Mind Association

The Re-Discovery of the Topics

Author(s): Otto Bird
Source: Mind, New Series, Vol. 70, No. 280 (Oct., 1961), pp. 534-539
Published by: Oxford University Press on behalf of the Mind Association
Stable URL: http://www.jstor.org/stable/2251609 .
Accessed: 26/10/2013 22:14

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

Oxford University Press and Mind Association are collaborating with JSTOR to digitize, preserve and extend
access to Mind.

http://www.jstor.org

This content downloaded from 128.143.23.241 on Sat, 26 Oct 2013 22:14:01 PM

All use subject to JSTOR Terms and Conditions
 VIII.-DISCUSSIONS

THE RE-DISCOVERY OF THE TOPICS

Professor Toulmin's Inference-Warrants

THE view is gaining some currency that laws of nature may be

profitably considered as inference-licences. Professor Ryle advanced
it in the metaphor of inference-tickets (R. 121).1 Professor Toulmin,
on his own admission, applied Ryle's ideas to the elucidation of
physical science in his Philosophy of Science (T. 260). Much the
same idea was put forward by Schlick, who told Professor Popper
that he owed it to Wittgenstein (P. 37, n. 7). Ramsey also said
much the same thing (RS. 241). More recently, Toulmin has launched
what amounts to a full-scale exposition of inference-warrants in his
study on the Uses of Argument.
This development, particularly in the form it takes in this last
book, has many similarities with the analysis of the Topics in
medieval logic. The resemblance is so close, as I hope to show, that
it appears we are witnessing something of a re-discovery of the Topics.
To make manifest the resemblance I shall describe first what Toulmin
has to say about inference-warrants, particularly in the third essay
of his book. I shall then present the medieval doctrine on the
Topics as found in the tradition that stretches from Abelard to
Albert of Saxony in the fourteenth century.

Professor Toulmin's Warrants

Concerned with what he calls " working logic " rather than with
the " idealized logic " of the formal logicians, Toulmin prefers to deal
with examples expressed in concrete terms. Thus for the analysis
of the inference-warrant he gives the following examples:
1. Harry's hair is red, so it is not black.
2. Petersen is a Swede, so he is almost certainly not a Roman
Catholic.
3. Harry was born in Bermuda, so he is a British subject.
In each of these we find what is called a Datum and a Claim. In
the second part of each statement we have the " claim or conclusion
whose merits we are seeking to establish (C) ", and in the first part
" the facts we appeal to as a foundation for the claim . . . our
data (D) " (T. 97).
However, of each of these statements we may still ask, how we
get from the Datum to the Claim. To answer this " we must bring
forward not further data, for about these the same query may
immediately be raised again, but . . . rules, principles, inference-
licences, . . . general, hypothetical statements which can act as
bridges, and authorize the sort of step to which our particular
These referencesare to the books listed at the end of this paper.
534

This content downloaded from 128.143.23.241 on Sat, 26 Oct 2013 22:14:01 PM

All use subject to JSTOR Terms and Conditions
 THE RE-DISCOVERY OF THE TOPICS 535
argument commits us " (T. 98). These are called Warrants(W).
The following statements are offered as warrants for the three
argumentsabove:
1*. If anything is red, it will not also be black.
2*. A Swede can be taken almost certainly not to be a Roman
Catholic.
3*. A man born in Bermudawill be a British subject.
In general then we get the following diagramfor the " layout of
arguments" (T. 99):
D -?So C
Since
W
However, for a full picture of the layout we must amplify the
diagram. We may be asked " why in generalthis warrantshouldbe
accepted as having authority ", and to this we will produce " other
assurances,without which the warrants themselves would possess
neither authority nor currency". Such assurances are called the
Backing(B) of the warrants(T. 103). As examples of such, he cites
for 2 and 3:
2**. The proportion of Roman Catholic Swedes is less than
2 %.
3**. The statutes and other legal provisionsgoverningthe nation-
ality of personsborn in the British colonies.
No backingis given for the first argument,presumablyon the ground
that the warrant itself is " self-authenticating" (T. 98). The
Backing, it is noted, can be expressed in the form of categorical
statements of fact, like D, and unlike W (T. 105-106).
Furthermore,we may want to qualify the degree of force which
the data confer on the claim in virtue of the warrant, which is
accomplished by such " modal qualifiers" as 'necessarily',
' probably', ' presumably'-which will be indicated by Q. It may
also be necessaryto state the conditionsfor the exception or rebuttal
(R) of the warrant as indicating the circumstancesin which the
authority of the warrantwould have to be set aside, i.e. statements
about its applicability(T. 100-102).
With these amplifications our diagram will now appear thus
(T. 104):
D- >So, Q, C
l I
Since Unless
W R

