“There are now standard and acceptable ways of transliterating Arabic letters in Latin script. I have not adopted any of these in their entirety. For a non-Arabist, it is not very helpful to be able to distinguish between the two types of h or s or t and readers who do know the language will in any case be aware of these.” (Kennedy, Hugh. The Great Arab Consquests: How the Spread of Islam Changed the World We Live In. Philadelphia: Da Capo Press. 2007. Pg. 9-10)
Transliteration often gets a note at the start of books dealing with Arabic subjects (or other areas where primary and major secondary sources are in non-Roman alphabets). In some books there are whole tables and/or two or three whole pages devoted to explaining how and why the author has rendered a phoneme in such and such a way. Highly specialized works frequently assume the reader’s familiarity with a particular system or give a fast explanation. Less specialized works assume the general reader’s disinterest or give a couple of paragraphs giving the overall idea of whatever system is used. In Watt’s Islamic Political Thought (Edinburgh: Edinburgh, 1968), for example, he explains:
The transliteration of Arabic words is essentially that of the second edition of The Encyclopaedia of Islam (London, 1960, continuing) with three modifications. Two of these are normal with most British Arabists, namely, q for k [with a dot below], and j for dj. The third is something of a novelty. It is the replacement of the ligature used to show when two consonants are to be sounded together by and apostrophe to show when they are to be sounded separately. This means that dh, gh, kh, sh, th (and in non-Arabic words ch and zh) and to be sounded together; where there is an apostrophe, as in ad’ham, they are to be sounded separately. The apostrophe in this usage represents no sound, but, since it only occurs between two consonants (of which the second is h), it cannot be confused with the apostrophe representing the glottal stop (hamza), which never occurs between two consonants. (pg. v-vi)
In another book of the same period, the explaination is even more brisk:
There remains a brief remark about the transliteration system of Arabic names. For reasons concerned mainly with printing difficulties, no diacritical marks appear with transliterated Arabic names. In other words, these names are spelled according to their Arabic pronunciation but without any diacritical marks. The hamza sign is ʼ and the ayn sign is ʿ ; however, they do not appear in an initial position. There should be no serious difficulties arising from this unavoidable system of transliteration. (Bishai, Wilson B. Islamic History of the Middle East: Backgrounds, Development, and Fall of the Arab Empire. Boston: Allyn and Bacon, 1968. pg. viii)
John D. Ruedy spends more time on it the preface to (the first edition of) his excellent Modern Algeria: The Origins and Development of a Nation (Bloomington: Indiana University Press, 1992 and 2005):
As anyone who has worked in Algerian history knows, the researcher is faced with daunting problems of transliteration. The terminology — names, places, institutions — comes from classical Arabic, Ottoman Turkish, and several Berber dialects. When they came to Europe and America, most of the names and terms arrived in French transliterations, which are not in themselves consistent from one area to another or even from one writer to another. I have decided that any attempt to impose total consistency would create more confusion than it would dissipate.
With regard to the names of Algerians before the twentieth century who did not speak or write French themselves, I have generally used a standard English transliteration of Arabic or Ottoman but without diacriticals. For well-known historical figures (e.g., Hussein Dey) and for modern Algerians, I have retained the French spelling. I have also retained the French transliteration of all place names, since this is the way readers will encounter them in almost every other source. The French colonists changed the names of many places during their 132-year occupation, and they also founded many towns and villages. After independence, however, Algerians changed them back or gave them new names. In the text, I use the name prevalent at the time under consideration. The appendix carries a name conversion chart. With regard to Ottoman and Arabic institutions and legal terminology, I have usually, though not invariably, employed English transliteration. (pg. xi-xii)
While introducing the third edition of A History of the Maghrib in the Islamic Period (Cambridge: Cambridge University Press, 1987), Jamil Abdun-Nasr writes:
The problem of spelling place and personal names, with which I was confronted at the time of writing A History of the Maghrib, remains one for which no completely satisfactory solution can be found. This problem arose form the fact that writing of place and personal names in this region has been much influenced by peculiarities in the languages of its various conquerors. Recently the problem has been further complicated by the policy of the governments of the Maghribi countries since independence of using Arabic or Arabized names for places which have been widely known by European adaptations which sometimes bear little or no resemblance to these names. […] However, in order not to confuse the reader unduly, the names by which important Maghribi towns are known to non-specialists have not been changed. Thus Casablanca has been preferred to al-Dar al-Baydaʼ, or Oran to Wahran, Algiers to al-Jazaʼir, Bizerte to Bin Zirt, and Tripoli to Tarablus al-Gharb. Arabic personal names have been transliterated systematically when they are not internationally known in a specific form when consistent transliteration does not lead to much confusion.
