Minimally-Signing Autism in Deaf Children of Deaf Parents: A Case Series

Abstract

Minimally-verbal autism is well described in hearing populations, but little is known about minimally-signing autism in deaf children with early, full access to sign. This case series presents seven deaf autistic children born to Deaf parents and exposed to American Sign Language (ASL) from birth who nonetheless remain minimally expressive signers. Participants were drawn from a nationwide cohort of 23 native-ASL deaf children with autism. Four children completed the Autism Diagnostic Observation Schedule–2 (ADOS-2); all caregivers completed the Social Communication Questionnaire (SCQ). Selected children also received the Test of Nonverbal Intelligence–4 (TONI-4) and the ASL Receptive Skills Test (ASL RST). Six of seven children scored above the SCQ cutoff, and all four ADOS-2 cases met diagnostic criteria. Across cases, hallmark autistic features were evident, including limited reciprocal interaction, reduced joint attention, and restricted/repetitive behaviors. Expressive signing ranged from absent to small repertoires of echolalic or idiosyncratic signs, rarely coordinated with gaze or affect; symbolic play was similarly constrained. Two children completed standardized testing: one showed average nonverbal cognition but ASL comprehension <3 years; the other showed below-average nonverbal cognition and minimal ASL comprehension. These findings demonstrate that minimally expressive autism occurs in Deaf children with full access to a natural signed language, ruling out language deprivation or auditory processing as necessary explanations. Instead, domain-general constraints (limited generativity, social–pragmatic integration, and sensorimotor planning) likely contribute across modalities. Documenting minimally signing autism underscores the need for modality-sensitive diagnostic tools, neurodiversity-affirming supports, and longitudinal research to better understand and serve this underserved group.

Autism spectrum disorder (ASD) is a neurodevelopmental condition defined by two main criteria: challenges in social communication and relatedness, and the presence of repetitive behaviors or restricted interests (APA, 2013). Current prevalence estimates suggest that one in 100 children globally meet criteria for ASD (Zeidan et al., 2022). Approximately 30% of autistic children are described as “minimally verbal,” meaning their expressive language is extremely limited (Tager-Flusberg & Kasari, 2013). Although definitions vary across studies—for example, by vocabulary size, standardized test cutoffs, or parent report—the term generally refers to children whose expressive speech is restricted to single words or fixed phrases (Bal et al., 2016).

The minimally verbal group is highly heterogeneous (Bal et al., 2016; Koegel et al., 2020). Levels of expressive language, receptive language, and nonverbal cognition vary widely, and some autistic individuals and their families prefer the term minimally speaking to highlight receptive and cognitive strengths despite limited speech (Baggs, 2012; Kedar, 2018). A group that is rarely considered in this framework is deaf ¹ autistic children, who do not speak but whose native language is a signed language. We refer to those who remain severely limited in expressive signing as minimally signing. By characterizing their communication and cognitive profiles, we aim to clarify factors that constrain expressive language development in autism across modalities.

Multiple factors have been linked to expressive language outcomes in autism; recent reviews have identified dozens (e.g., Muès et al., 2024). Here, we focus on five domains particularly relevant to minimally expressive language: social cognition, nonverbal cognition, receptive language, motor skills, and perceptual processing.

Domains Relevant to Minimally Expressive Autism

Social Cognition

Joint attention, especially response to joint attention, is a well-established predictor of language growth in autism (Bottema-Beutel, 2016; Mundy & Jarrold, 2010; Yoder et al., 2015). Deaf children, however, must coordinate attention differently than hearing children: they cannot rely on vocal bids and must visually shift from an object to a signer to receive language input. Research shows that Deaf infants engage in joint attention earlier and more frequently in visually scaffolded environments (Lieberman et al., 2014), potentially providing a strong foundation for language learning.

Nonverbal Cognition

Higher nonverbal intelligence is consistently associated with more favorable expressive language outcomes. For example, Wodka et al. (2013) found that nonverbal cognitive ability predicted attainment of phrase and fluent speech in school-aged autistic children with severe language delay, while Faerman et al. (2023) reported robust links between nonverbal reasoning and expressive performance in autistic youth. Systematic reviews confirm that stronger nonverbal skills increase the likelihood of phrase speech, though cognitive ability alone does not account for why some children remain minimally expressive (Byrne et al., 2024; Muès et al., 2024).

Receptive Language

Receptive language ability is often the strongest predictor of later expressive development. Longitudinal work shows that toddlers with higher receptive scores are more likely to achieve phrase speech (Luyster et al., 2007; Pickles et al., 2014). Reviews likewise identify receptive language, joint attention, and imitation as critical predictors of expressive growth in minimally verbal children (Muès et al., 2024; Tager-Flusberg & Kasari, 2013). Even when expressive output is limited, higher receptive skills often support alternative communication, such as gesture or single signs (Saul & Norbury, 2020).

