Michel Reddé (R.) tackles half a millennium or so of change, from roughly 300 BCE to roughly 200 CE, in his remarkable, dense, but readable personal view of the evidence and arguments for the period. Though the two-word main title encompasses the whole of Gaul beyond what became Narbonensis, the next four words show that the focus is really on what was to become Belgica and northern Lugdunensis, essentially today's France north of the Loire along with Belgium. The areas to the south are mentioned when helpful, so also is the Rhineland (technically part of Gaul down to the 80s CE) making use of the German-language literature, with which R. is very familiar (he is also well up on the English-language literature). To mention France in the same sentence as Gaul, as above, would provoke R.'s ire, since one of his aims, set out in the introductory Prologue, is to divorce Gaul from France, undoing the tradition established by Camille Jullian in his lecture of 1919 to the Collège de France, which R. sees as imposing modern nationalist agendas that cloud perceptions of the period, as exemplified in the final paragraph of the book. For Jullian, of course, text was the foundation for study of this period, above all Caesar's self-serving account of his conquest of the Gallic peoples in the 50s BCE. In the last 50 years, the quantity and quality of archaeological research has claimed for that discipline a voice at least as powerful as Caesar's, in fact probably more powerful, and one that can only get more insistent and independent. This has been driven by development, some of it state-engendered grands projets such as autoroutes, canals and TGV lines, some of it the result of commercial developments such as the ZACs and ZUPs and the rest of the alphabet soup beloved of French development bureaucracies. The data that this has all yielded have transformed the quantity and range of the evidence base and of the frames of reference and thought. R. himself has been a central figure in this process, especially through his major European Research Council-funded “Rurland” project, which sought to create a model for analysis and synthesis of a wide range of evidence types across northeastern Gaul.Footnote ¹ The historical hinge around which this period turns is obviously the Caesarian invasion of the areas of Gaul beyond what was then Transalpina. The archaeological hinge, or hinges, as becomes clear, either lie earlier, in the 4th and 3rd c. BCE, or they lie later, very often beyond the turn of the millennia, so that the creation of what we think of as Roman Gaul can seem to have little to do with the conquest. The work under review offers a wide-ranging conspectus of the archaeological evidence and of how it is transforming the subject and will continue to do so for the foreseeable future.

The book is a credit to the Presses of the University of Rennes. A major strength is its extensive use of high-quality illustrations, many in color. Several of these show off the results of the major excavation projects of the last half-century, with the extensive uncovering of structures and settlements and areas of landscape. Graphs, tables and pie-charts (to English-speakers, to Francophones camemberts; differences in national cuisines cannot be avoided) supplement this information. The text is clearly laid out and frequent subheadings divide what might on occasion be long and dense passages of evidence and argument into digestible sections. That said, this reviewer has two presentational gripes. One is a preferred chronological expression. Though calendar dates are given often enough and phases of sites sometimes situated in relation to periods such as the Gallic Wars or the reign of Augustus, all too often what is given is an alphanumeric subdivision of the “La Tène” period, such as La Tène D2b. To those not adepts of these mysteries, such formulations are simply an aid to incomprehension. In fact, this periodization can perfectly well be hooked onto absolute dates or events (La Tène D2b it transpires is “the period after the Gallic Wars”). Far better to go with calendar dates or, if necessary, periods. The other gripe is that there is no index, a great pity since the book is so full of sites and topics and arguments which the reader might want to revisit. The bibliography, though, is an impressive resource.

The first of the two Parts into which the book is divided concerns itself essentially with the nature of the archaeological evidence, in four chapters: Ch. I, Le pays et les hommes; Ch. II, Un monde de paysans; Ch. III, Le phénomène urbain; Ch. IV, Des dieux et des hommes. The main focus throughout is on structures, sites, and settlements, with which R. clearly feels most at home.

