1. The Question

The standard account of the law of 30 June 1838 enumerates its achievements: an asylum in every department, a legal framework for admission, a statutory position for the insane. What that enumeration passes over is the law's most peculiar feature. It made possible the deprivation of a citizen's liberty, for an indefinite period, without any judge examining the case — on the strength of a prefect's signature or of a physician's certificate accepted by the director of an establishment.

The peculiarity is sharpened by what had preceded it. The Revolution had denounced the arbitrariness of the lettre de cachet and abolished it by the law of 16–26 March 1790, whose ninth article provided that persons detained on grounds of insanity should, within three months, be interrogated by judges and examined by physicians, so that on the strength of the resulting sentence they might be released or treated in designated hospitals.[1] The revolutionary legislator, in other words, had begun by referring the fate of the insane to a joint examination by justice and medicine. Forty-eight years later the same country established a permanent apparatus of detention that dispensed with the court altogether.

This article asks how that reversal became possible, and it does so by refusing two familiar shortcuts. The first is the professional narrative, in which the law appears as the triumph of a medical speciality and Esquirol as its author. The second is the functionalist narrative, in which the law appears as the necessary institutional expression of a bourgeois social order. Each is illuminating; neither identifies who, at a given moment and under what domestic pressure, actually set the machinery of confinement in motion.

The argument proceeds on three levels, which are not three causes but three registers of a single process. Section 2 treats the actor: how psychiatric authority was constituted before the law, through hygienist sociability and ministerial appointment. Section 3 treats the demand from above, and in particular the judicial barrier that made the ordinary legal route unusable. Section 4 treats the demand from below, in the household economy of a peasant society. Section 5 returns to the parliamentary debate, where the two revolutionary legacies collided and one was chosen. Section 6 sets the French case against the English.

A word on the relation of this article to Robert Castel's L'Ordre psychiatrique (1976), on which the historiography of the 1838 law — including the present author's own earlier work — has rested heavily. Castel's central claim is that the new social order founded on contract had to devise, for those who could not enter into contract, “a status different from and complementary to the contractual status governing the whole body of citizens,” and that this second status — tutelle, guardianship — was invented around the problem of madness by way of its medicalization.[2] The claim is powerful, and it is not the intention here to displace it. But it explains why a tutelary status was required without explaining who invoked it, when, or in what domestic circumstances; and it presents the tutelary form as the answer a contractual society needed, when the parliamentary record shows that form to have been contested to the last article and adopted against a fully articulated alternative. This article accordingly treats Castel's thesis as a hypothesis to be tested — from below, against the evidence of the households that demanded confinement, and from within, against the debates in which the tutelary solution was argued into existence.

2. The Actor: Esquirol's Network and the Report of 1818

2.1 The first specialist

Jean-Étienne-Dominique Esquirol (1772–1840) is conventionally introduced as the theoretical author of the 1838 law. The description is misleading in a specific way: it suggests that a doctrine produced a statute. What the record shows is that a professional network produced an administrative fact, which the statute subsequently ratified.

Castel's observation on this point is the most useful thing he says about Esquirol. Esquirol was “the first specialist” in the modern sense — the first around whom a career devoted entirely to mental illness became possible, and around whom there gathered the men who would become the great figures of the alienist movement: Falret, Parisot, Ferrus, Georget, Voisin, Leuret. What is striking is the timing: this specialization occurred precisely when the reorganization of the medical profession was otherwise tending towards unification.[3]

Equally important is the channel through which the group reached the administration. It was not as psychiatrists that they were connected to government but as hygienists. Esquirol sat on the conseil de salubrité of the Seine, founded in 1802, alongside Parent-Duchâtelet, Marc, Parisot and Villermé; he became its president in 1822. When the Annales d'hygiène publique et de médecine légale was founded in 1829 on the initiative of Marc and Esquirol, seven of its twelve editors were members of that council.[4] The 1818 report, to which we now turn, issued from this nexus.

2.2 The double function of the report of 1818

In September 1818 Esquirol presented to the Minister of the Interior a memoir on the establishments for the insane in France, based on a tour he had made at his own expense. Its most quoted passage deserves to be given in full, because its rhetorical construction — first-person witness, accumulated by anaphora — is precisely what gave it its political force:

…plus maltraités que des criminels, et réduits à une condition pire que celle des animaux. Je les ai vus nus, couverts de haillons, n'ayant que la paille pour se garantir de la froide humidité du pavé sur lequel ils sont étendus. Je les ai vus grossièrement nourris, privés d'air pour respirer, d'eau pour étancher leur soif…[5]

Elsewhere in the same memoir Esquirol describes the insane as thrown, like state criminals, “dans des culs de basse-fosse, dans des cachots où ne pénètre jamais l'œil de l'humanité,” left to consume themselves in their own filth under the weight of chains that tear their flesh.[6]

That this indignation was sincere there is no serious reason to doubt; that the document simultaneously functioned as an instrument legitimating the extension of medical authority is equally clear. To insist on the second point is not to impugn the first. The interest of the case lies exactly in their inseparability — a structure this article calls, following the earlier Japanese version of this argument, Esquirol's paradoxical modernity, and which will reappear in Section 5 in the arguments of Dufaure.

