Songspiel – Blog

Listening to Nashville Skyline (the later, mellower Dylan) I cast my mind back to some of his earlier music, recalled The Times, They are A-changin’, and saw potential for a puzzle. I assumed there had been other puzzles on the theme of time but hoped this particular song hadn’t featured.

My initial idea was to include a number of anagrams of ‘time’ words. I drew up a list of the latter, ranging alphabetically from ENEMY/YEMEN to YEARS/EYRAS (a surprisingly short list). I wanted at least one that was figurative or idiomatic and was particularly taken by the transformation of PORRIDGE to PERIGORD. I then set about constructing the first grid. Early on, I decided to extend the theme, using words that would convert to other words when T’s were altered. Initially I wasn’t sure whether to hide BOB DYLAN in the grid or include him as an unclued entry, but after a few trials I saw the potential for the name to emerge in the top row of the grid by altering T’s, so those letters  and some thematic anagrams determined the shape of the grid.

In the course of experimenting with various grids I had my first serendipitous moment, which suggested the next thematic layer – Sympathy put ANZIANI into the top row, using the AN of DYLAN. I did not know the word, but after checking the meaning I knew I had some more thematic material and decided to incorporate the other groups in the grid. It was simple to find synonyms for ‘mothers’, ‘fathers’, ‘writers’ and ‘critics’ but ‘congressmen’ was difficult. This is where I decided to make congressmen a special group for solvers to add below the grid. Constructing the grid now became trickier, not least because I wanted to avoid multiple occurrences of MAS and PAS (at that stage I was anticipating a highlighting requirement).  After many failures using Sympathy, with MAS and PAS occurring all over the place, I began constructing the grid manually, using Tea rather more effectively, and going back to Sympathy for some final filling. In the end I ditched the highlighting requirement. The other difficulty, of course, was constructing a grid that did not use that rather common letter, T. I tried entering ‘*T*’ in the ‘kill list’ in Sympathy, but that didn’t work, so I simply added words containing T to the ‘kill list’ as they appeared.

Once I had a selection of grids, I scanned them to see where T could be substituted in a checked cell to form a new word, which would be the one to be clued. SOLDERING/TOTTERING in 15 down was a late discovery, though I wondered whether I should use it as a thematic item. With two unchecked letters and three clashes, solvers had only four checked letters to go on, though the most likely ending would be -ing, reducing the clashes/unchecked letters to four. I also reasoned that once solvers had spotted the theme, and realised that all changes made real words, it would be easier; failure to solve the clue would not, in any case, prevent solvers from completing the grid. I was reassured when the test solver didn’t object to the number of changes. In the end 39 down appeared to cause more problems.

I chose to convey the song title through definition misprints as a self-imposed discipline, since I’d not used this device before. Once I started looking at the words to be clued I found that it was easier if I alternated between misprints and correct letters (even if this made things simpler for the solver), but even allowing myself this latitude, 2 and 30 (both needing a misprint of H) presented particular problems.

Finally, I wanted to present solvers with an additional puzzle with the ambiguities in the top row. I didn’t want to be explicit, but some hint was needed. ‘Nominally’ seemed a suitably cryptic hint to lead solvers to the final penny-drop, though for some the penny dropped rather earlier.

There is an eerie postscript to the puzzle, published three days after the election of Senator Barack Obama as the next US president. I’m indebted to Derek Arthur for pointing out that in addition to  congressmen and senators, OBAMA features at various points in the puzzle. There is the coincidence of his initials, BO, in the NW corner; he can be traced via knight’s moves starting with the O of SOLDERING and ending with the A in cell 25; there is also a jumble of OBAMA involving the A of MASE, the B of EMBRUE and the AMO of AMOEBA. The Delphic Oracle works in mysterious ways.

The times, they are a’ changin’, in America at least.

Andy Stewart (Dysart)

This time I can be perfectly clear where the idea came from – my desk calendar, which recorded the 69^th anniversary of the broadcast last year.  ‘69,’ I thought, ‘how ungainly’ – followed by ‘But that means a more suitable anniversary next year’.  (If there’s anyone sniggering about 69, by the way, they can stop right now.)

The next thing that came was the Welles/Wells link – I couldn’t think of that being exploited anywhere else before and in the context of this radio show it seemed unmissable.  (It also seemed the sort of gimmick that might well fall out of the gridding process without too much work – choose that unchecked letter, and that light becomes a word you can either add or subtract an E to or from.)

