1If a newborn dies before its birth has been registered, should a birth certificate and a death certificate be issued, or just a single certificate for an infant who was “presented lifeless”? From what gestational age should an infant “deceased in the mother’s womb” be registered? Since the introduction of civil registers after the French Revolution, legislators have grappled with the issues of how to define and record stillbirths, and their responses have varied over time. Vincent Gourdon and Catherine Rollet examine the debates that surrounded the introduction of legislation in France during the nineteenth century, parts of which remained in force until the late twentieth century. The authors explain the implications of these definitions in fields as varied as law, medicine, religion and statistics. The issues explored in this article are still relevant today. The category of “false stillbirths”, for example, only disappeared from French statistics in 1993, with the introduction of birth certificates for infants who die before they are registered. The minimum gestational age for registering the birth of stillborn infants was lowered from 28 to 22 weeks in 2001, then lifted altogether in 2008, causing a discontinuity in the statistical series that would doubtless have displeased the statisticians of the nineteenth century.
“Civil registration such as France possesses today can be offered as a model to all peoples: nowhere else have the conditions that underpin the security of man in the state of society been determined with such wisdom and foresight: when a man is born, marries, or dies, the admirable concision of the law fulfils its designated purpose; these circumstances only have to be certified with the diligence that the law prescribes, and a civil status is created; where they are not certified, or are improperly certified, there is no civil status, and disorder and confusion take hold”.
3“The term ‘stillbirth’ is used to designate not only foetuses that die in their mother’s womb but also infants that are not born viable and those who die a few hours after birth” (Bayard, 1846, p. 3). Thus commented Dr Bayard, a medical expert to the courts, in 1846. Contrary to Mr Pontonnier’s lofty declaration cited above, and its depiction of an apparently simple French civil registration system, the category of stillbirths was far from straightforward: its very name – in French mort-né, literally “dead-born” – encapsulates its composite nature. Three groups of infants are enumerated, with reference to the processes of pregnancy, childbirth, and life immediately after birth. Indeed, a homogeneous definition was still a long way off, and the situation was typical of the vagueness characterizing many notions under French law (such as “the child’s interest” in the Civil Code). The category of stillbirths was also hazy from a statistical point of view, since it comprised both infants who were born alive and infants who were born dead.
4This article examines the construction of this ambiguous category in the nineteenth century. It does not attempt to deconstruct the statistical data to assess their usefulness as an indicator of demographic realities. Neither does it seek to correct them, as Etienne van de Walle (1974), Noël Bonneuil (1997), Jacques Vallin and France Meslé (1988, 2001), Victor Kuagbenou and Jean-Noël Biraben (1998) have done, and as Robert Woods has done (2008) for stillbirths. Nor is it intended to be a detailed demographic analysis of the phenomenon of stillbirths in terms of sex, social status or place of delivery.
5Our aim is to shed light on the very processes that shaped the construction of this composite category. These were complex processes, resulting from the interaction of several factors existing on quite different social planes. It is this complexity in relation to “stillbirths” that we investigate here. How did this category develop in law? Which administrative, legal, medical and religious constraints influenced the choices made? Under what circumstances were the measures adopted?
6Our analysis is conducted at three levels. First, at the level of the state as the producer of norms (the ministries represented by the prefectures), through an examination of laws, rulings, circulars, decrees, etc. (in the national archives and in journals such as the Annales d’hygiène publique et de médecine légale and the Revue d’hygiène et de police sanitaire). Second, at the level of the city/département, namely Paris and the Seine département, which applied those norms and took initiatives, as we shall see below (the Paris archives provide abundant material in this respect). Last, at the level of the statistics published by the statistical services for Paris, which cover the whole of the nineteenth century (Annuaire statistique de la ville de Paris). There is an intermediate level to which we did not have access, namely the practical application of the laws and prefectoral circulars on the ground. The Paris civil register was destroyed in the fire that gutted the Hôtel de Ville and its annexes during the Commune, ruling out any detailed study of the certificates issued for stillborn infants. However, the reports on the Paris civil registration system in the above-mentioned journals and analyses of the registers of neighbouring départements provide some useful pointers.
7Despite gaps in the archives, the case of Paris is interesting because the French capital led innovation in terms of the day-to-day management of citizens’ lives. Like other European capitals, Paris had the financial and human resources to establish sophisticated monitoring systems. The city was also the receptacle of contrasting developments in law and social practices: the elites strongly influenced behaviour related to births and deaths, and even baptisms and burials.
Our starting point is a presentation of the stillbirth rates in the city of Paris over the nineteenth century, based on data from the statistical yearbook of the city of Paris (Annuaire statistique de la ville de Paris). Next, background information is introduced to explain changing trends in the stillbirth rate over time, starting with the Ancien Régime, followed by a thematic plan analysing in more depth the issues of “false stillbirths”, the circumstances behind the improvement of records – the verification of deaths and births in the home – and last, the registration of foetuses. We will thus reveal the fault lines of the debate between legislators, physicians, statisticians, municipal functionaries, religious authorities and families on the issue of stillbirths, which remains sensitive to this day.
I – Statistically problematic trends in the stillbirth rate
8The stillbirth rate, calculated as the ratio of stillbirths (l’Annuaire statistique de la ville de Paris) to total births in a given year, was between 5% and 7% in the first half of the nineteenth century and well over 7% subsequently (Figure 1).
Stillbirth rate in Paris in the nineteenth century (stillbirths per 100 total births)
Stillbirth rate in Paris in the nineteenth century (stillbirths per 100 total births)
9The curve of stillbirth rates in the Seine département (not shown here) is parallel but slightly below that of the city of Paris (the difference is two percentage points). In the two arrondissements of Saint-Denis and Sceaux, still very rural, the registration of stillbirths is much less frequent than in Paris, especially at the beginning of the period.
10These figures came as a shock to observers at the time, as can be seen in the remarks of Adolphe Trébuchet (1801-1866), chief of the sanitary bureau at the prefecture, echoing those of Alexandre Moreau de Jonnès (1778-1870), director of Statistique générale de la France (SGF). The stillbirth rate in Paris was twice as high as in the country as a whole, a fact which was hard to explain; de Jonnès did not, however, accept an increase in “criminal acts” in the capital as an explanation in the 1840s. It did not occur to Trébuchet to question registration procedures (Trébuchet, 1851). In 1851-1853, according to the SGF, the département of Seine ranked first or second nationally for the proportion of stillbirths, with a rate of more than 6 per 100 total births. Here again, seeking an explanation, the two authors neglected registration procedures, instead pointing to excess stillbirths among infants born out of wedlock: “the circumstances [of illegitimate unions] are excesses of all kinds or excessive deprivation, the prolonged dissimulation of pregnancy by expedients most dangerous for the health of the child, attempted abortions, clandestine deliveries, etc.” (SGF, Mouvement de la population en 1851, 1852 et 1853, § 6, 7 and 8).
11When questioned by parliamentarians on 4 February 1874 about the rise in stillbirths in France, Dr Louis-Adolphe Bertillon (1821-1883) was prudent, because, as he explained, the registration of stillbirths began much later than the civil register, only really commencing in 1840: “The population is not yet used to the idea of registering a stillborn child separately. One can therefore be certain that in the countryside it is still customary to bury aborted foetuses and numerous stillbirths which, in the cities, are registered” (Rapport à l’Assemblée nationale, 1874, Annexes, p. 110). To avoid regrettable confusion, Bertillon suggested adopting the Belgian model, which distinguished clearly between the different categories of stillbirths, according to whether the child died before, during or after birth (Bertillon, 1876).
12As confirmed in 1905 by Jacques Bertillon (1851-1922), stillbirths in Paris in the nineteenth century increased in stages, reaching more than 8% after 1890, and nearly 9% just prior to 1900 (Bertillon, 1907).
13We note, however, that the percentage of stillbirths was in decline up to the mid-1820s, confirming the “constant decrease under the Consulate and the Empire” in Metz (Lhote, 1970, p. 459). The practice of institutions run by nuns, disinclined to report stillbirths occurring at an early stage of pregnancy, seems to have played a part in the quality of registration, as noted at the time by the prefect Chabrol and the hygienist and statistician Villermé (Chabrol, 1826; Villermé, 1826). It was not until after that period of decrease that the percentage of stillbirths rose sharply to a peak in 1871, and to another peak in the 1890s.
