GRAMMATICAL IMPAIRMENT IN DEVELOPMENTAL APHASIA Max Kerschensteiner and Walter Huber (Abteilung Neurologie der Medizinischen Fakultat der Rheinisch-Westfalischen Hochschule, Aachen)
INTRODUCTION
In his survey of developmental speech disorders, Ingram (1969) describes as unsatisfactory those classifications of speech disorders which are based more on studies of clinical findings than upon linguistic analysis. In the literature, surprisingly little attention has been given to the actual language disorders. Clinical classifications are based mainly on secondary symptoms such as speech and hearing disorders, mental and psychiatric disorders, and adverse environmental conditions (Leischner, 1967; Lenneberg, 1964; von Stockert, 1969). Even in more recent surveys (Ingram, 1969; Bauer, 197 3 ), speech disorders and language disorders are "hopelessly confused" (Critchley, 1970). The same can be said of the distinction between linguistic and mental retardation. In reviewing cases of receptive and expressive language retardation, Benton ( 1964) stressed the necessitiy of distinguishing a group of patients with specific language disorders from those patients with deafness, mental deficiency, and severe personality disorders. He suggested that for specific language disorders in children the term "develop mental aphasia" could be used. He found it highly probable that cerebral damage is implied as a causative factor in almost all cases, and he suggested that the basis for developmental aphasia is a higher level auditory perceptual deficit. Many empirical studies are concerned with such a deficit (Lowe and Campbell, 1965; Stark, 1967; Stark, Poppen and May, 1967; Weiner, 1969, 1972; Tallal and Piercy, 1973 a, 1973 b, 1974), whereas less consideration is given to disturbances in language production (Menyuk, 1964; Lee, 1966; Lackner, 1968; Morehead and Ingram, 197 3). Rees ( 197 3) showed convin cingly that there is only weak evidence for the assumption made by Eisenson ( 1968 a, 1968 b) and others, that developmental aphasia should primarily be explained by a failure in auditory processing. We had the opportunity to study the grammatical impairment of the language of a patient with developmental aphasia. His language impairment Cortex (1975) 11, 264-282.
Grammatical impairment in developmental aphasia
265
is of particular interest as it is still present at his age of 23. Thus we could study developmental language disorder independent of ongoing primary lan guage acquisition. CASE REPORT
In November 197.3, G. H., a 2.3 year old right-handed factory worker, presented at our clinic with a history of language problems since early childhood. There was no family history of hereditary speech disorders, of left handedness nor ambidexterity. His mother was 40 when he was hom via a complicated forceps delivery leading to a severe asphyxia in the infant patient. There was no later neurological or psychiatric illness. According to his parents and his two older sisters, his motor development was normal. His mental development is reported normal with the exception of a retarded and incomplete language development. Up to the age of 4, his speech production consisted only of babbling. At the age of 8 the patient was still not accepted at primary school, and so was hospitalized in a. psychiatric unit. On examination there were some soft neurological signs such as hypotonia and hyperextensibility of the joints of the upper limbs, and some clumsiness in skilled movements. EEG was normal. His language deficit was described as global stammering. At present there are no abnormalities in his general physical state and in his social behavior. Neurological examination shows no abnormal signs except a slight increase of the deep tendon reflexes in the right upper limb. EEG is normal. In our standard aphasia examination the patient's spontaneous speech was judged to be fluent with good prosody and normal rate of speech, but with hurried and bad articulation. He produced relatively few phonemic and even fewer semantic paraphasias. A few word finding difficulties were noticeable. The most obvious feature was a distortion in grammatical structure. In conversation, the communicative intention of the patient could easily be understood, though not always in precise detail. On repetition of mono- and hi-syllabic words the patient produced phonemic paraphasias of .3.3 in 100 words. There was also a moderately severe distortion on repetition of words of more than two syllables, and of sentences. The patient had considerable difficulties in discriminating phonemes which were arranged in minimal pairs in a special task for two-word repetition. Serial speaking was intact. In object naming the patient failed mainly on those objects with names of non-germanic origin, partly because he had not learnt these words at school, partly because he could not pronounce them. In a picture-matching task the patient's comprehension of words was unimpaired. In comprehending sentences there was a slight impairment (4 items were failed out of 40 ), and in comprehending spoken text there was a fairly severe impairment ( 1.3 items were failed out of 26 ). In a shorter version of the Token Test, the patient failed 14 items out of 50. This is equivalent to a moderate distortion of language comprehension in aphasic patients. In grammatical judgement tasks the patient was better in perceiving lexical than in perceiving syntactic deviations. In a card-sorting task he was able to select correctly those words belonging to a semantinc field.
