What is Relatedness?

[Larry Trask]

Rick Mc Callister writes:

[somebody else]

>> Hasn't Frisian also generalised a plural in -ar or -er, which also
>> corresponds with what Norwegian and Swedish (and formerly Danish) did?

>       Can you elaborate a bit more on the -ar/-er plural?
>  While the English strong plural was -en,
>  there seems to be some relic forms with -r-; actually -ren
>       e.g. child-r-en, eyren Middle English dialect plural for <eggs>
>  [unless the -r- of eyren was part of the singular stem and perhaps lost in
>  the presumed singular <ey>]

Old English, like a good Germanic language, had many different classes of
nouns exhibiting different sets of inflectional endings.  So far as I know, a
nominative plural in /r/ was found only in a rather small class of neuter
nouns which formed their plurals in /-ru/ or sometimes in /-ra/.  One of
these was /æg/ 'egg', whose plural was /ægru/ or /ægra/.  This plural
developed regularly to Middle English <eyer> (and variants).  But the rarity
of the r-plurals eventually allowed this to be assimilated into the (then)
much larger class of n-plurals, and so a new double plural <ey(e)ren> was
created.  This continued in use in the south of England until the northern
form <eggs> (of Scandinavian origin) displaced it.

And 'children' is rather similar.  In OE, this was a masculine noun with
plural <cild>, but the word went walkabout and wandered into two or three
other declensional classes.  Most particularly, it entered the 'egg' class
and acquired a new plural <cildru> or <cildra>.  This gave rise regularly to
ME <cildre> or <childer>, which again finally acquired an n-ending to yield
the double plural <child(e)ren>.

English is not alone here.  Roger Lass cites similar examples from other
Germanic languages: Dutch <ei> 'egg', plural <ei-er-en>; Afrikaans <eier>
'egg' (with incorporation of the old plural marker into the stem), plural
<eier-s>; Dutch <kind> 'child', plural <kind-er-en>; Afrikaans <kinder>
'child', plural <kinder-s>.  Compare standard German <Kind> 'child', plural
<Kind-er>.

So far as I know, the OE /-ru/ plurals are not cognate with the Scandinavian
/r/ plurals.  The OE forms appear to derive from PIE nouns with stems in
*/-es/ ~ */-os/.  The addition of a vocalic plural suffix allowed the */s/ of
the stem to undergo first Verner's Law (to */z/) and then rhotacism, yielding
the observed /-rV/.  But the Scandinavian /r/ plurals, on the whole, appear
to derive from PIE */-Vs/ (note Greek /-os/ and Latin /-us/, for example).
This became */-Vz/ in Proto-Germanic, and then /-r/ in Scandinavian, but the
final consonant was lost entirely in Old English.

But I'm no Germanist, and I'll be happy to be corrected if I've screwed this
up.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk