Vous êtes sur la page 1sur 20

I n t e r n a t i o n a l J o u r n a l o f P h i lo s o p h ic a l S t u d i e s Vo l .

1 0 ( 4 ) , 4 2 9 4 47

Brandom, Hegel and


Inferentialism
Tom Rockmore
Abstract
In the course of developing a semantics with epistemological intent, Brandom
claims that his inferentialism is Hegelian. This paper argues that, even on a
charitable reading, Brandom is an anti-Hegelian.
Keywords: Hegel; Brandom; Rorty; inferentialism; semantics

In the course of developing a semantics with epistemological intent,


Brandom claims several times, in veiled fashion in an earlier book
(Brandom, 1994), more clearly in a recent book (Brandom, 2000), that his
inferentialism is Hegelian. This suggests that his position is not only consistent with, but also improves on, Hegels. Rorty further claims that
through their Hegelianism Sellars and Brandom overcome the divide
between analytic and continental philosophy.
Let us suppose, for purposes of discussion, that we know what we mean
by continental and analytic philosophy in order to concentrate on these
two claims. In response, I will be making three points. To begin with, I
will be contending that Brandoms inferentialism is not, as he thinks,
Hegelian, but rather deeply anti-Hegelian on even a charitable reading.
Any resemblance between Brandom and Hegel is purely coincidental since
they are up to different things. Second, I will be arguing that where
Brandoms view differs from Hegels, Hegels view is often superior.
Finally, I will be contending that however interpreted, Sellars and
Brandom fail to overcome the divide between analytic and continental
philosophy. It is not that analytic philosophers are incapable of close
discussions of and contributions to continental concerns. Indeed, I will be
suggesting that McDowell gets further than Brandom into Hegels texts
and problems. The reason rather is that on at least one description analytic
philosophy and continental philosophy are pursuing discernibly different
projects, which, hence, cannot simply be substituted for one another.
Although this would be misunderstood as undermining analytic philosophy in whole or in part, it does suggest that it cannot simply look to
International Journal of Philosophical Studies
ISSN 09672559 print 14664542 online 2002 Taylor & Francis Ltd
http://www.tandf.co.uk/journals
DOI: 10.1080/0967255021016740 5

I N TE R NAT IO NA L J O U R NA L O F P H ILO S O P H ICA L ST U D IES

continental philosophy or continental concerns to reinvent itself at the


beginning of the new century.
Brandoms Inferentialism
In order to grasp the relation between Brandom and Hegel, we need to
lay out the main lines of Brandoms view. This is far from easy to do.
Although he mainly employs simple language, Brandoms writing style is
more than usually convoluted, even by elastic philosophical standards, and
often difcult to follow. He typically circles around a point more than
once, multiplying technical vocabulary. He tends to work from within
Sellars position on the apparent assumption that it is familiar to his
readers and has denitively settled a number of important questions, such
as the status of empiricism after about a century of analytic criticism,
which simply need not even be addressed. These expository shortcomings
often make it difcult to know what Brandom is claiming. In describing
his position, or at least enough of it to understand his approach to Hegel,
I will be simplifying somewhat.
As the name suggests, inferentialism features inference of all kinds and
for various purposes. As concerns knowledge, inferentialism is a second
best, potentially fallible approach, which appeals to inference if and only
if we cannot do better. Doing better would mean going beyond (the need
for) inference, which would then become superuous, unneeded, in any
case unnecessary for cognitive purposes. Descartes clearly means to go
beyond inference in his foundationalist epistemological strategy. He is
claiming not that we can infer that true and distinct ideas are true but
rather that (through deduction from a true principle, or the cogito) we
know that they are (in fact) true in telling us about the way the world is.
In virtue of their reliance on inference, inferentialist theories are typically anti-empiricist, non-realist, holist, and anti-representationalist .
Brandoms inferentialism differs from this (admittedly ideal) model
through at least a covert commitment to realism that, so far as I know,
he does not directly discuss anywhere.
His main interest lies in an inferential, non-referential approach to
semantics with an explicitly epistemological intent. In his recent book, he
develops this idea through a series of alternatives, initially by privileging
inference over reference in the order of semantic explanation (Brandom,
2000a: p. 1). Thus he rejects a Platonic strategy, or a prior understanding
of conceptual content, for a pragmatist (or functionalist) strategy in which
the use of linguistic expressions, or the functional role of intentional states,
confers conceptual content on them (Brandom, 2000a: p. 4). According
to Brandom, the implicit content of language with respect to stating, asking
for, and providing beliefs, or knowing how, can be rendered explicit in
the form of knowing that.
430

BR A N D O M , H E G EL A N D IN FE R EN T IA LI S M

Brandom links his inferentialism to a series of gures and movements,


such as analytic semantics and pragmatism, between which he tends to
detect similarities where other observers more often see differences, and
especially to Hegel. Imitating Rortys tendency to extend the term pragmatism to virtually everyone, Brandom sees analytic semantics and Hegel
as pragmatic.
Since, for Brandom, cognitive claims appeal to and hence rely on
semantic inference, in his view pragmatism is a form of semantics. His
take on semantics runs from Sellars, his hero, back to Frege, whom he
criticizes. He criticizes the mature Frege for holding that truth is based
on inference and not conversely. According to Brandom, the inversion of
the relation of truth and inference points to the priority of the propositional, which he also calls claiming. And what can play the role of premise
and conclusion of inference is a saying the sense of a claiming (Brandom,
2000a: p. 13). From Sellars, he takes the idea that grasping a concept is
mastering the use of a word (Brandom, 2000a, italics removed: p. 6).
Brandom links his account of concepts to rationalism, or asking for and
giving reasons within an inferential context, and so-called expressivism.
It is unfortunate, since he relies heavily on expressivism, that he never
claries his understanding of this concept. Sellars claims in Kantian fashion
that rules, or reasons, are causes. Brandom, who is less interested in
why or how we act as we do, is more interested in cognitive claims
with respect to thought and action. He glosses Sellars statement as
making explicit in a form that can be thought or said what is implicit
in what is done (Brandom, 2000: p. 56). His main concern seems to lie
in the standard epistemological goal of making true claims. This becomes
clearer, if not clear, in his reliance on expressivism as an alternative to
representationalism.
Representationalism is usually employed to indicate the relation of
something to something else, for instance in Platos denial that art or
poetry adequately captures the forms. In his famous letter to Herz (21
July 1772), Kant describes what later became the critical philosophy as
the analysis of the relation of representations to objects. Brandom similarly regards representationalism as a way of thinking about empirical
knowledge in which representations are caused by what is represented,
leading on to philosophy of language (reference, denotation, extension).
Inferentialism and expressivism appear to be synonymous for Brandom.
Inferentialism is expressivist, since it elicits commitments, and expressivism is inferentialist, since it gives up representationalism in favour of
making things clear through inference. Brandom applies this view in a
number of ways, for instance as concerns logic. He gives up the traditional
view of logic as proving claims in favour of an expressivist view of logic
as helping us, through a logical vocabulary, to make explicit implicit
inferential commitments.
431

