Various Works

  philosopher does not care. Nay, he may possibly be even anxious to

  secure axioms as familiar and as near to the question in hand as

  possible: for these are the bases on which scientific reasonings are

  built up.

  The sources from which one's commonplace arguments should be drawn

  have already been described:' we have now to discuss the arrangement

  and formation of questions and first to distinguish the premisses,

  other than the necessary premisses, which have to be adopted. By

  necessary premisses are meant those through which the actual reasoning

  is constructed. Those which are secured other than these are of four

  kinds; they serve either inductively to secure the universal premiss

  being granted, or to lend weight to the argument, or to conceal the

  conclusion, or to render the argument more clear. Beside these there

  is no other premiss which need be secured: these are the ones

  whereby you should try to multiply and formulate your questions. Those

  which are used to conceal the conclusion serve a controversial purpose

  only; but inasmuch as an undertaking of this sort is always

  conducted against another person, we are obliged to employ them as

  well.

  The necessary premisses through which the reasoning is effected,

  ought not to be propounded directly in so many words. Rather one

  should soar as far aloof from them as possible. Thus if one desires to

  secure an admission that the knowledge of contraries is one, one

  should ask him to admit it not of contraries, but of opposites: for,

  if he grants this, one will then argue that the knowledge of

  contraries is also the same, seeing that contraries are opposites;

  if he does not, one should secure the admission by induction, by

  formulating a proposition to that effect in the case of some

  particular pair of contraries. For one must secure the necessary

  premisses either by reasoning or by induction, or else partly by one

  and partly by the other, although any propositions which are too

  obvious to be denied may be formulated in so many words. This is

  because the coming conclusion is less easily discerned at the

  greater distance and in the process of induction, while at the same

  time, even if one cannot reach the required premisses in this way,

  it is still open to one to formulate them in so many words. The

  premisses, other than these, that were mentioned above, must be

  secured with a view to the latter. The way to employ them respectively

  is as follows: Induction should proceed from individual cases to the

  universal and from the known to the unknown; and the objects of

  perception are better known, to most people if not invariably.

  Concealment of one's plan is obtained by securing through

  prosyllogisms the premisses through which the proof of the original

  proposition is going to be constructed-and as many of them as

  possible. This is likely to be effected by making syllogisms to

  prove not only the necessary premisses but also some of those which

  are required to establish them. Moreover, do not state the conclusions

  of these premisses but draw them later one after another; for this

  is likely to keep the answerer at the greatest possible distance

  from the original proposition. Speaking generally, a man who desires

  to get information by a concealed method should so put his questions

  that when he has put his whole argument and has stated the conclusion,

  people still ask 'Well, but why is that?' This result will be

  secured best of all by the method above described: for if one states

  only the final conclusion, it is unclear how it comes about; for the

  answerer does not foresee on what grounds it is based, because the

  previous syllogisms have not been made articulate to him: while the

  final syllogism, showing the conclusion, is likely to be kept least

  articulate if we lay down not the secured propositions on which it

  is based, but only the grounds on which we reason to them.

  It is a useful rule, too, not to secure the admissions claimed as

  the bases of the syllogisms in their proper order, but alternately

  those that conduce to one conclusion and those that conduce to

  another; for, if those which go together are set side by side, the

  conclusion that will result from them is more obvious in advance.

  One should also, wherever possible, secure the universal premiss

  by a definition relating not to the precise terms themselves but to

  their co-ordinates; for people deceive themselves, whenever the

  definition is taken in regard to a co-ordinate, into thinking that

  they are not making the admission universally. An instance would be,

  supposing one had to secure the admission that the angry man desires

  vengeance on account of an apparent slight, and were to secure this,

  that 'anger' is a desire for vengeance on account of an apparent

  slight: for, clearly, if this were secured, we should have universally

  what we intend. If, on the other hand, people formulate propositions

  relating to the actual terms themselves, they often find that the

  answerer refuses to grant them because on the actual term itself he is

  readier with his objection, e.g. that the 'angry man' does not

  desire vengeance, because we become angry with our parents, but we

  do not desire vengeance on them. Very likely the objection is not

  valid; for upon some people it is vengeance enough to cause them

  pain and make them sorry; but still it gives a certain plausibility

  and air of reasonableness to the denial of the proposition. In the

  case, however, of the definition of 'anger' it is not so easy to

  find an objection.

