Romanisation of Arabic words

The 'Romanisation' of Arabic words in English texts is often required in translation. Romanisation is important for names of places and people, shari'a terms, or for words for which there exists no equivalent term in the target language.

Romanising Arabic words can prove problematic for the translator, as there are several different methods to approach the exercise of transcription or transliteration. The choice on which method to use is dependent on several factors, including the subject material, intent of the author, the target audience and the resources and equipment available to the translator.

Transcription

Transcription is the process of rendering what can be seen in the Arabic script into the the Roman alphabet, - in contrast to transliteration, which is the process of rendering what can be heard into an equivalent readable form in the target alphabet. Transcription involves taking the Arabic word and replacing the letters with the 'corresponding' letters in the target language. This comes with its own difficulties as there are 28 letters in the Arabic alphabet and only 26 in the Roman. Many of the letters in Arabic in fact have no corresponding sound in English, or corresponding letter in the Roman alphabet, - in fact only the characters ز، ر، م، ن، ل، ك، س، ف، ت and ب have clear equivalents in English (l, n, m, r, z, t, f, s,k and b)
As such non-standard letters and symbols are employed in transcription.
Rigorous methods need to be followed when transcribing Arabic into the Roman alphabet, as the process should be entirely reversible. That is to say, once transcribed into the Roman alphabet it should be able to be reconverted, by man or machine, back into the Arabic script without error.
The bulk of a translators work will not necessitate the use of transcription.
Correct transcription cannot be understood except among the specially trained, and as such would only be used for such an audience. Transcribed Arabic will appear meaningless to the reader who is not trained in the conventions of transcription, and who has no knowledge of the Arabic language. As transcription renders what is written, and not what is heard, and Arabic does not display short vowels, these will not appear in true transcription. Thus, the verb كتب would appear as ktb, and would be incomprehensible to those unfamiliar with the language and convention.
An alternative to strict transcription is to use expanded transcription- that is, inserting the undisplayed vowels into the target text. This allows for a better understanding for untrained readers - ktb would become kataba, and as such would be more easily read. However, expanded transliteration still leads to inaccurate pronunciation amongst those who are not familiar with Arabic as it leaves out many characteristics of the language. For example, the gemination that occurs with defined 'sun' letters is not displayed in transcription, so الشمس, transliterated would be rendered "alshams". Only a reader already familiar with the language would correctly pronounce this "ash-shams".

Phonetic Spelling

Phonetic spelling, using the international phonetic alphabet is the only way of assuring correct pronunciation of the Arabic source word. However, this is highly academic, and again, can only be understood by trained linguists and specialists. It also uses non-standard symbols and characters, and as such is impractical for most use.

Transliteration

Phonetic transliteration is not an exact science. Trends in transliteration have changed and evolved as the understanding of the Arabic language in the West has developed. Transliteration attempts to render Arabic into a readable form in the target language, whilst retaining an accurate pronunciation of the Arabic source word for the untrained reader.

The development of transliteration trends proves for interesting reading. An oft- cited example displaying the development is that of the transliteration of الْقُرْآنَ,- early Orientalists, unfamiliar with Arabic often wrote as the alcoran, mistaking the definate article as part of the word. Later koran became more widely used. Recently there has been increased use of the transliteration quran, which preserves more accurately it's Arabic counterpart. In addition, qur'an is also employed, which is even more accurate, as it uses the apostrophe to indicate the 'alif madda, and thus is more faithful to the Arabic written form

