Anomalous Dominance in Down Syndrome Young Adults

ANOMALOUS DOMINANCE IN DOWN SYNDROME YOUNG ADULTS Susan Giencke and Lawrence Lewandowski (Syracuse University)

A growing body of research suggests that a higher than normal proportion of individuals with developmental disorders demonstrates atypical lateralization patterns on measures such as handedness, dichotic listening, visual half field recognition, electrophysiological response, and cytoarchitecture. The disorders most often cited in regard to anomalous dominance are dyslexia (e.g., Galaburda, 1983), autism (e.g., Prior and Bradshaw, 1979), Turner Syndrome (e.g., Waber, 1979), and mental retardation, particularly Down Syndrome (e.g., Zekulin­ Hartley, 1978). The fact that certain developmentally disabled individuals may have compromised neurological integrity resulting in deviant brain organization and function is widely accepted. However, many questions remain regarding whether such compromised neurological integrity can be specifically attributed to or result in structural hemispheric dysfunction with consequent influences on the processing of information. One group of developmentally disabled persons which has received rather limited attention is that of Down Syndrome (DS) individuals. DS is a specific genetic disorder which encompasses a relatively homogeneous group. As a group, DS subjects have demonstrated atypical hem­ ispheric lateralization, particularly on dichotic listening tasks. An early study by Sommers and Starkley ( 1977) using dichotic listening (DL) demonstrated bilateralization of linguistic processing that was atypical from nonretarded subjects. A further analysis of the Down Syndrome sample into those with greater or lesser linguistic skill revealed that the more severe the linguistic deficit the more atypical the performance on the DL task (more of a tendency toward bilateral performance). Other DL studies (Hartley, 1981; Pipe, 1983; Zekulin-Hartley, 1978) have found atypical results as well, but not bilaterally. All of the above cited studies have found LEAs in the DS population of comparable magnitide to the REA's found with normal control subjects. The only exception to these findings is a study by Tannock, Kershner and Oliver (1984) who found REAs in a DS group. While there is great variability in age group tested (Hartley's subjects were 3-5 year olds, Zekulin-Hartley's subjects were 8-12 and Pipe's subjects were 9-15) and level of intellectual functioning (from mild to moderate levels of retardation) it is clear that an atypical pattern (LEA) is generally found for the DS population on DL tasks. Explanations of these atypical results are not clear. One explanation is Cortex, (1989) 25, 93-102

94

Susan Giencke and Lawrence Lewandowski

that functional memory retrieval strategies are deficient. Influences on retrieval include attention bias (Obrzut, Hynd, Obrzut and Pirozzolo, 1981), task com­ plexity such as memory load (Geffen, 1978) and type of information to be recalled (MoscoVitch, 1979). If such factors as attention and task complexity can be controlled, it is possible to ascertain their influence on DL performance differences with DS subjects. Another explanation for the atypical LEAs found with DS subjects suggests an underlying structural cerebral abnormality. To the extent that a structural deficit exists it should be relatively stable across variability in task influences such as attention bias and level of task complexity. Zekulin-Hartley (1978) attempted to address attentional bias by providing for both a free and cued recall condition. Under the free recall condition a clear atypical LEA was found with DS subjects, whereas the cued condition yielded bilateral performance for all subjects (nor­ mal, Down and nonDown retarded subjects). Tannock, Kershner and Oliver (1984), testing DS subjects with only a blocked cued condition, found typical REAs in their subjects. This is the one recent DL study which has found typical lateralization in DS subjects. Some information processing theorists argue that if atypical ear advantages to linguistic stimuli can be changed so easily to bilateral and/or more typicall right ear advantages, then it is unlikely that an underlying structural deficit exists (Friedman and Polson, 1981; Moscovitch, 1979). Other researchers such as Pipe (1983), and Pipe and Beale (1983) suggest that if subjects are adequately trained to equal entry level skills in such areas as understanding directions and attending to task, one can more clearly determine the relationship between atypical later­ ality and an underlying structural deficit. Using such pretest training they still found atypical LEAs with DS subjects. They also suggest that the bilateral results obtained by Sommers and Starkley (1977) were due to random guessing when task complexity became too difficult. This factor also may account for Zekulin­ Hartley's (1978) findings of bilateral performance with random cueing, but does not explain the REA results of Tannock, Kershner and Oliver (1984). Research with normal subjects also suggests a loss of ear advantage under conditions of greater task complexity (Geffen, 1978). The purpose of this study was to test the stability of the atypical ear advantage found with Down syndrome subjects under various demands of attention and memory. This was done by employing both free and cued recall conditions and presenting three levels of list length for a total of six experimental conditions. However, unlike Zekulin-Hartley (1978) who used random cueing, this study employed alternate cueing (predictable back and forth ear to ear). It was found on pilot testing that random cueing resulted in much confusion for all retarded subjects as they were unable to ancitipate to which ear they should attend. Alternate cueing provided subjects with more anticipatory attending but still controlled for attention bias since the opportunity to block one or the other ear totally (an option in cued recall) was not available to subjects under this con­ dition. Three levels of task difficulty were employed to ascertain the influence of memory on ear advantage. Early evidence by Zekulin-Hartley (1978) suggests that even Down syndrome subjects who could not respond to more than one digit

