Aristotle · a new plain-English translation from the original language
1| for once it is shown that every virtue is knowledge, or that none is, or that some other virtue is too, for example justice, or that practical wisdom itself is not knowledge, the thesis in question will have been demolished. It is also useful to look at particular cases, those in which something is said to hold or not to hold, just as with universal problems. Further, one must look among the genera, dividing according to species down to the individuals, as has been said before; for whether the attribute appears to hold of all or of none, after bringing forward many instances one may demand that the opponent agree to the universal, or bring an objection stating in what case it is not so. Further, in cases where the accidental attribute can be marked off either by species or by number, one should examine whether none of these apply, for example that time does not move and is not motion, by counting up how many species of motion there are; for if none of these belongs to time, it is clear that it does not move and is not motion. Likewise also that the soul is not a number, by dividing that every number is either odd or even; for if the soul is neither odd nor even, it is clear that it is not a number. This, then, is how one must proceed against the accidental attribute, by means of such arguments.
2| Book 4. After this we must examine the topics relating to genus and property. These are the elements of the topics relating to definitions; but the debaters who practice dialectic rarely conduct inquiries about these very things. If, then, a genus is laid down for one of the things that are, first one must look among all the things akin to what has been stated, to see whether it fails to be predicated of any of them, just as with the accidental attribute, for example, if the good is laid down as the genus of pleasure, whether some pleasure is not good; for if this is so, it is clear that the good is not the genus of pleasure, since a genus is predicated of everything that falls under the same species. Next, whether it is predicated not in the what-it-is but as an accident, as white is predicated of snow, or that which is moved by itself of soul. For snow is not precisely white, and so white is not the genus of snow, nor is soul precisely that-which-is-moved; rather, being moved is an accident of it, just as it is often an accident of an animal to walk and to be walking. Further, that-which-is-moved seems to signify not what a thing is, but rather that it does or undergoes something. Likewise also white; for it does not signify what snow is, but rather indicates some quality of it. So neither of these is predicated in the what-it-is.
3| One should look especially at the definition of an accident, to see whether it fits the genus stated, as with the examples just given. For it is possible for a thing both to move itself and not to, and likewise both to be white and not to be. So neither of these is a genus but an accident, since we said an accident is what can hold of a thing and also not hold of it. Further, whether the genus and the species are not in the same division, but the one is a substance and the other a quality, or the one is a relative and the other a quality, for example snow and swan are substances, while white is not a substance but a quality, so that white is not the genus of snow nor of swan. Again, knowledge is among relatives, while good and fine are qualities, so that good or fine is not the genus of knowledge; for the genera of relatives must themselves also be relatives, as in the case of the double; for the manifold too, being the genus of the double, is itself also among relatives. To put it generally, the genus must fall under the same division as the species; for if the species is a substance, so must the genus be, and if the species is a quality, so too the genus must be a quality, for example if white is a quality, so is color.
4| Likewise also with the other cases. Again, whether it is necessary or possible that the genus, of the thing placed within it, should partake of the genus. The definition of partaking is admitting the account of the thing partaken of. It is clear, then, that species partake of genera, but genera do not partake of species; for the species admits the account of the genus, but the genus does not admit the account of the species. One should examine, then, whether the assigned genus partakes, or can partake, of the species, for example if someone assigned some genus of being or of the one; for it will follow that the genus partakes of the species, since being and the one are predicated of all the things that are, and so is their account. Further, whether the assigned species is truly predicated of something of which the genus is not, for example if being or the knowable were laid down as the genus of the opinable. For the opinable will be predicated of what is not; for many things that are not are objects of opinion. But that being or the knowable is not predicated of what is not is clear. So being is not the genus of the opinable, nor is the knowable; for of whatever the species is predicated, the genus must also be predicated.
5| Again, see whether the thing placed in the genus can share in none of the species; for it is impossible for something to share in the genus while sharing in none of the species, unless it is itself one of the species arising from the first division — for these share in the genus alone. If, then, motion is posited as the genus of pleasure, we must examine whether pleasure is neither locomotion nor alteration nor any of the other kinds of motion that have been set out; for clearly, if it shares in none of the species, it plainly does not share in the genus either, since whatever shares in the genus must also share in some one of the species. So pleasure would not be a species of motion, nor would it be one of the individuals [nor] falling under the species of motion — for individuals too share in both the genus and the species, just as a particular man shares in both man and animal.
