Although, as we have seen, the Hippocratic doctor was in a good position to exercise control over the reproductive lives of his female patients, in a strictly medical sense he was often far from being in control of the misadventures of childbirth. Lacking the basic tools and methods of modern biomedicine—antibiotics, blood transfusions, forceps, and cesarean sections—the Greek doctor was all but helpless in the face of obstetric emergencies that today’s doctors meet successfully almost as a matter of routine. The situation as described for early modern Europe would apply to classical Greece as well:
"A whole variety of conditions, such as hemorrhage, pelvic deformity, disproportion between the sizes of the child’s head and the pelvis, severe abnormal presentations such as transverse lies, eclampsia and uterine inertia early in labour, are likely to have posed problems which were beyond the capacities of those attending the birth to alleviate.. . . Furthermore, attempts to remove a dead child, especially by the old- fashioned hooks and crotchets in general use before the eighteenth century, probably severely threatened the mother’s life".1
Records for modern developed countries before the 1930s, when maternal mortality began the steep decline associated with the introduction of antibiotics, show that 35 to 55 percent of maternal deaths were due to puerperal infection (an exception being the period of the early modern hospital when death rates in epidemics reached 80 to 90 percent), 20 percent to toxemia, 20 percent to hemorrhage, and the remainder to various causes, especially abortion.2 Yet the argument persists that childbirth for women in preindustrial societies posed relatively few risks. Thus Calvin Wells attributed women’s indisputably shorter life-span in such societies in comparison with that of men not to the risks of childbearing but to nutritional deficiencies resulting from the preferential feeding of males.3 In particular, he maintained that women in preindustrial societies, not exposed to the conditions that gave rise to rampant puerperal fever in the early modern hospital—the unwashed hands of physicians and medical students—did not suffer from that scourge. Well’s arguments seem to have gained considerable acceptance,4 and, as the preceding figures attest, there is something to be said for his judgment on puerperal fever. Nevertheless, his overall assessment of childbirth risk is far too low—and too reminiscent of the old cultural stereotype that “natives” do not suffer in the same way that their modern Western observers do.
Two sorts of evidence that one would assume would help in determining childbirth risks, skeletal remains and epitaphs, are more promising in prospect than in fact. Skeletal remains in particular would seem likely to offer direct evidence of death in childbirth in the form of fetuses lodged in pelvic bones. In fact, the skeletal evidence is disappointing. One reason for this is that attention has only recently begun to be paid to this type of evidence, and relatively few studies are available for classical Greece. The redirection of the archaeological focus to these and other mundane finds may add somewhat to our information in the future, but the outlook is still not promising for several reasons. Problems with the age- and sex-determination of skeletal remains, and especially with the use of pelvic scars to estimate numbers of births, discussed in chapter l,5 will continue to limit the usefulness of skeletal material in the study of childbirth even when more attention is paid to it in excavations. Moreover, the nature of the Greek soil makes it highly unlikely that fetuses will be found actually lodged in the pelvises of
female skeletons. In fact, there are remarkably few instances of this phenomenon from any area in antiquity: Calvin Wells reported only two cases from Egypt, one from early Saxon England, and two from medieval Europe, but none from Greece. Owsley and Bradtmiller in a study of native American burials in four well-preserved sites in South Dakota (a.d. 1600-1832) also found the phenomenon to be extremely rare, amounting to only two cases (0.9 percent of the burials), neither one of which provided evidence for cause of death.6 An interesting example of the ambiguity of such skeletal material, where it does exist, is provided by the Anglo- Saxon burial of a mother and infant reported by Wells and Hawkes in 1975: they could not determine whether the infant was placed upon the mother’s body after death or whether both died during an impeded delivery, perhaps as a result of “contraction ring,” cord round the neck—or whether both mother and child were quickly dispatched when the infant’s head emerged unpleasantly deformed as a result of placenta praevia.
Epitaphs are another form of evidence that one would think would be helpful, at least in providing demographic information. Unfortunately, Greek epitaphs seldom give age at death, and even the far more numerous Roman tombstones, which often do include this information and which might have offered at least an ancient parallel, reflect commemorative practices and economic realities more than actual death patterns. Thus widowers were more likely to commemorate spouses than were widows, who usually had fewer resources; parents were more likely to commemorate children than children parents, again, probably a matter of resources; and the very young are underrepresented while the very old are overrepresented.7 One form of funerary memorial, the pictorial representations of women in labor on tombstones, can be useful in assessing community attitudes, but it does not help in determining the type or degree of risk of childbearing, at least some of these monuments may not record childbirth deaths at all).
