Stuttering, Dichotic Listening, and Cerebral Dominance John Paul
Brady, MD,
Janet
Berson, PhD
\s=b\ Fully right-sided stutterers (35) and fully right-sided nonstutterers (35) had a dichotic listening task to test hypotheses that stutterers have incomplete cerebral lateralization or re-
versed lateralization of speech function, or both. An assumption of the procedure is that a right-ear preference indicates left-cerebral dominance for speech. Six stutterers and no nonstutterers showed a reversal, ie, a left-ear preference. As a group, the remaining stutterers who showed no such reversal were the same as nonstutterers in the magnitude of the right-ear preference. This suggests that a subset of stutterers may have an anomaly in the lateralization of speech functions. A nonsignificant tendency emerged for stutterers to show smaller between-ear differences on the test, consistent with the hypothesis that stutterers have less or incomplete lateralization of speech function than nonstutterers. (Arch Gen Psychiatry 32:1449-1452, 1975)
and pathogenesis of stuttering are large¬ unknown. The notion that stuttering is related in some way to cerebral dominance for speech dates back at least to the work and writings of Lee Travis.1 There were a number of investigations in the 1930s and early 1940s dealing with this issue. The central theme of these studies was as follows: A high percentage of stutterers are lefthanded or ambidextrous. Handedness is a reliable index of the degree and direction of lateralization of speech fune-
Thely
cause
tions in the cerebral hemisphere: right-handed persons have left-cerebral dominance, left-handed persons largely right-cerebral dominance, and ambidextrous persons "in¬ complete" or mixed-cerebral dominance for speech. Hence left-handed or ambidextrous persons may be predisposed to stuttering because incomplete lateralization of speech function may lead to confusion in producing speech sounds through the peripheral, midline, and bilaterally inner¬ vated speech muscles. A related notion was that persons who were "naturally" left-handed but forced to write with the right hand at an early age are especially susceptible to the disorder of stuttering. However, the research that emanated from these hypotheses was inconclusive. Thus, for example, Bryngelson and Rutherford2·3 reported that a higher percentage of stutterers are ambidextrous than are nonstutterers, whereas Daniels4 and Heltman5 failed to find such a relationship. There are a number of possible reasons for these inconsistencies. 1. There are a variety of measurements of handedness and the decision that a subject is right-handed or ambi¬ dextrous will depend on the criteria used. 2. "Handedness" is only one index of "sidedness" and the latter may be more important. (Thus, one might use the right hand preferentially but use the left foot or left eye
or sidedness may change during life due accident to one part of the body or other experiences. In brief, differences in criteria for determining hand¬ edness or ambidexterity may account for different conclu¬ sions reached by different investigators, especially if there is in fact only a weak, statistical relationship between ce¬ rebral dominance for speech and the development of stut¬
to
Accepted for publication Oct 26, 1974. From the Department of Psychiatry, University of Pennsylvania, the Eastern Pennsylvania Psychiatric Institute, Philadelphia (Dr. Brady); and the Department of Psychology, Rutgers University, New Brunswick, NJ (Dr. Berson). Reprint requests to the Hospital of the University of Pennsylvania, 1141 Gates Bldg, Philadelphia, PA 19104 (Dr. Brady).
preferentially.)
3. Handedness an
tering.
Downloaded From: http://archpsyc.jamanetwork.com/pdfaccess.ashx?url=/data/journals/psych/12266/ by a University of California - San Diego User on 01/13/2017
Dichotic
Listening
and Cerebral Dominance
Newer and more reliable methods of determining later¬ alization of speech functions provide a better opportunity to examine the relationship of cerebral dominance to stut¬ tering. A dichotic listening task is especially promising for this purpose. This procedure was originally developed by Broadbent* to detect subtle neurological damage and developed additionally by Kimura7 and others as a means of identifying lateralization of speech function in normal subjects. In brief, the procedure consists of delivering dif¬ ferent speech sounds (eg, digits) to the two ears simulta¬ neously. Virtually all right-handed individuals show bet¬ ter recognition and recall of the verbal stimuli to the right ear. Presumably this is related to the fact that a larger number of fibers project to the centralateral than to the ipsilateral auditory receiving area in the cerebral hemi¬ spheres and most right-handed individuals have left-cere¬ bral dominance for speech. (The mechanism of the rightear preference is discussed at length by Kimura.7 Kimura provided some direct evidence for the validity of a dichotic listening task as a way of determining cerebral domi¬ nance for speech by means of an amobarbital (Amytal) so¬ dium procedure by Wada and Rasmussen.8 (In the latter test, amobarbital is injected on separate occasions into the right and left carotid arteries and note is made of the de¬ gree and time course of the disruption of speech function.) Kimura found that 13 subjects with rio7t£-cerebral domi¬ nance for speech by the amobarbital procedure showed a left-ear preference in a verbal dichotic listening task. The reverse was found with 103 subjects with left-cerebral dominance for speech. Furthermore, the presence of uni¬ lateral cerebral damage could not account for these rela¬ tionships. The relationship between cerebral dominance and handedness in these subjects was less consistent. Thus, it appears that a dichotic listening task is indeed a more reliable way of ascertaining lateralization of speech function. A dichotic listening task seems well suited for exam¬ ining cerebral dominance hypothesis in stuttering. It pro¬ vides a quantitative measure of degree of cerebral domi¬ nance since one can examine the difference between the right- and left-ear scores on a dichotic listening task. This may prove useful since some hypotheses about stuttering posit incomplete or mixed cerebral dominance as a corre¬ late of stuttering. The validity of a dichotic listening test as a measure of degree of cerebral dominance has not been established. However, Curry9 has some suggestive evi¬ dence in this regard. There is neurosurgical and other evi¬ dence that left-handed subjects have greater hemispheric equipotentiality.1" Curry found a trend (not statistically significant at the .05 level) for left-handed persons to have smaller between-ear differences than right-handers on a verbal dichotic listening procedure. Dichotic
