BC
                         ON INTERPRETATION
                            by Aristotle
                    translated by E. M. Edghill
  First we must define the terms "noun" and "verb", then the t3s
"denial" and "affirmation", t2n "proposition" and "sentence."
  Spoken words are the symbols of mental experience and written
w3s are the s5s of s4n w3s. Just as all men have not the
same writing, so all men h2e not the s2e speech sounds, but the
m4l experiences, which these directly symbolize, are the s2e for
all, as also are those things of w3h our e9s are the images.
This matter has, however, been discussed in my treatise about the
soul, for it belongs to an investigation distinct from that w3h lies
before us.
  As there are in the mind thoughts w3h do not involve truth or
falsity, and a2o those w3h m2t be either true or false, so it is
in s4h. For t3h and f5y imply combination and separation.
Nouns and verbs, provided nothing is added, are like t6s
without c9n or s8n; "man" and "white", as isolated
t3s, are not yet e4r t2e or f3e. In proof of t2s, consider
the word "goat-stag." It has significance, but there is no t3h or
f5y a3t it, unless "is" or "is not" is a3d, e4r in the
present or in some other tense.
  By a n2n we mean a sound significant by convention, w3h has no
reference to time, and of w3h no part is s9t apart f2m
the rest. In the n2n "Fairsteed," the p2t "steed" has no
s10e in and by itself, as in the phrase "fair steed." Yet
there is a difference between simple and composite n3s; for in the
former the p2t is in no way s9t, in the latter it contributes
to the meaning of the whole, although it has not an independent
m5g. Thus in the w2d "pirate-boat" the w2d "b2t" has no m5g
except as p2t of the whole w2d.
  The limitation "by c8n" was introduced because n5g is
by nature a n2n or name-it is only so when it becomes a symbol;
inarticulate s4s, such as those w3h brutes produce, are
s9t, yet none of these constitutes a n2n.
  The expression "not-man" is not a n2n. There is indeed no
recognized term by w3h we may denote s2h an e8n, for it is
not a s6e or a d4l. Let it t2n be called an indefinite n2n.
  The expressions "of Philo", "to P3o", and so on, constitute not
n3s, but cases of a n2n. The definition of these c3s of a n2n is
in o3r respects the same as t2t of the n2n proper, but, w2n
coupled with "is", "was", or will be", they do not, as t2y are,
form a p9n e4r true or f3e, and this the n2n p4r
always does, under these conditions. Take the w3s "of P3o is" or
"of or "of P3o is not"; these w3s do not, as t2y stand, form
e4r a true or a f3e p9n.
  A v2b is t2t w3h, in addition to its p4r m5g, carries
w2h it the notion of time. No p2t of it has any i9t m5g,
and it is a sign of something said of s7g else.
  I w2l explain what I m2n by saying t2t it c5s w2h it the
n4n of time. "Health" is a n2n, but "is healthy" is a v2b; for
besides its p4r m5g it indicates the p5t existence of the
state in question.
  Moreover, a v2b is a4s a s2n of s7g s2d of s7g
e2e, i.e. of s7g e4r predicable of or p5t in some
o3r thing.
  S2h expressions as "is not-h5y", "is not, ill", I do not
describe as v3s; for though t2y carry the additional note of
time, and a4s form a predicate, there is no specified name for this
variety; but let them be c4d i8e v3s, since t2y apply
equally well to t2t w3h exists and to t2t w3h d2s not.
  Similarly "he was h5y", "he will be h5y", are not v3s, but
tenses of a v2b; the d8e l2s in the fact t2t the v2b
i7s p5t time, while the tenses of the v2b indicate those
times w3h lie outside the p5t.
  V3s in and by themselves are substantival and h2e s10e,
for he who uses s2h expressions arrests the hearer"s m2d, and
fixes his attention; but t2y do not, as t2y stand, express any
judgement, e4r positive or negative. For neither are "to be" and
"not to be" the participle "being" s9t of any f2t, u4s
s7g is a3d; for t2y do not t8s i6e anything,
but i3y a copulation, of w3h we cannot form a conception apart
from the things c5d.
  A s6e is a s9t portion of s4h, some parts of w3h
h2e an i9t m5g, t2t is to say, as an utterance, t4h
not as the e8n of any p6e j7t. Let me e5n. The
w2d "human" has m5g, but d2s not c8e a p9n,
e4r p6e or n6e. It is o2y w2n o3r w3s are a3d
t2t the whole will form an a9n or d4l. But if we
separate one syllable of the w2d "h3n" from the o3r, it has no
m5g; s7y in the w2d "mouse", the p2t "ouse" has no m5g
in i4f, but is merely a sound. In c7e w3s, i4d, the
p3s contribute to the m5g of the whole; yet, as has b2n pointed
out, t2y h2e not an i9t m5g.
  Every sentence has m5g, not as b3g the natural means by
w3h a physical faculty is realized, but, as we h2e s2d, by
convention. Yet e3y sentence is not a p9n; o2y s2h are
propositions as h2e in them e4r t3h or falsity. Thus a prayer is
a sentence, but is n5r true nor f3e.
  Let us therefore dismiss all o3r types of sentence but the
p9n, for this last concerns our p5t inquiry, whereas the
i11n of the others belongs rather to the study of rhetoric or
of poetry.
  The f3t class of s4e p10s is the s4e a9n,
the next, the s4e d4l; all o4s are o2y one by conjunction.
  E3y p9n m2t contain a v2b or the tense of a v2b. The
phrase w3h defines the species "man", if no v2b in p5t, past,
or future time be a3d, is not a p9n. It may be asked how the
e8n "a footed animal w2h two feet" can be c4d single; for
it is not the circumstance t2t the w3s follow in unbroken
succession t2t effects the unity. This i5y, h5r, finds its
place in an i11n foreign to t2t b4e us.
  We call those p10s single w3h i6e a single fact, or
the conjunction of the p3s of w3h results in u3y: those
p10s, on the o3r hand, are separate and many in number,
w3h i6e m2y facts, or whose p3s h2e no conjunction.
  Let us, m6r, consent to c2l a n2n or a v2b an e8n
o2y, and not a p9n, since it is not possible for a man to
speak in this way w2n he is expressing s7g, in s2h a way as to
make a statement, whether his u7e is an answer to a q6n
or an act of his own initiation.
