How HANDy are baby signs? A systematic review of the impact of gestural communication on typically developing,hearing infants under the age of 36 months

Abstract

Baby sign language is advocated to improve children’s communication development. However, the evidence to support the advantages of baby sign has been inconclusive. A systematic review was undertaken to summarize and appraise the research related to the effectiveness of symbolic gestures for typically developing, hearing infants with hearing parents. Eight electronic databases, reference lists, and websites were searched for relevant studies published from January 1990 to February 2013. Studies were included if they involved typically developing hearing children who were exposed to symbolic gestures before age 36 months. Of 1902 identified citations, 10 reports met the inclusion criteria. This review shows that the effectiveness of baby sign in improving communication development remains unclear. However, no evidence was identified to suggest that using baby sign interferes with typical child development.

Keywords

Baby sign development children gestures infants language review

Introduction

Parents want the best for their children and are interested in programs and products that support and promote early child development. Language acquisition is an exciting and crucial phase in children’s development as they learn to interact, communicate, and transmit their thoughts and desires. Typically, a child’s first words appear at approximately 12 months of age, while word combinations are observed during the second year of life (Jusczyk, 1997). However, communication intents can be evident as early as 6 months. A gap between a child’s desire to communicate and their ability to do so by using spoken words is well documented (Bates & Dick, 2002; Iverson & Goldin-Meadow, 2005; Rowe & Goldin-Meadow, 2009).

Global motor development occurs much earlier than oral language and the first gesture (generally pointing) occurs at approximately 8–10 months (Vallotton, 2008). In fact, gestures are inextricably connected to language development (Acredolo, Goodwyn, Horobin, & Emmons, 1999; Bates & Dick, 2002). For example, studies have shown an association between gesture use in early childhood and later verbal vocabulary size as well as between gesture–speech combined utterances, the age of two-word utterance production, and later sentence complexity (Iverson & Goldin-Meadow, 2005; Iverson, Capirci, Volterra, & Goldin-Meadow, 2008; Rowe & Goldin-Meadow, 2009). Recent research continues to lend support to the benefits of gestures in terms of facilitating parent–child interaction and early communication development. For example, Olson and Masur (2013) found that mothers responded more frequently and used more verbal responses when infants’ communicative attempts involved gestures. Parents’ positive responses to their infants’ gestures have been shown to increase infants’ attempts at communication through both gestures and vocalizations (Miller & Lossia, 2013). These studies suggest that infant gesture use and parents’ subsequent responsiveness lead to joint-attention opportunities, which have long been linked with positive language outcomes (Tomasello & Farrar, 1986).

Two types of early child gestures are commonly identified: deictic gestures and symbolic or representational gestures (Capirci, Iverson, Pizzuto, & Volterra, 1996). Deictic gestures, such as pointing and reaching, are among the first gestures produced by children (Locke, 2007). Symbolic gestures are generally defined as hand movements that take on a form or function of items and can stand alone in much the same way as words. They are representative of objects and actions and can be used to label, describe, and request (Goodwyn & Acredolo, 1998). They can also be observed used in combination with spoken words to form short utterances (Capirci et al., 1996). An example of a symbolic gesture is the motion of hands flapping to represent ‘bird’. Symbolic gestures, as considered in this review, are also often referred to as baby signs when caregivers deliberately and specifically provide enhanced gesture training to infants to promote early communication development. These baby signs differ from established sign languages used by individuals who are deaf in that they are not intended to be used as a language or as a replacement for oral vocabulary but rather only as a symbolic support to spoken language in early childhood.

Due to the early development of gross motor skills, the use and teaching of symbolic gestures to infants before their ability to communicate orally, often referred to as baby sign language, has grown in popularity in the media, on television, on the Internet, in daycare programs as well as in clinical settings. In recent years, a wide variety of programs, books, videos, and classes of baby signs have been made available to parents and caregivers (Doherty-Sneddon, 2008; Nelson, White, & Grewe, 2012). In addition to facilitating early communication, these products and websites claim numerous benefits such as greater understanding of the child’s needs, reduced frustration and stress, fewer tears and tantrums, better parent–child bonding, higher self-esteem, improved receptive and expressive language abilities, and enhanced cognitive development (Johnston, Durieux-Smith, & Bloom, 2005; Nelson et al., 2012; Paling, 2007).

Furthermore, in 2011, the American Academy of Pediatrics (AAP) recognized that baby sign language helps to improve communication (American Academy of Pediatrics, 2011). Despite this type of endorsement and the general support for baby signs in promoting language, there has been relatively little scientific evidence to support these claims. To date, three reviews have reported findings related to the benefits of teaching baby signs to young children with normal hearing. A systematic review in 2005 (Johnston et al., 2005) and a 2007 critical review (Paling, 2007) both reported inconsistent and equivocal results between studies. The Johnston et al. (2005) review, which included 17 reports published between 1980 and 2002, evaluated the effectiveness of signing in studies of children of parents with normal hearing (n = 5) and also included studies of children who were exposed to signing because their parents had hearing loss (n = 12). Paling’s (2007) critical review, which applied narrower inclusion criteria, reported on three studies published from 1980 to 2005, one of which was the Johnston et al. (2005) systematic review. These reviews both concluded that there was insufficient evidence to support the advancement in language development and that the included studies had serious methodological weaknesses. More recently, Nelson et al. (2012) conducted a review of 33 baby sign websites that promoted improved development through sign language, and argued that while these websites cite a wealth of reports, they offer limited information to support the benefits claimed. Of the cumulated 82 pieces of evidence cited as supporting research, only eight consisted of empirical research studies, while 90% of the citations were opinion articles.

Given the continued interest and support for baby sign, we undertook a systematic review to evaluate the current literature by identifying strengths, weaknesses, and gaps. This information will permit knowledge users such as early childhood educators, clinicians, and parents to make informed decisions about the use of and emphasis to place on baby sign language with typically developing children. The main objective of this project was to update the existing available information by systematically reviewing, evaluating, and synthesizing published literature on baby sign language intervention and its impact on developmental outcomes, particularly in language development as well as parent–child interaction and other related aspects of parent–child well-being. The specific research question addressed in this systematic review was: Does baby sign language help promote language, social, cognitive development, parent/caregiver–child interactions or other aspects of quality of life in typically developing hearing infants under the age of 36 months?

