E.G.

"Freedom," like so many other philosophically-significant ideas, is an idea which most of us use to explain and describe events in our daily lives.  But freedom differs from some other such ideas in an important way: Freedom is employed in descriptions and explanations specifically concerning human beings (and, occasionally, some non-human animals).  We routinely bring up "freedom" when we talk about persons’ intentions, or persons’ moral responsibility for their actions, or persons’ plans for the future; we don’t mention "freedom" when we describe or explain non-persons or inanimate objects, such as landslides, planets, or machines.

"Persons" are, therefore, the "wellspring" of the notion of freedom: they are what make the notion of freedom occur to us in the first place.  As a consequence, a philosopher who wants to understand freedom is likely to return again and again to this wellspring in order to shape and guide her approach to the subject.

For instance: Suppose we believe that, under a certain set of circumstances (call them "circumstances A"), Amanda is to be blamed for killing Tom, and we believe that Amanda must have freely performed her action in order to be blameworthy.  In that case, if a proposed conception of freedom is at all plausible, it must be such that, in circumstances A, Amanda’s action qualifies as being "free".  This is a test we can use to cross items off our list of proposed conceptions of freedom.  And if we can continue to generate other, similar tests by considering other, similar hypothetical scenarios involving Amanda and Tom, we might hope that we’ll eventually be able to cross every item off the list except one.  Once that happens, we expect we will have found the correct conception of freedom.  This, I take it, is the standard way (or a version of the standard way) of trying to arrive at a conception of freedom.

But what if persons and their circumstances aren’t enough to do this job?  What if no matter how thoroughly we scrutinize persons, choices, actions, responsibility, and so on, we still end up with a long list of still-plausible conceptions of freedom? 

As we’ve seen, one can draw from intuitions of the form "Person A in circumstance X is free," or "Person B in circumstance Y is not free," etc., in order to approach the correct notion of freedom.  But this leaves out another intuition, which is widely held: the intuition that landslides, planets and machines — inanimate objects — are not free. 

When we consider a proposed conception of freedom, then, we should ask all the usual questions about whether it matches our intuitions regarding persons’ responsibilities, intentions, and so on.  But we should also ask: Does the proposed conception of freedom allow us to say that a landslide does not slide freely?  Can it explain what it is about freedom which is absent in a landslide?  Or is it unable to say why a landslide isn’t free?  Or — and this, I suppose, would be the worst outcome — does the proposed conception of freedom commit us to say that a landslide is free?  If so, we’re entitled to scratch it off our list, even if it gives us all the answers we expect when we ask questions of the first type — questions about persons.

I don’t know whether versions of the second question are very often asked.  I also don’t know whether asking them will prove to be very helpful.  But I have the suspicion that it might be helpful in at least some cases.  To provide just one, outdated, example: I think that Kant is committed to say that inanimate objects are free.  The reason I think Kant has this commitment is quite simple (or perhaps just simple-minded).  Kant defines freedom (or, at least, that form of freedom which he calls "transcendental") as consisting in being "grounded in" the noumenal world.  But for Kant, no appearances are mere appearances; they are all grounded in the noumenal world.  This includes appearances of human beings, as we want — but it also includes appearances of landslides, planets and machines.   Thus all appearances are transcendentally free. 

In another venue I have put a fuller version of this argument to people who know much more about Kant than I do.  They were unconvinced.  I was unconvinced by their responses.  I suppose you would be well-advised to go with their opinion on the matter rather than mine.  But even if you decide that Kant does not have this obscene commitment to calling landslides "free," I think my larger point might still have some merit.  Philosophers who think about freedom have a responsibility to advance conceptions of freedom which give us reasons to say what we want to say about persons: that "Person A acts freely," or that "Person B does not act freely."  But a philosopher’s eagerness to fulfill that responsibility has the potential to blind her to another responsibility: The responsibility to give us reasons to say that Landslide C is not free.  It is possible, anyway, that attention to that latter responsibility can provide a way to cross some otherwise plausible conceptions of freedom off our list.

[Note: What you see above is a slightly modified version of the original post.]