Motor Skills

Motor development contributes to language via shared neural substrates and developmental cascades (Iverson, 2010). Gross motor milestones such as onset of walking predict language growth (Leonard et al., 2015), fine motor skills correlate with expressive outcomes (Bal et al., 2020; Choi et al., 2018), and oromotor difficulties, including childhood apraxia of speech, can severely constrain spoken output (Chenausky et al., 2019; Gernsbacher et al., 2008).

Perceptual Processing

Differences in perceptual processing also affect language. ² In hearing children, atypical auditory processing (especially speech perception, prosody processing, and neural encoding of speech) has been linked to speech delays (Alcántara et al., 2004; Bhatara et al., 2013; Khalfa et al., 2001; O’Connor, 2012; Russo et al., 2009; Schwartz et al., 2020). For Deaf children, visual processing is the relevant modality. Cortical visual impairment (CVI) is reported at elevated rates in developmental disabilities (Chokron et al., 2021), and difficulty interpreting dynamic visual input could interfere with processing of manual and facial cues in signed languages.

Manual Signs in Minimally Verbal Autism

For decades, manual signs have been used as augmentative or alternative communication strategies for minimally speaking hearing children with autism (Bonvillian et al., 1981; Carr, 1979). Bonvillian et al. (1981) reviewed over 20 studies in which signs were taught to more than 100 autistic children, finding that most children were able to learn a small number of signs, but did not acquire grammatical structures as a result of intervention. Similarly, more recent programs such as Key Word Sign combine core signs with speech, but these approaches prioritize functional communication and rarely yield fluent or syntactically structured signed language (Tan et al., 2014). In most cases, signs have been viewed as a possible bridge to speech, though the efficacy of this approach appears to be limited: for example, less than one-quarter of the children covered in Carr's (1979) review of seven studies were reported to develop any speech following sign intervention.

Minimally-Signing Deaf Autistic Children

Very little is known about Deaf autistic children who remain minimally expressive despite full early access to sign. Most reports involve children of hearing parents with limited language exposure (Jure et al., 1991; Meinzen-Derr et al., 2014; Roper et al., 2003). Only two published cases describe native-signing Deaf autistic individuals: one Greek child who communicated via Picture Exchange Communication System (PECS) (Malandraki & Okalidou, 2007) and one American adult who produced echolalic signs with limited generativity (Poizner et al., 1990). These cases raise the possibility that domain-general factors, rather than language deprivation, may underlie persistent expressive limitations.

The Current Study

This study presents the first multi-case series description of Deaf autistic children with native ASL exposure who remain minimally-expressive signers. By examining their linguistic, cognitive, and social-communicative profiles, we aim to clarify the mechanisms that constrain expressive language across modalities.

Methods

Participants

This study draws on data from a nationwide project in the United States investigating language, cognition, and social development in Deaf autistic children born to Deaf parents. All participants were Deaf children with full early access to American Sign Language (ASL) as a native language (n = 23; age range: 4;5–14;4, M_age = 9;5). Written informed consent was provided by a parent prior to data collection, and the Institutional Review Board of [Boston University] approved the study protocol prior to data collection. For the present case series, we selected seven children who showed minimal expressive signing at assessment. These children ranged in age from 4 to 10 years and were recruited through schools for the Deaf, parent networks, and community organizations across the United States. All participants had confirmed diagnoses of ASD and scored above the clinical cutoff on a standardized ASD screener or assessment. None had a history of language deprivation. This makes them the first systematically described cohort of Deaf autistic children with native ASL exposure who nonetheless remain minimally expressive signers. Table 1 summarizes key characteristics and assessment scores for these participants.

Table 1.

Participants With Minimal Expressive Sign Language.

ID	Age	Sex	Hearing status	Parental hearing status	SCQ	ADOS-2 module	ADOS-2 score	TONI-4 standard score	ASL RST raw score	ASL RST language equivalent
Ethan	4;0	M	Deaf	Deaf	27	1	21	-	-	-
Caleb	4;11	M	Deaf	Deaf	23	-	-	-	-	-
Noah	5;1	M	Deaf	Deaf	26	1	18	-	-	-
Owen	6;9	M	Deaf	Deaf	19	-	-	-	-	-
Maya	6;9	F	Deaf	Deaf	8	1	18	98	1	<3;0
Liam	9;5	M	Deaf	Deaf	31	1	18	80	2	<3;0
Sophia	10;5	F	Deaf	Deaf	17	-	-	-	-	-

Note. Blank cells indicate that the measure was not administered or the child did not respond to the task. SCQ = Social Communication Questionnaire; ADOS-2 = Autism Diagnostic Observation Schedule, Second Edition; TONI-4 = Test of Nonverbal Intelligence, Fourth Edition; ASL RST = ASL Receptive Skills Test.

Scores ≥ 11 for Module 1 on the ADOS-2 result in a classification of autism spectrum.