This trend in stages can be attributed to several processes: first, the civil government’s reluctance to legislate in this area, whereas the Ancien Régime had adopted a fairly simple logic on the issue; next, an improvement in the recording of stillbirths due mainly to the inclusion of infants that died shortly after birth; and last, the emergence in the 1830s and 1840s of a growing concern to register “miscarriages” with a view to preventing “criminal” acts, namely abortion and infanticide.
The category of stillbirths, as evidenced in the statistics for Paris, is of direct relevance for the recording and production of statistics. It raises questions relating to conception, or the first months after conception at least, and, further down the line, to the birth of a child, who may be born alive or dead. It is this intermediate state, between life and death, that will be explored here.
II – Changes to the legislative framework over the nineteenth century
The legacy of the Ancien Régime
14To understand the complex construction of the category of stillbirths, it is useful to look back at the registers of the Ancien Régime. The civil register in pre-revolutionary France did not record demographic events (births and deaths) per se, but the religious ceremonies associated with them. In the sixteenth century, royal ordinances required the clergy to keep records of baptisms, burials, and later, marriages. Likewise, the Council of Trent (1545-1563) ordained that registers of baptisms and marriages be kept, and in 1614 the Roman Ritual added the Liber mortuorum (Le Mée, 1999a, pp. 29-36). These registers did not have a statistical purpose; they provided the clergy with information on the parish population, enabling them to oversee marriages and identify any impediments due to consanguinity or spiritual incest, etc. Strictly speaking, there was no reason for stillbirths to appear in parish registers. Since these infants did not live, they were not baptized – it is sacrilege to baptize a deceased person – and since they were thus prevented from joining the Christian community, they were not supposed to be registered in the burial records. This non-registration represented a more crucial double exclusion. In the Catholic view of the time, not only were the souls of stillborns and children who did not live long enough to receive the sacrament of baptism deprived of the vision of God (a place – limbo – was nevertheless reserved for them), but their bodies were also excluded from the right to burial in consecrated ground among their forebears, and had to be interred in a separate, profane area (Gélis, 2006, p. 26).
15In practice, families were often spared that trauma. Many stillborn children were recorded in the Catholic registers because they were given the benefit of the doubt (provisional baptism), or because priests and parents generously recognized signs of life, which allowed emergency baptism (ondoiement en péril de mort) be performed either at birth or in “sanctuaries of respite”, where the child’s body was taken in the hope of a miracle (Gélis, 2006).
Nevertheless, with respect to registration, stillborn infants, even those who had received an emergency baptism, did not attract much attention before the eighteenth century. Although the monarchic state displayed an early concern to curb infanticide (the edict of Henry II of March 1556 required unmarried pregnant women to declare their pregnancies), it was not until 1736 that a royal declaration required every child receiving an emergency baptism by a lay person or a priest to be entered on the baptism registers (Le Mée, 1999a, pp. 38-45). Should this be interpreted as a new desire by the monarchy to account for these intermediate situations, or simply as a transposition of the church’s new rigour on this matter? In any case, the parish clergy were encouraged to draw clearer distinctions and to mention cases of baptism performed on the boundary between life and death. However, in strict canon law, the deceased who had received emergency baptism could not be equated with stillbirths, and the latter category, inconsequential in religious terms, remained absent from the legislation and therefore poorly assessed until the end of the Ancien Régime.
Contradictory legislation: 1792-1806
16The general reorganization of the law that began during the Revolution and culminated in the Napoleonic Code affected the definition and recording of stillbirths through three key developments: the introduction of declarations of birth, the definition of the conditions for recording deaths, and the issue of succession.
17The secularization of civil registration in September 1792 did not immediately dispel the vagueness surrounding stillbirths, which was a peripheral matter for the Assembly at the time. The main consequence was a transfer of authority for the keeping and custody of records, but with no significant change in registration practices. While the term used was now “birth”, the procedure was a virtual replication of the prescriptions of baptism. Birth certificates were to be drawn up within 24 hours, as in the royal declaration of 1698 on baptisms. Two witnesses were required, potentially of different sexes, much like a godfather and godmother. In cases of “imminent peril”, the public officer went to the home of the newborn to draw up the certificate, as did the priest for emergency baptism of infants in mortal danger. Lastly, the wording of the birth certificate was modelled closely on its Catholic antecedent (provision of law in Duvergier, 1834, pp. 482-488).
18While the Napoleonic Code made no reference to the recording of stillbirths, it did change the facts of the matter, however. Article 55 of the code stipulates: “Declarations of birth shall be made, within three days after delivery, to the civil officer of the place: the child shall be shown to him”  (Code civil, 1909, p. 24). The three-day deadline, stipulated by Bonaparte in person – while the first draft submitted to the Council of State maintained a 24-hour limit (Discussion, 1841, pp. 73-74) – seems to have been inspired by the time limits for baptism prescribed by the Archbishop of Paris (Synod of 1673), based on the logic of immediate baptism for Catholics as confirmed by the Council of Trent. While under the Ancien Régime, the obsession with baptism led to stillborn children being considered as having lived, the extended deadline introduced the opposite risk of having children who were born alive considered as stillborn if they died before their birth was declared. The expression “the child shall be presented to him” in practice created a requirement that was to create numerous problems relating to both live births and stillbirths. Article 56 completed the provision by requiring any person who witnessed a birth (such as the father or a physician) to declare the birth, or incur the penalties set forth in Article 346 of the Penal Code (Code d’instruction criminelle et code pénal, 1910, pp. 434).
19The Revolutionary, then Napoleonic, civil legislation was confronted more directly with the issue of stillbirths when registration of burials of the baptized were abandoned in favour of registration of deaths. The Act of 1792 made no reference to stillbirths, simply requiring that a death be declared to the public officer by the two closest relatives or neighbours of the deceased (Title V, Article 1), and that the officer pay a visit to make certain of the death before recording it in a double register (Title V, Article 2). However, the example of Montfort l’Amaury, a canton in Seine-et-Oise, suggests that on the ground, confronted with events on the boundaries between life and death, timid developments were taking place. Although individual civil officers drafted the certificates in different ways, they all issued only one death certificate in the event of death in utero, during delivery or within one hour of birth, but two certificates (birth and death) when the newborn child lived for at least an hour. The terms used to refer a stillbirth (the term “mort-né” was never used although it was known in medical and religious circles) were extremely diverse (19 cases): mort avant la naissance, né mort, déjà mort, mort avant l’accouchement, mort avant qu’il n’ait vu le jour, venu mort, aucun signe de vie, n’a donné aucun signe de vie, ni avant, ni après (dead before birth, born dead, already dead, dead before delivery, dead before seeing the day, arrived dead, no signs of life, gave no sign of life either before or after) (Malphettes, 1984, pp. 7-13). Yet although approximate, the certificates distinguished between infants truly born dead and those born alive, which was an improvement on the Ancien Régime, where this question received little attention.
20The Napoleonic Code continued the spirit of the Act of 1792. Article 77 asserts that “No interment shall take place without an authority on paper, free of all expense, from the officer of the civil power, who is forbidden to deliver it until he shall have been conducted to the deceased person to assure himself of the death, and that not later than twenty-four hours after decease, except in cases otherwise provided for by the regulations of the police”. Detailed recommendations are given concerning witnesses (Article 78) and the drafting of the certificates (Article 79), but nothing specific is said about stillbirths.
21A highly significant development occurred in Year III (1795), however. Aiming to clarify the form of acts of civil status, the authorities turned their attention for the first time to the declaration and registration of stillbirths: “a decision by the Legislation Committee dated 8 Thermidor Year III [26 July 1795] has simply prescribed that each municipality receive declarations of stillborn infants and record them on the civil registers, for the surety and peace of the families”.  That decision by the Legislation Committee is recorded in a document produced by the bureau of the administrative, civil and military police of the département of Paris, which cites an “excerpt from a letter of the representatives of the people who form the Legislation Committee dated 19 Fructidor of the current year” (5 September 1795) addressed to the administrators of the département: “The declarations of stillborn infants shall be recorded on the civil register, a precaution that is necessary either for the surety of the women or girls who were the mothers, or in the interest of the families” (true excerpt signed Marquet). 