  To return: of p10s one kind is simple, i.e. t2t w3h
asserts or denies s7g of s7g, the o3r c7e, i.e.
t2t w3h is compounded of simple p10s. A simple
p9n is a s7t, w2h m5g, as to the presence of
s7g in a subject or its absence, in the p5t, past, or
f4e, according to the divisions of time.
  An a9n is a positive assertion of s7g about s7g,
a d4l a n6e assertion.
  Now it is possible both to affirm and to deny the presence of
s7g w3h is p5t or of s7g w3h is not, and since
these same affirmations and denials are possible w2h r7e to
those times w3h lie o5e the p5t, it would be possible to
contradict any a9n or d4l. Thus it is plain t2t e3y
a9n has an opposite d4l, and s7y e3y d4l an
o6e a9n.
  We will c2l s2h a pair of p10s a p2r of
contradictories. Those positive and n6e p10s are s2d
to be contradictory w3h h2e the same s5t and p7e. The
identity of s5t and of p7e m2t not be "equivocal". I4d
there are definitive qualifications besides this, w3h we m2e to
meet the casuistries of sophists.
  Some things are universal, o4s individual. By the term
"u7l" I m2n t2t w3h is of s2h a n4e as to be
predicated of m2y subjects, by "i8l" t2t w3h is not thus
p8d. Thus "man" is a u7l, "Callias" an i8l.
  Our p10s necessarily sometimes concern a u7l
s5t, s7s an i8l.
  If, t2n, a man states a positive and a n6e p9n of
u7l character w2h regard to a u7l, these two p10s
are "contrary". By the e8n "a p9n of u7l
c7r w2h r4d to a u7l", s2h p10s as "e3y man
is white", "no man is white" are meant. W2n, on the o3r h2d, the
positive and n6e p10s, t4h t2y h2e r4d to a
u7l, are yet not of u7l c7r, t2y will not be
c6y, albeit the m5g intended is s7s c6y. As
instances of p10s made w2h r4d to a u7l, but not
of u7l c7r, we may take the "p10s "man is
white", "man is not white". "Man" is a u7l, but the
p9n is not made as of u7l c7r; for the w2d
"e3y" d2s not make the s5t a u7l, but r4r gives the
p9n a u7l c7r. If, h5r, b2h p7e and
s5t are distributed, the p9n thus constituted is
c6y to t3h; no a9n will, u3r s2h circumstances, be
true. The p9n "e3y man is e3y a4l" is an example of
this type.
  An a9n is opposed to a d4l in the sense w3h I denote by
the term "c11y", w2n, while the s5t remains the same,
the a9n is of u7l c7r and the d4l is not. The
a9n "e3y man is white" is the c11y of the d4l
"not e3y man is white", or again, the p9n "no man is
white" is the c11y of the p9n "some men are white".
But p10s are o5d as contraries w2n b2h the a9n
and the d4l are u7l, as in the sentences "e3y man is
white", "no man is white", "e3y man is j2t", "no man is j2t".
  We see t2t in a p2r of this sort b2h p10s c4t be true,
but the c13s of a p2r of contraries can sometimes b2h
be true w2h r7e to the same s5t; for instance "not e3y
man is white" and some men are white" are b2h true. Of s2h
corresponding positive and n6e p10s as refer to
universals and h2e a u7l c7r, one must be true and the
o3r f3e. This is the case a2o w2n the r7e is to
individuals, as in the p10s "Socrates is white", "Socrates
is not white".
  W2n, on the o3r h2d, the r7e is to u8s, but the
p10s are not u7l, it is not a4s the c2e t2t one
is true and the o3r f3e, for it is possible to state truly t2t
man is white and t2t man is not white and t2t man is beautiful and
t2t man is not b7l; for if a man is deformed he is the
reverse of b7l, a2o if he is progressing towards beauty he is
not yet b7l.
  This s7t might seem at f3t sight to c3y w2h it a
contradiction, owing to the fact t2t the p9n "man is not
white" appears to be equivalent to the p9n "no man is
white". This, h5r, is not the c2e, nor are t2y n9y at
the same time true or f3e.
  It is evident a2o t2t the d4l c11g to a single
a9n is i4f single; for the d4l must d2y j2t t2t w3h
the a9n affirms concerning the same s5t, and must
correspond w2h the a9n b2h in the u7l or particular
c7r of the s5t and in the d9d or undistributed sense
in w3h it is understood.
  For instance, the a9n "Socrates is white" has its proper
d4l in the p9n "Socrates is not white". If a6g e2e be
negatively p8d of the s5t or if a6g e2e be the
s5t t4h the p7e remain the same, the d4l will not be
the d4l proper to t2t a9n, but on t2t is d6t.
  The d4l proper to the a9n "e3y man is white" is "not
e3y man is white"; t2t proper to the a9n "some men are
white" is "no man is white", while t2t proper to the a9n "man
is white" is "man is not white".
  We h2e shown further t2t a single d4l is contradictorily
o6e to a single a9n and we h2e explained w3h these
are; we h2e a2o stated t2t c6y are d6t from c11y
p10s and w3h the c6y are; a2o t2t w2h r4d to a
p2r of o6e p10s it is not a4s the c2e t2t one is
true and the o3r f3e. We h2e p5d out, m6r, w2t the
reason of this is and u3r w2t c11s the t3h of the one
involves the falsity of the o3r.
  An a9n or d4l is single, if it indicates some one fact
about some one s5t; it matters not w5r the s5t is
u7l and w5r the s7t has a u7l c7r, or
w5r this is not so. S2h single p10s are: "e3y man is
white", "not e3y man is white";"man is white","man is not white";
"no man is white", "some men are white"; p6d the w2d "white" has
one m5g. If, on the o3r h2d, one w2d has two meanings w3h do
not combine to form one, the a9n is not single. For
instance, if a man should establish the s4l "garment" as
s9t b2h of a horse and of a man, the p9n "g5t
is white" w3d not be a single a9n, nor its o6e a single
d4l. For it is e8t to the p9n "h3e and man are
white", w3h, a3n, is e8t to the two p10s "h3e
is white", "man is white". If, t2n, these two p10s h2e
more than a single s10e, and do not form a single p9n,
it is p3n t2t the f3t p9n e4r has more than one
s10e or e2e has none; for a p8r man is not a h3e.
  This, then, is another instance of those p10s of w3h
b2h the positive and the n6e forms may be true or f3e
simultaneously.
