Tables and endnotes are numbered consecutively
through the three parts of this essay on Habermas' contribution
to practical philosophy.
Part 6b: Argumentation theory
and practical discourse – Habermas 2
We are still engaged in an effort to review
the practical philosophies of Aristotle, Kant, and Habermas,
to see what we can learn from them for the purpose of grounding
reflective practice philosophically. The discussions of Aristotle
and Kant were detailed but still found place within a single
essay each. The current discussion of Habermas, however, takes
more space and I have therefore decided to split it into three
parts (see the right-hand note). Before we continue with the
second part, it may help returning readers if I briefly sum up where we
stand; should you be new to the Bimonthly, I recommend
you read the previous Part 6a/7 to facilitate your reading of
the present Part 6b/7 (click on the "Previous" button
at the top right of this page).
We have found an essential aim of the practical
philosophy of Jurgen Habermas in his concern for strengthening
noninstrumental patterns of reasoning and social rationalization.
The central notion is "communicative rationality," the
idea that people can peacefully coordinate their interests
and actions through communication aimed at mutual understanding
and (where necessary) backed by argumentative rather than non-argumentative
means of conflict resolution. Further, inasmuch as argumentation
is employed, communicative rationality implies that it goes
along with a cooperative attitude rather than with a non-cooperative
stance of merely "strategic rationality." We have considered in some detail the language-analytical
and speech-act-theoretical foundation that Habermas proposes for comunicative
and have then moved to the second of the three levels of analysis
we distinguished in Table 1*, the level of "discourse." The
core issue at this second level is the question of what constitutes
a good argument; that is, we are entering the field of argumentation
have seen that Habermas analyzes the "general pragmatic
presuppositions" of argumentation from three complementary
perspectives, the perspectives of process, procedure, and product.
as process, we said, is about the effectiveness
of communication in achieving the telos of mutual understanding;
as procedure, about provisions for securing rationally
defendable agreement; and as product, about the
assessment of the strength of validity claims. The
first part of this introduction to the practical philosophy
of Habermas ended with a discussion of his reconstruction of
the "process" perspective in terms of the requirements
of "rational motivation" – the idea that a cooperative
attitude should orient the argumentative process – and
of the "procedure" perspective in terms of the general
symmetry conditions of an anticipated "ideal speech situation"
– the idea that discourses should be open to everyone concerned
and should allow a free and equal exchange of arguments. Although
these conditions are never fully given, or at least we
should not assume they are, they are nevertheless operative
as soon as we enter into a discourse; for we cannot reasonably
argue without assuming it is indeed possible to improve mutual
much for a short glance back at where we stand. If we now are
to move beyond these general presupposition, let us try and
see how exactly argumentation, assuming such conditions, can
play the role of a "court of appeal" (Habermas, 1984, p. 17)
that helps us settle differences peacefully, "with reason"
rather than "with force." To answer this question,
we now need
to turn to the "product" perspective, which deals
with the difficult issue of what constitutes (in traditional
rhethoric terms) a "convincing"
or (in pragmatic terms) a "cogent" argument. That is,
we have to clarify the argumentative logic of discourse
– the key issue of argumentation theory. Since it is a key issue,
and since it is at the same time a difficult issue that has
remained largely unresolved in the history of logic and argumentation
theory, I will dedicate a large portion of this essay to
it, before then turning to a much shorter discussion of
the fourth and final requirement of discourse, the need for
always being able and allowed to raise argumentation to
a higher level of self-reflection. This will lead us in
the end to a summary account of the central concept of "practical
discourse" – the employment of discourse for settling questions
of what we "ought" to do – and to a brief appreciation
main lessons that we might want to learn from this discussion
with a view to the project of promoting
'Cogent argumentation': the
a deductive logic
to a pragmatic logic of argumentation
If we want to settle our differences discursively
rather than strategically, the crucial questions becomes: How do we assess the validity (conclusiveness)
of arguments? This is a crucial issue – perhaps the most
crucial issue in any conception of communicative rationality
– and I will therefore discuss it in some detail, drawing not
only on Habermas but on a brief review of the development that
leads from Aristotle via modern logic and argumentation theory
to Habermas. Unless we clarify this issue, we cannot translate
the procedural notion of rationality that we have associated
with the ideal speech situation thus far into clear rules and
criteria of what it means to rely on the force of the "better
argument." If arguments are to be the only force that should decide
for or against disputed validity claims, we need to be clear
about the argumentative logic required – the logic of "good"
(i.e. conclusive) argumentation.
logic The traditional approach to
this question, of how we can assess the conclusiveness of arguments,
back to Aristotle's logical writings, the Organon, and
particularly to his work on the syllogism in Prior Analytics, an
early theory of the logic
of inference (Aristotle, 1984a). Logic (or analytics, as he
called it) was for him quite simply the science of valid inference.
The central concept is that
of a deduction, or in Greek: sullogismos
(a term that has a somewhat broader meaning to Aristotle than
the term "syllogism" has today in formal logic). In
a discourse in which, certain things being stated, something
other than what is stated follows of necessity from their being
so. I mean by the last phrase that it follows because of
them, and by this, that no further term is required from without
in order to make the consequence necessary. I call perfect a deduction which needs nothing other than
what has been stated to make the necessity evident. (1984a,
I.1, 24b18-24, italics added)
which is stated at the outset is the premises, and that
which follows is a conclusion. The deductive argument that leads
us "of necessity" from the premises to the conclusion
is what Aristotle calls a sullogismos; and when the deduction
is perfect, that is, requires no other backing than what has
been stated in the premises, he calls it
a demonstration. Note that Aristotle's definition allows
for logical (analytical) as well as causal (scientific) and
principled (rule-based) reasoning, which is to say, it relies
on an understanding of the "because of" behind "necessity"
which includes both analytic and substantial reasons.
vs. imperfect deduction
particular interest in the Prior Analytics is in the
question of "what sort
of deduction is perfect and what imperfect" (1984a, I.1,
24a13). The distinction allows him to analyze the special case
of merely analytic reasoning without losing sight of the general
case of conclusive reasoning that he associates with deductive
argumentation. Analytic reasoning is "perfect" in
the sense that it is self-contained, that is, it does not depend
on any evidence beyond what is stated in the argument. All other forms of deductive reasoning are "imperfect"
in that they may turn out to be not so self-contained, although
they still represent forms of conclusive reasoning. As an
example of a perfect deduction (or demonstration)
we may think of a mathematical equation. If we resolve it properly,
that is, according to the rules of mathematics, it yields a result that is correct of necessity (i.e.,
by definition) and thus requires no further backing of an empirical
or other kind. By contrast, we may think of an astronomer's
prediction of the next eclipse
of the moon as an example of an imperfect deduction.
next eclipse of the moon: an example of 'imperfect'
himself refers to this example in the Posterior Analytics (1994b,
I.8, 75b33). He does not detail it in any way though, so let me
do it for him. Like any forecast, predicting a lunar eclipse
depends on empirical premises in the form of a record of past
observations of the phenomenon in question (in this case, the
moon's moving through the shade of the earth) and moreover,
some insight into the statistical and/or causal patterns that
describe or explain this observational record. On this basis,
astronomers can calculate the exact time and location of the
next lunar eclipse (the conclusion) with a reliability
that is virtually beyond doubt. Most scientists will accordingly
tend to see the argumentative step from the premises to the
conclusion as embodying a rigorously deductive kind of reasoning,
quite along the lines of Aristotle's basic concept of deduction.
It is quite clear to them, however, as it was to Aristotle,
that the deduction is not "perfect"
in the same way as the mathematician's, in that it is never
a contradiction in itself to assume that such a prediction may
turn out to be wrong. However rigorous the argument is, we may
not possess sufficient knowledge of all the empirical conditions
on which it depends.
or 'after analytics' In
the case of astronomical forecasts, the success of past forecasts
gives us good grounds for assuming that the astronomers got
their records and calculations right. In fact we have so much
faith in their calculations that we tend to forget that the validity
basis of such astronomical forecasts, just like that of any other
forecasts, includes some inductive reasoning – a well-grounded
conclusion from particular observations of the past to general
propositions that will hold in the future. This logical step
is what we call "induction." It is different from
deduction in that the conclusion is not merely tautological
but adds new information to the premises (past observations).
This may be more obviously problematic
with other forecasts, say meteorologcial or economic
forecasts; but the crucial difficulty remains the same. It consists
in the unavoidable assumption that our premises capture all
the relevant phenomena, as well as the causal or statistical relations
between them, in a way that describes the future as well as
the past. On this assumption rests the (imperfectly) "deductive" character of
the conclusion as Aristotle understands it. But of course, since
the premises and the way we use them contain statements of an experiential (observational
and theoretical) nature, we may some day find them to describe
"some" rather than "all" of the relevant
phenomena, namely, if some previously unknown exceptions or
other restrictions emerge. Imperfect deductions may therefore always be challenged on rational
grounds, and may then require some additional evidence as to
why in the specific case the step from the premises to the conclusion
is warranted or else, on what additional conditions not previously
stated it depends – the "further terms from without"
to which Aristotle refers in his above-quoted definition of a deduction.6)
is different from perfect deductions or "demonstrations,"
which rely on premises that either are logically necessary (namely,
by definition, within an axiomatic system such as logic or mathematics)
or else have been established beyond any reasonable doubt to
represent truly universal propositions (say, laws of nature)
or principles (say, basic human rights). While it may be the
aim of science to "demonstrate" the nature
of things as an expression of the universal laws of nature (1984b,
I.2, 71b17–32); and of ethics, to "demonstrate" principles
of the virtuous life that hold good usually (though not necessarily
universally, 1985, I.3, 1094a22), Aristotle reminds us that
the normal methods of science and ethics nevertheless argue
towards, not from, universal propositions or first
principles. That is, inasmuch as they involve more than inductive
reasoning, they embody forms of imperfectly deductive reasoning,
in which the premises include some inductive elements. Demonstrations thus remain a special, ideal case
of deductive argumentation, and deductive argumentation a special
case of logical reasoning. Already for the founder of formal logic it
was thus clear that a satisfactory logic of argumentation could
be reduced to a logic of analytic (or "perfect") reasoning,
which is what is now generally understood by deductive logic. A broader notion of argumentative conclusiveness
is called for. Imperfect rather than perfect
deduction – substantial rather than analytical inference – is the daily bread of argumentative practice,
in the fields of science and ethics no less than in everyday
need for warranting principles
distinction between perfectly and imperfectly deductive argumentation
it itself imperfect, in that we cannot maintain it in argumentative
practice. Imperfect deduction always raises the issue
of how we are to establish universal propositions that can serve
as basic warrants (or in Aristotle's terms, "principles")
for conclusive argumentation. It thus depends
on a complementary logic of induction. This is the topic of
Aristotle's (1994b) Posterior Analytics. Its core difficulty
is that universal propositions or "appropriate principles"
(1984b, I.2, 72a6) cannot be deductively demonstrated, for perfect
deductions depend on such principles in the first place:
"one cannot demonstrate anything except from its own principles."
(1984b, I.9, 75b37, cf. 76a13-17 and II.19, 99b20f) Hence,
some alternative, non-deductive (or more exactly: non-demonstrative)
forms of argumentation are required,
which Aristotle describes in terms of episteme (theory
of science) and nous (theory of first principles) and
later, in his practical philosophy, also in terms of
the art of deliberation about the "right way" to orient
our practice towards eudaimonia (my definition in Ulrich,
2009a, p. 14). These alternative modes of argumentation
are all part of Aristotle's concept of reason (logos),
the rational faculty or activity of the soul that makes humans
aware of the good and the true. In more contemporary terms:
although they are fallible rather than "perfect,"
we can still assess their conclusiveness rationally.
sum, Aristotle's notion of conclusive argumentation was
not merely deductive, and his understanding of deductive argumentation
or sullogismos was not purely analytic. Instead, he allowed for
that a deductive conclusion might add new information to what was
the premises; and he gave a complementary role to deductive
and inductive reasoning in that each entailed elements of the
other. For the founder of deductive logic, the Prior
and the Posterior Analytics formed a whole just like
theoretical and practical philosophy, too, formed a whole.
