...it is almost certainly true...that the U.S. government, and all other governments,...base their foreign policies on the notion that these policies advance their national interests. This information does not help us analytically, however, for the concept of national interest is anything but a sure guide to policy. In fact, except within very broad limits, the national interest is no guide to policy at all. Within these limits -- which are not trivial but are not very discriminating either -- national interest means whatever different people want it to mean. As a symbol for justifying actions and for rallying or mobilizing support, the notion of national interest has considerable power.... As an analytic concept from which one can deduce or predict behavior, however, the concept has very little utility.
-- Robert A. Packenham, Liberal America and the Third World (pb, 1976), pp. 323-24 (footnotes omitted)
from R. Aron, Peace and War, trans. R. Howard & A.B. Fox (1966), pp.91, 93:
Not only are the historical objectives of political units not deducible from the relation of forces, but the ultimate objectives of such units are legitimately equivocal.... The plurality of concrete objectives and of ultimate objectives forbids a rational definition of "national interest".... The theory we are sketching here tends to analyze the meaning of diplomatic behavior, to trace its fundamental notions, to specify the variables that must be reviewed in order to understand any one constellation. But it does not suggest an "eternal diplomacy," it does not claim to be the reconstruction of a closed system.