DA1 This essay is presented as an example of 'Political Philosophy'. I should not normally be too concerned about the proper way to describe my efforts, but in this case I am expected to produce political philosophy and so I must be prepared to defend the claim that I am doing so. Philosophy is not so keen to fall into categories as people might like it to be, and I shall find it necessary to trespass upon much that is not properly 'political' in the course of my essay. Among these transgressions I expect to figure prominently methodological deliberations, metaphysics, and perhaps, moral philosophy. I find it helpful in my philosophical wanderings to have a focus or a home base, from which I shall wander, but from which I shall return from time to time in the hope of sewing the material into some sort of coherent whole; the philosophy department has a taste for titles, so my focus and my title will be simply 'liberty'1. So as not to spring too many surprises on the reader I shall here give a brief prognosis. I expect to begin with a touch of methodology, and proceed from there to a few words on 'free will', the metaphysical aspects, and thence into the political problem. ('metaphysical' may well be a misclassification there, in which case I beg the reader's indulgence.)
DA2 One of my methodological objectives is to plot a course between the Scylla of Platonism and the Charybdis of 'common usage' worship, and generally to evade as far as possible disputes about 'the meaning' of words. Language, I suggest, is as it is largely as a matter of fact rather than as a matter of necessity2. It is the business of science to establish the facts, philosophy should properly be concerned with the complementary notions of 'necessity' and 'possibility', and particular languages are the tools of philosophy not its subject matter. Would that the matter were so simple! However, I don't want to identify philosophy with logic either.
DA3 When I concern myself with political philosophy I most certainly am not primarily interested in the meanings of words. I am concerned with the possible ways in which human beings might organise themselves into societies, and with the pro's and con's of these various political schemes. Even the word 'political' is unimportant. I am not concerned with the political aspects of such societies alone, I am concerned with all that might be said for and against them, whether or not it strictly concerns the political organisation of the society. At no stage is it necessary for me to settle precisely 'the' meaning of a word. What matters is that I should try to be as clear as possible about what I mean. I need make no pretence to using language 'properly', so long as I use it precisely and unambiguously. The approach of both Plato and Wittgenstein to disputes about meanings are side issues, for I am not using a language fixed ideally in heaven, nor one strictly determined by 'common usage', but rather one adapted from reasonably common precedent to suit my own purposes.
DA4 The contingencies of language are not the only impediment to a study, from a philosophical standpoint, of political institutions. The sphere of logical possibility unrestricted by contingent assumptions is too large to illuminate our problem. We need to arm ourselves with a whole host of contingent assumptions (or 'facts') before we can arrive at any interesting conclusions. Normally these things are assumed more or less covertly, jsut as in mathematical proofs many steps are taken so naturally that a mathematician is often unaware of the full range of assumptions necessary to make his proof valid. A mathematical logician might see it as part of his business to encourage greater rigour in mathematical demonstrations, in the form of more precise definitions of concepts and more complete enumeration of premises. In producing more rigorous proofs in this way a mathematician makes the significance of his conclusions much clearer and the likelyhood of error much smaller. A philosophical logician might hope to achieve comparable results in philosophy by similar methods3.
DA5 No doubt any attempt to reduce philosophy to logic would suffer even more strenous opposition than did the logicist philosophy of mathematics, the view of pure truths of logic as being in themselves devoid of content and hence devoid of value is hard to dislodge. Nevertheless, I follow Descartes in wishing to see philosophy emulate more the merits of mathematics4, and though the logicists failed to convince all that mathematics is purely logic, the main grounds for their dismissal were the need of premises which were neither obviously true nor obviously logical, in order to prove the truths of mathematics. We need only supply these premises as hypotheses and we obtain genuine logical truths. The body of mathematics is deduced by pure logic from definitions and premises which are themselves justified, if at all, only on pragmatic or aesthetic grounds, the grounds that they give satisfactory results in economic or otherwise pleasing ways. And yet mathematics is by no means entirely vacuous and futile5. It seems that if we clothe a logical truth in suitably worldly attire it can be of the greatest practical utility5b. So if I emphasise that the core of philosophy is logic, I am not attempting to reduce it to the trivial.
