Periodical amnesia and dédoublement in case-reasoning: Writing psychological cases in late 19th-century France

Abstract

The psychoanalytical case history was in many ways the pivot point of John Forrester’s reflections on case-based reasoning. Yet the Freudian case is not without its own textual forebears. This article closely analyses texts from two earlier case-writing traditions in order to elucidate some of the negotiations by which the case history as a textual form came to articulate the mode of reasoning that we now call ‘thinking in cases’. It reads Eugène Azam’s 1876 observation of Félida X and her ‘double personality’—the case that brought both Azam and Félida to prominence in late 19th-century French science—against a medico-surgical case penned by the Bordeaux physician in the same decade. While the stylistics of Azam’s medical case mirror its epistemic underpinnings in the ‘vertical’ logics of positivist science, the multiple narratives interwoven in Félida’s case grant both Azam and his patient the role of knowledge-making actors in the text. This narrative transformation chimes with the way Azam reasons ‘horizontally’ from particulars to Félida’s singular condition, but sits in tension with his choice to structure the observation along a ‘vertical’ axis. Between the two, we glimpse the emergence of the psychological observation as a mode of writing and thus of thinking in cases.

When Eugène Azam penned his account of Félida X—her extraordinary neurosis, double personality and periodical amnesia—he wrote as a ‘physician, [who] relate[s] as best [he] can an observation which belongs more to psychology than to medicine’ (Azam, 1876a: 481). ¹ He wrote, in other words, at the intersection of two case-reasoning fields—medicine and psychology—each of which enter into John Forrester’s work on ‘thinking in cases’ (2017). As a professor at the Faculty of Medicine in Bordeaux, Azam (1822–99) was accustomed to documenting his interactions with patients in the form of cases, or ‘observations’, for such was the currency of medical knowledge in 19th-century France. As a physician invested in ‘the progress of the medical sciences’, ² Azam’s medical writings were also informed by principles of positive science, as exemplified in the medical sphere by Claude Bernard’s experimental physiology. But when it came to Félida’s case, Azam acknowledged his uneasiness: The observation did not quite fit under a medical framework, and Félida’s ‘periodical amnesia’ raised different challenges for scientific representation. In this article, I explore how Azam responded to these challenges by analysing the textual practices and styles of reasoning he mobilized as he documented Félida’s two ‘lives’ and related her case to wider themes in psychological science.

My choice to read Azam’s text closely for its narrative and epistemic dynamics is inspired by Forrester’s detailed studies of singular psychoanalytic cases, especially his analysis of Robert Stoller’s Sexual Excitement (Forrester, 2017: 65–88). When Forrester traces the convoluted development of Stoller’s case, as psychoanalytic encounter and textual object, he concurrently unravels larger epistemological questions around case-reasoning; similarly, this article—which in form is also inspired by Forrester’s study—does more than expand the relatively limited scholarship on Azam and Félida. ³ It notably uses Félida’s case to elucidate the broader issue of how thinking in cases can diverge from or overlap with writing in cases (or what are labelled as such). That is, I take seriously Azam’s sense of having to relate Félida’s case differently from his medical cases, and I examine how differences in writing style matter for what it means to ‘think in cases’. What textual and narrative practices support the kind of ‘thinking in cases’ studied by Forrester? How might cases be written to allow for other forms of scientific reasoning?

I therefore compare Azam’s writing practices in his medical cases, dealing with surgical complaints and therapeutic interventions, with his extended observation of Félida. As Azam’s narrative style shifts, so too does his mode of reasoning, and we see how the case as a textual form comes to articulate a particular model of scientific reasoning: the horizontal reasoning along a ‘chain of precedents’ described by Forrester (2017: 128–9), which contrasts with ‘vertical’ ways of organizing knowledge, like those in play in Bernardian experimental medicine. I take the distinction between ‘horizontal’ and ‘vertical’ organizations from Jean-Claude Passeron and Jacques Revel (2005: esp. 26), who associate vertical configurations with inductive, deductive, and classificatory practices in the ‘hard’ sciences. There, elements of a single category are largely interchangeable, capable of being stacked together with little regard for their particular features. On the other hand, analysis that proceeds through detailed description and comparison, as in clinical or historical reasoning, can be considered as organized horizontally.