Methods

Search strategy

A search strategy was designed with the assistance of two library information specialists with expertise in systematic reviews. These specialists helped define the terminology and guided and refined the strategy for the eight included electronic databases. The search strategy was developed in MEDLINE and then adapted for the other databases: CINAHL, Cochrane Library (Wiley), ComDisDome, EMBASE, ERIC, LLBA, and PsycINFO. Search terms included: sign language, baby sign, symbolic gesture, gestural communication, children, infant, preschool, hearing, child development, and language development. The search was restricted to the pediatric age group and the start date was limited to 1990; databases were searched for articles published from January 1990 to February 2013. Electronic searching of databases was followed by hand searching from reference lists of studies that met the inclusion criteria, review articles, and websites in order to collect relevant articles in English or French. Website searching involved a Google Scholar search using the keywords: baby sign language, signs, symbolic gestures, gestural language, cognitive, social child development.

Eligibility criteria

Population

Studies of children with typical development and normal hearing who were exposed to baby sign language in early childhood (before age 36 months) were eligible for inclusion. The hearing status of parents was important in order to exclude research with children using gestures or signs as a first or second language. Thus, children with hearing loss, and children born to deaf parents, and children with atypical development were not considered eligible for inclusion, as this was not the focus of this review.

Intervention

For the purpose of this review, the definition of baby signs was viewed as symbolic gestures representing objects (i.e. arm flapping for bird), actions (i.e. thumb to mouth for ‘drink’) or states (i.e. drag index finger down cheek for ‘sad’) that were relevant to children and that were taught to infants through adult modeling. These are the types of signs used in the commercially available baby sign programs. Therefore, studies related to deictic gestures (e.g. pointing) were not considered appropriate for this review. Any intervention meeting the description of baby signs using terms such as baby sign language, signs, symbolic gestures, and gestural language as well as any articles using commercial programs of signing with children were considered eligible for inclusion.

Comparison

We originally sought to retrieve studies that included a comparison or control group of children with typical development and normal hearing who did not receive baby sign language or other symbolic gesture intervention. However, given the limited number of studies, we also included studies of children exposed to baby sign language without a comparison group.

Outcomes

The primary outcomes of interest were measures of the effectiveness of baby sign on language acquisition including receptive and expressive language. Receptive language refers to the child’s understanding of language including both words and gestures, while expressive language involves language output or production. Studies involving social and cognitive development as well as parent–child interaction, parent responsiveness, and any other relevant child behavior or parent-related outcomes reported were also included.

Study design

Study designs included randomized controlled trials, prospective cohort, retrospective cohort, cross-sectional studies, and case series of greater or equal to five participants. The review was restricted to primary studies, therefore reviews, letters, editorials, and commentaries were excluded.

Timing

Given that there were few studies prior to 1990 and that these early studies were reported in previous reviews (Johnston et al., 2005; Nelson et al., 2012; Paling, 2007), the time period for this review was limited to studies conducted after 1990.

Study selection

The search results were downloaded into a RefWorks database where duplicates were removed. Inclusion and exclusion criteria were then applied independently by two investigators (JT and VG) at two levels of screening, first title and abstract and then full-text. Calibration exercises to ensure consistent interpretation of the inclusion criteria were carried out at each stage. Two authors reviewed all titles and abstracts (if available). Next, using a priori inclusion criteria, a detailed relevance assessment of the complete document was carried out on the full-text of all publications judged to be potentially eligible in order to select the final reports for data abstraction. Disagreements and uncertainties were resolved through consensus or through consultation with a third team member, one of the study authors (EF).

Data abstraction

A data extraction sheet was developed for the study to capture general study characteristics and details of interventions. Details were abstracted related to country of origin, study design, study population, comparison group(s), intervention, and outcomes for each included study. Two authors (JT and EF) independently abstracted data for one of the included articles to clarify and discuss any discrepancies in interpretation. One author (JT) completed data extraction sheets for the remaining articles and a second author (VG) verified all entries. Any discrepancies were verified by a third author (EF). Due to the time limits of the review, authors were not contacted for further information or clarification.

Study quality

Methodological quality of each included study was evaluated using the McMaster University Quality Assessment Tool for Quantitative Studies (National Collaborating Centre for Methods and Tools, 2008), which resulted in a classification of each publication as ‘strong’, ‘moderate’ or ‘weak’. This assessment tool, which includes a scoring dictionary, was developed to evaluate studies according to eight components: selection bias, study design, confounders, blinding, data collection methods, withdrawals and dropouts, intervention integrity, and analyses. The five first study components were then rated as strong, moderate or weak according to the dictionary and a global rating for the study was determined. This procedure was performed by three independent reviewers (JT, EF, and CJ) for one article to discuss any differences in interpretation of questions. Discrepancies between reviewers with respect to the component ratings were discussed until consensus was reached. Two authors (JT and EF or CJ) independently assessed the quality of the remaining articles in the review. Any discrepancies were discussed between at least two investigators.

Data synthesis

Given the methodological heterogeneity in the included studies, variability in the data, as well as the range of measurement tools, data were not combined across studies in a meta-analysis. Studies were grouped according to the outcome domains reported. Data were described narratively when reported as such in the studies. For studies that included summary scores, means, and standard deviations (or other statistics), data were collected as reported by authors. Due to this heterogeneity, data were summarized qualitatively in this review. Study characteristics and description of outcomes were summarized in table format.

Results

Overview of included studies

Of the 1902 reports reviewed after removing duplicates, 1747 were excluded at the initial screening of titles and abstracts and 155 were retrieved for a more detailed relevance assessment. A total of 10 of these reports met the inclusion criteria. Assessment of published reports and reasons for exclusion are summarized in Figure 1.

Figure 1.

Study selection.