6| Again, see whether what has been placed in the genus is said more widely than the genus, as the opinable is said more widely than being; for both being and not-being are opinable, so that the opinable could not be a species of being — for the genus is always said more widely than the species. Again, see whether the species and the genus are said of equal extension, as would happen if, among things that follow everything, one were posited as the species and the other as the genus, for example being and the one; for being and the one belong to everything, so neither is a genus of the other, since they are said of equal extension. The same holds if the primary and the starting point were posited as reciprocal to each other; for the starting point is primary and the primary is a starting point, so that either the two terms named are the same, or neither is a genus of the other. A general rule for all such cases is this: the genus must be said more widely than the species and the differentia, for the differentia too is said more narrowly than the genus. One should also see, for any of the things that are undifferentiated in species, whether the stated genus fails to belong to it, or would seem not to — when constructing an argument, whether it does belong to one of them. For the genus is the same for all things undifferentiated in species. If, then, it is shown for one, it is clear it holds for all; and if it fails for one, it is clear it fails for none — as when someone, positing indivisible lines, says their genus is the indivisible. For the stated genus does not belong to lines that admit of division, and these are undifferentiated in species from the others; for all straight lines are undifferentiated from one another in species.
7| One should also examine whether there is some other genus of the given species that neither includes the assigned genus nor falls under it — for example, if someone posited knowledge as the genus of justice. For virtue too is a genus, and neither of the two genera includes the other, so knowledge could not be the genus of justice; for it seems that, whenever one species falls under two genera, the one genus is included by the other. But this raises a difficulty in some cases. For some think that practical wisdom is both a virtue and a knowledge, and that neither genus is included by the other — though not everyone concedes that practical wisdom is a knowledge. Still, even if one were to grant that this claim is true, it would seem necessary that genera of the same thing should fall under one another or under the same thing, as indeed happens with virtue and knowledge; for both fall under the same genus, since each of them is a state and a disposition. One must therefore examine whether neither of these applies to the assigned genus. For if the genera neither fall under one another nor both fall under the same thing, the assigned genus could not be correct.
8| One must also examine the genus of the assigned genus, and so on always with the genus above it, to see whether all such genera are predicated of the species, and whether they are predicated in the what-it-is; for all the higher genera must be predicated of the species in the what-it-is. If, then, there is any discord at some point, clearly the assigned term is not a genus. Again, see whether the genus shares in the species — either the genus itself or one of the genera above it; for nothing lower shares in what is above it. In demolishing an argument, then, one must proceed as has been said; in constructing one, if it is agreed that the stated genus belongs to the species, but it is disputed whether it belongs as a genus, it is enough to show that one of the higher genera is predicated of the species in the what-it-is. For when one thing is predicated in the what-it-is, all the things both above and below it, if they are predicated of the species at all, will be predicated in the what-it-is; so the assigned genus too will be predicated in the what-it-is. That when one thing is predicated in the what-it-is, all the rest, if predicated at all, will be predicated in the what-it-is, must be grasped by induction. But if it is disputed simply whether the assigned genus belongs at all, it is not enough to show that one of the higher genera is predicated of the species in the what-it-is.
