I think there are real and serious problems in McDowell's position but the following extract indicates how he seeks to avoid (excuse the pun) 'spinning in a void'. McDowell's basic move to steer clear of the problem you ably expose is to claim that sense experience itself - the causal impact of the world on our senses - already has conceptual content. The move needs a clearer description this review might provide (sorry for the long quotation but McDowell as you know doesn't lend himself to easy exposition or critique) :
McDowell's basic idea is that we can
satisfactorily overcome the opposition between Coherentism and
the Myth of the Given only by recognizing, with Kant, that concepts
and intuitions, understanding and sensibility, must be integrated
together in every cognitive act or process - even in the mere intake
of experiential content characteristic of sense perception. There is
thus no room, according to McDowell, for either unconceptualized
sensory input standing in no rational relation to conceptual thought ("intuitions without concepts are blind") or purely intellectual thought operating independently of all rational constraint
from sense experience ("thoughts without content are empty").
McDowell characterizes the understanding, the sphere of conceptual thought, as the "space of reasons" (5). The understanding, for McDowell, is thus constituted by rational or inferential
relations ("relations such as implication or probabilification" (7)),
and it counts as a faculty of spontaneity in virtue of the Kantian
linkage between rational necessitation and freedom. The understanding is active rather than passive because of our freedom - and
accompanying responsibility - rationally to examine and to revise
all elements in our perpetually evolving conception of the world
(12-13). However, if the understanding can thus generate a conception of a truly independent empirical world, there must also be
some rational constraint from sense experience. The operations of
spontaneity cannot be entirely free, on pain of "degenerate[ing]
into moves in a self-contained game" (5) or "a frictionless **spinning
in a void" (11). This is precisely the threat posed by Coherentism,
and the opposing Myth of the Given then tries to alleviate this
threat by invoking bare (unconceptualized) sensory presences
somehow acting on the understanding from outside the conceptual
sphere. But the problem here, as Davidson in particular has made
especially clear, is that the relation between sense experience and
conceptual thought can now not be conceived of as a rational
one-as a genuine relation of justification. As Davidson himself
puts the point, "nothing can count as a reason for holding a belief
except another belief."
This last formulation leads Davidson himself to a "coherence
theory of truth and knowledge": specifically, to the view that sense
experience - the impact of the world on our senses - plays a causal role in the generation of belief rather than a justificatory role.
And precisely this consequence is the basis for McDowell's objection to Davidson. Since "Davidson's picture depicts our empirical
thinking as engaged in with no rational constraint, but only causal
influence, from outside," it does indeed pose the Coherentist
threat of "spontaneity as frictionless, the very thing that makes the
idea of the Given attractive" (14). For McDowell, the only way to
overcome this threat is to maintain that sense experience itself -
the causal impact of the world on our senses - already has conceptual content: "In experience one takes in, for instance sees, that
things are thus and so. That is the sort of thing one can also, for
instance, judge" (9). This does not mean, however, that sense
experience, for McDowell, just is a form of belief or judgment. For
the impact of the world on our senses is an expression of our
receptivity rather than our spontaneity. In sense experience the
world strikes us, independently of our control, as it were, as thus
and so: we are passively presented with the world's appearing to
be thus and so rather than actively judging (perhaps after reflectively deciding whether to accept this appearance or not) that the
world is in fact thus and so (10-11). (Michael Friedman, 'Exorcising the Philosophical Tradition: Comments on John McDowell's Mind and World', The Philosophical Review, Vol. 105, No. 4 (Oct., 1996), pp. 427-467 : 427-9.)
Problems for McDowell
Phenomena such as active thinking are placed in the space of
reasons, while the motions of inanimate particles must be located in the
realm of law. The two worlds have to be kept separate if we are to preserve
distinctively normative concepts such as justification.
Yet by McDowell's own account, the two worlds must also interact if we
are to have empirical knowledge, and more fundamentally, empirical
content. Human perception, insofar as it is 'permeated by spontaneity', is
unique in the animal kingdom. It involves the very conceptual capacities
that are involved in giving and asking for reasons, and therefore must be
placed in the space of reasons. But human perception (and therefore
human reason) is also conditioned by the physical processes that govern
the interaction between our sensory apparatus and our environment -
processes that belong to the realm of law.
McDowell acknowledges the obvious point that perception is enabled
only through physical, law-governed processes. He develops the concept of
second nature precisely to show how human reason can have 'enough of a
foothold in the realm of law', enough of a basis in the 'potentialities that
belong to a normal human organism,' to respect natural science (84). The
idea is that the biological human infant, as part of her first nature (located
squarely in the realm of law), has the potential to be initiated or moulded
into becoming responsive to rational demands, in just the way that she has
the potential to learn English or Chinese. The second nature of a mature,
adult human is precisely her moulded intellect, which allows her to
perceive these rational requirements.
In terms of the 'two-worlds' argument, however, second nature appears
to play the role of the overworked 'intervening entity'. Does our second
nature belong solely to the realm of law, solely to the space of reasons, or
to both worlds? Since second nature involves the possession of conceptual
capacities, i.e. spontaneity, it cannot belong exclusively to the realm of law.
Nevertheless, as emphasized above, even if we grant that an adult human,
equipped with second nature, has a species of perception different from
that of non-human animals, actual cases of perception - seeing another
person's face, for example - must involve interaction with the natural environment. An exercise of our perceptual capacities might belong to our
second nature, but it does not leave behind or become distinct from our
first, or biological, nature; rather, it just is that nature actualized in a determinate way. Insofar as second nature just is a particular shaping of our first
nature, it must also belong to the realm of law. This leaves us with the last
option, that second nature belongs to both worlds. Following Passmore,
we are then entitled to ask: if this is possible for second nature, then why
not for other things as well? We can no longer insist on the separateness of
the space of reasons and the realm of law. (Paul Bartha and Steven F. Savitt, 'Second-Guessing Second Nature', Analysis, Vol. 58, No. 4 (Oct., 1998), pp. 252-263 : 257-8.)
Reply to query
I'm not sure how it answers the question. It does redefine the problem, showing the tension between the passive "given" and the active "space of reason", but I still can't see how this tension is solved.
As I understand him, McD rejects the myth of the given on the grounds that nothing is 'given' pure in experience but 'all experience concepts and intuitions, understanding and sensibility, must be integrated together in every cognitive act or process - even in the mere intake of experiential content characteristic of sense perception'. This integration is inherent in human cognition and is not 'given' as an external input in relation to which we are passive. There is no 'unconceptualized sensory input'. Equally, however, 'if the understanding can ... generate a conception of a truly independent empirical world, there must also be some rational constraint from sense experience. The operations of spontaneity cannot be entirely free'. So if we retain 'a conception of a truly independent empirical world', there can be no 'purely intellectual thought operating independently of all rational constraint from sense experience'. No vocabulary we might adopt can terminate rational constraint from sense experience, leaving us to justify every belief we hold true only by using other beliefs - 'spinning in a void'.
However, if you read the new extract from Paul Bartha and Steven F. Savitt you may come to feel that your problems with McD don't derive from a failure of understanding on your part but from a defectiveness of argument on his.
