critical 
• review 

Taulminls Bald 
Experiment 

Ralph H. Johnson 
University of Windsor 

A Critical Review and 
Appreciation of 

An Introduction to Reasoning 
by Stephen TOulmin, 

Richard Rieke, and Allan Janik 

I. INTRODUCTION 

What would your first instinct be upon 
hearing of an introductory logic text that 
made virtually no mention of staples such as 
premise and conclusion, validity and invalid-
i~y, ~nductive and deduct~ve arguments? You 
~ght think that it ought to be condemned to 
the flames as containing nothing but sophistry 
and illusion. But hold! Were you to follow 
that inclination, you would be missing out on 
an important and challenging text. Its name 
is An Introduction to Reasoning; its authors 
are-Professor Stephen Toulmin and his cohorts, 
Richard Rieke and Allan Janik.l 

There have been precious few innovations in 
argument analysis since the time of Aristotle. 
Most logicians have been content to remain 
within the perimeter of the standard concep-
tual framework of logic, analyzing arguments 
into premises and conclusions; classifying 
them as either deductive or inductive, and 
upholding soundness as the ideal for deduc-
tive arguments. In the last 25 years, there 
are indications of changes underway that 
might serve to weaken the hold of the tradi-
tional framework. In another place, Profes-
sor Blair and I have attempted to identify 
some of these changes--most of them within 
the orbit of what we call "informal 10gic."2 

The main stimulus for these changes has, it 
seems, been pedagogical. Students in the late 
60s began to demand "relevance," and their 
logic teachers--at least some of them--were 
sensitive to these demands. Students wanted 
(sometimes for political reasons) an introduc-
tion to logic that equipped them to enter com-
bat with real argurnents--not the "Dick and 
Jane" var~ety found in so many logic primers. 

16 

For their part, we may presume that teachers 
were finding themselves increasingly uncom-
fortable teaching formal methods of analysis, 
and began to discover the enormous problems 
involved in the attempt to apply those meth-
ods to real arguments set in context of pres-
sing social and political affairs. Some of 
them began to tinker around, and because the 
only avenue available to them were textbooks, 
it turns out that textbooks have played a 
significant role in the recent development of 
informal 10gic. 3 

Toulmin's text continues this trend, for it 
is obvious that pedagogical concerns have in-
fluenced him, too. In the ~aching Guide, he 
says: 

This is little more than an updating 
of a part of the trivium of the time-
honored liberal arts so as to meet the 
contemporary challenge to philosophers 
and educators, that is, to educate a 
generation of students whom Anthony 
Burgess has aptly described as having 
"experienced everything and read 
nothing." C!Q, 1-2) 

Although pedagogical goals have fuelled the 
engine driving recent experiments, it must be 
noted that theoretical issues have quickly 
emerged. Indeed, it has become evident that 
if informal logic brackets the canons of for-
mal logic and addresses itself seriously to 
the task of analyzing and criticizing real 
arguments Cas our pedagogical values demand), 
then we will find that we lack anything sub-
stantially new in two important areas: Cl) 
~ theory of arsueent and (2) ~ theory of 
cr~ticism. So ~t ~s the strong ~nterplay 
between pedagogy and theory that has invested 
informal logic with much of its vitality and 
growing appeal to logicians. 

All this being so, it was with the greatest 
anticipation that I received, last spring, a 
copy of Toulmin's new text. Not only is he 
himself a highly respected philosopher~ but 
his earlier monograph--The ~ ~ Argument;-
was one of few serious attempts ~n the l~ter­
ature of logic to address itself squarely to 
the adequacy of the traditional framework. 

1 

That alone gives Toulmin at least partial 
claim to the title "the grandfather of us all." 
In that work, Toulmin wrote: 

In tackling our main problems about 
the assessment of arguments, it will 
be worth while to clear our minds of 
ideas derived from existing logical 
theory, and seeing by direct inspec-
tion what are the categories in terms 
of which we actually express our 
assessments, and what precisely 
they mean to us. (6-71 

How terribly sane and insightful the thought! 
In effect, Toulmin was saying that it was 
about time that logiCians became more empir-
ical and looked more carefully at the problem 
of analyzing real arguments. His own pro-
posal, forming the core of that work, was 
that we can best understand the process of 
critical assessment of arguments by means of a 
jurisprudential model rather than a mathe-
matical one. In An Introduction to Reasoning, 
some 22 years later, Toulmin has undertaken 
the task of translating that proposal from a 
theoretical to a pedagogical setting. 



In the intervening years, we have seen in-
itiatives from other logicians who (whether 
knowingly or not) were rallying to the cry 
Toulmin had issued in The ~ of Argument. 
What he predicted there-Kas come to pass, for 
at the end of his introduction, he wrote: 

But of one thing I am confident: 
that by treating logic as generalized 
jurisprudence and testing our ideas 
against actual practice of argument 
assessment rather than against a 
philosopher's ideal, we shall eventu-
ally build up a picture rather dif-
ferent than the traditional one. (10) 

Just such a picture has begun to emerge, al-
though its outline and finer details remain 
blurry. Informal logic now finds itself in 
the very sort of situation Kuhn describes in 
The Structure of Scientific Revolutions. The 
old parad~gm is-los~ng, or has lost, its grip; 
several new ones have been developed; none 
has thus far managed to carry the day and win 
allegiance. It is in this context, I believe, 
that we must place Toulmin's text. Is this 
the new paradigm for which ~ have been 
searCliTng? 

I hope to be able to give this text the 
careful scrutiny it deserves, though I very 
much fear that I shall not succeed. For the 
reviewer must be in a position to appraise 
both the theoretical and pedagogical adequacy 
of the text, and I'm afraid I must admit to 
limits on both counts. To take the matter of 
pedagogy first, I must confess that I have 
not had the opportunity of teaChing from this 
text and therefore have no solid basis for 
gauging its merits ~ ~ text; i.e., how do 
students respond to the text? what sorts of 
problems might one encounter in using it? On 
such questions, I shall have to pass, con-
tenting myself with the hope that someone 
who is qualified will, in a future issue of 
the Newsletter, take them up. Then there is 
the matter of the conceptual framework Toul-
min introduces, about which I make these pre-
liminary observations. It is a radical de-
parture from the traditional framework. It 
is fascinating, fresh, and anything but a 
rehash of the same old tired approach that 
has occupied centre stage for so long. 