9| For example, if someone assigned locomotion as the genus of walking, it is not enough to show that walking is a motion in order to show that it is a locomotion, since there are other motions as well; rather one must further show that walking shares in none of the species under the same division except locomotion. For whatever shares in the genus must also share in some one of the species arising from the first division. If, then, walking shares neither in growth nor in diminution nor in the other motions, clearly it would share in locomotion, so that locomotion would be the genus of walking. Again, in cases where the species posited as a genus is itself predicated, see whether the assigned genus too is predicated in the what-it-is of the very things of which the species is predicated, and likewise whether all the genera above the genus are as well. For if there is discord anywhere, clearly the assigned term is not a genus; for if it were the genus, then it and everything above it would be predicated in the what-it-is of just those things of which the species too is predicated in the what-it-is. In demolishing an argument, then, it is useful if the genus is not predicated in the what-it-is of the things of which the species is predicated;
10| in constructing one, it is useful if it is so predicated. For it will follow that the genus and the species are predicated of the same thing in the what-it-is, so that the same thing falls under two genera. The genera must therefore fall under one another. If, then, it is shown that the genus we wish to construct does not fall under the species, clearly the species would fall under it, so that it will have been shown that this is the genus. One should also examine the definitions of the genera, to see whether they fit both the assigned species and the things that share in the species. For the definitions of the genera must be predicated of the species and of the things that share in the species; if, then, there is discord anywhere, clearly the assigned term is not a genus. Again, see whether he has assigned the differentia as the genus — for example, if he made the immortal the genus of god. For the immortal is a differentia of animal, since of animals some are mortal and some immortal. It is clear, then, that he has erred; for a differentia is never the genus of anything. That this is true is clear: for no differentia signifies a what-it-is, but rather a certain quality, as do the terms footed and two-footed. And likewise if he placed the differentia in the position of the genus, as the odd, which is precisely a number.
11| For "odd" is a differentia of number, not a species. Nor does the differentia seem to participate in the genus; for everything that participates in the genus is either a species or an individual, but the differentia is neither a species nor an individual. It is clear, then, that the differentia does not participate in the genus, so that "odd" would not be a species either, but a differentia, since it does not participate in the genus. Again, check if he has placed the genus into the species — for instance, contact as being cohesion, or mixture as being blending, or as Plato defines locomotion as motion in respect of place. For it is not necessary that contact be cohesion; rather, the reverse, that cohesion be contact — for not everything that touches is held together, but what is held together touches. Similarly in the remaining cases too; for not every mixture is a blending (for the mixture of dry things is not a blending), nor is every change in place locomotion — for walking does not seem to be locomotion; for "locomotion" is used, roughly speaking, of things that change place involuntarily, as happens with inanimate things. And it is clear that in the definitions given, the species is being said more widely than the genus, when the reverse ought to happen.
12| Again, check if he has placed the differentia into the species, for instance, "immortal" as being god. For it will follow that the species is said either equally widely or more widely than the genus; for the differentia is always said either equally widely or more widely than the species. Again, check if he has placed the genus into the differentia, for instance, color as being what discriminates, or number as being odd. And check if he has stated the genus as a differentia; for it is possible for someone to put forward even such a position as this, for instance, mixture as the differentia of blending, or change of place as the differentia of locomotion. All such cases should be examined by the same means; for the topics overlap — for the genus must be said more widely than the differentia, and must not participate in the differentia. But if it is rendered in this way, neither of the things mentioned can hold; for it will be said of less, and the genus will participate in the differentia. Again, if none of the differentiae of the genus is predicated of the species that has been rendered, neither will the genus be predicated, as for instance neither "odd" nor "even" is predicated of soul, so that "number" is not predicated of it either. Again, check if the species is prior in nature and its removal removes the genus along with it; for the contrary seems to hold. Again, if it is possible for the stated genus or the differentia to be absent from the thing, as for instance for soul to lack being in motion, or for opinion to lack being true or false, neither of the things mentioned would be a genus nor a differentia of it;
13| for the genus and the differentia seem to accompany the thing as long as the species exists. One should also examine whether what is placed in the genus participates in something contrary to the genus, or whether it is possible for it to participate in it; for the same thing will then participate in contraries at once, since the genus never leaves it, while it also participates, or can participate, in the contrary of the genus. Again, check if the species shares in something that cannot possibly belong at all to the things falling under the genus. For instance, if soul shares in life, but no number can possibly live, then soul would not be a species of number. One should also examine whether the species is used homonymously with the genus, employing the elements stated for detecting homonymy; for genus and species must be used synonymously. And since every genus has several species, one should examine whether it is not possible for there to be another species of the stated genus; for if there is not, it is clear that what has been stated would not be a genus at all. One should also examine whether he has rendered as the genus something said metaphorically, for instance temperance as harmony; for every genus is predicated of its species in the strict sense, but "harmony" is predicated of temperance not in the strict sense but metaphorically — for all harmony is in sounds. Again, if there is something contrary to the species, one should examine it. And the examination proceeds in several ways: first, whether the contrary too is in the same genus, when there is nothing contrary to the genus;
14| for the contraries must be in the same genus, if there is nothing contrary to the genus. But if there is something contrary to the genus, one should examine whether the contrary of the species is in the contrary genus; for it is necessary for the contrary to be in the contrary, if there is indeed something contrary to the genus. Each of these points is clear by induction. Again, check whether the thing contrary to the species is in no genus at all, but is itself a genus, for instance the good; for if this is not in a genus, neither will its contrary be in a genus, but will itself be a genus, as happens with the good and the bad — for neither of these is in a genus, but each of them is itself a genus. Again, check whether, if both the genus and the species are contrary to something, there is an intermediate for the one but not for the other. For if there is an intermediate for the genera, there is one for the species too, and if for the species, then for the genera too, as in the case of virtue and vice, and justice and injustice; for there is an intermediate for each of these pairs. An objection to this is that there is no intermediate between health and disease, though there is between bad and good. Or check whether, though there is something between both pairs, both for the species and for the genera, it is not in the same way, but for the one it is by negation, for the other as an underlying subject.
