Karen and I like to listen to the puzzle on Weekend Edition Sunday. In case you’ve never heard it, the format is like this: first, New York Times puzzle editor Will Shortz reminds listeners of the previous week’s puzzle challenge, for which everyone has had a chance to submit a solution by Thursday. A respondent with the correct answer is chosen at random to solve a set of small word puzzles on the air (and gets some puzzle-related prizes), and then another challenge is given to listeners for the following week. It’s a fun 10-or-so minutes of public radio.

It’s not atypical for the on-air puzzles and puzzle challenges to involve thinking about how words sound as opposed to how they’re spelled; Will typically distinguishes the sound-based puzzles by noting that the relevant aspect(s) of the puzzle should be considered “phonetically”. Not that I’ve been paying a ton of attention — and of course, we often don’t solve the puzzle challenge and fail to listen the following week — but I can’t remember ever thinking that the sense of “phonetic” Will uses is misleading or anything like that. Until this morning.

The challenge given last week (Jan. 8) was:

Name an American Indian tribe. Somewhere inside this name, phonetically, is a kind of tree. Remove this tree, close up the remaining letters, and the remainder, phonetically, will name another kind of tree. What Indian tribe is this… and what are the trees?

When we heard the challenge last week, Karen and I figured that the tree-name-to-be-removed was probably simple and monosyllabic: pine or oak, say, but not spruce, palm, or maple. And in fact, that was the answer announced this morning: the Indian tribe is Cherokee, from which you can remove oak and have cherry left over.

But strictly speaking, I felt this was a little unfair (though maybe just because I couldn’t come up with the answer). There’s no oak [oʊk] in Cherokee [ˈʧɛɹəˌkiː], though I’ll grant that when you remove the unstressed [ək] part that is supposed to be the oak part, you do get cherry [ˈʧɛɹiː]. Clearly, this is an example of the type of “phonetic” puzzle where you have to think of the spelling of the word (in this case, Cherokee) and how each syllable-ish bit might be pronounced if it were individually stressed: Cher [ʧɛɹ], (r)ok [(ɹ)oʊk], and (k)ee [(k)iː]. And of course, if you ever felt the need to make a word with Cherokee as its base, such that the stress shifted to the relevant syllable — for example, imagine the adjective Cherokial — then I think we would all agree that the result would be [ʧəˈɹoʊˌkjəl]. But that’s kind of a stretch for someone like me who tends to keep his phonetic-thinking and his orthographic-thinking more separate.