266
M. Kerschensteiner and W. Huber
In reading aloud the patient made phonemic paraphasias, but fewer than in repetition of spoken words. The comprehension of written text was affected in the same way as the comprehension of spoken texts. The patient's writing was distorted for linguistic reasons, and not as a result of constructive apraxia. The patient's non verbali.Q. was 90, as assessed by administration of Raven's Coloured Progressive Matrices. The Wechsler-Bellevue scores were as follows: verbal I.Q. 76, performance I.Q. 88, global I.Q. 81. On a specially designed task requiring the recall of sequentially ordered digits and colors, the patient failed consistently on sequences of as few as 4 elements, and he performed sligthly better with non-verbally than with verbally recalled sequences. His abilities in elementary sensori-motor tasks was below average for his age. He improved with practice in most of the tasks, but he demonstrated no transfer of learning in general. The patient showed severe disturbances in imitation of rhythmic motor patterns. In Benton's Three Dimensional Constructional Task he performed well (29 out of 30). No signs of bucco-facial apraxia, ideo-motor apraxia, or signs of visual and auditory agnosia were found. Tonal audiometry revealed no hearing deficit. Linguistic analysis Material and method
The material for the linguistic analysis consisted of a 3 hour conversation, which was tape-recorded the day the patient first presented at our clinic. After transcription, about 200 sentences were selected for syntactic analysis. The sentence boundaries were chosen on the basis of pauses and the semantic context in which the utterances appeared. The syntactic analysis was carried out whitin the framework of transformational grammar (Chomsky, 1965, 1973; Ross, 1967; Huber and Kummer, 1974). Transformational analysis tries to extend the observation of linguistic segments and their distribution in the surface form of sentences to the deduction of a system of rules - so called transformations which derive surface structures from syntactic base structures. In base structures the semantic relations of one particular usage of a sentence are specified in an unambiguous way, whereas from surface structures different semantic interpretations may follow. The full derivation of a sentence proceeds through different components of the grammar: the lexicon, the syntactic base component, the transformational component (and the semantic and phonological components which are not refered to in our analysis). In addition to descriptive adequacy, the grammar is postulated to have explanatory adequacy in so far as the grammar is assumed to be a model for the tacit grammatical competence of a language user. As the languages used by children of different ages, or by aphasics of different clinical types and different levels of severity may be systematically different from the language of a normal adult speaker, it is heuristically necessary to describe their languages by grammars of different degrees of complexitiy. Basic sentence structures
· In the language of G. H. there is a variety of basic sentence patterns which result from different stages of language development. These basic patterns can be
Grammatical impairment in developmental aphasia
267
seen in simple declarative sentences which bear no special emphasis. A list of sentences consisting only of a verb phrase (VP) with the verb in the infinitive form is given in Table I. TABLE I
VP Sentences 1 (1) (2) (3) (4) (5)
lassen Ruh wohnen bier in SteinkaulstraEe acht Stunden schlafen GroEstadt au gern wohnen ich bin jetzt zur Zeit so mal sagen Maschinen fiihren
Some of the infinitive patterns correspond to the basic grammatical relation verb-object, one of the first acquired by normal children (Bloom, 1970). Notice that two different word orders are possible, as has also been reported in the language of German children (Grimm, 1973 ), and in the language of Broca aphasics (Myerson and Goodglass, 1972 ). Complement phrases may follow the infinitive as in sentences 1 and 2, or complement phrases may preceed the infinitive as in 3 to 5. In the following sentence both word orders can be found: ex.: (G. H. talks critically about foreign aid given by the government.) so ohne Zinsen dadei geben . . . zuri.ickbezahlen wenige Zinsen .. . without any interest lending ... paying back low interest .. . wollmer mal sagen ... acht Prozent zuri.ickbezahlen let me say . . . eight percent paying back This example shows that the object phrase preceeds the verb whenever the objects phrase bears special emphasis. In terms of syntactic rules this means that the basic word order is verb-object, and that the reversed word order object-verb is derived by a permutation transformation. In Table II the first three sentences exhibit the word order SVO (subject verb-object), and in the last three sentences the word order is SOV (subject· object-verb). TABLE II
SVO vs. SOV Sentences (1) (2) (3) (4) (5) (6)
Ja ich versteh die Satze ( +) viele Deutsche verstehn dat niemals ( +) ach der macht nich de Hande dreckig ( +) der zuhause n guten Auskommen hat FDP und SPD die zusammen ein Partei machen die immer mich vorsagt "geh in die Sprechschul"
1 In the following tables, dots indicate pauses, and "( +)" after a sentence indicates that this sentence is grammatical in adult grammar with respect to the critical grammatical features. Where possible corrected versions of ungrammatical forms are added in italics.
M. Kerschensteiner and W. Huber
268
Compare for instance the English translations of sentences 1 and 4: ex. 1: Yes I understand the sentences ex. 4: (G. H. tells about a friend whose parents have a fairly big income.) he at home a good income has Whereas the SVO sentence is grammatical in Standard German, the SOV sentence is ungrammatical. The SOV order would however be perfectly correct in subordinate clauses (a possibility which is clearly excluded because of the semantic and syntactic context). The equal occurrence of SVO ans SOV pattern demonstrates that a general syntactic differentiation of main and subordinate clauses is not yet maintained. In terms of syntactic rules this means that application of the rule Verb Shift which puts a finite verb onto the end of a verb phrase only in subordinate clauses is not yet restricted. This could predict that Verb Shift also does not operate generally in subordinate clauses, which is confirmed by the following sentence 3: ex.: (When asked whether some people dont't understand him, G. H. answers that he has only friends who do understand what he says.) ich Bekannten kenne die verstehn mich alle I friends know who understand me all The rule Verb Shift does not apply in the necessarily restricted relative clause which is extraposed to the end of the sentence. But in the main clause Verb Shift does apply. In adult grammar, both application and non-application results in ungrammaticality. However, in G. H.'s grammar the rule Verb Shift may apply unrestrictedly and optionally. By comparing this rule with the above mentioned rule which puts an object phrase into the intitial position in VP sentences, we see that rules of different stages of language development coexist in G. H.'s language. In VP sentences it is not the verb but the object phrase which is shifted, whereas in SVO sentences it is the verb which may be shifted. A specific function for Object Shift is already developed, viz. marking of emphasis, whereas the specific function of Verb Shift, viz. marking of subordination, is not generally observed. Syntactic marking
The grammatical relations subject of a sentence and object of a sentence are not generally marked by rules for person, case, gender, and number. In sentence 2 of Table III the application of the rule Subject-Verb-Agreement which
TABLE III
Subiect Verb Agreement (1)
(2) (3) (4) (5)
der sagen (sagt) wo kommen her ich kommt (komme) nich drauf dann sieh (seh) ich rot die is jetzt 26 ( +) ofmal verstehste nix von denen ( +)
-·'"""""..'
