Biblical Hebrew: רָאשָׁיו

[Ali Smith]

שלום

וַיִּקְרָ֤א יְהוֹשֻׁ֙עַ֙ לְכׇל־יִשְׂרָאֵ֔ל לִזְקֵנָיו֙ וּלְרָאשָׁ֔יו וּלְשֹׁפְטָ֖יו וּלְשֹׁטְרָ֑יו וַיֹּ֣אמֶר אֲלֵהֶ֔ם אֲנִ֣י זָקַ֔נְתִּי בָּ֖אתִי בַּיָּמִֽים׃

(יהושע כג ב)

What does רָאשָׁ֔יו mean and is it pronounced rashaw or roshaw? I mean, is the vowel a קמץ גדול or a קמץ חטוף?

אני מודה לכם מאוד

 

[Drink]

The plural of רֹאשׁ (rosh) is רָאשִׁים (rashim).

In context ראשיו refers to Israel's leaders.

 

[Ali Smith]

Thanks! But given the fact that the vowel was originally an o-class vowel (in the singular), wouldn't it make more sense for us to interpret it as a קמץ חטוף?

 

[Drink]

No, because that would make no sense. Why would it be shortened?

Rather it's the opposite. The word was originally ra’š-, but the singular form went through the Canaanite shift.

 

[Ali Smith]

Thanks! Can you tell me more about the Canaanite shift?

 

[Drink]

Long "a" became long "o" in Canaanite languages.

Examples:
- PS *šalām- > Hebrew šālōm (contrast Aramaic šəlām(ā), Arabic salām-)
- PS *kātib- > Hebrew kōṯēḇ (contrast Aramaic kāṯiḇ/kāṯeḇ/kāṯēḇ, Arabic kātib-)

It's worth noting that the qamatz qatan usually only comes from an original short "u", not from a former long "a" or diphthong "aw". The only time I know of that a Canaanite-shifted long "a" becomes a qamatz qatan is in the forms of the word שלש, such as שְׁלָשְׁתָּם.

 

[Ali Smith]

Thanks! But why didn't the Canaanite shift occur in רָאשָׁיו or, for that matter, the plural absolute, which is רָאשִׁים?

 

[Drink]

It's an open question, but in my opinion the most plausible explanation is that the א of the plural had not been elided yet at the time of the Canaanite shift.

 

[utopia]

Thanks! But why didn't the Canaanite shift occur in רָאשָׁיו or, for that matter, the plural absolute, which is רָאשִׁים?

Well, the singular was RA'SHU - and the A'> LONG O --- ROOSH

The plural was RA'SHUMA (or RA'SHIMA) and thus the stress was on the plural sign - UMA, IMA.

And the A' was not stressd. According to Yehoshua Blau, it's a prerequisite of the Canaanite shift.

 

[Ali Smith]

Well, the singular was RA'SHU - and the A'> LONG O --- ROOSH

The plural was RA'SHUMA (or RA'SHIMA) and thus the stress was on the plural sign - UMA, IMA.

And the A' was not stressd. According to Yehoshua Blau, it's a prerequisite of the Canaanite shift.

Thank you! Could you please tell us where Dr. Yehoshua Blau said this?

 

[utopia]

Thank you! Could you please tell us where Dr. Yehoshua Blau said this?

In his book, but I found a site about reconstructed ancient Hebrew, that gives you lots of Blaus’ insights.

Just write reconstruction of ancient Hebrew and find it.

 

[Ali Smith]

Got it. Thanks. However, the fact is that the Canaanite shift occurred in words like kātib, resulting in כּוֹתֵב. Note that the /ā/ was not stressed here either.

 

[utopia]

The original word was Kaatibu

The aa was stressed because it was long. Then it changed to KOOTIBU, stressed on KOO, then the stress shifted to kooTIbu, then the final sound was dropped, changing to kooTEv

 

[Drink]

And the A' was not stressd. According to Yehoshua Blau, it's a prerequisite of the Canaanite shift.

This is one of the old theories, however it fails to account for the fact that there are many, many counterexamples where an unstressed vowel went through the Canaanite shift (like the plurals of almost any other noun that went through the Canaanite shift). In fact, this seems to be the rare exception that didn't. What's different here? The fact that this isn't just an instance of the Canaanite shift, but also of elision of the alef. So perhaps it was the elision of the alef which was conditioned by the singular but not plural form, and it had nothing to do with stress. Also note that likely the plural form was originally ra’ašūma/ra’ašīma, which would be expected to yield רְאָשִׁים.