Ten reports contributed information about the benefits of baby sign language for this systematic review (Table 1). All reports were published in English. Of the 10 reports, seven related to studies conducted in the United States, two additional studies were conducted in the United Kingdom (UK), and one in Chile. These 10 included articles reported results from seven original research studies: two randomized controlled trials (RCT), two quasi-experimental studies, two cohort pre-post studies without control groups, and one retrospective cohort study. While seven articles were original research studies, two papers (Goodwyn & Acredolo, 1993; Moore, Acredolo, & Goodwyn, 2001) reported additional distinct data from the same study reported in detail by Goodwyn, Acredolo, and Brown (2000) and a third involved a subset of the same cohort assessed at age 8 years (Acredolo & Goodwyn, 2000). Therefore, data were abstracted from each of these reports. Six new reports meeting our inclusion criteria were published since the previous reviews of published studies in 2005 and 2007 (Johnston et al., 2005; Paling, 2007).

Table 1.

Characteristics of included studies according to outcome domains assessed.

Study	Design	Participants/Age/Intervention	Outcomes assessed/Measures	Results	Quality rating
Language
Goodwyn & Acredolo (1993) ^a	Cohort pre-post No control group data reported for this article	Exp n = 22* infants Age = 11 m (onset) Parents taught 8 symbolic gestures by video, encouraged to add others; biweekly telephone contact with investigators	• Sequenced Inventory of Communicative Development (SICD) at 15 m • Compared onset of first signs and first words	• No significant difference on SICD receptive or expressive for infants who acquired gestures first vs. infants with words first • Significant difference in mean age of 1st gesture vs. 1st word (11.9 m vs. 12.6 m; p = .0009); • No significant difference in mean age of 5th gesture vs. 5th word (13.6 m 14.3 m; p = .0778)	Weak
Goodwyn, Acredolo, & Brown (2000)	Quasi-exp longitudinal	Exp = 32 NoInt = 39 TxCtrl = 32* Age = 11 m (onset) 36 m (end) Parents taught 8 symbolic gestures by video, encouraged to add others; biweekly telephone contact with investigators *TxCtrl group did not perform better than NoInt group on any measure; analyses compared to Exp and NoInt groups	• SICD 15 m, 19 m • ROWPVT 24, 30, 36 m • Phonemic discrim. task 30 m • EOWPVT 24, 30, 36 m • MCDI (exp.) 15, 19, 42, 30 m • MLU/longest utterance 24 m	• Composite receptive language – significant difference for exp. group (.21 vs. –.10; p = .04) • 19 m SICD (receptive) – significantly better scores for exp. group (p = .01) • 24 m ROWPVT – significantly better scores for study group (p = .04) • 30 and 36 m: no significant difference between groups on receptive measures • Composite expressive score: significantly better for exp. group (.17 vs. –.17; p = .02); better score for study group at 15 m (.19 vs. –.26; p < .01) and 24 m (.33 vs. –.20, p < .0009) • 19, 20, 36 m: no significant difference between groups on individual expressive measures • 24 m MLU: significant difference for exp. group (2.26 vs. 1.94; p < .04) • Composite language score: significantly better for exp. group at 15 m (p = .004), 19 m (p = .04), and 24 m (p = .008), not at 30, 36 m	Moderate
Vallotton (2004)	Quasi-exp	Exp = 16 NoInt = 13 Age = mean 23 m (onset) (10–34 m); mean 29 m (end) (17–41 m) Children from Early Head Start program 7-month baby sign program at home and EHS center; baby sign instruction and supporting materials provided to parents and EHS providers; home visits to observe families	• MCDI • Gesturing Acquisition Interview (GAI) to measure participation in intervention, i.e. variety of gestures, number of routines, and frequency of gesture use	• MCDI: no significant difference between groups • Significant relationship between child’s sentence length and number of people using gestures as well as frequency of gesture use • Significant differences for exp. group: higher overall gesture use score (p = .002); greater variety of gesture use by child (p = .001), parent (p = .003), and provider (p = .0000); gestures used in more daily routines (p = .036) • No significant difference between groups: frequency of gestures (p = .121); frequency of vocal cues and behavior cues; rate of word and non-word vocalizations • Significant relationship between mother’s symbolic gesture use and child’s verbalizations (cue word rate [p = .019], vocal cue rate [p = .013], and vocalizing rates [p = .016])	Moderate
Vallotton (2008) ^b	Cohort pre-post No control group	Exp = 22 (total): 10 infants; 12 toddlers Age infants = 4.5–11.5 m (onset); 12.5–19.5 m (end) Toddlers = 17.3–24.8 m (onset); 20.8–28.3 m (end) Student caregivers used Baby Signs® program in university classroom	Qualitative observations/coding of representations of: emotions, feelings, and time concepts	• Use of emotion gestures by 6 children (earliest at 10.9 m) • Use of feeling gestures by 5 children (earliest at 14.7 m) • Use of time-related gestures by 11 children (earliest at 12.8 m)	Weak
Kirk, Howlett, Pine, & Fletcher (2013)	RCT longitudinal	Exp (gesture) = 10 Exp (BSL) =10 TxCtrl = 10 NoInt = 10 Age = 8 m (onset); 20 m (end) Teaching of targeted gestures (symbolic and BSL); training pack biweekly telephone contact and 4 home visits	• MCDI* 12, 16, 20 m • PLS-3* 12, 16, 20 m • GAPP 12, 16, 20 m • Video at 10, 12, 16, 20	• No significant effect of BSL or symbolic gesture: receptive vocabulary (p = .64); expressive vocabulary (p = .94); auditory comprehension (p = .94); expressive communication (p = .59), gestures, actions, pretend play (p = .83) • No effect of gesture learning on number of target words acquired (p = .81) but female gender was significant (p = .04) • Gesture training had a significant positive effect (p < .01) on expressive language only in boys who had low baseline scores	Strong
