Perceptual and Motor Skills, 1975, 40, 379-386. @ Perceptual and Motor Skills 1975
DICHOTIC W O R D - P E R C E P T I O N OF APHASIC AND NORMAL SUBJECTS1 W. H. MOORE, JR.' Auburn University
AND
WILLIAM E. WEIDNER Kent State University
Summary.-The present investigation examined the role of the right cerebral hemisphere in linguistic perception following a left cerebral insult which had resulted in aphasia. Auditory dichotic procedures were utilized to investigate the ear-preferences of 30 aphasic Ss, grouped relative to the amount of time since the onset of left cerebral insult, and a group of 10 normal control Ss. An oral-response task and a pointing-response task were given. Statistical analyses showed a significant left-ear preference for aphasic Ss who were more than 6 mo. post-cerebral insult; however, Ss less than 6 mo. post-cerebral insult did not demonstrate a significant ear-preference under auditory dichotic stimulation. In contrast, a significant right-ear preference was noted for the normal controls. Significant differences were not shown for the aphasic or control groups on the oral and pointing tasks. Studies which have utilized simultaneous presentation of auditory stimuli have indicated that the ear contralateral t o the dominant hemisphere is more efficient i n perception of words and digits than is the ipsilateral ear (e.g., Kimura, 1961a, 1961b, 1963; Bryden, 1963; Broadbent & Gregory, 1964; Dirks, 1964; Cooper, et al., 1967; Borkowski, et al., 1965; Berlin & Lowe, 1972). In most normal Ss a preference for the right ear (left hemisphere) has been indicated under dichotic stimulation. Kimura (1961a, 1961b, 1 9 6 3 ) has suggested that these findings may indicate that the crossed auditory pathways of the contralateral ear are stronger than the uncrossed or ipsilateral pathways and that the dominant hemisphere plays a more important role i n the perception o f spoken material than does the non-dominant hemisphere. The effects of temporal-lobe damage i n the perception of dichotically presented digits has been investigated by Kimura ( 1 9 6 1 b ) . Ss with temporal lobectomy demonstrated impairment a t the ear contralateral t o temporal lobectorny. Perception of dichotically presented melodies and digits was investigated by Shankweiler ( 1 9 6 6 ) for patients with temporal-lobe damage. Ss with right temporal lobectomy demonstrated impairment in perception of melodies; however, Ss with left temporal lobectomy showed impairment in the perception of digits. Oxbury and Oxbury ( 1 9 6 9 ) reported following left temporal lobectomy, which included Heschl's gyrus, more errors for the right-ear responses, but not 'This article is based, in part, on a doctoral dissertation completed by W. H. Moore, Jr. under the direction of William E. Weidner. The authors are grateful to R. K. Sommers. J. P. Millin, J. C . McNutt, W. Brady, E. Brown, and W. ~ i t t for s their assistance. he authors also wish to thank Mrs. Sherrie McArdle for her assistance in the preparation of the manuscript. aRequests for reprints should be directed to W. M. Moore, Jr., Speech and Hearing Clinic, Auburn University, Auburn, Alabama 36830.
380
W. H. MOORE, JR.
&
W. E. WEIDNER
for left-ear responses, occurred under dichotic stimulation with digits. When Heschl's gyrus was not removed the effect was an altered order of reporting such that stimuli to the left ear were reported first more frequently than those to the right ear. Pettit (1969) investigated the ear-preference of 25 aphasic adults and a control group of 25 normal adults by administering linguistic and non-linguistic auditory dichotic tests. Results indicated that the performance of the left ear for aphasic Ss was significantly better than the right on the linguistic dichotic tests. However, the normal control group demonstrated a significant preference for the right ear on the linguistic dichotic tests. Although a significant preference for the left ear was found for Pettit's aphasic Ss time between the onset of cerebral insult and testing was not investigated; thus, information about earpreference in regard to time post-cerebral insult was not available from Pettic's study. The increasing information on the perception of dichotically presented linguistic stimuli by normal Ss suggests that preference for the right ear by these Ss is related to the more direct connections between the right ear and the "language centers" of the left hemisphere, supporting a left cerebral-dominance theory for speech perception. The present investigation was designed to assess the ear-preference of Ss who had suffered a left cerebral insult which had resulted in aphasia relative to the time since the occurrence of the left cerebral insult. A second purpose was to study the effects of response modality on earpreference under dichotic stimulation of aphasic Ss.
