For what it’s worth, I know New York Times crossword editor Will Shortz slightly, and he says that modern crossword puzzles are symmetrical almost exclusively because of one person: Margaret Farrar (nee Margaret Petherbridge), who was the original crossword editor of the old New York World. She established almost all the basic conventions of modern American crosswords in the early 1920’s.
And why did she insist on symmetrical puzzles? According to Shortz, simply because she thought puzzles “looked nicer” that way!
Well, Will is an expert, but the fact remains that the first crossword puzzle in 1913 was symmetrical and as far as I know the ones that followed were similar.
Farrar didn’t join the New York World until 1920 (or possibly 1921). It’s possible that they were doing nonsymmetric puzzles until then, or Shortz might just have meant that she was responsible for the modern conventions of the puzzle, as state on this Crossword Puzzle history page:
And the writer, Merl Reagle, agrees that Farrar had her own unique sense of aesthetics for puzzles:
Interesting column. On a tangential note, I am curious if this puzzle is unique, as far as anyone knows. It’s certainly worth doing, for NYT crossword aficionados who might have missed it.
There are two possible solutions, depending on which candidate won the 1996 Presidential election. Both work for the same grid and set of clues.
Quint, the link you gave seemed to point to an explanation of the puzzle, but not to the puzzle itself. At least, the text said to stop reading if you want to solve the puzzle yourself, so I did. I haven’t read your spoiler either. Is there a link directly to the puzzle, without any spoilers?
If you follow the link you can download the puzzle, ‘tho you may need to install another program provided in the link. If you don’t scroll down the page the puzzle will not be entirely spoiled, tho’ the explanation on the top of the page may provide enough information for you to decipher part of the unique nature of the puzzle. If you want to solve it without any clue as to it’s nature, you can try to find a link to the NYT crossword of Nov.5 1996. The only link I found for this date required signing up, so that is another option.
It’s unusual, to be sure, but I doubt that it’s unique. Cruciverbalists love doing these challenges that undoubtedly have Margaret Farrar spinning in her grave. The circumstances of the puzzle may never be duplicated, however.
I wouldn’t be so sure of that. According to the linked article, another puzzle with the same gimmick was made 16 years earlier, but for some odd reason wasn’t published (maybe the editors didn’t figure out the gimmick, and thought that it was presumptuous?). If it’s been done twice, it could be done three times.
Another one I saw a few weeks ago (I think it was a Shortz) followed all of the rules, except that one of the answers extended a single space outside of the grid (complete with extra little square drawn outside the boundary to contain it). So it wasn’t symmetric, had an unchecked square, and wasn’t even a square. Given what the answer was, though, I think it justified breaking the rules.
In the late seventies/early eighties, back when i did crossword puzzles on my long bus ride home(Hour+), i once came across a magazine that had all it’s puzzles in the shape of cuddly animals. These puzzles weren’t particularly hard.
As a puzzler myself, I can comment on Cecil’s not figuring out why the sides of crosswords have odd numbers of squares, as a rule.
Most American-style crosswords these days are themed. That is, the longest entries in the crosswords have something in common, whether it’s that they’re all parts of a quote, or they’re related puns or references. In addition, since the crossword is 180-degree symmetrical, the theme entries pretty much have to be, too. This means it’s really handy to have a centre row for an odd number of theme entries.
Given all this, the standard for the industry is a crossword that has a side with an odd number of squares - specifically, 15x15, 19x19 or 23x23. It just makes life easier on an editor, since he’s going to get puzzles of this size anyway, and he only needs to set three prices based on size.
I make crossword puzzles myself and solve the NYT every day. I imagine most of these conventions would have come about anyway – certainly the ones about 2-letter words and having parts of the grid broken apart by black space (in effect creating more than one small puzzles).
You can only get the Election Day puzzle online with a paid subscription to the NYT puzzles.
I have a friend who claims his brother or brother–in-law created a puzzle where the entire grid had two different fills for one set of clues – I suspect he’s got it backwards, that there are two sets of clues for one grid. The other way around is evident to anyone who’s constructed a puzzle as very close to impossible. The election day puzzle only has 8 variable entires with one set of clues, and is remarkable for doing that.
I find it hard to believe a veteran solver would have “never noticed” that the grids were always symmetrical. If she saw one that wasn’t, she’d probably notice that something was weird about it.
That one will be back in season a week from today.
One April 1, Shortz ran a crossword that broke all the rules. There were 2 letter words, a little asymmetry, the same word used twice, and several errors like twice-clued entries and missing clues. Even his name was misspelled at the top! That is one of my personal favorites.
The ban on two-letter words in “major league” puzzles is often ignored. *Ai, em, en, pi * and possibly some others that I can’t think of off-hand have appeared in my daily newspaper’s puzzle, which is syndicated from the NY Times, I suspect.
Crosswords do have a lot of plurals of two-letter words, so you may have seen ems, ens, pis, and etc., but if you see two-letter words that are not part of a theme (and then exceedingly rare), your newspaper is not carrying the NYT crossword, or any of the other well-known syndicated puzzles (TMS, Universal, etc.). What newspaper is it?
It probably was the plural. I don’t recall exactly. The paper is the Charleston Post and Courier, and I’m sure it’s a syndicated NY Times puzzle. But plural of two-letter words shouldn’t count, either.
What is of importance is that the answer be of three characters, not the character of those characters, so to speak. Length is everything. Two character answers lead to weird placements and intersectors and you wind up with a lot of dumb clues. Try doing a puzzle with real two-character answers to see the difference.