Anomalous dominance in Down syndrome

95

pair per trial nonetheless demonstrated the atyPical left ear responses he found with multi-level responders. This study predicted that similar findings of left ear advantages would be demonstrated in DS subjects across conditions of attention and memory load. If such a prediction is stable across such factors, it is inferred that structural anomalous dominance is a more likely explanation for left ear advantages than either attention bias or other information processing demand such as memory. Secondarily, atypicallateralization in a young adult age group would not likely be explained as a function of cerebral immaturity, an explanation offered for pre­ vious findings with DS children. MATERIALS AND METHOD

Suf;Jjects Two groups of mentally retarded subjects and two groups of normal subjects were initially enrolled in the stuty. The group of age matched normals performed at or near ceiling on four of the six dichotic listening task conditions, and were excluded from the study. The remaining three groups were ten young adults with Down syndrome, ten retarded adolescents of comparable age, gender, handedness, and IQ level, and ten youngsters between the ages of five and eight years of normal intelligence and comparable to the other groups in terms of gender, handedness, and mental age. All subjects were caucasian, male, and right-handed. All retarded subjects had recent standardized intel­ ligence test scores ranging between 40 and 60. Because some of these subjects lived in group homes, socioeconomic status could not be adequately determined and matched. An attempt was made to select subjects from similar geograhic settings so as to balance the socioeconomic and geographic distributions. Handedness was determined by teacher report and through writing and drawing samples. The two groups of retarded young adults were quite similar withstanding the etiology of retardation. Etiologies of the NonDown (ND) group included such things as prenatal encephalopathy, infant seizure disorders, and early childhood encephalopathy. For these two groups the mean ages were 17.7 years forDS and 17.5 yearsforND, and the range of IQ scores was 40-55 forDS and 42-57 for ND. The mean age of the normal groups was 6.99 years (S.D. = 1.01). The chronological ages of these average children were comparable to the mental age equivalent of the retarded groups as measured on the Peabody Picture Vocabularuy Test- Revised (mean = 6.84 yr., S.D. = 1.26 forDS group; mean = 6.9 yr., S.D. = .94 for ND group). Only subjects with pure tone hearing within normal limits were included. In addition, retarded subjects were administered impedance tests to screen for their higher than normal number of hearing problems. Two retarded individuals were excluded from the study for this reason.

Stimuli The dichotic listening tape was designed for this study and developed by the exper­ imenters at Haskins Laboratories (New Haven, Ct.). The stimulus material consisted of computer generated, synthesized and randomized digit pairs (digits 1-6) presented at three levels of complexity (single, double, or triple digit pairs). The tapes were played on a TEAC X-3 stereo recorder. It had been calibrated for the particular dichotic tape by Stereo Lab (Cambridge, MA), to within one decibel (dB) difference between each channel. The recorder also was checked for range of dB differ­ ences between each output. This range was 0.3 dB between stimuli. At the end of all data collection the recorder was rechecked and the readings indicated instrument precision. Subjects received stimulus material through TDH headphones with MX9 cups. These cups were matched for equal calibration curves ensuring the same power level into each cup and

96

Susan Giencke and Lawrence Lewandowski

therefore equal sound pressure levels. Channel intensity was monitored via calibration tones at the beginning of each series of digits. Headphones were not reversed in this study as equal output had already been determined by calibration and pilot testing.

