One of the best discussions I know of on the difference between positive and negative conceptions of freedom is David Kelley’s in A Life of One’s Own: Individual Rights and the Welfare State. This was the final word for me on positive and negative freedom for some time. Looking again at Kelley’s argument, I find I am not as convinced as I once was.
Let’s look at a few passages:
Freedom always involves the capacity to choose among a range of alternative actions. In that sense, freedom is a positive concept. But it is also a negative concept: the freedom to choose existss as long as no one interferes with the choice coercively, using force to prevent the person from selecting one of the alternatives. … A diner at Joe’s Cafe has a more limited menu to choose from than does a diner at the Four Seasons, but both people are equally free to choose among the entrees available. The fact that Joe’s does not serve oysters on the half shell is not an issue of freedom.
OK. The question that arises for me, then, is why is the guy at Joe’s instead of the Four Seasons? If he (let’s call him Frank) just likes Joe’s, cool. But if it’s because he cannot afford the Four Seasons, or a place with an equivalently broad and high-quality menu, then the question is, Why not? The answer to that question is important. Suppose, for the sake of argument, that Frank, and every other person at his approximate level on the economic ladder, would have the means to choose from bigger menus if only his society’s rate of economic growth had been just fractionally greater each year for the past two decades. Relative to faster growth, slower growth takes things off the menu. But is this an issue of freedom from coercion? Maybe, maybe not.
Suppose that people in Frank’s society just like relaxing more than working, and so aren’t extremely productive, leading to unimpressive rates of growth. Now, Frank is a highly motivated, hard-working, and would like to order off a Four Seasons menu, but simply can’t because his entire society is too poor. The point of this is to emphasize the interdependence of opportunities. Things can be off your menu, not because you’re lazy, or being coerced, but because of the (non-coercive) patterns in which other people are coordinating their behavior.
This brings us to Kelley’s next paragraph:
To be sure, there is not always a hard and fast distinction between the number of alternatives one has and the degree of one’s freedom to choose among them. Theoretically, any obstacle, restraint, or limitation may be looked at in either of two ways: we may view it (1) as something that eliminates one or more alternatives a person would otherwise have available or (2) something that prevents the person from choosing one or more alternatives. The difference lies in whether we consider the limitation as affecting the range of alternatives he has or the process of choosing among them. Advocates of positive freedom have exploited this fact, insisting that lack of a certain opportunity because of poverty, illness, or disability deprives a person of the freedom to choose that opportunity. Conversely, we could in principle view overt coercion, physical force, or violence, not as something that prevents a person from choosing an alternative but as something that removes alternatives he would otherwise have.
OK! Then Frank’s case is confusing, right? It looks like a case of (1). He doesn’t already have the menu, so his inability to choose from it is a moot question—not an issue of freedom. But suppose we get the exact same result—a low rate of growth—not from the indolence of the population, but from a few bad government policies that, say, restrict international trade. It turns out that Frank trades only with locals, so the trade restrictions don’t coercively prevent him from trading with anyone he wants to. But by coercively limiting others’ trade opportunities, others have less means, and thus less with which to buy Frank’s services, which ends up badly limiting his trade opportunities. And that’s why he doesn’t already have the menu he’d like. Still a case of (1)?
Here’s how Kelley asks us to tell the difference:
There are real differences between (1) and (2). One difference is whether the obstacle or limitation is imposed by reality or by other people. When some fact of reality affects the range of alternatives we face, it is wishful thinking to regard it as an obstacle to what we would otherwise be free to do. Facts are facts. The world operates a certain way, according to causal laws, and the constraints imposed by nature are the foundation for human choice, not a barrier to it.
I now find this remarkably unhelpful. Are other people’s preferences and patterns of behavior, which create huge limitations on the alternatives open to me, “imposed by reality” or “by other people”?
Here’s an illustrative example Kelly offers:
If I cannot run a five-minute mile, my incapacity does not abridge my freedom to do so; it is simply a fact about my nature. But if I can run that fast, and somebody forces me to wear lead weights as a handicap, he is restricting my freedom.
Now, imagine the following possibilities:
(a) a network of completely voluntary choices leads to air pollution as a side-effect; I could have run a five-minute mile had the air been cleaner.
