The Death of the Urdu Script

[Behnam Esfahbod]

I just read this article: The Death of the Urdu Script: Can Microsoft and Twitter save the dying Urdu nastaliq script from the hegemony of the Western alphabet and an overbearing Arab cousin?

https://medium.com/writers-on-writing/9ce935435d90

Besides the technical issues, there's an important non-technical one that the author is talking about in this article: Naskh is alien to Hindi-Urdu speakers. And, on the other hand, we know that Nasta'liq doesn't look right to Arabic speakers either.

Now, the questions important to us, Persian speakers, is: how come we get along with both styles? how we balance it out? and how that affects our typography?

To answer the first two, I think it's because we learn writing in Naskh in the first few years of school, and then grow old into Nasta'liq afterwards.

But about the typography, again, going back to one of the first typefaces, Nazanin and Titr, I think these typefaces benefit from both worlds: the alignment of the letters is simply like Naskh, but the shape of the letters are closer to Nasta'liq. This is how our handwriting is, most of the type. Not totally Naskh, not totally Nasta'liq.

And maybe it's the same for ligatures. In Perso-Arabic font families, those with many ligatures give me a non-Persian feeling. This could be because of the fact that these ligatures are based on Naskh style, and rarely appear in Nasta'liq.

We talked about the size of letter Heh in another thread, and how the size of this letter comparing to the basic elements (teeth, alefs, etc) could be a deal-breaker. In Nasta'liq, the shape of letter Heh is very subtle, and IMHO this could be the root for that.

-Behnam

--

Behnam Esfahbod | بهنام اسفهبد
http://behnam.es/

GPG-FP: 3E7F.B4B6.6F4C.A8AB.9BB9.7520.5701.CA40.259E.0F8B

[Bahman Eslami]

On Monday, October 14, 2013 11:03:21 PM UTC+3:30, Behnam Esfahbod wrote:

I just read this article: The Death of the Urdu Script: Can Microsoft and Twitter save the dying Urdu nastaliq script from the hegemony of the Western alphabet and an overbearing Arab cousin?

https://medium.com/writers-on-writing/9ce935435d90

Besides the technical issues, there's an important non-technical one that the author is talking about in this article: Naskh is alien to Hindi-Urdu speakers. And, on the other hand, we know that Nasta'liq doesn't look right to Arabic speakers either.

Yeah, very good article. The need for a Nastaliq style for non-arabs is real. Because best writing systems well suited for running perso-arabic text and without any need for diacritics are Nastaliq and Shikasteh style. I know that most of urdu readers use their ow Nastaliq fonts which in Iran we call it Pakistani Nastaliq. It shows their need to use nastaliq style for publishing, Yet they have newspaper hand written in Nastaliq style.

 

Now, the questions important to us, Persian speakers, is: how come we get along with both styles? how we balance it out? and how that affects our typography?

We are forced to read naskh without diacritics because it has been implemented in horizontal, rigid metal typesetting environment, and nastaliq dos not fit in that environment. In digital era now we have the opportunity to bring nastaliq to publishing and web but we need proper tools and implementation.

To answer the first two, I think it's because we learn writing in Naskh in the first few years of school, and then grow old into Nasta'liq afterwards.

There are several years in Iran that elementary school books are published in a variation of Nastaliq style (تحریری), and it might be shocking for many that the children could get accustomed to the style in a short period of time without any hassle (if you are interested i'll send some photos)

But about the typography, again, going back to one of the first typefaces, Nazanin and Titr, I think these typefaces benefit from both worlds: the alignment of the letters is simply like Naskh, but the shape of the letters are closer to Nasta'liq. This is how our handwriting is, most of the type. Not totally Naskh, not totally Nasta'liq.

 I agree. Nazanin and Titr are very well designed typefaces and they are somehow influenced from nastaliq style. I think they are mostly adapted from the way latin typefaces are designed beside nastaliq script. For example isolated form of jim in Titr typeface has a reversed contrast that does not exist neither in nastaliq or naskh style but it's somehow between stroke type in latin and stroke type of nastaliq.

And maybe it's the same for ligatures. In Perso-Arabic font families, those with many ligatures give me a non-Persian feeling. This could be because of the fact that these ligatures are based on Naskh style, and rarely appear in Nasta'liq.

Exactly, ligature doesn't go with regular pattern of persian texts especially written in naskh style. But structure of those ligatures are not very different between nastaliq and naskh, the only difference is the slanting baseline that makes the harmony of letter fusions majestically consistent. Structure of connection between letters are basically same in nastaliq, thulth, naskh, etc.