Procedure The dichotic listening tasks were conducted in a single session by an experimenter in a quiet room of a school. Hearing testing and the PPVT-R were given one week prior to the experimental session. Prior to the dichotic task each subject was asked to count aloud from one to six, and to repeat a series of three digits presented monaurally. All subjects were able to recall up to three digits within ten seconds. There were three practice trials at the beginning of each digit series for each recall condition. Anyone who did not correctly identify stimuli on two of the three practice trials was excluded. One subject failed to reach this criterion. Instructions to the subjects were kept simple. They were told: "You are going to hear some numbers. Tell me what you hear". Then the single pair trials were presented. Next they were told: "Now listen to the numbers again. This time I will point to the ear that I want you to listen with. Pay attention because we will be changing ears back and forth". Then the single digit pairs were played again under this cued recall condition. On the two and three digit pair series the subject was told: "This time there will be more numbers. Listen to all the numbers and then tell me what you hear. It will be a little harder. Do your best". Each free recall segment :was followed by a cued recall se~ent of the same number of digits. In the cued conditions, cueing was alternated between ears and not randomized. During pilot testing this was found to be a more reliable and less confusing procedure for retarded subjects. Fifteen trials at each of the three levels of stimulus complexity were presented to each subjects twice, once requiring free recall and another time requiring alternate responses from each ear (cued recall). Subjects completed 90 trials, or 45 trials per recall condition. Digits in each pair were presented simultaneously to each ear with a one second gap between each pair and ten seconds between each trial. The verbal responses of the subjects were written down in the exact order given and later transformed to tally sheets. For the free recall condition the four possible responses were right ear correct, left ear correct, both ears correct, or error (no report or digit not in trial). Since an error and a "both correct" responses do not influence the laterality index, the free recall condition is considered a binomial response condition (Bryden and Sprott, 1981). For the cued recall condition the possible responses were: correct when cued right, correct when cued left, incorrect when cued right, incorrect when cued left, and error. This cued condition is a multinomial response condition as defined by Bryden and Sprott, since the sources of error, incorrect localizations, do influence the laterality measure. REsULTS

Pilot testing with nine retarded adolescents was performed at two week intervals. Retest reliability coefficient of .72 and .92 were obtained for free and cued conditions respectively. All subjects were capable of performing the task with no ceiling or floor effects. Based on these findings the stimulus materials were considered adequate for the purposes of the study. The raw experimental data were expressed as a function of log-odds ratios (lambda), following the procedures outlined by Bryden and Sprott (1981). A .5 correction factor was added as a constant to all raw scores in order to eliminate an ear score of zero. Fleiss (1981) suggests adding this correction factor to avoid an odds ratio of zero. The bias to one ear or the other was expressed as the difference

Anomalous dominance in Down syndrome

97

between the two lambda scores of the .right and left ears. If the proportion of right ear responses was greater, the laterality index was positive, and if the proportion of left ear responses was greater, the lambda index was negative. This index is less mathematically constrained than simple proportions as it is independent of overall levels of accuracy (Bryden and Sprott, 1981). Since this study involved the testing of normals, the dependent measure lambda (~.) was selected to minimize effects of performance level differences. Data from the free recall condition were subjected to a binomial procedure as follows: Xr }... =In­ XI where Xr is the number of correct right ear responses and XI is the number of correct left ear responses, and In is the natural log of right divided by left respones. The data from the cued recall condition were subjected to a multi­ nomial procedure as follows: }... = In Xr(n-Xl)

Xl(n-Xr)

where n-Xl is the number of incorrect right ear responses, and n-Xr is the number of incorrect left ear responses. These procedures are further detailed by Bryden and Sprott (1981 ). Mean lambda scores for each group across the three free recall conditions are presented in Table I. Note the negative lambda scores for the DS group, indi­ cating a left ear listening advantage. Also, note the high variance for the normal group under the single digit presentation. This is a result of ceiling effects which make those data unreliable. Chi square analyses for the stimulus presentations TABLE!

