[First off, links to the current puzzle and solving guidelines]

The other day, one of our regular solvers voiced an objection to this clue in last week’s puzzle:

CHARITABLE  Generous one moving in two pieces of furniture (10)

Here, the wordplay points the solver to two pieces of furniture (CHAIR, TABLE), with I (“one”) changing position in it to yield the answer (“generous”). This clue struck our friend as unfair, because he felt that it came too close to violating the taboo on “indirect anagrams.”

Before considering this complaint, let’s take a minute to rehearse what constitutes a forbidden indirect anagram. By longstanding convention, a constructor who asks a solver to anagram a group of letters is expected to supply those letters (the so-called “anagram fodder”) explicitly in the clue itself.

What’s not acceptable is to give the solver a synonym of the anagram fodder, and expect him or her to first find the intended synonym, and then put its letters in the right order. That’s an indirect anagram, and it’s not hard to see why it places an undue burden on the solver. Most words can have a number of different synonyms, so it’s not always easy to know when you have the right one; to then be asked to rearrange the letters of a word that might not even be right is a step too far.

There’s no question that our clue for CHARITABLE has a bit of this problem to it. But we would argue that the distinction between indirect anagrams and legitimate clues is not entirely clear-cut. Rather, they lie at the two ends of a spectrum that contains a graded series of wordplay strategies—and our clue was an attempt to see whether we could nudge the line of acceptability a little further without sacrificing fairness.

What all the clues on this spectrum have in common is this: the solver is asked to (a) find a synonym for a word or phrase in the clue, and then (b) perform some kind of operation on it. What varies along the spectrum is how strictly determined the operation is.

So at one extreme are clues in which no operation is performed at all; the synonym is simply placed directly into the answer. Here’s a simple example from last week:

CARNATION  Pink auto country (9)

There’s nothing for the solver to do here except find the appropriate synonyms for “auto” and “country” and glom them together. At the opposite end, the operation is anagramming, and there’s general agreement that that is too indeterminate a process to be fair.

But between these extremes, consider the range of things a solver might be asked to do to a synonym once it’s found:
1) reverse it
2) say it out loud (i.e., find a homophone)
3) remove the first or last letter
4) remove the nth letter, where n is specified, or where the letter is specified some other way
5) change a specified letter to another specified letter
6) remove the first or last half, or some other specified fraction
7) move a specified letter to a new and specified location
8) move a letter from a specific location to a new and unspecified location, as in these clues
(from Puzzles 3233 and 3236, respectively):
PISCES  Postpone first bit of seasoning for fish (6)
ABDOMEN  A black cat, perhaps, with head stuffed inside belly (7)
9) move a specific letter to a new and unspecified location
10) move an unspecified letter to a new and unspecified location
11) change an unspecified letter to another unspecified letter

Now, note that although these processes shade into one another, they vary in their acceptability in a cryptic clue. 1, 2, 3, and 4 are uncontroversially OK. 5 is fine as long as it’s quite clear, while 6 tends to be more or less accepted depending on the length of the word. 8 is unorthodox, but we’ve done it without complaint, and 11 is undeniably too ambiguous to be fair. In between comes 9, which is the CHARITABLE clue—and it’s just a hair further along the spectrum than 8. Frankly, we don’t rule out some day using 10 if the letter shift is interesting, and if instead of a general anagram indicator the clue makes it clear that the change is minimal.

The point, as ever, is that a lot of what solvers come to see as fair or unfair in a cryptic puzzle is a matter of convention and expectation, and often those conventions can be unnecessarily narrow.

The bottom line is that we consider moving a single letter to get from CHAIR TABLE to CHARITABLE to be fun and interesting. Likewise SPICES to PISCES, with an interesting change in pronunciation, and BAD OMEN to ABDOMEN. All three would be boring as anagrams, but as letter shifts they’re quite entertaining—and isn’t that what this game is all about?