Triadic Relationship Definition Essay

1. Two Concepts of Liberty

Imagine you are driving a car through town, and you come to a fork in the road. You turn left, but no one was forcing you to go one way or the other. Next you come to a crossroads. You turn right, but no one was preventing you from going left or straight on. There is no traffic to speak of and there are no diversions or police roadblocks. So you seem, as a driver, to be completely free. But this picture of your situation might change quite dramatically if we consider that the reason you went left and then right is that you're addicted to cigarettes and you're desperate to get to the tobacconists before it closes. Rather than driving, you feel you are being driven, as your urge to smoke leads you uncontrollably to turn the wheel first to the left and then to the right. Moreover, you're perfectly aware that your turning right at the crossroads means you'll probably miss a train that was to take you to an appointment you care about very much. You long to be free of this irrational desire that is not only threatening your longevity but is also stopping you right now from doing what you think you ought to be doing.

This story gives us two contrasting ways of thinking of liberty. On the one hand, one can think of liberty as the absence of obstacles external to the agent. You are free if no one is stopping you from doing whatever you might want to do. In the above story you appear, in this sense, to be free. On the other hand, one can think of liberty as the presence of control on the part of the agent. To be free, you must be self-determined, which is to say that you must be able to control your own destiny in your own interests. In the above story you appear, in this sense, to be unfree: you are not in control of your own destiny, as you are failing to control a passion that you yourself would rather be rid of and which is preventing you from realizing what you recognize to be your true interests. One might say that while on the first view liberty is simply about how many doors are open to the agent, on the second view it is more about going through the right doors for the right reasons.

In a famous essay first published in 1958, Isaiah Berlin called these two concepts of liberty negative and positive respectively (Berlin 1969).[1] The reason for using these labels is that in the first case liberty seems to be a mere absence of something (i.e. of obstacles, barriers, constraints or interference from others), whereas in the second case it seems to require the presence of something (i.e. of control, self-mastery, self-determination or self-realization). In Berlin's words, we use the negative concept of liberty in attempting to answer the question “What is the area within which the subject — a person or group of persons — is or should be left to do or be what he is able to do or be, without interference by other persons?”, whereas we use the positive concept in attempting to answer the question “What, or who, is the source of control or interference that can determine someone to do, or be, this rather than that?” (1969, pp. 121–22).

It is useful to think of the difference between the two concepts in terms of the difference between factors that are external and factors that are internal to the agent. While theorists of negative freedom are primarily interested in the degree to which individuals or groups suffer interference from external bodies, theorists of positive freedom are more attentive to the internal factors affecting the degree to which individuals or groups act autonomously. Given this difference, one might be tempted to think that a political philosopher should concentrate exclusively on negative freedom, a concern with positive freedom being more relevant to psychology or individual morality than to political and social institutions. This, however, would be premature, for among the most hotly debated issues in political philosophy are the following: Is the positive concept of freedom a political concept? Can individuals or groups achieve positive freedom through political action? Is it possible for the state to promote the positive freedom of citizens on their behalf? And if so, is it desirable for the state to do so? The classic texts in the history of western political thought are divided over how these questions should be answered: theorists in the classical liberal tradition, like Constant, Humboldt, Spencer and Mill, are typically classed as answering ‘no’ and therefore as defending a negative concept of political freedom; theorists that are critical of this tradition, like Rousseau, Hegel, Marx and T.H. Green, are typically classed as answering ‘yes’ and as defending a positive concept of political freedom.

In its political form, positive freedom has often been thought of as necessarily achieved through a collectivity. Perhaps the clearest case is that of Rousseau's theory of freedom, according to which individual freedom is achieved through participation in the process whereby one's community exercises collective control over its own affairs in accordance with the ‘general will’. Put in the simplest terms, one might say that a democratic society is a free society because it is a self-determined society, and that a member of that society is free to the extent that he or she participates in its democratic process. But there are also individualist applications of the concept of positive freedom. For example, it is sometimes said that a government should aim actively to create the conditions necessary for individuals to be self-sufficient or to achieve self-realization. The welfare state has sometimes been defended on this basis, as has the idea of a universal basic income. The negative concept of freedom, on the other hand, is most commonly assumed in liberal defences of the constitutional liberties typical of liberal-democratic societies, such as freedom of movement, freedom of religion, and freedom of speech, and in arguments against paternalist or moralist state intervention. It is also often invoked in defences of the right to private property. This said, some philosophers have contested the claim that private property necessarily enhances negative liberty (Cohen 1991, 1995), and still others have tried to show that negative liberty can ground a form of egalitarianism (Steiner 1994).

After Berlin, the most widely cited and best developed analyses of the negative concept of liberty include Hayek (1960), Day (1971), Oppenheim (1981), Miller (1983) and Steiner (1994). Among the most prominent contemporary analyses of the positive concept of liberty are Milne (1968), Gibbs (1976), C. Taylor (1979) and Christman (1991, 2005).

2. The Paradox of Positive Liberty

Many liberals, including Berlin, have suggested that the positive concept of liberty carries with it a danger of authoritarianism. Consider the fate of a permanent and oppressed minority. Because the members of this minority participate in a democratic process characterized by majority rule, they might be said to be free on the grounds that they are members of a society exercising self-control over its own affairs. But they are oppressed, and so are surely unfree. Moreover, it is not necessary to see a society as democratic in order to see it as self-controlled; one might instead adopt an organic conception of society, according to which the collectivity is to be thought of as a living organism, and one might believe that this organism will only act rationally, will only be in control of itself, when its various parts are brought into line with some rational plan devised by its wise governors (who, to extend the metaphor, might be thought of as the organism's brain). In this case, even the majority might be oppressed in the name of liberty.

Such justifications of oppression in the name of liberty are no mere products of the liberal imagination, for there are notorious historical examples of their endorsement by authoritarian political leaders. Berlin, himself a liberal and writing during the cold war, was clearly moved by the way in which the apparently noble ideal of freedom as self-mastery or self-realization had been twisted and distorted by the totalitarian dictators of the twentieth century — most notably those of the Soviet Union — so as to claim that they, rather than the liberal West, were the true champions of freedom. The slippery slope towards this paradoxical conclusion begins, according to Berlin, with the idea of a divided self. To illustrate: the smoker in our story provides a clear example of a divided self, for she is both a self that desires to get to an appointment and a self that desires to get to the tobacconists, and these two desires are in conflict. We can now enrich this story in a plausible way by adding that one of these selves — the keeper of appointments — is superior to the other: the self that is a keeper of appointments is thus a ‘higher’ self, and the self that is a smoker is a ‘lower’ self. The higher self is the rational, reflecting self, the self that is capable of moral action and of taking responsibility for what she does. This is the true self, for rational reflection and moral responsibility are the features of humans that mark them off from other animals. The lower self, on the other hand, is the self of the passions, of unreflecting desires and irrational impulses. One is free, then, when one's higher, rational self is in control and one is not a slave to one's passions or to one's merely empirical self. The next step down the slippery slope consists in pointing out that some individuals are more rational than others, and can therefore know best what is in their and others' rational interests. This allows them to say that by forcing people less rational than themselves to do the rational thing and thus to realize their true selves, they are in fact liberating them from their merely empirical desires. Occasionally, Berlin says, the defender of positive freedom will take an additional step that consists in conceiving of the self as wider than the individual and as represented by an organic social whole — “a tribe, a race, a church, a state, the great society of the living and the dead and the yet unborn”. The true interests of the individual are to be identified with the interests of this whole, and individuals can and should be coerced into fulfilling these interests, for they would not resist coercion if they were as rational and wise as their coercers. “Once I take this view”, Berlin says, “I am in a position to ignore the actual wishes of men or societies, to bully, oppress, torture in the name, and on behalf, of their ‘real’ selves, in the secure knowledge that whatever is the true goal of man ... must be identical with his freedom” (Berlin 1969, pp. 132–33).

Those in the negative camp try to cut off this line of reasoning at the first step, by denying that there is any necessary relation between one's freedom and one's desires. Since one is free to the extent that one is externally unprevented from doing things, they say, one can be free to do what one does not desire to do. If being free meant being unprevented from realizing one's desires, then one could, again paradoxically, reduce one's unfreedom by coming to desire fewer of the things one is unfree to do. One could become free simply by contenting oneself with one's situation. A perfectly contented slave is perfectly free to realize all of her desires. Nevertheless, we tend to think of slavery as the opposite of freedom. More generally, freedom is not to be confused with happiness, for in logical terms there is nothing to stop a free person from being unhappy or an unfree person from being happy. The happy person might feel free, but whether they are free is another matter (Day, 1970). Negative theorists of freedom therefore tend to say not that having freedom means being unprevented from doing as one desires, but that it means being unprevented from doing whatever one might desire to do (Steiner 1994. Cf. Van Parijs 1995; Sugden 2006).

Some theorists of positive freedom bite the bullet and say that the contented slave is indeed free — that in order to be free the individual must learn, not so much to dominate certain merely empirical desires, but to rid herself of them. She must, in other words, remove as many of her desires as possible. As Berlin puts it, if I have a wounded leg ‘there are two methods of freeing myself from pain. One is to heal the wound. But if the cure is too difficult or uncertain, there is another method. I can get rid of the wound by cutting off my leg’ (1969, pp. 135–36). This is the strategy of liberation adopted by ascetics, stoics and Buddhist sages. It involves a ‘retreat into an inner citadel’ — a soul or a purely noumenal self — in which the individual is immune to any outside forces. But this state, even if it can be achieved, is not one that liberals would want to call one of freedom, for it again risks masking important forms of oppression. It is, after all, often in coming to terms with excessive external limitations in society that individuals retreat into themselves, pretending to themselves that they do not really desire the worldly goods or pleasures they have been denied. Moreover, the removal of desires may also be an effect of outside forces, such as brainwashing, which we should hardly want to call a realization of freedom.

Because the concept of negative freedom concentrates on the external sphere in which individuals interact, it seems to provide a better guarantee against the dangers of paternalism and authoritarianism perceived by Berlin. To promote negative freedom is to promote the existence of a sphere of action within which the individual is sovereign, and within which she can pursue her own projects subject only to the constraint that she respect the spheres of others. Humboldt and Mill, both advocates of negative freedom, compared the development of an individual to that of a plant: individuals, like plants, must be allowed to grow, in the sense of developing their own faculties to the full and according to their own inner logic. Personal growth is something that cannot be imposed from without, but must come from within the individual.

3. Two Attempts to Create a Third Way

Critics, however, have objected that the ideal described by Humboldt and Mill looks much more like a positive concept of liberty than a negative one. Positive liberty consists, they say, in exactly this growth of the individual: the free individual is one that develops, determines and changes her own desires and interests autonomously and from within. This is not liberty as the mere absence of obstacles, but liberty as autonomy or self-realization. Why should the mere absence of state interference be thought to guarantee such growth? Is there not some third way between the extremes of totalitarianism and the minimal state of the classical liberals — some non-paternalist, non-authoritarian means by which positive liberty in the above sense can be actively promoted?

3.1 Positive Liberty as Content-neutral

Much of the more recent work on positive liberty has been motivated by a dissatisfaction with the ideal of negative liberty combined with an awareness of the possible abuses of the positive concept so forcefully exposed by Berlin. John Christman (1991, 2005, 2009), for example, has argued that positive liberty concerns the ways in which desires are formed — whether as a result of rational reflection on all the options available, or as a result of pressure, manipulation or ignorance. What it does not regard, he says, is the content of an individual's desires. The promotion of positive freedom need not therefore involve the claim that there is only one right answer to the question of how a person should live, nor need it allow, or even be compatible with, a society forcing its members into given patterns of behavior. Take the example of a Muslim woman who claims to espouse the fundamentalist doctrines generally followed by her family and the community in which she lives. On Christman's account, this person is positively unfree if her desire to conform was somehow oppressively imposed upon her through indoctrination, manipulation or deceit. She is positively free, on the other hand, if she arrived at her desire to conform while aware of other reasonable options and she weighed and assessed these other options rationally. Even if this woman seems to have a preference for subservient behavior, there is nothing necessarily freedom-enhancing or freedom-restricting about her having the desires she has, since freedom regards not the content of these desires but their mode of formation. On this view, forcing her to do certain things rather than others can never make her more free, and Berlin's paradox of positive freedom would seem to have been avoided.

