Schizophrenia a Conceptual History

A conceptual history of schizophrenia

Topics covered:

  • The ‘continuity' version
  • The ‘discontinuity' version
  • History of terms and concepts
  • History of behaviours
  • Conclusions
  • References


According to some, the history of schizophrenia consists of a progression of definitions culminating in the present. (1) However, instead of the ‘continuity' implicit in this view, historical research indicates that (a) the history of ‘schizophrenia' is a series of unconnected and contradicting research programmes, and (b) the current definition of schizophrenia is a patchwork of features.

The ‘continuity' version also includes only alienists making modern-sounding points. This would matter little were it not that it denies researchers access to important aspects of the history of schizophrenia. (2) For example, it is a moot point whether the Kraepelinian view would be as popular as it is had it not been for the untimely death of Wernicke, who by 1905 was developing an exciting neuropsychological classification for the psychoses.

Hence, rather than asking who were the alienists who ‘foresaw' the wonders of the present, this article will ask what historical factors made some views survive, the point being that ‘unsuccessful' views have contributed to the history of schizophrenia as much as ‘victorious' concepts. Also, clinical criteria are meaningless if taken out of historical context, an approach sometimes called ‘polythetic' (3) For example, ‘first-rank symptoms' mean little if separated from Schneider's theoretical views on the ‘disorders of the self'.

The ‘continuity' version

The ‘continuity' version states that for centuries ‘insanity' and ‘madness' referred to a melee of mental diseases which physicians were unable to separate. In the 1850s, Morel coined the term démence précoce to refer to states of cognitive deficit in adolescence. During the second half of the nineteenth century, ‘catatonia' was described by Kahlbaum and ‘hebephrenia' by Hecker. At the end of the century, Kraepelin realized that both disorders, together with ‘dementia paranoides' (which he had himself discovered), were manifestations of the same disease process. Kraepelin called this disease dementia praecox and based it on empirical data kept in follow-up cards. In 1911, Bleuler renamed it ‘schizophrenia' and, during the 1930s, Schneider listed diagnostic criteria which, owing to their ‘empirical' and ‘atheoretical' character, deserved to be enshrined in DSM-IV.

It is also part of the continuity story that European psychiatry was influenced by Kraepelin, while American psychiatry followed Adolph Meyer, Bleuler, and the psychoanalysts (4,5 and 6) (although Manfred Bleuler, in an interesting paper approved by his father, wrote: (7) ‘Since coming to the United States I have had the valuable experience of realizing that the conceptions of schizophrenia are very different here from those held in our clinic at Burghölzli'). This would explain the confused definitions offered in DSM-I and DSM-II (7) and the diagnostic disparities found between the United Kingdom and the United States (of course, this view does not explain the major differences in the conception of schizophrenia between Germany, Italy, France, Russia, Norway, the United Kingdom, etc.). In the event, Kraepelin and Schneider were discovered in the United States, psychoanalysis was eased out, and this paved the way for the advent of biological psychiatry. After some uncertainties (e.g. DSM-III), DSM-IV now offers the de facto official definition of schizophrenia.

The problem with this version of events is that it occurs in a historical vacuum. Neither the alternative definitions of schizophrenia nor the factors that explain the successful views are ever mentioned; indeed, the impression is given that there has been an ineluctable progress towards ‘the truth'. However, this flattering narrative is hollow for, given that the current definition of schizophrenia is still made on phenomenological grounds, it is necessary to ask how to decide which of the historical definitions were right and which were wrong.

The ‘discontinuity' version

Powerful forces shaped the construction of dementia praecox and schizophrenia during the nineteenth century. For example, ‘association psychology' provided the basis for the metaphor of ‘splitting' and ‘faculty psychology' offered a template (the mind as a bundle of intellectual, emotional, and volitional functions) in terms of which the new mental disorders were defined. ‘Neo-Kantianism', in turn, supplied a model of thinking which was to become crucial to ‘formal thought disorder'—a central ‘symptom' of schizophrenia. Lastly, ‘evolutionary theory' provided an explanatory framework. For example, according to Kraepelin, (8) the disease process underlying dementia praecox activated a set of ‘pre-formed reactions' (responsible for the clinical picture), all of which were biological and evolutionary in origin. Likewise, Bleuler (9) suggested that the Schnauzkrampf of schizophrenia was related to the protrusion of lips seen in chimpanzees (to ‘express dissatisfaction').