Cognitive
Acredolo & Goodwyn (2000) ^c	Quasi-exp	Exp = 19 (of 32 in original study) NoInt = 24 (of 39 in original study) Age = 11 m (onset); age at study: 8 yr Parents taught 8 symbolic gestures by video and encouraged to add others; biweekly telephone contact with investigators	• Cognitive development Wechsler Intelligence Scale for Children–III (WISC–III)	• Study group mean IQ = 114 vs. control group = 102 (p = .0012); sig. difference for both verbal (p = .0038) and performance sub-scales (.017)	Moderate
Parent–child interaction/responsiveness/behavior/stress
Moore, Acredolo, & Goodwyn (2001) ^a	Quasi-exp	Exp = 32 NoInt = 39 TxCtrl = 32 Age = 11 m (onset); 36 m (end) Parents taught 8 symbolic gestures by video, encouraged to add others; biweekly telephone contact with investigators	• Joint attention 19 and 24 m • Joint attention examined in relation to expressive language (EOWPVT 24 m)	• Exp. group initiated joint attention significantly more frequently and with more complex strategies (statistical significance not reported) • Infants with high infant-initiated joint attention demonstrated better expressive language at 24 m (statistical significance not reported)	Weak
Vallotton (2004)	Quasi-exp	Exp = 16 NoInt = 13 Age = mean 23 m (onset) (10–34 m); mean 29 m (end) (17–41 m) Children from Early Head Start program 7-month baby sign program at home and EHS center; baby sign instruction and supporting materials provided to parents and EHS providers; home visits to observe families	• Parent–child interaction video recorded • Child Behavior Checklist (CBCL); • Parenting Stress Index (PSI)	• No difference between groups in: mutual exchange (level of engagement, time engaged, and time in child-led play); children’s social cues (p = .059) or total cue frequency (p = .074); mother’s affect responsiveness; change in mother’s appropriate response to social cues • Significant relationship between: partial engagement (mother attending to/observing child) and overall gesture use (p = .028), (but not between child’s or mother’s gesture use, or number of routines where gesture was used); child’s social cue frequency and mother’s gesture use; mother’s affect responsiveness and overall gesture use (p < .05); fewer distress cues by child and overall gesture use, number of people using gestures, number of routines with gestures, frequency of gesture use • Higher percent of mother’s gesture use associated with higher percent of appropriate response to child’s distress cues (p = .039) • Less child distress was associated with more appropriate mother’s responses to distress cues (p = .007) but also with more inappropriate mother’s responses (p < .001) • No effect of group status on social skills (CBCL total behavior problems, internalizing and externalizing sub-scales) • No significant difference between symbolic gesture use (child use, mother use, overall use) and internalizing and externalizing behavior • No significant differences between groups in overall parent stress, or in parents’ perception of acceptability of child’s behavior • Parents in exp. group experienced less stress related to need for more positive reinforcement from their children (p = .023)	Moderate
Vallotton (2009)	Cohort pre-post No control group	18 student caregivers Exp = 10 infants (mean range 8.7 m onset – 13.3 m end) Caregivers used Baby Signs® program in university classroom Caregivers observed for 10 weeks interacting with infants (range 1–5)	• Video data: observations of student caregiver responsiveness to infants’ gestures during interactions	• No significant effect of frequency and variety of infants’ use of baby signs or conventional gestures on caregiver responsiveness • Greater responsiveness from caregivers when infants used gestures in response to caregivers’ gestures (p < .05) • As infants age, pointing (p < .05), signing frequency (p < .001), and sign variety (p < .01) increased caregiver responsiveness	Moderate
Góngora & Farkas (2009)	RCT	Exp = 7 NoInt = 7; 5–9 m (onset); 18–20 m (end) Spanish Baby Signs® program, 2 training sessions, biweekly home visits	• Early Interaction Analysis in Spanish (AIT Grid): vocal, visual, tactile, holding, and affective behaviors	• Exp. group showed significant difference: duration of visual interactions at 18–20 m (p < .05), frequency and duration of tactile behaviors (p < .05) • No significant difference between groups: frequency of visual interactions (p = .097); frequency or duration of vocal and affective interactions	Weak
Howlett, Kirk, & Pine (2011)	Retrospective cohort	Study (gesture) = 89 Control (no gesture) = 89 Age = mean 15.99 m (3–36 m) Control = mean 16.22 m Study group involved in variety of gesture classes –duration < 4 m to > 1 yr; control group in a variety of general activity classes	• Parent Stress Index (PSI)	• Study group mothers had higher mean total stress scores (p = .02) and significantly higher scores on sub-scales: Demandingness, Acceptability, Health, and Child Domain • No relationship between mothers’ stress level and how long (.17) and how often they used gestures (p = .34)	Moderate
Kirk et al. (2013)	RCT longitudinal	Exp (gesture) = 0 Exp (BSL) = 10 TxCtrl = 10 NoInt = 10 Age = 8 m (onset); 20 m (end) Teaching of targeted gestures (symbolic and BSL); training pack; biweekly telephone contact and 4 home visits	• Video-recorded interactions at 10, 12, 16, 20 m to examine maternal mind-mindedness (mother’s responses to child)	• No significant difference between groups in overall maternal mind-mindedness, mothers’ imitation of infant vocalizations, appropriate mind-related comments, or inappropriate mind-related comments (p > .05)	Strong

