It would be churlish (as well as difficult), when my own work is treated so generously in this article, to object to its thrust too strongly. But agreement does not make for much of a dialogue! Let me state my agreements briefly, then.

1. 1. Versluys has nailed the terminological impasse: ‘Romanization’ is far worse than Romanization, because it has all the sins of the former without its conviction. But I have less sympathy for those TRAC speakers ‘ordered’ not to use the concept by their supervisors. If they can answer the many criticisms made of the concept, and make it work for them on their material, let them demonstrate this. If not, they need to find something better.

2. 2. Versluys also seems to me quite correct that some postcolonial approaches have often ended up in an unsatisfactory anti-colonialism. Yvon Thébert (1978) made a similar objection when he asked whether Bénabou (Reference Bénabou1976) had decolonized the history of Africa in the Roman period or simply turned it on its head. Denouncing ancient imperialism, colonialism and racial and sexual abuse might make us feel more comfortable, but it does not improve our analysis. I would add that it has also allowed British Romanists to return to a very traditional preoccupation: rewriting the Roman chapters of ‘our island's story’ in dialogue with contemporary imperial preoccupations.

3. 3. Versluys argues that we should ‘focus on “cultural transformation taking place in the context of empire” rather than on “imperialism and colonialism”’ (p. 8). This too makes very good sense. But I wonder what the word ‘cultural’ adds to this programme? Does it operate to exclude the study of other kinds of change (economic? technological? agricultural?). I doubt that this is what Versluys advocates and cannot see the advantage of arbitrarily demarcating one sphere of life as ‘cultural’ and excluding discussion of other changes. And I doubt that it would be possible to do this in any case. How would we talk about bathing without aqueducts, engineering and hydrology, as well as euergetism, notions of the body and foodways? Or about wine without thinking about techniques of agriculture, exchange systems and so on. If the abundant recent literature on entanglement – along with Hodder (Reference Hodder2012) I am thinking particularly of Thomas (Reference Thomas1991), Dietler (Reference Dietler2010) and Garrow and Gosden (Reference Garrow and Gosden2012) – has taught us nothing else, it is that we cannot easily separate ‘the cultural’ from the rest of life. Or does ‘cultural’ give holistic accounts of change a particular flavour? Or does it designate some particular Schwerpunkte for study? I have more sympathy with this position, but I suspect that it now obstructs more than it illuminates our projects. Now that ‘culture’ is no longer the final chapter of a book which has already dealt with conquest, administration, politics, the army, agriculture, manufacture and trade, town and country, and late antique decline (the traditional format of volumes in the genre ‘provincial history’), perhaps we no longer need to signal so strongly that culture is all-encompassing and can simply study together the whole set of changes with which we are concerned?

All the same, I am wary of signing up at once to Romanization 2.0. My commentary is an attempt to articulate my reasons for this reluctance.

For a start, as Versluys says, the search for a substitute grand narrative to occupy the same space as the one we can no longer live with has been tried several times already in relation to Romanization and to Hellenization. Sophisticated critiques of both concepts date back to the 1970s at least (eg. Gallini (Reference Gallini1973) and Goudineau (Reference Goudineau1979), the latter giving the title to Janniard and Traina (2006)). Arguably the Romanization debate has consisted of 30 years of expressions of dissatisfaction. Rejection of Romanization has become as much a habit as its use, and the case has been made on different occasions for ‘creolization’ (Webster Reference Webster and Cooper2001), ‘Mediterraneanization’ (Morris Reference Morris and Malkin2005) and ‘globalization’ and/or ‘glocalization’. To those studies of the latter listed by Versluys we can add Martin and Pachis (Reference Martin and Pachis2004), Sweetman (Reference Sweetman2007), Hitchner (Reference Hitchner2008) and Vlassopoulos (Reference Vlassopoulos2012). Yet in many – not all – cases what these usages offer is not so much a new explanatory framework as a new descriptive one, and one that tactfully leaves undamaged most earlier scholarship as well as the many interpretations that have grown up in interdependence with notions like Romanization. The point that we often seem to be substituting new words for old ones is eloquently made for Romano-British studies by Millett (Reference Millett2007). If certain kinds of postcolonialism created histoire inversé – the same old narrative with the moral valencies reversed – then other new labels simply sanitize discredited brands. If so, then saying ‘Mediterraneanized’ is just as much an evasion as putting ‘Romanized’ in scare quotes. Naturally I hope to be proved wrong about all this, and I look forward in particular to Pitts and Versluys (Reference Pitts and Versluys2014), which promises a much more rigorous appraisal than we have had to date of the costs as well as the advantages of employing globalization theory in the study of antiquity.

