Automatic Shva-na and Meteg Based Upon Context

gohebrew's picture

Advanced OpenType Software Tool, like MS VOLT, can create an automatic replacement of a string of glyphs, and insert a needed glyph, such as a shva-na and a meteg.

As a results, intelligent students can better study the grammatical use and proper placement of the shva-na and the meteg.

At first. I was distracted by seemingly Hebrew experts as to the impossibility of doing this, as the use and placement of shva-na and meteg is purely grammatical, in their minds, and not at all contextual.

So, I consulted with two truly Hebrew language experts, Prof. Aron Dotan of Tel Aviv University in Ramat Aviv, Israel, and Rabbi Shmuel Rabin of Toronto, Canada, a Lubavitcher chossid. They confirmed that this indeed was possible. Rabbi Rabin warned though that this was very difficult according to the viewpoint of Rabbi Zalman Hanau.

There are basically three schools of thought based upon the writings of the Radak, Rabbi David Kimchi - the father of our understanding of Hebrew grammar. They are:

1. Rabbi Zalman Hanau and most of Jewish thinking until about 225 years ago. The Chabad Lubavitch printing house in the USA, Kehos follows this approach.

2. Rabbi Eliyahu Bachur, and later the Vilna Gaon. The ArtScroll printing house follows this approach.

3. Minchas Shai. Shai LeMorah publishers of Jerusalem, and the Chabad Lubavitch printing house in Israel, Kehos, follow this approach.

david h's picture

> There are basically three schools of thought...

...And one School of Headache.

Your last post on that subject was 2 years ago! What do you want now?

gohebrew's picture

I was distracted for two years.

Experts suggest that though the shva-na and meteg are based upon grammar rules, a different set of rules can be defined by context.

Hence, three sets of shva-na replacements are required to support these three school:

1. Rabbi Zalman Hanau and most of Jewish thinking until about 225 years ago. The Chabad Lubavitch printing house in the USA, Kehos follows this approach.

2. Rabbi Eliyahu Bachur, and later the Vilna Gaon. The ArtScroll printing house follows this approach.

3. Minchas Shai. Shai LeMorah publishers of Jerusalem, and the Chabad Lubavitch printing house in Israel, Kehos, follow this approach.

William Berkson's picture

Israel, Eli Fried at the end of this thread: http://typophile.com/node/51733 claims to do what you have been trying. But he had to include some kind of look-up for special cases. John Hudson argued that this belongs in an application rather than the font, but in any case it seems to have been done some way or other.

So what's wrong with his effort?

david h's picture

> I was distracted for two years.

Don't work and watch TV :)

So where's the problem?
You have 'two truly Hebrew language experts', 'three schools of thought' (maybe add 2 eggs, 1 cup sugar, 1/2 cup flour... and make Brownie)...
what do you need?

gohebrew's picture

Bill,

I haven't seen or used Eli's fonts. What does it mean: "some kind of look-up for special cases"? If there are three views of the Radak's rules, should there also be three versions of each font, to insert a shva-na according to R' Z. Hanau, or according to R' E. Bachur, or according to the Minchat Shai?

I think that John's opinion was expressed when david and you were convinced that this would not be done. If we simply extend what John did for the furtive patach, to the shva-na, and to the meteg, then why would John conclude that this is out the realm for MS VOLT? John, am I correct?

I was in contact with a Rabbi Shmuel Goldstein of Florida regarding the rules of shva-na, at rebbeg@gmail.com, who has ideas as well on this subject.

gohebrew's picture

David,

TV? A Lubavitcher?

Brownies? In the Sukkah?

I plan to post a sample with the Shva-na fonts soon.

As I am doing a few jobs for custom OpenType fonts now, I imagine the posts will only occur later than sooner. In Appletalk, "real soon now". :)

John Hudson's picture

Israel, I thought I had been adequately clear all along that my objection to doing this with glyph substitution lookups has nothing to do with whether or not it might be possible; indeed, it if were not possible, what would be the point of having an objection? My point has always been that if an author or publisher wishes to visually distinguish sheva and sheva-nah, then it is presumably because he considers the distinction to be semantically important. If something is semantically important, then it should be possible to capture it and preserve it in plain text, i.e. at the character encoding level, not at the glyph processing level where it is subject to variant results in different software and is reliant on particular fonts. That is what I have stated repeatedly, and it has nothing to do with whether or not it is possible to do it at the glyph level in OpenType: it is about whether it should be done at that level.

