Monthly Archives: September 2026

power: The ability to make people (or things) do what they would not otherwise have done. The purpose of the modern concept of power was recognized as early as 1748, with the publication of Hume’s essay, “Of the Original Contract.” Almost all of the governments, which exist at present,’ says Hume, ‘…have been founded originally either on usurpation of conquest, or both, without any pretense of a fair consent, or voluntary subjection of the people.’ Describing the process of political change—migration, colonization, and military victory—Hume demands, rhetorically, ‘Is there anything discoverable in all these events but force and violence?’

Hume’s comments offer one of the first clear versions of the assumptions of a ‘modern’ age, which seeks to study politics positively, eschewing theological justifications and moral evaluations in favor of a causal assessment of how the political world works in reality. Politics is seen to be about neither left nor right; indeed, in Hume, as in much social science, might is seen as creating right de facto because the seizure of power leads to the establishment of authority and the successful inculcation of belief. Power is the appropriate central concept for this world-view because, in its modern form, it is concerned with which groups or persons dominate, get their own way or are best able to pursue their own interests in societies. James March, in his 1966 essay ‘The Power of Power,’ stressed that the concept conveyed simultaneous overtones of the cynicism of Realpolitik, the glories of classical mechanics and the comfort of anthropocentric theology. In other words the ‘power’ world-view offers the would-be social scientist an immunity from moral evaluation and theoretical speculation, and the possibility of emulating the explanatory achievements of the physicist.

Bertrand Russell defined power as ‘the production of intended effects,’ but this serves better as an indication of what we want to mean when we talk about power than as a working definition. A large number of other writers have offered more complex definitions of power or paradigms of power relationships. The core of these, in respect of the expression ‘A has power over B’ are:

1. A has effects on B’s choices and actions.
2. A has the capacity to move B’s choices and actions in ways that A intends.
3. A has the capacity to override opposition from B.
4. The relationship between A and B described by propositions 1, 2, and 3 is part of a social structure (not necessarily the social structure) and has a tendency to persist.

Problems with any definition include:

1. If we do not include a condition of intentionality, then we are left with a paradoxical and useless concept of power. For example, the Victim has power over the Bully because his or her weakness vulnerability is provocative to the Bully’s action. On the other hand, there are intuitively satisfactory examples of power without intention: Subserviens may regard Superior as a power person and, therefore, try to please him, but he may respond in ways which are not according to Superior’s intentions or even contrary to them. If Subserviens is so in awe of Superior that the only reaction of which he is capable is to throw his arms around Superior’s ankles and kiss his feet, a practice which Superior detests, then Superior cannot be said to have power over Subserviens since he lacks the capacity to control him. We want the concept of power, ultimately, to tell us about who can get their own way, to distinguish between the Barrack Room Lawyer who appears to obey orders but is generally capable of manipulating structures and relationships, and the Formal Authority who appears to be obeyed to the letter, but ho has not close control over his relationships.

The solution to this paradox is to acknowledge that the possession of power can have unintended consequences, but that the test of whether a person has power or not must be conducted in terms of control, of the capacity to conceive intentions. If a person has power, the consequences of that power must be attributable to that person, who is responsible for those consequences. Without intentionality and attributability the concept of power becomes vague to the point of meaninglessness, not like the concept of energy in physics (which Russell wanted it to be), but more like the concept of the ether, the presence of which could not be distinguished from its absence.

2. Comparability and Quantifiability. If the concept of power is to be the central concept for understanding certain kinds of politics, its use must go beyond isolated remarks of the form ‘A has power over B’ and at least extend to comparative analyses: ‘A has more power than C in context x’ and “A has more power than anybody in context x.’ This raises issues of great complexity because the range of variables which might be used to compare the power of two people is considerable. Different writers have given different names to these variables, but an account of power must consider both the geographic and demographic range over which the power extends and the scope of issues affected. There is the question of the objective weighting of A’s power in comparison to C’s, that is, the extent to which the individuals affected care about the effect which A can control. There is also the question of subjective weighting, the extent to which A is able to control what he or she really cares about. This is complicated by the phenomena of anticipated reactions: many shrewd political actors modify their aims to the political environment. For them, the possession of power may be an end in itself and the question of comparing the extent to which they can modify events according to their own will become obscure and irresponsible.

