This is a sharp and well-focused study, largely of the Romulus and Remus story. N. contends that we should understand the story within the context of ‘dyadic rivalry’, that is two roughly equal individuals who co-operate at first to general benefit, but at some point cease to do so, and at that point, one is marginalised. This, it is then argued, can be set into the context of the balance between individual and shared power throughout Rome's history.

Although the material is familiar, N. brings a fresh and determined approach to some difficult challenges. She swiftly undermines the explanatory power of most anthropological approaches to the twins, preferring a historiographical account. N. has taken full advantage of the new versions of the fragmentary Roman historians and uses this to try to sketch the earlier versions of the myth.

N. summarises her approach as follows: ‘The institution of productive rivalry that is celebrated in the second century sources lays a foundation for the innovations in this discourse that appear in the first century, when such rivalry was harmful to the state’ (p. 27). Thus N. stresses signs of equality between Romulus and Remus in Ennius’ account, and suggests that Remus simply drops out of the account after the omen scene, and may not even have been killed by Romulus (which she admits is speculation). Similarly N. suggests that Hemina presented Romulus and Remus as effectively operating as joint magistrates. In the next chapter, ‘Invective’, N. argues that if Romulus and Remus were seen as predominantly models of co-operation as late as the 120s, even in the 60s they were not especially differentiated and only very late on, in the 40s, does the emphasis come to rest on the murder of Remus.

N. then moves to rituals; she argues that the Lupercalia was a ceremony of co-operative behaviour but that in 44 b.c. Caesar tried to wrestle it into something more personally focused; and that Caesar's association with Quirinus was another step in an assimilation to Romulus, which in turn led Cicero to be less favourable about Romulus-Quirinus in his later works. This train of thought is then carried through to art; the depiction of the twins is almost by definition a statement of equality, whereas in Augustan art Romulus becomes more present and on his own. The last chapter specifically on Romulus and Remus focuses on the desecration of the walls. N. emphasises the complexity of the stories, and the fact that one cannot make a facile move from the murder to any specific political parallel, but argues that the Augustan narrative tends to stress Romulus the king, however conflicted he might be.

The amount of available information on Romulus and Remus of course exceeds that for any other relevant story, but N. looks at Romulus and Titus Tatius, Numitor and Amulius, and Brutus and Collatinus to trace similar themes. In the first dyad, Romulus is a better king for being on his own, but we only have the later sources for this. N. suggests that the familiar rivalry between Numitor and Amulius emerges only very late. The Brutus and Collatinus story is especially tricky; the differences between the two become schematic but could be differently characterised. Brutus founds Republican collaboration but only by getting rid of a rival colleague. N. then ends with a discussion of tyrants, and her most illuminating example is that of Camillus and Manlius, who move from being partners in the Gallic affair to foes, with Camillus succeeding. N. draws out the problematic nature of Camillus as a role model, emphasising his tyrannical aspects. A short conclusion restates the value of dyadic rivalry as a way of thinking about competition. The book is well produced and well written.

This is a very helpful account and introduces a useful concept of dyadic rivalry. N. is acutely aware of her problems with scarce evidence, but the reader will be able to tell easily where she is speculating. My feeling is that there are some arguments in tension with each other. One argument runs that second-century historiography offers highly consensual readings – there is a lot of co-operation. The conflicts become more highly wrought in a brief period around the 40s b.c., and then the Augustan writers emphasise the value of sole rule. This would imply that dyadic rivalry is seen as positive, to become negative over time. N. also wants to have quite complex and nuanced individual readings, for instance of Livy's concern over the ambition for power. Not every Augustan reading has sole rule as unequivocally good; and Republican accounts had to kill off rivals too. The evidence permits more sophisticated readings for the later Republic than it does for the earlier period, but N. may be a little quick to underplay the sophistication of the earlier sources, and there are a couple of points where the silence of the fragmentary tradition is taken as telling rather than a product of our lack of knowledge.

N. comes close I think, without being explicit, to saying that the earlier historiography paints a much more community driven account, and that may very well be true. Another way of putting this might be that earlier historiography was more abstract in the way it posed the problems of power sharing and collaboration. Inevitably N. has much more to say about how the experience of government by Caesar and Augustus shaped the discourse, but her conclusions for second-century historiography would be worth exploring further. It is not quite clear where we should put the emphasis for this earlier period: on the positive outcomes of collaboration, or the negative outcomes of destructive competition, or on the whole narrative trajectory. If the latter, the difference between the Republican and the early imperial discourse might not be so far apart, and the number of such stories, and the different outcomes, suggests that whilst N. is right to see these stories as teaching tools, the whole history of Rome offered examples, and the shapes of the stories varied from the beginning. As well as the colleague-to-colleague rivalry in politics, including the annual rivalry between consuls, and the fraternal rivalry of Romulus and Remus, we should consider also father and son rivalries, and maybe even rivalries between ethnic groups such as the Romans and the Sabines, which allow contrasts to be played out. N.'s book has the merit of offering a series of examples which should prompt further thought.