David Hill readings reprinted from The Politics of Schizophrenia: psychiatric oppression in the United States 1983 (University Press of America 1983) with the permission of the author (currently not in print).
Chapter 10
EMIL KRAEPELIN: DEFINING THE NON-EXISTENT
Let me dismiss from the outset the idea that Kraepelin's contributions are of historical interest only and have little real impact on the conceptualizations and treatments of the 1980s. Shershow (1978) characterizes his work as "an approach to psychiatric disorders that is regarded as the foundation stone of the medical model in contemporary psychiatry" (p.22). Menninger (1963) believes that "Kraepelin's lifelong work represents probably the greatest nosological synthesis ever accomplished in psychiatry" (p.462). Arieti (1974) points out that "even today Kraepelian psychiatry is the best known in the world. Thousands and thousands of patients are still viewed and classified as Kraepelin taught, and until the middle 194Os in the United States, too, they were still labelled with the name dementia praecox" (p.12).
Kraepelin's "Psychiatrica" was published in eight editions from 1883 to 1913. His 'dementia praecox' notion first appeared, in a 10 page presentation, in the fourth edition, published in 1893. As his enthusiasm for the construct grew, so did the length of its description; to 31 pages in the fifth edition, 77 in the sixth, and 187 in the seventh. My critique is based on his eighth, and final, edition which included more than 300 pages on the subject of 'dementia praecox'. One of Kraepelin's major tasks in describing and supporting his new entity was to state its defining characteristic. In the absence of any evidence of a physiological etiology (i.e., cause) the task became essentially that of grouping various behaviors that seemed, in Kraepelin's thinking, to belong together. The criterion for the grouping by which he had defined the original (1893) version had been any behaviors associated with the onset, in adolescence, of a process of irreversible deterioration of the psychological function, resulting in dementia. He soon observed, however, similarities between behaviors grouped under the heading 'dementia praecox' and those supposed to be symptoms of two other diseases generated by the 19th century, 'catatonia' and 'dementia paranoides'. By 1913 he had lumped all three together and named this new, more inclusive construct 'dementia praecox, even though the problems of many of the people so diagnosed neither began in adolescence nor resulted in dementia.
By 1913 his attempts to define this new construct consisted of a mixture of the prognostic approach--still founded on the notion of progressive dementia, the symptomatic approach--attempting to document behaviors indicative of the disease, and the etiological approach--constituting little more than frequent references to the undiscovered, unspecified "morbid process". The three approaches are intertwined and cross-referenced in such a way as to create the convincing illusion that there is some theoretical substance to the entity he had created. A separate examination of each approach, however, reveals serious logical flaws and a lack of empirical data.
The Prognostic Approach
Definition of his new construct in terms of prognosis (eventual outcome) comes in the form of a division into two major categories of those individuals whose behavior could not be explained by disease entities such as dementia paralytica. The essential criteria for this new division was whether or not the individual 'got better'. Those who did were to be classified as 'manic.depressive',and those who did not as having 'dementia praecox . To be more specific and, perhaps, more fair, Kraepelin was focussed on the presence or absence of "Die Psychischen Entartungs" (the psychological degeneration processes), characterized by "the rapid development of a permanent state of psychological weakness" (Kraepelin, 1913, p.xiii). Ernest Harms, in his introduction to the 1971 translation of Kraepelin's eighth edition, makes the distinction between diagnostic and symptomatic approaches to diagnosis, applauding Kraepelin's emphasis on the former:
Pointing out clearly individual types of disturbances will lead us to the establishment of specific diagnoses . . The real character and meaning of such symptoms, however, are really not clear until we also observe the course and the end of the disease. . . He underlined that proper diagnosis needed first of all a careful prognosis. . . . Kraepelin demanded that prognosis ought to be so didactically developed that it would be possible to teach it as a specific method. (pp. xii,xiii)
The value of Kraepelin's introducing longtitudinal observation cannot be denied. Nevertheless, when outcome is employed as a defining characteristic of a disease, one is engaging in a dubious form of logic, especially if one cites a particular outcome as evidence for the existence of the disease. The tautology is obvious. First one excludes from the construct those individuals who do not meet the outcome criterion. Second, one gives the construct a title--x. Third, one cites as evidence for the construct the fact that all those who have disease x also have a common outcome--y. That disease x results in outcome y is a hypothesis which cannot be disproved if x is defined in terms of the presence or absence of y. The logical loophole is that the disconfirming evidence of any outcome other than y can result in the exclusion of that individual from the x construct. Kraepelin engages in precisely this process:
The undoubted inadequacy of my former classification has led me once more to undertake the attempt to make a more natural grouping, as I have in hand a larger number of possibly more reliable cases. . . "Recovered" cases were not taken into account because of the uncertainty of their significance which still exists, but only such cases as had led to profound dementia or to distinctly marked and permanent phenomena of decreased function. (pp.89,90)
Kraepelin appears to have made the additional error of confusing the process of increasing reliabilty (i.e excluding cases which don't fit his definition, of the construct) with the process of discovering 'natural groupings'. Kraepelin, like his predecessors, seems to believe that he has uncovered some natural phenomenon when he has done nothing more than add, albeit with a new approach, to the list of person-made categories of 'mental illnesses'. Inevitably the first step on the path to discovering a 'law of nature' is often the creation of a hypothesis designed to approximate the 'real world'. It is crucially important, however, to maintain the distinction between the two at least until we have tested our latest definition of reality by various processes which have come to be known as 'construct validation'. Some would argue that even our most thoroughly validated constructs represent nothing more than the closest approximation available. In this light Kraepelin's conclusion that "the proof is furnished that our picture of dementia praecox is in the main agreeable to natural law" (p.252), apart from being based on circuitous logic, seems extremely premature. In short, he creates the illusion of discovery where there is nothing more than a relatively crude form of invention.