William Berkson's picture

John, my understanding is that the distinction between the sheva and sheva-nah is not semantic, but only a matter of correct pronunciation. David will correct me if I'm wrong, but I think there are no words that become ambiguous in meaning when a distinction between the two marks is lacking.

John Hudson's picture

I was careful not to use the word ‘semantic’ in my last post because I understand that it can be variously interpreted. What I said is that the distinction, if it is made, is meaningful to the person using it, and hence not something that should be lost in text interchange. A desired plain text distinction isn't necessarily one that conforms to the definition of semantic that you suggest: Unicode includes a number of mechanisms for preserving visual distinctions, or at least the desire for such distinction (e.g. the ZWJ mechanism for indicating a ligature), in plain text in situations in which these are meaningful to particular users or user communities. Such mechanisms include control character sequences, free variant selector characters, and also optional disunifications. Since we already have a good example of the latter in Hebrew in the disunification of U+05B8 and U+05C6 for optional plain text distinction of a difference in pronunciation, there is precedence for a similar approach to sheva-nah. Having multiple mechanisms, some character based and some glyph based, for parallel text processing issues strikes me as a bad idea, and clever though I am sure the glyph processing solutions are, the time spent on them might have been better employed in crafting and submitting Unicode and ISO encoding proposals. [Israel mentioned furtive patach; this is not a direct parallel, because the distinction is one of mark positioning, not glyph shape; the distinction is not lost in text interchange, since it is always reliant on intelligent layout capabilities, and a font that cannot correctly position furtive patach is actually broken, as distinct from the very many that deliberately make no distinction between e.g. qamats gadol and qamats qatan, or between sheva and sheva_nah.

gohebrew's picture

David,

I am not a chef, or a chef's son - although I make a mean omelete, and have a secret recipe for techina (that I might tell John if he say l'chayim).

If you master MS Volt, you'll see that mixing in the right ingrediants is just a small part. Right, John?

gohebrew's picture

Bill,

Rabbinic Hebrew language experts in Israel pointed out certain words (even in the Bible) in the prayerbook which differ in meaning by pronouncing a word with or without the shva-na.

Hence, on certain words or prayers which are chanted, the distinction is key.

Yet, to this day, John, the Unicode Co. ignores this real practical need.

Plus we need to teach Jewish children these matters, and serious Christian Bible students needs these matters as well. Are your contacts at SBL, John, aware of this?

gohebrew's picture

>> My point has always been that if an author or publisher wishes to visually distinguish sheva and sheva-nah, then it is presumably because he considers the distinction to be semantically important. (John)

See above post.

gohebrew's picture

>> If something is semantically important, then it should be possible to capture it and preserve it in plain text, i.e. at the character encoding level, not at the glyph processing level where it is subject to variant results in different software and is reliant on particular fonts. (John)

Here is the error.

Hence, we wait, David. No Unicode code, no Volt. No ticky, no washy.

gohebrew's picture

John

>> ...it has nothing to do with whether or not it is possible to do it at the glyph level in OpenType: it is about whether it should be done at that level.

Because of this point, I had to banter with these for nearly a year. If you have said then: "...it has nothing to do with whether or not it is possible to do it at the glyph level in OpenType", then we could have argued about the Yankees and Red Sox - whose better? Yanks, Sox, Yanks, Sox...

gohebrew's picture

John,

I'll respond to your John Hudson 1.Oct.2010 8.03pm soon, but you must catch my drift.

david h's picture

> Rabbinic Hebrew language experts in Israel pointed out certain words (even in the Bible) in the prayerbook which differ in meaning by pronouncing a word with or without the shva-na.

Israel,

Such as...?

gohebrew's picture

David,

It's in an old email which I have (nothing gets lost on Gmail).

It's only a few words: one in Tanach; one is the prayerbook (even the Reform), and other is a very rare wierd word (it might be in Psalms).

John has software for every Hebrew word with all nikud taam etc in Tanach.

