What is attested
Roman household religion is documented above all by furniture. The eruption of AD 79 preserved, in Pompeii and Herculaneum, hundreds of fixed shrines: painted panels on kitchen walls, plastered niches, masonry aediculae with columns and pediments, and the traces of wooden cupboards. George K. Boyce catalogued the Pompeian examples in the Memoirs of the American Academy in Rome in 1937, and that corpus, with its later supplements and revisions, is still the base of the subject. The shrines are found in atria, in gardens and peristyles, and very often in kitchens and service corridors; a single house may have more than one, in rooms of very different status. That distribution is one of the few hard facts we have about where household cult was installed; what it implies about who used it is disputed, and is set out below.
The commonest painted scheme is remarkably standardised. Two youthful figures in short belted tunics stand or dance on tiptoe at either side, each holding a drinking horn and a bucket. Between them a togate figure stands with the fold of his toga drawn over his head, holding a shallow offering dish and often a small box; attendants and a sacrificial pig may accompany him. Below, on a separate register, one or two crested serpents approach an altar bearing an egg or a pine cone. The flanking figures are identified as Lares chiefly by comparison with the public compital cult, where dedications and reliefs name the Lares explicitly — a cross-reading from civic to domestic material of the kind this brief warns against elsewhere, and the identification should be weighted accordingly. The central veiled figure is identified as the Genius of the head of the household on the strength of inscriptions elsewhere that pair the Genius with the Lares — a strong inference, but an inference: the paintings themselves are almost never captioned.
Alongside the shrines are the statuettes. Bronze and occasionally silver figures were found in and near household shrines and in the cupboards and boxes associated with them: Lares, Mercury, Hercules, Venus, Isis and other Egyptian deities, small portrait busts of ancestors. No two assemblages are identical and there is no fixed household pantheon. What the finds show is selection — a family assembled the gods it wanted, and the mix in a given house is evidence about that house, not about Roman religion in general. Find-recording in the early excavations was poor, so for many statuettes the association with a particular shrine is looser than the published literature implies.
One point of terminology matters. The word lararium is unattested in any surviving text before late antiquity, where it appears in the imperial biographies of the Historia Augusta, most famously in the fictionalising account of Severus Alexander's private chapel; the terms in earlier use are sacrarium, sacellum and aedicula. 'Lararium' is a modern scholarly convenience projected backwards onto the Pompeian material, and the projection carries an assumption with it: that these shrines were primarily shrines of the Lares.
The written evidence is thin, scattered, and almost never descriptive of an actual occasion. The single most useful passage is Cato's instruction, in his mid-second-century BC farming manual, that the estate manager's wife hang a garland at the hearth on the Kalends, Nones and Ides and on holy days, make her petition to the Lar familiaris according to her means, and keep the hearth swept clean daily. That is prescription addressed to a slave overseer on a country estate, not observation of a household at prayer. Plautus opens a comedy with the Lar familiaris of a house speaking in his own person and mentioning the daughter's daily offerings; Horace addresses a poem to a countrywoman making a modest offering; Petronius has boys at Trimalchio's dinner set two bulla-wearing Lares on the table, to which Trimalchio gives three joke names, and describes silver Lares in a shrine in his atrium in a separate passage. Each is a literary construction. The contrast with the public crossroads cult is sharp: the compital shrines of Rome, reorganised under Augustus with Lares Augusti and annually appointed neighbourhood officers, produced dated inscriptions naming the men who served — the domestic shrines produced almost no text at all.
How the rite worked
The occasions we can name come almost entirely from Cato: the Kalends, Nones and Ides of each month, plus festival days, marked by a garland at the hearth and a petition to the Lar. The hearth is the focus of that instruction, and it is worth noting that in Pompeii the painted shrines are frequently in or near the kitchen — the one point where text and archaeology converge. Beyond this, household observance is reported for life-cycle moments: a boy's dedication of his bulla on assuming the adult toga, offerings by a bride entering a new house. These are attested in later antiquarian and lexicographical sources rather than in contemporary description, and should be handled as reports about practice rather than records of it.