2.3 The law ratifies an accomplished fact

There is a received image of the years after 1838 in which young alienists were dispatched to the departments as missi dominici of a new order. Castel's correction is decisive and is confirmed by the parliamentary record. The movement had begun long before: by the late 1830s every relevant post in Paris was held by a representative of the same school, and Esquirol, who controlled the initiative in appointments, had already placed his pupils at Rouen, Nantes, Toulouse, Auxerre and Rennes.[7]

The parliamentary record supplies the institutional counterpart. Vivien's report of 18 March 1837 records that the commission consulted “several men of the art whom their reputation designated to it”: Esquirol, “whose name recalls thirty years devoted to the insane”; Ferrus, who joined to his learning the experience acquired in the general inspectorate; Falret; and Scipion Pinel, “who bears with honour a name dear to science.”[8] The commission also notes, in passing, that the number of public establishments specially devoted to the insane, which in 1818 had stood at eight, now stood at thirty-four, not counting Charenton, Bicêtre and the Salpêtrière.[9] The twenty years between the report of 1818 and the law of 1838 were not a period in which a statute created alienist power. They were a period in which alienist power created accomplished facts, and the statute recognized them.

3. From Above: Contract, Tutelle, and the Judicial Barrier

3.1 The barrier described by those who met it

The Civil Code provided one regular route by which the legal situation of a person of unsound mind could be settled. Article 489: “The adult who is in a habitual state of imbecility, dementia or fury must be interdicted, even where that state presents lucid intervals.”[10] Articles 490 to 512 set out the procedure: petition by a relative or by the public prosecutor, referral to the tribunal de première instance, convocation of a family council, examination of the person, judgment.

It has become customary — and the present author has followed the custom — to describe this procedure as a threefold barrier: cumbersome, costly, public. The description is correct, but it has usually been supported by reference to Castel. It need not be. Esquirol's Examen du projet de loi sur les aliénés, published in 1838 while the bill was before the Chamber of Peers, describes the barrier from the inside, and does so more sharply than any modern summary.[11]

On the incoherence of practice before the law:

Dans tel département l'interdiction est nécessaire avant d'obtenir l'admission ; dans tel autre il suffit de traiter avec les administrateurs ; ici le maire délivre l'autorisation ; là c'est le préfet.[12]

On the consequence of judicial scruple — and this is the single most important sentence for the present argument:

Les magistrats, fidèles à la lettre du Code, n'envoient les aliénés dans les hospices qu'après avoir provoqué leur interdiction ; et, en attendant, pour ne pas violer la liberté individuelle, ils les font enfermer dans les prisons.[13]

The judicial route did not merely fail to deliver confinement quickly. It delivered imprisonment. Magistrates unwilling to violate individual liberty by an administrative confinement placed the sick in gaol while the procedure ran its course. Whatever else may be said about the alienist campaign, its central empirical claim was not a fabrication.

On the barrier's two other faces, named by Esquirol in a single phrase — les longueurs et la publicité de l'interdiction[14] — and on the solution he proposed:

L'interdiction est remplacée par des voies légales, moins lentes et moins solennelles que les formes judiciaires.[15]

And on the fiscal vacuum that the finance law of 18 July 1836 had begun to fill: “Avant la loi des finances de 1836, les préfets ne savaient où prendre les fonds pour subvenir aux frais de la séquestration.”[16]

Finally, the conceptual distinction on which the whole reform turned, stated by Esquirol at the opening of the pamphlet with complete clarity:

Les lois anciennes ordonnèrent le renfermement des furieux, parce qu'ils troublent l'ordre public ; elles autorisèrent et prescrivirent l'interdiction des fous pour la conservation de leur fortune.[17]

Confinement answers to public order; interdiction answers to property. Once the two are analytically separated, the question “why must a man be interdicted in order to be confined?” answers itself — and the answer is the law of 1838.

