Geography has for too long been jettisoned from international relations, owing in large part to the historical legacy of the Third Reich. But borders, land, ports, and peoples can each be contextualized by their location, which is neither random nor arbitrary. Rather, they result from and impinge upon the everyday ebb and flow of politics and economics. This welcome volume goes against the grain of much quantitative research in international relations and examines some aspects of the role of geography in the conduct of foreign policy and the attendant implications for international politics. It builds beyond a growing tradition in quantitative IR of including some geographical explanations in empirical investigations to actually focusing on a particular aspect of how the relation of states to territory affects their behaviors.

The primary argument of this volume focuses on the aftermath of territorial changes—changes that can come about in a variety of ways, only some of which are peaceful. The transfer of territory, succession, and unification each embody territorial changes, but Jaroslav Tir argues that they each leave in place the “land control oriented motivation” (p. 125) that can lead to conflicts. This is because, he argues, land has two sorts of value that leaders seek: tangible (resources, people, e.g.) and intangible. Conflict is more probable if there is a relatively high degree of intangible value to territory, since it is harder—he asserts—to divide intangible assets than tangible ones. Empirically, Tir finds that territorial changes are often unsuccessful in preventing subsequent conflicts. In particular, he reports that territorial changes are often followed by conflict over the changed borders, whether the changes are for unifications, transfers, or successions (p. 127). However, he notes that after changes, dyadic participants in the change are less likely to fight with each other. Based on this, he concludes that territorial change can serve as a conflict management tool, but recognizes the nuance that surrounds this conclusion. Much like Longfellow's “little girl,” territorial changes with good consequences, have very good consequences. When, however, they are “bad,” they are very, very bad, often leading to violent conflicts.

Part of Tir's conclusions is contextualized by the putative nature of the disputed territory. The author stipulates that territory has two fundamental properties. In the first instance, territory has tangible value, by which is meant strategic and economic values that emanate directly from the plot of land, sui generis. The other type of value is the symbolic meaning that territory has to a group of people. Tir's results suggest that tangible, economic aspects of the value held by a piece of land are those aspects most likely to lead to belligerent interactions after their transfer. In addition to the quantitative studies, he also includes a nicely framed case study in each substantive chapter illustrating how his hypotheses work out in specific ways. This is a very nice touch.

What are the limits of Tir's research? That it nicely builds on Paul K. Huth's data and conceptualization. But at the same time, Tir's approach may be limited in its ability to distinguish between different types of territorial conflict. This goes back to the conceptualization of territory as a material possession that is excludable and ignores, largely, the idea of joint possession of territory, on the one hand, and the notion of denial of territory to opponents, on the other. Admittedly, part of this stems also from a dyadic view of the world, but even with a dyadic filter one country might wish to deny control of a territory to its opponents without wanting necessarily to control the territory itself. Much of U.S. contemporary foreign policy might be so characterized. Moreover, apparently territory has either tangible or intangible value, but conceptually it is not clear why these cannot overlap or even interact. One implication of the truncation of the sample to countries that either gained or lost territory in conflicts is that Tir assumes that no other parties, let alone other states, were stakeholders in these conflicts. The nonignorable elephant here is, of course, Iraq during the first Persian Gulf War—on which he has also written. The United States did not lose any territory when the territorial dispute between Iraq and Kuwait erupted in 1990. Nor did it gain any when the territory in question returned to formal Kuwaiti control the following year.

I would like to see a broader conceptualization and examination of the role of territory, recognizing that looking at the fates of those countries gaining and losing territory according to Tir is an important piece of the puzzle. But it is still a big puzzle, with many pieces still on the table, despite all the numbers put in the tables. Political geography should be able to help here. One obvious place to look is beyond the dyad. In political geography, researchers typically focus on three different scales, the micro, meso, and macro, roughly corresponding to subnational, regional, and global labels in international relations. Hierarchical models can allow researchers to focus simultaneously on the influences in each of these levels, rather than concentrating on a single level. Fortunately, we are at a point in the development of IR, geography, and statistics where we can facilitate deep investigations into multilevel phenomena, such as territorial change.

In short, I like this book and welcome its contribution to the literature bringing together politics and geography. It comes closer than most of the international conflict literature to embracing an understanding of the role of geography in politics beyond the inclusion of a variable named “geography,” which has become a widespread if superficial take on multidisciplinary insight. The statistical models, as well as the focused case studies, provide a wealth of inductive hypotheses about different implications of territorial change that are challenges for future research, rather than providing a compelling answer to any of these questions.