About the presenter: Patricia M. Roberts, Ph.D., CCC-SLP, SLP(C), currently Associate Professor in speech language pathology at the University of Ottawa, in Canada's capital. She holds degrees from Queen's University (Kingston, Canada) and Florida State University and obtained her Ph.D. from the Universite de Montreal.In her first career as a clinical SLP, she worked with many bilingual clients and was privileged to have the late Marie Poulos as colleague and mentor. She is spending much of her 2nd career as a professor and researcher trying to understand the many unsolved puzzles of bilingual stuttering.

You can post Questions/comments about the following paper to the author before October 22, 2010.

Myths and Mysteries of Bilingual Stuttering

by Patricia Roberts
from Canada

In this presentation, I will focus on four mysteries (things we do not yet know) about stuttering in bilingual children and adults and some of the myths associated with these gaps in our current knowledge.

To make this essay easier to read, I won't say "bilingual or multilingual" each time the word "bilingual" comes up but in most places, what applies to bilinguals also applies to multilingual speakers - as far as we know - so far.

Mystery 1: How many bilingual people are there?

It is sometimes confusing to even try to discuss bilingualism because the word bilingual means different things to different people. For some people, bilinguals are people who speak two (or more - for multilinguals) languages equally and perfectly. People who speak two languages in their daily lives, and can do most things such as talking to people at work, reading the newspaper, understanding conversations with friends sometimes say "Oh yes, I can do all that. But I am not bilingual". Other people describe themselves as bilingual if they can communicate basic ideas, even if they make many errors in grammar and pronunciation and have a very small vocabulary in one language.

In research, both these kinds of people are seen as having different levels of bilingualism. Ratings from 1 to 7 or 1 to 9 are often used to estimate where each person falls along the continuous line that goes from "I really ONLY know one language" to "I am one of those rare people who feels equally at home in two languages, no matter what the task or topic". For speaking, hearing, reading, and writing, most of us are at slightly different levels of ability, in each of the languages we know.

For this essay - and the discussions I hope it will spark - let's think of bilingualism as being a continuum. We don't divide the world into tall people and short people. There is no rational cut-off to separate "tall" from "short". Same thing for "bilingual" and "unilingual". Everyone is at some point along the line that goes from strongly unilingual to very, very bilingual.

With a broad definition of bilingualism, some authors estimate that there are at least as many people in the world who need to use two or more languages in their daily lives as there are people who can only function in one language (see, for example, Bhatia & Ritchie, 2006). We cannot make precise estimates unless we first define what levels of bilingualism are included or excluded from the count (and where to divide dialects from languages).

Mystery 2: Is the incidence of stuttering the same in different languages?

There are studies of the incidence of stuttering in different countries. Some authors use these studies to say things like "the incidence of stuttering is higher in Country X than in Country Y". But, if each study used different ways of sampling and different ways of determining who stutters, it is not valid to compare across studies. For example, from one study to the next, different methods were used in deciding who is stuttering: parent reports? Teachers in schools or day care centres? Parents remembering what the child was like 5 or 10 years ago? There are also differences in what counts as stuttering: only for a few months at age 3? Only people who stuttered for more than a year? Only those who reach a given level of severity?

When people see these reports, they often speculate about why the incidence figures SEEM to be different (ignoring the differences in how the estimates were reached), often using their favourite aspect of stuttering as the explanation. Thus, we see explanations like:

1) "There is more stuttering in Country X than in Country Y because the grammar of the language spoken in Country X makes greater demands on memory...." The complexity of a language might be relevant, in some subtle ways including the location of moments of stuttering within a sentence. Concluding that the language itself influences the number of people who stutter requires a huge, dangerous leap of logic. There are other possible explanations that have to be ruled out before we can select one of them and reject the others.


2) "There is more stuttering in ___ because that culture views speaking well as very important and the pressure to speak well makes people stutter". This explanation now seems very unlikely, given what we know about the causes of stuttering.

Now that we understand the importance of genetics and the inherited nature of stuttering for many people, it seems logical to ask whether, in some ethnic groups, more people carry the genes that make them vulnerable to stuttering than is the case in other ethnic groups. Perhaps the genetic pool, not cultural or linguistic features, has the strongest influence on the incidence of stuttering. Or, more likely, perhaps several causal factors contribute, interacting with each others in ways we do not yet understand.