METHOD Sabjects Thirty aphasic Ss were selected from in-patient and out-patient clinical case loads of rehabilitation hospitals and speech and hearing clinics in norcheastern and central Ohio. These Ss were adults who had been diagnosed as having aphasia resulting from left cerebral insult (Ss were selected on the basis of medical reports of neurological examination which specified a left cerebral insult). An equal number of Ss were selected from the following groups relative to onset of cerebral insult: Group I, 1 to 6 mo. post-cerebral insult; Group 11, 7 to 12 mo. post-cerebral insult; Group 111, beyond 12 mo. post-cerebral insult. With the exception of 2 Ss in Group I11 (one of whom had traumatic head injury, the other had brain surgery for the removal of a tumor) all Ss had suffered cerebrovascular accidents. Group I Ss ranged in age from 41.9 to 74.3 yr., with a mean age of 60.2 yr. Time post-cerebral insult ranged from 1 mo. to 5.3 mo., with a mean of 2.64 mo. The 10 Ss in Group I1 ranged in age from 47.3 to 69.3 yr., with a mean age of 59.5 y r Time post-cerebral insult ranged from 6.1 ro 11.2 mo., with a mean of 8.37 rno. Group I11 Ss ranged in age from 22.8 to 72.7 yr., with a mean
WORD-PERCEPTION IN APHASICS AND NORMALS
381
age of 53.4 yr. Time post-cerebral insult ranged from 12.2 to 93 mo., with a mean of 38.2 mo. The aphasic Ss were required to meet the following criteria: (1) were right-handed prior to cerebral insult; ( 2 ) were monolingual, native English speakers; ( 3 ) had no reported speech/language problems prior to cerebral insult; (4) had hearing within 20 dB HL (re: ANSI) for the frequencies 500, 1000, 2000 Hz; ( 5 ) could repeat and point correctly to the 8 stimulus words utilized in the investigation when presented by the experimenter by live-voice in free field. Also, the aphasic Ss were required to meet a 50% average criterion of correct responses on a number of items on the following subcests of the Minnesota Test for Differential Diagnosis of Aphasia (Schuell, 1965) : ( 1 ) Auditory Disturbances, ( 2 ) Visual and Reading Disturbances, and ( 3 ) Speech and Language Disturbances. In the control group were 10 normal, young adults who were either graduate students or university staff. These Ss were individuals who had not participated in research utilizing dichotic procedures nor were they familiar with results of such research. These Ss ranged in age from 22.1 to 34.2 yr., with a mean age of 24.9 yr. They were required to meet the following criteria: (1) were right-handed; ( 2 ) were monolingual, native English speakers; ( 3 ) had no history of brain pathology; ( 4 ) had hearing within 20 dB HL (re: ANSI) for the frequencies 500, 1000, and 2000 Hz.
Constmction of Dichotic Tapes The word-pairs used in the investigation were: Mat-Bat, Bad-Dad, TearPear, and Pan-Tan. Each member of a pair of CVC words was recorded on a single channel of a two-channel Arnpex, Model 602-2, tape loop recording system with a moveable playback head on channel 1. Each pair was monitored on a storage oscilloscope while the moveable playback head was adjusted to align the onset of times of the CVC words within f2 msec. When proper onset alignment had been achieved, the pairs were then recorded onto a master tape of a two-channel Ampex, Model 606-2, tape recorder. Aligned pairs of CVC words from the master tape were used to generate two randomized dichotic audio tapes consisting of 24 pairs of CVC words. Each pair of CVC words was presented six times with each member of the pair presented to each ear an equal number of times. Root mean square intensities of the CVC word-pairs were monitored with a Bruel and Kjaer Sound Level Recorder. Word-pairs were within +2 dB of each other both within and between word-pairs. Expe~imentalTasks The first task required an oral response from Ss. Following the dichotic presentation of each of the 24 word-pairs Ss were instructed to respond verbally with the first word they had heard. During the second task, the pointing-response
382
W. H. MOORE, JR. & W. E. WEIDNER
task, a list of three words was placed on a reading stand in front of Ss who were instructed to point to the word which they heard. These word lists consisted of the word presented to the right ear, the word presented to the left ear and third CVC word. Subject-assignment for the oral and pointing tasks was counterbalanced across both the aphasic and control Ss.
Pro c e d u ~ e Simultaneous pairs of CVC words were presented over Telephonics, T D H 39 earphones mounted in MX 41/AR cushions and matched within -+2 dB from 100 to 3000 Hz. The test tapes were played on an Arnpex, Model 602-2 stereo tape recorder with internal setting of 90 dB (re: SPL) output per channel, relative to a 1000-Hz calibration tone, as measured at the earphones employing a Bruel and Kjaer, NPS 6-2/cc coupler and a Bruel and Kjaer audio frequency recorder, Type 3312. All CVC word-pairs were presented within 2 2 dB of each other, both within and between word-pairs. Two Hewlett-Packard, Model 305D, attenuator sets, calibrated for linearity, were utilized to present the stimulus material at 50 dB, SL (re: PTA) to each ear. Ss' three frequency pure-tone averages were converted back to SPL so that attenuator adjustments could be made for presenting the stimuli at 50 dB SL. Two impedance matching transformers, 500 to 8 ohms, were placed in line between the attenuator sets and the earphones. During the oral-response task Ss' oral responses to the simultaneously presented word-pairs were recorded on appropriate data sheets. Ss were instructed to point to the word which they had heard during the pointing-responses task. Immediately following each presentation, a list of three words was placed in front of Ss on a reading stand and Ss were instructed to point to the word which they had heard. Ss' pointing responses were recorded on appropriate data sheets. Reliability Reliability of the experimental procedures was estimated by readministering the test procedures to six of the aphasic Ss drawn at random. The intraclass correlation procedure, described by Winer ( 1971 ) , was employed to estimate the reliability for the right and left ears for the oral- and pointing-response tasks. Identical correlations of .99 were found for the right and left ears for the oral task, while identical correlations of .96 were obtained for the right and left ears for the pointing task. RESULTS