22The reliability of declarations remained problematic, however. In this regard, the prefect Nicolas Frochot (1761-1828), State Councillor and the first prefect of the Seine département (appointed on 1 Germinal Year XIII ) took an interesting initiative on the registration of stillbirths in Paris on 21 Vendémiaire Year IX (1801), which probably had an effect (Annales d’hygiène publique et de médecine légale, 1843, 30, pp. 118-159). We will return to this point below. Considering that relatives, neighbours and civil registrars did not have the skills required to establish the death of infants and prevent precipitate burials, Frochot implemented a system of verification of death in the home by medical officers. When a death was declared, the mayor dispatched a medical officer to the home of the person presumed dead. The death certificate was then drafted and signed on the basis of the medical officer’s report. Burial within 24 hours of the declaration by the family was prohibited, except in special cases. The death verification service in Paris dates from that time. Another letter by Frochot addressed on 4 Thermidor Year IX (1801) to the mayor of the 6th arrondissement is interesting because it shows the authorities’ interest in deaths of children aged under 15: the prefect asks the mayor to determine the exact status of all deceased children, indicating their name, age and sex and the cause of death. A standard form was appended that included a stillborn column (mort-né).  The term “mort-né” only appeared in statistics, however, not in the provisions of law.
23The law of succession also influenced the approach to stillbirths, by introducing, with the Civil Code, the important concept of viability. In the chapter entitled “Of the Qualities Requisite to Succeed” in the Napoleonic Code, Article 725 provides that “In order to succeed, the party must of necessity be in existence at the moment at which the succession is opened. Those incapable of succeeding are, 1st. He who is not yet conceived; 2d. The child who is born not viable; 3d. He who is civilly dead.” (Code civil, 1909, p. 180). Hence, a non-viable infant could not inherit, and none could inherit from him. Only a child born alive and viable, even if it only lived for a short while, was entitled to inherit from its parents or the surviving parent. A man whose wife died in childbirth and whose child died two days later, would inherit from his child the property left to it by the deceased mother; similarly, if the husband died, his widow would inherit from his child. If there was no child, however, the husband (or wife) could not inherit from his (her) spouse. Hence the importance of indications of the child’s “viability”,  the duration of life after delivery, and, in the event of the mother’s death, the order of the mother’s and child’s deaths. This was essential information for establishing the rights of the heirs. Viability was therefore a fundamental concept, even if Article 725 did not define its limits.
Other chapters of the Civil Code nevertheless suggest that a threshold of 180 days of gestation was favoured by the legislator, despite the reticence of eminent physicians of the day (Foderé, Marc), who considered such a view of “viability” to be a functional legal definition not based on scientific evidence (Betta, 2006, pp. 121-127). In the chapter “Of paternity and filiation”, Article 314 of the Code stipulates that 180 days (i.e. 6 months) represents the duration of the marriage below which a husband cannot disavow a child “if the child is not declared viable”. Similarly, under Article 312, a father can only disavow his child if it was impossible for him to have been living with his wife between the 300th and the 180th day prior to the child’s birth (Code civil, 1909, pp. 86-87).
So the public registration of stillborn infants, born at term or before term, was not simply a matter of civil records and statistics. It became a vital matter for families which could lead to disputes over inheritances and involve the judicial system. Yet the category of stillbirths and their mode of registration were still not clearly defined.
Towards the Decree of 4 July 1806
24This situation, still rarely debated, changed in 1806 in the context of a dispute over the place and role of national statistics in France. The public authorities seem to have been hesitating since 1803 between descriptive topographies based on the German model (i.e. continued production of voluminous statistical reports to obtain a comprehensive image of France) and the English model, closer to political arithmetic, i.e. a more critical use of record-keeping in the human sciences (Dupâquier, 1985). One of the proponents of such state “geometry” was Emmanuel-Etienne Duvillard, a Geneva-born mathematician (1755-1832) specialized in life annuities and a member of the bureau of statistics at the ministry of the interior (Thuillier, 1997). In late 1805 and early 1806, he produced copious notes, dissertations and reports to alert the authorities to the importance of calculation and verification, rather than mere accumulation. He wanted, in particular, to achieve greater clarity in the records of population movements sent by the prefects to the ministry of the interior and, to this end, he enumerated the desirable changes point by point.  But he was not always listened to, and in 1806 experienced a series of professional setbacks: after achieving his requested promotion to sub-director of the bureau of statistics at the Ministry of the interior in August 1805, he was not appointed bureau chief and returned to his earlier position of deputy in April 1806.
25It was against this backdrop that, in 1806, a discussion took place at the highest level on “how to certify the birth of a child who was born alive but died before there was time to present it to the town hall and have a birth certificate issued”. Some considered that only one act of civil status, a death certificate, should be issued in this case, “including a declaration by the witnesses that the child came into the world alive”. Others argued that two certificates should be issued, “a birth certificate and a death certificate, on the grounds that, since the two events occurred, both should be recorded in accordance with the formalities prescribed by the law”. The prefect Frochot decided to consult the grand judge, the Minister of Justice. The minister upheld the second opinion on 25 March 1806, declaring that there would be “infinitely grave consequences, both for the order of succession and for the peace of families, if registration were limited to a death certificate in such cases”. In his letter dated 9 April 1806 (Archives de Paris, VD41, pièce 66), Frochot asked the mayors to comply with this decision, while remarking that “there would be something repugnant” about requiring parents to present a dead infant at the town hall. That objection was examined and discussed, explains Frochot, the formality of a presentation being implicitly fulfilled by the verification of death. The “decision” of the minister of justice therefore required that two certificates – birth and death – be issued, with the former including the detailed declaration of the witnesses who attest that the child had lived, along with mention of the fact that a death certificate has also been issued “with every precaution taken to ensure that third parties can later claim their rights […]” (Archives générales de médecine, March 1826). Thus a concern to distinguish between infants born alive and those born dead emerged here, with the clearly stated objective of settling the issue of inheritance rights. It is reasonable to assume that Duvillard had something to do with this clarification.
26Yet barely four months after the grand judge’s “decision”, the decree of 4 July 1806 expressed the problem quite differently, since it required the civil officer to record on the death register infants who were “presented lifeless” to him. The decree concerning “infants presented lifeless to the civil officer”, which applied not only in France but throughout the empire, particularly in the Netherlands and Belgium, deserves to be cited in full: “Art. 1. When the corpse of an infant, whose birth has not been registered, is presented to the civil registry officer, that officer shall not pronounce that such child is deceased, but only that it was presented to him lifeless; he shall in addition receive the declaration of witnesses, and record the surnames, first names, occupations and domiciles of the child’s father and mother, and the year, day and hour at which the child left its mother’s womb. 2. The event shall be recorded on the death registers, without prejudice as to whether the child lived or not” (Code civil, 1909, Article 79, p. 32).
27Thus, in total contradiction with practices adopted since the secularization of civil registration and with all previous laws, if a child died before the birth was declared it was registered as “presented lifeless” on the register of deaths and was not recorded on the register of births. No distinction was made between true stillbirths and infants that Louis-Adolphe Bertillon later proposed calling “false stillbirths” (Bertillon, 1876, p. 6). In a letter dated 9 November 1806 to the mayors of Paris, the same prefect Frochot advised them to apply the new rules and sent them a “standard form for recording acts in accordance with the decree” (Document 1). Nothing was said at the time about how these infants “presented lifeless” were to be accounted for in the statistics.
The precise reasons that prompted the legislator to adopt this position are unknown. Was it out of a concern to prevent civil registrars from establishing a birth certificate when a child was presented to them dead? Or was it to circumvent the problem of false declarations? Indeed, the civil registrar, having no way of ascertaining himself whether the child had been born alive or dead, was obliged to rely on the declarations of the witnesses. Yet these declarations were likely to be false since parents, and indeed all Christians, had an interest in declaring that a child was born alive so that it could be baptized. Likewise, questions of inheritance provided a strong motive for false declarations. In any case, it was probably because of the authorities’ inability to establish the exact truth of the circumstances of the birth – which was only possible through verification of births and deaths in the home – that they preferred to introduce this procedure “without prejudice” as to whether the child had actually lived or not. This reflects a concern to separate matters pertaining to civil registration, in the strictest sense, from those that fall into the legal sphere, such as matters of succession. The debate between the Justice and Interior ministries would continue throughout the nineteenth century.