~~=--=-':'.~~...r.!l:"
-
..._
Grammatical impairment in developmental aphasia
269
copies the syntactic markers for person and number of the subject phrase onto the verb leads to a wrong output: ex.: ich kommt nich drauf I doesn't get it In other sentences the correct grammatical forms of Standard German show up. A failure in agreement could be caused on purely lexical ground. Within the lexicon, the paradigms of certain verb forms may have gaps, or may consist just of one form which is generally inserted on syntactic marking. The following sentences demonstrate that there is a stage of language development with paradigms for the pronouns 3rd pers. sing. and 1st pers. plur. consisting just of the forms "er" ("it") and "uns" ("us"), respectively. ex.: (Talking about his work at a rubber factory, G. H. describes how he and some colleagues had to pull the heated rubber out of the furnace.) mu.E er uns rausziehen ... na kommt der Gummi raus n Ofen da then comes the rubber out of the furnace ... then must it us pull out The pronoun "er" in the second clause is coreferent to the nominal phrase "der Gummi" of the first sentence, i. e. "er" is the logical object of the verb "rausziehen." However, syntactically the pronoun "er" appears to be a subject pronoun on the basis of standard word order and of its lexical form. Likewise, the pronoun "uns" is syntactically marked as object despite its logical function as a subject. One could get the impression that the patient reversed the logical order; but this is in contradiction to the fact that G. H. was able to talk about his work at the rubber factory in a reasonable logical order. The failure lies in the paradigmatic organization of his lexicon. The only form available here are the nominative form "er" and the accusative form "uns," so that the surface form of the sentence is misleading for a possible hearer. (The deviation in word order is less important as there is also relatively free word order in colloquial German.) That we are actually dealing with a lack of differentiation in the lexicon is supported by sentence 3 of Table IV showing the same accusative form "uns" instead of the appropriate nominative form "wir" ("we"). All these TABLE IV
Case, Gender, Number (1)
(2) (3)
(4) {5) (6)
die immer mich (mir) vorsagt so n Band gerissen mich (mir) ham uns ( wir) Mist gebaut die SPD so langsam die so . . . der (die) iiuberlegt nich ich weiB nich aile StraBe (Straf3en) mein Elter ist alt (meine Eltern sind alt)
examples demonstrate that parts of the patient's lexicon are less differentiated than in a lexicon required for Standard German. This leads to a situation in which rules for syntactic marking cannot apply. Similar failures are well known from the language of children and of aphasics.
270
M. Kerschensteiner and W. Huber
Lexical Differentiation
Not all lexical disturbances are purely grammatical in nature. Sentence 1 of Table V is a typical example of word finding difficulty, and sentence 2 shows that a failure in phonological discrimination might lead to an erroneous semantic paraphasia. Sentence 3 leads us back to grammar. There, rules of word formation apply in a semantically unrestricted way. The semantic differentiation of verbs is TABLE V
Disturbances in Lexical Differentiation
(1) (2) (3)
eine geht no n Schul Frankfurt Betriebswirschaft lernt ... net Betriebswirschaft ah ... wie heiBt n normal ... wie heiEt normal ... Sonder eh ... wie heiBt n normal ... Sonder faarbeit ... nie ... Faarbeit ... Sander (target word: "Sozialarbeiter") ich dachte das meiste Sprechen Nerven zusammenhanden (zusammenhiingen) ah zusammen packen nur n biEchen nervlich angespannt ... soli mer sagen ... verabeitet ... verhetzt.
achieved productively by the addition of prefixes in Standard German. In sentence 3 two prefixes are mixed up: instead of "ab" the prefix "ver" is used to give lexical derivations of the verbs "arbeiten" and "hetzen" ("work" and "chase"), respectively. By adding the prefix "ab" one can derive in German meanings similar to the English participles "worn out" and "exhausted." Thus, what the patient wanted to say in sentence 3 is: ex.: nur n bi[3chen nervlich angespannt. . . sollmer sagen ... so to say nervous only a bit abgearbeitet . . . abgehetzt exhausted worn out However, the use of the prefixes "ver" makes the meaning of the sentence semantically anomalous: ex.: nur n bi[3chen nervlich angespannt ... sollmer sagen ... verarbeitet ... verhetzt manufactured ... agitated This shows that certain prefixes in G. H. 's lexicon are not differentiated, which leads to a false application of rules for word formation. That the selection and combination of lexical items is less restricted than in adult grammar can be seen from sentences in Table VI. TABLE VI
Selectional Restrictions
(1) (2) (3)
ne Arbeit komm ich immer ausgeschimp "hast net verstanden ... du bist n Tuppes" so wa ... ich "is gut ich hab n Tuppes" Bayern ... da hab ich ja eben aufgefallen so Galekt so ich bin so Sprinder so ... ich muE mal da einsetzen muE ich da
Grammatical impairment in developmental aphasia
271
Take for instance sentence 1. In the dialect spoken around Aachen there is an insult "Tuppes" meaning "idiot," and there exists the syntactic constraint that "Tuppes" can only be used as a predicate noun. This constraint is observed in the first direct discourse in sentence 1, "You are a Tuppes," but not in the second one, "I have a Tuppes," which is ungrammatical. Similarly in sentences 2 and 3 the grammatical constraints for the insertion of transitive versus intransitive verbs are not observed, i.e. an intransitive verb is used where a transitive verb should be used in Standard German, and vice versa. Differentiation of syntactic categories