To give the reader some sense of what Toul-
min is about, I shall quote three passages 
from the Teaching Guide: 

Our conception of logic is based on 
the notion that reasoning is essen-
tially a matter of dialogue rather 
than monologue, in which certain 
techniques come to be mastered. ell 

Among the most important consequences 
of the notion that reasoning is to be 
taken as a human transaction rather' 
than as the analysis of arguments in 
terms of canons that establish their 
validity or invalidity for all time, 
is the correlative notion that argu-
mentation (and indeed rationality 
itselfL is an "open-textured" 
activity. Only in mathematics and 
formal logic do we encounter reason-
ing that can be said to reach a con-
clusion, a finally completed state. e21 

Because a Socratic approach is sug-
gested throughout, it will be helpful 
to look at some possible topics of 

17 

discussion that can be stimulated in 
the opening session and carried 
throughout the remainder of the 
course. An Introduction to Reasoning 
aims at an-articulation o~the stages 
througn which actual arguments and 
discussions go. e2l 

These passages make plain that Toulmin has, 
quite self-consciously, attempted to develop 
an approach to logic that is, for all practi-
cal purposes, independent of the assumptions 
and concerns of formal logic. We cannot, I 
think, underestimate either the boldness or 
the significance of this experiment. Most 
texts in the recent batch of informal logic 
texts have, for all their merits, in one way 
or anotner retained significant elements or 
assumptions of tne traditional framework, 
even when they have tried valiantly to over-
come them. 4 Toulmin has taken a different 
road, thougn one will surely see its philo-
sophical antecedents in the writings of 
Peirce and Wittgenstein, to name but two. 
Althougn Toulmin does not say it explicitly, 
I think what ne is providing us with here is 
a new paradigm for (informal) logic, some 
aspects of wnich are evident in the passages 
quoted. First, Toulmin has chosen to inves-
tigate reasoning and argumentation as prfi-
cesses (rather than as products) and so as 
devised a model for understanding them which 
is dialoyical (~ather than solopsistic) and 
d~nam~c rather than static). Second, Toul-
m~n has decided to look at the processes of 
reasoning and argument in law and science 
(principallyl, finding in them better ana-
logues for now reasoning is developed and 
criticized than he believes can be found in 
the traditional analogates: geometry, mathe-
matics, and formal logic. 

Hence, the appropriate matrix for this re-
view cannot be the conventional one: Is this 
a good introductory logic text? For built 
into that very question are a host of assump-
tions about what logic is! Instead, I pro-
pose that the right question to ask here is: 
How successful is this text in attempting to 
develop a new conception of what logic is, 
and how to go about the task of analyzing and 
criticizing arguments? 

Yet to answer this question, the reviewer 
(and here I finally return to the matter of 
my limits) must have the benefit of both time 
and practice. I have the benefit of neither. 

A final problem faced by the reviewer (it 
seems there is no end) is that this is a big 
text <.337 pages) of very broad scope. I 
shall have to be selective in my remarks. 
The very least that I can do, at this point, 
is to give the reader a rougn idea of the 
shape of the whole. Then I shall burrow in. 

An Introduction to Reasoning has four parts. 
Part I ~s the "General introduction," con-
taining the opening chapter. Part II, the 
core of the text, presents what Toulmin calls 
"the basic pattern of analysis." It contains 
six chapters, each of which comes with exer-
cises designed to allow the student to apply 
what he or she has read in the chapter. Part 
III is titled "Rational assessment," and con-
sists of 4 chapters dealing witn various 
aspects of criticism. The largest of these 
is Chapter 11, "Fallacies," about which I 
shall comment in some detail. Part IV shows 
how the basic pattern is applied in different 



fields. Chapter 13 deals with legal reason-
ing; Chapter 14 with argumentation in sci-
ence; Chapter 15 with aesthetic arguments; 
Chapter 16 with decision making in manage-
ment and business; and Chapter 17 with ethi-
cal reasoning. 

Here then is the plan for the review. In 
Section II, I begin with a brief discussion 
of Chapter 1 and then take a detailed look at 
each of the elements in the pattern of analy-
sis. In Section III, I will focus on Toul-
min's theory of criticism as set forth in 
Part III of the text. In Section IV, I pre-
sent my answers to the questions I have said 
constitute the appropriate matrix for this 
review. 

II • TOULMIN 's THEORY OF ARGtn-IENT (PARTS I , 
!!l 

A. Chapter 1: Reasoning and its goals 

The first chapter aims "to sharpen the stu-
dent's perception of just what it is to give 
reasons" (TG, 1). It begins with five exam-
ples of reasoning drawn from different areas, 
after which Toulmin asks: 

What does it mean to ask if someone's 
statement or argument or advice is 
sensible or well reasoned, sound or 
logical? .•• What do these demands 
for "good reasons" and "sound argu-
ments" amount to? And how are we to 
judge this kind of goodness and sound-
ness? That is what this book is all 
about. (4) 

An engaging and effective start, I thought. 
The topics dealt with in the rest of the 

chapter not only serve the aim declared above 
but also lay the foundation and indeed set 
the tone for the rest of the book. They are: 

--"The varied uses of language," in 
which Toulmin distinguishes between 
the instrumental and the argumentative 
uses of language, thereby putting some 
limits around the need for "giving 
reasons." 
--"Reasoning varies with situations," 
which makes the fundamental point that 
the kind of reasoning that is appro-
priate varies from context to content. 
--"Reasoning as a critical transaction," 
which presents the idea of reasoning 
as one whose essential locus is "a 
public, interpersonal or social one," 
(9) and of reasoning as "not a way 
of arriving at ideas but rather of 
testing ideas critically" (9). 
--"The structure of arguments," which 
makes the point that arguments, or 
trains of reasoning, are constructed 
out of their constituent parts. 
--"Some definitions," in which 
Toulmin presents his definitions of 
araumentation, reasoning, argument 
an rationality. 
--"The forums of argumentation," 
which again emphasizes that argu-
ments have different forums and must 
be judged accordingly. 

18 

Before moving on, I want to call attention 
to two commitments of Toulmin' s approach that 
emerge in this first chapter. The first is 
Toulmin's position on standards (or criteria) 
of assessment. He says: 

So reasoning involves dealing with 
claims with an eye to their contexts, 
to competing claims, and to the 
people who hold them. It calls for 
critical evaluation of these ideas 
by shared standards; a readiness 
to modify claims in response to 
criticism; and a continuing critical 
scrutiny both of the claim provisionally 
accepted and of any new ones that may 
be put forward subsequently. (9) 

I like Toulmin's emphasis here, but the re-
ference to "shared standards" was both in-
triguing and perplexing. For shortly after 
this passage, Toulmin writes: 

One of the central questions in our 
whole inquiry will be, indeed, just 
how far, and in just what respects, 
we can hope to state general or 
universal standards of judgement 
for telling the validity, relevance, 
and strength or weakness of "reasons" 
or "arguments"; just how far, and in 
what respects, these standards will 
inevitably vary in time or differ 
according to the context and circum-
stances of judgement. (10) 

The question of whether or not there are uni-
versal standards for the evaluation of argu-
ments is a crucial and vexing one for the 
theory of logical criticism. Between the 
lines here, I thought I detected a hint of 
relativism--the view that there are no uni-
versal standards of criticism, that all 
standards are context-dependent. With equal 
justice, such a view might also be called 
Eiagmatism. In any event, Toulmin returns to 

e quest~on of standards several pages 
further on: 

While certain very broad rules of 
"rational procedure apply to argu-
ments in all these forums, many of 
the more specific rules of procedure 
(or "due process") that govern argu-
ments in one area are relevant only 
to, say, the proceedings in a law 
court rather than a scientific 
meeting, or the other way around. 
One of the main aims of this book 
will in fact be to show the difference 
between: 

--Those universal ("field-
invariant") rules of procedure 
that apply to rational criticism 
in all fields and forums, and 

--Those particular ("field-
dependent") rules that are 
appropriate in law, or SCience, 
or business, but not everywhere. 
(15-16) 

And so Toulmin's theory of criticism shades 
into his theory of argument. That there are 
both field-invariant and field-dependent 
standards or rules clearly presupposes that 
arguments can, without insuperable difficulty, 
be parcelled out to various fields. Should 
this presupposition prove to be implausible, 
Toulmin's theory of argument and his theory 
of criticism alike will be in some jeopardy. 



