10b | CR fragment of remake of Book IV, Chapter 2

close to oneself merely because it happens to
happen in one's spatial neighbourhood. This
personal pattern is projected into objects
just as the other one was.

Another kind of unity which seems to contain an
ineradicable residue of objectivity is the uni-
ty of dependants. Events which depend on each
other seem to stand out as clusters which be-
long functionally together, relatively indepen-
dant of spatial arrangements. In recent years
the quantitative study of functions, correlat-
ions and co-variations allowed us to measure
degrees of dependance, and thereby of closen-
ess and intimacy in the interrelations of va-
rious facts, and to determine quantitatively
the relative weight of one given cause in a
multiplicity of conditions.

The units thus obtained may sometimes coinci-
de with the supposed limits of things and per-
sons, but very often they do not. Proper-
ties found in the same subject need not vary
in conjunction, i.e. the colour, density and
specific heat of a piece of copper. If some
2/3 of the variation of the wife's age (at mar-
riage) is accounted for by variations in the
husband's age (the square of r being 0.6683),
it does not follow that husband and wife are
the same person. In any case, in pattern
units of dependance are very unilike the fam-
iliar persons and things, not being arranged
in degrees of centrality, or closeness to a
supposed centre. They bring a and h together,
but not with reference to a common subject (A),
or to the self of anything. The relations of
dependance refer to statistical averages,
and other extreme abstractions which are not
meant to be put together into persons or
things. The connections of physical abstracti-
ons in particular, with the perceptual univer-
se is very obscure.

To describe a function which expresses a law
as a durable, lasting, permanent relation is a
rather misleading manner of speaking. We speak
of 'law' when 2 or more similar events a and b
(their dissimilarities being regarded as of no
importance) will, repeated in approximately the
same isolated set of circumstances or condit-
ions, exhibit practically the same