The only way we will ever know if the incidence of stuttering varies across languages or countries is to do international, collaborative studies where the same rigourous methods are used in everywhere THEN we can propose explanations for the similar or differing rates of stuttering in different languages and/or countries.

Mystery 3: Does speaking more than one language increase the risk of a child stuttering?

Many people think it does. For the general public, it seems logical. Sometimes people reason this way: 1) Speaking two languages is hard. 2) For children who stutter, speaking is hard. 3) Therefore, children who stutter (or those who are at risk of developing stuttering, because of a known family history) should not be expected to learn two languages.

In the research on stuttering, the Demands and Capacities model of stuttering seems to apply here. But is it "hard" to learn two languages or is this a myth? If learning two languages as a young child is neurologically or cognitively strenuous, why is it that tens of millions of children do so successfully? Are their brains slightly stressed if they have to sort out two languages during the best language-learning years in childhood? Most of the research on bilingualism says "no". However, bilingualism is seen as something very positive by most people who do research on it, and most studies are designed to detect advantages and not problems associated with bilingualism. Also, this research is based on children with no speech or language problems. In children with a genetic vulnerability to stuttering, is learning two sets of words and grammar rules, and two sets of speech sounds harder than it is for children without this vulnerability? If learning two languages as a child is much harder than learning one, is it all potentially bilingual children or only a sub-group of those who might be at risk for stuttering?

How should we interpret the recent and somewhat controversial study by Howell, Davis and Williams (2009) that found a higher incidence of stuttering in children if they began learning English (the language of their new country? before age 5? Were there other reasons for the finding that children who learned English before starting school were more likely to stutter than those who reportedly began learning English when they began school in London, England? (See Packman, et al, 2009 letter to the editor and Howell et al.'s reply.)

There are four other, older studies that have led some people to conclude that bilingualism is too great a strain for children who stutter. In each case, these studies have serious flaws that make it impossible to draw any conclusions from them. Travis, Johnson and Shover (1937) asked people with no training in communication disorders (such as priests and steel company personnel directors) to talk to young children and classify them as stuttering or not stuttering based on one interview. Stern (1948) interviewed children if their parents reported that they stuttered. In both these studies, we have little information about the type of speech sample obtained, how long it was or how the disfluencies were counted. Applying current standards to these studies, they would not be accepted for publication.

Dale (1977) reported that four Cuban-American teenagers reported feeling that being made to speak their weaker language made them more disfluent. Most bilinguals have a stronger and a weaker language. For these teens, their first language - Spanish - was their weaker language, since so much of their lives at school and with friends took place in English, their second language. This study "blames" bilingualism. But we have no information about real disfluency rates across different situations, and Dale does not distinguish between normal disfluencies and tense, stuttered disfluencies. There are studies showing that, in adults, the memory load of speaking in their weaker language may lead to a higher number of normal disfluencies (ums, uh's, revisionsŠ) in the weaker language than in the preferred language (e.g., Fehringer & Fry, 2007). Perhaps that is all that was happening in this study. Dale offers no data to support the notion that any of the four adolescents, in fact, stuttered.

Karniol (1992) described how stuttering appeared to increase and decrease in a young boy whose environment included exposure to various levels of English, Hebrew, and Hungarian during an extremely tense time that included a war going on around him. With the information provided, we cannot tell what his real level of exposure to each language was (siblings, friends, parents etc.) and whether his parents' attempts to expose him to only Hebrew had any impact on what is described as a recovery from stuttering. The parents' diaries cover a period of approximately one year (age 2 to age 3) when the boy was in the age group where the chances of spontaneous recovery from stuttering are very, very high.

There is (still) no clinical research to support the strategy of removing one language from a child's environment. Recent reviews of the literature do not find support for doing this routinely for all children (e.g., Bernstein Ratner, 2004; Roberts & Shenker, 2007; Van Borsel, Maes & Foulon, 2001). Some clinicians do this, however, if they work in a Demands and Capacities framework OR if the child also has delayed language and/or problems learning the speech sounds of his/her language. Until there is solid evidence on the impact of bilingualism in young children (i.e. a series of studies, done by different authors, ideally on different types of speakers and different pairs of languages), each clinician is left to try a particular strategy and assess its impact on a case by case basis.

Mystery 4: Do some bilingual people stutter in only one language?