This article is based on the view that ideas about mental symptoms and diseases originate in ‘convergences'. By this term is meant the coming together of a term (newly coined or recycled), a behaviour (putatively related to a brain disturbance or to an allegorical human action), and a concept (as a carrier of definitions and explanations and rules). (10,11) These components will be considered in relation to schizophrenia, starting with terms and concepts and then considering behaviour.

History of terms and concepts

Dementia praecox

The term ‘dementia' had participated in at least three ‘convergences' before it was incorporated into dementia praecox. Up to the seventeenth century, dementia referred to states of psychosocial incompetence, whether inborn or acquired, and had a ‘legal' connotation—age and irreversibility were not part of its meaning. By the eighteenth century, dementia became linked to states of acquired intellectual deficit at whatever age and of whatever aetiology, i.e. there was a shift towards clinical usage. By the end of the nineteenth century, dementia was redefined in terms of a loss of cognition (mainly ‘memory', since called the ‘cognitive paradigm'). (12) Most importantly, age of onset, reversibility, and evolution became important, so that cognitive deficits in children or acquired states in younger adults (e.g. after head trauma) were no longer called ‘dementia'. (13) By the turn of the century, senile and other forms of dementia had been described. (14,15)

Morel (16) coined the term démence précoce to refer to the mental state and behaviour of young patients with stupidité (stupor) (‘surdi-mutité, faiblesse congénitale des facultés, démence précoce', and ‘Une espèce de torpeur voisine de l'hébétude remplaça l'activité première, et lorsque je le revis, je jugeai que la transition fatale à l'état de démence précoce était en voie de s'opérer'), i.e. with a disorder of motility and the will secondary to melancholia. (17) By ‘dementia' he meant any state of psychosocial incompetence related to a mental disorder and occurring at any age—the criterion of irreversibility did not yet exist. In this sense, the term démence précoce has little relationship to the work of Kraepelin or Bleuler. In his brilliant analysis of the development of the concept of schizophrenia, Minkowski (18) stated: ‘An abyss separates Morel's démence précoce from that of Kraepelin where the streamlet has become a river which, having forgotten its humble origins, threatened to flood everything'. It would seem, however, that the French themselves encouraged the idea that the history of schizophrenia started with Morel. (19)

By the time Kraepelin used the term dementia praecox, the general concept of dementia had acquired a third meaning different from that of the time of Morel. For example, Gross (20) noted that: ‘The meaning of the term dementia has changed in current usage. We have grown accustomed to employing it to denote not only an end state but also a developing state, a process...'. Furthermore, Morel's old term had sunk into oblivion; indeed, there is no evidence that in 1896 Kraepelin knew of its existence. Kraepelin first used the term in the fifth edition of his textbook (21) in 1896. Under Verblödungsprocesse, Kraepelin lists three independent conditions:

‘Dementia praecox' (mild and severe forms, and hebephrenia), ‘catatonia', and ‘dementia paranoides'. Nowhere in this text is Morel's name mentioned; it only appeared three editions later. By the time that Kraepelin was writing the fifth edition, the term ‘dementia' had already changed its meaning; hence he saw the need to qualify ‘dementia' by the term ‘praecox', by which he meant ‘early', ‘not at the expected age', etc. He only ‘acknowledged' the Frenchman for the first time in the last edition of his book, where he wrote: ‘The name dementia praecox which had already been used by Morel'. (22) There is, however, the distinct possibility that Kraepelin borrowed it from Arnold Pick who used ‘dementia praecox' as early as 1891. (23) For an excellent comparison of the editions of Kraepelin's textbook with regard to the construction of ‘dementia praecox', see Hoff. (8)


