I INTRODUCTION

The death of Gaius in 4 c.e. was a double blow for Augustus, who lost not only his second grandson, but also his sole remaining heir. Through the example of Asinius Pollio, however, Seneca reminds readers that even in the face of deep grief, life can and should continue. The anecdote invites reflection on Roman responses to death. A long-standing tradition has established that a fear of death pollution dictated the rites of the funeral while placing a range of restrictions on mourners. Nevertheless, the precise nature of the fear remains unclear. To some modern commentators, death pollution was a contagious force, a miasma that leached its way out of the corpse and through mourners to threaten society as a whole.Footnote ³ Others reconstruct a less tangible danger, by which the dead made the living metaphysically unclean, offensive both to their own community and to the gods.Footnote ⁴ According to some, fears of death pollution had faded by the end of the republican period, even while practical concerns over corpses as vectors for disease continued to fuel a discomfort with the dead; following others, the religious anxiety retained its force well into the age of the emperors, disappearing only with the rise of Christianity in Late Antiquity.Footnote ⁵ Some argue that pollution was a threat only prior to the completion of funerary rites, while others contend that tombs remained objects of fear and revulsion.Footnote ⁶ Given these considerable disagreements over the nature of death pollution, past work finds concord only on its effects, agreeing that a fear of the dead is unmistakable in Roman funerary practices.

This article attempts to unravel some of the confusion that has surrounded the concept of Roman death pollution. My focus is the literary evidence derived from the city of Rome in the late republican and early imperial periods, generally corresponding to the final century b.c.e. (especially its second half) and the first two centuries c.e. Not only do the data concentrate in those eras and that city, but also earlier work most often grapples with the same material, time and place, whether explicitly or otherwise. A careful re-evaluation of the texts dealing with Roman funerals reveals that we know both more and less than past studies have assumed; we can trace specific contours in responses to the dead even while admitting the many gaps in our understanding of what Romans of these periods did following a death, as well as why they did so. Rereading this material, I argue that the temporary changes to life that resulted from death are better attributed to obligation than pollution. Behaviours were shaped not by fears of metaphysical uncleanliness, but by a variable blend of internal reactions and external expectations, dependent on factors such as gender and social status as well as on personal negotiation of the emotions aroused by mourning, grief above all.

II THE SYNCHRONIC FUNERAL

Regarding funerary rites, a substantial distance separates what we might think we know from what we actually can know of the residents of late republican and early imperial Rome. The modern reconstruction of the ‘Roman funeral’ originated in the late nineteenth and very early twentieth centuries, when a series of dictionaries published in various languages collected, organised and disseminated evidence for the ancient world.Footnote ⁷ Among many areas of focus, such dictionaries sought to reconstruct funerary ritual from the moment of death, through the funeral itself, to post-interment rites. The method was to compile relevant literary references, sometimes supported by iconographic and archaeological data, drawing them together into a narrative of so-called typical practices. The approach had value: it compiled the essential ‘sources’ while also legitimising the topic as an area of study. Less constructively, though, such works assumed that brief references and offhand mentions directly reflected ancient reality, with no need to question intentions of authors or the demands of genre. Likewise, the dictionaries ignored potential diversity in practice introduced by chronological, geographic or social variation, and relied heavily on uncritical readings of antiquarian texts, which often described activities of the vaguely distant past, as well as on scholiasts who dated centuries after the period under study. Ultimately, these early efforts presented a seemingly coherent funeral that derived from references scattered across time, space and genre. Despite its apparent solidity, this synchronic funeral might be a cipher, an amalgam that never existed and would have had little meaning to any given resident of the capital, much less to an individual living within the broader administrative purview of the Empire.

Although various scholars have attempted to point out its problems, the synchronic, nineteenth-century version of a standardised funeral continues to underpin work on Roman death.Footnote ⁸ The typical reconstruction — unsurprisingly concerned with elite men in the capital, although often taken as broadly applicable for others and elsewhere — goes something like this: the ritual began at the moment of death, upon which the family became funestus or funestatus, ‘polluted by death’.Footnote ⁹ Dressed in mourning clothes to mark their uncleanliness, family members were responsible for completing the following rites: they hung boughs at the door of the house to warn of pollution within, closed the eyes of the deceased, called out his name with a conclamatio and placed his body on the floor to mark his separation from life. They then washed the body, dressed it in a toga and displayed it on a couch in the atrium, with its feet facing towards the door. Following a period of lying in state, the funeral procession made its way to the grave. Included in the procession were the family and friends of the deceased, musicians, hired mourners and actors wearing imagines, wax ancestor masks. Torches and incense helped to contain the pollution of the cortège, while distinctive music cautioned outsiders to stay away. For notables, the procession stopped in the forum for a eulogy; others received graveside eulogies. Prior to cremation, earth was thrown on the body, a bone was removed for the ritual of os resectum and the eyes were opened. Gifts of food and incense were placed on the pyre, together with a coin in the mouth of the deceased so that he might pay the ferryman Charon to enter the underworld. After cremation, the pyre was drenched with wine or milk, and women of the family collected the bones and placed them in an urn. The urn was either buried in the ground or deposited in a tomb chamber; the sacrifice of a pig made the grave sacred. After sharing a meal that consisted of a special type of sausage (silicernium), mourners purified themselves through the os resectum and a suffitio of fire and water, then purified the house of the deceased with a sweeping ceremony (alternatively, this act could have occurred immediately following the departure of the procession). Nine days after the interment, the nouendialis (or nouemdialis) was celebrated with another meal eaten at the tomb. This ceremony ended the feriae denicales — the primary period of mourning — and removed the status of funestus from the family; the os resectum and suffitio might have been performed only at this point. Afterwards, family members returned to the tomb to leave offerings and eat communal meals during public festivals for the dead, as well as on days with private significance, such as the birthday of the deceased or the anniversary of his death, funeral or nouendialis.