3.2 Georget, 1825: the practice already existed

The separation Esquirol proposed in 1838 had been an accomplished practice for at least thirteen years. In 1825 Étienne-Jean Georget, reviewing the criminal trials in which mental alienation had been pleaded, appended to his account of the Jean-Pierre case a footnote rebutting the avocat général's suggestion that an acquitted defendant should be interdicted. The footnote is worth quoting at length, because Castel quotes only its first sentence and the remainder is more damaging:

Presque tous les aliénés sont enfermés sans être interdits, en vertu de la loi du 24 août 1790. Il est toujours possible de motiver la séquestration, sans même qu'il y ait des actes de violence ou de fureur ; car il n'est peut-être pas d'aliéné qui ne puisse devenir dangereux pour la sûreté publique. Un individu qui serait acquitté pour cause d'aliénation mentale, ne serait pas même mis en liberté ; l'autorité judiciaire n'aurait qu'à prévenir l'autorité municipale, qui ne manquerait pas d'en ordonner la détention dans une maison de force.[18]

Three things are established here at once, by a witness with no interest in exaggeration. First, confinement without interdiction was already the ordinary case in 1825. Second, the criterion of dangerousness was understood, by a physician sympathetic to the practice, to be indefinitely extensible — “there is perhaps no lunatic who could not become dangerous to public safety.” Third, a working circuit already existed by which the judicial authority passed the case to the municipal authority, which would not fail to order detention.

The law of 1838, on this evidence, did not invent administrative confinement. It legalized, standardized and financed a practice that had been running outside the law for a generation. Castel's formulation — that the achievement of the law was “the legalization of sequestration without interdiction” — is exactly right;[19] what Georget adds is that the sequestration without interdiction had not been waiting for legalization in order to occur.

3.3 A note on the figure of twenty-nine

At this point the literature customarily produces a statistic. In 1835, we are told, an average year, only twenty-nine judgments of interdiction were pronounced in the whole of France. The figure has passed into general use through Castel and has been repeated, including by the present author, without verification. It does not survive verification, and the manner in which it fails is instructive.

Castel's note refers the figure to a Paris law thesis of 1876, without a page.[20] The figure itself, with a fuller context, is found in a Paris law thesis of 1872 by Louis Lallement, in his narrative of the parliamentary struggle over the inspection clauses. Lallement is reporting the argument by which the peer Barthélemy defended the system of visits against those who, invoking Esquirol's authority, held that frequent inspection would harm the patients:

— Les aliénés, dit-il, ne sont interdits que dans une proportion très minime, et entrent dans des maisons de santé sans l'intervention des formes judiciaires. En 1833, d'après M. de Belleyme, plus de deux cents maisons de santé recevaient des aliénés, et cependant les jugements d'interdiction ne s'élevaient qu'à vingt-neuf.[21]

Three discrepancies follow. The year is 1833, not 1835. There is no scope: “the whole of France” does not appear. Neither does “an average year.” And the attribution matters: Louis-Marie de Belleyme had presided over the civil tribunal of the Seine since 1830, so that the judgments he was in a position to count were in the first instance those of his own jurisdiction.

The figure, in short, is not a measurement. It is an argument, advanced in debate by a speaker who had every reason to want the number to be small, on the authority of a magistrate whose competence was departmental. It should not be used as evidence of how many interdictions were pronounced in France. It can, however, be used as evidence of something else, and something more directly relevant: that the legislators believed interdiction to have fallen into desuetude, and legislated on that belief. The argument of this article does not depend on the number, and the number is retained here only as a specimen of what the Chambers took themselves to know.

3.4 Castel's thesis, stated

With the barrier established from contemporary sources, Castel's interpretation can be stated in its own right rather than used as a source. The new order established by the Revolution rests, on his account, on a juridico-administrative fiction: each citizen is at once subject and sovereign, submitted to obligations whose breach the state apparatus sanctions, and drawing his rights from the regulated activities whose exercise defines his liberty. The madman resists this reduction:

La véritable spécificité du fou est de résister à ce rabattement à tel point que, pour l'inscrire dans le nouvel ordre social, il va falloir lui imposer un statut différent et complémentaire du statut contractuel qui régit l'ensemble des citoyens.[22]

Madness thus exposes “a lacuna in the contractual order”; juridical formalism cannot regulate everything, and there exists at least one category of individuals who must be neutralized by means other than those available to the judicial-police apparatus.[23] What fills the lacuna is a new tutelle, and the law of 1838 is its institutionalization.

Two observations qualify this account without refuting it. The first is chronological and is Castel's own: the distribution of powers over the insane among justice, administration and family long predates the Revolution, and within it a class division was already sketched — judicial guarantees for the rich, repression by agents of the executive for the poor.[24] When the law of 1838 bypassed the courts it did not abolish that division; it opened the administrative route to everyone. The second observation is the subject of Section 5: that the tutelary form was not a solution the age reached for but one it argued over, and that the alternative was defended in the Chamber by a deputy who named the tutelary character of the proposal more precisely than any subsequent commentator.

4. From Below: The Household Economy and Its Defensive Limit

4.1 A society of smallholders

The Revolution's land settlement, through the sale of nationalized ecclesiastical and émigré property, rooted a broad class of peasant proprietors in French soil.[25] France in the first half of the nineteenth century remained a society in which roughly three quarters of the population lived in the countryside; the rural population reached its maximum, at about twenty-seven million, in the 1840s, and not until the census of 1931 would the urban population exceed the rural.[26] The massive appearance of wage labourers severed from the land, which enclosure and Poor Law reform produced in England, did not occur in France on a comparable scale or timetable.

