Ancient History & Civilisation


IV i

At the beginning of Book IV Aristotle’s discussion takes a more practical turn. The political theorist, he maintains, must investigate four kinds of constitutions:

1. The ideal or absolutely best conceivable, without regard to actual circumstances.

2. The best attainable in the circumstances that prevail.

3. Constitutions inferior to 1 and 2, which evidently may be worth putting into practice and preserving.

4. An all-purpose constitution which would suit all states. He must also appreciate that there are several different kinds of each of the main constitutions (that there are, e.g., several different varieties of democracy), and that each variety requires different laws. This programme of study is embarked upon in chapter iii.

To attain (1) and (2) would be the concern of a ‘utopian’ political theorist such as Plato: the Republic presumably describes, in his view, the first kind of constitution, the Laws the second (but note the wider horizons of Laws 739ab ff.). Aristotle, however, evinces a certain impatience with utopianism: he is prepared to offer help to constitutions far removed from the ideal – and, interestingly enough, not by sudden or radical constitutional change, nor by demanding a fresh start, but by ‘Fabian’ gradualism and piecemeal ‘social engineering’, and with the consent of the inhabitants of the state concerned. The whole chapter is a salutary reminder that in spite of all Aristotle’s theoretical discussion, and his keen interest in the utopias he describes in Book II, his Politics has an essentially practical purpose.

1288b10 In all the skills and sciences which do not operate piecemeal but give complete coverage to some class of objects, it is the task of a single skill or science to investigate what suits each individual class. For instance, what kind of training is advantageous for what kind of body, and what training is best? For the best must necessarily suit the body which is best endowed and best equipped. It is a further task of gymnastic to investigate what single form of training for everyone will serve the greatest number. And even if a man has no ambition to acquire the condition or the knowledge1 appropriate for athletic contests, it is none the less the business of the teacher or gymnastic trainer to impart that2 degree of ability too. We see the same principle at work in medicine, shipbuilding, clothing and every other skill.

1288b21 So it is clearly true also that it is the task of one and the same science to consider the best constitution, what it is and what it would be like if it were constructed exactly as one would wish, without any hindrance from outside. Another of its tasks is to consider what constitution is suited to what persons, because for many the best is perhaps impossible to attain; so the good lawgiver and the genuine statesman will have to bear in mind both the ‘absolutely best’ constitution and the ‘best in the circumstances’. There is also a third, which starts from an assumption – I mean he must be able to consider also a constitution which is given, both how it could come into being, and how once in being it may last longest. I am speaking particularly of a state which happens to operate neither the best constitution (being without provision even for its basic needs), nor the one possible in the actual circumstances, but a worse. Besides these there is a fourth he must recognize – the constitution which will suit pretty well all states. This is why the majority of those who give their views on constitutions, however well they may do it in other respects, fall down on questions of utility. For we must consider not only the best constitution but also the possible, and likewise also that which is easier and more within the reach of all states.

1288b39 But there are of course some today who concentrate on the search for the highest constitution, which needs ample resources; others talk rather about some common3 constitution, yet dismiss entirely the constitutions that actually exist and simply give their approval to the Lacedaemonian or some other. But what is needed is the introduction of a system which the people involved will be easily persuaded to accept, and will easily be able to bring in, starting from the system they actually have. Hence it is a no less difficult task to put a constitution back on its feet than to create one from the start, just as relearning a lesson is no less hard than learning it in the first place. Thus it is another of the duties of a statesman, in addition to those stated, to be able to render assistance to actually existing constitutions, as noted before.4 But he cannot do this without knowing how many forms of constitution there are Some people think that there is only one democracy and only one oligarchy. This is not true, and therefore one should not forget how many differences there are between constitutions and how many different ways there are of combining them.

1289a11 This same practical wisdom enables one to discern both which laws are best, and which of them suit each constitution. For one ought to lay down laws to fit constitutions (as indeed is always done), not constitutions to fit laws. A constitution is the arrangement which a state adopts for the distribution of offices, and for the determination of sovereignty in the constitution and of the end which the particular association aims at realizing. Laws distinguishable from descriptions of constitutions are those according to which the rulers shall rule and shall watch out for those that transgress them. This makes it quite clear that even for the purpose of laving down laws it is necessary to start with a grasp of the varieties of constitutions and the definition of each, because it is impossible for the same laws to be good for all oligarchies and all democracies, since there is more than one form of each, both of oligarchy and of democracy.

IV ii

It is difficult to know what to make of this chapter. It may be merely an alternative draft of the programme of inquiry outlined in IV i, to which its final paragraph is fairly similar. If however it was meant to be taken as a confirmation and development of it, then the train of thought would seem to be: ‘We envisage an inquiry into constitutions, and we have previously distinguished six. But by virtue of having already discussed the “best” constitution, we have in effect dealt with two of the three “correct” ones. (Implication: there is an order of merit among these.) Now the three “perverted” constitutions also may be graded; but in order not of merit, but demerit. But leaving questions of grading aside, we must, before discussing the various constitutions, distinguish the sub-varieties of each’ (cf. IV i ad fin.) The programme of inquiry is therefore now restated with variations of emphasis, to prepare afresh for IV iii ff.; in this revision, greater emphasis is given to the relevance of the sub-varieties, and one further and final item (e) is added to the list of topics to be investigated. But even if this reconstruction is correct, the programme of inquiry of IV i and ii is followed only approximately in the subsequent chapters and books: it refers primarily to IV iii–xiii, but topics (d) and (e) at the end of the chapter appear to look ahead to V and VI (see notes to text). On the significance of this for the composition and structure of the Politics, see the Select Bibliographies (esp. commentaries).

1289a26 In our first inquiry1 into constitutions we analysed them as follows: the right constitutions, three in number – kingship, aristocracy, and polity – and the deviations from these, likewise three in number – tyranny from kingship, oligarchy from aristocracy, democracy from polity. Aristocracy and kingship have been discussed already2 (because the inquiry into the best constitution is identical with discussion of these two words, as both have the same aim, that is virtue endowed with material goods), and the distinctions between them have been drawn;3 we have also determined when it is that kingship is to be adopted.4 It therefore remains to discuss polity,5 which is called by the name which is common to them all, and then the other three: oligarchy, democracy and tyranny.

1289a38 As to these three deviations, which is worst and which second worst – this is obvious, for the deviation from the first and most divine must be the worst. Now kingship, unless it is receiving a name to which it is not entitled, must exist in virtue of the great superiority of the person who reigns. Accordingly, tyranny is the worst, and is farthest away from polity; oligarchy comes second, for aristocracy is very different from this kind of constitution; while democracy is the most moderate of the deviations. One of my predecessors6 has expressed such views on these matters, but he takes a different standpoint from mine. He thinks that where all are reasonable enough (when oligarchy, for instance, and the rest are of a serviceable standard), then democracy is the worst of them; but when all are bad. democracy is the best. My view is simply that these constitutions are erroneous; and it is therefore improper to speak of one oligarchy as better than another, but only as less bad.

1289b11 But let us leave such estimates aside and go on to analyse: (a) How many varieties of constitutions there are, on the assumption that there is more than one form of oligarchy and democracy. (b) Which is the most universal7 constitution, and which is the most preferable constitution after the best; and which other, if any, though aristocratic and well-constructed, is even so suitable for most states – what constitution will answer to that description? (c) Among other constitutions too. which is preferable for whom, since probably for some democracy is essential, rather than oligarchy, for others oligarchy rather than democracy. (d) In what manner one who wishes to set up these constitutions should go to work in each several case, I mean each particular form both of democracy and of oligarchy.8 And lastly (e), when we have given as good and succinct an account of all these as we can, we must tackle the question which are those processes that destroy constitutions, and which are those that keep them stable, both in general and in each particular case; and also what causes most naturally give rise to these processes.9

IV iii

Somewhat unexpectedly, in view of the ‘programme’ of IV i–ii, Aristotle now tells us briefly why there are several different constitutions. States have various ‘parts’ (households, professional groups, etc.) in various combinations, and the resulting variety of economic and social pressures inevitably produces a variety of constitutions. Aristotle considers unsatisfactory the popular view that there are fundamentally only two constitutions, democracy and oligarchy, the others being grouped roughly under one or other of these headings; it would be truer, he suggests, to think of one or two ‘well-formed’ constitutions, the others being ‘deviations’ from them. He thus prefers a classification based on ‘value judgements’ to a purely descriptive one. In effect the chapter prepares us for the several different kinds of constitution we shall meet in subsequent chapters, by contrasting a popular classification with a philosophical, which looks for the reasons for the complexity of constitutional forms and seeks to grade them on moral criteria.