15| For it is the accepted view that it should be the same for both, as in the case of virtue and vice, justice and injustice; for in both cases the intermediates are by negation. Again, when there is nothing contrary to the genus, one should examine not only whether the contrary is in the same genus, but also the intermediate; for whatever genus contains the extremes contains the intermediates too, as in the case of white and black; for color is the genus both of these and of all the intermediate colors. An objection is that deficiency and excess are in the same genus (for both are in the bad), but the moderate amount, though intermediate between them, is not in the bad but in the good. One should also examine whether the genus is contrary to something while the species is contrary to nothing. For if the genus is contrary to something, the species is too, as virtue is contrary to vice and justice to injustice. Examining likewise in other cases too, this would seem to be evident. An objection arises in the case of health and disease; for, taken generally, all health is contrary to disease, but a particular disease, being a species of disease, is contrary to nothing, for instance fever, ophthalmia, and each of the others.
16| In destructive argument, then, one must examine in this many ways; for if what has been stated does not hold, it is clear that what was rendered is not the genus. In constructive argument, in three ways: first, whether the thing contrary to the species is in the stated genus, when there is nothing contrary to the genus; for if the contrary is in this genus, it is clear that the thing under discussion is too. Again, check whether the intermediate is in the stated genus; for whatever genus contains the intermediate contains the extremes too. Again, if there is something contrary to the genus, examine whether the contrary of the species is also in the contrary genus; for if it is, it is clear that the thing under discussion is in the genus under discussion. Again, in the case of the grammatical inflections and the co-ordinate terms, check whether they follow correspondingly, both in destructive and in constructive argument. For a predicate belongs, or fails to belong, to one and to all together, as for instance if justice is a kind of knowledge, then also "justly" is "knowledgeably," and the just person is knowledgeable; but if any of these does not hold, then none of the rest holds either.
17| Again, consider things that stand in a similar relation to one another—for instance, the pleasant stands to pleasure as the beneficial stands to the good, since each is productive of the other. If, then, pleasure is exactly the good, the pleasant too will be exactly the beneficial; for clearly it would be productive of a good, since pleasure is a good. And likewise in cases of coming-to-be and destruction: e.g. if building is being active, then having built is having been active; and if learning is being reminded, then having learned is having been reminded; and if being dissolved is being destroyed, then having been dissolved is having been destroyed, and dissolution is a kind of destruction. And likewise too with things productive of coming-to-be and of destruction, and with capacities and their exercises; and in general, whatever the similarity, one must examine it both in demolishing and in establishing, just as we said in the case of coming-to-be and destruction. For if what is destructive is dissolutive, then to be destroyed is to be dissolved; and if what is generative is productive, then to come to be is to be made, and coming-to-be is a kind of making. Likewise with capacities and their exercises: if a capacity is a disposition, then to be capable is to be disposed; and if the exercise of something is an actuality, then to exercise it is to be active, and to have exercised it is to have been active. If the opposite in species is a privation, there are two ways to demolish the proposed genus: first, if the opposite is not in the genus assigned—
18| for either the privation is, without qualification, in no genus at all the same as it, or it is not in the lowest genus—e.g. if sight is in perception as its lowest genus, blindness will not be a perception. Second, if the privation is opposed both in genus and in species, but the opposite is not in the opposite genus, then neither would the thing assigned be in the genus assigned. In demolishing, then, one must proceed as has been said; in establishing, only in one way: for if the opposite is in the opposite, the thing proposed would also be in the genus proposed—e.g. if blindness is a kind of non-perception, sight is a kind of perception.
