Aristotle · a new plain-English translation from the original language
1| Topics, Book 1. The aim of this treatise is to find a method by which we will be able to reason deductively about any problem set before us on the basis of generally accepted opinions, and by which, when we ourselves are sustaining an argument, we will avoid saying anything self-contradictory. First, then, we must state what a syllogism is and what its differentiae are, so that the dialectical syllogism may be grasped, since it is this that we are seeking in the treatise before us. A syllogism, then, is an argument in which, certain things having been laid down, something other than what was laid down follows of necessity through what was laid down. It is a demonstration when the syllogism proceeds from things true and primary, or from such things as have gained the origin of the knowledge about them through things true and primary; it is a dialectical syllogism when it reasons from generally accepted opinions. Things are true and primary when they carry conviction not through other things but through themselves; for in the case of the first principles of a science one should not go on asking the reason why, but each of the principles should be convincing in its own right. Generally accepted opinions, on the other hand, are those accepted by everyone, or by the majority, or by the wise -- and among the wise, either by all of them, or by the majority, or by the most notable and reputable of them. A contentious syllogism is one that proceeds from opinions that appear to be generally accepted but are not, or one that merely appears to proceed from generally accepted opinions, or from opinions that merely appear to be generally accepted.
2| For not everything that appears to be a generally accepted opinion actually is one. None of the things called generally accepted opinions has an altogether superficial appearance, of the sort that occurs at the starting points of contentious arguments; for there the nature of the falsehood is at once evident, for the most part, even to those with only a slight power of insight. Of the two kinds of contentious syllogisms just mentioned, let the first be called a syllogism as well; let the other be called a contentious syllogism, but not a syllogism, since it appears to reason but does not actually reason. Further, besides all the syllogisms mentioned, there are the false reasonings that arise from premises proper to particular sciences, as happens in the case of geometry and its kindred sciences. This kind of reasoning seems to differ from the syllogisms mentioned; for the one who draws a false figure reasons neither from things true and primary nor from generally accepted opinions. For he does not fall under the definition: he takes neither what everyone thinks, nor what the majority think, nor what the wise think -- and among the wise, neither what all, nor what the majority, nor what the most reputable of them think -- but constructs his syllogism from premises that are proper to the science in question but are not true. For he produces his false reasoning either by describing semicircles not as they should be described, or by drawing certain lines not as they would be drawn.
3| Let the kinds of syllogisms just described, then, stand as an outline account of them. To speak generally about all of these and those still to be discussed, let it be settled to this extent: that we do not propose to give a precise account of any of them, but wish only to go through them in outline, considering it wholly sufficient for the method before us to be able to recognize each of them in some way. Following on from what has been said, it would be appropriate to state for how many purposes, and which, this study is useful. It is useful for three: for mental training, for encounters with others, and for the philosophical sciences. That it is useful for training is obvious from the facts themselves; for once we possess a method we will be more easily able to argue about whatever is proposed. It is useful for encounters with people because, having enumerated the opinions held by the majority, we will deal with them not on the basis of other people's convictions but on the basis of their own beliefs, shifting whatever they seem to us to be saying incorrectly. It is useful for the philosophical sciences because, being able to work through the difficulties on both sides of a question, we will more easily discern in each case what is true and what is false. Further, it is useful with regard to the primary things among the principles proper to each science. For it is impossible to say anything about these on the basis of the principles proper to the particular science under discussion, since the principles are what is primary in relation to everything else, and it is necessary to work through them by means of the generally accepted opinions about each. This task is peculiar, or most especially proper, to dialectic; for being an art of examination, it holds a path toward the principles of all methods of inquiry.
4| We will possess the method completely when we are in the same position as we are with regard to rhetoric and medicine and abilities of that kind. This means producing, out of what is available, what we have chosen to produce. For neither will the skilled speaker persuade by every possible means, nor will the skilled physician heal by every possible means; but if he leaves out none of the available means, we will say that he possesses the science adequately.