1289b27 The reason for the plurality of constitutions lies in the plurality of parts in every state. We observe in the first place that all states are composed of households, and then that of this total number of persons some must be wealthy, others poor, others in the middle between them; furthermore the rich and the poor are respectively armed and unarmed. We also observe that the people are divided into three classes, agricultural, commercial, and mechanical workers. There are also differences among the upper classes, according to their wealth and the extent of their possessions. We ask, for example, how many horses a man keeps. (Horse-rearing is always difficult without wealth. Hence in ancient times the states whose power lay in their horses had oligarchies, and they made use of their horses in war against states whose borders were contiguous. We see this in Chalcis and Eretria and, on the Asiatic side, Magnesia on the Maeander, and other areas.) In addition to wealth there are other differentiae, of family, virtue, and any other similar feature we described as ‘part’ of a state, when in our discussion of aristocracy we were analysing the essential parts of any state.1

1290a3 Sometimes all these parts have a share in the constitution, sometimes a large number of them, sometimes fewer. It thus becomes clear that there must be several constitutions differing in kind from each other, since the parts differ in kind among themselves. For a constitution is the arrangement of the offices, which are everywhere distributed either according to the power of the participants, or on an equal basis, that is, equality as between the propertyless, for example, or as between the propertied, or some other equality as between them both. There must therefore be as many constitutions as there are arrangements of the superiorities and differences between the parts.

1290a13 But they are commonly reckoned to be two: in the same way as the winds are sometimes classified into northerly and southerly, the rest being deviations from these, so there are supposed to be two constitutions – democracy and oligarchy. For people treat aristocracy as a type of oligarchy, it being oligarchy in a way, and polity, as we call it, as a type of democracy, as the west wind is taken to be a sort of north wind and the east a sort of south wind. The same kind of duality some people find also in music; they lay down two types of mode, Dorian and Phrygian,2 and give one or other of these names to all the other arrangements of notes. People have therefore formed a firm habit of looking at constitutions in this way. But our own classification is better, as well as more accurate, because the well-formed constitutions are two (or perhaps only one), and all others are deviations, either from the absolutely best or from the harmonious and well-balanced mixture.3 These deviations we label oligarchical, if they are too strict and master-like,4 but democratic if they are loose and relaxed.

IV iv

This long and difficult chapter contains several repetitions and inconsistencies which suggest that the material which Aristotle left has fallen into disarray. Comment is devoted to elucidation of the structure of the chapter as it stands. I have divided it into five sections:

Definitions of democracy and oligarchy.

Democracy and oligarchy are commonly defined in excessively simple terms. Numbers are not in themselves decisive (cf. III viii): e.g. democracy is the rule of the poor, the dēmos, be they few or many; and they are usually the latter. Wealth and status, as well as numbers, are crucial for classification.

The parts of the state, and resulting variety among constitutions (1).

Constitutions are of several kinds, because every state has several parts which can be combined in various ways, just as different kinds of animals could be produced by different combinations of different varieties of ‘parts’ (i.e. different sorts of mouth, stomach, limb etc.). A list of the parts of a state is now embarked upon.

Plato on the parts of the state.

A digression in refutation of Plato’s views in the Republic on the minimum number of parts essential to a state if it is to be self-sufficient.

The parts of the state, and resulting variety among constitutions (2).

A resumption of the list of the parts. Often enough, one man can perform more than one function in the state; but no man can be both poor and rich, so it is easy to regard the large numbers of the poor and the small numbers of the rich as ‘opposites’ among the parts of the state. Hence constitutions are thought to be two: democracy and oligarchy.

Varieties of democracy.

Democracy (and oligarchy) will obviously take various forms according to the type of person dominant in it – traders, fishermen, artisans etc. A list of five kinds of democracy follows. The chapter closes with a discussion of the last and most extreme and tyrannical form, where the people decide everything by decree at the bidding of demagogues, and dispense with the rule of law.

Summarized thus, and surveyed as a whole, the chapter presents an understandable if ungainly sequence of argument. Space forbids discussion of the many points of political and philosophical interest, and I fasten on one only, which we have had occasion to notice before (see introduction to III viii): the constant assumption that dēmokratia means exactly, and starkly, what it says – power exercised by the dēmos, that is, by a particular social class, in its own interest (cf. p. 362).

Definitions of Democracy and Oligarchy

1290a30 It is a mistake, which some people habitually make today, to describe democracy and oligarchy in terms too simple and absolute, saying that there is a democracy where the mass of the people is sovereign (as if the majority1 were not also sovereign in oligarchies and everywhere), and an oligarchy where few are sovereign in the constitution. Suppose a total of 1,300; 1,000 of these are rich, and they give no share in office to the 300 poor, who also are free men and in other respects like them; no one would say that these 1,300 lived under a democracy. Or again, suppose the poor to be few, but to be stronger than the well-to-do, who are more numerous than they: no one would call such a constitution oligarchical, where a share in honours is not given to the others, the rich. Therefore we should say rather that it is a democracy whenever the free are sovereign, oligarchy when the rich are sovereign; but what actually occurs is that the former are many, the latter few: many are free, few are rich. Otherwise,2 if the offices were distributed on a basis of height, as is said by some to be done in Ethiopia, or of handsome appearance, then there would be an oligarchy, because the numbers of tall or handsome are small.

1290b7 Nevertheless even these considerations,3 on their own, are not adequate to define these constitutions. Both democracy and oligarchy consist of a number of parts; and whenever the free are not numerous, but rule over a majority who are not free, we still cannot say that it is a democracy, nor that it is an oligarchy where the rich rule in virtue of superior numbers. The former situation existed in Apollonia on the Ionian Gulf, and on Thera; for in each of these states honours were restricted to a minority – those persons of distinguished ancestry who had taken part in the original settlements. The other state of affairs existed at one time in Colophon, where before the war with Lydia4 the majority had amassed substantial possessions. A democracy exists whenever those who are free and are not well-off, being in the majority, are in sovereign control of government, an oligarchy when control lies with the rich and better-born, these being few.

The Parts of the State, and Resulting Variety among Constitutions (1)

1290b21 That constitutions are several, and the reason why they are several, has now been shown. Let us now start from the point mentioned before,5 and show that they are even more numerous than was stated, and say what they are and from what cause. We agree that every state is composed of many parts, not just one. Now if our chosen subject were forms of animal life, we should first have to answer the question ‘What is it essential for every animal to have?’ And among those essentials we should have to include some of the organs of sense-perception, and something for the processing and reception of nourishment, such as mouth and stomach, and in addition parts of the body which enable the animal in question to move about. If these were all that we had to consider and there were differences between them (several different kinds of mouth, for instance, of stomach, of sense-organs, and of parts to do with locomotion), then the number of ways of combining these will necessarily make a number of different kinds of animals. For it is impossible for one and the same animal to have several different kinds of mouth or ear. So when you have taken all the possible couplings of these one with another, they will produce forms of animal; and the number of forms of animal will be equal to the number of combinations of essential parts.

1290b38 We may apply this to the constitutions mentioned; for states too are made up not of one but of many parts, as has often been said. These are (a) the large numbers of people concerned with food-production, called tillers of the soil; (b) the part called mechanical, by which we mean people who follow those skills in the absence of which a state is uninhabitable (these skills are further divided into the absolutely essential and those who minister to luxury or the good life); (c) the commercial, by which we mean that section which spends its time on buying and selling, merchant commerce and retail trade; (d) the section comprising hired labourers; (e) the element which will defend in time of war. This last is just as indispensable as the others, if the population is not to be a slave to aggressors. For of course no state can possibly have a claim to the name, if it is naturally a slave to others: a state is self-sufficient, which is just what a slave is not.

Plato on the Parts of the State

1291a10 For this reason I find the treatment given to this topic in the Republic6 to be more clever than adequate. For Socrates says that a city is compounded of four absolutely essential elements, and that these are the weaver, the farmer, the shoemaker and the builder; then, deeming these to be not self-sufficient, he adds metalworkers and those in charge of indispensable livestock, then the merchant and the retailer. This becomes the full complement of the ‘first’ state, apparently on the supposition that every state is formed to satisfy minimum needs and not rather for a finer purpose; and also that its need for shoemakers is equal to its need for farmers. To defenders in war he assigns no part until territorial expansion involves them in contact with neighbours and war breaks out.

1291a22 But surely there also ought to be among the four, or any other number of associates, one whose duty it will be to decide upon and render justice. If the soul is to be regarded as part of a living creature even more than its body, then in states too we must regard the corresponding elements as being parts in a fuller sense than those which merely conduce to utility and necessity: I mean such things as the fighting force and all those connected with the judicial administration of justice;7 and over and above these, that deliberative element which represents political wisdom in action.8 It is irrelevant to my argument whether these qualities are to be found separately in several people or in the same; it is quite normal for the same persons to be found bearing arms and tilling the soil. If then those on both lists9 are to be counted as parts of the state, it is clear that at least10the heavy-armed military element is an essential one.

The Parts of the State, and Resulting Variety among Constitutions (2)

1291a33 Those who render service11 by their possessions are a seventh part; we call them the well-to-do. An eighth part is composed of those who are on official business, i.e. who render public service11 in connection with the offices – since a state cannot do without officials. So there must be persons capable of holding office and rendering service11 of this kind to the state, either continuously or by turns. There remain those who have in fact just been distinguished,12 the deliberators and those who give decisions where matters of justice are in dispute.

1291a40 Now if these elements must exist in states, and exist in a fine and just manner, it becomes essential that some of the citizens should be possessed of virtue. Many suppose that the other capacities may very well coexist in the same persons: the same people may be defenders, farmers, skilled workmen, and judges and deliberators too. All these have some claim to virtue also, and believe themselves to be capable of most of the offices. But the same people cannot be both rich and poor, and that is why the prime division of a state into parts seems to be into poor and the well-to-do. Further, owing to the fact that the one group is for the most part numerically small, the other large, these two parts appear as opposites among the parts of the state. So the constitutions are accordingly constructed to reflect the predominance of one or other of these, and there seem to be two constitutions – democracy and oligarchy.

