By Ehud Yairi, Ph.D.
Professor Emeritus at University of Illinois

Let me begin with a general truism: Language can exist without speech but speech cannot exist without language. That is, language is independent whereas speech depends on an internal language system to produce meaningful vocal sounds. Although stuttering is widely recognized by the general public, professional speech clinicians, and people who stutter, as a disorder of speech production, one cannot escape recognizing that language components are also involved. Perhaps the first to do so are people who stutter, both older children and adults, who, based on self-experiences, have realized that certain speech sounds, that are not necessarily more motorically demanding and/or multisyllabic words, are more difficult for them to speak. Indeed, there is strong evidence that a person’s instances of stuttered speech do not occur at random. Early modern research (e.g., Brown, 1937, 1945; also see review by Bloodstein et al., 2021) demonstrated strong influences of apparent linguistic features on the specific locations where stutter events, sometimes referred to as “moments of stuttering,” occur. For example, stuttering events tend to occur at the beginning of sentences and phrases, on words of certain grammatical classes, such as verbs and adjectives, etc. Still, some or much of this phenomenon can also be ascribed to motoric factors. 

Another line linking stuttering to language seems to be especially intuitive in relation to young children, as most begin to stutter by 36 months of age. An intensive longitudinal study reported that nearly 38% of stuttering onsets occur by 30 months of age (!), reaching over 65% by 36 months of age (end of 3rd year of life), increasing to 85% by 42 months of age and to 95% by age 48 months -- the 4th birthday (Yairi & Ambrose, 2005). 

The mean age at onset according to 5 robust studies is very young — 32 months (Yairi, Seery, & Singer, 2025). One study even reported a mean of only 29 months (Reily et al., (2009)! This is the period when children’s expressive and receptive language capacities rapidly expand (Levina, 1963; Ratner, 1997). Is it possible that the fast-growing language skills are taxing the child’s speech production system? Nearly 60 years ago Cheverkeva (1967) went so far as proposing that stuttering is a disorder of language development, a notion subsequently echoed by Bloodstein (2002). Perhaps. However, as pointed out by Alm (2004), this is also the peak period for dopamine receptors to occur in the basal ganglia affecting the ability to produce timing cues. In any event, considering the principle of parsimony, that is, applying the smallest number of variables to account for the largest number of facts, and based on available scientific information, my current view is that stuttering is basically a motor disorder* affected by other domains, such as emotionality and language, including the demands of social discourse, as communication does not occur in a vacuum. Interestingly, in this respect, Alm (2026) has pointed out that the overwhelming majority of children with low linguistic abilities do not stutter! Scientific views, however, should accommodate new information as presented later in this paper. Also, many years ago we suggested the possibility that stuttering, phonology, and language problems occurring simultaneously in young children arise independently from a common biological cause (e.g., E. Paden in Yairi & Ambrose, 2005). 

A growing interest in the language aspect of stuttering has also been reflected in several psycholinguistic models of the disorder as depicted in the Fault Line Hypothesis (Wingate, 1988), the Covert-Repair Hypothesis (Postma & Kolk, 1993), the Trade-Off Hypothesis (Ratner, 1997), the Cognitive Interference Model (Bosshardt, 2002), and others. They do not propose that people who stutter lack knowledge of the vocabulary or phonology of their language nor do they suggest that stuttering results from motoric interference. Rather, they propose that stuttered speech events result from subtle aberrancies in mental operations that plan and transform intended language messages into articulated words. Thus, a theory may assert that stuttering is a disorder of phonological encoding -- an impairment of the processes responsible for timely/precise retrieval of sequential phonological elements, primarily phonemes (the smallest speech sounds units of a language), for smooth speech outcome. That is, speakers have problems with timely assembly of their intended verbal message even if they know the intended words, the required speech sounds, and the rules for their use. Clearly, the psycholinguistic viewpoint shifts the focus from the motor system of speech production to central (brain) processes involved in the planning of intended speech. The clinical implications of the psycholinguistic theories are that people who stutter (1) should slow down when initiating and continuing to talk, allowing extra time for their systems to plan the correct sequence/timing of speech sounds, and (2) practice speech sound combinations to increase their speed and accuracy. Young children would be expected to benefit from phonological awareness activities. Here I note that a very recent NIH-sponsored study of single brain cells (Jing et al., 2026) appears to strongly support the role of the above-mentioned psycholinguistic processes, commending the practice of sequentially programmed phonological elements for running speech. They reported that certain neurons represented the detailed grammatical relationship between words; that neuronal recordings observed just before participants spoke predicted many properties of their subsequent speech production. There was a division of labor among neurons, with some reflecting basic information, such as the meaning and roles of specific words, while others tackled more complex tasks, including grouping phrases into structured sentences. Hopefully, such research will be expanded to stuttering, shedding light on the role of language in the dynamics of stuttered speech.