The prognostic approach to establishing a disease entity has additional problem. Its retroactive nature, as compared to the traditional etiological approach, led Arieti to declare it "incompatible with the scientific method" (1974, p.11). Sullivan, as early as 1927, noted:
We are but beginning to free ourselves of many misconceptions concerning "dementia praecox", and the majority still revise diagnosis when the patient recovers. The Kraeplinian diagnosis by outcome has been a great handicap, leading to much retrospective distortion of data, instead of careful observation and induction. (1927, pp. 759,760)
In the same paper, Sullivan suggests that:
Psychiatric prognosis may best be considered as a specialized technique in social psychology. Its problem is the prediction of the future adaptability of an individual within some more or less clearly envisaged milieu composed principally of people. (p. 759)
Sullivan's efforts to understand the social context in which behavior occurs is matched by Kraepelin's ability to ignore it. The latter makes no mention of the fact that the diagnosis, by itself, created such negative expectations as to risk constituting a self-fulfilling prophecy. The effects of institutionalization had been clearly stated by Pinel; Falret had suggested the effects of the labelling process. Kraepelin's label not only implies but is actually defined by hn outcome of irreversible dementia. Only by a rigid focus on the physiological could the self-fulfilling nature of such a label be ignored.
The Symptomatic Approach
Diagnosis could not, of course, be delayed until the outcome became known. Some form of symptomatic approach was inevitable. Diagnosing on the basis of present behavior presented Kraepelin with two serious problems--both of which are predictable from the premise that all he had really accomplished was the grouping of relatively extreme deviant behaviors of very different kinds. The problems were heterogeneity (i.e., variability) of outcome and heterogeneity of symptoms. Kraepelin realized that prognosis could not stand alone as the central assumption concerning the nature of the disease. Perhaps the number of patients whose improvement necessitated a change of diagnosis was becoming too great. By 1913, he acknowledged that "the general course of dementia praecox is very variable" (p.4). Such a realization--especially after attempts to exclude such evidence by rediagnosis--would make many researchers wonder about the validity of their Construct; some might reach the conclusion that this particular grouping of behaviors did not merit any further investigation. For Kraepelin, however, it is only the name of the construct, not the construct itself, which is challenged by such data:
It has since been found that the assumptions upon which the name rested are at least doubtful. . . . The possibility cannot . . . be disputed that a certain number of cases of dementia praecox at tain to complete and permanent recovery, and also the relations to the period of youth do not appear to be without exception. . . . If therefore the name which is in dispute, even though it has been already fairly generally adopted, is to be replaced by another, it is to be hoped that it will not soon share the same fate of so many names of the kind, and of dementia praecox itself in giving a view of the nature of the disease which will turn out to be doubtful or wrong. (p.4)
His next recommendation provides an even stronger demonstration of his dogmatism concerning this construct. He argues that "a name that as far as possible said nothing would be preferable" (p.4). To proclaim a disease entity in the absence of a proven etiology locates Kraepelin's work on the borders, at best, of medical science. To argue for a meaningless name in order to avoid any assumptions about the nature of the disease which might later be disproved appears to place his thinking beyond the realms of science altogether.
Variability of outcome might have been less problematic for Kraepelin's construct if he could have demonstrated a relatively homogenous symptomatology. Kraepelin repeatedly acknowledges, however, both the heterogeneity of symptoms within the dementia praecox construct and also the considerable overlap of symptoms with other diseases.
