Returning to classes this Fall, I explained (as many readers of this journal will have done many times) to yet another group of students in my Roman History survey class that Rome had an empire long before there were emperors. The nature of the connections between the two institutions is the big question raised at the beginning of this interesting book. The question, of course, is so large that this very interesting new book does not pretend to cover it all. Norena’s strategy is rather is to pursue a fairly narrow line of inquiry through most of the work (albeit one that relies on controlling a large mass of data). By the end, however, he has said some important and general things about the workings of power in the empire, and how the trajectory of ideological messaging may have altered its very foundations.
The book comprises a general introduction and conclusion, two chapters on concepts presented on imperial coinage (“values and virtues” and “benefits of empire and monarchy,” respectively), two on the use of similar themes in honorific inscriptions, general introductions to those two pairs of chapters discussing the samples of evidence on which they rely, and 15 appendices further cataloging evidence and discussing statistical technicalities. (The study treats principally the western empire from 69-235, but there are also wider views in time and space.) Before going on to further analysis, I will summarize the contents of the four core chapters and the conclusion.
Chapter 2. While Roman politics had always been framed in broadly ethical terms, its reduction (at least in a certain rhetorical register) to a list of virtues and vices was largely an imperial development, and one that does not really start to take hold until the Flavians. Personifications of virtues are one of the two most important reverse types on imperial coins. Looking at coins over time and across metals we can construct a set (never proclaimed as a canon in antiquity) of the most important of these: aequitas, pietas, virtus, liberalitas, and providentia (61). Noreña analyzes the meaning of these terms based on a variety of texts and images, and looks at their distribution. In several cases he makes a case that “the state was careful to ensure that its public claims correspond to some objective reality.” (70) I will have more to say about the substance of this claim below, but here I note that the “objective reality” in question is not typically individual episodes but rather broad trends in policy.
Chapter 3. Even more frequent than the virtues are purported “material, social, and juridical benefits of empire.” (106) Here the canon is slightly larger and has a fuzzier edge, but Noreña plausibly restricts detailed consideration to victoria, pax, concordia, fortuna and salus (111; several of the less common types are readily grouped with one or the other of the major ones). These concepts are framed in ways that allow the generalization of a benefit to a specific audience or at a specific time (e.g. annona for the urban populace; a particular military conquest) to slide over into more general ones (abundance, victory in general). The importance of victoria varies considerably over time, and seems to track the importance of aggressive war as measured by other indices. More constant over time, however, is an emphasis on pax and concordia in the silver coinage as opposed to fortuna and salus in base metals; Noreña reads this as reflecting a more elite audience for the former. Conspicuous by its relative rarity in any form and at all times is libertas.
Chapter 4. The emperor’s presence was projected throughout the empire by a number of media (especially sculpture). Among these, monumental inscriptions—both official and unofficial—stand out for their combination of articulateness and (in at least some sense) broad availability.1 The official inscriptions work with largely the same canon of virtues as the coins. Moreover, the frequent coincident appearance (or prominence) of particular terms or themes in both media suggest a centralized source (if not authority) for language that appears across the entire west. Severan inscriptions, however, show a strong tendency (only weakly represented in the coinage) to replace generalized virtue terms with more narrowly militaristic language.
Chapter 5. Inscriptions by unofficial actors again exhibit the same general canon of virtues and sensitivity to the reign-to-reign variation we have already seen. Noreña takes this as evidence that non-state actors appropriated the official language. More specifically, local aristocrats voluntarily reproduced the official line, but they did not do so in order to reflect it back to the central government, nor did they seek to display “loyalty” to the regime or attract/reciprocate imperial benefactions. Rather they sought to achieve local advantage. They were themselves beneficiaries of the gross inequalities of the imperial system, and this provided a ready-made ideological package to legitimate that system. It also allowed them to re-appropriate the same symbolic language to describe and legitimize themselves more directly. This became problematic under the Severans, as the emperor moved from “[ethical] model to master” (283), from optimus to dominus. The new, autocratic vision offered by the central state left little space for layered hierarchies. Under the Antonines, playing the emperor’s ideological game made it worthwhile to replicate (and in fact make much more concrete) his supposed beneficence. This became an engine for the local euergetism that made the system function. As the ideological motivation declined, that euergetism did as well.
Chapter 6. The argument just presented suggests a convergence of language (and symbols more generally) between the state and local aristocracies based on converging but separate individual interests. Noreña also shows paths by which this ideological convergence might have gone deeper in the population as well, though, as he notes, this is hard to prove and probably not directly relevant to the main argument. The ideological convergence of different elites was enough to generate a lot of power. In particular, the notional exchange of imperial generosity for various honors served as a model for similar but much more concrete exchanges at the local level. This is what motivated mechanisms of redistribution throughout the empire.