I N TE R NAT IO NA L J O U R NA L O F P H ILO S O P H ICA L ST U D IES

Expressivism is further important to Brandoms reading of Hegel.


According to Brandom, Hegel takes a pragmatic approach to conceptual
norms in adopting a rationalist form of so-called Enlightenment expressivism, which Brandom sees as a potentially very promising alternative to
representationalism. For Brandom, rationalist pragmatism and rationalist
expressivism are equivalent terms (Brandom, 2000: pp. 22, 34). Brandom
explicitly attributes this idea to Hegel.
[R]ationalist expressivism understands the explicit the thinkable,
the sayable, the form something must be into count as having
been expressed in terms of its role in inference. I take Hegel to
have introduced this idea, although he takes the minimal unit of
conceptual content to be the whole holistic system of inferentially
interrelated judgeables, and so is not a propositionalist.
(Brandom, 2000: p. 35)
We can understand Brandom to be suggesting that Hegel pioneers
rationalist pragmatism and expressivism, or rationalist pragmatism
and rationalist expressivism, as opposed either to expressivism or pragmatism.
The epistemological payoff of Brandoms materialist semantics is
visible in his critical remarks about naturalism. According to Brandom,
naturalism as ordinarily understood denies the human characteristic of
appealing to conceptual norms developed in social linguistic practices.
He may have in mind naturalized epistemology as it has developed in
Quines wake. Brandoms non-naturalist inferentialism-cum-expressivism
features a distinction between things that have natures and those that
have histories. His argument includes two specic claims. His rst claim
is that the objects of science (physics, chemistry, biology) have natures
which cannot be reduced to their histories. In saying this he seems to
mean that objects with natures can be studied empirically. He also has in
mind the ordinary distinction between the Naturwissenschaften and the
Geisteswissenschaften. His second claim is that the empirical properties of
things allow us to determine whether inferences based on them about
concepts are correct.
For the properties governing the application of those concepts [for
instance physical things such as electrons and aromatic compounds]
depends on what inferences involving them are correct, that is, on
what really follows from what. And that depends on how things
really are with electrons and aromatic compounds, not just on what
judgments and inferences we endorse.
(Brandom, 2000: p. 27)
432

BR A N D O M , H E G EL A N D IN FE R EN T IA LI S M

Brandoms Pragmatism
Brandoms suggestions that his inferentialism is pragmatist and Hegelian
may strike some readers as arbitrary. After all, it is not every day that
someone who traces his position to Sellars and Frege also insists on its
relation to pragmatism and to Hegel. Some observers might want to
emphasize the differences between such thinkers, or the differences
between pragmatism and Hegel. Brandoms conviction that pragmatism
is semantic since (all) cognitive claims appeal to, and hence rely on,
semantic inference at best shows that pragmatism relies, or ought to rely,
on semantics for claims to know. It suggests that pragmatism, which some
see as epistemological, has a semantic dimension since knowledge claims
of all kinds rely on semantics. We will return to this claim below. But it
does not show that all forms of semantics, or even those forms associated
with Frege and Sellars, are pragmatist.
One thing Brandom gets right about pragmatism is its often neglected
epistemological thrust (see Rescher, 2000). Broadly speaking, classical
American pragmatism, which was founded by Peirce, follows from the
continuing concern with knowledge after the rejection of Cartesian epistemological foundationalism. In current terminology, Peirces critique
amounts to a rejection of epistemological foundationalism while simultaneously avoiding epistemological scepticism. In different ways, this concern
unites Peirces theory of meaning, James theory of truth, Deweys evolutionary approach to knowledge through warranted true belief, C. I. Lewis
interest in a relativized a priori, and so on.
The evolution of the discussion in the hands of analytic thinkers turned
pragmatist (Quine, Putnam, Rorty, Brandom) has more recently broadened the understanding of pragmatism in a way that makes it
questionable whether the term is still meaningful. When Quine claimed
that his position was pragmatic, he meant that ideas could not be matched
up with the real (Quine, 1961). Rorty extends the appelation to Putnam,
supposedly the most important current pragmatist, Nietzsche, Davidson,
and others. His two favourite pragmatists at present are Davidson and
Dewey (Rorty, 1999: p. 24). In Brandom, the term has been extended, to
the best of my knowledge for the rst time, to apply even to Frege,
someone who until now has always seemed to fall outside pragmatism,
however understood.
Brandom has almost nothing to say, except in the most general terms,
about the classical pragmatists. In identifying his position as pragmatist,
he seems to have in mind specically cognitive forms of inferential practice as opposed to theory. Unlike Marxists, Brandom has in mind
specically cognitive practices, like asserting, making a claim, or giving or
asking for reasons, as opposed to theory. But that is simply not enough
to justify the claim for pragmatism, since beginning with Socrates I cannot
433