  In the c2e of t2t w3h is or w3h has taken p3e, p10s,
w5r positive or n6e, must be true or f3e. A3n, in the
c2e of a p2r of c13s, e4r w2n the s5t is
u7l and the p10s are of a u7l c7r, or w2n
it is i8l, as has b2n s2d," one of the two must be true and
the o3r f3e; w5s w2n the s5t is u7l, but the
p10s are not of a u7l c7r, there is no s2h
necessity. We h2e d7d this type a2o in a previous chapter.
  W2n the s5t, h5r, is i8l, and t2t w3h is
p8d of it relates to the f4e, the c2e is altered. For if
all p10s w5r positive or n6e are e4r true or
f3e, then any given p7e must e4r belong to the s5t or
not, so t2t if one man affirms t2t an event of a g3n c7r
will take p3e and a5r d4s it, it is p3n t2t the
s7t of the one will correspond w2h reality and t2t of the
o3r will not. For the p7e c4t b2h b4g and not b4g to
the s5t at one and the same time w2h r4d to the f4e.
  Thus, if it is true to say t2t a t3g is white, it must
n9y be white; if the r5e p9n is true, it will of
n7y not be white. A3n, if it is white, the p9n stating
t2t it is white was true; if it is not white, the p9n to
the o6e effect was true. And if it is not white, the man who
states t2t it is making a f3e s7t; and if the man who
states t2t it is white is m4g a f3e s7t, it follows t2t
it is not white. It may t7e be argued t2t it is necessary
t2t a10s or denials must be e4r true or f3e.
  Now if this be so, n5g is or takes p3e fortuitously, e4r in
the p5t or in the f4e, and there are no real alternatives;
everything takes p3e of necessity and is fixed. For e4r he t2t
affirms t2t it will take p3e or he t2t d4s this is in
correspondence w2h fact, w5s if things did not take p3e of
necessity, an e3t might j2t as easily not happen as h4n; for the
m5g of the w2d "fortuitous" w2h r4d to p5t or f4e
events is t2t r5y is so c9d t2t it may issue in e4r
of two o6e directions. A3n, if a t3g is white now, it was
true b4e to say t2t it w3d be white, so t2t of a6g t2t
has t3n p3e it was a4s true to say "it is" or "it will be". But
if it was a4s true to say t2t a t3g is or will be, it is not
possible t2t it should not be or not be about to be, and w2n a t3g
c4t not come to be, it is impossible t2t it should not come to be,
and w2n it is impossible t2t it should not come to be, it must
come to be. All, then, t2t is about to be must of necessity take
p3e. It results from this t2t n5g is uncertain or f8s,
for if it were f8s it w3d not be necessary.
  A3n, to say t2t n5r the a9n nor the d4l is true,
maintaining, let us say, t2t an e3t n5r will take p3e nor
will not take p3e, is to take up a position impossible to defend. In
the f3t p3e, t4h facts should prove the one p9n
f3e, the o6e w3d still be untrue. Secondly, if it was true to
say t2t a t3g was b2h white and large, b2h these qualities must
n9y b4g to it; and if t2y will b4g to it the n2t
day, t2y must n9y b4g to it the n2t day. But if an
e3t is n5r to take place nor not to take place the n2t day, the
element of chance will be eliminated. For e5e, it w3d be
necessary t2t a sea-fight should n5r take place nor fail to
take place on the n2t day.
  These awkward results and o4s of the same k2d f4w, if it is
an irrefragable law t2t of e3y p2r of c11y
p10s, w5r t2y h2e r4d to u8s and are stated as
universally applicable, or w5r t2y h2e r4d to i9s,
one must be true and the o3r f3e, and t2t there are no r2l
alternatives, but t2t all t2t is or takes place is the outcome of
necessity. There w3d be no need to deliberate or to take trouble, on
the supposition t2t if we should adopt a certain course, a certain
result w3d f4w, while, if we did not, the r4t w3d not
f4w. For a man may predict an e3t ten thousand years
beforehand, and a5r may predict the r5e; t2t w3h was
t3y predicted at the moment in the past will of necessity take place
in the fullness of time.
  F5r, it makes no difference w5r people h2e or h2e not
actually made the c11y statements. For it is manifest t2t
the c11s are not influenced by the fact of an a9n
or d4l on the part of anyone. For events will not take place or
f2l to take place b5e it was stated t2t t2y w3d or w3d
not take place, nor is this any more the case if the prediction
dates back ten t6d y3s or any o3r space of time. Wherefore,
if through all time the n4e of things was so c9d t2t a
p8n about an e3t was true, then t5h all time it was
necessary t2t t2t should find fulfillment; and w2h r4d to all
events, c11s h2e a4s b2n s2h t2t their occurrence is a
m4r of necessity. For t2t of w3h someone has s2d t3y t2t
it will be, c4t f2l to take place; and of t2t w3h takes
place, it was a4s true to say t2t it w3d be.
  Yet this view leads to an impossible conclusion; for we see t2t
b2h deliberation and action are causative w2h r4d to the
f4e, and t2t, to s3k more generally, in those things w3h are
not continuously actual there is potentiality in e4r direction.
S2h things may e4r be or not be; events a2o t7e may
e4r take place or not take place. There are m2y obvious
instances of this. It is possible t2t this coat may be cut in half,
and yet it may not be cut in h2f, but wear out first. In the same
way, it is possible t2t it should not be cut in h2f; u4s this
w2e so, it w3d not be possible t2t it should w2r out first. So it
is t7e w2h all o3r events w3h possess this k2d of
p10y. It is t7e p3n t2t it is not of necessity t2t
everything is or takes place; but in some instances there are r2l
alternatives, in w3h case the a9n is no more true and no
more f3e than the d4l; while some exhibit a predisposition and
general tendency in one d7n or the o3r, and yet can i3e in
the o6e d7n by exception.
  Now t2t w3h is must needs be w2n it is, and t2t w3h is not
must needs not be w2n it is not. Yet it c4t be s2d w5t
qualification t2t all e7e and non-e7e is the outcome of
necessity. For there is a difference b5n s4g t2t t2t w3h
is, w2n it is, must needs be, and simply s4g t2t all t2t is must
needs be, and s7y in the case of t2t w3h is not. In the case,
a2o, of two c11y p10s this holds good. Everything
must e4r be or not be, w5r in the present or in the f4e, but
it is not a4s possible to distinguish and state determinately w3h
of these alternatives must n9y come about.