5| First, then, we must consider from what elements the method proceeds. If we could grasp for how many purposes, and of what sorts, and from what materials arguments are constructed, and how we are to become well supplied with these, we would have an adequate grasp of the task before us. Now the things from which arguments proceed and the things about which syllogisms are concerned are equal in number and the same. For arguments arise out of premises, while the things syllogisms are about are problems. Every premise and every problem indicates either a genus, a property, or an accident -- for the differentia, being generic in character, should be classed together with the genus. But since a property may in one case signify the essence and in another case not, let property be divided into both of the parts just mentioned, and let the part that signifies the essence be called a definition, while the remainder be called, following the common usage established for it, a property. It is clear, then, from what has been said, that on the present division everything reduces to four things in all: property, definition, genus, or accident. Let no one suppose we are saying that each of these, taken by itself, is a premise or a problem; rather, it is from these that both problems and premises arise. A problem and a premise differ in their manner of expression. For when it is put this way -- 'Is 'a terrestrial two-footed animal' the definition of man?' or 'Is animal the genus of man?' -- it becomes a premise. But when it is put as 'Whether a footed, two-footed animal is or is not the definition of man' or 'whether animal is its genus,' it becomes a problem. And likewise with the others. So it is reasonable that problems and premises are equal in number; for from every premise you will produce a problem by changing its manner of expression.
6| We must now say what a definition is, what a property is, what a genus is, and what an accident is. A definition is an account that signifies the essence. It is rendered either as an account in place of a name, or as an account in place of another account -- for it is also possible to define some of the things signified by an account. Those who render their account by means of a mere name in any way whatsoever clearly do not render the definition of the thing, since every definition is some kind of account. Still, something of this sort should also be classed as definitional -- for instance, that the fine is the fitting. Likewise too the question whether perception and knowledge are the same or different; for concerning definitions as well, most of the discussion turns on whether they are the same or different. Let us, speaking generally, call definitional everything that falls under the same method as definitions. That everything just mentioned is of this kind is clear from the examples themselves. For if we are able to argue that two things are the same or that they are different, we will in the same way be well equipped to attack definitions as well; for by showing that a thing is not the same as something else, we will have demolished the definition. However, the converse of what has just been said does not hold: showing that two things are the same is not sufficient to establish a definition. But to demolish one, it is sufficient to show that they are not the same.
7| A property is something that does not show the essence, but belongs only to the thing and is counterpredicated of it — for example, it is a property of man to be capable of grammar; for if something is a man, it is capable of grammar, and if it is capable of grammar, it is a man. For no one calls a property something that can belong to something else as well, such as sleeping in the case of man, nor even something that happens at some particular time to belong to it alone. But if any such thing is called a property, it will not be called a property without qualification, but a property at some time or in relation to something: being on the right is at some time a property, while being two-footed happens to be called a property in relation to something, for example of man in relation to horse and dog. And it is clear that nothing that can belong to something else is counterpredicated of it; for it is not necessary, if something is sleeping, that it be a man. A genus is what is predicated in the category of what a thing is of several things that differ in species. Let those things be said to be predicated in the category of what a thing is which it is fitting to give in answer to the question what the thing before us is — just as, in the case of man, it is fitting, if asked what the thing before us is, to say that it is an animal. It also belongs to the genus to determine whether one thing is in the same genus as another or in a different one.