The most crucial problem with the symptomatic approach to validating a construct is the exact same criticism previously applied to the prognostic approach--that of circular logic. Having decided that certain behaviors are signs of some disease, one then 'demonstrates' the existence of that disease by pointing out the occurence of those behaviors in people who have been diagnosed as having the disease. As with the myriad of previous categorizations, nothing has been accomplished other than a grouping of behaviors and the application of a label. Such circular logic would be a problem even if only one or two well-defined behavior patterns were all that were being categorized. The greater the range of behaviors categorized as symptoms of the same disease, the greater becomes the risk of an arbitrary grouping process. Kraepelin described 36 types of "psychic symptoms' and 19 "bodily symptoms". Acknowledgement of the problem posed by such an extreme range of symptoms, and his attempt to solve it--once again, employing the most obvious type of circular logic--are evident in the following statement:
The complexity of the conditions which we observe in the domain of dementia praecox is very great, so that their inner connection is at first recognizable only by their occurring one after the other in the course of the same disease. (p.5)
On other occasions, when attempting to account for such an extreme range of symptoms, Kraepelin implies some underlying physiological process which, in some unexplained way, expresses itself differently in different people at different times:
We are justified in regarding the majority at least of the clinical pictures which are brought together here as the expression of a single morbid process, though outwardly they often diverge very far from one another. (p.3)
Despite his warning about making assumptions which might later be proved "doubtful or wrong,' Kraepelin was willing to speculate about which symptoms were more "fundamental" than others:
Everywhere the same fundamental disorders return again, the loss of inner unity in thinking, feeling, and acting, the dulling of higher emotions, the manifold and peculiar disorders of volition with the connected delusions of psychic constraint and influence, lastly the decay of the personality with comparatively slight damage to acquired knowledge and subordinate expertness. (p. 255)
He is then faced with the problem that even these most fundamental disorders are not found in all cases. His first response is to blame inadequate observation:
These features are certainly not demonstrable with full clearness in each individual case. But still the general view over a great number of complete observations teaches that nowhere can a state be discovered which is not connected by imperceptible transitions with all the others. (p. 255)
The message here would appear to be that if one looks hard enough one can perceive the imperceptible, one can find a disease where there is only an arbitrary grouping of undesirable behaviors. The same 'if you can't find it you're not trying hard enough' argument is employed in another attempted solution which takes the approach that it's not individual symptoms that matter anyway but the general impression:
Unfortunately there is in the domain of psychic disorders no single morbid symptom which is thoroughly characteristic of a definite malady. . . . On the other hand we may expect that the composition of the entire picture made up of its various individual features, and especially also the changes which it under goes in the course of the disease, could scarcely be produced in exactly the same way by diseases of a wholly different kind: at this or that point, sooner or later, deviations will be certain to appear, consideration of which makes possible for us the distinguishing of the morbid forms. It may in certain circumstances be very difficult, not only to judge correctly of the diagnostic significance of such deviations, but even to recognize their very existence. (p. 257)
We see here an attempt to integrate the symtomatic and prognostic approaches. In fact, Kraepelin's decision as to which symptoms are supposed to be 'fundamental' was not based, as the traditions of medical science would dictate, on a theoretical link to physiological processes, but on his original prognosis idea:
Those must therefore be regarded as fundamental disorders, which meet us in the picture of dementia simplex and in the terminal state called simple weak-mindedness. (p. 248)
Ignoring the problems--outlined above--of each of the approaches, Kraepelin attempts to integrate the two, thereby creating the illusion of a substantive theory. The pitfalls involved are evident when he explains problematic data generated by one approach via reference to the other. At one point, for instance, he argues that people who recovered (and thereby challenged his prognostic approach) were really still suffering from dementia praecox, since their symptoms had been similar to the symptoms of those who did not recover. He also argues, with similarly dubious logic, that individuals with markedly varying symptoms were suffering from the a same disease because they have the same prognosis! In fact his previously discussed conclusion that he had discovered some natural law is based on precisely this process of cross-referencing the two approaches.
Given the serious logical errors inherent in each of these two approaches, to explain either in terms of the other, or to defend against the accusation of "unjustified grouping" (p.252) by attempting to integrate the two, leads Kraepelin into what might best be described as a spiral of illogicality. An extension of this spiral occurs when, in his etiological approach, he attempts to use a completely unspecified physiological etiology to account for data that is problematic for either his prognostic or symptomatic theorizing.
The Etiological Approach
The etiological (causation) approach to establishing a disease entity had become the norm within medical science. Kraepelin was not overtly challenging the traditional approach so much as employing a new form of it. In his version, the physiological basis to the disease was something to be discovered later rather than something which, having already been demonstrated, formed the theoretical foundation of the construct. In the traditional etiological approach the symptoms and prognosis were required to make sense in relation to a given etiology; the latter somehow explaining the former and thereby creating testable hypotheses. With the physiological basis left undefined one is free to conjure up an etiology that accounts for any inconsistencies or problematic data generated by whatever approaches one has employed--in Kraepelin's case the symptomatic and the prognostic. Such hypothetical explanations cannot be tested.