In transliteration I have followed a simplified system. Of the usual diacritical marks I have used only ʼ for the Arabic hamza and ʿ for the Arabic gutteral ‘ayn. The ˉ over a vowel to indicate that it is long I have used only when its omission would make it difficult even for Arabic-speaking persons to recognize in transliteration the assimlation of the definite article al- before the qamariyya letters. Hence Sayf al-Nasr, although his name would be pronounced by an Arabic speaker as Sayfu ʼn-Nasr. (pg. xi-xii)
Arabic transliteration can be controversial: the great debate over how to render the Libyan leader’s name seems to never end. Journalists love the subject; it fits into the popular narrative around the dictator’s personal eccentricities. It is as if to say: And if you thought his outfits weren’t weird enough, we can’t even figure out how to spell his name! It makes an exciting add-on for some; for those who have better Arabic it can actually be an interesting bit of human interest. The resolution sometimes turns out blaming Arabic for being so difficult, what, with all its grammar and peculiar sounds and stops. How utterly fascinating.
In Kennedy’s case this is an Arabist — not a journalist. He has written substantive books before. Not only the popular When Baghdad Ruled the Muslim World: The Rise and Fall of Islam’s Greatest Dynasty (Da Capo, 2005), but also of The Prophet and the Age of the Caliphate: The Islamic Near East from the Sixth to the Eleventh Century (Longman: 1986; revised, 2004). In his other books there is hardly deep discussion of transliteration. But on page 6 of The Great Arab Conquests he writes: “Arabic sources use the term conquest (fath) to describe the taking over of the lands of the Byzantine and Persian empires.” He clarifies that fath (فتح) derives from the fth (ف ت ح) root; the root from which we get the title of the first sura of the Qur’an, (سورة الفاتحة) “the Opening.” Thus it has a special connotation in Arabic that conquest does not in modern English. Since Kennedy does not explain his transliteration, what stops the non-Arabist from reading his فتح (fath) as something where the t is actually a th (as in faith or the Arabic letter ث) rather than where the h (Arabic ح) is its own sound not linked to the t? None of that stops The Great Arab Conquests for being a great book for its own purposes, even if it does not bother with digraphs and macrons (or should it be macra?) and all that. In his other works he makes full use of these. There are many quick to moan about the irrelevance or stupidity of debates and discussions over transliteration — and they are usually right.
This blog hardly ever agonizes over transliteration; there are more important things to consider. What matters is that the word can be easily recognized and pronounced, not that it follows anybody’s rules. Especially in the Maghreb the problems associated with transliteration are manifest: the French (phonetic) system or some haphazard English-phonetic system used by journalists and scholars are often predominant. It is on occasion the case that two individuals with the same (Arabic) name spell them in wildly different ways using Latin letters. Two acquaintances are Abd el-Rahmans (عبد الرحمن). One spells it Abdelrahmane the other uses Abderrahman. Another, with the same name uses Abdalraham. These are minute differences, and they can produce their own results in certain circumstances. Use of one or the other spelling may put one in common with a completely different individual. There are Khalids and Khaleds; Mohameds and Muhammads; Belkasems and Belgacems and Belkacems; nowadays such confusion can be especially erroneous in a number of directions.
Imagine young people going to school in the United States or the United Kingdom writing their names with dots and macrons, somebody named Muḥammad asking his principal to make sure that his diploma had a dot below the h — ḥ. It always amuses to see these used in older books for personal names that do not otherwise make use of them; in the “Preface to the First Edition” of Philip Hitti’s History of the Arabs (as seen in the tenth edition), he acknowledges a former student named “Buṭrus ʿAbd-al-Malik” using a dot below the t to signify that it represents a ط rather than a ت (pg. xii). How this Buṭrus actually spelled his name is a mystery here — it is doubtful he used the ṭ. (Nowadays a capital T is used in technical texts; this is conventional for many of the “emphatic” letters.) In another of Hitti’s books (The Syrians in America, Gorgias: 2005, orig. published 1924) he notes in a footnote to Appendix A that “No uniform system of transliteration was used in the proper names of priests which occur in the Appendices. The names were rather spelled as the owners would spell them.” (pg. 125) Dots do not seem to have caught on among western Arabs, at least not in everyday usage. Individuals sometimes use the “modifier letter left hand ring” (ʿ) or other funny looking “apostrophes” for ayin (ﻉ) or ʼ for hamza (ء). As’ad Abukhalil, who pays close attention to transliteration, favors this, for instance. There seems to be no consensus around using “ou” or “oo” or “u” for waws (و) and dammas or how to represent shaddas. One can find Maqdisis who have something more Celtic-looking like MacDesey or the like. One becomes frustrated with the various ways ghayn (ﻍ) is represented: a kh or a gh or a g or something else with a dot or a macron or whatever. One sees multiple Arabic names with the same letters represented differently in the same article. This blog has alternated in usage, for the same individuals names; مختار is sometimes, carelessly, Moctar, Mokhtar, Moktar, etc. Some of those have been fixed, others have not. Muhammad for محمد is hardly used here — Mohamed is better, not for any real reason other than personal preference. This is really what it comes down to very often; utility and preference. Consistency matters, too.
UPDATE: Brian Whitaker, of the Guardian and al-Bab, offers his thoughts; Whitaker’s blog also has a summary of transliteration schemes that makes for useful and interesting reading. He also gives examples of some recent blunders in transliteration from the UK and elsewhere, as well as Colonel Lawrence’s goofy response to demands that he correct inconsistencies in his transliteration of place and personal names. Whitaker wrote the fun article on Qadhdhafi‘s name (or Qadhafi or Qaddafi, as it is sometimes written here) referenced above.