After Aristotle, argumentation theory did not develop much for
a long time, and when it did start to develop again, things
went somehow downhill. Aristotle's comprehensive conception
of logic was increasingly narrowed down; the discipline of logic
was transformed from
a theory of argumentation as he had envisioned it – a logic
of deductive and inductive inference that could be used as a
tool of argumentation in all fields of knowledge and practice
to a theory of analytic reasoning only. Since the 17th century, through the work of logicians and mathematicians
such as W. Leibniz, G. Boole, A. de Morgan, J. Venn,
C.S. Peirce, G. Frege, G. Peano, A.N. Whitehead,
and many others (for an introduction, see, e.g., Smith,
2009), the study of logic has developed into the highly formalized system
of contemporary mathematical or symbolic logic, which may be thought of as a kind of
"algebra of logic." Its main branch, propositional
calculus (also called propositional logic), tells us how
by means of logical operators such as "and," "or,"
"not," "if," "only if," and
"if … then," we can combine and transform
basic sentences or propositions (understood as strings
of symbols that are associated with some defined meaning) into
more complex propositions without changing their so-called
truth value, a proposition's relation to truth (i.e.,
its being true, false, probable, or conditional). As Aristotle
might have commented, had he experienced this development:
"perfection" won out over meaningfulness at the expense
new beginning When Stephen
E. Toulmin's (2003) book The Uses of Argument first appeared in 1958, it offered
an entirely new approach to the theory of argumentation. The book does not deal extensively with Aristotle;
but by returning to Aristotle's almost forgotten quest for a
logic of argumentation that would help us establish conclusions
in different fields of science and practice, it managed to challenge the established
of formal deductive logic more seriously than any other work
did since Aristotle's day. By trying to be relevant rather than "perfect,"
it made it painfully apparent to logicians how far their field
had moved away from any argumentative practice. It was accordingly unpopular with them, they called it "Toulmin's anti-logic
book"! Despite such unfriendly early reception, the book has long since become
a standard text for anyone studying the theory and practice
of argumentation, or what soon became known as "the Toulmin
model of argumentation." Meanwhile, due to the efforts
of Jurgen Habermas to integrate Toulmin's work with the speech act theory of Austin and Searle and with
his own formal pragmatics, it has found even wider recognition
as a pioneering outline of a non-analytic – more accurately: not
merely analytic – logic of argumentation. Accordingly,
it is now often referred to as the Toulmin-Habermas model
semantic and pragmatic turn of argumentation theory
In the light of our previous discussions of speech-act theory
and of deductive logic, the aim of an "imperfect" but
relevant logic of argumentation is clear: rather than operating
at a purely syntactic level of securing
well-formed propositions or chains of propositions (WWFs, well-formed formulae), it needs
to offer us a way of grasping
the semantic meaning and pragmatic relevance of arguments in specific
and changing contexts of argumentation. It must, in other words, not be blind
to issues of hermeneutics (How may we understand the situation?)
and practical philosophy (What would in a thus-understood situation
constitute rational action?). This becomes obvious as soon as one
thinks of the expressive (e.g., emotional) and normative (e.g.,
moral) content of speech acts: the form and "truth
value" (cf. note 6) of utterances and even their propositional
content may remain the same, yet the semantic and pragmatic
implications we associate with them may change. Hence, to secure
argumentative conclusiveness in a sense that considers the relevant contexts of meaning
and action at play, we need a richer concept of conclusiveness,
one that replaces deductive necessity by pragmatic cogency
as the central notion
(a term yet to be defined). In addition, a practically useful
model of rational argumentation might also need to consider
that the nature of the argumentative process is not irrelevant
for assessing the rationality of the outcome; that is, we may
need to adopt a partly procedural notion
of validity; for the argumentative practices by which a conclusion is reached matter
as much as its form and content. In the terms that today's logicians use, such an
account of argumentation would represent a piece of "informal"
rather than formal logic; in our own terms of reflective practice
and critical pragmatism, it would allow us to measure the strength
of arguments against varying contexts and procedures of argumentation
rather than just requirements of well-formedness.
a basic alternative model for informal logic, Toulmin (2003, pp. 7f, 10, 39,
235) boldly proposed a jurisprudential analogy, for two main
reasons as I understand him. First, in legal proceedings it is more clear than
in formal logic that valid conclusions are always the result of
credible argumentative practice. And second, legal practice
renders it more obvious than the study of formal logic that
the origin and target of argumentation
is always a disputed validity claim, the meaning and
validity of which depends on the specific circumstances. By contrast, the development of formal logic since Aristotle
has led away from such practical and empirical issues; it has
therefore also failed to study the
differences and similarities of conclusive argumentation in
applied fields of
argumentation such as science, law, or medicine. Toulmin does
not claim that judicial practice provides a perfect model for
all the other fields; but
at least, he argues, it leads us beyond the narrow perspective of modern
logic towards a broader, practically oriented framework:
claim implicit in an assertion is like a claim to a right or
to a title. As with a claim to a
right, though it may in the
event be conceded without argument, its merits depend on the
merits of the arguments which could be produced in its support.
Whatever the merits of the particular assertion may be – whether
it is a meteorologist predicting rain for tomorrow, an injured
workman alleging negligence on the part of his employer, a historian
defending the character of the Emperor Tiberius, a doctor diagnosing
measles, a businessman questioning the honesty of a client,
or an art critic commending the paintings of Piero della Francesca
in each case we can challenge the assertion, and demand to have
our attention drawn to the grounds (backing, data, facts, evidence,
considerations, features) on which the merits of the assertion
are to depend. We can, that is, demand an argument; and a claim
need be conceded only if the argument that can be produced in
its support proves to be up to standard. (Toulmin, 2003, p. 11f)
can be compared with law-suits, and the claims we make and argue
for in extra-legal contexts with claims made in the courts,
while the cases we present in making good each kind of claim
can be compared with each other. A main task of jurisprudence
is to characterize the essentials of the legal process:
the procedures by which claims-at-law are put forward, disputed
and determined, and the categories in terms of which this is
done. Our own inquiry is a parallel one: we shall aim,
in a similar way, to characterize what may be called "the
rational process," the procedures and categories by using
which claims-in-general can be argued for and settled. Indeed
… law-suits are just a special kind of rational dispute, for
which the procedures and rules of argument have hardened into
institutions." (Toulmin, 2003, p. 7)
personal experience with judicial practice may not exactly suggest
as close a "parallel between procedures of rational assessment
and legal procedures" as Toulmin (2003, p. 39) proposes; mechanisms of power and institutional selectivity play an
all too pronounced role for that. But then, is judicial practice
so different from other fields in this respect? As a matter
of principle (and indirectly also, as a critique of judicial
is indeed difficult to see why a sound argument in support
of a disputed legal right or title (say, to a property or a
professional qualification) should be fundamentally different
in nature (or better, logic) from a sound argument in support
any other disputed assertion or claim, including scientific,
moral, and philosophical claims. Toulmin's judicial metaphor
is thus not as odd or arbitrary as it may look at first. As the reader may remember from an
earlier essay of this series in which we discussed Kant's concept of practical
reason and the role of the principle of universalization in
it, it was in fact Kant (1787, Axif, Bxiii, and B779) who
first used the judicial metaphor to describe the aim of
his critical philosophy: the three Critiques were
to subject reason in all its employments to the "court
of pure reason" or to "reason's self-tribunal"
(see Ulrich, 2009b, pp. 2 and 14; cf. 1983, pp. 199 and 2003).
Toulmin does not mention Kant,8)
but his intent is similar:
is one special virtue in the parallel between logic and jurisprudence:
it helps to keep in the center of the picture the critical
function of the reason.… A sound argument, a well-grounded or
firmly backed claim, is one which will stand up to criticism,
one for which a case can be presented coming up to the standard
required if it is to deserve a favorable verdict. How many legal
terms find a natural extension here! One may even be tempted
to say that our extra-legal claims have to be justified, not
before Her Majesty's Judges, but before the Court of Reason." (Toulmin,
2003, p. 7f)
and changing elements of argumentative logic
Toulmin's "court of
reason" differs from Kant's in that it is constituted by practitioners of different fields
of professional practice such as law, medicine, science, business,
ethics, philosophy, mathematics, cultural criticism, and so
on, rather than by "pure reason." Consequently, since
argumentative practice takes place in such different
fields of argument
(2003, p. 14f), we have to expect that it will be couched in
different conventions or "canons" (2003, pp. 15f and 34) and thus
will employ changing, field-dependent
criteria or standards of assessment (2003, pp. 15,
28, and 33-35). That does not imply, however, that the basic
procedure by which argumentation reaches well-grounded conclusions
needs to be different in each field; Toulmin treats this issue
as an open empirical question. We may well be able to uncover
some general, basically
(2003, pp. 15 and 33-37) features, which we may then understand
and teach as a skeleton or basic layout of arguments
that applies to all fields or uses of argument (2003, pp. 40 and 87-134).
I understand Toulmin correctly, his core idea, then, is something
like this: taking into account the field-specific
characteristics of an argument will free us to focus on the
field-invariant logical patterns at play. By paying attention
to what changes, we can learn about what remains the same, namely,
the ways we combine field-invariant with field-specific
features to formulate strong arguments. Although Toulmin does not explicitly say so, it seems to me
he applies this core idea to the philosophical task of constructing
a general logic of argumentation as well to the practical
job we all do every day of assessing specific arguments in real-world
situations of problem solving and decision making. With his
"field of arguments," Toulmin makes sure the general
framework allows for the changing
semantic and pragmatic contexts of argumentation that we have
found missing in the deductive-logical model of rational argumentation.
is, we need not escape into abstract, formal logic to ensure
general applicability and validity! Taken together,
then, Toulmin's message is:
an argument can be made to the effect that we all may, in our
argumentative practice, consider particular argumentative contexts of meaning
and relevance and yet apply forms and procedures of argumentation
that are universally valid and rigorous. Whether the argumentative
contexts are adequately specified in
the disciplinary or institutional terms of different fields
of professional practice such as those we have mentioned is another matter that need not concern us at this point;
I rather doubt it.9)
comparative empirical approach
task that Toulmin mapped out for argumentation theory is then
clear. The main difficulty in developing a generic model of
argumentation consists in the great variety of argumentative circumstances
and purposes in different fields. In response to this difficulty,
Toulmin sees logic as a philosophical discipline that includes
comparative empirical analysis of the actual working logic
– the argumentative patterns – used in different fields of argumentation,
as distinguished from the idealized logic of logical theorists (2003,
pp. 9 and 135-194).
an example of such empirically generalizing analysis, Toulmin
(2003, pp. 17-40) analyzed the use of modal terms
such as "possibly," "might," "presumably,"
"chances are," "certainly," or "necessarily" in different fields of argumentation.
How do people use such terms to qualify claims or to criticize and
defend arguments? He found that although the criteria
(standards, grounds, reasons) for asserting or questioning such qualifications vary with the
field, the qualifications (or logical modalities) themselves have the same argumentative
(i.e., implications of use, p. 28) in all fields. For instance, taking the
example of qualifying a suggestion as "possible,"
order for a suggestion to be a "possibility" in any
context, ... it must "have what it takes" in order
to be entitled to genuine consideration in that context.
To say, in any field, "Such-and-such is a possible answer
to our question," is to say that, bearing in mind the nature
of the problem concerned, such-and-such answer deserves to be
considered. This much of the meaning of the term "possible"
is field-invariant. The criteria of possibility, on the other
hand, are field-dependent, like the criteria of impossibility
and goodness. The things we must point to in showing that something
is possible will depend entirely on whether we are concerned
with a problem in pure mathematics, a problem of team-selection,
a problem of aesthetics, or what; and features which make something
a possibility from one standpoint will be totally irrelevant
from another.… "Can" and "possible" are,
accordingly, like "cannot" and "impossible"
in having a field-invariant force and field-dependent standards.
This result can be generalized: all the canons for the
criticism and assessment of arguments, I conclude, are in practice
field-dependent, while all our terms of assessment are field-invariant
in their force. (Toulmin, 2003, p. 34f)
unchanging layout of argumentation
on this kind of comparative empirical analysis, Toulmin proposes a field-invariant
"layout" of argumentative procedure and logic that any sound arguments tends
to follow in practice. We can summarize it in a
basic and an expanded scheme. Figure 1 shows the basic scheme, Figure 2 the enlarged scheme.
1: The layout of arguments
adopted from Toulmin, 2003, pp. 92 and 97
basic scheme works with four components:
C = Claim: a conclusion to be justified. Example
of Toulmin (2003,
pp. 92-99, slightly adapted here): “arry is a British citizen.”
D = Data: an empirical observation or a statement of "fact"
that is offered as evidence for C. Also called
G = Ground (esp. in Toulmin et al, 1984). Example: "Harry was born in Bermuda,
a British overseas territory."
W = Warrant: a rule or principle
that justifies the step (transition)
from D to C. Example: "A person born
in a British overseas territory
will generally be a British citizen."
B = Backing: some evidence or a general reason
in support of W, to be supplied if citing
W is not sufficiently convincing to all those addressed.