19| Again, one must examine the negations taken in reverse, as was said in the case of the accident—e.g. if the pleasant is exactly the good, then what is not good is not pleasant. For if this were not so, there would be something not good that is pleasant. But that is impossible, if the good is the genus of the pleasant; for nothing can be pleasant of which the genus is not predicated, among things whose genus is predicated neither of them nor of any of their species. And in establishing one must examine it the same way: if what is not good is not pleasant, then the pleasant is good, so that the good is the genus of the pleasant. And if the species is a relative, one must see whether the genus too is a relative; for if the species is among the relatives, so is the genus—as with the double and the multiple, each of which is a relative. But if the genus is a relative, it is not necessary that the species be one too; for knowledge is a relative, but grammar is not. Or perhaps even the earlier statement would not seem true: for virtue is exactly the fine and exactly the good, and virtue is a relative, while the good and the fine are not relatives but qualities.
20| Again, one must see whether the species is not said in relation to the same thing both in its own right and according to the genus—e.g. if the double is said to be double of the half, then the multiple too must be said to be multiple of the half; if not, the multiple would not be the genus of the double. Further, one must see whether it is not said in relation to the same thing both according to the genus and according to all the genera above that genus. For if the double is a multiple of the half, then what exceeds will also be said of the half, and in general it will be said in relation to the half according to all the higher genera. An objection: it is not necessary that a thing be said in relation to the same thing both in its own right and according to its genus; for knowledge is said in relation to the knowable, but state and disposition are said not in relation to the knowable but in relation to the soul. Again, one must see whether the genus and the species are said alike with respect to their cases—whether both take 'to something' or 'of something,' or however many other ways there are. For as the species is said, so should the genus be said, as with the double and the genera above it: both the double and the multiple are said 'of something.' Likewise with knowledge: it too, and its genera—disposition and state—are said 'of something.' An objection: that in some cases it is not so—
21| for the different and the contrary are said 'of something,' while the other, being the genus of these, is said not 'to something' but 'of something,' since it is called other than something. Again, one must see whether relatives said alike with respect to their cases fail to convert alike, as with the double and the multiple. For each of these is itself said 'of something,' and is said the same way when converted, since the half too and the fractional part are said 'of something.' Likewise with knowledge and supposition: these too are said 'of something,' and convert alike, since the knowable and the supposable are each said 'to something.' If, then, in some cases they do not convert alike, it is clear that the one is not the genus of the other. Again, one must see whether the species and the genus are not said in relation to equal numbers of things; for each is thought to be said in a similar and equal number of ways, as with gift and giving. For a gift is said to be of something or to someone, and giving too is of something and to someone; and giving is the genus of gift, since a gift is a giving that is not given back. But in some cases it does not turn out that they are said in relation to equal numbers: for the double is double of something, while what exceeds, and the greater, are of something and to something—
22| for everything that exceeds, and the greater, exceeds something and exceeds it by some amount. So the terms mentioned are not the genus of the double, since it is not said in relation to equal numbers of things as the double is in species. Or else it is not universally true that the species and the genus must be said in relation to equal numbers of things. One must also see whether the opposite of the term is the genus of the opposite of the species—e.g. whether the multiple is the genus of the double, and the fractional part of the half; for the opposite ought to be the genus of the opposite. If, then, someone were to posit that knowledge is exactly perception, the knowable too would have to be exactly the perceptible. But this is not so, for not everything knowable is perceptible, since some intelligible things too are knowable. So the perceptible is not the genus of the knowable; and if this is not so, then neither is perception the genus of knowledge. Now, of the things called relatives, some are necessarily in, or concerned with, the very things in relation to which they happen to be spoken of—e.g. disposition and state and due proportion, for these cannot possibly belong in anything else; the things mentioned are in those very things in relation to which they are spoken of. Others need not be in the things in relation to which they happen to be spoken of, though they may be—e.g. if the soul is knowable—
23| for nothing prevents the soul from having knowledge of itself, though this is not necessary, since this same knowledge can also be in something else. Others simply cannot be in the things in relation to which they happen to be spoken of at all—e.g. the contrary cannot be in the contrary, nor knowledge in the knowable, unless the knowable happens to be soul or man. One must therefore examine whether someone has placed a thing of this non-corresponding sort into such a genus—e.g. if someone said that memory is a persistence of knowledge. For every persistence is in, and concerned with, the thing that persists, so that the persistence of knowledge too would be in the knowledge. Memory, then, would be in knowledge, since it is a persistence of knowledge; but this is not possible, for all memory is in the soul. The topic just mentioned is also common to the topic of accident; for it makes no difference whether one calls persistence the genus of memory or says that this belongs to memory as an accident, since if memory is in any way a persistence of knowledge, the same argument will fit the case.