So, I can't tell you now, but poor Bill has been pronouncing incorrectly for years.
But G-d is merciful. :)

gohebrew's picture

John,

Does the database know where a shva-na should go?

Scott-Martin Kosofsky's picture

You might take a look at the article "Shva" in Wikipedia, which is accurate so far as I understand the issue. As a practical matter of phonology, we might benefit more from differentiating the dagesh kal from dagesh hazak, than worrying about the shva-na and shva-nach, which need less practical guidance. People from the Middle East pronounce the doubled consonants of the dagesh hazak quite naturally, whereas Westerners have trouble. The only Western language I know that has a similar doubled consonant is Catalan, which has a doubled L (lowercase) separated by a raised dot. It may well be an artifact of pre-1492 Spain.

david h's picture

> If you master MS Volt, you'll see that mixing in the right ingrediants is
> just a small part

What you do speaks so loudly that I cannot hear what you say.

gohebrew's picture

david,

I don't understand.

I do a few things, some for parnossa, some for future earnings, and others as Christians say because I feel that I have a calling.

Some work that I do is to create Biblical Hebrew fonts, very high level OpenType fonts with MS Volt. This is mainly graphical programming, requiring intense logic. Few people, far less a dozen, perhaps a half a dozen or less.

John Hudson is one of them; I am another. Besides having brilliant logic, John is a fine mensch.

I have not met anyone like him, with the exception of Scott-Martin Kesofsky. They are giants, and others are great but not giants. I wash dishes.

Typograph's picture

Israel (gohebrew)

"This is mainly graphical programming, requiring intense logic. Few people, far less a dozen, perhaps a half a dozen or less."

Before intense logic and all the rest of the mombojombo you must know what you'r doing, why are you doing it,who you are doing it for and what are you aiming for.

my personal filling is that you are somehow confused and not all that clear on the subject of this thread.

before befor you get involved in the different opinions
first get dwon the basics
first stgae, the easiest on = Dagesh Kal & Dagesh Hazak
then work out the basic sheva na/nach
then work out the basic Qamats Rahav/Qatan

from here on you strat refining the project.
for now don't concern your self with all of the arguments, just build a solid basic fonctioning font.
only then involve your self with the rest of the issues and take it as far as your "Call" calls for you.

gohebrew's picture

Hi Eli,

I don't think that I am confused.

I use Volt mainly for the improvement of nikkud, meteg, and taam (and the combinations thereof) of Hebrews via GSUB look-ups, originally written by John Hudson.

Shortly afterwards, I used the Volt software professionally, I discovered serious bugs with it. After testing the results of dozens of different possibilities I pinpointed and defined the problems. I created work-arounds, which I currently use. I stopped using John's 'program' after I taught myself how to create 'look-ups', 'GPOS', 'GSUB', 'programs' etc., with initial help from Diane Collier, recommended by Microsoft Type.

I believe I no longer have these bugs, after months of frustration.

MS Volt is powerful software, even with the bugs. But its graphical interface appears for programmers, and not for designers, who seek something intuitive and easy to use.

Eli:
> before before you get involved in the different opinions, first get down the basics.
First stage, the easiest on = Dagesh Kal & Dagesh Hazak; then work out the basic sheva na/nach; then work out the basic Qamats Rahav/Qatan.

Israel:
I decided to start with shva-na and meteg (among nikkud) because they could be contextually defined, and used in MS Volt with the GSUB routine - as explained to me by Prof. Aron Doton, formerly a Professor Emeritus at Tel Aviv University in Israel, and Rabbi Shmuel Rabin of Toronto, Canada, the greatest Hebrew language expert that I've spoken to. I feel privileged to have spoken with Prof. Doton, who at first thought it was impossible to contextually define a shva-na, because this was never done before. But, as a true genius, he thought about it in less than a minute, and changed his opinion. Rabbi Rabin never had a doubt, but warned that according to the view of Rabbi Zalman Henna or Hanau, this was exceeding difficult, as his rules were more complicated than that of the Minchat Shai (who we have on the entire Tanach), and Rabbi Eliyahu Bachur.

