IE languages: consonant shift "t(t)>n"

[Whodunit]

Hello,

in one post of another thread, I claimed that in quite early times, the consonant shift t(t)>n must have taken place between some IE languages. The first evidence for this could be the development of our international Italian word piano with piatto (along with English plane and plate, German planieren and Platt[düütsch], French plan and plat, Czech pláň and plátek, Lithuanian planas and platus... the list is almost endless), which I would trace as follows:



IE *plāt- >

(1) AGr πλᾰτύς > VLat *plattus > RomLang. F plat-/I piatt/S not recorded/P chat-​

(2) ClassL plānus > RomLang. F plan-/I pian-/S llan-/P chã-​

(3) GermLang. *flata-/*plata > Goth. plat- > NHG flach, platt​

How we come from p-l-k to p-l-t should be another story, since there was placnus in Old Latin, flōkan in Gothic etc. It would be too long to include this development into my hypothesis here.

Anyway, do you think that there was a consonant shift that changed the voiceless plosive [p] into the nasal consonant [n]? I can't find any eveidence for this in any etymological dictionary, but I can try to prove it by examples like plāt- > piano and (from the other thread; see link above) *māter > nun.

Let's see what you have to say.

 

[Frank06]

Hi,

Not really a reply (I am not at home right now), but I am wondering what you meant by:

PIE *pelək- > IE *plāt- >

How do you see the difference PIE versus IE?

I can't look things up at this very moment, but wouldn't it be better to talk about an alternation t - n rather than a shift?

Groetjes,

Frank

 

[Whodunit]

How do you see the difference PIE versus IE?

Before I can answer the question, I should tell you that my *pelək was wrong (I will edit it). The correct root is *plāt only, since *pelək is the root for damp.

I just though *plāt (=flat) developed from *pelək, but that was wrong, so I chose PIE (apparently too carelessly ).

I can't look things up at this very moment, but wouldn't it be better to talk about an alternation t - n rather than a shift?

What is the difference between alternation and shift? In other words, what's the difference between the Germanic shift p>f as opposed to the Modern (?) Italian (not Italic!) alternation pl>pi?

 

[Frank06]

Hi,

Hello,

IE *plāt- >

I am still not sure what you by PIE as opposed (?) to IE, but I think the confusion starts with the PIE root *plāt- you suggest.

As far as I understand this explanation, the basic PIE root is *pelə2. *plā-no is a suffixed form of the variant *pleə2, while *plat- (and alternative *plad-) is an extension of *pelə2. Otherwise said, I still fail to see the shift. [edit]I mean the consonant shift t>n you suggested[/edit].
As for the examples you gave in this thread: I have the strong impression these words are derived from the different roots, or rather root+extension (*pla-t) and root+suffix (*pla-no), and that they concern loans, reloans and the like.

Groetjes,

Frank

 

[Whodunit]

Hi,

I am still not sure what you by PIE as opposed (?) to IE, but I think the confusion starts with the PIE root *plāt- you suggest.

Yes, it would be hard to explain, especially now that I had made a mistake!

As far as I understand this explanation, the basic PIE root is *pelə2. *plā-no is a suffixed form of the variant *pleə2, while *plat- (and alternative *plad-) is an extension of *pelə2. Otherwise said, I still fail to see the shift. [edit]I mean the consonant shift t>n you suggested[/edit].

Interesting. However, why were those suffixes added? A suffix is never added unintentionally (maybe every now and then in one language, but not in nearly all languages of the same language family), it has to signify something. I can't see a meaning in -no. And why should a word be extended without any reason (at least I can't see any)?

As for the examples you gave: I have the strong impression these words are derived from the different roots, or rather root+extension (*pla-t) and root+suffix (*pla-no).

Again, here are my two examples:

plane/plate (within one language; might have developed independently)
*mâtér (IE)/anna (Hittite) (if anna is really a descendant of *mâtér, some sound shift must have taken place; at least I can't find any root for English nun, which I assume could be *mâtér)

I will keep my eyes open for more examples with a possible t>n sift.

 

[Frank06]

Hi,

Interesting. However, why were those suffixes added? A suffix is never added unintentionally (maybe every now and then in one language, but not in nearly all languages of the same language family), it has to signify something. I can't see a meaning in -no. And why should a word be extended without any reason (at least I can't see any)?

The suffix -no was not added in all those languages separatedly, our suffix -no is part of a PIE stem or word.
What we find in a PIE dictionary is the PIE root, and a PIE root hardly ever equals a PIE word. That root is a rather abstract 'entity' with a very abstract meaning. In PIE, roots as such weren't used in speech (roots are abstractions made by scholars and academics), they got extended by one or more suffixes (=stem), and depending on the wordclass, with endings. So, it's not the word that gets extended, but the root. And that's a crucial difference. See here (paragraph 37). I can't find an exhaustive list of PIE suffixes, but here (pdf) you can find a few of them and some further explanation.

*mâtér (IE)/anna (Hittite) (if anna is really a descendant of *mâtér, some sound shift must have taken place

I don't know a lot about Hittite, but I would be highly surprised if Hit. anna- would go back to PIE *mater. It would involve some quite spectacular sound changes. But again, I don't know a lot about Hittite. Can you give some further explanation?

at least I can't find any root for English nun, which I assume could be *mâtér)

Sorry, but I don't understand this part. What do you mean by this?

