Barfield says that "in the days of Locke and Hume it was felt that science was the newcomer, requiring a foundation in philosophy; but since then the two have changed places". Despite the "ever-changing assumptions" of science, modern philosophers do not try to question the scientific assumptions of the day (seeing them as "given"), but seek to "justify the ways of science to man". Who can argue with success?
He notes that there is one assumption of science that has remained unchanged longer than the rest, and that is that the real world is a "somewhat" in "the construction which the mind of man does not participate; of which it is purely a detached observer" (here, Hume's philosophy is relevant). "It is of course in attempting to describe more precisely the nature of the 'somewhat' that science both parts company with the man in the street and keeps changing its ground". In the 19th century the real world was assumed to consist, in the last resort, or things. These things kept getting smaller and smaller, but they were at least there, and if you ad a powerful enough microscope, for example, you'd be able to see them (18). In like fashion, Hume had been content to say that the "'impressions' which were the material of knowledge were produced in the senses by 'objects'" (19).
20th century science though, complains Barfield
"has abolished the 'thing' altogether; and twentieth-century philosophy (that part of it, at least, which takes no account of imagination) has obediently followed suit. There are no objects, says the voice of Science, there are only bundles of waves - or possibly something else; adding that, although it is convenient to think of them, it would be naive to suppose that the waves or the something else actually exist. There is no 'referent', echoes the philosophy of linguistic analysis deferentially, no substance or underlying reality which is 'meant' by words. There are only descriptions, only the words themselves, though it 'happens to be the case' that men have from the beginning so persistently supposed the contrary that they positively cannot open their mouths without doing so" (19).
Barfield then quotes Logical Positivist A.J. Ayer saying "that we cannot, in our language, refer to the sensible properties of a thing, without introducing a word or phrase which appears to stand for the thing itself as opposed to anything which may be said about it."
He goes on: "Kant erected the Forms of Perception as a kind of impenetrable screen between the real world of 'things in themselves' and the mind of man. The Positivists have substituted syntax for the forms of perception, and scrapped the things as otiose" (19).
My summary: Barfield seems to be on the verge of concluding that given scientist's skepticism about finding out what is really real - and positivist's corresponding expulsion of the referent in language - we are stuck with not really being able to know anything (i.e. if we take their presuppositions to their logical conclusion).