Which view do you follow? Rabbi Zalman Hanau (Henna), Minchat Shai, or Rabbi Eliyahu Bachur.

gohebrew's picture

Eli:

> From here on you start refining the project.
For now don't concern yourself with all of the arguments, just build a solid basic fonctioning font.
Only then involve yourself with the rest of the issues and take it as far as your "Call" calls for you.

Israel:

How can a font be created in general, not according any particular view? I am not intending to do the shva-na work at the font level in FontLab Studio. Rather, the font is completed, and every shva is nach, then at the Volt level the shvas are changed in real time only to eith shva-nas or shva-nachs, like the furtive patach changes certain patachs to a furtive patach.

This is doable, but the look-ups, GPOS, and GSUB, is very time consuming, because pages (books really) of material must be reviewed carefully according to each viewpoint.

Then, I must do the same for the meteg.

Are you suggesting that this can be done for the dagesh kal/chazak (is there a visual difference), and komatz rabah/katan?

Typograph's picture

Israel, Some Sheva Na's And Qamats Qatan depend on the dageshim there for starting directly with sheva na/nach is a mistake.
Forget Meteg, your waisting your time.
You must first understand what builds on what.

Israel, regarding the 3 SHITOT, give it a rest.
First start on what they all agree on.
you are not there yet.

second, forget about the different views, and break it down to more detailed issues.
but again, you are not there yet.

after you hav a basic functioning font, then start talking about actual issues here on typophile every thing else is a bounch of mombojombo in the thin air and will lead you nowhere
for now, you are waisting yor own time and others who don't excactly understad what the hell you are talking about.

when you bump your self with an actual issue, post a thread regarding the particular issue.

Typograph's picture

btw, yor approach abount having the the defoult shevas as nachs is a mistake.
i sugjest you first make all shevas to na and the work it down to nach.

The main reason is because you cant check a first letter of a word that is in a begining of a line, and all shevas at the first letter of a word is allways NA

Typograph's picture

to start explaining to you why not to deal with meteg is diffcult.
but in general try avoiding inserting glyphs into text at an OT level that should absolutly be done at the charecter level, never ever by OT.

sheva na/nach ect, does not insert a sheva, just replaces them with alternates.

gohebrew's picture

Eli,

> The main reason is because you can't check a first letter of a word that is in the beginning of a line; all shevas at the first letter of a word is allways NA.

Yes. We used to identify it with the boldened shva.

But today, we do not identify it. Hence, It appears like a shva-nach, no different in any way. Only when a shva appears in the middle of a word, then a floating graphic is inserted by the special OT font - on the screen, at the printer or PDF, but not in the file.

Again, my way seems logical and simpler, no? GSUB sees a certain string which according to the defined rules in GPOS look-ups based upon Gloyph Groups, then inserts above some shvas a floating glyph.

It doesn't need to know if a dagesh is kal or chazak.

A komatz katan can help define a shva-na I was told. In fact, I believe that the kollel is called Revach, in three words, the komatz is katan and then the shva is nach, but when the komatz is rabah, then the shva is na.

So, the three letters could simply be a Glyph Group. And we can save search for komatz katan later, if it can be defined contexually.

gohebrew's picture

Eli,

> to start explaining to you why not to deal with meteg is diffcult.
but in general try avoiding inserting glyphs into text at an OT level that should absolutly be done at the charecter level, never ever by OT.

Here I disagree.

The who purpose of Volt and an advanced OT font is remove the difficult decision making about diacritic placement from the typeetter, and to place it in the responsibility of the type designer.

The user does not to know where to place the komatz or patach. The type design needs to know, and place it correctly. If it's an aleph, it goes here; if it's a reish, it goes here. If it's a patach at the end of a word, and a hei, then replace it with a furtive patach.

I want to do the same with a shva-na, and afterwards with a meteg if possible.

Why?

These are basic Hebrew grammar things that kids should learn, and teachers should teach. When a typesetter works on voweled Hebrew text, then these glyphs could be inserted automatically.

Baruch Gorkin's picture

You people are talking AS IF all the rules of shva-na / sheva hach are universally accepted by all authorities and all you have to now do is program them and, presto - you're done!