Thus, two important commitments have emerged 
in this first chapter. The first, belonging 
to the theory of argument, is that arguments 
can be assigned to fields. The second, be-
longing to the theory of criticism, is that 
some standards are field-invariant, while 
others are field-dependent. We s~ll want to 
monitor future developments witn these prin-
ciples in mind. 

B. 
ana YS1.S 

What Toulmin refers to as "the basic pattern 
of analysis" is described in a nutshell in 
Chapter 2, and then discussed step-by-step in 
the chapters that follow. The pattern has 
six nelements": the claim (Chapter 3); the 
ground (Chapter 31; t:Fi8'"Warrant (Chapter 4); 
the 6ackinCf (Chapter 5); the modality (Chap-
ter 6); an possible rebuttals (Chapter 6). 
I shall be looking Closely at each of them, 
for clearly Toulmin's novel approach will 
stand or fall largely on the attractiveness 
of the pattern of analysis. In particular, 
I shall be keeping two questions in mind. 
First, are the elements clearly explained? 
Second, is the pattern flexible enough to be 
universally applicable? Anything less than 
strong affirmatives means trouble for his 
theory of argument. 

Claim. By this term, Toulmin means "an as-
sertIOn put forward publicly for acceptance" 
(29). In the dialectical situation, the 
claim is implicitly linked to the next element 
in the pattern--the ground(s). Hence the 
claim plays roughly the same role in Toulmin's 
schema that the conclusion plays in the tra-
ditional one. (Toulmin's rejection of the 
term "conclusion" has an ideological basis, 
about which more later.) Since the claim 
stands at the core of the argument, and since 
arguments Cas we have seen) are to be assigned 
to various fields, it is no surprise that 
Toulmin holds that claims differ from one 
field to the next. The procedures required to 
make good on them differ accordingly. 

One can pose questions about Toulmin's con-
ception of a claim. What precisely counts as 
an assertion? What happens to what would be 
called, in the traditional schema, an implied 
conclusion? Will Toulmin have to swallow 
nimplied assertion"? But I do not think these 
are major problems. Perhaps more to the point 
would be to note that the section on ambiguous 
or unclear claims C31-321 was elliptical. In 
analyzing arguments, one is often confronted 
by a passage where the conclusion or claim is 
unclear. It is my experience that what stu-
dents need are some substantial pointers about 
how to clarify (and indeed identify in the 
first placel muddy or ambiguous statements. 
While it is true enough that context will of-
ten help clar ify Cas Toulmin says), I believe 
an excursus into the nature and the theory of 
meaning WOuld serve the aims of argument anal-
ysis better still. 

Grounds. By this term, Toulmin means to 
refer to "statements specifring particular 
facts about a s1.tuation rel1.ed on to clarifr 
and make <Jooo the previous claim" IT3). Thl.S 
category loS meant to capture what would, in 

19 

the traditional schema, be termed the evidence 
or support for the conclusion; i.e., the pre-
mises. It seems to me that there are several 
problems with this way of looking at the sup-
porting cast. 

First, it seems that grounds are, ~ their 
nature, factual. Not only the above text, 
but others as well, support this interpreta-
tion: 

The term "grounds n refers to the 
specific facts relied on to support 
a claim. Lm (emphasis added) 
Q's demands for grounds is ••• a 
request for A to put into the discussion 
some body of-specific "facts of the 
case" that can be agreed uponiis-a-
secure starting point acceptable to 
both sides and so "not in dispute." 
03} Cemphasis added} 

Unless Toulmin's conception of fact is a~tra­
ordinarily elastic, this requirement seems 
unduly restrictive. What happens to arguments 
in which one normative claim is put forth as 
support for another? Consider the following 
example: 

Girls should never touch alcoholic 
liquors. The reasons are obvious. 
It is for them to steady the young 
men, and so maintain their dignity, 
their beauty, and their intelligence. S 

on the traditional model, the third statement 
is the premise supporting the first statement 
as its conclusion. On Toulmin's model, the 
first statement is the claim, but can the 
third be categorized as its ground? Can it be 
interpreted as stating SlbcifiC facts? Only 
if we allow for a most 1 era! interpretation 
of that category can the statement be con-
strued as a ground. 

Second, some of the statements adduced by 
Toulmin as examples of grounds do not seem to 
me to fit his definition. Consider this exam-
ple: 

Just compare them (the Oakland 
Raiders) with the opposition. None 
of the other teams has such a com-
bination of offensive and defensive 
strength. C321 

Can this last statement be viewed as one which 
gives specific or particular facts? If we're 
talking football, the following statements 
strike me as more suitable candidates: 

Oakland allowed the fewest points 
per game of any team in the N.F.L. 

Oakland allowed the least yards per 
game of any team in the N.F.L. 
Oakland had the best 3rd down effi-
ciency rating of any team in the N.F.L. 

Toulmin's candidate strikes me as one or two 
dialectical jumps above the "facts of the 
case." It is a characterization or inter-
pretation of them. No doubt, such a statement 
might be offered as evidence for some other 
claim. aut can such a statement be cate-
gorized as a ground, given Toulmin's defini-
tion? 

Third, what happens to arguments whose evi-
dence consists in counter factual conditional 
statements? I do not know how Toulmin pro-
poses to handle such objections, though he is 
not unaware of the problems involved in set-



I 
I 
I 

tling on the grounds: 

Not all of the things that A 
initially offers as "facts"-need 
strike Q as unquestioned "facts" 
also •. - .. As a result, a sub-
stantial amount of time may be 
spent in the early stages of any 
argument, going over the material 
initially offered by A as his 
supporting "facts," for Q must 
satisfy himself which of-A's 
grounds he must indeed accept as 
data. (33-34) 

Evident in this passage is the influence of 
the jurisprudential (or dynamic) model, and 
the time has come to say more about it. 

The jurisprudential model. Although Toulmin 
does not here refer to~model as jurispru-
dential, still it is the same pattern as was 
presented in The Uses of Arrument, where he 
made a strong-case for-aump~g the geometrico-
mathematical model that had dominated logic 
for so long and replacing it with one patterned 
on the process of argument found in legal dis-
course. Such a shift has, as we are now see-
ing, a number of consequences. One is the 
new pattern of analysis. Another is that the 
context of argument will naturally be dia-
lectica1 6 and greater emphasis will be paid 
to the process of argumentation. Thus, in 
the text, Toulmin unveils his pattern of 
analysis as a series of transactions between 
~, the assertor, and ~, the questioner. 

There are clear pedagogical gains in this 
shift. Students can witness and become part 
of the process of argumentation. They learn 
what sorts of questions it is appropriate to 
ask, and at what stage. They develop an ap-
preciation of argumentation as an ongoing 
process, where claims can be modified and 
revised in light of criticism. They come to 
understand that criticism is an integral part 
of the process, not the enemy. All of this 
is preferahle to the idea that an argument is 
a one-shot, winner-take-all affair. 