 

[utopia]

Of course it’s all theory, as we can't know for sure what linguistic process was taking place at the time, but we can make an assumption about what had happened.

If the elision of the alef happened before the canaanite shift then it would explain it, pending on the stress. But I can only assume that there was more than one way to say the plural of ROSH, and eventually this one won.

As for examples where an unstressed vowel went through the Canaanite shift, and the opposite - that a stressed vowel did not - Blau writes about some of them in his book - the MAQOM>MAQOMOT (unstressed vowel) for example, and he explain this specific example by system equation (if that is the term).

Stress is a pwoerful tool, but also analogy.

 

[Drink]

But the thing to explain would be why in this word the stress conditioned the shift, and in all other words it did not. What's special about this word that the whole rule is seemingly based on this one word when there are other possible explanations? What's special about this one word that the powerful tool of analogy worked on all other words but this one?

There just doesn't seem to be enough evidence to postulate a stress conditioned Canaanite shift.

 

[utopia]

On the singular?

If you're talking about ROSH, then why wouldn't the stress stipulate the shift?

RAASHU>ROSHU, which is quite logical.

If you're talking about RASHIM, then you'd have to explain what you mean.

I suspect that RASHIM had both RASHIM and ROSHIM as plural, just as some say today. And that RASHIM eventually spread.

 

[Drink]

What I'm saying is that the plural רָאשִׁים is literally the only reason to hypothesize that the Canaanite shift was conditioned by stress. It seems strange to use such an example as evidence as there are numerous other possible explanations for why it is רָאשִׁים rather than רֹאשִׁים that do not require alleging such a widespread phenomenon that left no other evidence behind. Especially note that in other examples of where the Canaanite shift did not occur, the syllable is in fact stressed, e.g. קָם and the קַטָּל pattern.

 

[utopia]

What I'm saying is that the plural רָאשִׁים is literally the only reason to hypothesize that the Canaanite shift was conditioned by stress. It seems strange to use such an example as evidence as there are numerous other possible explanations for why it is רָאשִׁים rather than רֹאשִׁים that do not require alleging such a widespread phenomenon that left no other evidence behind. Especially note that in other examples of where the Canaanite shift did not occur, the syllable is in fact stressed, e.g. קָם and the קַטָּל pattern.

The קם example is actually (I don't know if you knew it) what Blau brings in his book to explain why the shift did not occur in some instances. He believes it's because of the analogy, a systematic equation of the A sound instead of the O that was expected there.

As for קטל, well, he brought DAVAR (I don't know if they are the same as for the ultimate closed syllable), and he stresses that according to his knowledge it was a short A: DABARU> DABAR>DAABAAR

He says that this A became a long AA only after the shift had already ceased, thus the reason the shift did not occur here.

If KATTAAL had the same procedure, it explains why it did not become KATTOOL.

 

[Drink]

דבר is not relevant here, it is from dabaru, not dabaaru. No relevance to the Canaanite shift.

qaam is a different story. It is explained a different way.

But my point is all of these cases can be explained another way, so there is really nothing point to a stress conditioning in the shift. This is why modern scholars generally do not accept this theory.

By the way, another great counterexample is the ־מו suffix, as in לָמוֹ and תִּמְלָאֵמוֹ, etc. This comes from the old dual suffix -mā.

 

[utopia]

דבר is not relevant here, it is from dabaru, not dabaaru. No relevance to the Canaanite shift.

qaam is a different story. It is explained a different way.

But my point is all of these cases can be explained another way, so there is really nothing point to a stress conditioning in the shift. This is why modern scholars generally do not accept this theory.

By the way, another great counterexample is the ־מו suffix, as in לָמוֹ and תִּמְלָאֵמוֹ, etc. This comes from the old dual suffix -mā.

If דבר explains קטל then it is relevant. The DAAVAAR of today comes from DABARU, that is, the second syllable changed from short to long only after the shift had already ceased.

QAM is explained by analogy to the forms where the shift did not occur.

If they are explained another way, you are well aware that you're free to post other ways of explanation. I would be very happy to read other ways to explain these phenomena.

