Aristotle
Aristotle chose to connect fallacious reasoning with dishonest
purposes, and to announce it as employed exclusively by dishonest
debaters, he seems to have found satisfaction in describing it as
something which no honest man of ordinary understanding could accept
as true: the Sophist being thus presented not merely as a knave but
as a fool.
[Footnote 19: Topic. I. i. p. 100, b. 24, seq.]
[Footnote 20: Soph. El. i. p. 164, a. 23-b. 27. [Greek: to\n au)to\n
de\ tro/pon kai\ sullogismo\s kai\ e)/legchos o( me\n e)/stin, o( d'
ou)k e)/sti me/n, phai/netai de\ dia\ tê\n a)peiri/an; oi( ga\r
a)/peiroi ô(/sper a)\n a)pe/chontes po/r)r(ôthen theôrou=sin.]]
I think it a mistake on the part of Aristotle to treat the fallacies
incidental to the human intellect as if they were mere traps laid by
Sophists and litigants; and as if they would never show themselves,
assuming dialectical debate to be conducted entirely with a view to
its legitimate purposes of testing a thesis and following out
argumentative consequences. It is true that, if there are infirmities
incident to the human intellect, a dishonest disputant will be likely
to take advantage of them. So far it may be well to note his
presence. But the dishonest disputant does not originate these
infirmities: he finds them already existing, and manifested
undesignedly not merely in dialectical debate, but even in ordinary
discourse. It is the business of those who theorize on the
intellectual processes to specify and discriminate the Fallacies as
liabilities to intellectual error among mankind in general, honest or
dishonest, with a view to precaution against their occurrence, or
correction if they do occur; not to present them as inventions of a
class of professional cheats,[21] or as tares sown by the enemy in a
field where the natural growth would be nothing but pure wheat.
[Footnote 21: Soph. El. i. p. 165, a. 19, seq.]
In point of fact the actual classification of Fallacies given by
Aristotle is far sounder than his announcement would lead us to
expect. Though he entitles them Sophistical Refutations, describing
them as intentionally cultivated and exclusively practised by
professional Sophists for gain, or by unprincipled litigants for
victory, yet he recognises them as often very difficult of detection,
and as an essential portion of the theory of Dialectic generally.[22]
The various general heads under which he distributes them are each
characterized by intellectual or logical marks.
[Footnote 22: Ibid. xi. p. 172, b. 7.]
His first and most general observation is, that language is the usual
medium and instrument through which fallacies are operated.[23] Names
and propositions are of necessity limited in number; but things named
or nameable are innumerable; hence it happens inevitably that the
same name or the same proposition must have several different
meanings. Since we cannot talk of things except by means of their
names, the equivocation inseparable from these names is a constant
source of false conclusions.[24]
[Footnote 23: Ibid. i. p. 165, a. 5.]
[Footnote 24: Ibid. a. 10: [Greek: ta\ me\n ga\r o)no/mata
pepe/rantai kai\ to\ tô=n lo/gôn plê=thos, ta\ de\ pra/gmata to\n
a)rithmo\n a)/peira/ e)stin. a)nagkai=on ou)=n plei/ô to\n au)to\n
lo/gon kai\ tou)/noma to\ e(\n sêmai/nein.]]
In dialectical procedure, the Sophist and the litigious debater aim
at the accomplishment of five distinguishable ends:--(1) To refute,
or obtain the false appearance of refuting, the thesis; (2) To catch,
or appear to catch, the opponent in affirming something false or
contradictory; (3) Or in affirming something paradoxical; (4) Or in
uttering incorrect and ungrammatical speech; (5) Or in tautological
repetition. The first of these five ends is what the Sophist most
desires; where that cannot be had, then, as secondary purposes, the
succeeding four, in the order in which they are enumerated.[25]
[Footnote 25: Soph. El. iii. p. 165, b. 12-22.]
