Schmitt C.B. (ed.) The Cambridge History of Renaissance Philosophy
Подождите немного. Документ загружается.
Traditional logic
157
view
was a seventeenth-century phenomenon. Both Hobbes and Locke
have
been praised for their innovation and insight,
85
whereas at least in the
case
of
Locke,
very close correlations can be found between
parts
of his
discussion
of language and the discussion found in the scholastic logicians
who
were read at Oxford.
86
In taking the conventionalist position and
rejecting
Adamitic language, Locke was aligning
himself
with
a continuous
tradition stretching back through the Middle
Ages
to Aristotle, whereas
Leibniz's
interest in Adamitic language was a relative novelty.
The
second crucial doctrine was
that
spoken (and hence written)
language
corresponds to an inner mental language which is naturally
meaningful,
and without which no utterance could be significant. This
doctrine had its origin in De interpretatione 1 (i6
a
3~8) and was reinforced
both by St Augustine's discussion in De trinitate
xv.io—11
and by Boethius'
two
commentaries 6n De inter pretatione. The doctrine is common to
logicians
throughout the Middle
Ages,
and it continued
well
into the
seventeenth century, being found in both Locke and Leibniz. In order to
understand the doctrine several issues have to be considered: the
nature
of
signification,
the relation of words to both things and concepts, and the
supposed structure of mental language.
First, an account must be given
of
signification.
The word significatio is all
too easily translated as 'meaning', but as Spade has noted, 'signification is a
psychologico-causal
property of terms'.
87
One early account of significa-
tion was drawn from Boethius' translation of De interpretatione m (i6
b
i9),
which
suggested
that
to signify something was 'to establish an understand-
ing'
of it.
88
This definition leaves one in doubt both about the status of
syncategorematic terms (such as logical connectives) and about the status of
mental terms themselves, since no mental term can be said to signify in the
sense
that
it causes or brings about a concept. For this reason, the early
sixteenth-century author Raulin (c.
1443-1514)
explicitly stated
that
the
Boethian
definition applied only to a sub-class of signs.
89
Another early
account
of
signification
was drawn from St Augustine: 'A sign is something
which
is
itself
sensed and which indicates to the mind something beyond the
sign
itself.'
90
Although St Augustine's definition was frequently quoted, it
85. See
Aarsleff
1982, pp. 63, 284. Grazia 1980, p. 326,
specifically
mentions
Hobbes'
view
that
language
was
conventional,
although
created
by God, as if
this
were
a
novel
doctrine.
Cf.
Martinich
1981, pp. 351-2.
86. For a
study
of
Locke's
scholastic
sources,
see
Ashworth
1981a and
1984.1
do not, of
course,
wish
to
suggest
that
all
Locke
had to say
about language
was
traditional
in
nature.
87. Cambridge History 1982, p. 188
(Spade).
88. Ibid. 89.
Raulin
1500, sig. g
v
ra
"
rb
.
90.
Augustine
1975, P- 87. The
most
common
source
for the
definition
is
Augustine,
De doctrina
Christiana
11.1.
Cambridge Histories Online © Cambridge University Press, 2008
158
Logic
and
language
gave
rise
to the
same difficulties
as the
Boethian definition.
Accordingly
in
the fifteenth
and
early sixteenth century,
we
find
two
developments. First,
there
is a
widespread
use
of
Pierre
d'Ailly's
definition
of
signijicare:
'Now
to
"signify"
is to
represent
(a)
something,
or (b)
some things
or (c)
somehow,
to
a
cognitive power
by
vitally changing it.'
91
The
presence
of the
term
'somehow'
(aliqualiter)
was
expressly intended
to
extend
the
definition
to
syncategorematic terms. Second,
there
is a
close attention paid
to
different
kinds
of
sign. Pierre
d'Ailly
himself
had
noted
that
some signs, namely
mental terms
or
concepts, affect
the
cognitive power
by
virtue
of
their
nature,
92
and in
some early sixteenth-century authors such
as
Soto
we
find
a
careful
classification
of
both signs
and
types
of
signification.
Spoken words
were
said
to be
instrumental signs, because
of
their causal properties,
and
mental terms were said
to be
formal signs because they represented
by
their
very
nature.
93
These classifications were elaborated
in
later sources, such
as
the Coimbra commentary;
94
and in the
seventeenth century
John
of St
Thomas took
up the
topic
in
considerable detail.