It remains to be seen, however, just what a state can do, in practice, to promote positive liberty in Christman's sense without encroaching on any individual's sphere of negative liberty: the conflict between the two ideals seems to survive his alternative analysis, albeit in a milder form. Even if we rule out coercing individuals into specific patterns of behavior, a state interested in promoting autonomy in Christman's sense might still be allowed considerable space for intervention of an informative and educational nature, perhaps subsidizing some activities (in order to encourage a plurality of genuine options) and financing this through taxation. Liberals might criticize this on anti-paternalist grounds, objecting that such measures will require the state to use resources in ways that the supposedly heteronomous individuals, if left to themselves, might have chosen to spend in other ways. Some liberals will make an exception in the case of the education of children (in such a way as to cultivate open minds and rational reflection), but even here other liberals will object that the right to negative liberty includes the right to decide how one's children should be educated.

3.2 Republican Liberty

Other theorists of liberty have remained closer to the negative concept but have attempted to go beyond it, saying that liberty is not merely the enjoyment of a sphere of non-interference but the enjoyment of certain conditions in which such non-interference is guaranteed (see especially Pettit 1997, 2001, 2014, and Skinner 1998, 2002). These conditions may include the presence of a democratic constitution and a series of safeguards against a government wielding power arbitrarily, including the separation of powers and the exercise of civic virtues on the part of citizens. As Berlin admits, on the negative view, I am free even if I live in a dictatorship just as long as the dictator happens, on a whim, not to interfere with me (see also Hayek 1960). There is no necessary connection between negative liberty and any particular form of government. On the alternative view sketched here, I am free only if I live in a society with the kinds of political institutions that guarantee the independence of each citizen from exercises of arbitrary power. Quentin Skinner has called this view of freedom ‘neo-Roman’, invoking ideas about freedom both of the ancient Romans and of a number of Renaissance and early modern writers. Philip Pettit has called the same view ‘republican’, and this label has tended to dominate in the recent literature (Weinstock and Nadeau 2004; Larmore 2004; Laborde and Maynor 2008).

Republican freedom can be thought of as a kind of status: to be a free person is to enjoy the rights and privileges attached to the status of republican citizenship, whereas the paradigm of the unfree person is the slave. Freedom is not simply a matter of non-interference, for a slave may enjoy a great deal of non-interference at the whim of her master. What makes her unfree is her status, such that she is permanently liable to interference of any kind. Even if the slave enjoys non-interference, she is, as Pettit puts it, ‘dominated’, because she is permanently subject to the arbitrary power of her owner.

Contemporary republicans therefore claim that their view of freedom is quite distinct from the negative view of freedom. As we have seen, one can enjoy non-interference without enjoying non-domination; conversely, according to Pettit, one can enjoy non-domination while nevertheless being interfered with, just as long as the interference in question is constrained, through republican power structures, to track one's interests. Only arbitrary power is inimical to freedom, not power as such. On the other hand, republican freedom is also distinct from positive freedom as expounded and criticized by Berlin. First, republican freedom does not consist in the activity of virtuous political participation; rather, that participation is seen as instrumentally related to freedom as non-domination. Secondly, the republican concept of freedom cannot lead to anything like the oppressive consequences feared by Berlin, because it has a commitment to non-domination and to liberal-democratic institutions already built into it.

It remains to be seen, however, whether the republican concept of freedom is ultimately distinguishable from the negative concept, or whether republican writers on freedom have not simply provided good arguments to the effect that negative freedom is best promoted, on balance and over time, through certain kinds of political institutions rather than others. While there is no necessary connection between negative liberty and democratic government, there may nevertheless be a strong empirical correlation between the two. Ian Carter (1999, 2008), Matthew H. Kramer (2003, 2008), and Robert Goodin and Frank Jackson (2007) have argued, along these lines, that republican policies are best defended empirically on the basis of the standard negative ideal of freedom, rather than on the basis of a conceptual challenge to that ideal. An important premise in such an argument is that the extent of a person's negative freedom is a function not simply of how many single actions are prevented, but of how many different act-combinations are prevented. On this basis, people who can achieve their goals only by bowing and scraping to their masters must be seen as less free than people who can achieve those goals unconditionally. Another important premise is that the extent degree to which a person is negatively free depends, in part, on the probability with which he or she will be constrained from performing future acts or act-combinations. People who are subject to arbitrary power can be seen as less free in the negative sense even if they do not actually suffer interference, because the probability of their suffering constraints is always greater (ceteris paribus, as a matter of empirical fact) than it would be if they were not subject to that arbitrary power. Perhaps this non-trivial probability is sufficient to explain the sense of exposure and precariousness of the ‘dominated’. In reply, Pettit (2008a, 2008b) and Skinner (2008) have insisted that what matters for an agent's freedom is the impossibility of others interfering with impunity, not the improbability of their doing so.

Much of the most recent literature on political and social freedom has concentrated on the above debate over the differences between the republican and liberal (i.e. negative) conceptions of freedom. Critiques of the republican conception that build on, or are otherwise sympathetic to, those of Carter and Kramer, can be found in Bruin (2009), Lang (2012) and Shnayderman (2012). Pettit himself has continued to refine his position, and has further discussed its relation to that of Berlin (Pettit 2011). Berlin's own conception of negative liberty, he argues, occupies an inherently unstable position between the more restrictive Hobbesian view and the more expansive view of freedom as non-domination.

Pettit's analysis of freedom has inspired a number of recent works by political theorists sympathetic to the republican tradition. Frank Lovett has developed an account of domination as a descriptive concept, and of justice as the minimization of domination (Lovett 2010). Several other authors have made use of the concept of domination in addressing more specific problems in normative political theory, such as disability rights, workplace democracy, social equality, and education policy (De Wispelaere and Casassas 2014; Breen and McBride 2015).

4. One Concept of Liberty: Freedom as a Triadic Relation

The two sides identified by Berlin disagree over which of two different concepts best deserves the name of ‘liberty’. Does this fact not denote the presence of some more basic agreement between the two sides? How, after all, could they see their disagreement as one about the definition of liberty if they did not think of themselves as in some sense talking about the same thing? In an influential article, the American legal philosopher Gerald MacCallum (1967) put forward the following answer: there is in fact only one basic concept of freedom, on which both sides in the debate converge. What the so-called negative and positive theorists disagree about is how this single concept of freedom should be interpreted. Indeed, in MacCallum's view, there are a great many different possible interpretations of freedom, and it is only Berlin's artificial dichotomy that has led us to think in terms of there being two.

MacCallum defines the basic concept of freedom — the concept on which everyone agrees — as follows: a subject, or agent, is free from certain constraints, or preventing conditions, to do or become certain things. Freedom is therefore a triadic relation — that is, a relation between three things: an agent, certain preventing conditions, and certain doings or becomings of the agent. Any statement about freedom or unfreedom can be translated into a statement of the above form by specifying what is free or unfree, from what it is free or unfree, and what it is free or unfree to do or become. Any claim about the presence or absence of freedom in a given situation will therefore make certain assumptions about what counts as an agent, what counts as a constraint or limitation on freedom, and what counts as a purpose that the agent can be described as either free or unfree to carry out.

The definition of freedom as a triadic relation was first put forward in the seminal work of Felix Oppenheim in the 1950s and 60s. Oppenheim saw that an important meaning of ‘freedom’ in the context of political and social philosophy was as a relation between two agents and a particular (impeded or unimpeded) action. This interpretation of freedom remained, however, what Berlin would call a negative one. What MacCallum did was to generalize this triadic structure so that it would cover all possible claims about freedom, whether of the negative or the positive variety. In MacCallum's framework, unlike in Oppenheim's, the interpretation of each of the three variables is left open. In other words, MacCallum's position is a meta-theoretical one: his is a theory about the differences between theorists of freedom.

To illustrate MacCallum's point, let us return to the example of the smoker driving to the tobacconists. In describing this person as either free or unfree, we shall be making assumptions about each of MacCallum's three variables. If we say that the driver is free, what we shall probably mean is that an agent, consisting in the driver's empirical self, is free from external (physical or legal) obstacles to do whatever he or she might want to do. If, on the other hand, we say that the driver is unfree, what we shall probably mean is that an agent, consisting in a higher or rational self, is made unfree by internal, psychological constraints to carry out some rational, authentic or virtuous plan. Notice that in both claims there is a negative element and a positive element: each claim about freedom assumes both that freedom is freedom from something (i.e., preventing conditions) and that it is freedom to do or become something. The dichotomy between ‘freedom from’ and ‘freedom to’ is therefore a false one, and it is misleading say that those who see the driver as free employ a negative concept and those who see the driver as unfree employ a positive one. What these two camps differ over is the way in which one should interpret each of the three variables in the triadic freedom-relation. More precisely, we can see that what they differ over is the extension to be assigned to each of the variables.

Thus, those whom Berlin places in the negative camp typically conceive of the agent as having the same extension as that which it is generally given in ordinary discourse: they tend to think of the agent as an individual human being and as including all of the empirical beliefs and desires of that individual. Those in the so-called positive camp, on the other hand, often depart from the ordinary notion, in one sense imagining the agent as more extensive than in the ordinary notion, and in another sense imagining it as less extensive: they think of the agent as having a greater extension than in ordinary discourse in cases where they identify the agent's true desires and aims with those of some collectivity of which she is a member; and they think of the agent as having a lesser extension than in ordinary discourse in cases where they identify the true agent with only a subset of her empirical beliefs and desires — i.e., with those that are rational, authentic or virtuous. Secondly, those in Berlin's positive camp tend to take a wider view of what counts as a constraint on freedom than those in his negative camp: the set of relevant obstacles is more extensive for the former than for the latter, since negative theorists tend to count only external obstacles as constraints on freedom, whereas positive theorists also allow that one may be constrained by internal factors, such as irrational desires, fears or ignorance. And thirdly, those in Berlin's positive camp tend to take a narrower view of what counts as a purpose one can be free to fulfill. The set of relevant purposes is less extensive for them than for the negative theorists, for we have seen that they tend to restrict the relevant set of actions or states to those that are rational, authentic or virtuous, whereas those in the negative camp tend to extend this variable so as to cover any action or state the agent might desire.

On MacCallum's analysis, then, there is no simple dichotomy between positive and negative liberty; rather, we should recognize that there is a whole range of possible interpretations or ‘conceptions’ of the single concept of liberty. Indeed, as MacCallum says and as Berlin seems implicitly to admit, a number of classic authors cannot be placed unequivocally in one or the other of the two camps. Locke, for example, is normally thought of as one of the fathers or classical liberalism and therefore as a staunch defender of the negative concept of freedom. He indeed states explicitly that ‘[to be at] liberty is to be free from restraint and violence from others’. But he also says that liberty is not to be confused with ‘license’, and that “that ill deserves the name of confinement which hedges us in only from bogs and precipices” (Second Treatise, parags. 6 and 57). While Locke gives an account of constraints on freedom that Berlin would call negative, he seems to endorse an account of MacCallum's third freedom-variable that Berlin would call positive, restricting this to actions that are not immoral (liberty is not license) and to those that are in the agent's own interests (I am not unfree if prevented from falling into a bog). A number of contemporary libertarians have provided or assumed definitions of freedom that are similarly morally loaded (e.g. Nozick 1974; Rothbard 1982). This would seem to confirm MacCallum's claim that it is conceptually and historically misleading to divide theorists into two camps — a negative liberal one and a positive non-liberal one.

5. The Analysis of Constraints: Their Types and Their Sources

To illustrate the range of interpretations of the concept of freedom made available by MacCallum's analysis, let us now take a closer look at his second variable — that of constraints on freedom.