By “household economy” is meant here a form in which production and consumption are integrated at the level of the household and the allocation of members' labour follows the logic of the household's survival rather than the market wage — the concept in the Chayanov–Brunner lineage, as formulated by Kriedte, Medick and Schlumbohm.[27] Within such a form, members lacking full labouring capacity — the old, the disabled, and the mildly disturbed — could retain a place through auxiliary tasks: fetching water, watching, minding children, tending animals. Olwen Hufton's “economy of makeshifts” included, as one of its components, the incorporation of marginal members.[28]

A caution is required here, and the earlier Japanese version of this argument stated it too weakly. Direct evidence that French rural households systematically deployed the mentally ill as auxiliary labour is thin. The claim this article defends is the more modest one of maintenance at home. For that, the evidence is good, if later than one would like. Alexandre Le Bras, working from the patient files of the Saint-Athanase asylum at Quimper, has shown that in later nineteenth-century Brittany the family remained the primary locus of care for the patient who was not dangerous, and that maintenance at home long preceded institutional admission.[29] To assume the same structure for households before 1838 is an inference backwards from survival; but it is an inference consistent with the evidence, and the direction of the error is conservative — if households were still absorbing this much after the institutional route was fully available, they were absorbing at least as much before it existed.

4.2 The limit, and why it was low

The Civil Code's principle of equal partition among heirs (article 745) fragmented holdings and thinned the economic reserve of the individual household.[30] Le Play's critique of this structure as the dissolution of the family is tendentious in its conclusions but accurate in its observation.[31] The absorptive capacity of the household economy was not generous. The incapacitation of a single member, still more his destructive behaviour, could place the household's reproduction in immediate jeopardy. Breadth of inclusion and lowness of the defensive limit were two faces of the same structure.

4.3 What the admission records record

Claude Quétel's lexical analysis of the admission registers of the Bon-Sauveur at Caen found that 46 per cent of recorded motives for admission were described in the vocabulary of violence, noise, agitation, fits, crises and attempted suicide, and he judged the true proportion higher.[32] This figure has been read, including in the present author's earlier work, as evidence of the logic of social defence. It bears a more specific reading. These words are not in the first instance the classificatory vocabulary of administrators. They are the verbalization of the threshold at which a family concluded that it could no longer keep the patient at home.

Le Bras's Quimper files allow that threshold to be observed in operation. What emerges is that families did not choose confinement lightly; that confinement functioned as a régulation en dernier recours; that families turned to the institution only once épuisées ou ruinées; and — the finding most consequential for the argument of this article — that even the placement d'office, which the law assigned to the prefect's own authority, was in practice set in motion by families presenting themselves before the mayor or the police commissioner.[33]

The limitation must be stated plainly. Le Bras's corpus is from the second half of the nineteenth century, after the 1838 regime was fully established; its application to the situation before the law is a structural analogy, not an extrapolation. But the analogy runs in a determinate direction, as noted above.

4.4 The consequence for Castel's thesis

Here the test of Castel's account can be stated. The demand for confinement did not issue from the rationality of administration. It issued from households that could no longer contain the sick. When the household's defensive limit was breached, inclusion inverted into a demand for exclusion, and the community sought to refer the case to a higher instance. What the tutelle was for, in Castel's sense, is one question; who activated it, and under what pressure, is another, and the second question has an answer that his framework does not supply. The prefect's power was not, in the daily reality reconstructible from the files, a power exercised upon families. It was very largely a power exercised by families through the prefect.

5. Two Revolutionary Legacies: The Debate of April 1837

5.1 The problem as the Chamber saw it

The Chamber of Deputies opened its debate on the bill on 3 April 1837.[34] The scale of the problem, as presented from the tribune, was contested even in its arithmetic. Calemard de Lafayette, opening against the bill, spoke of legislating on the fate of fifteen thousand persons “rejected by a painful exception outside the common law.”[35] The commission, reporting the returns analysed by Ferrus in 1834, arrived at eight or nine thousand — of whom some thousand were scattered in small hospices or in prisons and some two thousand five hundred were left in their families — while noting that Esquirol and Scipion Pinel were disposed to double that estimate.[36] Ferrus's survey had also found that only thirty-eight departments possessed special establishments, and that nineteen placed their insane in hospitals partially given over to their detention, where the patients were most often housed in what Calemard called “véritables cachots”: cells of a few feet of surface, on the ground floor, thick-walled and unheated.[37]

That the situation required legislation was, in the Chamber, common ground. What was in dispute was which authority should decide.

