Listener 4196, Here and There: A Setter’s Blog by Hedge-sparrow
Posted by Listen With Others on 24 July 2012
Having constructed a few Listener crosswords with scientific themes, I thought I’d like to try an alternative tack for a change. I found myself wondering one day how easy (or otherwise) it would be to construct a grid in which half the answers had a letter removed before entry, and the other half had that same set of letters inserted, always producing real words and with the usual rules for grid construction, symmetry, etc. “Here and There” was the result.
In fact, the construction of the grid was not as difficult as might be expected: perhaps I was just fortunate that it all came together reasonably quickly. As always, the process of making up the clues took much longer. I also had to think carefully about what information solvers would actually need to have in order to reach a solution: for example, should I clue the words forming the unmodified answers, or the modified grid entries, or both? At first, I was minded to clue the unmodified answers and then have one-word definitions of the actual grid entries randomly distributed throughout the clues as “redundant” words. I can’t remember why I decided to move away from this idea, but I ended up clueing just the unmodified answers, and leaving solvers to deduce the fact that each entry required one letter either to be added or removed. The necessary information that “All entries are real terms” was included as “letter insertion” or “letter deletion” misprints in the down clues (rather than just being stated in the preamble) in order to provide an additional hint of the required grid entry technique.
One consequence of this was that not one of the actual grid entries was clued. I was concerned that this might not be considered “fair” by solvers, but the editors seemed to be happy with it. I also had to make sure that the solution was unique: most of the amendments to grid entries are “forced” by crossing entries, but there are a couple (20 dn and 32 dn) which are ambiguous. Eventually I convinced myself that the fact that the sets of extracted and inserted letters are the same “forces” the amendments at these entries, and that the solution is indeed unique.
The infamous “special” clue was really a bit of an afterthought. Having constructed the grid and started on the clueing, I began to wonder if it wasn’t just a bit too “mechanical” and dull. The special clue with its possibility of two answers, either of which (after amendment) could form the entry at 28 ac, was really intended to provide a bit of additional interest to the solving process, but it also became the means by which the equivalence of the sets of extracted and inserted letters is established (see below), again rather than simply stating this fact in the preamble.
I think most solvers will have deduced early on that all entries are either one letter shorter or one letter longer than the clue answer. The preamble didn’t explicitly state that there are equal numbers of both types, but this could obviously be deduced by solving all the clues and checking answer length against grid entry space: also, as noted above, it is hinted at in the down clues which are also of “letter extracted” or “letter added” type, and in equal numbers. But the key to the crucial fact that the sets of extracted and inserted letters are the same is the “special clue” 28 ac: “Mental disorder: Hebrew woman rings NICE doctors for a pill”.
What I intended should happen, once solvers had identified this as the special clue, was: (i) solvers would identify the two separate definitions for possible entries at 28 ac: “Mental disorder” = MANIA, and “Hebrew woman” = MARIA, giving possible entries MANA or MARA once the letter “I” is removed (the preamble does state that all answers require amendment before entry in the grid, including this entry for the special clue); (ii) notice that the remaining phrase in the clue for 28 ac (“rings NICE doctors for a pill”) comprises a total of 24 letters, and since one of these is an “extra” letter not involved in grid entry amendments, there are therefore only a total of 23 letters involved in the amendments. The fact that there are only a total of 23 letters involved in amendments for the 46 entries forces the sets of “extracted” and “introduced” letters to be the same. The fact that these sets are the same, together with the information derived from the down clues that “all entries are real terms”, also “forces” the letters to be inserted at the “ambiguous” positions such as 20 dn and 32 dn.
As I say, that is what I intended should happen. I wrote the preamble in a somewhat “minimalistic” manner, believing that the additional information required to solve the grid, particularly regarding the equivalence of the sets of extracted and inserted letters, could be deduced as described above. In hindsight, however, I can see that the whole process is rather confusing (and particularly so if the two definitions at 28 ac., “Mental disorder” and “Hebrew woman” are not identified correctly), and that it might have been better either to make more information explicit in the preamble, or even to omit the “special clue” contrivance altogether. I also was not aware until very recently that the latest edition of Chambers does not include the list of “First Names”, which would have made identification of “Maria” more problematic for solvers using that edition. (It’s a shocking admission, but I’m afraid I don’t own a latest edition of Chambers: I’m hoping to win one through winning the Listener Crossword competition one week, but until that happens I have to take trips to our local library when constructing new crosswords to check that obscure words or definitions are still included!)
So apologies to all those solvers whom I inadvertently confused with “Here and There”. I find that it is never very easy to judge the level of difficulty of a particular crossword, nor how much additional information to provide either in the preamble or “hidden” in the clues. The Listener Crossword editors of course provide assistance with this aspect (as always, they were extremely helpful and thorough with getting “Here and There” ready for publication), but even then the result may not please everyone. But many thanks to all those solvers who have taken the trouble to get back to me with their thoughts and comments – it’s much appreciated.