Varieties of Democracy

1291b14 That constitutions are several and for what reasons they are several, has been mentioned before; let us now show that there are several forms both of oligarchy and of democracy. But this is clear even from what has been said already. There is on the one hand the people, on the other the notables, as we call them. Of each of these there are several kinds. For example, of the people one kind is engaged in agriculture, another in crafts, and yet another in commerce, in buying and selling. Another kind takes to the sea, and there they fight or trade or carry passengers or catch fish. (In many places the sections of the population engaged in one or other of these occupations are large: fishermen are numerous at Tarentum and Byzantium, traders at Aegina and Chios; at Athens many are engaged on triremes, at Tenedos in passenger traffic.) To these we may add the labouring class and those whose possessions are so small that they cannot have any time off, also those who are not of free birth on both sides, and any other similar kind of multitude. The distinguishing marks of the notables are wealth, good birth, virtue, education and the other things listed under the same heading.13

1291b30 The first variety of democracy is that which is so called because it is based chiefly on the principle of equality. In such a democracy the law interprets equality as meaning that the poor shall not enjoy any more advantage than the rich, that neither shall be sovereign, but both shall be exactly similar. For if, as is held by some, freedom is especially to be found in democracy, and also equality, this condition is most fully realized when all alike share most fully in the constitution. But since the people are a majority, and the decision of the majority is sovereign, this must be a democracy. Here, then, is one type of democracy.

1291b39 Another type has a property-qualification for office, but only a modest one, so that he who gets the requisite amount is allowed to share in office, but ceases to do so if he loses it. Another type is when all citizens share unless they fail to pass a scrutiny,14but the law rules. Another type is when everyone shares, provided only that he is a citizen, but again the law rules. Another type of democracy is the same in other respects, but the multitude is sovereign and not the law. This occurs when the decrees are sovereign over the provisions of the law.

1292a7 It is the demagogues who bring about this state of affairs. When states are democratically governed according to law, there are no demagogues, and the best citizens are securely in the saddle; but where the laws are not sovereign, there you find demagogues. The people becomes a monarch, one person composed of many, for the many are sovereign, not as individuals but as an aggregate. What kind of multiple rulership, collective or of several individuals, Homer meant,15 when he spoke of it as being a bad thing, I do not know. But at all events, such a people, in its role as a monarch, not being controlled by law, aims at sole power16 and becomes like a master, giving honour to those who curry its favour. Such a democracy is the counterpart of tyranny among monarchies. Hence its general character too is exactly the same: both play the master over the better sort of person, and the decrees of democracy are the directives of tyranny; the tyrant’s flatterer is the same as, or analogous to, the demagogue, each having special influence in his sphere, flatterers on tyrants, demagogues on peoples such as I have described. They are able to do this primarily because they bring every question before the people, and make its decrees sovereign instead of the laws. This greatly enhances their personal power because, while the people is sovereign over all, they rule over the people’s opinion, since the multitude follows their lead. Moreover, the accusers of the officials claim that the decision ought to belong to the people; the people need no second invitation, and so all the offices are brought low.

1292a31 So if you were to say that such a democracy is not a constitution at all, your strictures would seem to be perfectly right. Where laws do not rule, there is no constitution. The law ought to rule over all, in general terms, and the officials ought to make rulings in individual cases; then we can decide we have a constitution. So if democracy is one of the constitutions, it is clear that this kind of set-up. where everything is governed by decree, is not a democracy at all, in the real sense; for no decree can have general validity. So much for the classification of democracy.

IV v

Aristotle now fulfils the promise of the preceding chapter, to give a list of the various kinds of oligarchy, and appends a brief but valuable reminder that constitutions are not static, and do not always lend themselves to tidy classification: a democracy, for instance, may be a democracy in law, but an oligarchy in temper and operation; and these ‘mixed’ situations are produced particularly by changes of constitution. There is an important political truth here: a given formal or legal constitution may, in sufficiently determined hands, be ‘manipulated’ for ends radically different from its ostensible aims; this is indeed a danger inherent in the discretion allowed to officials under almost any constitution. Aristotle however barely begins to discuss this possibility, though the last sentence of the chapter suggests that he was aware of it.

1292a39 Of oligarchy there are four types. In (a), access to office is restricted by a property-qualification such that the not so well-off, though more numerous, have no share in the constitution, but a share is open to him who acquires property. In (b), there is a very high property-qualification, and they themselves1 choose those who are without.2 (If this choice is made from among all these,2 the practice is considered to be aristocratic, if from a limited group only, oligarchical.) (c) This is hereditary, son succeeding father in office, (d) This also is hereditary, but the officials rule, not the law. This type of oligarchy holds among oligarchies a position analogous to that of tyranny among monarchies, and among democracies to that extreme democracy of which we have just been speaking. An oligarchy of this type is in fact called a ‘power-group’.

1292b11 This completes our list of types of oligarchy and democracy; but it should not be forgotten that things often turn out differently in practice. There are plenty of instances of a constitution which according to its law is not democratic, but which owing to custom and training is democratic in its workings; conversely, there are in other places constitutions which according to law incline towards democracy, but by reason of their customs and training operate more like oligarchies. This is especially apt to happen after a change of constitution. The citizens do not at once discard their old ways, but are at first content to gain only moderate advantages from their victory over the opposing side, whichever that may be. The result is that the existing laws continue to be valid, but power is in the hands of those who have brought about the change in the constitution.

IV vi

After the digression in the second part of IV v, Aristotle now gives a list of four kinds of democracy and then four of oligarchy; this latter list reads like an expansion of the first part of IV v, and the chapter as a whole has many points of affinity with IV iv and IV xiii.

The classification of democracies and oligarchies in this chapter is made in terms of what Aristotle sees as an inverse ratio between the amount of property possessed by the politically dominant part of the state, and the rule of law: the greater the property, the greater the opportunity for that part to seize the reins of power for itself, and the less willingness to let law rule. Aristotle thinks this analysis holds good both for oligarchy and for democracy; in the latter case, it was notoriously a mark of extreme democracy actually to pay holders of office a kind of salary out of state revenues, a practice which facilitated control of government by the poorer sections of the population. Connections between wealth and political power fascinated Plato too (see e.g. Republic VIII–IX andLaws 756b–e, 758b).

On payments to office-holders, cf. II xi, V viii, and M. H. Hansen, ‘Misthos [pay] for magistrates in classical Athens’, Symbolae Osloenses, 54 (1979), pp. 5–22.

1292b21 The statements already made have listed clearly enough the forms of democracy and oligarchy. For either all the aforementioned parts of the people must have a share in the constitution, or only some and not others. When the farming element, and the element in possession of a moderate amount of property, is the sovereign in the constitution, the constitution is operated in accordance with the laws, because so long as they work they have enough to live on; but they cannot afford to take time off,1 so they put the law in charge and attend only the necessary meetings of the Assembly. But the rest of the population, as soon as they acquire enough property to qualify them according to the law, also have the right to participate. Thus all who have acquired it have that right. For where some do not enjoy the right to participate on any terms, that is a mark of oligarchy. But to have the right to take time off is not possible without revenue.2 This then is one form of democracy, and these are its causes.

1292b34 Another form of democracy is based on the next distinction: birth. Here office is open to all about whose birth there is no question,3 but only those participate who can take time off. In such a democracy the laws rule because there is no revenue. A third form of democracy allows participation to all who are free, although they do not in fact participate for the reason already stated; so here too the law inevitably rules. The fourth type of democracy is in point of time the last to develop in states. The reason for this lies in their growth. Not only are they much larger than they originally were, but they have much larger revenues. Thus all participate, because the mass of the people preponderates; and even the poor, being able to have time off, take part in the administration of the constitution, receiving pay for doing so. In fact, the mass of the poor take the most time off: they have no encumbrances, while the wealthy, who have private affairs to look after, often do not take part in the Assembly and courts of law. Thus in this fourth kind of democracy it is not the laws that are sovereign in the constitution, but the mass of the poor.

1293a10 So much for the list of kinds of democracy and their characteristics, and the constraints that produce them. Here is the succession of types that can be observed in oligarchies. First, where those who own possessions are many but the amount that they own is on the small side and not too much, they allow participation in government to everyone who acquires possessions; and because of the large numbers who thus become members of the citizen-body, the laws and not the men are necessarily sovereign. This is because the further they are removed from the exercise of rule by a single person,4 and possess neither so much that they can afford to neglect it and take time off, nor so little that they are maintained by the state, the more they are bound to think it best that the laws should rule for them, and not they themselves.

1293a21 Next, when the owners are fewer than in the previous case and their possessions larger, we have the second type of oligarchy. Having more power, they expect to make their position more profitable. So they restrict entry into the citizen-body from outside their number to those of their own choosing; but because they are not yet sufficiently powerful to rule without law, they lay down a law to fit the case;5 if they intensify the process, and become fewer in number and greater in possessions, the third stage of oligarchy is reached. In this they keep offices in their own hands, but do so in accordance with a law which provides for sons succeeding their fathers at their death.