The number of items correct at the right and left ears for aphasic and control groups on the oral- and pointing-response tasks constituted the raw data. A three-factor analysis of variance for repeated measurements (Winer, 1971) was employed. If F rests for any of the main or interaction effects were significant at the .05 level, a posteriori mean comparisons tests were employed,
WORD-PERCEPTION IN APHASICS AND NORMALS
383
where appropriate, utilizing the Newman-Keuls method (Winer, 1971) to determine whether differences between obtained means of the F tests were statistically significant. Results of this analysis of variance showed a significant main effect for ears ( F = 7.33, d f = 1/72, p < .01) and a significant groups X ears interaction ( F = 29.38, df = 3/72, fl < .01). The F ratios for the main effects of groups ( F = .04, df = 3.72) and tasks (F = .12, df = 1/72) were not significant ( p > .05) nor were significant F ratios shown for the remaining three interaction effects of groups X tasks (F = .33, df = 3/72), ears X tasks ( F = .26, d f = 1/72) and groups X ears X tasks ( F = .86, df = 3/72) when tested at the .05 level. The F test of the main effect for ears indicated that the over-all mean of 13.18 correct for the left ear was significantly different from that of 7.94 for the right ear. As there was an over-all significant interaction of groups X ears a posteriori mean comparisons of the between- and within-groups mean correct responses for right and left ears were made. The Newman-Keuls procedure, described by Winer (1971), was employed to test the differences. The following critical differences were required for significance at the .Ol level: two steps, 3.49; three steps, 3.86; and four steps, 4.14. AS shown in Table 1, Groups I1 and 111 obtained a significantly greater mean number correct for the left ears than for right ears, while Group I did not demonstrate such a significant difference. For the right ear mean correct of 16.10 for controls was significantly different from the mean of 7.90 for the left ear ( p < ,01; Table I ) , showing an over-all right ear-preference for the normal controls. When the mean correct for the right ear by the control group was compared with that for the right ear by the three aphasic groups all comparisons were significant ( p < .01), indicating more mean correct for the control group's right ears than for each of the aphasic groups' mean correct TABLE 1 ABSOLUTE DIFFERENCES BETWEEN MEAN NUMBERS OF RESPONSES FOR RIGHTAND LEFT EARSOF APHASIC AND CONTROL GROUPSFOR AUDITORY, DICHOTIC PROCEDURES Group
Group I Right Left
M Righ,t 10.70 Left 12.75 11 Right 8.15 Left 15.25 111 Right 8.85 Left 14.80 Cont Right 16.10 Left 7.90
I
10.70 12.75 2.05
Group I1 Right Left 8.15 15.25 2.55 2.50 7.10*
Group I11 Left
Control Right Left
8.85 14.80 1.85 2.05 .70 .4 5 5.95*
16.10 7.90 5.40* 4.85* 7.95* 7.35" 7.25* 6.90* 8.20*
Right
,
384
W. H.MOORE, JR. pc W. E. WEIDNER
for the right ear (Table 1 ) . These comparisons indicated that the normal control group obtained significantly more correct for the right ear than did each of the three aphasic groups. These analyses also showed significantly fewer correct for the left ear for the control group than for each of the three aphasic groups (Table 1 ) . Fig. 1 illustrates the mean number of correct items for the right and left ears for each of the four groups. There is not simply a decrement in responses for aphasic Ss but rather a shift in ear-preferences relative to that for the normal Ss. As indicated in Fig. 1 responses for the left ear for all three of the aphasic groups are higher than that of the normal control group.
)Right Ear '/
*-. Left Ear
I
II
III
Control
GROUP
FIG. 1. Profile of mean number of responses for right and left ears by experimental and normal control groups
DISCUSSION Results of the analysis of variance and a posterio~i comparisons of means showed a right ear-preference for controls, as a group, under dichotic stimulation using meaningful linguistic stimuli. These findings agree with other work which has employed syllables, words, and digits ( e.g., Kimura, 1961a, 1961b, 1963; Bryden, 1963; Broadbent & Gregory, 1964; Dirks, 1964; Cooper, et al., 1967; Borkowski, et al., 1965; ~ e i l i n& Lowe, 1972). The asymmetry in superiority of ears, in favor of the right ear, for linguistic material by normal Ss has been attributed to the more direct auditory pathways between the right ear and the "language centers" in the left hemisphere by Kimura (1961a, 1961b, 1963). These findings and explanation for the superiority of the right ear under dichotic stimulation parallels those for the superiority of the right visual field under bilateral tachistoscopic stimulation ( McKeever & Huling, 197la, 1971b; Hines, 1972; McKeever, et al., 1972). That auditory and visual superiority, under specified conditions is contralateral to the left hemisphere supports the view that a left cerebral-dominance underlies language perception and processing. However, the results from the aphasic Ss in the current investigation, utilizing auditory dichotic stimulation, do not support such a position. All three of the aphasic groups demonstrated more
WORD-PERCEPTION IN APHASICS AND NORMALS
385
