This article was downloaded by: [Universitaetsbibliothek Giessen]On: 27 January 2015, At: 06:11Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

AphasiologyPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/paph20

On the neurological status of speechautomatisms and its significance forneurolinguistic modelsClaus-W. Wallesch a , J. Christian Haas a & Gerhard Blanken aa Department of Neurology , Freiburg UniversityPublished online: 29 May 2007.

To cite this article: Claus-W. Wallesch , J. Christian Haas & Gerhard Blanken (1989) On theneurological status of speech automatisms and its significance for neurolinguistic models, Aphasiology,3:5, 435-447, DOI: 10.1080/02687038908249005

To link to this article: http://dx.doi.org/10.1080/02687038908249005

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilities whatsoever orhowsoever caused arising directly or indirectly in connection with, in relation to or arisingout of the use of the Content.

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

APHASIOLOGY, 1989, VOL. 3, NO. 5, 435-447

On the neurological status of speech automatisms and its significance for neurolinguistic models

CLAUS-W. W A L L E S C H , J . C H R I S T I A N H A A S , and G E R H A R D B L A N K E N Department of Neurology, Freiburg University

(Received 7 June 1988; revised 5 August 1988; accepted 25 August 1988)

Abstract

Forty-five aphasic patients, who had suffered a left middle cerebral infarction involving more than 2% of forebrain volume at least 4 months previously, were investigated for the presence of speech automatisms and their accompanying neurological and neuropsychological symptoms. Fourteen patients were found to produce automatisms. The presence ofthis symptom was correlated, although not closely, with degree of buccc-facial apraxia, but not significantly with lesion size. The correlation with Token Test scores was insignificant when the effects of ideomotor apraxia had been partialized. We interpret our observations on the basis of a theory that speech automatisms are a symptom of primary or secondary severe dysfunction in the periphery of the language production apparatus in the vicinity of processes of speech motor planning and execution.

Introduction

Speech automatisms have been defined as repetitive, invariant utterances consisting of neologistic syllables, arbitrary words or phrases which neither lexically nor syntactically fit into context, and which are produced contrary to the patient’s presumed intention (Huber, Poeck, Weniger and Willmes 1983). If automatisms predominate a patient’s speech production, as is frequently the case, the term ‘recurring (or recurrent) utterances’ may be preferred. Although speech automatisms have intrigued neurologists and speech pathologists since they were first described in detail by Hughlings-Jackson (1 879-80), controversy still remains concerning their neurological, as well as their neurolinguistic status.

In contrast to perseverations, echolalia, or parapraxias, speech automatisms have no obvious counterparts in motor disturbances, the closest resemblance occurring to uniforrne Unruheerscheinungen (stereotyped actions in pathological restiveness) as described by Kleist (1934), which are rarely encountered in clinical neurological practice today. Speech automatisms, on the other hand, are a frequent symptom in aphasic patients. Hughlings-Jackson (1879-80) specifically stated that patients with recurring utterances suffered from extensive left hemisphere destructions, which resulted in a disinhibition of right hemisphere nervous arrangements.

The next eminent neurologist who analyzed speech automatisms in some detail was Alajouanine (1956). He called them ‘permanent verbal stereotypes’, the

Address for correspondence: Dr C. W. Wallesch, Department of Neurology, Hansastr 9, 7800 Freiburg, FR Germany.

0268-7038/80 $3.00 0 1989 Taylor & Francis Ltd.

Dow

nloa

ded

by [

Uni

vers

itaet

sbib

lioth

ek G

iess

en]

at 0

6:11

27

Janu

ary

2015

436 C.- W. Wallesch, J . C. Haas, and G. Blanken

occurrence of which he saw as being closely linked to motor aphasia, hemiplegia, and bucco-linguo-facial apraxia. Alajouanine observed, as Hughlings-Jackson had done earlier, that patients with speech automatisms regularly suffered from severe deficits in other language modalities.

Neurolinguistic theory has advanced mainly three functional explanations for the production of speech automatisms:

(1) They were considered to result from a severe breakdown of the language system and thus to be a key symptom of global aphasia (Stachowiak, Huber, Kerschensteiner, Poeck and Weniger 1977, Brunner, Kornhuber, Seemiiller, Suger and Wallesch 1982).