8| For this question, too, falls under the same method as the genus; for by arguing that animal is the genus of man, and likewise of ox, we shall have argued that they are in the same genus. And if we show that the one is a genus of something and the other is not, we shall have argued that they are not in the same genus. An accident is what is none of these — neither definition, nor property, nor genus — but belongs to the thing, and is such that it can belong to any one and the same thing or not belong to it; for example, sitting can belong to some one and the same thing, or not belong to it. Likewise also white: for nothing prevents the same thing from being white at one time and not white at another. Of the two definitions of accident the second is better; for the first, if one is to understand it, requires one already to know beforehand what definition and genus and property are, whereas the second is self-sufficient for recognizing what is meant, in its own right. Let there also be added to accident the comparisons made between things with one another, said in whatever way from accident — for example, whether the fine or the advantageous is more choiceworthy, and whether the life according to virtue or the life according to enjoyment is more pleasant, and anything else that happens to be said in a similar way to these;
9| for in all such cases the inquiry is about which of the two the predicate belongs to more. It is clear from this that nothing prevents an accident from becoming at some time and in relation to something a property — for example, sitting, though it is an accident, will be a property when someone alone is sitting, but when he is not the only one sitting, it is a property in relation to those who are not sitting. So that nothing prevents an accident from becoming a property both in relation to something and at some time; but it will not be a property without qualification. We should not fail to notice that the points made regarding property and genus and accident will fit for being said with regard to definitions as well. For by showing that what is given does not belong only to the thing under the definition, as was also the case with property, or that what is rendered in the definition is not a genus, or that something of what is stated in the account does not belong, which could also be said with regard to accident, we shall have demolished the definition; so that, on the account given earlier, all the things enumerated would in a way be relevant to definitions. But this is not a reason to look for one single universal method applying to all cases; for it is neither easy to find such a thing, nor, even if it were found, would it be at all clear or serviceable for the inquiry before us. Once a method proper to each of the distinguished genera has been given, the way through the subject before us would be easier, proceeding from what is proper to each. So that in outline, as has been said before, one must divide the subject, and then, of the remainder, attach to each what most belongs to it, calling these points ones bearing on definition and ones bearing on genus. And in fact what has been said has for the most part already been attached to each.
10| First of all, one must determine in how many ways the same is spoken of. The same would seem, taken in outline, to be divided into three. For we are accustomed to call things the same in number, or in species, or in genus: the same in number, those of which there are several names but the thing is one, for example cloak and mantle; the same in species, things which, though several, are undifferentiated in respect of species, as man is the same as man and horse as horse; for such things are said to be the same in species, as many as fall under the same species. Likewise also things are the same in genus, as many as fall under the same genus, for example horse and man. But the water from the same spring, said to be the same, would seem to have a certain difference beyond the ways mentioned; nevertheless, let even such a case be classed together with the things called the same under one species, whatever the manner of speaking. For all such things seem to be akin and similar to one another. For all water is said to be the same as all water in species, because of having a certain similarity, and the water from the same spring differs from other water in no other way than in having a more intense similarity; hence we do not separate it from the things called the same under one species, however that is meant. But it is the one that is the same in number that seems, by the most general agreement, to be what everyone means by the same.
11| This too is customarily rendered in several ways: most properly and primarily, when the same is rendered by a name or a definition, as cloak is the same as mantle, and footed biped animal is the same as man; secondly, when it is rendered by the property, as capable of knowledge is the same as man, and what by nature moves upward is the same as fire; and thirdly, when it is rendered from the accident, as the one sitting, or the musical one, is the same as Socrates. For all these mean to signify the one that is numerically one. That what has just been said is true, one could best learn from cases where people change the designations. For often, when giving orders to call by name one of those sitting, we change the designation, when it happens that the person to whom we are giving the order does not understand, thinking he will understand better from the accident, and we bid him call to us the one sitting, or the one conversing — clearly supposing that we signify the same thing both by the name and by the accident.
12| The same, then, as has been said, is to be divided in three ways. That arguments proceed out of, and through, and in relation to the things spoken of before, one assurance is that from induction; for if one were to examine each of the premises and problems, it would appear to have arisen either from the definition, or from the property, or from the genus, or from the accident. Another assurance is that through syllogism. For it is necessary that everything predicated of something either be counterpredicated of the thing or not. And if it is counterpredicated, it would be either a definition or a property: for if it signifies the essence, it is a definition, and if it does not signify it, it is a property; for this was what a property was — what is counterpredicated but does not signify the essence. But if it is not counterpredicated of the thing, it either is or is not among the things stated in the definition of the subject. And if it is among the things stated in the definition, it would be a genus or a differentia, since the definition is made up of genus and differentiae; but if it is not among the things stated in the definition, it is clear that it would be an accident; for the accident was said to be what is neither definition nor genus nor property, but belongs to the thing.