correct responses for the left ear, exactly opposite that demonstrated by the normal control group. The between-groups main effect did not indicate significant differences between ears for the over-all correct scores of the three aphasic groups and the normal control group when scores were collapsed across ears. Thus, there was not simply a decrement in $Cores for the ear contralateral - . to the lesioned hemispheres of the aphasic Ss but rather a shift in dominance to the left ear. Although all three of the aphasic groups demonstrated greater mean scores for the left ear than the right ear, only Ss beyond 6-mo. post-cerebral insult, Groups I1 and 111, obtained significantly more mean correct responses to the left ear than to the right ear. These findings for aphasic Ss beyond 6-mo. postcerebral insult corroborate Pettit's (1969) findings on ear-preference for aphasic Ss. Our current findings did not show a significant ear-preference for Ss below 6-mo. post-cerebral insult, suggesting the importance of time in determining earpreference of our aphasic Ss. Indeed, when viewed together, these findings appear to support the inference that there is a shift in hemispheric dominance for linguistic perception and processing of auditory-linguistic stimuli following a left cerebral insult which has resulted in aphasia. Our current data on the perception of auditory-linguistic stimuli appears to compliment the already accumulated body of data which stresses the importance of the right hemisphere in linguistic processing following left cerebral insult resulting in aphasia (Smith, 1971). However, a temporal factor appears to be related to this shift in hemispheric perception and processing of auditory stimuli, in that aphasic Ss below 6-mo. post-cerebral insult did nor show a significant left ear-preference under auditory dichotic stimulation. This finding may suggest that the recovery of language processes seen in aphasic Ss during the first 6 or 7 mo. post-cerebral insult (Darley, 1972) may reflect this temporal factor which appears related to the shift in hemispheric dominance for auditory-linguistic stimuli in aphasic Ss. In the present investigation no attempt was made to match the ages of the aphasic and control Ss. It is possible that older control Ss who have a negative history of brain pathology would show a less pronounced preference for the right ear than did our control group of normal young adults. However, that Pettit's ( 1969) control group which ranged in age from 22 to 77 yr. ( M = 42 yr.) demonstrated the same significant superiority of the right ear as our control Ss indicates some consistency in findings for the right ear-preferences of normal nonpachological Ss regardless of age. Nonetheless, future researchers may explore the importance of matching aphasic and control Ss for age. Statistical analyses indicated that aphasic Ss obtained statistically equivalent mean numbers of correct responses for both the oral- and pointing-response tasks. The finding that aphasic Ss do comparatively better on non-verbal response tasks of recognition than on oral tasks of verbal recall (Keenan, 1968) was not supported here when dichotic listening procedures were utilized. These
386
w. H. MOORE,
JR
W. E. WEIDNER
discrepant findings may relate to the relatively concrete word meanings, high frequency of occurrence, and redundancy of the stimulus words employed in the current investigation. However, the present findings are supportive of Schuell, et al.'s (1969) observations that intensive auditory stimulation and repetitive sensory stimulation facilitate evocation of correct verbal responses from aphasic patients. REFERENCES BERLIN,C. I., & LOWE, S. S. Temporal and dichotic factors in central auditory testing. In J. Katz (Ed.), Handbook of clinical audiology. Balcimore: Williams & Wilkins, 1972. Pp. 280-312. BORKOWSKI,J. G., SPREEN, O., & S T U ~J., Z. Ear preference and abstractness in dichotic listening. Psychonomic Science, 1965, 3, 547-548. BROADBENT,D. E., & GREGORY,M. Accuracy for speech presented to the right and left ears. Quarterly Jot/rnal of Experimentaj Psychology, 1964, 16, 359-360. BRYDEN,M. P. Ear preference in auditory perception. Iournal of Experimental Psychology, 1963, 65, 103-105. COOPER,A,, ACHENBACK.K., SATZ, P., & LEVY,C. M. Order of report and ear asymmetry in dichotic listening. Psychonomic Science, 1967, 9, 97-98. DARLEY,D. L. The efficacy of language rehabilitation in aphasia. Journal of Speech and Hearing Disorders, 1972, 37, 3-21. DIRKS, D. Perception of dichotic and monaural verbal material and cerebral dominance in speech. Acta Otolaryngologica, 1964, 58, 73-80. UMW, D. Cerebral dominance and the perception of verbal stimuli. Canadian lournal o f Psychology, 1961, 15, 166-171. ( a ) KIMURA.D. Some effects of temporal-lobe damage on auditory perception. Canadian Journal o f Psychology, 1961, 15, 156-165. ( b ) 1963. KIMURA,D . A note on cerebral dominance in hearing. Acta Oto~aryngo~ogica, 56, 617-619. HINES,D. Bilateral tachistoscopic recognition of verbal and non-verbal stimuli. Cortex, 1972, 7, 313-322. MCKEEVW, W. F., & HULING, M. D . Bilateral tachistoscopic word recognition as a function of hemisphere stimulated and interhemispheric transfer time. Neuropsycbologia, 1971, 9, 281-288. ( a ) MCKEEVER, W. F., & HULING,M. D . Lateral dominance in tachistoscopic word recognition performances obtained with simultaneous bilateral input. Neuroprychologia, 1971, 9. 15-20. ( b ) MCKEEVER,W. F., SUBWI, J., & VAN DEVENTER,A. D . Fixation control in tachistoscopic studies of laterality effects: comment and data relevant to Hine's experiment. Cortex, 1972, 8,473-479. KEENAN,J. S. The nature of receptive and expressive impairments in aphasia. Journal of Speech and Hearing Disorders, 1968, 33, 20-25. OXBURY. J. M., & OXBURY, S. M. Effect of temporal lobectomy on the report of dichotically presented digits. Cortex, 1969, 4, 3-14. PETTIT, J. M. Cerebral dominance and the process of language recovery in aphasia. Unpublished doctoral dissertation, Purdue Univer., 1969. SCHUELL,H. Differential diagnosis of aphasia with che Minnesota test. Minneapolis: Univer. of Minnesota. 1965. SCHUELL,H., JENKINS,J. J., & JIMENEZ-PABON, E. Aphasia in adults: diagnosis, prognosir, and treatment. New York: Hoeber Medical Division, Harper, 1969. SHANKWEILER, D . P. Effects of temporal lobe damage on the erception of dichoticall presented melodies. Journal of Comparative and Pbysiofogical Psychology, 1966): 62, 115-119. SMITH, A. Objective indices of severicy of chronic a hasia in stroke patients. Journal of Speech and Hearing Disorders, 1971, 36, 12;-207. WINER, B. J. Statistical principles in experimental design. New York: McGraw-Hill, 1971.