Key: RCT: randomized controlled trial; Quasi-exp: quasi experimental study; Exp: experimental group; TxCtrl: treatment control group; NoInt: control group-no intervention; m: months; yr: years.

Key – assessments: SICD: Sequenced Inventory of Communication Development; ROWPVT: Receptive One-Word Picture Vocabulary Test; EOWPVT: Expressive One-Word Picture Vocabulary Test; MCDI: MacArthur Communicative Development Inventories; PLS–3: Preschool Language Scale–3; GAPP: Gestures, Actions, and Pretend Play Checklist; MLU: Mean Length of Utterance; WISC–III: Wechsler Intelligence Scale for Children–III; CBCL: Child Behavior Checklist; PSI: Parent Stress Inventory.

Study group drawn from longitudinal study reported in Goodwyn et al. (2000) and pilot study.

Study details regarding baby sign provided in Vallotton (2009).

Cohort is drawn from a longitudinal study reported in Goodwyn et al. (2000). Reported as 24 of 37 in study.

British version of tests.

Study characteristics

This review aimed to examine the effects of baby signs on several outcome domains: language, cognitive and social development, parent–child interaction, and aspects related to child or parents’ quality of life. The included studies examined a range of outcomes, which we have broadly summarized under three domains: language, cognition, and parent–child interaction/quality of life outcomes. Table 1 provides details of the population, intervention characteristics, and outcomes according to these domains as well as the quality ratings determined based on the McMaster tool. Given the wide range of outcomes and data collection methods reported, a meta-analysis was not possible.

The 10 reports retained for this review included five that focused on some aspect of language development (Table 1); two of these included both measures of language and interaction/responsiveness. Only one study (Acredolo & Goodwyn, 2000) addressed specific cognitive functioning and six of the reports included information on some aspect of parent interaction/responsiveness and/or quality of life.

As shown in Table 1, interventions of different durations and in different settings (e.g. home and/or childcare program) were implemented in the included studies. The interventions included the presentation of a specific infant signing intervention program that was comprised of a defined set of target gestures or signs taught to parents (n = 7 studies). In two cohort reports with no control group, student caregivers were taught to model symbolic gestures (Vallotton, 2008, 2009). One retrospective study involved parents of children who were participating in a variety of gesture classes including both symbolic gestures and British Sign Language (BSL) throughout the UK (Howlett, Kirk, & Pine, 2011).

Language related outcomes

Between 1993 and 2013, five reports, including one RCT (Kirk, Howlett, Pine, & Fletcher, 2013), two quasi-experimental (Goodwyn et al., 2000; Vallotton, 2004), and two cohort studies (Goodwyn & Acredolo, 1993; Vallotton, 2008), measured one or more aspects of receptive or expressive language. All were conducted in the United States except for the one RCT from the UK (Kirk et al., 2013). As shown in Table 1, the RCT study was rated as strong, the two quasi-experimental reports as moderate, and the two cohort studies (without comparison groups) as weak. Four articles (Goodwyn & Acredolo, 1993; Goodwyn et al., 2000; Kirk et al., 2013; Vallotton, 2004) reported on at least one standardized language assessment measure and one involved qualitative observations of children’s symbolic gesture use (Vallotton, 2008). These five reports involved 90 different participants who received gesture interventions with study group samples sizes ranging from 16 (Vallotton, 2004) to 32 infants (Goodwyn et al., 2000). Two studies reported results of different language outcome measures for some portion of the same group of study participants (n = 16 of 22 in the study) (Goodwyn & Acredolo, 1993) and (n = 32) (Goodwyn et al., 2000). Two reports (Goodwyn & Acredolo, 1993; Vallotton, 2008) did not include a control group. In addition to language outcomes, two of the reports (Kirk et al., 2013; Vallotton, 2004) also provided results related to parent–child interaction or related aspects and are presented in a subsequent section of this report and in Table 1.

The body of work undertaken by Goodwyn and colleagues in California made important early contributions to the questions related to the advantages of exposure to baby sign language for developing language. Goodwyn and Acredolo (1993) reported on a subset of 22 children, 16 of these were from the same longitudinal study described in detail in a later report by Goodwyn et al. (2000). The authors found that infants exposed to signs acquired their first gesture earlier than their first word but found that the difference between the mean age for the fifth gesture and fifth word was not statistically significant (13.6 vs. 14.3 months). There was also no significant difference at 15 months of age in receptive and expressive language (Sequenced Inventory of Communication Development – SICD) for infants who acquired gestures first compared to those who acquired words first.

The most comprehensive report came from Goodwyn et al. (2000) who compared results on several standardized and study-specific language measures between 32 children who were exposed to baby signs from their parents and 39 children who received no intervention. Another intervention group described as a treatment control group of 32 children receiving verbal intervention was also included in the study. The authors reported that there were no differences on any measure between the treatment control group and the no intervention control group and that all analyses were therefore conducted to compare outcomes between the experimental group and the no treatment control group. Composite receptive, expressive, and total language scores led the authors to conclude that there were significant differences between the two groups. As described by the authors, these composite scores included a compilation of the average of z-scores across all study ages for receptive language measures (composite receptive), expressive language measures (composite expressive), and all language measures (total language). However, no significant differences between experimental group and control group children were reported after age 24 months on any individual measure (details in Table 1).

In contrast to Goodwyn and colleagues’ findings, Kirk et al. (2013) examined outcomes on several standardized language measures (British versions) at ages 12, 16, and 20 months and found no significant difference between the experimental group of children (symbolic gestures, n = 10, BSL, n = 10) on receptive and expressive language measures and control groups of peers who received verbal training (verbal training control group [n = 10]) or no intervention (n = 10). In summary, as detailed in Table 1, the Goodwyn et al. (2000) report is the only one in this review that reported better language outcomes on standardized language assessments for children who received baby sign intervention compared to no sign intervention.

Two other studies examining language were conducted by Vallotton (2004, 2008). Vallotton (2004) used a Gesturing Acquisition Interview (GAI) to examine gesture use in 29 children in a Head Start program in the US (16 in the experimental group and 13 in the control group). Significant differences were found between groups for gesture-specific aspects including overall gesture use, variety of gesture use, and overall number of routines in which children used gesture. However, there was no significant difference between the two groups for frequency of symbolic gesture use, frequency of vocal and behavior cues, or rate of word and non-word vocalizations. In a cohort study of 22 children in a university research childcare program where infants were exposed to baby sign from student caregivers, Vallotton (2008) examined infants’ use of gestures to express emotions. Based on emotion gestures (n = 6 children), feeling gestures (n = 5), and time-related gestures (n = 11), the author suggested that infants using symbolic gestures could express these concepts. As there was no control group, it is unknown whether children without baby sign intervention would also use gestures to express these concepts.

Cognitive development

Only one report in this review (Table 1) specifically addressed cognitive development using a standardized measure. Following up on the cohorts from their 2000 study, Acredolo and Goodwyn (2000) examined cognitive development at 8 years of age in 19 of the original 32 experimental group children and 24 of 39 (noted as 37 in the report) original control group children. The authors reported a significant advantage for the intervention group on the verbal sub-scale, performance sub-scale, and IQ of the Wechsler Intelligence Scale for Children–III. They concluded that baby sign use in the early years was associated with higher intellectual functioning in the early school years.