The offerings visible in the iconography and implied by the finds are cheap and repeatable: incense, wine poured from a small dish, garlands, cakes, first fruits, eggs. The pig that appears in some paintings represents a much larger outlay and cannot have been an everyday event. Small altars, incense burners and lamps recovered near shrines fit the same picture — a cult of frequent small acts rather than periodic expense. The officiant in the paintings has his head veiled in the standard Roman sacrificial manner and is accompanied by a youth attendant conventionally identified with the camillus of state sacrifice — an identification the paintings do not label, and one that carries a status requirement a household whose cult may have been conducted by slaves need not have met.
Personnel is the area where the material evidence is genuinely informative. Shrines in service quarters, and dedications by slaves and freedmen to the Genius of their master, indicate that household cult was not confined to the free family. Whether kitchen shrines were used chiefly by the slave familia, or simply placed where the fire was, is disputed. Cato assigns the hearth observance to a woman of servile status — but only under the master's or mistress's authority, having first forbidden her to perform rites on her own initiative or to have anyone perform them for her. The passage complicates the assumption of a uniformly male-headed domestic cult without overturning it.
What cannot be reconstructed is the act itself. There is no surviving prayer, no formula, no order of service for a household shrine, and no source that tells us how often anything was done at one. The paintings show a moment of sacrifice in a highly conventionalised form; they are decoration on the shrine, not a record of the rite performed at it.
The evidence and its limits
Each class of evidence below is followed by what it cannot show. That second half is not a disclaimer; it is the part most accounts of ancient religion leave out.
Archaeology
Fixed shrines in situ at Pompeii, Herculaneum, Ostia and elsewhere, with their positions in the plan of the house, plus associated altars, lamps, incense burners and statuette assemblages.
What it cannot show — Furniture is not liturgy. Excavation can show where a shrine stood and what stood on it, never what was said or done, how often, or by whom. Nineteenth-century recording at Pompeii was inconsistent, so many object-to-shrine associations are less secure than they appear in print.
Images
The standardised painted scheme of paired Lares, veiled sacrificant, attendants and serpents gives a repeated visual formula across hundreds of examples.
What it cannot show — A repeated formula is evidence of a pictorial convention, not proof that the depicted act took place in that house. The paintings are almost never labelled, so the identification of the central figure rests on external argument.
Literary sources
Cato gives a prescriptive calendar of hearth observance; Plautus, Horace, Ovid and Petronius supply incidental scenes; the Historia Augusta supplies the word lararium.
What it cannot show — None of these authors is describing an observed domestic rite for the record. Cato prescribes, the poets construct, Petronius satirises, and the Historia Augusta is a late and unreliable source that here describes an emperor's private chapel, not an ordinary house.
Inscriptions
Dedications to the Genius of a paterfamilias and to the Lares, and the rich epigraphic record of the Augustan compital cult with its neighbourhood officers and Lares Augusti.
What it cannot show — The abundant inscriptions are public and neighbourhood-level; household shrines themselves are almost entirely mute. Using compital evidence to explain domestic practice imports the organisation of a state-reformed civic cult into the private house.
What the sources do not record
No prayer, invocation or formula from a household shrine survives — not one line. Everything ever written about 'the daily rite at the lararium' is reconstruction from images and from Cato's prescriptive sentence.
The evidence does not record how often anything happened. Cato names monthly and festival days for one estate; nothing establishes that any Pompeian household observed that calendar, or any calendar.
The paintings do not tell us whether they depict what was done there. They may show an ideal, a generic sacrifice, or a scene lifted from the compital repertoire.
The record is overwhelmingly Campanian and dates to a single moment in AD 79. Household religion in Republican Rome, in the provinces, and in later centuries is poorly documented by comparison, and the Pompeian picture is routinely and illegitimately generalised to all of it.