5.2 Isambert's case

The most articulate defence of judicial control came from the deputy François-André Isambert. It is important to see that he was not an opponent of the bill in general, and he said so in his opening sentence: he shared the opinion of the commission and of the government itself on essential points, and acknowledged that the legislation was insufficient as to the number and regime of establishments and as to the financial means necessary for the cure of the indigent sick.[38]

He conceded more than that. He conceded the central factual premise of the alienist campaign:

Les tribunaux… n'étant appelés à statuer que sur les interdictions et sur la capacité civile, n'osent avant l'interdiction statuer même provisoirement sur la liberté des personnes. D'un autre côté, l'urgence de la séquestration est le plus souvent telle qu'il est impossible d'attendre les résultats d'une procédure en interdiction.[39]

This is the leading opponent of administrative confinement stating that the courts dare not act before interdiction and that the urgency of sequestration usually makes it impossible to await the procedure. The dispute in the Chamber was therefore not about whether the judicial route was blocked. Both sides agreed that it was. The dispute was about what followed.

Isambert's answer was that the blockage called for a different judicial officer, not for the abandonment of the judiciary. The Civil Code assimilated the interdicted, in person and in goods, to minors; the president of the tribunal already exercised analogous powers over minors and wards; why displace that power?[40] What the bill proposed instead was, in his phrase, to erect the political magistrate into the magistrate of families:

l'esprit de centralisation et d'irresponsabilité fait chez nous de si rapides progrès, qu'on est allé jusqu'à ériger le magistrat politique en magistrat des familles, et à s'en remettre pour les garanties dues à la liberté individuelle, au pouvoir discrétionnaire du délégué direct du ministre de la police générale.[41]

And, on the substitution itself: “on substitue l'interdiction administrative à l'interdiction judiciaire”; and in this fever of innovation the Chamber does not fear to lay hands on two articles of the Civil Code, abrogating one and modifying the other.[42]

5.3 “An interdiction in miniature”

The commission, having decided that the formalities of interdiction were useless or dangerous, had immediately confronted the question of what should become of the patient's property during sequestration, and had been obliged to substitute for interdiction a provisional administration or guardianship whose limits it could not define. Isambert's verdict on this expedient is the single most valuable sentence in the debate:

C'est une interdiction au petit pied, moins les garanties voulues par le Code civil ![43]

An interdiction in miniature, less the guarantees required by the Civil Code.

What Castel would characterize, a century and a half later, as a new tutelle — a status distinct from and complementary to the contractual status of citizens — was named on the floor of the Chamber, at the moment of its creation, by the deputy who opposed it, and named more precisely: as a guardianship stripped of the guarantees that guardianship in the Code had carried. The tutelary reading of the 1838 law is not an interpretation imposed retrospectively by a sociologist of the 1970s. It is the reading its contemporary opponents gave it.

5.4 Two legacies of 1790

Isambert then made the argument that reveals the structure of the whole conflict. The principle he attacked, he said, was none other than that of the lettres de cachet — and he specified the equivalence institutionally, not rhetorically:

Qu'était-ce que les lettres de cachet, sinon des ordres du lieutenant de police de Paris, ou du secrétaire d'État de la maison du roi, aujourd'hui remplacés par le préfet de police et par le ministre de l'intérieur, pour éloigner de leur domicile… des individus qui étaient accusés secrètement de déshonorer leurs familles ou de troubler habituellement l'ordre public par leurs désordres, ou d'alarmer la société par des signes d'aliénation au moins temporaire, et qu'on voulait soustraire à l'action ordinaire des lois et de la justice ?[44]

He then observed that the government and the commission had themselves invoked the law of March 1790, which had abolished the lettres de cachet for ever — and that the strange thing was that they leaned on that law to claim for the administration a prerogative which the law conferred only on the judicial authority. He read article 9 in full, and concluded: “Far from rejecting that authority, it is that authority I invoke; it is its provision that, by an amendment to article 1, I wish to substitute for the government's project and the commission's.”[45]

Here the debate discloses its real structure. Both sides were arguing from the revolutionary legislation, and each had a text. The government and the commission stood on the law of 16–24 August 1790, which had committed to municipal bodies the care of obviating the mischief that might be occasioned by the insane left at liberty, and on the law of 19–22 July 1791 which had reinforced it. Isambert stood on the law of 16–26 March 1790, article 9, which had committed the examination of the detained to judges and physicians. The Revolution had bequeathed two legacies, a police legacy and a judicial one, and they were incompatible in this domain. The law of 1838 is the point at which the French state chose the police legacy.

This matters for the assessment of Castel's thesis. On his account the tutelary status is what a contractual society required; the account is functional, and functional accounts tend to make what happened look like what had to happen. The record of April 1837 shows a contested choice between two available juridical genealogies, decided by a majority, against an alternative that had been fully specified — Isambert's amendment would have had the insane received into public establishments “sur une simple ordonnance de justice, quand la sûreté ou l'ordre public l'exigeront.”[46] The amendment was withdrawn and the judicial system defeated.[47] That the tutelary solution was necessary is a proposition the Chamber did not accept; it adopted it as the better of two contested options.

