MacCallum, Gerald C. 1967. “Negative and Positive Freedom.” The Philosophical Review 76 (3). Duke University Press: 312–34. http://www.jstor.org/stable/2183622.
Disputes about freedom have been about what it constitutes, how its attainment relates to the attainment of other “social benefits” like “economic and military security, technological efficiency”, where it may be ranked among such benefits, and what consequences policies may have on the attainment of freedom.
Once one admits that freedom is not the only benefit a society may secure its members, disputes about reconciling it with other benefits or values may arise. We may legitimately ask whether reconciliation is possible; and if possible, whether it is desirable. However, in practice, these questions are often obscured by disputes about the implications of policy on these values.
It has also been common for “partisans” of all kinds to claim for themselves special affinity to freedom in light of the policies or forms of organisation that they advocate, while reserving the opposite treatment to their rivals. This is why freedom has come to be associated with so wide an array of social and individual benefits as to utterly obscure its meaning. This has suited the “purposes of the polemicist”.
The distinction between negative and positive liberty must be seen against this backdrop of confusion, and being influenced by it, the distinction itself is confused because it fails to fully understand the conditions under which the use of the concept of freedom is intelligible.
Freedom is always freedom of something (an agent or agents), from something, to do, not do, become, or not become something; it is a triadic relation which can be expressed as: “x is (is not) free from y to do (not do, become, not become) z,” where x ranges over agents, y ranges over such “preventing conditions” as constraints, restrictions, interferences, and barriers, and z ranges over actions or conditions of character or circumstance. Not all of the three terms need be explicitly stated because the missing term can often be inferred from the context.
It must not be construed that the claim that freedom is a triadic relation is about what we say. Rather, it is about conditions under which what we say may be intelligible. In other words, simply saying something is free or not free does warrant the application of this claim. To put it even more concretely, the claim does not apply to the statements such as “The sky is now free of clouds” or “His record is free of blemish”. The two statements are not intelligible as claims about freedom. The first does not deal with agents at all and in the second, it is not clear whether the statement is about the freedom of the agent or of something/someone else.
Let us look at some troublesome cases where not all the terms of the triadic relationship are clear.
(a) Cases where agents are not mentioned: Consider expressions of the form “free x” where x does not clearly refer to an agent — “free will” — or where x clearly does not refer to an agent — “free beer”. These cases, even if the agents not explicitly mentioned, nevertheless are concerned about agents and are intelligible only if they are understood as such. “Free will” for example is obviously concerned with the freedom of persons or selves. While not as obvious, “free beer” still refers to beer that “people are free from the ordinary restrictions of the market place to drink without paying for it.”
(b) Cases where it is not clear what corresponds to the second term: Consider the expression “freedom of choice”. The preventing conditions are usually clear from the context. In political matters, they are usually legal. In Mill, they were social pressures.
(c) Cases where it is not clear what corresponds to the third term: Consider the expression “freedom from hunger”. It could simply mean being rid of hunger. This doesn’t conform to the triadic schema. However, it could also mean that to be free from hunger is to be free to do the things one can’t do when hungry. Even more satisfactorily, it could mean “a world in which people would be free from barriers constituted by various specifiable agricultural, economic, and political conditions to get enough food to prevent hunger”. This last view of “freedom from hunger” both makes perfect historical sense and conforms to the triadic schema of freedom.
The conventional characterisation of the difference between negative and positive freedom as “freedom from” and “freedom to” does not distinguish between two genuinely different kinds of freedom. Rather, it serves to emphasise one or the other of two features or variables that are present in every kind of freedom.
The next problem is how, or if, the differing answers to the question “When are persons free?” survive the agreement that freedom is a triadic relation. For example, differing views on what is the “true” identity or desire of an agent or on what counts as a constraint or on the range of things agents might be free (or not free) to do (or become) might offer dramatically different accounts of when persons are free. Given the variables involved, accounts of freedom can diverge in many ways. It is therefore crucial to get the range of the variables quite clear.