13| After this we must mark off the kinds of predication in which the four things we mentioned are found. These are ten in number: what a thing is, quantity, quality, relation, place, time, position, having, doing, being affected. For the accident, the genus, the property, and the definition will always fall under one of these kinds of predication; for all the premises formed through these signify either what a thing is, or a quality, or a quantity, or one of the other kinds of predication. It is clear from the examples themselves that the person signifying what a thing is sometimes signifies substance, sometimes a quality, and sometimes one of the other predications. For when a man is set before us and someone says that what is set before us is a man, or an animal, he states what it is and signifies substance; but when a white color is set before us and he says that what is set before us is white, or a color, he states what it is and signifies a quality. Likewise if a magnitude of a cubit is set before us and he says that what is set before us is a magnitude of a cubit, he will state what it is and signify a quantity. It is the same in the other cases too; for each of these, whether it is said of itself or its genus is said of it, signifies what it is. But when it is said of something else, it does not signify what it is, but a quantity, or a quality, or one of the other predications. So these are the things about which and out of which our arguments are made, and this is how many they are; but how we are to grasp them, and by what means we shall become resourceful in them, must be stated next.
14| First, then, let it be marked off what a dialectical premise is, and what a dialectical problem is. For we must not take every premise, nor every problem, as dialectical; for no one in his senses would put forward as a premise what no one holds, nor would he propose as a problem what is evident to all or to most people; for the latter admit no difficulty, and the former no one would maintain. A dialectical premise is a question that is reputable, held either by all, or by most, or by the wise -- and among these, either by all, or by most, or by the most notable of them -- and is not paradoxical; for one would accept what seems true to the wise, provided it is not contrary to the opinions of the many. Dialectical premises also include things resembling reputable opinions, and the contraries of opinions that seem to be reputable, when they are put forward in contradictory form, and all opinions that accord with established arts. For if it is reputable that the knowledge of contraries is the same, it would appear reputable also that the perception of contraries is the same; and if grammar is one in number, flute-playing too is one, and flute-playing one; but if there are several grammars, then several flute-playing arts too; for all these seem to be alike and akin. Likewise the contraries of reputable opinions, put forward in contradictory form, will appear reputable too;
15| for if it is reputable that we ought to treat our friends well, it is also reputable that we ought not to treat them badly. Now to say that we ought to treat our friends badly is the contrary, while to say that we ought not to treat them badly is the contradictory negation. Likewise, if we ought to treat our friends well, we ought not to treat our enemies well. This too is the contradictory of the contrary; for the contrary is that we ought to treat our enemies well. And so on in the other cases. Also, by way of comparison, the contrary about the contrary will appear reputable, for instance, if we ought to treat our friends well, we also ought to treat our enemies badly. It might also appear that treating friends well is the contrary of treating enemies badly; but whether this is really so or not will be discussed among the remarks made about contraries. It is clear too that all the opinions that accord with the arts are dialectical premises; for one would accept the views of those who have studied these matters -- for instance, about matters of medicine, as the doctor holds them, and about matters of geometry, as the geometer holds them; and likewise in the other cases.
