Janet Testerman, of Naples, Fla., writes: “I am teaching a high-school world-history class for the first time this year, and I’m baffled. At the bottom of every page of the 1,000-page textbook, there are pronunciation guides for words on that page. The Sui dynasty is pronounced ‘Sway.’ Ogodei, an important descendent of Genghis Khan, is pronounced ‘Ergoday.’ Why don’t they print translations of foreign words phonetically in the first place?”

I’ve had the same thought. But what is a “foreign” word, and what’s a “phonetic” spelling? How to treat foreign words in general is a big subject, so maybe we should stick to discussing ones like your two examples: proper names. No matter how we define “foreign” names, the category ought to include Sui and Ogodei, which come from languages that don’t even use our alphabet: Chinese in the case of Sui, Mongolian for Ogodei.

Consider Sui. Chinese is written in a way that only hints at pronunciation, so to represent it in our alphabet, we’ve had to start with the language as it’s spoken—by somebody, though naturally there are dialects. Different systems of romanization have been tried. Early efforts were so unsystematic that today’s experts have a hard time explaining, for instance, why the names Peking and Canton were given to those cities. Then the Wade-Giles system was imposed, and it remained standard for most of the twentieth century, though its rules were often broken or bent. (Hardly anyone ever bothered with the strict Wade-Giles renderings of those city names, Pei-ching and Kuang-chou.) In the past few decades, scholars have been switching over to the Pinyin system, which gives us Beijing and Guangzhou. The spelling of Sui remained unchanged from Wade-Giles to Pinyin—even though everyone agrees that nonexperts could intuit the Chinese pronunciation more easily if the name were written the way you wrote it. But since the expert consensus is in favor of Sui, who are we to lobby for Sway?

How to phoneticize words from languages that are written with the Roman alphabet is murky for different reasons. In the 2,500 years or so that the Roman alphabet has been around, languages as varied as Swahili, Czech, Vietnamese, and Esperanto have adopted it. Then they’ve adapted it to their purposes and reached their own conclusions about what sounds the letters should represent.

Time was, we English speakers blithely respelled non-English speakers’ names when we felt like it. Certain Austro-Hungarians who signed their names Habsburg, for instance, have long been known to us as Hapsburgs. These days, though, such ethnocentricity is rightly considered disrespectful. And anyway, English itself is notoriously unphonetic, even unto people’s names. (A spectacular example would be the nineteenth- century British journalist Walter Bagehot, whose name is pronounced “Baj-it.”) If we want to tidy up the whole world of names as they appear in English- language texts, we probably ought to start by getting our own house in order.

Richie Babb, of Chesapeake, Va., writes: “Am I the only person who is driven to distraction by the unnecessary use of y at the end of words that end in -ence? ‘There is a danger of dependency (instead of dependence) on sleeping pills.’ ‘There is a growing insurgency (instead of insurgence) in Iraq.’ Am I missing something?”

Remind me, please: Did Ralph Waldo Emerson say a foolish consistency is the hobgoblin of little minds, or a foolish consistence? Foolish or not, looking for consistency in such suffixes is bound to drive anyone to distraction, because it doesn’t exist. I’ll admit that the word dependence is older than dependency, but both words were in use before the first permanent American settlement was established at Jamestown, in 1607. Dependent preceded them both. Similarly, insurgent was an English word before either insurgence or insurgency—but insurgency seems to be a few decades older than insurgence.

We have credence but clemency, turbulence but tendency—and emergence and emergency, which have become fully differentiated words. Dependency remains less common than dependence—at least, judging by how often the words appear in U.S. news sources. But insurgency is much more common than insurgence and is to be preferred.

Do you have a language dispute? Write to Word Court in care of The Atlantic Monthly, P.O. Box 67375, Chestnut Hill, MA 02467, or send e-mail to msgrammar@theatlantic.com. All letters become the property of Word Court. Ms. Grammar is also on the Web, at www.theatlantic.com/courtrecord.