There are several different transliteration styles and conventions in current use. Despite attempts by different bodies, there has been no universally accepted method of transcription. This is mainly due to the fact that the roman alphabet is not equipped to deal with the scope of sounds employed in the Arabic language. Thus the transliterator is faced with trying to employ the closest counterpart of letter clusters in the target language, or must create letters and shapes that symobolise the Arabic sounds.
The sounds created by the characters ي، و، ه، ن، م، ل، ك، ف، ش، س، ز، ر، ذ، د، ج، ث، ت، ب، and ا all have eqivalent approximate sounds in English and can be transliterated
through the use of individual letters or consonant clusters in order to mimic the sound in Arabic.
However, for the sounds of ص، (a voiceless alveolar fricative), ض، (a voiced velarised dental stop), ط، ( a voiceless velarised dental stop), and ظ، (a voiced apico-interdental velarised fricative), have no exact equivalent in English, and thus are harder to transcribe. However, they are, broadly speaking, emphatic versions of the s, d, t, and th sounds. For this reason, if diacritics are not being employed, these emphatics are sometimes represented by using a capital S, D, T, and TH respectively.
There are a further 5 letters for which there is no similar sound in the English language. These are the ones which pose the greatest challenge in transcription. They are are as follows:
The letter ح, a voiceless pharyngeal fricative;
The letter خ, a voiceless velar frictive sound, which has no equivalent in English but is similar to the Scottish 'ch' sound, as in 'loch';
The letter ع, a pharyngeal frictive;
The letter غ, a velar frictive;
The letter ق, a voiceless aspirated velar stop.
Without using diacritics, these sounds are hard to represent simply because they are not in the repertoire of the English language. Often they are represented as h, kh, `, gh, and q respectively, which is generally accepted, as is readable to those unfamiliar with Arabic, though causes the original sound to be somewhat lost. Finally the hamza, a glottal stop inexistant in English, is often portrayed using an apostrophe (').

Whilst transliteration of Modern Standard Arabic (MSA) is most common, the regional variations in pronunciation can cause discrepancies in transliteration (the 'g' sound in the Egyptian dialect, compared to the 'j' sound employed in the Levant, for example).
The nature of the target language will also lead to different transliterations of the same Arabic word. For example, شيخ would generally be rendered sheikh for speakers of English, but for french speakers would be cheikh.
Most Western media use different forms of phonetic transliteration, though often there is no rigor in there use, and variations can be seen used even within the same publication.

The main proposed transliteration standards are:
The American Library Association and Library of Congress Approved Romanization Tables and Transliteration Schemes for non-Roman Scripts. (ALA-LC).

The International Standards Organistion; ISO 233:1984, Transliteration of Arabic characters into Latin characters.

The British Standard BS 4280

The United Nations Romanization System for Geographical Names is useful reference for suggested transliterations of place names

All of the above transliteration standards involve the use of non-standard symbols or characters. This makes it difficult for those using a standard QWERTY keyboard

The Buckwalter transliteration standard is a useful reference as it is a proposed standard for transliteration using only characters found on a standard QWERTY keyboard. It is easily employed by those familiar with Arabic but suffers from the same problems as strict transcription, in that it replaces letters and not sounds, so can lead to incorrect pronunciation amongst those unfamiliar with the language.

An interesting 'Instant Messaging' Arabic transliteration is growing in popularity, particularly amongst the younger generations. This does not employ any non-standard symbols or characters, and instead uses numbers to symbolise those sounds which are foreign to English, as follows:
3 represents the letter ع .
5 or 7' represent the letter خ .
6 represents the letter ط .
6' represents the letter ظ .
7 represents the letter ح .
8 represents the letter ق .
9 represents the letter ص .
9' represents the letter ض .
2 is employed to represent the non-initial أ .

This seems to be an organic development, and whilst growing in popularity is unlikely to become part of the translators repertoire, as cannot be employed for formal use.

Exceptions

There will always be exceptions when it comes to transliteration, even if rigorously employing a certain method or convention. These are generally with names and places. When an individual indicates a preferred spelling of his or her name then this should be used, regardless of whether it fits the standard being employed. Likewise, when there is a widely accepted spelling of a name or a place this should be used, even if this does not conform to the convention- for example, ابن سينا is known in the West as Avicenna. Whilst Ibn Sina can be used as well, it is helpful to the Western reader to include Avicenna, even if only in parentheses.

The above information attempts to display a brief overview of the different forms of romanisation available for the translator. There are many different styles and methods available, all of which have their own benefits and faults. The best style to use is the one that will be best understood by the target audience, whilst simultaneously retaining as many qualities as possible of the original Arabic.