16| A dialectical problem is a subject of inquiry that bears either on choice and avoidance, or on truth and knowledge, either in itself or as contributing to some other such subject, about which people hold no opinion either way, or the many hold an opinion contrary to that of the wise, or the wise contrary to that of the many, or each group is divided within itself. For some problems it is useful to know for the sake of choosing or avoiding something, for instance whether pleasure is to be chosen or not, while others are useful only for the sake of knowing, for instance whether the universe is eternal or not; and others are, in themselves, of no use for either of these purposes, but contribute toward some problem of this kind; for there is much that we do not wish to know for its own sake, but for the sake of other things, so that through it we may come to know something else. There are also problems about which there are opposed arguments on both sides, raising a difficulty as to which way the matter really stands, because there are persuasive arguments on both sides; and problems concerning great matters for which we have no argument, because we think it hard to give the reason, for instance whether the universe is eternal or not; for such things too someone might inquire into. The problems, then, and the premises have been marked off as has been said;
17| a thesis is a paradoxical supposition held by some notable person in philosophy, for instance that contradiction is impossible, as Antisthenes said, or that all things are in motion, according to Heraclitus, or that being is one, as Melissus says; for to concern oneself with the view of some chance person who declares opinions contrary to the received ones would be foolish. Or a thesis may be about matters concerning which we have an argument contrary to the received opinions, for instance the sophists' claim that not everything that is has either come to be or is eternal; for a man who is musical is literate without either having come to be so or being so eternally. For this, even if someone does not accept it, would seem to be so because there is an argument for it. A thesis, then, is also a problem, but not every problem is a thesis, since some problems are of the kind about which we hold no opinion either way. That a thesis is also a problem is clear; for from what has been said, it is necessary either that the many dispute with the wise about the thesis, or that either group disputes within itself, since a thesis is a paradoxical supposition. Nearly all dialectical problems are now called theses. Let it make no difference whatever it is called; for it was not for the sake of coining names that we made this division, but so that it should not escape us what differences there in fact are among them.
18| We ought not to examine every problem or every thesis, but only one that someone might be puzzled about who stands in need of argument, and not of punishment or of the senses; for those who are puzzled whether one ought to honor the gods and love one's parents or not stand in need of punishment, while those puzzled whether snow is white or not stand in need of perception. Nor should we examine matters where the demonstration lies too close at hand, nor those where it lies too far off; for the former admit of no difficulty, while the latter involve more difficulty than belongs to mere exercise. Now that these points have been marked off, we must distinguish how many kinds of dialectical arguments there are. One kind is induction, the other syllogism. What a syllogism is has been stated before; induction is the progress from particulars to universals, for instance, if the pilot who has knowledge is the best pilot, and likewise the charioteer, then in general the person who has knowledge is the best in his own field. Induction is more persuasive and clearer and more familiar according to perception, and shared by the many, while syllogism is more compelling and more effective against argumentative opponents.
19| The classes with which arguments are concerned, and from which they are drawn, have been marked off as before described. As for the tools by which we shall be well supplied with syllogisms and inductions, there are four: first, to get hold of premises; second, to be able to distinguish the number of ways in which each thing is said; third, to find the differentiae; fourth, the study of what is similar. And in a way the last three of these are premises too, for it is possible to make a premise out of each of them, for instance, that the fine or the pleasant or the advantageous is choiceworthy, and that perception differs from knowledge in that it is possible to recover the one once it is lost while it is not possible with the other, and that the healthy stands to health as the fit for good condition stands to good condition. And the first sort of premise proceeds from what is said in many ways, the second from the differentiae, the third from things that are similar.
20| Premises are to be selected in as many ways as have been marked off concerning the premise: either by taking in hand the opinions held by everyone, or by most people, or by the wise—and among these either all of them, or most, or the most reputable—or the opinions contrary to those that appear to hold, and all the opinions that belong to the various crafts. And one must also propound the premises contrary to the reputable opinions that appear to hold, taken by way of contradiction, as was said before. It is also useful, in selecting them, to construct premises not only out of those that are actually reputable but also out of those similar to these, for instance, that of contraries the perception is the same (for so too is the knowledge), and that we see by receiving something in, not by sending something out—for it is the same way with the other senses too, since we hear by receiving something in, not by sending something out, and we taste likewise, and similarly with the rest. Further, whatever appears so in all cases or in most, this must be taken as a starting point and an accepted position; for those who do not see the exception in some particular case posit it as holding without qualification. One should also select premises from written discourses, and construct outlines concerning each genus separately, setting them down apart—for instance concerning the good, or concerning animal, and concerning the good as a whole, beginning from what it is.