(2) Already Broca (1861) had proposed that speech automatisms were generated in an articulatory cerebral component. More recently, Blanken, Dittman, Haas and Wallesch (1988) suggested that speech automatisms indicate a severe dysfunction in the periphery of the language production system close to speech motor functions. This dysfunction could result either from a primary dysfunction of the peripheral component or from a disconnection and disinhibition of this component due to damage to more central processes. The latter mode of automatism generation would correspond to the ‘mode of inhibition’ of Haag, Huber, Hiindgen, Stiller and Willmes (1985).

(3) In their more anatomically oriented interpretation, Haas, Blanken, Mezger and Wallesch (1 988) speculated-as one alternative hypothesis to explain their finding of automatism production being linked to age-that the presence of diffuse brain damage or dysfunction, together with the focal left hemisphere lesion, were prerequisites for speech automatisms to occur. They assumed, as had Haag et al . (1985) before that disinhibition and breakdown of control accounted for automatism production and release.

Code (1987) stressed a distinction between two types of speech automatisms, namely ‘non-meaningful recurring utterances’ (NMRU) and ‘real-word recurring utterances’ (RWRU). He speculates that the former may result from “a severely compromised left hemisphere phonological system without any hemisphere- subcortical input” whereas in the production of the latter phenomenon he considered the right hemisphere to be heavily implicated. For the production of NMRUs, which are more frequent, his position therefore is a variant of (2 ) . Considering Code’s line of argument, RWRUs could result from one of two pathological conditions, namely (a) severe damage to the left hemisphere language processing apparatus, or (b) a left hemisphere lesion producing nonfluent aphasia in cases of less pronounced lateralization. RWRUs were not a prominent symptom in the patients investigated in the present study.

Positions (1) and (2) are cognitive hypotheses, whereas (3) is basically a pathophysiological one. These positions are not mutually exclusive: a severe trauma to central linguistic processes may disinhibit peripheral production processes, and its large causative brain lesion may lead to diffuse dysfunction. This is obviously a frequent clinical constellation. As Blanken, Dittmann and Wallesch (in press) point out, the exceptions to this pattern, especially are able to shed light upon the cognitive pathogenesis of automatism production.

Each of the three main positions described above leads to prediction with respect to clinical symptomatology:

Dow

nloa

ded

by [

Uni

vers

itaet

sbib

lioth

ek G

iess

en]

at 0

6:11

27

Janu

ary

2015

Neurological status of speech automatisms 437

(a) if speech automatisms were the result of severe language breakdown, their presence should be regularly associated with other, e. g. receptive, language deficits;

(b) if speech automatisms reflected pathology in the very periphery of language production, then their presence should be linked with other deficits of cerebral output systems such as ideomotor apraxia or paresis;

(c) if they reflected overall cerebral pathology, the clinical neurological and neuropsychological examination should reveal evidence of right hemisphere or diffuse pathology in patients with speech automatisms.

The present study aimed at a comparison of the clinical neurological and neuropsychological symptoms of aphasic patients with and without speech automat- isms. The pattern of deficits associated with automatism production was expected to reflect to some extent its underlying functional pathology.

Patients and methods

Forty-five right-handed patients, who had suffered from a single left middle cerebral artery infarction at least 4 months previously, were investigated. Only patients were included for whom recent CT scans (performed more than 3 months post-stroke, but less than 12 months prior to assessment) were available and showed a single demarcated infarction, but no generalized cerebral atrophy. Lesion size was estimated by use of the method of Blunk, de Bleser, Willmes and Zeumer (1981). The lesions were required to exceed 2% of the forebrain volume. The patients investigated in the present study form part of the group described by Haas et al. (1988), namely those, for whom complete sets of neurolinguistic, neuropsychological, neurological and CT data could be obtained. Details of thepatients’ CT scans are being published elsewhere (Haas et al., 1988).

For neurolinguistic analysis and syndromatic classification the Aachen Aphasia Test (AAT, Huber, Poeck, Weniger, Willmes, 1983, Huber, Poeck and Willmes 1984) was applied. Syndrome classification was performed by the nonparametric discriminant analysis (ALLOC, Habbema, Hermanns and van den Broek 1974; compare Huber, Poeck and Willmes 1984) that is part of the AAT. Speech automatisms occurred only among globally aphasic patients, of whom 14 exhibited this symptom. If automatisms were present, they made up for more than 50% of syllables in spontaneous speech production (median 96%, range 54100%). In all automatism patients neologistic automatisms (‘non meaningful recurring utterances’; Code 1982, 1987) predominated the stereotypical productions (e. g. ‘widdi-widdi’, ‘werri-werri’, ‘do-do-do’, ‘natatatata’), and three patients produced real word automatisms in addition. Clinical and demographic data of the patients with and without automatism production are given in table 1.