Obviously, some of the major contributions of all the work over the past decades have related to the enormous amount of other evidence types yielded under modern excavation conditions, principally paleoenvironmental and artifactual. This has gone hand-in-hand with a commensurate expansion of the range and types of analyses and interpretations derived from such evidence. Paleoenvironmental results do figure, particularly in Ch. I and, to an extent, Ch. II. Artifacts figure much less. Even the numerous and well-studied coinages get little attention compared with what is usually the case in works of synthesis on this period. R. is well aware of all these classes of evidence and uses them on occasion, but for reasons of space and inclination he has chosen not to consider them in detail in this work. It does, though, mean that apart from named, high-status individuals, the Gallic peoples tend to be treated very much as a mass, undifferentiated by age, gender, or other characteristics.

Chapter I sets the scene, opening with a longue durée consideration of the climate and its variations since the last retreat of the ice, though with a focus on the moyenne durée of the period covered by this work. This is followed by summaries of the development of soils and of vegetation, all this dependent on palynological, pedological, and other proxies. Humans then intrude, with discussion of possible population parameters. The rest of the chapter moves onto essentially text-based topics, such as the names of Gaulish peoples and their geographical organization. This raises the acute question of what such “peoples” were. Influenced by Roman administrative divisions we tend to think of them in territorial terms, but this may be a misapprehension. Such episodes as the migration of the Helvetii suggest that territory was less important than social and political identity and cohesion. Given the scale of some Roman-period “rural sanctuaries,” religion and cult may have been more important than we recognize.

Chapter II traces the development of the countryside, its settlements, crops, and herds over the second âge du Fer (the second half of the 1st millennium BCE) based on a corpus of 750 sites (which in itself gives some idea of the data explosion in French archaeology). This was a period that saw significant deforestation, the introduction of new crop species, and across many regions studied, an increase in settlement numbers, with a peak from 200–100 BCE, then a decline from well before the Gallic Wars: not all change can be laid at the door of the balding proconsul.Footnote ² In some regions the archaeology is better documented than in others: for instance, Picardy or the areas comprised within the Rurland project. But as tends to be the case with such exercises,Footnote ³ there is more variation in density of sites, settlement morphology, building types, or agrarian regimes than there is uniformity, and the reasons for these variations remain very difficult to apprehend in the present state of understanding, likewise, the trajectories of development of individual sites or of regions. R.'s attempt at a synthesis of the continuities and changes from about 200 BCE down into the imperial period is a very thoughtful exploration of themes such as agrarian change, settlement mobility and building forms, but with a lively awareness of variations on these themes. Broadly, it is difficult archaeologically to see the Roman conquest in most rural areas. Continuity and change continue to operate apparently to their own rhythms rather than due to the irruption of the Roman army, though the army's demands in men and matériel may have changed the economic framework very significantly. Only in the Augustan period does the pattern begin to change, with the start of the climb in numbers of sites towards their peak (and subsequent decline) in the second half of the 2nd and the first half of the 3rd c. CE. The period from Augustus to the Antonines also sees the emergence of what is generally taken as the emblem of a new dispensation, the “villa,” and its possible meanings for land ownership and management. The classic area for this is of course Picardy, with the characteristic form, first identified from the air by Roger Agache in the 1970s, of a rectangular or trapezoidal complex with the residence at one short end and relatively long alignments of subsidiary buildings down the long sides, a form present also more widely across the northern half of Gaul.Footnote ⁴ Archaeology has subsequently shown that the long, rectilinear enclosure was well established before Caesar,Footnote ⁵ but it has also shown that the Roman-period structures came well after the conquest, often not until the 2nd c. CE and in new architectural forms and materials, so the story was more complex than the simple “Romanization” of the pre-conquest form.

