Apropos of Brian Doherty's reportage on the tensions between "purists" and "reformers" in the LP, I see that the arbiters of purity over at LewRockwell.Com ran at least three articles on that theme last week: "In Defense of Libertarian Purity," "Principles," and "Evicting Libertarian Party Principles."
Each seems to be built on a premise that I think is pretty radically wrong: that "libertarianism" is fundamentally a sort of moral philosophy rooted in the non-aggression axiom. The principle is definitive; the policy prescriptions that follow from it (low or no taxation, freedom to engage in "capitalist acts between consenting adults," minimal or no regulation, etc.) are accidental. On that view, then, there's just one "pure" libertarian position, which is either anarchist or damn close to it, and all deviations from that position constitute a watering down of the definitive principle, presumably for the sake of political expediency.
If we don't take the biological analogy too seriously, we could call this the view that libertarianism is a genotype rather than a phenotype. That's probably a bit too starkly binary, of course, because there's not really this sharp divide, but rather a continuum of principles people hold at different levels of abstraction, from the metaphysical view in which an ethical ideal is rooted, through general strucutral or procedural principles of government, and down to more or less specific policy-guiding principles. Still, just to have a term, let's call the view represented above the libertarianism-as-genotype view.
What's appealing about the idea, of course, is that obviously even a pretty broad political orientation is much more than just a set of policy prescriptions. A law banning some kinds of pornography might be defended on progressive feminist grounds or conservative "family values" grounds. And both philosophies are ways of approaching new issues, not just a random cluster of positions on existing ones. So the contrasting view—libertarianism as phenotype—shouldn't be that libertarianism is just a set of policy positions. Instead, let's take as our point of contrast the view that libertarianism is what Wittgenstein would've called a "family resemblance" concept, a cluster of both policy positions and lower-level abstract principles and dispositions for thinking about politics. These might include, among others, sensitivity for unintended consequences of state action, respect for the information aggregating power of the price system, an emphasis on the value of autonomy and people's desire to shape their own lives, and a regard for the importance of robust property rights as a means to that end. On this view, there isn't any one "foundational" or essential principle that's definitive of libertarianism, and so there's no thought of explaining variation among libertarians—more or less "radical" views, that is—along some single dimension defined by that principle, where the only question is how much particular people decide to dilute that one purportedly common ideal
There are two main reasons for preferring the phenotypic view: one conceptual, the other empirical. The conceptual reason is that the "non-aggression principle" is, by itself, vacuous. A socialist could claim to support it, provided he coupled it with a concept of property rights on which holding more than one's equal share of social resources counts as coercive infringement of the rights of others. To get particular libertarian conclusions out of that principle, you've got to load a lot of theory into the term "coercion" or "force" or "aggression," such that when I apply coercion to enforce performance of a contract, or to recover my wallet from someone who snatched it from my coat, that counts as "retaliatory" force, even if I'm the first one to employ aggression in any colloquial sense. But once you realize that, the apparently unitary principle is exposed as potentially encompassing a range of very different ones depending on the details of your theory of property.
The empirical reason is that if you talk to younger libertarians in particular, it's just not clear that the rhetoric of "initiation of force" plays a very large role in their thinking at all, at least not as a fundamental political principle from which all else is derived. This is a point I suspect people who were involved in "the movement" in the 70s, say, really don't yet appreciate fully: For a New Liberty is no longer required reading; there's a significant chunk of the younger generation that, however radical they might be, have never cracked a book by Rothbard or Rand and might never. You can, I suppose, just stamp your feet and insist that such people are ipso facto not really libertarians at all, like a kind of extreme prescriptivist grammarian who knows that a word "really means" whatever Webster's said it meant in 1806, even if everyone now uses it differently. This yields some pretty silly consequences: It entails that an anarcho-capitalist utilitarian whose ultimate criterion is aggregate happiness will not be a libertarian "really," nor will a radical minarchist who takes a Rawlsian "basic structure" approach, where as someone who espouses far more moderate views might be, so long as his foundational principle is the NAP, however watered down for practical reasons. This seems like pointless pedantry: If Rothbardians want to argue that, really, all libertarians ought to prefer their version of the theory, let them argue that, rather than getting tangled in a tedious debate over what "libertarian" really means. [Cross-posted @ NftL]