I have worn spectacles for more than 40 years*, so I have seen plenty of eye-test charts (to the annoyance of opticians I am very good at memorising the things). Anything structured with letters in it should be meat and drink to a crossword setter, but nothing came to my mind over the decades. And when it finally did I couldn’t recall anyone else ever using it. It did seem a missed opportunity.
I sort of know why now. The actual motivation for the puzzle was solving The World’s Most Ornery Crossword in GAMES magazine. One answer was SNELLEN CHART. Is that really what it’s called? thought I. I checked in the dictionaries. Chambers didn’t have it, and ODE offered Snellen Scale (which is what determines the relative heights of each row). Google was much better informed, and it does seem odd that so commonplace and familiar an item isn’t mentioned plainly in dictionaries.
Finally knowing what it was called (and, lookee, it’s 12 letters long – almost impossible not to think in 12×12 terms these days) seemed to be a spur to grid creation, and off I set. It proved to be an enormously intricate journey.
I had in mind a Listener puzzle many years ago (by Jago, I think it was) in which the only way to fit in the answers was to assume that some letters occupied not a single cell but a 2×2 block. I set out with the idea of a 4×4 block for the top letter, 3×3 for the pair below, 2×2 for row 3, and single cells at the bottom. And that – allowing for SNELLEN CHART to read as a single across entry – was the dimensions of the grid sorted.
Here’s my first attempt, after hunting round for an ophthalmological 10-letter word:
The original intention was for the squares to read something in order, and PRESBYOPIA was clearly not that something. You can also see I was trying and rejecting the usual rotational symmetry.
Turn the page and you see the final grid.
My memory tells me this sudden leap is wrong, and that I scribbled a lot of experiments on sheets of paper that haven’t survived. Having dropped PRESBYOPIA, I fetched up on ASTIGMATIC as having a much friendlier set of letters, and I’m pretty certain there were a few goes at trying for a version which spelled out ASTIGMATIC in the same way as the abortive first attempt spelled out PRESBYOPIA. But the final version saw all the consonants sink to the bottom where they mostly didn’t have to do anything very special in the way of intersecting with other answers!
What else? Well, I pulled the oddest words out top right for future reference – proper nouns, abbreviations and KOPF. And on the opposite page the title has emerged – can’t recall what ‘Word’ is doing underneath it.
This obviously wasn’t going to be a grid done in your average computer software package, and if solvers felt some perplexity in disentangling the tracks words took through the grid – well, so did I! But it did feel rather pleasing to get to the end of it.
Clueing was the normal agony. My original intent was to omit word lengths, leaving solvers to puzzle out even more how things fitted together, so my original submission was missing those. But the EV editor solved it in that form and eventually decided to reinstate them…and then moved on to the problems of presentation.
If it was hard to have done in (say) Sympathy, then the sort of typesetting needed to present the grid in a way that at least hinted at the potential oddities ahead was also hard. For all that crosswords may be ‘clickbait’, newspapers are still struggling with ways of presenting them for printing off (let alone interactive solving). We considered presenting the grid purely with tickmarks along the borders, and other ways to highlight the unusual nature of entry (without actually explaining it). But in the end we ended up back on my original submission. I’d toyed with barring the four letters off to hint at the squares motif – but the letters involved were parts of words, and it would have taken a couple of extra sentences in the preamble to spell that out, so I was content with shading.
This process may be a strong hint as to why a puzzle on this theme hasn’t been seen before – there are many stages on the route to completion where one can get beached. But I hope solvers weren’t.
Before I finish, let’s turn two more pages in my workbook.
If you solved my IQ puzzle Illusion (the blog is here) a while ago you should recognise this grid. It is very close to the final concept, even though it’s crossed out. The final version did indeed adopt the 11×13 shape noted in my scrawl. As you can see, once the idea of one letter in multiple squares took hold, it settled in another puzzle which also required different scales in the grid. My expectation was that Illusion would be the second to appear, and thus I could pontificate eloquently on the influence of Here’s Looking… on its lunar successor. But no such luck, as editorial intervention reordered the EV puzzles and hence the linked pair appeared in the opposite order to that expected. And hence I had to be a little circumspect when writing the blog for Illusion the other week. But all can now be revealed.
*Did I really just write that? 40+ Crivvens! No wonder my nose is that shape. I did try contact lenses once, but it resulted in two hours of continual tear production, and a reluctance ever to get close to strawberry blancmange.