There are two logically different kinds of B: If
B implies C, the argument is merely analytic, as in syllogistic
logic. If however C is not implied in B (the more important
case for argumentative practice), then the argument is substantial,
that is, it adds information and is not covered by syllogistic
logic. Example (of the substantial kind): "This is so on account of the following
statutes and legal provisions: … (e.g., the British Nationality Act
1981 and the British Overseas Territories
these four components, the first three
are required and are therefore usually explicit in any sound argument,
whereas the fourth is required only if someone challenges the warrant W,
and will thus remain implicit in many arguments.
But since a challenge is always possible, any argument consisting of the first
three components (D, W, so C) implies the availability of the
fourth (i.e., some B) and may, if doubted, need to make it explicit (D, W, B, so C). But what
happens if B is challenged in turn? Then the proponent of C
may either offer an alternative, hopefully more convincing backing (B'),
or else may argue why the original backing (B) is valid. In
the latter case, the "T" layout applies once again,
so that B then results as the conclusion of a preliminary argument
(D', W', so B), or in a short notation that Toulmin does not
a preliminary argument is possible if D is challenged (D',
W', so D):
"T" layout is in this sense recursive, that
is, it may be applied to its own components – an important
characteristic that renders its use very flexible and allows
to build entire chains of arguments. Some recursive loops – recurring
"rounds" of argumentation about an argument's components
– may indeed be very useful at the outset to prepare the ground,
as it were, and must obviously remain possible at all times as
the argument unfolds. In a sense, then, such recursiveness constitutes
the methodological core of what Habermas terms the step from
communicative action to discourse, as well as of the argumentative
principle in general. Although neither Habermas nor Toulmin
say it in these terms, the recursiveness of the "T"
layout seems crucial if discourse (the argumentative process)
is indeed to "bracket" (suspend) all issues except that of a disputed
claim's validity; for only thus can the assumptions and implications
of arguments be freely unfolded. On the other hand, if the participants take
this recursive business too seriously and keep challenging each
other's Bs and Ds from the outset, then the argument about the
original claim (C) never really starts. The good news is that
the danger of an infinite regress is only a theoretical risk; practically
if discourse is to play a role, the participants need to share
some basic assumptions, otherwise they have no basis
for reaching an understanding at all.
There are two more components, which the proponent
of an argument may, but need not, employ from the start. They are
useful whenever participants
question the force of basic arguments (i.e., arguments following
the basic scheme of Fig. 1), in that they may help to avoid endless recursive
loops or else, a breakdown of the argumentative process altogether:
Fig. 2: The layout of arguments
adapted from Toulmin, 1984, p. 98, and 2003, p. 97
two additional components are:
Q = Qualifier: a modality
expressing the force (strength or certainty) with which
is asserted, typically formulated with a term
such as "presumably," "surely,"
"probably," "necessarily," "in
general," "chances are," or “as
far as the evidence goes.” Qualifiers expressing incomplete strength recognize
the conditional character of an argument, allowing for
the possibility of rebuttals. Example: "Chances
are Harry is a British citizen, unless he has become
a naturalized American or neither of his parents was
a British citizen."
R = Rebuttal: a statement of some exceptional
circumstances that may limit or undermine the force of an
argument (specifically of Q, W and B) and thus the validity
of C, typically beginning with "unless,"
"except that" or "if and only if."
Example: "Someone born
in a British overseas territory may generally be assumed
to be a British citizen, except that in this case,
neither of Harry's parents was
a British citizen, so the British Overseas Territories
Act 2002 does not apply."
3 shows an example taken from meteorological practice.
For a hyperlinked overview of all issues
of "Ulrich's Bimonthly" and the previous "Picture of the
see the site map
This is the second of three parts reviewing the implications
of the work of Habermas for reflective professional practice.
The first part appeared in the Bimonthly of September-October
2009; the third and concluding part is planned for a later Bimonthly.
It seems to me Toulmin indeed offers us a generic model of argumentation.
It is generic in at least two senses: first, it is applicable
to practical questions (Fig. 4) as well as to theoretical
questions (Fig. 3); and second, it encompasses "logical"
issues not only of analytic but also of substantial reasoning.
It may thus help us to recover the broader notion of logic as
argumentative logic with which Aristotle started out
two millennia ago, prior to its subsequent reduction to formal
deductive logic. This historical curtailment of argumentative
logic (and ultimately, the logic of systematic thinking) still
hinders and impoverishes our contemporary notions of what rational
conclusions – rational argument and criticism – are all about.
To mention just two major examples, it is still prevalent in
the "exact" sciences in the form of the so-called Hempel-Oppenheim model of explanation,10)
and even in the "inexact" sciences it has remained
prominent in the form of Popper's earlier-discussed deductive
concept of "rational criticism." The unspoken ideal
of such a deductive notion of "rational" conclusion
is to eliminate from systematic thinking all elements that cannot
be entrusted to a machine or to a "propositional calculus."
To be sure, the advantage of analytic reasoning is that it can
do without considering the empirical, normative, and expressive
content of conclusions; but the price we pay for measuring the
rationality (or conclusiveness) of all thought and argumentation
against such an ideal is definitely too high – it begs the issue.
For as we have learned from both Aristotle and Kant, but also
from many other outstanding thinkers about the nature of thinking
(e.g., Dewey, 1910, and Bateson, 1972, 1979), the task of rational
thinking and argumentation consists precisely in establishing
the connections between things that experience alone
cannot give us; the pattern which connects or "metapattern,"
to use Gregory Bateson's (1979, Ch. 1) famous phrase. Only
reason can inform us about the basic principles that
connect things, both in experience (theoretical reason) and
in action (practical reason). Allow me to summon John Dewey
as an independent witness who is widely respected for his account
of How We Think:
is thus a double movement in all reflection: a movement from
the given partial and confused data to a suggested comprehensive
(or inclusive) entire situation; and back from this suggested
whole ... to the particular facts, so as to connect these with
one another and with additional facts to which this suggestion
has directed attention.… To think means, in any case,
to bridge a gap in experience, to bind together facts or
deeds otherwise isolated. (Dewey, 1910, p. 79f, my
teaches us how to bridge the gap rationally. It bursts through
the limits of a merely analytic concept of "conclusiveness."
Although it superficially resembles Hempel and Oppenheim's (1948)
model, it recognizes that the job of substantial reasoning is
to add new content to what is
previously known or assumed (the premises), and that merely analytic
schemes of conclusive argumentation cannot handle this task.
We are facing an epistemological rather than just a deductive-logical issue. The crucial question
is how we can justify knowledge (or in any case, the new content
in question). To reduce this question to a merely analytic issue
implies an error of category or in Toulmin's (2003, pp. 150,
153, 155, 212-216) term, a type-jump
– an impossible inferential leap from analytic
conclusiveness (a tautology) to substantial conclusiveness (new
content). Type-jumps are unavoidable, but they involve a non-analytic
transition from one type of logic to another, and thus burst
the framework of analytic conclusiveness. This does not imply,
however, that they are arbitrary, or that the arguments in question
cannot be conclusive; all it implies is that they are not analytic,
and in this sense non-trivial.
need for 'type-jumps' To require, as formal logicians do, that conclusions
must always (i.e., in any rational argument) follow analytically from the data and backing, amounts
to an inadequate handling of type-jumps. The error, to be sure, is not the attempt
to jump from D and B to C, but only the attempt to treat the
jump as a purely analytic issue. This attempt is bound to
lead us into an apparent logical gulf –
is, because it is merely the consequence of a narrow understanding
of what "logic" and "rationality" are all about. The
gap is an analytical gap, but not necessarily an argumentative
gap. Argumentative logic
is about rational argumentation; but rational argumentation
is not just about internal consistency, it is also and mainly
about the "strength" (relevance, force, cogency) of an argument
within specific contexts of meaning and action.
Although internal consistency of arguments is always a necessary
requirement for "strong" argumentation, it is not
a sufficient criterion, except of course in purely analytic judgments
– a special, particularly
simple case of conclusiveness that we must not mistake for all there is to argumentative logic.
If we do, and consequently try to define rational argumentation in purely
analytic terms, we are bound to end up with a bottomless epistemological
much for Toulmin's pioneering analysis. Let us now draw some
conclusions for the step from a deductive logic of inference
to a pragmatic logic of argumentation (step 3 in Table 3).
1: Farewell to 'Hume's problem' It
is difficult in this connection not to think of David Hume's
(1978, Book I) long-standing critique of empiricism and
inductive reasoning, which has remained an unresolved problem
for epistemology ever since. "Hume's problem" has remained unresolved, as
now begin to understand, because he defined it in a self-defeating
way. It was the inevitable consequence of his attempt to reduce the logic of inquiry
(i.e., of substantial argumentation) to one of analytic reasoning
only. Thus seen, it was indeed "Hume's problem"; an
artefact of his assumptions. To do justice to Hume, his attempt pursued a critical
purpose; it taught us that any such attempt is futile. Because something
like a language-analytical turn of argumentation theory was
out of sight then, he had no option but to try and explain substantial argumentation
in analytic terms – and had to fail. Understandably, neither Hempel and Oppenheim's
(1948) deductive model of scientific explanation nor Popper's
(1959) "falsificationist" use of deductive logic could really
solve Hume's problem, although, to do justice to Popper, he probably
came as close to a solution as a purely analytic framework,
without access to hermeneutic and pragmatic reasoning, could
get. The difference
is, Hume recognized that his experiment had failed!
Hume, as well as Hempel and Oppenheim's Hume, is definitely not Kant's Hume, the
managed to awake the great critical philosopher from his slumbers!
Nor is he Toulmin's Hume, who makes us understand that any attempt
to reduce rational argument to a deductive-logical concept of
rationality commits a petitio principii:
every step he rejected anything other than analytic criteria
and proofs. There is no certainty that a pinch of salt put in
water will dissolve. Why? Because, however much evidence I may
be able to produce of salt's dissolving in water in the past
or present, I may suppose that a pinch dropped in water tomorrow
will remain undissolved without contradicting any of this evidence.…
Throughout the Treatise Hume appeals repeatedly to considerations
of this kind: the understanding is to admit arguments
as acceptable, or "conformable to reason," if and
only if they come up to analytic standards. But, as he soon
discovers, all arguments involving a transition of logical type
between data and conclusion must fail to satisfy these
tests: however grotesque the incongruity produced by
conjoining the same data with the contradictory of the conclusion,
the very presence of a type-jump will prevent the result from
being a flat contradiction. (Toulmin, 2003, p. 152f)
a less self-defeating approach can begin with Toulmin's (2003,
p. 212) recognition that we should not "talk away"
the need for type-jumps, that is, simply eliminate them from
our concept of rationality. We better learn to handle them carefully!
Handling type-jumps carefully is what Toulmin's layout of argumentation
is all about. It teaches us how to take the step from D and
B to C in a way that deals explicitly and critically with the
warrant W or, in the earlier discussed terms of Aristotle, with
the "principles" on which we rely in taking this non-analytic
step. We can now formulate two essential guidelines to this
model accurately defines and locates "Hume's problem"
as the type-jumps involved in all non-trivial (i.e.,
not just analytic) argumentation. It makes us understand
that the argumentative force or cogency
of an argument depends essentially on the way we bridge
the analytical (but not argumentative) gap between
B and W, and consequently, the resulting gap between
D and C.
model tells us precisely how to handle the two
non-analytic transitions ("type-jumps") involved,
from B to W and from D to C. It calls for, and regulates,
a discursive validation of the "bridge principles"
we use, whether we are aware of them or not, for this
be sure (and here I seem to differ a bit from Habermas' understanding
of Toulmin, to which I will turn in a moment), we must never
forget that "bridge" principles are just that:
auxiliary principles that help us in making those non-analytic
transitions. They serve us to understand the type-jumps
involved, but not necessarily to justify them in any
definitive way; they are working hypotheses, as it were. The
point is, in substantial reasoning we cannot avoid relying on
some bridge principles; hence, from a critical point
of view, it is imperative that we make it clear to ourselves
and to everyone concerned what these principles are and how
they affect the perceived strength of an argument. Although
we need them for assessing arguments, they should not stop us
from considering, in each case, alternative transitions.
basic lesson concerns our understanding of the principle of
excluded contradiction, as the core principle of analytic
reasoning. Hume, Popper, Hempel, and Oppenheim all appear to
have overestimated how far it carries. Counter to them, I suggest
we understand it as a criterion
of meaningfulness rather than of validity: we cannot argue meaningfully
if we contradict ourselves, and that is why we need it. But
validity is a different issue. In purely analytic reasoning
we may take meaningfulness and validity to be congruent (propositions
that are logically true are logically meaningful and those which
are logically false are by definition not meaningful), which
is to say, we do not need a separate concept of validity at
all. In assessing the validity of substantial arguments, however,
it is never a contradiction in itself to imagine that the contrary
conclusion or claim might be true; sometimes it is a critical
necessity to do so! To put it differently: whether a
claim is logically implied or contradicted by its premises tells
us nothing about what difference it makes in specific contexts
of meaning and action. Insisting on analytic criteria for assessing
the validity of substantial claims is therefore beside
the point (Toulmin, 2003, pp. 156 and 216). The
third basic guideline, then, is something like this:
principle of excluded contradiction is not an adequate
bridge principle to ensure valid transitions from B
to W and from D to C. It is a necessary condition of
meaningful argumentation but not a sufficient condition
of cogent argumentation.
a forth and last lesson, we
may apply Toulmin's analysis to Hume's negative assessment of all
reasoning: although deductive-logically correct (by definition!), it is epistemologically
beside the point. An analogous conclusion obviously holds
for issues of practical philosophy. All Hume's rejection of
inductive reasoning really tells us is that
inductive logic is different
from deductive logic. That is, it calls for a richer concept
of conclusiveness, one that takes into account the specific
and changing contexts of argumentation, as well as probably different procedures of –
non-trivial – argumentation.
