The first rule of foreign policy, says the adage, is never to invade a country where they use q’s without u’s. Besides saving the Republic from overseas boondoggles, I like to think this chestnut exists to rescue American copy editors from endless niggling over Arabic names. If war is, as Paul Rodriguez quipped, God’s way of teaching Americans about geography, it is also a way of teaching his humble servants at the Atlantic copy desk how to cram foreign words and phrases into an alphabet that manifestly doesn’t suit them.
Since terrorism and Iraq began to dominate our coverage, The Atlantic has crammed many Arabic names and phrases into our text, and some have not gone gently. The simplest cases demand fussiness and attention: We spell Iraq’s second-largest city Basra. But why not Basrah, as writers in English called it for hundreds of years before? And why no dot under the s, as scholars seem now to prefer?
I discern three major currents of thought on transliteration. Our own policy, following that of other fastidious copy desks, is somewhere at the intersection of the three.
1. Spell Arabic names using a strict one-to-one correspondence between Arabic letters and English ones.
Generations of pedants have bequeathed us highly arcane and exact systems for putting Arabic into English, and for those who believe in consistency and precision, the temptation to use one is immense. But it is a temptation The Atlantic resists.
Adopting a strict system—one that truly honors the nuances of Arabic—would mean festooning our copy with strange apostrophes, with dots, with macrons, with seemingly unpronounceable clusters of consonants. English simply lacks the sounds of Arabic, and to mark the differences would demand an arduous regime of annotation. Because Arabic has two s’s, for example, a word that is easy on our eyes, like Basra, would require a disfiguring mark—usually a dot under the s—to distinguish it from its dotless, nonsense counterpart.
Similarly, the name of the leader of Libya could, according to one strict scheme, be spelled Qadhdhafiy. We spell it Qaddafi, which doesn’t do justice to the consonants in the middle of his name but is a much less forbidding jumble of letters. What we render as dd sounds in Arabic not like our d but like a hollow buzz, a th (“ this”) pronounced with the tongue arched down to open up the space of the mouth. The dh of strict transliteration is meant to capture that sound, and those who know Arabic will recognize it as the Arabic letter dhal.
But at what price consistency? H. W. Fowler wrote a full-throated protest against transliterative didacticism in 1926:
We owe no thanks to those who discover, and cannot keep silence on the discovery, that “Mahomet” is further than “Mohammed,” and “Mohammed” further than “Muhammad,” from what his own people called him. We want one name for the one man […] and the one should have been that around which the ancient associations cling. It is too late to recover unity; the learned, and their too docile disciples, have destroyed that, and given us nothing worth having in exchange.
The sole accomplishment in using the strict system is, as T. E. Lawrence pointed out to his editor in response to a query about inconsistent Arabic, to help people who know enough Arabic not to need helping. And what’s worse, for the ordinary reader the page just gets trashed with annoying and meaningless symbols, most of which bring one no closer to authentic Arabic sounds.