As for מו- - I checked what Blau says about it, and according to him it comes from *humuu. Nothing to do with the Canaanite shift.

 

[Drink]

If דבר explains קטל then it is relevant. The DAAVAAR of today comes from DABARU, that is, the second syllable changed from short to long only after the shift had already ceased.

What I meant was דבר did not have a long a originally, so the Canaanite shift has no relevance to it. I only mentioned qattaal, not qaataal.

QAM is explained by analogy to the forms where the shift did not occur.

Analogy to which other forms?

If they are explained another way, you are well aware that you're free to post other ways of explanation. I would be very happy to read other ways to explain these phenomena.

The conventional explanation these days is that qaam had a distinct vowel that derived from the elision of -aya- (qaam < qayama), and this vowel was distinct from other long-a vowels. Perhaps it was more like /æ:/.

And for qattaal, the theory is that this form is a borrowing from Aramaic, while the native form is qittool, as in גִבּוֹר.

As for מו- - I checked what Blau says about it, and according to him it comes from *humuu. Nothing to do with the Canaanite shift.

That does not seem likely, as then why is it -moo and not -muu? We do not have open uu that became oo in Hebrew.

But anyway, you have not answered the question: raashiim could be explained without postulating that the entire Canaanite shift only applied to stressed syllables. So why do we take one little piece of evidence and postulate this whole theory? Normally theories like this require multiple pieces of evidence.

 

[utopia]

What I meant was דבר did not have a long a originally, so the Canaanite shift has no relevance to it. I only mentioned qattaal, not qaataal.

Yes, and since I didn't know if QATTAAL had the same process I expressed my doubt if it applies here. As for DAVAR, it is relevant to the Canaanite shift as a contra example.

Analogy to which other forms?

Analogy of קמתי, קמת, קמנו, קמתם/ן to the third person forms.

The conventional explanation these days is that qaam had a distinct vowel that derived from the elision of -aya- (qaam < qayama), and this vowel was distinct from other long-a vowels. Perhaps it was more like /æ:/.

And how would you explain the fact that in the present forms it's still קָם, קָמָה, קָמים, קָמות?

And for qattaal, the theory is that this form is a borrowing from Aramaic, while the native form is qittool, as in גִבּוֹר.


That does not seem likely, as then why is it -moo and not -muu? We do not have open uu that became oo in Hebrew.

Yes, but he does say about final vowles that there seems to be anceps in many occurances. So long U sometimes occured as a short one. And that, with the fact that third person possession is OO (analogy), might do the work, but it's my speculation.

But anyway, you have not answered the question: raashiim could be explained without postulating that the entire Canaanite shift only applied to stressed syllables. So why do we take one little piece of evidence and postulate this whole theory? Normally theories like this require multiple pieces of evidence.

It was my logics to go and look for an answer in Blau's book. He made sense in past topis, so I don't see why not him. The theory of the Canaanite shift and its period seems to explain the case of RASHIIM too, that is that it did not apply to an unstressed AA sound. So why not. And if you have any other explanations, as I said, I'm never against it!

 

[berndf]

No, because that would make no sense. Why would it be shortened?

Are we sure the original a was long? I don't think so. Arabic, e.g., has a short a. So, it seems that the a was lengthened prior to the Canaanite shift and that may very well only have affected the singular.

 

[raamez]

Are we sure the original a was long? I don't think so. Arabic, e.g., has a short a. So, it seems that the a was lengthened prior to the Canaanite shift and that may very well only have affected the singular.

Perhaps it's gone through a de-hamza-ization like in spoken Arabic prior to the Canaanite shift. Thus it was ra'sh -> raash -> roosh

 

[berndf]

Perhaps it's gone through a de-hamza-ization like in spoken Arabic prior to the Canaanite shift. Thus it was ra'sh -> raash -> roosh

Maybe, but then the loss of the [?] must have been very old, perhaps older than then earliest Hebrew spelling. But the Aleph is still present. Do we have any idea when the Canaanite shift happened?

 

[Drink]

Yes, and since I didn't know if QATTAAL had the same process I expressed my doubt if it applies here.

I'm not sure what you mean. Qattaal is known to have had an original long vowel. That is why it is long in Arabic and Aramaic. Not like davar which had short vowels.

As for DAVAR, it is relevant to the Canaanite shift as a contra example.

It's not a counterexample. It's just a word where there is no applicability of the Canaanite shift. There are no long-a's in the word.