In the past, language research in relation to stuttering dealt with five distinct variables: (a) Language Complexity: the intricacy of vocabulary, sentence structure, and the broad range of ideas expressed,  (b) Pragmatics: how language is used in real-life communication, (c) Phonology: the proper use of the language sound system, (d) Loci of  stuttering events, and (e) Language Skills: the ability to understand and use language effectively.  Most of this work has been conducted with children. For example, investigators reported increased stuttering as a function of language complexity (Logan & Conture, 1995; Zackheim & Conture, 2003). A very comprehensive review can be found in Bloodstein et al. (2021). Particularly relevant to clinicians and parents is the question of whether children who stutter (CWS) are more likely than children who do not stutter (CWNS) to exhibit language development difficulties, or even impairments. Bloodstein et. al. (2021), and Yairi et al. (2025) noted some support to this notion, whether in data or scholarly opinion, published early in the current century, including Arndt & Healey (2001), Wingate (2001), and Blood et al. (2003), as well as more recently (e.g., Luckman et al. 2020 and Koenraads et al., 2025). The Blood et al. (2003) team reported that 26% of school-aged CWS revealed language disorders. Several years later, an analysis of 22 studies, mostly of school-age children of up to age 8, concluded that CWS scored lower than CWNS on standard measures of both receptive and expressive language (Ntourou et al., 2011). Note, however, that (a) lower scores do not necessarily mean a “disordered” classification, and (b) the participant children were years beyond the typical age at stuttering onset that occurs at a mean of 32 months as mentioned above. Hence, their results pertain to only a small minority -- 20% or fewer -- of all children who ever stuttered because so many of them have already stopped stuttering due to natural recovery (see Yairi & Ambrose, 1999).

A more recent analysis of 19 studies of general population samples, aged 1 to 16 years, totaling nearly 25,000 children (Hill et. al., 2023) revealed prevalence of low expressive language skills ranging from 7.7% to 25.2% and prevalence of low receptive language skills ranging from 5% to 16.9%! Also, the Language section at NIH’s link, https://www.nidcd.nih.gov/health/statistics/quick-statistics-voice-speech-language, reveals a 7% figure for incidence of language disorders in children nation-wide. Unfortunately, similar data for a stuttering population aged 1 to 16 are not available. One may, however, roughly infer the missing information from Blood et al.’s (2003) finding that 26% of school age CWS demonstrated language disorders. Note that by age 8 or so, their results pertain only to a minority of 20%-25% (likely only 20%) of the original population of the schoolchildren who stuttered – the ones who did not experience natural recovery.  Now, if we assume that the 75% of CWS who have already recovered did not have histories of language disorders, then 26% of schoolchildren who stuttered would correspond to only 6.5% of the initial total population of children who ever stuttered presenting with disordered language -- close to the 7% national population incidence as per the NIH summary. In this case, it is possible to infer that stuttering, in general, is not associated with a higher risk of language disorders than what occurs in the  and that only the persistent group may exhibit such problems. But, if this assumption is wrong, and some or many of the 75% of children who recovered from stuttering also had language disorders, then language skills are not a dependable risk predictor of persistency and recovery. Still, perhaps only a sub-population of children who stutter, those who persist in stuttering, present such problems. In this case, language status may have some prognostic value. 

