I'm told

[bwald]

----------------------------Original message----------------------------
Again I am indebted to Cindy for her further comments.  At this point, my
responses to her latest comments are simply to explain the reasons for my
interest -- without asking for or expecting further comment by Cindy.
 
>It sounds from these comments as though Benji has accepted the tradional
>view that the new passives were caused by a reanalysis of a fronted dative
>as nominative-or at least that the dative was still fronted when change to
>nominative took place.
 
The view strikes me as having some plausibility as a favoring factor (the
preposed position of the DAT in a "subject" position).  I do not take for
granted that it is the only favoring factor.  I perceive that Cindy comes
from the view that a stronger hypothesis identifying a SINGLE factor as
*decisive* in precipitating a change is the most useful one, because of its
maximised vulnerability to disconfirmation.  I see the point of this
approach, but I find myself less capable of committing to such a strong
hypothesis on the basis of what I know (and may not know).  In short, for
me it's not (yet) a matter of EITHER-OR (but NOT both -- or neither?).  Of
course, the issue is growing association of NOM with the grammatical
category called "subject", rather than directly with a set of semantic
roles.
 
She finishes this comment with:
 
I have tried to explain why I don't think that is
>the case.  As far as I can see, there is absolutely no difficulty sorting
>out the timing of verbal agreement.  I can only refer interested parties to
>chapter 9 of my book again.
 
That's sufficient for me.  I want to see if number agreement precedes DAT >
NOM.  I think there are some similar phenomena with number agreement in
Attic (Classical Greek) impersonal constructions (DAT allows number
agreement but the gender of the verbal noun remains neuter).
I wonder if at one time in OE > ME,  NOM was a sufficient but not necessary
condition for verb agreement toward the development of the indirect
passive.  I will certainly consult Chapter 9 of Cindy's book.
 
Later with regard to IO > SBJ (= ACC/DAT > NOM w pronouns) for the verb
"please" in early Modern English (or whatever you call it), Cindy comments:
 
>The 'reanalysis' is certainly too late to be lumped in with the cases where a
>preposed dative starts showing up as a nominative.  My basic point was that
>this could not have been an ambiguity-driven reanalysis.
 
By lexical diffusion of the process I allow it to continue over a long
period of time, jumping from one verb to another, which is necessarily
later to receive it.  I separate this from the fact that, as Cindy says,
there is an early period when the process is very common and is supposed to
be indicative of the 'reanalysis' shifts taking place in case-marking and
subject position.  I did not respond to her basic point about
ambiguity-driven reanalysis, because I was not attracted to such a
superficial hypothesis in the first place.  Again, in my view, ambiguity
could have been a factor favoring the change, but it seems like a minor and
hardly decisive factor in view of the other factors that have been proposed
-- by Cindy no less than by others.
 
>As for what the OED says-it is true that I oversimplified matters by not
>mentioning the brief life of nominative experiencers with 'please' followed
>by an infinitive, but anyone looking at the sources that I referred Benji
>to will know that I discuss these.
 
I hope to read this article soon, though I consider what I said about
"please" a minor point in the larger scheme of things.  I mainly meant to
suggest that certain lexico-grammatical processes involving verbs have
continued to operate over a long period of time, maybe showing that their
energy (or impetus) is not spent as quickly as grammatical changes which
encourage them.
 
Her comments on the OED are well-taken:
>I also discuss why the OED is simply
>wrong about 'please'.  The OED is a wonderful tool, but a necessary
>limitations of dictionaries, however good, is that they cannot give a
>really systematic account of the history of any lexeme because they cannot
>show how the lexeme is embedded in the grammatical system of the time.
 
That is quite true.  There are, of course, problems in its (traditional)
classifications and its dependence on texts for dating.  It's the problems
in the traditional classifications it uses and its necessary economy in
selecting grammatical contexts that prevents its examples from providing
sufficient information about their embedding in a grammatical system at any
point in time.  They merely cover a few contexts that any analysis does
well to consider, and suggest things that must be checked out by more
detailed information than the OED or any other dictionary will ever offer
-- because dictionaries are not detailed grammatical descriptions.
 