Listening
and
Stuttering
Curry and Gregory11 have published a report on dichotic listening in stutterers. They subjected 20 adult stutterers and 20 well-matched nonstutterers to four listening tasks: a monotonie (ie, to one ear only) word memory test, two dichotic tasks involving nonspeech sounds (one entailed
recognition of environmental sounds and the other pitch discrimination), and a dichotic task involving word recognition. Only the last test differentiated between stutterers and nonstutterers, so only this will be com¬ the
mented on. Their dichotic task involved the recognition and recall of pairs of consonant-vowel-consonant words of high fa¬ miliarity, presented in groups of six pairs with 0.5 seconds separating each pair. The task was for the subject to recall as many of the 12 words presented after each group of sixpaired words. The test consisted of 12 such groups. The data were analyzed by noting the number of words pre¬ sented to each ear that were correctly identified and re¬ called. Thus, they obtained a "right-ear score" and a "leftear score" for each subject and also calculated the differ¬ ence between the two scores for each subject. Fifteen of the 20 nonstutterers (75%) showed a rightear preference, ie, recalled more words presented to the right ear than to the left. Although this is the predicted direction, it is not clear why five nonstutterers actually showed a left-ear preference. They state that all 40 sub¬ jects were right-handed, which was apparently the only criterion used for sidedness. Of the 20 stutterers, only nine showed a right-ear preference (45%). They report this directional difference between stutterers and controls as significant at the .10 level (chi-square test for independent
samples). Curry and Gregory also subjected their data to another kind of analysis that bears of the hypothesis of "mixed" cerebral dominance in stutterers. They calculated the size of the absolute between-ears difference score for each sub¬
ject, ie, difference scores without regard for the direction of the ear superiority. They found that the mean absolute difference for the control group was more than twice that of the group of stutterers. This difference in means was significant at the .005 level (one-tailed t test). Curry and Gregory are admirably cautious in inter¬ preting the results of their study. The data do suggest dif¬ ferences between stutterers and nonstutterers in the neurophysiological organization of speech function, which is consistent with the hypothesis mentioned earlier that stutterers have incomplete or mixed cerebral dominance for speech. It is possible that both of the differences re¬ ported by Curry and Gregory could be accounted for by differences in degree of sidedness of the stutterers and nonstutterers.
Recently, Quinn12 reported on performance of 60 stut¬ terers and 60 nonstutterers on a dichotic word test that was very similar to that used by Curry and Gregory and
that partially confirmed the latter study. Again, the sub¬ jects were "right-handed" by their own report. Confirm¬ ing the Curry and Gregory study, he reported12 "rever¬ sals," ie, a left-ear preference, among the stutterers but only two among the nonstutterers. However, they did not find a significant difference between the two samples on the magnitude of the between-ear differences for each subject. METHODS
The present study differs from those of Curry and Gregory" and Quinn12 in several regards. First, a dichotic listening task was devised that generates a right-ear preference in virtually all
Downloaded From: http://archpsyc.jamanetwork.com/pdfaccess.ashx?url=/data/journals/psych/12266/ by a University of California - San Diego User on 01/13/2017
right-sided nonstutterers.
This is in keeping with the neurosurgical literature and results of the sodium amytal test that show that a very high percentage of normal (ie, not brain damaged) persons who are right-sided have left-brain speech representation.13·" Second, the dichotic listening test requires the subject to select, by circling with a pencil, between two written nonsense syllables rather than vocally report what he hears. This obviates the prob¬ lem that the stutterer's performance on the test may be influ¬ enced by his difficulty in speaking out loud in the test situation. Third, a number of criteria of sidedness were used to ascertain that the subjects were unequivocally right-sided individuals.
Subjects Thirty-five
adult stutterers served as experimental subjects. was obtained after the nature of the experi¬ mental procedure was explained.) All scored at least 3 on the seven-point Scale for Rating Severity of Stuttering developed by Johnson and others.13 The nonstuttering control group consisted of 35 adults who never considered themselves to be stutterers, were never told that they had a stuttering problem, and who scored zero on the severity scale of Johnson et al. All subjects had normal hearing in both ears, no history of brain damage, and were "fully right-sided," as defined below. The two samples did not differ sig¬ nificantly in sex, age, or educational level (Table 1).