A few studies explored language skills in young preschool children close to the time of stuttering onset, reflecting the status prior to natural recovery among a much larger stuttering population. Early results indicated typical language skills. For example, Bernstein Ratner, and Silverman (2000) studied 15 3-year-old CWS under 4 months post stuttering onset finding language scores within the normal range. Similar conclusions were reported by Ryan (1992), Anderson and Conture (2000; 2004), and Anderson et al. (2005). Of special interest are the Kloth et al., 1995a, 1995b studies that tracked 96 children at risk for stuttering because one or both of their parents stuttered. The 26 children who, after one year, began stuttering, did not differ from the CWNS on speech-language measures made prior to stuttering onset. In my judgment, due to the unique nature of this study, its findings are particularly convincing. Also, at the University of Illinois Stuttering Research Project, in a large longitudinal study of 84 children, ages 2 to 5, all tested close to stuttering onset, language skills data yielded mean scores within the norm (Watkins, Yairi & Ambrose, 1999). The minority of children who continued to stutter, however, presented greater variability and atypical patterns of language production compared to those who later recovered naturally. A German team also reported that preschoolers who stuttered had language skills at or above age norms (Hage, 2001; Rommel et al., 1999), as did an Australian team whose longitudinal study tapped children very close to stuttering onset. Their CWS had higher verbal and nonverbal language scores than CWNS (Reilly et al., 2013). Interestingly, it is not infrequent that parents report that their child had a spurt of language development just prior to the onset of stuttering including speaking in longer sentences and using new words. They often say that "his brain seemed to be working faster than his mouth" could manage.

In discussing the Illinois studies, Nippold (2012) concluded that a stuttering-language connection was not supported and posited that language difficulties in CWS, when existing, are but a reflection of a compromised motor control system. Initially R. Watkins of the Illinois program, concluded that “there is virtually no evidence that language development is vulnerable in any significant number of young children who stutter” (see her chapter in Yairi & Ambrose, 2005). Later analyses of our complete data, however, did conclude that at one or two years post-onset, both expressive and receptive language abilities in the subgroup of children who persisted in stuttering lagged behind those observed in children who recovered (Ambrose et al., 2015). Still, the mixed findings led us to question the predictive power of a child’s language skills regarding persistency or natural recovery. Note again that “lag behind” does not necessarily mean “disordered.” Five years later, a meta-analysis of 11 studies conducted by Singer et al. (2020) also concluded that children who persisted in stuttering performed lower than those who recovered on standardized language tests. Subsequently, Sugathan and Maruthy (2021), in reviewing 19 pertinent studies, also concluded that, at the time of stuttering onset, children who later persisted in stuttering exhibited lower expressive language skills compared to those who eventually recovered. I question, however, the validity of their statement regarding “…time of stuttering onset” – as most studies were not careful at establishing this all important parameter.  

Our question, then, is how strong the language predictor is, and how it compares with other predictors?  Of course, it all depends on the strength of the research concerned with natural recovery/persistency and two all-important parameters: (1) children’s age at the initial observation in a longitudinal study, and (2) the time interval between that initial observation and the stuttering onset as accurately determined. These two parameters determine the validity of the reported rate of natural recovery and how well it represents the general population of preschool children who ever stuttered. Clearly, the older the children, and the longer the elapsed time from stuttering onset at the beginning of a longitudinal study, the less accurate the findings will be because children who naturally recovered prior to the starting point of the study will not be accounted for. The potential recovery predictors for these already-recovered younger children also will be missing. Are such potential predictors also language related? Furthermore, this missing early recovered group may be significant in size. That natural recovery also occurs in two- and three-year old children who stutter has been repeatedly reported (Ryan 2001; Yairi & Ambrose, 1999, 2005; Mansson, 2005). Hence, unquestionably, studies that tracked stuttering from ages 4 or 5, or even later, missed all those cases of earlier recovery. Therefore, their reported rate of natural recovery is rather inaccurate for the stuttering population of young children.

