Date: Tue, 13 Feb 2001 11:53:04 -0500 (EST)
Subject: the epistemic fallacy


Hi all,

As is apparently my wont, I think that this epistemic fallacy thing is more
complicated than it seems.  Sorry for the long post to follow.  Skim it (or
skip it!) if you like.  

But here's the thing: Bhaskar goes at the concept of the epistemic fallacy
in a number of different ways.  And I don't think that they all have the
same meaning.

The first way that he presents the concept is that he gives us a definition:
the epistemic fallacy is when questions about being are transposed into
questions about our knowledge of being.  So, to give a crude and flawed
example, "Do dogs exist?" becomes "What is it possible to know of dogs'
existence?" or, as in Kant, "What are the necessary cognitive pre-conditions
of our having knowledge of dogs?"  [For the post-linguistic turn, po-mo type
examples that people have given in recent posts (e.g., "How are dogs
discursively constructed?") Bhaskar actually has a separate name: the
linguistic fallacy.]

In any case, in this first formulation the the mistake that has been alleged
to have been made is an EPISTEMOLOGICAL, or meta-theoretical one.  It's like
the person has made a methodological error, but at the level of meta-theory.

The second way that Bhaskar presents the concept of the epistemic fallacy is
via his discussion of the three "domains" of the stratified critical realist
ontology.  Here the epistemic fallacy is when the domain of the Real is
identified with the domain of the Empirical.  In this case, the problem is
not that the person (in this case an empiricist) has transposed a question
about what exists into a question about what &/or how we know things, but
rather that they have inadevertently lopped off most of reality by saying
that the only things that exist are either (a) the things that we perceive
through our senses or (b) our sensations themselves.  [Whether (a) or (b)
depends on what sort of general metaphysical implications the person wants
to draw from their empiricist theory of knowledge.]  So here, in this second
formulation, the mistake that has been made in an ontological one.  The
person has adopted a flawed, radically truncated ontology.

The third way that Bhaskar approaches the concept of the epistemic fallacy
is through the analysis of scientific laws.  Empiricists, Bhaskar says,
believe that natural scientific laws refer in the end to regularities in our
perceptual fields.  Properly conceived, however (according to Bhaskar),
scientific laws refer not to perceptual regularities but to properties of
independently existing, non-mental entities.  In this presentation, the
epistemic fallacy is when laws are mistakenly thought to refer to
"subjective" features of our experience rather than to "objective"
properties of entities.  Here we are back to it being a epistemological
error, or what I called above a problem at the level of meta-theory: the
person doesn't know what a certain kind of theoretical construct (viz., a
"law") is.

Finally, a fourth way that Bhaskar approaches the concept is through the
analysis of causality.  The epistemic fallacy here is when causality is
mistakenly believed to be either an expectation (of regularity), as in Hume,
or, as in Kant, an a priori category of reason itself.  Here the argument is
the same as that concerning laws, but the error is back to being an
ontological one:  the person has a flawed ontology because they don't
realize that causality is an objective feature of the world, operative in
virtue of the real powers of entities.       

The reason why I think that it is important to distinguish between each of
these different presentations is that a thinker such as Kant, for example,
who is the prime example of someone who commits version 1, doesn't commit
version 2 at all.  Meanwhile, at least in the form in which *do* see it in
Kant, version 1 is not at all the same sort of thing as the sort of
subjectivism that one might argue is the end-result of adherence to an
empiricist theory of knowledge.  Thus it is questionable whether empiricists
actually commit version #1.

In the end, I think that Bhaskar runs into trouble by trying to take on Hume
and Kant with the same concept.  Empiricism and transcendental idealism are
different enough from eachother that they require separate lines of
argument.  Personally, I tend to think of version 1 as "the epistemic
fallacy"; version 2 as "phenomenalism"; version 3 as either "positivism" or
"transcendental idealism," depending on who the target is; and version 4 as
either "That empiricists can't account for causality" or "transcendental
idealism," again depending on the target.  One could argue that version 1 is
just transcendental idealism too, but to me it's the formulation that's the
most compelling -- the one in which the introduction of a new term, a
fallacy with a proper name, seems the most appropriate.

So yes, I would say that the epistemic fallacy designates a purported error
in reasoning, in which an ontological question is transposed into an
epistemological question.

Ruth 

       


First of all, there's a real and I think unhelpful ambiguity at play
regarding what kind of a mistake the epistemic fallacy actually is.  Most of
the time, Bhaskar uses the term to designate a specifically ONTOLOGICAL
error: the epistemic fallacy refers to a flawed ontology, an ontology in
which the category of "that which exists" is taken to be exhausted by the
category of "that which is known."  In the empiricist case, "that which is
known" equals "that which is perceived through sensation"; in the radical,
or phenomenalist empiricist case, "that which is known"  equals "sensations
themselves."  In RTS, the focus is on empiricists of the purest,
phenomenalist variety.  Thus the reduction of reality to that which is known
is presented in RTS as a reduction of "the Real" to "the Empirical," where
the latter means "perceived by someone through their sense organs or through
a technological extension thereof."      

  


At 11:27 AM 2/11/01 -0700, you wrote:
>
>I fully agree with Greg's answer to the first question.
>The question was:
>
>>Bhaskar says that Humean empiricism commits the epistemic
>>fallacy: explain where it does this.
>
>Greg stresses that the error lies in the *implicit*
>ontology.
>
>Expicitly, overtly, Humeans think about it whether
>experience can be our only source of knowledge, or whether
>our knowledge of the world must also include an a priori or
>theoretical component.  But in all these deliberations they
>overlook that they let their conceptions pf what they
>consider to be "the world" be guided by sense-eperience.
>
>Let me say it again, elaborating a little: They ask: "Is my
>sense-experience of this chair here (and of other things and
>events which I can experience) sufficient to constitute my
>knowedge of this chair?"  They ask this not only about
>chairs but also about all the other objects of
>sense-experience, but they don't ask whether there are other
>things in this world, which are not the objects of
>sense-experience, worth knowing about too (for instance
>because without these things the scientific procedures which
>they are engaged in would not be possible).  I.e., they
>implicitly define that what is real by that what can be
>experienced, and this is where they commmit the epistemic
>fallacy.
>
>
>Here is a quote from RTS2, p. 41, just after the passage I
>sent to the list, where Bhaskar says this (I think):
>
>> But from Hume onwards the sole question in the philosophy
>> of science is whether our knowledge is exhausted by our
>> knowledge of such events and their conjunctions; it is
>> never questioned whether they in fact occur.  That is,
>> philosophy's concern is with whether our knowledge of the
>> world can be reduced to sense-experience as so conceived
>> or whether it must include an a priori or theoretical
>> component as well; not with whether experience can
>> adequately constitute the world.
>
>And here is a sentence of Greg's, with my additions in
>square brackets: is this how you meant it, Greg?
>
>> Since Hume sees knowledge as only [knowledge of] events of
>> sense-experience and scientific laws are induced from the
>> constant conjunction of these events, then there is no
>> reality beyond [these objects of] sense-experience.  This
>> is the epistemic fallacy because is reduces the idea of
>> what is real to what we can know about reality.
>
>
>I will stop here for now, because I don't want the emails to
>get too long.  More about the other questions later.  Others
>are invited to answer too, or to comment on these answers
>here (including my own).
>
>-Hans.
>
>
>     --- from list seminar-14-AT-lists.village.virginia.edu ---
>



     --- from list seminar-14-AT-lists.village.virginia.edu ---