95
Unless
one
realises
that
to
signify
is to
represent
or to
make known,
96
it is
impossible
to
understand
the
debate whether spoken words signify concepts
or things. This question first became popular
in the
late
thirteenth
century,
and
it
appears
in
commentaries
on the De
interpretatione through
the
Middle
Ages
and
into
the
seventeenth century.
97
It is
discussed
by
Toletus,
98
by
Sebastian Couto
in the
Coimbra commentary,
99
by the
Polish Jesuit
Martinus Smiglecius, whose Logica
was
first published
in
1618,
100
and in
John
of St
Thomas.
101
Echoes
of the
debate reappear
in
ordinary
logical
textbooks
such
as
Burgersdijk's
Institutionum
logicarum libri duo
102
and in
John
Locke's
Essay, when
he
writes 'Words
in
their primary
or
immediate
Signification
stand
for
nothing,
but the
ideas
in the
Mind
of him
that
uses
them.
. .'
(in.2.2). All
the
participants
in the
debate agreed
on
certain issues.
They
agreed
that
concepts play
an
essential role
in the
significative process,
for
we
cannot refer
to
objects
we do not
know;
nor can we
speak
91.
Ailly
1980, p. 16, §2. 92.
Ibid.,
p. 16, §1; p. 17, §8. 93. Soto 1529, ff.
v
ra
-vi
ra
.
94.
Collegium Conimbricense
1607 (part 11), cols. 7-33.
95.
John of St Thomas 1930, pp. 9—10, 646-722.
96. Pratus 1530, f. ii
va
, said that
signijicare
could be taken as equivalent to
manifestare,facereformare
or
facere cognoscere.
97.
For full
discussion
and references see Ashworth 1981a and forthcoming.
98. Toletus 1596, pp.
208-9.
99-
Collegium Conimbricense
1607 (part 11), cols. 34-45.
100. Smiglecius 1658, pp.
436-8.
101. John of St Thomas 1930, pp. 104-8.
102.
Burgersdijk 1637, p. no, wrote: 'Voces articulatae significant animi conceptus, primo scilicet,
atque
immediate: nam res etiam significant, sed mediantibus conceptibus.'
Cambridge Histories Online © Cambridge University Press, 2008
Traditional logic
159
meaningfully when
we
have nothing
in
mind.
103
Equally, everyone agreed
that
words,
at
least some
of
the time,
are
used
to
pick
out
things
in the
world.
If
I say
'Some human beings
are
running', what
I say is
true
(if at all) of
individual things
and not
of
my
concepts.
The
debate concerned
the
way
in
which
the
role
of
concepts
in the
significative process
was to be
described.
Some
authors
held
that
words primarily signified
or
made known concepts
and only secondarily signified things.
104
Others, such
as Eck,
held
that
words signified things alone while being subordinated
to
concepts.
105
Yet
others, including Couto, preferred
to say
that
while
the
signification
of
things
had a
certain primacy, both concepts
and
things were made known
by
words, which
thus
enjoy
a
double signification.
106
The
debate just described concerned categorematic words,
or
lexical
items.
The
role
of
syncategorematic terms, such
as
quantifiers
and
logical
connectives,
is
also
of
considerable interest,
but
only
a
handful
of
early
sixteenth-century logicians
at the
University
of
Paris seem
to
have discussed
the
matter
in any
detail.
One
important
writer is Fernando
de
Enzinas,
who
in
his
Tractatus
de
compositione propositionis mentalis discussed
not
only
the
role
of
syncategorematic
terms
in
mental language
but
also
the way in
which
the
relationship between
lexical
items
and
grammatical features
should
be
accounted
for at the
mental
level.
107
He
noted
that
different
languages dealt with case
and
number
in
different
ways,
and
suggested
that
these features should
be
represented
at the
mental
level
by
syncategorematic
acts,
analogous
to
those representing
the
force
of
quantifiers
and
logical
connectives,
rather
than
by
postulating different concepts
for
each
grammatical case
of a
lexical
item.
The
question
of
the
structure
of
mental language also arose
in
relation
to
the definition
of
a
standard
categorical proposition
as
having
a
subject
and
predicate, both
in the
nominative case,
and a
copula formed from
the
verb
'to
be'. The
main problems arose from sentences with so-called adjectival
verbs such
as
'runs',
or
indeed
any
ordinary verb other
than
the
verb
'to be',
103.