For a detailed analysis of neurological and neuropsychological symptoms two groups of 12 patients each were drawn, one of which comprised patients who produced speech automatisms and the other patients in whom this symptom was absent. These groups of patients were selected to be comparable with respect to age, sex, duration of illness, lesion size, and configuration of the cortical and basal ganglia lesion (table 2).

That such matching for lesion size and configuration was at all possible indicates

Dow

nloa

ded

by [

Uni

vers

itaet

sbib

lioth

ek G

iess

en]

at 0

6:11

27

Janu

ary

2015

438 C.- W. Wallesch, J . C. Haas, and G. Blanken

Table 1. Clinical and demographic data of 14 patients with or without automatism production.

Automatism (+) Automatism (-)

N 14 31 Sex (male female) 1113 19/12 ALLOC classification of syndrome 14 global 12 global; 14 Broca; 3 Wer-

nicke; 2 nominal/residual

Median Range Median Range

Age (years) 58.0 52-74 54.0 2+76*

Time since stroke (months) 16.0 4 8 0 20.0 4-84

AAT results (percentile ranks) Token Test 16 5-94 38 7-99 ** Written language 13 5-39 38 5-99 ** Comprehension 13.5 7-91 55 +loo**

lesion size (Yo forebrain volume) 8.0 4.5-18.0 8.1 2.2-21.0

Mann-Whitney U-test: P<O.O5; ** P<O.Ol.

Table 2. Clinical data of two matched collectives of 12 patients each with and without production of speech automatisms.

Automatism (+) Automatism (-)

Male/female 913 8/4

Median Range Median Range

Age years 56.5 52-68 56.0 43-76 Duration (months) 16.0 4-80 18.5 +60 Lesion size

Token Test (YO forebrain volume) 8.0 4.5-18.0 8.8 4.8-17.3

(percentile ranks) 18.0 5-94 39.0 7-81’

Lesions (number of patients in whom the structure was involved) Broca’s area 8 7 Wernicke’s area 9 8 h u l a 11 11 Basal ganglia 8 8

Mann-Whitney U-test: *P<O.O5.

that the CT lesions did not differ markedly between patients with and without speech automatisms (Haas et a l . , 1988; Figures l a , lb). With few exceptions the patients of both parallel groups suffered from large infarctions involving the anterior and middle area of supply of the middle cerebral artery. The lesions included structures in the depth of the hemisphere supplied by the lenticulostriate branches in all cases of patients with automatism production and, consequently, in their matches.

All patients included in the present investigation received a detailed neurological and neuropsychological assessment, performed by a trained neurologist no longer than 4 weeks before or after the application of the AAT. The neurological

Dow

nloa

ded

by [

Uni

vers

itaet

sbib

lioth

ek G

iess

en]

at 0

6:11

27

Janu

ary

2015

Neurological status of speech automatisms 439

c/~-F-= - -=.z:-- ~ -I.~-l> ?-~-'-=' -F+ - b -------- -~ -->

Figure l a . Superimposed CT lesions of 12 patients with middle cerebral infarctions and production of speech automatisms.

Figure lb. Superimposed CT lesions of 12 patients with middle cerebral infarctions comparable for lesion size and configuration, age, sex, without speech automatisms. Single-hatching indicates the lesion shared by four or more patients, cross-hatching by 8 or more patients.

investigation consisted of the analysis of cranial nerve functions, reflex status, detailed assessment of hemiparesis, sensitivity and coordination. The present study includes analysis of the promoters detailed below.

Cranial nerves Facial sensitivity: sensitivity to pinprick, temperature and light touch were

compared with the unaffected side. Scoring was performed in analogy to arm and leg sensitivity (see below).

Facial paresis: forceful closing of the eyes, baring of the teeth, and puffing of the cheeks were investigated. A score of 0 was given if there were no side differences, of 1 if differences could only be felt, and of 2 if they could be seen.

Motor symptoms The strength of the following movements of muscles and muscle groups was

assessed according to the methods and scaling of Riddoch, Bristow, Cairns, Carmichael, Critchley, Greenfield, Learmonth, Platt, Seddon, Symons, Young and Herrald (1 943) :

Upper extremity: ante- and retroversion, abduction and adduction of the shoulder, flexion and extension of the elbow, wrist, and fingers.