Parent–child interaction and quality of life

Six of the 10 studies, conducted from 2001 to 2013, specifically considered parent–child interaction and/or parent responsiveness/stress issues. Moore et al. (2001) examined joint attention in the cohort studied by Goodwyn et al. (2000). They reported that the experimental group of 32 children initiated joint attention more frequently and with greater complexity than the control group and suggested that symbolic gesture and joint attention both contributed to better expressive language at 24 months. Vallotton (2009), in a qualitative study of parent–child interaction of 10 children, documented better caregiver responsiveness for 10 infants after age 10 months. Góngora and Farkas (2009) showed a significant increase in the study group’s use of visual and tactile interventions. They concluded that parent–infant interactions using these types of stimuli are improved when baby sign is used. Although this was an RCT based on random allocation, our overall quality assessment resulted in a weak rating due to limited reporting of information about the method of randomization, validity of assessment methods, and withdrawals as well as the small number of study participants. Taken together, these three studies suggest that using baby sign language does not interfere with parent–child interaction and may, in fact, contribute to positive interactions.

Vallotton (2004) examined both parent perceptions of behavior and stress in a study of 29 parent–child dyads. Using a standardized tool to examine behavior, the author reported no significant difference on social skills either on total behavior or on the internalizing and externalizing behavior sub-scales of the Child Behavior Checklist (CBCL). In addition, no difference was noted between groups on the Parent Stress Index (PSI) on overall stress or on parents’ perception of the acceptability of their child’s behavior. However, the group of 16 parents in the experimental group did experience more stress related to the need for more positive reinforcement from their child. In contrast, using the PSI, Howlett et al. (2011) examined stress levels through a retrospective cohort study in the UK. Participants were 89 mothers who attended various baby sign or BSL programs and 89 comparison group mothers whose children attended other types of learning and play activities. They found that mothers in the baby sign exposure group were significantly more stressed than those in the comparison group. However, as a retrospective study, no information was collected regarding mother stress levels prior to enrollment in gesture programs. Finally, in a subset of 18 of 40 children in their study (nine in study group, nine in control group), Kirk et al. (2013) investigated maternal mind-mindedness, which refers to parent responsiveness to their child. The authors found no significant difference between the two groups of children. Overall, these studies present a mixed portrait of parent-related aspects in relation to the use of baby sign language.

Summary of outcomes reported

In summary, Table 2 provides an overview of all assessments/measures reported in the seven reports that compared a study group and a control/comparison group in the areas of verbal communication development or some aspect of parent–child interaction or related outcomes. For each outcome measured, the table displays whether there was a positive effect, no effect, or a negative effect of symbolic gesture use. As shown, the 10 reports included a total of eight different standardized measures, five of them specifically assessing some aspect of early language development. As seen in Table 2, only two of these individual language measures showed significantly better results at certain ages (19 and 24 months) for the group of children receiving sign intervention, both from the Goodwyn et al. (2000) study. Of the total measures reported, including the composite language scores, 10 showed a positive effect at any age for the baby sign intervention groups, eight showed no effect at any age reported by any study, and one measure, PSI overall stress, showed a negative effect in one study and a positive effect in another. Of the 10 that reported a positive effect, four were from Goodwyn et al. (2000) and showed an advantage of baby sign in communication development only up to 24 months of age. The only standardized language measure used in more than one study was the MCDI (Goodwyn et al., 2000; Kirk et al., 2013; Vallotton, 2004), a parent questionnaire that documents parents’ observations of the child’s early words and gestures. None of the three studies found a difference on this measure between children exposed and not exposed to sign programs.

Table 2.

Results for outcomes for studies with control groups.

Measure/tool	Goodwyn et al. (2000)	Acredolo & Goodwyn (2000)	Moore et al. (2001)	Vallotton (2004)	Góngora & Farkas (2009)	Howlett et al. (2011)	Kirk et al. (2013)	Comment
Composite lang-Rec^a	✓
Composite lang-Exp^a	✓							No effect overall, positive 15, 24 m only
Overall language^b								Positive 15, 24 m only
SICD-Rec	✓							Positive 19 m only
SICD-Exp	✓
ROWPVT	✓							Positive 24 m only
EOWPVT	✓							Positive 24, 30, 36 m
MCDI	✓			✓			✓*	All studies = no effect
PLS-3							✓*
GAPP							✓
Phon. discrim. task	✓
MLU	✓
Longest utterance	✓
WISC–III		✓
CBCL				✓
PSI overall stress				✓		✓		Howlett = negative effect
Joint attention			✓
Caregiver responses				✓	✓
Parent mind-mindedness							✓

Table includes results for language outcomes, parent–child interaction/caregiver responsiveness, and parent stress measures for studies with control groups.

All receptive or expressive measures across 5 study ages = average of z-scores.

All receptive and expressive measures across 5 study ages = average of z-scores.

British version of tests.

Key: SICD: Sequenced Inventory of Communication Development; ROWPVT: Receptive One-Word Picture Vocabulary Test; EOWPVT: Expressive One-Word Picture Vocabulary Test; MCDI: MacArthur Communicative Development Inventories; PLS-3: Preschool Language Scale–3; GAPP: Gestures, Actions, and Pretend Play Checklist; MLU: Mean Length of Utterance; WISC–III: Wechsler Intelligence Scale for Children–III; CBCL: Child Behavior Checklist; PSI: Parent Stress Inventory.

Discussion

This review aimed to update information to help guide early childhood educators, clinicians, and parents who have an interest in knowing whether baby sign language improves communication in young children. The results are important because they inform these individuals about the current state of evidence in this field and about the gaps in knowledge on this topic. This review indicates that despite the seemingly widespread interest in baby sign language, the pace of scientific contributions to understanding this intervention is relatively poor. While there is considerable literature describing the benefits of baby sign language, only a limited number of primary studies have actually investigated the effectiveness of using signs in early childhood. Only six studies that met our inclusion criteria have been published over the last 10 years since the Johnston et al. (2005) review.