The childbirth cases in the Epidemics are thus left as the most promising form of evidence for assessing childbirth risks in Greece. In considering them, perhaps the most obvious question is the one upon which Wells focused: did these women suffer from childbed fever? Retrospective diagnosis is, of course, difficult. We cannot automatically use the categories of modern biomedicine to analyze and diagnose illnesses described in terms of the medical system of a premodern culture.8 Nor can we assume that modern medicine has the “right” answers, and that we can fit the symptoms described by the Greek system into modern diagnostic patterns and come up with a “correct” diagnosis. The differences in diagnostic categories are such as to make an exact, one-to-one fit unlikely. Symptoms noted and described by a doctor who interpreted illness as an imbalance of humors, who saw the uterus as capable of wandering about the body causing trouble, and who lacked thermometers, instruments for measuring blood pressure, and other modern diagnostic tools are unlikely to fit neatly into the diagnostic categories devised by doctors who attribute illness to the actions of microorganisms and describe symptoms in terms of sophisticated modern laboratory tests. Nonetheless, attempts at retrospective diagnosis continue to compel interest, and scholars search for ways to overcome the bind created by the disparities in medical systems.9
A helpful strategy in dealing with this bind is offered by the distinction made by medical anthropology between illness and disease, and by the related concept of a culturally constructed illness. The term “disease” is used to refer to a specific physiological and/or psychological malfunction—an entity defined by the list of symptoms in the Merck Manual, if you will. It is what the doctor sees and diagnoses; it is viewed as some thing that a person has, like tuberculosis, and there is an assumption that it will be the same disease in any culture or society in which it appears. On the other hand, the term “illness” is used to apply to the disease as it is experienced by a particular patient in a particular cultural and social context; it encompasses the meaning that the disease has for the patient and for those around him, and the way he or she responds in seeking treatment.10
Intrinsic to the distinction between disease and illness is the concept of illness as a cultural construct:
"A given medical system in its socio-cultural context does considerably more than name, classify, and respond to illness.... In a real sense, it structures the experience of illness and, in part, creates the form disease takes. Disease occurs as a natural process. It works upon biophysical reality and/or psychological processes, as the case may be. But the experience of illness is a cultural or symbolic reality.. . . What is perceived as illness in one culture may not be so perceived in another.. . . any disease ... is in part a cultural construct."11
In terms of these concepts, we can identify puerperal infection as a disease defined by modern Western medicine as an infection of the genital track arising after childbirth; its definitive symptom is “a temperature of 35° C or above on any two successive days after the first 24 h postpartum and other causes are not apparent.”12 Other symptoms include chills, tenderness of the uterus, headache, malaise, and changes in lochial flow (diminished, or heavy and offensive). When this disease occurred in the specific cultural environment of the early modern hospital, in which male doctors and medical students moved from autopsies to deliveries without employing effective antiseptic procedures, the result was the illness of childbed fever in epidemic form. In these epidemics, often 80 to 90 percent of the patients died.13 This illness was a true cultural construct in that a pathogen converged with a specific cultural complex—the early hospital, its program of medical training, and its doctors’ ignorance of the sources and prevention of infection— to create an extremely dangerous culture-specific epidemic illness.
The tragic results of early hospital obstetric care have left a vivid impression of childbed fever as an illness of the premodern hospital, but the disease that is puerperal infection can and does occur at any time and in any situation in which women give birth, arising either from external contamination or from normal vaginal bacteria that become pathologic in such situations as anemia, preeclampsia, prolonged labor, repeated examinations, traumatic delivery, retention of placental fragments, or postpartum hemorrhage.14 Greek women indeed did not suffer from the culture-specific illness that was childbed fever, but from this we cannot infer that they did not suffer from the disease, puerperal infection. In order to determine this, we need to consider the evidence further.
The evidence that is available consists of the case histories in the Epidemics, descriptions of postpartum complications in the gynecological treatises, the conceptual framework—the theory of disease—within which the Hippocratic cases were recorded, and the definitions and descriptions of modem medicine (since it is, after all, for people in our own culture that we want to classify the Greek experience). It will also be helpful for comparative purposes to consider modern descriptions of cases for the period before puerperal infection became a treatable illness. For this purpose, descriptions of cases of childbed fever are useful,15 for, like the postpartum complications in the Hippocratic treatises, they developed without the intervention of antibiotics; moreover, they occurred in a medical system that still employed the basic conceptual framework of Hippocratic humoral medicine.16
A considerable literature from the eighteenth and nineteenth centuries on childbed fever exists, from which we will consider only two doctors’ accounts. The first of these describes three cases that arose after home birth and were treated in the patient’s home; the other provides a general description of a physician’s experiences with epidemic childbed fever, including accounts of both home and hospital births. Both authors distinguished between two forms of puerperal fever: sporadic and epidemic.
In 1768 Thomas Denman, a physician, published Essays on the Puerperal Fever, and on Puerperal Convulsions,17 in which he described the illness and provided accounts of three cases that he had treated. In describing these cases he followed the model of the Epidemics, and he argued for a Hippocratic treatment by evacuations (cathartics and emetics), for which he gave recipes. He listed the symptoms as an initial chill followed by fever, vomiting of green or yellow bitter matter, swelling and tenderness of the abdomen (the definitive sign), great pain in the back and hips, change or suppression of lochia, nausea, diarrhea, painful and scanty urination (the kidneys being affected), and anxiety. He states that the illness usually began two to five days after delivery, but notes as an exception Hippocrates’ wife of Philinus, who became ill on the fourteenth day. He notes that the patient usually died on the eleventh day, and that those who recovered did not undergo a crisis but improved gradually. He reports that mortality was high but that the three patients whose case histories he presents survived as a result of his method of treatment. The first fell ill in 1766 three days after an uneventful delivery; she suffered chills, fever, vomiting, tenderness and swelling of the abdomen, alteration in the lochia, and delirium, but was able to stop treatment at twelve days. The second patient fell ill in 1767 three days after a safe but laborious delivery; she suffered shivering, pains, swelling, and fever, but was able to stop medication at fourteen days. The third fell ill in 1768 five days after a favorable delivery, with vomiting, pain, and fever; she was finally well after eighteen days.