So let us remember that there are alternatives. Most obviously we could simply abandon the search for a grand narrative of change – cultural or not – consequent on, and contained within, the Roman oikumene.

This is not intended to sound pessimistic, and is certainly not an attempt to foreclose debate (one that would certainly fail). But it is intended to sound a note of scepticism about the prospects of developing any new grand theory that will successfully achieve all the goals for which Romanization theory was eventually evoked. Romanization theory was asked (among other things) to provide:

1. 1. an account of how and why material culture and social practices changed in areas controlled by Rome,

2. 2. an account of how cultural change was related to imperialism,

3. 3. an account of the formation of identities in Roman provinces and

4. 4. a contribution to our understanding of the ideological foundations of Roman power and the means by which the consent of Rome's subjects was either won or engineered.

Just possibly, asking one concept to do all of this was a bit ambitious. And in practice different studies have given different weight to different parts of the programme. So Millett (Reference Millett1990) focused mainly on interpreting the material record to reconstruct processes of economic and social change, while Mattingly (Reference Mattingly2006) aimed at rewriting a conventional narrative in ways that cohered better with modern scholarship on imperialism, and so on.

It seems to me that those who do not want to sign up for Romanization 2.0 have at least two strategies available, both of which look quite promising.

The first is to devise a series of less grand (or less grandiose) theories, each one directed more precisely at a smaller target. A good model is provided by a series of recent studies focused simply on identity formation, among them Goldhill (Reference Goldhill2001), Roymans (Reference Roymans2004), Howgego, Heuchert and Burnett (Reference Howgego, Heuchert and Burnett2005), Derks and Roymans (Reference Derks and Roymans2009) and Whitmarsh (Reference Whitmarsh2010). Many of these combine material from the East as well as the West, just as Versluys advocates. It is also striking that for this particular set of questions none of them attempts a purely archaeological debate but each instead combines epigraphy and numismatics, art history and philology, history and archaeology. That approach seems appropriate as well as successful. Indeed the very idea of trying to limit an analytical strategy so as to cohere with the contingent and contemporary limits of a modern academic discipline seems a bit bizarre.

A second alternative is to try to develop understandings of change that do not depend on (or at least defer to a later stage of the argument) the creation of higher-order abstractions like Romanization. This approach has been recently advocated in this journal on the basis of actor-network theory (Van Oyen Reference Van Oyen2013). Rather than starting from the conviction that there must be something like Romanization out there if we only had the ingenuity to find and describe it properly, we could simply look at how things are made, used, exchanged and consumed and see what patterns emerge.

Chris Gosden (Reference Gosden2005, 209) offers one such formulation in relation to Roman Britain that is worth considering, especially for those wishing to begin from things rather than texts:

This formulation does not provide Gosden with the basis for a new processual account of change – his interest in this essay is in how objects work on people, rather than in what they stand for or represent, or what they can be used (instrumentally) to achieve for the benefit of imperialist occupiers, profiteering merchants, collaborative elites, sexually rapacious soldiers or resistant natives. My point is not that issues of dominance and resistance should not be addressed – of course they should. But there are also other questions that may be asked about the mutually implicated transformations of material and social worlds, before, during and after episodes of conquest and contact. Many of these questions can be posed and answered without recourse to Romanization, creolization, globalization or any other high-order concept.