I N TE R NAT IO NA L J O U R NA L O F P H ILO S O P H ICA L ST U D IES

think of anyone willing to engage in dialogue who refuses asking for


reasons. All the action is about whether the reasons are worth taking seriously. In any case, Brandoms commitment to the non-traditional Sellarsian
approach to the analysis of meanings seems to have nothing particular in
common with, say, classical American pragmatism, which needs to be characterized in order to know what Brandom thinks it is, and not much more
in common with such analytic neo-pragmatists as Putnam, Rorty, and
Quine. Since Brandom nally says so little specically about pragmatism,
and what he does say is not clearly related to it, at least not as ordinarily
understood, one may doubt whether the term pragmatism ought to gure
in Brandoms vocabulary.
Brandoms Quinean Hegelianism
Brandoms unfamiliar form of pragmatism is relevant to his equally unfamiliar Hegel. He claims to be a Hegelian, and teases the reader with
repeated references to a forthcoming Hegel book. Yet the little he says
about Hegel only arouses suspicion. Brandom always, or almost always,
reads Hegel through the positions of leading analytic philosophers, in practice Quine and Sellars, and never, or almost never, reads Hegel directly.
To put it another way, he develops his reading of Hegel on the basis of
analytic doctrines he claims to nd in Hegel, who, at least in Brandoms
approach, appears as an early analytic philosopher, as someone who a
couple of hundred years ago was not only a pragmatist but also presciently
anticipated a number of ideas that would later become central to the
analytic debate.
For present purposes, we can divide Brandoms nascent Hegelianism into
two phases, an earlier Quinean phase in Making It Explicit and in a series
of articles leading up to Articulating Reasons, and a more recent Sellarsian
phase in the latter book. For the purposes of this paper, I will be supposing that Quine and Sellars hold very different theories, with different
emphases, and feature different approaches often to dissimilar problems.
It is for this reason that, say, Quine looms large in standard accounts of
holism, which do not even mention Sellars. Analytic philosophy can be
described as a big tent with room for people as diverse as Moore and
Russell, two thinkers about as far from each other as it is possible to be
and still be considered as representing the same general set of views. It is,
then, signicant, not irrelevant, whether Brandom approaches Hegel with
reference to Quine or rather with reference to Sellars.
These earlier Quinean and later Sellarsian phases in Brandoms successive readings of Hegel are connected by Brandoms conviction that Hegel
inverts the traditional approach to semantic explanation in a view of
experience as inferential in which concepts emerge from the roles they
play. In other words, Brandom begins from the idea that Hegels position
434

BR A N D O M , H E G EL A N D IN FE R EN T IA LI S M

is an inferentialist form of semantics, and hence the ancestor of his own


inferentialism. In the earlier, less developed, more tentative Quinean phase
of his Hegelianism, Brandom approaches Hegel through concepts and
problems he nds in Quine. In the later, slightly more developed, less
tentative Sellarsian phase of his Hegelianism, he approaches Hegel
through concepts and problems he nds in Sellars. In both cases, he silently
assumes that the concerns of leading analytic philosophers transpose
effortlessly into Hegels position.
Since Quine and Sellars hold very different theories, it is not surprising
that Brandom presents different readings of Hegel in his successive phases.
In his Quinean phase, Brandom reads Hegel through holism. In his
Sellarsian phase, he reads Hegel through analytic philosophy of language.
A generally holist approach to Hegel is plausible in that at least since the
early Lukcs attention has been focused on the category of totality
(Lukcs, 1971). Yet what does holism mean with respect to Hegel? That
is unclear since not enough attention has so far been focused on the precise
nature of Hegelian holism. It is tempting to think more closely about what
Hegelian holism entails and what can be said for it, although that cannot
be done here. Sufce it to say that at a minimum Hegel and Quine understand holism so differently that the analogy seems strained, uninformative,
and forced. One can detect a philosophy of language in Hegel, although
the differences between Hegels anti-semantics and analytic semantics
seem more important than the similarities.
In Making It Explicit, the most important reference to Hegel, later developed in a series of papers, is the unusual suggestion that he anticipates
Quines holism (see Brandom, 1994: p. 92). This suggestion is unusual since
there has been practically no attention given to a possible link between
Hegel and Quine, and, to the best of my knowledge, none at all to their
supposed common allegiance to holism in standard discussions (Fodor and
Lepore, 1992). In a chronological sense Hegelian holism obviously precedes
Quinean holism. Yet Brandom seems not to be aware that Hegel does not
invent or discover holism, which goes back at least to ancient Greece (in,
for example, Speusippus). Brandom might be taken to mean that Hegel
anticipates Quines specic form of holism. Yet Hegels and Quines forms
of holism are irreducibly different, and hence incompatible.
In Brandoms interpretation, Quines holism concerns the relativizing
of meaning to a total theory, or a theory for which the unity of meaning
is the whole (Brandom, 1994: pp. 477481). It is well known that for Hegel
the truth is the whole. Yet it is simply implausible to equate Quines essentially asocial and ahistorical view of the change of meaning and belief, to
use Brandoms language, with Hegels very different view of the constitution of knowledge and ethics for real human individuals within the social
and historical context. Or to put the same point in another way, Quine
focuses on the epistemic conditions of knowledge in general, in virtue of
435