In 1845, three years before Semmelweis’s discoveries about the causes of childbed fever were announced in England, William Harris published Lectures on Puerperal Fevers.18 He began his account with a history of the malady, reporting that it had been known from antiquity, calling attention to the case of the wife of Philinus and remarking (as had Denman) that it was unusual in beginning fourteen days after delivery. He then discussed etiology, attributing the illness to atmospheric conditions. But he also expressed strong suspicions that it was contagious, suggesting that it was carried by doctors, especially by their clothing, and noting that it frequently occurred when a doctor delivered a patient after participating in an autopsy or the delivery of another patient who had developed the fever. He even mentioned specific doctors who were contaminated and who had lost almost all of their patients to the fever while neighboring patients of other doctors escaped. Nevertheless, he also knew of cases in which patients escaped infection despite their doctors’ prior participation in autopsies, and he refrained from drawing definite conclusions as to contagion.
Harris presented a vivid picture of epidemic childbed fever from his own experience, noting its high mortality and more frequent occurrence in hospital deliveries. He listed as the most prominent symptoms a shivering fit followed by a hot fever and profuse sweating; severe pain in hypogastric and iliac regions; and great tenderness of the abdomen. Other common symptoms included distension of the abdomen, difficulties in urination, diarrhea, nausea and vomiting (in the last stage, black vomit), delirium, lochia diminished or badly odorous, hurried respiration, a cough (when the disease was complicated with pleuritis, the most common concomitant). He stressed, however, that the illness took many forms, that its symptoms depended upon the organs that were involved, and that the fever was sometimes missing in cases that proved fatal.
If we now turn to the Epidemics, we find descriptions of illnesses with similar symptoms (as both Denman and Harris had already noted). The gynecological treatises, however, attribute these symptoms to two illnesses rather than one: disorders of the lochial flow (flows of blood were evidence of apostasis), and inflammation of the uterus. In Diseases of Women i, chapters 35-41 deal with disorders of lochial flow and are organized in accordance (too much, too little, none, and various patterns of flow); these are identified by Grmek as puerperal fever.19 Symptoms common to these chapters include a swollen tender belly, pain in the hips and flanks, chills, fever, and headache. The writer’s frequent comment that such complications are serious and often fatal demonstrates the reality of the risk they posed for Greek women in childbirth. Nevertheless, the statement in chapter 35 that some cases occurred without fever shows that a complete correlation between the conditions described and the modern definition of puerperal infection is not possible (although Harris recognized afebrile cases in 1845). Diseases of Women 150-54 and 63 discuss inflammations of the uterus; the various chapters describe the symptoms in varying degrees of detail as an inflamed and swollen belly, fever, chills, thirst, anorexia, and headache. Again, however, fever is not seen as a necessary element in these illnesses: chapter 54 allows for both fever and lack of fever, and chapter 53 does not mention fever.
These passages in Diseases of Women i make it clear that Hippocratic doctors were quite familiar with postpartum complications whose symptoms were similar to those of childbed fever. These illnesses are diagnosable as puerperal infection in terms of modern biomedicine, although the Greek system of classification and conceptual framework differed from those of the early modern and modern doctor: the illnesses from which Greek women suffered were classified in Hippocratic terms either as disturbance of the lochial flow or as uterine inflammation. Thus one disease— puerperal infection—took the form of different illnesses in different historical conditions, its incidence, mortality, and even the course of its symptoms depending upon cultural conditions such as sanitary practices, the type of childbirth attendant, and the training methods of physicians who participated in deliveries.
Since the Hippocratic doctors observed the course of the illnesses that they reported in the case histories in terms of their own conceptual framework and recorded syndromes unmoderated by antibiotic treatment, retrospective diagnosis in particular cases is difficult. Thus, of the five cases that Grmek diagnosed as puerperal fever in Books l and in—Book i, cases 4, 5, and 11, and Book m, cases 2 and 14 (second series)—only one appears to be a clear-cut instance of puerperal infection.20 This case, reported in 14, involves the wife of Philinus, who had fever, pain in the genital area, and cessation of the lochia. The fact that her illness was recognized as puerperal fever, despite its atypically late onset,21 by doctors in the eighteenth and nineteenth centuries who were thoroughly familiar with the disease in the form of childbed fever is very persuasive.
Less convincing is 15, a case involving the wife of Epicrates, who began to have shivering two days before delivery, but whose fever is noted only after delivery. No change in the lochial flow is reported, but she had pain in the genital area. The doctor also reports that she suffered from a severe sore throat throughout the illness.22 The illness seems beyond positive identification. We noted earlier that the doctor seems to have edited this case to make it resemble i 4 by relegating the sore throat to a postscript. This suggests that he did not view the sore throat as the primary problem, and, since I 4 is recognized as puerperal fever, it gives weight to a similar diagnosis in this case, complicated by tonsillitis. Nonetheless, a malarial attack beginning as usual with chills also seems to be a possibility.
In 111, the wife of Dromeades developed a fever on the second day after giving birth; she had pain in the hypochondrium and delirium. The lochia were normal, and there is no mention of pain lower in the belly or of tenderness or swelling. The symptoms are typical of malignant tertian, or falciparum, malaria. The two cases that Grmek identifies in Book m are also more probably malaria and will be considered later. Three other cases identified as puerperal fever by Fasbender do not report changes in the lochial flow or pain in the belly.23 From the later books, v 13 may be a case of puerperal fever; it involves a woman of Larissa, who gave birth at full term to a stillborn infant with a birth defect, although her apparent rapid recovery seems atypical: the afterbirth was retained for three days, and six days after delivery she developed a fever that lasted forty-eight hours and was accompanied by pains in the belly (but did she recover after the forty-eight hours or did the doctor assume that explicit mention of her death was unnecessary?).
We can recognize elements in the Hippocratic handling of childbirth which would have exposed women to enhanced risk of postpartum infection, despite their freedom from hospital care: autogenic conditions (anemia, preeclampsia, and retention of the placenta) and the introduction of bacteria through the use of pessaries and manual interference in delivery.24 There is no evidence or reason to suggest that the resultant illness was epidemic, and, given the great difference in conditions of obstetric care between classical Greece and nineteenth-century Europe, we cannot identify it as childbed fever. Nevertheless, we cannot deny that Greek women suffered—and often died—from puerperal infections, which were diagnosed and treated according to Hippocratic categories as uterine inflammation or bad lochial flow.
Prior to the 1930s, 20 percent of modern maternal deaths were due to toxemia,25 although today preeclampsia occurs in only 5 percent of pregnant women, and 0.5 percent of these develop into eclampsia.26 Symptoms suggestive of toxemia (preeclampsia and eclampsia) do not appear frequently in the Hippocratic cases. The definitive symptoms of these conditions that could have been observed by the Greek doctor are swelling in the face or hands after the twentieth week of pregnancy through the first postpartum week and (progressing into eclampsia) convulsions.27 Although these symptoms are not the focus of the description (swelling is not mentioned and convusions are only indirectly indicated), Malinas and Gourevitch have identified the condition of uterine suffocation during pregnancy described in Diseases of Women as eclampsia;28 the symptoms noted by the Hippocratic doctor were sudden and violent suffocation, loss of voice, eyes rolled back into the head, and “everything that women suffer during uterine suffocation”— that is, chilling, hypersalivation, grinding of the teeth, and epileptic manifestations. In the Epidemics, only two cases report swelling soon after delivery,29 and one reports it in a woman pregnant for four or five months.30 None of these cases report convulsions, but convulsions are noted in four cases in which swelling is not mentioned.31
Hemorrhage, which caused 20 percent of maternal deaths prior to the 1930s,32 does not appear in the Hippocratic cases. The explanation for this is probably that it was rapidly fatal and a doctor was thus not consulted (or did not choose to try to treat it). Another omission, which underlines the selective nature of these cases, is the absence of any cases in which forcible intervention was used, as described in Excision of the Fetus, or the woman died undelivered.33
In addition to the risks presented by these specific complications of pregnancy, pregnant women are also at greater risk in a number of diseases that affect the general population. This is a result of the suppression of cell-mediated immunity during pregnancy, a condition that apparently serves to prevent the immunological rejection of the fetus.34Aristotle noted the shorter life-span of women who had several pregnancies,35 and the Hippocratic doctors were aware that certain illnesses had especially severe effects on pregnant women. Thus the author of Diseases i 3 states that in pregnant women phthisis (in most cases identifiable as tuberculosis) is usually fatal, as are pneumonia, kausos (remittant fever characterized by intense sensations of heat), phrenitis (remittant fever with pain in hypochondrium and delirium), and erysipelas of the womb (puerperal fever?). The author of the third constitution of Epidemics I notes that pregnant women suffered especially severe effects from kausos,36 and the woman in iv 25 aborted at seven months during the course of an epidemic that took the life of her husband.37
Malaria posed perhaps the most serious risk to women in childbirth. This was the result of the suppression of cell-mediated immunity in pregnancy, as Eugene Weinberg has established.38 In areas in which nearly all the population is affected, malaria is the most common cause of maternal death. The incidence of stillbirth and prematurity is also much higher among pregnant women with malaria, apparently because the parasites infest the placenta, decreasing its efficiency in supplying oxygen to the fetus.
There is evidence that malaria underwent a marked resurgence in Greece in the fifth century.39 Its presence is especially noteworthy in the Hippocratic treatises,40 where it is reflected in the concern with critical days,41 and in the important role assigned to black bile: black bile is the only humor that does not actually occur in the body under normal conditions, and the suggestion is persuasive that it was postulated on the basis of the black urine of black- water fever.42 Malaria comes to our attention in the very first pregnancy reported in the Epidemics, that of the daughter of Telebulus, who died six days after childbirth in a constitution marked by fever (kausos).43 Kausos as a medical term does not correspond with any one disease in the modern definition, although it frequently indicates malaria.44 In this case this sense is certain, however, since another patient included in the same disease constitution of kausos, Philiscus, is also the subject of the first case in Epidemics i and has been identified in a careful and detailed study by Grmek as a victim of the malarial complication blackwater fever.45 The writer comments that most women who gave birth in this constitution died, and that all the women who fell ill while pregnant miscarried.
Because the effects of malaria are frequently attributed to puerperal fever, it will be helpful to describe its symptoms in some detail. It is characterized by periodic attacks of fever caused by the reaction of the body to the presence in the blood of parasites of the genus Plasmodium carried by the Anopheles mosquito.46 In Greece in the classical period three species of Plasmodium were present, each with a characteristic reproductive cycle that determined the periodicity of the fever: P. max, the agent of benign tertian malaria (fever every third day, or at intervals of forty-eight hours); R falciparum, the agent of malignant tertian malaria (longer lasting than the benign form, with a low-grade fever usually persisting between attacks, so that the patient always feels miserable; in untreated cases, the mortality rate is high);47 and P. malariae, the agent of quartan malaria (fever every fourth day, or at intervals of seventy- two hours). Multiple infection is also possible; in such cases the cycles combine and the fever may occur daily, mock other fever patterns, or become continuous. In addition to the fever, chills and sweating occur in most forms of the disease, and secondary symptoms include headache, anorexia, nausea, abdominal pain, vomiting, enlargement of the spleen and liver, generalized swelling of the limbs, nosebleed, diarrhea, respiratory difficulties, jaundice, and pallor due to anemia. In untreated cases, if the body’s defenses are successful the attack ends; subsequent attacks add to immunity, and eventually the attacks cease. Those with immunity, however, often develop a chronic condition marked by malaise, listlessness, headaches, anorexia, fatigue, and low fever. Complications, which are often fatal, include cerebral malaria, in which the small vessels in the brain are blocked by enormous numbers of parasites, and which is most common in infants, pregnant women, and the non- immune; hyperpyrexia, in which the fever continues to rise until the patient dies; and blackwater fever (malaria hemoglobinuria), a complication of falciparum malaria whose distinguishing symptom is black-colored urine.48
The course of illness in untreated women after delivery was described in 1900 by F. H. Edmonds:
"I have seen young and healthy women pass to the last week of pregnancy in good condition, then fall off, become sallow-looking, owing to low remittent fever; during labour the temperature rises, the tongue gets thickly coated with a yellow fur, the patient becomes very restless, the pains weak and long drawn; after delivery there has usually been gushing of dark fluid blood; then an improvement for—usually— forty-eight hours, when a relapse (another paroxysm?) comes on with higher temperature, deeper jaundice, greater weakness and constipation, which, on relief being given by an enema, results in the passage of a large, black, stinking stool. After five or nine days’ alternations- - each marked by increasing weakness—the patient dies quietly, with many appearances of puerperal fever, but having had her lochia of good colour, quantity, and odour; and having had no uterine pain or tenderness. In these cases the child is frequently strong and healthy, but there is not a more dangerous condition for a woman than to be seized by a malarial remittent during her puerperium."49
Today the complications caused by malaria in pregnancy can be countered to a great extent by prophylaxis or treatment (and by the avoidance of malarial areas), but the Greeks had no effective means of either prevention or treatment. There is evidence that the Hippocratic doctors were aware of the dangers that malaria posed for pregnant women. Prognosis 20 mentions recurrent periodic fever in women after childbirth that follows the same pattern as it does in men, and a general warning about malaria in pregnancy appears in Aphorisms v 55. That the risk was recognized in lay circles as well is attested by Sophocles’ Oedipus Tyrannus 25: “the barren pangs of women, withal the Fever-god swooping down, is ravaging the city.”50
We can recognize eight probable or possible cases of malaria in pregnant women in the Epidemics. We have already discussed 111, which Grmek had diagnosed as puerperal fever. Among the other cases in i and hi that Grmek identified as puerperal fever, two are good candidates for a diagnosis of blackwater fever.51 In ill case 14 (second set), a woman in Cyzicus had a fever that began on the first day after delivery; she did not have pain or swelling in the abdomen and her urine was black, which should at least raise suspicions of blackwater fever. The second case (second set) in in, involving the woman who lay sick beside the cold water, is, however, the strongest in its suggestion of blackwater fever. The author notes that she had been feverish for a long time before the birth and that her urine was black throughout her long illness. She had pain in the hips, but not in the belly, nor is any change in the lochial flow noted. Moreover, she suffered from despondency and her mind was “melancholic,” a mental condition frequently associated with malaria in the Hippocratic treatises.52 Finally, the author notes that she was lying beside the cold water, perhaps a standing pool or a swampy place that provided a haven for malarial mosquitoes.
In addition to fevers involving black urine, other fevers may have been malarial, but in the absence of evidence of periodicity there is no clear indication.53 We can, however, at least rule out puerperal infection when the illness preceded delivery or abortion. Thus in 113, a woman three months pregnant was afflicted with a fever, pain in the hypochondrium, and delirium—symptoms characteristic of pernicious tertian malaria. She recovered, and there is no indication that she aborted (malaria is most dangerous in the later stages of pregnancy). Again in iv 6 a woman aborted on the sixth day of an attack marked by nausea, chills, and sweating that came to a crisis on the fourteenth day; no fever is noted, but the brevity of the account, and the likelihood that chills and sweating imply fever, make it tempting to assume its presence. The case in 15, in which shivering began before birth and fever after, and in which the lochia were normal, was discussed earlier; its parallelism with i 4 may indicate puerperal infection. In other cases, a lack of specific symptoms pointing to puerperal infection (changes in lochial flow, pain in belly) may suggest malaria. For instance, in in case 12 (first set), fever arose after birth and shivering was pronounced throughout; no changes in the lochial flow are reported, and only slight pain in the kardia early in the illness.54 Similarly the patient in in 11 (first set) suffered a fever after abortion; she had pain in the hip, but there is no indication of change in the lochia or pain or tenderness in the belly.
The role of malaria in pregnancy and childbirth in classical Greece has been all but ignored by modern scholars. Grmek does not even consider the subject when he discusses malaria, identifying, as we have seen, most cases of fever in the postpartum period as puerperal infection, even those in which the fever preceded delivery or in which black urine was a prominent symptom. Jones, however, writing in 1909, provided the most striking illustration of the way in which cultural values can shape perception. He realized full well the dangers the disease posed for pregnant women, for, in commenting on the passage about the Fever-god in the Oedipus Ty- rannus, he said that the reference to childbirth was strongly suggestive of malaria, and he cited the article by Edmonds that contains the case quoted earlier.55 Nevertheless, in his chapter on the effects of malaria on the general population he did not mention women; he even omitted women in discussing the reduction of population due to malaria and the effects of the disease on the newborn. In an appendix, however, he suggested that the role of women as nurses during malarial attacks increased Greek men’s appreciation of their wives. Thus for Jones the noteworthy effect of malaria on Greek women was not miscarriage or death, but their service to men, which provided “a new ideal of womanhood.”56
Phthisis is one of the most frequently used terms in the Hippocratic Corpus.57 In most cases it can be identified as tuberculosis, although the wasting that defined the condition for the Greeks occurred in other illnesses as well.58 It was recognized as especially dangerous in pregnancy: the author of Airs, Waters, Places says that it was frequent after childbirth because of the violence suffered by women.59 The wife of Simus (v 103), who died six months after being shaken in childbirth (succussion), provides a good, if rather extreme, example. Her symptoms were typical of tuberculosis: fever, wasting, pain in the chest, pus-filled sputum, with diarrhea at the end. The use of succussion was (as far as we can tell) not typical of normal deliveries, being used mainly in emergencies involving the death of the fetus.60 As Grmek points out,61 the succussion did not cause the consumption, but pregnancy and the stressed birth aggravated the patient’s latent tuberculosis.
While we will never have exact statistics, the evidence of the Hippocratic treatises leads us to conclude that Greek women who were attended by Hippocratic doctors, or by midwives following Hippocratic methods, faced all the risks of childbirth inherent in a premodern culture that lacks the resources of modern medicine. In addition, some of the practices of Hippocratic medicine—for example, the use of cathartics, succussion, and manual interference in delivery—were themselves the cause of (or contributed to) complications that sometimes proved fatal. There is ample evidence that women did in fact suffer from puerperal infection. Moreover, tuberculosis and malaria posed special risks for them, again augmented by the lack of modern prophylactics and treatment. There was every reason for Greek women facing pregnancy and childbirth to seize upon whatever promise of help that their culture and religion might offer, in addition to seeking the most up-to-date medical care.
Notes
1. Schofield 1986, 235.
2. Loudon 1991, 34.
3. Wells 1975.
4. See, e.g., Russell 1983, and Gdis 1991, 229. The evaluation of risk and the way in which statistics are presented lends itself to a degree of subjectivity. Hopkins (1987, 118, n. 10), in discussing deaths in the Roman Empire, says that maternal deaths in childbirth in preindustrial societies “may have been grossly exaggerated,” citing the evidence of Gutierrez and Houdaille 1983 and Schofield 1986 for eighteenth- century England and France. It is true that these studies revealed a lower death rate than their authors had expected, averaging about 100-110 per 10,000 births, but the figures for England are 100 times the death rate in 1980 and for France 70 times the rate in 1975-77, and the authors consider these rates to be very high. Whether we can apply the eighteenth-century figures to antiquity is also a question, in both England and France, the average age of marriage hovered around 25, and in England 12 percent of the women never married, while in classical Greece virtually all women married and most marriages occurred before the age of 20, most frequently around 15. Since early marriage exposes women to higher-risk early childbirth and to more occasions of childbirth in their childbearing career, it involves an overall higher risk (Gutierrez and lloudaille attribute the decline in death rate over the 130 years of their study to a rise in the average age of marriage from 24 to 26 [pp. 986-87]). Moreover, these figures were for rural areas, and there were large regional differences in death rates in the French study, reaching a high of 291 per 10,000 in Mogneneins; if the average figure of 100-110 did not apply to Mogneneins in the eighteenth century, one cannot assume that it would apply to Greece in the fifth century b.c.
5. See chapter 1.
6. Wells 1975,1237; Owsley and Bradtmiller 1983; 1 have been unable to find the source of the photograph published by Angeletti 1990, of an apparent eight-month fetus in breech position; she suggests that the situation was compatible with preeclampsia, which she believes to have been the cause of the unfavorable Hippocratic prognosis for eight-month infants; however, there is no information in the references she gives about the site where these remains were found (Greece?), or their date.
7. Hopkins 1966, 1987.
8. See, e.g., Grmek 1975.
9. See esp. Grmek 1989.
10. Eisenberg 1977; Kleinman 1980, 72-80; Helman 1984, 65-73.
11. Quotation from Kleinman 1973a, 208-9; see also Freidson 1975, part 2.
12. Merck Manual, s.v. “puerperal infection.”
13. On childbed fever, see Semmelweis 1983 [I860]; Penniston 1986; Gelis 1991, 249-52.
14. See Seligman 1991; an illustration of non-hospital-associated puerperal fever is provided by the case of Mary Wollstonecraft, who died of puerperal fever after a delivery that took place at home under the attendance of a midwife; the placenta was retained and extracted— apparently only partially and with contamination resulting—by a physician called to the home. See Godwin 1990 [1798].
15. This will not involve aspects of the epidemiology of childbed fever, which are not relevant to the disease in classical Greece.
16. Brief histories of puerperal fever before the introduction of antibiotics are provided by Peckham 1935, and G£iis 1991, 249-52.
17. Denman 1768.
18. W. Harris 1845; Semmelweis’s discoveries were announced in England in a lecture by C. H. F. Routh in 1848 (Semmelweis 1983 [I860], 174-75); Alexander Gordon in 1795 had preceded him by about half a century with his statement that puerperal fever is contagious and carried by the hands of doctors and medical attendants, but Harris was not entirely convinced. On the distinction that Semmelweis saw between his views and those held in Britain, see Semmelweis 1983 [I860], 133, 136-50,176 and 176 n. 12.
19. Grmek 1989, 391, n. 72.
20. Grmek 1989, 132; he omits three cases included by Fasbender 1897, 189-93: in 10, 11, and 12 (first series), and includes in 2 (second series), labeled as doubtful by Fasbender.
21. Fasbender (1897,189, n. 2) comments that this was remarkably late— if the earliest stage was not overlooked; however, in England and Wales in 1976, a rise in temperature to 38° C within fourteen days after birth was notifiable as puerperal infection; in Scotland the time period was twenty-one days, Walker, MacGillivray, and Macnaughton 1976,404.
22. Tonsillitis as a common complication of the postpartum period is noted by Reynolds and Newell 1902, 516.
23. Fasbender 1897,189-93; the cases are in 10, 11, and 12 (first series; pain is noted without site specified in in 10, in the hip in iii 11, and in the hardia in m 12.
24. On interference in live delivery, Super jet. 4 (8.478.4-16 Li.) and 15 (8.484.10 Li.); following the death of the fetus, Super jet. 5, 6, 7 (8.478.17-480.12 Li.); Dis.Wom. i 70 (8.146.21-148.2 Li.); Excision of the Fetus 1 (8.512.1-514.3 Li.); on the use of pessaries: Superjet. 36 (8.506.11-13 Li.) and 42 (8.508.7-8 Li.); Dis.Wom. i 78 (8.172.14- 198.2 Li.), with many recipes. Rousselle (1980,1097-98) argues that the use of garlic, a known antiseptic, in preparations, as in the fumigations of Dis.Wom. II133 (8.286.9 Li.) and Ster. 230 (8.440.2 Li.), effectively negated risk; even if this were so, it takes care of only a small proportion of the cases of internal application (forty-four instances in the gynecological books), which often featured dreck as a primary ingredient. Loudon (1991, 72) identifies the main determinant of maternal mortality as the care given by the birth attendant, including dangerous and unwarranted interference.
25. Loudon 1991, 34.
26. Ibid.
27. Merck Manual, s.v. “preeclampsia.”
28. Malinas and Gourevitch 1982, discussing Dis.Wom. l 32 (8.76.1-22 Li.) suffocation in pregnancy; see also Gourevitch 1984, 160—62; Angeletti argues that the difficulties attributed to an eighth-month delivery are to be attributed to “a mild toxemic disease” (Angeletti 1990, 89).
29. Epid. ii 2.20 (but of the belly), and v 11.
30. Epid. vii 6.
31. Epid. i cases 4,11,13; in 14 (second set).
32. Loudon 1991, 34.
33. See Gelis 1991, 226-36, for the situation in early modern Europe.
34. See Weinberg 1984, 1987; among the diseases whose effects are enhanced are malaria, tuberculosis, influenza, polio; Sallares (1991, 131-32) is probably right to attribute the excess female mortality between the ages of fifteen and forty-five in modern preindustrial populations, as well as in ancient Greece, to this weakening of immunity in pregnancy.
35. Arist. HA 582a21-24; 583b26-28; GA 775al3-16; Peri makrobiotos 466bl5-16; 467a31-32.
36. Epid. i 8 (2.646.9-13 Li.).
37. Littr£ suggested that this case belonged to the Cough of Perinthus (vi 7.1 [5.330.11-336.11 Li.]), which, according to Grmek 1980, 1989, 305-28, involved diphtheria and probably influenza, whooping cough, night blindness, and bronchopneumonia.
38. Weinberg 1984,814,818; see also Weinberg 1987; Diro 1982; Watkin- son and Rushton 1983. Rutman and Rutman (1976) employ the hypothesis of malaria to explain the high death rate for women in the childbearing years in the seventeenth century in the Chesapeake and in Salem, Massachusetts.