Advocates of negative conceptions of freedom typically restrict the range of obstacles that count as constraints on freedom to those that are brought about by other agents. For theorists who conceive of constraints on freedom in this way, I am unfree only to the extent that other people prevent me from doing certain things. If I am incapacitated by natural causes — by a genetic handicap, say, or by a virus or by certain climatic conditions — I may be rendered unable to do certain things, but I am not, for that reason, rendered unfree to do them. Thus, if you lock me in my house, I shall be both unable and unfree to leave. But if I am unable to leave because I suffer from a debilitating illness or because a snow drift has blocked my exit, I am nevertheless not unfree, to leave. The reason such theorists give, for restricting the set of relevant preventing conditions in this way, is that they see unfreedom as a social relation — a relation between persons (see Oppenheim 1961; Miller 1983; Steiner 1983; Kristjánsson 1996; Kramer 2003; Morriss 2012; Shnayderman 2013; Schmidt forthcoming). Unfreedom as mere inability is thought by such authors to be more the concern of engineers and medics than of political and social philosophers. (If I suffer from a natural or self-inflicted inability to do something, should we to say that I remain free to do it, or should we say that the inability removes my freedom to do it while nevertheless not implying that I am unfree to do it? In the latter case, we shall be endorsing a ‘trivalent’ conception, according to which there are some things that a person is neither free nor unfree to do. Kramer 2003 endorses a trivalent conception according to which freedom is identified with ability and unfreedom is the prevention (by others) of outcomes that the agent would otherwise be able to bring about.)

In attempting to distinguish between natural and social obstacles we shall inevitably come across gray areas. An important example is that of obstacles created by impersonal economic forces. Do economic constraints like recession, poverty and unemployment merely incapacitate people, or do they also render them unfree? Libertarians and egalitarians have provided contrasting answers to this question by appealing to different conceptions of constraints. Thus, one way of answering the question is by taking an even more restrictive view of what counts as a constraint on freedom, so that only a subset of the set of obstacles brought about by other persons counts as a restriction of freedom: those brought about intentionally. In this case, impersonal economic forces, being brought about unintentionally, do not restrict people's freedom, even though they undoubtedly make many people unable to do many things. This last view has been taken by a number of market-oriented libertarians, including, most famously, Friedrich von Hayek (1960, 1982), according to whom freedom is the absence of coercion, where to be coerced is to be subject to the arbitrary will of another. (Notice the somewhat surprising similarity between this conception of freedom and the republican conception discussed earlier, in section 3.2) Critics of libertarianism, on the other hand, typically endorse a broader conception of constraints on freedom that includes not only intentionally imposed obstacles but also unintended obstacles for which someone may nevertheless be held responsible (for Miller and Kristjánsson and Shnayderman this means morally responsible; for Oppenheim and Kramer it means causally responsible), or indeed obstacles created in any way whatsoever, so that unfreedom comes to be identical to inability (see Crocker 1980; Cohen 1988; Sen 1992; Van Parijs 1995).

This analysis of constraints helps to explain why socialists and egalitarians have tended to claim that the poor in a capitalist society are as such unfree, or that they are less free than the rich, whereas libertarians have tended to claim that the poor in a capitalist society are no less free than the rich. Egalitarians typically (though not always) assume a broader notion than libertarians of what counts as a constraint on freedom. Although this view does not necessarily imply what Berlin would call a positive notion of freedom, egalitarians often call their own definition a positive one, in order to convey the sense that freedom requires not merely the absence of certain social relations of prevention but the presence of abilities, or what Amartya Sen has influentially called ‘capabilities’ (Sen 1985, 1988, 1992). (Important exceptions to this egalitarian tendency to broaden the relevant set of constraints include Waldron (1993) and Cohen (2011), who demonstrate, for the sake of argument, that relative poverty is in fact empirically inseparable from, and indeed proportional to, the imposition of physical barriers by other agents, and Steiner (1994), who grounds a left-libertarian theory of justice in the idea of an equal distribution of social freedom.)

We have seen that advocates of a negative conception of freedom tend to count only obstacles that are external to the agent. Notice, however, that the term ‘external’ is ambiguous in this context, for it might be taken to refer either to the location of the causal source of an obstacle or to the location of the obstacle itself. Obstacles that count as ‘internal’ in terms of their own location include psychological phenomena such as ignorance, irrational desires, illusions and phobias. Such constraints can be caused in various ways: for example, they might have a genetic origin, or they might be brought about intentionally by others, as in the case of brainwashing or manipulation. In the first case we have an internal constraint brought about by natural causes; in the second, an internal constraint intentionally imposed by another human agent.

More generally, we can now see that there are in fact two different dimensions along which one's notion of a constraint might be broader or narrower. A first dimension is that of the source of a constraint — in other words, what it is that brings about a constraint on freedom. We have seen, for example, that some theorists include as constraints on freedom only obstacles brought about by human action, whereas others also include obstacles with a natural origin. A second dimension is that of the type of constraint involved, where constraint-types include the types of internal constraint just mentioned, but also various types of constraint located outside the agent, such as physical barriers that render an action impossible, obstacles that render the performance of an action more or less difficult, and costs attached to the performance of a (more or less difficult) action. The two dimensions of type and source are logically independent of one another. Given this independence, it is theoretically possible to combine a narrow view of what counts as a source of a constraint with a broad view of what types of obstacle count as unfreedom-generating constraints, or vice versa. As a result, it is not clear that theorists who are normally placed in the ‘negative’ camp need deny the existence of internal constraints on freedom (see Kramer 2003; Garnett 2007).

To illustrate the independence of the two dimensions of type and source, consider the case of the unorthodox libertarian Hillel Steiner (1974–5, 1994). On the one hand, Steiner has a much broader view than Hayek of the possible sources of constraints on freedom: he does not limit the set of such sources to intentional human actions, but extends it to cover all kinds of human cause, whether or not any humans intend such causes and whether or not they can be held morally accountable for them, believing that any restriction of such non-natural sources can only be an arbitrary stipulation, usually arising from some more or less conscious ideological bias. On the other hand, Steiner has an even narrower view than Hayek about what type of obstacle counts as a constraint on freedom: for Steiner, an agent only counts as unfree to do something if it is physically impossible for her to do that thing. Any extension of the constraint variable to include other types of obstacle, such as the costs anticipated in coercive threats, would, in his view, necessarily involve a reference to the agent's desires, and we have seen (in sec. 2) that for those liberals in the negative camp there is no necessary relation between an agent's freedom and her desires. Consider the coercive threat ‘Your money or your life!’. This does not make it impossible for you to refuse to hand over your money, only much less desirable for you to do so. If you decide not to hand over the money, you will suffer the cost of being killed. That will count as a restriction of your freedom, because it will render physically impossible a great number of actions on your part. But it is not the issuing of the threat that creates this unfreedom, and you are not unfree until the sanction (described in the threat) is carried out. For this reason, Steiner excludes threats — and with them all other kinds of imposed costs — from the set of obstacles that count as freedom-restricting. This conception of freedom derives from Hobbes (Leviathan, chs. 14 and 21), and its defenders often call it the ‘pure’ negative conception (M. Taylor 1982; Steiner 1994; Carter and Kramer 2008) to distinguish it from those ‘impure’ negative conceptions that make at least minimal references to the agent's beliefs, desires or values.

Steiner's account of the relation between freedom and coercive threats might be thought to have counterintuitive implications, even from the liberal point of view. Many laws that are normally thought to restrict negative freedom do not physically prevent people from doing what is prohibited, but deter them from doing so by threatening punishment. Are we to say, then, that these laws do not restrict the negative freedom of those who obey them? A solution to this problem may consist in saying that although a law against doing some action, x, does not remove the freedom to do x, it nevertheless renders physically impossible certain combinations of actions that include doing x and doing what would be precluded by the punishment. There is a restriction of the person's overall negative freedom — i.e. a reduction in the overall number of act-combinations available to her — even though she does not lose the freedom to do any specific thing taken in isolation (Carter 1999).

6. The Concept of Overall Freedom

The concept of overall freedom appears to play an important role both in everyday discourse and in contemporary political philosophy. It is only recently, however, that philosophers have stopped concentrating exclusively on the meaning of a particular freedom — the freedom to do or become this or that particular thing — and have started asking whether we can also make sense of descriptive claims to the effect that one person or society is freer than another or of liberal normative claims to the effect that freedom should be maximized or that people should enjoy equal freedom or that they each have a right to a certain minimum level of freedom. The literal meaningfulness of such claims depends on the possibility of gauging degrees of overall freedom, sometimes comparatively, sometimes absolutely.

Theorists disagree, however, about the importance of the notion of overall freedom. For some libertarian and liberal egalitarian theorists, freedom is valuable as such. This suggests that more freedom is better than less (at least ceteris paribus), and that freedom is one of those goods that a liberal society ought to distribute in a certain way among individuals. For other liberal theorists, like Ronald Dworkin (1977, 2011) and the later Rawls (1991), freedom is not valuable as such, and all claims about maximal or equal freedom ought to be interpreted not as literal references to a quantitative good called ‘liberty’ but as elliptical references to the adequacy of lists of certain particular liberties, or types of liberties, selected on the basis of values other than liberty itself. Generally speaking, only the first group of theorists finds the notion of overall freedom interesting.

The theoretical problems involved in measuring overall freedom include that of how an agent's available actions are to be individuated, counted and weighted, and that of comparing and weighting different types (but not necessarily different sources) of constraints on freedom (such as physical prevention, punishability, threats and manipulation). How are we to make sense of the claim that the number of options available to a person has increased? Should all options count for the same in terms of degrees of freedom, or should they be weighted according to their importance in terms of other values? In the latter, does the notion of overall freedom really add anything of substance to the idea that people should be granted those specific freedoms that are valuable? Should the degree of variety among options also count? And how are we to compare the unfreedom created by the physical impossibility of an action with, say, the unfreedom created by the difficulty or costliness or punishability of an action? It is only by comparing these different kinds of actions and constraints that we shall be in a position to compare individuals' overall degrees of freedom. These problems have been addressed, with differing degrees of optimism, not only by political philosophers (Steiner 1983; Carter 1999; Kramer 2003; Garnett 2016) but also, and increasingly, by social choice theorists interested in finding a freedom-based alternative to the standard utilitarian or ‘welfarist’ framework that has tended to dominate their discipline (e.g. Pattanaik and Xu 1991, 1998; Hees 2000; Sen 2002; Sugden 1998, 2003, 2006; Bavetta 2004; Bavetta and Navarra 2012, 2014).

MacCallum's framework is particularly well suited to the clarification of such issues. For this reason, theorists working on the measurement of freedom tend not to refer a great deal to the distinction between positive and negative freedom. This said, most of them are concerned with freedom understood as the availability of options. And the notion of freedom as the availability of options is unequivocally negative in Berlin's sense at least where two conditions are met: first, the source of unfreedom-creating constraints is limited to the actions of other agents, so that natural or self-inflicted obstacles are not seen as decreasing an agent's freedom; second, the actions one is free or unfree to perform are weighted in some value-neutral way, so that one is not seen as freer simply because the options available to one are more valuable or conducive to one's self-realization. Of the above-mentioned authors, only Steiner embraces both conditions explicitly. Sen rejects both of them, despite not endorsing anything like positive freedom in Berlin's sense.

7. Is the Distinction Still Useful?

We began with a simple distinction between two concepts of liberty, and have progressed from this to the recognition that liberty might be defined in any number of ways, depending on how one interprets the three variables of agent, constraints, and purposes. Despite the utility of MacCallum's triadic formula and its strong influence on analytic philosophers, however, Berlin's distinction remains an important point of reference for discussions about the meaning and value of political and social freedom. Are these continued references to positive and negative freedom philosophically well-founded?

It might be claimed that MacCallum's framework is less than wholly inclusive of the various possible conceptions of freedom. In particular, it might be said, the concept of self-mastery or self-direction implies a presence of control that is not captured by MacCallum's explication of freedom as a triadic relation. MacCallum's triadic relation indicates mere possibilities. If one thinks of freedom as involving self-direction, on the other hand, one has in mind an exercise-concept of freedom as opposed to an opportunity-concept (this distinction comes from C. Taylor 1979). If interpreted as an exercise concept, freedom consists not merely in the possibility of doing certain things (i.e. in the lack of constraints on doing them), but in actually doing certain things in certain ways — for example, in realizing one's true self or in acting on the basis of rational and well-informed decisions. The idea of freedom as the absence of constraints on the realization of given ends might be criticised as failing to capture this exercise concept of freedom, for the latter concept makes no reference to the absence of constraints.