Mean Lambda Scores, Variance and Chi Squares for the Free Recall Condition at the Three Levels of List Length

List length Single digit

Double digit

Triple digit

Lambda

Variance

Chi square

p

Down NonDown Normal Total

-0.22 0.25 0.17 0.05

.038 .031 .231 .016

20.18 14.55 2.29 40.25

.025

Down NonDown Normal Total

-0.15 .56 .14 .17

.019 .022 .032 .008

30.12 27.28 12.06 81.64

.005 .005

Down NonDown Normal Total

-0.33 .50 .33 .16

.014 .015 .017 .005

54.10 56.29 25.58 162.00

.005 .005 .005 .005

Group

NS NS .07

NS

.005

Susan Giencke and Lawrence Lewandowski

98

indicate that the overall subject pool cannot be considered as a homogeneous group, and therefore justify the investigation of between group differences. Mean lambda scores for each group across the three cued recall conditions are presented in Table II. The DS group produced negative lambda scores for the single and double digit presentations, indicating left ear listening adavantages. As with the free recall conditions the chi square analyses of all subjects for each presentation revealed heterogeneity of the sample, thus allowing for further analysis of subgroup differences. The mean lambda scores for all three groups, for each of the digit presen­ tations, and for both recall conditions are visually depicted in Figure 1. In four of the six conditions the DS group demonstrated reversed laterality, while in one of

TABLE II

Mean Lambda Scores, Variances and Chi Squares for the Cued Recall Condition at the Three Levels of List Length

List length Single digit

Double digit

Triple digit

Group

Lambda

Variance

Chi square

p

Down NonDown Normal Total

-21. .74 1.05 .13

.064 .081 .084 .052

13.89 19.48 14.42 50.41

NS .025 NS .01

Down NonDown Normal Total

-l.l2 .46 1.05 .27

.033 .043 .045 .027

22.13 37.13 21.51 77.19

.01 .005 .01 .005

Down NonDown Normal Total

.10

.026 .030 .025 .019

35.58 58.92 16.35 115.89

.005 .005 .05 .005

.71 .99 .44

the cued recall conditions the DS group demonstrated weak, if any, right ear advantage. The other groups demonstrated consistent right ear advantages. In order to determine the existence of group differences in dichotic listening performance the mean lambda scores were compared by use of z-score. Table III illustrates the z-scores for group comparisons for each stimulus presentation within each recall condition. Significant differences (p<.Ol) are noted between the DS and ND groups on four of the six comparisons. The Normal group differed significantly from the DS group on four of five comparisons. Only five comparisons could be made because the Normal group encountered ceiling effects on the single digit free recall condition. The Normal and ND groups did not differ significantly in any condition (at the .01 level). Because the free and cued conditions required different formulas to calculate lambda, the laterality index, an analysis of variance procedure was ruled out as a way of comparing performance under the two recall conditions. In order to compare free with cued recall performance, individual lambda scores were

99

Anomalous dominance in Down syndrome

2D \5 l()

~

.5

ID

:2:

:5

00

z

I~ I

~

- I()

I

l.ogond

-\5

O NOAMAL

-2D

t;S) DOWN

NON DOWN SINGLE

DOUBLE

TRIPLE

-FREE RECALL­

Fig. 1 -

SINGLE

DOUEl.E

TA;f'LE

-CUED RECALL-

Group mean lambda scores for each list length of both recall conditions.

TABLE III

Z-Scores of Mean Lambda Group Differences for the Free and Cued recall Conditions for Three Levels of List Length

Comparison Single

1.78* Down vs. NonDown 0.75 Down vs. Normal NonDowm vs. Normal -0.15 *p.<.05 **p.<.Ol

Free recall Triple Double 4.85** 3.48** 3.70** 1.26 -0.99 -1.82*

Single 2.52** 3.29** 0.75

Cued recall Double 1.20 3.30** 1.99*

Triple 2.55** 3.94 1.21

transformed to deviation scores by subtracting the grand mean from each sub­ ject's lambda score for each of the six experimental conditions. A correlated t-test between free and cued performance was determined by comparing the deviation scores for a given digit length within a group. The nine such t-tests are presented in Table IV. The only group that demonstrated a consistent and significant difference between recall conditions was the ND group. For the other two groups no significant effect of cueing was found.

100

Susan Giencke and Lawrence Lewandowski TABLE IV

Correlated t-Scores between the Free and Cued Recall Conditions at the Three Levels of List Length Group Down NonDown Normal

Single digit 1.81 -2.57* 2.50*

Double digit

Triple digit

1.97 -2.35* 1.20

0.57 -2.51* 1.29

*p.<.05 Note: The t-scores were determined based on the deviation scores of each subject relative to the total group for each experimental condition. A positive sign indicates an increase in right ear bias; a negative sign indicates a decrease in right ear bias.