Readers might perhaps have noticed that the preceding description includes no citations of ancient texts. The omission is deliberate; although scholarship recounting the synchronic funeral typically cites individual points, the treatment is too often uncritical and avoids reconsidering each element's validity within the narrative as a whole. The case of praeficae, hired mourners, provides a good example. Despite the considerable attention they receive in modern reconstructions, such mourners are poorly attested for the late republican and imperial periods.Footnote ¹⁰ Most citations point to Varro (Ling. 7.70), but that reference glosses a passage of Plautus, explaining that praeficae were women hired to sing the praises of the deceased outside their homes and citing as evidence Aristotle, along with early Latin authors Gnaeus Naevius and Appius Claudius Caecus. Apparently, Varro expected his readers to be unfamiliar with such women, who had played a role in earlier funerary practice. Another reference appears in a fragment of the second-century b.c.e. satirist Gaius Lucilius, further indicating that hired mourners were an element of mid-republican funerals (Lucil. 995–6). In fact, the only mentions of praeficae from the imperial period come from Festus (s.v. praeficae), who appears to draw directly on Varro, and from Aulus Gellius’ Noctes Atticae, where praeficae are invoked to represent useless wailing in a conflict between the orator Favorinus and the grammarian Domitius ‘the Insane’ (Gell., NA 18.7.1–3).Footnote ¹¹ Given the literary and antiquarian predilections of all involved (not least the author himself), the passing remark cannot be taken as good evidence for the employment of praeficae in Gellius’ day.Footnote ¹²

By the early twenty-first century, newly incorporated theoretical perspectives, above all those derived from the much earlier work of Robert Hertz and Arnold van Gennep, had secured death pollution's role as the primary force that dictated the events of the Roman funeral. Hertz focuses on the practice of secondary burial, describing several Indonesian tribes whose members (according to Hertz) believed that the deceased could not join the ancestors until his body had decomposed fully.Footnote ¹³ Stuck between two worlds, the unsettled spirit was malicious and repulsive, emitting a cloud of contagious pollution that required relatives — even distant relatives who had not been in contact with the corpse — to separate themselves from society. Close family members went to extreme lengths of isolation: they stayed at home, adopted special dress to warn others to avoid them, and even went so far as to refuse to answer questions for fear that their speech might spread pollution. Widows and widowers were most affected and subject to the greatest number of taboos. The family remained a source of dangerous pollution until full decomposition had been achieved, at which point the dry bones were buried in a second funeral and a series of elaborate ceremonies restored purity. Hertz's key interest is to show the interrelationship between the physical state of the corpse, emotions that arise from mourning, beliefs regarding the condition of the soul and social responses to death. His ideas find resonances in van Gennep's work, which establishes rites of passage as fundamental elements of human experience.Footnote ¹⁴ Following van Gennep, all such ‘rites’ — births, deaths, marriages, initiations etc. — include three stages: separation of the subject from the original state, liminality as the subject passes from one state to the other and reintegration of the subject into the new state. In the case of death, deceased and mourners undergo parallel (although not always strictly synchronised) paths through the rites; as the dead move from life to afterlife, those who mourn them navigate the transition from life with the deceased to a new existence without them.

Hertz and van Gennep both worked within an academic climate that sought to understand universal systems underpinning human actions, that is, to explain humanity's socio-cultural ‘evolution’.Footnote ¹⁵ According to ideas common at the time, certain concepts were shared across earlier human cultures as well as in contemporary ‘primitive’ groups. A fear of death pollution was thought to be one such concept, forming a standard step in the development of human reactions to death.Footnote ¹⁶ Notably, however, neither Hertz nor van Gennep advocate universal models of death pollution. Hertz presents the Indonesian tribes as an extreme illustration of his ideas on the interrelated forces that determine attitudes towards death and the dead. As for van Gennep, his work is relevant to all life-course transitions, events that might be greeted in diverse ways by various cultures, and regarding death pollution he is explicit: whereas once he had viewed all mourning behaviours as expressing taboos related to pollution, his theory of death as a rite of passage is more nuanced, allowing for flexibility in the attitudes that guided mourning.Footnote ¹⁷ Both authors demonstrate that mortuary practices reflect belief systems, but neither presents the fear of pollution as a necessary response to death.

Today, self-consciously theoretical approaches to Roman death pollution begin with Hertz and van Gennep; other key studies appear strongly influenced by both authors, but do not reference either specifically.Footnote ¹⁸ Such work, however, tends to treat their theories as universalising, in some cases to the point of citing them as evidence — rather than means of structuring evidence — for Roman fears of death pollution.Footnote ¹⁹ Certain elements of the synchronic funeral, such as the adoption of mourning clothes or the sharing of a meal at the tomb, appear comparable to practices Hertz documents in Indonesia, while the funeral's course as a whole can be mapped onto van Gennep's tripartite rite of passage, with practices like the conclamatio or the placement of the body on the floor as rites of separation, the cremation and other rituals performed at the tomb itself as rites of transition and post-funerary observances like the os resectum, the suffitio and the nouendialis as rites of incorporation. Current theoretical reconstructions amalgamate parts of both works — sometimes explicitly bolstered by Mary Douglas’ structuralist definition of pollution as ‘matter out of place’ — to argue that Roman corpses and mourners, being liminal objects in states of transition, were polluted by death.Footnote ²⁰ Even looking beyond the misapplication of Hertz and van Gennep, these ideas remain grounded in the nineteenth-century narrative of funerary practices, which has retained its primacy despite subsequent shifts in perspective that have revolutionised approaches to Roman religion.Footnote ²¹ If the synchronic funeral is an amalgam, then it provides an extremely rickety foundation on which to build an understanding of ancient beliefs.