Chapter III turns to matters “urban.” As R. says, the traditional model emphasized the ill-defined Caesarian term “oppidum” as a political center and extensive religious, residential, and economic focus, and its relationship to a Mediterranean-style successor. Perhaps the classic example of this was the transfer from the upland site of Bibracte/Mont Beuvray, with its imposing murus gallicus defenses, to the lowland, orthogonally planned site of Augustodunum/Autun, with its Roman-style walls and gates. In other cases, the oppidum remained static, with the “Gallo-Roman” civitas-capital succeeding the late Iron Age oppidum; the classic case of this is perhaps Avaricum/Bourges. Individual cases such as this, and the general Roman-period pattern of territorial civitates each endowed with a single urban center or “capital,” has led to this pattern being back-projected into the pre-conquest period, with territorial “tribes” (a term derived essentially from 19th-c. ethnography) each having a single, dominant oppidum, and thus transitioning unproblematically into the Roman system. But Caesar also mentioned vici, sites to him inferior to oppida yet clearly still important foci of population and activity. Archaeologists have created the class of “open” oppida, such as Aulnat or Levroux, which seem to have shared many features and functions with the classic oppida, save for the lack of defenses, though some lay on hilltops. Thanks to recent survey and excavation it is now appreciated that open agglomérations were common over most of Gaul, with evidence for non-agrarian activities.Footnote ⁶ But as with rural settlements, the more one looks and the more examples that are found, the more variety becomes the norm, making it very difficult to generalize about their social, economic, and religious functions, which differ by region. Importantly, they may also differ by date, with particular sites rising and falling in size, complexity, and presumably importance. In R.'s discussion, defenses, generally extensive and elaborate, remain definitive for his notion of what constitutes an oppidum. Within these can be found areas and structures indicative of “public” and religious functions, the very important recent excavations at the hilltop site of Corent (? the principal focus of the Arverni pre-conquest) with its religious enclosures and “proto-theater” furnishing an excellent example for R. (though its extensive evidence for ritual structures and activities is hived off into the following chapter, which is a pity). These sites are also characterized by being foci of population and of economic activity, receiving considerable quantities of Mediterranean goods, particularly wine amphorae.Footnote ⁷ By the time of the Caesarian conquest, many regions also display a range of oppida and agglomérations, suggesting complex relations rather than perhaps a simple hierarchy dominated by a single oppidum, as traditionally imagined. In the vicinity of Avaricum/Bourges were four other large sites with a murus gallicus and two with dump ramparts. Onto this spectrum of possibilities for pre-conquest focal (central?) places was grafted the network of “Roman” administrative cities – the familiar (to us) chefs-lieux de cité/civitas-capitals – with their orthogonal street grids, forums, basilicas, baths, amphitheaters, and so on.Footnote ⁸ There were foundations ex nihilo, but these tended to be in peripheral and relatively underdeveloped areas such as the Atlantic façade. For most of Gaul, it was more a process of choice and adaptation within the existing Late Iron Age networks of major settlements. R. sees as crucial to the development of chefs-lieux the “Roman” strategic road network, often a regularization of existing routes. Many sites lay away from these roads and failed to develop into political centers, others lay on them and prospered. In R.'s view, the abandonment of Bibracte in favor of Augustodunum came about not through the authorities’ distrust of its defended situation nor a wish to erase the Gaulish past, but simply because it was at a distance from the new transport axes, of which Augustodunum could take good advantage (though there is a certain amount of chicken and egg here as between roads and city). These axes were also crucial to the development of the agglomérations secondaires, the “small towns” of the anglophone literature. R. shows that, despite their name, many of these places were extensive and, as at Jouars-Pontchartrain, some developed a pseudo-street grid. Many also had versions of forums, for instance, Alesia. It would have been nice to see more about the role of the huge temple complexes generally distanced from the civitas-capital. Vandeuvre-du-Poitou is illustrated, but sites such as Allonnes close to Le Mans (for the Aulerci Cenomani), Sanxay (for the Pictones), or the astonishing Vieil-Evreux (for the Eburovices) would have added an important religious dimension to the choices about “urbanism” and the wider roles of buildings such as baths and theaters. One last message from this chapter is that the increasing evidence for dates suggests that many cities (e.g., Tongeren/Tongres) did not look very “Roman” for much of the first century of their existence. It was from the Flavian period on into the 2nd c. that many developed their monumental equipment, later than has often been thought.