Measures

Autism Diagnosis

Autism diagnoses were confirmed through either direct assessment, parent-report screeners, or both. The Autism Diagnostic Observation Schedule, Second Edition (ADOS-2; Lord et al., 2012) was administered by a research-reliable clinician fluent in ASL, using Module 1 for all observed children (n = 4). Because the ADOS-2 was not designed for use with deaf children, some adaptations were made by the assessor, a research-reliable clinician-researcher fluent in ASL (e.g., the “Response to Name” item was modified for visual communication, including a series of scaffolded visual attention-getters such as waving in the child's periphery followed by a gentle tap on the shoulder).

In addition, all caregivers completed the Social Communication Questionnaire (SCQ; Rutter et al., 2003), a validated ASD screener designed to assess social communication, play, and repetitive behaviors. A score of 11 or greater is commonly considered to be indicative of symptoms consistent with autism (Allen et al., 2007; Corsello et al., 2007). Like the ADOS-2, the SCQ was designed for use with hearing children and is written in language that makes reference to speech and hearing. Even if these items are not taken literally by the caregiver/rater, some items on the SCQ are inappropriate for deaf children and cannot be scored. For example, Item 5 asks if children ever mix up their pronouns, a phenomenon that appears to much more prevalent in autistic speech than in autistic signing (Shield et al., 2015). Most deaf children would be penalized on certain other items (e.g., Item 38, “If you come into a room and start talking to her/him without calling his/her name, does she/he usually look up and pay attention to you?”). Thus, although it is a valid and reliable instrument, SCQ scores should be understood as rough estimates of autistic symptomatology when used with deaf populations.

Nonverbal Cognitive Ability

The Test of Nonverbal Intelligence, Fourth Edition (TONI-4; Brown et al., 2010) is a standardized, language-free measure of general intelligence and problem-solving ability. Designed for individuals aged 6 to 89 years, the TONI-4 assesses abstract reasoning using a series of visual analogies and pattern completion tasks. It requires no spoken, signed, or written language, making it especially useful for evaluating cognitive abilities in individuals with speech, language, hearing, or motor impairments. The test yields a standard score with a mean of 100 and a standard deviation of 15, and is widely used in clinical and research contexts.

Sign Language Comprehension

ASL comprehension was assessed using the ASL Receptive Skills Test (ASL RST; Enns et al., 2013). This test presents signed sentences of increasing grammatical complexity and requires the child to select a matching picture from among several alternatives. It is a validated measure of grammatical comprehension in signing children that requires no expressive output. Importantly, the ASL RST has been validated with native signers of Deaf parents as well as early-exposed non-native signers, making it well-suited to our sample.

These measures were administered in ASL and/or in a visually supported format, and all sessions were conducted by researchers fluent in ASL and experienced in working with Deaf autistic children.

Other Measures

When available, we also report relevant results from prior medical or neuropsychological evaluations, as shared by parents.

Results

Cases with ADOS-2 Observations

Case 1: Ethan (Age 4;0)

Ethan, a 4-year-old deaf boy born to Deaf parents and exposed to ASL from birth, was diagnosed with autism at 18 months of age. He occasionally used hearing aids, though he did not have a cochlear implant. At the time of the ADOS-2 assessment, Ethan exhibited no functional expressive signed or spoken language and was not reported to use short phrases or sentences in sign, as confirmed by caregiver responses on the SCQ (total score = 27). Despite full access to a visual language environment since birth, Ethan's expressive communication was profoundly limited.

During Module 1 of the ADOS-2, Ethan did not produce any meaningful signs or gestures, though his mother indicated that he “signs the same thing over and over again” on the SCQ. He did not point, imitate, show, or give; but was observed to use physical manipulation of others’ bodies to communicate (e.g., guiding the examiner's balloon hand to her mouth). These instrumental actions were unaccompanied by eye contact or coordinated affect. Although the ADOS-2 does not include a formal tactile prompt protocol, even direct touch cues failed to elicit joint attention or response to name. Eye gaze was fleeting and not used functionally to regulate interaction.

Social engagement was markedly impaired. Ethan exhibited no spontaneous social overtures during the session, and his only social smile occurred during a familiar peek-a-boo routine with his mother. His facial expressions were minimal and never directed to others, though caregiver report on the SCQ indicated an appropriate use of facial expressions. He showed no interest in gaining or maintaining the examiner's attention, except when help was needed. Rapport was extremely limited, and the examiner had to work actively to sustain even brief episodes of engagement.

Ethan's play was highly restricted. He showed no functional or symbolic play and did not engage in imitation. He explored toys in a repetitive and sensory-seeking manner—most notably, by spinning the wheels of a toy car rather than using it in a representational way. He also attempted to eat non-edible materials (e.g., Play-Doh). His affect was flat throughout, and he did not participate in simple cooperative games or routines.