However, this defence of the positive-negative distinction as coinciding with the distinction between exercise- and opportunity-concepts of freedom has been challenged by Eric Nelson (2005). As Nelson points out, most of the theorists that are traditionally located in the positive camp, such as Green or Bosanquet, do not distinguish between freedom as the absence of constraints and freedom as the doing or becoming of certain things. For these theorists, freedom is the absence of any kind of constraint whatsoever on the realization of one's true self (they adopt a maximally extensive conception of constraints on freedom), and the absence of all factors that could prevent the action x is, quite simply, equivalent to the realization of x. In other words, if there really is nothing stopping me from doing x — if I possess all the means to do x, and I have a desire to do x, and no desire, irrational or otherwise, not to do x — then I do x. An equivalent way to characterize the difference between such positive theorists and the so-called negative theorists of freedom lies in the degree of specificity with which they describe x. For those who adopt a narrow conception of constraints, x is described with a low degree of specificity (x could be exemplified by the realization of any of a large array of options); for those who adopt a broad conception of constraints, x is described with a high degree of specificity (x can only be exemplified by the realization of a specific option, or of one of a small group of options).

What perhaps remains of the distinction is a rough categorization of the various interpretations of freedom that serves to indicate their degree of fit with the classical liberal tradition. There is indeed a certain family resemblance between the conceptions that are normally seen as falling on one or the other side of Berlin's divide, and one of the decisive factors in determining this family resemblance is the theorist's degree of concern with the notion of the self. Those on the ‘positive’ side see questions about the nature and sources of a person's beliefs, desires and values as relevant in determining that person's freedom, whereas those on the ‘negative’ side, being more faithful to the classical liberal tradition, tend to consider the raising of such questions as in some way indicating a propensity to violate the agent's dignity or integrity (Carter 2011a). One side takes a positive interest in the agent's beliefs, desires and values, while the other recommends that we avoid doing so.


Introductory works

  • Feinberg, J., 1973, Social Philosophy, New Jersey: Prentice-Hall, ch. 1 [article-length general introduction].
  • Flickschuh, K., 2007, Freedom. Contemporary Liberal Perspectives, Cambridge: Polity [introduction to Berlin and MacCallum together with analysis of the conceptions of freedom of Nozick, Steiner, Dworkin and Raz].
  • Carter, I., Kramer, M. H. and Steiner, H. (eds.), 2007, Freedom: A Philosophical Anthology, Oxford: Blackwell [large number of excerpts from all the major contemporary contributions to the interpretation of freedom, with editorial introductions. The first of its nine sections is specifically on positive vs negative liberty].
  • Gray, T., 1991, Freedom, London: Macmillan [comprehensive book-length introduction].
  • Kukathas, C., 1993, Liberty, in R. Goodin and P. Pettit (eds.), A Companion to Contemporary Political Philosophy, Oxford: Blackwell [article-length general introduction].
  • Pelczynski, Z. and Gray, J. (eds.), 1984, Conceptions of Liberty in Political Philosophy, London: Athlone Press [collection of essays on single authors, mostly historical].
  • Miller, D. (ed.), 1991, Liberty, Oxford: Oxford University Press. 2nd ed., The Liberty Reader, Boulder, CO: Paradigm Publishers, 2006 [representative collection of contemporary essays, including Berlin and his critics, with editorial introduction and a guide to further reading].
  • Plant, R., 1991, Modern Political Thought, Oxford: Blackwell, ch 1 [article-length general introduction].
  • Schmidtz, D. (ed.), 2017, The Oxford Handbook of Freedom, New York: Oxford University Press [collection of up-to-date essays by major contemporary authors].

Other works

  • Arneson, R. J., 1985, ‘Feedom and Desire’, Canadian Journal of Philosophy, 3: 425–48.
  • Baum, B. and Nichols, R. (eds.), 2013, Isaiah Berlin and the Politics of Freedom. “Two Concepts of Liberty” 50 Years Later, London: Routledge.
  • Bavetta, S., 2004, ‘Measuring Freedom of Choice: An Alternative View of a Recent Literature’, Social Choice and Welfare, 22: 29–48.
  • Bavetta, S. and Navarra, P., 2012, The Economics of Freedom. Theory, Measurement, and Policy Implications, New York: Cambridge University Press.
  • Bavetta, S., Navarra, P. and Maimone, D., 2014, Freedom and the Pursuit of Happiness. An Economic and Political Perspective, New York: Cambridge University Press.
  • Berlin, I., 1969, ‘Two Concepts of Liberty’, in I. Berlin, Four Essays on Liberty, London: Oxford University Press. New ed. in Berlin 2002.
  • –––, 1978, ‘From Hope and Fear Set Free’, in I. Berlin, Concepts and Categories. Philosophical Essays, ed. H. Hardy, London: Hogarth Press; Oxford: Oxford University Press, 1980. Reprinted in Berlin 2002.
  • –––, 2002, Liberty, ed. H. Hardy, Oxford: Oxford University Press, 2002.
  • Bobbio, N., 1955, ‘La libertà dei moderni comparata a quella dei posteri’, in N. Bobbio, Politica e cultura, Turin: Einaudi.
  • Bosanquet, B., 1899, The Philosophical Theory of the State, London: Macmillan.
  • Breen, K. and McBride, C. (eds.), 2015, ‘Freedom and Domination: Exploring Republican Freedom’, Special Issue of Critical Review of International Social and Political Philosophy, 18: 349–485.
  • Bruin, B. de, 2009, ‘Liberal and Republican Freedom’, Journal of Political Philosophy, 17: 418–39.
  • Carter, I., 1999, A Measure of Freedom, Oxford: Oxford University Press.
  • –––, 2004, ‘Choice, Freedom and Freedom of Choice’, Social Choice and Welfare, 22: 61–81.
  • –––, 2008, ‘How are Power and Unfreedom Related?’, in Laborde and Maynor 2008.
  • –––, 2011a, ‘Respect and the Basis of Equality’, Ethics, 121: 538–71.
  • –––, 2011b, ‘The Myth of "Merely Formal Freedom"’, Journal of Political Philosophy, 19: 486–95, Reprinted in S. Cahn and R. B. Talisse (eds.), Political Philosophy in the Twenty-first Century. Essential Essays, Boulder CO.: Westview Press, 2013.
  • –––, 2015, ‘Value-freeness and Value-neutrality in the Analysis of Political Concepts’, in D. Sobel, P. Vallentyne and S. Wall (eds.), Oxford Studies in Political Philosophy (Volume 1), 279–305.
  • Carter, I. and Kramer, M. H., 2008, ‘How Changes in One's Preferences Can Affect One's Freedom (and How They Cannot): A Reply to Dowding and van Hees’, Economics and Philosophy, 2008.
  • Christman, J., 1991, ‘Liberalism and Individual Positive Freedom’, Ethics, 101: 343–59.
  • –––, 2005, ‘Saving Positive Freedom’, Political Theory, 33: 79–88.
  • –––, 2009, The Politics of Persons. Individual Autonomy and Socio-historical Selves, Cambridge: Cambridge University Press.
  • Christman, J. (ed.), 1989, The Inner Citadel: Essays on Individual Autonomy, Oxford: Oxford University Press.
  • Cohen, G. A., 1988, History, Labour and Freedom: Themes from Marx, Oxford: Clarendon Press.
  • –––, 1991, Capitalism, Freedom and the Proletariat, in Miller 1991.
  • –––, 1995, Self-Ownership, Freedom and Equality, Cambridge: Cambridge University Press.
  • –––, 2011, ‘Freedom and Money’, in G. A. Cohen, On the Currency of Egalitarian Justice and Other Essays in Political Philosophy, ed. M. Otsuka, Princeton NJ: Princeton University Press.
  • Cohen, M., 1960, ‘Berlin and the Liberal Tradition’, Philosophical Quarterly, 10, pp. 216–27.
  • Crocker, L., 1980, Positive Liberty, London: Nijhoff.
  • Day, J. P., 1970, ‘On Liberty and the Real Will’, Philosophy, 45: 177–92, reprinted in Day 1987.
  • –––, 1987, Liberty and Justice, London: Croom Helm.
  • De Wispelaere, J. and Casassas, D., 2014, ‘A Life of One's Own: Republican Freedom and Disability’, Disability and Society, 29: 402–16.
  • Dimova-Cookson, M., 2003, ‘A New Scheme of Positive and Negative Freedom: Reconstructing T. H. Green on Freedom’, Political Theory, 31: 508–32.
  • Dowding, K. and van Hees, M., 2007, ‘Counterfactual Success and Negative Freedom’, Economics and Philosophy, 23: 141–162.
  • Dworkin, G., 1988, The Theory and Practice of Autonomy, Cambridge: Cambridge University Press.
  • Dworkin, R., 1977, Taking Rights Seriously, London: Duckworth.
  • –––, 2011, Justice for Hedgehogs, Cambridge, Mass.: Harvard University Press.
  • Flathman, R., 1987, The Philosophy and Politics of Freedom, Chicago: Chicago University Press.
  • Garnett, M., 2007, ‘Ignorance, Incompetence and the Concept of Liberty’, Journal of Political Philosophy, 15: 428–46.
  • –––, 2016, ‘Value-neutrality and the Ranking of Opportunity Sets’, Economics and Philosophy, 32: 99–119.
  • Gibbs, B., 1976, Freedom and Liberation, London: Chatto and Windus.
  • Goodin, R. E. and Jackson, F., 2007, ‘Freedom from Fear’, Philosophy and Public Affairs, 35: 249–65
  • Gorr, M., 1989, Coercion, Freedom and Exploitation, New York: Peter Lang.
  • Grant, C., 2013, ‘Freedom and Oppression’, Politics, Philosophy and Economics, 12: 413–25.
  • Gray, J., 1980, ‘On Negative and Positive Liberty’, Political Studies, 28: 507–26.
  • –––, 1995, Isaiah Berlin, London: HarperCollins.
  • Green, T. H., 1895, Lectures on the Principles of Political Obligation, London: Longmans, Green.
  • Hayek, F. A. von, 1960, The Constitution of Liberty, London: Routledge and Kegan Paul.
  • –––, 1982, Law, Legislation and Liberty, London: Routledge.
  • Hees, M. van, 2000, Legal Reductionism and Freedom, Dordrecht: Kluwer.
  • Hindricks, F., 2008, ‘The Freedom of Collective Agents’, Journal of Political Philosophy, 16: 165–83.
  • Hirschmann, N. J., 2003, The Subject of Liberty: Toward a Feminist Theory of Freedom, Princeton NJ: Princeton University Press.
  • Kramer, M. H., 2003, The Quality of Freedom, Oxford: Oxford University Press.
  • –––, 2008, ‘Liberty and Domination’, in Laborde and Maynor 2008.
  • Kristjánsson, K., 1996, Social Freedom: The Responsibility View, Cambridge: Cambridge University Press.
  • Laborde, C. and Maynor, J. (eds.), 2008, Republicanism and Political Theory, Oxford: Blackwell.
  • Lang, G., 2012, ‘Invigilating Republican Liberty’, Philosophical Quarterly, 62: 273–93.
  • Larmore, C., 2004, ‘Liberal and Republican Conceptions of Freedom’, in Weinstock and Nadeau 2004.
  • Lovett, F., 2010, A General Theory of Domination and Justivce, Oxford: Oxford University Press.
  • List, C. and Valentini, L., 2016, ‘Freedom as Independence’, Ethics, 126: 1043–74.
  • MacCallum, G. C. Jr., 1967, ‘Negative and Positive Freedom’, Philosophical Review, 76: 312–34, reprinted in Miller 1991.
  • Macpherson, C. B., 1973, Berlin's Division of Liberty, in C. B. Macpherson, Democratic Theory: Essays in Retrieval, Oxford: Clarendon Press.
  • Miller, D., 1983, ‘Constraints on Freedom’, Ethics, 94: 66–86. Partial reprint in 2nd ed. of Miller 1991.
  • Milne, A. J. M., 1968, Freedom and Rights, London: George Allen and Unwin.
  • Morriss, P., 2012, ‘What is Freedom if it is Not Power?’, Theoria, 59: 1–25.
  • Nelson, E., 2005, ‘Liberty: One Concept Too Many?’, Political Theory, 33: 58–78.
  • Nozick, R., 1974, Anarchy, State and Utopia, New York: Basic Books.
  • Oppenheim, F. E., 1961, Dimensions of Freedom: An Analysis, New York: St. Martin's Press.
  • –––, 1981, Political Conceps: A Reconstruction, Oxford, Blackwell.
  • Pansardi, P., 2012, ‘Power and Freedom: Opposite or Equivalent Concepts?’, Theoria, 59: 26–44.
  • Pattanaik, P. and Xu, Y., 1990, ‘On Ranking Opportunity Sets in Terms of Freedom of Choice’, Recherches Economiques de Louvain, 56: 383–90.
  • –––, 1998, ‘On Preference and Freedom’, Theory and Decision, 44: 173–98.
  • Pettit, P., 1997, Republicanism: A Theory of Freedom and Government, Oxford: Oxford University Press.
  • –––, 2001, A Theory of Freedom, Cambridge: Polity Press.
  • –––, 2008a, ‘Freedom and Probability. A Comment on Goodin and Jackson’, Philosophy and Public Affairs, 36: 206–20.
  • –––, 2008b, ‘Republican Freedom: Three Axioms, Four Theorems’, in Laborde and Maynor 2008.
  • –––, 2011, ‘The Instability of Freedom as Non-Interference. The Case of Isaiah Berlin’, Ethics, 121: 693–716.
  • –––, 2012, On the People's Terms. A Republican Theory and Model of Democracy, Cambridge: Cambridge University Press.
  • –––, 2014, Just Freedom. A Moral Compass for a Complex World, New York: Norton.
  • Pitkin, H., 1988, ‘Are Freedom and Liberty Twins?’, Political Theory, 16: 523–52.
  • Plamenatz, J., 1938, Consent, Freedom and Political Obligation, London: Oxford University Press.
  • Rawls, J., 1971, A Theory of Justice, Cambridge Mass.: Harvard University Press.
  • –––, 1991, Political Liberalism, New York: Columbia University Press.
  • Ricciardi, M., 2007, ‘Berlin on Liberty’, in G. Crowder and H. Hardy (eds.), The One and the Many. Reading Isaiah Berlin, Amherst NY: Prometheus Books.
  • Rothbard, M. N., 1982, The Ethics of Liberty, Atlantic Highlands: Humanities Press.
  • Ruggiero, G. de, 1925, Storia del liberalismo europeo, Bari: Laterza, English R. G. Collingwood, The History of European Liberalism, London: Oxford University Press 1927.
  • Sen, A., 1985, ‘Well-being, Agency and Freedom’, Journal of Philosophy, 82: 169–221.
  • –––, 1988, ‘Freedom of Choice: Concept and Content’, European Economic Review, 32: 269–94.
  • –––, 1992, Inequality Reexamined, Oxford: Oxford University Press.
  • –––, 2002, Rationality and Freedom, Cambridge, Mass.: Harvard University Press.
  • Schmidt, A., 2015, ‘Why Animals have an Interest in Freedom’, Historical Social Research, 40: 92–109.
  • Sharon, A., 2016, ‘Domination and the Rule of Law’, in D. Sobel, P. Vallentyne and S. Wall (eds.), Oxford Studies in Political Philosophy (Volume 2), New York: Oxford University Press.
  • –––, forthcoming, ‘Abilities and the Sources of Unfreedom’, Ethics, 126.
  • Shnayderman, R., 2012, ‘Liberal vs. Republican Notions of Freedom’, Political Studies, 60: 44–58.
  • –––, 2013, ‘Social Freedom, Moral Responsibility, Actions and Omissions’, Philosophical Quarterly, 63: 716–39.
  • –––, 2016, ‘Ian Carter's Non-evaluative Theory of Freedom and Diversity. A Critique’, Social Choice and Welfare, 46: 39–55.
  • Skinner, Q., 1998, Liberty before Liberalism, Cambridge: Cambridge University Press.
  • –––, 2002, ‘A Third Concept of Liberty’, Proceedings of the British Academy, 117(237): 237–68.
  • –––, 2008, ‘Freedom as the Absence of Arbitrary Power’, in Laborde and Maynor 2008.
  • Steiner, H., 1974–5, ‘Individual Liberty’, Proceedings of the Aristotelian Society, 75: 33–50, reprinted in Miller 1991.
  • –––, 1983, ‘How Free: Computing Personal Liberty’, in A. Phillips Griffiths, Of Liberty, Cambridge: Cambridge University Press.
  • –––, 1994, An Essay on Rights, Oxford: Blackwell.
  • –––, 2001, ‘Freedom and Bivalence’, in Carter and Ricciardi 2001.
  • Sugden, R., 1998, ‘The Metric of Opportunity’, Economics and Philosophy, 14: 307–337.
  • –––, 2003, ‘Opportunity as a Space for Individuality: its Value, and the Impossibility of Measuring it’, Ethics, 113(4): 783–809.
  • –––, 2006, ‘What We Desire, What We Have Reason to Desire, Whatever We Might Desire: Mill and Sen on the Value of Opportunity’, Utilitas, 18: 33–51.
  • Taylor, C., 1979, ‘What's Wrong with Negative Liberty’, in A. Ryan (ed.), The Idea of Freedom, Oxford: Oxford University Press, reprinted in Miller 1991.
  • Taylor, M., 1982, Community, Anarchy and Liberty, Cambridge: Cambridge University Press.
  • Van Parijs, P., 1995, Real Freedom for All, Oxford: Oxford University Press.
  • Waldron, J., 1993, ‘Homelessness and the Issue of Freedom’, in J. Waldron, Liberal Rights. Collected Papers 1981–1991, Cambridge: Cambridge University Press.
  • Weinstock, D. and Nadeau, C. (eds.), 2004, Republicanism: History, Theory and Practice, London: Frank Cass.
  • Wendt, F., 2011, ‘Slaves, Prisoners, and Republican Freedom’, Res Publica, 17: 175–92.
  • Williams, B., 2001, ‘From Freedom to Liberty: The Construction of a Political Value’, Philosophy and Public Affairs, 30: 3–26.
  • Young, R., 1986, Autonomy. Beyond Negative and Positive Liberty, New York: St. Martin's Press.
  • Zimmerman, D., 2002, ‘Taking Liberties: the Perils of "Moralizing" Freedom and Coercion in Social Theory and Practice’, Social Theory and Practice, 28: 577–609.
Charles Sanders Peirce