39. Although W. H. S. Jones (1907 and 1909) argued that malaria was a new disease in Greece in the fifth century, most scholars today agree that it was present earlier, although it underwent a resurgence at this time, possibly as a result of the entry of large numbers of infected individuals into Greece with the Persian army; see Grmek 1989, 275-83; Borza 1979, and contra, Hammond 1984, answered by Borza, 1987; Borza’s discussions are especially relevant because they deal with the situation in northern Greece, the area in which the doctors of Epidemics were practicing
40. E.g., Epid. i 10-11 (2.670.16-672.3 Li.); Aph. n 25 (4.478.1-3 Li.); Dis. ii 40 (7.56.3-4 Li.), 42 (7.58.22-23 Li.), 43 (7.60.6 Li.); Progn. 20 (2.168.6-14 Li.); Affect. 18 (6.226.5 Li.); fora full discussion of all the relevant passages, see W. H. S. Jones 1909, 61—73.
41. Another empirical source for the number lore, especially as it appears in the gynecological treatises, might have been experience with pregnancy: even today we distinguish between the first, second, and third trimesters. Numbers such as this serve as handy general mnemonic devices for recalling the stages of pregnancy or calculating delivery date, as in Epid. n 3.17 (5.116.12-13 Li.): “That which moves in seventy days is finished in thrice that.” Nevertheless, in some cases the concern with numbers seems to have its source in number magic. For example, the number seven had a special significance, as in the treatise Sevens and the widely held belief that the seven-months’ child survives while the eight-months’ child does not; see Hipp. Seven Months’ Child, Eight Months’ Child, and Hanson 1987.
42. Timken-Zinkann 1968; for an alternative theory, see Langholf 1990, 47-50.
43. Hipp. Epid. i 8 (2.646.12 Li.).
44. On the meaning of kausos, see Grmek 1989, 289-92.
45. Epid. l case 1 (2.682.4-684.9 Li.); Grmek 1989, 295-304.
46. The modern medical literature on malaria is abundant; useful brief discussions may be found in the Merck Manual and in Havard 1990, s.v. “malaria”; for a more extensive discussion, with bibliography, see Manson-Bahr and Apted 1982.
47. Zulueta (1973) has argued against the presence of falciparum malaria in classical Greece on the grounds that the species of mosquito most important for its transmission in modern times resists infection with some types of the parasite and that time was required for selection to overcome this; see too Bruce-Chwatt and Zulueta 1980, 14-19. But see Grmek 1989, 278-81, which reviews the evidence for the presence of falciparum.
48. On blackwater fever, see Manson-Bahr and Apted, 1982; on blackwa- ter fever in the Hippocratic writings, see Grmek 1989, 284-304; kausos with black urine, Epid. I 9 (2.652.8-9 Li.).
49. Edmonds 1900, 260 (emphasis added).
50. Progn. 20 (2.172.3-4 Li.); Aph. v 55 (4.552.6-8 Li.); Soph. Oed.Tyr. 25, trans. in W. H. S. Jones 1909, 33.
51. Foy and Kondi (1935) and Grmek (1989) discussed possible cases of blackwater fever in the Epidemics, but confined themselves to males; the apparent assumption was that women in the postpartum period could only suffer from puerperal infections.
52. Hipp. blat.Man 15 (6.68.8-9 Li.); on melancholia, see W. H. S. Jones 1907,45-47; 1909,98-101.
53. Other possibilities include typhoid; Grmek so diagnosed the wife of Hemoptolemos in vn 11 (5.382.13-386.22 Li.) from the characteristic color of the stools (toroKifipa), explaining the absence of reference to a rash by suggesting that the light rash of typhoid would not have been seen as significant by the doctor (Grmek 1989, 346-50). Scurlock identifies the kausos in which the daughter of Telebulus died (i 8 12.646.12 Li.]) as typhoid, apparently because of the emphasis on nosebleeds, but this was the epidemic in which Philiscus died of blackwater fever according to Grmek (Scurlock 1991, 57, n. 223).
54. Corvisier (1985, 110, n. 23) diagnoses it as puerperal fever.
55. W. H. S. Jones 1909,43 and 43, n. 1; he refers to Edmonds 1900; most of the footnotes are devoted to animals.
56. W. H. S. Jones 1909,123-26.
57. E.g., Dis. n 48-49 (7.72.6-76.13 Li.); Int.Ajf. 10-12 (7.188.26- 192.19 Li.); Epid. iv 24 (5.172.1-5 Li.); v 103 (5.258.9-12 Li.) = vu 49 (5.418.1-4 Li.), the wife of Simus after childbirth; vu 50 (5.418.5- 17 Li.), 51 (5.418.18-420.10 Li.); for a collection of the references, see Meinecke 1927.
58. According to the modern definition, tuberculosis is “a chronic granulomatous infectious disease caused by bacteria of the genus Mycobacterium" \ Manchester 1984,162. The Greeks, of course, were not aware of the bacterial agent responsible and therefore did not recognize the various forms of tuberculosis as the “same” disease, as, for example, tuberculosis of the spine, or Pott’s disease, characterized by a sharply angular curvature of the spine that can sometimes be identified in ancient representations.
59. Hipp. AWP 4 (2.22 Li.); on modern recognition of this, see Youmans 1979,203-5.
60. See chapter 3, n. 35.
61. Grmek 1989, 191.
Written by Nancy Demand in "Birth, Death and Motherhood in Classical Greece", The Johns Hopkins University Press, USA, 1994, excerpts chapter 4 pp-71-86. Digitized, adapted and illustrated to be posted by Leopoldo Costa.