A Brief History of Transliteration

Language is powerful; and its written word is like a weapon which when launched out into the world enlivens or maims. As cultures and civilisations have moved and evolved through history, the way they have chosen to write their language speaks volumes about their deep motives and aspirations. And of course as one civilisation morphed into another it had to borrow the old words but rewrite them in its new way. This has been going on since writing first began around 5000BC. This is a mighty subject and I can only hope to throw a few personal observations into this post. It is something which affects all of us profoundly, but imperceptibly so.

I want to start from the present and work backwards as this is the only way you will grasp this rather arcane subject. About 23 years ago I was working with the Islamic Text Society in Cambridge UK trying to design books in a computer, a task which hitherto had not been attempted as the criteria for book design were much more demanding than say your average newsletter or magazine, the most a business oriented PC of the time was capable of. One of the priorities was adapting a roman font, Bembo in this case, to accommodate all the different conditions that translated arabic demands and creating the vowels and consonants that don’t appear in the standard roman alphabet. Publishers and writers, mostly academics and translators, have come up historically with various systems of transliteration which typographically were inadequate, like the font Middle East Times, which some might remember from the early 1990s, loved by some academics, which automatically placed a thick line (a macron) above the letter a to correspond to a long aa in arabic. The problem is that you were stuck with the one typeface and it didn’t respect the fact that each letter with a diacritic mark has to be balanced, proportionate and readable at all sizes. Which is why editing a font with these marks is a lengthy process as each character has to be designed carefully and individually … you can’t average it out for all conditions. In other words each diacritic glyph has to be an individual work of art and importantly readable small or large.

Adapting arabic (or any hand-scripted foreign language come to that) into roman type (i.e. English, French etc,) is a recent development but reflects the convergence of cultures in this time, namely the adaptation of arabic words into roman type prompted initially by translations of the Qur’an and the general interest in Islam in the west. The earliest printed Qur’ans in English, e.g. the George Sale 1850 edition (of which I have an original copy here) did their best using marks common in the printer’s palate… eg the long aa would utilise the circumflex mark â and a long uu would utilise an accented ù, and so on. But no attempt was made to make new glyphs, i.e. to place a dot under a consonant which would indicate consonants unique to arabic or to indicate the ayn symbol. In the transliterated fonts which I worked on and which were started off by Monotype back in my time at ITS, we established an acceptable set of characters which have stood the test of time and which have been used in the many books I have worked on since then and by many other publishers and academics who have used the same fonts. We’ve also been able to usefully add arabic salutations (colophons) as part of the same font set. (see an edited version of Baskerville below)

What radically changed things was the introduction of Unicode fonts in the early 1990s which recognised many if not all of the different transliterations in the many roman and non-roman languages in use all around the world. A truly global undertaking. So suddenly most of the  marks we were having to create were available in these new fonts. Also these new font formats allowed for up to 15,000 characters per font whereas previously only 256 characters were available (the ascii set) which necessitated substitutions and juggling to fit in the extra characters. Unicode has revolutionised computer typesetting and particularly the availability of what are called pro fonts which give you every variation of a particular typeface in one font….previously necessitating several fonts to accomodate what you needed. But still the powers that put this together (Xerox, Microsoft, NeXT, Apple and Adobe et alia) unbelievably left off the dotted consonants necessary in arabic translations. (see illustration above) This has meant having to edit the new fonts to create the necessary new glyphs. Also the ayn symbol, a small superscript c, has to be edited into the set.

Mustafa Kamal Attaturk, in the spirit of modernisation after the fall of the Ottoman Caliphate, forced the Turks to change overnight, Nov 1 1928  to be precise, from an adapted arabic script to a roman text, reading from left to right, adding all sorts of diacritics to cope with the spoken Turkish language. He and his accomplices wanted to close the door to its 1,000 year Muslim past and open the door to the Western world. This must have been traumatic for the nation, even though Attaturk himself still went on handwriting his own letters in scripted old Turkish as did many Turks right up to the 1960s. The old scripted Turkish had made an adaptation of arabic to encompass the particular Turkish consonants and vowelling, so to then change into roman script was another huge jump for the populace. Confusing one might think, but in one generation people do adapt – if compelled to. My prediction is that these new transliterated letter forms which we have introduced into the roman alphabet in the last fifty years will in time, maybe 50 years, become absolutely the norm and will be learnt as part of the established alphabet. This must have happend in Iran where arabic was adapted to accomodate the Persian consonants and vowels which don’t occur in pure arabic and which are now absolutely part of the language and now used  by 26 other languages in Asia like Tajik and Urdu.  If we tried to write English in an arabic script (I’ve tried it) we would be faced with the same dilemmas with having to borrow some Persian conventions and even inventing new ones for some of our odd noises. If you have ever tried to read alhamiado, archaic Spanish written in arabic script you would appreciate this problem. (see below) So the form that a language chooses to write in has huge political, religious and social consequences.

Above you see a pilot project undertaken by the poet Emin Alzueta in Spain and myself, which is a page spread of the poem The Beautiful Names of God by Ibn Abbad of Ronda, (1333–1390) a famous Andalusian saint now buried in Fes. Top left is the original Arabic. Top right the Alhamiado version (archaic Spanish) and below right, the romanised old Spanish and below left the modern Spanish translation. The Alhamiado version is a copy from a manuscript found in Spain executed in Andalusian script. To own anything written in arabic script (even Spanish) in post Islamic Spain was severely punished by the Inquisition but such things are still surfacing having been buried in walls and floors of old houses for centuries. The Spanish have their own transliteration system for arabic translations, just slightly different from the English system.

A side subject to this, about which there is considerable disagreement, is the bearing that right left and left right scripts have on the brain and whether this preconditions anything. The Chinese wrote from top to bottom originally and the Ancient Egyptians wrote left to right, right to left and vertically. and the ancient Sabeans and Safaitic cultures wrote boustrophedonically, which means that they wrote from right to left, followed by a second line from left to right mirroring the letters and then back again on the next line in normal writing and so on changing alternately.  It would be an interesting way to write a book and quite easy given the capability of current computer layout programmes. As to why one would do this –  I have absolutely no idea!

These few words barely touch on this complex subject but I hope shows how utterly fascinating it is . When you begin to un-peel written language forms it is like making a postmortem of human history. You see where humanity went and where it is going.

About Ian Whiteman

see www.ianwhiteman.com
This entry was posted in language, miscellaneous, Publishing, typography / design, Uncategorized and tagged , , , , , . Bookmark the permalink.

3 Responses to A Brief History of Transliteration

  1. Iljas Baker says:

    Salaam: I was particularly interested in the side subject. Many cosmologies privilege the right over the left as does Islam and of course the sunna. I’ve no idea if it has an effect on the brain. Certainly implementing the privileging of the right in one’s life according to the sunna slows one down and fosters awareness. But don’t know if it’s more than that. I seem to remember that Henry Corbin wrote about this.


  2. Abdelkhalek says:

    As salamu aleykum,

    Thank you for the immense inspiration you bring here.
    As a fan of your works, and I do not say that fallaciously, I would be most grateful if you could generously inform me what is the typeface used by the ITS on the Ghazali series both for the cover and the text and how can add the diacritics to baskerville as pictured above.

    Jazzakallah khayr

    • Ian Whiteman says:

      The ITS Ghazali series utilised Monotype Bembo, a specially edited version, for the inner pages. The covers were hand lettered by Tom Perkins from Ely although the cover of Letter to a Student was faked by me using partly his lettering but also Bembo. Just goes to show what an authentic timeless kind of Roman lettering Bembo is. Hand lettering is always better but expensive.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s