(Informed
consent
Determination of Sidedness Three aspects of sidedness were ascertained: hand, leg, and eye. Thus, the subject was asked which hand he uses to write, throw a ball, use hand tools (hammer, screwdriver, or sewing needle), and brush his teeth. Footedness was determined by the foot used to kick a ball and operate a machine (eg, sewing machine). Finally, he was asked which eye he uses with a monocular optical instru¬ ment (microscope or telescope). Questions appropriate for detect¬ ing any congenital or acquired physical differences between the two sides were asked since any gross disparity might also account for a side preference (eg, differences in visual acuity of the right and left eyes, muscular weakness of an arm or hand). The subject was regarded as "fully right-sided" only if he indicated the pre¬ ferred use of the right side on all the above except where injury or defect could account for an exception. However, only one such ex¬ ception was allowed for each subject in this determination.
The Dichotic
Listening
labiés that differ only in the initial consonant (eg, KEP vs GEP). Great pains were taken in the development of this tape to syn¬ chronize the onsets of the paired syllables (within a few milli¬ seconds). Care was taken also in phonetic counterbalancing in the pairing of stimuli. Thus, each of six initial stop consonants (B, D, G, K, P, and T) is paired with each other an equal number of times. Also, each pair is presented twice, differing in which syl¬ lable is presented to which ear. There are a total of 60 such pairs. The use of this tape by its developers demonstrated a large and consistent right-ear preference.18 The verbal stimuli were presented dichotically to the subject by means of earphones connected with a two-track stereophonic tape recorder. Rather than have subjects vocally report what they hear, which is commonly done in this procedure, subjects were in¬ structed to choose between the two syllables that constitute each pair written out on a response sheet (Table 2). Each pair is sepa¬ rated by two seconds except that there is a 12-second break after the 15th, 30th, and 45th pairs. The exact instructions to the subject were as follows: "Check or circle the syllable of each pair which is closest to what you hear. If you are not sure do the best you can, guessing if necessary." Thus, the subject was in a forced-choice situation. After the subject listened to the entire tape once, the earphones were removed, the tape rewound, and the earphones re¬ placed but reversed (without the subject's knowledge). The entire task was then repeated. On this second trial, of course, syllables previously presented to the right ear were presented to the left ear and vice versa. This was to balance for any bias in any part of the entire system (ie, any inherent imbalance in the loudness of one sound of any pair, undetected differences in the ear-phones). This dichotic listening task yields data that are easily scored. One can determine the number of occasions out of 120 (two trials of 60 stimuli each) in which the subject selected the signal that was presented to the right ear (the right-ear score) and the left
Table
1.—Summary of Subjects Stutterers 35
Sex Women Men
Age, yr Range Mean Formal education,
Task
The study used the magnetic tape recording developed by Shankweiler and Studdert-Kennedy.16,; This tape presents simulta¬ neously to the two ears pairs of consonant-vowel-consonant syl-
Mean,
yr
Nonstutterers 35
29
28
17-46 25.5
17-52 26.5
15.0
15.2
Severity of stuttering, Mean*
4.7
Seven-point scale of Johnson et al (1963) .,s Table
Dip Bep Gip Gep Gip Kip Pep Dep Kep Dep Bep Pip Bip Tep
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.
Gip Gep Kip Dep Pip Pip Gep Bep Tep Kep Tep Tip Pip Dep
16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29.
2.—Response Sheet for Dichotic Listening Task* Tip Kep Pep Bip Gip Tip Dip Bip Dep Kep Kip Pep Tep Gep
Gip
Gep Kep Kip Bip Bip Tip Dip Pep Bep Dip Bep Pep Tep
31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44.
Tip Gep Pep Gip Gep Gip Pep Dip Gip Bip Pep Pip Pip Tep
Kip Tep Bep Kip Dep Pip Gep Tip Bip Kip Kep Tip Dip Dep
46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59.
Tip Kep Bip Bep Dep Kep Dep Bip Kip Dep Kep Kip Bep Tep
Gip Gep Pip Tep Kep Tep Bep Dip Pip Pep Bep Dip Gep Pep
15._Pip_Dip_30_Tip_Kip_45._Tip_Bip_60;_Dip_Gip There is
a
12-second break after the 15th, 30th, and 45th pairs.
Downloaded From: http://archpsyc.jamanetwork.com/pdfaccess.ashx?url=/data/journals/psych/12266/ by a University of California - San Diego User on 01/13/2017
Table
3.—Summary of Experimental Stutterers
(N No. of reversals
Right-ear score, mean Total sample Right-ear score, mean Excluding reversals Absolute between-ear difference score, mean
=
35)
Note
Data Nonstutterers
(N
=
35)
6*
0*
68.4f
73.5f
72.5
73.5
24.7
26.9
Difference significant at .012 level. t Difference significant at .025 level. Other differences between *
groups not ear
(Table 3) that the
mean
right-ear
score
was
nonstutterers than for stutterers. Although not large, this difference also is significant (P < .025 by a one-tailed Student t test for independent samples). Both
greater for
statistically significant.
(the left-ear score). The absolute value of the difference be¬
tween these scores, the absolute between-ears difference score, may reflect the degree of cerebral lateralization of speech func¬ tion in the subject. Performed and scored in this manner, normal,
fully right-sided
individuals almost all show a right-ear prefer¬ About 40% of left-sided, normal individuals show a left-ear preference. These findings are consistent with the neurosurgical literature on the relationship between sidedness and cerebral dominance for speech." ence.