I N TE R NAT IO NA L J O U R NA L O F P H ILO S O P H ICA L ST U D IES

the underdetermination of theories, without regard to the underlying social


context; but Hegel focuses on their social and historical constitution in
relation to the underlying (social and conceptual) context. Hegel, who
never distinguishes cleanly between theory and context, understands theories as indexed to the historical moment, and hence as relative to time
and place, in a word as intrinsically historical. Quine, on the contrary,
takes a behaviourist but not a historical approach to knowledge. There is
not the slightest hint in Quine that he understands cognition as historical.
As early as his rst published philosophical text, Hegel rejects a priori
forms of epistemological justication, exemplied in Kants critical philosophy. In its place, he argues for the progressive justication of claims
to know through the development of theories testable with respect to
conscious experience (Hegel, 1977) and with respect to the background
of the prevailing views held at a given point in time by the relevant section
of the community (Hegel, 1987: chapter 6). An analogy might be that the
theories about the top quark are justied by the general understanding
of quarks against the background of widespread acceptance of quantum
mechanics at present. Hegel explicitly denies that cognitive claims can
somehow escape from the limitations of the historical moment (Hegel,
1967: p. 11).
Brandoms view of Hegel later evolved. In a recent article on pragmatist themes in Hegelian idealism (Brandom, 1999), he elaborates his earlier
claim. At this stage he no longer emphasizes Hegels anticipation of Quine
in arguing (in more detail than before) for an explicit parallel between
Hegel and Quine. According to Brandom, Quine works out his pragmatist view, or the adjustment of meaning and belief within experience, in
the same way that Hegel works out his idealist view (Brandom, 1999:
p. 176). Brandoms claim, which depends on a distinction between socalled pragmatic and idealist theses, suggests an awareness of differences
between, not the supposed identity of, pragmatism and idealism. Through
a distinction between the alleged pragmatist thesis that the use of concepts
determines their content and the so-called idealist thesis, that the structure and unity of the concept is the same as the structure and unity of
the self (Brandom, 1999: p. 164), Brandom makes two points: Hegels
idealist thesis is his way of making the pragmatist thesis work, and this is
useful for contemporary semantics.
There are a number of things questionable or even plainly wrong about
the supposed parallel between Hegel and Quine. Since Hegel precedes
rather than succeeds the rise of pragmatism, Brandom appears here to be
reading the history of philosophy backwards. I confess that I cannot make
any sense of the so-called idealist thesis, which apparently suggests some
unspecied (structural?) similarity between concepts and selves. It is
obvious that at a sufciently abstract level, anything resembles anything
else. Yet I see no reason to hold that a concept is like a self or a self is
436

BR A N D O M , H E G EL A N D IN FE R EN T IA LI S M

like a concept. Clearly if there are concepts, selves have them; but selves
are not and cannot be reduced to concepts. Finally, as for Kant, though
in a wholly different way, concept (Begriff) is a technical term for Hegel.
To put the point quickly, Hegel uses concept to designate the form of
knowledge possible after abandoning representation in the Kantian sense
while restricting cognition to conscious experience. For Hegel, philosophical knowledge, which relies on concepts, grasps what is as it is given
in experience through the identity of the theory of the cognitive object
and the object of the theory. Brandoms understanding of concept, which
remains unclear, includes reference to the world as it is, which the concept
is intended to grasp. Since Hegel specically denies that we ever know
the world as it is, he cannot be using concept in that sense. Hence,
Brandom gives us no reason to believe that either he or Quine understands concepts in a way even remotely similar to Hegels technical
understanding of the same term.
Brandoms Sellarsian Hegelianism
Brandoms comparison between Hegelian idealism and Quinean pragmatism presupposes a similarity based on a shared commitment to holism,
but a distinction between idealism and pragmatism. I contest the similarity between Hegel and Quine, but I nd the distinction between
pragmatism and idealism helpful. They are not the same and need to be
distinguished. In different ways, the classical American pragmatists all
react for or against Hegel and other forms of idealism. Whatever idealism
is, it is not identical with pragmatism.
This useful distinction (between idealism and pragmatism) simply
disappears in Articulating Reasons, where Brandom takes a Sellarsian
approach in directly linking his own position to Hegels. The reason for
what certainly looks like a transition from a Quinean to a Sellarsian
approach to Hegel is mysterious. There is no indication that Brandom has
meanwhile become critical of Quine, and hence critical of his own earlier,
specically Quinean way of interpreting Hegel. In the later book,
Brandoms Hegel interpretation combines a series of complex claims about
the relation of analytic and continental philosophy, Brandoms relation
to Sellars, Sellars relation to Kant, and Brandoms relation to Hegel.
Following Rortys practice, each of Brandoms claims depends on
blurring a series of distinctions.
Brandoms recent Sellarsian reading of Hegel rests on two specic claims
by Rorty. First, through their proto-Hegelianism the text says prope
(Latin for near)-Hegelianism, but this looks like a misprint Sellars and
Brandom overcome the split between analytic and continental philosophy
(Rorty, 1997: p. 11). Second, Brandom provides a Hegelian extension of
Sellars Kantian approach to thought and action (Rorty, 1997: p. 11).
437

I N TE R NAT IO NA L J O U R NA L O F P H ILO S O P H ICA L ST U D IES

In Rortys version of the story, Brandoms Hegelianism depends on


characterizing Sellars as a Kantian. Is Sellars a Kantian? It would be more
accurate to say that at different points in his career and for different
reasons Sellars is both a Kantian and a Hegelian. Certainly in the period
he was attacking the myth of the given (Sellars, 1997), he was developing
a clearly Hegelian line of argument in what he himself described as
Hegelian meditations. This is true even if his Hegelianism later wanes. It
is inexact, then, to say or to imply that he is only a Kantian, although he
is more Kantian than anything else. Yet it is only if Brandom can legitimately be called a Hegelian that his Hegelianism improves on Sellars
Kantianism. In Articulating Reasons, Brandom undertakes to lead analytic
philosophy from its Kantian to its Hegelian stage in part through a distinction he attributes to Hegel between nature and culture, Hegels supposed
pragmatism about conceptual norms, and the alleged Hegelian link
between logic and self-consciousness (Brandom, 2000a: pp. 335).
Each of these attributions is doubtful. Brandoms attribution of his own
distinction between nature and culture to Hegel echoes Rortys reading
of Sellars scientism. Sellars famously distinguishes between so-called folk
views, which are unsubstantiated and do not count as knowledge, and
science, which is our only reliable cognitive source. Rorty applies
Sellars distinction in contending that philosophy defends the further
distinction between science and non-science, or culture (Rorty, 1979: p.
268). On this model, philosophy separates cognitive disciplines, the only
acceptable source of truth and knowledge, from folk disciplines, which fall
below acceptable cognitive standards. On the RortySellars view, the
philosopher functions like a gym teacher in choosing players for opposing
cognitive teams: those apt only for folk views, which are incorrigibly
subjective and depend on local conditions; and those apt for scientic
knowledge, the objective form of cognition, who win all the conceptual
games.
Brandoms reading of Hegel conates two basic distinctions, each of
which is important for Hegel: the relation between nature and science,
and the idea of science as a cognitive enterprise. In simplest terms, Hegel
relativizes the distinction between nature and science, but privileges
scientic philosophy, or philosophy as science, over natural science as a
cognitive source. Hegel, who systematically relativizes all distinctions, does
not, in fact cannot, distinguish sharply between nature and culture. He
detects a reciprocal relation between nature and culture. Culture is situated within nature, which is a cultural phenomenon, since what we mean
by nature is itself a historical variable.
Hegels preference for philosophy as the highest form of cognition is
incompatible with the scientism endemic in analytic philosophy. An
example is Carnaps famous view that henceforth science is the sole source
of knowledge (Carnap, 1937: p. xiii). For Hegel, the natural sciences
438

