J O U R N A L O F T H E S O U T H A F R I C A N L O G O P E D I C S O C I E T Y THE CORTICAL LOCALISATION OF SPEECH; 1. An Analysis of Preliminary Difficulties M. WRIGHT. M.Sc.. M.B.. B.Ch. (Formerly Registrar at the National Hospital for Nervous Diseases, Queen Square, London. At present electroencephalographer to Tara Hospital.) INTRODUCTION. The initial premise of the discussion is that a lack of an accepted and relevant analysis of language, a s an aspect of human behaviour and experience, has necessarily caused a diversity in the classification, description and explanation of partial l a n g u a g e defects pro- duced by localised cerebral d a m a g e . The formulation of the previous sentence reveals that a strong element of tautology may lie within it. Hence advocates of any particu- lar description of aphasic defects might agree that the premise essentially is a tautology, but assert that they possess the only relevant analysis of l a n g u a g e ; alternatively they could deny both the validity and the tautologous nature of the premise on the grounds that an accepted and adequate analysis of l a n g u a g e is not lacking (or perhaps is unattainable) and that diversity of terminology depends solely on disagreement concerning the physiological or psychological mechanisms whereby cerebral d a m a g e leads to partial failure of the l a n g u a g e process. The latter opinion seems untenable since the various terms applied to the aphasia reflect a divergence in the ap- proach to l a n g u a g e behaviour rather than a disagreement on cerebral mechanisms ; "sen- sory" (aphasia) and "motor" (aphasia) presup- pose an analysis of l a n g u a g e in the manner of clinical neurology, "expressive" and "recep- tive" a p h a s i a (Weisenberg and MacBride (10)) are a product of psychology, while "nominal" and "syntactic" are adjectives borrowed from the schoolroom study of grammar. Such funda- mental differences of approach preclude even a preliminary discussion of physiological mechanisms since the initial step in any scien- tific investigation, namely the formulation of a problem in relation to specific techniques, has a s yet been left undone. Two examples (one hypothetical, another historical) derived from the neurological description of sensation illus- trate and further emphasise the site of "patho- logy" in those clinical descriptions of language defects which presuppose on everyday familiar analysis of l a n g u a g e :— Imagine the incredible difficulty which would result if the ascending fibre tracts in the spinal cord should have to be described without the analysis of sensation offered by the everyday words "pain," "sight" (vision), "touch" etc. The outcome could only be an anatomical descrip- tion of nerve processes and pathological lesions, on which much useless philosophical comment could well be raised. However, con- fusion is averted b e c a u s e a n analysis of sen- sation does exist; moreover the analysis is both adequate and relevant since it is incor- porated within every l a n g u a g e and thus can only be criticised a s being too subjective if it b e supposed that the whole of humanity is deluded. The second example concerns the deleterious effect on the physiology of sensa- tion caused by the introduction of the two descriptive terms "epicritic" and "protopcrthic" sensation. Walshe (9) has displayed the logical non sequiturs that are implied in the original definition of their terms and also the errors in clinical and physiological interpreta- tions that may result from their use. The two terms are b a s e d on an evolutionary, and thus necessarily hypothetical, interpretation of cer- tain clinical findings following nerve lesions ; they belong to abstract theory and not to observation nor convention ; they must stand or fall on an empirical test of their usefulness to neurophysiology and they are discarded be- cause they are neither an adequate nor a relevant addition to the analysis of sensation. The argument that any particular classifica- tion of the ap hasia is b a s e d on an adequately instructed analysis of l a n g u a g e can only be refuted by showing that the analysis in ques- tion is not acceptable because it is not relevant. It is suggested that relevance in this c a s e can be equated with the notion "useful for physio- logical or experimental psychological investi- gations." Thus an analysis of l a n g u a g e must at least contain terms definable by direct observation or by precise logical deduction from direct observation, and the formal state- ment of the analysis must have intrinsic logical R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 2) J O U R N A L O F T H E S O U T H A F R I C A N L O G O P E D I C S O C I E T Y O C T O B E R cohesion; once the latter criteria are met, there remains the empirical test of whether there is enough useful correlation between the formal analysis and the processes of l a n g u a g e a s re- vealed in normal and aphasic subjects. The result of the empirical test