(b) the anti-technology norms of my society, transmitted through education and social opproprium (no coercion!), have ensured that new physical performance technologies that, but for those norms, would have been invented, and would have made me able to run a five minute mile.
(c) bad government policy that does not directly prevent me from doing anything at any particular time, decreases the rate of growth, decreasing the amount of capital available for R&D, ensuring that new physical performance technologies that would have been invented aren’t.
(d) If super-steroids were available, I could run a two minute mile, but the price is that I would die a week later, so the government bans them. It happens that nobody would have taken them, even if they weren’t banned.
We could go on. The point is, many incapacities are contingent, and are very, very often a side-effect, intended or unintended, of human action. I understand why having weights forcibly attached to one’s body is an especially vivid, salient, and emotionally compelling violation of freedom. But I can’t see a principled reason why the very concept of freedom should apply only to coercive interference with the excercise of present capacities when present capacities are so dependent on contingent patterns of human interaction. I simply cannot see the special normative salience of coercion as a method for trimming the feasible set. I can definitely see why robust restrictions on the exercise of coercion are completely necessary for achieving the range and kinds of opportunities that allow us to create, develop, and fully excercise our capacities. But then that’s why we should care about limiting coercion.
Kelley builds his argument on the Objectivist distinction between the metaphysical and the man-made, but it is hard to know what to make of the distinction once one makes note of the interedpendence of our capacities and opportunities. Are emergent social patterns—the consequences of human action but not human design—metaphysical or man-made? I think a little bit country and a little bit rock n roll.
The reason we fight so hard over the basic structure of our formal institutions and over the basic structure of our informal institutions (i.e., over our culture) is that those structures create patterns of behavior we as individuals have no choice but to engage in their actually existing form as if they were metaphysical, as if they were constraints of nature. We can’t rewrite the laws of nature, but we can be less bound by nature by bending the laws to our will. This is how reason, science, and the institutions of market production liberate. We can TRY to rewrite the local laws of social coordination, and this may change the overall pattern of behavior, but the pattern won’t be the one we were aiming at. Other people, other factions, are always also trying to rewrite the formal and informal laws, and the equilibrium that emerges (or doesn’t) from the clash of ideas is nobody’s idea. There is a sense in which science and technology make nature more tractable than people. Or perhaps it is better to see people as fully part of nature, and to see certain formal institutions and cultures—-like a system of strong individual rights and the beliefs and norms that back them—-as just other technologies of liberation from recalcitrant nature.
And the twentieth century history of ideas is full of cases where someone showed you just can’t find a solution satisfying a reasonable-sounding set of constraints. Close to the article, there’s Arrow’s theorem; further away, Goedel’s Incompleteness Theorem and the halting problem and quantum indeterminacy are high peaks in a complicated landscape that stretches quite a long way in all directions.
If one cares about reality, then it’s worth carefully distinguishing whether “the fact that Joe’s does not serve oysters on the half shell is not an issue of freedom” is one aspect of a rule the writer has in mind, or instead a constraint on what not-yet-specified rule the writer would be willing to accept.
Arguing for or against individual constraints can be well and good, but as Arrow’s theorem illustrates, it is quite possible to specify so many reasonable-sounding constraints that in fact no rule can satisfy the constraints, and even for the narrow technical problem of voting systems, people’s informal intuitions about how many constraints can be satisfied may not be very reliable. If an author embraces enough constraints that it’s not clear that they are mutually satisfiable, arguments about morality can be a form of intellectual pornography where extreme excitement might come with a drawback of conditioning the reader to unrealistic expectations.
And the twentieth century history of ideas is full of cases where someone showed you just can’t find a solution satisfying a reasonable-sounding set of constraints. Close to the article, there’s Arrow’s theorem; further away, Goedel’s Incompleteness Theorem and the halting problem and quantum indeterminacy are high peaks in a complicated landscape that stretches quite a long way in all directions.
If one cares about reality, then it’s worth carefully distinguishing whether “the fact that Joe’s does not serve oysters on the half shell is not an issue of freedom” is one aspect of a rule the writer has in mind, or instead a constraint on what not-yet-specified rule the writer would be willing to accept.