Which is what Toulmin's work is all about. Our fourth guideline,
therefore, may read:
is time to bid farewell to "Hume's problem":
Toulmin's analysis has freed us once and for all to see that
"non-analytic arguments also can
be conclusive" (2003, p. 216).
beyond Toulmin's model, we will want to embed his layout of
argumentation in a broader, hermeneutic and pragmatic framework
for critical discursive practice such as it has become available
through Habermas' work. Let us, then, return to Habermas "formal-pragmatic"
reconstruction of argumentation theory.
2: The Habermas-Toulmin model of argumentation
What we call the "Toulmin-Habermas model"
is simply the way Habermas adopts
Toulmin's model and embeds it in his
larger framework of formal pragmatics. As is to be expected,
he ties it to the "general pragmatic presuppositions"
of communicative rationality that we have discussed earlier.
The layout of arguments remains the same, only its interpretation
and use in discursive practice is partly different from Toulmin's
reading. There is no need to repeat our account of the "formal-pragmatic"
lens through which Habermas (e.g., 1973c, pp. 238-252;
pp. 22-27 and 31-42; 2009, pp. 243-259) reads Toulmin's layout
of argumentation; it is clear that he uses it to enrich
and operationalize his understanding of "rational"
discourse with concepts such as the telos
of mutual understanding and the general symmetry conditions of discourse,
as well as
with his analysis of the different types
of validity claims involved in all communication, with the resulting
notion of a universal validity basis of speech, and so on. It
may be more helpful, instead, to offer a short discussion of
those particular aspects of Toulmin's reading that he welcomes
and those which he wishes to revise.
with the "welcoming" part of Habermas' reception,
he finds it essential that Toulmin's conception of argumentative
logic includes issues of argumentative practice
that reach beyond formal logic. He acknowledges that by
considering different uses and contexts (or "fields")
of argumentation as well as the non-trivial transitions these
uses of argument may involve, Toulmin opened the discipline
of logic up to the wider concerns of a theory of argumentation
properly speaking, a theory that can deal with
the hermeneutic and pragmatic contexts of argumentation. Already
his early writings on communicative competence and on the need
for a consensus theory of truth made it clear that Toulmin's
analysis helped him in developing an adequate understanding
of argumentation theory in the first place, for example, as
it relates to his concepts of "rational motivation,"
of "discourse," and of "rational consensus";
in particular, it made him see more clearly that "the logic
of discourse is a pragmatic logic [that] examines the formal
properties of contexts of argumentation." (Habermas, 1973c,
p. 249). Later, in the Theory of Communicative Action,
Habermas (1984, p. 31) explicitly designates it as an "advantage
of Toulmin's approach" that "he allows for a plurality
of validity claims while not denying the critical sense of a
validity transcending spatio-temporal and social limitations." He
is similarly explicit about the value of Toulmin's empirical finding
of the field-invariance of both the layout of arguments and
the force of modal qualifications.
these many points of agreement, Habermas' finds
it necessary to expand Toulmin's perspective. For Habermas,
a proper theory of argumentation amounts to nothing less but
a theory of rationality in general, and such a theory can for
be a social theory of argumentation, that is, part of
a more encompassing social theory as he envisions it with his
theory of communicative action. At the other end of the scale,
Habermas thinks an adequate argumentation theory requires a
further-reaching basis in language theory. In addition to this broader
outlook, Habermas has a number of more specific methodological concerns that
do not allow him to adopt an empirically generalizing approach such
as Toulmin's without further ado;
myself to mentioning three of them.
back in the 'process' and 'procedure' perspectives
First of all, Habermas finds that Toulmin focuses one-sidedly
logical (or "product") perspective of argumentation
while rather neglecting the rhetorical (or "process")
and the dialectical (or "procedure") perspectives (cf.
Table 3). Especially the latter is of course essential to Habermas. In
his view, therefore,
does not push the logic of argument far enough into the domains
of dialectic and rhetoric. He doesn't draw the proper lines
between accidental institutional differentiations of argumentation
[read: fields of argument] on the one hand, and the forms
of argumentation determined by internal structure [read:
types of validity claims and processes required to substantiate
them, i.e., to reach rationally motivated agreement], on the other. (Habermas,
1984, p. 35)
example, much of the argumentation going on in the field of
legal practice is oriented towards success, negotiation, and
at best compromise, rather than towards reaching genuine agreement
(as, say, in the fields of science and moral discourse). However, "negotiating
compromises does not at all serve to redeem validity claims
in a strictly discursive manner, but rather to harmonize nongeneralizable
interests on the basis of balanced positions of power";
and furthermore, "arguments in a court of law … are distinguished
from general practical discourses through being bound to existing
law, as well as through the special restrictions of an order
of legal proceedings that takes into account the need for an
authorized decision and orientation to success of the contesting
parties." (1984, p. 35)
the pragmatic presuppositions of discourse
A consequent second concern relates to what Habermas sees as wanting clarification
of the pragmatic presuppositions of discourse in Toulmin's account.
just mentioned that there
are relevant differences of purpose between argumentation in court
(Toulmin's jurisprudential model) and argumentation in rational discourse properly speaking
(Habermas' discourse-theoretic model). In particular, argumentation
in court is not relieved from external pressures
such as the influence of power and the "need for an authorized
decision" (Habermas, 1984, p. 35). In
legal practice the participants
are usually pursuing a strategic rather than communicative orientation,
quite apart from arguing under heavy pressures of cost and time
as well as asymmetric distribution of decision authority. Toulmin's
account remains rather silent on such issues, which for Habermas
call for a methodological counterconception (or standard) such
as the "ideal speech situation."
the suppression of generalizable interests
A third and last concern that I want to mention here regards
the distinction of
nongeneralizable vs. generalizable interests. When we agree
or argue about a validity claim, we need to understand what
it means for the different parties concerned; to which extent
has it a bearing on everyone's interest or only on some particular
interests? If it is to address such issues, an adequate theory of argumentation cannot
do without giving a well-defined role to Kant's principle of universalization (or generalization).
Toulmin's framework, due to its empirically generalizing rather
than philosophically constructive approach, appears to offer
no systematic place to Kant's principle, or at least remains
largely silent on its role. For Habermas (1984, pp. 17,
35), argumentation and discourse
can in the end only lead us to valid conclusions if they address
audience of all those concerned, that is, are open to everyone who may have something
to contribute or to object.11)
Convincing a universal audience, so as to
gain general assent for one's claim, is for Habermas (1984,
p. 26) "the fundamental intuition connected with argumentation."
As he sees it, Toulmin does not distinguish clearly enough between
generalizable and nongeneralizable interests; in fact, Toulmin's
focus on the empirical analysis of a number of fields of argument
such as law, morality, science, management, and art criticism,
with their institutionally and professionally bounded audiences,
rather works against a universalist perspective. Habermas sees
a danger that with such an empirical and institutional orientation
of our notions of sound argumentation, our argumentative practice
may inadvertently rely on some
preexisting notion of rationality, rather than making rationality
its core subject (Habermas, 1984, pp. 33-35)
Habermas I would argue that an adequate framework for argumentative
practice should indeed give a more central place to the universalization
principle than it is given in Toulmin's work. This seems particularly
obvious when it comes to the normative implications that discursively
reached agreements may have for third parties. Without the Kantian idea of
testing and justifying our claims with a view to the generalizability
of underlying norms or principles of action, we risk losing sight of the "critical difference
between warranted and unwarranted consensually achieved decisions."
(Burleson, 1979, p. 113, quoted in Habermas, 1984, p. 35)
But similar conjectures are equally appropriate regarding the
procedures used in the sciences for generalizing observational
statements to hypotheses and nomological laws. It is the same
essential concern which led Peirce (1878,
par. 407), in the realm of theoretical discourse, to understand
truth as a the ultimate agreement of an indefinite community
of competent researchers; and Kant (1786, 1788; cf. Ulrich,
2009b), in the realm of practical discourse, to understand
morality in terms of moral universalization.12)
definition of pragmatic cogency
In consequence of these and other observations, Habermas wishes to give his theory of discourse a more clearly
pragmatic and discourse-theoretic twist
than he finds it in Toulmin's model of substantial argumentation.
Successful argumentation, apart from not exhausting itself in deductive-logical
inferences, amounts to what Habermas terms
cogent argumentation. Cogent argumentation is basically similar
to Toulmin's concept of conclusive argumentation in that it
involves "type-jumps" and for this reason entails
argumentatively non-trivial transitions from premises (D and
B) to conclusions (C, via W). Beyond
that shared understanding, it is essential for Habermas to insist
that a discursively reached agreement should count as rational
only the extent it is the result of a rationally motivated,
undistorted discourse. He therefore maintains that we can adequately
conceive of argumentative
cogency only in terms of communicative rather than strategic reason;
in addition to Toulmin's layout of cogent argumentation, such
a concept of cogency entails corresponding requirements of process
(communicative competence), procedure (undistorted discourse),
and product (rationally motivated agreement). The argumentative
process, procedure, and product must all live up to the general (or formal) pragmatic presuppositions of argumentation;
we have summarized these conditions, in Tables 1-3, in terms
of different core issues and requirements of communicative rationality
and types of validity claims concerned. Furthermore, since for
Habermas a proper logic of cogent argumentation is a pragmatic
logic, we need a clear understanding of how we define argumentative
conclusiveness in pragmatic terms. As Habermas explains:
terms of discursive modalities, an argument is unfitting
(or impossible) if W cannot be interpreted as a rule
of inference that allows the transition from D to C. An
argument is compelling (necessary) if C can be inferred
from B; in this case we have an analytic rather than substantive
argument, for W is not adding any information to B.
We call an argument cogent if and only if it is possible
in terms of discursive modalities.
This is the case if there is no deductive relation between B
and W, but B nonetheless provides sufficient motivation
for accepting W as plausible. We call such arguments substantive,
as they generate plausibility despite a logical discontinuity,
that is, a type-jump [Typensprung] between B
and W. (Habermas, 1973c, p. 243, and 2009, Vol. 2,
p. 249, my transl.; note:
in the second sentence of the German text, both in the 1973
original and in the 2009 edition, 'C' is misspelled
summary account of cogent argumentation is precise, but not
easy to handle. It may be advisable for later reference, therefore, to
it into the following definition.
Within a pragmatic logic of substantial argumentation along
the lines of Toulmin and Habermas, we
may define argumentative cogency as
follows. An argument is "cogent" if and only if:
step from D and B together to C is a substantial one
(i.e., D and B do not entail C analytically, or in other
words, C is not logically necessary);
is logically and theoretically possible
(i.e., it contradicts neither logic nor the facts); and
is redeemed discursively, that is, it effectively
meets with rationally motivated consensus (i.e., it
convinces everyone concerned to agree, under conditions
of basically unconstrained discourse).
conclusion, then, we may say that in the Toulmin-Habermas model
of argumentation, the layout of argument itself (as proposed
by Toulmin) does not change, but its understanding and use does.
of discourse': the step
from initial to higher levels of reflection
we now return to our starting point – the requirements of rational
argumentation as summarized in Table 3 – there remains
a fourth and last step we need to take. Its necessity follows
from the preceding discussion. A pragmatic concept of argumentative cogency does not alter
the fact that in all non-trivial,
substantial, argumentation there is an element of inductive reasoning involved. It is thus
always possible and meaningful to question the cogency of the
step from D (via W and B) to C, or quite simply to argue for
an alternative conclusion. Habermas responds to this issue
with two strategies. The first strategy builds on the idea
of bridge principles that should render
the step from D to C plausible, despite its inductive implications
(i). We have already encountered two such bridge principles,
indefinite community of researchers (when C stands for theoretical
claims) and Kant's concept of moral universalization (when C
stands for practical claims). In
addition, Habermas suggests the "principle of discourse"
as a third bridge
principle (we will discuss this in connection with his
discourse ethics). The second strategy builds on the idea that
a radicalization of discourse must always be an option,
in the sense that discourses may become their own subject (ii). That
is, whenever the
plausibility of the step from D to C becomes problematic, Habermas
suggests a practical need for taking the discourse to metalevels at
which the presuppositions of inductive reasoning can be analyzed.