But there are problems, too. First, dia-
lectical interchanges between Q and A will 
usually be public only in the restricted 
sense of being available to anyone within ear-
shot. Yet Toulmin has said that the essential 
locus of reasoning and argument is public, 
interpersonal and social: 

Wherever an idea or thought may 
have come from, it can be examined 
and criticized "rationally" . . . 
only if it is put into a position 
where it is open to public, col-
lective criticism. (9) 

Those who take the argument-as-product as 
their focus and try to teach students how to 
analyze and criticize arguments have no such 
tension to resolve. The argument on the Op-
ed page of the New York Times is by its very 
nature public--an objectrve-entity there for 
anyone who cares to read and/or lock horns 
with. But in engaging with that argument, 
the reader or critic will not have the bene-
fit of having shared in the process that led 
up to the formulation of the argument. Which 
brings me to the second problem. There are 
all sorts of moves open to one who is part of 
the process that will not he available to one 
confronted with only the distillate of that 

20 

process--the argument itself. In we're part 
of the process, we can ask the arguer to 
clarify her claims; or restate her grounds; 
or add to, subtract from or expand upon her 
statements. But when we've only got the ar-
gument itself to go on, none of this is pos-
sible. 

perhaps r can illustrate better what I'm 
driving at here by turning briefly to the 
Exercises for Chapter 3. There are 20 pas-
sages, calling for the student to, among 
other things, identify the claim and the 
grounds given for it. The first 10 examples 
are straightforward enough; the second 10 are 
more difficult, perhaps because they have the 
virtue of oeing real (as opposed to artifi-
cial) examples of arguments. However, because 
the text has not really prepared them for 
handling any but the most streamlined of ar-
guments, r think students may find the last 
10 examples difficult and frustrating. Toul-
min has said nothing about the problem of 
weeding extraneous material (what I call 
"clutter") from the argument, nor about de-
ciding precisely what the claim is and how 
best to formulate it. Partly, one is in-
clined to believe, these lapses are the re-
sult of his having employed the dialectical 
model. For if Q has any doubts, he or she 
can ask A just what the claim really is, and 
just which statements A takes to be grounds, 
But when the scene Shilts, and Q is the stu-
dent working on, let us say, #13, a product 
awaiting scrutiny, then frustration sets in. 

Let us, then, look at #13, which I shall 
reproduce here: 

Business doesn't draw up a contract 
with government; it tries to get the 
best deal it can in an increasingly 
coercive society. There is no such 
thing as voluntary planning. It 
compels somebody to do something 
he otherwise WOUldn't do. 

[Objection from Skeptic magazine 
interviewer 1 : 

What if we vote for it? 
It's still coercion. You can call 

it tlazism, or you can call it Com-
munism. You vote away my minority 
rights to disagre~; r don't find that 
anything but coercion. Majority rule 
is coercion if it tramples on minority 
rights. 

Robert M. Bleiberg, Editor Barron's 
in an interview with Skeptic magazine 

This is the very sort of passage that is like-
ly to give students fits. Just what is . 
Blaiberg's claim? It might have helped had 
Toulmin given us some information about the 
context in which this passage occurred. In-
deed, the failure to do so seems a serious 
oversight on the part of someone whose ap-
proach places such emphasis on context. With-
out such information, it is difficult to de-
cide which of Bleiberg's many claims is to be 
taken as the focal point. Is it: "There is 
no such thing as voluntary planning?" Or: 
"Business doesn't draw up a contract with 
government?" Or: "Majority rule is coercion 
if it tramples on minority rights?" If we 
knew the drift of the conversation leading up 
to this excerpt, we might better be able to 
decide what Bleiberg's point was, and, ~ndeed, 
whether this excerpt even qualifies as an ar-
gument in the first place. 



Here is Toulmin's analysis from the Teaching 
Guide: 

Claim: Business is subject to 
government coercion. 

Grounds; All government planning with 
regard to the economy amounts 
to coercion and can be compared 
with totalitarianism. 

I have reservations about Toulmin's inter-
pretation, but let's grant that this is an 
argument. Rleiberg did not, as far as I can 
see, assert that "Business is subject to 
government coercion." Be may have made state-
ments which implied or are equivalent to it, 
but he did not make that assertion. Yet Toul-
min defines a claim as "an assertion put for-
ward publicly for general acceptance." Per-
haps Toulmin can get around this by amending 
the definition of a claim. Even still, it is 
not at all clear to me that Bleiberg's intent 
here was to establish that big business is 
(as a matter of fact) sumject to government 
coercion. It seems rather more plausible to 
take Bleiberg to be using that fact as a 
point of departure on the road to claiming 
that this situation is tantamount to Nazism or 
Communism. I was also bothered by TOulmin's 
rephrasing of Bleiberg's statements. Why re-
phrase, unless it is necessary to do so in 
order to bring out the meaning more clearly? 
For such rephrasing runs the risk of distorting 
the arguer's position, while conferring no 
apparent benefits. 

Well, reasonable men and women may well 
disagree with one another about the analysis 
of any given passage. Toulmin realizes this. 
In the Teaching Guide, he says CI preserve 
the capltalizationr:-

IT IS IMPORTANT TO EMPHASIZE THAT 
THE CONCEPTION OF LOGIC UPON WHICH 
AN INTRODUCTION TO REASONING IS 
BASED DOES NOT ADMIT OF FINAL 
"CONCLUSIONS" TO ARGUMENTS, NOR 
DOES IT HOLD THAT THERE ARE ANY 
UNIQUE "SOLUTIONS" TO PROBLEMS 
rNVOLVING REASONING PRACTICE. 
THE ANALYSES IN THIS BOOK ARE NOT 
THE ONLY POSSIBLE ONES: EACH OF 
THEM REPRESENTS ONE AMONG MANY 
POSSIBILITIES. WE PRESENT THEM AS 
CONVENIENCES TO THE INSTRUCTOR. (A-51 

While I admire the forthrightness of this 
declaration, I must yet admit that I find my-
self distressed at not being able to fathom 
how he arrived at his analysis of the above 
passage! Toulmin is surely right, however, 
in stating that we cannot expect unique solu-
tions in informal logic. Admit this, and you 
are on the slope to a pernicious species of 
relativism, unless you are prepared to declare 
on what basis one analysis is better than an-
other. The problem is thorny enough to make 
one hanker for the rarefied climes of formal 
logic--almost! 

Warrants. Of all the elements in Toulmin's 
schema, the warrant is the one I had the 
greatest problem understanding. The intuitive 
idea seems clear enough, but las I will try 
to show) Toulmin's rather breezy style of 
exposition creates some of the confusion. 
Then, too, this is the element that departs 
most radically from the traditional schema, 
thereey forcing one to look at the structure 
of arguments in a different light. That takes 

21 

some getting used to. Grounds, after all, may 
be readily compared to premises, and claims to 
conclusions. But there is no counterpart for 
the warrant. I will, therefore, first attempt 
to set out as clearly as I can what Toulmin 
means by the term, and than present my diffi-
culties with it. 

SiDlply stated, a warrant is a link between 
the grounds and the claim. ToulmInwrites; 
"Put colloquially, the question at the level 
of the warrant is not, 'What do you have to 
go on?' but 'Bow do you get there?'" (44). 
The warrant then justifies passage from the 
grounds to the claim. 