What else?  Well, ‘broadcast’ was an obvious anagram indicator, so anagramming the title of the book/show was indicated, and with it the idea of ‘…the programme broadcast’ in the preamble.  The final component took a little longer to mature, but felt particularly pleasant when it came – the types of programme that might be interrupted, with the interruptions forming the same message picked up by listeners to the Welles broadcast.  (And it has only just this minute – here, in this blog – occurred to me that here I am blogging a Listener puzzle about the radio.  So now you know what didn’t go into my thinking.)

Once the interruption idea had been hit upon, it only remained to choose the types of programme – actually quite a limited list if you wanted generic one-word descriptions.  I was pleased to note, however, that the descriptions I chose all had ‘other’ definitions.  I fiddled a little with the mix of programme and interrupting letters till I’d got a selection that looked reasonably friendly for DLM purposes (they weren’t, of course).  And then it occurred to me that it might be fun to make each of the DLM clues refer to a different type of interrupted programme – which proved to be quite a tricky thing to do.  I kept struggling with a DLM and allowing myself a few ordinary cryptic clues for relaxation, and in the end I still had two DLMs as the last two to write.

Perhaps this is why the puzzle’s progress felt odd – a lot of the way through I felt rather gloomy about the puzzle, but as I finished typing it up for submission, I looked through it, and suddenly thought it held together pretty well.  Off it went, and the first editor sailed through it, spotting the theme perhaps a little too early, and asking for one clue to be modified.  The second editor emailed back to say it had been accepted.  When I replied saying I thought having to alter only the one clue was my smoothest acceptance yet, he said that I’d see that he’d had a much different solving experience from editor No. 1, principally through not latching on to the theme quickly (and having a less clear recollection of it when he did).  But as he’d filled the grid correctly before resorting to Google, it had got through. 

So that rather whetted my appetite for seeing what the broad mass of solvers would make of it…

Listener No 4004  –  Signal Boxes  by  Ploy

I’d been aware for some time that Chambers 2003 contained entries for five logic circuit elements or “gates” (AND, NAND, NOR, NOT, and OR), with clear explanations of their functions.  Having spent my working life in electronics research, this appealed to me as a promising theme for a Listener crossword.  If it were not for simple electronic items such as these, many of the everyday devices we now take for granted would not exist.

My idea was that the grid would contain a logic circuit built up from a number of gates, with the solver having to select a self-consistent set of binary signal values (0 or 1) to complete the grid.  I also wanted to include references to George Boole (logic) and Claude Shannon (circuits), both key figures in this subject.  I decided to use one of each of the five gates in Chambers, and exploit some of the O’s and I’s in Boole, logic, Shannon, and circuit as signal values zero and one.  Surprisingly quickly I arrived at a suitable basic layout containing the key components, both electronic and lexical.  This had ten O/I clashes, five of them resolved by thematic names/words, the remaining five being uniquely determined by the nature of the gates.  Any prospect of having a symmetrical bar pattern was ruled out by the amount of thematic material I was trying to include.  The grid fill was then completed using Ross Beresford’s excellent Tea & Sympathy software, and took eight attempts.

The choice of title and preamble wording was intended to (misleadingly!) suggest a railway connection, my previous Listener puzzle, “Travelling Light”, having had such a theme.  In the event, however, a fair amount of time passed before “Signal Boxes” was published, though a number of solvers did comment on an initial, apparent connection.

Clue writing was straightforward, there being no gimmicks to take care of, and I was able to introduce a scattering of railway terminology in them, just for fun!

While I was fairly certain that most solvers would make progress by initially noticing “logic” or “Boole” in the grid, I appreciated that there would be much less familiarity with the name “Shannon”.  I was relieved to learn later that there is an entry for Claude Shannon in Chambers Biographical Dictionary, so any complaints about obscurity would not be on very strong grounds.

I was delighted to find that overall the puzzle received a very good reception, and that nobody seemed to mind that a crossword required the solver to perform some mathematical processing.

The most common error was, perhaps predictably, incorrect assignment of one or more of the five “free” signal values.  Opinion was divided as to whether the final stage was straightforward or not, as is often the case with puzzles where there is some “business” to attend to once the grid has been filled.  Comments on the clues showed that they were regarded as being on the easy side by Listener standards.

Digital circuits exploit other logic gates apart from the five I used, but in fairness I stuck to those with entries in Chambers 2003.  A solver’s remark on the absence of XOR in the puzzle prompted me to check the new edition of Chambers (2008), when I found that the two “missing” gates, XOR and XNOR, have now been added.  The puzzle might have looked different had I been setting it in 2009!

Phil Lloyd  (Ploy)

10^th November 2008