DISCUSSION

The central hypothesis addressed by this study was that cerebrallateralization for linguistic information would be associated with the left hemisphere (REA) for both normal subjects and the ND retarded subjects, whereas cerebrallaterali­ zation for language would be associated with the right hemisphere (LEA) for the DS subjects. This hypothesis was confirmed. The findings of the usual processing patterns (REA) in a normal control group helped to validate this study's exper­ imental materials and procedure. The atypical (LEA) pattern found with the DS subjects was robust and consistent across most of the experimental conditions of attention and memory. This result could not likely be attributed to experimental error or statistical artifact. In addition, this study's design provided for a con­ servative test of the anomalous dominance theory, since only right handed subjects participated. These are individuals in the DS population who are least likely to show anomalous dominance for speech. Left handed individuals are more likely to demonstrate anomalous dominance and are found in greater proportion in the mentally retarded population when compared to the general population. All previous such studies of DS subjects had not eliminated left handed subjects. By only using right handed subjects, any anomalous dominance attributed to left handedness was eliminated, thus eliminating a characteristic bias in the mentally retarded population. Even with such a conservative approach in testing the DS retarded subjects, the anomalous dominance for language was readily apparent. If one supports the notion that the results of verbal dichotic listening tasks reflect cerebrallateralization for language, then one could infer that DS subjects possess a different or unique type of cerebral organization for language processing. These results are consistent with the findings in studies by Zekulin-Hartley (1978) for free recall conditions and Hartley (1981) and Pipe (1983) for cued recall. Possibly related to anomalous dominance, DS subjects have a higher inci­ dence of language disorders over other retarded persons (Keane, 1972; Rohr and Burr, 1978; Zisk and Bailer, 1967). These findings may indicate deficient left hemisphere functioning. DS subjects are poorer at stereognostic, kinesthetic, and tactile shape discrimination, yet have better visual spatial discrimination relative to other retarded subjects (Zekulin-Hartley, Gibson, Mosely and Brown, 1974). Findings such as these add to the correlational data supporting the Geschwind and Galaburda theory of anomalous dominance, which suggests relative

Anomalous dominance in Down syndrome

101

strengths in right hemispheric functioning often accompanying left hemisphere deficiencies. In the present study an attempt was to examine for differences in laterality performance as a result of different attention conditions (free or cued). Regard­ less of attention condition, the normal group demonstrated the greatest right ear Sllf'H'ri 'rity and the DS group demonstrated the greatest deviation from right ear sup,cn )rity. Thus, anomalous dominance as indicated by nonright ear asymmetry w.~" demonstrated across the free and cued conditions. Although task factors did cause some fluctuation in ear performance, subjects did not change in their relationship to one another. That is, DS subjects remained consistently lowest in right ear asymmetry when compared to other subjects. In addition to attention, the variable of stimulus list length which varied memory demand seemed to have little effect on the overall results. Group per­ formances and relative ear asymmetries were maintained across the three levels of stimulus complexity. The consistency of results across attentional and memory demand suggests that the finding of anomalous dominance in Down Syndrome is reliable. Consequently we feel that the reported findings of left ear asymmetry in DS individuals cannot be explained by information processing factors of atten­ tion and memory, but rather, are best explained by the "anomalous dominance" theory of Geschwind and Galaburda ( 1985) which assumes early structural and functional deviation in brain development and organization. This theory, as well as our data with young adults, rejects a maturational lag explanation for atypical lateralization in the retarded. It would seem that this theoretical approach may be fruitful in reaching a better neuropsychological understanding of Down Syn­ diOme and other developmental disorders. ABSTRACT

The ear advantages of groups of Down Syndrome and developmentally retarded (NonDown) young adults, and normal youngsters matched for mental age were compared on dichotic listening performance. The paradigm employed strings of single, double, and triple digits presented to each ear under both free and cued recall conditions. The developmentally retarded and normal groups demonstrated the typical right ear advan­ tage (REA), whereas the Down Syndrome group produced a significant left ear advantage (LEA) in four of the six experimental conditions. In addition, for the cued as compared to free recall conditions, all three groups demonstrated relatively better right ear perfor­ mance. These results indicate anomalous dominance in Down Syndrome young adults which is consistent across varying memory load and attentional demands. Furthermore, these results are not likely due to a maturational lag phenomenon, but more likely related to genetic, biologic, and neurologic, factors as suggested by Geschwind and Galaburda (1985). REFERENCES