1293a30 The final stage is reached when they overtop all the rest in possessions and in numbers of friends; this kind of power-group is near to single rule,4 the men being sovereign and not the law. This is the fourth form of oligarchy, corresponding to the last form of democracy.

IV vii (1293a35–1293b21)

Aristocracy in the strict sense has already been dealt with in Book III, to which reference is made in this chapter. But since the essential thing about an aristocratic constitution is that it is composed of the best men, the word is often loosely applied to any constitution, such as that of Carthage and Sparta, which officially or unofficially attaches importance to choosing the best men, even when it is oligarchical in its pursuit of wealth, or democratic in its concern for the common people. Such constitutions show a mixture of aims, but they are not formally mixed constitutions like polity, which also, when it inclines towards oligarchy, is apt to get the epithet ‘aristocratic’.

1293a35 There are besides democracy and oligarchy two constitutions, one1 of which is generally recognized and has been included as one type in the list of four – monarchy, oligarchy, democracy, and what goes by the name of aristocracy. The other makes a fifth on that list, and is called by the name which is common to them all, for men call it ‘polity’; but since it is rarely found in practice it is overlooked by the typologists of constitutions, who therefore, like Plato, give a list of only four in their Constitutions. Now the aristocracy which we dealt with earlier2 in this work is aptly named, for this name is justly given only to that constitution which is composed of those who are without qualification best in virtue, not simply to one composed of those who are good in relation to some assumed situation. For only in the former type of aristocracy are good man and good citizen one and the same without qualification;3 the good men in other cases are good only in relation to their own constitution.

1293b7 However, the name aristocracy is used of some constitutions to mark a distinction both from oligarchically run constitutions and from what we call polity – that is to say, it describes a constitution in which election to office depends on merit, not only on wealth. This constitution differs from the other two,4 and is called aristocratic. For even in constitutions which do not publicly promote virtue there are nevertheless persons of good reputation who are regarded as respectable men. So where, as at Carthage, the constitution has a threefold aim, wealth, virtue and the good of the people, it is aristocratic; so too is that of Sparta, where there is a dual purpose only, virtue and the good of the people, and thus a mixing of the two (democracy and virtue). So we may say that apart from aristocracy properly so called, which is the best, there are these two5 and also a third which occurs when what we call polity inclines rather towards oligarchy.6

IV viii (l293b22–1294a29)

The main burden of this subtle and difficult chapter is to present polity as a mixture of oligarchy and democracy, and to combat the common tendency to call ‘aristocracies’ such of those mixtures as lean towards oligarchy. Since in the latter case the influence of wealth makes it confusingly easy to regard the mixture as an aristocracy (see note 3), Aristotle evidently wishes to restrict the use of ‘polity’ to those mixtures which do not produce the features characteristic of aristocracy (e.g. education, good government, virtue), or at any rate do not count them as qualifications for ‘honours’. This thesis requires Aristotle to distinguish rather carefully not only between aristocracy and polity, but also between aristocracy and oligarchy.

The description of ‘mixture of oligarchy and democracy’ is not as it stands particularly illuminating, and in the next chapter Aristotle seeks to explain it in more detail.

1293b22 We have still to discuss what we call polity (and also tyranny). We have placed it here because it is not a deviation, and neither are the above-mentioned aristocracies. In strict truth they have all deviated from the most correct constitution, and so are counted among the deviations which we spoke of originally, and which are themselves deviations from them.1 It is reasonable to defer mention of tyranny till the end because in comparison with the rest it is least of all a constitution, which is what our inquiry is about. Having explained why this arrangement has been adopted, I must proceed to discuss polity; for now that oligarchy and democracy have been explained, the function of polity becomes clearer, since, to put it in a word, polity is a mixture of oligarchy and democracy.

1293b34 But it is those mixtures which lean more towards democracy which are generally called polities, and those which lean towards oligarchy are called aristocracies, because education and good birth belong more to the better off. Moreover the well-to-do appear to have those things for the sake of which malefactors commit crimes; hence the rich are called ‘people of quality’ and ‘notables’. Since therefore aristocracy aims at distributing the highest positions to the best of the citizens, it is said that oligarchies also are composed on the whole of ‘people of quality’. But one thing I think is quite impossible: that a state which is controlled not by the best but by the worst2 should be governed by good laws, and likewise that a state without such government should be ruled by the best. It is not government by good laws where the laws enacted are good but not obeyed. ‘Government by good laws’ should therefore be understood in two senses: obedience to the laws laid down, and well-enacted laws laid down by which people abide (it is quite possible to be obedient to badly enacted laws). And a further distinction is possible: obedience may be given either to the best laws available to them in the circumstances, or to the absolutely best.

1294a9 It is the especial mark of aristocracy that the distribution of honours should have been made in accordance with the virtue of the recipients. Virtue is the definitive principle of aristocracy, as wealth is of oligarchy, and freedom of democracy. The principle of majority-decision belongs to all three: in oligarchies and in aristocracies and in democracies, whatever has been decided by the larger part of those who participate in the constitution is sovereign. For most states, the form ‘polity’ is a misnomer, because the aim of the mixture is merely to have regard to the interests of both well-to-do and poor, both wealth and freedom, whereas almost everywhere the well-to-do and the people of quality are coextensive.3 But since there are the three grounds for claiming equality in a constitution, freedom, wealth and virtue (a fourth claim, called ‘good birth’, arises out of the two last of these three, for good birth is wealth plus virtue going back to one’s forbears), it is clear that the term polity should be applied to the dual mixture of well-to-do and poor, and the term aristocracy kept for the triple mixture. This is the most genuine of aristocracies after the true and primary.4

1294a29 I have now shown that there are other types of constitution besides monarchy, democracy, and oligarchy, and what their characters are. It is clear too how one aristocracy differs from another, and how polities differ from aristocracy; and that the two are closely related.

IV ix (1294a30–1294b41)

Aristotle now gives some practical advice on how a polity may be established by ‘mixing’ oligarchy and democracy. We may either (a) adopt the law of both, or (b) take neither’s law as it stands, but effect a compromise, or (c) take parts of the law of each. Elucidation of (a) comes in the second paragraph of IV xiii.

Once again we see the perennial fascination of the Spartan constitution: cf. II ix and Plato, Laws 682e ff. and 712d ff.

1294a30 Our next task is to continue on from what has been said and describe how what we call polity develops in relation to democracy and oligarchy, and what has to be done in order to establish it. The definitive features of democracy and oligarchy will emerge at the same time; for we have to grasp the differences between these two, and then take as it were a ‘tally’ from each of them to put together.

1294a35 Now one may mix or combine in one or other of three definitive ways. The first is to take the legislation of both. For example, in the matter of judging, in oligarchies they impose a fine on the wealthy for non-attendance as jurymen in the law-courts, and they do not pay the poor for their attendance; in democracies no fines for non-attendance are imposed on the wealthy, and the poor receive pay for their services. Both these together constitute something in common and midway between them, and would therefore, as a mixture of the two, be characteristic of a polity. Thai is one way of joining the two. A second way is to take something intermediate between the two sets of provisions. Thus for membership of the assembly in democracies there is no property-qualification (or only a very small one), whereas in oligarchies the property-assessment is high. Here something common is provided by neither, but by an assessment fixed midway between. The third method is from two sets of regulations to take one part from the oligarchical law, the other part from the democratic. For example, the filling of offices: to do this by lot is regarded as democratic, by selection oligarchic; a property-qualification is oligarchic, its absence democratic. Take therefore one from each, the oligarchical election of officials and the democratic freedom from property-qualification, and the result is both polity-like and aristocratic. So much for the method of mixing.

1294b14 A definitive feature of the well-mixed democracy and oligarchy is that it is possible to describe the same constitution either as democracy or as oligarchy. It is clearly the very excellence of the blending that creates this impression in those who thus describe it. A middle position has the same characteristic: each of the two extremes appears in it. This is exactly what happens in the case of the constitution of the Lacedaemonians. Many people try to describe it as a democracy, because the system has a number of democratic features: first the rearing of the children, under which the sons of the rich are reared in the same way as the sons of the poor and receive an education which the sons of the poor could also receive; then similarly in the next age group, and when they are grown up, the arrangements for feeding in the communal messes are the same for all, for thus there is no outward mark of distinction between rich and poor; and the rich wear clothing which any poor man could get for himself. There is also the fact that the people choose the members of the Council of Elders and share in the Ephorate, the two most important offices in the state. Others call it an oligarchy because of its many oligarchical features: the absence of the use of the lot, all officials being elected; the power of few to pronounce binding sentence of death or exile, and many other similar points. A constitution which is a really well-made combination of oligarchy and democracy ought to look like both and like neither.

1294b36 It should be kept stable by means of itself and not through outside agencies. It is not doing that when the number of those who wish it to continue make a majority (a condition which can equally arise in a bad constitution), but only when no section whatever of the state would even wish to have a different constitution. We have now mentioned the way in which polity ought to be established, and likewise the so-called aristocracies.

IV x

In the following short chapter on tyranny in relation to other types of monarchy Aristotle harks back to certain previous discussions (III xiv–xvii). That which is technically the rule of a tyrant need not necessarily be without law, and such rule has obvious resemblances to kingship. Discussion of tyranny, he claims, is therefore justified.