Illustrating the preceding comments is a study of 91 Swedish school children reporting only 47% recovery over 10 years (Fritzell, 1976). It is hopelessly flawed because many participants were first observed between ages 7 and 9, some even older, whereas many natural recovery cases occur several years before this age. Likewise, the Purdue University studies (e.g., Spencer & Weber-Fox, 2014; Leech et al., 2017; Gerwin et al., 2019; Walsh et al., 2021; 2022) were initiated with 52 children, 36 boys and 14 girls, aged 4 and 5. The absence of younger children who stutter (e.g., ages 2 and 3) is reflected in their appreciably low natural recovery rate, only 44%, when the expected rate is 75%-80% or even higher. This means that at least 30% of children who began stuttering were not covered in these studies! Should we accept that recovery/persistency predictors derived from 4- and 5-year olds’ language skills are applicable to all preschoolers who stutter, many of whom are only 2- and 3-years old when first observed to stutter? Should we assume that the language skills of children who recovered at earlier ages were above par? Also, inasmuch as children’s language performance is known to be higher for girls than for boys, the children’s sex distribution is another overlooked factor in early prediction studies that employed 4 and 5 year old preschoolers. Whereas the male/female ratio of 4:1 is recognized for the general population of people who stutter, it is only 2:1 at the initial stage of the disorder (Yairi & Ambrose, 2005; 1999a). By ages 4 and 5, however, that ratio is appreciably altered in favor of males due to the earlier natural recovery of girls. Thus, whereas the Illinois studies began with a 2:1 male to female ratio, it was 3:1 in the Purdue studies. Obviously, risk factors based on language skills of 4-5-year-olds cannot be applied to 2- or 3-year olds who stutter when the mean age of stuttering onset according to several studies is about 32 months. Therefore, in my view, a low language score as a risk factor for persistent stuttering, is insufficient. It may contribute to clinical decisions of older preschoolers only as a supplementary sign. There are no language-based risk signs for 28-month-old children who begin to stutter! One additional reason for the limitations of past language studies stems from the global nature of the language measures employed that yield superficial information, particularly regarding handling online demands of language use in context. Finer language tasks may contribute to clarifying the developmental relation between stuttering and language in general, especially at the early stages of the disorder.

Finally, more than 20 years ago, Yairi and Ambrose (2005) delineated 15 risk predictors for early childhood stuttering to become a persistent disorder. These were identified through their intensive longitudinal investigation that followed 84 children from close to their stuttering onset and for several years with a mean age at stuttering onset being 33 months. These predictors were grouped into three levels according to their predictive power. Risk factors in the strongest group in order are: (1) positive family history, (2) gender, (3) overt stuttering trends, (4) time since onset, (5) age at onset. They advocated for the inclusion of multiple factors in each case. Note that language data were not included. With the most recent advances in the genetics of stuttering (e.g., Polikowsky, et al., 2025) it would appear to me that transitory skill domains, such as motor and language, will become less significant in early prognosis of persistency in, or recovery from, stuttering.

*I expressed a similar view in a term paper presented in my first year as an M.A. student at the University of Iowa in 1965.

Ehud Yairi, Ph.D., is Professor Emeritus in the Department of Speech and Hearing Science at the University of Illinois Urbana-Champaign and an internationally recognized authority on stuttering. His research has focused especially on the onset, development and treatment of stuttering in young children. Yairi has received major honors for his contributions to speech-language research, including the American Speech-Language-Hearing Association’s Kawana Award for lifetime achievement.