Nothing records what happened to a household's gods when a family moved, died out, or sold the house, or how a new shrine was consecrated.
What the evidence supports
The canonical scheme in a well-preserved example: two dancing Lares with horn and bucket flanking a veiled sacrificant, with a crested serpent below approaching an altar.
Painted aedicula shrine in the House of the Vettii (Pompeii (region VI)).
The painted figures correspond to a widely produced statuette type, and that households owned portable images matching the wall paintings.
Bronze statuettes of Lares with drinking horn and situla, from Pompeian houses (Museo Archeologico Nazionale, Naples).
Household cult installations were sited in working areas as well as in display rooms, and that a single house could have several.
Painted shrines in kitchens and service corridors across the Pompeian housing stock (Pompeii).
The public counterpart of household cult, with named officiants, dated dedications and a formal iconography of Lares — the reason the domestic shrines' silence is so conspicuous.
Altars of the Augustan crossroads cult showing neighbourhood officers sacrificing to the Lares Augusti (e.g. the Belvedere altar) (Vatican Museums, Rome).
The scale and typological range of the evidence, and the shrine-by-shrine find data on which all later argument depends.
Boyce's catalogue of Pompeian household shrines, Memoirs of the American Academy in Rome vol. 14 (1937) (Documentary record of the Pompeian corpus).
Stories told about the rite
Ancient writers offered these as explaining the practice. They are evidence for what the tradition said about itself, and they are not the origin of anything: each is given with the author who tells it and the distance between the two.
- Evidence level: Mythological
Ovid, Fasti book 2 (c. AD 8), an aetiological poem on the Roman calendar.
The Lares were born to the nymph Lara, whose tongue Jupiter cut out for talking, and who was fathered upon by Mercury as he escorted her to the underworld.
Offered as explaining — Explains why the Lares are two, why their mother is silent, and connects them with Mercury, who often appears in household shrine paintings.
An Augustan poet's etymological invention playing on Lara/Larunda and lares. It postdates the cult by centuries and explains a name, not an origin.
- Evidence level: Mythological
Ovid, Fasti book 5.
The Lares Praestites, guardians of the city, were honoured with a dog at their feet because they and the dog both keep watch over the household.
Offered as explaining — Explains an iconographic attribute and the protective function of the Lares.
Ovid supplies several alternative explanations for the same festival elsewhere, which is itself a sign that he is reasoning from images rather than transmitting a known origin.
- Evidence level: Mythological
Vergil, Aeneid (Augustan), and the broader Trojan foundation tradition.
The Penates worshipped in Roman households were the gods Aeneas carried out of burning Troy and brought to Latium.
Offered as explaining — Makes every Roman household shrine a continuation of the founding act of the nation and links domestic cult to the state Penates at Lavinium.
A national charter myth attached to a domestic practice. It gives the cult a pedigree; it dates and explains nothing.
Vocabulary
- lararium
- Modern standard term for a Roman household shrine. The word is unattested before late antiquity; the terms in use in the Pompeian period are sacrarium, sacellum and aedicula.
- Lar familiaris
- The guardian deity of a particular household, singular in the older texts, usually paired in the imperial-period imagery.
- Genius
- The generative divine double of a living man, honoured in the household and, for the emperor, in public cult. The equivalent for a woman is generally called her Juno.
- Penates
- Gods of the household store, invoked with the Lares; the state also maintained public Penates connected to the Trojan foundation legend.
- capite velato
- With the head covered by a fold of the toga — the standard Roman posture for sacrifice, and the attribute by which the central figure in shrine paintings is identified as the officiant.
- camillus
- In state sacrifice, a freeborn boy attendant carrying the incense box or the jug. The youth attendants in the shrine paintings are conventionally called by this name, but the paintings do not label them and the status requirement of the public role need not have applied inside a house.