1295a1 It remained for us to say something about tyranny, not because there is much to say, but so that it might take its place in our inquiry; for even to it we assign a place of a sort in a list of constitutions. We defined kingship in an earlier part of this work and discussed whether, in the most usual sense of the word ‘kingship’, it was or was not a good thing for states, and also who is to be appointed king, and how and from what source. When dealing with kingship we also defined two forms of tyranny, because the power exercised by a tyrant in both cases in some sense overlaps with kingship too, both forms of rule being according to law. (For example, among certain non-Greeks, sole rulers with absolute personal power are elected to that office; and among ancient Greeks of long ago there arose sole rulers of this type, called aisumnētai.) In certain respects these two forms of tyranny differ from each other; but they were king-like because they were according to law, and because the sole rule was over willing subjects; yet they were tyrannical, because rule was exercised as by a master, according to the personal decisions of the tyrants.

1295a17 There is a third type of tyranny, thought to be the most extreme because it is the exact converse of absolute kingship. Any sole ruler, who is not required to give an account of himself, and who rules over subjects all equal or superior to himself to suit his own interest and not theirs, can only be exercising a tyranny of this third kind. Hence it is endured unwillingly, for no one willingly submits to such rule if he is a free man. These then are the kinds of tyranny, and the number of them; and they arise for the reasons given.


The ‘middle’ constitution which Aristotle recommends so enthusiastically in this important chapter seems to be the ‘mixed’ constitution of IV viii and ix (cf. Ill vii), considered from sociological, psychological and economic points of view rather than from a strictly constitutional one. He advocates that a large number of the citizens should each have moderate and adequate property, less than that of the rich and more than that of the poor. The effect of such a ‘middle’ population will be to diminish conflict between rich and poor, and to prevent the constitution from being an extreme one. In short, this chapter gives the conditions in which the technical, political and constitutional arrangements of the ‘mixed’ constitution or ‘polity’ are likely to prove acceptable and workable.

Aristotle attempts to justify the argument of this chapter by reference to one of his best-known philosophical principles, namely that virtue is a mean between two extremes. This ethical rule of thumb suggests the doctrine of a constitutional middle way easily enough, yet somewhat unconvincingly: what real parallel or affinity is there between a mean in conduct and a mean in the level of wealth of a section of the population or in the character of a constitution? Evidently Aristotle would make the connection by arguing that a middle class with a moderate income is disposed to ‘mean’ conduct, which suits a ‘mixed’ constitution. But this, if true, can be so only ‘for the most part’, to use one of Aristotle’s favourite expressions: rich men have been known to act moderately, and people of moderate wealth can be savage in its defence.

In order to avoid possibly misleading associations with the technical Marxist concept of ‘class’, or with modern middle classes, I have, at the cost of some slight awkwardness, excluded the term ‘middle class’ from the translation. Aristotle usually says in this chapter merely hoi mesoi, ‘the middle people’, or to meson (neuter singular), ‘the middle’ the nearest he comes to the notion of class is in his use of the word meros, ‘part’ or ‘section’, e.g. at the beginning of the third paragraph, and in IV xii. As he makes clear at the beginning of the third paragraph, his analysis in this chapter of the structure of the state is based firmly on the amount of wealth possessed by each ‘part’.

1295a25 What is the best constitution and what is the best life for the majority of states and the majority of men? We have in mind men whose virtue does not rise above that of ordinary people, and whose education does not depend on the luck either of their natural ability or of their resources; and who have not an ideally perfect constitution, but, first, a way of living in which as many as possible can join and, second, a constitution within the compass of the greatest number of states. The ‘aristocracies’, as they are called, that we have just been discussing1 do not fall within the competence of most states, but some of them do approximate closely to what we call polity (hence we ought to speak of both constitutions as though they were one and the same).

1295a34 The decision on all these points rests on the same set of elementary principles. If we were right when in our Ethics2 we stated that virtue is a mean, and that the happy life is a life without hindrance in its accordance with virtue, then the best life must be the middle life, consisting in a mean which is open to men of every kind to attain. And the same principles must be applicable to the virtue or badness of constitutions and states. For the constitution of a state is in a sense the way it lives.

1295b1 In all states there are three state-sections: the very well-off, the very badly off, and thirdly those in between. Since therefore it is agreed that moderation and a middle position are best, it is clear that, in the matter of the goods of fortune also, to own a middling amount is best of all. This condition is most easily obedient to reason, and following reason is just what is difficult both for the exceedingly rich, handsome, strong and well-born, and for their opposites, the extremely poor, the weak, and those grossly deprived of honour. The former incline more to arrogance and crime on a large scale, the latter are more than averagely prone to wicked ways and petty crime. The unjust deeds of the one class are due to an arrogant spirit, the unjust deeds of the other to wickedness. Add the fact that it is among the members of the middle section that you find least reluctance to hold office as well as least eagerness to do so; and both these attitudes. eagerness and reluctance, are detrimental to states.

1295b13 There are other drawbacks about the two extremes. Those who have a superabundance of good fortune, strength, riches, friends, and so forth, neither wish to submit to rule nor understand how to do so; and this is engrained in them from childhood at home: even at school they are so full of la dolce vita that they have never grown used to being ruled. Those on the other hand who are greatly deficient in these qualities are too subservient. So they do not know how to rule, but only how to be ruled as a slave is; while the others do not know how to be ruled in any way at all, and can command only like a master ruling over slaves. The result is a state not of free men but of slaves and masters, the former full of envy, the latter of contempt. Nothing could be farther removed from friendship or from partnership in a state.3 Sharing is a token of friendship; one does not want to share even a journey with one’s enemies. The state aims to consist as far as possible of those who are like and equal, a condition found chiefly among the middle people. And so the best-run constitution is certain to be found in this state, whose composition is, we maintain, the natural one for a state to have.

1295b28 It is the middle citizens in a state who are the most secure: they neither covet, like the poor, the possessions of others, nor do others covet theirs as the poor covet those of the rich. So they live without risk, not scheming and not being schemed against. Phocylides’ prayer was therefore justified when he wrote, ‘Those in the middle have many advantages; that is where I wish to be in the state.’4

1295b34 It is clear then both that the best partnership in a state3 is the one which operates through the middle people, and also that those states in which the middle element is large, and stronger if possible than the other two together, or at any rate stronger than either of them alone, have every chance of having a well-run constitution.

For the addition of its weight to either side will turn the balance and prevent excess at the opposing extremes. For this reason it is a most happy state of affairs when those who take part in the constitution have a middling, adequate amount of property; since where one set of people possess a great deal and the other nothing, the result is either extreme democracy or unmixed oligarchy, or a tyranny due to the excesses of either. For tyranny often emerges from an over-enthusiastic democracy or from an oligarchy, but much more rarely from intermediate constitutions or from those close to them. The reason for this we will speak of later when we deal with changes in constitutions.5

1296a7 The superiority of the middle constitution is clear also from the fact that it alone is free from factions. Where the middle element is large, there least of all arise factions and divisions among the citizens. And big states are freer from faction, for this same reason, namely that their middle element is large. In small states it is easy for the whole body of citizens to become divided into two, which leaves no middle at all, and nearly everybody either rich or poor. Democracies too are safer than oligarchies in this respect and longer-lasting thanks to their middle people, who are more numerous and take a larger share of honours in democracies than in oligarchies. For when in their absence the unpropertied preponderate in numbers, trouble arises and they soon come to grief. An indication of the truth of what we have been saying is to be found in the fact that the best lawgivers have come from the middle citizens – Solon, for example, whose middle position is revealed in his poems, and Lycurgus, who was not a king, and Charondas and most of the rest.6

1296a22 These facts also show why most states are either democratic or oligarchic; for the middle being frequently small, whichever of the two extremes is on top, those with possessions or the common people, abandons the middle and conducts the constitution according to its own notions, and so the result is either democracy or oligarchy. Then again, owing to constant strife and faction between the people and the wealthy, neither side, whichever of the two succeeds in gaining the mastery, ever sets up a constitution which is equally based and acceptable all round. Taking supremacy in the constitution as a prize of victory they proceed to set up a democracy or an oligarchy as the case may be. Also, those7 who came to exercise leadership among the Greek states installed democracies or oligarchies in them according to the constitution which each had at home, looking entirely to their own advantage, not to that of the states themselves. So for these reasons the middle constitution has never occurred anywhere, or only seldom and sporadically. Only one8 of a long succession of leaders was prevailed upon to allow a system of this kind. And to this day into whatever state you go, you will find that they have got into the habit of not even wanting equality: their aim is to rule, failing which they accept a condition of defeat.

1296b2 Which constitution is best, and why, will be clear from the above. As for the rest, the different kinds of oligarchy and democracy which we say there are, it is not difficult to arrange them in order of merit, this one better, that one worse, and so on; for now that the best is decided upon, proximity to it must denote better, and the farther away one moves from the middle, the worse; unless of course one judges by a reference to a given situation;9 for often enough, although one constitution is more desirable, there is nothing to stop another from being, for some people, more advantageous.