Accepted Decen ber 12, 1974.
DICHOTIC W O R D - P E R C E P T I O N OF APHASIC AND NORMAL SUBJECTS1 W. H. MOORE, JR.' Auburn University
AND
WILLIAM E. WEIDNER Kent State University
Summary.-The present investigation examined the role of the right cerebral hemisphere in linguistic perception following a left cerebral insult which had resulted in aphasia. Auditory dichotic procedures were utilized to investigate the ear-preferences of 30 aphasic Ss, grouped relative to the amount of time since the onset of left cerebral insult, and a group of 10 normal control Ss. An oral-response task and a pointing-response task were given. Statistical analyses showed a significant left-ear preference for aphasic Ss who were more than 6 mo. post-cerebral insult; however, Ss less than 6 mo. post-cerebral insult did not demonstrate a significant ear-preference under auditory dichotic stimulation. In contrast, a significant right-ear preference was noted for the normal controls. Significant differences were not shown for the aphasic or control groups on the oral and pointing tasks. Studies which have utilized simultaneous presentation of auditory stimuli have indicated that the ear contralateral t o the dominant hemisphere is more efficient i n perception of words and digits than is the ipsilateral ear (e.g., Kimura, 1961a, 1961b, 1963; Bryden, 1963; Broadbent & Gregory, 1964; Dirks, 1964; Cooper, et al., 1967; Borkowski, et al., 1965; Berlin & Lowe, 1972). In most normal Ss a preference for the right ear (left hemisphere) has been indicated under dichotic stimulation. Kimura (1961a, 1961b, 1 9 6 3 ) has suggested that these findings may indicate that the crossed auditory pathways of the contralateral ear are stronger than the uncrossed or ipsilateral pathways and that the dominant hemisphere plays a more important role i n the perception o f spoken material than does the non-dominant hemisphere. The effects of temporal-lobe damage i n the perception of dichotically presented digits has been investigated by Kimura ( 1 9 6 1 b ) . Ss with temporal lobectomy demonstrated impairment a t the ear contralateral t o temporal lobectorny. Perception of dichotically presented melodies and digits was investigated by Shankweiler ( 1 9 6 6 ) for patients with temporal-lobe damage. Ss with right temporal lobectomy demonstrated impairment in perception of melodies; however, Ss with left temporal lobectomy showed impairment in the perception of digits. Oxbury and Oxbury ( 1 9 6 9 ) reported following left temporal lobectomy, which included Heschl's gyrus, more errors for the right-ear responses, but not 'This article is based, in part, on a doctoral dissertation completed by W. H. Moore, Jr. under the direction of William E. Weidner. The authors are grateful to R. K. Sommers. J. P. Millin, J. C . McNutt, W. Brady, E. Brown, and W. ~ i t t for s their assistance. he authors also wish to thank Mrs. Sherrie McArdle for her assistance in the preparation of the manuscript. aRequests for reprints should be directed to W. M. Moore, Jr., Speech and Hearing Clinic, Auburn University, Auburn, Alabama 36830.
380
W. H. MOORE, JR.
&
W. E. WEIDNER
for left-ear responses, occurred under dichotic stimulation with digits. When Heschl's gyrus was not removed the effect was an altered order of reporting such that stimuli to the left ear were reported first more frequently than those to the right ear. Pettit (1969) investigated the ear-preference of 25 aphasic adults and a control group of 25 normal adults by administering linguistic and non-linguistic auditory dichotic tests. Results indicated that the performance of the left ear for aphasic Ss was significantly better than the right on the linguistic dichotic tests. However, the normal control group demonstrated a significant preference for the right ear on the linguistic dichotic tests. Although a significant preference for the left ear was found for Pettit's aphasic Ss time between the onset of cerebral insult and testing was not investigated; thus, information about earpreference in regard to time post-cerebral insult was not available from Pettic's study. The increasing information on the perception of dichotically presented linguistic stimuli by normal Ss suggests that preference for the right ear by these Ss is related to the more direct connections between the right ear and the "language centers" of the left hemisphere, supporting a left cerebral-dominance theory for speech perception. The present investigation was designed to assess the ear-preference of Ss who had suffered a left cerebral insult which had resulted in aphasia relative to the time since the occurrence of the left cerebral insult. A second purpose was to study the effects of response modality on earpreference under dichotic stimulation of aphasic Ss.