Thus it is very difficult to compare the power of two groups, or institutions. Often the difficulty is as logically simple as comparing a person who has apples to a person who has oranges and coming to a conclusion about which possessed more ‘fruit.’ It would be intuitively obvious that fifty apples was more fruit than five oranges, but a closer comparison might evoke alternative standards of market of subjective value, weight, volume, nutrient capacity, and so on. Thus ordinal comparisons of power are impossible in many circumstances, and dealing with the kind of cardinal numbers we need to make power ‘like energy in physics’ is usually out of the question.

3. Time and Causation. If A has the power to achieve x at time t and he or she wants x, does that mean x will necessarily occur? The answer must surely be no, because it must be possible for A to possess the power, but to fail to use it. This raises a profound doubt about the nature of power. How would we know at any one time what power a person had? Most exercises of power affect the possession of that power. The use of power may be self-diminishing, particularly where it ‘spends’ the resource (such as money or credibility) on which it is based. Equally, it may be self-increasing, as when actors ranging from teachers to the leaders of military coups establish control over their domain. In many cases there are contingent increases and decreases of power: for instance, macroeconomic conditions are bound to affect what can be achieved by entrepreneurs or trade union leaders.

Thus the instances of A exercising power at t₁, t_2, etc., are of little value in estimating A’s power at t. A series of exercises of power may be catastrophically self-diminishing: precisely because a person has succeeded n times, it may be that moves to thwart them are afoot on the (n 1)th occasion. It is no help to say, as Dahl does, that power is best explained in terms of probability: statistical probability is no use, because it only works for a series of identical instances like the spins of a fair roulette wheel and inductive probability is merely an estimate of the odds which takes everything into account (including, presumably, the ‘power’ of the actors). The fundamental problem is that the concept of power seeks to make static statements about a dynamic reality and the consequent doubt must be as to whether the concept ever really helps us understand or predict real events such as the fall of Margaret Thatcher in 1990.

Power is often classified into five principal forms: force, persuasion, authority, coercion, and manipulation. However, only coercion and manipulation are uncontroversially forms of power.

1. Force in its narrow sense implies a control of the body rather than the person. We may kill, bind, or render comatose without being able to get a person’s actions to conform to our will. Only when they comply because of the threat of force can the relationship be called power and this becomes, strictly, coercion.

2. Persuasion, by which the slave may persuade the emperor or the professor the Prime Minister. In other words, the powerless may persuade the powerful: the offering of ideas is not control until it creates a dependency and, therefore, the capacity to manipulate.

3. Authority is sometimes defined as ‘legitimate power.’ But it can also be understood as the existence (in various senses) of rights to command and corresponding duties to obey. Authority is therefore separate from power, though it constitutes a resource for power in the same way as does money and a capacity for rational persuasion. It can exist in a pure form, without power, as, for instance, the authority of a priest over his flock in a secular society.

4. Coercion is perhaps the paradigm form of power and is said to consist of controlling people through threats, whether overt or tacit. It is, though, extremely difficult to distinguish a threat from other forms of relationship. Is it a threat if we say we are going to make a person worse off than they expected to be? Or worse off than if we were not to act? Most modern relationships, whether children’s pocket money or promotions at work, seem to exist in a middle territory between threats and offers for which we have no established word in English (though Hillel Steiner recommends ‘throffers’).

5. Manipulation involves control exercised without threats, typically using resources of information and ideas. Usually people do not realize they are being manipulated or the process would not work. Arguably, it is a more durable form of power: Subserviens obedience to Superior is more securely founded on the belief that God wants him to obey than on the fear of being whipped. But arguments about manipulation can easily slide into unfalsifiable arguments about ‘false consciousness.’ Increasingly, as power has failed as a concept for the positive investigation of political systems, it has been taken over by writers like Foucault who see power as permeating all social relationships. This tradition of thought does not, generally, seek to measure or attribute power or to distinguish its forms, but is content to emphasize its transcendence and the effect of power in distorting social relations.

 In summary, the concept of power has not filled the central role in the study of politics which many pioneers hoped it would. It has proved much easier to believe generally that ‘Politics is about power,’ or, particularly, that individual P or group E possesses power, than it has been to clarify what such beliefs mean or what would constitute proof or disproof of them.”

Excerpted from: McLean, Iain, and Alistair McMillan, editors. Oxford Concise Dictionary of Politics. New York: Oxford University Press, 2003.