IV xii


This somewhat obscurely articulated chapter is a good example of the informal and semi-systematic way in which Aristotle’s thought is apt to move. I interpret as follows. Chapter xi (of Book IV) had concluded that in certain circumstances the best or preferable constitution, the ‘mixed’ one based on the ‘middle people’, may be abandoned in favour of one more ‘advantageous’. Aristotle now naturally, and in apparent abandonment of the subject of the ‘middle’ people, is prompted to ask the question: what constitution is ‘advantageous’ to what sort of person? Obviously democracy to the many poor, oligarchy to the wealthy few; but there will be no stability in such constitutions unless the element that wishes it to continue is ‘stronger’ than that which does not. Aristotle suggests, if I understand him aright, that the numbers of the poor may be ‘stronger’ than the ‘quality’ of the rich (their wealth, education, etc.); or this ‘quality’ may be ‘stronger’ (meaning ‘more influential’, presumably) than the numbers of the poor; in the former case, a democracy is natural, in the latter an oligarchy. There is some point, presumably varying from case to case, at which the ratio between number and quality shifts decisively one way or the other; and the greater the imbalance, the more extreme will be the democracy or oligarchy that results.

Now comes, at the beginning of the third paragraph, the crucial junction of thought: both democracy and oligarchy should always (sc. in spite of being, as surh, ‘advantageous’ to their respective adherents) somehow accommodate the ‘middle people’ – because thus the ‘stronger’ element will be even stronger, and the constitution will be more stable; and of course without stability the constitution will not be as advantageous to the ‘stronger’ element as it might be. Stability is therefore in a sense the key to the chapter, which turns out to have been concerned all along, like its predecessor, with the constitutional merits of an influential ‘middle people’. In effect it argues that even partisans of ‘extreme’ constitutions have or should have, on their own showing, an interest in cultivating the middle ground.

In the final paragraph, Aristotle notes that makers of aristocratic constitutions commit the mistake of trying to ensure stability not by ‘mixing’ their constitutions but by tricking the people. In the next chapter he proceeds to list some of the subterfuges employed.

1296b13 It is proper to follow what has been said by a discussion of the question of what constitution is advantageous for what persons, and what kind of constitution for what kind of men. First, we must grasp a principle which is universally applicable to them all: it is essential that that part of the state which desires the permanence of the constitution should be stronger than that which does not.

1296b17 Now every state can be measured either qualitatively – I mean by such qualities as freedom, wealth, education, and good birth – or quantitatively, that is by numerical superiority. Look at the parts which make up a state: it is possible that quality may be present in one, quantity in another (e.g. the non-noble may be numerically greater than the noble, the poor than the rich), but without quantitative superiority being enough to compensate for the qualitative inferiority. These must therefore be weighed one against the other. Where the number of the poor is sufficiently large to exceed the given ratio, there democracy naturally arises; and the type of democracy will depend on the type of people which has the numerical superiority in each case. Thus, if those who cultivate the soil make up the superior numbers, the democracy will come first1 on the scale; if those engaged in mechanical work and receiving pay for it predominate, then it will come last,1 and similarly with the rest in between. Where, on the other hand, the rich and notable people have a greater qualitative superiority than quantitative inferiority, there an oligarchy naturally arises, and once again its type will depend on the degree of superiority in those who form the oligarchical body.

1296b34 But at all times a legislator ought to endeavour to include the middle people in the constitution. If he is framing laws that are oligarchical in character, he should have the middle people always in view; if democratic, he should again make them attractive to those in the middle. Wherever the middle people outweigh a combination of the two extremes, or even one only, then there is a good chance of permanence for the constitution. There is no danger of rich and poor making common cause against them; for neither will want to be slaves to the other, and if they are looking for a constitution more acceptable2 to both, they will not find any better than this. Their mistrust of each other would make it impossible for them to accept alternation in office. But in all places the mediator is best trusted by the parties, and the one in the middle is a mediator.

1297a6 The better mixed a constitution is, the longer it will last. It is a mistake made by many, even by those seeking to make an aristocratic constitution, not only to give too great a preponderance to the rich, but to cheat the people. In the long run mistaken good inevitably gives rise to unmistakable evil; for the greedy grabbing of the rich does more harm to the constitution than that of the people.

IV xiii


In the preceding chapters Aristotle has dwelt on the importance to a constitution of gaining the adherence of the majority (cf. introduction to II x), and in closing IV xii he had mentioned how this may be achieved by trickery. A list of such tricks is now given. Oligarchy can be made to appear less objectionable to the people in five ways, which may even strengthen its position. It is evidently less easy for democracies to delude oligarchs, and the question becomes merged in the general question of mixing the two into a moderate constitution. But it is by no means clear in this chapter when Aristotle is using politeia in the general sense of constitution and when he means ‘polity’.

In the third paragraph Aristotle recommends a legitimate device, of allowing only arms-bearers to ‘share in the constitution’; since Greek soldiers commonly had to supply their own equipment, this is in effect to fix a property-qualification that will exclude the poor but still ensure that supporters of the constitution are in a majority. (The recommendation does not spring from any particular admiration for soldiers as such.) After a brief historical review of the role of the armed forces as it affected constitutions, Aristotle rounds off his present set of topics by a summary that only partly corresponds to the first three items in the ‘table of contents’ given in IV i–ii.

1297a14 There are five devices which are employed against the people in order to give a constitution a more attractive appearance. They concern the Assembly, the offices of power, the law-courts, the carrying of arms, and physical training. Thus first, membership of the Assembly may be open to all, fines for non-attendance being imposed only on the rich (or much larger fines on them than on others). Next the offices: those with a certain property-qualification are not allowed to decline office by oath,1 but the poor are. Thirdly the law-courts: fines are imposed on the rich if they do not serve as jurymen, but the poor are exempt (or else, as in Charondas’2 laws, there are small fines in the latter case, large in the other). Sometimes membership both of Assembly and of juries is thrown open to any persons who have their names put on the roll; but if, being thus enrolled, they fail to serve in the Assembly or the court, they incur large fines. It is the deliberate intention that the threat of a fine should cause people to avoid enrolment, and that non-enrolment should lead to their serving neither in law-court nor in Assembly. Similar regulations are made about carrying arms and gymnastic training: it is lawful for the poor not to possess arms, whereas the rich are fined if they do not have them. And if they fail to attend gymnasia, there is a penalty for the rich but none for the poor, so that the possibility of a penalty may cause the rich to attend; while the poor, having no penalty to fear, may absent themselves.

1297a34 These are oligarchical devices in the framing of laws; in democracies they are made in the opposite sense. They provide pay for the poor who serve in assembly or law-court, and impose no fine on the rich for non-attendance. That is why, if it is desired to make a just mixture of oligarchy and democracy, obviously the thing to do is to draw upon both sides and include both the payment for the one and the fine for the other. In this way all would have a share; otherwise the resultant constitution will be in the hands of one set or the other.

1297b1 Citizenship ought to be reserved for those who carry arms.3 This means imposing a property-qualification, whose amount cannot be settled absolutely and once and for all. The level must be fixed no lower than is absolutely necessary to ensure that the number of those who share in the constitution is larger than those who do not. For the poor are generally content enough, even if they do not share in honours, provided only that they are not liable to be ill-treated or deprived of any of their possessions. (That, however, is far from easy to achieve: members of a citizen-body do not always behave considerately.) And in time of war those who have no resources of their own are reluctant to serve as soldiers unless they are fed, but are quite ready to fight if rations are provided.

1297b12 In some places the constitution includes those who have served as hoplites4 as well as those serving currently; this was the case in Malis, but only those on active service were elected to hold office. The earliest constitution (after the kingships) among the Greeks was in fact composed of warriors, of the cavalry in the first place, because it was in them that strength and superiority in war were to be found (for without organized formations a hoplite force is useless, and the ancients had no fund of experience of such things and no tactical procedures5 for them, so that their strength rested with their cavalry). Then when states became larger and those with arms became stronger, the number of sharers in the constitution became larger. For this reason what we nowadays call ‘polities’ were formerly called democracies. But the constitutions of those early times were, understandably enough, oligarchical or royal; for owing to the smallness of the population their middle element was not large, and so, being neither numerous nor well organized, they were content to allow others to rule.

1297b28 We have now stated why constitutions are several in number, why there are others besides those commonly mentioned (there is not one kind only of democracy, and the same applies to the rest), and further what the differences are and the causes of them; and in addition which, speaking generally, is the best of the constitutions, and which kind among the others suits which kind of person.

IV xiv

Discussion of ‘stratagems’ in the preceding chapter now evidently prompts Aristotle to deal in the remainder of this book with various patterns of constitutional machinery that may be found ‘expedient’ in different constitutions – a topic which follows on here naturally enough, since there is practical value in it, and the book began with a reminder of the need to be practical as well as theoretical. Whatever be the form of government, it must provide for three things: (a) discussion and decision about what is to be done, (b) officers of state of all grades to carry out the policy, and (c) a judicial system. The combinations and permutations of the appointment of, and relationships between, the personnel of the deliberative, administrative and judicial functions of government are extremely complex and defy summary. On the assumption that Aristotle is not inventing them, or not all of them, out of his head as an academic exercise, they may serve to point up the immense variety of constitutional procedure evolved by the ancient Greek city-states, which we tend to forget through our customary concentration on some of the better-known examples (Athens in particular) of the major forms of constitution (democracy, oligarchy, etc.). As Aristotle often remarks, and as is abundantly clear from the Politics as a whole, none of these forms was monolithic.