  Moreover, formulate your proposition as though you did so not for

  its own sake, but in order to get at something else: for people are

  shy of granting what an opponent's case really requires. Speaking

  generally, a questioner should leave it as far as possible doubtful

  whether he wishes to secure an admission of his proposition or of

  its opposite: for if it be uncertain what their opponent's argument

  requires, people are more ready to say what they themselves think.

  Moreover, try to secure admissions by means of likeness: for such

  admissions are plausible, and the universal involved is less patent;

  e.g. make the other person admit that as knowledge and ignorance of

  contraries is the same, so too perception of contraries is the same;

  or vice versa, that since the perception is the same, so is the

  knowledge also. This argument resembles induction, but is not the same

  thing; for in induction it is the universal whose admission is secured

  from the particulars, whereas in arguments from likeness, what is

  secured is not the universal under which all the like cases fall.

  It is a good rule also, occasionally to bring an objection against

  oneself: for answerers are put off their guard against those who

  appear to be arguing impartially. It is useful too, to add that 'So

  and so is generally held or commonly said'; for people are shy of

  upsetting the received opinion unless they have some positive

  objection to urge: and at the same time they are cautious about


  upsetting such things because they themselves too find them useful.

  Moreover, do not be insistent, even though you really require the

  point: for insistence always arouses the more opposition. Further,

  formulate your premiss as though it were a mere illustration: for

  people admit the more readily a proposition made to serve some other

  purpose, and not required on its own account. Moreover, do not

  formulate the very proposition you need to secure, but rather

  something from which that necessarily follows: for people are more

  willing to admit the latter, because it is not so clear from this what

  the result will be, and if the one has been secured, the other has

  been secured also. Again, one should put last the point which one most

  wishes to have conceded; for people are specially inclined to deny the

  first questions put to them, because most people in asking questions

  put first the points which they are most eager to secure. On the other

  hand, in dealing with some people propositions of this sort should

  be put forward first: for ill-tempered men admit most readily what

  comes first, unless the conclusion that will result actually stares

  them in the face, while at the close of an argument they show their

  ill-temper. Likewise also with those who consider themselves smart

  at answering: for when they have admitted most of what you want they

  finally talk clap-trap to the effect that the conclusion does not

  follow from their admissions: yet they say 'Yes' readily, confident in

  their own character, and imagining that they cannot suffer any

  reverse. Moreover, it is well to expand the argument and insert things

  that it does not require at all, as do those who draw false

  geometrical figures: for in the multitude of details the whereabouts

  of the fallacy is obscured. For this reason also a questioner

  sometimes evades observation as he adds in a corner what, if he

  formulated it by itself, would not be granted.

  For concealment, then, the rules which should be followed are the

  above. Ornament is attained by induction and distinction of things

  closely akin. What sort of process induction is obvious: as for

  distinction, an instance of the kind of thing meant is the distinction

  of one form of knowledge as better than another by being either more

  accurate, or concerned with better objects; or the distinction of

  sciences into speculative, practical, and productive. For everything

  of this kind lends additional ornament to the argument, though there

  is no necessity to say them, so far as the conclusion goes.

  For clearness, examples and comparisons should be adduced, and let

  the illustrations be relevant and drawn from things that we know, as

  in Homer and not as in Choerilus; for then the proposition is likely

  to become clearer.

  2

  In dialectics, syllogism should be employed in reasoning against

  dialecticians rather than against the crowd: induction, on the other

  hand, is most useful against the crowd. This point has been treated

  previously as well.' In induction, it is possible in some cases to ask

  the question in its universal form, but in others this is not easy,

  because there is no established general term that covers all the

  resemblances: in this case, when people need to secure the

  universal, they use the phrase 'in all cases of this sort'. But it

  is one of the very hardest things to distinguish which of the things

  adduced are 'of this sort', and which are not: and in this connexion

  people often throw dust in each others' eyes in their discussion,

  the one party asserting the likeness of things that are not alike, and

  the other disputing the likeness of things that are. One ought,

  therefore, to try oneself to coin a word to cover all things of the

  given sort, so as to leave no opportunity either to the answerer to

  dispute, and say that the thing advanced does not answer to a like

  description, or to the questioner to suggest falsely that it does

  answer to a like description, for many things appear to answer to like

  descriptions that do not really do so.