III FUNERARY RITES

Significantly, the only explicit record of a Roman fear of death pollution comes from Servius’ commentary on the Aeneid, written in the late fourth or early fifth century c.e.Footnote ²² In reference to the Trojan matrons who held cypress branches at the improvised funeral of Polydorus, Servius says that Romans hung boughs of cypress outside their homes during the lying-in-state to warn priests of death pollution within (Ad Aen. 3.64, ‘Moris autem Romani fuerat ramum cupressi ante domum funestam poni, ne quisquam pontifex per ignorantiam pollueretur ingressus’).Footnote ²³ According to Pliny the Elder, the cypress tree was sacred to Dis, god of the underworld, and so displayed on houses as signs of mourning (Plin., HN 16.60, ‘Diti sacra et ideo funebri signo ad domos posita’), while Horace and Ovid both characterised cypresses as sad or funereal (Hor., Carm. 2.14, Epod. 5.18; Ov., Tr. 3.13.21).Footnote ²⁴ No text prior to Servius, however, connects the trees with pollution. The association might have been natural enough that earlier authors had no need to discuss it, but with a late antique commentary as our only source, that assumption is far from given.

Beyond Servius’ explanation of the boughs, a passage of Festus is among the most cited evidence for a Roman fear of death pollution. According to Festus (s.v. euerriator):

Many have taken this passage literally, reconstructing a funerary rite that involved sweeping the house with a special broom to remove the stain of death pollution.Footnote ²⁵ Nevertheless, if the sweeping ceremony were still common when Verrius, Festus’ source, was writing in the Augustan period, why would he specify that the term euerriator came from the word for ‘sweeping’, or go on to define the exuerrae?Footnote ²⁶ Separating the passage from the assumption that Romans feared death pollution, it seems best interpreted as an etymology for euerriator, an archaic name for the chief heir explained in terms of a long-outdated sweeping ritual. Indeed, the word euerriator appears nowhere else in any ancient text, not even within the extensive legal discussions of heirship, nor does any other author mention sweeping the house as a funerary rite. Perhaps the ancient ceremony indicates an early fear of death pollution that had faded later, but the passage introduces an additional problem: when the sweeping was performed. According to the synchronic narrative, it followed the carrying-out of the body and therefore purified the house of death's polluting presence. The use of the future passive participle (ferendus est), however, suggests that the sweeping preceded the carrying-out, possibly constituting part of the formal initiation of rites by the heir. If this were the case, the ceremony is unlikely to have removed pollution introduced by the corpse.Footnote ²⁷

Other elements often associated with funerals in the literary sources — such as torches, incense and distinctive music — seem inherently tied to ritual purity, and for that reason have been used as evidence that the dead spread pollution.Footnote ²⁸ Purifications, however, were standard elements of Roman religious activity, a category to which the funeral clearly belongs.Footnote ²⁹ In no other case did the purifications included in a Roman rite indicate that the proceedings themselves were sources of pollution.Footnote ³⁰ Even beyond their purifying functions, moreover, other concerns could draw such elements to funerals. Most notably, torches, incense and music marked the boundaries of the ritual space surrounding the corpse.Footnote ³¹ Furthermore, many of the ancient texts present incense as a means of displaying wealth and status, and the Twelve Tables had banned it from funerals entirely (Cic., Leg. 2.24.59–60). Epitaphs advertising quantities of incense donated for the funerals of prominent individuals stress its role as a luxury object, as do criticisms of its use, such as Pliny's complaint that the gods were happier receiving salted spelt than rich incense (HN 12.83).Footnote ³² The role of music, too, is less straightforward than we might assume. Some have seen the noise of processions as warning others to stay away and avoid death pollution; a law from Puteoli that required funerary workers to ring a bell while transporting the bodies of torture or execution victims has been interpreted along the same lines.Footnote ³³ Still, music could alert bystanders to the presence of the funeral, encouraging them to notice or even join the procession, a possible function also for the executioner's bell.Footnote ³⁴ In 17 b.c.e., horns were used in precisely this way to advertise Augustus’ Ludi Saeculares and attract participants.Footnote ³⁵ Propertius suggested that a crowded pompa was desired by many; his rejection of such a procession would have no meaning if it were not a common element of contemporary funerals (Prop. 2.13.17–24).Footnote ³⁶ The processions of great men could be extensive, and Pliny even recalled a funeral for a beloved talking bird that consisted of ‘countless’ mourners (HN 10.121, innumeris exequiis).Footnote ³⁷

The suffitio and os resectum, rites reconstructed in the synchronic narrative as purifications of the family, present no less complexity. The ceremony of the suffitio followed the interment and involved mourners stepping over fire and being sprinkled with water; as a result, it often is presented as the essential act that cleansed the family of death pollution.Footnote ³⁸ The rite, however, appears only in Festus, where it is described in the past tense, ‘and so those who had escorted a funeral procession, on returning stepped over fire after having been sprinkled with water; which type of purification they called suffitio’ (s.v. aqua et igni, ‘itaque funus prosecuti redeuntes ignem supergradiebantur aqua aspersi; quod purgationis genus uocabant suffitionem’). No other author refers to any aspect of this rite, nor can any casual references to funerary activity be interpreted as relating to it. As for the exuerrae, its presence in Festus suggests that the suffitio might have been considered ancient already in the Augustan period. Moreover, we cannot be certain that the purification of the suffitio was directed at death pollution, and fire and water might have played additional symbolic roles. When discussing a Roman marriage ritual of the bride touching fire and water, Plutarch lists four possible explanations for the practice, and purity factors into only one of those (Plut., Quaest. Rom. 263E). Likewise, Ovid sees purity as one of seven potential reasons for the use of fire and water together in the Parilia (Ov., Fast. 4.783–806). We should also remember that the denial of fire and water (‘aqua et igne interdici’) was a symbolic punishment for serious criminal offences against the state that resulted in exile and loss of citizenship, with the two elements representing essential provisions for life (Ov., Fast. 4.791–2).Footnote ³⁹ Any inclusion in funerary ritual, therefore, might have been related to more complex considerations than recognised in the synchronic narrative.