The “form” used to register infants “presented lifeless”
The “form” used to register infants “presented lifeless”
Infants “presented lifeless”: application of the decree of July 1806
29The evidence indicates that the users of the civil registers disagreed for years about the choice of method for constructing statistics. A comparison – too long to detail here – of the different sources of data on deaths by age for the city of Paris since 1812 (Trébuchet, 1849; Chabrol, 1826; Annuaire du Bureau des longitudes – BnF, cote V-21601) shows that while Chabrol (1773-1843, prefect of Seine from 1814) and the Annuaire du Bureau des longitudes published identical figures, those supplied by Trébuchet were often very different from 1816 onwards. For example, in 1818, Trébuchet counted 3,944 deaths before the age of three months in Paris, whereas Chabrol and the Annuaire des longitudes only counted 2,750. The difference is 1,194, which one suspects may be attributed to stillbirths. Indeed, stillbirths were already included in the Chabrol’s figures and those of the Annuaire, so it appears that Trébuchet counted stillbirths twice among the deaths. Evidently, the statisticians did not have a standard way of processing stillbirth data. Further inconsistency is revealed by a comparison of the data for the city of Paris and the Seine département, which suggests that stillbirths were included in the statistics for Paris but not in those for the département.
30Chabrol attempted to clarify the situation in a letter addressed to the mayors of the 11th and 6th arrondissements in February 1817 (Archives de Paris, VD6619, pièce 1; VD42, pièce 400). Referring explicitly to the decision of 8 Thermidor Year III (1795) and making no mention of the decree of 1806, he requested that mayors, without making any changes to the keeping of the civil registers, “in future kindly subtract stillborn infants from the summaries of deaths intended for statistical purposes. It is sufficient to continue to make numerical mention of them in the column intended for that purpose, notably in the statement of recorded acts that you send me each year”. Chabrol thus sought to cease “continually having to make corrections that take up a great deal of time”.
31There are several interesting aspects to the letter. Evidently the mayors of the 11th and 6th arrondissements had interpreted the decree of 4 July 1806 in the sense of including stillborn infants in the data on deaths sent to the statistical services. The prefect now asks them to “subtract” stillborn infants from the statistics. He also demonstrates a genuine concern for efficiency; his aim is to avoid the time-consuming task of correcting the statistics sent by the town halls. By recalling the decision of Year III (1795), Chabrol also explains the term “stillborn”: “But the child who died in its mother’s womb, who never saw the day, who received no name, and who is not included in births cannot, for that reason alone, be included in deaths”. In his eyes, this is a child who is truly born dead. But the allusion to naming suggests that Chabrol also has a social view of birth which clearly evokes the rite of baptism whereby a child traditionally received a Christian name. This accumulation of different restrictive criteria is by no means redundant for the definition of a category that remained somewhat elusive to the actors on the ground.
32A major contradiction thus emerges between the administrative application of the decree of 1806 and the use of civil registers for statistical purposes.
33The problem stems from the form of the decree, which, on the whole, was suited to civil registration but not to the needs of statisticians. The civil registrars simply had to record an event: a child presented dead or alive at the time of the declaration. However, this means that “a child who lived and who consequently inherited and can pass on an inheritance is placed in the same category as a child who was never entitled to inherit or bequeath” (Archives générales de médecine, March 1826, p. 632). In practice, families could ask for an enquiry to prove that the child was born alive and thus obtain a certificate of acknowledgement. This enabled them to claim the birth of a living child in order to settle a disputed inheritance. The practice created considerable problems for statisticians, however. “False stillbirths” were not included in births but were counted alongside true stillbirths in deaths. As Mr Gasc explained on behalf of the Commission de la police médicale of the Academy of Medicine, the prescribed form “only hampers statistical research, making it impossible to achieve absolute exactness in comparative tables of births and deaths” (Archives générales de médecine, March 1826, p. 631).
34In January 1826, a debate was sparked at the Academy of Medicine by a paper delivered by Louis-René Villermé (1782-1863), commenting on Parisian statistics for the 1817-1821 period compiled by Villot, chief of the bureaux of statistics for the Seine département, and published by the prefect Chabrol (Villermé, 1826, pp. 237-240).
35This debate is highly enlightening for our research. Villermé noted a large difference in the percentages of stillbirths depending on whether the child was born at home or not. When a child was born at home, “a certain period of time elapses between the birth and the declaration” – which was supposed to be made within three days of the birth. When a child was born outside the home, stillbirths were registered “immediately after delivery”. At the lying-in hospital, “any child who comes into the world with unequivocal signs of life is not included among stillbirths” (Villermé, 1826, p. 238). “Consequently, the designation of children born in the home as stillbirths is far too widespread, because it applies to a considerable number of newborns who die a short time after they are born” (Chabrol, 1826, Table 46). On average, excluding births at the lying-in hospital (in the 12th arrondissement), there were 33 stillbirths per 1,000 births outside the home, compared with 61 stillbirths per 1,000 births at home. Villermé also suspected that a small part of the difference could be ascribed to the tendency of “public charitable institutions” to scale down the numbers of stillborn children. Villermé shows great lucidity when he concludes: “Furthermore, of all the matters relating to the population, there is scarcely any that offers more uncertainty than that of stillborn infants” (Villermé, 1826, p. 239). Villermé thus pointed up the role of the declaration procedure: the requirement to present the newborn in person to the civil officer probably encouraged many families, fearing for the baby’s health, to wait until the last moment before transporting the child to be declared – a trend parallel to that observed for baptisms (Gourdon, 2006).
As it continued into the subsequent weeks, the debate at the Academy of Medicine revealed the extreme complexity of the issue of stillbirths. In February, the inconsistencies in the application of the legislation were denounced: “Civil officers apply and certainly must apply these erroneous methods: few of them record on the death certificate the declarations of the witnesses about the number of hours the child lived, and in failing to do so, act arbitrarily […]” (Archives générales de médecine, March 1826, p. 631). Some physicians reflected on ways to improve the system. The expert obstetrician Kergaradec lamented that many civil officers used the term “stillborn” rather than “presented lifeless”, “which expresses an altogether different idea”. Gasc, dubitative, went to the town halls of the arrondissements to check for himself the highly variable ways in which infants presented dead were registered. And Gardien, author of the article on “embryos” in the Dictionnaire des sciences médicales (1815), suggested that birth certificates should indicate “not only whether the child lived but if it was born viable”. This suggestion was immediately rejected by Adelon (1782-1862), a forensic medical specialist (Betta, 2006, p. 167) undoubtedly concerned to protect the interests of his profession, who thought that such an indication exceeded the competence of a mere civil officer and that it would be best to maintain the status quo: “the decree of 1806 is perfectly satisfactory”, he explained, “because it confers upon the interested parties the power to prove that the child lived”. The Academy of Medicine was finally won over by his arguments (Archives générales de médecine, 1826, p. 632).
Although the debate did not lead to visible change, it nevertheless demonstrated that the views of some physicians, as well as of statisticians, could be contrary to that of the administration. For the latter, the key issue was the settlement of successions, which could be done with a certificate of acknowledgement, independently of the civil officer. The debate also showed that local registration practices were variable, since some arrondissements of Paris went as far as indicating the number of hours a child “presented lifeless” had lived. This variability did not prevent the settlement of legal issues, but neither did it facilitate the task of statisticians.
Disputes between ministries
36The authorities themselves were divided and adopted fluctuating positions, as reflected in various circulars issued in the 1830s and 1840s. A circular issued by the Minister of the Interior in May 1837 again prescribed that a death certificate but not a birth certificate be issued for infants who died before the declaration of their birth (Le Mée, 1999b, p. 75). The debate between the Minister of Justice and the Minister of the Interior, visible in 1806, remained heated, and a new dispute arose between the Ministry of the Interior and the Ministry of Agriculture and Trade.
