<<

. 4
( 11)



>>

ch a p t er 3

Truth conditions for existential assertions




In chapter 2 I sketched a theory of the truth conditions for predicative asser-
tions which can be plausibly attributed to Aristotle. The goal of the present
chapter is to attain the corresponding result for existential assertions: to
outline a theory of the truth conditions for existential assertions which
captures Aristotle™s views concerning this subject. Aristotle concentrates
on existential beliefs and assertions of two types: those concerning simple
items, which are essences and incorporeal substances (i.e. God and, per-
haps, the intellects that move the heavenly spheres), and those concerning
material substances.1
The ¬rst two sections of the chapter focus on existential beliefs and asser-
tions concerning simple items. The main witnesses are the middle part of
Metaphysics 10 and the beginning and the end of de Anima 3.6. Section 1
offers truth conditions for existential beliefs and assertions concerning sim-
ple items. Simple items are essences and incorporeal substances, all of which
are everlasting, i.e. exist always. It follows that every existential af¬rmative
belief, or assertion, concerning a simple item is always true, and every
existential negative belief, or assertion, concerning a simple item is always
false. Section 2 addresses a scholarly issue: what does Aristotle have in mind
when in Metaphysics 10 he speaks of ˜non-composite substances™? I defend
the traditional interpretation, according to which Aristotle has incorporeal
substances in mind.
Section 3 addresses the truth conditions for existential beliefs and asser-
tions concerning material substances. Since most material substances are
not everlasting, one can refer to a material substance that once existed but
does not exist at the time when the belief, or assertion, is evaluated. It fol-
lows that an existential af¬rmative belief, or assertion, concerning a material
substance can be true at one time but false at another (in this respect existen-
tial beliefs and assertions concerning material substances differ from those

1 For ˜corporeal™, ˜incorporeal™, ˜material™, and ˜immaterial™ see n. 7 of the introduction.

99
100 Bearers of truth or falsehood
concerning simple items). Material substances are composite items: they
are composed of form and matter. Material substances therefore resemble
states of affairs, which also are composite items. It follows that existential
assertions concerning material substances have a lot in common with pred-
icative assertions, which concern states of affairs. An existential af¬rmation
(denial) concerning a material substance asserts that that substance™s form is
combined with (divided from) its matter, and is true when and only when
the combination (division) in fact obtains.

1 e x ist entia l a s sertions conc e rn i ng s i m ple i t e m s
Predicative vs existential assertions. Aristotle contrasts in various ways pred-
icative and existential assertions: he distinguishes ˜being simply™ (e²nai
‰pl¤v) from ˜being something™ (e²na© ti);2 he opposes knowing that some-
thing exists (e.g. that gods exist) to knowing that something is so-and-so
(e.g. that the moon is eclipsed);3 he sets assertions like ˜A man is™, ˜A man is
not™, etc. apart from assertions like ˜A man is just™, ˜A man is not just™, etc.4

From composite to non-composite items. After discussing, in the ¬rst part
of Metaphysics 10 (1051a 34“1051b 17 = T2 + T6), true and false beliefs
and assertions concerning composite items, in the chapter™s central part

2 APr. 1.38, 49a 37“8; APo. 2.2, 90a 3“5; 90a 9“11; 90a 32“4; SE 5, 166b 37“167a 6; 25, 180a 36“8; Ph. 1.3,
187a 5“6 (f ™s reading); Metaph. Z 1, 1028a 30“1 (with Bostock (1994), 57) (cf. APo. 2.1, 89b 33; Metaph. Z
4, 1030a 25“6; Maier (1896/1936), ii.ii 282; Ross (1924), i.lxxxix; (1949), 610; van Bennekom (1986),
1“2; Zaslawsky (1986), 247; Celluprica (1987), 169; Charles (2002), 112; Thom (2002), 298). On
˜‰pl¤v™“˜ti™ see Int. 11, 21a 18“33; APr. 1.38, 49a 27“8; Top. 2.11, 115b 11“35; Rh. 2.24, 1402a 2“4. Aristotle
also distinguishes ˜being simply™ from ˜being particularly™ (e²nai –pª m”rouv) (see APo. 2.2, 89b 39;
90a 2 “ cf. SE 5, 166b 38; Mignucci (1975b), 231; Goldin (1996), 18“19), ˜being™ (e²nai) from ˜being this™
(todª e²nai) (see APo. 1.10, 76b 6), and ˜not being at all™ (mŸ e²nai ‚lwv) from ˜not being something™
(mŸ e²na© ti) (see Ph. 1.3, 186b 9“10). Some commentators (e.g. G. E. L. Owen (1960), 165; Kahn
(1966), 249, 263) think that Aristotle™s contrast between ˜being simply™ and ˜being something™ is not
between the existential and the predicative use of ˜to be™.
3 APo. 2.1, 89b 23“35; 2, 89b 36“90a 23; 8, 93a 16“20 (cf. 1.1, 71a 26“7). I am following the traditional
interpretation of these passages (cf. Grote (1880), 238; Calogero (1927/68), 29; Ross (1949), 609“10;
Bolton (1976), 517; Kahn (1966), 264“5; Miller (1971), 59; Mignucci (1975b), 231; Kahn (1976), 326“
7; Hintikka (1986), 85; Demoss/Devereux (1988), 134; Upton (1991), 316; McKirahan (1992), 122“3;
Barnes (1994), 203“4; Goldin (1996), 17, 41“2, 52“8; Hintikka (1999), 793“4). For a different (but less
convincing) reading see G´ mez-Lobo (1980) (criticised by Hintikka (1986), 87“9 and Goldin (1996),
o
17, 52“8). According to Kahn (1973c), 7, the earliest examples in Greek of existential assertions of the
form ˜There are t™s™ are from the period of the Sophists, and they always concern the existence of
gods.
4 Int. 10, 19b 14“31. Most commentators regard ˜A man is™, ˜A man is not™, etc. as existential assertions,
and ˜A man is just™, ˜A man is not just™, etc. as predicative assertions (see DuLac (1949), 166; Prior
(1955/62), 164“5; Ackrill (1963), 142; Mignucci (1975b), 242; Weidemann (1994/2002), 331“2; Whitaker
(1996), 135“7; D. Frede (1998), 89).
Truth conditions for existential assertions 101
(1051b 17“1052a 4) Aristotle addresses true and false thoughts and linguistic
expressions concerning non-composite items:
T 23 But then, with regard to non-composite items, what are ˜to be™ or ˜not to be™
and truth and falsehood? For they are not composite, so as ˜to be™ when they
are combined and ˜not to be™ if they are divided, like the log™s being white
and the diagonal™s being incommensurable, nor will truth and falsehood hold
still in the same sense as in the case of those. Or, just as truth in the case of
these is not the same, so also ˜being™, but truth and falsehood are “ well, to
touch and to enunciate is true5 (for af¬rmation and enunciation are not the
same thing), while to be ignorant is not to touch, for it is not possible to err
with regard to the ˜what it is™ except incidentally. Similarly also with regard
to non-composite substances,6 for it is not possible to err. And they are all in
actuality, not in potentiality: for they would come to be and cease to be, but
now being itself does not come to be or cease to be (for it ought to come to
be from something). Thus, with regard to those items which are just what it
is to be something and actualities7 it is not possible to err, but either to think
or not to think: however, with regard to them, one does investigate the ˜what
it is™,8 whether they are such-and-such9 or not. By contrast, with ˜being™ as
truth and ˜not being™ as falsehood, one is true if it is composed, while it is
false if it is not composed, the other is if it is a ˜being™ in this sense, while if
it ˜is not™ in this sense it is not.10 But truth is thinking of these things, while
falsehood does not exist, nor does error, but there is ignorance, which is not


At 1051b 23“4 I read ˜t¼ m•n ˆlhq•v £ ye“dov “ t¼ m•n qige±n kaª j†nai ˆlhq”v™: punctuation apart,
5
this is the reading handed down by the main manuscripts, presupposed by William of Moerbeke™s
translation, and printed by Ross and Jaeger. Ps.-Alexander (in Metaph. 599, 27) and some more
recent manuscripts have ˜t¼ m•n ˆlhq•v t¼ d• ye“dov, t¼ m•n qige±n kaª j†nai ˆlhq”v™, the reading
printed by Bekker, Schwegler, Bonitz, and Tredennick, and endorsed by Cassirer (1932), 160. Christ™s
extensive emendation (˜t¼ m•n ˆlhq•v [£ ye“dov t¼ m•n] qige±n kaª j†nai [ˆlhq”v]™) is endorsed
by several translators and commentators (e.g. Bonitz (1890), 211; Maier (1896/1936), i 7; Calogero
(1927/68), 24). The initial ˜m”n™ at 1051b 23 is answered by the ˜d”™ at 1051b 33 (cf. Ross (1924), ii 276;
Engelhardt (1953), 40).
At 1051b 27 I read ˜t‡v mŸ sunqet‡v oÉs©av™ with Ab and all the editions I consulted. Variants:
6
˜t‡v sunqet‡v oÉs©av™ (J), ˜t‡v sunq”touv oÉs©av™ (E™s ¬rst hand, ˜mŸ™ is added by the second
hand). William of Moerbeke™s translation also presupposes the absence of ˜mŸ™. The readings ˜t‡v
sunqet‡v oÉs©av™ and ˜t‡v sunq”touv oÉs©av™ are dif¬cult to square with the following ˜They are
all in actuality, not in potentiality . . .™ (1051b 28“30).
7 At 1051b 31 I read ˜–n”rgeiai™ with Ross. All witnesses and most editors read ˜–nerge©a™.
At 1051b 32 I read ˜t¼ t© –sti zhte±tai™ with E, J, and most editions. Christ instead reads ˜t¼ ›sti
8
zhte±tai™ with Ab .
At 1051b 32“3 I read ˜e« toia“t† –stin™ with all the main manuscripts and most editions. Ps.-Alexander
9
(in Metaph. 600, 34) relies on the reading ˜oÉk e« toia“t† –stin™, which is adopted by some editors
and commentators (Bonitz (1848/49), ii 410“1; (1890), 211; Heidegger (1925/26), 177; von Fragstein
(1967), 147). For a defence of the reading of the main manuscripts see Maier (1896/1936), i 21; Ross
(1924), ii 278; Berti (1990), 102; (1992), 85“6; (1993), 22“3; (2000), 14“15.
At 1051b 35“1052a 1 I adopt Christ™s punctuation: ˜e­per ¿n oÌtwv, ›stin, e« d• mŸ oÌtwv, oÉk ›stin™.
10
Alternative punctuations: ˜e­per Àn, oÌtwv ›stin, e« d• mŸ oÌtwv, oÉk ›stin™ (Ross and most
editions); ˜e­per ¿n oÌtwv ›stin, e« d• mŸ oÌtwv, oÉk ›stin™ (Bonitz).
102 Bearers of truth or falsehood
like blindness: for blindness is as if one completely lacked the intellectual
faculty. (1051b 17“1052a 4)

Compare the following excerpt of de Anima 3.6:
T 24 The thinking of indivisible items concerns items about which there is no
falsehood. By contrast, where both falsehood and truth are present there is
already some sort of composition of thoughts as if they were a single one “
as, in the words of Empedocles, ˜where neckless heads of many sprouted™
afterwards they were composed by Love, so here too these items which
were apart are composed, e.g. incommensurable and diagonal “ and, if the
thinking concerns the past or the future, by composing and thinking in
addition of time. For falsehood always concerns composition: for, even if
of something white it thinks that it is not white, it composes non-white.11
(430a 26“430b 3)
An af¬rmation12 is something about something, as too is a denial,13 and each
one is either true or false. By contrast, not all intellect is such, but that which
is of the ˜what it is™ according to the ˜what it was to be™ is true, and it is not
something about something. However, just as the seeing of a special object
is true, but seeing that the white thing is a man or not is not always true, so
also with what is without matter. (430b 26“31)

T 23 and T 24 are close: T 23™s claim that ˜it is not possible to err with
regard to the “what it is” except incidentally™ (1051b 25“6) corresponds to T
24™s ˜intellect [. . .] which is of the “what it is” according to the “what it
was to be”™ and ˜is true™ (430b 28); T 23™s ˜non-composite items™ (1051b 17)


At 430b 2“3 I adopt the reading ˜‹n t¼ leuk¼n mŸ leuk¼n t¼ mŸ leuk¼n sun”qhken™, attested almost
11
unanimously by the manuscripts and printed by most recent editions. Several editors and commen-
tators (Trendelenburg (1833), 92, 502; Roeper (1852), 324, 768; Torstrik (1862), 93; Vahlen (1872),
427“8; Trendelenburg (1877), 84, 414; Maier (1896/1936), i 25; Theiler (1959), 60, 145; Ross (1961),
ad loc.; Oehler (1962/85), 156) emend the passage.
12 For ˜j†siv™ and ˜ˆp»jasiv™ meaning ˜af¬rmation™ and ˜denial™ see Bonitz (1870), 813a 19“23. For a
different use of ˜j†siv™ see Int. 4, 16b 27; 5, 17a 17; de An. 3.7, 431a 8; Metaph. 10, 1051b 25 (on which
cf. n. 52 below and the portion of the main text it pertains to).
At 430b 26“7 I read ˜›sti d¬ ¡ m•n j†siv ti kat† tinov, ãsper kaª ¡ ˆp»jasiv™ with Ross
13
(cf. Torstrik (1862), 196“8; Calogero (1927/68), 22“3; Hamlyn (1968/93), 145; Cashdollar (1973),
161; Horn (1994), 115“16). The words ˜ãsper kaª™ are only in W: the remaining manuscripts
have ˜ãsper™. Moreover, ˜ˆp»jasiv™ is Torstrik™s emendation: all manuscripts have ˜kat†ja-
siv™. The corruption of ˜ SPERKAIHAPOFASIS™ into ˜ SPERHKATAFASIS™ could easily have
occurred.
The text handed down by most manuscripts could be defended if ˜j†siv™ could be taken to mean
˜predicative assertion™ (one would then take ˜ãsper™ to mean ˜as for instance™: for this usage cf. APo.
2.10, 93b 36; 11, 94b 32; 95a 1; 16, 98a 36“7; Pacius (1611), 388; Bullinger (1882), 52; Horn (1994), 115“16).
Some commentators (e.g. Torstrik (1862), 196; Rodier (1900), ii 488; Bod´us (1993), 233“4) claim

that Aristotle elsewhere uses ˜j†siv™ to mean ˜predicative assertion™, but their evidence (Int. 12, 21b 19)
is far from clear cut.
Truth conditions for existential assertions 103
correspond to T 24™s ˜indivisible items™ (430a 26).14 I shall assume that T
23 and T 24 expound essentially the same theory. The bulk of the present
section is dedicated to explaining what this theory is.

Beliefs and assertions concerning non-composite items as existential beliefs and
10 (1051b 17“1052a 4 = T 23) Aristotle
assertions. In the central part of
continues the train of thought of the chapter™s ¬rst part (1051a 34“1051b 17 =
T 2 + T 6): he extends to non-composite items his approach to composite
items. Recall:15 an af¬rmative (negative) belief, or assertion, concerning a
composite item posits that this composite item ˜is™ in the sense of being true
(˜is not™ in the sense of being false), and it is true when and only when
the composite item in fact ˜is™ in the sense of being true (˜is not™ in the
sense of being false). Now, analogously, an af¬rmative (negative) belief, or
assertion, concerning a non-composite item posits that this non-composite
item ˜is™ in the sense of being true (˜is not™ in the sense of being false), and
it is true when and only when the non-composite item ˜is™ in the sense of
being true (˜is not™ in the sense of being false). But there is a difference. For
a composite item (like a state of affairs) ˜to be™ in the sense of being true is
for it to be combined, i.e. for its components to be reciprocally combined.
However, since a non-composite item lacks components, for it ˜to be™ in
the sense of being true cannot be for its components to be reciprocally
combined “ it must be something different. Aristotle™s view seems to be
that for a non-composite item ˜to be™ in the sense of being true is simply
to exist, i.e. ˜to be there™.16 This is what Aristotle seems to say near the end
of T 23, as the following expansion of my translation shows:
With ˜being™ as truth and ˜not being™ as falsehood, one [a composite item] is true
if it is composed, while it is false if it is not composed, the other [a non-composite
item] is [exists] if it is a ˜being™ in this sense [if it ˜is™ in the sense of being true],17
while if it ˜is not™ in this sense [if it ˜is not™ in the sense of being false] it is not [does
not exist]. (1051b 33“1052a 1)
It follows that beliefs and assertions concerning non-composite items are
existential beliefs and assertions. Note that this does not imply that every
14 Cf. the ˜simple items™ of Metaph. E 4, 1027b 27“8.
15 I am summarising some results of sect. 1 of ch. 2.
16 Sometimes Aristotle uses ˜true™ and its cognates to say of something that it exists or that it is real:
see Ph. 1.8, 191a 25; 8.8, 263a 18; Cael. 1.5, 271b 6; 271b 9; 3.1, 298b 13; Metaph. A 3, 983b 3; 7, 988a 20;
a 1, 993a 30; 993b 17; MM 1.1, 1182a 29; EE 1.4, 1215b 2; Rh. 1.7, 1364b 9; Protr. fr. 73 Gigon 304a 20;
304b 13; 310a 27; 310b 14; 310b 20 (= Iamb. Protr. 38, 2; 39, 9; 54, 4; 55, 2; 55, 7). Some commentators
(e.g. Wilpert (1940), 5) assume that at Metaph. a 1, 993b 19“31 ˜true™ and ˜truth™ mean ˜existent™ and
˜existence™, but B¨rthlein (1972), 25“33 shows that such an interpretation is questionable.
a
17 Cf. Metaph. E 4, 1027b 29; 1027b 31; Wilpert (1940), 12; de Rijk (1952), 23.
104 Bearers of truth or falsehood
existential belief or assertion should concern a non-composite item: there
may well be existential beliefs and assertions concerning composite items.18

An example of an existential assertion concerning a non-composite item. Sup-
pose that for Aristotle the universal man is a non-composite item. The truth
conditions for the existential assertion ˜Man exists™ will then resemble those
described by this passage from Categories 12:
T 25 Among things which convert as to implication of being, what is in whatever
way a cause of the being of the other might reasonably be called ˜prior™ by
nature.19 And that there are some such cases is clear: for that man is converts
as to implication of being with the true sentence about it (for if man is, the
sentence whereby we say that man is is true, and this is sure to convert: for
if the sentence whereby we say that man is is true, then man is), but the true
sentence is in no way the cause of the object™s being, while the object seems
somehow the cause of the sentence™s being true: for it is because the object
is or is not that a sentence is called true or false. (14b 11“22)
The interpretation of T 25 raises at least three questions. What do the occur-
rences of ˜object™ at 14b 19“21 denote? How should the crucial occurrences of
˜to be™ be understood? What does the occurrence of ˜it™ in ˜the true sentence
about it™ (14b 15) refer to? The foregoing subsection™s considerations suggest
the following (reciprocally connected) answers: the occurrences of ˜object™
at 14b 19“21 denote those (composite or non-composite) objects that are
crucial to the truth or falsehood of assertions concerning them; the crucial
occurrences of ˜to be™ should be understood as expressing the ˜being™ in the
sense of being true discussed in Metaphysics 10; the occurrence of ˜it™ in
˜the true sentence about it™ (14b 15) refers to the species man. Suppose these
answers are correct. Then in T 25 Aristotle ¬rst claims that ˜among things
which convert as to implication of being, what is in whatever way a cause
of the being of the other might reasonably be called “prior” by nature™
(14b 11“13). He then illustrates this claim by mentioning the species man
and the assertion ˜Man is™: (i) the species man ˜is™ in the sense of being true,
i.e. exists, when and only when the assertion ˜Man is™ is true (cf. 14b 14“18),
(ii) the fact that the assertion ˜Man is™ is true is in no way the cause of the

18 My interpretation of 10™s central part (= T 23) derives from Brentano: see his (1874/1925), ii
54“5; (1889), 20“1; (1914), 164; (1915b), 136; (n.d.), 191. Brentano™s interpretation was forgotten by
most twentieth-century commentators (there are exceptions: see Calogero (1927/68), 25, 27“9; von
Iv´nka (1932), 19; Sillitti (1966), 321“4; Sadun Bordoni (1994), 74; Fattal (1996), 434“5, 439“40). In
a
his (1862), 26“7 Brentano favoured a different exegesis of T 23, more in line with the Scholastic
tradition (cf. n. 58 below with the paragraph it pertains to; Krell (1975), 82).
19 This priority ˜by nature™ (˜j…sei™) seems different from the priority ˜by nature and being™ (˜kat‡
j…sin kaª oÉs©an™) of Metaph. 11, 1019a 2“4 (which excludes ˜conversion as to implication of
being™).
Truth conditions for existential assertions 105
fact that the species man ˜is™ in the sense of being true (cf. 14b 18“19), but
(iii) the fact that the species man ˜is™ in the sense of being true is somehow
the cause of the fact that the assertion ˜Man is™ is true (cf. 14b 19“20). Finally,
Aristotle makes a general claim which backs up points (ii) and (iii) of his
example: it is because the object ˜is™ in the sense of being true or ˜is not™ in
the sense of being false that an assertion is true or false, not vice versa (cf.
14b 21“2).20

All the non-composite items mentioned in ˜ 10 exist always. The non-
composite items mentioned by Aristotle in Metaphysics 10 are essences
and ˜non-composite substances™.21 For Aristotle, essences are natural kinds.22
Moreover, several passages show that in Aristotle™s view at least some natural
kinds exist always.23 So Aristotle can be plausibly credited with the view
that essences exist always. As I shall show (in section 2 of this chapter), the
˜non-composite substances™ mentioned in 10 are probably incorporeal
substances, i.e. God and (perhaps) the intellects that move the heavenly
spheres. In Aristotle™s view, incorporeal substances also exist always. So
Aristotle can be plausibly credited with the view that all the non-composite
items he mentions in 10 exist always.

Af¬rmative beliefs and assertions concerning non-composite items are always
true. In an earlier subsection I suggested that for a non-composite item
˜to be™ in the sense of being true is for it to exist. In the last subsection I
argued that all the non-composite items mentioned in 10 exist always.
If I am right, then all the non-composite items mentioned in 10 always
˜are™ in the sense of being true. This has two important consequences: ¬rst,
every af¬rmative belief, or assertion, concerning one of these non-composite
items is always true (because it concerns an item that always ˜is™ in the sense
of being true); second, every negative belief, or assertion, concerning one
of these non-composite items is always false (again because it concerns an
item that always ˜is™ in the sense of being true). I take it that Aristotle is in
fact committed to these consequences.

The infallibility of af¬rmative beliefs concerning non-composite items. As we
have just seen, Metaphysics 10 can be plausibly taken to commit Aristotle

20 I disagree with those commentators (Hamlyn (1962), 194; Nuchelmans (1973), 33“4; de Rijk (1987),
34“5; Gaskin (1998), 42“5) who take T 25 to be dealing exclusively with states of affairs.
21 For essences see 1051b 25“6; 1051b 30“2 (cf. de An. 3.6, 430b 27“31; Metaph. E 4, 1027b 27“8). For ˜non-
composite substances™ see 1051b 27. In the central part of de Anima 3.6 (430b 6“26) Aristotle discusses
non-composite items of other kinds.
22 23 See n. 10 of ch. 2 and the passage it pertains to.
Cf. n. 6 of the introduction.
106 Bearers of truth or falsehood
to the claim that every af¬rmative belief concerning a non-composite item
is always true. It is probably this everlasting truth of af¬rmative beliefs
concerning non-composite items that Aristotle has in mind when in 10
he remarks (i) that ˜truth is touching™ (1051b 24), (ii) that ˜to be ignorant
is not to touch, for it is not possible to err with regard to the “what it is”
except incidentally™ (1051b 25“6), and (iii) that ˜truth is thinking of these
items™ (1052a 1). In these three remarks, Aristotle focuses on an essential
aspect of af¬rmative beliefs concerning non-composite items, and he claims
that this essential aspect of such beliefs guarantees their everlasting truth.
(Aristotle leaves unmentioned the everlasting falsehood of negative beliefs
and assertions concerning non-composite items.) Here are some detailed
comments on Aristotle™s three remarks.
(i) ˜Touching™ is an essential aspect of every af¬rmative belief concerning
a non-composite item: every such belief must ˜touch™ the non-composite
item it concerns.24 But this guarantees that the af¬rmative belief in question
should be always true (because the non-composite item it concerns always
exists, and therefore always ˜is™ in the sense of being true). In this sense,
˜truth is touching™ (1051b 24): the everlasting truth of an af¬rmative belief
concerning a non-composite item is guaranteed by that essential aspect of
it which is its touching its non-composite item.
(ii) As we have just seen, ˜touching™ is an essential aspect of every af¬r-
mative belief concerning a non-composite item. Not to touch is to be
ignorant,25 i.e. to be in the blank condition of lacking any object to think
about.26 When one has a belief about a ˜what it is™, i.e. an essence, which
is a non-composite item, one cannot err ˜except incidentally™ (1051b 26), i.e.
24 For the thought“touch comparison cf. de An. 2.9, 421a 19“26 (where Aristotle claims that the accuracy
of touch is proportional to that of thought, cf. Freeland (1992), 234); 3.3, 427b 4; 8, 432a 1“3; Metaph.
L 7, 1072b 20“1; Pl. Phd. 79d6; Tht. 186d3“4; 190c6“7; 190d9“10; 209a8; Smp. 212a4“5; Phdr. 273a4;
Ti. 37a6; 52b2; Lg. xii 965b4; 969b6; Speus. fr. 30 Lang = fr. 35 Isnardi (= Procl. in Euc. 179, 18);
Thphr. Metaph. 9b 13“6 (with van Raalte (1993), 456“9); Plot. 1, 1, 9, 12“14; 5, 3, 10, 43; 5, 3, 17, 24“7;
Ross (1923), 25; Rosen (1961), 129“32. Aristotle also compares thought with sight (Top. 1.17, 108a 11;
de An. 3.5, 430a 14“25; Metaph. a 1, 993b 9“11; 10, 1052a 1“4; EN 1.4, 1096b 28“9), hearing (APo.
1.10, 76b 36“8), and perception (de An. 1.5, 410a 25“6; 3.3, 427a 17“29; 4, 429a 13“18; 429a 23“429b 5; 7,
431a 8; 8, 431b 26“432a 3; EN 6.12, 1143b 5, cf. Pl. Tht. 202b5“6).
See ˜ˆgnoe±n™ at 1051b 25 and ˜Šgnoia™ at 1052a 2, cf. ˜mŸ noe±n™ at 1051b 32.
25
26 Cf. Maier (1896/1936), i 7, 19“20; Ross (1924), ii 279; De Corte (1934), 238; Joachim (1948), 26;
Whitaker (1996), 29. At 1052a 2“4 Aristotle says that ignorance is unlike blindness because blindness
would correspond to having no intellect. The fact that he feels the need to emphasise that ignorance
and blindness are different indicates that in his view they resemble each other in important respects “
otherwise he would not think that somebody could be tempted to confuse them. The resemblance
is probably that ignorance is like not seeing when one can see (in this respect ignorance is unlike
blindness, which is the incapacity to see). But the analogy of ignorance with not seeing when one can
see suggests that ignorance should be the ˜blank™ condition of lacking any object to think about. A
kindred usage of ˜ignorance™ was probably displayed in a passage of Aristotle™s lost work On Opposites
reported or summarised in Simplicius™ commentary on the Categories (390, 19“25 < fr. 625 Gigon
726b 9“19).
Truth conditions for existential assertions 107
except insofar as the essence is treated as ˜being incidental to™ something or
something is treated as ˜being incidental to™ it,27 i.e. insofar as the essence
is treated as a predicate, i.e. as holding of something, or as a subject, i.e.
as having something holding of it.28 The following considerations support
this interpretation of the restrictive clause ˜except incidentally™ at 1051b 26.29
(ii.i) At the end of de Anima 3.6 Aristotle says:
T 26 An af¬rmation is something about something, as too is a denial, and each
one is either true or false. By contrast, not all intellect is such, but that which
is of the ˜what it is™ according to the ˜what it was to be™ is true, and it is not
something about something. However, just as the seeing of a special object
is true, but seeing that the white thing is a man or not is not always true, so
also with what is without matter. (430b 26“31 < T 24)
A ˜“what it is” according to the “what it was to be”™ (430b 28) is an essence.
Since essences do not contain matter, the phrase ˜what is without matter™
(430b 30“1) probably refers to essences.30 Hence, T 26 compares the truth
or falsehood of a belief concerning an essence with the truth or falsehood
of a perception of a special object.

27 Aristotle sometimes explains what it is for a certain event to happen ˜incidentally™ by appealing to the
circumstance that in it something ˜is incidental to™ something: see APo. 1.19, 81b 24“9; 22, 83a 1“23;
Top. 2.3, 110b 17“25; Ph. 5.1, 224a 21“3; Metaph. A 1, 981a 18“20.
Aristotle sometimes uses ˜to be incidental to™ (˜sumbebhk”nai™ + dat.) as equivalent to ˜to be pred-
28
icated of ™: see Top. 2.3, 110b 24“5 (with Brunschwig (1967), 141“2); SE 5, 166b 28“36 (with Hamblin
(1970), 84“6 and Dorion (1995), 233); 6, 168a 40“168b 2; PA 1.5, 645a 36“645b 2; Metaph. A 1, 981a 18“20;
Z 5, 1030b 20“1; N 1, 1088a 17“18.
29 Cf. Fonseca (1577/89), iii 668; Suarez (1597), xxviii; Maurus (1668), iv 480“1; Oehler (1962/85),
214“15; Kessler (1976), 180; Berti (1978), 149; Seidl (1989/91), ii 493; Berti (1993), 24“5; Vigo (1997),
39“40; Berti (2000), 12; Charles (2000), 136. Other commentators interpret differently the restrictive
clause ˜except incidentally™ at 1051b 26.
According to some (e.g. [Alex. Aphr.] in Metaph. 600, 16“17; [Phlp.] in Metaph. 39va ; Bonitz
(1848/49), ii 411; Grote (1880), 618; von Fragstein (1967), 147, 148), ˜incidentally™ here means ˜in an
improper sense™ (cf. Cat. 6, 5a 38“5b 10; Ph. 2.3, 195b 3“4; Metaph. 2, 1014a 7“8; I 1, 1052a 17“19; Waitz
(1844/46), i 442“3). Aristotle is therefore saying that it is only in an improper sense of ˜to err™ that
one can say that it is possible to err about the ˜what it is™.
According to other commentators (e.g. Aquinas in Metaph. 1908 Cathala/Spiazzi; Maier
(1896/1936), i 21; Ross (1924), ii 277; Keeler (1932), 245“6; De Corte (1934), 239; Tricot (1966),
ii 524; Grayeff (1974), 209), ˜incidentally™ here means ˜as a coincidence™. Aristotle is saying that
error about a simple item may coincidentally occur because the simple item itself (although simple
insofar as not exhibiting predicative composition) is complex insofar as composed of its genus and
its differentia: an error can occur in the attempt to explain the simple item in a de¬nition, but the
error cannot be imputed to the thought whereby one ˜touches™ the simple item, and it is therefore
˜incidental™ to this thought.
Yet other commentators (e.g. Grayeff (1974), 209; Volkmann-Schluck (1979), 272“4; Burnyeat
et al. (1984), 159) offer different interpretations which cannot be discussed here.
30 Cf. Hamlyn (1968/93), 145; Movia (1979), 387; Fattal (1996), 430. Other commentators (e.g. [Phlp.]
in de An. 557, 9“10 Hayduck; Aquinas in de An. 762 Pirotta) take the phrase ˜what is without matter™
(430b 30“1) to refer to intellects, which have no bodily organ (see de An. 3.4, 429a 18“27), yet others
(e.g. Pacius (1611), 388) take it to refer both to intellects and to essences.
108 Bearers of truth or falsehood
Since T 26 compares the truth or falsehood of a belief concerning an
essence with the truth or falsehood of a perception of a special object, it is
worthwhile considering here, albeit brie¬‚y, Aristotle™s views on the truth or
falsehood of perceptions of special objects. In de Anima 2.6 (418a 7“25) Aris-
totle distinguishes several kinds of objects of perception: ¬rst he separates
objects perceptible ˜in themselves™ from objects perceptible ˜incidentally™;
then, within objects perceptible in themselves, he separates ˜special™ from
˜common™ objects. A special object can be perceived by only one percep-
tual faculty. Thus, colours are special objects of perception because they
can be perceived by only one perceptual faculty, i.e. sight; similarly with
sounds with respect to hearing, odours with respect to smell, ¬‚avours with
respect to taste, and properties of temperature and solidity with respect
to touch.31 A common object is properly perceived by two or more per-
ceptual faculties32 collaborating with one another (for this reason it is also
described as an object of the faculty of perception as a whole, or ˜common
perception™).33 Movement, rest, number, ¬gure, and magnitude are among
Aristotle™s examples of common objects of perception.34 In a perception of
an incidental object one perceives something on the basis of an ˜interpre-
tation™ of a special object of perception, which one perceives in the strict
sense.35 For example, suppose one sees an instance of whiteness, which is
a special object of perception, and ˜interprets™ it in such a way that one
can be described as seeing Diares™ son: Diares™ son is then an incidental
object of perception. (The expressions ˜interpretation™ and ˜interprets™ here
must be handled with care because the operation involved in perceptions
of the sort described is often “ though not always “ subconscious, while
˜interpretation™ and ˜interprets™ tend to be reserved for conscious opera-
tions.)36 Aristotle thinks that some perceptions are true, other false. This
view could seem very strange if Aristotle thought that perceptions do not
have propositional content. However, Aristotle probably thinks that at least

31 De An. 2.6, 418a 9“14; Somn. Vig. 2, 455a 12“15.
32 In several passages (de An. 2.6, 418a 9“11; 418a 16“20; 3.1, 425a 27“30; 425b 4“11) Aristotle claims that
common objects are perceived by all perceptual faculties. On one occasion (Sens. 4, 442b 4“10) he
quali¬es this claim by saying that common objects are perceived by sight and touch.
33 Cf. Modrak (1987), 63“5, 68.
34 De An. 2.6, 418a 17“18; 3.1, 425a 15“16; Mem. 1. 450a 9“12; Insomn. 1. 458b 4“6.
35 Cf. de An. 2.6, 418a 24“5.
36 In ¬ve passages Aristotle uses ˜incidental™ and ˜to be incidental™ in talking of perceptions of incidental
objects: de An. 2.3, 414b 6“11; 2.6, 418a 8“25; 3.1, 425a 13“425b 11; 3.3, 428b 18“30; Metaph. 5, 1010b 14“
26. Aristotle, however, discusses perceptions of incidental objects also elsewhere: de An. 3.6, 430b 29“
31; Insomn. 1. 458b 31“3. Other relevant passages are APr. 1.27, 43a 33“6 and Metaph. M 10, 1087a 19“20.
My interpretation of Aristotle™s views on perceptions of incidental objects is close to Modrak (1987),
69“71, 77.
Truth conditions for existential assertions 109
some perceptions do have propositional content. For example, in de Anima
2.6 he says that a perception of a special object ˜does not err that there is
colour nor that there is sound™ (418a 15).37 Again, in de Anima 3.3 he says
that sight ˜is not mistaken that there is whiteness™ (428b 21).38 Aristotle™s
position is very reasonable: a dog having a perceptual experience involving
whiteness can be appropriately described as perceiving that there is white-
ness “ after all, the purpose of perception is to inform the animal of what
is there in the environment. Even when Aristotle expresses himself as if he
were saying that a perception is of a ˜thing™ (e.g. of whiteness), he probably
means that one is perceiving that the ˜thing™ is there.39 Aristotle believes that
perceptions of objects of different types are exposed to error in different
degrees: in his view, almost every perception of a special object is true,40
while perceptions of common and incidental objects are relatively often
false.41 Aristotle probably thinks that perceptions of special objects are false
only on those rare occasions when their circumstances are abnormal, e.g.
when the perceptual faculty is biased because of a recent intense perception
or damaged by illness, or when the medium has distorting effects.42 Aristo-
tle™s main reason for thinking that almost all perceptions of special objects
are true is probably that these perceptions are, or are based on, processes of
assimilation which in ordinary circumstances are accurate.43 According to
some commentators,44 Aristotle thinks that perceptions of special objects

37 Here I take the two occurrences of ˜‚ti™ as declarative and I understand an existential ˜›sti™ both
after ˜cr¤ma™ and after ˜y»jov™.
38 Here I take ˜‚ti™ as declarative and I understand an existential ˜›sti™ after ˜leuk»n™: cf. Everson (1995),
268; Cassin (1996), 286“7; Fattal (1996), 434, 439; Everson (1997), 190.
39 Cf. Cashdollar (1973), 161“7; Ben-Zeev (1984), 119, 122; Sorabji (1992), 196“203; (1996), 314“15;
Cassin (1996), 286“8; Birondo (2001), 62“4. According to other commentators (e.g. Phlp. in de
An. 88, 68“72 Verbeke) Aristotle thinks that perceptions do not have propositional content, which
requires some faculty superior to perception. Note that someone who denies that perceptions might
have propositional content could still claim that perceptions were true or false (cf. Block (1961), 8;
Ben-Zeev (1984), 119; Caston (1996), 44“5; Charles (2000), 121): one could claim that a vision of
whiteness is true just in case the object from which it is produced is white, and, in general, endorse
every instance of the schema ˜A perception of ness is true just in case the object from which it is
produced is ™.
40 De An. 3.3, 428b 18“19. Elsewhere (de An. 2.6, 418a 11“16; 3.3, 427b 11“12; 6, 430b 29; Sens. 4, 442b 8“10;
Metaph. 5, 1010b 1“26) Aristotle is more con¬dent “ he simply says that perceptions of special
objects are true. Once (de An. 3.3, 428a 11) he even says that perceptions ˜are always true™ (is he using
˜perception™ as a success-expression?).
41 De An. 3.3, 428b 19“25, cf. 2.6, 418a 8“25; 3.1, 425a 30“425b 4; 3, 428b 25“30; 6, 430b 29“31; Sens. 4,
442b 8“10; Insomn. 1, 458b 31“3; Metaph. 5, 1010b 14“26.
42 Cf. de An. 2.9, 421a 9“26; 10, 422b 5“10; 3.4, 429a 29“429b 5; Insomn. 2, 459b 5“23; PA 2.2, 648b 12“17;
GA 5.1, 779b 12“780a 25; Pr. 31.7, 958a 24“7; 11, 958b 11“15; Metaph. 5, 1010b 1“26; 6, 1011a 25“1011b 1;
K 6, 1062b 35“1063a 10; EN 3.6, 1113a 25“31; 10.2, 1173b 20“5; Kenny (1967), 193; Long (1981), 101;
Ben-Zeev (1984), 118; Vasiliou (1996), 127; Charles (2000), 118“19.
43 44 E.g. Irwin (1988), 314.
Cf. Burnyeat (2002), 45.
110 Bearers of truth or falsehood
are infallible because an object must have the perceptual quality with which
it appears to a perceiver. This interpretation must be wrong: otherwise Aris-
totle could not avoid the consequence (which, however, he rejects) that per-
ceptions of common objects also are infallible because an object must have
the perceptual quality with which it appears to a perceiver.45 Perceptions of
common objects are often false because they require the collaboration of
different perceptual faculties: this collaboration can easily go wrong.46 Per-
ceptions of incidental objects are often false because the ˜interpretation™ of
the special object perceived is ¬‚awed. In terms of Aristotle™s example, one™s
perception of Diares™ son may well be false because one™s ˜interpretation™
of the instance of whiteness one sees is ¬‚awed: Diares™ son is not inciden-
tally identical to (˜incidental to™) the individual that has the whiteness one
sees.
There are two aspects to T 26™s comparison of the truth or falsehood of
a belief concerning an essence with the truth or falsehood of a perception
of a special object: ¬rst, as almost every perception concerning a special
object (e.g. a case of seeing that there is whiteness) is true, so every belief
concerning an essence is true; second, as one™s perception of an incidental
object can be false because the object one perceives incidentally is not
incidentally identical to the individual which has the special object one
perceives (e.g. one™s perception of Diares™ son can be false because Diares™
son is not incidentally identical to the individual that has the whiteness one
sees), so can one falsely think that a certain essence holds of something.47
Since T 26 is closely connected to the central part of Metaphysics 10, T
26™s comparison endows with plausibility the suggested interpretation of
the restrictive clause ˜except incidentally™ at 1051b 26, i.e. the interpretation
according to which this restrictive clause is saying that error with regard to
an essence is possible insofar as the essence is treated as holding of something
or as having something holding of it.
(ii.ii) A few lines after saying that ˜it is not possible to err with regard to
the “what it is” except incidentally™ (1051b 25“6), Aristotle says:
With regard to those items which are just what it is to be something and actualities
it is not possible to err, but either to think or not to think: however, with regard
to them, one does investigate the ˜what it is™, whether they are such-and-such or
not. (1051b 30“3)

45 46 Sens. 4, 442b 4“10, cf. Modrak (1987), 79.
Cf. Block (1961), 3; Vasiliou (1996), 115, 117.
47 Cf. Phlp. in de An. 88, 66“78 Verbeke; [Phlp.] in de An. 553, 30“554, 2; 557, 1“6 Hayduck; Block
(1961), 2; Moreau (1961), 24; Cashdollar (1973), 161“2, 172; Movia (1979), 313, 387; Charles (2000),
135“6, 138.
Truth conditions for existential assertions 111
This remark makes it plausible to assume that one of the enterprises con-
cerning essence for which Aristotle allows ˜incidental™ error should be the
inquiry into the ˜what it is™ of ˜those items which are just what it is to be
something™, i.e. the search for a de¬nition of an essence.48 A de¬nition of
an essence is a predicative assertion where the subject is the de¬niendum
and the predicate is the de¬niens.49 It is because of this predicative structure
of de¬nitions that the error concerning essence which sometimes occurs
in them is described by Aristotle as ˜incidental™. One remarkable aspect of
Aristotle™s position is that having a true existential belief about an essence
does not automatically enable one to formulate correctly its de¬nition:
despite having a true belief about the essence, one must investigate its
essential characteristics (its ˜what it is™), and this investigation may well
result in a falsehood. This aspect of Aristotle™s position ¬ts well with his
choice of the image of touching: as touching a thing falls short of grasping
it, so referring to an essence in a true existential belief about it does not
automatically enable one to de¬ne that essence. It is worthwhile observing
that this aspect of Aristotle™s position seems correct “ philosophers keep dis-
cussing and disagreeing about the essential characteristics of many items.50
I suspect, however, that the search for an essence™s de¬nition is only one
of several cases where ˜incidental™ error concerning the essence is possible:
Aristotle™s position on this issue is probably that an ˜incidental™ error con-
cerning an essence occurs whenever the essence is somehow involved in a
false predicative belief (e.g. if I falsely believe that every man is uneducated
then I make an ˜incidental™ error about the essence man, with regard to
which I have a true existential belief ). Aristotle™s mention of the search for
the de¬nition of an essence immediately after his mention of an existen-
tial belief about an essence ¬ts well with a view he puts forward in the


48 Cf. Oehler (1962/85), 236. For a de¬nition ˜revealing™ or ˜signifying™ the essence see Top. 1.4, 101b 21“2;
5, 101b 39; 8, 103b 9“10; 7.3, 153a 15“16; 5, 154a 31“2; Metaph. 6, 1016a 33“4; 1017a 6; 8, 1017b 21“2; Z 5,
1031a 12; H 1, 1042a 17.
49 Cf. the paragraph to which n. 62 below is appended.
50 Cf. Vigo (1997), 32“3. Some commentators (e.g. Sorabji (1980), 218“19; (1981), 243; (1982), 298;
Denyer (1991), 204; Galluzzo (1997/98), 54) attribute to Aristotle the view that one cannot err in
the search for de¬ning characteristics. The key idea is that what might appear as a mistake in one™s
search for the de¬ning characteristics of x is really a correct step in one™s search for the de¬ning
characteristics of y, which is what one really had in mind: if Mrs Malaprop says that allegories are a
certain species of reptile, she is not erring in the search for the de¬ning characteristics of an allegory “
rather, she is making a correct step in the search for the de¬ning characteristics of what she really
had in mind, i.e. alligators. This position, however, is untenable. It is implausible to say that the
de¬nitions proposed by Theaetetus in Plato™s homonymous dialogue are correct because they give
the de¬ning characteristics (not of knowledge, but) of the items Theaetetus really had in mind.
Rather, Theaetetus seems to have made at least three mistakes in his attempt to de¬ne knowledge.
112 Bearers of truth or falsehood
Posterior Analytics, i.e. that it is after discovering that something exists that
one proceeds to search for what it is.51
(iii) ˜Thinking of ™ is an essential aspect of an af¬rmative belief con-
cerning a non-composite item: in every af¬rmative belief concerning a
non-composite item one must ˜think of ™ that non-composite item. But this
guarantees the everlasting truth of the belief (because the non-composite
item in question always exists, and therefore always ˜is™ in the sense of being
true): in this sense ˜truth is thinking of these items™ (1052a 1), i.e. the ever-
lasting truth of an af¬rmative existential belief concerning a non-composite
item is guaranteed by that essential moment of it which is its thinking of
its non-composite item.

The infallibility of af¬rmative assertions concerning non-composite items.
When Aristotle says that ˜truth is [. . .] enunciating™ (1051b 24), what he has
in mind is that af¬rmative existential assertions concerning non-composite
items are always true. Aristotle is again focusing on an essential aspect of
af¬rmative existential assertions concerning non-composite items, and he
is claiming that this essential aspect of them suf¬ces to guarantee their
everlasting truth. ˜Enunciation™ is the semantic task performed by names
within assertions,52 and is therefore the semantic task performed by the
subject of an af¬rmative existential assertion concerning a non-composite
item: in any af¬rmative existential assertion concerning a non-composite
item the subject must ˜enunciate™ the non-composite item. But this guaran-
tees everlasting truth (because the non-composite item in question always
exists, and therefore always ˜is™ in the sense of being true): in this sense
˜truth is enunciating™.

There is no need that all existential af¬rmative assertions or beliefs should
always be true. Aristotle can be plausibly credited with the claim that every
assertion (belief ) concerning a non-composite item is an existential asser-
tion (belief ). This however does not commit him to the converse claim that
every existential assertion (belief ) should be an assertion (belief ) concerning
a non-composite item. Since, in Aristotle™s view, it is af¬rmative assertions
(beliefs) concerning non-composite items that are always true, Aristotle is
not committed to the obviously false claim that every existential af¬rmative
assertion (belief ) should be always true: he could consistently acknowledge
that some existential af¬rmative assertion (belief ) is sometimes false “ but

51 APo. 2.1, 89b 33“5; 2, 89b 36“90a 23; 90a 31“4.
52 Int. 4, 16b 26“30; 5, 17a 17“19 (cf. Torstrik (1862), 196“7; Ross (1923), 60).
Truth conditions for existential assertions 113
he will have to grant that such an existential af¬rmative assertion (belief )
concerns a composite item. Indeed, in the ¬rst chapter of de Interpretatione
(at 16a 16“18) Aristotle considers the assertions ˜A goatstag is™ and ˜A goatstag
is not™, which he almost certainly regards as (respectively) an everlastingly
false existential af¬rmative assertion and an everlastingly true existential
negative assertion.
Aristotle does not even seem committed to the claim that one should be
able to tell whether an af¬rmative existential assertion (belief ) concerns a
non-composite item: after all, in Metaphysics 10, at 1052a 4“7, he acknowl-
edges that one might be unable to tell whether the item with which one™s
belief is concerned is of an unchanging nature.

No universal being. As we have seen, the thoughts and the linguistic expres-
sions concerning non-composite items discussed in T 23 and T 24 are
existential beliefs and existential assertions. This commits Aristotle to the
view that being is not a universal.53 For, if being were a universal (on a
par with, say, incommensurable and white), then every existential assertion
(belief ) would be a predicative assertion (belief ) concerning a composite
object, i.e. the state of affairs composed of the universal being and of the
thing spoken (thought) about. For example, if being were a universal, then
the assertions ˜Socrates is white™ and ˜Man exists™ should have the same
structure: as ˜Socrates is white™ asserts that the state of affairs of Socrates™
being white, a composite object, ˜is™ in the sense of being true, i.e. asserts
that the two components of this state of affairs, the universal white and
the individual Socrates, are reciprocally combined, so ˜Man exists™ should
assert that the state of affairs of man™s existence, a composite object, ˜is™ in
the sense of being true, i.e. should assert that the two components of this
state of affairs, the universal being and the universal man, are reciprocally
combined.
The attribution to Aristotle of the view that being is not a universal is
hardly surprising: it ¬ts well both with Aristotle™s well-known thesis that
being is not a genus54 and with his claim that ˜one man and an existent man
and a man are the same™ (Metaph. 2, 1003b 26“7).55 Note, however, that the
reason we now have for crediting Aristotle with the view that being is not a
universal is particularly strong. For this reason not only enables us to claim
that Aristotle rejects a single all-embracing universal, being, attributed by
every existential af¬rmation to whatever it is referring to. It also allows us to
53 Cf. Brentano (1889), 20; (1914a), 162; (1914b), 128; (1915b), 136.
54 APo. 2.7, 92b 14; Metaph. B 3, 998b 22; K 1, 1059b 33“4 (cf. APo. 1.32, 88b 1“2; EE 1.8, 1217b 33“5).
55 I follow Jaeger™s text. Cf. Metaph. I 2, 1054a 16“17.
114 Bearers of truth or falsehood
deny that for Aristotle every existential af¬rmation attributes some universal
or other (possibly different universals for different existential assertions) to
what it refers to. For instance, Aristotle cannot coherently endorse a theory
of existential assertions which some commentators56 attribute to him: that
every existential af¬rmative assertion asserts that the item it refers to falls
under a certain highest genus, where the highest genus introduced varies
from case to case (in one case it is the category of substance, in another the
category of quality, etc.).57

Alternative interpretations: (i) non-propositional truth. My interpretation
of Aristotle™s theory in Metaphysics 10 and de Anima 3.6 about non-
composite items, truth, and falsehood differs from the exegeses offered by
most commentators. In this subsection and in the next I discuss two rival
exegeses of this area of Aristotle™s thought.
The ¬rst rival exegesis is very widespread: it is already present in Scholastic
authors and is endorsed by most modern commentators.58 According to
this interpretation, in 10 and de Anima 3.6 Aristotle distinguishes two
senses of ˜true™. In its ¬rst sense, the most usual one, ˜true™ applies to items of
a propositional sort: assertions, beliefs, and states of affairs. The property
expressed by ˜true™ in this ¬rst sense is the property of corresponding to
reality, and the contrary of this property is falsehood. In its second sense,

56 Kahn (1978), 261; Hintikka (1986), 90“2; Grice (1988), 178“82; Hintikka (1999), 787“90.
57 For the categories as the highest genera of what exists see Cat. 8, 11a 37“8; 10, 11b 15“16; APr. 1.27,
43a 29“30; 43a 36“7; APo. 1.22, 83b 12“17; 32, 88b 1“3; 2.13, 96b 19“20; Ph. 1.6, 189a 14; 189b 23“4; 5.4,
227b 4“6; 7.1, 242b 35; de An. 1.1, 402a 22“5; 1.5, 410a 18“21; 2.1, 412a 6; Metaph. 6, 1016b 31“5; 28,
1024b 9“16; I 1, 1052b 18“19; 3, 1054b 27“31; L 5, 1071a 24“6 (cf. G. E. L. Owen (1965), 77; (1968a),
108“9; Leszl (1970), 60“6; Brakas (1988), 33“40, 42, 55; Loux (1991), 13; Smith (1997), 74“5; Fraser
(2002), 46“7). In one passage (Metaph. I 2, 1053b 22“4) Aristotle denies that substance is a genus, a
claim which raises problems that cannot be addressed here (cf. Brentano (1862), 100“1; Kahn (1978),
251“2; Fraser (2002), 75“81).
58 Ammon. in Int. 27, 28“33; Fonseca (1577/89), iii 666; Pacius (1611), 389; Brandis (1835/66), ii.ii.i
520“1; Bonitz (1848/49), ii 409“11; Brentano (1862), 26“7 (cf. n. 18 above); Torstrik (1862), 197; Grote
(1880), 618“19; Wallace (1882), 273“4; Dilthey (1883), 197“8; Maier (1896/1936), i 19“20; Joachim
(1906), 7, 124; Hicks (1907), 512; Jaeger (1911), 26“7; Lewinsohn (1911), 202“3; Jaeger (1912), 26“7; Ross
(1923), 25; (1924), ii 225, 275“6, 278; Heidegger (1926), 305“6; Calogero (1927/68), 24; Stenzel (1931),
176“7; Keeler (1932), 244“5; De Corte (1934), 237“8; Wilpert (1940), 10“11; Joachim (1948), 26“7;
Owens (1951/78), 412“13; de Rijk (1952), 9“10, 13“14; Engelhardt (1953), 37“8; Tugendhat (1956/88),
57; Reale (1961/67), 161; Aubenque (1962/83), 166; Oehler (1962/85), 182, 218, 241; Apostle (1966),
321; Luther (1966), 180“1, 186“7; Tugendhat (1966a), 256“7; von Fragstein (1967), 146“7; Hamlyn
(1968/93), 142; Reale (1968/93), i 101; Seidl (1971), 181“4; Grayeff (1974), 208, 209“10; Engmann
(1976), 261; Marx (1977), 7“9; Berti (1978), 144; Apostle (1981), 169, 171; Seidl (1989/91), i 428, ii
493“4; Volkmann-Schluck (1979), 286“9; Fleischer (1984), 25“6; Besoli (1985), 9, 41; Frede/Patzig
(1988), ii 318; Kal (1988), 47; Mignucci (1994), 152“3, 154“5; (1996a), 419“21; Whitaker (1996), 29;
Negro (1996/97), 344, 348“50; Vigo (1997), 32; Galluzzo (1997/98), 53“5; Detel (1998), 172; Pritzl
(1998), 178, 187“8; Szaif (1998), 521; Fiorentino (2001), 267“8, 275; Modrak (2001), 55, 64“6.
Truth conditions for existential assertions 115
which is somewhat unusual or even deviant, ˜true™ applies to items which
are not of a propositional sort: simple linguistic expressions (subjects and
predicates of predicative assertions) and simple thoughts (concepts). The
property expressed by ˜true™ in its second sense is the property of having
some object or other as content. More precisely: in the case of a simple
linguistic expression, the property expressed by ˜true™ in its second sense is
that of signifying some object; in the case of a simple thought, the property
in question is that of grasping some object. The contrary of the property
expressed by ˜true™ in its second sense is (not falsehood, but) lack of content,
i.e. the ˜blank™ condition of a simple linguistic expression which fails to
signify any object or of a simple thought which fails to grasp any object
(˜ignorance™).
The main dif¬culty with the Scholastic interpretation is that it is very
hard to see how ˜true™ could apply to simple linguistic expressions and
simple concepts. Simple linguistic expressions and simple concepts are not
the type of item to which ˜true™ naturally applies.

Alternative interpretations: (ii) de¬nitions. According to Silvestrus Maurus,
Sorabji, and Berti,59 the simple items discussed in Metaphysics 10 and de
Anima 3.6 are de¬nitions (sentences consisting of two linguistic expressions,
the de¬niendum and the de¬niens) and those thoughts that are their mental
counterparts. Moreover, these commentators assume that de¬nitions and
their mental counterparts are true (not in a ˜secondary™ or ˜deviant™ sense,
but) in the ˜ordinary™ sense. Since for Aristotle de¬nitions are identity
statements of a particular type,60 what he has in mind when he says that
a non-composite thought ˜is not something about something™ (de An. 3.6,
430b 27“8 < T 24) is that in such a thought a thing is attributed to itself.61
This interpretation faces an objection. For Aristotle a de¬nition is a
predicative assertion where the de¬niens is the predicate and the de¬niendum
the subject.62 (ii.i) In the Topics (1.4, 101b 11“25) he regards the link between
a de¬nition™s de¬niens and de¬niendum as one of four ways in which a
59 Maurus (1668), iv 480“1; Sorabji (1980), 218; (1981), 242; (1982), 298“9; Berti (1993), 28; (1994),
125“6; (1996a), 393“4; Wolff (1999), 55“6; Berti (2000), 10“11, 19, 24“5.
60 Top. 1.5, 102a 5“17; 7, 103a 23“39 (cf. Crivelli (2002), 243“5).
61 According to Wedin (1988), 125“6, Aristotle distinguishes two kinds of thoughts that cannot be false:
acts or states of grasping an item (whose truth consists in being in touch with this item) and acts or
states that correspond to the assertion of an item™s essence. Thus, Wedin seems to think that both the
traditional Scholastic interpretation and the exegesis going back to Silvestrus Maurus are partially
correct.
62 Aristotle™s view that a de¬nition is a predicative assertion where the de¬niens is the predicate and
the de¬niendum is the subject is compatible with his other view that a de¬niens does not exhibit
predicative structure in that in it no differentia is predicated of its genus (e.g. in the de¬niens
116 Bearers of truth or falsehood
predicate can attach to a subject. He even says: ˜It is necessary that whatever
is predicated about something should either be predicated in a convertible
way or not, and if it is predicated in a convertible way it will be either
a de¬nition or something peculiar™ (1.8, 103b 7“9).63 (ii.ii) Three passages
of the Posterior Analytics (1.14, 79a 24“9, 2.3, 90b 1“7, and 2.13, 97b 26“7)
presuppose that de¬nitions are universal af¬rmative assertions. (ii.iii) In
Posterior Analytics 2.13 (96a 32“4) Aristotle indicates that when one is looking
for a de¬nition one must ¬rst seek every attribute which holds essentially
of, but has a larger extension than, the item to be de¬ned, and then form
a compound of these attributes that does not extend beyond the item
to be de¬ned. Given his view that a de¬nition is a predicative assertion,
how could Aristotle claim that (the mental counterpart of ) a de¬nition ˜is
not something about something™? What he could say is that (the mental
counterpart of ) a de¬nition ˜is not something about something different™. One
might attempt to defend the interpretation in question by assuming that
the formula ˜something about something™ (˜tª kat‡ tin»v™) has a technical
meaning whereby it does not apply to de¬nitions. But even this will not
work: in Posterior Analytics 2.4 Aristotle, after claiming that ˜a syllogism
proves something about something™ (91a 14“15), goes on to consider the
consequences of this claim for the question whether the essence can be
demonstrated, and has no qualms about the idea that the conclusion of
a syllogism might be ˜something about something™ and none the less be a
de¬nition.64


2 non-composi t e s u bs ta nces
Two interpretations of Aristotle™s ˜non-composite substances™. T 23 men-
tions, alongside essences, certain ˜non-composite substances™ (˜mŸ sunqetaª
oÉs©ai™, 1051b 27). What are these ˜non-composite substances™? According to


˜biped walking animal™, ˜walking™ is not predicated of ˜animal™ nor ˜biped™ of ˜walking animal™): see
APo. 2.3, 90b 33“8; Metaph. Z 12, 1037b 18“21 (cf. Ross (1949), 614“15; Bolton (1976), 523; Barnes
(1994), 208; Wedin (1988), 130; Mignucci (1994), 141). In one passage (Metaph. H 3, 1043b 28“32)
Aristotle does seem to assume that a de¬niens exhibits predicative structure in that its differentia is
predicated of its genus.
63 Cf. Top. 7.5, 154a 36“154b 2; APo. 2.4, 91a 15“16; Mignucci (1975a), 38; (1994), 141; (1996a), 409.
64 There is a passage from Posterior Analytics 1.2 (72a 20“1) where according to some commentators
(e.g. Phlp. in APo. 35, 2“13; 37, 7“13) Aristotle says that de¬nitions do not display a predicative
structure. However, other commentators (e.g. S. Mansion (1946/76), 138; Ross (1949), 504“5) offer
an alternative, and more plausible, interpretation of the passage in question: Aristotle is saying (not
that de¬nitions do not display a predicative structure, but) that a de¬nition does not make an
existential claim.
Truth conditions for existential assertions 117
the ˜traditional™ interpretation,65 they are incorporeal substances, i.e. God
and (perhaps) the intellects that move the heavenly spheres; according to
the ˜modern™ interpretation,66 they are (or at least include) forms of material
substances.67 An assessment of these competing exegeses is best introduced
by two points about Aristotle™s language.

(i) ˜Non-composite™. ˜Non-composite™ can mean either ˜non-composed with
anything™, i.e. ˜isolated™, or ˜non-composed of anything™, i.e. ˜simple™. The
notion of non-composition occurs elsewhere in 10™s central part: indeed,
the passage™s main topic is the application of ˜true™ to non-composite items.
But the notion of non-composition occurring elsewhere in the passage is
not that of isolation, but that of simplicity.68 Hence our ˜non-composite
substances™ are probably simple substances. Now, when Aristotle says that
a substance is simple, the ¬rst thing coming to one™s mind is that the
substance should not be composed of matter and form: for Aristotle applies
˜composite substance™ to substances composed of matter and form.69 Thus,
our ˜non-composite substances™ are probably substances not composed of
matter and form.

(ii) ˜Similarly also™. Immediately after saying that ˜it is not possible to err
with regard to the “what it is” except incidentally™ (1051b 25“6), Aristotle
introduces ˜non-composite substances™ by saying: ˜Similarly also [¾mo©wv
d• kaª] with regard to non-composite substances, for it is not possible to
err™ (1051b 26“8).
The formula ˜similarly also™ is common in Aristotle. Its main use is
expansive: the items it introduces are not among those previously discussed.
65 [Alex. Aphr.] in Metaph. 600, 25“7; Ascl. in Metaph. 6, 20“1; Aquinas in Metaph. 1901 Cathala/Spiazzi;
Suarez (1597), liii; Schwegler (1847/48), iv 186, 187; Bonitz (1848/49), ii 410; Brentano (1862), 27;
(1889), 20; Jaeger (1923), 212; Ross (1924), ii 274, 276; Rolfes (1904/28), ii 378; Faust (1931/32), i
216, 359; Wilpert (1940), 11; Owens (1951/78), 413“14; Merlan (1953), 158“9; D´carie (1961), 161;
e
von Fragstein (1967), 148; Reale (1968/93), iii 464; Grayeff (1974), 208; Aubenque (1962/83), 374“
5; (1979a), 79; Wolf (1979), 111, 417. Maier (1896/1936), i 7 and Engelhardt (1953), 45 take 10™s
˜non-composite substances™ to include God and the celestial bodies.
66 Jaeger (1912), 25; Oehler (1962/85), 183“4, 221“2; Seidl (1971), 183, 194; Kessler (1976), 181; Marx
(1977), 6; Berti (1978), 147“50; Volkmann-Schluck (1979), 274“5; Sorabji (1982), 298“9; Burnyeat
et al. (1984), 157; Liske (1985), 424; Kal (1988), 126; Seidl (1989/91), i 428; Berti (1990), 115“19; (1993),
25; (1994), 131“2, 134; (1996a), 399“400, 402; Pritzl (1998), 189“90; Berti (2000), 12, 20.
67 Some commentators support neither the ˜traditional™ nor the ˜modern™ interpretation: e.g. Mignucci
(1994), 150“1 and (1996a), 418“19 takes our passage™s ˜non-composite substances™ to be ˜the ontological
counterparts of our concepts™.
68 Cf. the ˜thinking of indivisible items™ in de Anima 3.6 (430a 26) and the thought of ˜simple items™ in
Metaphysics E 4 (1027b 27“8).
69 De An. 2.1, 412a 6“9; Metaph. H 3, 1043a 30. Aristotle distinguishes three kinds of substance: matter,
form, and their compound (see e.g. de An. 2.1, 412a 6“9; Metaph. Z 3, 1028b 36“1029a 7; 10, 1035a 1“2).
118 Bearers of truth or falsehood
For instance, in ˜Breathing occurs in all mammals, similarly also with birds™
the items introduced by ˜similarly also™ are not among those discussed
previously (birds are not mammals).70 Another use of ˜similarly also™ is
restrictive: the items it introduces are among those previously discussed. For
instance, in ˜Breathing occurs in all mammals, similarly also with whales™
the items introduced by ˜similarly also™ are among those previously discussed
(whales are mammals). The restrictive use presupposes either that the items
introduced could be expected not to rank among those previously discussed
(one could expect whales not to be mammals) or that the items introduced
could be expected to lack the property ascribed to those previously discussed
(one could expect whales not to breathe).71
If the use of ˜similarly also™ at Metaph. 10, 1051b 26“7 is expansive, the
˜non-composite substances™ discussed in 10™s central part are not among
the items previously discussed, i.e. essences. If instead the use of ˜similarly
also™ at Metaph. 10, 1051b 26“7 is restrictive, 10™s ˜non-composite sub-
stances™ are among the items previously discussed, i.e. essences. In this case,
however, ˜non-composite substances™ could be expected either not to rank
among essences or to lack the property ascribed to essences, i.e. the property
of being foreign to error.72

Assessment of the two interpretations. As for the ¬rst point about Aristotle™s
language, the ˜traditional™ and the ˜modern™ interpretation fare equally:
10™s ˜non-composite substances™ may well be (neither God nor the heavenly
intellects, which are incorporeal substances not composed with matter, but)
substantial forms which, though not composed of matter, are immersed in
(hence composed with) it.
The second point about Aristotle™s language creates more of a dif¬culty
for the ˜modern™ interpretation. For suppose the ˜modern™ interpretation is
right. Then 10™s ˜non-composite substances™ are the formal components
of ˜composite™ substances. They therefore are essences in the category of
substance. Then they constitute a proper subclass of the essences (the ˜what
it is™™s) mentioned in the preceding sentence (for there are essences not

70 All the Metaphysics occurrences of the formula ˜similarly also™ which are relevantly similar to the one
at hand seem to be instances of its expansive use: see B 2, 996b 21; 997b 20; 6, 1002b 21; 2, 1003b 30;
1004b 4; 4, 1014b 30; 7, 1017a 30; 1017b 6; 12, 1019a 26; Z 1, 1028a 22; 17, 1041b 27; H 1, 1042b 1; 1,
1046a 25“6; 3, 1046b 32; 1046b 35“6; 1047a 22; 8, 1050a 16; 1050a 33; I 2, 1053b 26; 1054a 1; 6, 1056b 3; 7,
1057a 26; K 9, 1066a 34; L 2, 1069b 17; 3, 1070a 28“9; M 6, 1080a 29; 1080b 23; 9, 1085a 7; 1085b 27; N 1,
1087b 36.
71 I know no examples of the restrictive use of ˜similarly also™ in Aristotle™s works.
72 Discussions of the formula ˜similarly also™ can be found also in Berti (1978), 147 and Burnyeat et al.
(1984), 157.
Truth conditions for existential assertions 119
only in the category of substance, but also in other categories).73 Hence
the expansive use of ˜similarly also™ at Metaph. 10, 1051b 26“7 is excluded.
Aristotle can only have in mind its restrictive use. This presupposes that
˜non-composite substances™, i.e. essences in the category of substance, could
be expected either not to rank among essences or to lack the property just
ascribed to essences in general, i.e. the property of being foreign to error. But
this presupposition strikes one as odd: any reader who understands enough
of Aristotle™s jargon to make sense of the expression ˜non-composite sub-
stances™ as denoting the formal components of ˜composite™ substances will
know very well that they are essences, and it is hard to see why such a reader
should expect that these essences in the category of substance should have
a special reason for lacking the property just ascribed to essences in gen-
eral, i.e. being foreign to error. No such dif¬culty arises for the ˜traditional™
interpretation. For the ˜traditional™ interpretation can take the use of the
formula ˜similarly also™ at Metaph. 10, 1051b 26“7 to be expansive: so far
Aristotle has shown that essences of things are foreign to error, now he
adds that the same holds for incorporeal forms, i.e. God and (perhaps) the
heavenly intellects. So the second point about Aristotle™s language favours
the ˜traditional™ interpretation over the ˜modern™.
An objection raised by some commentators against the ˜traditional™ inter-
pretation is that it is hard to see why Aristotle should mention incorporeal
substances at this point.74 This objection is easily answered: since Aristotle
does believe in incorporeal substances, it is natural that he should mention
them in a discussion of non-composite items. Moreover, Aristotle™s reason
for mentioning incorporeal substances at this point, at the end of book ,
might be that he is building up the Metaphysics towards the treatment of
them (in book L).
Another objection raised against the ˜traditional™ interpretation is that
human beings do not have a direct and immediate grasp of incorporeal sub-
stances (Aristotle himself needs to argue for their existence).75 This objec-
tion can be answered in two ways. First, Aristotle need not be discussing
a human epistemic state: he might well be addressing a divine epistemic
state. Second, a few lines after our passage (at 1051b 32“3) Aristotle clearly
says that one™s belief in a non-composite item does not require one™s ability
to articulate its de¬nition. So, even if Aristotle were to claim that human

73 For items from categories other than substance having essences see Top. 1.9, 103b 27“37; de An. 1.1,
402b 16“21; 3.4, 429b 10“12; Metaph. 6, 1016b 1“3; Z 4, 1030a 17“27; Berti (1978), 146“7; (1994),
131“2; (1996a), 399; Vigo (1997), 38.
74 75 Cf. Seidl (1971), 183.
Cf. Mignucci (1994), 150; (1996a), 418.
120 Bearers of truth or falsehood
beings have an immediate grasp of incorporeal substances, this would not
amount to a form of ˜religious intuitionism™.
A strength of the ˜traditional™ interpretation is its ability to explain why
Aristotle uses the formula ˜being itself ™ (˜t¼ ¿n aÉt»™, 1051b 29), which has
a strong Platonic ¬‚avour.76 If, as the ˜traditional™ interpretation claims, the
˜non-composite substances™ of 10™s central part are God and (perhaps)
the heavenly intellects, Aristotle might be using the formula ˜being itself ™ to
suggest a comparison of the role played in his system by God and (perhaps)
the heavenly intellects with the role played in Platonic philosophy by the
Good and (perhaps) Forms.77
Thus, all considered, the ˜traditional™ interpretation has the edge on the
˜modern™.

Properties of ˜non-composite substances™. In the last part (1051b 28“30) of his
discussion of ˜non-composite substances™, Aristotle indicates some of their
properties.
(i) Incorporeal substances are completely in actuality (cf. 1051b 30“1).78
This means that they contain no matter (which Aristotle equates with
potentiality and contrasts with actuality).79 This con¬rms the previously
suggested explanation of why ˜non-composite™ applies to these substances “
it is because they are not composed of form and matter.80
(ii) The fact that an incorporeal substance cannot come or cease to be is
linked to its containing no matter. For all coming to be is ˜from™ something,
i.e. from matter,81 and matter persists through the process of coming to be,
i.e. remains present in what has come to be: thus, only material entities
can be the result of coming to be, while ˜being itself ™ (1051b 29), i.e. what is


76 Cf. Metaph. B 4, 1001a 23; 1001a 27; 1001a 30; N 2, 1089a 3. Aristotle reports that the Platonists use the
word ˜itself ™ to compose the name of a form: see Metaph. Z 16, 1040b 33“4; EN 1.4, 1096a 34“1096b 5;
EE 1.8, 1218a 11.
Elsewhere Aristotle does compare his God with Plato™s Good: see Metaph. L 10, 1075a 38“1075b 1;
77
1075b 34“7 (with D¨ ring (1966), 221“2). Ps.-Alexander (in Metaph. 600, 25“7) identi¬es the ˜being
u
itself ™ of 1051b 29 with ˜the simple substance which is above being and moves without being in
motion, God worthy of great admiration™. The expression ˜above being™ (˜Ëpero…siov™) used by
ps.-Alexander to describe God recalls the expression ˜beyond being™ (˜–p”keina t¦v oÉs©av™) used by
Plato to describe the Good (see R. vi 509b8“10).
Cf. Metaph. L 6, 1071b 17“20; 7, 1072a 25“6; 1072b 8.
78
De An. 2.1, 412a 9; 3.5, 430a 10“11; Metaph. L 5, 1071a 8“11; etc. (cf. Oehler (1962/85), 222).
79
80 Cf. the subsection to which n. 68 above is appended.
81 For Aristotle™s use of the preposition ˜from™ (˜–k™) to indicate matter see Ph. 1.7, 190b 24“7; 9, 192a 31“2;
2.3, 194b 23“6; Metaph. 2, 1013a 24“6; Z 7, 1032a 12“14; 8, 1033a 25“6; 1033b 1“3; 1033b 8“9. However,
at Metaph. Z 7, 1033a 19“22 Aristotle observes that on careful examination it is inappropriate to use
˜from™ to indicate matter.
Truth conditions for existential assertions 121
purely essence or form with no material component, cannot have come to
be.82

3 singul a r exis tent ia l a s se rt i ons concerni ng
materia l s ub s ta nc es
Aristotle™s silence. In the last two sections I discussed Aristotle™s views on true
and false existential assertions concerning non-composite items. I argued
that among these non-composite items there are essences and incorporeal
substances. Does Aristotle have views on true and false singular existen-
tial assertions concerning material substances (e.g. assertions like ˜Socrates
exists™ and ˜Bucephalus does not exist™)? If he does have views in this area,
what are they?
Aristotle does not explicitly discuss true and false singular existen-
tial assertions concerning material substances. My strategy is therefore to
attempt to reconstruct what views in this area Aristotle is committed to.
The results gained with regard to existential assertions readily translate into
points regarding existential beliefs.

Two possible developments. The theory expounded in Metaphysics 10 could
be developed in two directions. On the one hand, Aristotle might say
that material substances are non-composite items alongside essences and
incorporeal substances; on the other, he might say that material substances
are composite items alongside states of affairs. Some commentators83 think
that Aristotle takes the ¬rst direction. However, several passages strongly
suggest that he favours the second: he would probably say that material
substances are composite items alongside states of affairs.