Os resectum is better attested than the suffitio; references to the rite appear in Cicero, Festus and Varro. None of these authors, however, clarify what it entailed, and all describe it using slightly different language. The term os resectum comes from Cicero, unfortunately included in a list of rituals he chooses not to explain in detail:

Cicero goes on to describe inhumation as the most ancient form of burial, citing this as the reason why the pontifices required an earth throwing ceremony to accompany cremation (Cic., Leg. 2.22.56–7). Festus, meanwhile, suggests that a bone removal ritual accompanied cremations and allowed proper rites to be maintained: ‘a limb was said to be cut from the deceased when his finger was cut off, and because it had been saved, the rites could be carried out when the rest of the body had been cremated’ (s.v. membrum, ‘membrum abscidi mortuo dicebatur, quum digitus eius decidebatur, ad quod seruatum iusta fierent reliquo corpore combusto’). Together, the two passages have guided the standard idea that os resectum was necessary for legitimising cremations, allowing one small part of the body to be buried in keeping with the traditional practice of inhumation.Footnote ⁴¹

Cicero ties os resectum to the end of the period of mourning, an idea that Varro supports. Varro's account, furthermore, associates this act with a ‘purging’ of the family:

Beyond making the grave legitimate, therefore, the ceremony has been reconstructed as a cleansing rite that transitioned the polluted familia funesta back to purity.Footnote ⁴³ Another passage from Cicero has been taken as confirmation for this theory:

Drawing from these texts, the synchronic narrative has reconstructed os resectum as a ceremony that legitimised the grave, concluding the proper interment of the deceased and cleansing the family of death pollution.

This tidy reading of os resectum fuses Cicero, Festus and Varro, smoothing some significant rough edges in order to make each piece complement the next. In fact, none of these texts describes the rite in straightforward or fully complementary terms, to the point that all use different vocabulary: os resectum in Cicero, os exceptum in Varro and membrum abscidi in Festus. Corruptions in Cicero's text, furthermore, could mean that the phrase most often used in modern scholarship is a misreading. John Scheid has argued that os relectum (gathered, collected), reiectum (thrown back) or receptum (taken back) are equally likely, any one of which could alter interpretation.Footnote ⁴⁵ Taken together, Cicero, Festus and Varro suggest some kind of ceremony that involved one of the deceased's bones. They leave unclear, however, how widespread the practice was, what exactly was done with the bone, for how long the rite was a relevant part of funerary ritual, and what variations might have existed. Without these basic pieces of information, attempts to reconstruct the beliefs that guided the act seem unlikely to be successful.Footnote ⁴⁶

By piecing together disparate evidence for funerary rites into the synchronic narrative of the ‘Roman funeral’, past work has reconstructed an overarching fear of death pollution that guided nearly every aspect of the proceedings. When removed from that framework, however, only a few elements indicate pollution concerns, and all of those invite questions. Regarding the corpse itself, we have Servius’ statement on cypress boughs and the sweeping ceremony mentioned in Festus, neither of which can be tied firmly to the late republican or early imperial periods. The incorporation of torches, incense and music might also indicate the corpse as a source of pollution, but are more likely to have purified mourners undertaking religious action, and other concerns could have guided their use. Noting the weakness of evidence surrounding the corpse, others have proposed that pollution fears can be tracked more reliably through rituals focused on mourners, whose status as polluted or pure was of particular concern since they posed the greatest risk of spreading contamination through living society.Footnote ⁴⁷ The rites surrounding the familia funesta, however, are only slightly less problematic than those associated directly with the corpse. Like the sweeping ceremony, the suffitio is difficult to reconstruct as a typical practice of the second half of the first century b.c.e. and later. As poorly understood as it is, os resectum comprises our best evidence: Cicero and Varro both expected their readers to be familiar with the ritual, and together those passages suggest that it was tied to a ‘purging’ of the family to remove the funestus. None of this evidence, however, clarifies what it meant to be funestus — how did the status effect mourners, and is that effect best conceived in terms of pollution?

IV THE FAMILIA FUNESTA

Following the synchronic narrative, members of the familia funesta were polluted by death and therefore unable to maintain typical activities as they separated themselves from daily life.Footnote ⁴⁸ The standard interpretation contends that their pollution was marked clearly by special dress, which allowed others to avoid them. The relevant texts suggest that dark clothes were common, and the bereaved also might have soiled themselves with ashes or dirt, torn their cheeks or pulled out their hair.Footnote ⁴⁹ Nevertheless, no ancient authors associated mourning clothes with a warning against polluting danger, but instead presented them as outward signs of grief. When Propertius imagined his mistress Cynthia returning from the dead to berate him for missing her funeral, she explicitly connected mourning dress to his (lack of) emotion, accusing him of never having bent over her grave, crying into his black toga (Prop. 4.7.27–8). Likewise, Juvenal lamented the difficult lives of the poor by stating that no one cares when a small hovel catches fire, taking in one sweep all that a poor man has, but when a rich man's house burns the city despairs, dresses in black and rushes to build him an even swankier mansion (Juv. 3.203–20). Perhaps mourning clothes also warned against pollution, but their use in public life points away from that interpretation. During the late republican period, criminal defendants and their family and friends dressed as mourners to arouse pity and draw attention to their lot, as did condemned persons and legates or plebeians appearing before the Senate.Footnote ⁵⁰ As in funerary contexts, so in public life mourning clothes communicated emotion. In the Pro Caelio, for example, Cicero describes the depth of Caelius’ parents’ grief by equating the mourning dress of his father to the tears of his mother (Cic., Cael. 2). When Cicero encountered his own legal troubles, his supporters protested his exile by adopting mourning clothes (Cic., Dom. 113, Red. pop. 13, Sest. 32). Later, when Mark Antony faced a public-enemy charge, his friends dressed in mourning and threw themselves before the doors of the curia. Cicero, noticing that many senators were moved to pity by the spectacle, delivered a biting speech against Antony, but failed to secure a conviction (App., B. Civ. 3.8.51–61).