Eugen Bleuler

The term ‘schizophrenia' first appeared in print in 1908. (24) In 1911, Bleuler  (9) wrote: ‘Ich nenne die Dementia praecox Schizophrenie, weil, wie ich zu zeigen hoffe, die Spaltung der verschiedensten psychischen Funktionen eine ihren wichtigsten Eigenschaften ist' (‘I call dementia praecox schizophrenia because, as I hope to show, the splitting of the different psychic functions is one of its most important features'). And then: ‘In jedem Falle besteht eine mehr oder weniger deutliche Spaltung der psychischen Funktionen: ist die Krankheit ausgesprochen, so verliert die Persönlichkeit ihre Einheit... (‘In each case there is a more or less clear splitting of the psychological functions: as the disease becomes distinct, the personality loses its unity'). This seems straightforward, but is not. The meanings he gave to Spaltung and to psychischen Funktionen are ambiguous and need further historical clarification. By ‘splitting', Bleuler meant (a) a deep and general ‘primary loosening of associational networks' (‘primäre Lockerung des Assoziationsgefüges'), leading to irregular breaking (‘unregelmäbigen Zerspaltung') of ‘concrete concepts', and (b) a more apparent ‘systemic splitting of idea-complexes' (‘systematischen Spaltung in bestimmte Ideenkomplexe'). (9)

Behind these views there is a new model of the mind and hence, despite Bleuler's claims to the contrary, (9) there is a marked difference between Kraepelin's dementia praecox and Bleuler's schizophrenia. Thus the change was not just one of name. Indeed, a disciple of Bleuler wrote: ‘the concepts of Bleuler and Swiss psychiatry are looser than those of Kraepelin and German psychiatry'. (25) Gruhle (26) also felt that there was ‘no full correspondence' between the views of Kraepelin and Bleuler. This is confirmed by the fact that in France dementia praecox and schizophrenia were treated as different diseases up to the 1920s! (27)


Originating in early nineteenth century Romantic psychology and the work of Herbart, (28) the mechanism of separating, dividing, breaking, dissociating, divorcing, or splitting of mental functions was a common explanation (in popular literature and in psychology) for any unpredictable or strange human behaviour. For example, it is present in Wigan's two-brains model, Stevenson's Dr Jekyll and Mr Hyde, Hartmann's model of the unconscious, Jackson's hierarchical model of the brain, Azam's dissociation, Charcot's hysteria, Freud's splitting of the ego, and Wernicke's ‘sejunction'.

The mechanism of ‘splitting' (Spaltung) was popular in German psychiatry at the time when Bleuler coined the term ‘schizophrenia'. Indeed, there were a number of rival words more or less based on the same idea: intrapsychic ataxia, dementia dessecans, dementia sejunctiva, dysphrenia, discordance, etc. However, none was adopted. (29) Neither the new term (schizophrenia) nor its associated concept (splitting) were accepted by everyone. Freud expressed some reservation about both, for splitting ‘does not belong exclusively to that disease', (30) and Jaspers observed that splitting could not be observed in some schizophrenic patients. (31)

Kurt Schneider

Schneider is listed as the third major alienist in the history of schizophrenia, for example by Hoenig. (32) Study of Schneider's writings shows that there is a discontinuity between his views on schizophrenia and those proposed by Bleuler. For Schneider, the ‘first-rank symptoms' were not pathognomonic but suggested a diagnosis of schizophrenia only if there was no evidence of any other organic psychoses. The 11 first-rank symptoms only gain meaning when sought in the context of three diagnostic perspectives: course, symptomatology, and interaction. (33) Because, as Jaspers (34) proposed, the endogenous psychoses result from a process, there is little point in studying the ‘course' of schizophrenia. (35) All that was needed was to find out whether there had been some prodromal symptoms of schizophrenia. ‘Symptomatic comprehension' included the search for symptoms resulting from a defect in the integration of the self (hence it is not true to claim that Schneider believed that his first-rank symptoms were ‘empirical' and ‘atheoretical'). ‘Comprehension by interaction' referred to the way in which the patient is perceived by the clinician. In this regard, and without naming it, Schneider described the ‘praecox feeling' years before Rümke. (25)

Because Schneider had a cross-sectional view of diagnosis, the notion of ‘course' (in Kraepelin's sense) was alien to his thinking. Likewise, his conception of schizophrenia included all the paraphrenias, paranoias, marginal psychoses of Kleist, etc. He also believed that, in addition to schizophrenia and cyclothymia (manic–depressive insanity), the ‘endogenous psychosis' encompassed a large number of yet undiscovered diseases. (36)

In summary, there is no continuity between Schneider's notion of schizophrenia and earlier views; hence it is nonsense to choose some criteria from Kraepelin (i.e. course and duration), others from Bleuler (formal thought disorder), and yet others from Schneider (first-rank symptoms). It is nonsense because each of these alienists had a different (and non-additive) definition of schizophrenia, and hence the clinical features that each described only make sense in terms of their own conception and not in a decontextualized form. The DSM-IV definition happens to be a composite of the type that we have just described.