RESULTS
The experimental results are given in Table 1. The find¬ that none of the 35 nonstutterers showed a reversal, ie, none showed a left-ear preference, was anticipated, since almost all normal, right-sided individuals have leftcerebral dominance for speech. In contrast, six of the 35 stutterers (17%) showed a left-ear preference. This differ¬ ence is statistically significant (P=.012 by the Fisher Exact Probability Test19). The six subjects who showed a reversal did not differ clearly from the rest of the sample of stutterers in age, educational level, severity of stutter¬ ing, proportion of men, or in any other manner apparent to the authors. Both the Curry and Gregory11 and the Quinn12 studies reported reversals in nonstutterers as well as stutterers but with more in the latter group. Thus, the present study is consistent with their finding on this mea¬ sure but demonstrated a sharper division between the two samples. This difference may be related to the more strin¬ gent criteria for sidedness used in the present study or differences in the dichotic listening task itself, or both.
the studies of Curry and Gregory and of Quinn reported similar results. In the present study, the mean right-ear scores for stutterers and nonstutterers are nearly identi¬ cal if the stutterers who showed a reversal are omitted from the comparison (Table 3). This suggests that the dif¬ ference between stutterers and nonstutterers in this mea¬ sure is due to the subset of stutterers who showed a leftear preference in the dichotic listening task. Quinn12 found a similar result when he excluded subjects who showed a left-ear preference from his calculations. The absolute between-ear difference scores tended to be smaller for stutterers than nonstutterers. However, the mean difference between the two samples is small and not
statistically significant (Table 3), whether or not the sub¬ jects with a left-ear preference are excluded from the cal¬ culations. Recall that Curry and Gregory did find that stutterers had statistically significant smaller betweenear differences. The Quinn study, like the present one, found a difference in the same (predicted) direction but it was not statistically significant.
ing
CONCLUSIONS
keeping with the recent research of other investiga¬ stutterers in this study did show a less strong right-ear preference and smaller between-ear differences than nonstutterers in a dichotic listening task. However, these differences were small in magnitude and cast lim¬ ited light on the role of cerebral dominance in most stutter¬ ers. The most striking finding of this investigation is that six (17%) stutterers did show a left-ear preference while no nonstuttering control subjects evidenced such a rever¬ In
tors, the
sal. This suggests a subset of stutterers in whom an anom¬ aly of cerebral lateralization of speech functions may be important. A promising area of research entails the inten¬ sive study of this special group and its comparison with stutterers who do not show this phenomenon. This investigation was supported in part by Pub'ic Health Service search grant 24507 from the National Institute of Mental Health.
re¬
References 1. Travis L: Speech Pathology. New York, Appleton-Century-Crofts Inc, 1931. 2. Bryngelson B: Sidedness as an etiological factor in stuttering. J Gen Psychol 47:204-217, 1935. 3. Bryngelson B, Rutherford B: A comparative study of laterality of stutterers and non-stutterers. J Speech Disorders 2:15-16, 1937. 4. Daniels EM: An analysis of the relation between handedness and stuttering with special reference to the Orton-Travis theory of cerebral dominance. J Speech Disorders 5:309-326, 1940. 5. Heltman HJ: Contradictory evidence in handedness and stuttering. J Speech Disorders 5:327-331, 1940. 6. Broadbent DE: The role of auditory localization in attention and memory span. J Exp Psychol 47:191-196, 1954. 7. Kimura D: Functional asymmetry of the brain in dichotic listening. Cortex 3:163-178, 1967. 8. Wada J, Rasmussen T: Intracarotid injection of sodium amytal for the lateralization of cerebral speech dominance. JNeurosurg 17:266-282,1960. 9. Curry FKW: A comparison of left-handed and right-handed subjects on verbal and non-verbal dichotic listening tasks. Cortex 3:343-352, 1967. 10. Subirana A: The prognosis of aphasia in relation to cerebral dominance and handedness. Brain 81:415-425, 1958.
11. Curry FKW, Gregory HH: The performance of stutterers on dichotic listening tasks thought to reflect cerebral dominance. J Speech Hear Res 12:73-82, 1969. 12. Quinn PT: Stuttering, cerebral dominance and the dichotic word test. Med J Aust 2:639-643, 1972. 13. Satz P, Achenbach K, Pattishall E, et al: Order of report, ear asymmetry and handedness in dichotic listening. Cortex 1:377-396, 1965. 14. Rossi GF, Rosadini G: Experimental analysis of cerebral dominance in man, in CH Millikan, FL Darley (eds): Brain Mechanisms Underlying Speech and Language. New York, Grune & Stratton Inc, 1967, pp 167-184. 15. Johnson W. Darley FL, Spriesterbach DC: Diagnostic Methods in Speech Pathology. New York, Harper & Row Publishers Inc, 1963. 16. Shankweiler D, Studdert-Kennedy M: An analysis of perceptual confusion in identification of dichotically presented CVC syllables. J Acoust Soc Am 41:1581, 1967. 17. Shankweiler D, Studdert-Kennedy M: Identification of consonants and vowels presented to left and right ears. Q J Exp Psychol 19:59-63,1967. 18. Liberman AM, Cooper FS, Shankweiler DP, et al: Perception of the speech code. Psychol Rev 74:431-461, 1967. 19. Siegel S: Nonparametric Statistics. New York, McGraw-Hill Book Co
Inc, 1956, pp 96-101.