Best,

Bahman 

[Behnam Esfahbod]

Dan,

(CCing P-T.)

Dan Parvaz <dpa...@gmail.com> wrote:

Persian went to Naskh (albeit flavored our way) a while back, but Nast'aliq is still our script, I think. I remember sitting in class in Iran with the reed pens we shaved, copying out phrases in calligraphy class (not so well, in my case). And my grandparents' shekasteh looks like a lost art. It'd be nice to get Persian-flavored Nast'aliq produced more reliably on computer. Theoretically, LaTeX would be great for this, but I've never really seen it in action. Farsi Negar comes close... 

I think there's no doubt that Persian Calligraphy is a fine art and Nasta'liq is a beautiful and well-developed style. We grow up learning Nasta'liq and do practice Nasta'liq rules in our hand-writing.

But when it comes to typography, legibility is the main goal, and of course there are other values at stake, including user preferences. I prefer reading Hafez poems in Nasta'liq, and to be frank, they look so weird in anything else! But when it comes to reading a western novel in Persian (like a translation of a J.D. Salinger book), I really prefer Persian Newspaper style (read: Nazanin, Titr, Roya, etc) to any form of Nasta'liq.

Yes, there's a very good chance that such a behavior would not be acceptable just a century ago, but that doesn't mean that me and you, as the users of this writing system in the current time, should not be allowed to have our own preferences.

I believe each of these styles have their own use cases, and every one of them deserved to be alive: be practiced and evolve. I don't see how this issue could be any different from the typography/calligraphy diversity of the Latin world: having many styles--roman, italic, black-letters, etc--and many variants--sans, sans serif, etc--each coming to good use in its own place.

[Khaled Hosny]

On Monday, October 14, 2013 9:33:21 PM UTC+2, Behnam Esfahbod wrote:

I just read this article: The Death of the Urdu Script: Can Microsoft and Twitter save the dying Urdu nastaliq script from the hegemony of the Western alphabet and an overbearing Arab cousin?

https://medium.com/writers-on-writing/9ce935435d90

I find that article very unfair to Naskh, and it draws most of its image about Naskh from hideous typefaces like Arial; whatever Arial has is anything but Naskh.
 

Besides the technical issues, there's an important non-technical one that the author is talking about in this article: Naskh is alien to Hindi-Urdu speakers.

Are you sure about this? all Indian/Pakistani mushafs I have seen are written in so-called Indian Naskh (hand written, not typeset).
 

And, on the other hand, we know that Nasta'liq doesn't look right to Arabic speakers either.

In my experience, Nastaliq (simply called Farsi here) is the second most common style for titling among calligraphers here in Egypt, when I walk in my small town, it is like one in every three or four signs is in Nastaliq.

[Khaled Hosny]

On Monday, October 14, 2013 10:27:53 PM UTC+2, Bahman Eslami wrote:

On Monday, October 14, 2013 11:03:21 PM UTC+3:30, Behnam Esfahbod wrote:

I just read this article: The Death of the Urdu Script: Can Microsoft and Twitter save the dying Urdu nastaliq script from the hegemony of the Western alphabet and an overbearing Arab cousin?

https://medium.com/writers-on-writing/9ce935435d90

Besides the technical issues, there's an important non-technical one that the author is talking about in this article: Naskh is alien to Hindi-Urdu speakers. And, on the other hand, we know that Nasta'liq doesn't look right to Arabic speakers either.

Yeah, very good article. The need for a Nastaliq style for non-arabs is real. Because best writing systems well suited for running perso-arabic text and without any need for diacritics are Nastaliq and Shikasteh style. I know that most of urdu readers use their ow Nastaliq fonts which in Iran we call it Pakistani Nastaliq. It shows their need to use nastaliq style for publishing, Yet they have newspaper hand written in Nastaliq style.

 

Now, the questions important to us, Persian speakers, is: how come we get along with both styles? how we balance it out? and how that affects our typography?

We are forced to read naskh without diacritics because it has been implemented in horizontal, rigid metal typesetting environment, and nastaliq dos not fit in that environment.

 
Bulaq Press in Cairo had 3 (very beautiful, IMO) Nastaliq typefaces by 1873 (and had only two Naskh typefaces), so I don’t buy the argument that Nastaliq does not fit in metal typesetting, it might be hard, but not impossible:
http://typophile.com/node/34937#comment-488511

To answer the first two, I think it's because we learn writing in Naskh in the first few years of school, and then grow old into Nasta'liq afterwards.