Newsflash – nothing can be further from the truth! There are two major approaches to this subject alone. One the common one – (accepted by Academia LeLashon HaYivrit); the other one is that of Rabbi Zalman Henau. The differences between the two are so many, it will make your head spin. Like "malchuth" vs. "malechuth"; "ethchem" vs. "ethechem", etc., etc., etc.

Secondly, there are various masoretic traditions and one man's sheva-na is another man's chataf-patach, etc.

The same applies to quamatz-koton.

SO EVEN IF you wish to create a graphic differentiation between these vowels (and all attempts in this direction have thus far been hideously UGLY), you best leave this at up to the editor.

Lastly, the Tiberian system we use today doesn't distinguish between these vowels GRAPHICALLY. Why not just well enough alone?

raphaelfreeman's picture

I think you have to separate the various issues:
1) should the font even include automatic substitution.
2) is the font trying to be 100% correct
3) is the font trying to do this for a Tanakh with taamim or regular text
4) is the font trying to cover all situations.

1) Personally I don't think this should be encoded but if you are going to do it then...
2) Clearly to anybody who actually understands the subject, this is impossible. However, there is value in being 90% accurate and then the proofreading will be quicker. If you imagine a text probably has to go through at least 4 professional passes, if there is less things on the line to correct, this will be more accurate. Here it could be useful.
3) to those who understand the subject very well, then this point is crucial.
4) I think that obviously each publishing house will have it's own set of rules and no font can cover each system. And frankly a publishing house wouldn't need such a font! However, there is value in creating a font with a system for those who actually want this for a different reason and that is for using it as a basis of transliterating the text. I can definately see a teacher or school buying such a typeface for use and then they can easily make a transliteration for kids. As my time as freelance typesetter for major Jewish publishing houses in the USA, I saw that they had "experts" transliterating texts with sooooo many mistakes, clearly demonstrating their lack of ability to read the Hebrew language correctly by any system. Having inspected Typograph's sample of Psalms that he posted, a font such as this, even though "only" 95% accurate out of the box, would be fantastic for hundreds of thousands of students where the people behind the project of that particular book don't have the knowledge or more importantly the budget to do the job 100% correctly and do a job that's at best 60% good.

However, it is critical before embarking on such a mission to have the knowledge of grammar. It's clear that Typograph has this depth and breadth of knowledge, but with respect to gohebrew, I'm not sure a meeting with a Professor here (even it is Dotan) and Rabbi there is quite enough. I know that at Koren, the qualifications for doing this are at least a Masters in Hebrew grammar, preferably a Ph.D. so I would say to gohebrew, before continuing make sure you grammar knowledge is at the level of Typograph's otherwise I'm not sure what the point of the font is.

Typograph's picture

Rephael:
at the begining of my first response to this thread i wrote

"Before intense logic and all the rest of the mombojombo you must know what you'r doing, why are you doing it,who you are doing it for and what are you aiming for."

Before intense logic and all the rest of the mombojombo you must know what you'r doing.
++++ knowing the basics of the job at hand and what it takes.
why are you doing it
++++ what are you going to solve
who you are doing it for
++++ who is going to use it and does my work needed for them
and what are you aiming for.
to what level of accuarcy are you trying to get to.

No ISRAEL (gohebrew), i don't car who you spoke with and what was said to you.
right now you are not in a position to agree or disagree with me because you are clearly laking the basic knoledge.

I'm the only one (for now) who actualy programed this functions in a font (In VOLT) and actualy put it to test with sevral editors and print houses around the world who used it to print sidurim and chumashim. i have studied their problems and what can i solve automaticly and what i allow them manualy to change.

Proffesor Dotan With all do respect does not know programing nor volt.
So he can give you a theoretic opinion, I am sharing with you my knoledge after actual experience and know exactly what can be programed and what are the limitations of such a project.

so me to have no opinion.
I AM TELLING YOU ABSOLUTLY WHAT CAN AND CANNOT BE DONE!!!
I know that my writing right now aound arragent and all knowing, but my intention is to put you on the right path and save you houres of work.

Now, If you want to be a scientist then go right ahead and try your self what can and can not be done, be my guest. otherwise what i have said is absolute with no place of argument from your side.

When i tell you to give up on Metegs, I know what i am talking about.