For a discussion of the role of speaker intentions, see
Collegium Conimbricense
1607 (part
11),
cols.
34-5 and 42-4.
Contrary
to Martinich 1981, pp.
127-8,
Hobbes was not the first to notice the
imporance of speaker intentions.
104.
This
seems
to have been a Thomist position: for some references see Ashworth forthcoming, n. 55.
John of St Thomas 1930 holds both that concepts are more immediately signified and that things
are more principally signified: see pp. 106-7. 105. Aristotle
1516—17,
1, f. lxxii
ra
.
106. Collegium Conimbricense
1607 (part
11),
cols.
39-40.
The notion of double signification was attacked
by
John of St Thomas 1930, pp. 105—6.
107.
Enzinas 1528. It was also printed in Paris in 1521, 1526 and 1528. For discussion see Ashworth 1982.
Cambridge Histories Online © Cambridge University Press, 2008
IÓO
Logic
and
language
sentences with impersonal verbs such
as
'pluit'
('It is
raining')
108
and
sentences containing deictic words such
as
'this'
or T.
109
There were
two
levels
of
discussion.
At the
practical
level,
the
problem concerned
how to
rewrite these sentences
so
that
they could
be
converted
for use in
syllogistic.
Medieval
logicians discussed conversion
at
great length.
One
good example
is
found
in
Marsilius
of
Inghen's commentary
on the
Prior Analytics,
in
which
he
considers such examples
as 'I am
Socrates', whose conversion
will
have
to be
'Socrates
is
I.'
110
Many
of
Marsilius'
examples were taken
up by
Nifo
over
a
century later
in his
commentary
on the
Prior Analytics.
111
However,
this
is one of the
topics
that
dropped from
view
during
the
sixteenth century. Later authors content themselves with
the
occasional
remark about problems
of
conversion,
112
and
tend
to
present their
syllogisms
already
in
standard form.
At
the
theoretical
level,
the
problem concerned
the
structure
that
such
sentences would have
in
mental language. Some debate took place during
the fourteenth
and
fifteenth centuries,
but
once more
the
most interesting
discussions
are
found
in
Fernando
de
Enzinas
and
other early sixteenth-
century writers
at the
University
of
Paris.
113
One
of
the important features
of
the debate
is
that
although they thought
that
mental sentences
did
have
a
structure, they also thought
that
logically
equivalent sentences could have
different structures. There
was
little indication
of
a
belief
that
any of
these
structures mirrored reality.
114
These
remarks have
by no
means dealt with
all the
problems
in the
philosophy
of
language
that
were discussed between 1350
and
1600.
I
shall
conclude
by
mentioning
two
other issues, both
of
which were treated
at
length
in
Gregory
of
Rimini's influential commentary
on the
Sententiae
of
Peter Lombard, which
was
written
in the
1340s.
One of the
questions
he
discussed
was how it was
that
a
mental proposition functions
as a
united
whole,
with
a
force
that
its
apparent
parts
taken separately
do not
possess.
Gregory
argued
that
there
were
no
parts
to
be
joined
together
at the
mental
108.
Natura
pluit
or some such locution is a standard expansion of
pluit:
see Marsilius of Inghen 1516, f.
n
ra
; Dorp in Buridan 1499, sig. a 5
rb
. These expansions were later taken up by humanist
grammarians:
see Padley 1976, p. 54.
109.
For a discussion and references, see Ashworth 1982.
no. Marsilius of Inghen 1516, f. n
ra
. His discussion of conversion is found in ff. 5
va
-n
ra
.
in. Nifo
1553a,
ff. 7
va
-2i
va
. He did not like Marsilius' treatment of'Ego sum Sortes': see f. i8
ra_rb
.
112.
Du Trieu 1662, first published 1614,
notes
on p. 3 that 'Deus est', 'Johannes studet', 'Petrus orat
Deum'
are implicitly equivalent to 'Deus est existens', 'Johannes est studens', 'Petrus est orans
Deum.'
113. See Ashworth 1982.
114.
This is important because there is a tendency in the literature to suggest that the very notion of a
mental language carries with it a belief that mental language mirrors the world and hence offers a
special insight into phenomena: cf. Padley 1976, pp. 63-5; Slaughter 1982, pp. 87-8.
