Many crossword themes imply the smallest change alters a word’s whole definition. MOUSETRAPS are real, but MOOSE TRAPS are not. A Fairbanks, Alaska couple did build a giant “moose trap,” but only as a joke. It’d still be silly and unfair to clue M-O-_-S-E with “___ trap” and expect a solver to fill in “MOOSE.”
Cluing a Schrödinger puzzle works from the opposite premise. The goal is to come up with a clue that can apply to both possible answers. Or more—I have seen a few puzzles in which a Schrödinger square has more than two possibilities. For simplicity, though, we’ll assume two here.
You could clue MOOSE/MOUSE as “Animal” or “Mammal,” but that’s dull, vague, and old-fashioned. A better clue would be “Mammal with a comical plural.” The correct plurals of “mouse” and “moose” are “mice” and “moose,” but wags have been calling them “meeces” and “meese” for ages.
Some change-a-letter pairs cannot be assigned one clue that covers both, at least not in a satisfying way. Take QUINN and QUINT. There are famous and semi-famous people named Quinn, but I don’t think any of them is a quintuplet or related to any other “quint” definition.
But it can take a lot of thinking to separate the impossible from the challenging. When I wrote Schrödinger clues, I could stare at a single answer-pair for close to an hour before the best way to indicate both possibilities fell into place.
On the other hand, some pairs are a little too easy. UHF and VHF could both be clued as “Band on an old TV dial.” Likewise, GRAY and GREY are just variant spellings, with one feeling a bit more “British.” Eh. These pairs will do in a pinch, but they’re not too exciting.
What impresses me about Schrödinger puzzles are longer entries and more elegant meaning jumps. For instance, “Superpower like the House Corrino in Dune.” You could answer that with SPACE EMPIRE or SPICE EMPIRE. “Challenge to handle silently for history class” could be WRITTEN TEXT or WRITTEN TEST. “Likelihood of doing something sudden and a bit annoying” could be JERKINESS or PERKINESS.
For the Ubercross V, I knew I could fit in a lot of long variable answers. At first I thought, “What if I could do a puzzle containing one of each possible Schrödinger square—(A/B), (A/C), (A/D), and all the way down to (X/Y), (X/Z), and (Y/Z)?” But that would have added up to 325 squares…a little much even for me.
Partial spoiler paragraph: I settled on 52 squares, two for each letter. The grid’s longest answer said, LOOK FOR SOME CHOICES THAT AVOID THE OLD PATTERN YOU KNOW BACKWARD AND FORWARD TO LEARN WHY FLIPPING A COIN IS A GOOD DECISION-MAKING TRICK. The answer to that poser was found in the Schrödinger square options that did not go A, B, C, and so on, then back to C, B, A.
Schrödinger squares vary in difficulty. Changing a vowel to a consonant is harder than vowel-to-vowel or consonant-to-consonant. Q is the hardest to work with, but J and Z are also tough nuts. By randomizing the 52 squares but ensuring every letter would be represented, I could incorporate the full range of difficulty—for the times I wanted to climb a mountain and the times I wanted to take it easy.
Next weekend, I’ll do a quick cross-study of other puzzles’ Schrödinger clues and discuss a special subtype of Schrödinger puzzle (with a nod to Darryl Francis, who reminded me of it!). Tomorrow, I’ve got the first of a few previews from the next Journal issue. Till then!