But not everybody wants alternatives. Let me finish with some comments on the contemporary use and value of the Romanization debate, and on who wins and who loses from its prominence.

One of the values of Romanization as a concept is that via one or another of its many concerns it has the potential to bring so many of us together. Versluys began his piece with the excellent point that ‘the Romanization debate’ is very local, something almost entirely confined to Roman archaeologists working in the UK and the Netherlands (and, one might add, to particular generations in those two countries). The wide-ranging nature of the discourse of Romanization allows a range of positions and specialisms to be accommodated. In addition, the very diffuseness of which we often complain prevents the discipline becoming fragmented into doctrinaire factions. Those involved might even be thought of as what Stock (Reference Stock1983), writing about medieval monastic scholarship, calls a textual community, a group united by their common devotion to the exegesis, commentary on and critique of a small number of key texts. As with biblical exegesis this is not really a zero-sum game; there is no real risk that one side will lose or win, or that any of the core questions will be definitively solved. Either we will go on forever or we will lose interest. No other outcome is very likely.

Versluys is quite right to say that ‘every scholar has his or her own view on the historiography of the subject’ (p. 4). Let me offer one of my own. What the Romanization debate has done is to install a new canon of reference works and key concerns, one that replaces a common familiarity with a group of Latin and Greek texts and ancient historical problems (cruces) referring to the northern provinces. The Romanization debate, and the texts, conferences and papers through which it is reproduced, thrive precisely because they serve the needs of a generation of Romanists who, by virtue of their training (mainly in archaeology departments) have much more to do with archaeologists of other periods, and much less to do with classicists, than did their predecessors. This is not meant as an allegation of sectarianism. It is simply that if one lives, works and teaches alongside specialists on Bronze Age and Iron Age Europe or the empires of the Inka, if one goes to seminars and examines papers and theses on a range of archaeological rather than classical topics, one may come to think of the Roman period in a different way than if one works alongside those teaching Virgil's Aeneid and the history of the Augustan Principate. It is a generalization, of course, but Roman archaeology in the UK at least is now far more often found in archaeology departments than Classics ones. This has transformed the field for the better in all sorts of ways. But it has meant that Roman archaeologists have needed a new intellectual common ground and new canonical texts. This, I suggest, is what the Romanization debate has provided. If I am right then we already have ‘an archaeological perspective on Romanization’ (p. 15), as Versluys puts it, although not in the sense in which he means it.

My guess – a prediction rather than a prescription – is that the capaciousness of the Romanization debate means that it will continue to be able to accommodate approaches based on globalization and material culture, and also the ‘institutional archaeology’ advocated by Gardner (Reference Gardner2013) and the historicizing agenda which Terrenato (Reference Terrenato, Hurst and Owen2005) argues is needed to clear the ground of unhelpful analogies between ancient and modern imperialisms. After all, it has absorbed a very wide range of postcolonial approaches over the last decade without much trouble, and the institutional conditions within which Roman archaeology is taught and studied have not changed recently. ‘Let us continue the debate!’ as Versluys declares in his acknowledgements.

But maybe we should pause for a moment and consider the costs, as well as the benefits, of prolonging the conversation, comfortable and familiar as it is to many of us. Those who bear the heaviest costs are the new entrants to the debate who encounter an ever-growing bibliography of deuterocanonical and exegetical works that must be mastered before they can be full members of the textual community. Perhaps this is one reason for the discomfort Versluys noted among speakers at TRAC? And then there are costs too, which we all bear, if we devote energy to repairing and refitting the vast bulk of Romanization theory rather than exploring some of the alternatives available.

Romanization 2.0 sounds wonderful, with glittering new globalization and a host of additional features imported from entanglement theory and the like. I am definitely looking forward to the release. But I also feel I should get out of my comfort zone and take a good look at other opportunities before I take out a subscription.