Cambridge Histories Online © Cambridge University Press, 2008
Traditional logic
161
level;
and his opponents, who included Fernando de Enzinas nearly two
centuries later, argued
that
there
were parts, which were bound together by
syncategorematic acts, the mental correlates
of
syncategorematic terms.
115
This
topic did not altogether disappear during the sixteenth century, for it is
discussed by Martinus Smiglecius in his Logica;
116
and in Robert Pinke's list
of
questions on
logic,
metaphysics and ethics published in 1680 for the use of
Oxford
students, one of the logic questions was 'Is a mental proposition a
simple quality?'
117
The
other question popularised by Gregory of Rimini had to do with
complexe significabilia, i.e., 'propositions' in the sense adopted by some
modern philosophers.
118
According to medieval and post-medieval
logicians,
a proposition was an indicative sentence capable of being
true
or
false
which was the object
of
an act
of
assertion. As such it was firmly tied to
a particular occasion of utterance or thought. The question then arose
whether
there
was something which propositions signified, or something
which
was named by the dictum in an accusative and infinitive construction,
such as 'Petrum legere' in 'Verum est Petrum legere.' Gregory of Rimini
(following
Adam Wodeham) postulated complexe signijicabilia to
fulfil
this
function. He described them as eternal beings, neither mental nor physical
nor yet identical with God, which were at once the significates
of
indicative
sentences, the bearers of
truth
and falsity and the objects of knowledge or
belief
One opponent of Gregory's
view
was
Jean
Buridan, and many
subsequent authors including Paul of Venice followed Buridan in arguing
that
propositions could only have significates in so far as their subject terms
picked
out some objects in the world. Such a
view
has many difficulties,
particularly because many propositions
will
turn
out to have no adequate
significate.
Another
view
opposed to
that
of Gregory is found in some
Parisian authors, including Enzinas. He argued
that
propositions have no
significates
in the sense of things named or referred to. Instead they signify
aliqualiter; they function as syncategorematic
rather
than
categorematic
terms.
119
In accordance with this interpretation Enzinas recommended
replacing accusative and infinitive constructions with a
'that'
clause. Thus
'
"Human beings are animals" signifies human beings to be animals' should
be rewritten as ' "Human beings are animals" signifies
that
human beings
are animals', which removes the temptation to interpret the dictum as a
115.
For
details
of the
discussion,
see
Ashworth
1981b.
See
also
Nuchelmans
1980, pp.
27-44,
94—101.
116.
Smiglecius
1658, pp.
453-6.
See
also
John
of St
Thomas
1930, pp.
150-7.
117.
Pinke
1680, p. 12.
118. For
details
of the
discussion
see
Ashworth
1978c.
See
also
Nuchelmans
1980, pp.
45-73.
119.
Enzinas
1528, f.
xxxv
rb
~
va
.
Cambridge Histories Online © Cambridge University Press, 2008
IÓ2
Logic
and
language
name.
120
After
Enzinas,
the
question
of
what
a
proposition signifies seems
virtually
to
have disappeared from philosophical discussion.
THE
PERIOD
1530-1600
A
useful insight into
the
changes
in
logic
which took place during
the
sixteenth century
can be
obtained from
a
consideration
of
two groups
of
textbooks
—
four transitional texts
and two
texts which appeared
in the
Jesuit Ratio studiorum
at the end of
the century.
One
interesting transitional
text is
Nifo's
Dialéctica ludiera, which was first published
in
Florence
in
1520.
It
is an
introduction
to
logic
by a
skilled humanist
and
commentator
on
Aristotle
who was
also
well
acquainted with late medieval
logic.
All the
typically
medieval doctrines
are
discussed,
but
some, such
as
insolubles,
are
presented
in a
classical light; others, notably supposition theory,
are
truncated
in
order
to
serve
the
purposes
of
syllogistic.
121
A
more influential
text
is the
Logicae compendium
of
Javelli,
which
is
also called Commentarii
in
logicam Aristotelis.
122
Its
purpose
was to
introduce
the
student
to
Aristotle,
but
in
order
to do
this Javelli retained certain medieval doctrines, such
as
supposition theory, proofs
of
terms
and, to a
lesser extent,
the
theory
of
consequences.
