GUNTHER S. STENT
37
Holton refers to these as "analytic") and the empirical propositions
which "ultimately boil down to meter readings and other public
phenomena" (to which Kant referred as "synthetic" but which Holton
prefers to call "phenomenic"). Although this late eighteenth century
two-dimensional image of scientific knowledge represented a con–
ceptual advance over the two rival one-dimensional views of ra–
tionalism (true knowledge is analytic) and empiricism (true know–
ledge is synthetic) that preceded it in the seventeenth century, it is
still inadequate. Holton asks why, if the two-dimensional image
were true , "is science not one great totalitarian engine taking every–
one relentlessly to the same inevitable goal?" How can the two-di–
mensional image account for "the easily documented existence of
[scientific] pluralism at all times ," which is historically manifest in
the "fundamental antagonisms in terms of programs, tastes and be–
liefs, with occasional passionate outbursts between [highly compe–
tent, well-matched] scientific opponents?"
To resolve this quandary , Holton introduces a third axis, the
"thematic," which, in consort with the analytic and phenomic axes,
defines a three-dimensional space of scientific knowledge . Thus par–
ticular pieces of knowledge correspond to particular volume ele–
ments rather than to particular areal elements . Along the thematic
axis "are located these fundamental presuppositions, often stable,
many widely shared, that show up in the motivation of the scientist's
actual work, as well as in the end-product for which he strives. Deci–
sions between them, insofar as they are consciously made, are judge–
mental (rather than, as in the phenomic-analytic plane , capable in
principle of algorithmic decidability) ." Holton calls these presupposi–
tions (which require a philosophical magnifying glass to be distin–
guished from Fleck's thought styles or Kuhn's paradigms)
themata.
Thus, two equally competent scientists who arrived at the same
areal element in the phenomic-analytic plane would reach different
volume elements in the space above that plane if they brought dif–
ferent themata to their work.
It is not surprising that Holton's spatial analogy of scientific
knowledge with its thematic axis has been largely ignored by
philosophers of science. First, it is hard to imagine what kinds of
metrics are supposed to be represented along the phenomenal,
analytic and thematic axes. For Holton's three-dimensional know–
ledge space to have mnemonic value, it ought to incorporate the con–
cept of semantic distance between the particular volume elements: A
geometric analogy is not much help for remembering knowledge if it
