topics: |  linguistics

corpus based analyses of usage and grammar conventions.  the fact that the
examples are taken from corpora (SBC, LSAC) and occasionally from the web,
grounds the work and permits frequency analyses.
For example:  "40/55 (72.7 percent) of the shaving activities in LSAC depict
self-shavings."

LSAC:  Longman Spoken American Corpus (LSAC),
SBC:   Santa Barbara Corpus of Spoken American English
	(both compiled by Jack Du Bois at UC Santa Barbara. )

Excerpt

the synchronic grammar itself encompasses coexisting layers of conventions,
some of them old, some of them new, even when the two sets are incompatible
with each other, as when they govern the same form but make different
predictions about its use (e.g. for English reflexive pronouns -
grammaticization processes associated with English reflexive pronouns chapt
6). This is a panchronic view of grammar, then.

People normally wash themselves rather than others. This is supposed to
motivate the typologically well-known fact that wash and other verbs of
grooming tend to take less-marked reflexive markings.  Indeed, English wash,
shower, and bathe are lexically reflexive verbs (according to Reinhart and
Reuland, 1993). As such, they are zero marked, as compared with the very
marked reflexive pronoun used in she talks to herself (LSAC).

[But not clear how this is so for wash]. 184/299 (61.5 percent) of its
occurrences describe nonself-washings (Haspelmath, 2004c has similar findings
for the British National Corpus).  In fact, while 115 (38.5 percent) of the
cases describe self-washings, the majority of these are not reflexive, in
fact, because the washee is a specific body part (hands, hair, face), which
is only partially identical in reference with the agent. It is not hard to
explain why wash actually tends to depict nonself-washings. Since English has
(take a) shower/bath and bathe, dedicated to self-washings (and there were
230 of those), wash can have a different distribution. Interestingly, even if
we focus on intransitive wash occurrences (30), the understood ‘washee’ is
not invariably the self: 11 of these (37 percent) are interpreted as ‘wash’
something else (dishes, clothes).

In other words, alienable objects can also be zero marked with wash. We
cannot then assume an automatic, transparent translation of a cognitive fact
(‘people mostly wash themselves’) into a linguistic fact (verbs of washing
will show this bias).

[AM: alienability: grammatical or semantic concept?]

hide vs conceal

[verb actions alone do not determine argument structure]

Smith (2004) treats the event of ‘hiding/concealing’ as one concept,
depicting a self-directed activity (namely, an activity one usually performs
on the self and not on others).  However, the two verbs do not have the same
grammatical structure Only hide can be intransitive
    1.  She’s hiding (SBC: 023)
    	* She’s concealing.
Obviously, it cannot be the event in the world in itself which determines the
status of verbs as self versus other-directed. Rather, it is the world as
filtered not only by our conceptualizations, but also by the grammatical
options available to speakers. p.167

[has a chapter on the structural form for reflexives (herself etc), and how
it used to be an emphatic in 15th c. English. ]

   Old English speakers at some point started adjoining an independent
   emphatic form (self) to their pronouns in order to counteract the default
   pragmatic inference to disjointness.  Gradually, these became the
   reflexive pronouns as we know them today. ...
   Emphatics start out as optional adjuncts, but often enough, they gradually
   grammaticize and become obligatory in certain contexts (to various degrees).