Analogy of קמתי, קמת, קמנו, קמתם/ן to the third person forms.

Then why is there a qamatz instead of patach in קם? Most other pa'al verbs have a patach whether with or without the suffixes: אָמַר, אָמַרְתִּי. So why in קם and קמתי do we have different vowels, if the vowel in קם arose through "analogy" to קמתי?

And how would you explain the fact that in the present forms it's still קָם, קָמָה, קָמים, קָמות?

Perhaps same way, that -aayi- perhaps was elided to the same vowel? But it's also trickier because we know that there are different forms of the present tense. In some verbs, the present tense doesn't gave a long vowel at all (mostly stative verbs), and so perhaps that factored into the development here as well.

Yes, but he does say about final vowles that there seems to be anceps in many occurances. So long U sometimes occured as a short one. And that, with the fact that third person possession is OO (analogy), might do the work, but it's my speculation.

But there are no other examples of final -uu becoming -oo. It doesn't make any sense. If the long -uu became short -u, then it would have been dropped like the other final short vowels.

It was my logics to go and look for an answer in Blau's book. He made sense in past topis, so I don't see why not him.

I don't see why you are so attached to Blau. This theory is well known in Hebrew linguistics, and is considered outdated.

The theory of the Canaanite shift and its period seems to explain the case of RASHIIM too, that is that it did not apply to an unstressed AA sound. So why not.

It doesn't "explain" it, but rather the whole theory is based on it. There is no other example of the Canaanite shift not applying to unstressed syllables. ראשים is the only one. You can't make a theory like this based off of one word.

And if you have any other explanations, as I said, I'm never against it!

I've already given them. And I will mention again below.

Are we sure the original a was long? I don't think so. Arabic, e.g., has a short a. So, it seems that the a was lengthened prior to the Canaanite shift and that may very well only have affected the singular.

My explanation was as follows:

ra’š > rāš (loss of glottal stop) > rōš (Canaanite shift)

Note this also happened with some verb forms:

ya’miru > yāmiru > yōmiru > yōmar

While in the plural the glottal stop was present until after the Canaanite shift was no longer operative.

Maybe, but then the loss of the [?] must have been very old, perhaps older than then earliest Hebrew spelling. But the Aleph is still present. Do we have any idea when the Canaanite shift happened?

The spelling of the singular was etymological.

We know the Canaanite shift was complete before the consonantal text of the Bible was established, as words that went through the Canaanite shift are very often spelled with vav (I'd estimate they tend to be alternately spelled with and without vav about 50% of the time each).

Note also that יאמר (from the example above) retained the א in spelling as well.

 

[berndf]

The spelling of the singular was etymological.

For an "etymological spelling" to exist, we would have to assume that there was an established spelling system for Hebrew before the Canaanite shift and before the preceding loss of [?]. Therefore my question:

Do we have any idea when the Canaanite shift happened?

 

[Drink]

You are asking the question backwards, because our ideas of when the Canaanite shift happened are based on this sort of evidence itself.

Strictly speaking you are right that etymological spelling implies that scribal schools must have existed before and after the loss of the glottal stop. However, I was including analogical spelling under the umbrella of etymological spelling. That is if the glottal stop persisted in the plural, then the question of when the Canaanite shift happened becomes disconnected from the alef in this case.

PS: There are some words, such as כוס, for which we know that the original glottal stop (c.f. Arabic كأس) was lost much earlier, as even in Ugaritic, which preserves the glottal stop in ra’šu <rỉš> has lost it in kāsu <ks>. Thus כוס cannot be used to date the loss of glottal stop in ראש.

 

[berndf]

You are asking the question backwards, because our ideas of when the Canaanite shift happened are based on this sort of evidence itself.

Not really. Main evidence is comparison with other Semitic languages, not so much internal reconstruction.

 

[utopia]

I'm not sure what you mean. Qattaal is known to have had an original long vowel. That is why it is long in Arabic and Aramaic. Not like davar which had short vowels.

All fine, I just didn't know if it's similar or not.

It's not a counterexample. It's just a word where there is no applicability of the Canaanite shift. There are no long-a's in the word.

I meant that it's an example of when the shift could not occur, because there used to be a short vowel.

