Listen With Others

Some time ago I filed away the term “Ship of Theseus” under “Crossword ideas” considering it had the makings of an interesting theme. My initial thoughts were that the eventual puzzle might be a kind of SofT in its own right. Perhaps the grid, clues and entries could change somehow, as part of the solving, to transform the crossword into something that barely represented the original. (In the event, it didn’t take me long to realise that I would have to settle for just the entries changing; this in itself would be a tall order.) Thinking of the paradox thrown up by the concept, I wondered whether it might be possible to get the solver to question whether the title should still stand after all the changes had been made.

The first file I create when starting on a new puzzle is an assemblage of notes and references, well before I consider the grid shape. So here, looking back, I find various descriptions of the philosophical problem, the problem of identity. (As it transpired, I was unable to deal satisfactorily with this in the construction of the puzzle.) An important discovery I made was that an alternative (modern) scenario that illustrates the paradox is that of Trigger’s Broom. And the bonus was that this and SofT contain exactly the same number of letters.

I gave some thought as to how the design of the grid could incorporate this fact. Could the letters of the first phrase be systematically replaced so as to eventually reveal the second phrase? I had various attempts to achieve this but I realised it wouldn’t work. I would have to leave TB till later, and try to weave it into the solution process somehow. I had noted that SHIP OF THESEUS could be easily altered to SHIP OF THE SEAS. So to the grid construction with this last feature in mind. Let’s just see what happens:

Then:

But where is all this leading? It becomes clear that the loop of alterations has to finish somewhere within SHIPOFTHESE(AU)S, so I try SHIPO(FN)THESE(AU)S and SH(IO)POFTHESE(AU)S. But trial and error with these fails to get a complete set of 26 changes. Should I abandon Ship? Thankfully, the breakthrough came when I came up with S(HK)IPOFTHESE(AU)S. And I was able to associate SHIP with SKIP, both being “vessels”, albeit quite different ones. Also I needed a plan as to where the replacements would occur throughout the grid. This should provide a route which covers all entries:

Still a long way to go – entry in row 5 is missing an index. Trying to be more organised I got:

Eventually, after more trial and error, I arrive at:

… which is the final grid (Eureka).

I had the feeling that this could well be a unique solution because of the number of constraints I had been working with. But could I be absolutely sure that solvers would not find other solutions? So I’m prompted to introduce a further constraint that guarantees there are no more. And at the same time, I can introduce Trigger’s Broom. It was time to think about the clues and hiding the letters T,R,I,G,… therein. I was hoping for a method that would not provide solvers with a way of determining the “replacements” directly and so came up with the triples idea. I felt this would help solvers to check their working but contribute only a little to their determining it. Even with this extra factor built into the down clues, I have to say that clueing was the easy part of the whole process.

I must say that the exercise I’d set for myself was a real challenge. It turned out to be a lot like attempting a difficult Listener. And I do enjoy a good puzzle to solve.

Pointer.