37A letter by Rambuteau (1781-1869, Councillor of State, prefect of Seine from 1833 to 1848) addressed on 16 April 1841 to the mayor of the 1st arrondissement of Paris (Archives de Paris, VD42, pièce 404; the same letter to the mayor of the 6th arrondissement (VD41, pièce 230) detailed the first conflict. While the Minister of Justice advocated strict application of the decree of July 1806 (which “prohibits the civil registry officer from mentioning any circumstance other [underlined in the letter] than the presentation of the dead body of the child”, the Minister of the Interior wanted a “distinction to be made between infants that were born dead and those that died before the declaration of birth” as well as a distinction by sex and legitimacy. Confronted with this dilemma, the prefect asked the mayor to distinguish stillborn infants by sex and legitimacy in accordance with the decree of 1806. The addressee of the letter notes in the margin: “I do not fully understand what is asked of me”. That is hardly surprising since the position of the two ministries was quite simply contradictory. Provisionally, Rambuteau favoured the opinion of the Ministry of Justice, but in order to satisfy the Minister of the Interior, he added a paragraph stipulating that “the reports of the examination of the bodies of these infants [deceased before the birth was declared] shall mention any declarations by the parents as to the length of the life of infants who were not already deceased when leaving their mother’s womb”. The prefect observed that some certifying physicians “already take pains to do this”. In future such useful information was to be collected so as to enable his services to “establish the number of true stillbirths, if not authoritatively, then at least with sufficient probability”. Thus, the prefect divided the procedure into two parts: the civil registrars would apply the decree of 1806 to the letter, but the certifying physicians would add an indication of the duration of life, thus enabling the bureau of statistics to establish reliable or at least sufficiently credible statistics. Over time, the decree of 1806 was applied more effectively, but the verification of deaths in the home, thanks to the death certificates filled out by physicians, offered a way of circumventing the decree for the benefit of statisticians.
38The second dispute was between the statistics services of different ministries. Should false stillbirths be distinguished from true stillbirths? Should only false stillbirths be included in birth statistics? Or should both be excluded? On 15 June 1839, a circular from the Ministry of Agriculture and Trade (bureau of Statistique générale de la France, SGF) indicated that stillbirths should not be counted among deaths but mentioned in a special column, while “taking care to avoid” counting as stillbirths those infants who died before their birth was declared. The latter, it explained, “must not be separated from infants that were born viable” (Archives nationales, F20552). Two sets of statistics were thus produced, one at the Ministry of the Interior, which applied the decree of 1806, and the other at the Ministry of Trade (SGF) which requested that a distinction be made between true and false stillbirths. The subsequent circulars and dispatches (26 January 1844 and 29 November 1847) contradict each other, due to the existence of these two statistical departments at the Ministry of the Interior (Bureau d’administration générale headed by Alfred Legoyt), and at the Ministry of Agriculture and Trade (Bureau de la Statistique générale de la France, headed by Moreau de Jonnès from 1835 to 1850).
39The amalgamation of the two departments in 1852 under Legoyt (Le Mée, 1999b, p. 72) was a step towards clarifying the situation, but not without hurdles. In a circular dated 6 July 1852, the administrative services of the new Minister of the Interior, Agriculture and Trade asked the prefects to check the two series of tables sent for the 1846-1850 period, a first check having already been performed for 1841-1845 (Archives nationales, F204405 and F20513).  The prefects’ responses were suspiciously reassuring: the tables comply with the instructions; there are no mistakes (Archives nationales, F20552)! Yet, the contradiction was still as patent as ever between the requirements of the administration (the application of the decree) and those of the statistical bodies (the need for an accurate count of births and deaths). In his letter dated 3 March 1853 (Archives de Paris, VD42, pièces 406-407), the prefect of Seine invited the mayors, while applying the decree of 1806, to “calculate separate totals” of stillborn children and children deceased before the declaration of birth, so as not to include the latter category of deaths with the others (circular dated 29 December 1852; Archives nationales F204407, and F20749). On 25 December 1853, the prefect of Seine sent excerpts of the circular from the Minister of Agriculture, Trade and Public Works dated 24 September that confirmed the provisions of 1852: “we shall class stillborn children separately” (Archives de Paris, VD42, pièce 408). However, he did not cite certain interesting details of the circular: “in several départements,” explained the minister, “in particular in mountainous regions and rural areas, stillborn infants are declared to the church authority – in order for baptism to take place – but not to the civil register, whence a sometimes minimal number of stillbirths”. There are copies of the circular in several mountain départements (Brière, 1983, p. 18; Malphettes, 1984, p. 31), which suggests that practice in rural areas differed greatly from that in cities, particularly Paris, where there was a system of verification of deaths in the home.
Ten years later, in a circular dated 15 June 1863, the Minister of the Interior complained about the rise in stillbirths, and highlighted another source of error: “only infants that died before, during and after delivery and whose birth has not yet been registered should be recorded as stillbirths” (Dansette, 1985, p. 21).
The inconsistency of the two administrations, which confused local actors, necessarily distorted the statistical results, and clouded comparisons of different départements and localities. Throughout the century, Paris had a higher stillbirth rate than the rest of the country. That disparity can be attributed largely to much better recording of stillbirths in the capital, owing mainly to the verification of births and deaths in the home. But another explanation was the unnaturally high percentage of “false stillbirths” compared with other départements.
III – Understanding the stillbirth statistics for Paris
The percentage of “false stillbirths”
40The presence of false stillbirths in the statistics is, of course, due to the ambiguous definition of stillbirths, originating in the decree of 1806. At the beginning of the Third Republic in Paris, a question was added to the reports drafted by certifying physicians that made it possible to distinguish between infants who had breathed and other stillborns (Table 1). According to the statistics for Paris published from 1877 onwards, it is possible to calculate that between one-fifth and one-quarter of stillbirths with a gestational age above six months were in fact infants who had breathed, in other words false stillbirths. The percentage of these false stillbirths in total stillbirths (regardless of gestational age) ranged between 13% and 21% from 1877 to 1898 (but the series is not complete).
41In fact, one “viable” stillborn child in four or five was born alive. This high percentage is comparable to those calculated by Bertillon (1876, p. 222) for Belgium (23.3% for 1851-1860 and 22.3% for 1861-1865). This rate gives an idea of the true extent of neonatal mortality.  It also testifies to the powerful anxiety still felt by parents confronted with a premature death that might deprive their child of baptism or at least of emergency baptism. In nineteenth-century Paris, church baptism was often postponed, but the strong growth of the practice of emergency baptism at home when the baby’s health was at risk shows a persistent concern to ensure its salvation (Gourdon et al., 2004). However, a child who received emergency baptism but who died before being presented alive to the civil officer was very likely to be considered stillborn (and recorded under the category of “false stillbirths”), depending on the family’s declaration and on whether or not it “genuinely” lived for a few hours or days.
42This close link between “false stillbirths” and religious practice is confirmed by the map (Map 1) of “stillborns that had breathed” in 1907-1910 (all gestational ages),  based on data from Statistique générale de la France (Rollet, 1990). This map, which highlights Brittany, the southern Massif Central and several départements of eastern France, tallies fairly closely with Boulard’s famous map of 1947 on the conformity of religious practices in rural France (Cholvy and Hilaire, 2000, p. 182 et seq.): the Catholic regions tended to declare more frequently that the child had breathed, enabling baptism or at least emergency baptism to be performed.  The percentages were as high as 25% to 30% in Brittany and in the two départements of Savoie, compared with 15% for the Seine département. The rates were only around 10% in the least religious départements (Statistique générale de la France, 1907, 1908, 1909 and 1910).
Percentage of stillborns who had breathed (1907-1910)
Percentage of stillborns who had breathed (1907-1910)
Percentage of infants who breathed among total stillbirths and among stillbirths with a gestational age above six months (city of Paris, 1877-1898)
Percentage of infants who breathed among total stillbirths and among stillbirths with a gestational age above six months (city of Paris, 1877-1898)
43The relationship between baptism and false stillbirths is complex, however, and could conceivably work in both directions. The high percentage of “false stillbirths” might also reflect a decline in religious conformity. In Paris (Gourdon, 2006) and many other French cities and regions (Cholvy and Hilaire, 2000) in the nineteenth century, there was an observed tendency to postpone baptism in a church for as long as possible for reasons of convenience, particularly in wealthy families where the clergy often gave permission for emergency baptism at home (ondoiement par permission) so as not to put the baby at risk (Gourdon et al., 2004). Such postponement might have led to an increase in false stillbirths, to an extent that is unfortunately difficult to measure, because the declaration of the birth to the civil registry might also have been postponed slightly (in order to avoid taking the child outside), thus increasing the risk that the child would die before it was civilly registered (Le Mée, 1999b, p. 74).
Last, the high percentage of “false stillbirths” is not unrelated to the implications, in terms of succession, of the concept of “viability” in the sense of the Civil Code, which, as demonstrated above, was primarily a legal concept. In order to resolve matters of inheritance, it was important to define the precise gestational age of the foetus expelled lifeless, but also that of live births. From the vague category of “presented lifeless”, it was necessary to determine which infants were born dead and which were born alive, and this implied being able to recognise signs of life and of death. Whence the importance of verifying births and deaths in the home, which was promptly and efficiently organized in Paris.