Motor behaviors were notable for their intensity and frequency. Ethan engaged in frequent stereotypies, including hand-flapping, jumping, and unusual finger posturing, observed during at least two different tasks. He also demonstrated self-injurious behaviors, including head-banging and face-slapping. According to parent report on the SCQ, these behaviors were part of Ethan's regular repertoire and included visually self-stimulating actions (e.g., hand-flapping, moving fingers near his eyes) and repetitive jumping. Parents also reported that he preferred isolated activity, did not engage in cooperative play, and showed no imaginative behaviors.

The ADOS-2 yielded a Social Affect (SA) score of 18 and Restricted and Repetitive Behavior (RRB) score of 3, for a total of 21. His calibrated severity score was 7, indicating a high-moderate level of symptom severity consistent with a classification of autism.

The examiner attempted to administer the TONI-4 and ASL-RST, but Ethan did not respond to any of the items on either assessment, so no data could be collected.

Despite the overall absence of communicative sign use, an unexpected behavior emerged that may warrant future investigation. Ethan was asked to imitate signs corresponding to images on an iPad after the examiner produced the sign. Initially, he exhibited stim-like behaviors and did not sign. However, when the examiner gently touched Ethan's left forearm in an effort to support his signing movement, he began to produce correct signs with his right hand, even initiating signs independently before the examiner prompted him. This phenomenon is discussed further in the Discussion section.

Case 2: Noah (Age 5;1)

Noah, a 5-year-old deaf boy of two Deaf parents, was diagnosed with autism at age 3 following earlier concerns identified at 12 months. Noah had no cochlear implant or hearing aids and was also diagnosed with Waardenburg Syndrome Type 2, a genetic condition characterized by sensorineural hearing loss, pigmentation abnormalities, and typical cognitive development. In addition to native ASL exposure, he was raised in a home that incorporated the PECS (Bondy & Frost, 1994) as an augmentative support strategy.

At the time of his ADOS-2 (Module 1) assessment, Noah demonstrated a limited but emerging expressive signing repertoire, producing three non-echoed signs (MORE ³ , PLEASE, COOKIE). He produced several echolalic signs during the ADOS-2, including immediate repetitions of BUBBLE, MORE, and THANK-YOU. According to the SCQ (total score = 26), caregivers reported that Noah used some short phrases or sentences, but these were not employed for conversational purposes or to convey socially appropriate statements or questions. His signing was noted to include pronoun confusion and idiosyncratic, invented signs. Most of Noah's communicative bids during the ADOS-2 were physical in nature (e.g., patting the examiner's hand with a bubble wand to request play and manipulating the examiner's hands without coordinated eye contact). These behaviors were consistent with caregiver reports that Noah often used physical actions to communicate, even in the presence of accessible visual language.

Noah displayed inconsistent eye contact that was at times used to initiate interaction but was poorly modulated and infrequently coordinated with other communicative behaviors, though caregivers indicated on the SCQ that he would use eye contact while signing or wanting help. He responded to joint attention in one structured trial, orienting toward the target following the third prompt in the “look at that!” sequence. However, he did not exhibit any responsive social smiling, either toward the examiner or his mother, a pattern that aligned with parental report. While he occasionally initiated interaction, these overtures were often idiosyncratic or repetitive in nature and lacked a clear social function. He did not give or show objects during the session.

Despite these limitations, Noah showed some capacity for affective engagement in structured contexts. During bubble play, he smiled and appeared to enjoy the activity, though his affect was rarely directed toward others. Parents reported that he sometimes imitated their actions at home, such as vacuuming or gardening, though these were more ritualized routines than instances of flexible social imitation.

Play behavior was divided between typical and atypical patterns. Noah demonstrated imitated symbolic play with support—for example, mimicking pretend actions such as feeding or hugging a doll. However, no spontaneous pretend play was observed, and much of his play was characterized by sensory-seeking and perseverative interests. He stared at the ceiling for extended periods, repeatedly returned to pop-up toys, and fixated on spinning the wheels of a toy car, holding it close to his face. These observations were corroborated by parent responses on the SCQ, which noted a preference for parts of objects over their conventional use.

Socially, Noah demonstrated attempts to engage with others, but these efforts were unusual in quality or inconsistent with typical social behavior. Although he did not engage in imaginative or cooperative play with peers, parents described him as responsive to other children's approaches and occasionally willing to join in group games. During the ADOS-2, he required substantial adult support to engage and transition between tasks. His affect was occasionally bright but rarely shared.

Noah's restricted and repetitive behaviors included hand-flapping, jumping, and repetitive vocalizations, all observed during multiple activities. He also engaged in self-injurious behavior, slapping himself during the session. These behaviors were confirmed by his caregivers on the SCQ. Additionally, Noah displayed ritualistic behavior, not in the domain of language but in other domains (e.g., routines, physical patterns), reflecting a broader need for sameness and predictability.

The ADOS-2 yielded a SA score of 15 and RRB score of 3, for a total of 18. His calibrated severity score was 6, consistent with a classification of autism with moderate symptom severity.

An attempt was made to administer the TONI-4 and ASL-RST, but were discontinued because Noah did not respond to any of the training or test items on either assessment.