DA6 Here then is my methodological starting point. Philosophy is, like mathematics, a matter of finding suitably clothed logical truths, it differs from mathematics in the sort of clothes it deals in. It deals with contingent truths (i.e. those truths which are not truths of logic), and falsehoods either by assuming them or by deducing them from other assumed contingent propositions. It should allways be clear what assumptions are made and by what logic the conclusions follow from these conclusions so that the reader might construct for himself from the text a suitable but perhaps lengthy statement which distills all that the philosopher has said in his argument, and is a truth of pure logic. As far as the language which the philosopher uses for this purpose, no-one need dictate to him, it is up to him to balance the advantages to be derived from using common terms in slightly uncommon ways, against the disadvantages. A study of common usage may be of some assistance in teaching a philosopher when he needs to clarify his own usage, and it may help him better to know when his language may create more mist than it dispels, but it is not his primary business any more than the mating calls of the Blue Tit. The 'method' then consists in selecting suitable definitions, and adopting various assumptions, plausible or otherwise, and proceding therefrom by logical deduction. That is the format of all that I will admit as proof.6
DA7 I must allow myself a little more lattitude than that, for I have certainly been exercising it. Licence indeed for unbridled speculation. Provided only that we do not pretend to a proof when we have none. Speculation may well often be a fine thing, so long as we recognise it for what it is.7
DA8 Now I shall take a brief look at the 'free will' issue. One might reasonably expect that there is some relationship between the problems discussed under that head and the problems associated with 'freedom' or 'liberty' (which are at least roughly synonymous in common use) as political concepts. The free will issue is a debate about meanings, and hence about accepted proper usage of the language. (I am taking sides with 'Wittgenstein') Questions of freedom in the political contexts depend upon the same meanings, and generally presuppose that the freewill issue has been favourably settled. The question essentially is under what conditions is an action done freely. At the one level the main objective is to arrive at an answer which will give us the dignity of genuinely being free under most of the normal circumstances when we are accustomed to think we are.8 At the other level this result is presumed and also the result that the sort of external circumstances generally supposed to limit our freedom (e.g. physical restraint, or threat of same) does indeed do so. We are then concerned to consider what role such external limitations do or might play in the functions of society.
DA9 The dialectics of the free-will problem, as commonly handled by philosophers, fall outside what I have above suggested as the province of philosophy. That is because they rely on the settlement of linguistic disputes, and linguistics is not philosophy. As far as my interest in freedom is concerned, I can quite easily sidestep the problem, for I am not particularly interested in common language and after the manner suggested previously I am content to use the concept in my own particular way. Of course that still leaves me with the problem of clarifying for the reader what I mean by freedom, and I am sure I will find that difficult enough without trying to answer for the man on the Clapham omnibus. This allows me to say, by fiat as it were, that what I mean by freedom is at least not logically tied to the issue of determinism. This does not settle the meaning of freedom, but it cuts down on the options somewhat.9
DA10 In fact I have little intention of even attempting to formulate a definition of freedom adequate to my purposes. For I think that no such single concept will be adequate for the enquiries I have in mind. It would serve me better to replace that bald dichotomy with something a little more subtle. We might express this by saying that a man is not free or unfree, but rather that he is more or less free, and we would then be bound to suitably alter all enquiries about freedom. Instead of 'is freedom a polical right?' we would have 'how much freedom is a political right?'.10 But even this is too crude a usage to get us very far. For it assumes that freedom can be suitably totted up to give us an overall figure for any man, but any scheme which allowed for that would be bound to turn out rather artificial and I doubt it would serve us well. When we enquire after someones freedom a general overall judgement is not the best we can hope for. Instead of being told he is just so free, we would be better served if we were told he is free in this respect and in that respect but not free in these other matters.11 Any overall scheme would have to rely on weighting each freedom in some way (at its crudest counting them all equal) and has the same implausible ring about it as the suggestion that we can reasonably be expected to tot up the total pleasure accruing to mankind as a consequence of our actions. To get a proper appreciation of the measure of someones freedom we really need an enumeration of the particular freedoms he enjoys, and of the constraints he suffers.12
DA11 We are not yet at an end to the complications though. Even this enumeration of freedoms does not tell us all we might like to know. Our freedom may be limited in a variety of ways some of them more forceful than others. If I claim that I am not free to take a certain course you may be better placed to assess my grievance if you enquire of me what it is that prevents me. Only in the simplest cases are we deprived of our freedom by causes which physically incapacitate us from pursuing our course. More usually our problem is that the prohibited course would be followed by consequences to grave to be willingly incurred. These consequences might be the subtle consequences of social disapproval or they might be the cruder ones of legal action. When such consequences are sever enough to deprive us of our liberty is a very fluid matter. Even ordinary discourse can give little light to that problem13, for it seems we are wont to claim or disclaim freedom depending upon whether it is our desire to emphasise the absence or presence of a certain sort of constraint, the sort of constraint under consideration being determined by the context rather tan any notion of what kind of constraint is invariably sufficient to deprive us of freedom14. If I am engaged in discussing politics I might claim to be free to do something on the ground that there is no law against it, if we discuss morality I may claim to be not free to do that same action because my conscience dictates otherwise, and if we discuss metaphysics I may claim to be free in no respect whatsoever, because the universe is deterministic and so I could not do otherwise than I do. In these ways the meaning of the concept shifts, and in yet more subtle ways.15