Registration of stillbirths by verification in the home
44The undeniable increase in the percentage of stillbirths in Paris from the 1830s onwards can be explained in part by the enlargement of the administration in charge of verifying deaths, followed by the introduction of verification of births in the home.
45The verification of deaths and the determination of their causes, the object of Prefect Frochot’s initiative in Year IX (1801), once again became a topical issue after 1830. Out of the same concern to prevent precipitate burials and camouflaged crimes, a highly detailed report published in 1843, whose author we were unable to identify, retraces a chronology of the development of this service, detailing its successive improvements (Annales d’hygiène publique et de médecine légale 1843, pp. 118-159). According to a report submitted by Pontonnier, chief of the first division of the prefecture of Seine, on 26 November 1836, there were 35 certifying physicians in Paris, distributed very unevenly across the arrondissements. The process was well established: the mayor sent the physician “an order, entitled examination warrant, to convey himself to the home of the deceased, to be presented with the body, to certify death, to determine the causes of death, and to address a report to him forthwith” (Annales…, 1843, p. 128). But Pontonnier wanted to go further and proposed that an inspection committee be set up (Annales…, 1843, p. 151).
46This dual coverage of the city (the certifying physicians and the verification inspectors) predictably improved the recording of deaths, both quantitatively and qualitatively, including that of stillbirths. The 1843 report contains examples of precipitate burials, which were usually of “newborn infants: the certifying physician very often issued an urgent certificate, and the funeral services collected the body to convey it to the cemetery only a few hours after the declaration at the town hall” (Annales…, 1843, p. 153). Under the new system, premature burials of infants ceased. Through inspections, the administration could systematically record “the number of hours that the child had lived, however short the length of life after birth” (Annales…, 1843, p. 155) and thus distinguish true stillbirths from false stillbirths, while also determining the age of the foetus. Indeed, the indication of the age of expelled foetuses (“the duration of the child’s intra-uterine life”) was collected in accordance with the “Instruction on the verification of deaths in the city of Paris” published in the Recueil des actes administratifs, no. 14 in 1844 (p. 11). These data were subsequently used by Bayard for his studies of foetuses (Bayard, 1846). Last, the administration introduced an investigation technique to determine whether deaths were criminal by requiring that the child be examined unwrapped (Recueil des actes administratifs, 1844, p. 10). Here again, examples are given of infanticides or serious maltreatment that resulted in infant death (Annales d’hygiène publique et de médecine légale, 1843, pp. 156-158).
47Paris and the Seine département were, in this regard, well ahead of the rest of the country. According to a report submitted to the Academy of Medicine around 1863, 25,000 out of an approximate total of 30,000 municipalities of France did not certify deaths in this way (Brière, 1983, p. 13). Furthermore, the ministerial circular of 24 December 1866 included no more than a strong recommendation to set up a service for verification of deaths in all the départements, without making it compulsory (Malphettes, 1984, Annexe X; Brière, 1983, p. 12).
48Alongside this system of reporting deaths, the certification of births in the home was introduced in Paris in the nineteenth century. The recommendation for physicians to verify births was made in the early nineteenth century by Villermé, Baudelocque, Trébuchet and others, particularly to avoid the risks associated with taking a newborn to the town hall. The idea was taken up again in the 1840s by Joseph-Napoléon Loir, who published De l’exécution de l’article 55 du Code Civil relatif à la constatation des naissances (On the implementation of article 55 of the Civil Code relative to certification of births) in 1846 (Gourdon, 2009, pp. 111-117). According to Loir, Douai and Versailles were the first French municipalities to implement such a system, in 1846 (Loir, 1846, pp. 22-23), and he was eager to convince the mayors of the 12 arrondissements of Paris to adopt his proposal. A majority of them rejected the idea, observing that declaration in the home – costly for municipal finances – would not fulfil all of its aims, since the great majority of Parisian newborns were still baptized and sent out to wet nurses shortly after birth (Archives de Paris, VD41, pièce 260).
But this initial resistance under the July Monarchy (1830-1848) did not last, and the certification of births in the home, promoted relentlessly by hygienist physicians (Monot, 1872, pp. 29-34), was soon implemented in Paris.  In the early 1860s, the system was applied only in rare cases in a few arrondissements of Paris, but expanded rapidly up to the war of 1870. Its implementation was slower in the municipalities of the arrondissements of Saint-Denis and Sceaux, which explains why the stillbirth rates there were systematically lower. An innovation was introduced on 1 January 1869 with the prefect’s approval: the presentation of infants at the town hall was no longer compulsory, and the authority permitted civil registrars to delegate a physician who would visit the home to certify the birth. In Paris, presenting newborns at the town hall soon became rare.  One birth in five in Paris was certified at a town hall in 1881, and fewer than one in twenty by 1891 (Table 2). The number of hospital births began to increase slowly, but the practice did not become widespread until after the Second World War. Shortly before 1900, most births (four out of five) were certified in the home by an appointed physician.
Distribution of births by place of certification in 1881, 1886 and 1891 in Paris (%)
Distribution of births by place of certification in 1881, 1886 and 1891 in Paris (%)
49The new system for registering births probably did not have the same impact on the recording of stillbirths in Paris as the medical verification of deaths, but it undoubtedly contributed to improving the public records of these events on the boundary between life and death.
The registration of foetuses to combat “crime”
50The last factor explaining the surge in stillbirth rates in Paris in the nineteenth century was an obsession with abortion, which would soon lead to assiduous registration of premature deliveries and foetuses of all gestational ages. In the mid-nineteenth century, the issue of abortion was a crucial concern in wide-ranging circles – political, legal, scientific, medical and religious – as indicated by the debates on therapeutic abortion at the Academy of Medicine in 1827 and 1852, and the hostile reaction of the Catholic Church over subsequent decades (Betta, 2006; Le Naour and Valenti, 2003). The new competition between medical knowledge and theological reasoning with regard to the limits of life, and the growing affirmation of state power in the domains of pregnancy and childbirth from the late eighteenth century – through the supervision and training of midwives, and even the legal obligation of post-mortem caesarean sections (Filippini, 1995) – are not the only explanation for this growing interest. There was also the social question. The conviction that a stillbirth might conceal a “criminal” act of abortion or infanticide, which already existed under the Ancien Régime (edict of King Henry II), became much stronger under the July Monarchy, in connection with industrialization, the pauperization of a section of the working classes, and urbanization. People believed or wished to believe in a “demoralization” of society, conducive to violence and crime. With respect to stillbirths, such a view diverged from that of a man like Duvillard, who, under the First Empire, interpreted the high number of “stillborn or aborted foetuses” in the département of Seine in 1815 thus: “The same probable causes of shock and fright in these women may have given rise to these premature births”.  Causes that would today be seen as psychological, were replaced by a suspicion of voluntary abortion or infanticide.
51One of the key texts for this demonstration was a presentation by Dr Bayard in 1846. Henri Bayard, a medical physician and deputy inspector of the verification of deaths for the city of Paris, was well placed to study the question of foetuses and newborns in the capital. Drawing on published statistics and his own counts, he highlighted, like others, the considerable increase in the ratio of stillbirths to live births – from 1 in 27 to 2 in 30 – (Bayard, 1846, p. 3), but he also calculated for 1845 the number of stillbirths by number of months elapsed since conception (gestational age). With no delusions as to the quality of the data, he showed that of 1,971 stillbirths recorded in 1845 (according to the definition of 1806), 882 occurred before term and 223 were non-viable. He rightly noted that prior to four months, there were practically no declarations of foetuses expelled prematurely. These “miscarriages as they are commonly called,” he wrote, “are nevertheless extremely frequent: the size of a three-month foetus,” he explained, “makes it easy to dispose of, and the families have no interest in declaring a non-viable foetus (less than six months). The more advanced the pregnancy, the more difficult it is to dispose of the foetus. At four months, foetuses measure 20 cm and weigh 250 grams. While families continue to leave them in the street or at the morgue, or discard them in cesspits, some declare premature delivery. Women cannot hide the signs of their pregnancies and any physician or midwife who witnesses a delivery is bound by law to declare it. The number of declarations therefore increases with the duration of pregnancy and it is in the seventh month that registered premature deliveries are most common”. Bayard highlights the difficulty of distinguishing infants born alive from those born dead: here again the question of false stillborns arises. The rest of Dr Bayard’s demonstration concerns exposure (abandonment in the street) of foetuses and infants delivered at term, whose number increased between 1836 and 1844, perhaps in connection with the abolition of foundling wheels in 1837 (Bayard, 1846, p. 13). This takes us straight to the question of the registration of foetuses, which increased visibly from the late 1830s in Paris.