Then why is there a qamatz instead of patach in קם? Most other pa'al verbs have a patach whether with or without the suffixes: אָמַר, אָמַרְתִּי. So why in קם and קמתי do we have different vowels, if the vowel in קם arose through "analogy" to קמתי?

Maybe I'll cite him:
In all third person forms in the past tense there was a long vowel: *qaamu, *qaamat, *qaama, while in all the other verbs a short one: *qamtaa. These first and second person verbs influenced the third person verbs and the aa never changed to oo.

He continues: בינוני קל was built according to a parallel mishqal of qatal, in PAAL: קָם-קָם. That might be because most עי"ו verbs in PAAL are stative. Here the qamats did not change to holam both because of the QATAL and because of analogy to the plural forms, where there used to be an unstressed qamats. There are some exceptions though like הקומים (מלכים ב טז, ז)

Perhaps same way, that -aayi- perhaps was elided to the same vowel? But it's also trickier because we know that there are different forms of the present tense. In some verbs, the present tense doesn't gave a long vowel at all (mostly stative verbs), and so perhaps that factored into the development here as well.

Perhaps.

But there are no other examples of final -uu becoming -oo. It doesn't make any sense. If the long -uu became short -u, then it would have been dropped like the other final short vowels.

There are rare forms like אליהֵמָה so why other rare forms would be so strange that you wouldn't believe such etymology can exist. Not everything goes by the book.

I don't see why you are so attached to Blau. This theory is well known in Hebrew linguistics, and is considered outdated.

But from what I saw there are relevant parts.

It doesn't "explain" it, but rather the whole theory is based on it. There is no other example of the Canaanite shift not applying to unstressed syllables. ראשים is the only one. You can't make a theory like this based off of one word.

It might apply to אנוש - אנשים too. But I'm not that fanatic, though it works here. And I don't know why not raise it in such a case.

I've already given them. And I will mention again below.


My explanation was as follows:

ra’š > rāš (loss of glottal stop) > rōš (Canaanite shift)

Note this also happened with some verb forms:

ya’miru > yāmiru > yōmiru > yōmar

While in the plural the glottal stop was present until after the Canaanite shift was no longer operative.

Why would the singular go through elision but the plural wouldn't? Is it a matter of stress?

 

[Drink]

Not really. Main evidence is comparison with other Semitic languages, not so much internal reconstruction.

Not necessarily. The shift may have operated in different time frames in different languages/dialects.

 

[Drink]

There are rare forms like אליהֵמָה so why other rare forms would be so strange that you wouldn't believe such etymology can exist. Not everything goes by the book.

But who said this is the dual? Perhaps it is from -huma, as the original plural, just like הֵמָּה and הֵנָּה, and thus had no original long vowel.

I didn't say that rare forms are bad examples. In fact rare forms are often great examples. But you actually need to have more than one solid example to make a case like that.

It might apply to אנוש - אנשים too. But I'm not that fanatic, though it works here. And I don't know why not raise it in such a case.

Except that אנשים is never seen as the plural of אנוש. It seems to be the plural of ’inš, which became איש.

Why would the singular go through elision but the plural wouldn't? Is it a matter of stress?

Either because of stress, or because they are not the same form, as I have already said, the plural may have been ra’ašiima rather than ra’šiima.

 

[berndf]

Not necessarily. The shift may have operated in different time frames in different languages/dialects.

We would know about the shift even without internal reconstruction and it doesn't just concern Hebrew but all Canaanite languages. That it happened independently in different time frames is an ad hoc assumption, i.e. something to be frowned upon from a methodological perspective.

 

[Drink]

We would know about the shift even without internal reconstruction and it doesn't just concern Hebrew but all Canaanite languages. That it happened independently in different time frames is an ad hoc assumption, i.e. something to be frowned upon from a methodological perspective.

I never said independently. But on the contrary, there is very good reason to assume that it did not happen at the same time in all languages/dialects. And that reason is that the so-called "second Canaanite shift" (which has little to do with Canaanite languages) operated as an areal phenomenon over a very long period of time impacting different parts of the region at different times and to different extents.

By the "second Canaanite shift", I refer to the shift of Aramaic ā, and the secondary Hebrew and Phoenician ā to ō or a more ō-like vowel. You could maybe even lump Persian into this areal phenomenon as well.

Anyway, the point is not that this second shift presents any direct evidence of how the first shift happened, but rather it presents a very plausible possibility for how the first shift happened.