Matter“form and substance“accident. Aristotle regards the relation of matter
to form within a material substance, or ˜matter“form relation™, as analo-
gous to the relation of a whole material substance to those universals from
categories other than substance that hold of it, or ˜substance“accident rela-
tion™.84 He does not regard these relations as identical “ in fact, when he
states that matter and the matter“form compound ˜are subjects in different

82 Aristotle also claims that corporeal forms (forms combined with matter within a material substance)
neither come nor cease to be: see Metaph. E 3, 1027a 29“30; Z 8, 1033a 24“1033b 19 (with Oehler
(1962/85), 218“19 and Liske (1985), 424); 9, 1034b 7“16; 10, 1035a 28“30; 15, 1039b 20“7; H 1, 1042a 26“
31; 3, 1043b 14“16; 5, 1044b 21“4; L 3, 1069b 35“1070a 4; 1070a 15“17.
83 E.g. de Rijk (1952), 15“16.
84 Cf. Leszl (1971), 337; Lewis (1991), 146, 147, 246“8, 259“61.
122 Bearers of truth or falsehood
ways™,85 part of his point is probably that these relations are different. How-
ever, for present purposes, what counts is that Aristotle does consider the
matter“form and the substance“accident relations as analogous. Here is
the evidence for crediting Aristotle with this view.
(i) Aristotle thinks that as a non-substance universal is predicated of a
substance, so form is predicated of matter within a material substance.86
(ii) Aristotle seems to ˜discover™ the distinction between matter and form
by extending his analysis of quali¬ed coming-to-be (coming-to-be-white
and, in general, the phenomenon described by ˜to come to be™ followed by
an adjectival complement) to absolute coming-to-be (birth and, in general,
the phenomenon described by ˜to come to be™ on its own): this is how he
introduces the distinction between matter and form in Physics 1.7,87 which
is probably one of the earliest texts where it appears.88
(iii) Aristotle uses the same technical term, ˜compound™ (˜s…nolon™), for
matter“form compounds89 and for incidental compounds that consist of a
substance and a non-substance universal predicated of it.90
(iv) In several passages Aristotle draws the analogy explicitly. One such
passage is from Metaphysics H 4:
T 27 Those things that are by nature but are not substances have no matter, but
substance is what underlies them. For example, what is the cause of an
eclipse? What is its matter? For there is none, but the moon is what suffers
an affection. What is the cause in the sense of what moves and destroys the
light? The earth. Perhaps there is no purpose. And the cause in the sense of
form is the account. (1044b 8“12)91
In T 27 Aristotle says that the moon (a substance)92 plays in a lunar eclipse
(an event) a role analogous to that of matter in a material substance.

Existence, form, and matter. Since Aristotle regards the matter“form relation
as analogous to the substance“accident relation, one expects him to believe
that for a material substance to exist is for its form to hold of its matter.

85 Metaph. Z 3, 1029a 2“3; H 1, 1042a 26“31 (cf. 7, 1049a 27“36).
86 De An. 2.1, 412a 16“19; Metaph. B 1, 995b 35; 4, 999a 33“4; Z 3, 1028b 36“1029a 5; 1029a 23“4; 13, 1038b 4“
6; H 1, 1042a 26“31; 2, 1043a 5“6; 7, 1049a 34“1049b 2 (cf. Ross (1924), ii 164, 257; Leszl (1971), 337;
Brunschwig (1979), 148“50; Lewis (1991), 152“3, 266).
190b 10“17, cf. GC 1.4, 320a 2“5; Metaph. Z 7, 1032a 13“25; H 1, 1042a 32“1042b 3; L 1, 1069b 3“20;
87
G. E. L. Owen (1966), 148“9; (1978/79), 6; Lewis (1991), 223“8, 245, 248.
88 Cf. Graham (1987), 119, 133“7; Lewis (1991), 245, 247.
89 PA 1.1, 640b 23“8; Metaph. B 1, 995b 35; 4, 999a 33“4; Z 3, 1029a 3“5; 10, 1035a 17“21; etc.
90 Metaph. M 2, 1077b 8“9.
91 Cf. Z 3, 1029a 23“4; 13, 1038b 5“6; H 2, 1043a 4“7 (with Brunschwig (1979), 133“4); 7, 1049a 27“36.
92 For the moon being a substance see Metaph. Z 2, 1028b 8“13 (cf. APo. 2.2, 90a 12).
Truth conditions for existential assertions 123
This expectation is con¬rmed by Metaphysics Z 17 (1041a 10“1041b 11). In this
passage (on one plausible interpretation of it, an interpretation I shall not
endeavour to substantiate here) Aristotle draws an analogy between three
cases of knowledge of a thing™s existence. The ¬rst concerns a lunar eclipse:
to know that a lunar eclipse exists is to know that the moon (a substance)
has the property of being eclipsed. The second case concerns a house: to
know that a house (a substance, perhaps) exists is to know of certain bricks
and stones (a portion of matter) that they have a certain form. The third
case concerns a man: to know that a man (a substance) exists is to know of
a certain lump of ¬‚esh and bones (a portion of matter) that it has a certain
form.93
Aristotle™s view that for a material substance to exist is for its form (or
essence) to hold of its matter seems connected with another view of his: that
a thing™s essence is the cause of its existence.94 It also seems connected with
a passage from Metaphysics H 2 (1042b 15“1043a 14) where he distinguishes
various ˜differences™ which by being present in something make it into a
certain individual: he associates a signi¬cation of ˜to be™ with each of these
˜differences™, and he says that for an individual to exist is for its peculiar
˜difference™ to hold of its matter-like component.95

93 Cf. Charles (2002), 120. At the end of this passage, at 1041b 9“11, Aristotle says that the inquiry into
simple items must be different from that into the items previously mentioned. Several commentators
(Bonitz (1848/49), ii 360; Ross (1924), ii 225; von Iv´nka (1932), 21; D´carie (1961), 151; Grayeff
a e
(1974), 208; Aubenque (1979a), 69; Kahn (1985), 326“7; Frede/Patzig (1988), ii 318) note that this
brief allusion to simple items links Metaphysics Z 17 to the central part of 10, where beliefs and
assertions concerning non-composite items are discussed.
94 De An. 2.4, 415b 12“14 (cf. GC 1.3, 318b 25; GA 2.1, 731b 28“30; EN 9.4, 1166a 4“5; 7, 1168a 5“7; Protr.
fr. 73 Gigon 311b 39“40 = Iamb. Protr. 58, 11); Metaph. 8, 1017b 14“16; Z 17, 1041a 9“10; 1041a 27“8;
1041b 8“9; 1041b 27“8; H 2, 1042b 32“3; 1043a 2“4. Aristotle sometimes refers to x™s essence by means
of the formulae ˜to be for x™ and ˜what it was to be for x™ (see Bonitz (1870), 221a 34“9, 764b 11“17),
formulae which seem to allude to the cause of x™s existence (cf. Kahn (1976), 332“3; Loux (1991),
73“5; Bostock (1994), 50; Hintikka (1999), 786; B¨ck (2000), 107). However, the formulae ˜to be for
a
x™ and ˜what it was to be for x™ could indicate (not the cause of x™s existence, but) the cause for x to
be x (cf. Anscombe/Geach (1961), 24“6, 34“5): e.g. in Categories 7 Aristotle seems to identify ˜to be
for relatives™ (cf. ˜o³v t¼ e²nai™ at 8a 32 and ˜t¼ e²nai to±v pr»v ti™ at 8a 39) with ˜to be a relative for
relatives™ (cf. ˜t¼ pr»v ti aÉto±v e²nai™ at 8a 34“5) (cf. Top. 5.4, 133b 33“6).
95 Cf. Leszl (1971), 337, 339; Grice (1988), 182; Barnes (1995a), 80“1; Matthews (1995), 236“7; Shields
(1999), 234“5; Charles (2002), 121“3. G. E. L. Owen (see his (1960), 165, (1965), 76“7, (1966), 145,
and (1978/79), 11“12) interprets differently the passage from H 2: Aristotle would be claiming that
for a non-substantial individual to exist is for its peculiar ˜difference™ to hold of the non-substantial
individual itself (not of its matter-like component). The H 2 passage surely can be interpreted as
Owen does, and many commentators agree with him (see Cresswell (1971), 98, 106; Ackrill (1972),
210; C. J. F. Williams (1981), 134; Dancy (1986), 65; Hintikka (1986), 86; Upton (1988), 385; Loux
(1991), 28, 41, 73; Bostock (1994), 49“50, 52, 254“7; L. Brown (1994), 235; Whitaker (1996), 30“2,
136“7). However, the evidence from other texts makes it more plausible to understand the H 2
passage as claiming that for the non-substantial individual to exist is for its form-like component to
hold of its matter-like component.
124 Bearers of truth or falsehood
An account of existential assertions that appeals to form and matter. The con-
siderations of the last subsections show that Aristotle is more likely to regard
a material substance as a composite item than as a non-composite item. This
result, in turn, makes it plausible to assume that Aristotle would endorse an
account of true and false singular existential assertions concerning material
substances based on the idea that every such assertion claims that a form is
combined with, or divided from, a portion of matter:
[24] A singular af¬rmative existential assertion concerning a material sub-
stance asserts that that substance™s form is combined with that sub-
stance™s matter. Accordingly, it is true (false) when and only when that
form is combined with (divided from) that matter.
[25] A singular negative existential assertion concerning a material sub-
stance asserts that that substance™s form is divided from that substance™s
matter. Accordingly, it is true (false) when and only when that form is
divided from (combined with) that matter.96
A singular existential assertion latches on to a form and a portion of mat-
ter by passing through a material substance which at some time or other
exists. None the less, a singular af¬rmative (negative) existential assertion
concerning a material substance can be false (true) at some time or other,
i.e. when the material substance it concerns no longer exists.
Does the form or the matter of a material substance survive forever the
material substance™s destruction? The form of a material substance is an
essence, i.e. a natural kind, and therefore probably survives forever the
material substance™s destruction.97 On the other hand, the matter of a
material substance is unlikely to survive forever that material substance™s
destruction. It therefore seems sensible to construct the truth conditions for
a singular af¬rmative (negative) existential assertion concerning a material
substance in such a way that they do not require the form or the matter of
the material substance concerned to exist when the assertion is false (true).
The best way to achieve this result is to characterise the division appealed
to by the truth conditions in such a way that the portion of matter and the
form in question should be reciprocally divided whenever either of them
does not exist:
[26] For every form f, every portion of matter m, and every time t, if f is
combined with m at t then both f and m exist at t. For every form

96 Cf. Leszl (1971), 337“40; van Bennekom (1986), 11“12; Matthews (1995), 236“7.
97 Cf. the subsection to which n. 21 above is appended.
Truth conditions for existential assertions 125
f, every portion of matter m, and every time t, f is divided from m
at t just in case m is other than everything with which f is combined
at t.
Note the asymmetry between singular af¬rmative existential assertions con-
cerning material substances and af¬rmative existential assertions concern-
ing non-composite items: a singular af¬rmative existential assertion con-
cerning a material substance may well be false, when the material substance
concerned no longer exists, but an af¬rmative existential assertion concern-
ing a non-composite item must be always true because the non-composite
item it concerns is everlasting, i.e. always exists.98

Limits. Propositions [24] and [25] cannot handle singular existential asser-
tions about what never exists, e.g. an utterance of ˜Pegasus does not exist™.
The singular existential assertions addressed by [24] and [25] are those
af¬rming or denying the existence of a material substance that does exist
at some time or other (possibly distant from the time when the assertion is
uttered). The singular existential assertions addressed by [24] and [25] are
those involving a certain speci¬c use of ˜to exist™, a use one of the features of
which is that the verb can be accompanied by temporal adverbs like ˜still™
or ˜no longer™.99
98 99
Cf. the subsection to which n. 21 above is appended. Cf. G. E. L. Owen (1965), 78“9.
pa rt ii
˜Empty™ terms
ch a p t e r 4

Truth as correspondence




Can Aristotle™s theory of truth for assertions be regarded as a correspondence
theory of truth? Section 1 argues that on certain conceptions of truth as
correspondence it cannot, but on at least one other it can. Speci¬cally,
Aristotle™s theory of truth can be regarded as a correspondence theory of
truth in that it can be regarded as taking the truth of an assertion to
amount to a relation of isomorphism to reality. In particular: it relies on a
classi¬cation of assertions, with each class it associates some characteristic
that can hold of the items an assertion is about, and it claims that an
assertion is true when and only when the characteristic associated with its
class holds of the items it is about.
Section 2 addresses Aristotle™s reaction to the Liar, which creates a puzzle
for correspondence theories of truth like Aristotle™s. It is not clear whether
Aristotle addressed the Liar, and, in case he did, what version of it he
confronted. On the somewhat optimistic assumption that he did address
a robust version of the paradox, his solution is that the assertion on which
the paradox turns (an utterance of the sentence-type ˜I am speaking falsely™)
is sometimes neither true nor false.


1 a corres pond en ce th e ory of tru th?
Focusing on truth for assertions. The theory of truth we are in the best position
to attribute to Aristotle is a theory of truth for assertions (utterances) (it was
described in chapters 2 and 3).1 It is therefore with regard to this Aristotelian
theory of truth for assertions that it is worthwhile asking whether it is a
correspondence theory of truth. To answer this question, one must ¬rst
characterise what a correspondence theory of truth for assertions is.