On account of
B

This content downloaded from 128.143.23.241 on Sat, 26 Oct 2013 22:14:01 PM

All use subject to JSTOR Terms and Conditions
536 0. BIRD:

For concrete illustration Toulmin returns to the case of Harry

who was born in Bermuda (T. 105). However, it should also hold
for the example from geometric optics which he exploited in his
Philosophyof Science (T* 24-25). Applying these distinctions, we
would then get somethinglike the following:
4. (D) The sun is at an elevation of 30? and a wall is 6' high,
(C) So the wall will cast a shadow of 10' 6".
4*. (W) By the Principle of the Rectilinear Propagation of Light,
the sun's rays can be representedas a straightline and the
shadow of the wall can then be computedby trigonometry.
4**. (B) Particular cases in which this method of computation has
been shown to work-Newton's " renderinga proposition
generalby induction ", which ProfessorToulmininterprets
as using our observationsof regularitiesand correlations
CC

as the backing for a novel warrant" (T. 121).

4. (Q) In this case, a statement of the margin of errorthat would
be tolerated.
4. (R) The exceptions, e.g. diffraction, refraction, in which the
warrant(4*) does not hold.
This is a more elaborate analysis than Tou'min made in his earlier
work. Yet as ProfessorCooleyhas indicated, it would still seem to
be an over-simplificationwhen applied to the analysis of even so
simple a law as that of Snell; since it is not immediately clear
whether such relevant items as the success of comparablelaws, its
tie-up with the wave-theoryof light, and its predictionrecordare to
be counted as D, W, or B (CL, 312-315). However, this is not
immediately relevant to our concern of the similaritv of Toulmin's
analysis with the Topics, to which we can now turn.

The Topics
The examples make it clear that Toulmin is primarilyconcerned
with argumentswhich derive at least some of their argumentative
force from relations of meaning among the non-logical words-e.g.
Swedes and Catholics, native Bermudans and British citizens. In
fact, he declaresthat the backingof an argumentis " field-dependent"
in that it " varies from one field of argumentto another" (T. 103).
This is to say, in terms of the medieval logical analysis, that he is
concerned with material rather than with formal consequence.
'Formal' in this connection has to do with the syncategorematic
terms, such as the connectives, ' and ', ' or)', ' if . . . then ', ' not ',
and the quantifiers' all' and ' some', whereas ' material' refers to
the categorematicterms (AS. f. 24rb). The logical study of material
consequence,i.e. of logical consequencethat depends in some way
upon the categorematic terms, was for medieval formal logic
primarilythe study of the Topics.

This content downloaded from 128.143.23.241 on Sat, 26 Oct 2013 22:14:01 PM

All use subject to JSTOR Terms and Conditions
 THE RE-DISCOVERY OF THE TOPICS 537
The medieval Topics tradition, which stems from Boethius, drew
a two-fold distinctionin a Topic:
(1) The Topical Maxim, or rule accordingto which one can make
an inference-the maximapropositio.
(2) The Topical Difference, or the 'matter' from which the
Maximis composed-the differentiamaximaepropositionis.
What these two are and how they function may be seen from the
analysis of a typical Topical argument:
5. If it is a man, it is an animal.
(Abelard regularly cites Topical arguments as conditional proposi-
tions with 'if ', although later, e.g. in Peter of Spain, Albert of
Saxony, it was customaryto cite them as two propositionsconnected
by ' therefore').
5*. TopicalMaxim: Of whatever the species is predicated,so is
the genus (De quocumque
predicaturspecies,
et genus).
5**. TopicalDifference: 'Man' is related to 'animal' as
species to genus.
The secondof these is calledthe ' Difference' becauseit differentiates
one Topic from another. So in this case we have the Topic of Genus
and not the Topic of Definition, Whole, Part, Opposites,etc. Both
Maxim and Difference are called 'Topics', Peter of Spain writes,
"because both confer validity (firmitatem) on an argument"
(PS. 5.07). A Topic itself is most brieflydescribed,after Cicero,as a
'basis of argument' (sedesargumenti).
Now I would suggest that in the Topical Maxim and Difference
we have the traditional logical counterpartsof Toulmin'sWarrants
and Backing. The Maximperformsthe same function as a warrant
and to describeit as " a confirmatoryrulethat provesa consequence",
as Albert of Saxony does (AS. f. 33ra) is nothing Toulmin should
object to, except perhapsfor its departurefrom the sound of ordinary
language. The Differenceby its name indicatesthat diverseMaxims
rest on differentrelations. Toulmin himself cites the genus-species
relation as backingfor a warrant,when he says that 'All whales are
mammals' as a warrant for an argument between two singular
statements has as its backing, 'The class of B's includes taxonomi-
cally the entire class of A's (T. 112).
Furthermore,the first of Toulmin'sexamples cited above (i.e. 1)
is itself an instance of one of the traditional Topics, that namely
from opposites. The medieval logician, however, would push
the analysis of it further along the same lines than Toulmin seems
willing to go, since as we have seen, he apparently holds that its
warrant, in being " self-authenticating", needs no backing. A
Topical analysis of this examplewould appearsomethingas follows:
1'. If Harry'shair is red, it is not black.
1*. Maxim: Positing one contrary removes the other from the
same.