5.5 Dufaure: “isoler, ne pas interdire”

The decisive counter-argument came on 7 April from Dufaure, and its form deserves attention. It was a medical argument deployed to exclude the judiciary:

…voyez l'inconvénient qu'ils trouvent à ce que, lorsqu'un homme est atteint d'aliénation mentale, on provoque son interdiction, on l'appelle devant les tribunaux, on le soumette à des interrogatoires, on l'environne de tout cet appareil judiciaire qui souvent augmente le danger des symptômes qui s'étaient manifestés en lui, et quelquefois rend son état incurable.

[Les médecins] ont reconnu, d'un autre côté, que, dans presque tous les cas, l'isolement produisait d'excellents effets : isoler, ne pas interdire, deux idées nouvelles que la loi doit encourager et qui permettent de modifier le Code civil.[48]

The judicial apparatus is here not merely slow or expensive. It is pathogenic. Judges, interrogations, publicity aggravate the symptoms and may render the condition incurable; isolation, by contrast, produces excellent effects. On this footing the removal of the courts becomes an act of therapeutic beneficence, and the modification of the Civil Code an obligation of humanity. Castel's formulation that legal legitimacy was displaced by a legitimacy medically rather than juridically founded describes precisely what Dufaure was doing;[49] the parliamentary record lets us watch him do it. Dufaure also spelled out the property consequence: with the bar of article 504 lifted, relatives would no longer need to provoke interdiction in order to preserve the patient's fortune, and could avoid “cette fatale mesure qui compromet, pour la conservation d'une fortune, une raison, une volonté, une vie morale tout entière.”[50]

Isambert rose immediately in reply, and his objection was jurisdictional rather than substantive: the government itself had proposed nothing on this head; it was a new proposition tending to change an order of things established by the Civil Code, which contained separate provisions for imbecility, dementia and fury.[51] He was answering a therapeutic argument with a legislative one, and he lost.

5.6 Charamaule and the tainted witness

Judicial control was not the only axis of contention. On 5 April 1837, in the article-by-article debate on the commission’s article 5 — article 8 of the enacted statute — the deputy Charamaule fastened on a different defect: the text nowhere said who might petition for an admission. The commission required the head of the establishment to be furnished with four documents only: a request signed by “the person who will effect the placement”; a medical certificate as to the person’s mental state, issued not more than fifteen days before delivery; a birth certificate, passport or other proof of identity; and, if an interdiction had been pronounced, an extract of the judgment. Nothing required the petitioner to be a relative. “If one takes the trouble to examine the commission’s drafting,” Charamaule said, “one will be struck by this lacuna: nothing whatever is required by way of an expression of the family’s opinion.” Hence: “thus, by means of this certificate, the first comer may seize another and have him detained as insane.”[52]

What made the lacuna dangerous, on his account, was precisely the household economy this article has been describing. Charamaule graded the kin. For a spouse, descendants or ascendants one might still rely on “the bond of the affections.” Where only collaterals remained, however, “their interests are often divergent and opposed”:

Ceux qui espèrent des dispositions favorables inclinent à penser que leur parent jouit de toute sa raison ; ceux qui craignent d’être exhérédés inclinent à penser que leur parent, s’il a déjà quelques infirmités attachées à la vieillesse, est bien près de retomber dans l’enfance. Ainsi la famille se divise ; on voit souvent, après la mort des testateurs, des demandes en nullité des dispositions testamentaires, sous prétexte que le testateur n’était pas sain d’esprit.[53]

His fear was that the new statute would bring these quarrels forward into the testator’s lifetime: “Je crains que la loi ne hâte ces discussions et ne les fasse éclater souvent du vivant du testateur.” His amendment accordingly proposed a fifth requirement: the assent of the spouse, adult children or ascendants; failing them, a deliberation of the family council; and, where the family council was itself divided, an ordonnance de référé of the president of the civil tribunal of first instance. “Will you deliver the fate of the alleged lunatic to a plurality of votes? Better, it seems to me, to refer the matter to an impartial magistrate and to subordinate the divergent opinion of the collaterals to the conscientious opinion of the magistracy.”[53]

The amendment was rejected. Over the following day two further attempts to write family consent into the conditions of admission — La Rochefoucauld-Liancourt’s (“admission shall not become definitive except upon the consent signed by the family or by the spouse”) and Glais-Bizoin’s milder proposal that the petition be accompanied by the consent, or the stated grounds of refusal, of those liable for maintenance — were likewise put and likewise rejected.[54] Vivien’s grounds of refusal are worth separating out. The first was the family name: “mental alienation is a kind of blemish; it is a sort of humiliation cast upon the family to which the sufferer belongs… To spare these legitimate scruples, to reassure families, the law must not require formalities that might invite publicity.” The second was urgency: since the physicians had declared with one voice that “isolation was the first remedy,” delay arising from formalities “injures the situation of these individuals, risks compromising their state, and may degenerate an alienation of no determined gravity into a violent and sometimes fatal fury.” The third was the most candid of all:

