From: stlatos
Message: 65831
Date: 2010-02-10
>There was and ev. is seen to exist for it by observing irregular nominatives, such as -n in -lainen < *-RWaitYeÑ (the nom. was uvular -qH \ -X \ -R \ -Ñ \ ->j (any uvular could turn into any other uvular by free variation) and the common -X \ -R fricatives disappeared and lengthened the V (later unstressed -V: > -V (as seen in *wodo:r > *wede:R > *wete: > vete-)) so usually there is no dif. between nom. and stem that would indicate an added morpheme in the present language).
> > < *so:-ma: is certainly not "nearly certain" in the sense of being the accepted option, nor in the sense of being problemless (no regular process based on this etymology explains the final *-i, and Finland Proper is not particularly swampy).
> >
> > > > The stem is Suoma-, and several Finnish words ending in -i
> > > > have stems with -a- (this is because of an old nom -y
> >
> > > No, words from *-Vj have stems with -i- (but may derive from a
> > > root in -a, which I suppose you meant).
> >
> > That's not what I meant. The nom. of a-stems was -a-j,
>
> There's no nominativ marker, ancient or modern.
> > For example, all words ending in -mpi '-er' have stems in -mpa-.But the superlative is itself irregular, and should hardly be analyzed as anything but -i-mpa- (that is, the superlative is formed from the comparative). I'd say it's likely the nom * -impi dis. > -imp > -im > -in.
>
> IIRC that's been analyzed as analogy of the superlativ, or as a sound change *a > i word-finally in -CCV suffixes (perhaps extensible to all CC_# environments when in 3rd or later syllable.)
> An assumption this strong certainly can explain a lot, but predicts nothing. What decides whether a word was analogized to the form without *-j or to one with it?Nothing; and no prediction is necessary (in the Romance languages, for example, analogy usually favors the Latin stem, but sometimes the nom., and can differ among them, and is completely unpredictable).
>
>What of duplicates where both variants occur as distinct roots? Why don't we see any actual roots that have an -a- stem against an -i nominativ?
>I said -mpi (and suoma- should be heavy evidence for the same alt. before analogy).
>If it walks like a suffix and it talks like a suffix, you'll need more than an ad hoc appeal to analogy to argue for it not being a suffix.
>The nom. -j was a type of suffix. What are you saying is a suffix but wasn't nom. or -j? For which word(s)?
>(this can also be corroborated against other Uralic languages), and "Suomi" belongs in the latter declension class. Perhaps an example helps:
> > Also, some Finnish words in -i correspond to other Uralic words that are a-stems (in Hungarian, etc.);
>
> Certainly the change *aj > ej > i exists. But you need to keep in mind there is an observable distinction between words with final -i that derive from *-V-j, and those that derive from original *-i
>
> *kota-j > *kotej > koti "home", gen. *kota-j-n >> kodinThose are likely related to S kut.í-s. (f) 'hut, cottage' and kud.ya-m (n) 'wall' < *kèr.- (like Greek kurtía 'wickerwork').
> (note also retention of the original root in kota, kodan "hut"; Samic *koaté)
> *käti > käsi "hand", gen. *käte-n > käden (Samic *kietë)It gives no ev. either way. There are many known variants for 'fen', and more in other IE branches, and no way of knowing which existed with the exactly equivalent meaning at the time (or why one would be chosen over others if their meanings overlapped) as the name for 'Finland' if it did mean 'Fenland' before its original meaning was forgotten.
> > but since suoma-lainen means 'Finnlander' and suo-maa-lainen would mean 'fenlander', I'm not willing to accept coincidence.
>
> Even if it's not, it doesn't mean those are correct etymologies: given "Finn" for northern barbarians in general (attested much before "Finland"), and a tribe "suomalaiset" who call fens "suo", there's a good motivation to then link the two, regardless of the actual etymologies.
>
> Correct me if I'm wrong here, but that we have no i-variants of "fen" also supports the conclusion that the association of fens and Finns is a folk etymology.
>
>It's in the Gmc languages that *fan()- meant both (and more); I used the best English equivalent rather than whatever it would have been at the stage of Gmc because I don't know exactly when borrowing occurred, etc., and for simplicity. I used E to show the coincidence, but it is not only in E that it exists.
> Did I mention yet that while "suo" is the most general word for a swamp, "fen" is not? "Suomaalainen" would *not* literally translate as "fenlander", but "swamplander". Given that "fen" rather translates as "letto", perhaps you should be arguing for an identity of Latvians and historical Finns ;)
>