BR A N D O M , H E G EL A N D IN FE R EN T IA LI S M

comprise a collection of cognitive disciplines which, since they lack a


reexive dimension, fall below the level of philosophical science. His preference for philosophy as the highest form of cognition conicts with
Sellars scientism, more specically with the preference for natural science
as the main and indeed nal source of claims to know. In other words, as
concerns knowledge a Sellarsian commitment to scientism conicts with
the specically Hegelian commitment to philosophical science.
One can further doubt that Hegel is a pragmatist about conceptual
norms in anything like Brandoms sense. Brandom points to Haugelands
idea that transcendental constitution is social institution (Haugeland, 1982,
referenced in Brandom, 2000a: p. 34). Yet Hegels position, which relies
on social institutions, is not transcendental in any ordinary way, say in
Kants sense. It is unclear what about Hegel indicates a transcendental
dimension. It is difcult to know what transcendental pragmatism means.
And unless refusing a transcendental approach equals a commitment to
pragmatism, it is unclear why Hegel should be described as a pragmatist
at all. Hegels view that conceptual norms depend on the society in which
they emerge is not specically pragmatist. It sounds more like common
sense, which even non-pragmatists exhibit.
Hegelian Criticism of Inferentialism
I have been arguing that Brandoms claim that his position is Hegelian is
implausible on even a charitable interpretation. The deep, irreconcilable
differences between Brandom and Hegel open the door to Hegelian criticism of Brandom. Brandom presents inferentialism as a semantics with
epistemological intent. His inferentialism comes down to two claims: rst,
concepts are norms determining what counts as reasons for particular
beliefs; and second, the mind-independent real makes the conceptual structures which we construct true or false by telling us how it is with the
world. The latter claim means that under appropriate conditions we can
and do cognize the mind-independent world as it is.
Brandom illustrates his view in a passage where he distinguishes physical, chemical, and biological things that have natures, and presumably are
the stuff of science, from those, like English romantic poetry, that have
histories. Brandom holds that our concepts about such objects can be
known to be either true or false:
For the properties governing the application of those concepts [that
is, electrons and aromatic compounds] depend on what inferences
involving them are correct, that is, on what really follows from what.
And that depend on how things are with electrons and aromatic
compounds, not just on what judgments and inferences we endorse.
(Brandom, 2000a: p. 27)
439

I N TE R NAT IO NA L J O U R NA L O F P H ILO S O P H ICA L ST U D IES

This passage, which clearly indicates Brandoms intent, raises more questions than its answers. Left unclear are such crucial claims as how can we
ever know which inferences are in fact correct? Do quarks exist because
we infer that they do? What does it mean for inferences about electrons
or aromatic compounds, in Brandoms language, really to follow? Does
that mean we can check them against the way the world is?
Brandom is aware of the consequences of his claim with respect to
currently popular views. He criticizes Rorty, who is annoyed by any claim
for truth, in contending that facts make claims true for they make claimings true and that in a central range of perceptual experience, the facts
are the reasons that entitle perceivers to their empirical beliefs (Brandom,
2000c, p. 162).
Brandoms epistemological intention is clear in the claim that through
concepts we know how it is with electrons and aromatic compounds, and,
more generally, through facts we know how it is with the mind-independent world. His suggestion that in appropriate conditions it just is the case
that natural science is correct about the world is a version of the traditional realist claim to know. Realism of all types, which goes back in the
Western tradition at least to Plato, can be understood as the claim to
know what is as it is, to know the way the mind-independent external
world is (Devitt, 1997). In modern philosophy, Descartes typically claims
that certain ideas about the mind-independent world must necessarily be
true. Versions of realism are still widely popular after a century of analytic
criticism of empiricism.
Rorty rejects the idea of getting it right in any form about the world
in favour of more conversation, since no amount of discussion will show
that a given belief cuts reality at the joints (Rorty, 2000, in Brandom,
2000c). His observation that there is no way to know if beliefs are true
merely because reasons can be asked for and given concerning a particular belief without matching it up with the world effectively disposes of
Brandoms inferentialist realism at the cost of endorsing epistemological
scepticism. For Rorty, either we know how the world is, in which case
discussion comes to an end, or we do not and cannot know how the world
is, and the discussion is endless and not worth conducting since we nally
cannot know anything other than that we cannot know anything.
Hegel offers a promising alternative to Brandom and Rorty. It has been
known at least since Kant that, other than by accident, there is no way
to know the mind-independent external world as it is. Brandom and Rorty,
who in this respect are pre-Kantians, still make the possibility of knowledge depend on somehow getting it right about external reality. Hegel,
who accepts Kants point, constructs a view of knowledge that makes no
pretence to knowing the way the world is, which does not gure in his
account, in conning knowledge claims to conscious experience. Unlike
Brandom, he does not claim to know, or that it is possible to know, the
440