C’est pour obtenir ce résultat que nous n’avons pas voulu faire une loi judiciaire, de procédure, une loi de chicane, que nous n’avons pas voulu imposer des formalités désastreuses, onéreuses, contraires aux vues que nous nous proposons. Nous avons considéré d’abord l’intérêt du malade, parce que c’est dans cet intérêt que la loi est faite.[55]

Pressed twice on the identity of the petitioner — “Par qui ?”, “Et quelle sera cette personne ?” — Vivien declined to define it, and made the refusal itself a principle: the petitioner “will be the person who is called, by the ties of blood and, in certain cases, by the circumstances in which the unfortunate sufferer finds himself, to take this measure with regard to him. It is impossible to foresee all cases.”[55] It should also be recorded that the counter-interest was named in the same debate. Glais-Bizoin, explaining why Charamaule had gone too far, observed that relatives might unite in refusing consent “whether from certain prejudices or from the fear of having to furnish a maintenance allowance” — a possibility that is anything but abstract once one reads it against the chain of liability in articles 25–28.[54]

What, then, did the legislator put in the place of the safeguard it had refused? Vivien’s answer determined the character of the statute: not a magistrate and not a family council, but a plurality of mutually independent physicians. “I ask whether it is possible to suppose that the physician who must give the certificate before admission, the physician attached to the establishment, and finally the one who will be commissioned by the prefect to make the last visit, should all attest a state of mental alienation which did not exist.” On that footing he could dismiss the fear of a sane person being confined as “dérisoire.”[55] The threefold exclusion written into article 8 — the certificate is inadmissible if issued more than fifteen days before delivery, if signed by a physician attached to the establishment, or if the signatory is a relative or connection within the second degree of the proprietors of the establishment or of the person effecting the placement — is the residue of that design.

Two qualifications should be entered against Vivien’s threefold gaze, both of them visible in the same debate. The second paragraph of article 8 — “public establishments may dispense with requiring the production of the physician’s certificate before admission” — was adopted without objection, so that the prior certificate from an outside physician was obligatory only for private houses; and the second of the three examinations, that of the establishment’s own physician, was made by a man whose institution drew its revenue from the patient’s continued residence. The one examination that was both prior and disinterested was the first. Nevertheless the design outlived the statute that embodied it: the law of 27 June 1990, which repealed the law of 1838, required for hospitalisation à la demande d’un tiers “two detailed medical certificates dated less than fifteen days earlier,” provided that “the first certificate may be established only by a physician not practising in the establishment receiving the patient,” and barred both physicians from being relatives or connections within the fourth degree of each other, of the directors of the establishment, of the petitioner or of the patient. The fifteen days, the outside physician and the exclusion by degree of kinship recur, word for word, a hundred and fifty-two years later.[56]

The Chamber, in short, did not fail to see the danger that testimony about a person’s sanity might be corrupted by an interest in that person’s property. It saw it, debated it over two days, and chose to meet it with medical plurality rather than judicial guarantee. The cost of that choice is worth stating plainly. Removing the physician from the interest is not the same as conferring on him the power to adjudicate interests. Along the route of placement volontaire there is examination but no hearing: several physicians see the patient, but no procedure anywhere in the statute takes his account as the submission of a party and determines which of two conflicting depositions is true. The function Charamaule wished to entrust to “an impartial magistrate” — deciding the facts as between divergent testimonies — was left, in the law of 1838, assigned to no one.

5.7 The shape of the statute, and who pays

The law as enacted rested on two pillars. Departments were required to establish an asylum or to contract for provision (article 1). Admission was regulated on two tracks: placement d'office, ordered by the prefect on grounds of danger to public order or the safety of persons (articles 18–19), and placement volontaire, effected on the petition of a private person supported by a medical certificate and accepted by the head of the establishment (article 8).[65][57] Neither track passed through a court. Interdiction was detached from admission as a requirement, and the judiciary was relegated to ex post supervision — periodic visits by the public prosecutor (article 4) and jurisdiction over applications for discharge, which anyone might bring (article 29).

It is worth recording that Vivien's report had already formulated the separation as a positive doctrine rather than a mere convenience:

L'objet direct et principal de la demande en interdiction est étranger aux mesures dont la personne de l'aliéné peut être l'objet… l'interdiction a surtout en vue les intérêts matériels, la gestion des biens ; les mesures dirigées contre la personne se rattachent au traitement de la maladie, à la sûreté de l'aliéné, à celle de ses parents.[58]

The report also records that the commission struck out the government's requirement of prior prefectoral authorization for private admissions, on the ground that a family effecting the placement itself remained charged with the whole responsibility of the act and answerable for it to the justice of the country, whereas if the prefect had authorized it the responsibility would pass entirely to the administration.[59] The point is not trivial for the present argument: at the very moment when the state was taking the courts out of the admission process, the commission was taking care to leave the family in it — as the bearer of responsibility.

