believes that Galatians 2.11–4 assumes that a number of house churches existed in Antioch. Peter was accustomed to eating with the Gentiles. When challenged by the Jewish believers from Jerusalem, he began to have table fellowship exclusively with the Jews, a practice made possible by Christians gathering in different homes (Gehring, 2004, p. 112).
Antioch was the third or fourth largest city in the empire and had 18 ethnic quarters (Drane, 2011, p. 160). Jewish synagogues were separated according to ethnic background (Gehring, 2004, p. 113), and it is reasonable to assume that this was true also of the house churches. Ethnic and other social groupings would have lived in different parts of the city, and it would have been natural – if only for convenience – for each house gathering to have drawn in people from its vicinity and from its family and friendship networks.
Gehring does not argue this last point. Indeed, he believes that such separation was improbable because it would have created social barriers in a church that, from its beginning, had been anxious to pull barriers down. Yet as he himself points out (p. 113), the confrontation between Paul and Peter took place in front of the whole Antiochene church – ‘in front of them all’ (Gal. 2.14). This suggests that, as in Jerusalem, all the house churches gathered together from time to time. It was in this setting that people from different ethnic and social contexts most likely met together and expressed their oneness in Christ.
Rome
In Rome there was a similar pattern of scattered house-based churches. Robert Jewett has suggested that the bulk of the early Christians lived in the slum districts of Rome. Two-thirds of the names in Romans 16 indicate a Greek rather than Latin background, which suggests they were immigrants. Of the 13 names about which something certain can be said, nine have slave origin (Jewett, 2007, p. 63).
Christians in slum neighbourhoods would have met either in one of the workshop areas on the ground floor of a tenement block, or in a temporarily cleared space on one of the upper floors. The rooms on these floors were too small to accommodate a church, but – according to Jewett – the partitions between rooms were flimsy. Neighbouring families, each occupying one tiny room, may have temporarily removed the partitions to create a large enough space to hold perhaps 10 to 20 people,14 who would typically have lived nearby and whose social and ethnic backgrounds would have had much in common.
In the wealthier areas of Rome were luxury apartments and for the very rich entire houses. Some churches met in these parts of the city (Jewett, 2007, p. 64–5), drawing on different networks than the tenement churches. In addition to slaves and family members, the head of the household would have had a number of clients, reflecting the ubiquitous system of patronage. Some of these almost certainly would have attended the church.
There is no evidence that the house churches periodically met together, but they were clearly well networked. Paul could address his letter ‘to all in Rome’ (Rom. 1.7), which implies that the letter was passed from gathering to gathering. If they were not already meeting together, did Paul encourage them to do so when he arrived in Rome?
Paul’s churches
Certainly the churches that he founded continued the Jerusalem and Antioch practice of both meeting separately in homes and meeting together. ‘The household was much broader than the family in modern Western societies, including not only immediate relatives but also slaves, freemen, hired workers, and sometimes tenants and partners in trade or craft’ (Meeks, 2003, pp. 75–6). Patronage networks were important, too. New house churches, therefore, were ‘inserted into or superimposed upon an existing network of relationships’ (p. 76). In other words, they were largely comprised of homogeneous people groups, formed around hierarchical ties.
It is clear from 1 Corinthians 14.23, where Paul writes of the ‘whole church’ coming together, that these separate house gatherings periodically met as a single group, probably in a large home such as the one apparently owned by Gaius. In the concluding section of Romans, most probably written in Corinth, Paul sends greeting from Gaius, whose hospitality ‘the whole church here’ enjoyed (Romans 16.23). ‘In the Greek Old Testament this expression consistently refers to an assembly of all Israel; thus it must be the totality of Christians in Corinth which is in view’ (Banks, 1980, p. 38).
The meeting of the Ephesian elders in Acts 20 is suggestive. Luke says that ‘Paul sent to Ephesus for the elders of the church’ (v. 17). The reference is to the church as a single unit, a thought echoed in verse 28, where Paul urged the elders to keep watch over ‘the flock’, again singular. Evidently the house churches in Ephesus, which presumably emerged from the homes Paul visited in verse 20, had a sense of themselves as a citywide entity, a self-perception that would be natural if they periodically met together. The elders summoned by Paul may well have been the leaders of house churches who were used to assembling together (perhaps to plan the wider meetings).
Gehring maintains that recent scholars tend to agree that the early Christians gathered in two church forms: the house church and the whole church at any given location. While these two forms of church were geographically fixed, he suggests that the experience of individual gatherings could have been quite fluid. Congregations may often have lived in an in-between sphere hard to define (Gehring, 2004, p. 173).
New Testament believers appears to have found a way of combining homogeneous people groups with the potential to bridge social divides. This both/and approach held together sameness and diversity, small meetings and larger ones, and intimacy plus exposure to different ideas. Achieving this was far from easy, as the divisions at Corinth demonstrated, but Paul saw it as a priority.
For reflection
Some might say that Paul’s house churches were not strictly homogeneous groups. They contained for example slaves and heads of households, who were to relate to each other not in hierarchical ways but as ‘brothers and sisters’. Each house church modelled diversity. The only thing that members had in common was that they lived in the same area. Others might question the assumption that members drawn from the same area are not a homogeneous group. Don’t geographical divides often reflect social ones?
Is this a question about what counts as a homogeneous group? Or is it about whether, theologically, some forms of homogeneity are to be preferred than others?
Sustainable leadership
A key issue for church-starts is their sustainability, and central to this is encouraging local leadership. Does Paul’s practice contain lessons for how quickly and how best this can be done?
Delegation with support
Passing on leadership is a central theme in the writings of Roland Allen. An early twentieth-century missionary, Allen carried on a sustained polemic against the missionary methods of his day and contrasted them with Paul’s. His discussion sufficiently reflects the New Testament for Schnabel to commend him (Schnabel, 2008, p. 13).
Unlike missionaries who remained for several generations, Allen maintained, Paul never stayed in one place for more than a few months, or at the most two years. Once a congregation had been established, he selected