Boruch:
There are basics way befor all the SHITOT.
All this issus are left to the editor to decide.
I have pre-made lookup for different issues wich i attach to the project according to different needs.

Typograph's picture

Gohebrew:
You can weist your time with all kind of proffesors.
My advice Again, Jump in to the cold water and then start dealing with problems and diffculties that you encounter as you go along.

and another tip.
knowing the grammer alone is not enough
you must understand them as well.
as you swim in to deep water, you will se that alot of rules you are going to make up yourself (The rules of grames i cal KLALIM GEDOLIM And the one i made up i call KELALIM KETANIM).

gohebrew's picture

Eli,

First, I want to thank you for your advice and tips.

Second, realize that much of my seeming opposition is very like that of my brainy brother-in-law, the devil's advocate.

It seems that Boruch my friend, Raphael, and you, have not seen my lone tree in perceiving the vast forest of Hebrew grammar.

I am a simple Jew with limited knowledge and experience in these matters. I know Volt (which after one masters it, it seems very simple), how to define a glyph group, and the two kinds of look-ups, GPOS and GSUB.

And the great elderly Prof. Dotan, and the even greater young man, Rabbi Rabin, affirm that the shva-na can be defined contextually - and Volt can replace things contextually, so I feel that Volt can do the desired task.

I have no MA in Hebrew grammar, or even a BA in Hebrew grammar, Raphael.

I simply have a lot of Shai L'Morah books with nikkud and shva-nas, based upon the Minchat Shai's approach to the shva-na, two lots of ArtScroll books with nikkud and shva-nas, based upon Rabbi Eliyahu Bachur's approach to the shva-na, and three lots of Kehot USA books with nikkud and shva-nas, based upon Rabbi Zalman Henna's approach to the shva-na.

So, I believe that I have all the research material staring me in the face. I have the necessary knowledge of Volt. 1 + 1 = 2.

My intention is not financially motivated. I simply want to take the knowledge of when a shva-na is needed away from the scholars with MAs or PhDs, and to return it to children and their teachers.

Eli, your approach seems logical and tested, but I feel my approach works best for the contextual approach.

John Hudson, who has this database of every verse with nikkud taam and meteg? Do they have the Chapters of the Fathers with nikkud and meteg, and the Jewish prayerbook with nikkud and meteg as well? Can you hook me up with them?

If Unicode had a shva-na encoded, and professional scanning software supported nikkud taam and meteg, then I could create my own database.

So, Eli, Boruch, and Raphael, do you grasp my intent now?

Typograph's picture

Gohebrew.
to start learning a subject just to pass it on to children dos'nt seem logical to me.
second, if thats is your only intention so the ansawr allready exists why work a year on it????
as you stated before there are 3 main shitot, you want to teach it via a font??? that seems farfeched.

as i said befor, you are totaly not clear with your self, what you are doing? why? and for whom.

when ever you decide to land on planet earth, and talk about defined issues, I would be more than happy to discuse them with you, but the way things are going now, your "shtik" has no interest for any one on typophile.

gohebrew's picture

Eli,

>> learning a subject just to pass it on to children dos'nt seem logical to me.
second, if thats is your only intention so the ansawr allready exists why work a year on it????

>> 200 plus years ago, our (very big) rabbis decided to limit teachers and students from understanding Hebrew grammar well, as they did in generations earlier. Why did they do this?

Israel:

At that time, the Maskilim (The Early Reform) knew Hebrew grammar very well, dressed as Chassidic Jews, and tried to lead actual Chassidic youth away from Torah observance. Hence, our (very big) rabbis decided that to identify and kick out Maskilim, that generation's frum youth should not know Hebrew grammar.

This went on for generations.

Today, there no longer are Maskilim. Even the modern reform seek Jewish values to enrich their lives, because they see their path leads to widespread assimilation.

I seek to bring Hebrew grammar back to the fold.

My primary occupation in Volt-based Hebrew type design. I plan to return to full-time ePublishing through GoHebrew at www.GoHebrew.com and other web sites. I am also coordinating the massive Schneersohn at www.GoTalmud.com - both of which are a year away from earning money.

This endeavor is pro-bono.

Eli:

> as you stated before there are 3 main shitot, you want to teach it via a font??? that seems farfeched.

