>
> On second thought, you must be talking about hypercorrection, I
> guess:
> "Don't say <lembir>, say <lamb> (etc.)" leads to the belief that if
> <lamb> is correct, then all words that were the same in the plural,
> like <wort>, must be corrected forms for original forms with <-ir>,
> thus a rebellious folk starts to believe that the folksy forms must
> be forms with <-ir>, so rebelliously they start adding <-ir> to all
> nouns that were identical in the plural. So instead of retaining
> the inherited plural in only a few nouns, they come to believe that
> this was the true inherited plural in many other nouns, even though
> it was not, and therefore adopt it out of a sense of purism or
> desire not to be corrected. Is this more or less what you are
> saying? It sounds so complicated and therefore unlikely to me. I
> prefer my idea of a simple desire to have a distinct plural form
> for neuter nouns (and any nouns that were identical to the singular
> in the plural), so they adopted and dispersed the s-stem plural
> <-ir>. Maybe someone has written a paper that addresses this
> question?
Actually what you describe is similar to what happened in Danish,
among the neuter plurals which were endingless in ON, the plural
ending -e has spread (et hus, to huse, cf Sw. ett hus, två hus),
mostly by back formation from the def. pl. husene, but also because
it's practical. In eg. 'tog' "train" the plural -e is spreading, but
still isn't considered correct, which means that a loudspeaker warning
'Der kommer tog!' "Train(s) is/are coming!" might referring to one or
several trains. I used to argue with my younger brother about it,
being the correct one (and he is now an engineer by
profession).
But it was not what I had in mind, I was talking about the s-plural of
the a-stems, the thematic stems, the one that became the standard for
almost all nouns in English and for those words in Dutch which are
bisyllabic, ending in -en, -el, -er, or which are foreign (mostly
French); the ending which has become almost obliterated in German,
with exception of foreign words and 'low' words that are marked as
specifically Northern and part of a see culture (eg. Jungs "mates").
Now my idea was that the -s plural in England was felt to be a mark of
the foreigner who spoke bad middle English and messed up his
conjugations, specifically the French, since generalizing the standard
case is what you do when you speak a foreign language badly. Thus in
German, the a-stem s-plural would be so stigmatized by this tendency
that people avoided it, to the point of choosing the complicated,
perceived correct form whenever they were in doubt.
Suppose 'ain't I' for 'aren't I' spread in English, some people might
come up with *'amn't I'. That's the type of hypercorrection I'm
talking about.
Torsten