24| Again, see whether he has put the state under the activity, or the activity under the state — for example, calling perception 'motion through the body': for perception is a state, while motion is an activity. Likewise, see whether he called memory 'a retentive state of a supposition': for no memory is a state, but rather an activity. Those also go wrong who place a state under the capacity that follows on it — for example, calling gentleness 'mastery over anger,' and courage and justice 'mastery over fears and gains.' For the man called courageous or gentle is the one unaffected by these things, while the man called 'in control' is one who is affected but not led by it. Perhaps, then, a capacity of this sort does follow on each — such that, if he were affected, he would not be led but would master it; but this is not what it is to be courageous or gentle, but rather to be wholly unaffected by such things at all. Sometimes, too, people set down as the genus whatever happens to accompany the thing in some way — for example, distress as the genus of anger, and supposition as the genus of conviction. For both of the things mentioned do in some way accompany the species assigned to them, but neither of them is their genus. For the angry man feels distress, the distress arising in him first; for it is not anger that is the cause of the distress, but distress that is the cause of anger, so that anger simply is not distress.
25| On the same grounds, conviction is not supposition either: for it is possible to hold the same supposition without believing it. But that would not be possible, if conviction were a species of supposition: for the same thing cannot remain the same once it has changed entirely out of its species, just as the same animal cannot be a human being at one time and not at another. And if someone says that whoever holds a supposition is necessarily also convinced, then supposition and conviction will be said to hold of exactly the same things, so that not even on this view could it be a genus: for a genus must be predicated more widely. See also whether both are naturally found in the same thing: for wherever the species is, the genus is there too — for example, wherever white is, color is there too, and wherever grammar is, knowledge is there too. If, then, someone calls shame a species of fear, or anger a species of distress, it will not turn out that the species and the genus belong to the same thing. For shame is in the rational part, while fear is in the spirited part; and distress is in the appetitive part (since pleasure too is in this part), while anger is in the spirited part. So the genera assigned are not genera at all, since they do not naturally occur in the same part as their supposed species.
26| Likewise, if friendship is in the appetitive part, it could not be a kind of wish: for all wish is in the rational part. This commonplace is useful for accidents too: for an accident is in the same thing as that to which it belongs, so that if it does not appear to be in the same thing, clearly it is not an accident of it. Again, see whether the species partakes of the stated genus only in some respect: for a genus is not thought to be partaken of only in some respect — a human being is not an animal only in some respect, nor is grammar knowledge only in some respect. And likewise in other cases. So examine whether in some instances the genus is partaken of only in some respect — for example, if 'animal' is said to be exactly what is perceptible or visible. For an animal is perceptible or visible only in some respect: it is perceptible and visible in respect of its body, but not in respect of its soul; so 'visible' and 'perceptible' could not be the genus of 'animal.' Sometimes people also fail to notice that they are putting the whole under the part — for example, defining animal as an ensouled body. But a part is in no way predicated of the whole, so body could not be the genus of animal, since it is a part of it.