21| One should also note down the opinions held by individual thinkers, for instance that Empedocles said the elements of bodies are four; for one might posit as a premise something said by some reputable person. And to give a rough outline, the parts of premises and problems are three: some premises are ethical, some are natural, some are logical. Ethical premises, then, are of this kind: whether one ought to obey one's parents or the laws, if they conflict; logical premises are of this kind: whether the knowledge of contraries is the same or not; natural premises are of this kind: whether the universe is eternal or not. And the problems are similar. As for what each of the aforementioned kinds is, it is not easy to render this in a definition; one must try to recognize each of them through the familiarity gained by induction, examining them in light of the examples given above. Now for philosophy one must deal with these matters with a view to truth, but dialectically with a view to reputable opinion. One should take all the premises in as universal a form as possible, and turn the one into many, for instance that the knowledge of opposites is the same, then that this holds of contraries, and that it holds of relatives. And in the same way these in turn must be divided, for as long as division is possible, for instance into the good and the bad, and the white and the black, and the cold and the hot.
22| And similarly with the rest. Concerning the premise, then, what has been said is sufficient. As for the question of the number of ways a thing is said, one must deal not only with the things said in a different way but must also try to render the accounts of them, for instance not merely that the good is said in one way of justice and courage, and in another way of what is fit for good condition and healthy, but also that some of these are so called because they are themselves things of a certain quality, while others are so called because they are productive of something and not because they are themselves things of a certain quality. And likewise with the rest. Whether a thing is said in many ways or in one way alone in respect of form must be studied by the following means. First, one should look at the contrary and see whether it is said in many ways, whether it differs in form or in name. For some things are different straightaway even in name, for instance the contrary to the sharp is, in the case of sound, the flat, but in the case of a solid mass, the blunt. It is clear, then, that the contrary to the sharp is said in many ways. And if this is so, so is the sharp; for in respect of each of those the contrary will be different. For the same thing will not be sharp as contrary to both the blunt and the flat, but the sharp is contrary to each of them separately. Again, contrary to the flat, in the case of sound, is the sharp, but in the case of a solid mass, the light; so that the flat too is said in many ways, since its contrary is.
23| Similarly too, contrary to the fine, in the case of an animal, is the ugly, but in the case of a house, the shoddy, so that the fine is said in many ways, applied to different things by mere similarity of name. In some cases the names do not differ at all, but the difference in form is at once apparent in the things themselves, for instance in the case of white and black. For a sound is called white and black, and so is a color, likewise. In name, then, there is no difference at all, but the difference in form is at once apparent in the things themselves; for a color is not called white in the same way as a sound is. This is clear too through perception: for of things that are the same in form the perception is the same, but the white as applied to sound and to color we do not judge by the same perception, but the one by sight, the other by hearing. Similarly too the sharp and the blunt as applied to flavors and to solid masses—but the one by touch, the other by taste. For these too do not differ in name, either in themselves or in their contraries; for blunt is the contrary to each of them as well. Further, if to one thing there is something contrary but to another, absolutely nothing—for instance, to the pleasure that comes from drinking, the pain that comes from thirst is contrary, but to the pleasure that comes from contemplating that the diagonal is incommensurable with the side, nothing is contrary—then pleasure is said in more than one way.
24| And to loving in the sense of an affection of the mind, hating is contrary, but to loving in the sense of bodily activity, nothing is contrary; it is clear, then, that loving is used homonymously. Further, in the case of intermediates: if to some things there is an intermediate but to others none, or if there is one to both but not the same one—for instance between white and black, in the case of colors, gray is the intermediate, but in the case of sound, none, or if there is one after all, it is the muffled, as some say a muffled sound is the intermediate—then white is used homonymously, and likewise black. Further, if for some things the intermediates are several, but for others only one, as in the case of white and black; for among colors the intermediates are many, but for sound the muffled is the one intermediate. Again, one should look, in the case of what is opposed by contradiction, whether it is said in more than one way. For if this is said in more than one way, then what is opposed to it will also be said to be spoken of in more than one way, for instance not to see is said in more than one way, in one sense not having sight, in another not exercising sight. And if this is said in more than one way, it is necessary that to see too be said in more than one way; for to each sense of not seeing something will be opposed, for instance to not having sight, having it, and to not exercising sight, exercising it.