His
work, first published posthumously
in
1551,
received
a
number
of
editions
in
Italy, France
and
Germany.
123
A
third
transitional
text is
the
Dialéctica
of
John
Seton, published
in
London
in
1545,
and
later
to
become
popular with
the
annotations
of
Peter Carter.
In his
preface Seton
refers
to
Agricola
and
Melanchthon,
and he
says
that
his purpose
is to
make
Aristotle
easier
for
adolescents.
124
He
discusses supposition theory
and he
mentions insolubles.
It is
interesting
to
note
that
book iv, which dealt with
the topics,
was not
included
in the 1545
edition.
The
fourth text,
the
Summulae
by the
Spaniard Domingo
de
Soto,
is
transitional
in the
fullest
sense,
for its
editions themselves exhibit
an
internal change which mirrors
the general changes taking place
in the
sixteenth century.
The
first edition,
in 1529,
was a
loose
but
lengthy collection
of
treatises.
It
opened with
Introductiones dialecticae, which began with
the
three
modi sciendi, definition,
division
and
argumentation,
and
then gave
a
full
treatment
of
terms
and
supposition theory. There
followed
a
commentary
on
Peter
of
Spain's first
tract,
on
nouns, verbs
and
propositions;
a
work
on
exponibles;
a
120. Ibid.,
ff.
xxxv
vb
-vi
ra
.
121. For discussion, see
Ash
worth
1976.
122.
I have inspected the 1560 Venice edition of the
Commentarii.
I assume the other editions do not
differ.
123.
For a list of editions see
Lohr
1977, pp.
731-2.
124.
Seton 1545, sig. A ii
v
: 'ut Aristotelem meo studio redderem faciliorem, et adolescentibus magis
familiarem'.
Insolubles are discussed under 'Dilemma', sigs. K vi
v
-viii
r
, and supposition on sigs. E
iiii
r
-v
v
.
Cambridge Histories Online © Cambridge University Press, 2008
Traditional logic
163
commentary
on
Peter
of
Spain's fourth tract,
on
syllogisms;
a
work
on
insolubles,
and a
work
on
obligations.
The
second edition (Salamanca 1539)
is
much briefer.
125
The
material
is
organised into
five
books.
The
first
is on
terms;
the
second, which includes remarks
on
definition, division
and
argumentation,
is on
propositions
and
supposition theory;
the
third
is on
oppositions,
i.e., the
relations between propositions;
the
fourth
is on
exponibles;
and the
fifth
on
syllogisms. There
follow
the
rewritten works
on insolubles
and
obligations.
In his
preface Soto explained
that
while
the
sophistic excesses
of
late medieval logic
had
made
it
horrifying
and
inaccessible
to
boys, there
was
still
a
need
for a
fairly detailed introductory
text.
It was not
enough
to
take
a few
examples from Cicero
and
think
that
one
had
given sufficient treatment
to
logic.
Aristotle himself was hard
to
read, making some preparation necessary. Moreover, preliminary exercises
were
essential,
and to that end
Soto
had
retained 'some quibbles
and a few
sophisms'.
126
The
point
that
Aristotle, while paramount, needed
an
introduction
was
implicit
in the
Jesuit Ratio studiorum
of
1586, which recommended
the
Summula
of
Fonseca
for its
breadth, clarity, relevance
to
Aristotle
and
lack
of
sophistry.
127
In the
Ratio studiorum
of
1599
the
name
of
Toletus
was added
to
that of
Fonseca.
128
Both authors were extremely popular throughout
Europe.
The
Introductio
in
dialecticam
of
Franciscus Toletus
was
first
published
in
1561
in
Rome,
and the
last
of
its eighteen editions appeared
in
Milan
in
1621.
129
The
Institutions dialecticae
of
Pedro Fonseca
was
first
published
in
Lisbon
in
1564,
and the
last
of
its fifty-three editions appeared
in Lyons
in
1625.
130
Both works included material
on
supposition theory
and consequences,
but in
tone
and
approach they
are
much closer
to the
textbooks
of the
first half
of the
seventeenth century than
to
those
of the
early
sixteenth century. Whether medieval material
was
included
in the
seventeenth-century texts,
as it was in Du
Trieu
131
and
Sanderson,
132
or
125.