DA12 Let me now try to recapitulate and clarify some of the preceding material. It is my task to delve into some of the problems associated with concept of freedom or liberty, while not concerning myself too closely with the niceties of common usage. It might be supposed from my previous remarks about 'methodology' that I should proceed first of all by adopting some definition of freedom which though not precisely sqaring with common usage yet comes close enough to be worth consideration. The last two paragraphs were intended to illustrate in some measure why this procedure is unsatisfactory. What I am trying to suggest is that the variety of uses of the concept is so great that to select any one will not do. The criterion for judgements as to freedom vary from occasion to occasion just as do our judgements about whether something is hot or cold. Science makes progress in its investigation of the consequences of heat and cold by abandoning any important use of the concepts hot and cold, instead of giving these concepts a precise scientific definition it is found better to deal with the concept of termperature, and make that precise.16 In the same way with the concept of freedom we can say that we would do better to rest little weight on the word 'free' and look instead to these underlying factors. It is of little avail to consider the question 'am I free'17, for there is so much weighing and balancing to be done, and there is so much variation in the criteria applied to such questions that we could in the majority of cases make a good case for either answer to such a question based on common usage. And yet the fluidity of the concept is so much a part of its ordinary usage that any definition which made it useful would be but a mockery of common usage. The option adopted with regard to 'hot' amd 'cold', that of replacing a dichotomy by a continuous one dimensional spectrum, will not help much here, even though it is to some extent licenced by the legitimacy of comparative judgements in ordinary language (i.e. saying that someone has more or less freedom than another). Unlike temperature freedoms cannot readily be quantified. So if we want to talk about someones freedom or lack of it, to say 'he is free' or 'he is not free' would tell us very little indeed, and the possibility of saying something like 'he has 45 British standard freedom units' is uninviting.18 The best we can do is to enumerate the particular respects in which he is free and those in which he is not, and as I suggested even these particular judgements may mean little unless we attach to them an account of the factors which enforce these freedoms or lack of freedom.
DA13 A good start, if we want to know how free members of a society are, is to look at the laws of the society.19 The existence of any law at all is sufficient to tell us that the individual has less than perfect freedom (though the absence of laws would be no means confer absolute freedom).20 Many other factors are of equal or greater importance. The way in which the laws are enforced is important, some laws may be virtually neglected and others may be enforced by punishments slight enough to constitute poor deterrent. There may be laws vague enough to give almost unlimited scope for prosecution, the effect of which upon the freedom of individuals will depend entirely upon the police and the courts (for example the law against consiracy in Britain, under which the police may obtain convictions for conspiracy to do things which are not themselves offences). Apart from the legal system there are many other things which may extend or restrict the freedom of some or all individuals. Almost anything may be subject to discriminatory practice, which may or may not be considered to effectively limit the freedom of those discriminated against, but which whether or not it qualifies as a loss of freedom will often have equally distressing results, and may equally well claim to be 'unjust'. An example of such discrimination is in the access to education, which is heavily weighted in favour of the intellectually gifted. We may argue that the organisation of the educational system is reasonable and just despite such 'discrimination' but we cannot deny that a man whose I.Q. is 90 is as effectively barred from taking advantages of the facilities offered by our Universities as he is from taking his clothes off in public, and that the former deprivation may be the greater cause of concern to him.21
DA14 My problem remains, how to get this onto a philosophical plain. Now although I have said that it will not do for me to define the concept of freedom for myself, I think it might be helpful to consider ways in which that might have been attempted. The reason why no definition is satsifactory is that there are many possible criteria which we might select, none of which is indisputably superior to its fellows.22 We might take a persons freedom to be reduced when he is prevented from taking some course of action which he would be physically capable of taking, through fear of sanctions which would be imposed on him subsequently by the rest of society (that being suggested as a sufficient but not a necessary condition of such a reduction.) This partial definition doubtless has many flaws. What is to count as a sanction for the purposes of this definition? It would be unsatisfactory to count any social response which effectively changed someones behaviour, since any potential manifestation of dissaproval, from the right quarters, might be more effective than even legal sanctions, but we don't usually consider these to deprive us of our freedom.23 No conclusive case is likely to be possible in favour of any classification of sanctions into those depriving of freedom and those not doing so, because of the fluidity of the concept of freedom. However, even though such a classification may not give rise to a satisfying definition of freedom, it may nevertheless assist us in our investigation of the philosophical problems associated with that concept. The problem of investigating the role of freedom or lack of it in the functioning of society is opaque due to the unsatisfactory nature of the concept of freedom, but we may effectively make some progress in this area if we consider various classifications of the various influences in society on the behaviour of its members, and the role which these various sorts of influence might have in the functioning of society. Each class of influence might be considered as representing a distinct definition of freedom, i.e. that sort of freedom which is eroded by the influences in that class, provided that we tolerate a splendidly broad notion of 'sort of freedom'.