Lower extremity: flexion, extension, abduction and adduction of the hip, flexion and extension of the knee, foot, and toes.

For the overall performance over each joint a total maximal score of 20 could be

Dow

nloa

ded

by [

Uni

vers

itaet

sbib

lioth

ek G

iess

en]

at 0

6:11

27

Janu

ary

2015

440 C.- W. Wallesch, /. C. Haas, and G. Blanken

obtained. If only two directions of movement were assessed, the paresis scores were multiplied by 2 (e.g. right knee: extension 3, flexion 2, total s c o r e = 3 ~ 2 + 2 x 2 = 10). Similarly, for each extremity a total maximal score of 80 could be achieved.

Sensory symptoms Sensitivity for pinprick, temperature, vibration, light touch, and joint position

was assessed at distal locations of each extremity. For each modality a score of 0 was given if the patient indicated that there was no discernible difference between normal and affected side, of 1 if there was a discernible difference, and of 2 in the case of a severe deficit. The total maximal score for each extremity was 10.

Ref exes We included the following: biceps, triceps, brachioradial, knee, ankle, and

adductorjerks, Babinski’s, Gordon’s, and Oppenheim’s signs. A score of 0 was given if there was no side difference, of 1 if there were discernible differences (e.g. silent plantar response on one and flexor response on the other side), and of 2 if there were gross differences (e.g. plantar response flexor on one and extensor on the other side, or side differences of more than 2 values on the generally used six-point reflex scale (absent, diminished, below average, average, above average, and increased)). The scores for upper extremity reflexes, lower extremity reflexes, and signs of the Babinski group, respectively, were combined.

Coordination The only measure of coordination included in this analysis was hand-tapping. At

the beginning and at the end of the neuropsychological examination the patient had to tap or beat a lever as fast as possible for 15 seconds of which the last 10 seconds were scored. The hand to start with was randomly assigned in the first assessment, and in the second assessment the patient began with the other. The number of taps for each trial were combined for either hand.

The following tasks were included in the neuropsychological assessment.

Standard Progressive Matrices A and B (Raven 1956).

was scored (maximum 24). These were used as a test of general intelligence. The number of correct solutions

Benton’s Visual Retention Test (Benton 1964), Form E, instruction A. This was used for the assessment of nonverbal short term memory functions and

perseverative tendencies in drawing. We scored the numbers of correct solutions (maximum lo), of total errors, of elisions, and of perseverative errors.

Assessment of ideomotor apraxia according to Poeck (1982). The tasks were: sticking out the tongue; whistling; hissing; smacking; touching

upper lip with the tongue; touching the forehead with the back of the hand; touching the breast with the fist; pointing with the index finger to the knee; positioning of the left foot behind the right; turning the left foot on its tip; and pantomimes ofsmoking a cigarette, brushing the teeth, and kicking a ball. The patient was only required to imitate these movements and gestures. No item-specific verbal instructions were given. The number of parapraxically distorted actions was scored.

Dow

nloa

ded

by [

Uni

vers

itaet

sbib

lioth

ek G

iess

en]

at 0

6:11

27

Janu

ary

2015

Neurological status of speech automatisms 441

Testing for graphic perseveration As devised by Luria (1980) the patient was asked to continue the drawing of a

sequence of alternating squares and triangles. The number of perseverative elements among the first 10 shapes was scored.

For statistical analysis of the data the Mann-Whitney U-test, Kendall’s Tau B and Kendall’s partial rank correlation coefficient were employed (Siegel 1956). Whereas significance of Tau B may be tested by assuming a normal distribution of this value (Siegel 1956), there is no method for testing the significance of Kendall’s partial rank correlation.

Results

Non-oral-expressive language functions

Table 1 includes the results of AAT Token Test, Written Language (which consists of writing to dictation, word assembly to dictation, and reading aloud) and Language Comprehension performance for 14 patients with, and 31 patients without, speech automatisms. As a group, the automatism patients performed significantly worse in those language measures which do not rely on oral expression and were older (for discussion of age effects, see Haas et al. 1988). However, as the large ranges in the Token Test and Comprehension subtests indicate, there were patients among the automatism group who were only mildly impaired in receptive language functions.

Neuro log ica 1 assessment Details of neurological symptoms and neuropsychological impairment were

analyzed by comparison of two groups of 12 patients each, with and without automatism production, who were comparable for lesion size, distribution, age, sex and illness duration (table 2). Table 3 gives the results of the neurological investigation.