Chapter IV considers the study of religious phenomena, or more accurately epiphenomena, since what we mostly have is sites and materials used in the expression of religious beliefs but hardly ever do we have a grasp of the beliefs themselves that framed the use of these things.Footnote ⁹ The chapter again focuses on sites, though with much more consideration of objects and osteologyFootnote ¹⁰ because of their important roles in what anglophones would now call “structured deposition” at many of these sites. This does, though, mean that there is only limited consideration of deposition away from sanctuary sites, for instance, “wet place” deposition, or of religious evidence from domestic sites. Nor does burial archaeology get discussed, save for a few pages in Ch. V and in the Épilogue, despite the clear and datable impacts of Mediterranean material and practices across Gaul from before Caesar.

Over the last generation and more, perceptions of pre-Roman religious sites and practices in Gaul have been framed by two sites in northern Gaul, Gournay-sur-Aronde and Ribemont-sur-Ancre, geographically proximate and excavated one after the other.Footnote ¹¹ These gave an image of enclosures defined by ditches, with internal structures and evidence for “structured deposition” of human and animal remains, weaponry, coins, and other material – above all, the remarkable ossuaries or charnels of dismembered human and horse remains along with weaponry at Ribemont, dating from the 3rd c. BCE and interpreted as offerings or trophies after battle. At both sites, the Roman period saw the construction of a Gallo-Roman/Romano-Celtic sanctuary, that at Ribemont being especially extensive and complex (and crossed by trenches of the 1915 Battle of the Ancre with a nearby war cemetery: plus ça change…).

Gournay and Ribemont offered a range of Iron Age activities at ditched enclosures containing no “temple” structure but showing evidence for “ritual” behavior in the form of offerings of human and animal remains, metalwork (including armaments), ceramics (feasting?), and coins. This became the model for pre-conquest religious activities, but though some similar elements were found at a small number of sites, they were all in what would become the province of Belgica. Further afield, important well-excavated sites such as Mirebeau, Nitry, or the major oppidum of Corent had ditched enclosures, internal timber structures, and deposition of objects and human and animal remains, but differed far more from the structures, materials, and practices of the Gournay/Ribemont model than they resembled them. But ditched enclosures, internal structures, deposition, and often a succeeding Roman-period temple show a degree of similarity, though this may well risk focusing too much on a particular type of site at the expense of others.Footnote ¹² Of course, individual sites developed through time, as did practices at them, with, for instance, the later introduction of votive offerings including model wheels and miniature weapons.

The Roman conquest for a long time saw a shift in how religions were studied by antiquaries, then archaeologists and historians. One reason for this was the visibility of masonry structures, but perhaps even more important was the introduction of the “epigraphic habit,” with inscriptions naming divinities, sometimes assimilated to better-known Roman ones. This, along with statuary and other representations, transformed the supposedly “atectonic and aniconic” Iron Age practices.Footnote ¹³ R. lays out the evidence for these sites, and divinities and cults, emphasizing the changes in aspects such as the types of material offered, how it was offered, and what this may mean. At Ribemont-sur-Ancre, the Roman period saw the sanctuary elaborated by the end of the 2nd c. CE into a huge complex stretching for the better part of a kilometer down the hillside. All in masonry, and with Roman-style architecture and architectonics, a massive temple sat at the summit, then a succession of courtyards and open spaces flanked by buildings, with a theater and large baths on axis. This equipment resembles that at sites such as Sanxay or Vieil-Evreux, the huge “rural sanctuaries” of northern and central Roman Gaul. There are sufficient of them now known that one may argue they show that many Gallo-Roman civitates were bi-polar, with sites such as Ribemont complementing or rivalling the Roman-style urban centers, in this case, Samarobriva/Amiens. What does this tell us about how their inhabitants saw civitates as working? How much were they religious as well as politico-administrative constructs?

