by Will Eisenberg
Will is a French horn player, music teacher, and crossword constructor located in the Twin Cities. He is the current 3rd horn of Orchestra Iowa and a founding member of the Minnesota Horn Quartet. Will is a member of the leadership team for Lil AVCX, the mentorship arm of AVCX. More of his puzzles can be found on the blog Half-Baked Puzzles.
What is a crossword puzzle? It’s easy to point at the grid on the page, the captivating black-and-white visuals, the empty space progressing steadily towards a solution and say: there. That’s the puzzle. The grid is tangible, visual, enticing. But without the clues, there is no puzzle, just an artful arrangement of squares (and letters, if you had any way to be told what they were). Yet all too often, for crossword constructors, clues are relegated to the status of a chore, a repetitious slog the constructor must wade through to get back to creating grids. Considering just how much of a solver’s experience comes down to the clues, I’m here to argue in their defense. Cluing puzzles is not just fun, it’s mission critical.
We live in the age of access. Blogs have discussed puzzles for a long time, but since the pandemic-era crossword boom, the discussion has turned to what lies behind the puzzles: the craft of crossword construction, and the constructors who make them. Many miles from the days when crosswords ran in the newspaper without bylines, the amount of information about how a puzzle gets built, and who built it, has never been greater for solvers. Constructor notes are commonplace—AVCX, Puzzmo, and even the New York Times run thoughts from the constructor. Often, these notes talk about the process of how a puzzle came to be. And in newsletters and interviews, editors explain their process, along with their likes and dislikes, and constructors do the same.
Of late, I’ve become an avid fan of the podcast “Crosstalk,” a fine example of this phenomenon. Host Daniel Grinberg thoughtfully interviews a range of guests comprising some of the smartest and most insightful minds in the crossworld. If you haven’t been listening, stop reading and go do that instead. Though the constructors Daniel has interviewed so far vary in many ways, most of his subjects agree on one thing: making grids is their favorite part of constructing, and the part to which they devote the most energy and care.
Through all the interviews, constructor notes, and forum discussions, it’s clear that gridding is having a moment. By gridding, I mean the process of creating a crossword grid: placing theme entries or seed entries, building a black square pattern around them, and finding words to fill the rest of the grid, until, like magic, a crossword puzzle has been constructed from the ether.
Magic might be the key word there. Gridding is the thing that feels most intriguing to someone entering the world of crossword solving (and then, perhaps, constructing). While it’s easy enough to come up with a better alternative to a clue that bugs you, it’s harder for most people to imagine how a crossword grid is created. I think the appeal of creating that somehow perfectly interlocked grid is what hooks the vast majority of constructors. Even if we have a killer clue in our pocket, we’re likely to find our hopes of using it dashed as we iterate through the grid’s options, losing the entry we wanted to clue along the way. Or we find, upon publication, that the editor has killed our darling clue anyway.
Clues are like a tour guide for the crossword puzzle. Having a tour guide doesn’t change the city you’re exploring, or the museum, but the guide can mediate the experience in a way that shifts things drastically.
Crafting a grid is intense, complex, endlessly beguiling, and intriguing. The process offers endless opportunities for improvement and refinement. I’m not here to disparage, or minimize, the art of grid construction. And clues can certainly feel like they occupy a lesser position, because they’re easier for editors to tweak later in the process. But to me, that doesn’t take away from how integral clues are to a solver’s experience. I’m here to defend and celebrate the joy of crafting crossword clues, and to emphasize how important cluing is to the constructor’s job.
I’m tempted to do this by simply listing great clues—doubly tempted, perhaps, to only list clues from my own puzzles. But if you’re reading this, you already solve puzzles: you’ve seen the clues. The great ones—the ones that make you sweat, groan, and begrudgingly crack a smile, are no doubt easy enough to call to mind. That’s kind of the point. When we remember a puzzle, we remember the experience of solving the clues. Rather than list memorable clues, I’d rather lay out my philosophy of cluing.
Clues are like a tour guide for the crossword puzzle. Having a tour guide doesn’t change the city you’re exploring, or the museum, but the guide can mediate the experience in a way that shifts things drastically. A bad tour guide—bored, low-energy, and seemingly unhappy to be there—can reduce even the most interesting subject to a forgettable slog. But a great tour guide—one who’s knowledgeable, funny, and infectiously happy—can turn even a mediocre exhibit into something special. Clues are like that. They usher the solver through the grid, showing off what’s interesting and providing context.
A tour guide can make their tour group laugh, make them think, show them nuance and history they might otherwise have missed. It’s that same wide variety of options that excite me when cluing a puzzle. Rather than worrying about each individual clue, I like to think about the puzzle as a whole. In total, what can I achieve with the clues? If I can make the solver laugh once, learn something new, look at something familiar in a new light, and maybe even groan a bit, I feel like I’ve done my job. I’m not hoping to write seventy brand-new clues, but I am hoping to write a good handful that have never been done and do my best to make sure the solver leaves with those in their head.