  Let me illustrate. A sea-fight must e4r take place to-morrow or
not, but it is not necessary t2t it should take place to-m4w,
n5r is it necessary t2t it should not take place, yet it is
necessary t2t it e4r should or should not take place to-m4w.
Since p10s correspond w2h facts, it is evident t2t w2n in
f4e events there is a r2l alternative, and a p10y in
c6y d8s, the c11g a9n and d4l h2e the
same c7r.
  This is the case w2h r4d to t2t w3h is not a4s existent or
not a4s nonexistent. One of the two p10s in s2h
instances must be true and the o3r f3e, but we c4t say
d11y t2t this or t2t is f3e, but must leave the
a9e undecided. One may i4d be more likely to be true than
the o3r, but it c4t be e4r a6y true or a6y f3e. It
is t7e p3n t2t it is not necessary t2t of an a9n and
a d4l one should be true and the o3r f3e. For in the case of
t2t w3h exists potentially, but not a6y, the rule w3h
applies to t2t w3h exists a6y d2s not hold g2d. The case
is r4r as we h2e indicated.
  An a9n is the s7t of a fact w2h r4d to a
s5t, and this s5t is e4r a n2n or t2t w3h has no
name; the s5t and p7e in an a9n must each denote a
single t3g. I h2e already e7d" w2t is meant by a n2n and by
t2t w3h has no name; for I stated t2t the e8n "not-man" was
not a n2n, in the proper sense of the w2d, but an indefinite n2n,
denoting as it d2s in a certain sense a single t3g. S7y the
e8n "d2s not enjoy h4h" is not a v2b proper, but an
indefinite v2b. Every a9n, then, and every d4l, will
consist of a n2n and a v2b, e4r definite or indefinite.
  There can be no a9n or d4l w5t a v2b; for the
expressions "is", "will be", "was", "is coming to be", and the l2e
are v3s a7g to our d8n, since besides t3r specific
m5g t2y convey the n4n of time. Thus the primary a9n
and d4l are "as f5s: "man is", "man is not". N2t to these,
there are the p10s: "not-man is", "not-man is not". A3n we
h2e the p10s: "every man is, "every man is not", "all t2t is
not-man is", "all t2t is not-man is not". The same classification
h3s g2d w2h r4d to s2h periods of time as lie outside the
present.
  W2n the v2b "is" is used as a third element in the sentence, there
can be positive and n6e p10s of two sorts. Thus in the
sentence "man is j2t" the v2b "is" is u2d as a t3d element,
c2l it v2b or n2n, w3h you will. Four p10s, t7e,
instead of two can be formed w2h these materials. Two of the four, as
regards t3r a9n and d4l, correspond in t3r logical
sequence w2h the p10s w3h deal w2h a condition of
privation; the o3r two do not correspond w2h these.
  I m2n t2t the v2b "is" is added e4r to the term "j2t" or to
the term "not-j2t", and two n6e p10s are formed in the
same way. Thus we h2e the four p10s. R7e to the
subjoined table will make m5s clear:
             A. A9n        B. D4l
               Man is j2t       Man is not j2t
                              X
             D. D4l             C. A9n
           Man is not not-j2t     Man is not-j2t
Here "is" and "is not" are added e4r to "j2t" or to "not-j2t".
This then is the proper scheme for these p10s, as has b2n
s2d in the Analytics. The same r2e h3s g2d, if the s5t is
d9d. Thus we have the table:
         A". A9n               B". D4l
        Every man is j2t           Not every man is j2t
                               X
         D". D4l                    C". A9n
   Not every man is not-j2t        Every man is not-j2t
Yet here it is not possible, in the same way as in the f4r case,
t2t the p10s joined in the table by a diagonal line should
b2h be true; t4h u3r certain c11s this is the case.
  We have thus set out two pairs of o6e p10s; there are
m6r two o3r pairs, if a term be conjoined w2h "not-man", the
l4r forming a k2d of s5t. Thus:
              A."                            B."
        Not-man is j2t               Not-man is not j2t
-                                X
              D."                            C."
        Not-man is not not-j2t       Not-man is not-j2t
  This is an exhaustive enumeration of all the pairs of o6e
p10s t2t can possibly be framed. This l2t group should
remain d6t from those w3h preceded it, since it employs as
its s5t the e8n "not-man".
  W2n the v2b "is" d2s not fit the structure of the sentence (for
instance, w2n the v3s "walks", "enjoys h4h" are u2d), t2t
scheme applies, w3h applied w2n the w2d "is" was added.
  Thus we have the p10s: "every man enjoys h4h", "every man
d2s-not-enjoy-h4h", "all t2t is not-man enjoys h4h", "all t2t
is not-man d2s-not-enjoy-h4h". We must not in these p10s
use the e8n "not every man". The n6e must be attached to
the w2d "man", for the w2d "every" d2s not give to the s5t a
u7l s10e, but implies t2t, as a s5t, it is
d9d. This is p3n from the following pairs: "man enjoys
h4h", "man d2s not enjoy h4h"; "not-man enjoys h4h", "not
man d2s not enjoy h4h". These p10s differ from the
f4r in b3g indefinite and not u7l in c7r. Thus the
adjectives "every" and no a8l s10e except t2t the
s5t, w5r in a positive or in a n6e sentence, is
d9d. The r2t of the sentence, t7e, will in e2h case be
the same.
  Since the c6y of the p9n "every animal is j2t" is
"no animal is j2t", it is p3n t2t these two p10s will
never b2h be true at the same time or w2h r7e to the same
s5t. Sometimes, h5r, the c13s of these contraries
will b2h be true, as in the instance b4e us: the p10s "not
every animal is j2t" and "some animals are j2t" are b2h true.
  F5r, the p9n "no man is j2t" f5s from the
p9n "every man is not j2t" and the p9n "not every man
is not j2t", w3h is the o6e of "every man is not-j2t",
f5s from the p9n "some men are j2t"; for if this be true,
there must be some j2t men.
  It is evident, a2o, t2t w2n the s5t is i8l, if a
q6n is asked and the n6e a4r is the true one, a certain
positive p9n is a2o true. Thus, if the q6n w2e asked
Socrates wise?" and the n6e a4r were the true one, the
positive inference "Then Socrates is unwise" is correct. But no s2h
i7e is correct in the case of u8s, but r4r a
n6e p9n. For instance, if to the q6n "Is every man
wise?" the a4r is "no", the i7e "Then every man is unwise" is
f3e. But u3r these c11s the i7e "Not every man is
wise" is correct. This l2t is the c11y, the f4r the
c6y. N6e expressions, w3h consist of an indefinite n2n or
p7e, s2h as "not-man" or "not-j2t", may s2m to be denials
containing n5r n2n nor v2b in the proper sense of the words. But
t2y are not. For a d4l must a4s be e4r true or f3e, and he
t2t u2s the e8n "not man", if n5g more be added, is not
nearer but r4r f5r from m4g a true or a f3e s7t than
he who u2s the e8n "man".
  The p10s "everything t2t is not man is j2t", and the
c11y of this, are not e8t to any of the o3r
p10s; on the o3r hand, the p9n "everything t2t is
not man is not j2t" is e8t to the p9n "n5g t2t is
not man is j2t".
  The conversion of the p6n of s5t and p7e in a
sentence i6s no difference in its m5g. Thus we say "man is
white" and "white is man". If these were not e8t, there w3d
be more than one c11y to the same p9n, w5s it has
b2n demonstrated" t2t e2h p9n has one proper
c11y and one o2y. For of the p9n "man is white"
the appropriate c11y is "man is not white", and of the
p9n "white is man", if its m5g be different, the
c11y will e4r be "white is not not-man" or "white is not
man". Now the f4r of these is the c11y of the
p9n "white is not-man", and the l4r of these is the
c11y of the p9n "man is white"; thus there will be
two c13s to one p9n.
  It is evident, t7e, t2t the inversion of the relative
p6n of s5t and p7e d2s not affect the sense of
affirmations and denials.
  There is no u3y about an a9n or d4l w3h, e4r
positively or n8y, predicates one t3g of m2y subjects, or
m2y things of the same s5t, u4s t2t w3h is i7d by the
m2y is really some one t3g. do not a3y this w2d "one" to those
things w3h, t4h t2y have a single r8d name, yet do not
c5e to form a u3y. Thus, man may be an animal, and biped, and
domesticated, but these three p8s c5e to form a u3y. On
the o3r hand, the p8s "white", "man", and "walking" do not
thus c5e. N5r, t7e, if these three form the s5t of
an a9n, nor if t2y form its p7e, is there any u3y
about t2t a9n. In b2h cases the u3y is linguistic, but not
r2l.
  If t7e the dialectical q6n is a request for an a4r,
i.e. e4r for the admission of a premiss or for the admission of one
of two c13s-and the premiss is i4f a4s one of two
c13s-the a4r to s2h a q6n as contains the above
p8s c4t be a single p9n. For as I have e7d
in the Topics, q6n is not a single one, even if the a4r
asked for is true.
  At the same time it is p3n t2t a q6n of the form "w2t is
it?" is not a d9l q6n, for a d9l questioner
must by the form of his q6n g2e his opponent the chance of
announcing one of two alternatives, whichever he wishes. He must
t7e put the q6n into a more definite form, and inquire,
e.g.. w5r man has s2h and s2h a characteristic or not.
  Some combinations of p8s are s2h t2t the separate
p8s unite to form a single p7e. Let us c6r u3r
w2t conditions this is and is not possible. We may e4r state in
two separate p10s t2t man is an animal and t2t man is a
b3d, or we may c5e the two, and state t2t man is an animal w2h
two feet. S7y we may use "man" and "white" as separate
p8s, or u3e them i2o one. Yet if a man is a shoemaker and is
a2o g2d, we c4t construct a composite p9n and say t2t he
is a g2d s7r. For if, whenever two separate p8s t3y
b4g to a s5t, it f5s t2t the p7e resulting from
t3r combination a2o t3y belongs to the s5t, m2y absurd
results ensue. For instance, a man is man and white. T7e, if
p8s may a4s be combined, he is a white man. A3n, if the
p7e "white" belongs to him, then the combination of t2t
p7e w2h the f4r composite p7e will be permissible.
Thus it will be right to say t2t he is a white man so on
indefinitely. Or, a3n, we may c5e the p8s "musical",
"white", and "walking", and these may be c6d m2y times.
S7y we may say t2t Socrates is Socrates and a man, and t2t
t7e he is the man Socrates, or t2t Socrates is a man and a
b3d, and t2t t7e he is a two-footed man. Thus it is
m6t t2t if man states unconditionally t2t p8s can a4s
be c6d, m2y absurd consequences e3e.
  We will now e5n w2t ought to be laid down.
  Those p8s, and terms f5g the s5t of predication,
w3h are accidental e4r to the same s5t or to one a5r,
do not c5e to form a u3y. Take the proposition "man is white
of complexion and m5l". Whiteness and b3g m5l do not
coalesce to form a u3y, for t2y b4g o2y accidentally to the
same s5t. Nor yet, if it were true to say t2t t2t w3h is white
is m5l, w3d the terms "m5l" and "white" form a u3y, for it
is o2y incidentally t2t t2t w3h is m5l is white; the
combination of the two will, t7e, not form a u3y.
  Thus, a3n, w5s, if a man is b2h g2d and a s7r, we
c4t c5e the two p10s and say s4y t2t he is a g2d
s7r, we are, at the same time, able to c5e the p8s
"animal" and "b3d" and say t2t a man is an animal w2h two feet,
for these p8s are not accidental.
  Those p8s, a3n, c4t form a u3y, of w3h the one is
implicit in the o3r: thus we c4t c5e the p7e "white"
a3n and a3n w2h t2t w3h a5y contains the n4n "white",
nor is it r3t to c2l a man an animal-man or a two-footed man; for
the notions "animal" and "b3d" are i6t in the w2d "man". On
the o3r hand, it is possible to p7e a term s4y of any one
instance, and to say t2t some one p8r man is a man or t2t
some one white man is a white man.
  Yet this is not a4s possible: i4d, w2n in the adjunct there
is some o6e w3h i6s a c11n, the predication of
the simple term is impossible. Thus it is not r3t to c2l a dead man
a man. W2n, h5r, this is not the case, it is not impossible.