The intellectual temper of our times is to privilege continuities and the contributions of indigenous forms and practices to what came after. For Gaul this received huge impetus from the discoveries at Gournay and Ribemont, with the subsequent discoveries of other pre-conquest sanctuaries. Nevertheless, though a map of east-central France (Fig. 116) shows many Roman-period sites with Iron Age antecedents or material present, it also shows there were substantial numbers of rural temples with no such evidence. Religions and religious practices and manifestations continued to develop to a rhythm and for reasons which need not have depended on what went on before the conquest.Footnote ¹⁴ Moreover, the experience of the worshipper at Ribemont or elsewhere in the 2nd c. CE as regards the physical appearance of the site and the beliefs, the rituals, and the other practices that went on at it would have been utterly different than in the 2nd c. BCE. Perhaps we should be paying more attention to discontinuities and innovations, and to what they may mean for the beliefs and practices of religions.

Like Part 1, Part 2 also comprises four chapters: Ch. VI, La Gaule indépendante; Ch. VI, La Gaule conquise; Ch. VII, Le temps de l'organisation; Ch. VIII, Les soubresauts de la «romanisation», along with a concluding Épilogue. The brief Chapter V examines two aspects of pre-Caesarian Gallia Comata. One is the increasing involvement of some Gallic peoples, notably the Aedui and the Arverni, with the Mediterranean world and Roman politics. This is particularly evidenced by the spread from the 2nd c. BCE of quantities of Italian wine amphorae into the interior of Gaul and up to the Channel. This intensified into the first half of the 1st c. BCE, as shown by Fig. 136. The importance of access to and control of wine for social hierarchy and reproduction has been much explored over the last 40 or so years and has become a familiar trope for the period.Footnote ¹⁵ Clearly this, at least in part, propelled the rise of major groupings and dominant actors in Central Gaul in the half-century prior to Caesar. There is also evidence from the coinage types and inscriptions in these regions for the influence of Roman practice and personalities. The other aspect considered is the nature of Gallic societies on the eve of the Caesarian conquest. The traditional view, derived from a passage in Gallic Wars Bk. VI, is of a hierarchical society of kings, nobles, and the rest, one recognizable to Caesar's aristocratic Roman audience. R. is clearly attracted by the innovative interpretation proposed by Stéphane VergerFootnote ¹⁶ that rather than describing stable, hierarchical societies in these regions, it in fact gives away their segmentary character, with individuals and groups and their territories and centers rising and falling as they contest with each other. Such an interpretation of the texts would accord with the archaeological evidence mentioned above for Ch. III, with the multiplicity of settlements such as defended and undefended oppida and the various agglomérations.

Chapter VI is also quite brief and looks at the generation or so between the close of the Caesarian conquest and the formal restructuring of the administration of Gaul by Augustus and Agrippa. Over much of Gaul, the wars of conquest were bloody and profoundly shocking to the established orders. Some peoples, most famously the Eburones, simply disappear from the record (such as it is) after what seems to have been a Caesarian ethnocide. Others lost their autonomy, for instance, the Mandubii of Alesia, or were relocated, as with the Ubii, who were moved wholesale from one side of the Rhine to the other, all as suited the new rulers. Roman hold over Gaul in these decades was by no means secure, especially with Rome itself distracted by the Civil Wars down to Actium in 31 BCE. Evidence for the location of Roman military formations in the aftermath of the conquest has long been scanty. As more emerges, it seems that several were cantoned in oppida, such as the Titelberg (Luxembourg), presumably amongst other things to ensure the compliance of those subject to the political and religious authority of such places. The tenuous reality on the ground of Rome is also demonstrated by the very weak penetration of Roman coinages into the north of Gaul. By contrast, there were coinages struck by Gallic aristocrats, sometimes imitating issues by particular Roman nobles. Surprisingly, these were not all of the Caesarian party; the well-known ARDA emissions of the area of the Treveri initially copied Roman types of the Pompeian party, suggesting considerable knowledge of internal Roman politics and the ability of a Gallic notable to establish patronage networks. Even so, his latest strikings imitate Caesarian models. So immediately post-conquest Gaul's general stability, a few revolts apart, depended as much on the manipulation by both sides of personal links as on direct military presence. The impact of the conquest on the bulk of the population is harder to assess, though Mediterranean wine amphorae and other ceramics continued to arrive.

