Empire, Economics, and the New Testament

By Peter Oakes

Peter Oakes has long been recognized for his illuminating use of Greco-Roman material culture and social-scientific criticism to interpret the New Testament. This volume brings together his best work and introduces a substantial new essay that challenges current scholarly approaches to paradoxical teachings of the New Testament. 

Of special interest to Oakes throughout this book is the concrete impact of economic realities and Roman imperialism on first-century Christian communities meeting in house churches. To address this, Oakes considers an array of textual and archaeological resources from first-century non-elite life, including extensive archaeological evidence available from Pompeii. Readers will find here a deep trove of wisdom for understanding the New Testament in the context of the Greco-Roman world.

• Language: English
• Publisher: Eerdmans
• Release date: Nov 12, 2020
• ISBN: 9781467460033

Peter Oakes

Peter Oakes is Rylands Professor of Biblical Criticism and Exegesis at the University of Manchester. His other books include Reading Romans in Pompeii: Paul's Letter at Ground Level.

Book preview

PART I

HOUSE CHURCH

1 A HOUSE-CHURCH ACCOUNT OF ECONOMICS AND EMPIRE

Anyone of you who does not give up all their own possessions cannot be my disciple.

—Luke 14:33

If someone is hungry, let them eat at home.

—1 Corinthians 11:34

Let every person be subject to the governing authorities.

—Romans 13:1

And the ten horns that you saw, and the beast … will burn her up with fire … the great city that has kingship over the kings of the earth.

—Revelation 17:16–18

The surprising disparity among New Testament texts such as these has had a range of explanations. One is to argue that New Testament books have differing views on economic and imperial subjects. We can see this, for instance, in the work of Klaus Wengst.¹ A second explanation that has had long-term influence is Gerd Theissen’s classic sociological conclusion on the economic texts. He ascribes the differences to the effects of different groups among the Christians: itinerant charismatic preachers of the early Palestinian Jesus movement and settled Hellenistic house-church communities. He sees the economically radical texts as relating to the former groups, while many of the less radical ones relate to the latter.² A third explanation that has been prominent uses James C. Scott’s concept of hidden transcripts. It argues that many of the apparently conformist and subservient New Testament texts mask a radical underlying agenda, designed to be invisible to the authorities.³

This chapter argues that most of the diversity of New Testament rhetoric on economic and imperial issues can be satisfactorily accounted for by the situations of members of first-century house churches. Each end of the rhetorical spectrum on both economic and imperial issues makes considerable sense in first-century house-church contexts. Although not denying that there is some diversity of views among New Testament texts, some diversity of lifestyle among early Christians, and some textual presence of hidden transcripts, the diversity of New Testament rhetoric can be accounted for with far less need to appeal to such factors than has often been the case. Circumstances in the house churches make sense of both the more radical and the less radical rhetoric on economics and empire.

We will begin with an economic account of situations within first-century house churches, as represented by the model craftworker house church developed in Reading Romans in Pompeii.⁴ We will move from there to considering what relationships the members and groups would have to the Roman Empire. We will use these analyses to consider economic and empire-focused New Testament texts in relation to members of the Christian groups. We will end by considering structural issues underlying our arguments and conclusions.

ECONOMIC SITUATIONS IN THE MODEL CRAFTWORKER HOUSE CHURCH

The model craftworker house church is a device to help us understand socioeconomic locations, diversity, and potential hierarchies within first-century Christian groups, and to use that understanding for analysis of various issues and texts.⁵ The starting point for modeling a house church is the host’s household. Even just the exercise of writing down the list of people one might expect in the household of someone with space to host a reasonable-sized group immediately makes us aware that there is quite a range of social situations within a household, and hence within even the most basic house church. This itself should affect the way in which we analyze New Testament texts, alerting us to the fact that, for every New Testament text, there are important issues to consider about how it intersects with the lives of various types of people within the hierarchy of a first-century household.

