Excavations focused upon Greek housing has mainly been dedicated towards the oikos of the Prehistorical, Classical and Hellenistic eras and the subsequent interpretation of spatial segregation and architecture within. However, many Scholars readily admit that there has been far too little analysis of Greek housing during the period of Roman rule from the late Republican era of the middle to late second century BC and onwards into Imperial rule; the impression that is given is that it had turned into a ‘cultured backwater’ (McKay 1975: 211). The aim in this project is to see how much Roman rule influenced the development of Greek housing in the Aegean. I shall do this by analysing the Roman domus and domestic setting before comparing it with its Greek counterpart and then attempt to make light of the changes the oikos underwent.
Much of the evidence we have for Roman housing comes from the ancient literary sources. Pliny’s (the younger) letters contain a wealth of information about housing in the Early Imperial period. Andrew Riggsby’s work comments upon the fact that Pliny’s letters mention up to thirty-eight rooms distinctive rooms with thirty different room names but the description of room shape and content are minimal (2003: 170). Many of these room names are given in Greek as well as Latin. The use of Greek terminology may point towards extravagance and a sign of upper-class luxury (Riggsby 2003: 170). So what can Pliny’s work tell us about the way that different parts of the household were considered on or off limits to the outsider?
Modern thought tends to divide the household into definitive spaces, like ‘public’ and ‘private’; rather, we should try to consider a more qualitative aspect, such as what activities could be performed in a certain room and at what time (Riggsby 2003: 171)? Vitruvius distinguishes between rooms which common people were allowed into (communia) without invitation and rooms that required an invitation, such as the baths, dining rooms and cubicula (Treggiari 1998: 4). So, in his eyes, access to rooms depended on social standing. Cubicula are singled out in Pliny’s work as being isolated from the noise and activity from the rest of the house (Letters 2.17.22, 24; 5.6.21). Pliny’s exaggeration on this point implies that the cubiculum was a ‘secret space’ (Riggsby 1997: 43-47; 2003: 172). Therefore, it is plausible that different rooms were considered by way of how much privacy they could provide. That said, it may well be that Pliny’s cubicula happened to have been away from the activity in the house at that time. Spaces furthest away from the exterior were not always the most segregated (Grahame 1999: 59).
Pliny remarks casually that a wing of his villa was given over to his slaves and freedmen, which implies that they were thought of as part of the household (Letters 2.17.9; 5.6.30). There are also rooms where guests are allowed to sleep in and that friends dine with Pliny himself (2.17.9; 5.6.21). We could also consider those remarks in a qualitative way: different people will occupy those spaces during different times of day or year (Riggsby 2003: 175-6). Pliny describes his house as viewed by him, but the visitor would encounter certain places in which different activities took place in a set order (Riggsby 2003: 176). In some ways, we are able to view Pliny’s work as a visitor’s itinerary and there are similar instances in other works (See figure one (Grahame 1999: 83)).
The visitor is lead through the front door, into the house and ending in the most intimate rooms that they are allowed to see. Visitors are led through the rather mundane rooms in Pliny’s house (Letters 5.6.16) and through storerooms and horrea in Sidonius Apollinaris’ home (Sidonius Letter 2.2.9; Sidonius Carmen 22.169). The visitors then pass through more intimate rooms, such as Pliny’s library (Letters 2.17.7-10) and triclinium (Letters 5.6.21), whereas in Sidonius’ home, they pass through the richly decorated winter quarters (Carmen 22.187-193). The visit ends in the most special reception rooms. For Sidonius, this is his cenatiumculam with dining couch and shining dinner table (Letters 2.2.11) and for Pliny, this is his heliocaminus: a sun room (Letters 2.17.20-21), which shared views over the sea (this was also a way of saying to his guests that ‘all this is mine’, thus demonstrating what he wanted to show them (Grahame 1999: 84).
So what rooms in the Roman household offered different degrees of intimacy or privacy and when? Let us come back to the cubiculum (see above). A more specialised version of the cubiculum exists: the cubiculum dormitorium (Pliny (Elder), NH 30.52; Pliny, Letters 5.6.21), which defines this particular room for sleeping. However, the cubiculum in general hosted a variety of activities. There is also a distinction between the ‘normal’ cubiculum and those that housed the slaves, the cellae (Varro, Ling. 5.162; Vitruv. 6.7.2). The cellae, like their socially superior counterpart, probably were used for a variety of functions (Riggsby 1997: 54). It wasn’t rare for slaves to sleep outside their master’s bedroom door (Treggiari 1991: 416)!
Other than sleeping, there is the more sexual element. Sexual activity within the household was appropriated to the cubiculum; Seneca the Younger criticises the man who ‘does not even wait for the cubiculum’ when engaging in sex (Ep. 83.20). There was sex of the illicit variety too – adultery. If this took place in cubicula, then it was considered a violation of the household as well as a sexual offense (Riggsby 1997: 38), with the adulterer potentially originating from outside the household and breaking the bond of trust with the paterfamilias. It was also akin with the Classical Athenian to keep the female within the confines of the house in fear of being cuckolded and drawing the public eye into the private sphere of the oikos (Patterson 1998: 108). If adultery occurred, then the cubiculum was thought to be where it was carried out (Riggsby 1997: 38).
The cubiculum was also a place for guest reception and business transactions, as demonstrated by Cicero’s surprise that some parties are not admitted to cubiculae for meetings (Riggsby 1997: 41). In his oration against Verres, Cicero says that he is so perverse that the only person allowed in his cubiculum was the disgustingly smelly Apronius (Verres 3.23). Given this, it is of no surprise that the cubiculum was deemed an appropriate place to display paintings and sculpture for both the owner and for a ‘privileged’ audience. The presence of painting is at least verified by a multitude of high quality findings in many locations (Riggsby 1997: 38).