REFERENCES
Alm P. (2004). Stuttering and the basal ganglia circuits: a critical review of possible relations. Journal of Communication Disorders, 37, 325-369.
Alm, p. (2026) Personal communication.
Ambrose, N., Yairi, E., Loucks, T., Seery, C., & Throneburg, R. (2015). Relation of motor, linguistic and temperament factors in epidemiologic subtypes of persistent and recovered stuttering: Initial findings. Journal of Fluency Disorders, 45, 12–26.
Anderson, J., & Conture, E. (2000). Language abilities of children who stutter: A preliminary study. Journal of Fluency Disorders, 25, 283-304.
Andrews, G., Craig, A., Feyer, A., Hoddinott, S., Howie, P., & Neilson, M. (1983). Stuttering: A review of research findings and theories circa 1982. Journal of Speech and Hearing Disorders, 48, 226-246.
Arndt, J., & Healey, C. (2001). Concomitant disorders in school-age children who stutter. Language, Speech and Hearing Services in Schools, 32, 68-78.
Bernstein Ratner, N., & Silverman, S. (2000). Parental perceptions of children’s communicative development at stuttering onset. Journal of Speech, Language, and Hearing Research, 43, 1252–1263.
Blood, G., Ridenour, V. J., Qualls, C. D., & Hammer, C. S. (2003). Co-occurring disorders in children who stutter. Journal of Communication Disorders, 36, 427-448.
Bloodstein, O. (2002). Early stuttering as a type of language difficulty. Journal of Fluency Disorders, 27, 163-167.
Bloodstein, O., Bernstein Ratner, N., & Brundage, S. (2021). A handbook on stuttering (7th ed.). Plural Publishing.
Bosshardt, H. (2002).  Effects of concurrent cognitive processing on the fluency of word repetition: Comparison between persons who do and do not stutter. Journal of Fluency Disorders, 27, 93-114.
Brown, S. (1937). The influence of grammatical function on the incidence of stuttering. Journal of Speech Disorders, 3, 223-230.
Brown, S.  (1945). The locus of stuttering in the speech sequence. Journal of Speech Disorders, 10, 181-192.
Chevekeva, N. (1967). About methods of overcoming stuttering: A survey of the literature. Spetsial Shkola, 3, 9-15.
Fritzell, B. (1976). The prognosis of stuttering in schoolchildren: A 10-year longitudinal study. In Proceedings of the XVI Congress of the International society of Logopedics and Phoniatrics (pp. 186–187). Karger
Gerwin, K., Brosseau-Lapre, F., Brown, B., Christ, S., & Weber, C. (2019). Rhyme production strategies distinguish stuttering recovery and persistence. Journal of Speech, Language, and Hearing Sciences, 62(9), 3302–3319.
Häge, A. (2001). Können kognitive und linguistische Fähigkeiten zur Verlaufsprognose kindlichen Stotterns beitragen? (Cognitive and linguistic abilities in young children: Are they able to predict the further development of stuttering?).  Sprache Stimme Gehör, 25, 20-24.
Hill, E., Calder, S., Candy, C.  Truscott, G., Kaur, J., Savage, B., & Reilly, S. (2023). Low language capacity in childhood: A systematic review of prevalence estimates. First published: 10 August 2023. https://doi.org/10.1111/1460-6984.12944 Digital Object Identifier (DOI)
Jing Cai et al. (2026). Mapping the neuronal building blocks of human language with language models. Nature.  DOI: 10.1038/s41586-026-10691-5 
Koenraads, S., Jansen , P., Labuschagne, J., van der Schroeff, M., & M.C. Franken, M. (2025). Risk of stuttering onset and persistence linked to early language skills: results from the Generation R Study. Journal of  Fluency Disorders, 106-145.
Leech, K. A., Bernstein Ratner, N., Brown, B., & Weber, C. M. (2017). Preliminary evidence that growth in productive language differentiates childhood stuttering persistence and recovery. Journal of Speech, Language, and Hearing Research, 60(11), 3097–3109. 
Levina, R. (1966). Study and treatment of stammering children. In A. D’yachov (Ed.), Specialized Schools, 120, (4). Moscow: Education Publishing House. (Translation published in Journal of Learning Disabilities (1968), 1, 26-29.)