METHOD Sabjects Thirty aphasic Ss were selected from in-patient and out-patient clinical case loads of rehabilitation hospitals and speech and hearing clinics in norcheastern and central Ohio. These Ss were adults who had been diagnosed as having aphasia resulting from left cerebral insult (Ss were selected on the basis of medical reports of neurological examination which specified a left cerebral insult). An equal number of Ss were selected from the following groups relative to onset of cerebral insult: Group I, 1 to 6 mo. post-cerebral insult; Group 11, 7 to 12 mo. post-cerebral insult; Group 111, beyond 12 mo. post-cerebral insult. With the exception of 2 Ss in Group I11 (one of whom had traumatic head injury, the other had brain surgery for the removal of a tumor) all Ss had suffered cerebrovascular accidents. Group I Ss ranged in age from 41.9 to 74.3 yr., with a mean age of 60.2 yr. Time post-cerebral insult ranged from 1 mo. to 5.3 mo., with a mean of 2.64 mo. The 10 Ss in Group I1 ranged in age from 47.3 to 69.3 yr., with a mean age of 59.5 y r Time post-cerebral insult ranged from 6.1 ro 11.2 mo., with a mean of 8.37 rno. Group I11 Ss ranged in age from 22.8 to 72.7 yr., with a mean
WORD-PERCEPTION IN APHASICS AND NORMALS
381
age of 53.4 yr. Time post-cerebral insult ranged from 12.2 to 93 mo., with a mean of 38.2 mo. The aphasic Ss were required to meet the following criteria: (1) were right-handed prior to cerebral insult; ( 2 ) were monolingual, native English speakers; ( 3 ) had no reported speech/language problems prior to cerebral insult; (4) had hearing within 20 dB HL (re: ANSI) for the frequencies 500, 1000, 2000 Hz; ( 5 ) could repeat and point correctly to the 8 stimulus words utilized in the investigation when presented by the experimenter by live-voice in free field. Also, the aphasic Ss were required to meet a 50% average criterion of correct responses on a number of items on the following subcests of the Minnesota Test for Differential Diagnosis of Aphasia (Schuell, 1965) : ( 1 ) Auditory Disturbances, ( 2 ) Visual and Reading Disturbances, and ( 3 ) Speech and Language Disturbances. In the control group were 10 normal, young adults who were either graduate students or university staff. These Ss were individuals who had not participated in research utilizing dichotic procedures nor were they familiar with results of such research. These Ss ranged in age from 22.1 to 34.2 yr., with a mean age of 24.9 yr. They were required to meet the following criteria: (1) were right-handed; ( 2 ) were monolingual, native English speakers; ( 3 ) had no history of brain pathology; ( 4 ) had hearing within 20 dB HL (re: ANSI) for the frequencies 500, 1000, and 2000 Hz.
Constmction of Dichotic Tapes The word-pairs used in the investigation were: Mat-Bat, Bad-Dad, TearPear, and Pan-Tan. Each member of a pair of CVC words was recorded on a single channel of a two-channel Arnpex, Model 602-2, tape loop recording system with a moveable playback head on channel 1. Each pair was monitored on a storage oscilloscope while the moveable playback head was adjusted to align the onset of times of the CVC words within f2 msec. When proper onset alignment had been achieved, the pairs were then recorded onto a master tape of a two-channel Ampex, Model 606-2, tape recorder. Aligned pairs of CVC words from the master tape were used to generate two randomized dichotic audio tapes consisting of 24 pairs of CVC words. Each pair of CVC words was presented six times with each member of the pair presented to each ear an equal number of times. Root mean square intensities of the CVC word-pairs were monitored with a Bruel and Kjaer Sound Level Recorder. Word-pairs were within +2 dB of each other both within and between word-pairs. Expe~imentalTasks The first task required an oral response from Ss. Following the dichotic presentation of each of the 24 word-pairs Ss were instructed to respond verbally with the first word they had heard. During the second task, the pointing-response
382
W. H. MOORE, JR. & W. E. WEIDNER
task, a list of three words was placed on a reading stand in front of Ss who were instructed to point to the word which they heard. These word lists consisted of the word presented to the right ear, the word presented to the left ear and third CVC word. Subject-assignment for the oral and pointing tasks was counterbalanced across both the aphasic and control Ss.
Pro c e d u ~ e Simultaneous pairs of CVC words were presented over Telephonics, T D H 39 earphones mounted in MX 41/AR cushions and matched within -+2 dB from 100 to 3000 Hz. The test tapes were played on an Arnpex, Model 602-2 stereo tape recorder with internal setting of 90 dB (re: SPL) output per channel, relative to a 1000-Hz calibration tone, as measured at the earphones employing a Bruel and Kjaer, NPS 6-2/cc coupler and a Bruel and Kjaer audio frequency recorder, Type 3312. All CVC word-pairs were presented within 2 2 dB of each other, both within and between word-pairs. Two Hewlett-Packard, Model 305D, attenuator sets, calibrated for linearity, were utilized to present the stimulus material at 50 dB, SL (re: PTA) to each ear. Ss' three frequency pure-tone averages were converted back to SPL so that attenuator adjustments could be made for presenting the stimuli at 50 dB SL. Two impedance matching transformers, 500 to 8 ohms, were placed in line between the attenuator sets and the earphones. During the oral-response task Ss' oral responses to the simultaneously presented word-pairs were recorded on appropriate data sheets. Ss were instructed to point to the word which they had heard during the pointing-responses task. Immediately following each presentation, a list of three words was placed in front of Ss on a reading stand and Ss were instructed to point to the word which they had heard. Ss' pointing responses were recorded on appropriate data sheets. Reliability Reliability of the experimental procedures was estimated by readministering the test procedures to six of the aphasic Ss drawn at random. The intraclass correlation procedure, described by Winer ( 1971 ) , was employed to estimate the reliability for the right and left ears for the oral- and pointing-response tasks. Identical correlations of .99 were found for the right and left ears for the oral task, while identical correlations of .96 were obtained for the right and left ears for the pointing task. RESULTS