Noah's case reflects a distinct profile of a minimally expressive Deaf autistic child with partial sign language use, yet minimal social-functional application of that language. He showed the ability to produce a few single signs, but was otherwise imitative, echolalic, and ritualized, mirroring broader social-communicative difficulties.

Case 3: Maya (Age 6;9)

Maya, a 6-year-old deaf girl with Deaf parents and early exposure to ASL, had a complex medical and developmental history. She received a diagnosis of pervasive developmental disorder not otherwise specified (PDD-NOS) in early childhood and treatment for a right cerebellar desmoplastic medulloblastoma diagnosed at age five. Her treatment included high-dose chemotherapy, bone marrow transplant, and ventriculostomy for hydrocephalus. Prior to her medical diagnosis, her parents observed an early regression in sign production and eye contact around age two, accompanied by increased social withdrawal and motor stereotypies. These early signs led to a diagnosis of autism through her regional center.

At the time of a slightly earlier neuropsychological evaluation (age 6), Maya had recently returned to kindergarten after medical treatment. She was described as having age-appropriate nonverbal intellectual abilities, with a Performance IQ of 93 on the Wechsler Abbreviated Scale of Intelligence (WASI; Wechsler, 1999). Matrix Reasoning performance fell within the average range, and Block Design was in the low average range. However, her expressive language in ASL remained markedly delayed and immature and were characterized by signs that were described by her school interpreter as “toddlerese,” with minimal coordination of gaze and sign and poor intelligibility. She could name only a limited set of familiar objects on a Picture Naming task, placing her in the borderline-impaired range.

Clinically, she showed diminished eye contact, difficulty sustaining attention to visual language input, and significant challenges with social reciprocity. Her behaviors included solitary play, impulse control difficulties, preference for routine, and minimal responsiveness to spoken or signed instructions. Despite this, she demonstrated interest and facility with puzzles and visual-spatial tasks, consistent with relatively preserved nonverbal reasoning.

According to her parents, Maya was socially friendly but often missed out on full participation due to her being the only Deaf child in her educational placement. While they reported strengths in number recognition, problem-solving, and use of an iPad, they remained concerned about her expressive language, frustration tolerance, and ability to benefit from classroom instruction without communication peers. The neuropsychologist recommended a transition to a school with a Deaf peer environment, ongoing behavioral and educational support, and close neurodevelopmental monitoring due to her medical history.

Maya completed Module 1 of the ADOS-2 at age 6;9. Her total score of 18 (SA = 14; RRB = 4) fell in the range for autism classification. During the assessment, Maya used a small but recognizable repertoire of signs and gestures, including CANDY, LOVE, HAVE, and FINISH. However, her language was primarily single-sign utterances, often accompanied by limited or poorly modulated eye contact. Immediate echolalia was present in the form of repeated color and object signs. Gestures were more frequent than in some of her peers, and she did demonstrate some spontaneous pointing with gaze coordination.

Maya showed more pretend play than some of the other children in this case series, including rocking and feeding a baby doll, playing peekaboo, and pretending to cut cake. Nonetheless, her social overtures to the examiner were limited, and her responses to joint attention and name were inconsistent. The examiner noted a clear preference for visual puzzles and symbolic imitation, as well as stereotyped motor behavior (e.g., arm-crossing and rocking) that occurred across several contexts.

Her SCQ total score was 24, consistent with clinical concerns. Parents endorsed challenges in reciprocal communication, social referencing, and use of gestures, as well as repetitive interests and behavioral rigidity. Despite these challenges, parents also indicated several positive behaviors, including talking to be friendly, producing short phrases and sentences, engaging in to-and-fro conversation, producing appropriate facial expressions, nodding yes and shaking her head no, pointing to show, looking at interlocutors while signing, and sharing enjoyment in interaction.

Maya was also given the TONI-4, obtaining a standard score of 98, indicating normal intellectual functioning and corroborating prior performance on the WASI. However, on the ASL RST, she only responded correctly to one item (a practice item intended to ensure that children understand the task). In accordance with the ASL RST manual, the test was discontinued after five consecutive erroneous responses. Her comprehension of ASL corresponded to an age equivalent of under three years.

Maya represents a complex medical case of an autistic deaf child with a history of brain tumor. Despite typical nonverbal cognitive ability and native exposure to ASL, she was minimally expressive in ASL and showed minimal evidence of comprehension.

Case 4: Liam (Age 9;5)

Liam, a deaf boy of two Deaf parents, was observed by our team at age 9;5. His parents also furnished us with two prior reports: a multidisciplinary consultation at age three and a school-based observation at age six. At age three, the clinician (a developmental psychologist) reviewed the Childhood Autism Rating Scale, Second Edition (CARS-2; Schopler et al., 2010) completed by his parents and added her own direct observations. She noted that Liam had a history of eye gaze avoidance, immature social interests, fascination with cause-and-effect toys, and frequent motor stereotypes such as flapping and jumping. His behavior was described as consistent with autistic disorder, and concerns were raised about his minimal use of expressive sign language, despite lifelong exposure.