Cranial nerves No differences concerning face sensitivity and facial paresis could be demons-

trated. We suspect that the scoring of facial paresis was not precise enough, as in consideration of the results of the assessment of arm and leg paresis and of the quite regular pattern of hemiparesis in middle cerebral infarction with aphasia we would have expected more marked differences between the groups.

Motor symptoms Both the right-sided paresis of arm and leg were more pronounced with the

automatism patients. Differences between groups with respect to intensity and duration of physiotherapy or medication were not present. Signs of extrapyramidal motor disorder such as dystonic or hyperkinetic associated movements were observed in no single patient.

Sensory symptoms

to the patients’ hemisensory deficit. There were only minor differences and wide overlap between groups with respect

Dow

nloa

ded

by [

Uni

vers

itaet

sbib

lioth

ek G

iess

en]

at 0

6:11

27

Janu

ary

2015

442 C.-W. Wallesch, J . C. Haas, and G. Blanken

Table 3. Results of neurological assessment in two groups of 12 patients with and without automatism production. Normal values in bold

Automatism( +) Automatism( -)

Mean Range Mean Range

Cranial nerves facial sensitivity (0-6) facial paresis (0-6)

arm Paresis (0-80)

1%

1%

1%

Sensitivity (C-10) arm

Reflexes (0-6) arm

Babinski group

tapping right hand (20 sec) tapping left hand (20 sec)

Coordination

3.9 3.2

16.5 46.4

7.6 8.1

5.5 4.2 2.5

5.4 35.1

0-6 2-6

C-80 2 6 8 0

0-10 2-10

0-6 0-6 (M

1-48 13-50

3.0 2.8

52.5 62.0

5.9 5.6

3.7 2.0 1.1

24.4 43.0

0-6 0-6

C-80 17-80

0-10 6 1 0

0-6* w * 0-6’

1 4 3 * 22-67

Mann Whitney U-test: = P<0.05; ** = RO.01

ReJexes

group were present. Significant differences in both arm and leg reflexes and the signs of the Babinski

Tapping As could be expected from the results of the paresis assessment, the automatism

patients performed worse in tapping with the right hand. There were only insignificant differences between groups in tapping with the left hand.

Neuropsychological assessment Table 4 gives the results of the neuropsychological investigation for the two

groups of 12 patients with and without automatism production.

Nonverbal intelligence The average Progressive Matrices scores were slightly and insignificantly lower in

the automatism patients. The ranges demonstrate a wide operlap. With few exceptions, the patients’ performance was markedly impaired (norms according to Kratzmeier, 1979; 10th percentile for 60-year-old subjects: 12, 25th percentile: 14).

Benton Visual Retention Test Most patients’ performances were markedly subnormal. There were no signi-

ficant differences between groups. Patients with speech automatisms tended to be more impaired, but again there was a wide overlap. Perseverations constituted only a small proportion of errors in all groups of patients and were not encountered with remarkable frequency in any single case.

Dow

nloa

ded

by [

Uni

vers

itaet

sbib

lioth

ek G

iess

en]

at 0

6:11

27

Janu

ary

2015

Neurological status of speech automatisms 443

Table 4. Results of neuropsychological assessment in two groups of 12 patients with and without speech automatisms.

Automatisms( +) Automatisms( -)

Mean Range Mean Range

Progressive Matrices Benton Test

correct responses errors elisions perseverations

face arm

Assessment of ideomotor apraxia:

1% Graphic perseveration (Yo)

9.7

3.5 17.7 1.7 2.6

1.9 0.6 0.5

28.8

1-22

G 8 4 2 7 0-4 0-5

0-5 G 3 c-3 c-100

12.4

5.0 13.0 2.8 1.2

0.4 0.0 0.1

17.2

7-18

5 8 11-19 0-4 0-3

0-2 0-0’ G1 * 0-1 00

U-test: = P<0.05; ** = P<O.Ol

Assessment of ideomotor apraxia Buccofacial apraxia was more common than ideomotor apraxia with limb actions.

Patients with speech automatisms produced more frequent parapraxic errors when imitating movement sequences involving face or arm.

Graphic perseveration

and no statistically remarkable effects occurred between groups. There were gross inter-individual differences in performing Luria’s (1980) task

Correlation analysis

Kendall’s Tau B was computed and its significance tested for the correlations among the following variables in the complete group of 45 patients: percent of automatisms among syllables of spontaneous speech; patient’s age at infarction; lesion size in percent of forebrain volume; Progressive Matrices, number of correct solutions; token test, percentile rank; bucco-facial apraxia, number of actions performed incorrectly; right arm paresis score (see above); left hand tapping.