The number of items correct at the right and left ears for aphasic and control groups on the oral- and pointing-response tasks constituted the raw data. A three-factor analysis of variance for repeated measurements (Winer, 1971) was employed. If F rests for any of the main or interaction effects were significant at the .05 level, a posteriori mean comparisons tests were employed,
WORD-PERCEPTION IN APHASICS AND NORMALS
383
where appropriate, utilizing the Newman-Keuls method (Winer, 1971) to determine whether differences between obtained means of the F tests were statistically significant. Results of this analysis of variance showed a significant main effect for ears ( F = 7.33, d f = 1/72, p < .01) and a significant groups X ears interaction ( F = 29.38, df = 3/72, fl < .01). The F ratios for the main effects of groups ( F = .04, df = 3.72) and tasks (F = .12, df = 1/72) were not significant ( p > .05) nor were significant F ratios shown for the remaining three interaction effects of groups X tasks (F = .33, df = 3/72), ears X tasks ( F = .26, d f = 1/72) and groups X ears X tasks ( F = .86, df = 3/72) when tested at the .05 level. The F test of the main effect for ears indicated that the over-all mean of 13.18 correct for the left ear was significantly different from that of 7.94 for the right ear. As there was an over-all significant interaction of groups X ears a posteriori mean comparisons of the between- and within-groups mean correct responses for right and left ears were made. The Newman-Keuls procedure, described by Winer (1971), was employed to test the differences. The following critical differences were required for significance at the .Ol level: two steps, 3.49; three steps, 3.86; and four steps, 4.14. AS shown in Table 1, Groups I1 and 111 obtained a significantly greater mean number correct for the left ears than for right ears, while Group I did not demonstrate such a significant difference. For the right ear mean correct of 16.10 for controls was significantly different from the mean of 7.90 for the left ear ( p < ,01; Table I ) , showing an over-all right ear-preference for the normal controls. When the mean correct for the right ear by the control group was compared with that for the right ear by the three aphasic groups all comparisons were significant ( p < .01), indicating more mean correct for the control group's right ears than for each of the aphasic groups' mean correct TABLE 1 ABSOLUTE DIFFERENCES BETWEEN MEAN NUMBERS OF RESPONSES FOR RIGHTAND LEFT EARSOF APHASIC AND CONTROL GROUPSFOR AUDITORY, DICHOTIC PROCEDURES Group
Group I Right Left
M Righ,t 10.70 Left 12.75 11 Right 8.15 Left 15.25 111 Right 8.85 Left 14.80 Cont Right 16.10 Left 7.90
I
10.70 12.75 2.05
Group I1 Right Left 8.15 15.25 2.55 2.50 7.10*
Group I11 Left
Control Right Left
8.85 14.80 1.85 2.05 .70 .4 5 5.95*
16.10 7.90 5.40* 4.85* 7.95* 7.35" 7.25* 6.90* 8.20*
Right
,
384
W. H.MOORE, JR. pc W. E. WEIDNER
for the right ear (Table 1 ) . These comparisons indicated that the normal control group obtained significantly more correct for the right ear than did each of the three aphasic groups. These analyses also showed significantly fewer correct for the left ear for the control group than for each of the three aphasic groups (Table 1 ) . Fig. 1 illustrates the mean number of correct items for the right and left ears for each of the four groups. There is not simply a decrement in responses for aphasic Ss but rather a shift in ear-preferences relative to that for the normal Ss. As indicated in Fig. 1 responses for the left ear for all three of the aphasic groups are higher than that of the normal control group.
)Right Ear '/
*-. Left Ear
I
II
III
Control
GROUP
FIG. 1. Profile of mean number of responses for right and left ears by experimental and normal control groups
DISCUSSION Results of the analysis of variance and a posterio~i comparisons of means showed a right ear-preference for controls, as a group, under dichotic stimulation using meaningful linguistic stimuli. These findings agree with other work which has employed syllables, words, and digits ( e.g., Kimura, 1961a, 1961b, 1963; Bryden, 1963; Broadbent & Gregory, 1964; Dirks, 1964; Cooper, et al., 1967; Borkowski, et al., 1965; ~ e i l i n& Lowe, 1972). The asymmetry in superiority of ears, in favor of the right ear, for linguistic material by normal Ss has been attributed to the more direct auditory pathways between the right ear and the "language centers" in the left hemisphere by Kimura (1961a, 1961b, 1963). These findings and explanation for the superiority of the right ear under dichotic stimulation parallels those for the superiority of the right visual field under bilateral tachistoscopic stimulation ( McKeever & Huling, 197la, 1971b; Hines, 1972; McKeever, et al., 1972). That auditory and visual superiority, under specified conditions is contralateral to the left hemisphere supports the view that a left cerebral-dominance underlies language perception and processing. However, the results from the aphasic Ss in the current investigation, utilizing auditory dichotic stimulation, do not support such a position. All three of the aphasic groups demonstrated more
WORD-PERCEPTION IN APHASICS AND NORMALS
385