The model then builds on the host’s household by asking both how many others could fit into a meeting in the house and who these others are likely to be. If the host has the largest house in the group, the other members (maybe typically two to three times the number in the host household: anything beyond about that implies an amount of unallocated space unlikely in a non-elite household) will predominantly come from families poorer than the host family: any odd group members from wealthier families would, on these assumptions, not be householders. The host household would probably all, to some extent and in some sense, be members of the Christian group meeting there. Other households joining the group could be complete or partial.

The above produces the essential shape of the model house church. Adding details from studies of housing and households from any of the parts of the Mediterranean world where we know of first-century Christian groups would produce a model of the same general shape as that below (with the main variations probably being in the social level of the host, the ratio of host household to others, and the presence or absence of slaves). Reading Romans in Pompeii gives specific substance to the model by focusing on craftworker house churches, a category directly evidenced by Rom 16:5 (cf. Acts 18:3), supported by, for example, 1 Thess 4:11. It then uses evidence from a town where there is direct evidence of craftworker housing and where scholars have been able to make some useful estimates relating to the socioeconomic structure of the population.⁶ Our model comprises forty people, as follows:⁷

craftworker, wife, children, a few (male) craftworking slaves, (female) domestic slave, dependent relative

several other householders (smaller houses), some spouses, children, slaves, other dependents

a few members of families with non-Christian family heads

a couple of slaves with non-Christian owners

a couple of free or freed dependents of non-Christians

a couple of homeless people

We will take the six sets of members in turn.

Group 1: The Host Craftworker Household

Many of the larger Pompeian craftworker houses could have been used to construct the model. Drawing on the major analysis of the Insula of the Menander led by my Manchester colleague, Roger Ling, and others, especially Penelope Allison,⁸ Reading Romans in Pompeii took the instance of the Casa del Fabbro (House I.10.7) as the focus for construction of the model. Although the Italian archeologists’ name for the house means something like House of the Smith, the very large number of loose finds of woodworking tools and furniture parts suggests that the principal activity there was cabinetmaking.⁹

The ground-floor plan size is 310 square meters, including a rear garden. There are also several upstairs rooms. Comparing this house with others in Andrew Wallace-Hadrill’s survey of housing in Pompeii and Herculaneum, the House of the Cabinetmaker has a larger plan size than 70 percent of other houses.¹⁰ On the other hand, it is only about 30 percent of the size of houses at the lower end of elite housing.¹¹ Most rooms are reasonably but fairly simply decorated. Two rooms have very high-quality decoration, probably predating the house’s use by the cabinetmaker.¹² The loose finds include a few pieces of jewelry: a gold necklace, a pair of gold and pearl earrings, two gold earrings, and two gold rings. Two bodies of people who died in the eruption were found. One had twenty-six silver coins and forty-eight bronze ones. The other, younger person had twenty-six bronze coins. Upstairs there were a further fourteen silver and ten bronze coins. A number of further decorative items of moderate value were also discovered.¹³

Figure 1.1. House of the Cabinetmaker (I.10.7) floor plan, with I.10.6 and part of I.10.4 (the House of the Menander) to the right. Differing shading relates to building phases, indicating location of prior connections between atrium of I.10.7 and Rooms 5 and 6 of I.10.4. Source: Roger Ling, The Insula of the Menander at Pompeii, vol. 1: The Structures (Oxford: Clarendon, 1997), detail from fig. 26 (drawn by S. Gibson, J. S. Gregory, R. J. Ling, D. Murdoch, with thanks to Roger Ling for permission to reproduce the plan).

One clue to the likely key economic relationships of the householder of I.10.7 is from blocked doorways (see house plan), which show that, at an earlier period, the house had been linked with the very large elite House of the Menander (I.10.4). A further clue to the potential power relationship involved with that link is that, about thirty years before the eruption, a small workshop and living space (I.10.6) was cut out of Houses I.10.4 and I.10.7.¹⁴ That took out a fairly insignificant corner of the House of the Menander but a very prominent part of the House of the Cabinetmaker. The new workshop looks likely to have been a dependency of the House of the Menander. All this adds up to a likelihood that the cabinetmaking householder rented I.10.7 from the owner of the House of the Menander. This is not a surprise. In the first century the great majority of property was owned by the elite. Extremely few craftworkers would have owned their workshops.