Rather than relying on general bridge principles and reconstructive
analysis alone, we might say,
Habermas puts his
faith in the discourse participants themselves, by entrusting
them with the task of ensuring to their argumentative efforts
a self-reflective dimension.
the present context, I am mainly interested in the second strategy,
as it completes the idea of a progression of discursive
steps by which we try to understand the meaning of a "good"
argument, and accordingly the rationality requirements of discourse
(cf. Table 3). Habermas does not discuss the role of bridge
principles together with his notion of a radicalization of discourse,
yet it seems to me that the two strategies
are to some extent interdependent, in that the need for radicalization
arises partly from the somewhat precarious nature of the "bridge
principle" strategy. In other words, I believe the bridge
principle strategy cannot stand alone; only together with the
"radicalization" strategy is it credible. It makes sense, therefore, to begin
with a brief discussion of the first strategy.
(i). The role of bridge principles To better understand the nature and role of bridge principles, and
of methodological "reconstruction" in general, Habermas (1973c,
pp. 246-252; 1979a, pp. 14-22; 1979b; pp. 73f
and 77-82; 1984, pp. 2f, 67-69, 138-140; 2009,
Vol. 2, pp. 252-259) turns to Noam Chomsky's
(1965) analysis of linguistic competence, according to which
linguistic grammar is not conceivable without a corresponding
mental grammar, and to Jean Piaget's
(1932, 1970) research on the cognitive (intellectual and moral) development
of children. Further important sources are George Herbert Mead's (e.g., 1913,
1925, 1934) work on "symbolic
interactionism," with its central question of how we form our
sense of identity as members of society, our "social self";
and Lawrence Kohlberg's (1968, 1976, 1981,
and 1984) work on the stages
of moral development – two sources that we have discussed
earlier in this series (Ulrich, 2009b). In all these approaches,
"formal explication of the conditions of rationality and
empirical analysis of the embodiment and historical development
of rationality structures mesh in a peculiar way." (1984,
p. 2). The essential idea is
that all our cognitive capabilities, and thus also the bridge principles
on which we have to rely (and usually do rely intuitively) in inductive reasoning, embody
linguistic and cognitive
schemata that form in the course of our intellectual and
the basic predicates available in the languages we use for argumentation
do indeed express such cognitive schemata, induction means something
rather trivial: namely, the exemplary repetition of exactly
that type of experience which previously formed these cognitive
schemata themselves.… Induction thus loses its mysterious character,
although the limits of what it can achieve become equally apparent.
data [read: D and B] available for inductive confirmation or
rejection [of propositions] are unavoidably preselected
by our linguistic and conceptual framework [Sprachsystem],
so much so that "experience" cannot represent an independent
instance of validation.… It is, then, an entire framework rather
than any particular proposition which is effectively confronted
with reality; and this framework is regulated by our cognitive
development. (Habermas, 1973c, p. 246f; 2009, Vol. 2, p. 252f ,
my simplified transl. and my italics)
this is so, Habermas appears to suggest, we can indeed
have some basic faith in the adequacy of the cognitive schemata that we have
learned to apply to different domains of
experience and argumentation; for these object-domains shaped
our cognitive schemata in the first place. They act in
this sense as "guarantors" (1973c, p. 246; 2009,
Vol. 2, p. 252) for the adequacy of our argumentative languages,
although not of course for the validity of our claims; the latter
can only be redeemed discursively, and such redemption must
now include the dimension of the larger cognitive frameworks
The argument looks
rather similar to Kant's (1787, B193-197; cf. Ulrich,
1983, p. 208) famous "highest principle of all synthetic
judgments," according to which we ultimately cannot help but
presuppose that there exists a fundamental convergence between the
(cognitive) conditions of possible experience and the (ontological)
conditions of the objects of experience. But there is an important
difference: we can no longer unproblematically assume
today that the conditions of objective experience are at the
same time sufficient conditions for truth, as Kant could still
assume. "Objectivity" and "truth" have fallen
apart, or as Habermas (1973b, pp. 382-293, cf. Ulrich, 1983,
pp. 113-115) explained in his famous "Postscript"
to Knowledge and Human Interest, Kant's transcendental
a priori has dissolved into an empirical a priori
of experience and a discursive a priori of argumentation.
This is why Habermas, in addition to acknowledging the (unavoidable)
assumption of a basic adequacy of our cognitive apparatus, needed
to introduce all his "formal-pragmatic" provisions
for argumentative cogency. Ever since the "Postscript,"
he has therefore focused mainly (and as I have always felt,
all too one-sidedly; see the discussion in Ulrich, 1983, pp. 153-166,
esp. pp. 158 and 163) on the a priori of argumentation.
Only with Truth and Justification, he has recently (2004)
turned back the wheel a bit.
fact, it is because the two sets of conditions – concerning
the constitution of experience and the validation of claims
– are interrelated and must come together, that induction
may lose some of its mysterious character, as Habermas writes
in the above-quoted passage. Inasmuch as our cognitive schemata
are conditioned by our social and intellectual
development (both as a species and as individual), inductive reasoning is perhaps, as Habermas seems
to suggest, more trivial than we tend to think, namely, in that
it need not start from scratch with each argument but
has a history of maturation, a past record of probation as it were. I
may not be thoroughly convinced, nor do I assume the reader
is; the important point for me is, rather, that in any case
we should not take our cognitive schemata (including bridge
for granted. We better watch carefully how they influence both
the meaning and the validity we attribute to an argument – which
leads us to
the second strategy.
(ii). 'Radicalizing' discourses The cognitive schemata in question express themselves
and become effective through the specific linguistic and conceptual
frameworks that we use in argumentation. Consequently, communication and discourse take on an additional
role: they are not only means to exchange information
and arguments but also means to make us aware of, and "enlarge,"
our linguistic and conceptual frameworks. The substantial critique
of validity claims unfolds into a substantial critique of language.
That is, adequate argumentative procedures must allow for
a revision of the conceptual framework of a discourse, so that facts (D),
backings (B), norms or principles (W), and conclusions
(C) can all be reinterpreted and questioned
in a different light.
In the field of theoretical questions, this may also mean that
the theoretical framework used is questioned; in practical questions,
that the assumed ethical or political framework is questioned. For
example, in environmental discourses (say, about an environmental
impact assessment), participants may want to question whether
the wide-spread practice of measuring the value of natural resources,
as well as people's concern for nature, in financial terms,
is adequate; this may lead to a critique of the dominating framework
of cost-benefit analysis (CBA) and its theoretical, ethical,
and political implications as to what counts as "rational"
environmental policy. A satisfactory logic of substantial argumentation depends on
this possibility of a metalinguistic, metatheoretical, and metaethical or metapolitical radicalization of discourse
(1973c, p. 253f, 2009, Vol. 2,
ultimately unfolds into a
critique of knowledge, in which the
normative foundation of knowledge becomes problematic. At this highest
of reflection, the boundaries between theoretical and practical
questions become blurred, in that it is no longer
possible to distinguish sharply between them; we encounter,
in a famous formulation of Habermas (1971b, p. 61), a "dialectic
of potential and will," that is, an ultimate, unavoidable
interdependence of what we can know and do on the one hand,
and what we may want to do and ought to do on the other hand.
In the example of environmental discourses, what counts as "rational"
environmental action depends on a complex interplay between
our conceptions of environmental expertise (how do we identify
and assess risks and what do we know about the efficacy of alternative
environmental protection policies) and environmental ethics
(what place do we give to market values, aesthetic and spiritual
values, the options of future generations, and so on). Counter
to what advocates of "green" politics sometimes appear
to assume, there is no such thing as a straightforward conception
of "right" environmental action and "true"
possibility of a progressive radicalization of discourse to
increasingly reflected levels is therefore indispensable. Together,
these levels constitute the self-reflective dimension of the Toulmin-Habermas
model of argumentation (Table 4).
Table 4: The
self-reflective dimension of the Toulmin-Habermas
model of argumentation: levels of progressive radicalization
from Habermas, 1973c, p. 254; 2009, Vol. 2,
p. 262; and Ulrich, 1983, p. 141)
Entry into discourse
(assertions of fact, predictions, nomological
(action proposals, evaluations,
commands, prohibitions, etc.)
Substantial critique of validity claims
of theoretical discourse
of practical discourse
Substantial critique of language
revision of language
/metapolitical revision of language
Critique of knowledge and will
the interdependency of theoretical and practical
(reflection on what ought
to count as knowledge)
(reflection on what ought to count
as right interest or action)
view of the dialectic of potential and will
2009 W. Ulrich
brief analysis of the step from discourse to metadiscourse concludes
our discussion of the rational structure of discourse according
to Table 3. Four crucial steps have led us from everyday
practice to communicative action and on to discourse, to a pragmatic
concept of argumentative cogency, and to the option of metalevel
discourses. Each step embodies a self-reflective turn of the
previous conception of communicative rationality. It is time
to turn from the theory of communicative rationality to its
practical discourse, discourse ethics, deliberative democracy,
and social theory
So what? What is all this detailed analysis of the
formal-pragmatic conditions of competent speech,
meaningful communication, and cogent argumentation good for?
It is obviously not an end in itself but is to supply a theoretical
and methodological foundation for Habermas' larger project.
We have characterized this project at the outset as a quest for overcoming
"the jagged profile
of modernization" – the selective patterns of rationalization
that historically have developed in the course of an increasing
differentiation of competing "complexes
of rationality" and which threaten to undermine the project
of modernity, that is, the vision of an open and enlightened
from this initial characterization, we have not considered the
social theory of Habermas strictly speaking. I have preferred
in this introductory discussion to focus on the methodological foundation
on which Habermas aims to base his social theory as well as
his political vision, that is, formal pragmatics and what I
consider to be its methodological core, the Toulmin-Habermas
model of argumentation. On it rest our hopes, if we are to follow
Habermas, for strengthening noninstrumental patterns of reasoning and societal
rationalization, as against the current prevalence of one-dimensionally
instrumental patterns of rationality in many domains
The importance of the Toulmin-Habermas
model of argumentation derives from the fact that it extends
the range of rational discourse from questions of analytical,
theoretical, and instrumental reason to questions
of practical (ethical, moral, and political) reason. This is
so, we have understood through Habermas' analysis of the universal
validity basis of speech, because not
only claims to truth (assertion of facts) and to truthfulness
(expression of motives) but also claims to rightness (stipulation
of norms) admit of argumentative vindication and challenge.
Accordingly, the basic vehicle for extending the reach of communicative
rationality becomes what Habermas calls practical discourse.
While cogent argumentation is a generic concept that applies
to theoretical discourses as well, it is in the domain of practical
questions that we most urgently need new conceptions of rational
practice. Science has long since found ways to implement theoretical
discourses successfully and in this way to ensure (imperfectly)
rational research practices. But when it comes to applying such
rationality to applied science and expertise, as well
as to everyday problem solving and decision
making, we seem to be at our wits' end. The core questions we
then face have such a strongly normative side (What should
we do?) that they do not lend
themselves to the same "rational" treatment. That something
has gone awry with
this conception of rational practice becomes
clear, however, once one considers that theoretical and
practical discourse are ideal types that cannot be practiced
in pure form, except perhaps in some limiting cases of
"pure" science for which no application is on the
horizon. More usually, we cannot answer questions of "fact"
and "value" separately. Within a context of application,
what we consider a relevant "fact" is not independent
from what we think ought to count as relevant
fact; and what we consider an adequate "value"
is not independent from what we know or believe to know. There
is not only a close parallel but an inextricable interdependency
between theoretical and practical discourse (Table 5).