25| Again, one should look at things spoken of in terms of privation and possession: for if one of the pair is said in several ways, so is the other. For example, if being perceptive is said in several ways, both of the soul and of the body, then being unperceptive too will be said in several ways, both of the soul and of the body. And that the terms now mentioned are opposed as privation and possession is clear, since animals naturally have each kind of perception both in respect of the soul and in respect of the body. Again, one should look at the inflected forms. For if 'justly' is said in several ways, then 'just' too will be said in several ways; for corresponding to each sense of 'justly' there is a sense of 'just' — for example, if 'justly' means both judging according to one's own opinion and judging as one ought, then 'just' will be said in the same two ways. Likewise too, if 'healthy' is said in several ways, then 'healthily' too will be said in several ways — for example, if one thing is called healthy because it produces health, another because it preserves it, another because it indicates it, then 'healthily' too will be said in the sense of producing, or preserving, or indicating. And similarly in other cases: whenever the thing itself is said in several ways, the inflected form derived from it will also be said in several ways, and if the inflected form is said in several ways, so is the thing itself.
26| One should also examine the genera that fall under the categories signified by the name, to see whether they are the same in every case. For if they are not the same, then clearly the term used is homonymous. For example, 'good' in the case of food is what produces pleasure, in the case of medicine what produces health, but in the case of the soul it is being of a certain quality, such as temperate or courageous or just; and likewise in the case of a human being. In some cases it signifies a time, as in 'the good' meaning what is timely; for what is timely is called good. And often it signifies a quantity, as in the case of the moderate; for the moderate too is called good. So 'good' is homonymous. Likewise 'white' signifies, in the case of a body, a color, but in the case of a sound, what is pleasant to hear. Similarly too with 'sharp': for it is not said in the same way in every case; a sound is called sharp when it is fast, as the musical theorists who work by numbers say, an angle is called sharp when it is less than a right angle, and a knife is called sharp when it has an acute-angled edge.
27| One should also examine the genera falling under the same name, to see whether they are different and not subordinate to one another — for example, 'donkey' means both the animal and the implement. Their account corresponding to the name is different: in one case it will be spoken of as a certain kind of animal, in the other as a certain kind of implement. But if the genera are subordinate to one another, it is not necessary that the accounts be different. For example, both animal and bird are genera of the crow. So when we say the crow is a bird, we are also saying that it is a certain kind of animal, so that both genera are predicated of it. Likewise, when we call the crow a winged, two-footed animal, we are saying that it is a bird; and so in this way too both genera, and their account, are predicated of the crow. But in the case of genera that are not subordinate to one another this does not happen: for we do not, when we say 'implement,' also say 'animal,' nor, when we say 'animal,' say 'implement.' One should look not only at whether the genera of the term in question are different and not subordinate to one another, but also at the case of its contrary; for if the contrary is said in several ways, clearly so is the term proposed. It is also useful to look to the definition that results when the terms are combined, for example of a white body and a white sound —
28| for once the special feature is stripped away, the same account ought to remain. This does not happen in the case of homonyms, as in the examples just given. For in one case it will be 'a body having such-and-such a color,' in the other 'a sound pleasant to hear'; so once 'body' and 'sound' are stripped away, what remains is not the same in each case. Yet it ought to be, if 'white' as said of each were used in one and the same sense. Often, too, the homonymy escapes notice even within the accounts themselves; hence one should examine the accounts as well. For example, if someone says that what indicates and produces health is what stands in due proportion to health, one should not let the matter drop but should examine what he means by 'in due proportion' in each case — whether in the one case it means being such as to produce health, in the other being such as to indicate what sort of state one is in. Again, one should look to see whether the things are not comparable in terms of more or in terms of alikeness — for example, a white sound and a white cloak, or a sharp flavor and a sharp sound; for these are not called white or sharp alike, nor is one more so than the other. So 'white' and 'sharp' are homonymous. For everything spoken of in one and the same sense is comparable: it will be said either alike, or one more than the other.