Certain syntactic categories are not yet differentiated in some of the grammatical stages of G. H. 's language. In some sentences of Table VII, we find prepositions fully developed, as in sentence 1, "Don't you come from Aachen?". In other sentences prepositions are not used at all, like in sentence 3, "October I will also go Tunisia." At a third stage of grammatical development the position of prepositions is marked by the lexically neutral morpheme "n," for instance in sentence 5, "Well, you gave me an appointment n four o' clock." The morpheme "n" can also indicate the position of preposition plus article as in sentence 4, TABLE VII
Prepositions vs. Articles
(1) (2) (3) (4) (5) (6)
ja kommen Sie nich aus Aachen? ( +) dann bin ich Dr. Bolten hingegangen Oktober fahr ich noch Tunesien ich hab mein Zimmer n Dach oben ja ham Sie gestellt mich n 4 Uhr die meiste kommen aus n Dorf so
"I have my room n roof upstairs," which shows that there is a further stage of grammatical development in which the syntactic categories are not distinguished from each other. However, there exist also other grammatical forms which show clearly a distinction between preposition and article as in sentence 6, "Most of them came from n village." Here the preposition is lexically fully realized whereas the article is still realized by the neutral form "n." Some more examples of syntactic categories which are not differentiated at all stages of H. G. 's language development are nonseparable vs. separable verb particles, separable verb particles vs. predicate bound adverbs and predicate bound adverbs vs. verbs. This is demonstrated by the sentences of Table VIII. TABLE VIII
Adverb, Verb Particle, Verb
(1) (2) (3) (4) (5) (6)
passen die Worter nie mehr zusammen ( + ) ich mach immer n Fehler rein ( +) die viel meiste ausziehn da (ziehn da aus) die Katze hinter lauf hinter her (die Katze lauf hinterher) ich morgens 6 Uhr raus hab ich au die Worter nich richtig raus ( + )
M. Kerschensteiner and W. Huber
272
All these examples show that there are stages of grammatical development in the language of the patient at which the differentiation of syntactic categories and/or their lexical realization is not accomplished, although there are other stages at which both are accomplished. Adverb placement
The placement of certain minor syntactic categories in sentence structures is less restricted than in Standard German. In sentence 1 of Table IX, "alles so hetz mich," with a meaning similar to the colloquial American expression "everything is such as hassle," the sentence adverb "so" is in the preverbal position which is correct for many adverbs in English. In Standard German, however, there is a very strong word order constraint which reserves the second position TABLE IX
Adverb Placement (1) (2) (3) (4) (5) (6)
alles so hetz mich (alles hetzt mich so) so mal sagen (mal so sagen) n Mensa au kenn ich (n Mensa kenn ich au) de neue Arbeitgeber vielleicht gibt net sofort n Urlauh ( ... gibt vielleicht .. . ) de Pelz zusamm wird sortiert (de Pelz wird zusammensortiert) muE einer besonders n ganz gute Film sein (mu/1 einer n ganz besonders gute Film sein)
in a single declarative sentence exclusively for the finite verb. Sentence adverbs are normally put at the end of such sentences. Sentence 1 shows however that adverb placement in G. H. 's language does not follow this language specific constraint. Another case with an adverb showing up at a position different from that in adult language can be seen in sentence 6, the meaning of which is "It must be a really far out movie." Like the English adverb of degree "far" in the idiomatic expression "far out," the German adverb "besonders" is semantically bound to the following adjective. In German as well as in English, bound adverbs can not be set free from their modifying position. However in the language of our patient this constraint is not observed: ex.:
mu~ einer besonders n ganz gute Film sein must it far a really out movie be
. In addition to the free placement of the adverb "besonders!' it is the inversion of the subject pronoun and the finite verb in a declarative sentence which makes this sentence ungrammatical in Standard German. Permutation transformations
There are only a few environments in Standard German where inversion has to apply, viz. Yes-No-Questions and Imperatives. Under these environmental conditions, inversion is also triggered in G. H. 's language, as can be seen from sentences 1 and 2 of Table X.
Grammatical impairment in developmental aphasia
273
TABLE X
Subject Verb Inversion (1) (2) (3) (4) (5) (6) (7) (8)
dar£ ich eine rauchen? ( +) rauchen Sie auch eine? ( +) is nix mehr los (coli.: +) nee kann ich nich (coli.: +) hab ich keine Ausdauer bin ich so durcheinander kommen die n Monat 400 Mark ham uns frillier Wirtschaft gehabt
The other sentences of this table, however, give evidence for unrestricted application of inversion. Changing the normal word order by permutation trans formations does not lead in all cases to a violation of standard conditions for rule application. An example for the correct application of the rule Topicalization can be seen in sentence 1 of Table XI. TABLE XI
Topicalization, Dislocation (1) (2) (3) (4) (5) (6) (7) (8)
selten geh ich ins Kino ( + ) n Mensa liiuft au oft mal so n Film ( +) die arbeite immer zusamm die SPD und FDP ( +) ja n Arbeit da kann ich net mitkommen ( +) FDP und SPD die zusammen ein Partei machen ( +) genau zusammen Farben passen (genau zusammen passen Farben) his die 64 n Schule war ich (his die 64 war ich n Schute) denn Arbeitsamt hingegangen bin ich (denn bin ich Arbeitsamt hingegangen).