determines the ade- quacy and relevance of the analysis. It follows from the two preceding para- graphs that no purely introspective or "men- talistic" description of l a n g u a g e is acceptable b e c a u s e it must lack definitions b a s e d upon impersonal observation and b e c a u s e it has not peculiar authority, possessed by the terms for subjective sensations, of incorporation within the everyday language of all people. The apparent insight wisdom and complexity of these theories represent a very real intellectual danger since, like all "supraterrestrial edifices," their castles in the air can be suitably moulded to fit successively the earthy environment of diverse and discrete observations. No less dangerous are theories which openly eschew "introspection" but forthwith rise to the even more rarefied atmosphere of "total cerebral function," "oragnism a s a whole" and "integ- rated personality." A second premise may now be stated a s follows; the analysis of the content, logical form and everyday use of l a n g u a g e belongs to the field of modern logic. It is probably true that many neurologists and speech therapists have never even con- sidered the significance of the specific analysis of l a n g u a g e a s a separate field of investiga- tion ; the term language behaviour may indeed seem unreal, unnecessary or even misleading. The clinical neurologist has approached speech via the spinal cord and brain-stem, thus, not unnaturally, talks of "motor" and "sensory" aphasia. Unfortunately the terms "motor" and "sensory" possess but a fraction of their "spinal usefulness" for even the simplest problems of cortical physiology; they are necessarily even less effectual a s descrip- tive elements of the cortical mechanisms under- lying languarge. Modern electrophysiological research on the cerebral cortex has produced a terminology which is so specific to the ex- perimental operations employed that it can have no application to the analysis of aphasia. Moreover electrophysiology is an aspect of biophysics rather than clinical neurology and its techniques and terms are even less com- prehensible to the clinician (who is not trained specifically a s a scientist nor a s a physiolo- gist) than are the methods of logic. The speech therapist, on the other hand, approaches cerebral mechanisms from the more detached discipline of psychology (which assertion is not derived from first hand experience and is open to correction). The logopaedic approach, while more promising, would do well to avoid introspective analyses of l a n g u a g e and especi- ally to avoid the presumption that clinical neurologists are in any way better equipped to analyse language in the normal or the abnormal; to which, surely, the medical litera- ture on ap hasia is testimony enough. There are several relatively recent publica- tions by competent logicians dealing especi- ally with the analysis of l a n g u a g e processes. Those of Morris (6, 7, and 8), Carnap (2, 3 and 4),Bloomfield (1) and Longer (5) are par- ticularly relevant to the theme. Within their works there is general agreement on the frame- work of the logical analysis of l a n g u a g e ; the scope and the precision of development of ideas show convincingly that this field of logic is not one into which an amateur may stray without much purposeful effort. To quote from Bloomfield (1, p. 54-55):— "The subject matter of linguistics, of course, is human speech. Other activities, such a s writing, which serve a s substitutes for speech, concern linguistics only in their semiotic as- pect, a s representations of phonemes or speech-forms. Since the meanings of speech cover everything (designcrta, including deno- tata ; syntactic relations; pragmatic slants), linguistics, even more than other branches of science, depends for its range and accuracy upon the success of science a s a whole. For the most part, our statements of meaning are makeshift. Even if this were not the case, linguistics would still study forms first and then look into their meanings, since l a n g u a g e con- sists in the human response to the flow and variety of the world by simple sequences of a very few typical speech-sounds. Linguistics is the chief contributor to semio- tic. Among the special branches of science, it intervenes between biology, on the one hand, and ethnology, sociology, and psychology, on the other : it stands between physical and cul- tural anthropology. L a n g u a g e establishes, by means of sound waves and on the basis of communal habit, an ever ready connection between the bodies of individuals—a connection between their nervous systems which enables each person to respond to the stimuli that act upon other persons. The division of labour, civilization, and culture arise from this interaction. Popu- larly and even, to a large extent, academic- R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 2) J O U R N A L O F T H E S O U T H A F R I C A N L O G O P E D I C S O C I E T Y ally, we are not accustomed to observing