Rex Parker Does the NYT Crossword Puzzle: Grid great Greasy

Constructor: Martin Ashwood-Smith

Relative difficulty: Medium-Challenging

THEME: none

Word of the Day: Larry KERT (47A: Larry of the original "West Side Story") —

Larry Kert (December 5, 1930 - June 5, 1991) was an American actor, singer, and dancer. He is best known for creating the role of Tony in the original Broadway version of West Side Story. (wikipedia)

• • •

THE SATURDAY PUZZLE: It's a quad stack. Two, actually. Two quad stacks. I knew that's what I'd be dealing with before I even opened the puzzle. OK, fine. And it's much like all the others I've encountered. Looks the same. Has the same combination of dull / forced long stuff and awkward short stuff—although today I will say the short stuff is better than usual, possibly because the long stuff is so full of RSTLNE. Phrases with super-common letters abound: DENTALASSISTANT, STEMLESSGLASSES. Lots and lots of easy-to-work-with letters.There are a few interesting answers—most notably OBSCENE GESTURES (great) (51A: They're usually pixelated on TV) and PALESTRINA (27D: Renaissance composer of "Missa Papae Marcelli"), which only dimly rings a bell, but is lovely, and was pretty fairly crossed (unless NEALE got you, I guess (34A: Grid great Greasy)). But this issue of fair crossings brings me to the puzzle's one truly terrible element—the WAUKESHA / KERT crossing. Here is the one thing in its favor: I guessed correctly, so that "K" must have some tenuous claim to inferability. I have no idea what/where WAUKESHA is, but that missing letter *felt* like it had to be a "K" (my friend Amy says this was probably the influence of "MilWAUKee"…). But I had zero confidence in the "K." Was considering "B." Never heard of KERT, which is one of the dumbest-looking names I've ever seen, no offense (47A: Larry of the original "West Side Story"). He's been in the puzzle (per the cruciverb database) … twice. WAUKESHA, once. All three instance of these words had very fair crosses. You just can't cross one oddly spelled and not fantastically familiar proper noun with another at a letter than can't readily be inferred. We have a name for that: NATICK. (For a definition of "the Natick Principle," go here and scroll down)

Now, the last thing I care about, the Last thing, is whether any individual knew one or both of these. That is not the issue. Of course there are people in the world who know both things. Saying you knew WAUKE$HA so the crossing must be fair is like saying your aunt smoked and lived to 93 so smoking must not be harmful to your health. From a construction standpoint, you never, ever, ever want to have a solver end with one square where it's a know-it-or-you-don't situation (which is almost always proper noun x/w proper noun). It's god-awful form. If I knew neither answer and had to guess, it's a good bet others will be in same boat. And taking a random stab is not "solving." It's not fun. It's a drag. And it's especially Not the way you want solvers concluding their experience with your puzzle.

And that's the thing—these are answers generated by a database. They are deemed OK because someone already deemed them OK for an earlier puzzle so they must be OK. Only a computer thinks this way. A human being has to be able to say "yes, they were in other puzzles, but they're kind of obscure and, look—when the *other* puzzles used them, the crosses were ordinary words. So maybe these two shouldn't cross." No, Computer Say, 'Work Fine Boss.' There is a difference between constructing and database management. Subtle, sometimes, but very real.

Signed, Rex Parker, King of CrossWorld