correct responses for the left ear, exactly opposite that demonstrated by the normal control group. The between-groups main effect did not indicate significant differences between ears for the over-all correct scores of the three aphasic groups and the normal control group when scores were collapsed across ears. Thus, there was not simply a decrement in $Cores for the ear contralateral - . to the lesioned hemispheres of the aphasic Ss but rather a shift in dominance to the left ear. Although all three of the aphasic groups demonstrated greater mean scores for the left ear than the right ear, only Ss beyond 6-mo. post-cerebral insult, Groups I1 and 111, obtained significantly more mean correct responses to the left ear than to the right ear. These findings for aphasic Ss beyond 6-mo. postcerebral insult corroborate Pettit's (1969) findings on ear-preference for aphasic Ss. Our current findings did not show a significant ear-preference for Ss below 6-mo. post-cerebral insult, suggesting the importance of time in determining earpreference of our aphasic Ss. Indeed, when viewed together, these findings appear to support the inference that there is a shift in hemispheric dominance for linguistic perception and processing of auditory-linguistic stimuli following a left cerebral insult which has resulted in aphasia. Our current data on the perception of auditory-linguistic stimuli appears to compliment the already accumulated body of data which stresses the importance of the right hemisphere in linguistic processing following left cerebral insult resulting in aphasia (Smith, 1971). However, a temporal factor appears to be related to this shift in hemispheric perception and processing of auditory stimuli, in that aphasic Ss below 6-mo. post-cerebral insult did nor show a significant left ear-preference under auditory dichotic stimulation. This finding may suggest that the recovery of language processes seen in aphasic Ss during the first 6 or 7 mo. post-cerebral insult (Darley, 1972) may reflect this temporal factor which appears related to the shift in hemispheric dominance for auditory-linguistic stimuli in aphasic Ss. In the present investigation no attempt was made to match the ages of the aphasic and control Ss. It is possible that older control Ss who have a negative history of brain pathology would show a less pronounced preference for the right ear than did our control group of normal young adults. However, that Pettit's ( 1969) control group which ranged in age from 22 to 77 yr. ( M = 42 yr.) demonstrated the same significant superiority of the right ear as our control Ss indicates some consistency in findings for the right ear-preferences of normal nonpachological Ss regardless of age. Nonetheless, future researchers may explore the importance of matching aphasic and control Ss for age. Statistical analyses indicated that aphasic Ss obtained statistically equivalent mean numbers of correct responses for both the oral- and pointing-response tasks. The finding that aphasic Ss do comparatively better on non-verbal response tasks of recognition than on oral tasks of verbal recall (Keenan, 1968) was not supported here when dichotic listening procedures were utilized. These
386
w. H. MOORE,
JR
W. E. WEIDNER
discrepant findings may relate to the relatively concrete word meanings, high frequency of occurrence, and redundancy of the stimulus words employed in the current investigation. However, the present findings are supportive of Schuell, et al.'s (1969) observations that intensive auditory stimulation and repetitive sensory stimulation facilitate evocation of correct verbal responses from aphasic patients. REFERENCES BERLIN,C. I., & LOWE, S. S. Temporal and dichotic factors in central auditory testing. In J. Katz (Ed.), Handbook of clinical audiology. Balcimore: Williams & Wilkins, 1972. Pp. 280-312. BORKOWSKI,J. G., SPREEN, O., & S T U ~J., Z. Ear preference and abstractness in dichotic listening. Psychonomic Science, 1965, 3, 547-548. BROADBENT,D. E., & GREGORY,M. Accuracy for speech presented to the right and left ears. Quarterly Jot/rnal of Experimentaj Psychology, 1964, 16, 359-360. BRYDEN,M. P. Ear preference in auditory perception. Iournal of Experimental Psychology, 1963, 65, 103-105. COOPER,A,, ACHENBACK.K., SATZ, P., & LEVY,C. M. Order of report and ear asymmetry in dichotic listening. Psychonomic Science, 1967, 9, 97-98. DARLEY,D. L. The efficacy of language rehabilitation in aphasia. Journal of Speech and Hearing Disorders, 1972, 37, 3-21. DIRKS, D. Perception of dichotic and monaural verbal material and cerebral dominance in speech. Acta Otolaryngologica, 1964, 58, 73-80. UMW, D. Cerebral dominance and the perception of verbal stimuli. Canadian lournal o f Psychology, 1961, 15, 166-171. ( a ) KIMURA.D. Some effects of temporal-lobe damage on auditory perception. Canadian Journal o f Psychology, 1961, 15, 156-165. ( b ) 1963. KIMURA,D . A note on cerebral dominance in hearing. Acta Oto~aryngo~ogica, 56, 617-619. HINES,D. Bilateral tachistoscopic recognition of verbal and non-verbal stimuli. Cortex, 1972, 7, 313-322. MCKEEVW, W. F., & HULING, M. D . Bilateral tachistoscopic word recognition as a function of hemisphere stimulated and interhemispheric transfer time. Neuropsycbologia, 1971, 9, 281-288. ( a ) MCKEEVER, W. F., & HULING,M. D . Lateral dominance in tachistoscopic word recognition performances obtained with simultaneous bilateral input. Neuroprychologia, 1971, 9. 15-20. ( b ) MCKEEVER,W. F., SUBWI, J., & VAN DEVENTER,A. D . Fixation control in tachistoscopic studies of laterality effects: comment and data relevant to Hine's experiment. Cortex, 1972, 8,473-479. KEENAN,J. S. The nature of receptive and expressive impairments in aphasia. Journal of Speech and Hearing Disorders, 1968, 33, 20-25. OXBURY. J. M., & OXBURY, S. M. Effect of temporal lobectomy on the report of dichotically presented digits. Cortex, 1969, 4, 3-14. PETTIT, J. M. Cerebral dominance and the process of language recovery in aphasia. Unpublished doctoral dissertation, Purdue Univer., 1969. SCHUELL,H. Differential diagnosis of aphasia with che Minnesota test. Minneapolis: Univer. of Minnesota. 1965. SCHUELL,H., JENKINS,J. J., & JIMENEZ-PABON, E. Aphasia in adults: diagnosis, prognosir, and treatment. New York: Hoeber Medical Division, Harper, 1969. SHANKWEILER, D . P. Effects of temporal lobe damage on the erception of dichoticall presented melodies. Journal of Comparative and Pbysiofogical Psychology, 1966): 62, 115-119. SMITH, A. Objective indices of severicy of chronic a hasia in stroke patients. Journal of Speech and Hearing Disorders, 1971, 36, 12;-207. WINER, B. J. Statistical principles in experimental design. New York: McGraw-Hill, 1971.
Accepted Decen ber 12, 1974.