l a n g u a g e and its effects: these effects are generally explained instead by the postula- tion of "mental" factors. In the cosmos, langu- a g e produces human society, a structure more complex than the individual, related to him somewhat a s the many-celled organism is re- lated to the single cell." Evidently there has been no lack of effort on the part of logicians. Therefore it seems to me significant that I can find no reference to such work in clinical descriptions of speech defects, and that in teaching neurology to logo- paedic students there have been none who have heard of the logicians quoted above. At least an attempt at correlation would be in- teresting if not fruitful. FORMULATION SUITABLE FOR THE EMPIRICAL INVESTIGATION The development of semiotic, the science of signs and languages, has been extended by Morris (1938, 1946) in the vigorous manner of mathematical logic and also in the empirical tradition of objective psychology. In the following paragraph some of *he terms from semiotic are explained briefly, but there will b e no attempt to follow the precision charac- teristic of Morris or Carnap (1943). The description of l a n g u a g e processes (i.e., semiosis) may be divided into the related spheres of semantics, syntactics and pragma- tics. Semantics deals with any empirical observation or logical analysis concerning the relationships (i.e., the semantic relationships) existing between a linguistic expression and the object or event to which that expression refers; syntactics is concerned with the rela- tions (i.e., syntactic relations) between two or more of the numerous symbols (generally words) within a l a n g u a g e ; pragmatics describes ι relationships (i.e., pragmatic rela- tion) between linguistic expressions and the overt behaviour of such individual(s) who may utter or respond to the expressions. It is sub- mitted that these notions constitute a pre- liminary analysis of language which does not outrage common sense and which presumably satisfies the logical criteria for a potentially "useful" analysis since it is a produce of com- petent logicians. It gives power, within limits, of abstracting three groups of entities (i.e., "objects of the physical world," "words," "human behaviour") and providing objective descriptions of rela- tions existing within or between the groups. That a process of abstraction is involved is admitted, indeed it is axiomatic, but the ab- stractions are made deliberately and are well controlled; surely it is unreasonable to insist always that the process of l a n g u a g e "must be considered a s a whole" simply be- cause it is impossible to do so. Applying the three primary divisions of semiotic to descriptions of partial l a n g u a g e defects resulting from localised cerebral dam- age,it seems that any description in which specific reference can b e confined to the re- lations between words and objects to which they refer belongs to the field of descriptive semantics; any description in which specific reference can be confined to the relations be- tween linguistic expressions (e.g., in the formu- lation of sentences) belongs to the sphere of descriptive syntactics, while descriptions in which specific and necessary reference must b e made to the patient's response to words (e.g., his understanding of words) belong to descriptive pragmatics. In any discussion of a partial l a n g u a g e defect, it is, therefore, im- portant to decide whether reference to the patient (by name or personal pronoun) is in- cidental or whether it is a necessary part of the description. The' test situation should prompt the decision; if, for example, an object is shown and the patient names it, then the fact that the name is pronounced by the patient could be noted by several independent indivi- duals and could b e described without refer- ence to a particular patient; if, on the other hand, the patient's response to a written or spoken word is such that he apparently does not understand that world, then the defect can only be described by reference to the patient's behaviour. It is submitted that the common tests for ap hasi a and verbal agnosias can be classi- fied into three groups which test the integrity of semantic, syntactic and pragmatic relation- ships respectively. Further, the character of the test (the stimulus) and the observed re- sponse can be formulated so that an objective description may be given of any defect which is revealed. Hence the terms semantic aphasia, syntactic aphasia and pragmatic aphasia are suggested on the grounds that they are readily and rather precisely defined from the test situation, and because they are derived from an adequate analysis of langu- age. It then is necessary to decide if the analysis of language, and the terminology of l a n g u a g e defects derived from it, is not only adequate but also relevant. This is a matter for empirical investigation which must deter- mine whether the proposed nomenclature "fits" defects actually encountered in clinical neurology. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 2) J O U R N A L O F T H E S O U T H A F R I C A N L O G O P E D C S O C I E T Y O C T O B E R ELEMENTARY EMPIRICAL TESTING This section must be the least complete with- in itself, for the field of possible amplification and application of terminology to clinical con- ditions is immense. One amplification will be introduced at once; each of the three primary types of application can be revealed by stimuli directed to the patient by one sensory channel (i.e., sight, hearing, touch) alone). It is thus pos- sible to recognise visual, auditory and tactile varieties of semantic a p h a s i a ; visual and audi- tory subdivisions of syntactic aphasia, and visual and auditory varieties of pragmatic aphasia. The mode of subdivision is very similar to that used by Nielsen (8) for types of agnosia and in the latter application has been proved very useful. A further amplification on the basis of the site of pathology will not be attempted since there is no s p a c e to present the evidence for cerebral localisation of the lesions producing each type of aphasia ; an- other amplification b a s e d on the type of "words" used in the test situation (i.e., "nouns" or "abstract words") suggests itself but I am not competent to apply the idea. The one application, which is chosen be- cause it seems conclusive, is simply that types of a p h a s i a which would fall under the cate- gories semantic, syntactic and pragmatic do occur in clinical neurology. Semantic a p h a s i a s are described by Nielsen (8) a s amnesic a p h a s i a s ; syntactic a ph a s ia is equivalent to the latter author's formulation aphasia, while pragmatic aph a s i a is called semantic aphasia by Nielsen. The use of "semantic" by Niel- sen and his predecessors is indeed unfortun- ate and depended on the older and wider sense of the word semantic by which it in- cluded the whole of semiotic (i.e., it included semantics, syntactics and pragmatics). The case for renaming the older semantic aphasia, now calling it pragmatic aphasia, rests on the more authoritative claim of the modern an- alysis of l a n g u a g e and hence of "meaning." In my experience, the use of the proposed terminology gives a neater view of speech defects; it gives also further insight into the physiological mechanism of speech in the cerebral cortex, which insight is a powerful weapon upon the view that speech processes possess no precise cortical localisation; while finally it gives the most hopeful promise of clearing the "jargon" by which many descrip- tions of a p h a s i a induce a marked degree of pragmatic ap ha s i a among the audience. "Pure" a p h a s i a s are very rare, but surely then, when they do occur, they should be studied by the best possible techniques among which the logical analysis of l a n g u a g e is certainly to be numbered. BIBLIOGRAPHY. 1. Blomfield, L.—"Linguistic Aspects of Science." In- ternational Encyclopaedia of Unified Science, Vol. 1, No. 4. Chicago. University of C h i c a g o Press, 1939. 2. C a r n a p , R.—"Foundations of Logic and Mathema- tics.' International E n c y c l o p a e d i a of Unified Science, Vol. 1, No. 3. Chicago. University o! C h i c a g o Press. 1939. 3. Carnap, R.—"Formalisation of Logic.' C a m b r i d g e . Harvard University Press, 1943. 4. C a r n a p , R.—"Meaning and Necessity.' C h i c a g o : University of C h i c a g o Press, 1947. 5. Langer, S. K.—"Philosophy in a New Key." Cam- bridge : Harvard University Press. 6. Morris, C.—"Foundations of the Theory of S i g n s . " International Encyclopedia of Unified Science. Vol. 1, No. 2. C h i c a g o : University of C h i c a g o Press, 1938. 7. Morris, C.—"Signs, L a n g u a g e and Behaviour." New York : Prentice-Hal!, Inc., 1946. 8. Nielsen, J. M.—"Agnosia, Apraxia, A p h a s i a . " New York : Paul B. Hoeber, Inc., 1946. 9. Walshe, F. M. R.—"The Anatomy and Physiology of Cutaneous Sensation. A critical review." Brain. Vol. 6 5 : 48-112, 1942. 10. Weisenberg, Τ. H. and MacBride.—"Aphasia.' Brattleboro : E. L. Hildreth and Co. Inc., 1935. Books on Speech Defects IMPROVING THE CHILD'S SPEECH : Virgil A. Anderson. Price : 34/-. Post 1/3 SPEECH CORRECTION THROUGH STORY TELLING UNITS : E. McG. Nemoy. Price : 35/-. Post 1/3. LANGUAGE and LANGUAGE DISTURBANCES : Aphasic Symptom Complexes and their Significance for Medicine and Theory of Language. K. Goldstein. Price : 81/3. Post 1/6. SPEECH H A B I T A T I O N IN CEREBRAL PALSY : Marion T. Cass. Price : 25/6. Post I / - . CEREBRAL PALSY : J. F. Pohl. Price: 46/3. Post I / - . APHASIA THERAPEUTICS : Longerlch and Bordeaux. Price : 32/-. Post I / - . DIAGNOSTIC MANUAL IN SPEECH CORRECTION : Wendell Johnson et Al. Price: 23/3. Post 1/3. SPEECH IMPROVEMENT CARDS : Bryngelson and Glaspey. Price: 55/6. Post 1/6. STUTTERING : E. F. Hahn. Price : 25/6. Post I / - . VANGUARD B O O K S E L L E R S 23, JOUBERT STREET, JOHANNESBURG Telephone 23-3511 R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 2)