1 See [17], [24], and [25] on pp. 93“4 and 124 above.



129
130 ˜Empty™ terms
(i) Facts. According to a widespread, ˜classical™ conception, the hallmark of
a correspondence theory of truth for assertions is the claim that an assertion
is true (false) just in case there is some (no) fact to which it corresponds.2
On this conception, Aristotle™s theory does not qualify as a correspondence
theory of truth for assertions, for it does not mention facts, so that there
is no question of assertions corresponding to them.3 It is also worthwhile
noting that although Aristotle sometimes4 seems to use ˜object™ (˜prŽgma™)
to denote facts, there appears to be no Greek count-noun which can be
used like the English ˜fact™ in a Greek construction which corresponds word
for word to the English ˜the fact that . . .™5

(ii) States of affairs. According to a second conception, the hallmark of a
correspondence theory of truth for assertions is the claim that an assertion
is true (false) just in case the corresponding state of affairs obtains (does
not obtain). A fact is then de¬ned as a state of affairs that obtains.6
This conception provides better prospects for crediting Aristotle with
a correspondence theory of truth for assertions: after all, Aristotle does
analyse the truth of predicative assertions in terms of states of affairs that
˜are™ or ˜are not™ in the sense of being true or false. But a brief re¬‚ection
shows that Aristotle™s views on the relationship between truth and states of
affairs do not match the second conception of a correspondence theory of
truth for assertions. For Aristotle allows only ˜af¬rmative™ states of affairs, so
that while an af¬rmative predicative assertion is true when and only when
the corresponding state of affairs ˜is™ in the sense of being true, a negative
predicative assertion is true when and only when the corresponding state of
affairs ˜is not™ in the sense of being false (e.g. ˜You are not seated™ is true when
and only when the state of affairs of your being seated ˜is not™ in the sense
of being false):7 by contrast, the second conception of a correspondence
theory of truth for assertions requires that every assertion (negative as well

2 Cf. Russell (1912), 75; Moore (1953), 277; Prior (1967b), 226; David (1994), 17“18; Neale (2001), 86.
The correspondence theory of truth advocated by Russell and Moore concerns beliefs, but it can be
easily translated into one for assertions.
3 Cf. David (1994), 18 (contra see Ringbom (1972), 12). In Aristotle™s remark ˜He who is in a state
contrary to that of the objects is wrong™ (Metaph. 10, 1051b 4“5), ˜objects™ denotes not facts but
states of affairs.
4 APr. 2.15, 64b 10; GC 1.8, 325a 18; Ph. 8.8, 263a 17; Metaph. A 3, 984a 18 (cf. Pritzl (1998), 184).
Cf. Geach (1963), 21. Aristotle™s technical expression ˜t¼ ‚ti™ (APo. 1.5, 75a 16; 2.1, 89b 24; EN 1.7,
5
1098b 1; etc.) comes close, but is never followed by a clause expressing the ˜content™ of ˜‚ti™.
6 Cf. Boeth. in Int. Pr. Ed. 110, 16“18; Tarski (1933), 155; (1944), 118; Chisholm (1966), 103“4; Sommers
(1969/70), 267“8; Kirkham (1992), 119; David (1994), 31“4.
7 Cf. [4] and [5] on p. 55 above (which, however, talk of af¬rmative and negative predicative beliefs, not
assertions).
Truth as correspondence 131
as af¬rmative) should be true just in case the corresponding state of affairs
˜is™ in the sense of being true.8

(iii) Isomorphism. According to a third conception, the hallmark of a cor-
respondence theory of truth for assertions is explaining the truth of an
assertion in terms of a relation of ˜isomorphism™ to reality. Suppose that
one has introduced a classi¬cation of assertions that singles out n classes
C1 , . . . , Cn of assertions. Suppose, further, that one has established a one-
to-one mapping of the classes of assertions singled out by this classi¬cation
onto certain attributes (properties or relations) A1 , . . . , An . One can then
set out a correspondence-as-isomorphism theory of truth for assertions by
saying that an assertion a is true just in case either a belongs to C1 and the
item (or items) a is about has (or have) attribute A1 or . . . or a belongs to
Cn and the item (or items) a is about has (or have) attribute An .
On this correspondence-as-isomorphism conception, Aristotle™s theory
does rank as a correspondence theory of truth for assertions.9 In fact, the

8 However, recall (cf. the paragraph to which n. 18 of ch. 1 pertains) that in the Categories Aristotle
seems to postulate both ˜negative™ and ˜af¬rmative™ states of affairs. The theory of truth for assertions
which most probably accompanies such a conception of states of affairs will claim that an af¬rmative
(negative) assertion is true when and only when the corresponding af¬rmative (negative) state of affairs
obtains. Such a theory would count as a correspondence theory of truth for assertions according to
the second conception.
9 Thus, the widespread view that Aristotle endorses a correspondence theory of truth is vindicated: see
Ammon. in Int. 20, 32“21, 16; Aquinas in Int. 29, 31“2 Spiazzi; in de An. 760 Pirotta; Trendelenburg
(1836/92), 54“5; (1846), 13“14, 17“18; Bonitz (1848/49), ii 409; Brentano (1862), 26, 28“9 (with Krell
(1975), 84“6); Steinthal (1863/90), i 246; Ueberweg (1868), 25; Brentano (1874/1925), ii 54; Trendelen-
burg (1876), 78, 83; Dilthey (1883), 197“9; Brentano (1889), 7, 15, 18“19, 24“5 (with Srzednicki (1965),
18); Maier (1896/1936), i 17“18, 34“5, 37“40, 207; Joachim (1906), 124“6; Lukasiewicz (1910a), 20;
(1910b), 18; Jaeger (1912), 26; Herbertz (1913), 214“15; Brentano (1914a), 163; (1915a), 219“20; Geyser
(1917), 54, 56; Ross (1923), 26; Linke (1924), 407; Calogero (1927/68), 24; Keeler (1932), 247; Popper
(1934/59/75), 274; Cooke/Tredennick (1938), 7; Tarski (1944), 117“18; S. Mansion (1946/76), 235“7;
Baudry (1950), 8; Scarpat (1950), 24“5; de Rijk (1952), 13; Engelhardt (1953), 41; Viano (1955), 28;
Lenz (1957), 773“4; Saunders (1958), 377; Moreau (1961), 22; Reale (1961/67), 161; Oehler (1962/85),
176“9, 245“6; Ackrill (1963), 90, 140; Burrell (1964), 46; Brandt (1965), 14“15; Linke (1965), 307“9;
Cahn (1967), 29; Sainati (1968), 213, 225; Tarski (1969), 102“3; D. Frede (1970), 19“20; Abbagnano
(1971), 914; Morpurgo-Tagliabue (1971), 32; McKim (1971/72), 103; Patzig (1973), 923; Capozzi (1974),
320“1; Dickason (1976), 20; Graeser (1977), 374; Marx (1977), 9; Graeser (1978), 447; K. J. Williams
(1978), 67, 72; Aubenque (1979a), 69; Volkmann-Schluck (1979), 260, 286; M. J. White (1981), 231“2;
Bormann (1982), 1; Edel (1982), 242; Seel (1982a), 211; Berti (1983a), 276; Graeser (1983b), 201; Kreiser
(1983), 96; Fleischer (1984), 16, 22; Oehler (1984), 222; Craig (1988), 23“4, 32, 236; Hallett (1988), 8;
Irwin (1988), 5“6; Donini (1989), 11“12; Halper (1989), 217, 218; Wolenski/Simons (1989), 393; Crivelli
(1996), 158“9; Negro (1996/97), 342; Knabenschuh de Porta (1997), 200, 207; Vigo (1997), 7“10,
13“14, 27, 30; Galluzzo (1997/98), 43, 48“9, 58; Pritzl (1998), 177, 180; Szaif (1998), 520“1; K. Taylor
(1998), 115; Enders (1999), 8, 118; Hafemann (1999), 130“1; Wolff (1999), 51, 55; Berti (2000), 3, 8“9;
Fiorentino (2001), 280“5; Modrak (2001), 4, 55“62.
Some commentators doubt or deny that Aristotle™s theory of truth is a correspondence theory of
truth: see Adorno (1961), 407; Luther (1966), 182“3, 188; Davidson (1996), 268; Engel (2002), 15“16.
132 ˜Empty™ terms
condition introduced by the correspondence-as-isomorphism conception
is met at two levels by Aristotle™s theory.10

(iii.i) Isomorphism with respect to an item corresponding to the whole assertion.
In Metaphysics 7 Aristotle de¬nes truth and falsehood:
T 28 [That nothing can be in the middle of a contradictory pair, but it is necessary
either to af¬rm or to deny any one thing about one thing] is clear to whoever
de¬nes what truth and falsehood are. For, to say that what is is not, or that
what is not is, is false; to say that what is is, and that what is not is not, is
true. (1011b 25“7)11
In 8 Aristotle refers back to this de¬nition:
T 29 [. . .] one must argue on the basis of a de¬nition by assuming what ˜true™
and ˜false™ mean. But if what is true is nothing but af¬rming or denying,
and what is false too,12 it is impossible that everything should be false: for
it is necessary that the other13 member of the contradictory pair should be
true. (1012b 7“11)
T 29 shows that T 28™s de¬nition of truth and falsehood is an account of
what ˜true™ and ˜false™ mean.14 T 28™s de¬nition of truth and falsehood is
a premiss of an argument for Excluded Middle. This argument, which is
highly compressed, has been variously reconstructed. I shall not discuss it
here.15
T 28™s de¬nition of truth and falsehood can be interpreted in at least
four different ways.16

According to Heidegger ((1924/25), 127“8, 162“4, 166; (1926), 164; (1927), 226; (1930/31), 138; (1946),
163) Aristotle wavers between a correspondence theory of truth and a conception of truth as ˜uncon-
cealedness™. On Heidegger™s account of Aristotle™s views on truth see Luther (1966), 180; Krell (1975),
90“1; Berti (1990), 101; Vigo (1994), 75; Wolff (1999), 43; Berti (2000), 5, 17“18.
10 Cf. Wolenski/Simons (1989), 393.
11 Cf. Pl. Sph. 240d9“241a2; 263b4“10; Leal Carretero (1983), 41“3; Szaif (1998), 407. The order of the
cases in T 28 matches that at Int. 6, 17a 26“9: false denial, false af¬rmation, true af¬rmation, true
denial.
12 At 1012b 8“9 I adopt the reading ˜mhd•n Šllo t¼ ˆlhq•v £ j†nai £ ˆpoj†nai kaª t¼ ye“d»v –stin™,
which according to Alexander (in Metaph. 339, 18“20) is attested by some manuscript. Editors and
commentators prefer different readings or emend the passage.
13 For ˜q†teron™ (1012b 11) meaning ˜the other™ (rather than ˜one or the other™) see Int. 9, 19a 22; Po. 3.10,
1281a 27.
14 Cf. Metaph. 7, 1012a 3; 1012a 21“4. Aristotle holds that both in scienti¬c demonstrations and in
dialectical arguments some of the premisses can be de¬nitions ¬xing the meaning of terms: for
scienti¬c demonstrations see APo. 1.1, 71a 15“16; 10, 76b 15“16; 76b 19; 76b 20; 2.7, 92b 26“8; 10, 93b 29“
32 (cf. Demoss/Devereux (1988), 134“6; Goldin (1996), 44); for dialectical arguments see Top. 2.4,
111b 12“16 (cf. Cavini (1998), 7).
15 Cavini (1998), 7“14 expounds the earlier interpretations of Aristotle™s argument and offers a new
one.
16 Cf. Kirwan (1971/93), 117; Graeser (1981), 86; Ledda (1990), 36.
Truth as correspondence 133
(i) ˜To be™ is used in the existential sense, whereby it means ˜to exist™.
The de¬nition can then be paraphrased as follows: ˜To say of a thing which
exists that it does not exist, or of a thing which does not exist that it exists,
is false; to say of a thing which exists that it exists, or of a thing which does
not exist that it does not exist, is true.™17
(ii) ˜To be™ means ˜to hold of ™ and is properly applied to universals. The
de¬nition can then be paraphrased as follows: ˜To say of a universal which
holds of something that it does not hold of it, or of one which does not
hold that it holds, is false; to say of a universal which holds of something
that it holds of it, or of one which does not hold that it does not hold, is
true.™18
(iii) ˜To be™ is used in the predicative-elliptical sense, whereby it functions
like a schematic expression that could be expanded to ˜to be so-and-so™.19
The de¬nition can then be paraphrased as follows: ˜To say of what is so-and-
so that it is not so-and-so, or of what is not so-and-so that it is so-and-so,
is false; to say of what is so-and-so that it is so-and-so, or of what is not
so-and-so that it is not so-and-so, is true.™20
(iv) ˜To be™ is used in the veridical sense, whereby it means ˜to be true™
or ˜to obtain™, and is properly applied to states of affairs.21 The de¬nition
can then be paraphrased as follows: ˜To say of a state of affairs which in fact
“is” in the sense of being true that it “is not” in the sense of being false, or
of a state of affairs which in fact “is not” in the sense of being false that it
“is” in the sense of being true, is false; to say of a state of affairs which in

<<

. 4
( 11)



>>