52Knowledge of foetuses at different gestational ages became more advanced, as testified by the large number of papers on this theme published in forensic medical journals from the 1830s onwards: signs of infanticide on a foetus, relationship between volume of the foetus and age of the mother, examination of the blood and bones of the foetus. Images of foetuses in their physical reality also became more precise, liberated from theological presumption (Morel, 2009).
53The registration of foetuses, which reflected anxiety about the “crimes” of abortion and infanticide, but also signalled the growing medicalization of the civil registers, was not unconnected to a study of abortion by two physicians, Paul Lecomte and Ambroise Tardieu in 1850.
54The contribution of these two men, one a functionary at the prefecture of Paris, and the other a medical physician and assistant inspector of the certification of deaths, sheds light on the logic behind the registration of foetuses, even at early gestational ages, a practice that began in the large cities, especially Paris. The two authors demonstrate that the legislation becomes extremely complex when “birth and death merge in the case of individuals who are born dead”. They describe “a great confusion” and “a kind of arbitrary routine” in relation to stillbirths. The issue they sought to resolve was whether or not to declare stillborn foetuses when “delivery is premature”. Examining the legislation and jurisprudence, they noted the contradictory positions of two courts, with one judging the recording of stillborn foetuses on the civil register to be not compulsory (ruling of the court of appeal of Nancy of 17 September 1839), and the other recommending that the declaration and prior authorization for interment of stillborn foetuses become mandatory (ruling of the court of cassation of 2 August 1844). Explicitly concerned to prevent the “crime of abortion”, the two authors recommended compulsory reporting “of all deliveries, regardless of the outcome and the stage of gestation at which they occur” (Lecomte and Tardieu, 1850, p. 415). They nevertheless appealed for application to be flexible to avoid “vexatious” or “iniquitous” measures (with respect to burial).
55That extreme proposal led eventually to the Act of 13 May 1863 amending Article 345 of the Penal Code: “the product of all pregnancies shall be declared and represented, regardless of the degree of its development” (Annales d’hygiène publique et de médecine légale, 1897, p. 147). The strict application of the law was subsequently abandoned because, while it facilitated “repression of the crimes of abortion and infanticide”, it engendered “distress and worry in families” and could “offend public modesty”, since the actions stemming from the legislation, starting with the intervention of the funeral services, made any early pregnancy public. The legislator nevertheless remained firm about prosecuting crime. The last paragraph of the ministerial circular of 24 December 1866 on the danger of precipitate burials stipulated: “a special examination should be performed when newborn infants have died, with the purpose of ascertaining whether the death resulted from a crime” (Malphettes, 1984, Annexe X).
56In practice, the application of the new rules regarding foetuses was problematic. There was hesitation about the definition of a minimum gestational age. The prefectoral circular of 26 November 1868 considered that the decree of 1806 applied only to “products of conception that are at least four months of age”; foetuses of six weeks to four months also had to be declared, but it was sufficient to transcribe the certificate of the certifying physician onto a “special register” (Lutaud, 1896, pp. 165-166). The prefectoral circular of 15 January 1869 recalled that foetuses of less than four months only needed to be recorded on a simple police register initialled by the mayor,  but in the same year the Society of Forensic Medicine opposed the recording of foetuses below four or four-and-a-half months (depending on the source). On 20 January 1875, the prefect of Seine wrote that the above circular was incorrectly applied. The definition of an “aborted foetus” (“avorton”, a term that Duvillard had used for the years 1815 and 1816 when he equated stillborns with aborted foetuses) was not precise: the limit ranges from four to six months’ gestation, depending on the document.
57Despite these fluctuations, Bertillon senior reported in 1874 that “the habit […] of certifying and registering aborted foetuses is gaining ground, especially in the large cities,” (Rapport à l’Assemblée nationale, 1874). His son Jacques observed that municipal statistics had indicated the gestational age of stillbirths since 1866 for Paris and “we see that, from that time (and probably previously) a great number of stillbirths were declared that had only five months of intra-uterine life or even less” (Bertillon, 1907).
As Gustave Lagneau commented in 1878, “the compulsory declaration of all foetuses constituted a broader application of Article 56 of the Civil Code […] and failure to declare them would have incurred the penalty stipulated by Article 346 of the Penal Code” (Lagneau, 1878). In fact, according to Lagneau, registration was “highly incomplete” owing, he explained, to women’s desire to conceal pregnancies out of wedlock: “from 1 June 1871 to 31 December 1873, the Bulletin de la statistique municipale reported that in Paris there had been only three declarations of foetuses aged one month, six aged two months, 76 aged three months, etc.”. Other authors, such as Dr Lutaud in 1896, indicate broader motives: the declaration of miscarriages would infringe the privacy of women and families, a reticence understood by physicians and midwives who were bound by medical secrecy; furthermore, declaration would theoretically entail the intrusion of the funeral services into families, since the removal of “foetal products” of six months and below was regulated by the prefectoral circular of 26 January 1882. Such a prospect would have offended many families (Lutaud, 1896; Annales d’hygiène publique et de médecine légale, 1902, 1903).
While low in the first months after conception, these numbers nevertheless increased the totals of stillbirths. Their proportion became significant by the 1880s. It is highly likely that the introduction of home certification of births in Paris contributed to this increase in stillbirths recorded at an early gestational age. In 1893, the requirements for the statistical recording of these miscarriages changed again: from that date onwards, embryos of less than four months’ gestation were to be counted in the statistics of stillbirths for Paris.
The shape of the curve of “aborted foetuses” with a gestational age below six months under the Third Republic can be attributed largely to these administrative developments. Some 11% to 12% of “aborted foetuses” had a gestational age below six months in the 1870s, which is close to the percentage calculated from Dr Bayard’s data for 1845 (11.3%). Subsequently, the percentages increased very rapidly, doubling between 1879 and 1880 then, after a pause, surging late in the century to 45% of stillbirths in 1893 (Figure 2). Other French cities applied the same legislation: 40% of stillbirths recorded in Saint-Etienne had a gestational age below six months in 1887; while in Lyons, foetuses were recorded with information about the infant’s legal status and sex (Rollet, 1990).
Percentage of stillbirths with a gestational age below six months in Paris, 1872-1898 (per 100 stillbirths)
Percentage of stillbirths with a gestational age below six months in Paris, 1872-1898 (per 100 stillbirths)
58The limits evoked in the debate on the registration and counting of foetuses are interesting. The age of six months can be connected to the concept of “viability” in the Civil Code. The age of four months’ gestation seems to have originated in medical descriptions that, despite certain nuances, described this threshold as, in the words of Dr Ollivier in 1840, the stage when “all the parts of the foetus are highly distinct, and become increasingly pronounced” (cited in Betta, 2006, p. 118). But this age also seems to be a product of the debates about the animation of the foetus, rekindled in the second half of the nineteenth century by the heated polemics between physicians and theologians over therapeutic abortion, the admissible threshold for post-mortem caesarean section, and the recognition by the Church authorities of intra-uterine baptism. The increasingly early registration of foetuses, including before four months, inevitably brings to mind the growing pressure from proponents of the immediate animation of the embryo (at conception) who, although their position was contested by some theologians who supported the theory of delayed animation, became predominant in the Catholic church in the nineteenth century (Betta, 2006).
59Nevertheless, there can be absolutely no doubt that this new development in relation to premature stillborns reflects the importance attached by the police and judicial authorities to the verification of corpses when voluntary abortion was suspected.