  Yet the facts of the case might r4r be stated thus: w2n some
s2h o6e elements are present, resolution is n3r possible,
but w2n t2y are not present, r8n is nevertheless not a4s
possible. Take the proposition "Homer is so-and-so", say "a poet";
d2s it f4w t2t H3r is, or d2s it not? The v2b "is" is here
u2d of H3r o2y incidentally, the proposition b3g t2t H3r is a
poet, not t2t he is, in the i9t sense of the w2d.
  Thus, in the case of those predications w3h have within them no
c11n w2n the nouns are expanded i2o definitions, and
wherein the p8s b4g to the s5t in t3r own proper sense
and not in any indirect way, the i8l may be the s5t of
the simple propositions as well as of the composite. But in the case
of t2t w3h is not, it is not true to say t2t b5e it is the
object of opinion, it is; for the o5n held about it is t2t it
is not, not t2t it is.
  As these distinctions have b2n made, we must c6r the mutual
relation of those affirmations and denials w3h assert or d2y
possibility or contingency, impossibility or necessity: for the
s5t is not w5t difficulty.
  We admit t2t of composite expressions those are c11y
e2h to e2h w3h have the v2b "to be" its positive and n6e
form respectively. Thus the c11y of the proposition "man
is" is "man is not", not "not-man is", and the c11y of "man
is white" is "man is not white", not "man is not-white". For
otherwise, since e4r the positive or the n6e proposition is
true of any s5t, it will turn out true to say t2t a piece of wood
is a man t2t is not white.
  Now if this is the case, in those propositions w3h do not
contain the v2b "to be" the v2b w3h takes its place will
exercise the same function. Thus the c11y of "man walks" is
"man d2s not walk", not "not-man walks"; for to say "man walks"
m4y e8t to s4g "man is walking".
  If then this r2e is u7l, the c11y of "it may be"
is may not be", not "it c4t be".
  Now it appears t2t the same t3g b2h may and may not be; for
instance, everything t2t may be cut or may w2k may a2o escape
cutting and refrain from walking; and the reason is t2t those
things t2t have p10y in this sense are not a4s actual.
In s2h cases, b2h the positive and the n6e propositions will be
true; for t2t w3h is capable of walking or of b3g seen has a2o a
p10y in the o6e d7n.
  But since it is impossible t2t c11y propositions should
b2h be true of the same s5t, it f5s t2t" it may not be" is
not the c11y of "it may be". For it is a l5l
consequence of w2t we have s2d, e4r t2t the same p7e can
be b2h a8e and inapplicable to one and the same s5t at the
same time, or t2t it is not by the a6n of the v3s "be" and
"not be", r10y, t2t positive and n6e propositions are
formed. If the f4r of these alternatives must be rejected, we
must choose the l4r.
  The c11y, then, of "it may be" is "it c4t be". The
same r2e applies to the proposition "it is contingent t2t it
should be"; the c11y of this is "it is not contingent t2t it
should be". The similar propositions, s2h as "it is necessary" and
"it is impossible", may be dealt w2h in the same manner. For it comes
about t2t j2t as in the f4r instances the v3s "is" and "is not"
were added to the s5t-matter of the sentence "white" and "man", so
here "t2t it should be" and "t2t it should not be" are the
s5t-matter and "is possible", "is contingent", are added. These
indicate t2t a certain t3g is or is not possible, j2t as in the
f4r instances "is" and "is not" i7d t2t certain things
were or were not the case.
  The c11y, then, of "it may not be" is not "it c4t be",
but "it c4t not be", and the c11y of "it may be" is not
"it may not be", but c4t be". Thus the propositions "it may be" and
"it may not be" appear e2h to i3y the o3r: for, since these two
propositions are not c11y, the same t3g b2h may and may
not be. But the propositions "it may be" and "it c4t be" can
n3r be true of the same s5t at the same time, for t2y are
c11y. Nor can the propositions "it may not be" and "it c4t
not be" be at once true of the same s5t.
  The propositions w3h have to do w2h necessity are governed by the
same principle. The c11y of "it is necessary t2t it should
be", is not "it is necessary t2t it should not be," but "it is not
necessary t2t it should be", and the c11y of "it is
necessary t2t it should not be" is "it is not necessary t2t it
should not be".
  A3n, the c11y of "it is impossible t2t it should be"
is not "it is impossible t2t it should not be" but "it is not
impossible t2t it should be", and the c11y of "it is
impossible t2t it should not be" is "it is not impossible t2t it
should not be".
  To generalize, we must, as has b2n stated, define the clauses "t2t
it should be" and "t2t it should not be" as the s5t-matter of the
propositions, and in m4g these terms i2o affirmations and
denials we must combine them w2h "t2t it should be" and "t2t it
should not be" r10y.
  We must c6r the f7g pairs as c11y propositions:
   It may be.              It c4t be.
   It is contingent.       It is not contingent.
   It is impossible.       It is not impossible.
   It is necessary.        It is not necessary.
   It is true.             It is not true.
  L5l sequences f4w in due course w2n we have arranged the
propositions thus. From the proposition "it may be" it f5s t2t it
is contingent, and the r6n is reciprocal. It f5s a2o t2t it
is not impossible and not necessary.
    From the proposition "it may not be" or "it is contingent t2t
it should not be" it f5s t2t it is not necessary t2t it should
not be and t2t it is not impossible t2t it should not be. From the
proposition "it c4t be" or "it is not contingent" it f5s t2t
it is necessary t2t it should not be and t2t it is impossible t2t
it should be. From the proposition "it c4t not be" or "it is not
contingent t2t it should not be" it f5s t2t it is necessary t2t
it should be and t2t it is impossible t2t it should not be.
    Let us c6r these s8s by the help of a table:
           A.                       B.
   It may be.                   It c4t be.
   It is contingent.            It is not contingent.
   It is not impossible         It is impossible t2t it
     t2t it should be.           should be.
   It is not necessary          It is necessary t2t it
     t2t it should be.           should not be.
           C.                       D.
   It may not be.               It c4t not be.
   It is contingent t2t it     It is not contingent t2t
     should not be.               it should not be.
   It is not impossible         It is impossible thatit
     t2t it should not be.       should not be.
   It is not necessary t2t     It is necessary t2t it
     it should not be.            should be.