Beyond illuminating the textual underpinnings of case-reasoning as a scientific activity, the Félida case bears especial significance for those reflections on ‘thinking in cases’ initiated by Forrester’s work. As Forrester enunciated in ‘Inventing Gender Identity’ (2017: 127), his investigations of cases fan out from a central interest in the psychoanalytic case, especially its Freudian instantiations. ⁴ Yet psychoanalysis is not without its own textual forebears, and Azam’s study of Félida constitutes a critical step in that lineage. Félida’s was one of the defining cases of late 19th-century psychology, influencing enquiry into hypnotism, suggestion, and the very language and form of French pathological psychology (Carroy, 1992, 2001). Freud, in turn, can be viewed as ‘a critical and creative heir’ of these fields (Carroy, 2005: 206). ⁵ To uncover the dynamics of Azam’s case-writing and case-reasoning is thus to shed light on how this ‘particular kind of writing’ (Forrester, 2017: 127) evolved.

What made Félida’s case remarkable was both the rapidity of initial responses and the persistence and volume of subsequent discussion. From its first major diffusion in the Revue scientifique in May 1876 to the international hypnotism congresses of the fin de siècle, the case was circulated, critiqued, and evoked as an exemplar for comportments ranging from ‘double personality’ in general, through the capacities of altered psychical states, to the ‘fainting’ that accompanied switching states (see, for instance, Janet, 1889: 46; Liégeois, 1889: 184–5, 342). ⁶ This interest was sustained by some 24 communications—overlapping observations, reflections, and updates on Félida’s condition—presented by Azam between 1876 and 1893 (see Jullian, 1901). By the end of the century, the status of ‘the famous observation’ was such as to accord Azam ‘special mention’ among the precursors of scientific hypnotism (Dumontpallier, 1889: 24), and Félida semi-humorous recognition as ‘founder’ of the Chair of Experimental and Comparative Psychology at the Collège de France (Carroy, 1991: 103). ⁷

Even before Azam began issuing updates to the case, his observation of Félida had a complicated publication history, with several overlapping versions of the ‘original’ observation published on the basis of Azam’s communications to learned societies. In this article, I use the most widely diffused version of the case, as it was read before the Académie des sciences morales et politiques in May 1876, and quickly reproduced in the popularizing periodical, La Revue scientifique. Although the Academic communication did not appear in print until September 1877, I work from this ‘official’ version in preference to that of the Revue scientifique, except where the two texts differ. ⁸ Before undertaking a close reading of Félida’s case, however, I first examine the structure and textual form of Azam’s medical case-writing, in his most significant medical work of the mid-1870s: his 1874 ‘Nouveau mode de réunion des plaies d’amputation et de quelques autres grandes plaies’ on the treatment of amputation wounds (published in 1875). My analysis centres on the textual fabric of Azam’s ‘cases’—their structure, their narrative strategies, the presence of the subject. I argue that such textual practices are linked together with the modes of reasoning mobilized to make sense of cases, and I describe shifts in these interlinked features between medical and psychological ways of writing/thinking in cases. It is precisely in valorizing fine textual analysis, for its potential to elucidate epistemological questions, that this essay proposes to contribute—both to broad ongoing investigation of ‘thinking in cases’, and more narrowly to related scholarship on psychological or psychoanalytic case-writing. Whereas Jacqueline Carroy (2005) has surveyed ways the psychologist–subject pair is written into French cases, and Anne Sealey (2011) has proposed a set of formal features characterizing Freudian psychoanalytic case histories, here I attend closely to narrative and textual detail in a limited number of cases. Concomitantly, this article provides the first careful study of the Félida case as text and its textual relation to Azam’s previous work, thus extending Carroy’s earlier investigations establishing the case’s impact on philosophical/psychological discourse and highlighting the multiple collaborators involved in its production (Carroy, 1991: 103–9; 1992, 2001).