It is tempting, but mistaken, to think of a 
warrant as nothing more than a formal rule of 
inference. Toulmin addresses this point ex-
plicitly in the Teaching Guide (it would pro-
duce needless confusion to deal with it in 
the textl; 

The rule~ of inference appealed to 
in formal logic (modus ponens, for 
examplel are deSigned to guarantee 
merely the "formal consistency" of 
the premises and conclusion, in an 
argument whose structure has been 
set out in the manner required, but 
the warrants that serve as "rules" 
for connecting data and claims in 
our present analysis are intended to 
insure the substantive soundness of 
the steps in question ("From smoke, 
you may infer fire"). So substantive 
warrants are both more specific and 
concrete in their content and also 
more open to challenge than formal 
rules of inference. (~71 

At the very least, Toulmin thinks that there 
is a categorical difference between warrants 
and the rules of inference of formal logic. 
In this, I think he is right. 

It is also tempting to think of a warrant as 
nothing more than a missing or suppressed 
premise, conditional in form, whose antecedent 
contains the data or grounds, and whose con-
sequent contains the claim. I doubt that 
Toulmin would accept this view, but I'm not 
sure Cgiven what he says about warrants and 
some of his examples} that he can forestall 
it. Let us see. 

Toulmin first introduces the term in Chapter 
1: 

(Wl)Steps from grounds to claims are 
"warranted" in different ways in 
law, in science, in politics, and 
elsewhere. The resulting warrants 
take the form of laws of nature, 
rules of thumb, engineering formulas 
and so on. But in a practical case, 
some appropriate warrant will be 
needed if the step from the grounds 
to the claim is to be trustworthy. 
[261 

Here warrants are spoken of as steps. They 
are also Oby implication if not direct state-
ment) field-dependent; i.e., restricted in 
scope to some specifiable field or rational 
enterprise. And it is further clear that 
what makes something a warrant is its function 
rather than any intrinsic feature. 

To further verify this most recent point, 
one has only to notice the variety of descrip-
tions Toulmin has given of warrants. In Chap-



ter 4, he introduces them this way: 
(W2)NOW the questioner asks for warrants, 

that is, statements indicating the 
~eneral ways of arguing being applied 
~n eash particular case and implicitly 
relied on as ones whose trustworthiness 
is weIl-established. (431 

He then goes on to make all of the following 
claims about warrants: 
(W3}such a general, step-authorizing 

statement will here be called a 
warrant. C441 

(W4)A license to argue from grounds 
to conclusion • • •• (45) 

(WS)The questioner, Q, must now inquire 
about the generar rule or procedure 
that the assertor, A, is relying on 
in presenting the step from £ to C 
as a trustworthy step. (44) 

(W6}Many kinds of general. statement 
authorize th.e inferences by which 
different collections of specific 
information . • • are put forward 
as rational support for the claim. 
(53) 

In the Teaching Guide we find these state-
ments abOut warrants: 
(W7). • . the strength of reasoning 

depends upon the rules or "warrants" 
that stipulate just what kinds of 
information are relevant to answering 
questions about the subject in hand. 
LTG, 17) 

(WS). • . warrants . • . serve as rules 
for connecting data and claims 
(TG, 171 

(W91Warrants are rules that enable us 
to "make sense" of situations. 
LTG, 18) 

(WIO)Narrants are often merely assumed 
tacitly--or, as we say, "taken for 
granted. " 

This medley of descriptions might be con-
fusing to someone who is trying to master the 
meaning of a familiar term, here being put to 
special use. (W2I., eW3), and eW6} all char-
acterize warrants as reneral statements; 
eW5} , (W71, eN8) and m) all describe war-
rants as rules; (W4) refers to a warrant as a 
license: ana-TWIO) introduces the notion of 
an assumption. One suspects that, in the 
long run, all of these descriptions may prove 
quite compatible. In the short run, however, 
the sheer variety is more likely to engender 
confusion than understanding. 

Toulmin may indeed be aware of this point, 
for he says: 

Historically speaking, the term has 
always had close associations both 
with the notion of a license or 
permit and also with that of a 
warranty or guarantee. (45l 

Whether this usage note can provide the much-
needed unifying thread for Toulmin's various 
statements about warrants remains to be seen. 

Let us look at some examples of warrants. 
The first one Toulmin produces is this: 
ell Wherever there t s smoke, there's a 

fire. (44) 

22 

About this statement, Toulmin says: 
This last statement has the effect 
of authoriSing the step from G 
(the smokel to C (the fire). -We 
can in fact read it as meaning, 
"Wherever smoke is visible, it can 
De concluded that there is fIre---
iIso. • (44)--

This translation gives me pause. It is well 
known that any general statement can be given 
that sort of reading, and hence is potentially 
a warrant. But what'S to prevent us from 
giving Singular statements the same sort of 
reading? Thus "Ronald Reagan is the 40th 
President of the United States" can be read 
as meaning "Wherever Ronald Reagan is visible, 
it can be concluded that there is the 40th 
Pres~dentalso.h And if this move is allowed, 
then warrants need not be general statements 
or rules at all. 

Before looking at additional examples, I 
need to mention one important point aDout war-
rants: they are field-dependent. This is 
clear Doth from the way that Toulmin has 
introduced the notion, and from passages like 
this one: 

Wherever a fully established and 
articulated body of knowledge exists 
. . • we commonly find such warrants 
recognized and put to use. rn other 
fields, however, it may be harder to 
articulate all the warrants employed 
in argument, in the form of laws, 
rules or principles. (491 

In looking at the examples, it will be useful 
to keep these two questions in mind. First, 
which of the many descriptions of warrant 
does this example fall under? Second, what 
field does the warrant belong to? 
(2) If a woman is seen walking down 

the street with a man whom her 
husband does not know, it may be 
concluded that she is having a 
clandestine affair with that man. (45) 

This is a general statement in the sense of 
CW2), but it is difficult to assign it to any 
field. 
(31 [Toulmin refers to] a general formula 

that relates the breaking strain of 
a girder to its shape and dimensions 
. . •• L47) 

No problems with the field here: engineering. 
And (3) seems covered by LW6). 

(4) Anyone who leaves a car in a 
metered parking space without 
putting money in the meter can 
be found guilty of an offense. (48) 

This statement sounds as much like a warning 
as it does an inference ticket. Toulmin cites 
it as an instance of a warrant from the field 
of law. 
L5} It is unjust to expect your wife 

to give up her spare time to baby-
sit without ever taking a turn 
yourself. (49) 

This is cited as an instance of a warrant from 
ethics. It is a normative ethical claim, but 
it lacks the generality that one expects of a 
warrant •. 



r 

(6) Pallor, lethargy, and a low fever 
often means viral or bacterial 
infection, or exhaustion from 
overwork, or, in a few cases, 
neurotic stress. (50) 

This ia a piece of medical knowledge, one 
supposes. 
(7) This particular combination of 

pallor, low fever, lethargy, 
etc., which Dr. Bernard's 
experience enables him to recognize, 
generally means (i.e. entitles us 
to have some confidence in inferringl 
a virus infection. (511 

The field, once again, is medicine, but the 
highly specific nature of this claim makes it 
difficult to fit it under any of the descrip-
tions offered by Toulmin. 
(8) Anyone who seeks to work only at 

what is rewarded is not conscientious. 
(521 ' 

Toulmin produces this example as °a warrant 
from the field of psychology. But it seems 
to me rather to be a tautology, or analytic 
truth. 