B:x: Brent, Joseph (1998), Charles Sanders Peirce: A Life, 2nd edition, page x[1]

CDPT: Commens Dictionary of Peirce's Terms
CP x.y: Collected Papers, volume x, paragraph y
EP x:y: The Essential Peirce, volume x, page y

W x:yWritings of Charles S. Peirce, volume x, page y

Charles Sanders Peirce began writing on semiotics, which he also called semeiotics, meaning the philosophical study of signs, in the 1860s, around the time that he devised his system of three categories. During the 20th century, the term "semiotics" was adopted to cover all tendencies of sign researches, including Ferdinand de Saussure's semiology, which began in linguistics as a completely separate tradition.

Peirce adopted the term semiosis (or semeiosis) and defined it to mean an "action, or influence, which is, or involves, a cooperation of three subjects, such as a sign, its object, and its interpretant, this trirelative influence not being in any way resolvable into actions between pairs".[2] This specific type of triadic relation is fundamental to Peirce's understanding of "logic as formal semiotic". By "logic" he meant philosophical logic. He eventually divided (philosophical) logic, or formal semiotics, into (1) speculative grammar, or stechiology on the elements of semiosis (sign, object, interpretant), how signs can signify and, in relation to that, what kinds of signs, objects, and interpretants there are, how signs combine, and how some signs embody or incorporate others; (2) logical critic, or logic proper, on the modes of inference; and (3) speculative rhetoric, or methodeutic, the philosophical theory of inquiry, including his form of pragmatism. His speculative grammar, or stechiology, is this article's subject.

Peirce conceives of and discusses things like representations, interpretations, and assertions broadly and in terms of philosophical logic, rather than in terms of psychology, linguistics, or social studies. He places philosophy at a level of generality between mathematics and the special sciences of nature and mind, such that it draws principles from mathematics and supplies principles to special sciences.[3] On the one hand, his semiotic theory does not resort to special experiences or special experiments in order to settle its questions. On the other hand, he draws continually on examples from common experience, and his semiotics is not contained in a mathematical or deductive system and does not proceed chiefly by drawing necessary conclusions about purely hypothetical objects or cases. As philosophical logic, it is about the drawing of conclusions deductive, inductive, or hypothetically explanatory. Peirce's semiotics, in its classifications, its critical analysis of kinds of inference, and its theory of inquiry, is philosophical logic studied in terms of signs and their triadic relations as positive phenomena in general.

Semiotic elements[edit]

Here is Peirce's definition of the triadic sign relation that formed the core of his definition of logic.

Namely, a sign is something, A, which brings something, B, its interpretant sign determined or created by it, into the same sort of correspondence with something, C, its object, as that in which itself stands to C. (Peirce 1902, NEM 4, 20–21).

This definition, together with Peirce's definitions of correspondence and determination, is sufficient to derive all of the statements that are necessarily true for all sign relations. Yet, there is much more to the theory of signs than simply proving universal theorems about generic sign relations. There is also the task of classifying the various species and subspecies of sign relations. As a practical matter, of course, familiarity with the full range of concrete examples is indispensable to theory and application both.

In Peirce's theory of signs, a sign is something that stands in a well-defined kind of relation to two other things, its object and its interpretant sign. Although Peirce's definition of a sign is independent of psychological subject matter and his theory of signs covers more ground than linguistics alone, it is nevertheless the case that many of the more familiar examples and illustrations of sign relations will naturally be drawn from linguistics and psychology, along with our ordinary experience of their subject matters.

For example, one way to approach the concept of an interpretant is to think of a psycholinguistic process. In this context, an interpretant can be understood as a sign's effect on the mind, or on anything that acts like a mind, what Peirce calls a quasi-mind. An interpretant is what results from a process of interpretation, one of the types of activity that falls under the heading of semiosis. One usually says that a sign stands for an object to an agent, an interpreter. In the upshot, however, it is the sign's effect on the agent that is paramount. This effect is what Peirce called the interpretant sign, or the interpretant for short. An interpretant in its barest form is a sign's meaning, implication, or ramification, and especial interest attaches to the types of semiosis that proceed from obscure signs to relatively clear interpretants. In logic and mathematics the most clarified and most succinct signs for an object are called canonical forms or normal forms.

Peirce argued that logic is the formal study of signs in the broadest sense, not only signs that are artificial, linguistic, or symbolic, but also signs that are semblances or are indexical such as reactions. Peirce held that "all this universe is perfused with signs, if it is not composed exclusively of signs",[4] along with their representational and inferential relations. He argued that, since all thought takes time, all thought is in signs:

To say, therefore, that thought cannot happen in an instant, but requires a time, is but another way of saying that every thought must be interpreted in another, or that all thought is in signs. (Peirce, 1868[5])

Thought is not necessarily connected with a brain. It appears in the work of bees, of crystals, and throughout the purely physical world; and one can no more deny that it is really there, than that the colors, the shapes, etc., of objects are really there. Consistently adhere to that unwarrantable denial, and you will be driven to some form of idealistic nominalism akin to Fichte's. Not only is thought in the organic world, but it develops there. But as there cannot be a General without Instances embodying it, so there cannot be thought without Signs. We must here give "Sign" a very wide sense, no doubt, but not too wide a sense to come within our definition. Admitting that connected Signs must have a Quasi-mind, it may further be declared that there can be no isolated sign. Moreover, signs require at least two Quasi-minds; a Quasi-utterer and a Quasi-interpreter; and although these two are at one (i.e., are one mind) in the sign itself, they must nevertheless be distinct. In the Sign they are, so to say, welded. Accordingly, it is not merely a fact of human Psychology, but a necessity of Logic, that every logical evolution of thought should be dialogic. (Peirce, 1906[6] )

Sign relation[edit]

Signhood is a way of being in relation, not a way of being in itself. Anything is a sign — not as itself, but in some relation to another. The role of sign is constituted as one role among three: object, sign, and interpretant sign. It is an irreducible triadic relation; the roles are distinct even when the things that fill them are not. The roles are but three: a sign of an object leads to interpretants, which, as signs, lead to further interpretants. In various relations, the same thing may be sign or semiotic object. The question of what a sign is depends on the concept of a sign relation, which depends on the concept of a triadic relation. This, in turn, depends on the concept of a relation itself. Peirce depended on mathematical ideas about the reducibility of relations — dyadic, triadic, tetradic, and so forth. According to Peirce's Reduction Thesis,[7] (a) triads are necessary because genuinely triadic relations cannot be completely analyzed in terms of monadic and dyadic predicates, and (b) triads are sufficient because there are no genuinely tetradic or larger polyadic relations—all higher-arityn-adic relations can be analyzed in terms of triadic and lower-arity relations and are reducible to them. Peirce and others, notably Robert Burch (1991) and Joachim Hereth Correia and Reinhard Pöschel (2006), have offered proofs of the Reduction Thesis.[8] According to Peirce, a genuinely monadic predicate characteristically expresses quality. A genuinely dyadic predicate — reaction or resistance. A genuinely triadic predicate — representation or mediation. Thus Peirce's theory of relations underpins his philosophical theory of three basic categories (see below).