One common complaint from constructors is that cluing can get tedious, especially when they’re faced with writing a clue for a common bit of fill like OREO for the hundredth time. Some entries, like EKE or ALAI, leave the constructor without a wide range of clue options. But the tour guide need not wax rhapsodical about the water fountains or the bathroom. A simple set of directions (“The bathroom’s over that way”), delivered quickly, is exactly what’s needed. Clues are like this too, and it’s easy to spend far too long agonizing over every single clue in a puzzle. The solver’s never going to remember every one of the seventy-plus clues from your puzzle, so there’s no need to reinvent the wheel every single time. In attempting to avoid retreading well-worn territory, many a newer constructor has ended up with clues that are unclear, inaccurate, or just un-fun to solve.
But just as the goal may be to draw the solver’s eye quickly past some entries, every puzzle has opportunities for clues to shine. Wordplay clues, frequently denoted by a question mark, are often my favorite part of a puzzle. There’s no need to recap all the ways in which a constructor can write a clever question-mark clue—simply pick up a puzzle and you’ll find them. For outlets like USA Today that eschew question-mark clues almost entirely (though they’ve recently started allowing one or two in most puzzles!), there’s always the wordplay clue that is literally true and doesn’t require a question mark—for example, [It’s spotted on a safari] for CHEETAH. Question mark or not, my favorite puzzles are loaded up with wordplay. I’ll note that puzzles like USA Today tend to be much more approachable for newer solvers, which is fantastic. But newer solvers, ideally, turn into veteran solvers, who generally have a higher appetite for harder puzzles with fresh twists and wrinkles. A wordplay clue can highlight the longest entry in a grid, making it a marquee moment, but I’m also partial to shorter entries like ANTE and ELOPE that seem to have near-limitless possibilities when it comes to new wordplay clues. Far from feeling tapped out or stale, many of the typical words that fill a grid provide endless new avenues to be explored by constructors.
Clues are also one of the best avenues for bringing personality to a puzzle. A grid’s fill can certainly impart the feel of a certain constructor, like one who tends to pick higher Scrabble-valued entries, or eschews wordlists, and so on. But cluing is a spot where any constructor can put their distinct stamp on a puzzle. Even at highly edited venues, it’s almost impossible for a constructor’s unique voice to not have some sway over the puzzle via the clues. I notice sports angles popping up in my clues even when I’m not intending it—that’s just a part of who I am. Even relatively straightforward entries might provide an opportunity to highlight a musician, an actor, a TV show, or a book that’s been on your nightstand lately. But I also think that “voice” in puzzles can be more than just a collection of pop culture the constructor is into. Personally, I’d prefer to be remembered as someone who writes clever, concise, and fair clues over just “the guy who likes disc golf and the Timberwolves.” While I’ve learned a lot about the things my favorite constructors love, what ultimately made them my favorite is their ability to look at almost anything in a fresh and interesting way.
Does downplaying the skill of clue writing, and shifting it to the hands of the editor, lead to better puzzles overall? Or does it simply concentrate the potential range of clues and voices that wind up in published puzzles into a smaller pool?
Most of this essay has been aimed at constructors, but I’ll add a few thoughts here directed towards editors. It’s not uncommon for venues to evaluate themeless puzzles simply based on the grid alone. I’ve always found this an odd choice. A grid’s constructor and someone evaluating that grid entirely without clues might well be speaking different languages. Take the potentially disappointing fill ORI: an editor’s [“Do it, ___ will!”] might be a constructor’s [“___ and the Will of the Wisps” (2017 platformer)]. Moreover, the editor learns nothing at all about a constructor’s skill in cluing when evaluating an unclued grid. Given how deeply important quality clues are to the solving experience, this has always felt like a missed opportunity to me. Sure, editors who use a heavier hand might feel confident that they can wind up with a good set of clues no matter what the constructor hands them. But does downplaying the skill of clue writing, and shifting it to the hands of the editor, lead to better puzzles overall? Or does it simply concentrate the potential range of clues and voices that wind up in published puzzles into a smaller pool?
I don’t think I’ll necessarily flip your viewpoint overnight if you’re a constructor who views cluing as a slog. Likewise, I don’t expect to change the minds of any editors about their standards and specs, which I’m sure were developed with a lot of thought and care. But in a perfect world, what might change? For one, I wish venues that typically only evaluate unclued grids would consider having constructors clue 5–10+ of their favorite entries when submitting a themeless grid. I think this would be a nice middle ground between requiring a complete set of clues for a puzzle written on spec, which might end up needing to be reclued for a different venue, and simply evaluating a completely unclued grid. On the constructing side, I wish more constructors would set aside the anxiety of repeating prior clues, approach clues as an integral part of the puzzle rather than an afterthought, and embrace the myriad possibilities the cluing process has to offer every puzzle. We are puzzlemakers—and writing clues means making the puzzle.
Special thanks to Matthew Gritzmacher and Jess Shulman for their invaluable help improving this essay.