Still somewhat confused, I thought it ad-
visable to work through the Exercises for 
Chapter 4, which require the student to iden-
tify the warrants appealed to in the arguments. 
In order to save space, I shall not reproduce 
the original argument here, but warn the 
reader that he or she will not be able to fol-
low the discussion without the example. 

Examale #1. Toulmin and I came up with 
quite 1fferent warrants here: 

CToulmin) A society that conducts its 
public affairs in the style 
of a surrealistic painting 
can be regarded as worse than 
totalitarian. 

(Johnsont When the leading persons of 
the government, the party and 
the armed forces are coerced, 
then such a society can be 
called a terror-society--
something worse than being 
called totalitarian. 

I am not sure what field either of these al-
leged warrants belongs to. The difference in 
our formulations may, in part, stem from some 
disagreement over the grounds. Toulmin has 
focused on the second paragraph of the exam-
ple, whereas I have focused on the third. I 
believe my focus is more appropriate since 
(a) grounds are supposed to be factual in 
character and (b) the third paragraph seems 
more factual than the second; i.e., it would 
be easier to substantiate. Hence our differ-
ent ways of formulating the warrant here seem 
to be traceable to variant interpretations of 
how the argument works. 

Exam~e 2. Once again, Toulmin and I agree 
aboute claim but disagree about the grounds, 
which will naturally affect the formulation 
of the warrant. Here is Toulmin's view: 

Many conscientious, energetic 
individuals do not get ahead. 
The great minds of the world 
have net been able to assist 
them. C~, 20) 

23 

I agree with this, but would add: 
Astrology takes account of the 
complex structure of the human 
soul, whereas other social and 
economic formulas do not. 

In fact, I do not see how Toulmin can justify 
omitting this statement as a ground, since it 
is clearly meant to support the claim. 

This is how we each formulated the warrant: 

(Toulmin1 If conscientious individuals 
fail financially and if the 
great minds of the world can-
not find some solution in 
ordinary economics, then those 
who seek financial success ought 
to try astrology to find a way 
to financial security. C~, 20) 

Mine is even more cumbersome: 
(Johnson) If conscientious individuals 

fail financially and if the 
great minds of the world can-
not find some solution in 
ordinary economics, and since 
astrology takes account of the 
complex structure of the human 
soul, whereas other social. and 
economic formulas do not, and 
since the solution to financial 
problems requires an understanding 
of the complex structure of the 
human soul, then those who seek 
financial success cught to give 
astrology a try. 

Neither warrant is easily identified as be-
longing to any field and neither seems read-
ily classifiable as either a rule or general 
statement. 

Example #7. I found Toulmin's analysis of 
this very unsatisfactory, for it seems to 
distort the original argument. Once again, 
the problem lies in how the grounds (and 
hence the warrant) are to be formulated. 
Here is Toulmin's version: 

Larry J. Hillis bought eleven new 
Uniroyal steel-belted radials, all 
of which cracked and split. lTG, 21) 

The original text of the argument reads quite 
differently: 

"All the time with the tires it 
has been the same problem," Hillis 
wrote, "cracking and splitting 
around the rims, rounding on the 
edges and wearing improperly." (55) 

This is further clarified by the lead-in: 
One blowout three days after he 
bought the car; two more replaced 
within 12,000 miles; four tires 
cracoked and split in the sidewalls 
and rim. C55) 

SO apparently only four of the eleven tires 
cracked and split in the sidewalls and rim--
a quite different state of affairs than Toul-
min's version suggests. A more accurate for-
mUlation would be: 

Larry J. Hillis bought eleven new 
Uniroyal steel-belted radials, 
seven of which proved to be defective. 

But this is not all, for the original alsp 
contains this information: 



He (Hillis) insisted that he kept 
the tires "properly inflated and 
rotated, balanced and aligned." 

Fully-expressed, then, the grounds would be as 
follows: 

Larry J. Hillis bought eleven new 
Uniroyal steel-belted radials, 
seven of which proved to be 
defective. Hillis took proper care 
of the tires. 

Again differences in the ways the grounds 
are formulated will spawn different warrants: 

CToulminl When an individual buys a lot 
of new tires from one manufacterer, 
all of which turn out to be 
defective, the manufacturer 
must be at fault. 

(Johnson) When an individual buys eleven 
new tires from one manufacturer 
and takes proper care of them 
and seven turn out to be de-
fective, the manufacturer is 
at fault. 

I must admit that it is not clear to me how 
far the differences between Toulmin's war-
rants and mine are to be explained by dif-
ferent interpretations of what is being ar-
gued and how far by my difficulties in under-
standing the very concept of a warrant. But 
let me try to bring this section to a close 
by mentioning, once again, the main difficul-
ties r have encountered. 

Toulmin's explanation of the concept of a 
warrant is loose rather than tight. He gives 
us a number of different descr.iptions of, or 
ways of looking at, warrants which may all be 
compatible but whose unifying thread is not 
apparent. Some of the examples are not eas-
ily sorted out under any of the available 
descriptions. These problems may be only ex-
pository in nature. 

r think this is not the case with the idea 
that warrants are field-dependent. ~any of 
the warrants Toulmin cites as examples do not 
seem to be to belong to any identifiable 
field. Rut this only raises another problem. 
What is to count as a field? Law and science 
qualify,-of-course7 -Does-astrology? Does 
common sense? Does philosophy? And what 
happens when, as is often enough the case, an 
arguer provides grounds from different fields 
(cf. Example 2 on page 25)? Obviously no one 
warrant will suffice to link the grounds 
with the claim, unless the already nebulous 
concept of a field is to be extended to allow 
for super-fields! 

Well, these are problems I would gladly em-
brace, if only it were clear that Toulmin's 
new schema pays big dividends. But that is 
far from clear. If one looks carefully at 
the warrants Toulmin provides in the Teaching 
Guide for the exercises, one cannot help but 
be struck by the fact that 7 out of 10 of 
them CLe., 112, #3, #4, #5, #6, #7 (arguably), 
and #101 turn out to be the sort of "if • . . 
then" conditional proposition that a quasi-
deductivist would supply as missing premises 
in reconstructing the argument using the 
traditional schema!7 That is bad enough, but 
matters are worse still. Toulmin gives very 
little advice about how to go about formula-
ting warrants. In recent years, however, 

24 

informal logicians still operating more or 
less within the traditional schema have made 
some strides in handling missing premises. 
They have, for example, articulated the Prin-
ciple of Charity, which requires that the 
critic be scrupulous not to overcommit the 
arguer when filling in miSSing premises. Not 
only does Toulmin say nothing whatsoever 
about this important principle, but his own 
examples contain repeated violations of it, 
as we have seen. A new paradigm may well be 
expected to cause changes in the very nature 
of what is seen as a problem. In this case, 
we seem to be confronted with the loss of a 
problem (missing premises and how to formu-
late them), but I am not sure that its disap-
pearance should be construed as a step for-
ward. 

