assumably etc. (was Re: "last stitch effort")

[Arnold M. Zwicky]

putting aside the "supposably" topic, i'll focus on "assumably" (which
jonathan lighter and john baker, probably among others, dislike).

background: given the verbs "assume" and "presume" (of very similar
meaning) and the combinations of derivational suffixes "-ed-ly" and
"-ab-ly" (again, of very similar meaning), four perfectly well-formed,
and semantically overlapping, adverbs are predicted (given here with
their raw google web counts):

   assumedly:  14,200           presumedly:          7,310
   assumably:    6,340                  presumably:  3,050,000

that is, all four are attested, in respectable (and non-ignorable)
numbers.  "presume" heavily outnumbers "assume", but "-edly" is
preferred for "assume", "-ably" for "presume".

it's no great surprise that the preferences mirror history.  here are
the NSOED (1993) dates for the four words:

   assumedly: L19                       presumedly: M19
   assumably: not listed                presumably: M17

that is, "presume" got in there first, and with "-ably".  "assume" got
into the action much later, with "-edly".  i'd guess that there's no
deep reason for this order of appearance.  presumably, all four words
were possible creations for centuries, and probably were created many
times by individual speakers or writers.  as soon as one derived adverb
spread for "assume" and another for "presume", however, there was
little motivation to create either of the other two possibilities:  as
usual, the existence of one widely used expression blocks the creation
of a synonym.

however, this pre-emption effect depends on speakers knowing something
about the currency of particular expressions in the language of the
time, and it would be easy for particular speakers to be ignorant of
some of these facts, in which case the way is open for innovating,
using the productive derivational patterns of the language.  but it
would be hard to miss the high frequency of "presumably".  as a result,
"presumedly" is a possible innovation only if you want to make a
semantic distinction between "-edly" and "-ably", but "assumably" is a
*very* likely innovation on analogy with "presumably".  the result:
"assumably" is a serious competitor for "assumedly".

a somewhat similar case, though with very different details, is the
"seasonable"/"seasonal" competition.  these two adjectives now have
clearly different (but somewhat overlapping) meanings.  according to
NSOED, they appeared at different times: "seasonable" and "seasonably"
in middle english, "seasonal" and "seasonally" not until M19.
nevertheless, "seasonal" and "seasonally" have it all over "seasonable"
and "seasonably", presumably as a consequence of the frequency with
which people want to convey one meaning vs. the other:

   seasonal:  8,440,000         seasonally:  785,000
   seasonable:   44,900         seasonably:   23,600

(in fact, if you do a google search for "seasonable", you're asked if
you meant "seasonal", and similarly for "seasonably" vs. "seasonally".
google takes its numbers seriously.)

but now consider their negative versions.  MWDEU points out that
"unseasonal" and "unseasonable" are essentially synonymous (ditto, i
add, for "unseasonally" and "unseasonably"), and that "unseasonal" "is
a very rare word".  well, "unseasonal" (which i had thought, for years,
was the *only* acceptable word, until i realized that "unseasonable"
was all over the place) definitely lags behind "unseasonable", and
"unseasonally" is *hugely* behind "unseasonably":

   unseasonal:     13,300               unseasonally:     2,380
   unseasonable: 42,000         unseasonably: 67,900

so here, where the meaning difference is leveled, we see history
mirrored again: the earlier word continues to outnumber the later
synonym.

arnold (zwicky at csli.stanford.edu)