Downloaded From: http://archpsyc.jamanetwork.com/pdfaccess.ashx?url=/data/journals/psych/12266/ by a University of California - San Diego User on 01/13/2017
Brady, MD,
Janet
Berson, PhD
\s=b\ Fully right-sided stutterers (35) and fully right-sided nonstutterers (35) had a dichotic listening task to test hypotheses that stutterers have incomplete cerebral lateralization or re-
versed lateralization of speech function, or both. An assumption of the procedure is that a right-ear preference indicates left-cerebral dominance for speech. Six stutterers and no nonstutterers showed a reversal, ie, a left-ear preference. As a group, the remaining stutterers who showed no such reversal were the same as nonstutterers in the magnitude of the right-ear preference. This suggests that a subset of stutterers may have an anomaly in the lateralization of speech functions. A nonsignificant tendency emerged for stutterers to show smaller between-ear differences on the test, consistent with the hypothesis that stutterers have less or incomplete lateralization of speech function than nonstutterers. (Arch Gen Psychiatry 32:1449-1452, 1975)
and pathogenesis of stuttering are large¬ unknown. The notion that stuttering is related in some way to cerebral dominance for speech dates back at least to the work and writings of Lee Travis.1 There were a number of investigations in the 1930s and early 1940s dealing with this issue. The central theme of these studies was as follows: A high percentage of stutterers are lefthanded or ambidextrous. Handedness is a reliable index of the degree and direction of lateralization of speech fune-
Thely
cause
tions in the cerebral hemisphere: right-handed persons have left-cerebral dominance, left-handed persons largely right-cerebral dominance, and ambidextrous persons "in¬ complete" or mixed-cerebral dominance for speech. Hence left-handed or ambidextrous persons may be predisposed to stuttering because incomplete lateralization of speech function may lead to confusion in producing speech sounds through the peripheral, midline, and bilaterally inner¬ vated speech muscles. A related notion was that persons who were "naturally" left-handed but forced to write with the right hand at an early age are especially susceptible to the disorder of stuttering. However, the research that emanated from these hypotheses was inconclusive. Thus, for example, Bryngelson and Rutherford2·3 reported that a higher percentage of stutterers are ambidextrous than are nonstutterers, whereas Daniels4 and Heltman5 failed to find such a relationship. There are a number of possible reasons for these inconsistencies. 1. There are a variety of measurements of handedness and the decision that a subject is right-handed or ambi¬ dextrous will depend on the criteria used. 2. "Handedness" is only one index of "sidedness" and the latter may be more important. (Thus, one might use the right hand preferentially but use the left foot or left eye
or sidedness may change during life due accident to one part of the body or other experiences. In brief, differences in criteria for determining hand¬ edness or ambidexterity may account for different conclu¬ sions reached by different investigators, especially if there is in fact only a weak, statistical relationship between ce¬ rebral dominance for speech and the development of stut¬
to
Accepted for publication Oct 26, 1974. From the Department of Psychiatry, University of Pennsylvania, the Eastern Pennsylvania Psychiatric Institute, Philadelphia (Dr. Brady); and the Department of Psychology, Rutgers University, New Brunswick, NJ (Dr. Berson). Reprint requests to the Hospital of the University of Pennsylvania, 1141 Gates Bldg, Philadelphia, PA 19104 (Dr. Brady).
preferentially.)
3. Handedness an
tering.
Downloaded From: http://archpsyc.jamanetwork.com/pdfaccess.ashx?url=/data/journals/psych/12266/ by a University of California - San Diego User on 01/13/2017
Dichotic
Listening
and Cerebral Dominance
Newer and more reliable methods of determining later¬ alization of speech functions provide a better opportunity to examine the relationship of cerebral dominance to stut¬ tering. A dichotic listening task is especially promising for this purpose. This procedure was originally developed by Broadbent* to detect subtle neurological damage and developed additionally by Kimura7 and others as a means of identifying lateralization of speech function in normal subjects. In brief, the procedure consists of delivering dif¬ ferent speech sounds (eg, digits) to the two ears simulta¬ neously. Virtually all right-handed individuals show bet¬ ter recognition and recall of the verbal stimuli to the right ear. Presumably this is related to the fact that a larger number of fibers project to the centralateral than to the ipsilateral auditory receiving area in the cerebral hemi¬ spheres and most right-handed individuals have left-cere¬ bral dominance for speech. (The mechanism of the rightear preference is discussed at length by Kimura.7 Kimura provided some direct evidence for the validity of a dichotic listening task as a way of determining cerebral domi¬ nance for speech by means of an amobarbital (Amytal) so¬ dium procedure by Wada and Rasmussen.8 (In the latter test, amobarbital is injected on separate occasions into the right and left carotid arteries and note is made of the de¬ gree and time course of the disruption of speech function.) Kimura found that 13 subjects with rio7t£-cerebral domi¬ nance for speech by the amobarbital procedure showed a left-ear preference in a verbal dichotic listening task. The reverse was found with 103 subjects with left-cerebral dominance for speech. Furthermore, the presence of uni¬ lateral cerebral damage could not account for these rela¬ tionships. The relationship between cerebral dominance and handedness in these subjects was less consistent. Thus, it appears that a dichotic listening task is indeed a more reliable way of ascertaining lateralization of speech function. A dichotic listening task seems well suited for exam¬ ining cerebral dominance hypothesis in stuttering. It pro¬ vides a quantitative measure of degree of cerebral domi¬ nance since one can examine the difference between the right- and left-ear scores on a dichotic listening task. This may prove useful since some hypotheses about stuttering posit incomplete or mixed cerebral dominance as a corre¬ late of stuttering. The validity of a dichotic listening test as a measure of degree of cerebral dominance has not been established. However, Curry9 has some suggestive evi¬ dence in this regard. There is neurosurgical and other evi¬ dence that left-handed subjects have greater hemispheric equipotentiality.1" Curry found a trend (not statistically significant at the .05 level) for left-handed persons to have smaller between-ear differences than right-handers on a verbal dichotic listening procedure. Dichotic
