Having posted about
(1) (a) half (of) a Nom
(2) a half Nom,
focusing on the use of articles in such examples, I was moved to return to another aspect of these variants, the variation between plain (no of) and of constructions following half in the patterns in (1).
As I noted in that posting, some usage writers recommend against things like a half an hour on the grounds that it’s “redundant” or “pleonastic”; the advice is to omit one of the indefinite articles, as unnecessary. Some handbooks also recommend against
(3) half of NP (e.g., half of an hour, half of the shrubbery, half of the bushes)
as having an unnecessary of; the advice is to omit the of : half an hour, half the shrubbery, half the bushes. Similar advice is given for
(4) both/all of NP (e.g., both/all of my assignments, all of the shrubbery)
where once again we are told to omit the of: both/all my assignments, all the shrubbery.
All of has gotten more attention than the others. A few handbooks are resolute on this advice:
Weseen, Words Confused, p. 10: use “all my friends,” not “all of my friends”
Flesch, The ABC of Style, p, 19: a good writer or editor automatically changes all to all of
Some merely say that the of is unnecessary:
Bernstein, Dos, Don’t, & Maybes, p. 12: except with pronouns, the of following all is superfluous and may be omitted
At least one connects the of to informal style:
Garner, Garner’s Modern American Usage (3rd ed.), p. 33: All (of). The more formal construction is to omit of
Several handbooks (Burchfield, The New Fowler’s Modern English Usage, p. 41; Evans & Evans, A Dictionary of Contemporary American Usage, p. 25; MWDEU) note that the all of variant is more recent than the plain all variant.
And some (Evans & Evans, pp. 12, 25; MWDEU again) simply say that both variants are acceptable, as Swan’s Practical English Usage does for all, both, and half.
So far, this is a story of Omit Needless Words, with some writers advocating omission and others admitting both variants. There might also be some additional prejudice against the of variants on the basis of their relative recency (variants that are, or are perceived to be, innovations are often disfavored) or — a probably related consideration — their relative informality (innovations are often perceived to be more informal than corresponding older variants).
But there’s more: a prejudice, plain and simple, against the word of.