This choice was made to test between the three main neurolinguistic hypotheses of automatism generation outlined in the introduction. If hypothesis (1) was correct, where automatisms result from a severe breakdown of central language processes, then the Token Test scores should correlate most closely with automatism production. Hypothesis (2)-automatisms indicating a severe dysfunction in speech motor programming-would predict a correlation between automatism production and bucco-facial apraxia and/or hemiparesis. The last hypothesis (3), where diffuse brain damage is a prerequisite of automatism production, would predict strong effects of the Progressive Matrices scores and tapping with the intact left hand. Age and lesion size were included because of the findings of Haas et al . (1988) that age interacts with automatism production but lesion size does not.

Table 5 gives the Kendall’s Tau B correlation coefficients. The strongest associations with automatism production (P<0.01) occurred with bucco-facial apraxia and the Progressive Matrices scores. The production of automatisms was correlated with more severe bucco-facial apraxia and lower level of performance in

Dow

nloa

ded

by [

Uni

vers

itaet

sbib

lioth

ek G

iess

en]

at 0

6:11

27

Janu

ary

2015

444 C.-W. Wallesch, 1. C . Haas, and G. Blanken

Table 5. Kendall's Tau B correlation coefficients and related Z-scores for interactions among symptoms in 45 aphasic patients.

Age Lesion Progressive Token Apraxia Arm Tapping size matrices Test of face paresis left hand

Kendall's Tau B: Automatisms 0.2135

Lesion size Progressive Matrices Token Test Buccofacial Apraxia Right arm paresis Tapping hand

Age

Z-scores Automatisms 2.07'

Lesion size Progressive Matrices Token Test Buccofacial apraxia Right arm paresis Tapping left hand

Age

0.1224 -0.2986 -0.2586 -0.1418 -0.3956 -0.0747

-0.1233 -0.2648 -0.3734

1.19 2.89** 2.50' 1.37 3.83** 0.72

1.19 2.57** 362*'

0.3910 -0.1480 0.2498 -0.0728 0.1844 -0.2046

-0.4249 -0.2588 -0.4292 0.3006

0.3689

3.79** 1.43 2.42' 0.71 1.79 1.98' 4.11" 2.51' 4.16" 2.91"

3.57'*

-0.1085 -0,1186 -0.1321 -0.2652 -0.1655 -0.2151

0.2570

1.05 1.15 1.28 2.58** 1 6 0 2.08* 2.49'

* = P<0.05 (2 >1.96), ** = P<O.Ol (Z>2.57).

Table 6. Kendall's partial rank correlation coefficients; partialization of Token Test scores. Significance testing not possible (Siege1 1956).

Age Lesion Progressive Facial Arm Tapping size matrices apraxia paresis left hand

Automatisms 0.2016 0.0579 -0.2255 -0.3209 -0.0762 -0.0690 Age -0.3532 -0.3976 -0.2684 -0.0529 -0.1080 Lesion size -0.0244 -0.0812 -0.1359 -0.0928 Progressive Matrices -0.3158 -0.1657 -0.2223 Buccofacial apraxia -0.2794 -0.1618 Right arm paresis -0.2204 Tapping left hand

the nonverbal intelligence test. The correlation between automatism production and Token Test score is only significant at P<0.05. If bucco-facial apraxia and degree of right arm paresis, which are strongly related with the Token Test score, are partialized, Kendall's partial rank correlation coefficient between automatism production and Token Test score drops to 0.11.

If, on the other hand, the Token Test score is being partialized (table 6), there remains a strong correlation between automatism production and bucco-facial apraxia, whereas the negative correlation between automatism production and arm paresis is greatly diminished. Numerically, as significance testing cannot be made, the correlation between automatism production and facial apraxia remains in the range of the significant associations of table 5, even if the Progressive Matrices scores are partialized too (Kendall's partial rank correlation coefficient for automatism

Dow

nloa

ded

by [

Uni

vers

itaet

sbib

lioth

ek G

iess

en]

at 0

6:11

27

Janu

ary

2015

Neurological status of speech automatisms 445

production x bucco-facial apraxia, Token Test and Progressive Matrices partialized = 0.2700). Finally, if Token Test and bucco-facial apraxia scores are partialized, the correlation between automatism production and Progressive Matrices Scores drops to 0.1382.