Listening
and
Stuttering
Curry and Gregory11 have published a report on dichotic listening in stutterers. They subjected 20 adult stutterers and 20 well-matched nonstutterers to four listening tasks: a monotonie (ie, to one ear only) word memory test, two dichotic tasks involving nonspeech sounds (one entailed
recognition of environmental sounds and the other pitch discrimination), and a dichotic task involving word recognition. Only the last test differentiated between stutterers and nonstutterers, so only this will be com¬ the
mented on. Their dichotic task involved the recognition and recall of pairs of consonant-vowel-consonant words of high fa¬ miliarity, presented in groups of six pairs with 0.5 seconds separating each pair. The task was for the subject to recall as many of the 12 words presented after each group of sixpaired words. The test consisted of 12 such groups. The data were analyzed by noting the number of words pre¬ sented to each ear that were correctly identified and re¬ called. Thus, they obtained a "right-ear score" and a "leftear score" for each subject and also calculated the differ¬ ence between the two scores for each subject. Fifteen of the 20 nonstutterers (75%) showed a rightear preference, ie, recalled more words presented to the right ear than to the left. Although this is the predicted direction, it is not clear why five nonstutterers actually showed a left-ear preference. They state that all 40 sub¬ jects were right-handed, which was apparently the only criterion used for sidedness. Of the 20 stutterers, only nine showed a right-ear preference (45%). They report this directional difference between stutterers and controls as significant at the .10 level (chi-square test for independent
samples). Curry and Gregory also subjected their data to another kind of analysis that bears of the hypothesis of "mixed" cerebral dominance in stutterers. They calculated the size of the absolute between-ears difference score for each sub¬
ject, ie, difference scores without regard for the direction of the ear superiority. They found that the mean absolute difference for the control group was more than twice that of the group of stutterers. This difference in means was significant at the .005 level (one-tailed t test). Curry and Gregory are admirably cautious in inter¬ preting the results of their study. The data do suggest dif¬ ferences between stutterers and nonstutterers in the neurophysiological organization of speech function, which is consistent with the hypothesis mentioned earlier that stutterers have incomplete or mixed cerebral dominance for speech. It is possible that both of the differences re¬ ported by Curry and Gregory could be accounted for by differences in degree of sidedness of the stutterers and nonstutterers.
Recently, Quinn12 reported on performance of 60 stut¬ terers and 60 nonstutterers on a dichotic word test that was very similar to that used by Curry and Gregory and
that partially confirmed the latter study. Again, the sub¬ jects were "right-handed" by their own report. Confirm¬ ing the Curry and Gregory study, he reported12 "rever¬ sals," ie, a left-ear preference, among the stutterers but only two among the nonstutterers. However, they did not find a significant difference between the two samples on the magnitude of the between-ear differences for each subject. METHODS
The present study differs from those of Curry and Gregory" and Quinn12 in several regards. First, a dichotic listening task was devised that generates a right-ear preference in virtually all
Downloaded From: http://archpsyc.jamanetwork.com/pdfaccess.ashx?url=/data/journals/psych/12266/ by a University of California - San Diego User on 01/13/2017
right-sided nonstutterers.
This is in keeping with the neurosurgical literature and results of the sodium amytal test that show that a very high percentage of normal (ie, not brain damaged) persons who are right-sided have left-brain speech representation.13·" Second, the dichotic listening test requires the subject to select, by circling with a pencil, between two written nonsense syllables rather than vocally report what he hears. This obviates the prob¬ lem that the stutterer's performance on the test may be influ¬ enced by his difficulty in speaking out loud in the test situation. Third, a number of criteria of sidedness were used to ascertain that the subjects were unequivocally right-sided individuals.
Subjects Thirty-five
adult stutterers served as experimental subjects. was obtained after the nature of the experi¬ mental procedure was explained.) All scored at least 3 on the seven-point Scale for Rating Severity of Stuttering developed by Johnson and others.13 The nonstuttering control group consisted of 35 adults who never considered themselves to be stutterers, were never told that they had a stuttering problem, and who scored zero on the severity scale of Johnson et al. All subjects had normal hearing in both ears, no history of brain damage, and were "fully right-sided," as defined below. The two samples did not differ sig¬ nificantly in sex, age, or educational level (Table 1).
(Informed
consent
Determination of Sidedness Three aspects of sidedness were ascertained: hand, leg, and eye. Thus, the subject was asked which hand he uses to write, throw a ball, use hand tools (hammer, screwdriver, or sewing needle), and brush his teeth. Footedness was determined by the foot used to kick a ball and operate a machine (eg, sewing machine). Finally, he was asked which eye he uses with a monocular optical instru¬ ment (microscope or telescope). Questions appropriate for detect¬ ing any congenital or acquired physical differences between the two sides were asked since any gross disparity might also account for a side preference (eg, differences in visual acuity of the right and left eyes, muscular weakness of an arm or hand). The subject was regarded as "fully right-sided" only if he indicated the pre¬ ferred use of the right side on all the above except where injury or defect could account for an exception. However, only one such ex¬ ception was allowed for each subject in this determination.