The 10 included reports provided a wide range of outcome measures and presented findings in various formats, making it difficult to statistically compile study results. The quality of the majority of the studies was assessed as weak to moderate with one RCT rated as strong. Based on the included studies, our review has shown that there is no convincing evidence that exposure to symbolic gesture intervention is associated with benefits in language acquisition for typically developing children.

Communication and cognitive development

One of the primary cited benefits of baby sign language is to enhance spoken language development (Acredolo & Goodwyn, 1988; Goodwyn et al., 2000). The paucity of new literature examining the effectiveness of baby sign language on communication development, despite the numerous claims of benefits, is troubling. In the area of language acquisition, only one study published more than 13 years ago documented advantages on some standardized language assessments up to 24 months of age for a group of 32 children who received baby sign intervention (Goodwyn et al., 2000). In contrast, one recent RCT that used standardized language measures (Kirk et al., 2013) concluded that there was no advantage of early exposure to either symbolic gestures or BSL for typically developing children. Any improvements documented specifically in language by Goodwyn et al. (2000) were measurable only until age 24 months with no differences in the third year of life, suggesting that if there are benefits in language development, there is a relatively short time window after which any effects seem to be negligible. We therefore conclude that there is no robust evidence to support the assumptions that using baby sign language enhances early communication development in typically developing children. Further research is required to determine whether baby sign language has benefits and particularly any lasting benefits in the area of spoken language and cognitive development.

Parent–child/caregiver outcomes

Similar to the findings for communication development, evidence for the effectiveness of baby sign language in fostering parent–child interaction and greater parent or caregiver responsiveness is inconclusive because of the small number of published studies and their mixed findings. Vallotton (2004) found no difference between gesturing and non-gesturing mother–child dyads in mutual exchange (e.g. level of engagement) or in mothers’ affect responsiveness, however, mothers’ gesture use was associated with a higher number of appropriate responses to child’s distress cues. While both pointing and symbolic gestures produced by infants increased responsiveness of student caregivers in a childcare center, the frequency and variety of gestures had no effect on responsiveness (Vallotton, 2009). In the one study that examined mother responsiveness, Góngora and Farkas (2009) noted that mother–infant dyads had longer visual engagement and more tactile interaction behaviors but no significant difference in the frequency and duration of vocal and affective interactions. In the RCT that investigated maternal mind-mindedness (mothers’ verbal and non-verbal responsiveness to the child) (Kirk et al., 2013), there were no significant differences related to overall maternal mind-mindedness in mother–infant dyads who had or had not been taught gestures. However, only nine mother–infant dyads in each group were coded for these behaviors. Given the inconsistent findings within and across studies, it is difficult to draw any specific conclusions about the overall effects of symbolic gesture on parent or caregiver–infant interactions.

Likewise, this review found that studies provided little conclusive evidence on whether using baby sign language offers advantages for the child or parents with respect to quality of life related issues. The two studies addressing this issue produced strikingly different results ranging from no effect to negative effects. While one relatively large study of 89 parents showed that mothers enrolled in childcare sign programs experienced higher stress levels than those who were not involved, an important weakness, as pointed out by the authors, was that baseline information on stress levels was not available and may have accounted for the differences between groups (Howlett et al., 2011). The other study (Vallotton, 2004) revealed no difference between groups on parent stress levels except on parent’s perception of stress related to the need for more reinforcement from their child. The same study also did not find any differences in parents’ perceptions of their child’s social skills as measured by a standardized assessment tool. Further evaluation of the effects of using baby sign programs on parents’ stress levels and their perception of their child’s behavior is required before stronger conclusions can be reached.

The mixed findings reported in the literature highlight the need to further examine the benefits of baby sign language before clear recommendations can be made to parents or providers. Caution is also recommended when comparing any positive results from the studies in this review to other typical populations as parent–child dyads in the included studies were primarily enrolled in baby sign programs with regular follow-up from investigators. Such follow-up and coaching may not reflect typical baby sign learning conditions for most infants. Based on this review, we are unable to draw any conclusions regarding differences in effectiveness that may be related to instruction mode such as type of instructional program. As noted in other reviews (Johnston et al., 2005; Nelson et al., 2012), parents in the published studies received instruction in baby sign language from researchers whereas many available products for parents require self-instruction. No studies were retrieved that evaluated outcomes for parents using these self-taught methods. Since the only beneficial information at this time comes from children involved in experiments where parents received specific guidance in baby sign, it is important to assess the benefits of baby sign in real-life settings that are more representative of the types of programs parents are likely to encounter. Further controlled studies conducted in multiple programs are required to generate more robust information to help inform decision-making.

Limitations

It is important to underscore that this review was limited to studies involving typically developing children and did not examine the effects of baby sign language on children with developmental concerns. Therefore, no conclusions can be drawn about the benefits of sign language for children with special needs. Another limitation is that authors were not contacted to verify study details, therefore the study quality rating and data abstraction are based entirely on the information provided in the published report.

Conclusion

The notion of advantages for children who are exposed to baby sign language has been proposed for more than 20 years. It is common for parents to ask pediatricians and other providers whether baby sign can enhance their child’s development. The literature provides only limited evidence to help with this question. Our updated systematic review showed that the topic of the effectiveness of baby sign language has garnered relatively little interest in terms of scientific inquiry despite the many remaining gaps in knowledge. Although there is no strong evidence to support the claimed benefits of baby sign language, no studies reported any adverse effects on typical language development. Until research demonstrates otherwise, it seems reasonable to advise parents and care providers that baby sign language does not seem to benefit long-term development but that it also does not interfere with typical development or with parent–child interaction. Therefore, based on this review, there is no reason to advocate for policies to include baby sign in early childhood education programs or to encourage parents to enroll their children in baby sign programs. However, there is also no basis for discouraging parents who may wish to use symbolic gestures in early communication with their children.

Footnotes

Acknowledgements

We gratefully acknowledge the expertise of two librarian scientists who provided ongoing guidance in developing the search strategy and literature search. We also thank Marie-France Perrier for assistance with screening during the first stage of the study selection.

Funding

This work was partially supported through a Canadian Institutes of Research New Investigator Award and a Canadian Child Health Clinician Scientist Program Award held by E. Fitzpatrick.

