Reconstruction talk:Proto-Semitic/ṭāb-

The adjective form

So there are two separate words in Proto-Semitic, @Wikitiki89? Starling lists all under ṭayb-. Palaestrator verborum (loquier) 15:38, 7 December 2017 (UTC)

I don't think ṭayyib- or ṭayb- existed at all in Proto-Semitic. The root, which was ṭ-y-b or ṭ-w-b certainly existed. The Arabic form ṭayyib- was probably an innovation that occurred within Arabic, from the root and the qattil- pattern used in similar adjectives (جَيِّد (jayyid), سَيِّد (sayyid), سَيِّئ (sayyiʔ), مَيِّت (mayyit), etc.). People are often lazy with Proto-Semitic and list forms that are related by root as being derived from the same word, and try to incorporate elided root letters into the word just to make them visible. This is of course is due to the fact that we really don't know as much about Proto-Semitic as we'd like to. --WikiTiki89 15:53, 7 December 2017 (UTC)

@Wikitiki89 What I thought is that ṭayb- existed and the more northern Semitic languages had compensatory lengthening after having (y⟩ fall off. This would explain all forms. Note that that innovation would also have to take place in Old South Arabic for which Beeston lists ṭyb as adjective “sweet-smelling”, but Old South Arabic is claimed to be less related to Arabic than Hebrew and Ugaritic (In Ethiopic there is sadly nothing to compare). With setting ṭayb there is no such problem. Palaestrator verborum (loquier) 16:07, 7 December 2017 (UTC)

But why was the y dropped? This gets too far into the realm of speculation without evidence. Regarding OSA, I think it should also be removed from this list, I just wasn't paying close attention. --WikiTiki89 16:41, 7 December 2017 (UTC)

I don’t know. But why was the Proto-Semitic word ṭāb- or ṭāyb- or however it was dropped in Arabic, and later reinvented from the Arabic root, as well as dropped and then reinvented in Old South Arabian? For such a basic concept like “good” it is easier to assume that all languages have inherited the same word.

The dropping however does not need to be assumed if one say that the word for “good” was something entirely different and all languages have invented the word by themselves. Do you know an alternative word for the meaning “good” in Proto-Semitic? In Arabic ṭayyib is now the basic one, the most abstract one, i. e. for example the first and most used one in Nabil Osman’s German-Arabic dictionary for translating “gut”.

Or isn’t the word for “good” in language trees that stable even? Proto-Germanic *gōdaz is inherited in every German language, but not from Indo-European. And in Slavic we have *dobrъ but that is not used for basic “good” in Russian. And Italic has another entirely different word and the Indo-Iranians have all kinds of words.

But if the things are unstable, a Proto-Semitic word *ṭāb- did not exist either, which is unlikely considering the spread of this word ṭāb – if the things are stable, Arabic and Old South Arabian would have inherited the word. Except there is an other word for “good”, so which is it? @Wikitiki89 Palaestrator verborum (loquier) 17:40, 7 December 2017 (UTC)

Who said it was "dropped in Arabic, and later reinvented from the Arabic root"? The two forms could have existed in parallel until ṭāb- died out later. And there is a very good reason to say that this happened: the fact that this happened with a large majority of Arabic adjectives. Most Arabic adjectives, that correspond to various forms in other Semitic languages, in Arabic all have the same form: qatīl-. In hollow roots it seems that the patter qattil- was used instead. Aramaic did this to a lesser extend with the qaṭṭīl pattern. Some examples: Hebrew has ḥāḏāš < *ḥadaθ-, qārōḇ < qarub-, ʿāṣūm- < *ʿaθ̣ūm-, raḥūm < *raḥ(ḥ)ūm-, while Arabic has ḥadīṯ-, qarīb-, ʿaẓīm-, raḥīm-.

Arabic also shares many features with OSA. Even though they are not genetically related, they must have had a good amount of contact. I don't know much specifically about OSA, so I can't give you many examples, but one example is the "broken plural" system. It could be an innovation or a retention, who knows.