The Dichotic
Listening
labiés that differ only in the initial consonant (eg, KEP vs GEP). Great pains were taken in the development of this tape to syn¬ chronize the onsets of the paired syllables (within a few milli¬ seconds). Care was taken also in phonetic counterbalancing in the pairing of stimuli. Thus, each of six initial stop consonants (B, D, G, K, P, and T) is paired with each other an equal number of times. Also, each pair is presented twice, differing in which syl¬ lable is presented to which ear. There are a total of 60 such pairs. The use of this tape by its developers demonstrated a large and consistent right-ear preference.18 The verbal stimuli were presented dichotically to the subject by means of earphones connected with a two-track stereophonic tape recorder. Rather than have subjects vocally report what they hear, which is commonly done in this procedure, subjects were in¬ structed to choose between the two syllables that constitute each pair written out on a response sheet (Table 2). Each pair is sepa¬ rated by two seconds except that there is a 12-second break after the 15th, 30th, and 45th pairs. The exact instructions to the subject were as follows: "Check or circle the syllable of each pair which is closest to what you hear. If you are not sure do the best you can, guessing if necessary." Thus, the subject was in a forced-choice situation. After the subject listened to the entire tape once, the earphones were removed, the tape rewound, and the earphones re¬ placed but reversed (without the subject's knowledge). The entire task was then repeated. On this second trial, of course, syllables previously presented to the right ear were presented to the left ear and vice versa. This was to balance for any bias in any part of the entire system (ie, any inherent imbalance in the loudness of one sound of any pair, undetected differences in the ear-phones). This dichotic listening task yields data that are easily scored. One can determine the number of occasions out of 120 (two trials of 60 stimuli each) in which the subject selected the signal that was presented to the right ear (the right-ear score) and the left
Table
1.—Summary of Subjects Stutterers 35
Sex Women Men
Age, yr Range Mean Formal education,
Task
The study used the magnetic tape recording developed by Shankweiler and Studdert-Kennedy.16,; This tape presents simulta¬ neously to the two ears pairs of consonant-vowel-consonant syl-
Mean,
yr
Nonstutterers 35
29
28
17-46 25.5
17-52 26.5
15.0
15.2
Severity of stuttering, Mean*
4.7
Seven-point scale of Johnson et al (1963) .,s Table
Dip Bep Gip Gep Gip Kip Pep Dep Kep Dep Bep Pip Bip Tep
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.
Gip Gep Kip Dep Pip Pip Gep Bep Tep Kep Tep Tip Pip Dep
16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29.
2.—Response Sheet for Dichotic Listening Task* Tip Kep Pep Bip Gip Tip Dip Bip Dep Kep Kip Pep Tep Gep
Gip
Gep Kep Kip Bip Bip Tip Dip Pep Bep Dip Bep Pep Tep
31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44.
Tip Gep Pep Gip Gep Gip Pep Dip Gip Bip Pep Pip Pip Tep
Kip Tep Bep Kip Dep Pip Gep Tip Bip Kip Kep Tip Dip Dep
46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59.
Tip Kep Bip Bep Dep Kep Dep Bip Kip Dep Kep Kip Bep Tep
Gip Gep Pip Tep Kep Tep Bep Dip Pip Pep Bep Dip Gep Pep
15._Pip_Dip_30_Tip_Kip_45._Tip_Bip_60;_Dip_Gip There is
a
12-second break after the 15th, 30th, and 45th pairs.
Downloaded From: http://archpsyc.jamanetwork.com/pdfaccess.ashx?url=/data/journals/psych/12266/ by a University of California - San Diego User on 01/13/2017
Table
3.—Summary of Experimental Stutterers
(N No. of reversals
Right-ear score, mean Total sample Right-ear score, mean Excluding reversals Absolute between-ear difference score, mean
=
35)
Note
Data Nonstutterers
(N
=
35)
6*
0*
68.4f
73.5f
72.5
73.5
24.7
26.9
Difference significant at .012 level. t Difference significant at .025 level. Other differences between *
groups not ear
(Table 3) that the
mean
right-ear
score
was
nonstutterers than for stutterers. Although not large, this difference also is significant (P < .025 by a one-tailed Student t test for independent samples). Both
greater for
statistically significant.
(the left-ear score). The absolute value of the difference be¬
tween these scores, the absolute between-ears difference score, may reflect the degree of cerebral lateralization of speech func¬ tion in the subject. Performed and scored in this manner, normal,
fully right-sided
individuals almost all show a right-ear prefer¬ About 40% of left-sided, normal individuals show a left-ear preference. These findings are consistent with the neurosurgical literature on the relationship between sidedness and cerebral dominance for speech." ence.