History of behaviours

During the 1980s, and encouraged by the assumption that schizophrenia was a recognizable, real, unitary, and stable ‘brain disease', questions were asked about the fact that clear descriptions of the disease seem hard to find before 1800.

Did schizophrenia exist before the eighteenth century?

This question remains ‘unresolved'. (37) Based on an ‘epidemiological' claim, for which there is no historical evidence, supporters of the so-called ‘recency hypothesis' suggested that ‘some change of a biological kind occurred about 1800 such that a particular type of schizophrenia thereafter became much commoner'. (38) (Gruhle (26) and Cranefield (39) have reviewed ‘possible' earlier cases of schizophrenia.) In the wake of this claim, efforts were made to rediagnose earlier ‘cases of schizophrenia' as something else. (40) Interestingly, the ‘recency hypothesis' seemed also to be supported by those who believed that the emergence of ‘schizophrenia' coincides with the birth of some ‘modern episteme' (which Foucault identified with the Kantian revolution). (41) Foucault defined ‘episteme' as the set of discursive practices that make possible the emergence of scientific disciplines. Constituted towards the end of the eighteenth century, ‘modern episteme' included a view of man as an ‘autonomous being' and is responsible for the development of the so-called human sciences.

However, others believed that schizophrenia 'had existed as long as mankind', (42) and cases were rediagnosed in the relevant direction. For example, Macalpine and Hunter (43) suggested that the seventeenth-century painter Christoph Haitzmann (diagnosed by Freud as a ‘demoniacal neurosis' in 1923) (44) was suffering from ‘schizophrenia'. Likewise, after re-reporting the case of James Tilly Matthews as ‘the earliest clear description of schizophrenia', Carpenter (45) explained that the absence of earlier cases resulted from a ‘different selection and description of cases for publication'. Jeste et al. (46) ferreted out, although not as successfully as Gruhle, (26) ‘a substantial number of clinical descriptions resembling modern conceptions of schizophrenia'.

A pseudo-problem

Looking back, it seems clear that, in terms of the conceptual parameters accepted by the participants, the debate could not have been resolved. Firstly, there were the issues of what counted as evidence, how many cases would falsify ‘the recency hypothesis', (47) or what level of diagnostic clarity was required for a case to qualify as evidence? Secondly, both sides were making unprovable claims; for example, ‘schizophrenia has always existed' and ‘there was a biological change that brought schizophrenia into life during the nineteenth century'.

The issues here are at the same time simpler and harder than anything that the participants in the debate seem to have envisaged. Some issues concern the clinical focus, i.e. what controls the perception and description of mental symptoms and diseases, others the ontology itself of the RRUS, i.e. what claims are being made about the existence of the entities in question, and yet others the rules of the epistemiological game, i.e. what counts as evidence etc. In the actual debate these issues appeared in various combinations and permutations. In this article, there is only space for one example.

Let us say that, because of their biological basis, the units of analysis (mental symptoms) have always existed. The current definition of schizophrenia is, at best, the result of the belief that some of these mental symptoms occur more frequently together. The way in which research is carried out today makes it very difficult to decide how often such mental symptoms affect individuals who are not considered as suffering from ‘schizophrenia'. Let us now take both extremes—that no case or that thousands of cases of schizophrenia can be found before the nineteenth century. Would these lead to the conclusion that schizophrenia did not, or did, exist before the nineteenth century? In fact, neither inference follows; absence of reports can be explained away by using arguments taken from ‘clinical focus' or ‘rules of the epistemological game' issues (as mentioned above). However, multiple reports create a problem: Why was the disease not recognized and named before if it has such a frequent and stereotyped presentation?

Therefore it can be concluded that it is not possible to write a sensible history of ‘behaviours redolent of schizophrenia' for the period before the nineteenth century. (48) This is because both the current concept of mental symptoms and disease and that of schizophrenia are nineteenth-century constructions. Hence clinical reports before this period will always lack epistemological ‘distinctness' (from our perspective). However, such reports make a great deal of sense if assessed in terms of categories such as madness, insanity, lunacy, vesania, melancholia, and mania.


