Correction: I'm told

[bwald]

----------------------------Original message----------------------------
I appreciated Cynthia's further correction of my correction, and I look
forward to her further comments.  She notes:
 
>The OE was parallel to the German, with the theme causing subject-verb
>agreement (whether it was pre-or-post-verbal).  I think perhaps Benji may
>have had some confusion with passives of verbs like *deman* to judge, which
>took a single object in the dative case.  The verb had passive morphology
>but did not agree with anything, remaining in the neutral third singular,
>as in 'them was judged'.
 
I'm still haven't checked, but my understanding of the "impersonal passive"
in some earlier I-E languages is that there was no marked SUBJ (NOM)
(hence, "impersonal" = no subject/NOM marker).  It either occurred to me or
I read that in the case of neuters, there was formal ambiguity of between
"impersonal" and regular (ACC => NOM) passivisation for TWO reasons.  (1)
there is no formal distinction between neuter NOM/ACC, (2) the neuter may
not require number agreement on verb inflection, remaining singular
regardless of the number of the neuter NOM/ACC.  The situation with *deman*
mimics the "impersonal" passive, except that the verb does not allow an
additional (non-neuter) object, in order to judge whether it would be
marked as SUBJ or OBJ.  From Cynthia's comments I gather that by OE, the
"impersonal" passive has been lost in favor of the ACC-Passive (i.e., ACC
-> NOM), assuming that it ever figured in the prehistory of English or
earlier Germanic.
 
I think it would be instructive to know what the function of the
"impersonal passive" was which distinguished it from the "regular" passive
in languages which have both, in order to determine whether it contributed
anything to the development of the ACC*=DAT* passive in Middle English.
This might be the case, if the impersonal passive favored topicalisation
(pre-posing) of the DAT, i.e., a grammatical process which cooperated with
word order strategies.  Otherwise, it is simply the topicalisation strategy
itself that identified DAT, like ACC, with subject (NOM), when the marking
difference between DAT, ACC *and NOM* (!), was obscured.  Unmarked preposed
arguments came to be interpreted as subjects (= "objects") of
passivisation.  Non-passive residual phenomena associated with DAT
preposing persisted for quite a while, but lexically controlled (with some
increasingly dimly perceivable semantic motivation), as in
"perception/cognition" ("psych") verbs like "methinks" (i.e., "it seems to
me")  and "melikes" (note the eventual spelling at the threshold of
extinction in Early Modern E).
 
(post-OE DAT/NOM variation has had various *lexical* fates.  "seem" is an
interesting case in which a DAT > NOM innovation has not survived e.g., we
*seme* it no shame (1485) = consider, cf. we DEEM it no shame.   "what
*lacked* yew" (1175) eventually gave way to "both they *lakken* (1320 )
(still "he lacks both"), but survives marginally in "both are lacking
*in/to/for* him"  The masochistic "bitter griefs taste me best" (1586)
continues only as the *marked* "DAT" "bitter griefs taste best *to me*
cntr.  "I taste bitter griefs best", where "best" is an adv not an adj.
French "please" into ME got caught up with "like", but seems to retain only
limited use with DAT > NOM, e.g., "they do what they please, when they
please, if they please", compared to the range of contexts in which the
experiencer remains the object, e.g., it pleases them to listen, listening
pleases them, etc.  Lexical peculiarities of individual verbs mitigate any
exclusive direction of syntactic development, cf. I fear X, X
scares/frightens me, etc., though  X fears= scares me rests in a fresh
grave awaiting rebirth through the "transitivisation" pattern, e.g., X
fears and disappears me.)
 
In any case, the lexical longevity of the "methinks" type of verb contrasts
with the relatively rapid thoroughness with which DAT *passives* were
reanalysed as NOM=Object passives.  The German strategy for approximating
the English/Bantu "indirect" passive attracted my attention because it
shows a cooperation of word order and morphological marking strategies
(passivisation) that is suggestive of similar strategies in OE as a similar
case-inflected language.  English DAT/ACC passivisation continues those
word order strategies in the absence of case-marking (except for those
famously conservative pronouns.)  They are preferred to preposing
prepositional phrases in recent and current English, i.e., "I was given a
book" is preferred "to me a book was given".  More precisely, "I was given
a book" is used where the old DAT is "topical" and the clause "focus" is on
the old ACC.  "to me a book was given" denies post-verbal clause focus to
either argument, with an effect that serves more restricted purposes. ("to
me (there?!) was given a book" does not rate serious consideration as an
alternative "theme focus" strategy to "I was given..." in current English.)
 
I also appreciated Paul Hopper's additional observations on the neuter
singular pronoun in various Germanic languages, including the observation
about the avoidance (ungrammaticality??) of "that's *its*", parallel to
"that's mine".  In fact, I suspect similar avoidance of "that's *theirs*"
iff "their-" refers to an inanimate plural.  If that's the case (and I'm
not sure), it would detract from the suggestion that the avoidance of
*stress* on "its" (required by the construction) is the "primary"
motivation (as opposed to the inanimacy of the reference) -- not that Paul
was weighing alternative motivations.  It remains tempting to somehow
relate this avoidance of *its* to the disfavoring of case-marking (= GEN)
with neuter possessives to the advantage of two alternative strategies
(which I first read about in Sapir), the prepositional possessive
construction with *of*, e.g., "the leg of the table" vs. "compounding",
e.g., "the table leg".  I wonder if "it belongs to the table" is favored
over "it's the table's", again avoiding a GEN inanimate.
 
NB  Beyond the Present-Day pale is "WHOSE is it?" for an inanimate, cf. A:
whose keys are those?
B: Mine/??My car's (but A: WHAT/WHICH keys are those?  B: My car's, cf. the
car keys, the keys TO the car, ? the car's keys).  Discomfort remains
"non-standard" for the relative "a table WHOSE legs were broken" (pref
reduction to "a table WITH broken legs" when possible, expect "a table
THAT's legs were broken" when not).  GEN, the surviving OE case inflection,
is under severe and continuing  attack in the context of inanimates.
 
NNB.  The surviving irregular plurals in English are *animates*.  They
allow "it's the (wo)men's/
children's/cattle's/mice's" etc.  "it's the girls'/boys'/horses'/etc" only
works in print, though in practice intended number is probably rarely a
problem.  Still, the surviving humans remain useful, cf. "it's the man's
room" vs. "it's the men's room".   Inanimates lose on all counts, cf. ???
it's the tables'(s).