Although past work has stressed official restrictions on the familia funesta, the idea is built on very little evidence. Aulus Gellius is most often cited as support; when discussing ancient practices for declaring war, he recalls that a soldier could delay arriving for service if there were a death in his family or if he were observing the feriae denicales (Gell., NA 16.4.4). The allowance has been interpreted as a sign of the mourner's pollution and the necessity that he be purified prior to joining the army.Footnote ⁵¹ Significantly, however, Gellius specifies that the exception from mustering was not valid if the funeral or the feriae denicales had been scheduled for the express purpose of delaying a report for duty, ‘as long as these have not been planned on that day so that he might be absent’ (‘quae non eius rei causa in eum diem conlatae sint, quo is eo die minus ibi esset’). Presumably in that case, the soldier was required to appear even before the completion of funerary rites and any associated purification. Although we might question how the authorities enforced such a rule, certainly it does not suggest a fear of pollution. In this light, the funeral exception seems better linked to others on the list that allow for the proper maintenance of religious activity; absences also were permitted to perform rites in response to an omen or to celebrate an anniversary sacrifice within the household cult. Of course, this discussion overlooks a further complication — Gellius describes a historic practice, citing the first-century b.c.e. antiquarian Cincius as his source.Footnote ⁵² Notably, Cincius himself characterises the rules as governing ancient levies (dilectus antiquitus), and so they seem to have been outdated already by his time.

Other texts treated as evidence for restrictions on mourners are similarly difficult to relate to pollution fears. Among these is the Lex Ursonensis, a late republican foundation charter for the colonia Iulia Genetiva (modern Osuna in Spain) that excuses those celebrating the feriae denicales from appearing in court.Footnote ⁵³ Although recovered far from Rome itself, the law sometimes has been taken to indicate that the same restriction applied in the capital.Footnote ⁵⁴ Even if this were the case, however, the allowance might have concerned correct religious practice, much like the mustering restriction recorded in Gellius. Similarly, Columella says that the pontifices forbade harnessing mules during the feriae denicales (Columella, Rust. 2.21), but it is unlikely that those in mourning were prohibited from work due to fears that they might spread pollution, as some have imagined.Footnote ⁵⁵ Cato specifies that the pontifices permitted harnessing mules, horses and donkeys only during public festivals; during all domestic rites — not only those following a death — harnessing was forbidden (Cato, Agr. 138). Surely not all household religious activity resulted in pollution; mourners, therefore, were barred from working not due to their uncleanliness, but because of some particular aspect of domestic ritual, possibly the necessity of gathering the familia of slaves who might normally be in the fields.Footnote ⁵⁶ Furthermore, although Plutarch reports that Numa's laws required a widow to remain in mourning for ten months before remarrying, nothing indicates that she was considered polluted during that time (Plut., Num. 12.2).Footnote ⁵⁷ Widowers faced no restrictions on remarriage (Sen., Ep. 63.13), and the rule for widows has clear practicality; a delay of ten months would remove any doubts over the paternity of future children.Footnote ⁵⁸ Indeed, when Ovid discusses regulations on mourning, he attributes the prohibition explicitly to paternity concerns (Ov., Fast. 1.33–6).

The synchronic narrative has emphasised that becoming funestus required mourners to separate themselves from society for a period of time that is most often conceived as beginning with the moment of death and continuing for nine days following the interment, until the celebration of the nouendialis.Footnote ⁵⁹ A series of anecdotes involving elite men who had lost their sons challenges that reconstruction.Footnote ⁶⁰ Several authors state that Tiberius convened the Senate immediately after Drusus’ death, prior to the funeral and so presumably before any purificatory rites had been conducted (Tac., Ann. 4.8; Suet., Tib. 52.1; Cass. Dio 57.22). No account references pollution, instead focusing on Tiberius’ denial of grief to prioritise his responsibility to the state.Footnote ⁶¹ The dichotomy between private grief and public duty, along with praise for a quick return to normal life, emerges elsewhere as well. Valerius Maximus records that the consul Quintus Marcius Rex left his only son's pyre to convene the Senate, placing the state ahead of his personal pain (Val. Max. 5.10.3). Seneca the Younger reports that Marcus Calpurnius Bibulus, the consular colleague of Caesar, carried out his normal public duties the day after he heard that both of his sons had been murdered in Egypt (Sen., Ad Marc. 14.2). We are told that Aemilius Paulus, moreover, celebrated a triumph for his victory in Macedonia only days after the death of one son and lost his second son immediately following (Liv. 45.40–1; Plut., Aem. 35–6; Val. Max. 5.10.2). He then called an assembly of the plebs and delivered a rousing speech, saying that he had feared bad fortune would balance the greatness of his victories, but was glad that it had come at his own expense, rather than the state's. The point at which we began this investigation — Seneca's story of Asinius Pollio reporting that he dined on the day his son had died — also points away from any strict injunction against renewing social life while in mourning (Sen., Controv. 4.5). Seneca clearly admires Pollio's fortitude in the face of adversity, noting that he had declaimed within three days of his son's death (Sen., Controv. 4.6).

These anecdotes should not be taken as evidence that death did not impact the living; the very existence of stories praising those who continued public duties in the face of private loss implies that such men had behaved extraordinarily.Footnote ⁶² Nevertheless, the passages show no sign that mourners were restrained by pollution, either as vectors of miasmic contagion or as sources of more general offence. Roman attitudes towards excessive mourning also are relevant here. In contrast to the Indonesian tribes studied by Hertz (or at least, in contrast to his observations and interpretations of their behaviour), the late republican and early imperial texts indicate that the bereaved could violate acceptable norms by separating themselves too far from society and indulging too deeply in their grief, a common theme of consolation literature.Footnote ⁶³ One example comes from Seneca the Elder's Controversiae: the imagined case involves a mourning father who was dragged from the tomb of his children, forced to remove his mourning clothes and compelled to attend a party (Sen., Controv. 4.1). Seneca's discussion highlights the difficultly of balancing grief and proper social conduct, while indicating that norms could be negotiated internally and subject to individual interpretation. In this case, the father felt even worse upon being thrust back into normal life, while the young man who had compelled him to leave the necropolis believed that he had been acting as a friend. Note, too, that the only Roman texts discussing lengths of mourning dictated maximums rather than minimums, suggesting that their goal was to truncate excessive bereavement, not to require set periods in which contagion necessitated separation.Footnote ⁶⁴ Finally, this discussion overlooks an essential point: Plutarch's Numa taught the first pontifices explicitly that burial rites were not polluting, a passage that has yet to be reconciled against the synchronic funeral's emphasis on death pollution (Plut., Num. 12.1).