Medical ‘observations’ and amputation wounds

Although I have been discussing Azam’s ‘cases’, it was more properly as an ‘observation’ that Azam described his communication about Félida (Azam, 1876a: 481), like his writings about surgical patients. This terminological choice reflects the epistemological status of what I shall continue to call ‘cases’ (following Forrester’s usage) in 19th-century medicine and psychology. The French cas appears only once in Azam’s account of Félida and refers neither to the account as a textual entity, nor to the set of Azam’s interactions with his subject; rather, it occurs in a context of classification, as Azam considers how best to characterize Félida’s state in terms of existing medico-psychological nomenclature (Azam, 1877a: 382). In this, Azam conforms to French medical usage of the time; ‘cases’ in medical writing were almost always ‘cases of’, followed by the name of some disease entity. ⁹ The particulars that physicians observed in their interactions, or communicated to their colleagues were observations, faits (facts), and sometimes histoires (natural histories/stories). ¹⁰ Accordingly, it is in these terms that Azam and his colleagues refer to Azam’s writings about his surgical and psychological patients; the exception is histoire, which appears only in Félida’s case, usually paired with verbs of telling or publishing (Azam, 1876a, 1877a, 1893: 37–8; Bouchut, 1877; Dufay, 1876; Janet, 1876). Since histoire has connotations of storytelling, as the word for both ‘history’ and ‘story’, this divergence is suggestive of a more narrative approach in the psychological case. ‘Observation’ and ‘fact’, in contrast, bear strong associations with a particular model of medical reasoning.

The two terms resonate with long-standing traditions in medical writing, as well as with the framework of mid-19th-century scientific medicine, each tending to establish a distinction between the facts as observed and any interpretation or theorizing related to them. Gianna Pomata has traced the early modern emergence of the ‘observation’ (observatio) as a distinctive medical genre, characterized by a clear demarcation of the ‘case narrative’ from its ‘learned commentary’, coupled with a generalized ‘suspicion of theory’ (Pomata, 2011: 56–7, 67). Similar themes recur in the scientific ambitions of mid-to-late 19th-century medicine, for which Claude Bernard’s experimental physiology was perhaps the most influential model. Although Bernard himself disparaged clinical medicine as ‘conjectural’ (Bernard, 1865: 374), at best only a starting point for experimental investigation (ibid.: e.g. 247, 256–8, 351), many French physicians echoed his positivist precepts to mark their activities—clinical, as well as more properly experimental—with the stamp of scientific progress. An essential point for Bernard was to distinguish between ‘the experimental fact and its interpretation’ (ibid.: 332): Facts, once observed, could never be destroyed (ibid.: 306, 310), whereas theories were much less stable or dependable, being always susceptible to correction or rejection by the facts (ibid.: e.g. 23, 63, 287). Observation as a process implied ‘the plain noting [la constatation pure et simple] of a fact’ (ibid.: 29), undertaken ‘without any preconceived idea’ (ibid.: 41).

The medical observations presented by Azam in 1874 to the Société de chirurgie de Paris partake of the characteristics of both ‘fact’ and ‘observation’. These 26 short ‘observations’ recount his treatment of amputation wounds (and some other large wounds) by means of a new protocol. ¹¹ In his communication, Azam takes care to preserve a distinction between the observations and their bearing on his new method. ‘Here are the facts’, he announces, ‘After a succinct narration I will outline and discuss the method’ (Azam, 1875: 298). Such ordering and separation exhibit forms of reasoning present in traditional medical observations and in Bernardian science; they signal that Azam’s observations of amputation wounds are uncontaminated by theoretical preconceptions, and reciprocally incite the reader to refrain from prejudging them. They furthermore set up a vertical relation between the cases and Azam’s method in general. In an earlier attempt to promote the method, Azam had met with criticism for failing to underpin his ideas with ‘sufficiently precise observations’ (ibid.: 297). The 1874 observations are intended to provide precisely that missing support, that is, to bolster Azam’s claims for his method with concrete evidence. They do so primarily through counting, as Azam tallies the number of cures, or lists the duration of stages in the procedure (ibid.: esp. 307–8). Only rarely do the particular features of an observation matter for justifying the method. ¹² Thus in Passeron and Revel’s terms, Azam’s amputation observations relate to his method along a vertical axis: The observations are textually separate, and essentially undifferentiated for the purposes of interpretation.