The adverb "selten" ("rarely") is put into initial pos1t1on. This operation is followed by the rule Subject Verb Inversion which puts the subject phrase into the position immediately after the verb. It is interesting to note that Subject Verb Inversion does not follow Topicalization in English. Translated into English, sentence 1 would read "Very rarely, I go to the movies" and not "Very rarely go I to the movies." Sentence 5 is another example of a sentence in which permutation transformations apply correctly. Here the rule Left Dislocation puts a costituent into pre-sentential position and· leaves a pronominal copy in its original position. The rule operates in English in the same way: ex.: FDP und SPD die zusammen ein Partei machen FDP and SPD they into one single party merge Left Dislocation, unlike Topicalization, is not followed by Subject Verb Inversion in German. In G. H. 's sentence, Inversion is not applied either as the sentence pattern SOV does not provide the structural condition for Inversion. So far we have seen that permutation transformations apply in G. H. 's language
M. Kerschensteiner and W. Huber
274
under the same conditions as in Standard German. The sentences 6 to 8 show however that this does not hold for all sentences. In sentence 6 Topicalization applies without being succeded by the application of Subject Verb Inversion, resulting in the following sentence: ex.: genau zusammen Farben passen fit exactly together colors and not in, "genau zusammen passen Farben" which, from the point of view of word order alone, would be a grammatical sentence in Standard German. Another example of the violation of conditions under which Topicalization can apply can be seen in sentence 7. die 62 n Schule war ich ex.: bis until the 62 at school was I In this sentence two independent constituents are topicalized which is ruled out in Standard German as well as in many other languages by the general constraint that Topicalization can be applied in a simple sentence only once. Deletion transformations Rules which operate under more conditions than generally accepted in the standard language are said to have an overproductive application. Examples of such rules in G. H. 's language were Subject Verb Inversion, and Topicalization. Other examples come from sentences in the derivation of which deletion rules apply overproductively. The copula (German "sein," English "be") can be deleted in German if certain embedded sentences are to be reduced, for example if relative clauses are to be reduced to appositional phrases. However, the copula can not be deleted if "be" is the verb of a main clause. The sentences of Table XII TABLE XII
Copula Deletion
(1) (2) (3) (4) (5) (6)
die is jetz 26 ( +) der Film war ganz gut ( +) ich einzge Sohn noch zuhause so 30 Prozent no Deutsche da n Pelz so n Mangel ach n groEn Teil miese
show that this restr1ct1on is not generally observed. Of particular interest is sentence 3, because the condition under which the copula can be deleted in Standard German is observed, but not the condition under which the copula cannot be deleted. Another deletion rule which is overproductively applied is the deletion of subjects pronouns. The only derivation in which subject pronouns can be deleted in main clauses of German is the derivation of Imperatives. In G. H. 's language there are cases of such a subject pronoun deletion. However, there are also
Grammatical impairment in developmental aphasia
275
sentences, viz. Yes-No-Questions and Declaratives, in the derivation of which this deletion constraint is not observed (cf. Table XIII). TABLE XIII
Subject Pronoun Deletion
(1) (2) (3) (4) (6)
die sagen "nee bleib zuhause" ( +) "bist blod so! ... geh zur Seite!" ( +) tut rna leid offen Studenten wa (coiL: +) ja viel Spa.6 bekommen hab is nur hier in Karneval was los
Interference of different syntactic derivations
Most of the grammatical stuctures we presented so far gave evidence for the claim that different systems of grammatical rules are co-existent in G. H. 's language. As a consequence different systems interfere with each other resulting in redundant sentences in which different sentence patterns are mixed up. Sentences with such intermingled patterns are listed in Table XIV. TABLE XIV
Interference of Different Sentence Derivations
(1) (2) (3) (4) (5) (6) (7) (8)
na war ich bis 72 war ich da ja ich bin 1958 in Schule gekonnen bin ich ich mu.6 mal da einsetzen mu.6 ich da hab ich an Maschine gearbeitet hab ich ich hab meine 8 Stunden Schlaf hab ich gehabt noch ja in Aachen tanzen geh ich nich hier in Aachen jetzt hab ich vor vor n halbes Jahr hab ich gesehen "de Regenbogen" und zuhause hat n genug Geld zuhause
Take for instance sentence 4: ex.: hab ich an Maschine gearbeitet hab ich have I at machine worked have I The participle phrase "an Maschine gearbeitet" belongs simultaneously to two different sentence patterns. The first one results from a SVO sentence in which the rule Subject Verb Inversion is overproductively applied, and the second one results from a SVO sentence in which the embedded participle phrase is topicalized. Sentence 4 as well as the other sentences of Table XIV were spoken by the patient with normal sentential intonation. There were neither any hesita tions which would indicate self-correction, nor any rising or falling intonation at the end which would indicate constructions like tag-questions in English. Furthermore, these sentences are structurally quite different from sentences in which errors in linear sequencing occur, such as the sentences of Table XV.
276
M. Kerschensteiner and W. Huber TABLE XV
Disturbances in Linear Sequencing (1) (2) (3) (4) (5) (6)
so n Band gerissen mich . . . na wieder na wieder neu angefangen is so so n so n bis morgens Uhr 5 Uhr auf wa heute is ja heute is ja heute jeder aufgeklart wa hab ich au gesagt hier n Doktor Kerschensteiner ... gesagt ich kann die Worter zusamm s aussprechen wa da hab ich seh ich ja rot voriges Jahr n FuBgangerball war ganz Kopf Mensa auf n Kopf
Here we find immediate repetltton of single segments, the interruption of a sentence and continuation with a new one, and the anticipation of words within longer idiomatic expressions. None of these features is found in the sentences of Table XIV.
Complex sentence structures In addition to the interference of different sentence patterns each of which is derived by grammars of a relatively advanced level, there are sentence patterns of different stages of language development co-occuring is complex sentences (cf. Table XVI). Typically, different basic sentences such as VP sentences and SVO sentences are combined into complex sentences. Complex sentences do not differ from simple sentences in principle; some are fully correct in adult grammar; others show the same deviations (in syntactic marking, lexical organization and lexical insertion, in word order, and in deletion of syntactic elements) as simple sentences. The complex sentences include a variety of semantic types of embedded clauses: adverbial clauses (among them most frequently conditional clauses), verb complements, relative clauses, direct discourses. The semantic function of sub ordination is not in all cases lexically and syntactically marked. As mentioned before, the conditions of the rule Verb Shift which generates SOV patterns only in subordinate clauses are not generally observed. The insertion of conjunctions into basic sentences is lacking in many instances. Only in cases like sentences 22 and 23 can the absence of conjunctions be attributed to a rule of adult grammar. The conjunction "daB" ("that") can be optionally deleted from unreduced verb complements if the complements are governed by certain verbs in the main clause like "wissen" ("know"). Contrary to these cases, there is no lexical exception to the constraint that conjunctions cannot be deleted in sentences 27 to 29. From the deletion of "da{r' in the derivation of these sentences we see again that a syntactic rule may be overproduct ively applied because certain grammatical contraints are neglected. Similarly, the syntactic marking of reduced complements by "zu" ("to") is not generally observed, despite the fact that sentence reduction itself is accomplished (cf. sentences 31 to 34 of Table XVI).