60This obsession with the criminal dimension of stillbirths inspired the majority of the legal texts produced from the July Monarchy onwards. It was not until the work of Jacques Bertillon, which began in 1884 (at the International Congress of Hygiene and Demography in The Hague), that the concept was deconstructed on the basis of demographic analyses. Bertillon used historical examples, such as the famine in Finland in 1866-1867 and the siege of Paris, to demonstrate the influence of “poverty” (Bertillon, 1889, pp. 227-229; 1893, p. 548; 1896, pp. 480-484). His highly detailed analyses confirmed the peak of 1871 (Figure 1): “We observe an increase in stillbirth rates on two occasions, first among children born during the first siege of Paris (which began in September 1870) and then among children who were conceived not at the outset of the siege but between December 1870 and January/February 1871; excess mortality among newborns became considerable from the seventh month of pregnancy” (Bertillon, 1896, p, 477-484). Bertillon also studied stillbirths by duration of pregnancy and concluded that voluntary abortion had only a small influence. Similarly, by analysing the infants that had breathed according to whether they were legitimate or illegitimate enabled him to verify that, although more likely to voluntarily abandon their children and therefore to commit infanticide, unmarried mothers did not exhibit a higher percentage of “false stillbirths” than married women (Bertillon, 1893, pp. 546-551).
According to Bertillon, therefore, periods of deprivation, poverty and what today would be called stress were conducive to spontaneous abortions, both early and late. This finding agrees with the observations Duvillard had made several decades earlier about the high number of “stillbirths or aborted foetuses” in 1815 in the Seine département. But between Duvillard and Bertillon, physicians, administrators and jurists had been infused with more than half a century of discourse on the link between stillbirths and “crime”.
61This analysis of stillbirth rates in Paris in the nineteenth century has ranged far beyond demographic questions alone to delve into wide-ranging debates – such as the role of physicians in civil registration and the definition of the boundaries of life, or the criminal causes of abortion – encompassing aspects that are at once administrative, legal, political, religious and moral. As has been shown, many themes have to be addressed – be it the consistency of legislation, the definition of stillbirths, or the pressure from various segments of society (families, statisticians, jurists, physicians and the clergy) regarding this category of newborns – before we can grasp the true meaning of statistical data published at the time.
62We have seen that the legislation was highly contradictory: two texts in the same year prescribed two different practices. It is hard to comprehend the logic behind the decree of July 1806, which created much subsequent confusion: one hypothesis is that the administration of the time did not have the resources, especially at the national level, to verify whether a child had been born alive or dead, since the Civil Code of 1804 stipulated that the father should take the child to the town hall to have the birth certified. To make up for the lack of staff assigned to this task, and to prevent civil officers from overstepping the neutral role of recording declarations, it seemed preferable to allow families that so wished to obtain a certificate of acknowledgement endorsed by witnesses of the birth. The certification of deaths by physicians, followed by the certification of births in the home, which began locally towards the end of the July Monarchy and started in Paris in the late Second Empire, might have provided an opportunity to change the legislation, but this did not occur until the twentieth century.
63Uncertainty persisted in the 1830s and 1840s, with contradictory circulars from different ministries and courts that upheld conflicting opinions. The very definition of stillbirths was a subject of controversy. Approached from different levels, different traditions and different viewpoints, the issue led to divergence and discord.
64For the church and the majority of families, a stillbirth was not a statistical category but a child excluded from the sacrament of baptism that a sign of life, however feeble or doubtful, would allow to join the Christian community. For lawyers, a stillborn infant could jeopardize the “surety of families” in terms of succession. For judges, social observers or physicians looking for evidence of crime, a stillborn child was a potential victim of abortion or infanticide. But medical practitioners also represented a group of actors engaged in a process of professionalization, anxious to mark official practices with the stamp of their expertise. In the case at hand, the report of death verified by a physician, which offered a way out of the strictures of the 1806 decree by progressively adding indications of duration of life, age of the foetus, sex, place of delivery and signs of life, transfigured civil registration practices defined by legislators still largely confined within the regulatory bounds inherited from the Ancien Régime.
65The last stratum was constituted by statisticians wishing to establish clearly defined categories in order to observe demographic and social phenomena. They argued for a distinction to be made between infants that were born dead and those that died before their births were declared, so as to record the former under deaths only, and the latter under births and deaths. This viewpoint eventually prevailed, but not until the twentieth century! Initially, the statisticians succeeded in having stillbirths recorded separately from births and deaths, and the Statistique générale de la France did likewise in the 1850s by analysing retrospectively the statistics back to 1836. They also sought to distinguish between “infants that were born dead” and “infants that were born before the declaration of birth” (retrospective survey for 1841-1845 conducted by the Statistique générale de la France) and, in the 1907-1910 period, infants that had breathed from those that had not. But it was only in 1920 that a question made it possible to record whether the infant had breathed or not in all French départements, and that foetuses with a gestational age below six months were no longer counted in stillbirth statistics (Vallin, 1973, p. 34). It was not until 1993 that France applied the recommendation of the World Health Organization to record the birth of any child that shows a sign of life as a live birth.
66In Paris in the nineteenth century, there was a convergence of viewpoints that led to the recording of stillbirths at an ever earlier gestational age and ultimately regardless of that age. At the end of the century, a prefectoral circular also instructed officers to record foetuses of less than four months. The percentage of aborted foetuses at that time represented between one-third and half of all stillbirths in Paris!
It is not surprising that, under these conditions, the percentage of stillbirths in total births rose steadily to almost one in ten by the end of the century. This was a much higher rate than that observed in the countryside, since for France as a whole, stillbirths accounted for only 4% to 5% of births in the second half of the nineteenth century, a percentage that remained relatively stable until the First World War. However, Paris was by no means unique. The capital illustrated a trend observed in other French and foreign cities that had introduced civil registration and that were undoubtedly confronted with the same issues of public management. The example of Vienna is interesting in this regard, because it exhibited the same high proportions of stillbirths at the end of the nineteenth century (10 per hundred births on the eve of World War One, Statistiches Jahrbuch der Stadt Wien). 
Translated by Madeleine Grieve.
Centre Roland Mousnier, CNRS.
University, Saint-Quentin-en-Yvelines; Laboratoire Printemps, CNRS.
Correspondence: Catherine Rollet, Université de Versailles, Laboratoire Printemps, CNRS, 47 boulevard Vauban, 78047 Guyancourt cedex, France, e-mail: firstname.lastname@example.org
Citations in English are taken from a translation of the Code civil attributed to George Spence. Published by William Benning, Law Bookseller, 1827. xix, 627 pages (online version available at www.napoleon-series.org).
Cited in the letter addressed by the prefect Chabrol to the mayors of Paris in 1817, Archives de Paris, VD6619, n° 1, VD42, pièce 400.
Archives de Paris, DL1 Article 3. We are very grateful to Vincent Tuchais and Sandrine Aufray for referring us to this document.
Archives de Paris, VD4 2, pièce 393.
In the Dictionnaire des sciences médicales (t. LVII, Paris, Panckoucke, 1812, p. 412), Foderé proposes the following definition of viability in forensic medicine: “state of a newborn infant declared sufficiently strong and perfect to sustain the hope that it will live” (cited in Betta, 2006, p. 121).
BnF, N.a.f. 20588, folio 25-27, January 1806. See also Reinhard, 1950, pp. 103-120; Perrot, 1977, p. 64; Dupâquier, 1988, pp. 22-25; Thuillier, 1997.
For this reason, it seems unlikely that Duvillard himself was the initiator of the decree. Moreover, he did not have that power.
In 1841-1845, the proportion of infants who died before being declared, per 100 stillbirths, varied considerably between départements: from 13% in Calvados to 51% in Cantal.
Jacques Bertillon in 1893 remarked that these 600 to 700 “infants expelled alive from their mothers’ wombs” could benefit from Tarnier’s incubator (Bertillon, 1893, p. 544).
For this period, the SGF commissioned highly detailed reports concerning births and deaths, but also stillbirths: gestational age, place of delivery and legal status of the child, medical assistance and the answer to the question: “Did the child breathe?”.
We will not go as far as to claim, as did many nineteenth-century physicians (from Villermé to Budin), that taking a child out of the home for immediate baptism at church – a practice that continued in the most Catholic regions – was dangerous and increased the risk of neonatal mortality, and therefore the percentage of “false stillbirths” (Gourdon, 2009).
This development can be followed through the various editions of the book by Dr E. Bouchut, Hygiène de la première enfance, 1862, 1866, 1879.
We are very grateful to Sandra Bree, a doctoral student in demography, for this information.
BnF, N.a.f. 20589, Notes du tableau 1, 1815.
We have evidence of these registers for a later period in the suburbs of Paris: a “register of foetuses” was kept by the town hall of Neuilly-sur-Seine from 1914 to 1954 (Archives municipales de Neuilly, 5E2).
We would like to thank Peter Ward for providing us with these data.