Importantly, Azam’s medical observations echo this logic in their formal features; the way he writes these cases reflects the way he reasons from them. The treatments of amputation wounds are narrated briefly, in what seems a bare minimum of words, a result that is achieved chiefly by systematically removing articles, pronouns, and extraneous verbs. As described by Harriet Nowell-Smith (1995) in her analysis of late 19th-century Canadian medical cases, such stylistic features have the effect of excising the patient from the text. They also act to elide the presence of a mediating observer, such that the observations provide the illusion of a transparent, ‘plain noting’ of the facts. In Azam’s observations, the surgeon thus tends to appear only when it is a matter of identifying who performed the operation. Sometimes his role is voiced actively, as in observation 11, concerning an amputation performed by Azam—‘I undertook the amputation of her thigh’ (Azam, 1875: 302)—though more often, the information is conveyed in the passive voice, in sentences of the form ‘amputated in town by M. Denucé in November 1873’ (ibid.: 303). Here, the patient (a man) is the object of the verb ‘to amputate’, but even in this passive form, his presence is only implicit, in the absence of both the personal pronoun and auxiliary verb (i.e. ‘he was’) from the statement. ¹³ Indeed, patients are rarely individuated beyond enumeration of their sex, their age, and the condition requiring amputation. Rather, as in Nowell-Smith’s examples, the observations focus on procedures—like the dressing used—and symptoms—‘it arises a deep abscess’ (ibid.: 302)—and their narratives are structured chronologically with reference to these impersonal protagonists—for example, days elapsed since the amputation. This impersonal approach is furthered, in Azam’s case, by the greater capacity of French verbs to take an impersonal form, as in ‘il survient’ (it arises), or the tortuous (even in French) ‘il est procédé à l’amputation’ (it is proceeded to the amputation; ibid.: 300). Nonetheless, some trace of a reasoning observer persists amid the depersonalized prose and narrative of therapeutic actions: often merely the remark that there is something to be learnt from a given fact (ibid.: 302–3, 306), but occasionally more extended speculation around causes (ibid.: 306) or avenues for further investigation (ibid.: 302). As we turn to Félida’s case, where the object of study is memory and ‘personality’, such interventions increase; with the change in narrative complexity, we find a reciprocal shift in mode of reasoning from the case.

Interwoven narratives: The case of Azam and Félida

At first glance, Azam retains the conventional vertical logic of the medical observation in his psychological writing, by establishing a typographical separation between his ‘exposé’ of Félida’s case and an extended set of (theoretical) ‘reflections’. The exposé, or ‘observation’ proper, ¹⁴ exhibits further characteristics of a Bernardian fact—once observed, not to be altered—in that it is identical, word for word, in the two versions of the case history, as published in the Revue scientifique and the Academie’s Séances et travaux. ¹⁵ This may seem unremarkable—after all, they are supposed to be reproductions of the same case—but for pronounced differences in the introductions to the two texts, and in the commentaries that follow them. For Azam took advantage of the long period that elapsed between the reading of his communication at the Académie des Sciences morales et politiques in May 1876 and its publication in the Séances et travaux in 1877 both to provide an update on Félida’s condition and to address initial critiques of the case (Azam, 1877a: sections 3–4). That he did not simultaneously revise the exposé itself implies that he conceived the observation through a Bernardian framework—the same kind of reasoning that led staunch positivist Rudolf Virchow to insist that any extracts taken from an autopsy report be reproduced ‘using the exact words’ of the original (Virchow, 1880[1876]: 144).