We should beware of supposing that Aristotle’s threefold division of powers corresponds more than very roughly to the fairly sharp modern ‘separation of powers’ as between legislature, executive and judiciary: he evidently envisages considerable overlapping of functions. Here again he reflects ancient Greek practice.

1297b35 In dealing with our next topic we must speak again about the constitutions, both in general terms applicable to all and also severally, using the appropriate starting point. There are three elements in each constitution in respect of which every serious lawgiver must look for what is advantageous to it; if these are well arranged, the constitution is bound to be well arranged, and the differences in constitutions are bound to correspond to the differences between each of these elements. The three are, first, the deliberative, which discusses everything of common importance; second, the officials (here we ask what they ought to be, the limits of their sovereign powers, and the methods by which they are selected); and third, the judicial element.

1298a3 The sovereign powers of the deliberative element cover decisions as to war and peace, the making and dissolving of alliances, legislation, the penalties of death, exile and confiscation of goods, the choosing of officials, and the scrutiny of their conduct on expiry of tenure. The right to decide on all these matters must either be given to all the citizens, or all to some (e.g. to one or more officials, or some to some officials, others to others), or some of them to all, others to some. To allow everyone to decide, and on all matters, is democratic;1 for such equality is what the common people seek.1

1298a11 The principle that everyone should deliberate may be applied in various ways: first, they may all perform the work but by turns rather than collectively (as in the constitution of Telecles2 the Milesian): and in others there is an arrangement whereby the boards of officials deliberate jointly, but everyone enters upon office in turn, from the tribes and the smallest parts until the whole process is complete, coming together only for legislation and for constitutional matters, and to hear pronouncements from the officials. A second method involves the collective action of everyone; but they come together only for the purposes of electing officials, making laws, deciding on war and peace, and holding scrutinies, the remaining matters being deliberated on by officials appointed for the purpose in each case, these officials being appointed either by lot or by election from among all. A third method is for the citizens to meet about offices and scrutinies, and to deliberate on wars and alliances, but for the other matters to be managed by the officials, as many of them as possible elected; and for such officials expert knowledge is an essential prerequisite. Under the fourth method all come together to deliberate about everything, and the officials take no decisions but only carry out preliminary investigations. This is the way a modern extreme democracy works, which we maintain corresponds to an oligarchy that is run by a power-group, and to a monarchy that is a tyranny.3

1298a33 All these four methods are democratic; but where only some deliberate about all matters, the principle is oligarchic. This principle likewise can be applied in several different ways: first, when those eligible to participate in deliberation are elected on the basis of a property-qualification that is not too high, and are on that account fairly numerous, when they obey the law and do not change what it forbids them to change, and when the attainment of the required property level permits one to participate – then, though an oligarchy, such a constitution by reason of its moderation is like a polity. Next, when not all participate in deliberation but only those elected, and they rule according to law no less than in the previous case, then that is oligarchic. When again those who control the deliberating themselves choose their own members, and son succeeds father, and they control the laws, that arrangement is necessarily and in the most extreme degree oligarchical. Finally, when certain persons have control of certain matters, such as when all control war and peace and scrutinies, and the remaining business is controlled by the officials who are appointed by election and not by lot, the constitution is an aristocracy. When some matters are under the control of elected persons, others of those chosen by lot, and these latter are taken either from all or from a pre-elected few, or the elected and the lot-chosen function together, then some of these features are characteristic of an aristocratic constitution,4 others of polity itself.

1298b11 We have now distinguished in this manner the deliberative element in relation to the constitutions, and the administration of each constitution is as indicated by our analysis. But for what is nowadays reckoned to be a democracy in the fullest sense (I mean one in which the people is sovereign even over the laws) it is advantageous to adopt, in order to encourage better deliberation, a procedure used by oligarchies for their courts. I am referring to their practice of prescribing fines for failure to serve as jurymen as a means of making sure that those whom they want to serve really do appear in court (while democrats provide pay for the poor). To do this in relation to the assembly would, as I say, be beneficial, as it will ensure the presence of the notables as well as of the people, and they will deliberate the better when each side does so with the other. It is a good thing also that the membership of the deliberative element should be determined either by election or lot, and, if by lot, then an equal number from the sections of the population; and also, whenever the democrats among the citizens are numerically much superior, it is a good thing either not to provide pay for them all for their attendance, but to pay only a number equal to the number of notables, or to eliminate the excess5 by lot.

1298b26 In oligarchies, on the other hand, it will be found expedient to pick certain extra members from the mass of the people, or else, as is done in some constitutions, to set up a committee (pre-councillors, law-guardians, are the names given), and then to deal only with such business as has already been pre-deliberated by them. In this way the people will participate in deliberation without being able to set aside any of the provisions of the constitution. Alternatively, they should pass resolutions that are identical to, or at any rate not inconsistent with, the recommendations.6 A further alternative is to bestow only advisory powers on everyone, reserving full deliberative powers for the officials. One should in fact do just the opposite of what is done in constitutions:7 one should make the people sovereign when it rejects, not sovereign when it comes to approval, in which case there must be a reference back to the officials. In constitutions’ they act in the reverse manner: the small body is sovereign in rejection, but not in approval, in which case the matter is always referred back to the larger body. This then is the way in which an analysis should be given of the deliberative and sovereign element in the constitution.

Iv xv

Aristotle now turns to the second of the three ‘elements’ he had enumerated at the beginning of IV xiv, the executive (or, as he puts it, the archai, the ‘offices’ or ‘officials’). He raises a large number of issues: (a) What kinds of official are there? (b) Who appoints them? (c) From whom are they appointed? (d) How are they appointed? (e) What is their tenure of office? (f) What are their function and powers? (g) How do the various officials operate in relation to each other? (h) Which officials, modes of appointment, powers and tenure etc. are appropriate for which constitutions?

Aristotle’s long and discursive treatment of these topics is rich in observation and analysis, and two features of the chapter are of special interest: (1) His attempt, in the second paragraph, to isolate just what we mean by ‘official’, as distinct from persons merely ‘in charge’ of something. As can be seen from Newman’s comments and references ad loc. (cf. also III i and Plato, Laws 767a, 768c) the controversy was a live one. Aristotle briefly tries various lines of approach and plumps provisionally for the rather widely drawn criterion that officials deliberate, decide and give ordersthe last function being crucial. But after these inconclusive remarks he drops the search for definition as being of academic interest only. (2) His elaborate analysis, towards the end of the chapter, of the ways in which the various features of appointment may be combined. Admittedly, this matter is of practical interest, as particular constitutions find it expedient to combine them in particular ways; yet, as he virtually admits, he carries the mechanical analysis beyond the point where it could serve any useful purpose. One wishes he had devoted less energy to this enterprise and more to the philosophically interesting question, ‘What is an official as distinct from a mere functionary?

1299a3 Next, we turn to classify the officials. Great variety is also to be found in this element in the constitution: questions arise about their number and the scope of their sovereign powers, about the length of tenure of office of each (it is made six months in somecases, in others less, in others a year, but in yet others longer periods). Should tenure be perpetual, or for a very long term? And if neither of these, then the question arises whether the same persons should hold office repeatedly, or for one period only, being ineligible for a second. Then there is the officials’ appointment – from whom are they to be drawn, by whom appointed, and in what manner? We ought to be able in all these matters to determine how many different ways are possible and then match them up, looking to see what sorts of officials are best for what sorts of constitution.

1299a14 Another question – and even this is not easy to answer – is, which kinds ought to be called ‘officials’? The association which we call the state1 needs a great many people to take charge of various things, so many that we should not designate them all as officials, neither those appointed by lot nor those elected. I am thinking in the first instance of the priests, whose position is quite different from the offices of state. Then there are the heralds and the trainers of choruses,2 and envoys to be sent abroad, all these being appointed by election. Some responsibilities affect the state, and of these some affect all the citizens, for the purposes of a particular activity (e.g. a general, while they are under arms), others a section only (e.g. a controller of women or children). Other responsibilities affect the household;3 for example, they frequently elect persons to measure out corn. There are also ancillary responsibilities which, when resources permit, will be performed by slaves. Roughly speaking, we may say that officials are those empowered to deliberate and make decisions on certain matters and to issue orders; and particularly the last, since this is the essence of rule. All this makes hardly any difference in practice; disputes about the terms have not resulted in any decision. But it is a question which is open to further intellectual investigation.

1299a31 Turning to more difficult questions, we ask what kind of officials, and how many, are essential – essential, that is, if there is to be a state at all; we ask also what kind, while not essential, are yet useful for a sound constitution. Such questions need to be discussed in relation to every constitution, and in particular in relation to small states. For in the large states it is both possible and necessary to assign separate tasks to separate officials, one to each man because the large number of citizens makes it possible for many of them to sit on the committees, so that one man never holds the same office twice, or only after a long interval, and because every job gets done better when it is the responsibility of one official, not of many. In the smaller states, on the other hand, a number of offices have to be concentrated in a few hands (on account of the small population it is not easy for many to hold office, for who will there be to succeed?). However, sometimes the small states require the same officials and laws as the large; but whereas the small need the same persons to serve frequently, in the others this need arises only at long intervals. Thus there is really no reason why a number of different responsibilities should not be assigned together, because they will not hinder each other, and as a way of counteracting lack of manpower it is essential to make committees serve more than one purpose. (A hook will hang a roast as well as a lamp.) If therefore we can say how many officials are essential for any state at all, and how many are not essential but still ought to be there, then knowing all this we shall find it easier to deduce what sorts of offices are fit to be amalgamated into one.