The subsequent editions do not differ significantly from the second edition. I have inspected the
fourth edition of
1571
and the last edition of 1582, both published in Salamanca. See Soto 1554: the
title page says that it is the second edition and gives the date 1554; f.i6
r
has the heading 'Editionis
Tertiae
Summularum . . . Liber Secundus'; the end page, f.i59
v
, gives the date 1555.
126.
The introduction to the 1539 edition is reproduced in Soto 1554, sig. A 2
r_v
.
127. Monwnenta Germaniae paedagogica
1887, p. 131: 'Logicae Summula praemittatur, et Summula
quidem P. Fonsecae
esset
forte magis ad rem: quia latior, clarior, accommodatior Aristoteli, et
sine
tricis, quae et inutiles sunt, et deterrent tirones.'
128. Ibid.,
p. 332. 129. For details see Lohr 1982, p. 200.
130.
For full details see Fonseca 1964, 1, pp. xxxv-xlvi.
131.
Philip Du Trieu taught at Louvain before becoming a Jesuit. His
Manuductio
ad
logicam
first
appeared in 1614, and was frequently reprinted up to 1748: see Risse 1965.
132.
Robert Sanderson's
Logicae artis compendium
was first printed in Oxford in 1615, and was very
popular in England.
Cambridge Histories Online © Cambridge University Press, 2008
164
Logic
and
language
excluded
as in
Burgersdijk
133
and
Jungius,
134
it was
made clear
that
the
main business
of
logic
was to
discuss categories, propositions, syllogisms,
demonstration, topics
and
fallacies.
The
humanist
and
Ramist attempts
to
exclude
some
of
this material
had
failed,
135
and the
medieval material
which
remained
had
been integrated
in a way
which changed
its
significance.
There
are
three
features
of the
late sixteenth-century textbook which
particularly distinguish
it
from
the
early sixteenth-century
and
late
medieval
texts. First,
the
syllogism is presented as
the
focal point
of
the study
of
valid inference.
It is
not just
one
type
of
consequence among others,
as it
had been
in
Albert
of
Saxony's Perutilis logica,
136
or to be
passed over
rapidly
as it had
been
in
Paul
of
Venice's Logica parva. Second,
there
is a
complete absence
of
the sophisms which
had
formed
so
prominent
a
feature
of
earlier texts, whatever
the
topic.
For
instance,
in
early sixteenth-century
Parisian treatises
on
syllogisms,
the
standard procedure
was to
test each
syllogistic
mode
by
examining elaborate counter-examples.
To
quote just
one case,
in his
treatment
of
Barbara,
Dolz
discussed
the
apparent counter-
example:
'Only what begins
to be
every man's begins
to be a
man's donkey,
and every braying thing except Brownie
is
beginning
to be a
man's
or
will
be immediately after this; therefore every braying thing except Brownie
begins
to be a
man's donkey.'
137
Third,
there
is a
completely
new
attitude
to language, which
is
signalled both
by the
general disappearance
of
sophisms
and by the
disappearance
of
special problem cases from
the
discussion
of
such matters
as
exponibles
138
and
conversion.
139
Logicians
in
the medieval tradition
had
been particularly concerned with
the
effects
of
133.
Franco Burgersdijk's
Institutionum
logicarum libri
duo first appeared in 1626. In 1635 the Estates
ordered that it should be used by all schools in the Netherlands.
134.
Joachim Jungius'
Logica Hamburgensis
was published in 1638. Risse 1965 also
lists
a 1635 edition,
which contained the first three books.
135.
See Ashworth
1974a,
pp. 16-17 for a brief account of the Philippo-Ramist school of the 1590s,
which aimed at a reconciliation of Melanchthon's Aristotelianism with the doctrines of Ramus,
and which reintroduced first the categories, then the equivalence, conversion and opposition of
propositions, together with the standard syllogism.
136.
Albert
of Saxony 1522, f. 28
vb
: 'Postquam visum est de consequentiis simplicibus formalibus, nunc
videndum est de consequentiis formalibus syllogisticis.'
137.
Dolz
1511,
sig. c i
vb
: 'tantum cuiuslibet hominis incipiens
esse
incipit
esse
hominis asinus et omne
rudibile prefer brunellum hominis incipiens
esse
est vel immediate post hoc erit ergo omne rudibile
preter brunellum incipit
esse
hominis asinus'. Another good example is found in Antonio Coronel
1517,
f.
xvi
vb
,
where he gives this counter-example to Ferio: 'Cuiuslibet hominis b. asinus et asinus
non sunt asini alicuius angeli, omnes asini hominis quilibet asinus et quilibet asinus sunt: ergo
alicuius angeli omnes asini non sunt asini.' (Note the use of the special quantifier 'b'.) This is the
kind
of thing that made humanists froth at the mouth. Cf. Vives
1979a
p. 53.
138.
For a succinct account of the treatment of exponibles in the sixteenth century, see
Cambridge History
1982,
p. 792 (Ashworth). See also Ashworth
1973b.
139.
For references to the discussion of conversion, see nn. no-n above.
Cambridge Histories Online © Cambridge University Press, 2008
Traditional
logic
165
word
order
and of
various
syncategorematic
terms,
including
exponible
terms
themselves
as
well
as
quantifiers
of
one
sort
and
another,
on
sentence-
meaning.
They
had
tried
to
express
semantic
differences
through
different
syntactic
structures.
For
instance,
'Papam
vidi'
('The
pope
I
saw')
was
said
to
differ
in
meaning
from
'Vidi
Papam'
('I saw the
pope').
140
Partly
as a
result
of
humanist
attacks,
141
these
attempts
ceased
completely,
and the
only
examples
admitted
to
logic
texts
were
already
couched
in a
standardised
form.
I
shall
now turn to
consider
the
process
of
integration
as it
affected
both
specifically
medieval
doctrines
and
Aristotelian
syllogistic
itself.
Some
doctrines
were
never
integrated.
For
instance,
insolubles
were
not
seen
to
have
any
role
in
syllogistic,
or the treatment
of
topics
and
fallacies;
and
most
authors
did not
mention
them.
Those
who did
tended
to use a
classical
vocabulary,
speaking
of
inexplicabiles
rather than of
insolubilia,
and
citing
such
classical
puzzles
as the
crocodile
or the
story
of
Protagoras
and
Euathlus.
142
Soto
had
given
a
fairly
full
account,
but had
remarked
that
there
was
little
utility
in the
material.
143
At the end of the
century
the
only
interesting
discussion
is
found
in
Thomas
Oliver
who,
perhaps
out of an
antiquarian
rather than a
philosophical
interest,
gave
a
survey
of the
literature
including
Heytesbury,
Buridan
and
Marsilius
of
Inghen.
144
One
reason
for the
general
disappearance
of
insolubles
may
well
be that
medieval
logicians
themselves
had
tended
to see
them
only
as
interesting
puzzles,
rather
than as
paradoxes
which
strike
at the
very
heart of our
semantic
assumptions.
145
They
were
thus
thrown
out
together
with
the
other
sophisms.
Obligations
are another
doctrine
which
failed
to be
integrated
in the
late
140. For this example, drawn from appellation-theory, see Ashworth 1974a, p. 95. For Vives' reaction,
see Vives 1979a, p. 137. For More's reaction, see
ibid.,
p. 173. Soto did admit the case, even in later
editions: see Soto 1554, f. 5i
vb
: 'Nec sequitur, Papam vidi, ergo vidi Papam.'
141.
For instance, Nifo 1553a, f.
I7
ra
~
va
,
attacked the rules of the
Sorticolae
for dealing with negative
propositions in which the predicate was said to be undistributed, because it preceded the negative
term,
e.g., 'Quidam homo animal non est' (f. 17"). He denied that the syntactic order affected the
semantic structure: 'In hac autem, quidam homo animal non est, licet negatio
situ
sequatur ilium
terminum animal, intellectu praecedit: quia, si
ille
terminus animal est praedicatum, oportet ut
intellectu sequatur' (f.
I7
va
).
Cf. Vives on word order, in Vives 1979a, p. 123.
142.
For a full account and references see Ashworth 1972, pp. 35-7. See also Nifo 1521b, ff.
I56
r
-63
r
,
'De
dilematibus sive insolubilibus'.
143.
Soto 1554, f. 15i
rb
: 'Sed Moderni, praeter rei dignitatem, dilataverunt istam materiam: in qua
tamen re
vera,
parum est utilitatis, et ideo nos earn breviter transcurremus.'
144. Oliver 1604, pp. 8-14. Oliver is famous for his illustration of the written liar on p. 9: see Ashworth
1974a, p. 114 for a reproduction. For discussion, see I. Thomas 1965.