She continues:
The
>OED is a terrific place to start a historical investigation in English, but
>a bad place to stop.  I hope Benji will look at the evidence which I have
>presented concerning this matter.
 
I agree whole-heartedly.  Since I am not a specialist in the textual
history of English, I rely on scholars like Cindy to tell me what I want to
know, or to introduce me to a new hypothesis.
 
She notes:
>I am pleased that Benji and I agree in believing that case marking and
>grammatical relations must be kept separate.  But I have to correct the
>comment about 'deem'.  There has been no shift DAT>NOM here, because the
>'deemer' (judger) was always nominative in OE.  The judged thing, not the
>experiencer, was dative.
 
Right.  I was too superficial in grouping 'deem' with the others.  'deem'
is about the indirect passive as impersonal passive, i.e., DAT> NOM in that
context.  I think 'deem' is like 'help', where it wouldn't occur to English
speakers today to consider the object of 'help' a "recipient" rather than a
"direct object" -- since there' s no prepositional equivalent to the
unmarked object, e.g., I helped (*to/for) him.  Nevertheless, the pattern
for a verb like 'help' in many older I-E languages is to mark this object
as DAT, as if the verb were equivalent to English 'give-help', in which
case a preposition occurs as expected in English, since the NP 'help'
evidently fills the "object position" (as if a "theme" ACC were a "cognate
accusative", e.g., we gave/*helped help *(to) them = we gave/*helped (?to)
them help.)  'deem' presumably meant something like
'make/pass-judgment-on', where 'on' has somehow come to express the earlier
DAT role in this context.  Again, it does not seem to be simply an
arbitrary lexical matter in the relevant time period, but motivated by the
roles covered by DAT.  I could expect disagreement here, since Cindy could
insist that the change involved grammatical categories only, and that
semantic roles are not directly relevant.
 
Cindy ends with an interesting comment, which I can appreciate for its
principles.
 
>I think that we have different views of what a list
>like this is for.  It is great for raising questions such as the one which
>sparked this discussion and for finding out where to go to read what has
>been published on the subject, or for getting information that is not in
>print.  But I think that once someone has been directed to literature on a
>topic (such as the history of please) they ought to read it so they have
>all the evidence at their before making further comments on the subject.
 
I think it is OK to negotiate that with any particular user of the list,
because both time and interest are legitimate factors that might work at
cross-purposes in the real world.  The things I say on the list are usually
and admittedly off the top of my head.  However, in the case of the
indirect passive I found myself motivated to check some older standard
references that I had readily on hand.   One question leads to another.
(In fact, one of my ultimate interests in this topic has to do with how
facts about other languages might improve my understanding of the various
ways in which the Bantu passive has evolved, and my ability to reconstruct
those paths of change.)
 
As I said, it is negotiable whether a subscriber wants to stop at
recommending relevant literature or not, and all choices are acceptable.
However, maybe I should say that I would not go so far as to take for
granted discussing ON THE LIST with any authors detailed points of their
articles -- except if a review were expected by the list (this list doesn't
do that, and I'm not volunteering).  That, I think, is more formal than
what this list is designed for, and what it is currently expected to
accomplish.  By the same token, for example, the ling.list has even
experimented with conferences (I have not participated), but this list has
not, and does not currently have the person-power to organise such an
event.  In any case, I appreciate that Cindy was willing to take her
comments as far as she did.
 
She concludes:
>If someone reads
>my arguments and wants to say why they don't find them convincing, that's a
>different matter and I'll be glad to respond.
 
That's a gracious invitation.  As I said, I would first be inclined to
contact Cindy off-list (or any author who invites me to read their work).
That might simply be because I don't initiate discussions on the list, but
respond to the ones that attract my interest.  That way I know I'm not the
only one who's interested.  I treat the list something like a newspaper
(since it approaches "daily" change).  Topics generally don't go as deep or
last as long as the specialised interests of any reader, but there's a lot
of suggestive information for specialised concerns to pursue, and
discussion sometimes goes in unexpected directions.  (I, for one, did not
expect to have some of the ideas I now have when I first started reading
the "I'm told" discussion.)  Thanks again to Cindy, and I will continue to
value her expertise as a resource both in print and e-mail.  -- Benji