There are several years in Iran that elementary school books are published in a variation of Nastaliq style (تحریری), and it might be shocking for many that the children could get accustomed to the style in a short period of time without any hassle (if you are interested i'll send some photos)

I self-taught myself Nastaliq calligraphy at the age of 15, Nastaliq is the easiest form of Arabic calligraphy, believe me :) People who think otherwise are misguided, IMO.

(I wrote those when I was 16, not the pieces of art I thought they were back then, but not very bad either:
https://twitter.com/KhaledGhetas/status/235149342716071936)

[Khaled Hosny]

 
And my father has a Nastaliq handwriting style that he uses for important documents, never seen anyone having a difficulty reading it (that is Egypt, remember).

 

[John Hudson]

On 15/10/13 4:33 PM, Khaled Hosny wrote:

> And my father has a Nastaliq handwriting style that he uses for
> important documents, never seen anyone having a difficulty reading it
> (that is Egypt, remember).

It is often overlooked just how closely related ruqah script, so common
in daily handwriting, is to nastaliq.


With regard to typography, it should be noted that what commonly gets
called naskh is a kind of bastardised, mechanical compromise, which
barely corresponds at all to the complexities of traditional naskh
script. Compared to those complexities, nastaliq is simple indeed: it is
only mechanically complicated because of limitations of typesetting
systems. As a writing style, it is much more regular and involves far
fewer contextual and stylistic variants than naskh.

JH

[Bahman Eslami]

@Khaled 

You pointed a very shocking news to me. I don't think there are many in Iran whom are aware that Nastaliq has been implemented in metal type setting systems. I saw some images in the typophile thread, but could you post more pictures here. I'm sure there are a lot of enthusiasts in Iran to see the publishings.

@JH

The bastardised naskh is called with so many terms like: simplified arabic, 4-paradigm and recently Eurabic (by Thomas Milo). Personally I don't know which definition fits better for the style.

Nastaliq is not interpreted with a right method in opentype. I think the seemingly complexity comes from different type of connection between letters. The problem could be solved by decomposing every letter to a base glyph and connection/terminals. Connection/terminals is a form which is defining the state of the letter in the word. If the letter is in the final of the word we see a terminal after it, if it's in the middle of the word we substitute the terminal to a connection. Form of letter connection is different based on the context after the letter. That way we could reduce number of glyphs designed for a letter but we end up with different type of connections for the letter combinations and many contextual rules. In my head it's very easy to design with this method but as I mentioned in the opentype list, still there is a doubts to implement it in opentype and also generating the code for various connections needs some scripts and tools.

I stumbled on this page and It's an First Grade Elementary text book published circa 1945 and you could see that children were able to read Nastaliq and pass the grades.

Best,

Bahman

[Saadat M.]

On Tuesday, October 22, 2013 5:32:15 AM UTC+5, Bahman Eslami wrote:

@Khaled 

You pointed a very shocking news to me. I don't think there are many in Iran whom are aware that Nastaliq has been implemented in metal type setting systems. I saw some images in the typophile thread, but could you post more pictures here. I'm sure there are a lot of enthusiasts in Iran to see the publishings.

[snip]

On archive.org, there are two works of Turkish poetry that are, according to their metadata, published by the Bulaq Press, and that employ nastaliq. They are:

* Divan-e-Izzet Mola: https://archive.org/details/divanizzethazana00izzeuoft
* Divan-e-Fuzuli: https://archive.org/details/divanfuzuli00fuzuuoft

(I actually got the idea of searching for Bulaq Press's publications on archive.org after reading this page (which I found while googling for "Bulaq Press" some time last year): http://www.loc.gov/rr/amed/guide/nes-turkey.html . It mentions: "The section has acquired many works in Turkish published by the Bulaq Press during the nineteenth century in Egypt. Among these are the Divan-i Izzet Molla (Collection of Izzet Molla's poetry) and the Hamse-yi Nergisi (The Five works of Nergisi), both published in Cairo in 1840. They were printed using the beautiful Nastaliq (cursive script) typefont developed especially for literary works by that press.")

- Saadat

[Saadat M.]

On Tuesday, October 15, 2013 1:27:53 AM UTC+5, Bahman Eslami wrote:

Yeah, very good article. The need for a Nastaliq style for non-arabs is real. Because best writing systems well suited for running perso-arabic text and without any need for diacritics are Nastaliq and Shikasteh style. I know that most of urdu readers use their ow Nastaliq fonts which in Iran we call it Pakistani Nastaliq. It shows their need to use nastaliq style for publishing, Yet they have newspaper hand written in Nastaliq style.