27| See also whether he has put something blameworthy or to be avoided under a capacity or 'the capable' — for example, defining the sophist, or the slanderer, or the thief as one capable of taking what belongs to others, or capable of slandering, or of sophistry. For none of the men named is called such simply by being capable of doing these things: even god and the good man are capable of doing base things, yet they are not that sort of person; for all base people are so called in virtue of their choice. Further, every capacity is among the things worth having: for even the capacities for base things are worth having, which is why we say god and the good man have them too — since they are capable of doing base things. So capacity could not be the genus of anything blameworthy. Otherwise, it will follow that something blameworthy is worth having: for there will be some capacity that is blameworthy.
28| And see whether he has put something valued or worth having in its own right under a capacity, or 'the capable,' or 'the productive.' For every capacity, and everything capable or productive, is worth having for the sake of something else. Or whether he has put something that falls under two genera, or more, into only one of them. For some things cannot be put into a single genus — for example, the cheat and the slanderer. For neither the man who chooses to do it but is unable, nor the man who is able but does not choose to, is a slanderer or a cheat, but only the man who has both. So he must not be put into one genus, but into both of the ones mentioned. Further, sometimes people assign, in reverse, the genus as differentia and the differentia as genus — for example, defining astonishment as 'an excess of wonder' and conviction as 'an intensity of supposition.' But neither 'excess' nor 'intensity' is a genus — it is a differentia: for astonishment seems to be wonder that is excessive, and conviction seems to be supposition that is intense, so that wonder and supposition are the genus, while excess and intensity are the differentia. Further, if anyone assigns excess and intensity as the genus, inanimate things will be convinced and will be astonished.
29| For the intensity of a thing, and its excess, belong to that very thing of which they are the intensity and the excess. If, then, astonishment is an excess of wonder, astonishment will belong to wonder, so that wonder will be astonished. Likewise conviction will belong to supposition, if it is an intensity of supposition, so that supposition will be convinced. Further, it will follow for anyone who assigns things in this way that he calls intensity intense and excess excessive. For conviction is intense; if, then, conviction is an intensity, then intensity would be intense. Likewise, astonishment is excessive; if, then, astonishment is an excess, excess would be excessive. But neither of these is thought to be so, just as knowledge is not something known, nor is motion something moved. Sometimes people also go wrong by putting the affection under the genus of the thing affected — for example, all who say that immortality is 'life everlasting.' For immortality seems to be some affection or attribute of life. That what is being said is true would become clear if one granted that someone becomes immortal from being mortal: no one will say he takes on a different life, but rather that some attribute or affection comes to this very life he has. So life is not the genus of immortality. Again, see whether they say that the thing of which something is an affection is the genus of that affection — for example, defining wind as 'moving air.'
30| For wind is rather a motion of air: for the same air remains, whether it is moving or at rest. So wind is not air at all: for there would be wind even when the air was not moving, if the very air that was wind remains the same. And likewise in other such cases. But even if in this case one must grant that wind is air in motion, still this sort of account must not be accepted for all cases in which the genus does not hold true, but only for those in which the genus assigned is truly predicated. For in some cases it does not seem to hold true — for example, in the case of mud and snow. For they say snow is 'water frozen solid,' and mud is 'earth mixed with liquid'; but snow is not water, nor is mud earth, so neither of the things assigned could be the genus: for the genus must always hold true of the species. Likewise, wine is not 'water gone rotten,' as Empedocles says, 'water rotted in wood': for wine simply is not water.
31| Again, see whether the thing assigned is in general the genus of nothing at all; for clearly then it is not the genus of the thing in question either. Examine this from the fact that the things sharing in the assigned genus do not differ in species — for instance, white things: these do not differ from one another in species. But every genus has species that differ from one another, so that white would not be the genus of anything. Again, see whether he has stated as genus or as differentia something that follows everything. For there are several things that follow everything — for instance, being and unity are among the things that follow everything. If, then, he assigned being as the genus, clearly it would be the genus of everything, since it is predicated of them all; for a genus is predicated of nothing except its species. So unity too would then be a species of being. It follows, then, that of everything of which the genus is predicated the species is also predicated — since being and unity are predicated of absolutely everything — though the species ought to be predicated of fewer things than the genus. But if he stated as differentia something that follows everything, clearly the differentia will be said to hold of an equal or a greater number of things than the genus. For if the genus too is among the things that follow everything, it will hold of an equal number; but if the genus does not follow everything, the differentia would be said to hold of more things than it.