145.
Cambridge
History
1982, p. 253 (Spade): 'the medievals did not
seem
to have had any "crisis
mentality" about
these
paradoxes. Although they wrote a great deal about them there is no hint
that
they thought the paradoxes were crucial test
cases
against which their whole logic and
semantics might fail . . . [they] did not draw great theoretical
lessons
from the insolubles.'
Cambridge Histories Online © Cambridge University Press, 2008
Logic
and
language
sixteenth-century textbook.
The
traditional theory
had
been presented
by
various
authors
in the
first decades
of the
century, some
of
whom were
notable
for
their adherence
to
Marsilius
of
Inghen.
146
The
latest exponents
were
Sbarroya
and
Soto.
147
In the
first editions
of
his Summulae, Soto
had
already
omitted
a
number
of
standard sophisms, especially those
to do
with
the imposition
of
terms,
i.e., the
introduction
of
arbitrary
new
mean-
ings,
148
and at the end of the
treatise
he
described
the
topics discussed
as
trifling.
149
In the
third
edition
he
was considerably more forthright.
At the
beginning
of the
much shortened treatise
he
described obligations
as 'a
game
for
boys';
150
and at the end he
said
'It
is shameful
to
pursue this kind
of
filth and these dregs
of
nursery songs, which
can
produce only nausea.'
151
In order
to
understand what replaced
the
theory
of
obligations
in at
least
some
textbooks,
we
must
go
back
to the
works
of
Josse
Clichtove,
who
described
two
quite different kinds
of
disputation.
152
In
his commentary
on
the obligationes
of
Jacques Leievre d'Etaples,
he
presented
the
standard
doctrines,
albeit completely shorn
of
their sophisms,
and
with
no
reference
to
the
imposition
of
terms.
153
He
made
it
quite clear
that
the
purpose
of
obligational
disputations
was to
test
the
ability
of
boys
to
handle
logical
inferences.
At the
same time,
in his own two
works
In
terminorum
cognitionem introductio
and De artium
scientiarumque divisione introductio
he
described
what might
be
called
a
doctrinal disputation.
154
The
difference
between
the two
kinds
of
disputation
can be
briefly summarised.
In an
obligational
disputation
the
opponent begins
by
putting forward
a
false
proposition.
In the
most usual kind
of
obligational disputation (positio)
the
respondent
had to
grant this initial positum.
155
The
opponent then
put
146.
E.g., Thomas Bricot, John
Mair
and Domingo de Soto
himself.
The hallmark of Marsilius'
doctrine was his division into three types of obligational disputation,
positio, depositio
and
dubiepositio,
in accordance with the three possible ways in which a proposition could be presented.
Marsilius'
Tractatus
de
arte obligandi
was printed in Paris in 1489 under the name of Pierre d'Ailly.
147.
Sbarroya 1533, ff. lxviii
rb
-ix .
148.
Soto
1529,
f. cl
ra
, wrote 'Nempe quamvis doctrina haec non nihil possit summulistae conducere
(nam de praedestinatione, caeterisque id genus disputanti vix scias
nisi
obligatorie respondere),
impositiones tamen et casus qui hie plurimi multiplicari solent, indigni revera sunt, in quibus
extricandis et tempus et operam perdas.'
149.
Ibid.,
f. cliii
vb
. His closing words are 'Sed satis nobis sit, huiusmodi naeniis hactenus
lusisse.'
150.
Soto 1554, f.
I56
rb
:
'Est enim ars haec velut puerorum ludus.'
151.
Ibid.,
f.
159
vb
.
His closing words are: 'At vero, pudet id genus sordes ac naeniarum faeces persequi:
quae non
nisi
nauseam movere possunt.'
152.
For a full discussion and references, see Ashworth 1986.
153.
See Clichtove's Ars
obligationum
in Lefevre d'Etaples 1520, ff. ioi
v
-34
r
.
154.
Ibid.,
ff. 2
r
-i
9
v
.
155.
Strictly speaking, the respondent had to rule on the admissibility of the
positum
before he accepted
it. Clichtove omitted this step of
admissio.
In other forms of obligational disputation, the
respondent could require that the initial proposition be denied
(depositio)
or doubted
(dubiepositio).
Cambridge Histories Online © Cambridge University Press, 2008