The cabinetmaker’s renting of his house linked him to the owner of I.10.4 in a relationship of economic dependency. The relationship probably had something of a patron-client nature to it, albeit maybe in informal terms. As Bruce Malina writes, across the first-century Mediterranean world, the relationship between landlord and tenant often took on the character of favor-giving dependency via events such as deferring payments in times of hardship and the conversion of some rental payments to loans.¹⁵ The relationship also meant that the cabinetmaker probably sought commissions for work via a network related to the owner of the House of the Menander. In return, the cabinetmaker might, for instance, paint up supportive election slogans if his patron ran for office, a pattern studied in depth by Henrik Mouritsen.¹⁶

Craftworkers would also have relationships with suppliers, in this case particularly for furniture-quality wood and material for inlays, and with a range of customers. One feature of I.10.7 is a bench just outside the front door. In other contexts, such benches are often viewed as being placed for the sake of clients of a householder.¹⁷ For instance, benches on either side of the entrance to I.10.4 would tend to be seen that way. This raises the possibility that the cabinetmaker might himself act like a patron to clients. This sounds strange. Certainly, the cabinetmaker is unlikely to have been formally a patronus. However, he did rent more space than 70 percent of Pompeian householders, and some others beyond his household are quite likely to have had some long-term dependency on his favor. One instance to consider is the poorer households within the house church. There could be a two-way effect in which clients of the cabinetmaker are likely to join the house church, an event that then reinforces the patronal link. One particular type of patronage that the cabinetmaker would be likely to exercise is that of acting as a broker, requesting favors of his patron (the householder of I.10.4) on behalf of people without a direct connection to him.¹⁸ One more type of link that craftworkers frequently had was in an association/club. That would give social links with others in the same trade, including periodic banquets and, often, contribution to a fund to pay for burial.¹⁹

Some householders were women, but the great majority of householders with sufficient resources to rent a fair-sized house were men. The cabinet-maker’s wife broadly shares his social status, as viewed from beyond the household, but her status within the household was lower than his and, usually, she would have few financial resources. At wealthier levels of society, a wife’s father could effectively place some constraints on the husband’s ability to control the wife’s resources²⁰ but this would generally not operate at the level of a cabinetmaker’s wife. She would usually have very little opportunity to use significant amounts of money except on her husband’s say-so. The wife would have a certain status within the household, with primary responsibility for domestic cultic activities, care of children, and management of domestic slaves (in a house too small to have a steward), although in this case there would probably be only one domestic slave, if any at all. The wife would not directly participate in the mechanisms of patronage or of a trade association. However, in a modest-sized craftworking family the distribution of tasks and of outside contacts would be more subject to cooperation and to variation of roles than would happen further up the economic scale.

Children in a craftworking family would work from quite a young age. There could, however, be elementary schooling.²¹ Children, like wives, would not usually have independent access to financial resources. This would no doubt differ in practice for grown-up children who were still at home. However, even for them, the father still officially controlled everything.

It is surprising to us that even quite small craftworking households often had slaves, mainly to participate in the craftwork. This is contrary to our expectation that any workers from beyond the family would be employed, rather than owned. A key dynamic in this pattern is that few of the slaves would have been bought (as adults). They would more frequently have been acquired directly or indirectly through abandonment (exposure) of infants. Raising a slave child cost money but produced both work and a saleable asset. Slaves could (like, in fact, other family members) by custom acquire money that the householder was expected to let them keep and use, the slave’s peculium, although technically it remained the property of the family head. The most well-known use of peculium was for the slave to eventually purchase freedom, in cases where that was possible.²²

The cabinetmaker’s household as a whole was no doubt frequently in a financially precarious situation, a circumstance endemic among the great majority of the first-century non-elite. Adverse occurrences could easily blow the household disastrously off course. The ten members of the household were dependent on the economic resources controlled by the householder—the economic resources of this small craftworking enterprise. These resources needed to be nurtured and used carefully in a precarious economic environment.