Arguing for or against individual constraints can be well and good, but as Arrow’s theorem illustrates, it is quite possible to specify so many reasonable-sounding constraints that in fact no rule can satisfy the constraints, and even for the narrow technical problem of voting systems, people’s informal intuitions about how many constraints can be satisfied may not be very reliable. If an author embraces enough constraints that it’s not clear that they are mutually satisfiable, arguments about morality can be a form of intellectual pornography where extreme excitement might come with a drawback of conditioning the reader to unrealistic expectations.
I’ve spent some time thinking about this too. If coercion is to have some special normative salience, I think it will probably have something to do with the Kantian injunction never to treat persons as mere means but rather as ends in themselves. Typically, coercion will treat a person as mere means unless it is somehow justified to the person who is being coerced. But this doesn’t necessarily single out coercion as special: manipulation and perhaps certain kinds of exploitation are also problematic in this regard.
At first glance at least, all of these are intentional forms of disrespect for persons as ends in themselves. The same kind of wrong is not involved when we don’t work to provide someone with a larger range of choice than she would have had otherwise.
For a treatment of this issue from the Kantian perspective, I would recommend Korsgaard’s “The Right to Lie: Kant on Dealing With Evil” and Robert Taylor’s “A Kantian Defense of Self-Ownership.”
I’ve spent some time thinking about this too. If coercion is to have some special normative salience, I think it will probably have something to do with the Kantian injunction never to treat persons as mere means but rather as ends in themselves. Typically, coercion will treat a person as mere means unless it is somehow justified to the person who is being coerced. But this doesn’t necessarily single out coercion as special: manipulation and perhaps certain kinds of exploitation are also problematic in this regard.
At first glance at least, all of these are intentional forms of disrespect for persons as ends in themselves. The same kind of wrong is not involved when we don’t work to provide someone with a larger range of choice than she would have had otherwise.
For a treatment of this issue from the Kantian perspective, I would recommend Korsgaard’s “The Right to Lie: Kant on Dealing With Evil” and Robert Taylor’s “A Kantian Defense of Self-Ownership.”
What if we view things not in terms of freedom but in terms of justice? Then it seems to me it’s easy to see the special normative salience of coercion, and as Javier suggests, it’s rooted in the axiom that justice requires we treat persons as ends in themselves.
First of all, I take justice to mean the fulfillment of legitimate expectations; an injustice is done to you if and only if someone else has violated a legitimate expectation you had of them. So that incorporates the man-made/fact-of-reality distinction (we can’t legitimately expect anything from reality; its facts in and of themselves have no bearing on justice) but sharpens it to say: what can we legitimately expect of other people?
And here’s where the power of Kant is clear. We can’t legitimately expect of other people that they act in such a way as to provide us with the largest possible range of choice– because to do so would be to treat them as means to our ends! But we can expect that they not coerce us, i.e. that they not treat us as means to their ends either.
What if we view things not in terms of freedom but in terms of justice? Then it seems to me it’s easy to see the special normative salience of coercion, and as Javier suggests, it’s rooted in the axiom that justice requires we treat persons as ends in themselves.
First of all, I take justice to mean the fulfillment of legitimate expectations; an injustice is done to you if and only if someone else has violated a legitimate expectation you had of them. So that incorporates the man-made/fact-of-reality distinction (we can’t legitimately expect anything from reality; its facts in and of themselves have no bearing on justice) but sharpens it to say: what can we legitimately expect of other people?
And here’s where the power of Kant is clear. We can’t legitimately expect of other people that they act in such a way as to provide us with the largest possible range of choice– because to do so would be to treat them as means to our ends! But we can expect that they not coerce us, i.e. that they not treat us as means to their ends either.
I think we’d all do well to let go this idea – long pedigree though it has in western thought – which seems implicit in Kelley’s discussion as well as yours, that nature and human action are somehow opposed, that nature is something it makes sense to talk about being “liberated from”. Planetary systems are more tractable than judicial systems because they’re less mind-dependent, less response-dependent, and therefore less dynamic and complex from our vantage point as partial determiners of judicial systems. But neither system is less natural than the other. A social constraint is a species of natural constraint. We’re embodied minds: to be socially constrained is to be physically constrained. What we need here – in lieu of this wooly man vs. nature talk – are distinctions between various levels and types of control we enjoy over different natural systems, plus distinctions between what’s it like to participate in and/or observe different types of systems. Not sure how this bears on the questions about coercion, but it probably bears.
