But territory is not entirely destiny, even when it is the starting point for constituency definition as well as the residence-based distribution of one vote to every citizen. The history of race-based districting in the United States can be understood as attempts to mold geographical constituencies in ways that encompass nongeographical issues, and to do so through the inclusion of racial minorities in decision-making bodies. Quotas and reserved seats also compensate for the inflexibilities of geography, although each arrangement comes with costs to other dimensions of representation (Guinier 1994; Williams 1998, chs. 3, 7; James 2004). Functional role adjustments, even if ad hoc, may sometime compensate. Mansbridge (2003) notes that empirical political scientists increasingly identify forms of representation that are not based on standard “promissory” mechanisms, whereby candidates make promises to voters and are then judged in subsequent elections by the results. In “surrogate representation,” for example, a representative claims a constituency beyond his or her electoral district, as when Barney Frank (a member of the US House of Representatives from Massachusetts) represents gays beyond his district, or Bill Richardson (Governor of New Mexico) represents Latinos beyond his state. These functional adjustments testify not just to the inadequacies of territorial constituency, but also to its malleability. A key challenge for democratic theorists is to imagine how this malleability might be harnessed beyond the borders of nation-states.