Group 2: Other Householders and Their Households

In our model, other householders and their households form the bulk of the house church, half of the forty group members (a quarter of the group members are the host’s household; categories 3–6 add up to the remaining quarter). As discussed above, all these householders rent smaller spaces than the house-church host and, while some of these households are entirely in the house church, other households are a mixture of church members and nonmembers.

The household head is still most likely to be male but is increasingly likely to be female as we go further down the scale of economic viability: for instance, households headed by widows are particularly likely to be facing financial hardship. Given the links of contact and friendship, the household are fairly likely to be engaged in similar work to that of the cabinetmaker: Richard Ascough pushes this scenario furthest, seeing the Thessalonian Christians as an association of leatherworkers who have taken on worship of Christ.²³ As argued above, some of these other households may have some economic dependency on the host householder. Links of dependency operate at all levels of society. In particular, the house-church host may be a broker giving indirect access to resources controlled by his patron. Some households joining a house church might gain some kinds of access that they had not had previously. Having said that, some householders other than the host would no doubt also have had other networks of dependence, on patrons and others, that did not run via the cabinetmaker.

Everything written above about the economic precariousness of a craftworker household applies more so to these households, craftworker or otherwise, further down the economic scale. The poorer the household, the more exploitative any patronage toward them was likely to be. At the lower end of the range, the households would be perennially in financial difficulty. Another factor that complicates matters for some of the households is that not everyone in the household is part of the house church. The householder is financially responsible for everyone but, in these mixed households, some people in the household are not party to any decisions and commitments—including financially risky ones—that the church members may make.

Groups 3–5: Members of Households with Non-Christian Heads

Our model has about a fifth of the house-church members from households with non-Christian heads, that is, about eight people. This represents the various New Testament texts that discuss the implications of being in this situation (e.g., 1 Cor 7:13–16; 1 Pet 2:18–3:6). The eight could, for instance, be made up of four people from one household, the head of which is not a Christian, along with four other members of different households: people who have joined the Christian group through contact with a friend or through hearing preaching. The precise numbers are not important, but the situation is one that we need to consider.

The economic considerations above mainly apply (but see an exception below). As the heading of the subsection signals, groups 3–5 of our model could be taken as a single category. They are household members and, as such, share, to varying degrees, the economic ups and downs of the household as a whole. However, that observation also shows one value of distinguishing groups 3–5. Members of the biological or adoptive family of a householder (group 3) would share the household’s ups to an extent that slaves and other dependents (groups 4 and 5) generally would not. There was also a reason for distinguishing groups 3–5 in the analytical development of the model. Members of a householder’s family (in an English sense, rather than the Latin familia, which means something more like household) shared, to a significant extent, the status of the householder. The four in the model were, roughly speaking, drawn from a random sample of Pompeians. Since the cabinetmaker has a larger house plan size than 70 percent of other householders, conversely 30 percent have larger houses than he does. This means that, in a random sample of four Pompeian householders, we would expect one, or possibly two, to have larger houses than the cabinetmaker. Similarly, in a random sample of four Pompeian members of householders’ families, we would expect one, or possibly two, to be from families with higher incomes than that of the cabinetmaker (assuming that biological and adoptive family size is constant across the economic spectrum, which Egyptian census evidence shows to be not unreasonable).²⁴ The point of all this is that our model has one of the forty house-church members being from a family wealthier than that of the cabinetmaker. That person’s family might well not be in the economically relatively precarious position occupied by the other thirty-nine people in the model. (New Testament texts such as Rom 16:23 suggest that there were also wealthier householders in some Christian groups. That would produce another model that would be interesting and valuable to analyze. However, this chapter will stay with the craftworker house church, as a model likely to be representative of more Christian groups.)

The particular complication for groups 3–5 is that they do not technically—and to a significant extent practically—have control over any economic resources. This has three direct effects. The first is that they are likely to be sharply limited in any financial contributions that they can make to the activities of the Christian group (e.g., meals) or to the support of other people, either within or beyond the group. Moreover, any money that they can give is likely to be at a disproportionate cost to themselves, since it will be from a rather limited reserve from which they may be permitted to spend. This would be starkest for slaves. If their owner allowed them to contribute to the Christian group from their peculium, every denarius that the slave donated could well be an extra period added to the expected period of servitude before they could gain their freedom by paying most of their peculium back to their owner.