Phonologically *ṭāb- < *ṭayab- could make sense, but not *ṭāb- < ṭayb- or *ṭayab- > *ṭayyib-. --WikiTiki89 19:29, 7 December 2017 (UTC)

@Wikitiki89. Thanks, I understand this; but isn’t in that case of parallel existence of two forms until the older one died out the form ṭayyib a witness of the Proto-Semitic, i. e. we have to list the Arabic as descendant of ṭāb here? As this being from earlier ṭāb in analogy to other adjectives resp. pursuant to new rules. If that generally happens … Palaestrator verborum (loquier) 20:16, 7 December 2017 (UTC)

I mean that's when it gets tricky and we just have to decide what are rules are about this. If we list it, we have to note that it was adapted to a different pattern and is not a direct descendent. --WikiTiki89 20:24, 7 December 2017 (UTC)

@Wikitiki89 That sounds fine. We have |der=1 in {{desc}} for this: “Use this when the term derives by the addition of a morpheme (affix) or by analogical levelling, rather than by regular sound change.” The analogical levelling seems to be a regular sound change, even better. (Or have I misused a technical term “analogical levelling”?)

So we have a regular sound change and semantic reasons why the word has not died out and layout means for showing the deviation. Looks like nothing stands against counting the Arabic to ṭāb. Palaestrator verborum (loquier) 21:05, 7 December 2017 (UTC)

Yes, you could call this analogical leveling. The analogical leveling is not a sound change, but a morphological change, so you can't call it a regular sound change. Also, it's not actually regular, even though it is common. But yes, we could list it with that. However, the OSA case is trickier because we don't really know exactly what kind of form it represents (like whether it is also ṭayyib or something else).

The only problem with using |der=1 is that it formats it in a way that looks like what we use for borrowings. --WikiTiki89 21:28, 7 December 2017 (UTC)

The formatting of the parameter is of course something that you can mention at its place – the intention of the introduction of the parameter at least seems to fit here. About doubts in form, I don’t know: What is the best practice to note words that should be noted but have some quirks in their being related to the common stem? There would be parameters helpful to state in a normalized way that some derivations are doubtful. As the Akkadian and Eblaite at *ʿigl-. Those words are typically adduced and shall be so here too but cannot be claimed to be descendants because there are questions. I am disconcerted about having to list forms not at all being able to discuss each. @Wikitiki89 Palaestrator verborum (loquier) 22:05, 7 December 2017 (UTC)

I did actually bring it up here. And we can in fact add comments to descendants. --WikiTiki89 22:14, 7 December 2017 (UTC)

The noun

I have also a noun, @Wikitiki89. Which form has to be set?

==Proto-Semitic==

===Noun===
{{head|sem-pro|noun}}

# {{l|en|goodness}}
# {{l|en|good}} substance

====Descendants====
=====East Semitic=====
* {{desc|akk|tr=ṭābtu|t=goodness; salt}}

=====West Semitic=====
* {{desc|ar|طِيب|t=goodness; perfume}}
* {{desc|he|טוּב|tr=ṭuḇ|t=goodness}}

=====South Semitic=====
* {{desc|shv|ṭíb|t=gold}}
* {{desc|sem-srb|𐩷𐩺𐩨|tr=ṭyb|t=incense}}

Palaestrator verborum (loquier) 15:50, 7 December 2017 (UTC)

I would say each language created their own nominalized version of the adjective or the root. I don't know if we can reconstruct a single Proto-Semitic one. --WikiTiki89 15:54, 7 December 2017 (UTC)

I find this also a reasonable conclusion. But the form CīC/CiyC is somewhat unexpected for own creations in Arabic. But maybe it is borrowed from South Semitic – the meaning is too often specialized. Palaestrator verborum (loquier) 16:07, 7 December 2017 (UTC)

What do we know about verbal nouns in Proto-Semitic? The Semitic languages generally form them, and it would be unbalanced to claim of all of them that they are independently formed. I have thought this about *ḥapr- where *ḥapar- surely existed. Palaestrator verborum (loquier) 16:10, 7 December 2017 (UTC)

It's not that all verbal nouns are necessarily independently formed, but the ones that had different forms must have been independently formed. I would not have created an entry for *ḥapr- given its current contents. It's not uncommon for languages to change verbal nouns from one form to another, this is why so many languages have so many different forms for verbal nouns. --WikiTiki89 16:45, 7 December 2017 (UTC)