BR A N D O M , H E G EL A N D IN FE R EN T IA LI S M

mind-independent world as it is. Unlike Rorty, he does not conclude


that failure to know the mind-independent world leads to epistemological
scepticism.
Traditional views of knowledge often turn on variations on the theme
of matching up claims about the world to the world as it is. This view,
which is strongly illustrated in Descartes, is sharply challenged later. Kants
critical philosophy turns on the insight that we simply cannot understand
how what can be called representations relate to the way the world is,
since cognition is limited to the empirical real as distinguished from things
in themselves.
Following Kant and Fichte (Fichte, 1991: p. 246), Hegel also holds that
we can know nothing about the way the world is. Knowledge cannot
consist in matching up a concept, or even a theory to mind-independent
reality, which is uncovered or discovered as it is. In the Introduction to
the Phenomenology, Hegel famously develops a theory of knowledge as
a process turning on the reciprocal interaction, and eventual correspondence of, subject and cognitive object within consciousness (Hegel, 1977:
pp. 4657). Since, like his great German idealist colleagues, he denies that
we have experience of, cognitive access to, or knowledge of things in themselves, he never claims, nor could he claim, that the mind-independent
reality makes our concepts true. At most, he claims that in principle the
knowing process would come to an end when subject and cognitive object
as is given in experience coincided within consciousness.
Though he is an epistemological sceptic, Rorty rejects Brandoms view
of knowledge for about the same reason that Kant, Fichte, and Hegel
earlier reject any claim to know the world as it is. Rorty notes that we
never know that independent reality makes our concepts true. In this
respect the difference between Hegel and Rorty lies in the very different
conclusions they draw from very similar arguments. Hegel sees the
possibility of working out a theory of knowledge which does not depend
on somehow getting it right about the world, but Rorty holds that if we
cannot get it right about the world, then knowledge is impossible and
scepticism is the only alternative. The virtue of Hegels approach lies in
limiting cognition to what is given in conscious experience while avoiding
epistemological scepticism.
On Semantics and Epistemology
To the best of my knowledge no one seriously denies the split between
analytic and continental philosophy. Over the past century, this division
between two of the three leading movements which arose around the
beginning of the twentieth century the other is pragmatism has been
fraught with mutual suspicion. This relation has more often been ignored
than studied. In a recent book, Michael Friedman suggests that the
441

I N TE R NAT IO NA L J O U R NA L O F P H ILO S O P H ICA L ST U D IES

encounter between Cassirer and Heidegger at Davos in 1929 (Friedman,


2000) was centrally important for this separation, although it seems
that the split between the representatives of both movements was effectively completed long before this encounter. It has been well said that the
rejection of idealism is one of the founding acts of analytic philosophy.
Frege sharply criticized Husserls psychologism. Russell and Moore,
who were initially attracted to idealism, or at least to McTaggart, later
sharply rejected (British) idealism for denying the existence of the external
world.
The vague character of this objection has not kept generations of
analytic philosophers from accepting it to varying degrees as a reason to
reject idealism. This objection is problematic on several grounds. It is
unclear what is meant by idealism, but at least clear that no one identied as an idealist ever denies the existence of the external world. At
most some idealists deny that we can ever have a cognition relation to it,
in a word know it as it is. This suggests a fundamental opposition between
analytic thinkers, who are committed to knowing how, in one formulation, words hook onto the world, and idealist thinkers (and an increasing
number of analytic thinkers), who deny that we can ever come into cognitive contact with the world as it is. Hence Quine, whom no one would
confuse with an idealist, suggests that we simply cannot go beyond a
behaviourist model to develop a theory of meaning (Quine, 1974).
Doubts about reference and the naissant analytic turn (or return) to
Hegel (Sellars, Rorty, Brandom, McDowell) have not so far succeeded in
overcoming the split between analytic and continental philosophy.
Brandom begins from the idea that Hegels position is an inferentialist
form of semantics, and hence the ancestor of his own inferentialism. Yet,
if the above discussion of Brandoms reading of Hegel is correct, Brandom
arguably misconstrues Hegel, who is not obviously doing anything which
falls under the heading of inferentialist semantics. And, unless one gives
him more credit than should be due to the fact that in being interested
in Hegel he is interested in a so-called continental gure, there is simply
no reason to think that Brandom somehow overcomes the split between
continental and analytic philosophy.
Brandoms failure in this regard is not unexpected. It is due to his inattention to the conditions of serious philosophical dialogue. Rorty simply
disregards the conditions under which analytic philosophy could usefully
dialogue with continental philosophy in proclaiming that through their
Hegelianism Sellars and Brandom overcome the split between the two
traditions. But do they? Could they?
If analytic and continental philosophy were both distinct philophical
traditions, their very differences would impede their being taken up into
a single overarching conceptual structure. It would seem that analytic
concerns could be effortlessly transposed into continental concerns if and
442

BR A N D O M , H E G EL A N D IN FE R EN T IA LI S M

only if the analytic and continental traditions overlapped wholly or at least


in large part.
The analytic and continental traditions overlap within the wider framework of Western philosophy, more precisely as two of the many reactions
to Kant, which, over time, have diverged enormously. One can imagine
two things that are initially close together but later grow further and
further apart. When this happens the differences may become more important than the similarities. English is now spoken in different parts of the
world in ways that often make it difcult even for a native speaker of the
language from, say, the United States to understand a native speaker of
English from India.
Brandoms inferentialism draws on both epistemology and semantics.
These traditions are different and should not be conated. Since Descartes
and above all since Kant, epistemology typically concentrates on making
out cognitive claims for truth and knowledge. Semantics since Frege has
generally concentrated on questions of meaning and reference. Brandom
is obviously within his rights in crossing this informal divide by linking
his inferentialist semantics to epistemology, reference to truth. Other
analytic gures routinely do this as well, often in order to defend realism.
For instance, in opposition to Dummetts anti-realism, which suggests
that meaning cannot reject use (Dummett, 1978: pp. 2235), John
McDowell has tried a number of times to make sense of realism as the
idea that understanding a language transcends possible verication of
any specic claims about the world (McDowell 1998, 295365; see also
ibid., 27591).
These general considerations suggest three related points which I think
are lurking in the background in recent claims by Rorty, Brandom, and
possibly others about the relation of analytic and continental philosophy:
rst, suitably understood semantics can replace or stand in for epistemology; second, Hegel is either a semanticist or at the very least his theory
has potential semantic uses; and third, the semantic elaboration of Hegel
as an inferentialist will overcome the split between analytic and continental philosophy which took up nearly the whole of the twentieth century.
My scepticism about this entire programme concerns the deep difference between semantics and epistemology as usually understood, which
in turn undermines any effort to develop a specically Hegelian semantics or to resolve epistemological concerns through semantic means. I
believe, though I do not have the space to develop my view here, that
semantics and epistemology have different agendas, different reasons for
being, different strategies, different forms of argument. Semantics is
usually understood to concern problems of reference or denotation.
Russell typically held that denotation requires a correlation between the
words in a sentence and something outside it (Russell, 1964: pp. 4852).
Epistemology is concerned with a wider theme, to which reference
443