V MOURNERS AND THE GODS

Just as removing the scaffolding of the synchronic narrative reveals how little we know of Roman funerary rites, examining mourners without the expectation of pollution undermines standard assumptions on their behaviours. Lest we despair of ever finding solid ground on the topic, however, a variety of authors agree on one point: death could interrupt certain interactions with the gods. Livy tells us that women were unable to conduct rites for Ceres while in mourning, necessitating the Senate to limit mourning periods during the Second Punic War to ensure that some women were available to honour the goddess (Liv. 22.56.4–5, 34.6.15). Another instance comes from Plutarch's Life of Sulla. According to Plutarch, Sulla once had dedicated one-tenth of his property to Hercules and provided a massive, multi-day feast for the people (Plut., Sull. 35). During the feasting, his wife, Metella, grew seriously ill. With her death immanent, the pontifices forbade Sulla from seeing her, and the dictator went so far as to divorce her and have her carried to a different house before she died. Sulla loved Metella; after the festival had concluded he gave her an elaborate funeral, breaking — Plutarch notes — the sumptuary laws that he himself had passed. The text specifies that Sulla's callous behaviour during his wife's final days was not due to his feelings towards her, but to religious concerns (δεισιδαιμονία).Footnote ⁶⁵ It seems that Metella's death would have obstructed Sulla's ability to host the festival for Hercules, requiring him to take extraordinary measures to separate himself from her.

Some priests also avoided the dead. According to Aulus Gellius, the Flamen Dialis was forbidden from touching a body or approaching a pyre and required to leave office in the event of his wife's death (Gell., NA 10.15.23–5).Footnote ⁶⁶ These rules, however, are difficult to fit into a larger belief system, being part of a long list of strange restrictions specific to that office. For example, the Flamen Dialis also was prohibited from walking under a grape arbour and was required to cover the foot of his bed with clay and bury cuttings from his nails and hair under a fruit-bearing tree.Footnote ⁶⁷ Better evidence comes in the form of constraints on members of imperial family who acted as public priests. Here we might think of Tiberius’ disapproval of Germanicus overseeing the burial at the Teutoburg Forest while he was serving as augur (Tac., Ann. 1.62; cf. Suet., Calig. 3), instances in which both Augustus and Tiberius used curtains to prevent themselves from viewing corpses while delivering eulogies (Cass. Dio 54.28.4–5, 54.35.4; Sen., Ad Marc. 15.3), and the Senate's pardoning of Tiberius for touching the corpse of Augustus (Cass. Dio 56.31.3). Curiously, several texts debate the reasons for such behaviours, possibly indicating inconsistencies or changes in practice through the early imperial period.Footnote ⁶⁸ Tacitus, for example, does not fully understand why Tiberius saw Germanicus’ actions as inappropriate, and as Dio recalls, various authorities argued over which office had constrained Augustus and Tiberius: pontifex maximus or censor.Footnote ⁶⁹ Nevertheless, all expect that certain priesthoods might require particular actions in the face of death.

The best known example of death's potential to interrupt religious activity concerns the Temple of Jupiter Optimus Maximus on the Capitoline (Liv. 2.8.7).Footnote ⁷⁰ As the story goes, in the early days of the Republic the temple had yet to be dedicated. The consuls, Publius Valerius Publicola and Marcus Horatius Pulvillus, drew lots to decide who would receive the honour, and Horatius won. Offended, the friends of Publicola attempted to stop the dedication. Finally, they interrupted the ceremony itself with news that Horatius’ son had died and declared that while his family was funestus, he could not complete the dedication. Rather than halting the ceremony, however, Horatius paused only to order that the corpse be carried out for burial and then continued his prayers. The story suggests that the death should have prevented the dedication, but Horatius was able to deny the loss of his son to complete his public duty. As noted already by Servius (Ad Aen. 6.8, 11.2), Horatius might have avoided becoming funestus by refusing to indulge in his grief; according to Livy, he described his son's remains as a cadaver, a term that typically implied an abandoned body and so might have signified his emotional separation from the loss.Footnote ⁷¹ Seneca's version, too, stresses the denial of grief (Ad Marc. 13.1–2), recording that Horatius mourned upon returning home, but quickly turned his eyes back to the capital. The story implies that readers of the early imperial period were familiar with the idea that mourners — or at least, certain mourners — were expected to separate themselves from standard religious interactions, while the setting at the very dawn of the Republic indicates that those readers conceived of the practice as ancient. The dramatic gestures of the first two emperors, both of whom had keen interests in preserving or reviving ‘traditional’ religious life, further suggests that contemporary Romans saw their acts as part of a venerable custom.