Discussion

The aim of the present study was to seek evidence, derived from clinical symptomatology, that would throw light on the question of whether the production of speech automatisms by an aphasic patient reflected: (1) a severe breakdown of the central language system (Stachowiak et al. 1977, Brunner et al. 1982), (2) a severe dysfunction in the periphery of the language production apparatus (Broca 1861; Blanken et al. 1988), or (3) the combined presence of an aphasiogenic focal and a disinhibiting diffuse brain damage (as discussed by Haas et al. 1988). AS these hypotheses are not mutually exclusive, the present investigation cannot constitute a decision experiment on the basis of which two of the hypotheses can be rejected.

Our data confirm the results and impressions of Alajouanine (1956): speech automatisms indeed occur together with severe aphasia in most instances. There are, however, exceptions (Kremin 1986, Blanken et al. in press) and the large ranges of performance in non-oral-expressive language tasks found in the present investigation indicate that they are not uncommon. As has also been pointed out by Alajouanine (1956), the presence of speech automatisms correlates with ideomotor apraxia and degree of hemiparesis, but again there are exceptions, as is reflected in the wide ranges of the corresponding parameters in the present study.

Accordingly, each of the hypotheses faces problems in fully explaining automatism production:

(1) The first hypothesis is correct insofar as speech automatisms occur only in the context of a severe disturbance of certain language processes proper, namely certain processes of language production. As has been pointed out, they also do not have an obvious counterpart in motor pathology. In the present study they were not correlated with perseveration in nonlanguage modalities. Speech automatisms differ from non-aphasic symptoms, such as iterations and palilalia, in that their composition is invariant. These arguments support the assumption that damage to the language system causes speech automatisms. However, automatisms do not necessarily co-occur with severe aphasia, and in the correlation analysis of the present study, the Token Test scores-considered a valid measurement of severity of aphasia-did not correlate with automatism production, if other influences were controlled. Both these observations favour the hypothesis that damage to only a restricted number of functional components of the ‘language production apparatus’ may suffice as a cause.

(2) The present study gave no convincing evidence of diffuse brain pathology in patients with speech automatisms. When age, lesion size and configuration were controlled, the patients’ performance in the Progressive Matrices and in left hand-tapping did not indicate diffuse or additional right hemisphere pathology. Intragroup variability in these parameters was far in excess of intergroup differences. In the correlation analysis, the Progressive Matrices scores were associated with a number of other parameters, including tapping with the intact hand and age, indicating that this score indeed reflects to some degree the patient’s overall impairment. Partialization of the results of the Token Test and apraxia testing,

Dow

nloa

ded

by [

Uni

vers

itaet

sbib

lioth

ek G

iess

en]

at 0

6:11

27

Janu

ary

2015

446 C.-W. Wallesch, J . C . Haas , and G . Blanken

however, led to a sharp fall in the correlation between automatisms and Progressive Matrices performance.

(3) Significant correlations could be established between ideomotor apraxia and the production of speech automatisms. These correlations are numerically rather low, and cases in which these symptoms do not co-occur indicate differences in the underlying functional pathology. Blanken et al . (1988) point out that the assumption of a severe disruption of function in the periphery of the language production system (‘prearticulatory segmental processing’ in their model) could result from two different types of pathology. Firstly, in the case of severe damage to the peripheral component, more central language processes may be spared. This explanation could account for the rare cases of automatism production with preserved writing abilities (Kremin 1986, Blanken et a l . in press). Secondly, in the majority of automatism patients the central language functions are severely impaired, and the peripheral components are either included in the damage or are disconnected. Severe central damage including the controlling component would account for automatism production with intact prearticulatory programming (Blanken et a l . 1988). These are the patients who are able to complete highly overlearned sequences, such as sayings, prayers and popular songs. If Blanken et a l . (1988) are correct, their preserved ability is in the chronic stage of aphasia-a bad rather than a good sign for the outcome of therapy.

This investigation has evaded an analysis of the interrelation between speech automatisms and speech apraxia. A major reason for this restriction was the difficulty of assessing speech apraxia quantitatively with clinically practicable methods. The correlation between automatism production and ideomotor apraxia and the functional explanation given for automatism production may suggest that these two symptoms are closely connected. Further investigation of this question may shed light on the interaction and functional organization of the interface between linguistic and articulatory processes.