  If one has made an induction on the strength of several cases and

  yet the answerer refuses to grant the universal proposition, then it

  is fair to demand his objection. But until one has oneself stated in

  what cases it is so, it is not fair to demand that he shall say in

  what cases it is not so: for one should make the induction first,

  and then demand the objection. One ought, moreover, to claim that

  the objections should not be brought in reference to the actual

  subject of the proposition, unless that subject happen to be the one

  and only thing of the kind, as for instance two is the one prime

  number among the even numbers: for, unless he can say that this

  subject is unique of its kind, the objector ought to make his

  objection in regard to some other. People sometimes object to a

  universal proposition, and bring their objection not in regard to

  the thing itself, but in regard to some homonym of it: thus they argue

  that a man can very well have a colour or a foot or a hand other

  than his own, for a painter may have a colour that is not his own, and

  a cook may have a foot that is not his own. To meet them, therefore,

  you should draw the distinction before putting your question in such

  cases: for so long as the ambiguity remains undetected, so long will

  the objection to the proposition be deemed valid. If, however, he

  checks the series of questions by an objection in regard not to some

  homonym, but to the actual thing asserted, the questioner should

  withdraw the point objected to, and form the remainder into a

  universal proposition, until he secures what he requires; e.g. in

  the case of forgetfulness and having forgotten: for people refuse to

  admit that the man who has lost his knowledge of a thing has forgotten

  it, because if the thing alters, he has lost knowledge of it, but he

  has not forgotten it. Accordingly the thing to do is to withdraw the

  part objected to, and assert the remainder, e.g. that if a person have

  lost knowledge of a thing while it still remains, he then has

  forgotten it. One should similarly treat those who object to the

  statement that 'the greater the good, the greater the evil that is its

  opposite': for they allege that health, which is a less good thing

  than vigour, has a greater evil as its opposite: for disease is a

  greater evil than debility. In this case too, therefore, we have to

  withdraw the point objected to; for when it has been withdrawn, the

  man is more likely to admit the proposition, e.g. that 'the greater

  good has the greater evil as its opposite, unless the one good

  involves the other as well', as vigour involves health. This should be

  done not only when he formulates an objection, but also if, without so

  doing, he refuses to admit the point because he foresees something

  of the kind: for if the point objected to be withdrawn, he will be

  forced to admit the proposition because he cannot foresee in the

  rest of it any case where it does not hold true: if he refuse to admit

  it, then w
hen asked for an objection he certainly will be unable to

  render one. Propositions that are partly false and partly true are

  of this type: for in the case of these it is possible by withdrawing a

  part to leave the rest true. If, however, you formulate the

  proposition on the strength of many cases and he has no objection to

  bring, you may claim that he shall admit it: for a premiss is valid in

  dialectics which thus holds in several instances and to which no

  objection is forthcoming.

  Whenever it is possible to reason to the same conclusion either

  through or without a reduction per impossibile, if one is

  demonstrating and not arguing dialectically it makes no difference

  which method of reasoning be adopted, but in argument with another

  reasoning per impossibile should be avoided. For where one has

  reasoned without the reduction per impossibile, no dispute can

  arise; if, on the other hand, one does reason to an impossible

  conclusion, unless its falsehood is too plainly manifest, people

  deny that it is impossible, so that the questioners do not get what

  they want.

  One should put forward all propositions that hold true of several

  cases, and to which either no objection whatever appears or at least

  not any on the surface: for when people cannot see any case in which

  it is not so, they admit it for true.

  The conclusion should not be put in the form of a question; if it

  be, and the man shakes his head, it looks as if the reasoning had

  failed. For often, even if it be not put as a question but advanced as

  a consequence, people deny it, and then those who do not see that it

  follows upon the previous admissions do not realize that those who

  deny it have been refuted: when, then, the one man merely asks it as a

  question without even saying that it so follows, and the other

  denies it, it looks altogether as if the reasoning had failed.

  Not every universal question can form a dialectical proposition as

  ordinarily understood, e.g. 'What is man?' or 'How many meanings has

  "the good"?' For a dialectical premiss must be of a form to which it

  is possible to reply 'Yes' or 'No', whereas to the aforesaid it is not