The syllogism whereby the Sophist appears to refute without really
refuting, is either faulty in form, or untrue in matter, or
irrelevant to the purpose. The Fallacies that he employs to bring
about this deceitful appearance of refutation are various, and may be
distributed first under two great divisions:--
I. _Fallaciæ Dictionis_.
II. _Fallaciæ Extra Dictionem_.
I. The first division--_Fallaciæ Dictionis_--includes all those cases
wherein, under the same terms or propositions, more than one meaning
is expressed. Six heads may be distinguished:--
1. Homonymy (Equivocation): where the double meaning resides in one
single term--noun or verb.
2. Amphiboly: where the double meaning resides, not in a single word
but, in a combination of words--proposition, phrase, or sentence.
3. Conjunction (hardly distinguishable from that immediately
preceding--Amphiboly).
4. Disjunction: where what is affirmed conjunctively is not true
disjunctively, or the reverse. (_E.g._, Five are two and three; but
you cannot say, Five are even and odd. The greater is equal and
something besides; but you cannot say, The greater is equal.)
5. Accentuation: where the same word differently accentuated has a
different meaning.
6. _Figura Dictionis_: where two words, from being analogous in form,
structure, or conjugation, are erroneously supposed to be analogous
in meaning also.[26]
[Footnote 26: Ibid. iv. p. 165, b. 23-p. 166, b. 19.]
Such are the six heads of _Fallaciæ Dictionis_--Fallacies or
Paralogisms arising from words as such, or something directly
appertaining to them.
II. Under the second division--Fallacies or Paralogisms _Extra
Dictionem_--there are seven heads:
1. _Fallacia Accidentis_.
2. _Fallacia a dicto Secundum Quid ad dictum Simpliciter_.
3. _Ignoratio Elenchi_.
4. _Fallacia Consequentis_
5. _Petitio Principii_.
6. _Non Causa pro Causâ_.
7. _Fallacia Plurium Interrogationum_.[27]
[Footnote 27: Soph. El. v. p. 166, b. 20-27.]
1. The first of these varieties, called _Fallacia Accidentis_, arises
when a syllogism is made to conclude that, because a given predicate
may be truly affirmed of a given subject, the same predicate may also
be truly affirmed respecting all the accidents of that subject: as
when Koriskus is denied to be a man, because he is not Sokrates, who
is a man; or is denied to be Koriskus, because he is a man, while a
man is not Koriskus.
In the title given to this general head of Fallacy,[28] we must
understand Accident, not in its special logical sense as opposed to
Essence, but in a far larger sense, including both Genus when
predicated separately from Differentia, and Differentia when
predicated separately from Genus; including, in fact, every thing
which is distinguishable from the subject in any way, and at the same
time predicable of it--every thing except the Definition, which
conjoins Genus and Differentia together, and is thus identical and
convertible with the _definitum_.
[Footnote 28: Ibid. b. 29: [Greek: oi( para\ to\ sumbebêko\s
paralogismoi/]. Every man is an animal; but, because a predicate is
true of the subject man, you cannot infer that the same predicate is
true of the subject animal. This title comprehends within its range
another, which is presently announced as distinct and
separate--_Fallacia Consequentis_.]
2. The second general variety arises when a proposition is affirmed
with qualification or limitation in the premisses, but is affirmed
without qualification, simply and absolutely, in the conclusion. The
Ethiopian is white in his teeth and black in his skin; therefore, he
is both white and not white--both white and black. In this example
the fallacy is obvious, and can hardly escape any one; but there are
many other cases in which the distinction is not so conspicuous, and
in which the respondent will hesitate whether he ought to grant or
refuse a question simply and absolutely.[29] One example given by
Aristotle deserves notice on its own account: _Non-Ens est
opinabile_, therefore _Non-Ens est_; or, again, _Ens non est homo_,
therefore, _Ens non est_. This is one among Aristotle's ways of
bringing to view what modern logicians describe as the double
function of the substantive verb--to serve as copula in predication,
and to predicate existence.[30] He regards the confusion between
these two functions as an example of the Fallacy now before us--of
passing _a dicto Secundum Quid ad dictum Simpliciter_.[31]
[Footnote 29: Ibid. b. 37, seq. [Greek: o(/tan to\ e)n me/rei
lego/menon ô(s a(plô=s ei)rême/non lêphthê=|--to\ de\ toiou=ton e)p'
e)ni/ôn me\n panti\ theôrê=sai r(a/|dion--e)p' e)ni/ôn de\ lantha/nei
polla/kis.]]