As of 2010, I am still not aware of any documented case of this occurring. Like the Loch Ness monster, there are reported sightings from time to time, but no real proof that this is possible. In my years working with bilingual adults who stutter, I never assessed or treated a case of "unilingual stuttering". (Note: if you know of someone who stutters in only one of their two (or more) languages, I would be interested in exploring this with you. Just because there are no documented cases does not mean that it never occurs !)

Nonetheless, given the roles of genetics and motor processes in stuttering, it is highly unlikely that someone would stutter in only one of their languages. Van Riper (1971) cites second- and third-hand reports of two people reported to stutter in only one of their two languages, but offers no data. Howell, Davis and Williams (2009) report that 2 of the 38 children in their bilingual group stuttered in only one language, but there is no supporting data about the rates of disfluencies, or the level of proficiency in each language and the children were not assessed using a range of speaking tasks.

Roberts and Shenker (2007, Table 1) outlined the steps that would be required to show that someone with a working knowledge of two languages stutters in only one language. Sometimes, when someone appears to stutter in only one language it could be for one or more of the following reasons:

  1. They view using the L2 as playing a role. The stuttering is "there" but temporarily masked by the well-known phenomenon that allows actors who stutter to speak fluently on stage.
  2. They have such a low level of ability in L2 that most utterances are one or two words or learned by heart. They are not yet speakers of that language. They are beginning to learn it.
  3. Through word substitutions, language switching and other strategies, they manage to mask the existing stuttering (reading aloud would reveal the stuttering in these cases).
  4. The speech sample obtained was too short to accurately reflect their real speaking patterns. Stuttering varies from day to day, and even within a single day. This is why all the textbooks on stuttering recommend obtaining more than one speech sample AND using different types of tasks such as free play, speaking with a parent, speaking with an unfamiliar person, making phone calls, reading aloud, etc.

There are more and more studies about stuttering in different languages and some studies (and soon, a new book edited by Howell and Van Borsel) that focus specifically on stuttering in bilingual speakers. This is a very welcome change. Ten or fifteen years ago, there was little awareness that bilingual stuttering was a topic that needed exploring. Perhaps in a future ISAD forum, there will be articles about the answers to the questions raised in this one.


Bernstein Ratner, N. (2004). Fluency and stuttering in bilingual children. In B.Goldstein (Ed.) Bilingual language development and disorders in Spanish-English speakers (pp. 287-308). Baltimore: Paul H. Brookes.

Bhatia, T.K., & Ritchie, W.C. (2006). Introduction. In T.K. Bhatia & W.C. Ritchie (Eds.) The Handbook of Bilingualism (pp. 1-2), Oxford: Blackwell Publishing.

Dale, P. (1977). Factors relating to disfluent speech in bilingual Cuban-American adolescents. Journal of Fluency Disorders, 2, 311-314.

Fehringer, C., & Fry, C. (2007). Hesitation phenomena in the language production of bilingual speakers. Folia Linguistica, 41, 37-72.

Howell, P., Davis, S. Williams,R. ( 2009). The effects of bilingualism on stuttering during late childhood. Archives of Disease in Childhood, 94, 42-46

Karniol, R. (1992). Stuttering out of bilingualism. First Language, 12, 255-283.

Packman,A., Onslow, M., Reilly, S. et al. (2009). Stuttering and bilingualism. Archives of Disease in Childhood, 94, 248. (a letter to the editor re the Howell, Davis and Williams study)

Roberts, P.M. & Shenker, R.C. (2007). Assessment and treatment of stuttering in bilingual speakers. In R.F. Curlee & E.G. Conture (Eds). Stuttering and related disorders of fluency 3rd edition (pp. 183-209). New York: Thieme Medical Publishers.

Stern, E. (1948). A preliminary study of bilingualism and stuttering in four Johannesburg schools. Journal of Logopedics, 1, 15-25.

Travis, L.E., Johnson, W., & Shover, J. (1937). The relation of bilingualism to stuttering: a survey in the east Chicago, Indiana, Schools. Journal of Speech Disorders, 12, 185-189.

Van Borsel, J. Maes, E., & Foulon, S. (2001). Stuttering and bilingualism: A review. Journal of Fluency Disorders, 26, 179-205.

You can post Questions/comments about the above paper to the author before October 22, 2010.

SUBMITTED: September 19, 2010
Translate this page into your language

Return to the opening page of the conference