32| Again, see whether the assigned genus is said to be present in the species as in a subject, the way white is present in snow; for then clearly it would not be a genus, since a genus is said of the species as of a subject, not in it. Also examine whether the genus is not synonymous with the species; for a genus is predicated synonymously of all its species. Again, see when, there being a contrary both in species and in genus, he places the better of the two contraries under the worse genus; for it will follow that the remaining one is in the remaining one, since contraries fall under contrary genera, so that the better will be in the worse and the worse in the better — but the genus of the better is held to be itself better. Also see whether, though the same species stands alike in relation to both, he assigned it to the worse and not the better genus — for instance, calling soul motion, or a thing moved. For the same capacity seems alike to produce rest and to produce motion, so that if rest is the better, the genus ought to have been placed there. Again, argue from the more and the less: when refuting, see whether the genus admits of the more while the species admits it neither itself nor in what is said in accordance with it.
33| For instance, if virtue admits of the more, so do justice and the just man; for one man is called more just than another. If, then, the assigned genus admits of the more, but the species admits it neither itself nor in what is said according to it, the thing assigned would not be a genus. Again, see whether what is thought to be more so, or equally so, is not a genus — clearly then neither is the thing assigned. This topic is most useful for cases where several things appear to be predicated of the species in the 'what it is' and it has not been settled, nor can we say which of them is the genus — for instance, both pain and the supposition of being slighted seem to be predicated of anger in the 'what it is'; for the angry man both feels pain and supposes he is being slighted. The same examination applies also when comparing the species to something else: for if what seems to hold more, or equally, is not in the assigned genus, clearly the assigned species would not be in that genus either. In refuting, then, one should use the topic as has been said; but in establishing, if both the assigned genus and the species admit of the more, the topic is not useful;
34| for nothing prevents both admitting of the more and yet neither being the genus of the other. For both the fine and the white admit of the more, and neither is the genus of the other. But the comparison of genera and of species with one another is useful — for instance, if this and that are genera to the same degree, then if one is a genus, so is the other. Likewise too if one is less a genus and the other more — for instance, if capacity is more a genus of self-control than virtue is, and virtue is a genus, then so is capacity. The same things will fit to be said about the species too. For if this and that are alike species of the thing proposed, then if one is a species, so is the remaining one; and if the one thought to be less a species is a species, so is the one thought more so. Again, for establishing, one should examine whether, of the things of which the genus was assigned, it is predicated in the 'what it is,' there being not one assigned species but several and different ones; for then clearly it will be a genus. But if the assigned thing is one species, examine whether the genus is also predicated in the 'what it is' of other species; for then again it will turn out to be predicated of several different things. Since some think that the differentia too is predicated of species in the 'what it is,' one must separate the genus from the differentia by using the elements stated: first, that the genus is said of more things than the differentia; and next, that in giving the account of the 'what it is' the genus is a better fit to state than the differentia;
35| for the man who says 'animal' shows better what a human being is than the man who says 'footed.' And note that the differentia always signifies a quality of the genus, while the genus does not signify a quality of the differentia; for the man who says 'footed' states a certain quality of animal, while the man who says 'animal' does not state what quality 'footed' has. The differentia, then, is to be separated from the genus in this way. Since it seems that the musical, insofar as it is musical, is a kind of knowledge, and music is a kind of knowledge, and since, if the walking thing is moved by its walking, walking would be a kind of motion — in whatever genus you wish to establish something, examine it in the way stated: for instance, whether knowledge is a kind of conviction, by examining whether the man who knows, insofar as he knows, is convinced; for then clearly knowledge would be a kind of conviction. The same manner applies in the other such cases too. Again, since what always accompanies something without converting is hard to separate from being a genus — when this follows that in every case, but that does not follow this in every case, as calm follows windlessness, and the divisible follows number, but not conversely (for not everything divisible is a number, nor is rest calm) — one should oneself use what always follows as if it were a genus, when the other term does not convert; but when someone else proposes it, one should not concede it in every case.