The distinction between negative and positive freedom has made this difficult by encouraging the wrong questions. It is often asked which one of the two is correct, or desirable. Instead, what should be asked is what the range of the variables are. In other words, “[i]t would be far better to insist that the same concept of freedom is operating throughout, and that the differences, rather than being about what freedom is, are for example about what persons are, and about what can count as an obstacle to or interference with the freedom of persons so conceived”.
This insistence is necessary. Consider the differences between negative and positive freedom. Once the distinction between them as “freedom from” and “freedom to” is debunked (see above), the differences appear to be the following.
- Writers adhering to the concept of “negative” freedom hold that only the presence of something can render a person unfree; writers adhering to the concept of “positive” freedom hold that the absence of something may also render a person unfree.
- The former hold that a person is free to do x just in case nothing due to arrangements made by other persons stops him from doing x; the latter adopt no such restriction.
- The former hold that the agents whose freedom is in question (for example, “persons,” “men”) are, in effect, identifiable as Anglo-American law would identify “natural” (as opposed to “artificial”) persons; the latter sometimes hold quite different views as to how these agents are to be identified (see below).
These differences break down or, at least, become less dramatic when probed. With respect to the first, would proponents of “negative” liberty be disallowed from saying that a chained man is unfree because he lacks a key (absence of something), and not only because he is chained (presence of something)? Or would proponents of “positive” liberty be disallowed from saying that an untrained person failed to get a job because of existing economic or educational systems (presence of something) which led to the person being deprived of training (absence of something)? The answer to both is: no. They can, and do, give those answers. The point of difference, then, is not as dramatic as it is made out to be.
Also, the organisation of thinkers into two camps is ill-considered. Locke’s views on liberty as those actions that man “himself wills it” (Essay Concerning Human Understanding, Bk. 11, ch. xxi, sec. 15) or his view that law should not just “abolish or restrain, but … preserve and enlarge freedom” (Second Treatise of Government, sec. 57) should make him a serious candidate for inclusion in the “positive” camp but instead, he is made the poster boy of the “negative” camp.
Who is the proper agent whose freedom is in question? What is the proper range of obstacles or constraints? And what is the proper range of what that agent may be free (or not free) to be (or become)? For the adherents of negative freedom, the proper agent is a “person” understood in the most ordinary sense of the word. The proper range of obstacles is populated by just what we would ordinarily call “obstacles” which are arrangements made by human beings. The proper range of what an agent may be free (or not free) to be (or become) are what he “wants” to do or be, and what he wants to do or be can be “determined by what he says he wants to do, or by what he manifestly tries to do, or even does do.”
For adherents of “positive” liberty, the answers to the three questions are anything but ordinary. The proper agent is not the ordinary person but the “real”, or “moral”, or “rational” person who is often hidden within the ordinary person. The agent could also incorporate “the institutions and members, the histories and futures of the communities” of which he is an extricable part. This contraction of the meaning of the proper agent is the result of a worry that what we ordinarily want may not be what we “really want” in so far as they may be detrimental to our own interests. It’s expansion, on the other hand, stems from the idea that what we “really are” may be determined in part by our association with our families, communities, and so forth.
Given the radical departure in understanding who the proper agent is, what counts as an obstacle to that agent, unsurprisingly, is very different from that of “negative” freedom. While adherents of “negative” freedom see as obstacles only those arrangements which are the “made by human beings”, their “opponents” might not consider this qualification as relevant. In other words, the presence of obstacles, whether placed by humans or otherwise, is quite inessential. What is important for them is whether human arrangements can remove them.
As regards the third variable, proponents of “positive” liberty “emphasise conditions of character rather than actions”. The range of character conditions and actions are necessarily bound up with the idea of who an agent is and what obstacles are, of which there are, as already seen, many.
All of these divergences can be managed only if they are seen as disagreements on the range of variables that are part of the same idea of freedom as a triadic relationship.
This approach has been neglected because philosophers have made the mistake of asking unadorned questions like “When are men free?” or, alternatively, “When are men really free?”. These questions take it for granted that persons can be simply free or not free.
“One might suppose that, strictly speaking, a person could be free simpliciter only if there were no interference from which he was not free, and nothing that he was not free to do or become.” Given that societies invariably exercise some form of coercion and given the disputes regarding the proper range of the variables of the triadic relation, it should be obvious that persons in cannot be free or unfree simpliciter.
Perhaps, “in certain (conceivable) societies there is no activity in which men in that society are not free to engage, and no possible restriction or barrier from which they are not free.”
The burden of such an argument is to demonstrate that what is ordinarily considered as an interference or a barrier is actually not so, and that everything a person is ordinarily considered not free to do or become is actually irrelevant to freedom. However, other pitfalls remain. Often, questions regarding the legitimacy of interference are reduced to questions concerning genuineness as interference. Also, questions concerning the desirability are reduced to questions about possibility.
‘Perhaps, however, the claim that certain men are free simpliciter is merely elliptical for the claim that they are free in every important respect, or in most important respects, or “on the whole.”’ This, however, does not remove the need of asking, in “the most important aspects”, for example, what they are free from and what they are free to become. And straightforward answers to these questions will enable evaluation of whether men are free as claimed.
“Freedom is always and necessarily from restraint; thus, in so far as the adherents of positive freedom speak of persons being made free by means of restraint, they cannot be talking about freedom.” Let us examine the implications of this argument made by friends of “negative” freedom by investigating how we can we can make sense of the alleged claim of adherents of “positive” liberty that, for example, Smith is (or can be) made free by restraining (constraining, coercing) him.
The first interpretation is that “restraining Smith by means a [say, a regulation] from doing b [that prevents his crossing the streets wherever he likes] produces a situation in which he is now able to do c [but allows him to have a right of way over automobile traffic at pedestrian crossings] because restraint d [while abolishing the automobiles having general right of way over pedestrians] is lifted. He is thereby, by means of restraint a, made free from d to do c, although he can no longer do b.”
This interpretation is straightforward. It presents problems only if it is assumed that persons are free or not free simpliciter and also that the claim in question is that Smith can be made free simpliciter. If these assumptions are made, the following interpretation might be appropriate. Smith is not being “restrained” but being helped to do what he really wants to do or what he would do if he were reasonable (moral, prudent). The “constraint” put on him actually lifts a genuine constraint (ignorance, passion) that was upon him.
This is not at all straightforward. However, it can be disentangled by insisting on the specifications of the triadic relationship being advanced. What, for instance, is Smith being made free from? Perhaps he is made free from the constraint produced by the arbitrary uncontrolled actions of other residents, or perhaps it is the “constraint” arising from his own ignorance or passion, or perhaps it’s both. If it’s the former, the specification is straightforward. If it’s the latter, further argument will be needed for it is difficult to find the range of passion or ignorance that might limit freedom.
Who, for another, is the “true” Smith? The answer will be met if the third specification of the triadic relationship, what Smith is made free to do, is examined. Apparently, he is made free to do as he wishes, really wishes, or would wish if he were reasonable. But there is obviously something he is not free to do. That is the whole point of restraining Smith. But what is he not free to do? The problem with this question is realised when we realise that what usually appears as a “restraint” is not a restraint at all.
These comments do not seek to analyse in depth the claim made by “friends” of negative liberty. Rather, they are being made to examine and draw attention to the variety of interpretations that the analysis of freedom as a triadic relationship throws up. And these are interpretations that the “friends” of negative liberty do not consider or anticipate.
“In the end, then, discussions of the freedom of agents can be fully intelligible and rationally assessed only after the specification of each term of this triadic relation has been made or at least understood. The principal claim made here has been that insistence upon this single “concept” of freedom puts us in a position to see the interesting and important ranges of issues separating the philosophers who write about freedom in such different ways, and the ideologies that treat freedom so differently.”
 “Identified [by Isaiah Berlin in Two Concepts of Liberty]… as adherents of “negative” freedom, one finds Occam, Erasmus, Hobbes, Locke, Bentham, Constant, J. S. Mill, Tocqueville, Jefferson, Burke, Paine. Among adherents of “positive” freedom one finds Plato, Epictetus, St. Ambrose, Montesquieu, Spinoza, Kant, Herder, Rousseau, Hegel, Fichte, Marx, Bukharin, Comte, Carlyle, T. H. Green, Bradley, Bosanquet.”