on pronouns

[Frans Plank]

Dear David,

my hunch is that, whenever a language has relevant agreement, agreement
will reflect the primary/basic meaning of pronouns.

It's really too bad, the languages you've chosen to work on, English and
suchlike.  Consider German or French.

German DU has generic meanings like English YOU, but as subject always
controls 2SG agreement:

	Da frisst du 20 Karotten am Tag und wirst doch blind.

MAN 'one', whoever it is intended to refer to (addressee(s), perhaps
speaker included, perhaps others too), only controls 3SG agreement.  So,
its primary/basic meaning, whatever it is, is *not* 1PL or 2SG/PL.
Analogously for the personal pronoun UNSEREINER (and I leave it to you,
David, to find out what it means).

French VOUS, despite referring to a single addressee, takes 2PL verb
agreement.  Archaic German ER, also referring to a single (socially
inferior?) addressee (apart from a single non-SAP), nonetheless takes 3SG
verb:

	Wohin rennt er, Woyzeck?
	(meaning, 'Where are you running, Woyzeck?')

Interestingly, with nouns halfway grammaticalized as terms of address, you
sometimes get variation in number (though not person) agreement in German:

	Kriegen/Kriegt der Herr noch ein Bier?
	get-3PL/3SG the gentleman yet another beer?

There is a single addressee, but the term of address referring to it is
still nominal enough to control *3* person agreement (SG, appropriately);
but then it is also sufficiently pronominal, SIE-like to alternatively
control 3PL agreement.  When the semi-grammaticalized (pro-)noun of address
is too superior, I imagine only the plural agreement could be used (but
then this is not how I normally talk):

	Durchlaucht haben/??hat gerufen?
	Highness have-3PL/??3SG called?

Analogously for such speaker forms:

	Euer untertaenigster Diener erlauben/?erlaubt sich den Einwand ...
	your lowliest servant permit-3PL/?3SG 3.REFL the objection ...
	(meaning, 'I permit myself to object ...')

I've mentioned these agreement facts in passing in my 'Die Ordnung der
Personen' (Folia Linguistica 1985), but I'm sure so have others, and in
greater detail.

The point that there are languages where it may not make much sense to call
nominal forms (such as 'servant') pronouns but where there are no better
pronouns to be found, has been made by Householder in his review of
Forchheimer's The Category of Person in Language (1953), Lg 31, 1955.  I'm
sure it has also been made by many others, describing Japanese and suchlike
languages.

Best
Frans