The synchronic narrative of the Roman funeral reconstructs the familia funesta as polluted by death and therefore required to remove themselves from the presence of the divine.Footnote ⁷² I remain unconvinced, however, that pollution provides the best means of understanding this phenomenon. Beyond the accounts discussed just above, two passages of Ovid form the primary evidence that the dead were distasteful to the gods. The first is a statement in the Fasti that temples should be closed and weddings avoided during the Parentalia, when the dead wander freely (Ov., Fast. 2.557–70).Footnote ⁷³ Nevertheless, we should be cautious; the Fasti was a work of literature, and its narrator a literary construct whose timidity and naivety — traits that lead him to trust even in demonstrably unreliable sources throughout the text — have been well documented.Footnote ⁷⁴ The divinised dead known as the manes were available to assist the living throughout the year and in any location; their influence was in no way limited to their tombs or to the days of the Parentalia.Footnote ⁷⁵ The closure of temples and avoidance of weddings, therefore, should not be attributed to some atypical presence of death in the realm of the living.Footnote ⁷⁶ Likewise difficult to reconcile is Ovid's story of Alcyone, who upset Juno by praying with ‘funestus hands’ (‘manus funestas’) for the safety of a husband who — unbeknown to her — had died at sea (Ov., Met. 11.573–91). In this case, the story's setting in Trachis as well as in the deep mythological past complicates any attempts to read the comment as a direct reflection of beliefs common among Ovid's audience. We cannot know, for example, to what degree Ovid was influenced by fifth-century Athenian ideas surrounding death pollution, preserved in tragedies such as Sophocles’ Antigone or in the purification rites that accompanied festivals such as the Chytroi.Footnote ⁷⁷ Death pollution might have seemed natural within the world of the text, without any corresponding fear shared by contemporary readers at Rome. Finally, and looking for a moment beyond the literature, the legal relegation of burial to outside the city has often been taken as a clear marker of separation between the gods and the dead, the latter of whom were required to remain beyond the ritual boundary of the pomerium.Footnote ⁷⁸ Nevertheless, the so-called Servian wall of the fourth century b.c.e., not the pomerium, is best reconstructed as the boundary for internment at Rome.Footnote ⁷⁹ As the city extended past that wall, a process that began already by the second century b.c.e. and accelerated rapidly in the late republican and early imperial periods, tombs mixed with other development of all types: shops and workshops, apartment complexes and elite estates, major entertainment and other public buildings and, notably, extramural shrines and sanctuaries.Footnote ⁸⁰ None of these data demonstrate that the death offended the gods.

VI DEATH AND OBLIGATION

Bearing these points in mind, I prefer a more neutral translation for funestus: rather than polluted, mourners are better conceived of as obligated, bound to conduct the proper rites for the dead. This understanding draws funestus in line with other specifically charged or contracted religious states, such as inauguratus, institutus, imperatus, sanctus or even the damnatus of an individual who had received a divine request without yet fulfilling his vow.Footnote ⁸¹ None of these conditions was inherently positive or negative; each brought requirements and constraints, responsibilities as well as rights. Likewise, becoming funestus released mourners from other duties so that they might navigate familial responsibilities. The possibility of circumventing the state underlines its formal, delineating quality, a characteristic common to Roman religious thought.Footnote ⁸² The funestus could be rejected, as in the story of Horatius, or avoided on technicality, as for Sulla. In extreme situations, a third party could even forbid mourners from fully indulging in the state, as when the Senate limited women's mourning during the Second Punic War. Upon completing funerary rites, the funestus lifted. Varro's use of the word purgo to describe this process could indicate that the mourners had been considered metaphysically tainted, but does not necessitate that reading (Varro, Ling. 5.23).Footnote ⁸³ In some instances, that term implied a ceremonial purification, but in others it indicated a clearing, as when clouds part to reveal Aeneas to Dido,Footnote ⁸⁴ or even a freeing from troubles, as in Lucretius’ reference to the teachings of Epicurus removing various passions from his followers’ spirits.Footnote ⁸⁵

Nor does Cicero's opposition of the familia funesta and familia pura require that the former be seen as polluted (Cic., Leg. 2.22.57). An object that was purus could be ritually pure, but the word also was used to denote an original or unembellished state, as when Vitruvius discussed unadorned cornices (Vitr., De arch. 4.5.11, puras coronas), Ovid referred to a woman's undressed hair (Ov., Ars am. 3.137, capitis puri) or Juvenal described a greedy client who desired ‘only’ plain silver cups (Juv. 9.141, argenti uascula puri).Footnote ⁸⁶ Perhaps most relevant to the familia pura, however, is the use of purus in legal contexts, where the term signified a person or object that was not subject to condition.Footnote ⁸⁷ Indeed, purus was used in precisely this way in the civil laws regarding tombs, which differentiated a locus religiosus, land on which a body had been interred, from a locus purus, land without interred human remains.Footnote ⁸⁸ In this context purus did not indicate ritual purity, but neutrality; the law separated land that was bound by religio — with all accompanying obligations on the living community — from land that was unbound.Footnote ⁸⁹ Cicero, moreover, reports that the land on which a tomb was located did not become religiosus unless the proper rituals were performed (Cic., Leg. 2.22.55). Following Cicero, the presence of a corpse did not naturally impose this status on a tomb; the completion of rites removed it from its former state, and more importantly, bound the living to it through a network of obligations. In this context, a locus purus was not free from a polluting body, but free from specific religious ties — that is, not ‘pure,’ but unencumbered. This understanding explains practices surrounding exhumation. Moving or destroying a grave required expiation only in the case of a legitimate burial, since only then did the act destroy a locus religiosus.Footnote ⁹⁰ Illegitimate burials, located on land unbound by religio, could be moved freely without accompanying rites. Cicero's statement also clarifies why uninvited burial on public land or on land belonging to another did not legally change the status of that land; despite the presence of an interment, the land remained purus if the grave violated the rights of the owner freely to sell or transfer the property.Footnote ⁹¹ Similarly, public burial grounds used to dispose of individuals without the means for a funeral appear to have been considered loca pura despite the presence of corpses, since no rites had been performed to bind the area with religio.Footnote ⁹²

Interpreting the familia funesta as obligated, however, does not explain why that obligation interrupted certain communications with the gods. Here, we might benefit from reincorporating the work of van Gennep and Hertz, both of whom stress death's tendency to separate mourners from their communities. As we saw above, van Gennep recognises that mourners, like the dead themselves, navigate a tripartite rite of passage involving separation from society, transition through a liminal phase and reincorporation. Mourners in the second stage — like all undergoing a rite of passage — become ‘sacred,’ a term van Gennep uses to indicate a state of separation from normalcy, and which we might compare to the Roman funestus.Footnote ⁹³ Hertz helps to clarify the cause of their separation. He sees the process as primarily internal, driven by emotions like grief and feelings of duty, in a balance that varies between individual mourners.Footnote ⁹⁴ Following Hertz, the momentary nature of biological death challenges human perceptions. Mourners’ connections to the deceased leave them unable to comprehend the sudden loss of their loved ones, a phenomenon that is strongest for those whose relationships were closest. Hertz argues that emotional responses pull mourners towards the dead, separating them from normal life until funerary rites — and, importantly, the passage of time — allow them to process the death, disentangle themselves from the deceased and return to the world of the living.

This liminality best explains the constraints on Roman mourners’ interactions with the gods. Nearly all attestations for such restrictions relate to elites — and especially, elite men — who communicated with the gods on behalf of the state.Footnote ⁹⁵ Great public rituals both reflected and structured Roman society, serving to stabilise the social system. Led by elite priests, the rites reinforced hierarchies. Those lower on the social spectrum played lesser roles; when they were present, they most often served as audience members.Footnote ⁹⁶ The religious activities of the elite, therefore, were most likely to be impacted by a death, since they held the responsibility for conducting public ceremonies. Significantly, moreover, their duties were contingent on their place at the top of the living community. As both Hertz and van Gennep demonstrate, the dead not only pull mourners away from the living but also upset their social positions. Death literally overturns roles; the loss of a pater familias, for example, sent status changes reverberating through a Roman family, especially on the male side. The death of a public priest or member of the imperial family might have had similar reverberations through a larger portion of society. At the same time, mourners can throw off social positions more figuratively. We should remember that Roman mourning clothes not only signalled grief, but also denied status.Footnote ⁹⁷ Dark clothes marked foreigners and the poor; by exchanging his white toga for a dirty black robe, a Roman magistrate symbolised the upheaval of death, which separated him from his normal life.Footnote ⁹⁸ Setting aside the toga, he also suspended his official duties and at least figuratively turned away from public life and towards the private world of his own family. I believe that this turning away had the greatest impact on an individual's dealings with the gods. By symbolically stepping out of his place at the top of society, a Roman priest temporarily rejected his ability to undertake religious action on behalf of the state. For this reason, the unusual act of refusing to step aside (as both Horatius and Sulla did in slightly different ways) allowed him to continue in his normal roles, even in the face of death.

VII DEATH POLLUTION: A LATE ANTIQUE PHENOMENON?

There is no reason to conclude that fears of death pollution shaped Roman funerary practice of the late republican or early imperial eras. By Late Antiquity, however, the situation might have changed. As noted above, Servius states unequivocally that Romans of Virgil's day used cypress branches to warn against death pollution, a comment that has influenced nearly all subsequent work on Roman death (Serv., Ad Aen. 3.64). A full study of attitudes towards the dead in Late Antiquity lies beyond the scope of this article. Nevertheless, we should note that Servius wrote in a time during which many pagan leaders and intellectuals presented corpses as vectors for metaphysical contagion, and that attitude might have influenced his interpretation of earlier practices. Various authors of the fourth and fifth centuries stress the repulsive character of corpses, probably attempting to undermine the growing cult surrounding martyrs’ relics.Footnote ⁹⁹ We might think, for instance, of Eunapius’ statement that Christians visit martyrs’ tombs to ‘become better by staining themselves at graves’ (Eunap., VS 6.11.8–9, κρείττους ὑπελάμβανον εἶναι μολυνόμενοι πρὸς τοῖς τάφοις). References to death pollution also appear in law for the first time in the mid fourth century; these include a ruling of Constantius II and Julian dated to 356 or 357 c.e. that penalised tomb destruction on the grounds that the act both disrespected the dead and polluted the living, as well as Julian's edict of 363 c.e., which required the dead to be removed at night to prevent their polluting passers-by (Cod. Theod., 9.17.4–5).Footnote ¹⁰⁰ This last example might itself indicate how contemporary life impacted Servius’ interpretation of earlier antiquity; he reports that Romans originally had held their funerals at night, a view that remains entirely unsupported (Serv., Ad Aen. 6.224).Footnote ¹⁰¹ Mindsets of the fourth and fifth centuries appear to have influenced later interpretations of late republican and early imperial practice: for example, the seventh-century glossary preserved in the Vatican Library (Vat. lat. 3321) that explicitly defines funestus as ‘polluted by death’, the first known instance of the definition most often applied today.Footnote ¹⁰² Clearly, much work remains to be done, but for the time being I would caution against using late antique material as evidence for earlier beliefs and practices. Such texts provide insight into their authors’ own experiences but have far more tenuous connections to preceding centuries.

Whatever the case in Late Antiquity, the concept of death pollution has been unhelpful for those seeking to understand Roman responses to death in the late republican and early imperial periods. The texts are less equipped to represent even earlier beliefs and practices, but there is likewise little support for the idea of a traditional fear of the dead that had faded by the first century b.c.e.Footnote ¹⁰³ Chasing the evidence through the rites of the funeral and the behaviours of the familia funesta suggests neither fears of a metaphysical danger surrounding corpses nor strict injunctions separating survivors from the rest of society. Suffering a death could remove mourners from standard interactions with the gods — or at least, certain mourners from certain types of interaction — but nothing indicates that situation to have resulted from pollution. Rather than polluted, the familia funesta is better conceived as obligated, bound by a combination of emotions and conventions to behave in certain ways following a death. Above all, mourners were responsible for overseeing the transition of the dead into manes, but determining how that was done requires more consideration than any brief treatment I could provide here.Footnote ¹⁰⁴ I therefore leave the question to future work, with the warning that attempting to connect scattered evidence into a coherent narrative can lead us astray. How the dead were ushered into their new existence almost certainly varied across time and space, as well as according to factors like social status, wealth or even the distinct traditions maintained by individual families. Whatever their form, such rites lifted the funestus not by cleansing mourners of pollution, but by allowing them to process their loss and reintegrate into a new reality without the deceased. Ultimately, the degree to which death impacted a mourner's life varied according to his or her social roles and was dictated by internal responses as well as external expectations. This is a new way of understanding Roman responses to the dead, and it accords well with the surviving texts. As crucially, it allows for space and flexibility, acknowledging all that we do not and cannot know of how Romans grappled with the complexity of death.