Israel:

No, I am no teacher. The Volt-based shva-na fonts are the tools for the classroom.

Eli:

>> As I said before, you are totaly not clear with yourself, what you are doing? why? and for whom?

Israel:

I was and am clear. Bill, don't you think that this was clear? Don't you think every Jewish school would want this?

Typograph's picture

Gohebrew, again this is all talk in the air, no one is interested to read long threads about you wanting to do X and going on and on about it.

I think that after 2 years it's time to talk Haloche Lemaase.

Tachles, do you have any questions about X,Y,Z??? somthing people can relate too and discuse??? if yes so lets talk, if not it starts to be enoying.

gohebrew's picture

As I explained, I am not ding this Hebrew grammar stuff as a job. So, it's slow.

First, I must complete research on what strings cause a shva to be shva-na.

This can be done according to the Minchat Shai, Rabbi R' Shlomo Yedidiah Nortzi or Rabbi Shmuel Yitzhak Schor of Munstritz, since the Shai L'Morah books are extensive, while according to Rabbi Zalman Henna, used by much of Chabad, this is less. Rabbi Eliyahu Bachur, used by ArtScroll at times, this is even much less.

Maybe, something will be tachlis on the tisch in another year or too.

Thank you for your help until now.

William Berkson's picture

>Bill, don't you think that this was clear?

It's clear that you want to use open type to automatically set the kamatz katan and sheva na. It's not at all clear why and how you want to do this, especially now that Eli Fried has already done it—though Raphael Freeman thinks his effort is, and can be, only partially successful.

I think what Eli is reacting to in his comments is that you don't seem to understand the recursive character of the special grammatical rules here. Eli is telling you you need to first set what is a strong and regular dagesh before you can automate other stuff, because other stuff is dependent on that. Recursive here means there are rules built on top of other rules, and if you try to define only from immediate context it can't be done at all. So I think Eli Fried was telling you there is no way to avoid learning the grammar very well, to do this programming at all successfully.

As to whether Religious schools will want this, I think that depends on the religious school.

The history here is informative to me. Thanks to David Hamuel sending it to me, I have read William Chomsky's article "The Pronunciation of the Shewa." There Chomsky, as I read him, explains the history thus: these current "rules" were first laid out by David Kimhi, also known by the acronym Radak, who died in 1235. These rules were not only unknown before this, but were contradicted by other previous great scholars. Kimhi introduced the idea of long and short vowels, and a new view of when to pronounce the sheva. Chomsky argues that in fact, these rules were never followed before Kimhi. In other words, Kimhi didn't get Hebrew grammar right. He invented some complexities that never existed in spoken Hebrew, or the system of nikkud.

However, Radak's innovations—some of which were genuine and correct new insights—were accepted by later grammarians. Thus the partially incorrect ideas were accepted as "correct" and became traditional, even though incorrect. Now you have a situation in which, I gather, there are competing diverse traditions built upon Radak's mistakes, all of which are thought to be holy and correct by their adherents. But even they cannot be reduced to rules that don't have exceptions. So Raphael Freeman has four authorities who go over text to get it right by their lights—and another authority will probably disagree.

So what rules going to be helpful to a religious school or synagogue will depend on what tradition and authority they adhere to as relates to this.

Personally I find most helpful is what Chomsky explains as the current rules of spoken Hebrew for the sheva na, which he says were in fact followed by both Ashkenazim in Europe and Sephardi style in Israel. And they are simple. If there are two shevas in succession the first is silent, the second pronounced. You also get a spoken sheva at the beginning of the word, when the first letter is a "liquid consonant" (yod, lamed, men, nun, resh) with a sheva, or when it is a prefixed preposition (bet, kaf, lamed), or a conjunctive vav. Otherwise, all shevas are silent, including at the beginning of a word. There are some more complications, but that's basically it. These rules make total sense to my mouth and my brain.

For those of us who are struggling with learning Hebrew just an visual indication of those rules would be convenient. And as far as God understanding the prayers, I think He'll do just fine in any case. David tells me that Chomsky also got some stuff wrong, but in any case for me the actual rules of spoken Hebrew is what I'd be very happy with.