Racking. The next element in the pattern 
is the backing for the warrant. The dialec-
tical situation is this: A has made a claim 
for which grounds have been produced. q has 
challenged A's move from G to C. In response, 
A has adduced warrant, W.- But; as Toulmin 
says, "warrants are not-self-validatin~r (58). 
Q may rightly seek justification for t e war-
rant in either of two directions: 

1. "Is that warrant reliable at all?" 
2. "Does that warrant really apply to 

the present specific case?" (58) 

To answer either of these questions is to pro-
vide the warrant with its backing. 

The relationship between a warrant and its 
backing is close: 

A warrant and its backing are 
related in very similar ways in 
many different contexts of argu-
mentation. But the kinds of substantive 
considerations that actually support 
our warrants vary greatly between 
different enterprises and fields 
of argument: in scientific, medical 
and legal arguments, in discussions 
about sport, art or business, in 
abstract discussions of pure mathe-
matics, our warrants derive their 
foundation and authority from 
bac king 0 f quite dif feren t sorts. (.62) 

Again we see the influence of the assumption 
that arguments can be sorted out into dif-
ferent fields, though it must be clear that 
discussions about sport constitute a field in 
a very different sense than do abstract 
discussions of pure mathematics. Since the 
concept of backing is so closely tied with 
the concept of a warrant, I shall say no more 
about it here. 

Modality. The fifth element in Toulmin's 
pattern is the modality, or modal qualifier. 
By this term, Toulmin means "phrases showing 
what kind and degree of reliance is to be 
placed on the conclusions, given the arguments 
available to support them" (69). The modal 
qualifier indicates how strong the arguer 
thinks the connection is between the grounds 
(plus warrant and backing) and the claim. 
Examples of modal qualifiers are words and 
phrases such as: "necessarily," "certainly," 
"presumably," "in all probability," and n for 
all we can tell." 

Toulmin shows how modal qualifiers operate 
in fields such as law, medicine, and science, 
but insists Cand rightly so) that modals play 
an important role in everyday arguments as 



well: 
Whatever other differences there are 
between the modes of argumentation 
appropriate to our different activities 
and enterprises, we frequently have 
occasion 

1. To present our claims tentatively, 
without staking our whole credit 
on them. 

2. To put tha~ into debate in an 
uncommitted way, merely for 
purposes of discussion. 

3. To treat them as serious but 
conditional conclusions. 

4. To offer them simply as a good 
bet. 

As a result, the relevant modal qualifiers 
. . . have a part to play in all kinds 
of arguments. (74) 

This is well said, though the last statement 
is weaker than one made earlier: "In a word, 
every argument has a certain mOdalitri" (70). 
Toulmin's position here is probably est con-
strued as one which prescribes an ideal rath-
er than describing what is the practice. In 
my experience, medals show a relatively low 
incidence of occurrence in everyday arguments. 
People generally do not say what sort of 
strength their evidence is meant to provide 
for their claims. It would be better if they 
did. Surely one of the purposes of an intro-
ductory logic course is to acquaint the stu-
dent with the use and the importance of modal 
qualifiers. Toulmin's text does an excellent 
job on this count and deserves much credit 
for its fine presentation of this much-
neglected facet of argument analysis. 

Rebuttals. The last element in the pattern 
is the rebuttal; that is, the mention of "the 
extraordinary or exceptional circumstances 
that might undermine the force of the sup-
porting arguments" (75). These may be built 
into the argument to indicate what sorts of 
conditions might vitiate it. I have no prob-
lems with this idea, and found Toulmin's 
presentation Doth innovative and lucid. 

The question of how far one should go in 
qualifying one's claims is not something 
Toulmin believes can be given a fixed answer. 
Rather it is a matter for pragmatic choice: 

To avoid an excess of small print, 
we must spell out at sufficient 
length the kind of exceptions, 
exclusions and other rebuttals 
that limit the force of our argu-
ments. To avoid gobbledygook, we 
must prevent the recital of exceptions 
from getting too long. • . . Where is 
the line to be drawn? That decision 
can be made only when we know enough 
about the audience Claymen or attorneys), 
the forum of argumentation (law court 
or office) and the general purpose of 
the particular discussion in question. 
(81) 

Notice that here Toulmin has switched gears 
and is speaking not so much of the analysis 
and criticism of arguments as of their con-
struction. One of the best features of Toul-
min's approach is that it is serviceable in 
both departments: analysis and construction. 
Those instructors who, like myself, have be-
come convinced that introductory logic courses 

25 

should deal with both departments will very 
much appreciate this feature of Toulmin's 
text. 

Of course, students will have to be advised 
that skill in argument construction forces 
one outside the realm of logic proper; they 
will need to have both an adequate grasp of 
the issue and a sufficient supply of informa-
tion. The Exercises for Chapter 6 will help 
underscore these points. Students will find 
it difficult to formulate possible rebuttals 
to arguments without some working knowledge 
of the issues. For example, the rebuttal to 
the claim that water should be thrown on 
burning materials to extinguish them is: 
"Throwing water on burning oil is very dan-
gerous, for it will spread the fire" LTG, 30). 
The average person should know this, but the 
point is that logic will not confer this 
knowledge. 

All of the elements are now in place, and 
the chart below shows how one would diagram 
an argument using this pattern of analysis: 

G 

Grounds 

Backing 

Warrant 

Modality 

Rebuttal 

Given grounds, G, we may appeal to 
warrant, W (which rests on backing, 
al, to justify the claim C--or, at 
any rate, the presumption-CM) that 
C--on the absence of some specific 
rebuttal or disqualification (R). 
1.78) -

C. Summary 

Claim 

To bring this part of the review to a con-
clusion, I shall state what I take to be the 
cardinal features of Toulmin's theory of ar-
gument along with my reservations about them. 
In essence, Toulmin's theory of argument has 
three components: 

A) The conception of reasoning and 
argumentat~on as dialogical and 
pragmatic in character rather 
than static and syntactic; 

Bl The new schema or pattern of 
analysis for arguments, whose 
elements are these six: claim, 
ground, warrant, backing, modality, 
and rebuttal; 



C) The assumption that each and every 
argument can be assigned to some 
specific field or enterprise. 

B) and Cl reinforce one another, since many 
of the elements in the schema (notably, war-
rants and backing) are intelligible only under 
the assumption that arguments are field-
related. 

About A: I think it is high time that 
someone tried this experiment. As I stated 
in Section I, our conceptions of logic and 
argumentation have undergone very little ex-
pansion or revision since the time of Aris-
totle. The reign of the geometrico-mathe-
matical model has gone virtually uncontested. 
Indeed, it has if anything been strengthened 
by the rise to prominence of mathematical 
logic, following Frege's pioneering work. 
The philosophical assumptions behind the old 
logic have been subjected to vigorous chal-
lenges in the philosophical works of Peirce, 
Dewey and Wittgenstein, to name but a few. 
Close connections between the assumptions and 
ideals of traditional logic and the perennial 
specter of skepticism have been noted by many, 
Toulmin among them. We may like to think 
that logic, since it is only an instrument, 
is free of ideological or metaphysical influ-
ence. But that is an illusion, it seems to 
me. Kence it is healthy to have Toulmin 
actively promoting a new conception of rea-
soning and argument (and logic) within the 
matrix of this experiment. Indeed, Toulmin 
notes how the old te~inology reinforces the 
ideal behind the old logic: 

We have stated the elements of our 
analysis independently of the tra-
ditional terminology of formal logic 
(premise-conclusion, for example), 
not merely to set our mode of analysis 
apart from the traditional one, but 
also to emphasize that what we are 
doing involves a different conception 
of logic from either syllogistic or 
symbolic logic. These latter have 
more in common with each other than 
they do with the sort of analysis we 
are developing. Thus we speak of 
arguments as "supporting claims" 
rather than as "leading to conclu-
sions," because the nature of claims 
is such that they can be reformulated. 
The term conclusion suggests the 
reaching of an end point, rather like 
the ~ erat demonstrandum of the 
geometers:--TTG, 81 

Aside from the few reservations mentioned 
earlier (pp. 14-18), this aspect of ToulmL~'s 
approach works well. 