If Azam’s psychological communication is consistent with medical models when taken as a structural whole, its textual form and length transform the exposé from a depersonalized chronicle into a rich narrative fabric, in which Azam’s own narrative of scientific discovery is interwoven with details of Félida’s life. Whereas physical symptoms defined the amputation patients, Félida’s physical manifestations are a matter of small interest for Azam: He expounds them in a single paragraph and unproblematically designates hysteria as the disease entity underlying her various troubles (Azam, 1877a: 368). What counts in the observation is Félida’s ‘singular life’ (Forrester, 2017: 24)—or singularly ‘doubled’ life—as ‘a young woman whose existence is tormented by an alteration of memory which presents no analogy in science’ (Azam, 1877a: 363). In brief, from her teenage years, Félida suffered from what Azam described variously as ‘double personality’, ‘double life’, and ‘periodical amnesia’. With little warning, she would switch from her ‘normal state’ (état normal) into a second, altered state, her ‘condition seconde’. Serious, hard-working, and with morose tendencies in her normal state, Félida became cheerful, vivacious, and sociable in the condition seconde. To this modification in ‘personality’ was added an alteration of memory: When in her normal state, Félida had no memory of what occurred during her time in the condition seconde, whereas in the latter state she recalled events from both ‘lives’. When Azam first examined Félida in 1858, she spent around three to four hours every day in the condition seconde (ibid.: 365, 367), but by the time he recommenced his observation in the mid-1870s, it had expanded to fill the major part of her life (ibid.: 375–6).

The observation’s thematic focus on Félida’s singular-but-doubled existence is paralleled by narrative features that present her as an active character in the story. Far from stripping out subject pronouns, Azam makes very frequent use of his subject’s name and correspondingly curtails his recourse to passive or impersonal constructions. For example, in the first six paragraphs of the observation (which relate Félida’s antecedents), Félida’s name occurs five times as an active grammatical subject, ‘she’ appears thrice, and there are only four impersonal or passive clauses (Azam, 1877a: 363–4). An even more radical shift lies in the way the narration accords a significant place to Félida’s interior view, and also to her own account of her condition. On the one hand, Azam employs indirect style to portray Félida’s affective states, such as when ‘in [her] second life, her pregnancy didn’t worry her and she bore it quite cheerfully’ (ibid.: 370). ¹⁶ On the other hand, Félida’s sentiments and experiences are narrated in what are ostensibly her own words, sometimes in direct speech (ibid.: e.g. 369, 379–81), and also in an extended section in indirect style in the second half of the observation. ¹⁷ Here, the reader enters into Félida’s experience of dealing with gaps in her memory, notably what happened when she switched states during a funeral (ibid.: 377–8). Her individual narrative has epistemic value, in Azam’s view, even though including it means he risks losing his status as authoritative narrator with control over the coherence of his text: ‘Here I think I should report certain episodes in the existence of our patient [malade], related by her. They will grant an excellent and complete idea of her state’ (ibid.: 377). While Félida’s narrative provides details to complete those accessible to the psychologist-researcher, her contribution to the observation also extends to her particular naming system for the two states. The particularity resides in the fact that ‘she has always held that the state, whichever it is, in which she is at the moment of speaking to her is the normal state, which she calls her reason, by opposition to the other state that she calls her fit/attack [crise]’ (ibid.: 366; emphasis in original). ¹⁸ As a result, Azam is frequently concerned with ‘unravelling’ Félida’s terms to determine her ‘true’ state during a given interaction (ibid.: e.g. 375–6, 380). This interpretative work is visible in the text, but does not supersede Félida’s account: ‘I question her, and I learn that she is in her reason (she speaks correctly today)’ (ibid.: 380; emphasis in original). Indeed, the observation privileges including Félida’s view over providing a unitary, coherent narrative; the reader is exposed to a certain confusion, which perhaps mirrors Félida’s disorientation on being confronted with unpredictable gaps in her memory. ¹⁹