The history of schizophrenia can be best described as the history of a set of research programmes running in parallel rather than seriatim and each based on a different concept of disease, mental symptom, and mind. However, only few of these programmes have been discussed in this short chapter.

Historical research shows that there is little conceptual continuity between Morel, Kraepelin, Bleuler, and Schneider. Two consequences follow from this finding. One is that the idea of a linear progression culminating in the present is a myth. The other is that the current view of schizophrenia is not the result of one definition and one object of inquiry successively studied by various psychiatric teams, but is a patchwork made out of clinical features plucked from different definitions. More research is needed to find out what led to this sorry state of affairs.

The continuity story should be rejected because its main role has been not to illuminate the past but to justify the present. Hopefully, the discontinuity history will offer uncommitted researchers alternative ideas, for example that there is no such a thing as a unitary disease called schizophrenia but only a collection of mental symptoms, some congenital, some relics from evolution, and others acquired.


1. American Psychiatric Association (1994). Diagnostic and statistical manual of mental disorders (4th edn), p. 187. American Psychiatric Association, Washington, DC.

2. Berrios, G.E. (1995). Conceptual problems in diagnosing schizophrenic disorders. In Advances in the neurobiology of schizophrenia (ed. J.A. den Boer, H.G.M. Westenberg, and H.M. van Praag), pp. 7–25. Wiley, Chichester.

3. Andreasen, N.C. and Akiskal, H.S. (1983). The specificity of Bleulerian and Schneiderian symptoms: a critical reevaluation. Psychiatric Clinics of North America, 6, 41–54.

4. Lewis, N.D.C. (1936). Research in dementia praecox, p. 29. Northern Masonic Jurisdiction, New York.

5. Raskin, D.E. (1975). Bleuler and schizophrenia. British Journal of Psychiatry, 127, 231–4.

6. Fullinwider, S.P. (1982). Technicians of the finite. The rise and decline of the schizophrenic in American thought 1840–1960. Greenwood Press, London.

7. Bleuler, M. (1931). Schizophrenia. Archives of Neurology and Psychiatry, 26, 610–27.

8. Hoff, P. (1994). Emil Kraepelin und die Psychiatrie als klinische Wissenschaft. Springer-Verlag, Berlin.

9. Bleuler, E. (1911). Dementia praecox oder Gruppe der Schizophrenien. Franz Deuticke, Leipzig.

10. Berrios, G.E. (1996). The history of mental symptoms. Descriptive psychopathology since the 19th century. Cambridge University Press.

11. Berrios, G.E. and Porter, R. (ed.) (1995). The history of clinical psychiatry. Athlone Press, London.

12. Berrios, G.E. (1989). Non-cognitive symptoms and the diagnosis of dementia. Historical and clinical aspects. British Journal of Psychiatry, 154, 11–16.

13. Berrios, G.E. (1987). Dementia during the 17th and 18th century: a conceptual history. Psychological Medicine, 17, 829–37.

14. Berrios, G.E. (1990). Alzheimer's disease: a conceptual history. International Journal of Geriatric Psychiatry, 5, 355–65.

15. Berrios, G.E. and Freeman, H. (1991). Alzheimer and the dementias. Royal Society of Medicine, London.

16. Morel, B.A. (1860). Traité des maladies mentales, pp. 516, 566. Masson, Paris.

17. Berrios, G.E. (1981). Stupor: a conceptual history. Psychological Medicine, 11, 677–88.

18. Minkowski, E. (1925). La génése de la notion de schizophrénie et ses caractères essentiels. Evolution Psychiatrique, 1, 193–236.

19. Lanteri-Laura, G. and Gross, M. (1992). Essai sur la discordance dans la psychiatrie contemporaine, p. 13. Epel, Paris.

20. Gross, O. (1904). Dementia sejunctiva. Neurologische Zentralblatt, 23, 1144–6.

21. Kraepelin, E. (1896) Psychiatrie (5th edn), pp. 425–71. Barth, Leipzig.

22. Kraepelin, E. (1919). Dementia praecox and paraphrenia (transl. R.M. Barclay and G.M. Robertson), p. 4. Livingstone, Edinburgh. (Extracted from 8th and last edition, published in 1910.)

23. Pick, A. (1891). Über primäre chronische Demenz (so. Dementia praecox) intramuscular jugendlichen Alter. Prager Medicinische Wochenschrift, 16, 312–15.