Coordinate sentence structure There is still further evidence for our claim that in G. H. 's language some rather complex syntactic operations can ben accomplished, yet with one typical
Grammatical impairment in developmental aphasia
277
TABLE XVI
Complex Sentences
1. Conditional clauses (1) wenn ein so oniir so is . . . dann sieh ich rot ( + ) (2) wenn was ran is ... liiBt er immer de Werkstatt machen ( +) (3} wenn anrufst viele GriiBe stellen (4) n ganzen Tag n Beinen ... dann bin ich nich richtig dabei (5) n Wagen hast dann kannste immer in die Stadt fahren (6) icb selbst sprech jetzt . . . versteh ich das ganz (7) mit Unkosten alles recbnen ... teuer 2. Adverbial clauses (8) eben bier n Professor unterhalten hab ich ... die Worter nicht sam ... fassen ich richtig (9) ich mit mein Eltern noch gewohnt bah da bah ich immer im Wohnzimmer geschlafen (10) weil das Freundin Urlaub war konnte nich Freundin besuchen (11) ofmal verstehste nix von denen . . . wenig deutscb kann (12) ja genau Menscben wir
3. Relative Clauses (13) offen n Arbeiter so Vorarbeiter oder Meister die n Erlekt sprechen die wa ... die so . . . unfein . . . sowa so Oescber Platt ... (14) ich kenn eine der hassen Auslander wa (15) ich kenn n paar ... die musse bezahl so 400 Mark Miete (16) ich Bekannten kenne die verstehn mich aile (17) dort kommt der Mann den die Tage gesehen hab icb (18) der sagen wo kommen her so Drecklochern 4. Direct (19) (20) (21)
Discourses komm so n Student rein ... so n Bart oder ... "Wat n dat n einer" gut angezogen is "ja bisse mein Mann" mein Eltern mein Schwester gesagt wa "jetzt mach n Fiihrerschein" vor fiinf Jahren wa ... nee sag ich "jetzt kein Lust bah Fiihrerschein macben"
5. Verb Complement Clauses (22) · ich weiB ganz genau ich hab recht ( +) (23) ich weiB Sie verhoren mich so ( +) (24) hoffen daB Sprechen besser halbes Jahr ... dann mach ich n Priifung (25) er wundert daB schon Wetter war heute (26) es is so heiB daB muB ich das Fenster aufmachen (27) so verriickt bin au nicht ... so n Autorennen hinfahren (28) nocb stinkt so ... de Arbeiter so richtig Druck stehen (29) ich bin frob n Sommer wird (30) die Araber drehten das 01 zu damit Europiier kein 01 gibt 6. Complement Clause Reduction (31) da haste kein Lust n Kino zu gehen ( +) (32) ach bier in Aachen gibt n net viel unterhalten oder so (33) na hast so 10 Minuten Presse sein (34) eine geht so n Schul Frankfurt Betriebswirtschaft lernen.
deviation from adult grammar, namely the neglect of specific grammatical constraints under which alone these operations can apply. Sentences 2 to 7 of Table XVII exhibit reduced coordinate structures. The conditions under which deletion in coordinate sentences can occur is illustrated by the following non English sentence, "Father beer and mother drank wine." Translated word for word into German, this sentence would be just as ungrammatical. The structural
278
M. Kerschensteiner and W. Huber TABLE XVII
Coordinate Sentences (1) (2) (3) (4) (5) (6) (7)
ja mein Freund hat n Wagen so n Ente und der braucht ganz wenig Besin ( +) ich hab n paar Blumen gekauft und ich schenk meiner Freundin ich bin frob ... ich hierhin komm und mich handeln und alles ... und sprechen und alles da hab ich n Anfang Tankwar ... 1966 2 Jahre spiiter n Pelz Vogel angefangen wern dann so Metallplatte rauf . . . auch so n Presserei geschoben so n Kiirschner mich angelernt und geschnitten· hab ich erst n Mantel aufgezeichnet und dann werne abgeglichen
reason for the ungrammaticality is that in a coordinate sentence structure one of two lexically and structurally identical verbs can only be deleted in the second and not in the first sentence conjunct. That this constraint is violated in G. H. 's language can be seen from sentences 6 and 7. In both sentences, one of two identical verbs is deleted in the first conjunct. Notice however, that there are structures in Standard German in which such a deletion is possible, namely in subordinated clauses, which can be seen from another sentence made up for illustration, "when father beer and mother wine drank ". This would be a perfectly grammatical subordinate clause in German. This comparison shows that in G. H. 's language even rather complex rules are acquired, but that - so to say - the tacit knowledge of the conditions under which these rules apply in the grammar is not generalized.