I N TE R NAT IO NA L J O U R NA L O F P H ILO S O P H ICA L ST U D IES

arguably belongs as a subset, of how we can reliably claim to know the


world and ourselves.
The aims of semantics and epistemology are different, and should not
be conated; and neither can be reduced to the other. Semantics can no
more replace epistemology than reliabilism can replace reasons. Semantics
could replace epistemology if and only if the problem of knowledge could
be reduced to that of reference. I do not want to deny that in at least
some cases reference plays a role in knowledge. The more relevant question is whether all knowledge relies on reference and whether, if this is
the case, reference can, or at least under certain circumstances conceivably could, replace other approaches to knowledge. It seems obvious that
reference, or at least reference to the world, plays a role in knowledge
claims, such as in reporting that, say, Otto is here now. Yet this does not
seem always to be the case, for instance if one claims, to take a wellknown example, that the vericationist criterion of meaning is not itself
veriable. If that is the case, then semantics, which is not epistemology,
is at most a part of epistemology, as reference arguably belongs to the
justication of claims for truth and knowledge.
As concerns the second claim, Brandom deserves credit for perceiving
the epistemological interest of Hegels position. This is all too often overlooked in the belief, fostered by Kant, that epistemology worthy of the
name reaches a peak and an end with the critical philosophy. Yet
Brandoms effort to develop a Hegelian form of inferentialism, or a semantics with epistemological intent, enlists Hegels theory of cognition for
aims inconsistent with it.
Brandom is not alone in nding a theory of reference in Hegel. He
suggests that Hegels rationalist expressivism is rich enough to do real
semantic work (Brandom, 2000: p. 34), hence function as an alternative
to the representational semantic paradigm. It is sometimes overlooked
by those who nd a theory of reference in Hegel that Hegel is not a
semanticist, concerned with denite reference, but an anti-semanticist
(Stekeler-Weithofer, 1992; Grau, 2001). He is opposed to semantics, understood as a theory of denitive reference, on the grounds that reference
can only be developed informally, or ostensively. Hegels famous examination of sense-certainty is based on the idea that, since words are general,
they cannot serve to pick or otherwise identify singular items (Hegel, 1977:
pp. 5867). He specically insists that, as concerns sense-certainty, we
cannot say what we mean or mean what we say. If Hegel is correct, then
saying and meaning are separated by the intrinsic generality of language
that identies the true on the level of generality, whereas our immediate
intention is to pick out a single item given in sensation. For representationalism Brandom substitutes a reliabilist form of externalism. Yet in
Kants wake, Hegel gives up any reference to anything lying beyond
conscious experience.
444

BR A N D O M , H E G EL A N D IN FE R EN T IA LI S M

Conclusion: Inferentialism, Hegel and Brandom


Brandoms Hegelian inferentialism is not basically similar but rather basically dissimilar to, and incompatible with, Hegels view. The difculty is
not that Brandom still has much to learn from Hegel (Hsle, 2000) but
rather that he and Hegel are up to very different things.
Brandoms interest in Hegel is symptomatic of the fact that this is a
time of philosophical change. The gures who dominated philosophy in
the twentieth century have mainly disappeared from the scene. Quine, an
opponent of mainline analytic philosophy, was also the strongest analytic
gure to appear since the War. His recent death is likely to have philosophical consequences that cannot yet be foreseen.
Analytic philosophy had already greatly changed before Quine died.
That some analytic thinkers (Sellars, Rorty, McDowell, Brandom) have
recently been returning to Hegel, one of the greatest idealist thinkers,
shows the extent of change in a movement that began by rejecting idealism
in all its forms at the beginning of the twentieth century.
Brandoms effort to press his novel form of semantics into epistemological service is another in a long line of analytic attempts to do better
than so-called continental thinkers, including Hegel, in solving (or
resolving) the latters problems. Analytic concern with continental problems is certainly laudable. One would hope that in time it will be matched
by increased continental attention to analytic problems.
Yet it would be mistaken to believe that an analytic interest in continental themes, or an appropriation of selected continental doctrines for
analytic ends, will overcome the deep differences between two traditions
largely concerned, or mainly concerned, with different approaches to
different problems. Rortys suggestion that Sellars and Brandom overcome the split between continental and analytic philosophy is mainly
rhetorical. Even if analytic thinkers are now more interested in the
continental and pragmatic traditions than before, the analytic and continental traditions are obviously very different, irreducibly dissimilar,
incompatible.
Brandom, who claims that his inferentialism is Hegelian, simply does
not do enough with Hegel in any of his main writings to justify his public
show of interest. This leads one to speculate about the sources and the
future of his Hegelianism. One wonders whether Brandoms interest in
Hegel will continue to develop, whether he will go on to write a book on
Hegel, and whether he will nd a way to bring his German predecessor
into the centre of his position. The philosophical discussion is awash with
overly rapid declarations of victory, with revolutionary views announced
with fanfare that are later withdrawn, and more recently replete with
efforts on all sides to overcome the lengthy split between analytic and
continental philosophy (Bernstein, 1971).
445