DA15 We must still be clear of the difference between sociology, anthropology, political science and philosophy. In their own ways each of these disciplines might adopt similar tactics and so we must be clear of the distinctions to be drawn between the ways such disciplines use such devices. It is sufficient to draw the line between the philosophical and the rest. It is obvious, the philosophical considers what might be, the scientific considers what is. So from a scientific viewpoint the problem is to establish what role each of these classes of influence plays in the regulating of society, while from the philosophical viewpoint the problem is to establish what role such classes of influence might play. And how does the philosopher approach this nebulous problem? Well, I have suggested an approach earlier, he plays games with assumptions, he sees what propositions follow from what assumptions, or he considers what assumptions are necessary to establish some desirable result; this seems all that he can do, for if he once ceases to deal in tautologies (meaning by that logical truths) then by trespassing on the contingent he trespasses on that which stands in need of experimental confirmation, and experiment is the province of science. This business of selecting assumptions and deducing from them is certainly not going to be an easy matter, the difficulty not so much lying in conforming to these prescriptions as in producing something of interest in the process; we could easily enough draw a million subtle inferences and feel non the wiser for it in the end. What is the point of it? Well it ought to be of some use to scientists. Before one can reasonably assess the extent to which experimental evidence confirms a hypothesis it is necessary to have some appreciation of the range of alternative hypotheses which are consistent with the evidence. Viewed in this way philosophy is invariably a necessary prerequisite of any scientific activity.
DA16 In the area which concerns me science has not had its greatest successes, social sciences generally have failed to produce the sort of solid predictive successes we have grown accustomed to in the physical sciences. Consequently political science has never been so clearly severed from philosophy as have such sciences as physics and chemistry. Writings on political philosophy have often looked more like armchair science, and when they do not fall to making contingent claims in the sphere of political science, they often avoid that flaw by engaging solely in linguistic studies.
DA17 The plight of political science is of course of the greatest relevance to political philosophy. If we are to accept my simple account of what political philosophy should be engaged in, i.e. a consideration of what is logically possible which will help the political scientist by enabling him more clearly to see what possibilities are consistent with the evidence he has at hand, then this will certainly be much more helpful to the scientist if the possibilities are also plausibilities. The philosopher, to be of any use must not only be equipped to determine what is possible, but also to estimate what is likely. It is here that the main problem arises, not only in politics but in any science which involves predicting people; people are difficult to predict. Through a long evolutionary process we have evolved a very sophisticated central nervous system, the purpose of which is to give a more flexible response to the problems which the organism meets in a more or less hostile environment. This makes the organism much more capable of falsifying hypotheses than most things. In addition to this there have evolved tendencies which tend to support the survival of the species rather than and possibly at the cost of the individual. So individuals cannot even be relied upon to act in the pursuance of their own interests, they may sacrifice their well being or even their lives for 'greater ends'. Worse even than the idealist (from the point of view of predictability) is the madman, and we must take account of him also in our schemes.
DA18 In the light of that difficulty it might be best to make our first 'assumption', that people are unpredicatable. It needs to be carefully formalised becase it is their psychology which is unpredictable, not their physiology.24 There are further difficulties in separating the psychological from the physiological but these are not important enough to deserve lengthy consideration at this stage. So I shall assume that it is understood what the phrase 'physically capable' means and I shall use this concept to formulate my assumption about predictability. I want to make no assumption at all about the physical capabilities of human beings. What I want to assume is that any true law-like statement about the behaviour of human beings is true in virtue of the physical capabilities of human beings and the laws of logic, with no other assumptions being necessary. This is a complicated way of saying that it is a law about physical capabilities of human beings. I am supposing this to be equivalent to the claim that it is never possible to exclude in advance the adoption of any course of action of which a human being is physically capable.; this equivalence follows from the need for any significant psychological law to be falsifiable, and for some of the possible counter instances to be occurrences not to be covered by laws about physical capabilities(if it makes sense to suppose that a law may be a psychological law while not excluding any course of action which the human being is physically capable then I am not concerned with such laws as distinct from laws about physical capabilities and so when I talk about psychological laws I want to exclude those oddities). One more problem regarding the equivalence of the two formulation is the possible existence of statistical laws about psychology. Because these laws are not categorically falsifiable the second formulation may be true even if there are true statistical laws, though the first would not. The first formulation may be taken as definitive and the second though entailed by it is not for the reason just given equivalent.