Groetjes,

Frank

 

[Whodunit]

What we find in a PIE dictionary is the PIE root, and a PIE root hardly ever equals a PIE word. That root is a rather abstract 'entity' with a very abstract meaning. [...]

I know that. However, I don't think suffixes were added unintentionally (if a language like PIE had ever existed!). In German (you should know that from Dutch), suffixes like -lich, -keit, -ung, and prefixes like er-, be-, aus- etc. have a specific meaning (most of them have more than one). If we can find the one for -no and -to, too I might give up on this point and withdraw my assumption about a possible shift (or alternation, extension or whatever).

See here (paragraph 37). I can't find an exhaustive list of PIE suffixes, but here (pdf) you can find a few of them and some further explanation.

If I can find some time, I will read those pages. In fact, they look very interesting.

I don't know a lot about Hittite, but I would be highly surprised if Hit. anna- would go back to PIE *mater. It would involve some quite spectacular sound changes. But again, I don't know a lot about Hittite. Can you give some further explanation?

Here's the explanation I published in another thread (just a hypothesis, though):

According to this, IE *māter might have developed to Proto-Hittite (?) *matta>*mana, from which the m- was later dropped. See Serbo-Croation nena "mother", Sanskrit nona "mother", Persian nana "mother", Greek ναννα (?) "aunt", English nun and nanny, German Nonne "nun" ...

Sorry, but I don't understand this part. What do you mean by this?

I haven't been able to find a common root for German Nonne, English nun, Greek nanna, (for more languages, see here), although all of them mean something like mother (in church). Couldn't they be derived from *mâtér?

 

[Frank06]

Hi,

If we can find the one for -no and -to, too I might give up on this point and withdraw my assumption about a possible shift (or alternation, extension or whatever).

I keep searching for a list.

According to this, IE *māter might have developed to Proto-Hittite (?) *matta>*mana, from which the m- was later dropped. See Serbo-Croation nena "mother", Sanskrit nona "mother", Persian nana "mother", Greek ναννα (?) "aunt", English nun and nanny, German Nonne "nun" ...

I found three PIE roots here:
1. PIE *nan-/*nen: with exactly the same series of examples you mention nana, nona, etc), except Hittite.
2. PIE *an- (grandmother): from which Hittite hanna- and anna- (mother) are derived (and Latin anus, old woman, and German Ahn)
3. PIE *mā-t-er-
If I understand well, the explanations given in that database rule out PIE *mater- > ...> Hit. anna.
But please doublecheck, and if you come up with other data, please post them!

Take good care,

Frank

 

[Whodunit]

I keep searching for a list.

So do I.

I found three PIE roots here:

Could you help me with these databases, please*? I never happen to find anything useful there unless you direct me to the correct entries.

1. PIE *nan-/*nen: with exactly the same series of examples you mention nana, nona, etc), except Hittite.

Assumption: Hittite is not an Indo-European language. It may even be an isolated language, since even in the other language families, I can find a similar word (to the PIE *mâtér) for mother, but not any in Hittite.

2. PIE *an- (grandmother): from which Hittite hanna- and anna- (mother) are derived (and Latin anus, old woman, and German Ahn)

Interestingly, my cousins call my grandmother Oma Nanna (but only because her name is Johanna).

3. PIE *mā-t-er-
If I understand well, the explanations given in that database rule out PIE *mater- > ...> Hit. anna.

Are you sure? I can't find any connection between them in the database, but maybe I haven't had the closest look at it yet. However, if mater and anna are really connected, my earlier assumptions might be true.

But please doublecheck, and if you come up with other data, please post them!

Pokorny's is down again. And Koebler's doesn't include Hittite at all. Well, I'll keep my eyes open.

*I will send you a PM later. Maybe tonight or tomorrow.

 

[Frank06]

Hey,

Could you help me with these databases, please*? I never happen to find anything useful there unless you direct me to the correct entries.

You're really asking the right guy! Those databases surely do have a character of their own. What I do in this one is to select the encoding (utf-8), to toggle on the options under "include in report", give the 'meaning' in English. The rest of the butons and fields (query options, whole words) I don't touch at all. Then I press the 'search' button and then... well, it's a matter of good luck: I hope something pops up. It took me (well, still takes me) quite some time to get useful data out of it.

Assumption: Hittite is not an Indo-European language. It may even be an isolated language, since even in the other language families, I can find a similar word (to the PIE *mâtér) for mother, but not any in Hittite.

The Hittite vocabulary that can be connected with PIE is rather limited, so I don't think that the lack of one word (i.c. *mater) should lead us to dramatic conclusions. In this context, I want to point out that Hit. anna- seems to go back to PIE any which way.
Babaev states that "[t]he dictionary of Hittite has got only about 22% of purely Indo-European words, other ones are of Hatti and other origins." Other estimates I read about also range between 1/4 and 1/5.

Hittite is normally included in the IE family (there have been other ideas about the status of the Anatolian languages), but that inclusion is also based upon sound changes and grammatical features. It also played a mayor role in the acceptance of the theory of PIE laryngeals. More information here.

I can't find any connection between them in the database, but maybe I haven't had the closest look at it yet. However, if mater and anna are really connected, my earlier assumptions might be true.

I can't find anything in Beekes or Szemerényi that could justify PIE *mater > hit. anna.

Groetjes,

Frank