Stuttering and bilingualism: A reply to Packman, Onslow, Reilly, Attansio & Shenker
Packman et al. responded to Howell et al.’s article (doi 10.1136/adc.2007.134114) on bilingualism and stuttering. We showed that speaking two languages in the preschool years increased the chances of children starting to stutter and decreased the likelihood of recovery from stuttering relative to speakers who learned English when they started school.
We do not consider, as Packman et al. suggest, that studying the general population of bilinguals would benefit the study of how speaking a second language affects stuttering. First, statistically speaking, demonstrating that bilingualism has no effect on stuttering, as the authors wish to do, is impossible. Also, investigating the general population of bilinguals is a flawed approach as bilingualism is massively heterogeneous within and between societies (including the US, Australia and Canada, where the authors work). Consequently, stringent controls (not biased sampling) have to be imposed in studies on bilingualism and stuttering. The controls we imposed excluded children who were brought up bilingual for social, educational or financial advantage. This reduced the heterogeneity in the bilingual sample. We compared bilinguals who had to use an additional language in the home with a matched sample who only spoke a language other than English in the home. The two groups of children selected for our study were directly comparable with respect to socio-economic status etc., and provided an appropriate sample on which to test the hypothesis of whether deferring learning English as a second language affected onset of stuttering and whether stuttering recovered. The same controls are appropriate for selecting similar groups for study across countries (unlike a sample from the general population of bilinguals at large).
They go on to state that our cohort is biased, and refer to a clinical study that followed up children from about age 3 to 8. 1 They say that the generally accepted age of stuttering onset is 3 years. However, the authoritative handbook in the area notes that the onset of developmental stuttering can occur at any age up to 13 years. 2 Consequently, studies that cease examination at age 8 bias estimates of age of onset to lower ages by excluding cases where onset occurs after age 8. They also incorrectly state that the gender ratio is 1:1 (there was a higher ratio of males/females in our study). All studies we are aware of (including the one they cited 1) show more males stutter than females. For example, the classic study by Andrews and Harris 3 from 2 to teenage reported a ratio of 2.4:1. A gender ratio of 1:1 suggests they are including children who do not stutter as stutterers (they have many false positives) and there is about an equal chance of misdiagnosis for each gender. The reasons for such misdiagnoses could be because they 4 use parental nominations for identifying cases and employ symptoms for diagnosing stuttering (whole word repetitions) that are excluded by several authorities. 5 6 7 The inclusion of these would lead to fluent children being classed as stutterers.
Packman et al. suggest that the Lidcombe program of therapy may be appropriate to treat bilingual children who stutter. However, the study they cited excluded children who did not have “proficiency in English” and there is no translation of the procedure into several of the languages spoken by our participants. In addition to the fact that they may be treating children who do not stutter, the statistical treatment in the study they cite has been faulted and long-term follow-up indicates that some “children [treated by Lidcombe] may start to stutter again”. 8
They state that an epidemiological study is appropriate to determine whether there is a relationship between stuttering and bilingualism. We have warned about the problems of grouping together heterogeneous bilinguals for study, as is necessary in epidemiological work. They overlook the fact that our work reported on the risk of starting to stutter and chance of recovery, using longitudinal data through to teenage. They suggest the onset age is about three years and that most recovery occurs in the first three years after onset. Conceivably, an epidemiological study at ages less than 8 could provide information about risk factors for onset of stuttering. However, it would not be informative about recovery since 50% of cases still stuttering at age 8 will recover 9 10, and recovery is not resolved until teenage. 3 9 10 Although epidemiological work does not require a longitudinal dimension, this is necessary when studying recovery, where results on the same speakers are needed in order to avoid biases. As in our study, they will need to use standardized instruments to establish persistence and recovery at teenage. If they intend to asses the impact bilingualism has on school performance, they will again need to extend examination to teenage.
In their last paragraph, they state that 4% stutter-like dysfluencies suggests our children would still be stuttering. Frequencies around this value have been used to differentiate young children who stutter from normally fluent speaking peers for English 1 and Dutch. 11 As discussed earlier, relatively high rates of supposed stuttering events may be due to inclusion of monosyllabic whole word repetitions, which are a common feature in fluent children’s and adults’ speech. 12
In summary, we consider our conclusions to be valid. Our experimental groups were matched appropriately and thoroughly documented with respect to language background and assessment for stuttering. The authors of the letter make statements that are contradicted in the literature. There are major faults in the way the authors of the letter propose to study the relationship between bilingualism and stuttering.
Acknowledgement This work was supported by grant 072639 from the Wellcome Trust to Peter Howell. Address Correspondence to: Peter Howell, Division of Psychology and Language Sciences, University College London, Gower Street, London WC1E 6BT, England. Email: firstname.lastname@example.org
1Yairi E., Ambrose NG. Early childhood stuttering. Austin, TX: Pro- Ed, 2005.
2Bloodstein O., Bernstein Ratner N. A handbook on stuttering, 6th ed. Clifton Park, NY: Thomson, 2007.
3Andrews G., Harris M. The syndrome of stuttering. Clinics in Developmental medicine (No 17). London: Heinemann, 1964.
4Reilly S., et al. Predicting stuttering onset by the age of 3 years. Pediatrics, 2009;123:270-277.
5Conture EG. Stuttering. 2nd ed. Englewood-Cliffs: Prentice Hall; 1990.
6Ryan B. A longitudinal study of articulation, language, rate, and fluency of 22 preschool children who stutter. Journal of Fluency Disorders, 2001;26:107-127.
7Wingate ME. Foundations of stuttering. San Diego: Academic; 2002.
8Packman A., Kuhn L. Looking at stuttering through the lens of complexity. International Journal of Speech-Language Pathology, 2009;11:77– 82.
9Fritzell B. 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, 186-187. Basel: Karger, 1976.
10Howell P, Davis S, Williams R. Late childhood stuttering. Journal of Speech Language and Hearing Research, 2008;51:669-687.
11 Boey et al., Characteristics of stuttering-like disfluencies in Dutch speaking older children, adolescents and adults. Journal of Fluency Disorders, 2007;32:310-329..
12 Levelt W. Speaking: From intention to articulation. Cambridge, MA: Bradford Books, 1989.
Division of Psychology and Language Sciences, University College London, United Kingdom
Department of Language and Communication Science, City University London, United Kingdom