Listen With Others

I was fortunate enough to be the second ever winner of the Radix Auditorium Jug, awarded to the newcomer with the most correct Listener submissions, so I’ve had the pleasure of attending the last couple of Listener dinners as a humble solver. It was absolutely fascinating to talk to setters about their setting process and to learn more about the inner workings of my favourite crossword, and several of them encouraged me to try my hand at clue-writing. They’re all a lovely bunch (well, apart from one chap who cheated on the quiz, but I won’t name names – mostly because I can’t remember what his was). The following year, having very much enjoyed entering clue-writing competitions in the interim, I was further encouraged to try my hand at creating a whole puzzle. So I thought I’d give it a whirl.

The theme of Murder on the Orient Express came to mind fairly quickly, perhaps because I was a devotee of Agatha Christie in my youth, and this was the first of her books that I read – appropriately enough, it was given to me by my late grandfather, who also got me interested in crossword-solving. My first idea was to get solvers to change ‘RATCHETT’ to ‘CASSETTI’ in the grid (minor spoilers for the book, although it is fairly early on that it is revealed they are the same person), as I liked the notion of one hiding as the other. However, that felt like a fairly weak theme that needed expanding, so I also wanted to put ‘TWELVE STABS’ (further minor spoilers) in the grid, as well as HERCULE POIROT / ORIENT EXPRESS on the diagonal. On this one I got lucky that the two phrases were the same length. As you’ll know, all but the last of these three ideas was jettisoned along the way, mainly because I found it impossible to create a grid that worked with these restrictions – and also because I realised I was over-complicating the puzzle for little benefit.

Creating a grid where HERCULE POIROT and ORIENT EXPRESS on the diagonal would both give legitimate entries throughout was not easy – it was only when I was partway through trying to do it that I realised that it’s actually surprisingly rare to see that in the endgame – but it always appeals to me, so I persevered. One of the problems was the 180 degree symmetry, which meant that a neatly constructed top-left would then prove impossible to replicate in the bottom-right (at one point I had the words in a chevron, from top-left to centre to bottom-left, which was easier but much less elegant). Another problem was that the two phrases only had one letter in common, but by happy chance it was the middle E, and it struck me – eventually – that I could isolate that cell. In my draft of the solution notes I even made the bold claim that the E was in fact a way to ‘express Orient’.

For the clues, I tried to think of a way of indicating letters that I hadn’t seen in the Listener before. I won’t reveal the first idea I tried, because I’ve repurposed it for my next Listener submission (I’m hooked now!); the second idea – which I ended up using – was to have jumbles of suspects’ names plus an additional letter. I was then a bit annoyed to see this method come up in a puzzle earlier this year, but never mind, I’m sure it’s been done before anyway.

The next thing to do was make a list of all the possible letters that could be added to a jumble of each of the suspects’ names to give a real word, and then work out what useful phrase I could construct from them – I soon discounted surnames as being impossible to work with, and unnecessarily difficult for the solver, but forenames looked much more hopeful. I was limited to S (Natalia / Alsatian) and E (Rudolph / upholder) for two of them, several others gave lots of options, but Hildegarde unsurprisingly gave me nothing. Working manually I came up with a few ideas of words that could be feasibly hyphenated without sounding too many alarms: herd-leading, large-dished and dear-delight amongst them, but griddle-heat was the winner for me – and ended up working nicely in the oven-based clue for LEAR.

My initial idea was to give clues to both what needed to be removed and what needed to be added to the diagonal – BELGIAN TRAIN was on the cards until I realised I needed an S in there somewhere, based on that Natalia / Alsatian pair-up – but eventually I enjoyed ‘Eastern state’ too much as a misleading clue to ‘Orient express’, especially if I referred to the latter as a location. A slightly woolly claim, perhaps, but since the train is stationary (as it were) for most of the book, I thought it was legitimate enough.

For the finishing touches to the grid, the unclued entries of AGATHA CHRISTIE happened fairly organically when I realised that AGATHA fit one of the spaces I had left, and a bit of re-working would allow CHRISTIE too. Finally, it struck me as amusing to circle letters that I would describe as identifying “useful cells”, and then having that phrase spell out ‘little grey cells’. In the end that particular wording in the preamble was lost in editing and I forgot to ask why – presumably because the word ‘cells’ in that context was either inappropriate terminology, or gave away too much. Still, the bit about proving useful to the solver remained, and one of the circled letters helped (I think) with the fact that 1 down was over-unched.

Speaking of which, the first comment I got from Shane was to point out (very kindly!) that my average grid length was too low and two of my entries were over-unched, but that he’d reserve further comment on this until he’d test-solved the grid. This showed two elements of naivety on my part – firstly that I hadn’t appreciated that these were a little bit more than merely recommendations, and secondly that I had thought it was a strength rather than a weakness to have entries with no unches (and this was bringing down my average word length). Keen-eyed solvers might have noted that the words TIAN, HALT, ESSEN & SASSE were all in the grid, but got barred off – and rightly so – during the editing process. The majority of the other edits were for things like Scottish words that hadn’t been identified as such, or the fact that I had used ‘about’ in the same way in multiple clues.

When writing clues my first priority tends to be a good surface, and while not all the clues were perfect in that respect – ‘ergate’ is difficult to conceal – I was happy with how most of them turned out, and particularly glad that the cheeky ‘tired riders’ in 9 down seemed to go down well. Speaking of going down well, it did strike me that Shirley would be displeased by the relative paucity of alcoholic allusions in the clues (what can I say, I’m teetotal myself), but I honestly didn’t think about that poor hare until his fate was brought to my attention following publication.

I’m absolutely delighted to have joined the ranks of the Listener setters (only my second ever published crossword, the first coming when I deputised for a friend in my university’s newspaper over a decade ago), and I have to thank Shane & Roger for their assistance in making the puzzle publishable, as well as all the setters who encouraged me to have a stab at submitting something. Let me pass on the favour by encouraging any solvers reading this who are thinking about trying their hand at setting – go for it!