[Footnote 30: The same double or multiple meaning of _Est_ is
discriminated by Aristotle in the Metaphysica, but in a different
way--[Greek: to\ o)\n ô(s a)lêthe/s, kai\ to\ mê\ o)\n ô(s
pseu=dos]--[Greek: D].] vii. p. 1017, a. 31; E. iv. p. 1027, b. 18-36.
Bonitz (ad. Metaphys. Z. iv. p. 310) says:--"Quid quod etiam illud
_esse_ huc refert, quo non existentiam significamus, sed predicati
cum subjecto conjunctionem." Aristotle is even more precise than
modern logicians in analysing the different meanings of [Greek: to\
o)/n]: he distinguishes _four_ of them.]
[Footnote 31: Soph. El. v. p. 167, a. 1: [Greek: oi(=on ei) to\ mê\
o)/n e)sti doxasto/n, o(/ti to\ mê\ o)\n e)/stin; ou) ga\r tau)to\n
ei)=nai te/ ti kai\ ei)=nai a(plô=s.]
Compare Metaphys. Z. iv. p. 1030, a. 25, and De Interpretatione, p.
21, a. 25-34: [Greek: ô(/sper O(/mêro/s e)sti/ ti, oi(=on poiêtê/s;
a)=r' ou)=n kai\ e)/stin, ê) ou)/? kata\ sumbebêko\s ga\r
katêgorei=tai tou= O(mê/rou to\ e)/stin; o(/ti ga\r poiêtê/s e)stin,
a)ll' ou) kath' au(to/, katêgorei=tai kata\ tou= O(mê/rou to\
e)/stin.]
It is clear from the above passages that Aristotle was thoroughly
aware of the logical fact which Hobbes, James Mill, and Mr. John
Stuart Mill, have more fully brought out and illustrated, as the
confusion between the two distinct functions of the substantive verb.
Many excellent remarks on the subject will be found in the 'System of
Logic,' by Mr. J. S. Mill (Bk. I. ch. iv. s. 1); also in the
'Analysis of the Human Mind,' by James Mill, especially in the recent
edition of that work, containing the explanatory notes by Mr. J. S.
Mill and Dr. Findlater (Vol. I. ch. iv. p. 174, seq.). Mr. J. S.
Mill, however, speaks too unreservedly of this confusion as having
escaped the notice of Aristotle, and as having been brought to light
only by or since Hobbes. He says (in a note on the 'Analysis,' p.
183):--"As in the case of many other luminous thoughts, an approach
is found to have been made to it by previous thinkers. Hobbes, though
he did not reach it, came very close to it; and it was still more
distinctly anticipated by Laromiguière, though without any sufficient
perception of its value . . . in the following words:--'Quand on dit,
l'être est, &c., le mot _est_, ou le verbe, n'exprime pas la même
chose que le mot _être_, sujet de la définition. Si j'énonce la
proposition suivante: Dieu est existant, je ne voudrais pas dire
assurément, Dieu existe existant: cela ne ferait pas un sens: de
même, si je dis que Virgile est poète, je ne veux pas donner à
entendre que Virgile existe. Le verbe _est_ dans la proposition
n'exprime dont pas l'existence réelle; il n'exprime qu'un rapport
spécial entre le sujet et l'attribut, &c.'" The passages above cited
from Aristotle show that he had not only enunciated the same truth as
Laromiguière, but even illustrated it by the same example (Homer
instead of Virgil). I shall in another place state more fully the
views of Aristotle respecting _Existence_.]
3. The third of these heads of Fallacy--_Ignoratio Elenchi_--is, when
the speaker, professing to contradict the thesis, advances another