Note on authors’ contributions

JT and EF developed the initial study protocol. JT carried out the initial study selection and wrote a preliminary report. EF oversaw the review, was involved in all stages of the research, and prepared the first draft of the manuscript for publication. VG was involved in reviewing manuscripts and verifying data extracted. CJ provided guidance on the research protocol, conducted quality assessment reviews, and served as third reviewer for quality assessments as required. All authors read and provided input into the final manuscript.

References

Acredolo

L. P.

Goodwyn

S. W.

(1988). Symbolic gesturing in normal infants. Child Development, 49, 450–466.

Acredolo

L. P.

Goodwyn

S. W.

(2000, July). The long-term impact of symbolic gesturing during infancy on IQ at age 8. Paper presented at the International Society on Infant Studies, Brighton, UK.

Acredolo

L. P.

Goodwyn

S. W

Horobin

Emmons

(1999). The signs and sounds of early language development. In Tamis-LeMonda

L. B. C.

(Ed.), Child psychology: A handbook of contemporary issues (pp. 116–139). New York, NY: Psychology Press.

American Academy of Pediatrics. (2011). These hands were made for talking. In American Academy of Pediatrics (Ed.), Heading home with your newborn (2nd ed.). Available from http://www.healthychildren.org

Bates

Dick

(2002). Language, gesture, and the developing brain. Developmental Psychobiology, 40, 293–310. doi:10.1002/dev.10034

Capirci

Iverson

J. M.

Pizzuto

Volterra

(1996). Gestures and words during the transition to two-word speech. Journal of Child Language, 23, 645–673.

Doherty-Sneddon

(2008). The great baby signing debate. The Psychologist, 21, 300–303.

Góngora

Farkas

(2009). Infant sign language program effects on synchronic mother-infant interactions. Infant Behavior & Development, 32, 216–225. doi:10.1016/j.infbeh.2008.12.011

Goodwyn

S. W.

Acredolo

L. P.

(1993). Symbolic gesture versus word: Is there a modality advantage for onset of symbol use? Child Development, 64, 688–701.

10.

Goodwyn

S. W.

Acredolo

L. P.

(1998). Encouraging symbolic gestures: A new perspective on the relationship between gesture and speech. New Directions for Child Development, 79, 61–73.

11.

Goodwyn

S. W.

Acredolo

L. P.

Brown

C. A.

(2000). Impact of symbolic gesturing on early language development. Journal of Nonverbal Behavior, 24, 81–103.

12.

Howlett

Kirk

Pine

K. J.

(2011). Does ‘wanting the best’ create more stress? The link between baby sign classes and maternal anxiety. Infant and Child Development, 20, 437–445.

13.

Iverson

J. M.

Capirci

Volterra

Goldin-Meadow

(2008). Learning to talk in a gesture-rich world: Early communication in Italian vs. American children. First Language, 28, 164–181. doi:10.1177/0142723707087736

14.

Iverson

J. M.

Goldin-Meadow

(2005). Gesture paves the way for language development. Psychological Science, 16, 367–371. doi:10.1111/j.0956-7976.2005.01542.x

15.

Johnston

J. C.

Durieux-Smith

Bloom

(2005). Teaching gestural signs to infants to advance child development: A review of the evidence. First Language, 25, 235–251.

16.

Jusczyk

(1997). The discovery of spoken language. Cambridge, MA: The MIT Press.

17.

Kirk

Howlett

Pine

K. J.

Fletcher

(2013). To sign or not to sign? The impact of encouraging infants to gesture on infant language and maternal mind-mindedness. Child Development, 84, 574–590.

18.

Locke

J. L.

(2007). Bimodal signaling in infancy: Motor behavior, reference, and the evolution of spoken language interaction studies. Interaction Studies. Social Behaviour and Communication in Biological and Artificial Systems, 8, 159–175.

19.

Miller

J. L.

Lossia

A. K.

(2013). Prelinguistic infants’ communicative system: Role of caregiver social feedback. First Language, 33, 524–544. doi:10.1177/0142723713503147

20.

Moore

Acredolo

L. P.

Goodwyn

S. W.

(2001, April). Symbolic gesturing and joint attention: Partners in facilitating verbal development. Paper presented at the Biennial Meetings of the Society for Research in Child Development, Minneapolis, MN.

21.

National Collaborating Centre for Methods and Tools. (2008). Quality assessment tool for quantitative studies (Updated 13 April 2010). Retrieved from http://www.nccmt.ca/registry/view/eng/14.html

22.

Nelson

L. H.

White

K. R.

Grewe

(2012). Evidence for website claims about the benefits of teaching sign language to Infants and toddlers with normal hearing. Infant and Child Development, 21, 474–502.

23.

Olson

Masur

E. F.

(2013). Mothers respond differently to infants’ gestural versus nongestural communicative bids. First Language, 33, 372–387. doi:10.1177/0142723713493346

24.

Paling

(2007). Critical review: Does baby sign language have a positive affect on the language development of typically developing, hearing infants of hearing parents? School of Human Communication Sciences and Disorders, University of Western Ontario, London, ON.

25.

Rowe

M. L.

Goldin-Meadow

(2009). Early gesture selectively predicts later language learning. Developmental Science, 12, 182–187. doi:10.1111/j.1467-7687.2008.00764.x

26.

Tomasello

Farrar

M. J.

(1986). Joint attention and early language. Child Development, 57, 1454–1463.

27.

Vallotton

C. D.

(2004). Effects of symbolic gestures as a caregiving tool: Children’s social and language development and mothers’ perceptions and behavior. Dissertation Abstracts International Section B: The Sciences and Engineering, 65(9-B), 4877.

28.

Vallotton

C. D.

(2008). Signs of emotion: What can preverbal children ‘say’ about internal states? Infant Mental Health Journal, 29, 234–258.

29.

Vallotton

C. D.

(2009). Do infants Influence their quality of care? Infants’ communicative gestures predict caregivers’ responsiveness. Infant Behavior & Development, 32, 351–365. doi:10.1016/j.infbeh