The whole Kimhi epicycles within epicycles stuff about the sheva na is, if William Chomsky is right, a wrong turn in Hebrew grammar anyway. However, Kimhi's work was I believe an inspiration for William Chomsky's son, Noam, to come up with "transformational-generative grammar" as a theory of the grammar of all languages, and that has made a big splash in the world of linguistics.

Typograph's picture

William:
Mainly, what i was triying to get to, is bringing this discution down to actual.
all of our talk abov and in general is a weist of time.

A project like this is based upon grammatical rules, however not every rule can be applied, so you must work your way arround it.

programing a project to work 100% is not possible.
the idea is not to make a project that one simply uses and there he as it, but to provide a powerfull tool for people who know the grammer to produce sidurim and chumashim with sheva na\nach faster.

so the font does its thing to 99.98% with taamei mikra and let say 98% without Taamey mikra. wich is acceptable.

furthermor, it is imposible to program this kind of project base only on grammer rule.
i cannot put in all the Shorashim Hatayot and so on in the font, I can only make disitions based on what i have before and after the sheva, wich means, i can not make desitions based on SHORASHIM, PEALIM and so on. (sometimes yes, but not every where).

any how, what i was trying to do, is bring this thread down to earth, discusing particular issues and not tawsing around opinions that have no actual perpose.

I read alot on typophile but hardly write.
one reason is because my english spelling is horrible.
Second, for most issues there are enough people here to ansawr them.
here i wrote because i thought that this discution could be interesting.
but as thing are going here i am getting bord.
I thing i have what to say or add in this topic, i studied this issue, but all of this high talk, is not for me

gohebrew's picture

Eli,

>> A project like this is based upon grammatical rules...

I asked Prof Aron Dotan of North Tel Aviv, and Rabbi Shmuel Rabin of Toronto, two world class experts in Hebrew grammar, if the shva-na - normally defined by grammatical rules - could also be defined by context?

They both said yes.

Dotan advised to study the nikkud very carefully.
Rabin warn that Henna's system is very complex, but can be done.

gohebrew's picture

> programing a project to work 100% is not possible.

they disagree. who knows more?

gohebrew's picture

Eli,

Please explain why programming or even knowledge of dagesh chazak or kal is necessay to define a shva-na - as Bill interpreted you to say?

Typograph's picture

Gohebrew, I said before i dont care who you spoke with.

Tell me.
If i hav a word with a qamts and the next word with a sheva
how do i know if the qamats is qatan and the sheva is nach, or the qamats is gadol and the sheva is na???
So if i have a meteg next to the qamats it means that the qamats is gadol and the sheva is na, but if you alone are going to insert metegs then you are going to get in to a big mishqebabel

Just one example.
Israel, you can do what ever you want, but you are banging your head against the solid wall.

instead of talking as if you know your stuff, why not be huble and simply ask the right questions, question i can ansawr you in detail????

Typograph's picture

gohebrew in sefer Ester the word רִצְפַת the sheva is na or nach?????? and why???
ansawr that and the will talk further

gohebrew's picture

Eli,

>> instead of talking as if you know your stuff, why not be huble and simply ask the right questions, question i can ansawr you in detail????

Do I claim to know anything from my own pupick?
Whatever I know is from what people told me or taught me.
I try to verify everything.

gohebrew's picture

>> a word with a qamts and the next word with a sheva
how do i know if the qamats is qatan and the sheva is nach, or the qamats is gadol and the sheva is na???

words or letters.

john hudson, what is (if any) is the upward limit in a string of glyphs that Volt can handle.

gohebrew's picture

Eli,

>> ... if i have a meteg next to the qamats it means that the qamats is gadol and the sheva is na ...

Where is the shva?

=========================
THIS IS AN IMPORTANT RULE
=========================

ISRAEL, PLEASE VERIFY IT

++++++++++++++++
----------------
****************

Typograph's picture

sorry, i mean a letter with qamats followd by a letter with sheva

Typograph's picture

gow hebrew, one of the usages of meteg aftar a qamats is to indicate that the qamats is GADOL and the sheva at the following letter is a sheva na.

by the way, you should know, that the qamats qatan is popular, but did you know that thers is also a segol qatan or patah gadol accourding to some MEDAKDEKIM???

Syndicate content Syndicate content