Cubiculae were often located immediately off the atrium (with or without peristyle). Cubiculae that did not open up onto the atrium often belonged to suites that did connect to the atrium (Riggsby 1997: 51). Large open areas, like the atrium allowed for a 360 view of all the rooms attached to them. The cubiculum was an area of confinement (Riggsby 1997: 51) and hence they were arranged so as to inhibit the confinement they produced. Therefore, the privacy that the cubiculum allowed depended on the time and place; it can’t really be designated as a wholly ‘private’ space.
In Classical Athens (and Classical Greece to some extent), the right to do what you want in your own home was a fundamental institution (Cohen 1991: 230). In stark contrast, Roman state and society did not want an unwatched private sphere that could bring potential trouble, at least, certainly for the houses of the elite (Riggsby 1997: 52). Hence, the control of the domus area allowed the paterfamilias to be able to control acts of secrecy that could possibly threaten him because in the Romano-Greco world, there was a public interest in the private lives of their leaders – Roman politicians were scrutinised on their home-life – the man who ran his home well would also run his office well.
It was the paterfamilias who determined how far visitors could reach in the house. The friendlier his relationship with the visitor was, the further he was allowed in and thus the more intimate setting the meeting took place in (Grahame 1999: 49). Differently decorated space indicated his social status to the visitor. More intimate rooms were more likely to be decorated and therefore were appropriate to show off to the closest or most important visitors. The house was a very important tool in impressing clientele (Hales 2007: 335).
So what about the rest of the Roman style domus? Vitruvius’ commentary provides a basic outline and we see that the Roman domus, both rural and urban, was symmetrical and lay upon an axis (McKay 1975: 32), so that the visitor, from the doorway, had a good view of the grandest rooms, the atrium, tablinium and the peristyle. The vestibulum allowed the visitor to take refuge from the street and/or to await the paterfamilias for morning salutation. Clients, who would come specifically for this and had to perform this ritual daily, as the paterfamilias was their patron (McKay 1975: 32). This sort of visitor was often poor and the architecture around him was rich; it was also as far as he would reach for much of the time (Ellis 1999: 81-2).
Vitruvius also mentions the triclinium which he considered to be a broad room (VI. 3. 8). The excavated House of Livia has three such triclinia and they were very similar to Vitruvius’ requirements (McKay 1975: 69). A number of houses also contain rear entrances and corridors to the triclinia so that visitors awaiting outside could not see the slaves bring food into the triclinia (Ellis 1999: 86).The triclinium also played host to banquets which were used by noble families to display their wealth via the extravagant food and surroundings (Dunbabin 2003: 13), it was the ideal opportunity to impress visitors of similar status. The environment spoke of the greatness of the host, with the food plentiful, the beautiful architecture and with artefacts reflecting the host’s learning (see figure two (Ellis 1999: 82). Literary sources from the first centuries BC and AD acknowledge that more personal dinners contained no more than nine guests, with three per couch (Dunbabin 2003: 39)
Most Roman houses provided a court/peristyle (despite special constriction (Ellis 1999: 85) and this area contained porticoes which provided a shade of privacy to an extremely open area. Despite this, visitors to the peristyle could be seen by the occupants. These two groups could never be isolated from each other within the household (Ellis 1999: 85). It was also an area of traffic in the circulation of the household, since the dearth of direct connections between rooms forced visitors to pass through controlled or watched areas (Ellis 1999: 86). However, the centrality of the peristyle was such that it was hard to restrict inhabitants to one part of the house (Ellis 1999: 77).
The atrium allowed the paterfamilias to receive his family, friends, clients and other visitors (Graham 1966: 17). Household doors were made extremely high and fitted with iron locks/bolts (McKay 1975: 32). Presumably, these would have been fitted in strategic locations and provided a dual function: to ‘remind’ visitors as to which areas were not accessible and to provide security against those on the outside who had ill intention. Marcus Valerius’ house was constructed so that his doors opened outwards and thus restricted door-callers from seeing inside (Plut. Publicola. 20).
So how did Greek houses from the Roman period compare to Roman households? Evidence for Greek housing during the Roman period, unfortunately, pales in comparison to the data we have for Roman housing; there are a lot less excavated examples (McKay 1975: 211). The lack of literary sources also makes it hard to gauge the effect of the Roman occupation on households and social relations within it (Nevett 1999a: 100). Our understanding, therefore, comes from houses with rich decoration and/or shared Roman features and so we might miss out on some houses because they lack features that we see in Roman housing (Nevett 1999a: 101).
The Classical Greek house was organised around a courtyard (to shut off domestic activity from the street) and contained the andron, which entertained friends of the house-owner and for drinking parties (symposion). The andron is identified archaeologically by ‘decorated walls, plaster/mosaic pavements, drainage facilities and reclining couch emplacements’ (Nevett 1999a: 102). The Hellenistic era house covered a larger area and contained a second court, so as to separate domestic and public functions into either (Nevett 1999a: 105). The rooms around the second court contained a number of decorated rooms. The domestic area of the house remained mostly undecorated, increasing the divide between ‘public’ and ‘private’ areas. The emphasis on the household graduated towards displaying wealth, as attested by increasingly elaborate mosaic decoration (Nevett 1999a: 105) and towards social interaction.