Speech- and Language-Processing Skills and syntactic violations in spoken sentences were related to
Infants’ ability in processing speech sounds provides the basis SES (Pakulak, Sanders, Paulsen, & Neville, 2005). Both groups
for later developing language abilities. Foundational speech and of children showed a negative ERP component for semantic vio-
language-processing skills include the ability to differentiate lations (N400), but its amplitude and distribution varied accord-
phonetic contrasts of the native language, lexical processing effi- ing to SES. Both groups of children displayed a late positivity
ciency, and the ability to repeat words and nonwords. These (P600) for syntactic violations, but only children with higher
early skills predict later vocabulary and syntactic development SES displayed an early negativity, suggesting that the effects of
(Gathercole & Baddeley, 1990; Hurtado, Marchman, & Fernald, SES might be more pronounced for syntactic processing than for
2008; Junge, Cutler, & Hagoort, 2010; Roy & Chiat, 2013). SES semantic processing. This early negativity reflects immediate
may affect early-developing speech- and language-processing use of incoming language input to guide online syntactic parsing
skills. In one study, SES-related differences were seen in and thus greater syntactic proficiency (Pakulak & Neville,
18- and 24-month-olds’ lexical processing efficiency during 2010). A similar relation between larger amplitude negativity to
real-time language comprehension (Fernald, Marchman, & syntactic violations and childhood SES has been seen in adults
Weisleder, 2013). Lexical processing efficiency was measured (Pakulak & Neville, 2010). In a study that used functional mag-
by how quickly children turned toward the picture of a target netic resonance imaging, SES correlated with greater hemi-
object, as opposed to an alternative, when they heard the spheric specialization of the left inferior frontal gyrus during a
object’s name. Children with higher SES looked at the named rhyming task for 5-year-olds, even after controlling for behav-
target sooner and were more accurate. In another study, 3- to ioral measures of language ability (Raizada et al., 2008). These
5-year-olds with lower SES performed less accurately than their studies suggest that children with higher SES might engage lan-
higher SES peers on tasks in which they repeated words and guage-related regions of the brain more robustly and efficiently
nonwords (Roy & Chiat, 2013). during language tasks.
A small but growing body of research uses neuroimaging to
explore how SES affects the neural mechanisms underlying Attentional-Control Skills
speech and language processing. These studies reveal SES- In addition to domain-specific speech and language-processing
related differences in brain structures and neural processes rele- skills, domain-general attentional-control skills provide a basis
vant to speech and language processing, especially in frontal for language development. Language learning requires selective
(specifically inferior frontal and prefrontal areas) and temporal attention to particular verbal stimuli for further processing while
(specifically superior temporal and hippocampal regions) cortices filtering out distracting environmental stimuli (NICHD Early
Child Development Perspectives, Volume 0, Number 0, 2014, Pages 1–6