Chapter VII is the most substantial in Part 2 and deals with the transformation of post-conquest Gaul into three Roman provinces, each with its internal structure of civitates. The first three-fifths of the chapter depend heavily on texts and epigraphy, covering ground which has long been argued over, precisely because of the availability of these sources. The standard, long-held image is of three provinces (Aquitani[c]a, Belgica, Lugdunensis), each with its own governor and divided into bounded, territorial civitates, each with its own “capital” in the form of an urban center (though we should also remember the great “rural sanctuaries”). But this must have involved a lot of choices over such matters as which peoples were (or were not, as with the Mandubii) to constitute a civitas, what the boundaries were to be, where the “capital” was to be sited, and who were to be the office holders and priests. As we have seen, the pre-conquest archaeology shows nothing so fixed and stable, with individuals, groupings, and sites rising and falling; what Caesar's peoples such as the Arverni meant in terms of territory is very far from clear. R. shows that the visit to Gaul by Augustus from 16 to 13 BCE was the crucial event. Over such an extended period the princeps would have met with many Gallic notables seeking to influence him to their advantage. Augustus must have conducted a great deal of raw politics, favoring some and rejecting others, weaving and pulling the strings of patronage networks: indeed, Cass. Dio (54.25) is explicit about what went on, and how some prospered and others lost out. We have glimpses of the process in the favorable results such as those civitates given superior juridical status, such as foederata or libera, in preference to the general run of stipendiaria, or those cities with Augusto-toponyms. Unfavorable outcomes are passed over in silence. The year following Augustus’ departure, 12 BCE, Drusus carried out the indispensable census to regularize the fiscal affairs of Gaul and inaugurated the federal Altar of Rome and Augustus at Lyon. Drusus was not the last imperial prince to exercise power in Gaul and Germany, and down at least to Quinctilius Varus, other governors were high-ranking senators. Some glimpse of the forging of bonds of loyalty can be seen, for instance in early inscriptions to imperial personages such as Gaius and Lucius. In 12 BCE also, Drusus departed to commence the great series of campaigns to subdue Germania east of the Rhine. The effects of this are the focus of the latter part of the chapter, looking first at questions of recruitment from among the Gallic peoples and second at the linked question of the gradual diffusion of Roman coinages into northern Gaul, which latter shows no evidence for a program of monetization or purposive economic development. The recruitment of auxiliary forces from amongst Gallic and Germanic peoples is well attested, particularly those along the western banks of the Rhine; the remarkable case of the Batavi of the lower Rhine has recently been exceptionally well studied.Footnote ¹⁷ Clearly this was a way for peoples to win the favor of the Roman authorities, individuals also with their personal names embodied in unit titles such as the Ala Indiana or the Ala Petriana, the eponyms of both of these units were Treveran.