Placing Azam’s text in a lineage of case-writing, we find another kind of mirroring here, in that Azam’s narrative choices prefigure what are more often perceived as ‘formal departures’ (Sealey, 2011: 42) taken by later psychological/psychoanalytical case-writing. I refer specifically to the textual presence of what Forrester calls a ‘distinctive function’ of the case history as a genre: writing into the case ‘the unique psychoanalytic experience of both patient and analyst’ (Forrester, 2017: 65; also Sealey, 2011: 42), such that knowledge is made out of the interaction of both parties in the psychological observation. Félida’s contribution to knowledge-making appears most strikingly in her idiosyncratic present-centred terms for the two states: Confusing to the external observer, they also grant the reader insight into the logic of her interior perspective. Moreover, their inclusion has much the same effect as Freud’s making his cases ‘reflect the disjointed narratives offered by patients’ (Sealey, 2011: 43)—although Azam probably includes Félida’s terms, like her longer account, primarily for reasons of completeness. But a desire for completeness does not prevent Azam from interpreting Félida’s terms, that is, from doubling the narrative to present ‘a psychological tale different to that of the patient’ (Carroy, 2005: 219). If Azam’s parenthetical corrections are of limited extent, compared to the kind of intertwining, theoretically driven narratives Carroy remarks in Pierre Janet’s psychotherapeutic work, they are nonetheless indicative of the way Azam, before Janet or Freud, enters the observation as a character and narrator with his own views and comportments (ibid.; Sealey, 2011: 42).

Particularly in its early stages, Azam’s observation of Félida is as much a narrative of his scientific activity as of her condition and comportment: Interwoven with descriptions of Félida’s amnesias are narratives of Azam’s research actions, his thought processes, and various influences on his scientific development. There are the colleagues who supported Azam’s efforts and whose suggestions shaped his observation, inciting him to investigate both hypnotism and questions of memory (see, for example, Azam, 1877a: 367–8, 370, 374). ²⁰ Of greater interest, however, are the ways that Azam’s presence in the text as reasoning observer flattens the vertical logic written into his medico-surgical observations. While Félida’s narrative of her interior perspective might conceivably count as ‘noting’ the facts for the purposes of completeness—though its literary form would remain troublingly far from ‘plain’—Azam’s intrusions into the observation tend to intermingle ‘fact’ and interpretation. Least problematic are the occasions when Azam narrates not only the observational situation, but also the considerations and contingencies that inform his actions, such as his thoughts on encountering Félida in her then-rare normal state one day in July 1875: ‘Making the most of an occasion perhaps difficult to meet again, I study her with care’ (ibid.: 380). ²¹ Just like his mentions of the number of times Azam witnessed various phenomena (ibid.: 365, 368), this remark functions to model the traits of a good observation, on the one hand, and to legitimate his findings by demonstrating his reliability as an observer, on the other hand. ²² The ‘research narrative’ thickens, with Azam also intervening as narrator to elucidate the clues that let him interpret Félida’s confusing terminology—‘the memory I had of the past had thus already enlightened me’ (ibid.: 376)—or to propose alternative explanations for some phenomena—‘I could have taken for hallucinations of hearing and smell certain hyperaesthesic states’ (ibid.: 369). He even evaluates which of Félida’s two ‘lives’ he considers ‘superior’: the condition seconde (ibid.: 367). ²³ At these moments, Azam’s text collapses the separation between ‘observation’ and ‘interpretation’, and thereby flattens the vertical logic of Bernardian science that structures his medical cases and organizes Félida’s case into exposé and reflections. Correspondingly, as the observation weaves multiple perspectives into a complex narrative fabric, the psychological case comes to accommodate knowledge configured horizontally.