Table 5: Theoretical
and practical discourse
from Habermas, 1973c, p. 243 and 2009, Vol. 2,
248, cf. Ulrich, 1983, p. 139)
propositional content of statements)
or instrumental efficacy
causes of events,
motives of actions
to "norms" or "reasons":
standards of evaluation
to nomological hypotheses or statistical regularities
to moral principles, human rights, or other basic
standards of evaluation
observations regarding cause-effect relations
observations regarding needs
people, and consequences
2009 W. Ulrich
further seminal contributions of Habermas, in particular his
discourse ethics, his political philosophy with
its core ideas of the "public sphere" and of "deliberative
democracy," and his critical social theory centered
around the core concepts of "social action," "life-world"
and "system," depend on this concept of practical
In continuing our review of Habermas, I will concentrate mainly on the topic of discourse
ethics, along with brief considerations of the concepts of
deliberative democracy and "system vs. lifeworld."
way of proceeding is analogous to the way we earlier discussed
the contributions of Aristotle and Kant, namely, with a clear
focus on their contribution to ethics.
and appreciation To
some readers who are used to associate Habermas with Marxism
and "grand" social theory, it may have come as
a surprise that an introduction to
Habermas' practical philosophy should focus so much
on his theory of argumentation. Such a focus is obviously
of personal judgment and to some extent arbitrary; but more
importantly, it corresponds to the aim of this series of
reflections on reflective practice. I believe that a theory of substantial argumentation
is indeed key to a practical philosophy that is to help
us promote reflective professional practice. It is equally
important to Habermas theoretical aim, of developing the
"communicative turn" that
he has pioneered, along with a few other key contributors, in
contemporary philosophy and in the humanities.13)
there a way to summarize, in three or four sentences, the core
ideas of the new methodological
foundation that Habermas proposes for practical philosophy? I am not sure
– it may mean oversimplification – but it seems to me the
central concern of the "communicative turn" of practical
philosophy is as simple
to understand as it is powerful:
- It is only as social beings, through communication
and cooperation with others, that we can deal reasonably with
the inevitable limitations of our human condition, and with
these limitations shape our individual experiences and frameworks.
- Hence, with a view to improving the human condition, the place
to look for untapped rationality potentials – as well as for
sources of deception to be avoided – lies in the communicative
conditions that we create in this world of ours, our social
world, rather than
(as previous generations of philosophers assumed) in the ontological
constitution of the natural world, including our own biological
constitution (naturalism), or in the psychological constitution
of the human mind (mentalism), or in a transcendental-logical
conception of reason (transcendentalism).
practical philosophy needs to be grounded in an effort
to elucidate the communicative conditions that are conducive
to "rational" practice; basic to this task
are, in particular, a grounding in language theory
and, building on it, argumentation theory.
live up to the task, language theory needs to be developed
into a pragmatic theory of communicative competence,
and argumentation theory into a pragmatic theory of argumentative
cogency. Formal pragmatics is the framework that
Habermas proposes to this end; in it he sees the methodological
foundation not only for an overarching social theory
but also for the practical vision of promoting discursive
practices in all domains of society, and thus for the
communicative rationalization of society.
appreciation (1): the argumentative turn
has recently celebrated his 80th birthday. Still, it is probably
too early to assess what will ultimately remain of his work.
I would not be surprised though, if posterity will remember
him in the first place as one of the great argumentation theorists
of our epoch, along with or even prior to some of his many other
outstanding contributions, among which I would certainly count
his contribution to the revival of ethics as a subject of academic
discussion; his relentless defense of enlightenment ideas against
their postmodern "destruction"; or the model he has
provided through his work as to how we may overcome the
gap between the "two cultures" of the empirical sciences
and the humanities, just to mention a few examples.
implications of his work that interest me most at present concern its
methodological potential for the pursuit of rational professional
practice. I suspect it centers around what I am tempted to call the argumentative turn of our notion
of "sound" professional practice, towards a more open
and participatory, less elitist and expertise-driven, concept
of professional competence. Our concept of what
constitutes cogent argumentation, we have learned through the
work of Habermas, is the crux of all matters communicative,
scientific, moral, and political. To put it differently:
without a clear understanding of what mutual understanding
means and how we achieve it, we cannot hope to be competent
speakers, to communicate successfully, and to discourse and
act rationally. Argumentation under
fair conditions is the concept that replaces Kant's abstract
notion of the "court of reason," and which unfolds
into the participatory motto: Let arguments decide, not authority!
appreciation (2): 'enlarging' our thinking
A critical appreciation of Habermas'
work must wait for the end of the third and last part of this discussion.
At this point, I have only one major concern that I would like
to share, concerning the important but (I feel) still somewhat
unclear role of "bridge principles" in the Toulmin-Habermas
model of argumentation. My impression is that Habermas
burdens such bridge principles – in particular, the Kantian
principle of universalization – with a methodological role that
is still too weighty. Whether we like it or not, universalization
is an ideal; and ideals have this nasty tendency of resisting
reality. The attempt to relieve
the burden with the option of metalevel discourse looks rather theoretical to me, in the sense
that it risks putting ordinary discourse participants in a situation
of incomprehension and incompetence. After all, discourse (particularly
practical discourse) is to provide an argumentative opportunity
to all of us, not just to philosophers and academics. In addition,
my work on critical heuristics and boundary critique suggests
to me that an essential self-reflective dimension of discourse is
not well captured with Habermas' major focus on "metalinguistic," along
with "metatheoretical" and "metaethical,"
reflection. I believe
there are other, equally meaningful yet much more down-to-earth
ways to mobilize the idea of self-reflection. Without meaning to question
the need for metalinguistic discourse as such, I think we need
to enlarge our notion of what self-reflective discursive practice
is all about ... in practice!
suggest it is about the self-limitation of the validity
claim of discourse itself! To explain what I mean, we can go
back once again to what we have learned from Toulmin and Habermas, namely,
that the unity of argumentative
logic (the field-invariant "layout" of arguments,
cf. Figures 1 and 2) goes hand in hand with varying contexts of meaning and action
that shape the propositional, normative, and subjective contents
of our arguments (cf. Table 2). What the bridge principles in
question need to achieve, then, is (in Kantian terms rather than those
of Habermas or Toulmin) that they should guide us in "enlarging" our
thought beyond the subjective contexts of meaning and action
in which we always find ourselves, even in the most rationally
motivated discourse, towards perspectives that are less narrowly dependent
on our current individual views and needs. Earlier in this series
we have encountered Kant's beautiful formulation of much the
same idea in his Critique of Judgment:
the sensus communis [i.e., well-understood common sense]
we must include the idea of a sense common to all, that is,
ability of reflection that considers the ways all other humans may
think, in an effort to compare one's own judgment to the collective
reason of humanity, as it were, and thus to avoid the trap [orig.:
of allowing one's private conditions of thought, which one might
easily mistake for objective, to inform [orig.: affect in a harmful
way] one's judgment.…
The following maxims
of common human reasoning … may serve to elucidate the basic propositions
I associate with well-understood common sense]. They are: (1) to
think for oneself; (2) to think [as if one found oneself] in the
place of everyone else; and (3) to always think consistently with
oneself. The first is the maxim of unprejudiced thought;
the second of enlarged thought; the third of consequent
thought. (Kant 1793, B157f, my simplified transl.; similar formulations can
be found in Kant, 1798, § 43, and 1800, end of Sec. VII)
Kant, then, "enlarging" our thinking properly means
to unfold common sense into
community sense (cf. Kant 1793, B157f; discussed
in Ulrich, 2009b, p. 10).
If we apply this thought to our understanding of bridge principles,
we find that adequate bridge principles will help us to "enlarge" the
contexts that shape our notions of relevant facts and norms,
so that we may recognize their limitations and can reconsider them systematically
with others. Since in substantial argumentation we cannot avoid
relying on some bridge principles to take the inductive
steps from D to C; and since, at the same time, we cannot assume
that such principles ever represent indubitable guides to universalization,
it seems to me we need another, self-limiting kind
of metadiscourse, the focus of which would lie on the limitations
of any principle of "enlargement" assumed in
an argument, rather than on an attempt at universalization strictly
this way, it seems to me, we might ease the burden that our
bridge principles (whatever they are) need to carry, namely,
by taking what I call the critical turn of our concept of rationality,
or simply put: by a deliberate self-limitation of what we expect from rational
discourse, and a consequent focus on the idea of reflective
practice. Thus understood, discourse will be a valuable means
of reflective practice, rather than superseding it with yet
another version of supposedly superior rationality. We must
never allow the motto: let arguments decide, not authority!
to put people once again in a situation of incompetence. A theoretically
satisfactory conception of rational discourse is at risk of
doing just that. But at the end of the day, it is still ordinary
people, rather than any reference to the methodological ideas
of philosophers, which have to carry the burden of responsibility
for their actions. Rational discourse and Socratic self-limitation
must somehow go hand in hand. It is with this final reflection
that I will try to continue the discussion of Habermas in a
coming Bimonthly, then with a particular focus on the
idea of discourse ethics. See you later!
It may help readers not familiar with Aristotle's understanding
of deductive logic to briefly hint at the way it is tied to
his distinction of "universal" and "particular"
propositions or assertions. As we have noted, an argument for
Aristotle is deductive if its conclusion results of necessity from its premises,
that is, there can be no question about it inasmuch as contesting it would lead us into an immediate contradiction.
This is the case whenever an argument can be shown to move from
some universal proposition (such as "all men are
mortal") to a particular one (such as "Socrates is a man,
hence he is mortal"). This yields the classical syllogistic model of deductive-logical inference
(X is an A; all A's are B's; so X is a B). Hence, Aristotle
explains, "the propositions
on which the deduction depends are universal"; for "one
cannot demonstrate anything except from its own principles"
(1984b, I.8, 75b21f and I.9, 75b37). By contrast, when
the conclusion results not necessarily but only possibly, Aristotle
speaks of a dialectic argument. Such an argument leads to questions
and debate about the right kind of conclusion, or differently
put, about the right principles to be applied. It is of an inductive
rather than deductive kind, in that it works the other way round;
it attempts to infer universal from particular propositions (e.g.,
scientific theories, or basic principles of science and ethics).
This latter form of argumentation was already used by Socrates
and is central to Aristotle in the Posterior Analytics,
an early kind of "theory of science" (Aristotle, 1994b),
as well as in the Nicomachean Ethics, his theory of the
good and virtuous life (Aristotle, 1985). As we will see,
it is essential for establishing the "warrants" (scientific
or ethical principles) that make conclusive argumentation possible
beyond the reach of merely analytic reasoning or, with Aristotle,
"perfect" deduction (deductive-logical demonstration
in the narrower of Aristotle's two understandings of deduction). [BACK]
To give a
simple example, if the two propositions "p" (it rains)
and "q" (the road is wet) are both true, then the
proposition "p implies q" is equally
true whereas "p rules out q" is false. Note that
whether the new proposition is true or not depends solely on
the truth values of the original sentences along with the logical operation applied
to them; it does not depend on the content (meaning)
of the original sentences. For example, if the meaning of
"q" changes to "carbon dioxide is heavier
than air" (true), "p implies q" is still
true and "p rules out q" is still false (example taken
from Bochenski and Menne, 1965, p. 28; I.M. Bochenski was
in the late 1960s my logic teacher at the University
of Fribourg). Clearly, then,
syntactic well-formedness does not secure semantic meaningfulness,
much less pragmatic validity, without further ado. That is,
a meaningful and practically relevant logic of argumentation cannot be reduced to a
logic of syllogistic
inference. As I am tempted to say, using Aristotle's term: "perfection"
does not supersede relevance, in logic as little as elsewhere.