 

 
Perhaps, I can comment on the handwritten newspapers bit, since I am from Pakistan. :-)

Almost all Urdu newspapers in Pakistan use InPage[1] these days, a proprietary desktop publishing system developed for Urdu language by an Indian company. It uses (among others) the Noori Nastaliq[2] font, designed by Mirza Jameel Ahmad and licensed/distributed by Monotype, which uses a ligature-based system for its execution of nastaliq (there are more than 20,000 ligatures in Noori Nastaliq, if I am not wrong). InPage uses its own format for its fonts, and while it can use standard OpenType fonts in its latest version, its Noori Nastaliq (and the recent Faiz Lahori Nastaliq) still relies mostly on ligatures and its own formats. The underlying, proprietary text-setting engine handles all spacing/kerning issues for nastaliq. InPage is very ubiquitous in Pakistan's print world, mainly because it is one of the earliest WYSIWYG Urdu desktop publishing systems on Windows, even before the widespread adoption of Unicode (InPage itself started supporting Unicode only in 2008), and also because other typesetting systems cannot handle nastaliq kerning/spacing as optimally as InPage, thanks to the limitations of OpenType. (An exception is DecoType's Tasmeem, but I hear that Pakistani publishers find it too expensive -- besides, the years long usage of InPage has cultivated habits and workflows that make it difficult to switch.) This has resulted in the usage of Noori Nastaliq for literally everything in the print world including newspapers, magazines, books, posters, and others (and frankly, I am quite tired of seeing this single nastaliq design everywhere).

Examples of Urdu newspapers set in Noori Nastaliq through InPage can be seen at below URLs:
* http://express.com.pk/epaper/
* http://www.nawaiwaqt.com.pk/

[1] http://inpage.com/
[2] http://www.monotypeimaging.com/ProductsServices/wt_fontsample.aspx?type=Arabic

[Saadat M.]

On Wednesday, October 16, 2013 4:15:01 AM UTC+5, Khaled Hosny wrote:

 

Besides the technical issues, there's an important non-technical one that the author is talking about in this article: Naskh is alien to Hindi-Urdu speakers.

Are you sure about this? all Indian/Pakistani mushafs I have seen are written in so-called Indian Naskh (hand written, not typeset).
 

True, naskh is certainly not alien to Pakistanis, specially for the Quran. But for Urdu texts, there is very low tolerance for anything other than nastaliq, specially in print (and increasingly on the screen, too -- as is evident from the article this thread is about). There have been some attempts by enthusiasts to develop OpenType fonts for the Pakistani/Indian naskh and Pakistani/Indian "Quranic" naskh style, but I am not aware if any of the Quran publishers here in Pakistan have used them, or any other form of typesetting, for the mushafs.

- Saadat

[Behdad Esfahbod]

On 13-10-22 02:32 AM, Bahman Eslami wrote:
> Nastaliq is not interpreted with a right method in opentype. I think the
> seemingly complexity comes from different type of connection between letters. The

Nope. The truely hard part about doing Nastaliq in OpenType is how to
position connected pieces in the visual space of the line.

--
behdad
http://behdad.org/

[John Hudson]

On 22-Oct-13 04:35, Behdad Esfahbod wrote:

> Nope. The truely hard part about doing Nastaliq in OpenType is how to
> position connected pieces in the visual space of the line.

Yes. The weakest aspect of OpenType Layout architecture is the
interaction of different kinds of GPOS (glyph positioning). In the case
of nastaliq, we are working with three different GPOS lookup types:
cursive connection, kerning, and mark positioning (with the latter
involving both dot- and mark-to-base positioning, and mark-to-dot and
mark-to-mark positioning). There needs to be contextual interaction
between these different kinds of GPOS, which is something that is both
difficult and expensive (in terms of performance) in the OpenType lookup
model. The contextual interaction is further complicated by the need to
take into account relationships of visual adjacency between elements
that are not logically adjacent in either the character or glyph string
ordering (the barree yeh is the most extreme problem in this regard).

JH

[Khaled Hosny]

Good finding, Saadat.

Also, the first book typeset in the, then, new typeface is the Persian book Gulistan of Sa'di http://books.google.com.eg/books?id=hsJWAAAAcAAJ (1249 A.H., 1833 A.D).

Regards,
Khaled