The school-based observation at age six confirmed that Liam exhibited strong engagement with visual schedules, environmental routines, and visual supports. He was observed to use his part-day individual schedule effectively for transitions and could follow visual cues and structured classroom routines with some prompting. His engagement with both peers and adults was noted, though his communication continued to rely heavily on adult scaffolding. Visual tools such as picture-based choice boards and name sequencing tasks with sensory bins were used successfully to support his participation.

Despite gains in regulation and engagement, Liam remained heavily reliant on adult prompting for communicative initiation, sentence-level signing, and independent task completion. Staff noted that he benefited from dramatic reinforcement (e.g., high-fives with stickers) and would often require several prompts to sign complete sentences or initiate new tasks. The team recommended expanding the use of visual supports (e.g., sentence starters, task strips, and folders) to reduce his dependence on adult guidance.

Liam completed Module 1 of the ADOS-2, yielding an overall score of 18, with sub-scores of 15 in SA and 3 in RRB, placing him within the classification range of autism. His SCQ total score was 31, reflecting significant caregiver-reported impairments in reciprocal social interaction, communication, and restricted behaviors.

During the ADOS-2, Liam demonstrated minimal use of spontaneous, non-echoed sign language, with only isolated recognizable signs produced throughout the session. Much of his expressive output was echolalic, with a high frequency of repeated signs (e.g., MORE) and vocalizations with limited communicative intent. Although he was able to make requests for preferred items like bubbles and snacks, these often occurred without eye contact or coordinated gaze, suggesting limited integration of communicative behaviors. Immediate echolalia and motor stereotypies such as hand-flapping and jumping were prominent. His pointing was absent, and although he used a small set of gestures (e.g., reaching, handing), they were infrequent and not well-coordinated with gaze or facial expression.

Socially, Liam showed unusual and poorly modulated eye contact. He did not respond to the examiner's social smiles or attempts to engage her in joint attention. Facial expressions were infrequent and not typically directed to others. Although he was generally interested in the toys and activities presented, his interactions were largely one-sided and instrumental. He required considerable adult effort to sustain engagement and was primarily motivated by sensory aspects of materials, such as flapping, jumping, or manipulating toys in repetitive ways.

Liam's symbolic and functional play was limited. While he fed a baby doll and pretended to eat during the birthday party activity, these actions were either prompted or extremely brief. His pretend play was characterized more by functional mimicry than by spontaneous or creative use of symbols. Motor stereotypies occurred frequently throughout the assessment and included flapping, body rocking, and smelling objects. No self-injurious behavior was noted. The session was marked by a mildly uncomfortable and effortful rapport, with persistent reliance on the examiner's prompts to initiate or sustain interaction.

Liam's parents’ SCQ responses further supported these findings. Caregivers endorsed that Liam did not use short phrases or engage in back-and-forth communication in sign, rarely made eye contact, and did not engage in pretend play or peer interactions. Items reflecting shared enjoyment, gestures, imaginative play, and response to others’ emotions were largely negative, underscoring pervasive delays in social-pragmatic development despite full access to ASL since birth. However, parents also reported that Liam would show objects of interest to him, indicating shared enjoyment and social awareness.

On the TONI-4, Liam received a standard score of 80, indicating below-average nonverbal cognition. He responded correctly to only two items on the ASL RST (PIGTAIL BOW-BOW RED-CHEEKS “the girl with red cheeks and two bows in her hair” and ICE-CREAM NOTHING “there is no ice cream”). The test was discontinued after five consecutive errors. Liam thus demonstrated minimal comprehension of ASL and a language-age equivalent of under three years.

These findings reinforce the ADOS-2 and SCQ results and highlight a persistent, cross-setting profile of minimally expressive signing in the context of full ASL access and multimodal intervention.

Cases Without ADOS-2 Observations

The following three participants did not complete the ADOS-2, but caregiver responses on the SCQ provide insight into their social-communicative profiles.

Case 5: Caleb (Age 4;11)

Caleb, a deaf boy from a Deaf family, received an autism diagnosis at 24 months. His mother reported that he sometimes wore hearing aids. On the SCQ (score = 23), his mother reported that Caleb was not able to speak or sign using short phrases or sentences or have a conversation. She noted, however, that he sometimes invented new signs. She also reported that he had a tendency to go through rituals and had preoccupying interests that were unusual in their intensity and might seem odd to others. He had unusual sensory interests, and was more interested in parts of a toy or object rather than using the object as it was intended. He engaged in odd finger movements and whole body stimming such as spinning and bouncing. She also reported that he had special objects that he would carry around. Also absent were pointing, nodding his head yes, shaking his head no, and spontaneous imitation of others. In terms of social relationships, he was noted to struggle with peer interactions and rarely initiated social contact, and did not engage in pretend or cooperative play. However, he also was reported to respond positively to other children when they would approach him. At the time of our observation, Caleb did not respond to any of the tasks presented, so the TONI-4 and ASL-RST were discontinued without results.