As trivial as it may look, this insight had been
all but lost in the development of the theory of argumentation
from Aristotle's original conception of logic to the modern
calculus – until Stephen E. Toulmin (2003, orig. 1958) published
his seminal book on The Uses of Argument. [BACK]
Perhaps a reason why Toulmin does not mention Kant is
that the judicial metaphor has long since become part of our
everyday vocabulary of argumentation, no less than the propriety
or building metaphor: when we argue, we not only "claim"
to have "solid" reasons and "grounds" and
then try to "support" these with firm "backings";
we also talk about the sort of "case" we "present"
in defense of our claims and about the "procedures"
by which we try to convince the "parties." Even so,
I find it useful to associate the judicial analogy with Kant's
critical philosophy. Doing so reminds us that any relevant logic
of argumentation ultimately "ties up with the business
of rational criticism." Toulmin (2003, p. 6) [BACK]
The issue is essential, though, when it comes to promoting
reflective practice. As a preliminary reflection on this
issue, my work on critical systems heuristics (CSH) and critical
pragmatism (cf., e.g., Ulrich, 1983, 1987, 1996, 2000a, 2002,
2006a, b) suggests to me it is upon the discourse participants
themselves, whoever they are, rather than any prior "field-dependent"
(i.e., disciplinary or institutional) conventions,
to reach some mutual understanding as to what in a specific
situation are the relevant contexts of argumentation to
be considered. This is so because, in the terms of CSH, the
definition of relevant contexts is a normative-practical issue
of boundary critique by all those affected or concerned,
rather than one of theoretical-empirical justification by the
"experts" (professionals) and decision-makers involved
in a situation. The role of boundary critique will be in the
center of the final essay of this series. [BACK]
reader may have observed that Toulmin's basic scheme is superficially
similar to Hempel and Oppenheim's (1948) scheme of syllogistic
explanation in science, where C = explanandum (description of the empirical
phenomenon to be explained), D = initial or antecedent conditions (minor premise), W = general
laws or nomological hypotheses (major premise),
and B = empirical basis for W or basic statements; W and D together
are also called the explanans. But of course, the essential
difference consists in the fact that in Toulmin's scheme,
the step from B to W is no longer a merely analytic one; which
is to say, from the perspective of the Hempel-Oppenheim scheme,
it raises Hume's problem of induction – a problem we'll discuss
in a moment. Note that if the Hempel-Oppenheim scheme is indeed
to serve as a model of scientific explanation, then the problem
of induction is bound to come up again; symptomatically, with
its reference to "general laws," the model glosses
over the fact that it does indeed presuppose the validity of
some prior inductive inference from particular observations
to nomological hypotheses or "laws." That is
to say, the Hempel-Oppenheim model does not solve but merely
avoid the problem of a logic of substantial (or "inductive"),
rather than merely analytic (or "deductive"), argumentation.
– Similarly, Popper's (1959) attempt to avoid the need
for substantial argument by using deductive logic merely as
the "organon of criticism," is bound to avoid rather
than solve the problem. Popper's model is logically based on
the modus tollens (modus tollendo tollens) of classical logic,
according to which "the falsification of a conclusion
entails the falsification of the system from which it is derived"
(Popper, 1959, par. 18). Thus, if a statement
p says that A ("it rains") implies B ("the
street is wet") and we have ¬B [not B, the street is dry],
then ¬A [not A, it doesn't rain] should hold true. If
A still holds, p is "falsified." In Popper's
famous example: A=swan and B=white. While unproblematic as a
tool of analytic reasoning, Popper's attempt to use this scheme
substantial reasoning – more accurately, as the only
rational form of critical substantial argumentation – amounts
to a narrowing down of the concept of rational criticism
to the uncovering of logical inconsistencies, at the price of
excluding from the realm of rational criticism any considerations
of substantial inadequacy, e.g., regarding a claim's semantic
context of meaning and its pragmatic context of relevance. For
more detailed discussions of Popper's narrow concept of criticism
with a view to reflective research and professional practice,
see Ulrich, 2006c and 2008). [BACK]
The concept of the "universal audience" (or "ideal
audience") was coined by the Polish-Belgian philosopher
of law Chaim Perelman, who in cooperation with Lucie Olbrechts-Tyteca
attempted to extend classical rhetoric to an (informal) logic
of value judgments. The auditoire universel comprises
"all men who are rational and competent with respect to
the issues that are being debated" (Perelman, 1968, p. 21,
quoted in Alexy, 1978, p. 206). That is, it is the largest
possible audience which has an interest to hear and to agree.
Consequently, the value of an argument is to be measured by
the audience that it convinces, or in other words, by the
extent to which it convinces a particular rather than a universal
audience. A convincing, as distinguished from a merely
persuading, argument is "one whose premises are unversalizable,
that is, acceptable in principle to all members of the universal
audience" (Perelman and Olbrechts-Tyteca, 1969, p. 35). [BACK]
We may understand the fundamental importance
of the universalization principle in even more basic terms, without
presupposing (with Habermas) the language-pragmatic and discourse-theoretical
turn in the first place; namely, by relating it to Kant's general
principle of reason (which,
as its name suggests, applies to both theoretical as well as
practical reason). According to this principle, it is reason's
intrinsic necessity to always look for the general, that is,
for completeness on the side of the conditions on which its
conclusions depend (cf. Ulrich, 1983, p.
219f; Kant, 1787, B364). In simpler, less Kantian and more pragmatic
terms, a "reasonable"
argument must consider all the circumstances that may have a
bearing on the conclusion in question, now and in future. This explains why the Kantian
principle of generalization (or universalization) is indeed
fundamental to theoretical-empirical
as well as practical-normative reasoning, before and beyond the
language-pragmatic and discursive turn. [BACK]
Among these other key contributors I should mention his long-time
colleague and friend Karl Otto Apel (e.g.,
1967-70, 1972, 1981), to whose influence and importance
I have not even tried to do justice in this article.
I have been similarly selective with regard to some key
concepts of Habermas that have played an important role
in the development of his thought but are no longer so central
to him today. This concerns, for example, his "consensus
theory of truth," his work on "technology and
science as ideology" (1971b), and his abandoned early focus
on "knowledge-constitutive interests" (Habermas,
1971a). As explained, I have preferred instead to concentrate on a
ideas that I find of fundamental methodological interest
not only for Habermas' work but equally for our own current
Although empirical and contextual
considerations have recently gained more weight in Habermas' conception
of discursive rationality (see particularly Habermas, 2004),
as far as I can see his reading of bridge principles
still tends to be more strictly universal than what I consider
feasible for practical purposes. It seems
to me that any conception of "enlarged"
thought (whatever bridge principles it may imply)
entails a quest for comprehensiveness in our knowledge of relevant
circumstances and understanding of normative issues that is epistemologically
as unfeasible as it is unavoidable. The philosophical dilemma
we encounter here is the unresolved problem of
holism. An alarm bell is ringing: we
must not allow the talk of "bridge principles" to
deflect our attention away from the precarious nature of any
holistic claims. There is, symptomatically, no natural end to "universalization,"
"discourse," and so on; or in more technical terms:
any stopping rule that might end the quest for comprehensiveness
is arbitrary. In my work on critical
heuristics, I have therefore found it necessary to limit the burden that
any conceivable bridge principle can carry. I try to
achieve this by employing bridge principles – or as I prefer
to say, methodological guidelines or principles for "enlarged"
thought, including Kant's universalization principle but also, for example, Peirce's pragmatic
maxim and Singer's (1959) and Churchman's (1982) "sweep-in"
principle – in systematic combination with a counterprinciple that
I call the principle of boundary critique (for an introductory
discussion, see Ulrich, 2001, pp. 11-15 and 23f). It
will be in the center of my attempt, in the final essay of this series,
to sketch the outlines of a "philosophy in practice"
rather than of practice, that is, a practical philosophy
properly speaking. [BACK]
(for both Parts 6a/7 and 6b/7)
Adorno, T. W.,
Dahrendorf, R., Habermas, J., Pilot, H., & Popper, K. R. (eds.) (1976). The
Dispute in German Sociology. London: Heinemann, and New York: Harper.
R. (1978). Theorie der juristischen Argumentation: Die
Theorie des rationalen Diskurses als Theorie der juristischen
Begründung. Frankfurt am Main, Germany: Suhrkamp (English
transl.: A Theory of Legal Argumentation: The Theory of Rational
Discourse as Theory of Legal Justification,
Oxford, UK: Clarendon Press, 1989).
Apel, K.O. (1967-70). Introductions to
the German translation of C.S. Peirce's writings on pragmatism, in K.O. Apel:
Charles S. Peirce, Schriften, 2 vols., Frankfurt am Main,
Germany: Suhrkamp, Vol. 1, 1967, and Vol. 2, 1970. Subsequently published together
as Apel (1981).
K.O. (1972). The a priori of communication and the foundation of
the humanities. Man and World, 5,
pp. 3-37. Partly reprinted in F.R. Dallmayr and T.A. McCarthy, eds., Understanding and Social Inquiry,
University of Notre Dame Press, Notre Dame, Indianapolis, and London, UK, pp.
Apel, K.O. (1981). Charles S.
Peirce: From Pragmatism to Pragmaticism. Amherst, MA: University
of Massachusetts Press (German orig.: Der Denkweg von
Charles S. Peirce, Eine Einführung in den amerikanischen Pragmatismus,
Frankfurt am Main, Germany: Suhrkamp, 1975).
(1984a). Prior Analytics. In The Complete Works of Aristotle: The Revised
Oxford Translation, ed. by J. Barnes, Vol. 1, Princeton, NJ:
Princeton University Press, pp. 39-113.
(References are to the Book and Chapter according to Aristotle's
numbering [e.g., I.1 = Book 1, Chapter 1], followed by the standard
form of references to the classic ed. of the Greek text by I.
Bekker, Aristotelis Opera, Berlin, 1831 [e.g., 24a13f
= page 24, column a, lines 3-4].)
(1984b). Posterior Analytics. In The Complete Works of Aristotle: The Revised
Oxford Translation, ed. by J. Barnes, Vol. 1, Princeton, NJ:
Princeton University Press, pp. 114-166. (References are to
the classic Bekker edition of the Greek text, as explained for
(1985). Nicomachean Ethics. Translated, with Introduction,
Notes, and Glossary, by Terence Irwin. Indianapolis, IN: Hackett.
(References are to the Book and Chapter according to Aristotle's
numbering [e.g., I.1 = Book 1, Chapter 1], followed by the standard
form of references to the classic ed. of the Greek text by I.
Bekker, Aristotelis Opera, Berlin, 1831 [e.g., 1094a1-3
= page 1094, column a, lines 1-3].)
J.L. (1962). How To Do Things With Words. The William James Lectures,
delivered in Harvard University in 1955. Oxford, UK: Clarendon
Press/ Oxford University Press. (2nd ed. 1976).
Bateson, G. (1972). Steps to an Ecology of Mind.
Nasville, TN: Chandler. Pb. ed, New York: Ballantine Books,
1972. (New ed. Chicago, IL: University of Chicago Press, 2000.)
G. (1979). Mind and Nature: A Necessary Unity. New York:
E.P. Dutton. Reprint ed., Cresskill, NJ.: Hampton Press, 2002.
I.M. and Menne, A. (1965). Grundriss der Logistik, 3rd
ed. Paderborn, Germany: Ferdinand Schöningh (French orig. 1949,
1st German ed. 1954).
B.R. (1979). On the foundations of rationality: Toulmin, Habermas,
and the a priori of reason. Journal of the American
Forensic Association, 16, Fall Issue, pp. 112-126.
N. (1965). Aspects of the Theory of Syntax. Cambridge, MA:
C.W. (1982). An appreciation of Edgar Arthur Singer, Jr. In
C.W. Churchman, Thought and Wisdom, Seaside, CA: Intersystems
Publications, pp. 116-135.
P. (ed.) (1992). Autonomy & Solidarity: Interviews with Jurgen
Habermas. London: Verso, 2nd ed. (orig. 1986).
Dewey, J. (1910). How We Think.
MA: D.C. Heath & Co. Facsimile ed., Champaign, IL: Standard
J. (1971a). Knowledge and Human Interests. Boston, MA:
Beacon Press, paperback ed. 1972.
J. (1971b). Toward a Rational Society: Student Protest, Science,
and Politics. Boston, MA: Beacon Press.
J. (1971c). Vorbereitende Bemerkungen zu einer Theorie der kommunikativen
Kompetenz. In J. Habermas and N. Luhmann, Theorie der Gesellschaft
oder Sozialtechnologie – Was leistet die Systemforschung? Frankfurt am Main, Germany: Suhrkamp,
J. (1973a). Theory and Practice. Boston, MA: Beacon Press.
J. (1973b). "Nachwort (1973)," in J. Habermas, Erkenntnis und Interesse,
2nd. ed., Frankfurt am Main, Germany: Suhrkamp, pp. 367-416 (Postscript
to Habermas, 1972, not contained in that earlier English edition
but published as "A postscript to Knowledge and Human Interests"
in Philosophy of the Social Sciences, 3, No. 1,
1973, pp. 175-189).
J. (1973c). Wahrheitstheorien. In H. Fahrenbach (ed.), Wirklichkeit
und Reflexion, Walter Schulz zum 60. Geburtstag, Pfullingen,
Germany: Neske, pp. 211-265. Reprinted, with additional footnotes, in
J. Habermas, Vorstudien und Ergänzungen zur Theorie
des kommunikativen Handelns, Frankfurt am Main: Suhrkamp,
1984, pp. 127-183; and in J. Habermas, Philosophische
Texte, Vol. 2, Frankfurt am Main: Suhrkamp, 2009, pp. 208-269.