24. Bleuler, E. (1908). Die Prognose der Dementia praecox—Schizophreniegruppe. Allgemeine Zeitschrift für Psychiatrie, 65, 436–64.

25. Rümke, H.C. (1990). The nuclear symptom of schizophrenia and the praecox feeling. History of Psychiatry, 1, 331–41 (first published in Dutch in 1941).

26. Gruhle, H.W. (1932). Geschichtliches. In Bumke's Handbuch der Geisteskrankheiten. Part 5, Die Schizophrenie (ed. K. Wilmanns), pp. 1–30. Springer, Berlin.

27. Rodiet, A. and Heuyer, G. (1931). La folie au XXe siècle. Ètude médico-sociale, pp. 90–121. Masson, Paris.

28. Scharfetter, C. (1975). The historical development of the concept of schizophrenia. In Studies of schizophrenia (ed. M.H. Lader), pp. 5–8. Headley, Ashford.

29. Berrios, G.E. (1987). The fundamental symptoms of dementia praecox and Bleuler. In The origins of modern psychiatry (ed. C. Thompson), pp. 200–9. Wiley, New York.

30. Freud, S. (1953). Psychoanalytical notes upon an autobiographical account of a case of paranoia (dementia paranoides). In Collected papers (ed. E. Jones), Vol. III, pp. 387–470. Hogarth Press, London.

31. Jaspers, K. (1963). General psychopathology (7th edn) (transl. J. Hoenig and M.W. Hamilton). Manchester University Press.

32. Hoenig, J. (1983). The concept of schizophrenia: Kraepelin–Bleuler–Schneider. British Journal of Psychiatry, 142, 547–56.

33. Schneider, K. (1944). Conferencias psiquiatricas para médicos, Orfila, Madrid. (Translation of second German edition published in 1936.)

34. Jaspers, K. (1910). Eifersuchtswahn. Ein Beitrag zur Frage: ‘Entwicklung einer Personlichkeit' oder ‘Prozeb'. Zeitschrift für die gesamte Neurologie und Psychiatrie, 1, 567–637.

35. Häffner, H. (1993). Prozeb und Entwicklung als Grundbegriffe der Psychopathologie. Fortschritte der Neurologie Psychiatrie, 31, 393–437.

36. Huber, G. (1987). Kurt Schneider. The man and his scientific work. Zentralblatt für Neurologie und Psychiatrie, 246, 177–91.

37. Ellard, J. (1987). Did schizophrenia exist before the eighteenth century. Australian and New Zealand Journal of Psychiatry, 21, 306–14.

38. Hare, E. (1988). Schizophrenia as a recent disease. British Journal of Psychiatry, 153, 521–31.

39. Cranefield, P.F. (1961). A seventeenth case of mental deficiency and schizophrenia: Thomas Willis on ‘stupidity or foolishness'. Bulletin of the History of Medicine, 35, 291–316.

40. Hare, E. (1988). Schizophrenia before 1800? The case of the Revd. George Trosse. Psychological Medicine, 18, 279–85.

41. Foucault, M. (1970). The order of things. Tavistock Publications, London.

42. Strömgren, E. (1982). Development of the concept of schizophrenia. In Psychoses of uncertain aetiology (ed. J. Wing and I. Wind), pp. 17–26. Cambridge University Press.

43. Macalpine, I. and Hunter, R. (1956). Schizophrenia 1677. Dawson, London.

44. Freud, S. (1925). A neurosis of demoniacal possession in the seventeenth century. In Collected papers (ed. E. Jones), Vol. IV, pp. 436–72. Hogarth Press, London.

45. Carpenter, P.K. (1989). Descriptions of schizophrenia in the psychiatry of Georgian Britain: John Haslam and James Tilly Matthews. Comprehensive Psychiatry, 30, 332–8.

46. Jeste, D.V., del Carmen, R., Lohr, J.B., and Whyatt, R.J. (1985). Did schizophrenia exist before the eighteenth century? Comprehensive Psychiatry, 26, 493–503.

47. Turner, T.H. (1992). Schizophrenia as a permanent problem. History of Psychiatry, 3, 413–29.

48. Allen, D.F. (1995). Vers une perspective axiologique de la schizophrénie (éléments pour une histoire critique). Thèse de la Université de Paris 7.