Extension × intension = information.[citation needed] Two traditional approaches to sign relation, necessary though insufficient, are the way of extension (a sign's objects, also called breadth, denotation, or application) and the way of intension (the objects' characteristics, qualities, attributes referenced by the sign, also called depth, comprehension, significance, or connotation). Peirce adds a third, the way of information, including change of information, in order to integrate the other two approaches into a unified whole.[9] For example, because of the equation above, if a term's total amount of information stays the same, then the more that the term 'intends' or signifies about objects, the fewer are the objects to which the term 'extends' or applies. A proposition's comprehension consists in its implications.[10]

Determination. A sign depends on its object in such a way as to represent its object — the object enables and, in a sense, determines the sign. A physically causal sense of this stands out especially when a sign consists in an indicative reaction. The interpretant depends likewise on both the sign and the object — the object determines the sign to determine the interpretant. But this determination is not a succession of dyadic events, like a row of toppling dominoes; sign determination is triadic. For example, an interpretant does not merely represent something which represented an object; instead an interpretant represents something as a sign representing an object. It is an informational kind of determination, a rendering of something more determinately representative.[11] Peirce used the word "determine" not in strictly deterministic sense, but in a sense of "specializes", bestimmt,[11] involving variation in measure, like an influence. Peirce came to define sign, object, and interpretant by their (triadic) mode of determination, not by the idea of representation, since that is part of what is being defined.[12] The object determines the sign to determine another sign — the interpretant — to be related to the object as the sign is related to the object, hence the interpretant, fulfilling its function as sign of the object, determines a further interpretant sign. The process is logically structured to perpetuate itself, and is definitive of sign, object, and interpretant in general.[13] In semiosis, every sign is an interpretant in a chain stretching both fore and aft. The relation of informational or logical determination which constrains object, sign, and interpretant is more general than the special cases of causal or physical determination. In general terms, any information about one of the items in the sign relation tells you something about the others, although the actual amount of this information may be nil in some species of sign relations.

Sign, object, interpretant[edit]

Peirce held that there are exactly three basic semiotic elements, the sign, object, and interpretant, as outlined above and fleshed out here in a bit more detail:

  • A sign (or representamen) represents, in the broadest possible sense of "represents". It is something interpretable as saying something about something. It is not necessarily symbolic, linguistic, or artificial.
  • An object (or semiotic object) is a subject matter of a sign and an interpretant. It can be anything discussable or thinkable, a thing, event, relationship, quality, law, argument, etc., and can even be fictional, for instance Hamlet.[14] All of those are special or partial objects. The object most accurately is the universe of discourse to which the partial or special object belongs.[15] For instance, a perturbation of Pluto's orbit is a sign about Pluto but ultimately not only about Pluto.
  • An interpretant (or interpretant sign) is the sign's more or less clarified meaning or ramification, a kind of form or idea of the difference which the sign's being true or undeceptive would make. (Peirce's sign theory concerns meaning in the broadest sense, including logical implication, not just the meanings of words as properly clarified by a dictionary.) The interpretant is a sign (a) of the object and (b) of the interpretant's "predecessor" (the interpreted sign) as being a sign of the same object. The interpretant is an interpretation in the sense of a product of an interpretive process or a content in which an interpretive relation culminates, though this product or content may itself be an act, a state of agitation, a conduct, etc. Such is what is summed up in saying that the sign stands for the object to the interpretant.

Some of the understanding needed by the mind depends on familiarity with the object. In order to know what a given sign denotes, the mind needs some experience of that sign's object collaterally to that sign or sign system, and in this context Peirce speaks of collateral experience, collateral observation, collateral acquaintance, all in much the same terms.[16]

"Representamen" (properly with the "a" long and stressed: ) was adopted (not coined) by Peirce as his blanket technical term for any and every sign or sign-like thing covered by his theory. It is a question of whether the theoretically defined "representamen" covers only the cases covered by the popular word "sign." The word "representamen" is there in case a divergence should emerge. Peirce's example was this: Sign action always involves a mind. If a sunflower, by doing nothing more than turning toward the sun, were thereby to become fully able to reproduce a sunflower turning in just the same way toward the sun, then the first sunflower's turning would be a representamen of the sun yet not a sign of the sun.[17] Peirce eventually stopped using the word "representamen."[18]

Peirce made various classifications of his semiotic elements, especially of the sign and the interpretant. Of particular concern in understanding the sign-object-interpretant triad is this: In relation to a sign, its object and its interpretant are either immediate (present in the sign) or mediate.

  1. Sign, always immediate to itself — that is, in a tautologous sense, present in or at itself, even if it is not immediate to a mind or immediately accomplished without processing or is a general apprehended only in its instances.
  2. Object
    1. Immediate object, the object as represented in the sign.
    2. Dynamic object, the object as it really is, on which the idea which is the immediate object is "founded, as on bedrock"[19] Also called the dynamoid object, the dynamical object.
  3. Interpretant
    1. Immediate interpretant, the quality of the impression which a sign is fit to produce, not any actual reaction, and which the sign carries with it even before there is an interpreter or quasi-interpreter. It is what is ordinarily called the sign's meaning.
    2. Dynamic interpretant, the actual effect (apart from the feeling) of the sign on a mind or quasi-mind, for instance the agitation of the feeling.
    3. Final interpretant, the effect which the sign would have on the conduct of any mind or quasi-mind if circumstances allowed that effect to be fully achieved. It is the sign's end or purpose. The final interpretant of one's inquiry about the weather is the inquiry's purpose, the effect which the response would have on the plans for the day of anybody in one's shoes. The final interpretant of a line of investigation as such is the truth as the ideal final opinion and would be reached sooner or later but still inevitably by investigation adequately prolonged, though the truth remains independent of that which you or I or any finite community of investigators believe.

The immediate object is, from the viewpoint of a theorist, really a kind of sign of the dynamic object; but phenomenologically it is the object until there is reason to go beyond it, and somebody analyzing (critically but not theoretically) a given semiosis will consider the immediate object to be the object until there is reason to do otherwise.[20]

Peirce preferred phrases like dynamic object over real object since the object might be fictive — Hamlet, for instance, to whom one grants a fictive reality, a reality within the universe of discourse of the play Hamlet.[14]

It is initially tempting to regard immediate, dynamic, and final interpretants as forming a temporal succession in an actual process of semiosis, especially since their conceptions refer to beginning, midstages, and end of a semiotic process. But instead their distinctions from each other are modal or categorial. The immediate interpretant is a quality of impression which a sign is fitted to produce, a special potentiality. The dynamic interpretant is an actuality. The final interpretant is a kind of norm or necessity unaffected by actual trends of opinion or interpretation. One does not actually obtain a final interpretant per se; instead one may successfully coincide with it.[21] Peirce, a fallibilist, holds that one has no guarantees that one has done so, but only compelling reasons, sometimes very compelling, to think so and, in practical matters, must sometimes act with complete confidence of having done so. (Peirce said that it is often better in practical matters to rely on instinct, sentiment, and tradition, than on theoretical inquiry.[22]) In any case, insofar as truth is the final interpretant of a pursuit of truth, one believes, in effect, that one coincides with a final interpretant of some question about what is true, whenever and to whatever extent that one believes that one reaches a truth.

The significant example of the Semiotic theory of Charles Sanders Peirce appears in the event of storm of 1859 that determined the meaning of Peirce's understanding of "logic as formal semiotic". By "logic" called philosophical logic .He eventually divided (philosophical) logic, or formal semiotics, into speculative grammar, or stechiology on the elements of semiosis (sign, object, interpretant) : Sign : From our Lord Creator of the universe object : storm event of 1859 interpretant : create a technological platform necessary for human scientific development related to: Depending on what happened to our planet in 1859, which is the technological platform for what will be achieved by humans in various life sciences where it was Helped

  • The proper functioning of artificial satellites
  • Good communication between continents
  • The stabilization of the distribution of the electrical energy High tensionête_solaire_de_1859

Classes of signs[edit]

Peirce proposes several typologies and definitions of the signs. More than 76 definitions of what a sign is have been collected throughout Peirce's work.[23] Some canonical typologies can nonetheless be observed, one crucial one being the distinction between "icons", "indices" and "symbols" (CP 2.228, CP 2.229 and CP 5.473). The icon-index-symbol typology is chronologically the first but structurally the second of three that fit together as a trio of three-valued parameters in regular scheme of nine kinds of sign. (The three "parameters" (not Peirce's term) are not independent of one another, and the result is a system of ten classes of sign, which are shown further down in this article.)

Peirce's three basic phenomenological categories come into central play in these classifications. The 1-2-3 numerations used further below in the exposition of sign classes represents Peirce's associations of sign classes with the categories. The categories are as follows:

Name:Typical characterizaton:As universe of experience:As quantity:Technical definition:Valence, "adicity":
Firstness.[25]Quality of feeling.Ideas, chance, possibility.Vagueness, "some".Reference to a ground (a ground is a pure abstraction of a quality).[26]Essentially monadic (the quale, in the sense of the such,[27] which has the quality).
Secondness.[28]Reaction, resistance, (dyadic) relation.Brute facts, actuality.Singularity, discreteness, “this”.Reference to a correlate (by its relate).Essentially dyadic (the relate and the correlate).
Thirdness.[29]Representation, mediation.Habits, laws, necessity.Generality, continuity, "all".Reference to an interpretant*.Essentially triadic (sign, object, interpretant*).

 *Note: An interpretant is an interpretation (human or otherwise) in the sense of the product of an interpretive process.

The three sign typologies depend respectively on (I) the sign itself, (II) how the sign stands for its denoted object, and (III) how the signs stands for its object to its interpretant. Each of the three typologies is a three-way division, a trichotomy, via Peirce's three phenomenological categories.

  1. Qualisigns, sinsigns, and legisigns . Every sign is either (qualisign) a quality or possibility, or (sinsign) an actual individual thing, fact, event, state, etc., or (legisign) a norm, habit, rule, law. (Also called tones, tokens, and types, also potisigns, actisigns, and famisigns.)
  2. Icons, indices, and symbols. Every sign refers either (icon) through similarity to its object, or (index) through factual connection to its object, or (symbol) through interpretive habit or norm of reference to its object.
  3. Rhemes, dicisigns, and arguments . Every sign is interpreted either as (rheme) term-like, standing for its object in respect of quality, or as (dicisign) proposition-like, standing for its object in respect of fact, or as (argument) argumentative, standing for its object in respect of habit or law. This is the trichotomy of all signs as building blocks of inference. (Also called sumisigns, dicent signs, and suadisigns, also semes, phemes, and delomes.)

Every sign falls under one class or another within (I) and within (II) and' within (III). Thus each of the three typologies is a three-valued parameter for every sign. The three parameters are not independent of each other; many co-classifications aren't found.[30] The result is not 27 but instead ten classes of signs fully specified at this level of analysis.

In later years, Peirce attempted a finer level of analysis, defining sign classes in terms of relations not just to sign, object, and interpretant, but to sign, immediate object, dynamic object, immediate interpretant, dynamic interpretant, and final or normal interpretant. He aimed at 10 trichotomies of signs, with the above three trichotomies interspersed among them, and issuing in 66 classes of signs. He did not bring that system into a finished form. In any case, in that system, icon, index, and symbol were classed by category of how they stood for the dynamic object, while rheme, dicisign, and argument were classed by the category of how they stood to the final or normal interpretant.[31]

These conceptions are specific to Peirce's theory of signs and are not exactly equivalent to general uses of the notions of "icon", "index", "symbol", "tone", "token", "type", "term" (or "rheme"), "proposition" (or "dicisign), "argument".

I. Qualisign, sinsign, legisign[edit]

Also called tone, token, type; and also called potisign, actisign, famisign.

This is the typology of the sign as distinguished by sign's own phenomenological category (set forth in 1903, 1904, etc.).

