Patterns of time change only very slowly, as sociologists have shown. The same goes for one of the corner stones of our rhythm of life, urban or rural: mealtimes. The main meal of the day, for example, has moved from midday to the evening. Whereas most people of my generation use the word dinner to refer to the evening meal, my mother-in-law (celebrating her 90th birthday soon), consistently understands dinner to meal the midday meal – in the evening one has ‘tea’. Another novelty is brunch, this mid-morning meal, often eaten out or on holidays, that combines all the good things of breakfast and lunch.
In my research of Roman mealtimes, I have found some similar phenomena. In texts, we find two standard meals, referred to in texts the ariston and the deipnon. These daily meals were eaten in this order and are hence often rendered as ‘lunch’ and ‘dinner’. Dictionaries will also give the option of translating ariston as ‘breakfast’, but this mostly refers to the early archaic period, when the ariston was eaten in the early morning and the deipnon at midday. Just like in the late 20th century, the meaning of deipnon/dinner shifted by the end of the archaic period to the evening meal.
My research, however, looks at a second shift in late antiquity, and is based on papyri from Egypt. Let me start with a typical papyrological genre: the account. The short account below (P. Wash. Univ. II 98) dates from the fifth century CE and lists the daily expenditure of wine for consecutive days. I copy here the text of lines 5-7 from the edition (the only lines specifying meals) and translate them:
β εἰς ἄρ(ιστον) δῖπν(ον) δι(πλᾶ) β
γ εἰς ἄριστο(ν) δῖπνο̣ν δι(πλᾶ) β
εἰς δῖπνον̣ δι(πλοῦν) α
Day 2: for ariston (and) deipnon: 2 jars
Day 3: for ariston (and) deipnon: 2 jars; for deipnon: 1 jar
The observant reader may find something odd: why would the accountant give a combined amount of wine for lunch and dinner for day 3 and then an additional amount for dinner on the same day? The more logical practice would be to either separate the count for each meal, or to give only a combined amount.
The problem lies, I think, not with the accountant, but with the reading. The scribe wrote ‘ar. dipn.’ for day 2 and ‘aristodipn.’ for day 3. For each point marking an abbreviation in the transcription, there is an abbreviation sign in Greek. There is no additional abbreviation mark suggesting the ‘and’ implied by the translation, however. Therefore, I think that on both days, jars of wine were in fact served with an aristodeipnon, a ‘dinner eaten at lunchtime’.
 |
| Fifth-century invitation for a dinner around midday © Princeton University Library, original at https://dpul.princeton.edu/papyri/catalog/gb19f931j |
The term aristodeipnon is extremely rare, but not out of place in a late-antique context. It is also attested in the chronicle of Malalas (18.474 (71)); he mentions an aristodeipnon for military heroes hosted on 13 January, 532 CE by Emperor Justinian in his palace. Further confirmation that ‘dinners at lunchtime’ also make sense in the context of fifth-century Egypt is provided by P. Oxy. IX 1214. This is an invitation to ‘dine’ at the seventh hour, that is very shortly after midday and thus too early for a normal elite deipnon – which was served at the ninth hour in the high Empire, but moved somewhat later again in late antiquity.
The temporary fashion of aristodeipna or ‘dinners at lunchtime’ fits in the context of a late-antique overhaul in meal times. Up until the fourth century, the deipnon was the standard ‘main meal’. Public banquets, an important aspect of sociability in the ancient world, always revolved around the deipnon. In an upcoming article on the temporality of meals in Roman and late-antique Egypt, I argue that, from the fourth century onward, the ariston accrued a social function as well. At the same time, important features of the culture of commensality connected to the deipnon declined in the fourth century, as the cult associations which had organized public banquets disappeared, as did many festivals which had featured them. Other scholars, such as Hudson, have already argued that the end of public banquets coincided with a transformation in HOW people dined, namely with a decline in eating while reclining on couches. I would like to add that it also changed WHEN people dined. From the fifth century on, the ariston became the main meal. In late-antique papyri, it would therefore be more accurate to translate ariston and deipnon as ‘dinner’ and ‘supper’ than as ‘lunch’ and ‘dinner’.
Aristodeipna belonged to the intermediate phase in which the ariston had in practice already become the main meal, but people still remembered that ‘to banquet’ was ‘to eat deipnon’. In terms of anchoring innovation: the novelty of inviting guests for a midday meal was anchored on the practice of inviting guests for a deipnon. Once the practice had lost it novelty, they did not need this anchor anymore, and just spoke of an ariston, which had now acquired all the required connotations.
Sofie Remijsen (University of Amsterdam)
Sofie Remijsen is associate professor of Ancient History at the University of Amsterdam. Her research focuses mainly on the culture and society of the Roman Empire, including Late Antiquity. She is the director of a VIDI project funded by the Netherlands Organization for Scientific Research on “Lived Time in Late Antique Egypt “ which aims to explain how time and temporality, from annual festivals to what time to have dinner at, contributed to the shaping of multicultural communities in late antique Egypt.
Additional Reading
Hudson, Nicholas. Dining at the End of Antiquity. Class, Status, and Identity at Roman Tables (Berkeley: University of California Press, 2024).This piece is a revised version previously published on the website of Sofie’s NWO project Lived Time: https://www.livedtime.humanities.uva.nl/uncategorized/news/blog-please-be-invited-to-my-lunch-dinner/