RESULTS
The experimental results are given in Table 1. The find¬ that none of the 35 nonstutterers showed a reversal, ie, none showed a left-ear preference, was anticipated, since almost all normal, right-sided individuals have leftcerebral dominance for speech. In contrast, six of the 35 stutterers (17%) showed a left-ear preference. This differ¬ ence is statistically significant (P=.012 by the Fisher Exact Probability Test19). The six subjects who showed a reversal did not differ clearly from the rest of the sample of stutterers in age, educational level, severity of stutter¬ ing, proportion of men, or in any other manner apparent to the authors. Both the Curry and Gregory11 and the Quinn12 studies reported reversals in nonstutterers as well as stutterers but with more in the latter group. Thus, the present study is consistent with their finding on this mea¬ sure but demonstrated a sharper division between the two samples. This difference may be related to the more strin¬ gent criteria for sidedness used in the present study or differences in the dichotic listening task itself, or both.
the studies of Curry and Gregory and of Quinn reported similar results. In the present study, the mean right-ear scores for stutterers and nonstutterers are nearly identi¬ cal if the stutterers who showed a reversal are omitted from the comparison (Table 3). This suggests that the dif¬ ference between stutterers and nonstutterers in this mea¬ sure is due to the subset of stutterers who showed a leftear preference in the dichotic listening task. Quinn12 found a similar result when he excluded subjects who showed a left-ear preference from his calculations. The absolute between-ear difference scores tended to be smaller for stutterers than nonstutterers. However, the mean difference between the two samples is small and not
statistically significant (Table 3), whether or not the sub¬ jects with a left-ear preference are excluded from the cal¬ culations. Recall that Curry and Gregory did find that stutterers had statistically significant smaller betweenear differences. The Quinn study, like the present one, found a difference in the same (predicted) direction but it was not statistically significant.
ing
CONCLUSIONS
keeping with the recent research of other investiga¬ stutterers in this study did show a less strong right-ear preference and smaller between-ear differences than nonstutterers in a dichotic listening task. However, these differences were small in magnitude and cast lim¬ ited light on the role of cerebral dominance in most stutter¬ ers. The most striking finding of this investigation is that six (17%) stutterers did show a left-ear preference while no nonstuttering control subjects evidenced such a rever¬ In
tors, the
sal. This suggests a subset of stutterers in whom an anom¬ aly of cerebral lateralization of speech functions may be important. A promising area of research entails the inten¬ sive study of this special group and its comparison with stutterers who do not show this phenomenon. This investigation was supported in part by Pub'ic Health Service search grant 24507 from the National Institute of Mental Health.
re¬
References 1. Travis L: Speech Pathology. New York, Appleton-Century-Crofts Inc, 1931. 2. Bryngelson B: Sidedness as an etiological factor in stuttering. J Gen Psychol 47:204-217, 1935. 3. Bryngelson B, Rutherford B: A comparative study of laterality of stutterers and non-stutterers. J Speech Disorders 2:15-16, 1937. 4. Daniels EM: An analysis of the relation between handedness and stuttering with special reference to the Orton-Travis theory of cerebral dominance. J Speech Disorders 5:309-326, 1940. 5. Heltman HJ: Contradictory evidence in handedness and stuttering. J Speech Disorders 5:327-331, 1940. 6. Broadbent DE: The role of auditory localization in attention and memory span. J Exp Psychol 47:191-196, 1954. 7. Kimura D: Functional asymmetry of the brain in dichotic listening. Cortex 3:163-178, 1967. 8. Wada J, Rasmussen T: Intracarotid injection of sodium amytal for the lateralization of cerebral speech dominance. JNeurosurg 17:266-282,1960. 9. Curry FKW: A comparison of left-handed and right-handed subjects on verbal and non-verbal dichotic listening tasks. Cortex 3:343-352, 1967. 10. Subirana A: The prognosis of aphasia in relation to cerebral dominance and handedness. Brain 81:415-425, 1958.
11. Curry FKW, Gregory HH: The performance of stutterers on dichotic listening tasks thought to reflect cerebral dominance. J Speech Hear Res 12:73-82, 1969. 12. Quinn PT: Stuttering, cerebral dominance and the dichotic word test. Med J Aust 2:639-643, 1972. 13. Satz P, Achenbach K, Pattishall E, et al: Order of report, ear asymmetry and handedness in dichotic listening. Cortex 1:377-396, 1965. 14. Rossi GF, Rosadini G: Experimental analysis of cerebral dominance in man, in CH Millikan, FL Darley (eds): Brain Mechanisms Underlying Speech and Language. New York, Grune & Stratton Inc, 1967, pp 167-184. 15. Johnson W. Darley FL, Spriesterbach DC: Diagnostic Methods in Speech Pathology. New York, Harper & Row Publishers Inc, 1963. 16. Shankweiler D, Studdert-Kennedy M: An analysis of perceptual confusion in identification of dichotically presented CVC syllables. J Acoust Soc Am 41:1581, 1967. 17. Shankweiler D, Studdert-Kennedy M: Identification of consonants and vowels presented to left and right ears. Q J Exp Psychol 19:59-63,1967. 18. Liberman AM, Cooper FS, Shankweiler DP, et al: Perception of the speech code. Psychol Rev 74:431-461, 1967. 19. Siegel S: Nonparametric Statistics. New York, McGraw-Hill Book Co
Inc, 1956, pp 96-101.
Downloaded From: http://archpsyc.jamanetwork.com/pdfaccess.ashx?url=/data/journals/psych/12266/ by a University of California - San Diego User on 01/13/2017