References

ALAJOUANINE, T. (1956) Verbal realization in aphasia. Brain, 79, 1-28. BENTON, A. L. (1964) D e r Benton Test. Huber: Ben Stuttgart. BLANKEN, G., DITTMAN, J., HAAS, J . C., and WALLESCH, C. W. (1988) Producing speech automatisms

(recurring utterances): looking for what is left. Aphasiology, 6, 545557. BLANKEN, G., DITTMANN, J., and WALLESCH, C. W. (in press) Studies on the ‘speechless man’. The case of

speech automatisms. In: Festschrift 0. Szemerenyi. Benjamins: Amsterdam. BLUNK, R., DE BLESER, R., WILLMES, K. and ZEUMER, H. (1981) A refined method to relate morphological

and functional aspects of aphasia. European Neurology, 20, 69-79 BROCA, P. (1861) Remarques sur le siege de la faculte du langage articule, suives d’une observation

s’aphemie (perte de la parole). Bulletin of the Society of Anatomy (Paris), 36, 330-357. BRUNNER, R. J. KORNHUBER, H. H., SEEMULLER, E., SUGER, G. and WALLESCH, C. W. (1982) Basal ganglia

participation in language pathology. Brain and Language, 16, 281-199. CODE, C. (1982) Neurolinguistic analysis of recurrent utterances in aphasia. Cortex, 18, 141-152. CODE, C . (1987) Language, Aphasia and the Right Hemisphere. London: John Wiley. HAAG, E., HUBER, W., HUNDGEN, R., STILLER, U., and WILLMES, K. (1985) Repetitives Sprachverhalten bei

schwerer Aphasie. Nervenarzt, 56, 543-553. HAAS, J. C., BLANKEN, G., MEZGER, G., and WALLESCH, C . W. (1988) 1s there an anatomical basis for the

production of speech automatisms? Aphasiology, 6, 557-567. HABBEMA, J . D. F. , HERMANNS, J., and VAN DEN BROEK, K. (1974) A stepwise discriminant analysis program

using density estimations. In: G. Bruckmann (ed.), Compstat 1974, Proceedings in Computational Sfafistics. Vienna: Physica.

Dow

nloa

ded

by [

Uni

vers

itaet

sbib

lioth

ek G

iess

en]

at 0

6:11

27

Janu

ary

2015

Neurological status of speech automatisms 447

HUBER, W., POECK, K., WENIGER, D., and WILLMES, K. (1983) Der Aachener Aphasie Test. Hogrefe:

HUBER, W., POECK, K., and WILLMES, K. (1984) The Aachen Aphasa Test. In: F. C . Rose (ed.), Progress in

HUGHLINGS-JACKSON, J. (1879-80) O n affections of speech from disease of the brain. Brain, 2, 205222. KLEIST, K. (1934) Gehirpathologie. Barth: Leipzig. KRATZMEIER, H. (1979) Manual Raven-Matrizen-Test. Weinheim: Beltz. KREMIN, H. (1986) Is there more than ah-oh-oh? Alternative strategies for writing and repeating lexically.

In: M. Coltheart, R. Job, and G. Sartori (eds.), The Cognitive Neuropsychology of Language. London: Lawrence Erlbauni Associates.

Gottingen.

Aphasiology, New York: Raven.

LURIA, A. R. (1980) Higher Cortical Functions in Man . New York: Basic Books. POECK, K. (1982) Apraxie. In K. Poeck (ed.), Klinische Neuropsychologie. Stuttgart: Thieme. RAVEN, J. C . (1956) Standard Progressive Matrices. London: Lewis. RIDDOCH, G., BRISTOW, W. R., CAIRNS, H . W. B., CARMICHAEL, E. A,, CRITCHLEY, M. GREENFIELD, J. G.,

LEARMONTH, J. R., PLATT, H., SEDDON, H. J., SYMONS, C. P., YOUNG, J. Z., ~ ~ ~ H E R R A L D , F. J . Z. (1943) Aids to the investigation of peripheral nerve injurie3.H. M. Stationary Office, London.

SIEGEL, S. (1956) Nonparametric Statisticsfar the Behavioural Sciences. Tokyo: MacGraw Hill. STACHOWIAK, F. J.. HUBER, W., KERSCHENSTEINER, M., POECK, K., and WENIGER, D. (1977) Die globale

Aphasie. Journal of Neurology, 214, 75-87.

Dow

nloa

ded

by [

Uni

vers

itaet

sbib

lioth

ek G

iess

en]

at 0

6:11

27

Janu

ary

2015

本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具