In other words, it cannot be assumed a priori that the shift happened at the same time in the same way in all languages that it affected. In fact we know that that is never how sound changes work in practice.

PS: It is very clear that this is not something that can be looked at as "the Canaanite shift happened once in the common ancestor of Canaanite languages". There is good evidence for this: Ugaritic, which did not have the Canaanite shift, must be considered in the same clade as the Canaanite languages (by this I mean that there are Canaanite languages that are closer to Ugaritic than to other Canaanite languages), and yet, the other Canaanite languages had already diverged by the time of Ugaritic. So either the change happened before Ugaritic, and Ugaritic escaped from it (thus it did not impact all Canaanite languages equally) or it happened after or during Ugaritic, at a time when we know for sure the Canaanite languages had already diverged.

 

[berndf]

And that reason is that the so-called "second Canaanite shift" (which has little to do with Canaanite languages)

Ok, that is fine. So it is like the 1st Germanic shift, which applied to all Germanic languages and the 2nd, which only applied to High German dialects. But than we should keep the terminologically distinct.

 

[utopia]

But who said this is the dual? Perhaps it is from -huma, as the original plural, just like הֵמָּה and הֵנָּה, and thus had no original long vowel.

I understood that the original male plural form is *HUMU(U).

I didn't say that rare forms are bad examples. In fact rare forms are often great examples. But you actually need to have more than one solid example to make a case like that.


Except that אנשים is never seen as the plural of אנוש. It seems to be the plural of ’inš, which became איש.


Either because of stress, or because they are not the same form, as I have already said, the plural may have been ra’ašiima rather than ra’šiima.

Listening to people (as far as it might be from the sounds of ancient Hebrew) articulate the word RA'ASH (noise) RAASH, and in many cases I hear the plural RAASHIM, not RE'ASHIM and not RA'ASHIM, but RAASHIM eliding the ע, it's hard for me to understand how people in ancient times would say RAASHU >ROOSHU, but will refrain from elision in the plural.

It just looks as if we miss something here, and get lost in too many theories.

 

[Drink]

Ok, that is fine. So it is like the 1st Germanic shift, which applied to all Germanic languages and the 2nd, which only applied to High German dialects. But than we should keep the terminologically distinct.

No you misunderstand.

As I said:

“Anyway, the point is not that this second shift presents any direct evidence of how the first shift happened, but rather it presents a very plausible possibility for how the first shift happened.”

 

[Drink]

I understood that the original male plural form is *HUMU(U).

You make a good point, that does in fact seem to be what it originally was. I'll have to consider this again.

Listening to people (as far as it might be from the sounds of ancient Hebrew) articulate the word RA'ASH (noise) RAASH, and in many cases I hear the plural RAASHIM, not RE'ASHIM and not RA'ASHIM, but RAASHIM eliding the ע, it's hard for me to understand how people in ancient times would say RAASHU >ROOSHU, but will refrain from elision in the plural.

It just looks as if we miss something here, and get lost in too many theories.

This is not a good way to go about this. In modern Hebrew, the glottal stop barely exists as a phoneme at all. In ancient Hebrew, we know it was a solid phoneme. It was only elided in certain situations, and the farther back in time you go, the fewer those situations. Hebrew sounded very different back then.

I recommend listening to Arabic speakers that pronounce ق as a glottal stop. For them, this sound is very distinct.

 

[Ihsiin]

PS: There are some words, such as כוס, for which we know that the original glottal stop (c.f. Arabic كأس) was lost much earlier, as even in Ugaritic, which preserves the glottal stop in ra’šu <rỉš> has lost it in kāsu <ks>.

Not necessary, the shift away from glottal stops is something that has happened again and again in Semitic languages across a variety of words and over a large period of time. Case in point: modern Arabic has kās not ka's - if it we didn't know that Classical Arabic had ka's, we might be concluding that Proto-Semitic had kās on the basis of comparison. Ugaritic may well have made the shift independently as well.

Similarly, and as you've noted above, the Canaanite shift is something which has occurred a number of times across a number of languages and over a large period, being found not only in Hebrew and Phoenician , but also occurs in dialects of Aramaic and Arabic.

Bearing all this in mind I think goes quite a way to answering the question of this thread.