DA19 This assumption will not get us very far, and we might reasonably hope for its falsity, but it will serve as a starting point and give us some idea of what may be had from this approach to political philosophy. The assumption is negative in character and we may reasonably expect it to give us negative results. We are concerned with the relationship between freedom, or lack of it, and the functioning of society. If we translate this problem in the light of the previous discussion of 'freedom' we arrive at a host of related problems concerning, for example, the relationship between the punitive sanctions (or any sort of sanction) and the behaviour these sanctions are intended to prevent or the consequences they seek to avert. Whenever sanctions fall short of physically incapacitating poeple from committing the proscribed acts they rely for their effectiveness upon psychological laws. Our unpredictability assumption would therefore entitle us to conclude that no such system which relied in part upon measures depending on the psychology of human beings could be relied upon to be completely effective, and on the other hand, any system whatever or possibly no system at all might have the desired result (any desired result not excluded by the laws about physical capabilities.) To emphasise the trivial nature of the inference, I point out that if either of the conclusions were false we would be in a position to derive from the evidence which showed them false evidence which supported equally strongly some psychological law, which our assumption excludes.
DA20 This conclusion might be encouraging to anarchists, they being people more vociferously concerned with liberty than most, since the first conclusion superficially at least casts suspicion upon the effectiveness of the 'repressive mechanism of the state' in securing social well being, and in the second endorses the possibility of the anarchist society in which the state has been abolished. But the suspicion cast is not very substantial, for though we conclude that they might not work (repressive mechanisms or what have you), we do not conclude that they will not work, and though we affirm the possibility of a working anarchist state, we do not affirm that there ever will be such a state or that any course of reform or revolution would lead to one. In fact we have to affirm that a dictatorial state might result in the best of all possible worlds and to deny that any course whatever can be relied upon to secure a working anarchist state. The results are entirely negative consisting of the denial of necessities and the affirmation of possibilities. (you may dispute that that counts as 'entirely negative', being as it is a mixture of denials and affirmatives, but since any negation of possibility or necessity is equivalent to an affirmation of a necessity or a possibility we can give the phrase entirely negative no meaning at all unless we use it either to describe collections containing only the denials of necessities and affirmations of possibilities or to collections containing only the affirmations of necessities and the denials of possibilities. Since we are concerned with laws, i.e. necessities, the former usage is more apt.)
DA21 Of course, the alleged corrolaries in paragraph 20 will depend for their validity upon chosing suitable definitions for the various extra concepts used in stating them. I took no special care in wording them because I do not propose to carefully substantiate them. They were intended only to be suggestive of ways in which the results might be extended, or to point to other areas where they might be thought of some relevance. I don't propose to go much farther into the consequences of this hypothesis because it is clear that they will be of a limited nature and that we will obtain more interesting results if we can find an alternative hypothesis which claims human psychology to be something less than totally unpredictable. I shall follow the procedure of axiomatic mathematics by adopting the weakest possible alternative hypothesis first, which is in this case simply the negation of our original hypothesis. The advantage in this procedure is that if we first of all assume that psychologically human beings are not totally unpredictable, then whatever results we obtain with this weak hypothesis will automatically apply whenever we adopt a hypothesis which entails the weak hypothesis, i.e. whenever we assume the truth of any psychological law. By dealing with this hypothesis first we stand to avoid duplication of effort.
DA22 Our hypothesis, being the negation of the original, is that it is not the case that every true law-like statement about the behaviour of human beings is a law about the physical capabilities of human beings. This hypothesis does not entail, in virtue of the comments at the end of para. 18, the claim that we can sometimes, given sufficient knowledge of the circumstances, exclude in advance the adoption of some course of action by a human being who would in those circumstances be physically incapable of adopting that course, since the psychological laws may all be of a statistical nature, in which case behaviours would only be rendered probably or improbable.