1299b14 Another point not to be overlooked is fittingly mentioned here. What sort of matters ought to be supervised by many boards for different places, and which other sorts by a single official, whose authority is sovereign everywhere? I am thinking of such matters as keeping order in the market. Should there be a separate market-controller in every place, or one for everywhere? Should divisions be made according to the work or according to the persons? Is all regulation of good order a task for one man, or should there be a separate one for women and children? Similar questions arise when we consider offices in relation to constitutions. Does each constitution make some difference, or none, to the type of offices? Thus, are the same offices sovereign in democracy, in oligarchy, in aristocracy, and in monarchy, even though they are filled not from equal nor even similar people, but from different people in different constitutions – in aristocracies from the educated, in oligarchies from the rich, in democracies from the free? Or are there some offices that correspond precisely to the differences in the officials,4 and in some places do the same offices serve best, in others different ones? For it may happen that here an extensive role, there a minor, for the same office its the situation best.

1299b30 There are however offices peculiar to certain situations and not found everywhere, for example that of ‘pre-councillors’. This is not a democratic institution, but the council itself is; for a body of this kind whose responsibility is to deliberate on business beforehand is necessary so that the people shall get through their work. If this council consists of only a few men, it is oligarchical. But pre-councillors must necessarily be few in number; therefore they are always an oligarchical feature. Where both offices exist, there the pre-councillors stand over the members of the council; for the councillor is a democratic institution, whereas the ‘pre-councillor’ is an oligarchic one.

1299b38 The function of even the council may be weakened at the other end also, as in those democracies in which the people itself meets and takes a hand in everything. This is apt to happen wherever members of the assembly are well paid for their attendance, because this gives them time off to come to frequent meetings and decide everything themselves. A controller of children, a controller of women, and officers with sovereign powers to discharge responsibilities similar to these, are an aristocratic feature, not a democratic one (for who could prevent the wives of the poor from going out when they want to?). It is not oligarchic either, for the wives of oligarchs are rich and pampered.

1300a8 So much for the present on these matters; I must next try to describe from the beginning the ways of appointing officials. The alternative ways in which this can be done fall into three definitive groups, the combination and permutation of which will necessarily be found to include all the methods. First we ask who are those who make the appointments, second from among whom do they make them, third in what manner it is done. To each of these three questions there are three different answers: either all citizens appoint or only some do; appointments are made either from all citizens or from some specified persons,5 the qualification for appointment being property-group or birth or virtue or some similar thing (as at Megara, where only those qualified who had been among the returning exiles who had fought together against the people);6 and all this may be either by lot or by election. These coupled alternatives yield further couplings as follows: some appoint to some offices, all appoint to others; some offices are filled from all, others from some; some are filled by lot, others by election.7

1300a22 8For each of these alternatives there is a choice of four methods: all appoint from all by election or all from all by lot; or all appoint from some by election or all from some by lot (and if from all, either by sections, for instance tribe, deme, or brotherhood, and so on through all the citizens, or else from all on every occasion); or partly one way and partly the other.9 Again, if the appointers are only some, not all, the appointments may be made either by election from among all or by lot from among all, or by election from some, or by lot from some, or partly one way and partly the other – I mean partly by election from all, partly by lot, and partly by election from some, partly by lot. This gives twelve methods, apart from the two couplings.

1300a31 Of these modes of appointment, three are democratic – all appoint from all either by lot or by election or both, that is, to some offices by lot, to others by election. But that not all should appoint at one time,10 and either from among all or from some, by lot or election or both, or some officials from all and others from some, either by lot or election or both (that is some by lot, some by election) – that is a characteristic of a polity. When some appoint from all, by lot or election or both (some officers by lot and others by election), that is oligarchical, and the more so when they appoint from both.11 To appoint12 some officials from all, and others from only some, or some by election and others by lot, is characteristic of a polity, but with a bias towards aristocracy. But it is oligarchic when some appoint from some by election, and when some appoint from some by lot (it makes no difference that this is not in fact done), and when some appoint from some by both methods. When only some appoint from among all, and when all elect from some, that is aristocratic.

1300b5 Such then is the number, and classification in relation to the different constitutions, of the methods of appointing officials. Which method is best for what people, and how the appointments should be made, are questions which can only become clear in conjunction with decisions about what offices there are and what powers they have. By ‘power’ of an office I mean such things as sovereign control of revenue and of security; for the power of a general is different in kind from the function of controlling commercial contracts.

Appendix to IV xv

As Rackham remarks in the Loeb translation, 1300a22–b5 is a ‘dizzy passage’ (locus vertiginosus). If I understand it aright, the text down to a31 may be schematized as shown in the diagram. (In 1300a23 I read ‘four’, not ‘six’.)

These are the ‘twelve methods’: 1–4 and 7–10 each describe a method of appointment to all offices; 5–6 and 11–12 each describe a method of appointment coupling election and lot, in the sense that some offices are filled by election, others by lot. (Line 5 is in effect a conflation of 1 and 2, 6 of 3 and 4, and so on.) The ‘two couplings’, which are not represented in the diagram below, are Ai + Aii (all apoint to some offices, some to others), and Bi + Bii (all are appointed to some offices, some to others). Each of the six ‘althernalives’, all couplings apart, has ‘four methods’, represented by the four ticks in each column which have two companions in the same horizontal

‘These (six) alternatives’


line; e.g. there are four ways of conducting an election by choosing officials from ‘some’ (Bii) of the persons qualified. Hence, the two missing couplings apart, A, B, and C have a total of ‘twelve methods’ (p. 286, middle).

In the second paragraph of the passage, Aristotle isolates which methods are characteristic of which constitutions, and complicates the picture still further by making use of all couplings, Ai + Aii and Bi + Bii as well as Ci + Cii (which is the most frequent): see footnotes for the less obvious instances.

IV xvi

Finally, Aristotle treats the third of the ‘elements’ enumerated in IV xiv, the judicial system. This chapter exhibits broadly the same features as the preceding two: it contains an analysis of types of court, a description of how appointments of jurymen are made, and notes on which modes of appointment suit which constitutions.

Aristotle assumes that there will be a separate court for each of the eight main types of suit. The first of these to be mentioned is the euthuna, a scrutiny or investigation made of an outgoing official’s conduct. It is strange that there is no mention of suits about legacies and about property generally, which we know to have been of frequent occurrence. But the whole chapter is only a series of scraps and unfulfilled intentions.

1300b13 Of the three elements, only the judiciary now remains to be discussed. Here too the various procedures may be reviewed on the same principle. Courts of law differ in three definitive respects: (a) from whom the members are drawn, (b) the matters about which they have jurisdiction, (c) the manner of appointment. In regard to the first we ask whether members are drawn from all or from some; in regard to the second we ask how many types of court there are; in regard to the third we ask whether appointment is by lot or by election.

1300b18 I begin with the second of these, the number of types of court. There are eight of these, concerning (1) scrutinies, (2) offences against the public interest, (3) matters relating to the constitution, (4) disputes between officials and private persons about the imposition of fines, (5) private transactions of some magnitude, (6) homicide, (7) foreigners, (8) minor transactions involving sums between one drachma and five, or a little more (for even these require a legal decision, though not a body of jurymen). Under (6). there are four kinds of homicide court (whether with the same or a different set of judges), concerning (i) deliberate slaying, (ii) unintentional slaying, (iii) cases where the offence is admitted but justification is claimed, and (iv) cases where the charge is brought against persons exiled for murder, with a view to their return. A reputed example of this is the court in Phreatto1 at Athens; but such cases are rare at all times, even in large states. Under (7), there are two kinds, one for foreigners disputing with foreigners, the other for foreigners disputing with citizens. I will say no more about these,2 or about the homicide courts, or about those concerning foreigners, but will speak of those relating to the state; for if these are mismanaged, divisions follow, and changes in constitutions.

1300b38 If all judge all the matters listed, they must be appointed either (i) by lot or (ii) by election, or (iii) partly one way and partly the other, or else, although dealing with the same matters, (iv) some by lot and others by election.3 Thus there are four procedures, and four others where the appointments to the jury are made on a sectional basis. For once again, those who judge on all matters may be taken from some only, and may be appointed either by election or by lot (or partly by election and partly by lot); or some courts dealing with the same matters may consist both of elected members and of members appointed by lot. These then, as stated, are the procedures corresponding to those already listed.4 But we may in addition have the same pairings, part of a court being appointed from among all, part from some, or from both together, as happens for instance if members of the same court are appointed some from all, others from some, and either by election or by lot or by both. The possible procedures relating to courts of law have now been stated; of these the first (from all and about all matters) is democratic, the second (from some and about all matters) is oligarchic, the third (partly from all and partly from some) is characteristic of aristocracy and polity.

You can support our site by clicking on this link and watching the advertisement.

If you find an error or have any questions, please email us at Thank you!