This example also indicates the second effect, which is that, in many cases, the group member would need their householder’s permission in order to give money or goods to the house church. Of course, this is also part of a wider issue that the group members are likely to need their householder’s consent to participate in Christian meetings. Since the householder does not share the group member’s Christian allegiance or views, these issues of permission could be seriously problematic. In particular, the householder would have little motivation to permit the loss of some of the household’s financial resource—and hence security in a perilous economic world—to this alien religious group.

The third effect is that the group members are economically dependent on their householders. Actions contrary to the householder’s wishes could leave a group member cut off from economic support. A householder could oppose their actions either because of viewing the Christian groups as disreputable for one reason or another or, more simply, because the group member’s time and energy spent with the Christians took away time from their economically important activities in the household. A further reason for likely householder concern about participation in the house church is that it sets up an alternative allegiance (in Greek terms, an alternative structure of pistis: trust and loyalty).²⁵ A householder would probably be worried by a household member owing some allegiance to a group hosted by the head of another household. A householder may well also be worried by allegiance to Christ, who was not a figure in the household’s regular pattern of allegiances given to benefactors such as patrons and deities, including the imperial family.

Group 6: Homeless Group Members

There was no mathematical basis for settling on two as a number of homeless people in our model house church. The homeless leave little archeological or indeed textual evidence: certainly not enough to quantify them in the pattern of society in Pompeii or elsewhere. However, we know they were there. We also have little reason to exclude them from the house-church model, especially given the rhetoric of various New Testament texts (e.g., Luke 14).

People unable to afford even the lowest rentals, or who were excluded from settled housing for other reasons, existed on the most marginal and unreliable of economic resources and were vulnerable in a wide range of other ways too. Everything we have said above about precariousness applies to them in extreme measure, made even sharper by these people’s lack of access to the support networks of a household or of embedding in social structures of a neighborhood or of relatively settled social relations.

HOW MEMBERS OF THE MODEL CRAFTWORKER HOUSE CHURCH RELATE TO THE ROMAN EMPIRE

New Testament texts that relate to the Roman Empire in some way have many types of textual driver and are written for many different situations. A story about Jesus’s encounter with a centurion (Matt 8:5–13) may be driven by comparison between responses of Jews and gentiles to the Christian message. A lurid depiction of Rome’s downfall may be evoking motifs from a scriptural pattern of God’s dealings with wicked cities (Rev 17–18).²⁶ The situation in heavily colonial Roman Philippi lends a particular force to Paul’s description of his situation in the Praetorium (Phil 1:13).²⁷ The reading of various gospel passages to do with Roman institutions and characters will be affected by how the writing of each book relates to the events around the 70 CE fall of Jerusalem.

Factors such as these make inevitable some variation between Rome-related texts in various New Testament books. As footnoted above, I myself have written about some of these variations. However, our aim here is to consider how much variation in New Testament rhetoric is explicable just from the set of situations of people in our general model craftworker house church. Although it would be interesting to construct further models of Christian groups in, say, Philippi or post-70 Galilee, our current question is how much can be accounted for just by the socioeconomic situations of early non-elite house churches in general, rather than in particular places. As before, we will work though the groups of our model in turn.

Group 1: The Host Craftworker Household

The cabinetmaker’s patron (whether formally or informally) is probably the owner of the House of the Menander (I.10.4). When the cabinetmaker visits his patron and waits in the atrium, he sees what may be evidence of his patron’s loyalty to Rome. It is not evidence of explicit engagement with the political order of the day: there is no portrait of the emperor or anything similar. What there is is extensive depiction of the fall of Troy, a key event in Rome’s ideology: the point of origin of the family that founded the city and an important element of Roman identity. However, the cycle of Troy pictures here is also part of the house’s extensive collection of Greek mythological and cultural art, focused for the archeologists in the named portrait of the poet and playwright Menander. The collection expresses the owner’s taste and rootedness in culture. Also in the atrium is a very prominent lararium, a shrine to the Lares and other household deities. Further back in the