It is also interesting to note that I think all modern Arabic dialects have راس for 'head' with no glottal stop, but some dialects have the plural رؤوس, which still has the glottal stop. Another example from Classical Arabic would be the word ملك meaning 'angel', which is the elided form of ملأك - however, the plural is always ملائكة, with the glottal stop still there. Perhaps this is more expected in Arabic with its broken plural system, but these are still examples of similar phenomena in another Semitic language.

 

[utopia]

Actually I do listen to them every day. And what did catch my ear was one time when someone said BTIIRI instead of BTIQRI and DAAYE instead of DAQAYEQ
That actually brings me back to my aforementioned observation with which I can’t do much.

 

[Drink]

Not necessary, the shift away from glottal stops is something that has happened again and again in Semitic languages across a variety of words and over a large period of time. Case in point: modern Arabic has kās not ka's - if it we didn't know that Classical Arabic had ka's, we might be concluding that Proto-Semitic had kās on the basis of comparison. Ugaritic may well have made the shift independently as well.

I think we are mostly in agreement. All I was trying to say was this:
- we know that כוס once had a glottal stop, because Classical Arabic retains it
- we cannot assume that ראש lost its glottal stop at the same time as כוס did, because Ugaritic provides us with a case where know the latter had lost its glottal stop while the former had not

 

[berndf]

No you misunderstand.

As I said:

“Anyway, the point is not that this second shift presents any direct evidence of how the first shift happened, but rather it presents a very plausible possibility for how the first shift happened.”

Why do you say "no"? Either they are distinct events that happened at different times or they aren't. And if they are then different labels should be used. That is a prerequisite for a sensible discussion about how they relate or if they relate at all.

 

[raamez]

The very same thing can be witnessed in Arabic. The Quraishi dialect spoken by the prophet Muhammad then in which the Quran was revealed was known to have lost the Hamza. Nonetheless, hamzas were added later to the Quran because the Arabs somehow realized that this was the etymological or perhaps more archaic pronunciation. So the "newer" pronunciation of Quraish predates the more archaic spelling of classical Arabic.

 

[Drink]

Why do you say "no"? Either they are distinct events that happened at different times or they aren't. And if they are then different labels should be used. That is a prerequisite for a sensible discussion about how they relate or if they relate at all.

I said no because you are misunderstanding, and still misunderstanding. They The "second Canaanite shift" is a completely independent and disconnected from the first one. I said that already. I brought it as a parallel, not as a connection.

The very same thing can be witnessed in Arabic. The Quraishi dialect spoken by the prophet Muhammad then in which the Quran was revealed was known to have lost the Hamza. Nonetheless, hamzas were added later to the Quran because the Arabs somehow realized that this was the etymological or perhaps more archaic pronunciation. So the "newer" pronunciation of Quraish predates the more archaic spelling of classical Arabic.

My understanding is a little different. It's true the Quraishi dialect had lost the hamzas. But other dialects at the time had not. And it was those dialects that are represented in the Quranic diacritics (including vowels and hamzas).

 

[berndf]

The "second Canaanite shift" is a completely independent and disconnected from the first one.

So, what is your claim concerning ראש? When did the shift ra'š>rāš>rõš happen? With the first or with the second Canaanite shift.

 

[Drink]

So, what is your claim concerning ראש? When did the shift ra'š>rāš>rõš happen? With the first or with the second Canaanite shift.

The second Canaanite shift is unrelated to the processes discussed here. This is what I mean when I said you misunderstood by reference to it. When we say "Canaanite shift" we always refer to the first one, because that is the only one deserving of the name "Canaanite" shift.

Incidentally, the so-called "second Canaanite shift" affected the plural ראשים. But that's not directly relevant here.

Anyway, I urge you to go back to my original comment where I brought this up and reread it more carefully.

 

[berndf]

But then we are back at my problem: How can there be such a thing as an "etymological spelling" for a shift that happened before there was a written tradition in Hebrew?

 

[Drink]

But then we are back at my problem: How can there be such a thing as an "etymological spelling" for a shift that happened before there was a written tradition in Hebrew?

I've already answered this in comment #29.

PS: We do not actually know when a written tradition began either, so that doesn't help us with dating anyway.

 

[berndf]

I've already answered this in comment #29.

Yes, and I disagreed with it and instead of answering directly you evaded by talking about the second shift, which you now say is irrelevant. Can you please address the question directly instead of talking about what you said earlier?