- compitum
- A crossroads, and the site of the neighbourhood shrine of the Lares Compitales — the public cult that Augustus reorganised as the cult of the Lares Augusti.
- focus
- The hearth, which Cato makes the site of household observance, and around which many Pompeian kitchen shrines are placed.
Contested points
Are the Lares spirits of the dead — ancestors — or gods of place?
An older line of interpretation, drawing on Roman writers who connect Lares with Larvae and with the dead, treated them as deified ancestors watching over their descendants. Against this, Georg Wissowa argued that the Lares are gods of the farmland and of the boundaries and crossroads, and that the ancestral interpretation is a late confusion; Harriet Flower's The Dancing Lares and the Serpent in the Garden (2017) argues at length that the Lares were never ancestors and that the ancestral reading has distorted a century of scholarship. The dispute matters directly because it determines whether the household shrine was a place of contact with the family dead or not.
Who used the shrines in kitchens and service quarters?
One reading takes the frequent siting of shrines in kitchens and slave corridors as evidence of a distinct servile household cult, with the slave familia making its own offerings apart from the owners. The other holds that the kitchen simply contained the hearth and the fire, so the location reflects practicality rather than a separate congregation, and that the same shrines served the whole household. The evidence — position in the plan, and inscriptions by slaves to their master's Genius — is compatible with both.
What is the serpent in the lower register?
It has been read as the Genius of the paterfamilias in animal form, as an agathos daimon of Greek and Egyptian derivation signifying prosperity of the place, and as the guardian spirit of the location itself independent of any person. Where two serpents appear, some read a Genius and a Juno as a male-female pair. No painting labels the creature, and all three readings continue to be defended.
Where this comes from
Cato the Elder, De agri cultura 143
Forbids the estate manager's wife to perform religious rites, or to have anyone perform them on her behalf, without the orders of the master or mistress; then instructs her to hang a garland at the hearth on the Kalends, Nones and Ides and on holy days, to petition the household Lar so far as her means allow, and to keep the hearth swept clean each day before retiring.
Plautus, Aulularia prologue
The Lar familiaris of the house speaks the prologue in his own person, states that he guards the household, and says that the daughter honours him constantly with incense, wine or garlands, which is why he favours her.
Horace, Odes 3.23
Tells a countrywoman that if she raises her hands to the household gods at the new moon with incense, grain and a pig, her crops and vines will be safe — and that a small offering from clean hands placates the gods as well as a costly victim.
Petronius, Satyrica 60.8, with the atrium shrine at 29
Two boys set bulla-wearing Lares on the table while a third carries round a bowl of wine; the host gives the household gods joke names (Cerdo, Felicio, Lucrio) and an image associated with him is passed round. The silver Lares belong to a different passage, in the description of the shrine in his atrium.
Ovid, Fasti book 2 (the Lara episode, around 2.583-616; exact range medium confidence)
Narrates the punishment of the talkative nymph Lara and the fathering of the twin Lares by Mercury, presenting this as the origin of the household guardians.
Anonymous (Historia Augusta), Vita Alexandri Severi 29 (and 31); section numbers medium confidence
Describes the emperor's private shrine and the images kept in it, and is the source of the word lararium. Written long after the reign it purports to describe and widely judged unreliable, including on this passage.
How to read the evidence labels
- Evidence level: Documented
- Supported by a named primary text or by excavated material evidence.
- Evidence level: Probable
- Supported by strong inference that most specialists accept.
- Evidence level: Disputed
- Substantial specialist disagreement exists; competing readings are stated.
- Evidence level: Literary
- Belongs to poetic or mythological construction rather than to history.
- Evidence level: Mythological
- Belongs to shared religious or mythic tradition rather than to any single text.
- Evidence level: Unknown
- The available evidence is insufficient to decide.
These labels describe the status of the evidence, not our confidence in a conclusion. A claim marked Literary is not a claim we doubt; it is a claim about what kind of thing a poem is.