Case 6: Owen (Age 6;4)

Owen, a deaf boy of two Deaf parents, obtained an SCQ score of 19, above the autism screening threshold. He did not have a cochlear implant or use hearing aids, and his parents reported minimal expressive language and the use of PECS to communicate. His caregiver reported a lack of appropriate facial expressions, the use of others’ hands as a tool, odd finger and hand mannerisms, and complicated body movements. He did not nod his head yes or shake his head no, nor did he point to indicate objects of attention or look others in the face during interaction. In terms of social interaction, Owen did not appear to have any particular friends or a best friend. Also lacking was an engagement in pretend play, imaginative or cooperative social games, and peer engagement. He was noted not to comfort others, and not to initiate or respond to joint activities. However, several positive behaviors were also noted, including sharing enjoyment and reciprocal social smiling. While detailed language data are unavailable, his profile suggests limited spontaneous social communication, consistent with the patterns observed in the ADOS-assessed group. Also consistent with limited receptive language, Owen did not respond to any items when an attempt was made to administer the TONI-4 and ASL-RST, so no scores were obtained on these measures.

Case 7: Sophia (Age 10;5)

Sophia, a deaf girl of two Deaf parents, scored 17 on the SCQ. She had been diagnosed with autism at age 3 and had comorbid diagnoses of epilepsy and intellectual disability. She did not wear hearing aids or have a cochlear implant. Her caregivers reported some strengths, including interest in peers, shared enjoyment, and occasional pretend play. However, difficulties were noted in cooperative and imaginative play, emotional reciprocity, and facial expressiveness. She also was reported to engage in odd hand and finger mannerisms as well as complicated whole-body movements. She did not spontaneously point, nod her head yes, shake her head no, or spontaneously copy others. While her parents indicated that she was able to produce single signs, she did not engage in to-and-fro conversation and would sometimes sign the same thing over and over again. Her parents also reported that she showed unusual and intense interest in parts of objects and a preoccupation with ritual. Sophia's SCQ profile suggests a less severely affected presentation, but one still marked by atypical social-pragmatic behavior. Sophia did not respond to the examiner's attempts to administer the TONI-4 and ASL-RST, so both measures were discontinued.

Summary

Across these seven cases, all participants demonstrated limited expressive sign use despite full ASL exposure from birth by their Deaf parents, along with consistent impairments in joint attention, pretend play, and gesture integration. While some children (e.g., Maya, Noah) produced isolated signs, these were mostly echolalic and were not embedded in longer sentences. Five of the seven children did not respond to any items on the ASL RST or TONI-4, limiting our ability to comment on their comprehension of ASL or nonverbal cognition and suggesting the possible presence of language impairment and/or intellectual disability. For the two children who were able to comply with those tasks, comprehension of ASL was minimal, while nonverbal cognition was typical (for Maya) or below average (for Liam). Self-injurious behaviors and repetitive motor actions were common. While developmental trajectories varied, there was a consistent presence of minimally expressive signing in this subgroup of deaf children.

Discussion

This case series described seven Deaf children with autism, all born to Deaf parents and exposed to ASL from birth. Four were assessed with the ADOS-2 and all had SCQ data; six met or exceeded the SCQ threshold, and all four ADOS-2 cases met diagnostic criteria. Critically, each child showed minimal expressive signing despite full ASL access, suggesting modality-independent factors constrain language in some autistic children.

This is a novel finding. Prior work on minimally-verbal autism has centered on spoken language (e.g., Tager-Flusberg & Kasari, 2013), while reports in deaf populations often involve children from hearing families with limited sign exposure (Jure et al., 1991; Meinzen-Derr et al., 2014; Roper et al., 2003). By contrast, all children here were native ASL signers. Their limited expressive signing cannot be attributed to deprivation, highlighting the need to investigate core neurodevelopmental mechanisms. The children in this case series, with the exception of one syndromic case (Maya), appear to reflect what Mottron and Gagnon (2023) describe as prototypical autism—that is, autism characterized by consistent, modality-independent differences in social communication, generativity, and perception rather than by secondary factors such as language deprivation or intellectual disability. Their convergence with established autistic phenotypes across modalities underscores that the mechanisms constraining expressive language are intrinsic to autism itself, not contingent on language modality or access. Across cases, hallmark autistic features were observed: reduced reciprocal communication, limited joint attention, and restricted/repetitive behaviors. Integration of gaze, facial expression, and manual signs was sparse, and coordinated pointing to share attention was rare. Expressive output ranged from none (Ethan) to small, context-bound repertoires with echolalia or idiosyncratic signs (Noah, Maya). Symbolic play tracked these differences: the child with the most signs (Maya) showed the most pretend play, while the child with no signs (Ethan) also showed no pretend play, supporting accounts of reduced generativity in autism (Eigsti et al., 2011).

