Last edit: 21:54, 28 May 2014
I've noticed that, in terms of the Proto-Indo-European demonstratives, only the masculine and feminine nominative singular forms begin with *s (*só and *séh2, respectively) and that all other forms begin with *t (e.g. the corresponding accusatives *tóm and *téh2m). Could these forms have possibly derived from earlier **stó and **stéh2, which came from original *t- forms via s-mobile?
I doubt it, but I am not an Indo-European linguist so I can't really say more. I do know that a form like *stó would not have lost its s-, and so it really just shifts the problem rather than solve it: if those two forms had *s- why didn't the other forms, and if all of them had *s- why did those two forms keep it while the others lost it? I think the more likely explanation is suppletion, which isn't an uncommon thing with pronouns. The modern Slavic pronouns for example have exactly the same kind of suppletion, with derivatives of *onъ in the nominative but descendants of *jь in all the other cases. This suppletion didn't exist in early Proto-Slavic but developed during the late Common Slavic period, and is still attested in an incomplete way in OCS, so we know it was an innovation and not an archaism. And presumably if this could happen in Slavic 1500 years ago, it could happen to PIE 5000 years ago too.
I was thinking that the two forms with *s would have triggered s-mobile because of their frequency. Alternatively, the original onset could have been *st-, and then only the forms which most frequently occurred at the beginning of the sentence (thereby having no legatamente preceding *-s) preserved this fricative onset, while the rest became deleted (similar to *h1es-si > *h1esi). However, this is just my guess. Also, I don't think I fully understand suppletion. Could you explain how it affects pronouns?
Ok, now I understand. So, do you think there were two Pre-Indo-European demonstratives *só/séh2/sód and *tó/téh2/tód that came to form a single paradigm?
I don't know. *só is kind of odd to begin with because it doesn't have the normal animate nominative ending -s. Some other pronouns also lack it. It's likely that it is a very old word, maybe a holdover from a time before the nominative case existed in its later form. But what the language was like at that stage is anyone's guess.
That's another point; I often wonder where inflectional languages get their inflections, and I tend to convince myself that they developed from more ancient, agglutinative languages. For example, it could be that, in Pre-Indo-European, *só and the nominative ending *-s both contained the morpheme *-s- (e.g. in Ubykh (agglutinative) the same morpheme indicates the first person singular).
I think that the case endings were formed at different times, but the nominative and accusative are probably very old. There is a hypothesis called the Nostratic theory, which doesn't really have much support among linguists, that suggests that the Indo-European and Uralic nominatives are cognate, since they both end in -m.
There are also some possibilities for internal reconstruction, though. It is often suggested that the accusative plural ending -ns was formed from the singular -m with an additional plural ending -s, in which -ms was assimilated to -ns. The ablative ending -ead is thought to have been formed from the thematic vowel -e followed by an adverbial particle *ad (modern English at). Other case endings may have similar explanations, but the more "basic" cases seem harder to explain.
Finnish has a very rich set of endings but some of them are clearly related:
- partitive -(t)a
- essive -na
- inessive -ssa
- elative -sta
- adessive -lla
- ablative -lta
It is generally believed that the latter four were formed from an infix -s- or -l- of some origin, suffixed by the old partitive and essive cases, which originally were ablative and locative in meaning: -s-na > -ssa, -s-ta, -l-na > -lla, -l-ta.
(Did you mean: "that suggests that the Indo-European and Uralic accusatives are cognate, since they both end in -m."?)
Yep, I've heard all about macro-language-family theories, though I don't believe nearly any of them are true. My personal favorite (the one I find most interesting) is Dené-Caucasian, combining Basque, Sino-Tibetan, Dené-Yenisian, and many others. As for Nostratic/Eurasiatic, I believe that the following language families were for a time in very close contact, even when grammatical elements were forming:
- Proto-Pontic/Proto-Northwest Caucasian,
- Proto-Caspic/Proto-Northeast Caucasian,
- Proto-Lyndian, and
However, I believe that none of them are genetically or fundamentally derived from a single proto-language. I have heard of, and I do follow, the internal reconstructions you mention, both in Finnish and in PIE.
Hi again (sorry for bothering you so much), Can you give me some examples of third-person singular neuter pronouns in some Germanic languages (preferably more ancient ones)? Thanks,
Yes, but I forgot to add "personal" to my classification. It seems that *iz is the only one out of those widely used as an "animate" or "personal" pronoun. Is this correct?
The personal pronouns have four sources. West Germanic used *iz and *hiz, Gothic used *iz, while Old Norse used a combination of *sa/þ- and *hanaz which is unattested anywhere else.
Ok, so if you had to guess (or if you know), what would be the hypothetical reflexes of the masculine declension of *iz into Old English (and then to Modern English)?
Well if *hiz became hē, then presumably *iz would have developed into just ē?
I was thinking the same thing, but I thought that if such forms existed side by side then they would develop phonological changes to better distinguish them than the sole presence of the *h would have.
Either that or one or the other would have fallen out of use. Which is... well, what happened. :P
Unfortunately, yes :)
But as for the former, do you know what that process is called? I'm having trouble thinking of known examples.
I'm not sure if I know of any examples. When two words fall together and can no longer reliably distinguished, then either that single word takes on the meaning of both, or the old meaning is retained and new word is used to cover the gap. I'm not aware of a name for instances where another word was changed to make it less similar to another.
Can you give me an example of when a new word is used to cover the gap?
I see. So, if a phonological change happened to **ē to make it dissimilar, could you guess what it potentially would become?
Phonetic changes almost never take such considerations into account. They are kind of at a lower, more basic level, and are not normally affected by the meanings of words. It does sometimes happen, though, that when sound changes make certain words too similar, a language borrows an alternative form from another dialect or even another language. they was borrowed from Old Norse in that way, to replace the original word which had become too similar to he.