The winding paths of analogy

It is not that Azam deploys horizontal forms of reasoning when he reflects explicitly on the epistemic value of Félida’s case. On the contrary, he construes the observation as contributing to psychological knowledge through a vertical process of generalization, enacted through the intermediary of the hypothesis. Hypotheses, contends Azam in his reflections, are ‘the more or less winding paths which lead to the truth’ (Azam, 1877a: 408), and he follows Crookes, whom he cites, in viewing them as crucial to advancing scientific knowledge. ²⁴ To privilege hypotheses in knowledge-making is also to echo Bernard’s conception that they are ‘indispensable’ in ‘carry[ing] science forward’. They do so, moreover, though a vertical movement, as Bernard (1865: 285) enunciates: ‘Their utility is precisely…to draw us out of the fact’. In a later update to the case, Azam accordingly justifies, or rather ‘excuses’, ‘the care, the meticulousness [minutie]’ of his observations in terms of the ‘importance’ of ‘the questions that this study raises, from the point of view of cerebral physiology and psychology’ (Azam, 1877b: 577). ²⁵ His apologetic tone implies that the minutia of his observation have little epistemic value in themselves, beyond the requirement to record the fact ‘sincerely and clearly’ (Azam, 1876a: 481, 488). Instead, Azam frames the importance of his (very lengthy) ‘fact’ in terms of its exceptionality: Félida’s condition ‘presents no analogy in science’ (ibid.: 481; Azam, 1877a: 363). This declaration not only opens both versions of the case history, but, significantly, is the only part of the introduction to be reproduced word for word. Azam further designates his observation as ‘the narrative of an anomaly’, when he remarks on having provided ‘a certain number of hypotheses’ (the reflections) and affirms the value of hypotheses to scientific progress (Azam, 1877a: 407; emphasis in original). Indeed, despite recognizing the problematic status of the ‘isolated’ fact in positive science (in a later update, Azam, 1878: 194), ²⁶ Azam persistently positions Félida’s case as exceptional. Thus, even as he explores links between Félida’s condition seconde and other psychological phenomena, like those of somnambulism, he concludes that ‘the condition seconde…is not of the same nature as the analogous states already observed, or rather already published’ (Azam, 1877a: 385).

There is a dissonance in Azam’s statements, as he continues to stress Félida’s exceptionality, or at least difference (‘not of the same nature’), in the midst of drawing connections with other phenomena through analogy. His discomfort is, I propose, expressive of a tension between the way he frames the process of knowledge-making, and the mechanism by which his text relates Félida’s case to existing science. For, as Azam’s remark signals, it is to analogy that he turns when he wishes to situate his observation of Félida. ‘Shall we seek out analogies?’ (Azam, 1877a: 384), he asks in the reflections section, as he moves to consider the significance of Félida’s troubles of memory. But analogy makes connections along winding and contingent paths, along horizontal ‘chains of precedents’ (Forrester, 2017: 128), in a play of difference and similarity between particulars. It is precisely the mode of reasoning that characterizes ‘thinking in cases’ for Forrester and others: ‘Case-based disciplines reason analogically, creating complex networks of similarity and dissimilarity relations…with no guarantee of self-consistency’ (ibid.: 51; also Passeron and Revel, 2005: 26). As a mode of reasoning, it configures knowledge horizontally, and thus functions orthogonally to the hypotheses valorized by Azam in advancing knowledge.

Strikingly, Azam draws on analogy not only in formulating his reflections; it also enters his case-writing at a more fundamental level, through the narrative richness of the observation proper (the exposé). As the ‘research narrative’ of Azam’s enquiry unfolds, we see how analogy guides his investigation and interpretation of certain phenomena. For instance, he is prompted to re-examine the limits of Félida’s amnesia by considering how ‘forgetting’ plays out in other ‘famous facts of double life’, especially in the case known in French as ‘MacNish’s American woman’ (Azam, 1877a: 368). ²⁷ A more complex example of analogical reasoning seemingly inspires Azam to construe Félida’s condition seconde in terms of phenomena of hypnotism. He goes on to develop this analogy at length in his reflections and in later instalments of the case (Azam, 1876a, 1877a, 1878), yet introduces it first in the exposé, where it provides something between additional justification for testing hypnotism on Félida, and insight into how he came to understand Félida’s transition state (Azam, 1877a: 372). What matters is not that Félida can be hypnotized, but the similarity of her ‘spontaneous’ transition between states to various observations of spontaneous hypnotism. Tellingly, Azam refrains from any explicit mention of ‘resemblance’ or ‘analogy’ here, remarking only that Félida’s spontaneous transition ‘naturally made me think about hypnotism’, before listing a number of ‘examples’ (Azam, 1876a: 483; 1877a: 372). ²⁸ These examples take the form of conventional medical case histories in miniature: first, brief identification of the observer and what he observed, then a typographically demarcated interpretation. If individually the examples exhibit a vertical splitting of observation from interpretation, when read together, they constitute a ‘chain of precedents’, linking self-hypnotism provoked by sewing, through self-hypnotism at the discretion of the subject, to predictable but spontaneous ‘sleep’ (Azam, 1876a: 483). Azam leaves it to his reader to forge the last link in the chain, to wit, Félida’s spontaneous ‘sleep-like’ transition into the condition seconde. Indeed, he explicitly denies that he is reasoning from particularities: ‘I will draw no consequences from these facts’ (Azam, 1877a: 373). Yet despite Azam’s reluctance to admit to reasoning by analogy, his text nonetheless configures Félida’s case into a horizontal chain, and invites its readers to do likewise.