Because ASL relies on coordinated gaze, facial grammar, and pointing, early social-attentional differences may carry heightened costs in signing contexts. Behaviors that are ancillary in speech become grammatical and pragmatic in sign; difficulties with gaze/gesture integration may therefore disproportionately affect fluency. While non-social pathways to language acquisition have been documented in hearing autistic children (i.e., so-called unexpected bilinguals; Kissine, 2021), it is currently unclear if such non-social pathways to language exist in analogous ways for deaf, sign-exposed children.

Importantly, these expressive limitations cannot be explained by auditory processing deficits. Unlike minimally verbal hearing cohorts, where atypical auditory processing has been implicated (e.g., Alcántara et al., 2004; Bhatara et al., 2013; Russo et al., 2009), our Deaf cohort had no functional auditory input. Their outcomes point instead to domain-general mechanisms such as generativity, social-cognitive integration, or sensorimotor planning. Future work should also examine visual processing (e.g., CVI, visual crowding, competition between object- vs. signer-focused attention) as potential barriers to signed language.

Diagnostic and Interpretive Challenges

The ADOS-2 and SCQ require careful adaptation for Deaf children. Certain items (e.g., response to name) are inappropriate, likely lowering scores compared to hearing peers. Moreover, gaze, gesture, and facial expression are not merely paralinguistic in ASL but linguistic. Reduced eye contact, for example, may disrupt syntax as well as social engagement.

In one case (Ethan), a light touch to the forearm coincided with improved sign imitation. Though anecdotal, this suggests proprioceptive input may transiently support motor planning for signing. Prior work shows light touch can reduce stress and enhance responsiveness (Coan et al., 2006; Escalona et al., 2001; Holt-Lunstad et al., 2008). Prospective single-case designs could test this mechanism.

Clinical and Research Implications

Several conclusions emerge:

Minimally expressive sign language occurs in Deaf autistic children with native ASL exposure. This challenges assumptions that minimal language in Deaf autistic children reflects deprivation and questions the assumption that signed languages may be inherently more accessible to autistic children. In our sample, seven of 23 children (30.4%) fit this profile. While this proportion should not be taken as a population estimate, it is noteworthy that it falls within the range reported for minimally verbal autism in hearing cohorts (Tager-Flusberg & Kasari, 2013).

Core autistic impairments manifest across modalities. Generativity, joint attention, and symbolic play deficits appeared regardless of language modality.

The visual-manual modality poses unique challenges. Difficulties with gaze, gesture, and facial expression may disproportionately disrupt ASL fluency, though we did not find a higher rate of minimally expressive signing in this sample than what has been reported for hearing children exposed to a spoken language.

Stimming may directly interfere with sign. Hand-flapping, finger posturing, and body rocking may obstruct the articulators required for signed communication.

Standard tools remain useful but limited. Clinicians should interpret scores cautiously and adapt items for Deaf signers (see Phillips et al., 2022).

Strengths-based, neurodiversity-affirming approaches are critical. Several children showed preserved nonverbal cognition, visual-spatial strengths, and engagement with structured routines. Interventions should build on these competencies while ensuring exposure to ASL alongside AAC and speech.

Future research should prioritize cross-modal comparisons, develop tools normed for signing autistic populations, and use longitudinal designs to track developmental change.

Limitations

Conclusions are constrained by the small sample, incomplete standardized data for several children (noncompliance/feasibility), and absence of a matched comparison group (e.g., native-ASL autistic children who are not minimally expressive). Diagnostic instruments required adaptations that may affect scores, and our cross-sectional snapshots limit inferences about change. Finally, because many participants were unable to complete the TONI-4 or the ASL-RST, our assessment of nonverbal cognition and ASL comprehension is incomplete; future work could incorporate alternative nonverbal measures such as Raven's Colored Progressive Matrices, which minimizes linguistic demands and is commonly used in both Deaf and autistic populations (Courchesne et al., 2015). These limitations underscore the need for longitudinal, multimethod designs with modality-appropriate measures.

Conclusion

This study contributes to a growing understanding of autism in diverse linguistic and cultural contexts. The presence of minimally expressive sign language in Deaf autistic children from Deaf families with full access to ASL from birth demonstrates that language deprivation is not a necessary condition for profound expressive language delay in the deaf autistic population. Indeed, the fact that such children exist in such numbers (seven of 23 deaf children with autism from Deaf families) is a testament to the profound challenges that some autistic children have with accessing language, be it signed or spoken. Documenting this phenomenon also yields critical insights into the nature and origin of expressive language difficulties in autism: for the children in this study, for example, auditory processing deficits cannot be responsible for the expressive language difficulties observed. It also underscores the need to reevaluate longstanding assumptions about communication in autistic individuals, particularly in non-speaking populations. By recognizing both the challenges and the strengths of deaf autistic children, we can begin to build more inclusive models of development, diagnosis, and support.