[hist-analytic] Russell and Eddington on Structure

steve bayne baynesrb at yahoo.com
Wed Jan 28 12:28:40 EST 2009

I will be continuing to comment on Aune's _Reason 
and Action_. There is, presently, an impasse. I 
can't agree with some of his remarks on "Doing 
something with the intention..," but I can't find 
a way around some of them. So I need to think on 
this for a few days. In the meantime I will introduce 
a couple of comments on some work of mine, developing, 
on Russell's metaphysics and philosophy of science.

Ever since M. H. A. Newman reviewed Russell's _Analysis 
of Matter_ there has been uncertainty over the exact 
source of Russell's proclamation that we can only know 
structure. ("Mr. Russell's 'Cauasal Theory of Perception'," 
_Mind_, April 1928) Russell would retract this statement 
in a letter to Newman (housed at the Russell Archives at 
McMaster University, Canada). But the mystery as to the 
source of Russell's views has been unresolved, although 
there are intimations of it elsewhere, including Introduction 
to Mathematical Philosophy and, elsewhere. But Russell 
scholars have failed to examine in any detail an important 
source drawn upon by Russell. What I will do, over time, is 
demonstrate the importance of Eddington in the development 
of Russell's metaphysics and in particular his treatment of 
causal structures in Human Knowledge Its Scope and Limits.

Russell moved away fron logic in his middle/late years 
but he retained an interest in how to make use of his 
theory of relations which, as anyone who has read Carnap 
knows, was exercised considerable influence on the logical 
positivists' philosophy of science. I believe that the 
theory of relations interacts in Russell while he was under 
Eddington's influence. In particular, the application of 
group theory to physics (Weyl) possessed an allure Russell 
could not resist. Eddington was a regular "soldier" when it 
came to advocating this view, and it shows in his 1934 
Messenger Lecture. This lecture comes well after the 
Russell/Newman exchange, but I believe its spirit and content 
were incorporated into Russell exapanded treatment of structure. 
When in his correspondence with Newman he rejects his claim 
(without citation) of the primacy of our knowledge of structure 
I think he was correcting only the error that comes with 
saying that ALL we know is structure. After all, knowledge 
by acquaintance need not be knowledge of structure. His 
emerging views on structure interface with his departure 
from his earlier views on neutral monism. But more on this 
latter. For now consider this interesting quotation from 

"The knowledge we can acquire is knowledge of a structure 
or pattern contained in actions...But whatever is derived 
in the actual (highly difficult) study of the atom is knowledge 
of the same type, i.e., knowledge of structure of a set of 
unknown operators."
("The Theory of Groups" in _The World of Mathematics_ vol. I
V. ed. James Newman. Simon and Schuster, 1956).

I want to propose something a bit radical: heretofore, 
Russell's metaphysics has been pursued by way of his logic 
and, mainly his logic; but his logic as it relates to, say, 
proof theory, is not what moves him the most. What moved 
him were the ideas of people like Veblen, Hausdorff and 
Eddington. A new look at the way Russell studies is 
conducted must include a close look at Eddington. This is 
my intention.

Steve Bayne

More information about the hist-analytic mailing list