[hist-analytic] Reichenbach, Carnap, Positivism

Roger Bishop Jones rbj at rbjones.com
Sun Sep 20 04:01:11 EDT 2009

On Saturday 19 September 2009 14:49:11 Baynesr at comcast.net wrote:
>"It would be interesting to know which of Carnap's theories he was
>talking about. The Aufbau?"
>In all likelihood, Reichenbach was talking about Carnap's view that the
> protocol language consisted of statements about sense-data and that these
> sense-data are immediately given and, so, in some sense incorrigible. This
> incorrigibility throughout the history of philosophy has been compared to
> Descartes's grounding science in the incorrigibility deriving from "I think
> therefore I am."

The comparison seems pretty weak to me.
I would much more readily embrace a theory of perception in which
sense data are incorrigible (I think such a theory is tenable,
if you  get the details right) than accept the Cartesian cogito.

> Historically, I don't believe there can be any question
> about this.

About what?

> Reichenbach by contrast with the logical positivists averred
> that instead of talking about truth we talk about probability. This would
> rule out certainty with respect to the protocol language, whence the need
> to revise the doctrine.

I don't see that it necessarily would.
Even if sense data are incorrigible, the inference to the external
objects or to future sense data is not logical, and so one might
wish to retreat on these to probabilistic statements, though as
Hume observed, this doesn't solve the problem that the inference
is not sound.  The inference to probability claims isn't sound either,
unless they are just claims about the observed sense data rather than
claims about external objects or future sense data.
Whether or not you prefer probability statements for conclusions
you do not have to treat sense data as probabilistic, and I can't
see the merit in doing so.

>Interestingly, Reichenbach notes that it was the issue of probability that
> distanced the Berlin logical positivists from the Viennese logical
> positivists. The Berlin people thought that *prediction* could not be
> addressed within the framework of a logic that reduced to tautologies.

I think this is a mistake, and I would guess that this is a point
which Carnap never conceded.

> What
> Reichenbach did not appreciate, in my opinion, was that not only are
> statements of the future at issue, but so are statements about the past;
> and once you attempt to deal philosophically with the study of history you
> are no longer in the realm of strict causation. There are symptoms of that 
> which the positivists glided over.

But even if you were "in the realm of strict causation", causal
inference is not logical inference.

>One symptom is that contexts of discovery and contexts of justification were
> sharply distinguihed.

Sounds like a good idea to me!

> The former became a purely psychological matter.
> Discoveries could not in principle be predicted and might be arrived at in
> dreams etc.

Which surely is in fact the case?

> But I think this is symptomatic of a weakness in positivism as
> providing us with a world view.

What is the weakness?

> When we construct a theory we give reasons for making certain moves.

Which presumably belong to the context of justification?

> This is particularly evident in "miniature" when
> an investigator constructs a theory as to how a crime occurred. He gives
> reason for why a person MIGHT have done such and such. Coming up with a
> prediction of the investrigator's inferences does not fall within the
> purview of physical science, whence the need for the positivists to
> jettison the signficance of the context of discovery. Herein lies the real
> substance of what I am after; it cannot be artlessly dismissed without
> great loss.

I'm afraid I don't understand the point here.
I don't understand what you mean when you say that positivists "jettisoned"
the significance of the context of discovery, possibly because I don't
know what you think its significance is.

What is this "real substance" that you are after?

>According to Reichenbach, Carnap accepted the criticisms of the protocol
> language and made it a subdomain of the physical language; "logical
> positivism" became "logical materialism," just another variant of an old
> idea.

I would not myself think Carnap a materialist.
Even if his physicalistic language was purely materialistic, it was
for him just one way of talking among many.
He also endorsed a "theoretical" language, and had a well articulated
position admitting talk about abstract entities.

To the question of whether Carnap's mature philosophy should be called
positivism I have given some thought, since I have taken up the
term in my own philosophy even though my own views are on their
face even less positivistic than Carnap.
It is a topic I intend to address in some detail, and have started
a web page at:
Which, does not yet say much useful, but perhaps give a hint of direction.

I decided to give the name "Liberal Positivism" to what you get
be stripping negative dogmas out of the less liberal kind of
positivism, and this then provides an account of the positivistic
elements in Metaphysical Positivism.
However, I have yet to come up with a description of what you have
left when you do that, and why it is worth having.

> I don't think Carnap was "eager" to do this; he was just being
> honest.

I think he was pretty eager to get things right, and to fix any
problems in his philosophy whoever discovered them.

> In their later years, I think both Russell's and Carnap's
> philosophies became a wee bit sterile owing to their faith in materialism.

Can't speak for Russell here, but I can't see the legitimacy of
this criticism of Carnap.

>I'm going to defer comment on the synthetic/analytic distinction. I'm not
> sure how important it actually is. What is important to philosophy is the
> alleged nonexistence of the synthetic a priori. Retaining or rejecting the
> synthetic/analytic distinction is not as important as, say, definining
> numbers in terms of classes etc. inmy opinion.

How can you regard the synthetic/a priori as an important problem
without recognising the importance of the analytic/synthetic distinction?


More information about the hist-analytic mailing list