J. (1975). Legitimation Crisis. Boston, MA: Beacon Press.
J. (1979a). What is universal pragmatics? In J. Habermas, Communication
and the Evolution of Society, Boston, MA: Beacon Press, pp.
1-68 (text) and 208-219 (notes).
J. (1979b). Moral development and ego identity. In J. Habermas, Communication
and the Evolution of Society, Boston, MA: Beacon Press, pp.
69-94 (text) and 219-222 (notes).
J. (1979c). Historical materialism and the development of normative
structures. In J. Habermas, Communication
and the Evolution of Society, Boston, MA: Beacon Press, pp.
95-129 (text) and 225-230 (notes).
J. (1984). The Theory of Communicative Action. Vol. 1: Reason
and the Rationalization of Society. Cambridge, UK: Polity Press.
J. (1985) Dialektik der Rationalisierung, in J. Habermas, Die
Neue Unübersichtlichkeit, Frankfurt am Main, Germany: Suhrkamp,
Orig. published as "Dialektik der Rationalisierung: Jürgen
Habermas im Gespräch mit Axel Honneth, Eberhard Knödler-Bonte und
Arno Widmann" in Aesthetik und Kommunikation, No. 45/46,
1981. Engl. transl.: The dialectics of rationalization: an
interview with Jurgen Habermas, Telos, 49, Fall Issue, 1981;
reprinted in Dew (1992), pp. 95-130.
J. (1984). The Theory of Communicative Action. Vol. 1: Reason
and the Rationalization of Society. Cambridge, UK: Polity Press.
J. (1987). The Theory of Communicative Action. Vol. 2: The
Critique of Functionalist Reason. Cambridge, UK: Polity Press.
J. (1993). Justification and Application: Remarks on Discourse
Ethics. Cambridge, UK: Polity Press.
J. (1996a). Between Facts and Norms: Contributions to a Discourse
Theory of Law and Democracy. Cambridge, UK: Polity Press.
J. (1996b). Modernity: an unfinished project. In M. Passerin d'Entrèves
and S. Benhabib (eds.), Habermas and the Unfinished Project of
Modernity: Critical Essays on the "The Philosophical Discourse
of Modernity." MA: MIT Press, pp. 38-58.
J. (1998). The
Inclusion of the Other: Studies in Political Theory. Cambridge,
MA: MIT Press.
J. (2004). Wahrheit und Rechtfertigung: Philosophische Aufsätze.
Frankfurt am Main, Germany: Suhrkamp; enlarged paperback ed.
(orig. 1999). (Note: the English translation Truth and Justification
[Boston, MA: MIT Press, 2003], due to its being based on the
original 1999 ed., does not contain the two important chapters
2 and 5 of the enlarged German ed., which is why I use
the newer German ed.).
J. (2009). Philosophische Texte. 5 vols. Frankfurt am Main, Germany:
C.G., and Oppenheim, P. (1948). Studies in the logic of explanation.
Philosophy of Science, 15, No. 2, pp. 135-175.
M., and Adorno, T.
(2002). Dialectic of Enlightenment:
Philosophical Fragments. New
transl. by E. Jephcott. Stanford, CA: Stanford University Press, 2002.
orig.: Dialetik der Aufklärung, Amsterdam: Querido, 1947.
D. (1978). A Treatise of Human Nature. Second ed., with
text revised and variant readings by P.H. Nidditch, ed. by L.A.
Selby-Bigge. Oxford, UK: Oxford University Press (1st Oxford
ed. 1888; orig. 1739-40).
I. (1786). Groundwork of the Metaphysic of Morals. 2nd ed.
[B] (1st ed. [A] 1785). Transl. by H.J. Paton. New York:
Harper Torchbooks, 1964. German orig.: Grundlegung zur Metaphysik
der Sitten, 1st ed. [A] and 2nd ed. [B], in: W. Weischedel
(ed.), Werkausgabe Vol. VII, Kritik der praktischen Vernunft,
Grundlegung zur Metaphysik der Sitten, Frankfurt am Main, Germany:
Suhrkamp, 1977 (orig. 1968), pp. 9-102.
I. (1787). Critique of Pure Reason. 2nd ed. [B] (1st ed.
[A] 1781). Transl. by N.K. Smith. New York: St. Martin's Press,
1965 (orig. Macmillan, New York, 1929). German orig.: Kritik der
reinen Vernunft, 1st ed. [A] 1781, 2nd ed. [B] 1787, in: W. Weischedel
(ed.), Werkausgabe Vols. III and IV, Frankfurt am Main, Germany:
Suhrkamp 1977 (orig. 1968).
I. (1788). Critique of Practical Reason and Other Writings in Moral Philosophy,. Transl. by L.W.
Beck. Chicago, IL: University of Chicago Press, 1949. German orig.:
Kritik der praktischen Vernunft, 1st ed. [A] in: W. Weischedel
(ed.), Werkausgabe Vol. VII, Kritik der praktischen Vernunft,
Grundlegung zur Metaphysik der Sitten, Frankfurt am Main, Germany:
Suhrkamp, 1977, pp. 105-320.
I. (1793). Critique of Judgment. 2nd ed. [B] (1st ed. [A] 1790).
Transl. by T.H. Bernard. New York: Hafner, 1951. German orig.: Kritik
der Urteilskraft, in: W. Weischedel (ed.), Werkausgabe Vol.
X, Kritik der Urteilskraft, Frankfurt am Main, Germany: Suhrkamp,
L. (1968). Moral development. In D.L. Sills (ed.), International
Encyclopedia of the Social Sciences, Vol. 10, New York:
Macmillan and Free Press, pp. 483-494.
L. (1976). Moral stages and moralization: the cognitive-developmental
approach. In T. Lickona (ed.), Moral Development and Behavior;
Theory, Research, and Social Issues, New York: Holt, Rinehart
& Winston, pp. 31-53.
L. (1981). Essays on Moral Development, Vol. 1: The Philosophy
of Moral Development – Moral Stages and the Idea of Justice. San Francisco: Harper &
L. (1984). Essays on Moral Development, Vol. 2: The Psychology
of Moral Development – The Nature and Validity of Moral Stages. San Francisco: Harper &
H. (1964). One-Dimensional Man: Studies in the Ideology of Advanced
Industrial Society. Boston, MA: Beacon Press.
(1913). The social self. Journal of Philosophy,
Psychology and Scientific Methods, 10, No. 14, pp. 374- 380.
G.H. (1925). The genesis of the self and social control. International
Journal of Ethics, 35, No. 3, pp. 251-277.
G.H. (1934). Mind, Self, and Society from the Standpoint of a
Social Behaviorist. Ed. by C.W. Morris. Chicago, IL: University
of Chicago Press.
C.S. (1878). How to make
our ideas clear. Popular Science Monthly, 12, January, pp. 286-302. Reprinted
in C. Hartshorneand P. Weiss (eds.), Collected Papers of Charles Sanders Peirce, Vol. V: Pragmatism and Pragmaticism, Cambridge,
MA: Harvard University
Press, 1934, para. 5.388-5.410.
C. (1968). Rhetoric and philosophy. Philosophy and Rhetoric,
1, No. 1,
C., and Olbrechts-Tyteca, L. (1969). The New Rhetoric: A
Treatise on Argumentation. Notre Dame, IN: University of
Notre Dame Press (French orig.: La nouvelle rhétorique: Traité
de l'argumentation, 2 vols., Paris, 1958).
(1932). The Moral Judgment of the Child. London: Routledge
& Kegan Paul (also New York: Free Press, 1965).
Piaget, J. (1970).
Genetic Epistemology. New York: Columbia University Press, 1970.
Popper, K.R. (1959). The Logic of Scientific Discovery.
London: Hutchinson. New ed. London: Routledge 2002 (German orig. 1935).
Popper, K. R. (1962). Die Logik der Sozialwissenschaften.
Kölner Zeitschrift für Soziologie und
No. 2, pp.
Reprinted version in T.W. Adorno, R. Dahrendorf, H. Pilot, A. Albert,
J. Habermas, & K.R. Popper (eds.), Der Positivismus-
der deutschen Soziologie, Darmstadt, Germany: Luchterhand, 1969, pp. 103-123.
Engl. transl. in Adorno et al. (1976), pp. 87-104.
K.R. (1968). Epistemology without a knowing subject. In B. van Rootselaar and J.F. Staal (eds.), Proceedings of the
Third International Congress for Logic, Methodology and Philosophy
of Science, 25 Aug. to 2 Sept. 1967, Amsterdam, pp. 333–373;
reprinted in Popper (1972), pp. 106-152.
K.R. (1972). Objective Knowledge: An Evolutionary Approach.
Oxford, UK: Clarendon Press.
Searle, J.R. (1969).
Speech Acts: An Essay in the Philosophy of Language. Cambridge,
UK: Cambridge University Press, 1969.
Singer, E.A. (1959).
Experience and Reflection. Ed. by C.W. Churchman. Philadelphia,
PA: University of Pennsylvania Press.
Smith, R. (2009)
Aristotle's Logic. In E.N. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Spring 2009 Edition).
First published March 18, 2000; last modified Jan. 30, 2009.
S.E. (1972). Human Understanding, Vol. 1: The Collective Use and Evolution of Concepts.
Princeton, NJ: Princeton University Press, and Oxford, UK: Clarendon.
S.E. (2003). The Uses of Argument. Updated edition. Cambridge, UK: Cambridge
University Press (orig. 1958).
S., Rieke, R., and Janik, A. (1984). An Introduction to Reasoning.
Second edition. London: Collier, and New York: Macmillan (orig. 1979).
W. (1983). Critical Heuristics of Social Planning: A New Approach
to Practical Philosophy. Bern, Switzerland: Haupt. Pb. reprint ed. New
York: Wiley, 1994.
W. (1987). Critical heuristics of social systems design. European
Journal of Operational Research, 31, No. 3, pp. 276-283.
W. (1996). A
Primer to Critical Systems Heuristics for Action Researchers. Hull, UK:
Centre for Systems
Studies, University of Hull, 31 March 1996.
W. (2000a). Reflective practice in the civil society:
the contribution of critically systemic thinking.
No. 2, pp. 247-268.
W. (200b). Critically Systemic Discourse, Emancipation and
the Public Sphere. Lincoln, UK: Faculty of Business and
Management Working Paper No. 42, October 2000.
W. (2001). The quest for competence in systemic research and practice.
Systems Research and Behavioral Science, 18, No. 1, pp. 3-28.
W. (2002). Boundary
critique. In H.G. Daellenbach and R.L. Flood (eds.),
Informed Student Guide to Management Science, London:
Thomson, 2002, pp. 41-42. Rev, version: A mini-primer of boundary
critique, Werner Ulrich's Home Page, Section Critical Systems
Heuristics (CSH), 17 Oct 2005,
W. (2006a). The
art of boundary crossing: another introduction to boundary critique. Ulrich's
Bimonthly, May 2006.
W. (2006b). Critical
pragmatism: a new approach to professional and business ethics.
In L. Zsolnai (ed.), Interdisciplinary Yearbook of
Business Ethics, Vol. I, Oxford, UK, and Bern, Switzerland:
Peter Lang Academic Publishers, 2006, pp. 53-85.
W. (2006c). Rethinking critically reflective research practice:
beyond Popper's critical rationalism. Journal of Research Practice,
2, No. 2, article P1.
W. (2008). Reflections on reflective practice (2/7): Applied science
and expertise, or the art of testing and contesting practical claims.
Ulrich's Bimonthly, March-April 2008.
W. (2009a). Reflections on reflective practice (4/7): Philosophy
of practice and Aristotelian virtue ethics. Ulrich's
Bimonthly, January-February 2009.
W. (2009b). Reflections on reflective practice (5/7): Practical reason
and rational ethics: Kant. Ulrich's
Bimonthly, March-April 2009.
M. (1978). Economy and Society, ed. by G. Roth and C. Wittich,
2 vols., Berkeley, CA: University of California Press (expanded
version of Wirtschaft und Gesellschaft: Grundriss der verstehenden
Soziologie, 4th ed. Tübingen, Germany: J.C.B. Mohr, 1956, 1st
J.W. (1992). Perspectives on argument. In W.L. Benoit, D. Hample,
and P.J. Benoit (eds.), Readings in Argumentation: Pragmatics
and Discourse Analysis, Berlin, Germany, and New York: Foris
Publications /Walter de Gruyter, pp. 121-143. Orig. published
in J. Rhodes and S. Newell (eds.), Proceedings of the 1979 Summer
Conference on Argument, Falls Church, VA: SCA, 1980.