Chapter VIII traces the incorporation of the Gallic nobility into the imperial system through the 1st c. CE. This was a process not without its hiccups and setbacks, and the chapter opens with a discussion of the serious revolt of 21 CE, the Revolt of Florus and Sacrovir. Provoked by oppressive exactions to pay for and supply the eight-legion army (plus auxiliaries) on the Rhine and their campaigns, it was led by the two named individuals from the Treveri and the Aedui, both with the nomen Iulius and both former auxiliary commanders. Opposed to Florus was another Treveran, Iulius Indus, who raised an auxiliary unit which became the regular ala Indiana. Sacrovir interestingly attacked the schools at Autun, where the youth of the Gallic nobility were being integrated into the cultural networks of the Roman world, as well as those of politics and patronage. Clearly, rather than seeing this simply as a revanchist “Gallic” uprising, we should also acknowledge it as part of the working-out of the relations of factions of the Gallic nobility to each other and to Rome. After this, and through the 1st c. CE, a number of the civitates of Gaul were granted ius Latium, Rome rewarding those who had been loyal. Then there is the famous speech of Claudius to the Senate, requesting that Gallic nobles be admitted among the patres conscripti, a speech preserved for us by Tacitus and on the bronze tablets set up at the metropolis of the Three Gauls, Lyon. It was at Lyon in 67/68 that a Gallic senator and governor of Lugdunenses, Iulius Vindex, raised a revolt against Nero, attaching himself to Galba's usurpation. Several powerful civitates threw in their lots with Vindex, while others remained loyal to Rome, and he was defeated by the Rhine armies. This was then followed by the much more serious revolt of Iulius Civilis (another Iulius, another auxiliary commander) amongst the Batavi and others on the lower Rhine, supporting the claims to the purple of Vespasian, though it was so dangerous that Vespasian himself had the revolt suppressed rather than rewarded. Again, we see the factionalism of the primores Galliarum mixed with the factionalism of the imperial elite. Gauls had entered into the mainstream of Roman politics.

The book closes with an Épilogue entitled Le poids du passé, les prémices de l'avenir, which is partly a drawing together of themes from the preceding chapters and partly a consideration of some topics not dealt with before. For the former, the emphasis is on the long-drawn-out nature of much of the change from the pre-conquest situation in the mid-1st c. BCE towards the developed Gallo-Roman systems of the first half of the 2nd c. CE. Among the topics not previously dealt with is a brief treatment of the importance of funerary archaeology for tracing the chronology and processes of the impact of Mediterranean practices on Gaul, as always very much variable by time and area. The final subject addressed concerns languages, first the meanings of the spread of the epigraphic habit down to the late 2nd/early 3rd c.; second, the evidence for the survival of Gaulish language/s and the interaction with Latin, which was clearly not just a matter of social status.

En guise de conclusion…. How to condense and summarize a work that, it should be clear, is so hugely rich in evidence, thought, and argument? In many ways, the two Parts come over as semi-independent because of their evidential foci and thus their temporal patternings. Part 1 is based largely on archaeology, Part 2 largely on written sources, though each is frequently informed by the other. Because of this, there is the difference in temporal resolution. R.'s view of the development of the archaeology of Part 1 is one of long-drawn-out changes, moving more to the rhythms of the moyenne durée than evincing the sudden soubresauts of histoire événementielle. By contrast, Part 2 can often be more concerned with the time scales of single lives or the duration of lifetimes, showing how the agency of individuals can act in counterpoint to the structures of the medium term. One way of reconciling these chronological frames is to note that the century and a half from the Caesarian conquests to the appearance of what we might regard as fully fledged Roman Gaul (archaeologically and historically) covers some six generations, maybe more. What to us may seem gradual, almost imperceptible, change may to an individual on the ground have seemed full of change (for good or ill). An enduring message from the book is that there was no simple, linear progression over the period under consideration. Archaeologically, we have seen the variability and fluidity of the settlement evidence, already established before Caesar and continuing through into the Augustan period and down to the Flavians and beyond. To an extent, such Roman introductions as territorial civitates and single “civitas-capitals” distorted and fossilized a multiplicity of trajectories, also creating new dynamics down through the high Empire (and later). Likewise, the textual and epigraphic evidence shows the importance of individuals, their actions, and their relationships (especially with Romans) in forging new patterns that would also last deep into the development of Roman Gaul. Simple linear and programmatic narratives of culture change miss much, so much in both the archaeology and the history was contingent. A vital next stage is for a generation and more of upcoming archaeologists and historians to internalize R.'s arguments and insights and then pose new questions and mount new projects to test these in new ways. A crucial development will be to deploy the types of evidence less considered here to complement, question, and revise R.'s ideas and conclusions. That is the measure of the book, that it will become a touchstone against which future work is to be assayed.