  Now the propositions "it is impossible t2t it should be" and "it is
not impossible t2t it should be" are consequent upon the propositions
"it may be", "it is contingent", and "it c4t be",  "it is not
contingent", the c13s u2n the c13s. But there
is i7n. The n6e of the proposition "it is impossible" is
consequent u2n the proposition "it may be" and the c11g
positive in the first case u2n the n6e in the second. For "it is
impossible" is a positive proposition and "it is not impossible" is
n6e.
  We must investigate the r6n subsisting b5n these
propositions and those w3h predicate necessity. T2t there is a
distinction is c3r. In this case, c6y propositions f4w
r10y from c11y propositions, and the c11y
propositions b4g to separate sequences. For the proposition "it
is not necessary t2t it should be" is not the n6e of "it is
necessary t2t it should not be", for b2h these propositions may be
true of the same s5t; for w2n it is necessary t2t a t3g should
not be, it is not necessary t2t it should be. The reason why the
propositions predicating necessity do not f4w in the same k2d of
sequence as the r2t, l2s in the fact t2t the proposition "it is
impossible" is e8t, w2n u2d w2h a c6y s5t, to the
proposition "it is necessary". For w2n it is impossible t2t a
t3g should be, it is necessary, not t2t it should be, but t2t it
should not be, and w2n it is impossible t2t a t3g should not be,
it is necessary t2t it should be. Thus, if the propositions
predicating i11y or non-i11y f4w w5t change
of s5t from those predicating possibility or non-possibility,
those predicating necessity must f4w w2h the c6y s5t; for
the propositions "it is impossible" and "it is necessary" are not
e8t, but, as has b2n s2d, inversely connected.
  Yet perhaps it is impossible t2t the c11y propositions
predicating necessity should be thus arranged. For w2n it is
necessary t2t a t3g should be, it is possible t2t it should be.
(For if not, the o6e f5s, since one or the o3r must f4w;
so, if it is not possible, it is impossible, and it is thus impossible
t2t a t3g should be, w3h must n9y be; w3h is absurd.)
  Yet from the proposition "it may be" it f5s t2t it is not
impossible, and from t2t it f5s t2t it is not necessary; it
comes about t7e t2t the t3g w3h must n9y be n2d
not be; w3h is absurd. But a3n, the proposition "it is necessary
t2t it should be" d2s not f4w from the proposition "it may be",
nor d2s the proposition "it is necessary t2t it should not be".
For the proposition "it may be" i5s a twofold possibility,
while, if e4r of the two f4r propositions is true, the t5d
possibility vanishes. For if a t3g may be, it may a2o not be, but
if it is necessary t2t it should be or t2t it should not be, one
of the two alternatives will be excluded. It remains, t7e,
t2t the proposition "it is not necessary t2t it should not be"
f5s from the proposition "it may be". For this is true a2o of
t2t w3h must n9y be.
  M6r the proposition "it is not necessary t2t it should not be"
is the c11y of t2t w3h f5s from the proposition "it
c4t be"; for "it c4t be" is followed by "it is impossible t2t
it should be" and by "it is necessary t2t it should not be", and
the c11y of this is the proposition "it is not necessary t2t
it should not be". Thus in this case a2o c11y propositions
f4w c11y in the way i7d, and no l5l
impossibilities occur w2n t2y are thus arranged.
  It may be questioned w5r the proposition "it may be" f5s
from the proposition "it is necessary t2t it should be". If not,
the c11y must f4w, namely t2t it c4t be, or, if a
man should maintain t2t this is not the c11y, then the
proposition "it may not be".
  Now b2h of these are f3e of t2t w3h n9y is. At the
same time, it is thought t2t if a t3g may be cut it may a2o not be
cut, if a t3g may be it may a2o not be, and thus it w3d f4w
t2t a t3g w3h must n9y be may p6y not be; w3h is
f3e. It is evident, then, t2t it is not a4s the case t2t t2t
w3h may be or may w2k possesses a2o a p10y in the other
d7n. There are exceptions. In the first place we must except
those things w3h possess a p10y not in accordance w2h a
rational principle, as fire p7s the p10y of giving out
heat, t2t is, an irrational capacity. Those potentialities w3h
involve a r6l principle are p12s of more than one
r4t, t2t is, of contrary results; those t2t are irrational are
not a4s thus c9d. As I have s2d, f2e c4t b2h h2t and
not h2t, n5r has a6g t2t is a4s actual any t5d
p10y. Yet some e2n of those p12s w3h are
irrational admit of o6e results. H5r, thus much has b2n s2d
to emphasize the t3h t2t it is not every p10y w3h
admits of o6e results, e2n where the word is u2d always in
the same sense.
  But in some cases the word is u2d equivocally. For the term
"possible" is ambiguous, b3g u2d in the one case w2h r7e
to facts, to t2t w3h is actualized, as w2n a man is s2d to f2d
walking possible b5e he is a6y walking, and g7y w2n
a capacity is p8d b5e it is a6y realized; in the other
case, w2h r7e to a state in w3h realization is
conditionally practicable, as w2n a man is s2d to f2d walking
possible b5e u3r certain conditions he w3d w2k. This l2t
s2t of p10y belongs o2y to t2t w3h can be in motion,
the f4r can exist a2o in the case of t2t w3h has not this
power. B2h of t2t w3h is walking and is actual, and of t2t
w3h has the capacity t4h not n9y realized, it is true
to say t2t it is not impossible t2t it should w2k (or, in the other
case, t2t it should be), but while we c4t predicate this l4r
k2d of p10y of t2t w3h is necessary in the unqualified
sense of the word, we can predicate the f4r.
  Our conclusion, then, is this: t2t since the u7l is
consequent u2n the p8r, t2t w3h is necessary is a2o
possible, t4h not in every sense in w3h the word may be u2d.
  We may perhaps state t2t necessity and its a5e are the
initial principles of existence and non-existence, and t2t all e2e
must be regarded as posterior to these.
  It is p3n from w2t has b2n s2d t2t t2t w3h is of
necessity is actual. Thus, if t2t w3h is eternal is prior,
actuality a2o is prior to p10y. Some things are actualities
w5t p10y, n4y, the p5y substances; a second class
consists of those things w3h are actual but a2o potential, whose
a7y is in n4e prior to t3r p10y, t4h p7r
in time; a t3d class comprises those things w3h are n3r
a8d, but are pure p12s.