I N TE R NAT IO NA L J O U R NA L O F P H ILO S O P H ICA L ST U D IES

This task is harder than it looks. Analytic writers could usefully become
more interested in continental gures and conversely. Analytic philosophers might insist less on doing philosophy as opposed to doing the history
of philosophy. And continental philosophers could become more
concerned with problems than people. Many of the differences can indeed
be relativized, but it is unlikely that they can all be overcome. For the
analytic and the continental traditions are substantively different. I suspect
that in time Brandom will realize this point and talk less and less, not
more and more, about Hegel. For Brandoms position has to succeed or
fail on its own merits, not on any claimed relation to Hegels, which he
has not so far justied.
Duquesne University
References
Ayer, A. J. (n.d.) Language, Truth and Logic, New York: Dover.
Bernstein, Richard (1971) Praxis and Action: Contemporary Philosophies of
Human Activity, Philadelphia: University of Pennsylvania Press.
Brandom, Robert (1994) Making It Explicit, Cambridge, Mass.: Harvard University
Press.
Brandom, Robert (1999) Some Pragmatist Themes in Hegels Idealism:
Negotiation and Administration in Hegels Account of the Structure and
Content of Conceptual Norms, in European Journal of Philosophy 7(2)
(August): 16489.
Brandom, Robert (2000a) Articulating Reasons: An Introduction to Inferentialism,
Cambridge, Mass.: Harvard University Press.
Brandom, Robert (2000b) Vocabularies of Pragmatism: Synthesizing Naturalism
and Historicism, in Brandom, 2000c.
Brandom, Robert (ed.) (2000c) Rorty and His Critics, Oxford: Blackwell.
Brandom, Robert (2001), Holism and Idealism in Hegels Phenomenology
unpublished.
Carnap, Rudolf (1937) The Logical Syntax of Language, trans. Amethe Smeaton,
New York: Harcourt, Brace and Co.
Devitt, Michael (1997) Realism and Truth, 2nd ed., Princeton, NJ: Princeton
University Press.
Dummett, Michael (1978) Truth and Other Enigmas, Cambridge, Mass.: Harvard
University Press.
Fichte, J. G. (1991) The Science of Knowledge, trans. Peter Heath and John Lachs,
New York: Cambridge University Press.
Fodor, Jerry and Lepore, Ernest (1992) Holism: A Shoppers Guide, Oxford:
Blackwell.
Friedman, Michael (2000) A Parting of the Ways: Carnap, Cassirer, and Heidegger,
Chicago: Open Court.
Goldman, Alvin (1976) Discrimination and Perceptual Knowledge, Journal of
Philosophy 73(20): 77191.
Grau, Alexander (2001) Ein Kreis von Kreisen. Hegels postanalytische
Erkenntnistheorie, Paderborn: Mentis.
Haugeland, John (1982) Heidegger on Being a Person, Nous 16: 1626.

446

BR A N D O M , H E G EL A N D IN FE R EN T IA LI S M

Hegel, G. W. F. (1967) Hegels Philosophy of Right, trans. with notes by T. M.


Knox, New York: Oxford University Press.
Hegel, G. W. F. (1977) The Difference Between Fichtes and Schellings System of
Philosophy, trans. H. S. Harris and Walter Cerf, Albany: SUNY Press.
Hegel, G. W. F. (1987) Phenomenology of Mind, trans. A. V. Miller, New York:
Oxford University Press.
Hsle, Vittorio (2000) Inferentialism in Brandom and Holism in Hegel: A Reply
to Richard Rorty and Some Questions for Robert Brandom forthcoming.
Kuhn, Thomas (1970) The Structure of Scientic Revolutions, Chicago: University
of Chicago Press.
Lukcs, Georg (1971) History and Class Consciousness, trans. Rodney Livingstone,
Cambridge, Mass.: MIT Press.
McDowell, John (1998) Meaning, Knowledge, and Reality, Cambridge, Mass.:
Harvard University Press.
McDowell, John (1999) Comment on Robert Brandoms Some Pragmatist
Themes in Hegels Idealism , European Journal of Philosophy 7(2) (August):
1903.
Moore, G. E. (1958) The Refutation of Idealism, The Monist (1903), reprinted
in G. E. Moore, Philosophical Studies, London: Routledge & Kegan Paul, 1922,
1958.
Putnam, Hilary (1981) Reason, Truth and History, New York: Cambridge
University Press.
Putnam, Hilary (1998) Renewing Philosophy, Cambridge, Mass.: Harvard
University Press.
Quine, W. V. O. (1961) Two Dogmas of Empiricism, in From A Logical Point of
View, Boston, Mass.: Beacon, 2046.
Quine, W. V. O. (1974) The Roots of Reference, La Salle, Il: Open Court.
Ramsey, F. P. (1931) Knowledge, in F. P. Ramsey, The Foundations of Mathematics
and Other Essays, ed. R. B. Braithwaite, New York: Harcourt Brace, 1931.
Rescher, Nicholas (2000) Realistic Pragmatism: An Introduction to Pragmatic
Philosophy, Albany: SUNY Press.
Rorty, Richard (1979) Philosophy and the Mirror of Nature, Princeton: Princeton
University Press.
Rorty, Richard (1997) Introducion, in W. Sellars, Empiricism and the Philosophy
of Mind, Cambridge, Mass.: Harvard University Press, 1997.
Rorty, Richard (1999) Philosophy and Social Hope, London: Penguin.
Rorty, Richard (2000) Response to Brandom, in Brandom, 2000c: pp. 1838.
Russell, Bertrand (1964) Principles of Mathematics, New York: Norton.
Sellars, Wilfrid (1991) Philosophy and the Scientic Image of Man, in Sellars,
Science, Perception and Reality, Atascadero, Calif.: Ridgeview, 141.
Sellars, Wilfrid (1997), Empiricism and the Philosophy of Mind, Cambridge, Mass.:
Harvard University Press.
Stekeler-Weithofer, Pirmin (1992) Hegels analytische Philosophie, Die Wissenschaft
der Logik als kritische Theorie der Bedeutung, Paderborn: Schningh.
Wittgenstein, Ludwig (1969) On Certainty, ed. G. E. M. Anscombe and G. H. von
Wright, New York: Harper & Row.

447