This content downloaded from 128.143.23.241 on Sat, 26 Oct 2013 22:14:01 PM

All use subject to JSTOR Terms and Conditions
538 0. BIRD:

1**. Difference: 'Red' and 'black' are related as contrary

opposites. This involves recognizingthat they are opposites,
sorting out what kind of opposites they are (contrary,
relative, privative, contradictory),so as to know that they
are such that both cannot be present at. once in the same
subject; hence that the Maxim applies.

The FormalInterestof the Topics

Enough has been said to show the similarity between Toulmin's

analysis of warrants and backiing and medieval, Topical logic.
However,a short word may be in orderabout the differencebetween
them, even though this is to broachwhat is perhapsthe main thesis
of Toulmin'sbook, i.e. the over-formalizationof logic; and this is
indeed anotherand bigger story.
The Topics during the great period of medieval logic, say from
Abelard to Paul of Venice, were studied primarilyfor their formal
interest. In fact, a Topic as a 'material consequence' is to be
understood with reference to a 'formal consequence'. Thus the
Pseudo-Scot writes that whereas a formal consequence is " one
that holds for all terms when there is a similardispositionand form
of the terms," a material consequenceis " one which does not hold
for all terms retaining a similar disposition and form but with a
variation of the terms " (S. 287B). However, and this provides
the centre of formal interest in the Topics, a material consequence
" can be reduced to a formal consequencethrough the assumption
of a necessary proposition" (S. 288A). How this is done can be
seen from our example No. 5:
5. If it is a man, it is an animal.
If we add to this the proposition:
5'. Man is a species of animal
we get the following ' formal consequence':
5'. If man is a species of animal, then if it is man, it is animal.
This consequenceis ' formal ' because it holds good in virtue of
the form regardlessof the terms that realize it. It can accordingly
be formalizedwith variables. Interpretingthe relation in terms of
class-inclusionand class-membership,5' then becomesthe following
thesis of the logic of classes:
5"*. ocC : D: (x) . xEoc. xEp
In effect, we now have an abstract statement of the Topical Maxim
5*. Taking this and the Difference (5** or its equivalent 5') as
asserted,we may detach and assert5 by itself. In this way the whole
Topical structure can be viewed as an inference-scheme,as in fact
Abelarddoes.
Abelardthroughouthis long study of the traditionallist of Topics
(A) is mainly concernedto discoverthose which exhibit such formal

This content downloaded from 128.143.23.241 on Sat, 26 Oct 2013 22:14:01 PM

All use subject to JSTOR Terms and Conditions
 THE RE-DISCOVERY OF THE TOPICS 539
relations and to distinguish the correspondingnecessary Maxims
from those which are probable only. John of Salisbury, while
admiringhis logical acumen, complainsthat he is interested only in
necessity (J. III. 6). Toulmin sounds an echo of 'this complaint,
though with something less than admiration for the formal possi-
bilities. But that more has been done in the analysis of warrants
and backings than Toulmin seems to realize is clear from even so
brief a view of the medieval Topical tradition.

Reference8

A Abelard, Dialectica,ed. L. M. De Rijk, Assen, 1956.

AS Albert of Saxony, Perutili8 Logica, Venice, 1523 (courtesy of the
Bodleian).
CL J. C. Cooley, " On Mr. Toulmin's Revolution in Logic " in Journal
of Phiio8ophy, lvi (1959), No. 7, 297-319.
J John of Salisbury, MIetalogicon,ed. C. C. J. Webb, Oxford, 1929.
P K. R. Popper, The Logic of Di8covery. London, 1959.
PS Peter of Spain, SummulaeLogicale8,ed. I. M. Bochenski, Rome, 1947.
R G. Ryle, The Concept of Mind, London, 1949.
RS F. P. Ramsey, The Foundation8 of M3athematic8,London, 1931.
S Pseudo-Scot, Quae8tione8 super libro8 Priorum Antalyticorum, in Duns
Scotus, OperaOmnia,ed. Durand, Lugduni, 1639, T.1.
T S. E. Toulmin, The U8e8of Argument,Cambridge,1958.
T* S. E. Toulmin, The Philo8ophy of Science, London, 1953.
OTTo BIRD
Univer8ity of Notre Dame,
Indiana

This content downloaded from 128.143.23.241 on Sat, 26 Oct 2013 22:14:01 PM

All use subject to JSTOR Terms and Conditions