Logan, K. & Conture, E. (1995).  Relationships between length,grammatical complexity, rate, and fluency of conversational utterances in children who stutter. Journal of Fluency Disorders, 20, 35-61.
Luckman, C., Wagovich, S., Weber, C., Brown, B., Chang, S., Hall, N. & Bernstein Ratner, N. (2020). Lexical diversity and lexical skills in children who stutter. Journal of Fluency Disorders, 63, 105747
Månsson, H. (2005). Stammens kompleksitet og diversitet [Therapy groups for preschool children who stammer]. Dansk Audiologopædi, 41, 13–33.
Miles, S., & Ratner, N. B. (2001). Parental language input to children at stuttering onset. Journal of Speech, Language and Hearing Research, 44, 1116-1130
Nippold, M. (2012). Stuttering and language ability in children: Questioning the connection. American Journal of Speech-Language Pathology, 21, 183–196.
Ratner, N. (1997). Stuttering: A psycholinguistic perspective. In R. F. Curlee and G. M. Siegel (Eds.) Nature and Treatment of Stuttering: New Directions. Boston: Allyn & Bacon.
Rommel, D., Hage, A., Kalehne, P., & Johannsen, H. (1999). Developmental, maintenance, and recovery of childhood stuttering: Prospective longitudinal data 3 years after first contact. In K. Baker, L. Rustin, & K. Baker (Eds.), Proceedings of the Fifth Oxford Disfluency Conference (pp. 168–182). Windsor, Berkshire.
Polikowsky , H. G., et al. (2025). Large-scale genome-wide analyses of stuttering. Nature genetics , 57., https://doi.org/10.1038/s41588-025-02267-2. S
Postma, A., & Kolk, H. (1993). The covert repair hypothesis: Prearticulatory repair in normal and stuttered disfluencies. Journal of Speech and Hearing Research, 36, 472-487.
Singer, C., Hessling, A., Kelly, E., Singer, L., & Jones, R. (2020). Clinical Characteristics Associated With Stuttering Persistence: A Meta-Analysis. Journal of Speech, Language, and Hearing Research, 63, 2995-3015.
Spencer, C., & Weber-Fox, C. (2014). Preschool speech articulation and nonword repetition abilities may help predict eventual recovery or persistence of stuttering. Journal of Fluency Disorders, 41, 32–46. 
Starkweather, W. (1987). Fluency and Stuttering. Englewood Cliffs, NJ: Prentice-Hall.
Walsh, B., Bostian, A., Tichenor, S. E., Brown, B., & Weber, C. (2020). Disfluency characteristics of 4- and 5-year-old children who stutter and their relationship to stuttering persistence and recovery. Journal of Speech, Language, and Hearing Research, 63, 2555–2566.
Walsh, B., Christ, S., & Weber, C. (2021). Exploring relationships among risk factors for persistence in early childhood stuttering. Journal of Speech, Language, and Hearing Research, 64(8), 2909–2927. https://doi.org/10.1044/2021_JSLHR-21-00034
Watkins, R. , & Johnson, B. (2004).  Language abilities in young children who stutter: Toward improved research and clinical applications. Language, Speech and Hearing Services in Schools.   
Watkins, R., Yairi, E., & Ambrose, N. (1999).  Early childhood stuttering III: Initial status of expressive language abilities. Journal of Speech, Language, and Hearing Research, 42, 1125-1135.
Wingate, M. (2001). SLD is not stuttering. Journal of Speech, Language, and Hearing Research, 44, 381-383.
Yairi, E. (1983). The onset of stuttering in two- and three-year-old children: A preliminary report. Journal of Speech and Hearing Disorders, 48, 171-177.
Yairi, E., & Ambrose, N. (1999). Early childhood stuttering: I. Persistency and recovery rates. Journal of Speech, Language, and Hearing Research, 42, 1097–1112.
Yairi, E. & Ambrose, N. (2005).  Early Childhood Stuttering.  Austin: Pro Ed. 
Yairi, Seery, C., & Singer, C. (2025). Stuttering: Foundations and Clinical Applications (4th edition) San Diego: Plural.
Zackheim, C. & Conture, E. (2003). Childhood stuttering and speech disfluencies in relation to children's mean length of utterance: A preliminary study. Journal of Fluency Disorders, 28, 115-142.  

Posted Oct. 23, 2026