About B: There are two questions that must 
be asked in evaluating the success of Toul-
min's new pattern of analysis: Bl): Is the 
pattern viable? Are its elementS-clearly con-
ceived and explained in the text? B21: Is 
the pattern adequate? Can it be applied to 
all sorts of arguments? 

Throughout Section II of this review, I 
have indicated the problems I have enCO~i­
tered in attempting to understand some of the 
elements of Toulmin's schema, principally 
(though not exclusively) with warrants. Part 
of the problem may lie in Toulmin's rather 
breezy style of exposition, but I am not sure 

26 

that this is the whole problem. Then, too, 
I had some troubles with his conception of 
grounds, which seemed to me too restrictive. 
~y answer to Bll then is ~hat there are 
grounds for wondering whether the pattern is 
viable, but I would not want to ~ke any 
final judgement until Toulmin has had a 
chance to tighten some of the conceptual 
screws. 

I have rather more serious doubts about 
whether the pattern can be applied to all 
sorts of arguments. These comes from two 
different directions. First, the sort of 
argument that Toulmin chooses to illustrate 
his pattern is streamlined; thad is, it is 
short and has very b.ttle internal complexity. 
So it is not clear to me how this pattern 
will be deployed in the analysis of what r 
call, following Kahane 8 , extended arguments. 
It seems to me that it will be both awkWard 
and tedious to attempt to decipher the struc-
ture of, say, a 5,000 word editoFial using 
this pattern and method of diagramming. Per-
haps it will be possible to decompose such 
extended arguments into smaller, more manage-
able chunks which can then be digested by the 
pattern. But should the tail wag the dog? 
Since, as I have argued elsewhere, the analy-
sis of extended arguments must be viewed as 
the primary target,9 it seems desirable to 
devise a procedure for dealing with them in 
their integrity. So Toulmin's pattern runs 
into some problems here. 

The other doubt I have about the universal 
applicability of the pattern dovetails with 
reservations I have about C)-- the assumption 
~~at arguments can be assigned to fields. I 
shall cite an argument which is, I think, 
quite typical of the sort found in everyday 
discourse, but to which it is difficult to 
apply Toulmin's pattern. 

In the 1980 elections, Californians had to 
vote on Prooosition 10, which would have lim-
ited drasti~ally the places where people could 
smoke in public. The debate was heated. 
Here is a typical argument presented in 
favour of the Proposition: 

Californians should vote for 
Proposition 10 because (1) 
the medical evidence is clear that 
smoking represents a health threat 
to non-smokers and (21 it is clear 
to anyone who knows smokers that 
they will not themselves freely 
choose to limit their smoking and 
(3) such a proposition would entail 
minimal costs. 

In the first place, I find it difficult to 
assign this argument about a social and po-
litical issue to any specifiable field. Per-
haps this is because Toulmin never says pre-
cisely what counts as a field. He gives ex-
amples like law, medicine, science, and engi-
neering. He uses cognate terms like forum of 
argumentation, rational enterpris~, and ~­
text. But so far as r can tell, he never 
gIVes a precise definition of any of these 
terms. Since so much of Toulmin's approach 
depends upon the concept of a field, the 
failure to define it carefully seems to me a 
serious lacuna. 

Second, as to the analysis of this argument, 
there is no problem with the claL~ or the 
grounds. gut is there to be one warrant, or 

l 



three? If one, then I fail to see how that 
warrant can be located within any specifiable 
field. Suppose, then, that we connect each 
ground to the claim with a warrant of its 
own. How would the warrant for (1) be for-
mulated? We might suggest: 

Whenever there is a health threat 
to members of the public, there 
should be a law protecting members 
from that threat. 

What field underwrites this warrant? Similar 
questions will, I believe, arise as warrants 
for l2l and l31 are fleshed out. But there 
is another, potentially more serious, problem. 
In this particular argument, it seems to me 
that none of the grounds is meant to function 
independently of the rest. The arguer is re-
lying on their cumUlative weight. Kence any 
attempt to tie the grounds individually to 
the claim will result in a distortion of the 
argument. And so we are back to the first, 
but equally unsatisfact~ry, alternative of 
attempting to formulate some one warrant that 
will link all three grounds to the claim. 
But I doubt that this can be done, for here 
we have an argument which straddles several 
fields (speaking loosely}, crosses over bor-
ders, and is otherwise geographically messy. 
Yet the very fact that this argument seems 
to me quite typical of those found in every-
day argumentation causes me to have real 
reservations about the universal applicability 
of Tou~in's pattern. For the assumption on 
which the model rests depends for its cred-
ibility on a concept [field of argument) for 
which Toulmin has not provided a satisfactory 
elucidation. 

For the reasons indicated, therefore, I 
must conclude that Toulmin's theory of argu-
ment, as intriguing and exciting as it may 
appear to be, faces some severe challenges 
before it can be deemed successful.~ 

EDITORS' NOTE 

This is the first part of Professor Johnson's 
critical review of An Introduction to Rea-
soning. The second-aria fInal part ;flr--
appear in the next issue of ~ (iii. 3). 

27 

analysis of 
examples 

'l'IIlo analyses of examples appearing in !B!, 
ii/Supplement (~une, 19801 have been re-
ceived. They are printed below. We will 
continue to print such analyses throughout 
the year, as they arrive and as space permits, 
and we emphatically reissue our invitation to 
readers to suDmit their own critiques of 
these or other examples from the Examples 
Supplement. 

THE ARGUMENT 

A retired deputy inspector of the New York 
Police Department wrote the follOWing letter 
in November 1978 to the New York Times in 
response to a Times editorIar-aDout discrim-
ination againstlnOmosexuals: 

I am in complete agreement with the 
last paragraph of your November 10 
editorial on discrimination against 
homosexuals as a generality. However, 
unless a specific exc.eption is made 
relative to "on the street" police 
work, we might be opening Pandora's 
box. 
It is a well-accepted fact in police 
administration that recruitment of 
stable personnel is hampered by the 
inability of science to determine an 
adequate psychological test. In 
order to prevent the employment of 
an unstable police officer, with all 
its concomitant woes, every effort 
must be made to recruit only those 
whose personality traits fit within 
the parameters of normalcy, whatever 
that might be. Throughout the years, 
the news has been replete with in-
stances of recruitment failures in 
this area. No one has yet been able 
to determine what kind or degree of 
pressure will trigger a flawed per-
sonality. 
Unless we are ready to accept homo-
sexuality as totally normal, we must 
consider it as being at least a 
flawed personality trait. Other 
flawed personality traits, if known 
to the recruiter, have been suffi-