Conclusion

In one sense, it is not surprising that Azam should either evade or fail to articulate the epistemic implications of his psychological case-writing. As he writes horizontal connections between Félida’s condition and other psychological cases into his observation, in a narrative incorporating the perspectives of both psychologist-narrator and subject, he departs from the principles by which he, as a physician and surgeon with scientific ambitions, frames what it means to reason scientifically. In a medical context inspired by Bernardian positivism, researchers like Azam configured their clinical observations to function similarly to positive ‘facts’: Broader knowledge was to be drawn out vertically from sets of discrete observations, or alternatively, medical cases could be deployed to undergird some overarching idea. Azam accordingly establishes a textual separation between his surgical observations of 1874 and his general ideas on treating amputation wounds, and uses the former to support the latter. The textual style of his medical cases mirrors this mode of reasoning: The impersonal and symptom-focused language of the observation-as-text provides the illusion that it can be identified with the observation-as-plain-fact, quite separate from any interpretation.

But when Azam pursues psychological enquiry—into the periodical amnesia manifested by Félida—the textual features of his observation cannot sustain this illusion. Under the challenges of representing psychological phenomena, the depersonalized medical style gives way to a narrative richness. Alongside reports of Félida’s comportment, the text depicts Félida’s affective states through the immediacy of indirect discourse, presents her singular experience in her own terms, and weaves in narratives of Azam’s research activity. It is at points where the narrative becomes doubled that Azam’s observation most definitively collapses the distinction between ‘plainly noted’ fact and interpretation: when it doubles Félida’s relativist terms for her two states with Azam’s externalized perspective, or when Azam’s conjectures and conclusions add depth to his observational actions. Moreover, this shift in the way Azam writes his psychological case, a move that prefigures the literary features of the Freudian psychoanalytic case, also provides space in the observation for horizontal forms of thinking in cases. Azam, in the textual character of reasoning observer, is notably seen to use analogy to understand Félida’s case in relation to other psychological ‘facts’, and hence to deploy the characteristic form of case-based reasoning, according to Forrester.

What lends particular significance to Azam’s case history of Félida is that such horizontal reasoning appears within a text that is structured vertically (into exposé and reflections), that circulates unchanged like a Bernardian fact, and whose author remains committed to a vertical model of scientific knowledge-making (via hypotheses). There is a disjunction, in other words, between practices on a textual level and what happens on a structural or conceptual level. It resembles and extends the discrepancy (décalage) between methodological language and clinical practice that Passeron and Revel (2005: 27–9) remark in works by Durkheim or Freud. And it is precisely in this disjunction, in the shifts between Azam’s medical and psychological cases, that we see unfolding the textual negotiations by which a model of ‘thinking in cases’, in Forrester’s sense of the term, could emerge from a form of case-writing—medico-scientific cases—that articulates an orthogonal mode of reasoning. Ultimately, Félida’s case prompts us to attend to the textual foundations of case-reasoning as theorized by Forrester—to the way thinking in cases has historically implied certain forms of case-writing.