BERMAN, A. The problem of assessing cerebral dominance and its relationship to intelligence. Cortex, 7: 372-386, 1971. BRYDEN, M.P., and SPROTI, D.A Statistical determination of degree of laterality. Neuropsychologia, 19: 571-580, 1981. FLEISS J.L. Statistical Methods for Rates and Proportions. New York: John Wiley, 1981. FRIEDMAN, A, and PoLSON, M.C. Hemispheres as independent resource systems: Limited capacity processing and cerebral specialization. Journalof Experimental Psychology: Human Perception and Performance, 5: 1031-1058, 1981. GALABURDA, A.M. Developmental dyslexia: Current anatomical research. Annals of Dyslexia, 33: 41-53, 1983.

102

Susan Giencke and Lawrence Lewandowski

GEFFEN, G. The development of the right ear advantage in dichotic listening with focused attention. Cortex, 14: 169-172, 1978. GESCHWIND, N., and GALABURDA, A.M. Cerebrallateralization: Biological mechanisms, associa­ tions, and pathology: I. A hypothesis and a program for research. Archives of Neurology, 42: 428-459, 1985. HARTLEY, X. Hemispheric asymmetry in Down's Syndrome children. Canadian Journal ofBehavioral Science; 13: 210-217, 1981. KEANE, V.E. The incidence of speech and language problems. Mental Retardation; 10: 3-8, 1972. KERSHNER, J.R. Lateral preference and ability to conserve multiple spatial relations by mentally retarded children. Perceptual and Motor Skills, 35: 151-152, 1972. McMANUS, I.C. The interpretation of laterality, Cortes, 20: 421-426, 1984. MoscoviTCH, N. Information processing. In M.S. Gazzaniga (Ed.), Handbook of Neurobiology­ Neuropsychology. New York: Plenum Press, 1979. 0BRZUT, J.E., HYND, G.W.M 0BRZUT, A., and PiROZZOLO, F.J. Effect of directed attention on cerebral asymmetries in normal and learning disabled children. Developmental Psychology, 17: 118-125, 1981. PICKERSGILL, M.J., and PANK, P. Relation of age and mongolism to lateral preference in severely subnormal subjects. Nature, 228: 1342-1344, 1970. PIPE, M.E. Dichotic listening performance following auditory discrimination training on Down's Syndrome and developmentally retarded children. Cortex, 19: 481-491, 1981. PIPE, M.E., and BEALE, I.L. Hemispheric specialization for speech in retarded children. Cortex, 21: 91-98, 1983. PRIOR, M.R., and BRADSHAW, J.R. Hemispheric functioning in autistic children. Cortex; 15: 73-81, 1979. RENGSTORFF, R.H. The types and incidence of hand-eye preference found among mentally retarded children. American Journal of Optometry andArchives ofthe American Academy of Optometry, 45: 657-659, 1968. RoHR, A., and BURR, D.B. Etiological difference in patterns of psycholinguistic development of children of IQ 20 to 60. American Journal of Mental Deficiency, 82: 549-553, 1978. SOMMERS, R.K., and STARKLEY, K.L. Dichotic verbal processing in Down's syndrome children having qualitatively different speech and language skills. American Journal of Mental Deficiency, 82: 44-53, 1977. TANNOCK,R., KERSHNER, J., OLIVER, J. Do individuals with Down's Syndrome possess right hem­ isphere language dominance? Cortex, 20: 221-231, 1984. WABER, D.P. Neuropsychological aspects of Turner Syndrome. Developmental Medicine and Child Neurology, 21: 58-70, 1979. ZEKULIN-HARTLEY, X. Hemispheric asymmetry in Down's Syndrome children. Doctoral dissertation, University of Toronto, 1978. ZEKULIN-HARTLEY, X., GIBSON D., MOSELY, J.L., and BROWN, P.L. American Journal of Mental Deficiency, 78: 571-577, 1974. ZISK, P., and BAILER, I. Speech and language probh;ms in mongolism: A review of the literature. Journal of Speech and Hearing Disorders, 32: 228-241, 1967. L.J. Ll!wandowski, Ph. D., Dept. Psychology, Syracuse University, Syracuse, N.Y. 13244-2340, U.S.A.