  1. A qualisign (also called tone, potisign, and mark) is a sign which consists in a quality of feeling, a possibility, a "First."
  2. A sinsign (also called token and actisign) is a sign which consists in a reaction/resistance, an actual singular thing, an actual occurrence or fact, a "Second."
  3. A legisign (also called type and famisign) is a sign which consists in a (general) idea, a norm or law or habit, a representational relation, a "Third."

A replica (also called instance) of a legisign is a sign, often an actual individual one (a sinsign), which embodies that legisign. A replica is a sign for the associated legisign, and therefore is also a sign for the legisign's object. All legisigns need sinsigns as replicas, for expression. Some but not all legisigns are symbols. All symbols are legisigns. Different words with the same meaning are symbols which are replicas of that symbol which consists in their meaning but doesn't prescribe qualities of its replicas.[32]

II. Icon, index, symbol[edit]

This is the typology of the sign as distinguished by phenomenological category of its way of denoting the object (set forth in 1867 and many times in later years). This typology emphasizes the different ways in which the sign refers to its object — the icon by a quality of its own, the index by real connection to its object, and the symbol by a habit or rule for its interpretant. The modes may be compounded, for instance, in a sign that displays a forking line iconically for a fork in the road and stands indicatively near a fork in the road.

  1. An icon (also called likeness and semblance) is a sign that denotes its object by virtue of a quality which is shared by them but which the icon has irrespectively of the object. The icon (for instance, a portrait or a diagram) resembles or imitates its object. The icon has, of itself, a certain character or aspect, one which the object also has (or is supposed to have) and which lets the icon be interpreted as a sign even if the object does not exist. The icon signifies essentially on the basis of its "ground." (Peirce defined the ground as the pure abstraction of a quality, and the sign's ground as the pure abstraction of the quality in respect of which the sign refers to its object, whether by resemblance or, as a symbol, by imputing the quality to the object.[33]). Peirce called an icon apart from a label, legend, or other index attached to it, a "hypoicon", and divided the hypoicon into three classes: (a) the image, which depends on a simple quality; (b) the diagram, whose internal relations, mainly dyadic or so taken, represent by analogy the relations in something; and (c) the metaphor, which represents the representative character of a sign by representing a parallelism in something else.[34] A diagram can be geometric, or can consist in an array of algebraic expressions, or even in the common form "All __ is ___" which is subjectable, like any diagram, to logical or mathematical transformations. Peirce held that mathematics is done by diagrammatic thinking — observation of, and experimentation on, diagrams.
  2. An index* is a sign that denotes its object by virtue of an actual connection involving them, one that he also calls a real relation in virtue of its being irrespective of interpretation. It is in any case a relation which is in fact, in contrast to the icon, which has only a ground for denotation of its object, and in contrast to the symbol, which denotes by an interpretive habit or law. An index which compels attention without conveying any information about its object is a pure index, though that may be an ideal limit never actually reached. If an indexical relation is a resistance or reaction physically or causally connecting an index to its object, then the index is a reagent (for example smoke coming from a building is a reagent index of fire). Such an index is really affected or modified by the object, and is the only kind of index which can be used in order to ascertain facts about its object. Peirce also usually held that an index does not have to be an actual individual fact or thing, but can be a general; a disease symptom is general, its occurrence singular; and he usually considered a designation to be an index, e.g., a pronoun, a proper name, a label on a diagram, etc. (In 1903 Peirce said that only an individual is an index,[35] gave "seme" as an alternate expression for "index", and called designations "subindices or hyposemes,[36] which were a kind of symbol; he allowed of a "degenerate index" indicating a non-individual object, as exemplified by an individual thing indicating its own characteristics. But by 1904 he allowed indices to be generals and returned to classing designations as indices. In 1906 he changed the meaning of "seme" to that of the earlier "sumisign" and "rheme".)
  3. A symbol* is a sign that denotes its object solely by virtue of the fact that it will be interpreted to do so. The symbol consists in a natural or conventional or logical rule, norm, or habit, a habit that lacks (or has shed) dependence on the symbolic sign's having a resemblance or real connection to the denoted object. Thus, a symbol denotes by virtue of its interpretant. Its sign-action (semeiosis) is ruled by a habit, a more or less systematic set of associations that ensures its interpretation. For Peirce, every symbol is a general, and that which we call an actual individual symbol (e.g., on the page) is called by Peirce a replica or instance of the symbol. Symbols, like all other legisigns (also called "types"), need actual, individual replicas for expression. The proposition is an example of a symbol which is irrespective of language and of any form of expression and does not prescribe qualities of its replicas.[37] A word that is symbolic (rather than indexical like "this" or iconic like "whoosh!") is an example of a symbol that prescribes qualities (especially looks or sound) of its replicas.[38] Not every replica is actual and individual. Two word-symbols with the same meaning (such as English "horse" and Spanish caballo) are symbols which are replicas of that symbol which consists in their shared meaning.[32] A book, a theory, a person, each is a complex symbol.

 * Note: in "On a New List of Categories" (1867) Peirce gave the unqualified term "sign" as an alternate expression for "index", and gave "general sign" as an alternate expression for "symbol". "Representamen" was his blanket technical term for any and every sign or signlike thing covered by his theory.[39] Peirce soon reserved "sign" to its broadest sense, for index, icon, and symbol alike. He also eventually decided that the symbol is not the only sign which can be called a "general sign" in some sense, and that indices and icons can be generals, generalities, too. The general sign, as such, the generality as a sign, he eventually called, at various times, the "legisign" (1903, 1904), the "type" (1906, 1908), and the "famisign" (1908).

III. Rheme, dicisign, argument[edit]

This is the typology of the sign as distinguished by the phenomenological category which the sign's interpretant attributes to the sign's way of denoting the object (set forth in 1902, 1903, etc.):

  1. A rheme (also called sumisign and seme*) is a sign that represents its object in respect of quality and so, in its signified interpretant, is represented as a character or mark,[40] though it actually may be icon, index, or symbol. The rheme* (seme) stands as its object for some purpose.[41] A proposition with the subject places left blank is a rheme; but subject terms by themselves are also rhemes. A proposition, said Peirce, can be considered a zero-place rheme, a zero-place predicate.
  2. A dicisign (also called dicent sign and pheme ) is a sign that represents its object in respect of actual existence and so, in its signified interpretant, is represented as indexical,[42] though it actually may be either index or symbol. The dicisign separately indicates its object (as subject of the predicate).[43] The dicisign "is intended to have some compulsive effect on the interpreter of it".[41] Peirce had generalized the idea of proposition to where a weathercock, photograph, etc., could be considered propositions (or "dicisigns", as he came to call them). A proposition in the conventional sense is a dicent symbol (also called symbolic dicisign). Assertions are also dicisigns.
  3. An argument (also called suadisign and delome) is a sign that represents its object in respect of law or habit and so, in its signified interpretant, is represented as symbolic (and was indeed a symbol in the first place).[44] The argument separately "monstrates" its signified interpretant (the argument's conclusion); an argument stripped of all signs of such monstrative relationship is, or becomes, a dicisign.[43] It represents "a process of change in thoughts or signs, as if to induce this change in the Interpreter" through the interpreter's own self-control.[41] A novel, a work of art, the universe, can be a delome in Peirce's terms.

 *Note: In his "Prolegomena To an Apology For Pragmaticism" (The Monist, v. XVI, no. 4, Oct. 1906), Peirce uses the words "seme", "pheme", and "delome" (pp. 506, 507, etc.) for the rheme-dicisign-argument typology, but retains the word "rheme" for the predicate (p. 530) in his system of Existential Graphs. Also note that Peirce once offered "seme" as an alternate expression for "index" in 1903.[35]

The three sign typologies together: ten classes of sign[edit]

The three typologies, labeled "I.", "II.", and "III.", are shown together in the table below. As parameters, they are not independent of one another. As previously said, many co-classifications aren't found.[30] The slanting and vertical lines show the options for co-classification of a given sign (and appear in MS 339, August 7, 1904, viewable here at the Lyris peirce-l archive[45]). The result is ten classes of sign.

Words in parentheses in the table are alternate names for the same kinds of signs.

Phenomenological category:
Sign is distinguished by phenomenological
category of...
1. Quality
of feeling.
Reference to
a ground.

2. Reaction,
Brute fact.
Reference to
a correlate.

3. Representation,
Habit, law.
Reference to
an interpretant.
(Tone, Potisign)
(Token, Actisign)
(Type, Famisign)
II. ...the sign's way of denoting its OBJECT:ICON
(Likeness, etc.)
(General sign*)
III. ...the sign's way —
as represented in the INTERPRETANT —
of denoting the sign's object:
(Sumisign, Seme;
e.g., a term)
(Dicent sign, Pheme;
e.g., a proposition)

 *Note: As noted above, in "On a New List of Categories" (1867) Peirce gave the unqualified word "sign" as an alternate expression for "index", and gave "general sign" as an alternate expression for "symbol." Peirce soon reserved "sign" to its broadest sense, for index, icon, and symbol alike, and eventually decided that symbols are not the only signs which can be called "general signs" in some sense. See note at end of section "II. Icon, index, symbol" for details.
Note that a term (in the conventional sense) is not just any rheme; it is a kind of rhematic symbol. Likewise a proposition (in the conventional sense) is not just any dicisign, it is a kind of dicent symbol.

Sign classed
by own
Specificational redundancies
in parentheses
Some examples
(I)QualisignIconRheme(Rhematic Iconic) QualisignA feeling of "red"
(II)SinsignIconRheme(Rhematic) Iconic SinsignAn individual diagram
(III)IndexRhemeRhematic Indexical SinsignA spontaneous cry
(IV)DicisignDicent (Indexical) SinsignA weathercock or photograph
(V)LegisignIconRheme(Rhematic) Iconic LegisignA diagram, apart from its factual individuality
(VI)IndexRhemeRhematic Indexical LegisignA demonstrative pronoun
(VII)DicisignDicent Indexical LegisignA street cry (identifying the individual by tone, theme)
(VIII)SymbolRhemeRhematic Symbol (–ic Legisign)A common noun
(IX)DicisignDicent Symbol (–ic Legisign)A proposition (in the conventional sense)
(X)ArgumentArgument (–ative Symbolic Legisign)A syllogism














  1. ^Brent, Joseph (1998), Charles Sanders Peirce: A Life, 2nd edition, Bloomington and Indianapolis: Indiana University Press (catalog page); also NetLibrary.
  2. ^1906, EP 2:411 and CP 5.484. Peirce went on to say: "Σημείωσις [Sêmeíôsis] in Greek of the Roman period, as early as Cicero's time, if I remember rightly, meant the action of almost any kind of sign; and my definition confers on anything that so acts the title of a 'sign.'" See Σημείωσις in the Liddell & Scott Ancient Greek lexicon at the Perseus Digital Library.
  3. ^For Peirce's definitions of philosophy, see for instance "A Syllabus of Certain Topics of Logic", CP 1.183-186, 1903 and "Minute Logic", CP 1.239-241, 1902. See Peirce's definitions of philosophy at CDPT under "Cenoscopy" and "Philosophy".
  4. ^Peirce, C.S., CP 5.448 footnote, from "The Basis of Pragmaticism" in 1906.
  5. ^"Questions Concerning Certain Faculties Claimed for Man" (ArisbeEprint), Journal of Speculative Philosophy vol. 2 (1868), pp. 103-114. Reprinted (CP 5.213-263, the quote is from para. 253).
  6. ^"Prolegomena To an Apology For Pragmaticism", pp. 492–546, The Monist, vol. XVI, no. 4 (mislabeled "VI"), Oct. 1906, see p. 523. Reprinted CP 4.530–572; see para. 551 Eprint.
  7. ^See "The Logic of Relatives", The Monist, Vol. 7, 1897, pp. 161-217, see p. 183 (via Google Books with registration apparently not required). Reprinted in the Collected Papers, vol. 3, paragraphs 456-552, see paragraph 483.
  8. ^* Burch, Robert (1991), A Peircean Reduction Thesis: The Foundations of Topological Logic, Texas Tech University Press, Lubbock, Texas
    • Anellis, Irving (1993) "Review of A Peircean Reduction Thesis: The Foundations of Topological Logic by Robert Burch" in Modern Logic v. 3, n. 4, 401-406, Project Euclid Open Access PDF 697 KB. Criticism and some suggestions for improvements.
    • Anellis, Irving (1997), "Tarski's Development of Peirce's Logic of Relations" (Google Book SearchEprint) in Houser, Nathan, Roberts, Don D., and Van Evra, James (eds., 1997), Studies in the Logic of Charles Sanders Peirce. Anellis gives an account of a Reduction Thesis proof discussed and presented by Peirce in his letter to William James of August 1905 (L224, 40-76, printed in Peirce, C. S. and Eisele, Carolyn, ed. (1976), The New Elements of Mathematics by Charles S. Peirce, v. 3, 809-835).
    • Hereth Correia, Joachim and Pöschel, Reinhard (2006), "The Teridentity and Peircean Algebraic Logic" in Conceptual Structures: Inspiration and Application (ICCS 2006): 229-246, Springer. Frithjof Dau called it "the strong version" of proof of Peirce's Reduction Thesis. John F. Sowa in the same discussion claimed that an explanation in terms of conceptual graphs is sufficiently convincing about the Reduction Thesis for those without the time to understand what Peirce was saying.
    • In 1954 W. V. O. Quine claimed to prove the reducibility of larger predicates to dyadic predicates, in Quine, W.V.O., "Reduction to a dyadic predicate", Selected Logic Papers.
  9. ^Peirce, C. S. (1867), "Upon Logical Comprehension and Extension" (CP 2.391-426), (W 2:70-86, PEP Eprint).
  10. ^Peirce, C.S and Ladd-Franklin, Christine, "Signification (and Application, in logic)", Dictionary of Philosophy and Psychology v. 2, p. 528. Reprinted CP 2.431-4.
  11. ^ abPeirce, letter to William James, dated 1909, see EP 2:492.
  12. ^Peirce, C.S., "A Letter to Lady Welby" (1908), Semiotic and Significs, pp. 80-81:

    I define a Sign as anything which is so determined by something else, called its Object, and so determines an effect upon a person, which effect I call its Interpretant, that the latter is thereby mediately determined by the former. My insertion of "upon a person" is a sop to Cerberus, because I despair of making my own broader conception understood.

  13. ^See "76 definitions of the sign by C.S.Peirce", collected by Professor Robert Marty (University of Perpignan, France).
  14. ^ abA Letter to William James, EP 2:498, 1909, viewable at CDPT under Dynamical Object
  15. ^A Letter to William James, EP 2:492, 1909, viewable at CDPT under "Object".
  16. ^See pp. 404-409 in "Pragmatism", EP 2. Ten quotes on collateral observation from Peirce provided by Joseph Ransdell can be viewed here. Note: Ransdell's quotes from CP 8.178-179, are also in EP 2:493-4, which gives their date as 1909; and his quote from CP 8.183, is also in EP 2:495-6, which gives its date as 1909.
  17. ^"A Syllabus of Certain Topics of Logic", EP 2:272-3, 1903
  18. ^A Draft of a Letter to Lady Welby, Semiotic and Significs, p. 193, 1905
  19. ^In EP 2:407, viewable at CDPT under "Real Object"
  20. ^See Ransdell, Joseph, "On the Use and Abuse of the Immediate/Dynamical Object Distinction" draft 2007, ArisbeEprint
  21. ^See Peirce's 1909 letter (or letters) to William James, CP 8.314 and 8.315, and Essential Peirce v. 2, pp. 496-7, and a 1909 letter to Lady Welby, Semiotic and Significs pp. 110-1, all under "Final Interpretant" at CDPT. Also see 1873, MS 218 (Robin 379) in Writings of Charles S. Peirce v. 3, p. 79, on the final opinion, and CP 8.184, on final opinion as final interpretant, in a review of a book by Lady Welby.
  22. ^"Philosophy and the Conduct of Life", 1898, Lecture 1 of the Cambridge (MA) Conferences Lectures, published CP 1.616-48 in part and in Reasoning and the Logic of Things, Ketner (ed., intro.) and Putnam (intro., comm.), pp. 105-22, reprinted in Essential Peirce v. 2, pp. 27-41.
  23. ^See "76 Definitions of The Sign by C. S. Peirce" collected and analyzed by Robert Marty, Department of Mathematics, University of Perpignan, Perpignan, France, With an Appendix of 12 Further Definitions or Equivalents proposed by Alfred Lang, Dept of Psychology, University of Bern, Bern, Switzerland, ArisbeEprint.
  24. ^"Minute Logic", CP 2.87, c.1902 and A Letter to Lady Welby, CP 8.329, 1904. See relevant quotes under "Categories, Cenopythagorean Categories" in Commens Dictionary of Peirce's Terms (CDPT), Bergman & Paalova, eds., U. of Helsinki.
  25. ^See quotes under "Firstness, First [as a category]" in CDPT.
  26. ^The ground blackness is the pure abstraction of the quality black. Something black is something embodying blackness, pointing us back to the abstraction. The quality black amounts to reference to its own pure abstraction, the ground blackness. The question is not merely of noun (the ground) versus adjective (the quality), but rather of whether we are considering the black(ness) as abstracted away from application to an object, or instead as so applied (for instance to a stove). Yet note that Peirce's distinction here is not that between a property-general and a property-individual (a trope). See "On a New List of Categories" (1867), in the section appearing in CP 1.551. Regarding the ground, cf. the Scholastic conception of a relation's foundation, Google limited preview Deely 1982, p. 61
  27. ^A quale in this sense is a such, just as a quality is a suchness. Cf. under "Use of Letters" in §3 of Peirce's "Description of a Notation for the Logic of Relatives", Memoirs of the American Academy, v. 9, pp. 317–78 (1870), separately reprinted (1870), from which see p. 6 via Google books, also reprinted in CP 3.63:

    Now logical terms are of three grand classes. The first embraces those whose logical form involves only the conception of quality, and which therefore represent a thing simply as “a —.” These discriminate objects in the most rudimentary way, which does not involve any consciousness of discrimination. They regard an object as it is in itself as such (quale); for example, as horse, tree, or man. These are absolute terms. (Peirce, 1870. But also see "Quale-Consciousness", 1898, in CP 6.222–37.)

  28. ^See quotes under "Secondness, Second [as a category]" in CDPT.
  29. ^See quotes under "Thirdness, Third [as a category]" in CDPT.
  30. ^ abFor the reasons why, see CP 2.254-263, reprinted in the Philosophical Writings of Peircepp. 115-118, and in EP 2:294-296.
  31. ^See CP 8.343-75, from a 1908 partial draft of a letter to Lady Welby.
  32. ^ ab"New Elements (Kaina Stoicheia") MS 517 (1904); EP 2:300-324, ArisbeEprint, scroll down to /317/, then first new paragraph
  33. ^Cf. the Scholastic conception of a relation's foundation, Deely 1982, p. 61 (Google Books)
  34. ^On image, diagram, and metaphor, see "Hypoicon" in the Commens Dictionary of Peirce's Terms.
  35. ^ abIn 'A Syllabus of Certain Topics of Logic', EP 2:274, 1903, and viewable under "Index" at CDPT.
  36. ^In "A Syllabus of Certain Topics of Logic", EP 2:274, 1903, and viewable under "Subindex, Hyposeme" at the CDPT.
  37. ^MS599 c.1902 "Reason's Rules", relevant quote viewable under "MS 599" in "Role of Icons In Predication", Joseph Ransdell, ed. ArisbeEprint.
  38. ^"A Syllabus of Certain Topics of Logic", EP 2:274, 1903, and "Logical Tracts, No. 2", CP 4.447, c. 1903. Relevant quotes viewable at the CDPT, under "Symbol".
  39. ^"A Syllabus of Certain Topics of Logic", EP 2:272-3. Relevant quote viewable at CDPT, under "Representamen"
  40. ^A Letter to Lady Welby, Semiotic and Significs pp. 33-34, 1904, viewable at CDPT under "Rhema, Rheme".
  41. ^ abcPeirce, 1906, "Prolegomena To an Apology For Pragmaticism", pp. 506-507 in 492-546, The Monist, v. XVI, n. 4 (mislabeled "VI"), Oct. 1906, reprinted in CP 4.538
  42. ^A Letter to Lady Welby, Semiotic and Significs, pp. 33-34, 1904; also "A Syllabus of Certain Topics of Logic', EP 2:275-276 and 292, 1903; all three quotes viewable at CDPT under "Dicent, Dicent Sign, Dicisign".
  43. ^ ab"New Elements (Kaina Stoicheia)", Manuscript 517 (1904), and EP 2:300-324, see 308, viewable in ArisbeEprint, scroll down to /308/
  44. ^"A Syllabus of Certain Topics of Logic", EP 2:296, 1903, quote viewable at CDPT under "Argument".
  45. ^the image was provided by Bernard Morand of the Institut Universitaire de Technologie (France), Département Informatique.
  46. ^See peirce-l post by Anderson Vinicius Romanini "Re: representing the ten classes of signs (corrected)" 2006-06-16 Eprint and peirce-l post by Joseph Ransdell "Re: 1st image of triangle of boxes (MS799.2)" 2006-06-18 Eprint. The manuscript can be viewed (and magnified by clicking on image) here at the Lyris peirce-l archive. The image was provided by Joseph Ransdell, Professor Emeritus, Philosophy, Texas Tech University.

References and further reading[edit]

For abbreviations of his works see Abbreviations

Pieces by Peirce on semiotic
  • Peirce, C.S. (1867), "On a New List of Categories", Proceedings of the American Academy of Arts and Sciences 7 (1868), 287–298. Presented, 14 May 1867. Reprinted (Collected Papers (CP), v. 1, paragraphs 545–559), (Writings of Charles S. Peirce: A Chronological Edition, v. 2, pp. 49–59), (The Essential Peirce (EP) v. 1, 1–10). ArisbeEprint.
  • Peirce, C.S. (1867), "Upon Logical Comprehension and Extension", Proceedings of the American Academy of Arts and Sciences, pp. 416-432. Presented 13 November 1867. Reprinted CP 2.391-426, Writings v. 2, pp. 70–86. Eprint.
  • Peirce, C.S. (c.1894 MS), "What Is a Sign?". Published in part in CP 2.281, 285, and 297-302, and in full in EP 2:4-10. Peirce Edition Project Eprint.
  • Peirce, C.S. (1895 MS), "Of Reasoning in General". Published in part in CP 2.282, 286-91, 295-96, 435-44, and 7.555-8, and in full in EP 2:11-26.
  • Peirce, C.S. (1896), "The Regenerated Logic", The Monist, v. VII, n. 1, pp. 19-40, The Open Court Publishing Co., Chicago, Illinois, 1896, for the Hegeler Institute. Reprinted (CP 3.425-455). Internet Archive, The Monist 7, p. 19.
  • Peirce, C.S. (1897), "The Logic of Relatives", The Monist, v. VII, pp. 161-217. Reprinted in CP 3.456-552.
  • Peirce, C.S. (c.1902 MSS), "Minute Logic", CP 2.1-118.
  • Peirce, C.S. (c.1902 MS), "Reason's Rules" Eprint
  • Peirce, C.S. "A Syllabus of Certain Topics of Logic", EP 2:
    • Peirce, C.S. (1903) "Sundry Logical Conceptions", EP 2:267-88.
    • Peirce, C.S. (1903) "Nomenclature and Divisions of Triadic Relations, as Far as They Are Determined", EP 2:289-99
    • Peirce, C.S. (1904 MS) "New Elements (Kaina Stoicheia)", pp. 235–63 in Carolyn Eisele, ed., The New Elements of Mathematics by Charles S. Peirce, Volume 4, Mathematical Philosophy. Reprinted (EP 2:300-24). Eprint.
  • Peirce, C.S. (c.1903 MS), "Logical Tracts, No. 2", CP 4.418–509.
  • Peirce, C.S. (1904 Oct 12), A Letter to Lady Welby, CP 8.327–41.
  • Peirce, C.S. (1905), A Draft of a Letter to Lady Welby, Semiotic and Significs p. 193
  • Peirce, C.S. (1906), "Prolegomena To an Apology For Pragmaticism", pp. 492-546, The Monist, vol. XVI, no. 4
Categories: 1

0 Replies to “Triadic Relationship Definition Essay”

Leave a comment

L'indirizzo email non verrà pubblicato. I campi obbligatori sono contrassegnati *