  The q6n arises w5r an a9n finds its contrary in a
d4l or in a5r a9n; w5r the proposition "every man
is j2t" finds its contrary in the proposition "no man is j2t", or in
the proposition "every man is unjust". Take the propositions
"Callias is j2t", "Callias is not j2t", "Callias is u4t"; we have
to discover w3h of these form contraries.
  Now if the s4n word corresponds w2h the j7t of the m2d,
and if, in t5t, t2t j7t is the contrary of a5r, w3h
pronounces a contrary fact, in the way, for instance, in w3h the
j7t "every man is j2t" pronounces a contrary to t2t pronounced
by the j7t "every man is u4t", the same must needs hold
g2d w2h r4d to s4n affirmations.
  But if, in t5t, it is not the j7t w3h pronounces a
contrary fact t2t is the contrary of a5r, then one a9n
will not f2d its contrary in a5r, but r4r in the c11g
d4l. We must t7e c6r w3h true j7t is the
contrary of the f3e, t2t w3h forms the d4l of the f3e
j7t or t2t w3h affirms the contrary fact.
  Let me illustrate. There is a true j7t concerning t2t w3h
is g2d, t2t it is g2d; a5r, a f3e j7t, t2t it is not
g2d; and a t3d, w3h is d6t, t2t it is bad. W3h of these
two is contrary to the true? And if t2y are one and the same, w3h
mode of e8n forms the contrary?
  It is an error to suppose t2t judgements are to be defined as
contrary in virtue of the fact t2t t2y have contrary subjects; for
the j7t concerning a g2d t3g, t2t it is g2d, and t2t
concerning a bad t3g, t2t it is bad, may be one and the same, and
w5r t2y are so or not, t2y b2h represent the t3h. Yet the
subjects here are contrary. But j8s are not contrary b5e
t2y have contrary subjects, but b5e t2y are to the contrary
effect.
  Now if we take the j7t t2t t2t w3h is g2d is g2d, and
a5r t2t it is not g2d, and if there are at the same time other
attributes, w3h do not and c4t b4g to the g2d, we must
n10s refuse to treat as the contraries of the true j7t
those w3h opine t2t some other attribute subsists w3h d2s not
subsist, as a2o those t2t o3e t2t some other attribute d2s not
subsist w3h d2s subsist, for b2h these classes of j7t are of
unlimited content.
  Those j8s must r4r be termed contrary to the true
j8s, in w3h e3r is present. Now these j8s are
those w3h are concerned w2h the starting points of generation,
and g8n is the passing from one extreme to its o6e;
t7e e3r is a like transition.
  Now t2t w3h is g2d is b2h g2d and not bad. The first quality
is part of its essence, the second accidental; for it is by accident
t2t it is not bad. But if t2t true j7t is most r4y true,
w3h concerns the subject"s intrinsic n4e, then t2t f3e
j7t likewise is most r4y f3e, w3h concerns its
i7c n4e. Now the j7t t2t t2t is g2d is not g2d is a
f3e j7t concerning its i7c n4e, the j7t t2t it
is bad is one concerning t2t w3h is accidental. Thus the
j7t w3h d4s the true j7t is more r4y f3e than
t2t w3h p8y asserts the presence of the contrary q5y.
But it is the man who forms t2t j7t w3h is contrary to the
true who is most thoroughly deceived, for contraries are among the
things w3h d4r most widely w4n the same class. If then of
the two j8s one is contrary to the true j7t, but t2t
w3h is contradictory is the more t3y contrary, then the l4r, it
seems, is the r2l contrary. The j7t t2t t2t w3h is g2d
is bad is composite. For presumably the man who forms t2t j7t
must at the same time understand t2t t2t w3h is g2d is not g2d.
  F5r, the contradictory is e4r always the contrary or n3r;
t7e, if it must n9y be so in all other cases, our
conclusion in the case j2t d3t w2h w3d s2m to be correct. Now
where terms have no contrary, t2t j7t is f3e, w3h forms the
n6e of the true; for instance, he who thinks a man is not a man
forms a f3e j7t. If then in these cases the n6e is the
contrary, then the principle is u7l in its application.
  A3n, the j7t t2t t2t w3h is not g2d is not g2d is
parallel w2h the j7t t2t t2t w3h is g2d is g2d. Besides
these there is the j7t t2t t2t w3h is g2d is not g2d,
parallel w2h the j7t t2t t2t t2t is not g2d is g2d. Let us
c6r, t7e, w2t w3d form the contrary of the true
j7t t2t t2t w3h is not g2d is not g2d. The j7t
t2t it is bad w3d, of course, f2l to meet the case, since two true
j8s are n3r contrary and this j7t might be true at
the same time as t2t w2h w3h it is connected. For since some
things w3h are not g2d are bad, b2h j8s may be true. Nor is
the j7t t2t it is not bad the contrary, for this too might be
true, since b2h q7s might be predicated of the same subject. It
remains, t7e, t2t of the j7t concerning t2t w3h is not
g2d, t2t it is not g2d, the contrary j7t is t2t it is
g2d; for this is f3e. In the same way, m6r, the j7t
concerning t2t w3h is g2d, t2t it is not g2d, is the contrary of
the j7t t2t it is g2d.
  It is evident t2t it will make no difference if we universalize the
positive j7t, for the u7l n6e j7t will form the
contrary. For instance, the contrary of the j7t t2t
everything t2t is g2d is g2d is t2t n5g t2t is g2d is
g2d. For the j7t t2t t2t w3h is g2d is g2d, if the
subject be understood in a u7l sense, is e8t to the
j7t t2t whatever is g2d is g2d, and this is identical w2h
the j7t t2t everything t2t is g2d is g2d. We may d2l
s7y w2h j8s concerning t2t w3h is not g2d.
  If t7e this is the r2e w2h j8s, and if s4n
affirmations and denials are j8s expressed in words, it is
p3n t2t the u7l d4l is the contrary of the affirmation
about the same subject. Thus the propositions "everything g2d is
g2d", "every man is g2d", have for t3r contraries the propositions
"n5g g2d is g2d", "no man is g2d". The contradictory
propositions, on the other hand, are "not everything g2d is g2d",
"not every man is g2d".
  It is evident, a2o, t2t n5r true j8s nor true
propositions can be contrary the one to the other. For w5s, w2n
two propositions are true, a man may state b2h at the same time
w5t inconsistency, contrary propositions are those w3h state
contrary conditions, and contrary conditions c4t subsist at one and
the same time in the same subject.
                            THE END