DISCUSSION
Two hundred different sentences were selected from a sample of the spon taneous spee~h of a patient with developmental aphasia and analyzed in terms of generative grammar. The main results are: (1) There are different degrees of complexity including those which belong to the grammar of Standard German. These different grammatical forms partially interfere with each other. (2) Sentences with appropriate use of minor syntactic categories, like prepositions, articles, conjunctions, etc. are found together with sentences in which these categories are not yet differentiated. (3) Paradigmatic organization of inflectional endings, verb prefixes, and pronouns does not exist in many cases, or is incomplete in comparison to Standard German. (4) Some selectional restrictions which govern the insertion of lexical material into sentence structures in Standard German are not generally observed. (5) Many transformations of Standard German - even fairly complex ones are applied. Interestingly, however, only the very general form of the transformations is observed in many cases, whereas certain conditions of ap plications and certain grammatical constraints which are specific for German are neglected. In terms of models of language acquisition (Me Neill, 1970; Bloom,
Grammatical impairment in developmental aphasia
279
1970), the different grammatical forms in the language of our patient reflect several stages of language development. In fact, his language share& many formal features with that described for German children at the age of 3 to 6 (Grimm, 1973). Just as the language of a five year old child includes many of the syntactic structures of adult grammar, the language of our patient contains a variety of forms of Standard German. These similarities should not be understood as implying that any five year old child with normal language development will show the same number and the same combinations of grammatical irregularities as this 23 year old patient with developmental aphasia. The patient's general grammatical impairment results in a language in which several structurally different grammatical forms occur together. There are forms present which are standard in the adult language but they are not predominant. This shows that the patient's linguistic deficit does not lie in a general inability to acquire any particular grammatical form of the adult language. Rather it can be assumed that the patient's capacity for language acquisition was largely preserved. However, the process of language acquisition in our patient came to an end at an earlier stage of grammatical competence than in the normal adult speaker of German. In his investigation of language production of brain-damaged children, Lackner ( 1968) observed a language acquisition process which was delayed in onset and slowed down, but structurally similar to normal acquisition up to the point of arrest. Menyuk (1964) and Morehead and Ingram (1973) also reported considerable similarity between normal children and children with language disturbances in the development of syntactic transformation. These authors gave different explanations for the observed disorders in language production during language acquisition. Menyuk (1964 ), and later Lee (1966) suggested that the failure in syntactic development is due to a disability in applying increasingly differentiated syntactic rules. On the other hand Lackner (1968) assumed that all syntactic rules are fully .observed according to the level of syntactic development at which the arrest occurs. Finally, Morehead and Ingram (197 3) stressed again that there are some differences in language production between normal and linguistically delayed children. They suggest that the delayed children "were restricted in their ability to handle less general properties specified by the lexicon." In our case, different rule application because of a difference in lexical marking was also observed. (Chomsky, 1969, found the same for normal children at the age of 5 to 10). However, the more striking observation was that the range of application of rules is less restricted than in adult grammar. This leads to sentences with surface structures which, while complex, are nevertheless not to be found in the grammar of adult language. Although the critical period for the end of language acquisition is not later than puberty (Lefl!leberg, 1964, 1967), the language of our 23 year old patient includes grammatical forms from early stages of language develop
280
M. Kerschensteiner and W. Huber
ment as well as grammatical forms of the adult language. What sort of explanatory model could account for this? The non-existence of certain grammatical constraints specific for German could well be the linguistic reason for the incomplete development of the patient's language. Had the patient been able to acquire these grammatical contraints, he might have been able to make all the generalization necessary for the full acquisition of the syntactic competence of an adult speaker. From a neurological point of view the lack of ability to generalize those grammatical forms which belong to the adult grammar should be related to an incomplete maturation of underlying biolo gical mechanisms for language acquisition. As the neurological substrate for this incomplete maturation one may assume an insufficient development of certain inhibitory functions. These inhibitory functions would be the correlate of those grammatical constraints which are language specific in adult German. From the history of the patient, we consider the insufficient development of these inhibitory functions best be explained by a brain lesion as a result of birth injury. The maturation of the human brain can be regarded as constituting mainly the inhibition of earlier physiological levels of function. Such inhibitory processes are well known from many aspects of human deve lopment (e.g., the inhibition of grasping and groping reflexes of hand and mouth during maturation (Ajuriaguerra, Rego and Tissot, 1963; Binz and Brunner, 1973 ). Therefore one can easily assume that the deficient acquisition of language stems from a deficient development of inhibitory functions in the brain. Those parts of the brain which are responsible for the ability to gene ralize the grammatical forms of adult language cannot have reached that stage of maturation which makes possible the development of certain linguistic inhibitions. Many of the characteristic features we found in the language of our patient seem to parallel features of the language of certain aphasic patients. The common clinical expressions for those language deficits are agrammatism and paragrammatism (Pick, 1913; Kleist, 1934; Isserlin, 1936). Of course, there is no single aphasic patient and there is no clinical type of aphasia where these features c~m be found in the same combination as in the language of our patient. However, there are structural similarities both in the language of adult patients with developmental aphasia and in the language of adult patients with acquired aphasia. This leads us to the following tentative con clusion: Referring to Jackson's classical model of dissolution of function (Jackson, 1958), one may assume that there is a dissolution of certain gram matical function in adult aphasia, whereas in deveiopmental aphasia the inhibition of equivalent grammatical functions is not developed. In adult aphasia grammatical constraints and differentiations are reduced, thus leading to a language form in which grammatical structures of early stages of language development are set free. This hypothesis was alluded to before by Alajouanine (1956) in his discussion of agrammatism. In contrast to acquired adult aphasia,
Grammatical impairment in developmental aphasia
281
in developmental aphasia of adults such as our patient, inhibitory functions were not sufficiently developed during maturation. This leads to a language form in which many grammatical structures of the most advanced stages of development are acquired, but not as the predominant structures. This hypothesis is a weaker claim than the equation of language acquisition in the normal child and language reduction in the aphasic made by Jakobson (1969, originally in 1944) and later by Wepman and Jones (1964).
SuMMARY
The language production of a 23 year old patient with developmental aphasia caused by birth injury was analyzed in detail within the framework of Chomsky's generative grammar. It was shown that different stages of grammatical development occur simultaneously, without the predominance of those grammatical forms which are identical with or close to adult grammar. Thus the language of this patient with developmental aphasia is quite comparable to the language of children from 3 to 6, and shares many features with grammatical impairment well known from agrammatism and paragrammatism in adult aphasia. It was suggested that in developmental aphasia incomplete linguistic generalizations are related to an incomplete maturation of inhibitory functions during language acquisition, whereas in adult aphasia comparable forms of less advanced grammars are set free by dissolution of function.