If you’re saying, “perhaps it is better to see people as fully part of nature,” you’ve got to be off-base, or at best, not approaching the issue in the clearest way. ~Perhaps~? Should we be entertaining the possibility that people aren’t fully natural?
I think we’d all do well to let go this idea – long pedigree though it has in western thought – which seems implicit in Kelley’s discussion as well as yours, that nature and human action are somehow opposed, that nature is something it makes sense to talk about being “liberated from”. Planetary systems are more tractable than judicial systems because they’re less mind-dependent, less response-dependent, and therefore less dynamic and complex from our vantage point as partial determiners of judicial systems. But neither system is less natural than the other. A social constraint is a species of natural constraint. We’re embodied minds: to be socially constrained is to be physically constrained. What we need here – in lieu of this wooly man vs. nature talk – are distinctions between various levels and types of control we enjoy over different natural systems, plus distinctions between what’s it like to participate in and/or observe different types of systems. Not sure how this bears on the questions about coercion, but it probably bears.
If you’re saying, “perhaps it is better to see people as fully part of nature,” you’ve got to be off-base, or at best, not approaching the issue in the clearest way. ~Perhaps~? Should we be entertaining the possibility that people aren’t fully natural?
Ah, liberty. As Bentham put it:
“Liberty therefore not being more fit than other words in some of the instances in which it has been used, and not so fit in others, the less the use that is made of it the better. I would no more use the word liberty in my conversation when I could get another that would answer the purpose, than I would brandy in my diet, if my physician did not order me: both cloud the understanding and inflame the passions.”
Ah, liberty. As Bentham put it:
“Liberty therefore not being more fit than other words in some of the instances in which it has been used, and not so fit in others, the less the use that is made of it the better. I would no more use the word liberty in my conversation when I could get another that would answer the purpose, than I would brandy in my diet, if my physician did not order me: both cloud the understanding and inflame the passions.”
Aside from Objectivists, no-one uses the word “metaphysical” to mean “natural”. Any intelligent, well-read person can see that Rand used this word mainly as an empty signal that she was attempting a philosophical thought. Another example is her misuse of “existential” in place of the more simple and correct “existing”. In both cases it is simply bad, amateurish writing.. and take it from someone who knows.
Aside from Objectivists, no-one uses the word “metaphysical” to mean “natural”. Any intelligent, well-read person can see that Rand used this word mainly as an empty signal that she was attempting a philosophical thought. Another example is her misuse of “existential” in place of the more simple and correct “existing”. In both cases it is simply bad, amateurish writing.. and take it from someone who knows.
I think the problem is that negative liberty isn’t just about choice. For example, if someone were to punch me in the face unprovoked in such a manner that it was momentarily painful but had no lasting consequences, he didn’t limit my choices in any way. Negative liberty specifies protection against a harm, not a restriction of choice. An action that causes harm may restrict choice too – breaking someone’s arm does them both harm and restricts their choices. In Kelly’s example of the 5 minute mile and the weights, the weight policy reduces both possitive and negative liberty. Positive, because the subject can no longer run a five minute mile, negative, because it is implied that somebody would harm the subject in a sufficiently grevious way to deter them from running unencumbered if not obeyed.
I think the problem is that negative liberty isn’t just about choice. For example, if someone were to punch me in the face unprovoked in such a manner that it was momentarily painful but had no lasting consequences, he didn’t limit my choices in any way. Negative liberty specifies protection against a harm, not a restriction of choice. An action that causes harm may restrict choice too – breaking someone’s arm does them both harm and restricts their choices. In Kelly’s example of the 5 minute mile and the weights, the weight policy reduces both possitive and negative liberty. Positive, because the subject can no longer run a five minute mile, negative, because it is implied that somebody would harm the subject in a sufficiently grevious way to deter them from running unencumbered if not obeyed.
Pingback: Will Wilkinson / The Fly Bottle » Blog Archive » Positively Heretical?