29| Since things belonging to different genera not subordinate to one another also have differentiae different in kind — for example, animal and knowledge, since these have different differentiae — one should examine whether the things falling under the same name are differentiae of genera that are different and not subordinate to one another. For example, sharpness of sound and sharpness of a solid angle: a sound differs from a sound by being sharp, and likewise a solid angle differs from a solid angle by being sharp. So 'sharp' is homonymous; for they are differentiae of genera that are different and not subordinate to one another. Again, one should ask whether, among the things falling under the same name, the differentiae are themselves different — for example, color as found in bodies and color as found in complexions. For color in bodies is what distinguishes and gathers the sight, whereas in the case of complexions the differentiae are not the same. So 'color' is homonymous; for of the same things there are the same differentiae. Again, since form is never a differentia of anything, one should examine, among things falling under the same name, whether one is a form and the other a differentia — for example, 'white,' which as applied to a body is a form of color, but as applied to a sound is a differentia; for a sound differs from a sound by being white.
30| Concerning the ways in which a thing is said in several senses, then, one should investigate by these and similar means. As for differences, one should study them within the genera themselves, in relation to one another — for example, how justice differs from courage, and practical wisdom from temperance (for all these belong to the same genus) — and also as between one thing and another drawn from different genera that are not too widely separated, for example how perception differs from knowledge; for in the case of things widely separated the differences are entirely obvious. Likeness should be examined both among things in different genera, on the pattern 'as one thing stands to another, so a different thing stands to a different other' — for example, as knowledge stands to the object of knowledge, so perception stands to the object of perception — and on the pattern 'as one thing is in some other thing, so a different thing is in a different other,' for example as sight is in the eye, so intellect is in the soul, and as calm is in the sea, so stillness is in the air. One ought to train oneself chiefly on things that are widely separated; for we shall then more easily be able to discern the likenesses among the remaining, closer cases. One should also examine things belonging to the same genus, to see whether something the same belongs to all of them — for example, to a human being, a horse, and a dog; for insofar as something the same belongs to them, to that extent they are alike.
31| It is useful to have examined in how many ways a thing is said, both for the sake of clarity (for one will better know what one is positing once it has been made plain in how many ways it is said) and so that our deductions may bear on the matter itself and not on the mere name. For as long as it is unclear in how many ways a thing is said, it is possible that the answerer and the questioner are not directing their thought to the same thing; whereas once it has been made plain in how many ways it is said, and to which of them the answerer is directing his position, the questioner would look ridiculous if he did not frame his argument with reference to that. It is useful too both for not being misled by a fallacy and for producing one. For knowing in how many ways a thing is said, we shall not be misled, but shall recognize it if the questioner is not directing his argument at the same sense; and we ourselves, when questioning, shall be able to produce a fallacy, if the answerer happens not to know in how many ways the thing is said. This is not possible in every case, but only when, among the several senses, some are true and some false. This procedure, however, does not belong properly to dialectic; hence dialecticians should be thoroughly wary of it — of arguing merely about the name — unless one is otherwise altogether unable to argue about the subject proposed. Finding the differences is useful both for deductions about sameness and difference, and for recognizing what each thing is.
32| It is clear, then, that this is useful for arguments about sameness and difference: for once we have found any difference at all between the things proposed, we will have shown that they are not the same. It is also useful for coming to know what a thing is, because we are accustomed to mark off the account proper to each thing's substance by means of the differentiae proper to it. The study of likeness is useful for inductive arguments, for arguments from a hypothesis, and for the rendering of definitions. For inductive arguments, because it is by induction from particulars in the case of similar things that we claim to induce the universal; for it is not easy to induct without knowing the similar cases. For arguments from a hypothesis, because it is a reputable view that as it holds in one of a set of similar cases, so it holds in the rest. So, whichever of the similar cases we are well supplied with arguments about, we will get prior agreement that as it holds in these, so it holds in the matter proposed; and having shown that, we will have shown the matter proposed as well, on the basis of the hypothesis, for by hypothesizing that as it holds in these cases, so it holds in the matter proposed, we have made our demonstration. For the rendering of definitions, because if we are able to see what is the same in each case, we will not be at a loss as to what genus we ought to place the matter proposed in when defining it,