Acknowledgment. The authors wish to thank Prof. Dr. K. Poeck, B. Orgass, Dr. W. Hartje and F.-J. Stachowiak for helpful discussion and criticism. REFERENCES AJURIAGUERRA, ]. DE, REGO, A., and TissoT, R. (1963) Le reflexe oral et quelques activites orates dans les syndromes dementiels du grand age, "Encephale," 52, 189-219. ALAJOUANINE, T. (1956) Verbal realization in aphasia, "Brain," 79, 1-28. BAUER, H. (1973) Klinik der Sprachstorungen, in Phoniatrie und Padoaudiologie, ed. by P. Biesalski, Thieme, Stuttgart. BENTON, A. L. (1964) Developmental aphasia and brain damage, "Cortex," 1, 40-52. BINZ, P., and BRUNNER, E. (1973) Ober Greifreflexe der Hand und des Mundes als Zeichen der allgemeinen Hirnschiidigung, "Z. Neurol.," 206, 61-77. BLOOM, L. (1970) Language Development, Form and Function in Emerging Grammars, M.I.T. Press, Cambridge, Mass. CHOMSKY, C. (1966) The Acquisition of Syntax in Children from 5 to 10, M.I.T. Press, Cambridge, Mass. CHOMSKY, N. (1965) Aspects of the Theory of Syntax, M.I.T. Press, Cambridge, Mass. (1973) Conditions on transformations, in A Festschrift for Morris Halle, ed. by S. R. Anderson and P. Kiparsky, Holt, Rinehart and Winston, New York. CRITCHLEY, M. (1970) Aphasiology and Other Aspects of Language, Edward Arnold, London. ErsENSON, ]. ( 1968a) Developmental aphasia (dyslogia): a postulation of a unitary concept of the disorder, "Cortex," 4, 184-200. (1968b) Developmental aphasia: a speculative view with therapeutic implications, "J. Speech Hear. Dis.," 3, 3-13. GRIMM, H. (1973) Strukturanalytische Untersuchung der Kindersprache, Hans Huber, Bern, Stuttgart, Wien. HuBER, W., and KuMMER, W. (1974) Transformationelle Syntax des Deutschen, Fink, Miinchen.
282
M. Kerschensteiner and W. Huber
INGRAM, T. T. S. (1969) Developmental disorders of speech, in Handbook of Clinical Neurology Vol. 4, ed. by P. J, Vinken and G. W. Bruyn, North-Holland Publ. Co., Amsterdam. IssERLIN, M. (1936) Aphasie, in Handbuch der Neurologie, Vol. 6, ed. by 0. Bumke und 0. Foerster, Springer, Berlin. JACKSON, J. H. (1958) The factors of insanities, in Selected Writings, Vol. 2 p. 411, ed. by Taylor, Staples Press, London. }AKOBSON, R. (1969) Kindersprache, Aphasie und Allgemeine Lautgesetze, Suhrkamp, Frankfurt. KLEIST, K. (1934) Gehirnpathologie, Barth, Leipzig. LACKNER, J. R. (1968) A developmental study of language behavior in retarded children, "Neuro psychologia," 6, 301-320. LEE, L. L. (1966) Developmental sentence types: A method for comparing normal and deviant syntactic development, "J. Speech Hear. Dis.," 31, 311-330. LEISCHNER, A. (1967) Die Sprachstorungen im Kindesalter, "Zeitschr. f. Kinderheilk.," 100, 137-159. LENNEBERG, E. H. (1964) Language disorders in childhood, "Harv. Educ. Rev.," 34, 152-177. (1967) Biological Foundations of Language, Wiley, New York. LowE, A. D., and CAMPBELL, R. A. (1965) Temporal discrimination in aphasoid and normal children, ''J. Speech Hear. Res.," 8, 313-314. McNEILL, D. (1970) The Acquisition of Language, Harper and Row, New York, Evanston, London. MENYUK, P. (1964) Comparison of grammar of children with functionally deviant and normal speech, "J. Speech Hear. Res.," 7, 105-108. MoREHEAD, D. N., and INGRAM, D. (1973) The development of base syntax in normal and linguistically deviant children, "Speech Hear. Res.," 16, 330-352. MYERSON, R., and GooDGLAss, H. (1972) Transformational grammars of three agrammatic patients, "Speech Hear. Res.," 15, 40-50. PICK, A. (1913) Die Agrammatischen Sprachstorungen, Springer, Berlin. REES, N. S. (1973) Auditory processing factors in language disorders: The view from Procuste's bed, ''J. Speech Hear. Dis.," 38, 304-315. Ross, J. R. (1967) Constraints on Variables in Syntax, Unpubl. Ph. Dissertation, M.I.T. STARK, J. A. (1967) Comparison of the performance of aphasic children on three sequencing tests, "J. Commun. Dis.," 1, 31-34. -, PoPPEN, R., and MAY, M. Z. (1967) Effects of alterations of prosodic features on the sequencing performance of aphasic children, ''J. Speech Hear. Res.," 10, 849-855. STOCKERT, F. G. VON (1969) Sprachstorung des Kindesalters mit Einschlu/3 der Aphasien, in Handbuch der Kinderheilkunde, ed. by H. Opitz und F. Schmid, Vol. VIII/1. Springer, Berlin. TALLAL, P., and PIERCY, M. (1973a) Defects of nonverbal auditory perception in children with developmental aphasia, "Nature," 241, 468-469. -, - (1973b) Developmental aphasia: impaired rate of nonverbal processing as a function of sensory modality, "Neuropsychologia," 11, 389-398. -, - (1974) Developmental aphasia: rate of auditory processing and selective impairment of consonant perception, "Neuropsychologia," 12, 83-93. WEINER, P. S. (1969) The perceptual level functioning of dysphasic children, "Cortex," 5, 440-457. (1972) The perceptual level functioning of dysphasic children: a follow-up-study, "J, Speech Hear. Res.," 15, 413-422. WEPMAN, J. M., and JoNES, L. V. (1964) Five aphasias. A commentary on aphasia as a regressive linguistic phenomenon, "Ass. Res. Nerv. Ment. Dis.," 42, 190-203.
Dr. Max Kerschensteiner and Walter Huber, Abt. Neurologie der Med. Fakultiit der RWTH Aachen, 51 Aachen, Geothestr. 27-29, West Germany.