Andrew Louis and Eli Horowitz both have extensive and thoughtful responses to my recent post, Social constructions and Libertarianism.
I want to first respond to them indirectly; my original piece was not, I think, a model of clarity, so I want to try again.
Our primary task is to justify the normative aspects of ethical and political ideas. Without too much controversy (excluding the more extreme postmodernists and epistemic relativists), we can justify descriptive aspects of any set of ideas using the scientific method. It's a more problematic task to justify normativity.
To capture these problems, I want to introduce some terms with specific meanings. The first term is consistently determinable; an statement is consistently determinable if, given a context, i.e. some statements and inference rules, everyone will consistently assent or dissent to some statement. A statement for which assent is consistently determinable in some context is valid relative to that context; a statement for which dissent is consistently determinable is invalid; and a statement for which neither assent nor dissent is consistently determinable is indeterminate. Thus, given the axioms and inference rules of arithmetic, for example, everyone will consistently assent that "2 + 2 = 4"; it is a valid statement of arithmetic.
The next terms are objective and subjective. The use of these terms in general philosophical use varies, so I want to be precise about how I'm using them: the objective pertains to the world outside our minds, and the subjective pertains to our minds directly. The union of the objective and the subjective constitute the real: although no unicorns objectively exist, a thought about a unicorn is still a real thought. Similarly, gravity objectively exists, and the law of gravity, i.e. our thoughts about gravity, also subjectively exists.*
*I don't want to get into a theory of truth, but we can say, glossing over a lot of philosophical problems that are not pertinent here, that because our subjective thoughts about gravity seem in some sense to "match" the objective nature of gravity, our subjective thoughts seem in some sense to be "true."
We can combine these ideas. A statement that is consistently determinable without introducing any properties of any minds is objectively determinable. A statement that requires introducing properties of minds to be consistently determinable is subjectively determinable in this sense. This construction can be extremely problematic: it is not at all obvious how to operationalize objective determinability. But I think we can handwave over these objectives for the present moment; my terminology captures, I think, some deep intuitions about the world works, and how our minds conceptualize the world.
Using this terminology, I can express my earlier point more succinctly: all consistently determinable normative statements are subjectively determinable. We must in some way reference the contents of people's minds to consistently determine normative statements.
To illustrate this concept, I divide our notions about property into two classes: the objectively determinable components, which I label more-or-less arbitrarily as possession, and our normative, (potentially) subjectively determinable components as ownership. Thus the statement "Alice possesses that specific blue 2010 Honda Civic" is by definition objectively determinable: we would expect that even a space alien with no knowledge at all about our minds could tell whether or not Alice possesses that specific car: they might observe, for example, that Alice uses the car often and maintains consistent control over it. On the other hand, "Alice owns that specific car" is by definition subjectively determinable: our hypothetical space aliens would necessarily have to learn something about our minds to determine Alice's ownership. Again, these are just arbitrary distinctions for the purpose of clarity.
Given these distinctions, there seem to be two arguable positions about the relationship between ownership and possession. The first is that ownership directly matches possession: given some objectively determinable definition of possession, anyone who possesses an object always owns it. The second is that ownership does not directly match possession: we must in some sense know a lot of things about people's minds to determine whether or not someone owns something; for any objectively determinable definition of possession, it will be the case that people can possess things they do not own and own things they do not possess. Note that the specific definition of possession is irrelevant; what is relevant is only that possession is by definition objectively determinable.
(We can also exclude the case where, by dint of careful examination, we might be able to objectively determine ownership, but only at the "cost" of having enough information about the world to make good inferences about people's minds. Minds are just as real as rocks and trees, and all of reality is causally interconnected; the distinction between objectivity and subjectivity is a distinction of convenience, not a fundamental dualism.)
So finally, after a lot of tedious philosophy, we come to my two main objections to Libertarianism.
The first is that we have a lot of intuitions about ownership that are based directly on possession, and Libertarians use those intuitions as normative justification. If I really possess something, it takes objectively determinable initiation of physical force to expropriate the object without my consent. We need the things we use and possess, and we resent people forcing us to do things without our individual consent. These intuitions are not, by and large, problematic. Even a committed communist such as myself admits that I possess my car, for example; I use it, and I maintain control over it, I need it; and I would be quite peeved if anyone, including the state, arbitrarily expropriated it by force. So these intuitions are not problematic. However, the Libertarian argument then typically "subjectivizes" what appear to be the objectively determinable notions of possession and the initiation of coercion: the subjectively determinable legitimate employment of coercion, to enforce an agreement, for example, is not the "initiation" of coercion. Because Alice subjectively determinably owns something, she thereby possesses it; because she owns something, her use of coercion cannot be initiation. This justification is disingenuous because it intentionally confuses objective and subjectively determinable considerations.
The point is only that this particular justification is untenable. The point is not that people should or should not only own what they possess, nor is the point that a bad justification renders a proposition false. The point is simply that this particular justification, using our intuitions about objectively determinable states to justify states that can be only subjectively determinable, is disingenuous.