First, get the dictionary. Mine is the handsome Far East Chinese-English Dictionary, published in 1992 by the Far East Book Company, Ltd., of Taiwan. It was purchased at the Caves bookstore on the North Road in Taipei, while I was living and working there in 1993-94. My copy is stamped in red as No. 203481.

I had an idea of becoming fluent in Mandarin, but that dream went glimmering.

Still, I get the dictionary out now and then when I want to look up a character. And recently I used it to “back translate” into English a Chinese version of one of my own poems, translated by a Taipei friend, France Yu, and published in the Taipei Journal of Dentistry.

But this is not explaining how to use the dictionary—and to explain requires another short detour into the mysteries of Chinese orthography. Chinese is one of the world’s oldest written languages, evolving from ancient marks on tortoise shells and “oracle bones.” It is a “broken” syllabic language; the Chinese character stands for a sound rather than being a pure hieroglyphic. Several of these characters may go together to make up a word. The characters (letters) are based on 214 radicals or roots, which is what makes it possible to have a Chinese dictionary.

Not that this makes anything easy. Some radicals are extraordinarily complex. Radical 214, “ywè” (flute), is a large character in itself and combines with only two other characters in my dictionary. Other radicals are hidden within their characters, so only a fluent reader can see instantly where the word will be found.

And then there’s stroke count. Under each radical, my dictionary arranges characters in ascending order of count. (I believe the most complex Chinese character has 64 strokes.) Counting strokes might seem easy, but it’s not, at least for a Westerner. Certain conventions dictate the count—unless you know this code, you may count a character and be several strokes high or low. It has to do with the action of the brush—the actual stroke used by the calligraphic artist.

Discouraging? Not entirely. The Far East Dictionary has helps for Westerners. To begin with, it numbers every character in its 1,759 pages, from 1 (yī) to 7,331 (yù). The characters themselves are in large type and easy to scan.

In the front of the book is a list showing each of those 214 radicals and the page on which its characters begin. So if you can spot the radical, you’re now down to looking at perhaps 3 percent of the 7,331 characters. If you know the stroke count, you can narrow this still more. There is also an index listing every character under its radical, although this type is rather small. A good magnifying glass is helpful

But suppose, like me, you’re a rotten stroke counter and spotter of radicals. The dictionary offers help here, too. In the back, there’s the Mandarin Phonetic Symbol Index. There are also two lists of all 7,331 characters arranged phonetically by their sounds in English—the U.N. Mandarin Phonetic Symbol Index  and the Gwoyeu Romatzyh Index. (My dictionary omits the famous and formidable Wade-Giles index, for which I’m grateful.)

If I have heard the word pronounced, I can go to one of these lists and with luck (sometimes a lot of it) find the anglicized word and next to it the Chinese character I’m looking for, with its number in the dictionary.

How does this work in practice? Here, in brief, is how I tackled France Yu’s translation of my little poem, “Birds of Taiwan,” in 43 Chinese characters.

The title was easy. France had stuck to “Taiwan Birds” (“niăo”). Her opening words also followed my scheme. But then, as good translators do, she moved away from a slavish translation of the original and created her own fine poem in Chinese, based on mine but unique.

A few characters I knew. I could spot the radicals in others and find the words in the dictionary. But then I came to two characters that baffled me. I could locate one of them and get the phonetic “kāi,” which is character 6,533 in my dictionary. But “kāi” is an extensive character with definitions that go on for several pages. So I wondered if it was paired with the second character in my puzzling phrase? Yes, it was. The full word was “kāi-kĕn,” meaning “to open wasteland to farming.”

With the phonetic for this second character, I could then go to the lists at the back of the book to identify “kĕn” as character 883, under the radical “tŭ” with a stroke count (I think) of 12.

Some characters were tougher. I recall marching for a long time through all possible permutations of the sound “zhi,” trying to guess what the creators of the phonetic indices were thinking of.

In the end, I found all 43 characters and could read France’s nice translation and even speak it aloud in Chinese. And this all took only two days of solid work.