Kraepelin makes no claims to have discovered a physiological etiology. By his own admission:
The obscurity that hangs over the causes of dementia praecox has been a frequent motive for the examination of the blood-picture and of metabolism, but the findings up to now are not very satisfactory. (p. 85)
Kraepelin was, nevertheless, quite willing to speculate about etiology although he is careful to avoid being too specific--possibly for the same reason that he advocated the use of a meaningless name:
The absence of external causes in this disease . . . might perhaps suggest that we have an illness in which the final cause must be sought in the metabolism disturbances of the tissue. (1904)
He postulated various changes in cells in various layers of the cortex but then acknowledges that:
As the significance of the cortical layers is at present still almost wholly unknown, it will scarcely be possible to set up hypotheses with regard to the influence of the site of the morbid processes in definite layers (p. 220)
Kraepelin concludes that "the causes of dementia praecox are at the present time still wrapped in impenetrable darkness" (p.224). Note that Kraepelin, and Bleuler in his turn, repeatedly include in their admissions of ignorance such terms as 'at present" and up to now", clearly implying the expectation that such ignorance will be short lived. Never does Kraepelin even consider the possibility that there is no physiological etiology, that there is no disease entity. Instead, with absolutely no data to support his claim, Kraepelin asserts that:
We are justified in regarding the majority at least of the clinical pictures which are brought together here as the expression of a single morbid process. (p. 3)
Repeated reference is made to this unspecified and undiscovered "morbid process" in order to account for the otherwise inexplicable variety of behaviors and outcomes that Kraepelin is determined to group together under the heading 'dementia praecox'. It is as if there were some invisible cement filling in the gaps in his theory, binding together behaviors as dissimilar as "involuntary obedience and "negativism", "mutism" and "verbigeration", and outcomes as divergent as "profound dementia" and "complete recovery".
Before considering Bleuler's contributions we should note that Kraepelin himself was aware of at least some of the shortcomings discussed here. Many of his chapters include warnings about the tentativeness of his hypotheses. We have already noted his admission of ignorance concerning a physiological etiology, and acknowledgements of the heterogeneity of both symptoms and outcomes. His chapter on "Clinical Forms", considered to be one of his more valuable contributions, opens with the following disclaimer:
The presentation of clinical details in the large domain of dementia praecox meets with considerable difficulties, because a delimitation of the different clinical pictures can only be accomplished artificially. There is certainly a whole series of phases which frequently return, but between them there are such numerous transitions that in spite of all efforts it appears impossible at present to delimit them sharply and to assign each case without objection to a definite form. (p. 89)
Once again there is a disclaimer within the disclaimer; its not that its impossible to find meaningful types of dementia praecox, its "impossible at present."
Kraepelin's own introduction to his eighth edition account of dementia praecox begins with:
A series of morbid pictures are here brought together, under the term "endogenous dementias" merely for the purpose of preliminary inquiry. (p.1)
In psychiatry's eagerness to establish itself as a bona fide medical specialty, however, the warnings and disclaimers were all too readily forgotten or ignored. Bleuler notes that:
After paresis was excluded from among the "functional psychoses" and the other organic forms followed of themselves, for seventy years theoretical psychiatry stood entirely helpless before the chaos of the most frequent mental diseases. (1924, p.372)
Psychiatry could no longer survive if it 'continued to admit that the best it could come up with was "preliminary inquiry" and "impenetrable darkness". Rather than consider the possibility that their failures were due to attempting to impose a medical paradigm on a non-medical problem, psychiatry simply chose to ignore the absence of data, the circuitous logic and even the warnings of their founding parent Emil Kraepelin.
Even a most sympathetic biographer documents that "Kraepelin's supremely controlling object was the furtherance of psychiatry" (Jelliffe, 1932, p.764). There simply was no motivation to consider the possibility that all that was really being accomplished was one more sophisticated rationalization for the social control of unusual behavior. The medical model was, indeed, tightening its stranglehold on the market of deviancy, and for this process there was considerable motivation, notably in the form of reinforcements from those who had most to gain from controlling deviance and thereby maintaining the status quo:
Kraepelin succeeded in obtaining from the Rockefeller Foundation the
sum of $250,000 for his project, the first gift to be made by this organization
to research work in Europe. A further gift from the Rockefeller Foundation
increased this sum to $325,000 under the stipulation that there should be
strict adherence in the institute to Kraepelin's fundamental idea (Jelliffe, 1932, p.768)