The financial provisions (articles 25–28) established a chain of liability: the costs of maintenance fell first on the patient's own property; failing that, on those liable to maintain him under articles 205 and following of the Civil Code — that is, the family; and only where both were wanting or insufficient upon the department, through centimes added to its ordinary expenditure, together with a contribution from the commune of domicile. The law of 1838 was thus, among other things, an apparatus for transferring costs, with a means test built into it: only the maintenance costs of the indigent patient whom the household economy could no longer carry were transferred to the department. Exclusion and relief were two aspects of one fiscal operation. And the central government paid nothing, leaving the distribution to be negotiated between the departmental and communal councils.

Castel's summary of the legislative process is worth retaining here, because it corrects a widespread assumption about the order of events. The minister's initial draft was inspired by a purely administrative logic and would have given the prefect the bulk of the powers; what transformed it were the repeated interventions of the alienists' spokesmen in the two Chambers. “Thus the law of police and finance became a law of beneficence and public relief.”[60] The administrative solution was the starting point, not the destination; the medical justification was, in significant part, a product of the parliamentary process. And Castel is careful not to call the result a victory: it was a compromise laboriously negotiated on several levels, which secured the possibility of psychiatry's future development and simultaneously foreclosed it.[61]

6. The Comparative Test: Scull Reframed

Andrew Scull tied the growth of institutional confinement in England to the establishment of a labour market: where labour power is commodified, households lose the capacity to carry members who cannot work, and the demand for “disposal” is directed at the institution.[62] The argument of Sections 4 and 5 above might appear to be the mirror image of this — in France the survival of the household economy delayed confinement — but the symmetry does not hold, and it is worth saying why.

Two lines of criticism have qualified Scull's thesis. Akihito Suzuki has shown that in England too the family remained the primary agent of care, and that domestic management persisted well into the period of asylum expansion.[63] Peter Bartlett has located the driving force of confinement not in capitalism in general but in the specific structure of Poor Law administration.[64] Taken together, these corrections suggest that the Franco-English difference is not adequately described as the presence or absence of proletarianization. It is better described as a difference in the administrative channel that received the cases exceeding the household's defensive limit: in England the Poor Law apparatus of parish and union; in France the departmental administration constituted by the law of 1838.

On this reformulation the French particularity consists in two things. First, the persistence of smallholding society meant that breaches of the household's defensive limit occurred as discrete events rather than as a mass phenomenon — the demand arrived case by case, which is precisely the pattern the Quimper files display. Second, the channel that received those cases was neither judicial nor a poor-relief bureaucracy but a novel conjunction: the prefect's police power joined to the physician's certificate. It is that conjunction, and not the volume of demand, that distinguishes the French case; and it was, as Section 5 has shown, chosen rather than given.

7. Conclusion

The law of 30 June 1838 is, before it is a monument of psychiatric progress, the document of a bargain struck between the household economy of a smallholding society and the state at the point where that economy reached its defensive limit. The substance of the bargain, written in the language of therapy, was the transfer to the department of the danger and the cost that the household could not carry; and the courts were bypassed so that the transfer could be executed quickly.

But the transfer was possible only because the contractual society had needed, from its foundation, a second status for those incapable of contract — a guardianship neither judicial nor familial. Castel was right that this need was already outlined in the 1790s. What the parliamentary record adds is that the form the guardianship took in 1838 was not deduced from the need. It was chosen, against a fully specified judicial alternative, from between two incompatible revolutionary legacies; and the man who defended the losing legacy described the winning one, in advance and exactly, as an interdiction in miniature stripped of its guarantees.

What made the choice possible was not the force of the alienists' arguments alone. It was that, from below, the households of a fragmented peasant society were demanding a route to exclusion that would be quick, cheap, and free of dishonour, and that the interdiction procedure could offer none of the three. The law of 1838 is the record of these two demands, from above and from below, meeting at a single point; and at that point stood the prefect's signature and the physician's certificate.

The subsequent history is one of persistence. The law governed French psychiatric practice for a hundred and fifty-two years and shaped mental health legislation across Europe. Its successors — the laws of 27 June 1990, 5 July 2011 and 27 September 2013 — represent a gradual democratization in the direction of patients' rights, and the introduction in 2011 of mandatory review by the juge des libertés et de la détention marked the first systematic judicial control over psychiatric committal in French law since 1838. It is difficult not to read that reform as the return, after one hundred and seventy-three years, of the legacy Isambert defended and lost. The institution of compulsory psychiatric treatment, however, has persisted through every one of these transformations.