For ease of handling during checking and updating, and to ensure a fair draw, all entries are cut about 1mm outside the grid/address panel perimeter (or rectangular equivalent if an irregular shape).

This statement, taken from an addendum to the Listener Crossword checker's statistics for 2001, should make it clear to both solvers and setters what the boundaries of a puzzle solution are. However, Listener Crossword No. 4422, Buried Treasure by Poat, expected solvers to highlight four consecutive letters in the preamble below the grid/address panel.

I can appreciate that the Listener crossword editors had a problem here. Poat must have put an awful lot of work into this puzzle, and they would have been loath to simply refuse to accept it. But surely they could have tweaked the wording to make it clear that the solution required solvers to think, quite literally, outside the box. Or they could have tweaked the wording another way to make it clear which of the possible solutions figuratively outside the box, but literally inside it, they preferred.

The checker has sent those who wrote to him about this puzzle a breakdown of all the different solutions submitted. Less than a third of the solvers who submitted entries came up with the expected solution, and of course that doesn't take into account solvers who simply failed to think of any solution at all. But while some of the other solutions would, I think, be hard to defend, at least two are in my opinion hard to deny, given that some thinking outside the box (either literally or figuratively) is required.

The essence of my own solution is elegantly illustrated by Dave Hennings here, the only significant difference being that my letters were upright. (For those of you not familiar with Listen With Others, it's worth going there each week to see the latest example of Dave H's artistry.) The checker's reasons for rejecting this solution are: "Uncalled-for letter repositioning, the ER pair thus not side by side. Drawing a line is not really highlighting." However: 1) there's nothing to say that letters have to be in the centre of their cells, especially when thinking outside the box is called for; 2) the first definition of "side by side" in Chambers is "close together" (you'd have thought the checker would have bothered to look that up before raising such an obviously bogus objection!); and 3) although the checker illustrated this particular solution with a line through the letters, the highlighting in my solution (and I suspect Dave H's) completely surrounded the letters. The checker grudgingly adds that "This is the only solution that might be considered to be 'in' the SEARCH AREA string of letters in the grid." (But see below.)

The second hard-to-deny solution is slightly further outside the box (figuratively speaking), but is so wonderfully inventive that anyone with a heart would have to accept it. If you look again at Dave H's illustration, you will see that the word LIFT appears vertically in column 8 (at the start of 19dn) with the L splitting the word HARE in row 5. What several solvers did was to cut out LIFT (except for a hinge at the top or bottom), so that obeying that instruction not only made the letters of HARE consecutive but also added a new branch to the SEARCH AREA, which now included this HARE. I'm slightly annoyed that I didn't spot this delightful solution myself. I must have come tantalisingly close because I was certainly looking for some way of excluding the offending L. If I had, it's just possible that I'd have chosen that solution instead of the one I did; but perhaps it's just as well I didn't, since the checker suggested that those who did might be guilty of collusion. (Does he have any evidence for this, I wonder, or is it pure speculation? I'm reluctant to go grubbing round the murkier parts of the Web myself, but if anyone does have any hard evidence of collusion relating to this or any other solutions - including the expected one - I'd be interested to hear it.)

The checker also posits collusion when attempting to deal with the question of the boundaries of a puzzle solution:

Frequently mentioned in pre- and post-solution comments were the requests in the annual statistics concerning preferred submission presentation. These, which only reach about one third of all of any year's solvers, are not rules. The heading on the appropriate section states 'It saves the checker much time if ...'. Were the requests regarded as rules, the weekly error rate would be in the order of 90%. That this was indeed so often mentioned again suggests it may have derived from comments in chatrooms.

This is simply disingenuous. The preferred submission presentation makes it quite obvious (as if common sense wasn't enough) what the boundaries of a puzzle solution are, and the paragraph from the addendum to the 2001 statistics cited at the top of this posting confirms it. Since this is the main objection to the expected solution, it is hardly surprising that it was frequently mentioned.

The checker's addendum to the 2001 statistics concludes with a brief section on "alternative answers":

These are very rarely allowed, whatever the arguments in their favour. Decisions are always made well in advance of solution publication, with which it would be hoped to indicate accepted alternatives, but this cannot be guaranteed.

Appeals are therefore futile.

I don't know about you, but there's something about this I find disquieting. To me it feels mean-spirited, and somehow not very British. OK, so it's only a crossword, but there is such a thing as "fair play", which ought to include the right of appeal, particularly when the arguments being put forward are as misconceived as the checker's. Anyway I'm not going to be cowed into suppressing this rebuttal of the checker's arguments, if only to give the setter and editors the chance to consider the possibility that they may have been wrong.

Current Location: Ealing
Current Mood: worried