At the core of the book is, as Noreña points out, a lot of quantitative analysis of the deployment of various ideas on coins and elsewhere. In particular, he makes three kinds of quantitative argument. (1) Frequencies of appearance on coinage of particular ideas are compared to frequencies of appearance on coinage of other ideas. This gives us some sense of what messages were most important to the state on a reign-by-reign basis. (2) Frequencies of appearance on coinage are compared to those on inscriptions, both official and unofficial. The former illustrates the coherence of the messaging coming from the center; the latter shows the extent to which those messages were picked up and echoed (for whatever reason) by those outside the state. (3) Frequencies of appearance on coinage are compared to various statistics about external phenomena (purity of the coinage, frequency of imperial victory celebrations). This is used to argue for a connection between the official rhetoric and the realities of policy.
The arguments of the first and second categories (coins vs. coins, coins vs. inscriptions) are for the most part clear and straightforward, despite a few quibbles that could be offered. The connections between coin types and real- world factors are necessarily more complicated and more speculative, and deserve special attention. Let me start with a general remark, before considering two particular cases. When Noreña ties frequency of coin types to external phenomena, he calculates a correlation coefficient for the two variables, then does a further test to confirm the statistical significance of the first result. Within my limited ability to discern, these calculations have been done in a formally correct way. We should keep in mind, however, that “statistically significant” need not mean significant to the subject matter, and that correlations (even when “real”) can be caused by many things. Let me give two examples in which more explicit attention might have been paid to alternative interpretation of the data in light of these considerations.
Noreña points to a correlation between production of aequitas types and the silver content of contemporary coinage to argue that the aequitas types were used to emphasize actual production of good coin rather than (as some might have suspected) serving to conceal real debasement. He notes a strong correlation between his variables for the period from Nerva to Caracalla (r = -.73; pp. 69, 357) and an even stronger one for just the period from Marcus Aurelius to Caracalla (r = -.96; pp. 70, 357). It might be useful to take a look at a plot of the data involved.
See Graph
Noreña sees a single phenomenon, starting as a vague trend, then sharpening under Marcus Aurelius. It seems to me that we have, rather, two different linear phenomena, one from Nerva to Antoninus Pius and the other from Marcus Aurelius to Caracalla. In fact, the correlation coefficient for the first half (-.9687) is slightly higher than that already calculated by Noreña for the second half. If both series show high correlation, there is no reason to write either off as a poor approximation of the other. How then can we account for both halves? Frankly, I have no answer, but I would be cautious of any real-world interpretation of this graph that did not offer a clearer account of the change.
Or consider the correlation (r = .95) between the frequency of liberalitas types and frequency of congiaria over the reigns of the emperors from Hadrian to Caracalla. Here, a graphic plot would simply emphasize Noreña’s statistical findings, but we might want to take a closer look at the underlying data. Most rates of congiaria fall into a small cluster between .32 and .38 per year. In real world terms this seems like a fairly subtle difference. Suppose Hadrian or Septimius Severus had given out just one more over the course of their entire reigns or Antoninus Pius one fewer. This would send the former two to the top half of the cluster and the latter to the bottom.
See Table
One difficulty with any quantitative study of broad social phenomena is coming up with statistics to represent the phenomena in which we are really interested. How does one measure imperial generosity or social welfare or the like? Finding such statistics for the ancient world is, of course, doubly difficult. Counting congiaria is an entirely reasonable move to make, but it perhaps turns out not to give particularly robust results. And in any case, this is an area in which we might want multiple indices.
Noreña contrasts his opportunistic view of what provincial elites are up to with Ando’s account of loyalty (or at least loyalism).2 The interpretation that local imperial honorifics were in the first instance “public expressions of loyalty to the imperial regime… is not tenable.” (270-271, cf. 272) But elsewhere (19n61) he allows “these perspectives [including Ando’s] are not wholly incompatible with mine,” although “the emphases are very different.” I would suggest taking the latter formulation as more than simply polite. The case against the loyalist interpretation hinges on the argument (probably correct) that there were few or no ties between direct encounters with a traveling emperor and either imperial benefactions or dedications, and so the emperor (or even his senior officials) cannot be thought of as the audience for the dedications (271). But this is perhaps too literal a reading. This was after all a world in which the emperor maintained a considerable virtual presence, not least through the coins and statuary this book analyzes so well. Moreover, if we look beyond the kinds of communications studied here, it is also a world in which all manner of requests were thought suitable to be passed on to the emperor, with some real expectation of a response.3 Noreña makes a very strong case for the important local effects of local honorific regimes (and their consequences for the structure of the empire as a whole), but that need not constrain our account of their motivation. One should maintain the normal doubts of the sincerity of such gestures, but that does not mean that upward display is not an important part of what they were meant to do.
As often happens in reviews, I concentrated on points of disagreement, but I should stress both that I found most of the work convincing, and that even beyond that I have found it profitable to think through these issues with Noreña.
Notes
1. “Official” inscriptions here are those put up by any element or agent of the central government. (214) A text put up by a town council, then, would be “unofficial.”
2. C. Ando Imperial ideology and provincial loyalty in the Roman empire. Berkeley: The University of California Press, 2000.
3. F. Millar, The Emperor in the Roman World (Cornell 1992
1977.), esp. Part Three; S. Connolly, Lives Behind the Laws: the World of the Codex Hermogenianus (Indiana 2010).