The Stack

While we’re on the subject of puzzle books, there’s one other I’d like to describe. Tachyon, by Stephen Boughton and Daniel Sage, is very much in the vein of Journal 29: each two-page spread presents a single puzzle (which, in most cases, doesn’t really need that much space), and solving it gives you a word that feeds into a later puzzle. It’s essentially the same format as Codex Enigmatum and its sequel, Codex Mysterium, except that in those, solving a puzzle gives you the key directly, and in Tachyon, as in Journal 29, it’s indirect: solving a puzzle gives you a password to enter into a web site to get the actual key. Still, all four books are similar enough in format to clearly belong to a distinct mini-genre, and I find myself wondering about its origins. Journal 29 is the oldest of these books that Amazon recommends when I look at any of them — is it in fact the thing that the others are all imitating, or does it go back further in time?

Of all these books, Tachyon is the one with the closest thing to an actual story. It’s still mostly suggestive theming rather than narration, but there is some narrative, at least, spread out over both the pages of the book and, occasionally, the web site that provides the keys. Basically, it’s about the narrator’s attempts to decipher the notes left by his missing father, a physicist. Find the codes to activate his machines and possibly you can turn back time and stop his disappearance. Not that the puzzles that occupy most of your attention are particularly related to this story.

The puzzles, as in all these books, are short, and consist mostly pictures and diagrams presented without explanation, or with instructions consisting of a couple of key words from previous puzzles. The whole challenge in each is figuring out what the author wants from you. In that respect, it more resembles an adventure game than a puzzle hunt. There’s some reliance on technological knowledge: one page expects you to know how to send text messages with a numeric keypad, which was a stage of phone technology that I personally skipped over. And honestly, I found some of the puzzles unsatisfying, either because they were too facile or made what I felt to be unreasonable assumptions, causing me to seek hints even though I had got the central insight right.

But I have to give it kudos for its ending. This book plays with form in a way that none of the other books I’ve mentioned have attempted. Spoilers ahoy.

The web site that provides the keys prompts you for a solution to the final puzzle just like all the others, even though there’s no possible use for a key from there. When you do, it gives you the story’s climax: the narrator’s discovery that he made a mistake in interpreting the key word from the very first puzzle. This produces an entire chain of revisions: change a key word, and you have to redo the puzzle that uses it, yielding another altered key word, until you reach the final puzzle. As in the story’s premise about going back in time, you get a chance to do things over and make them right.

Now, this means less than it probably sounds like. The nature of these puzzles is that there isn’t much to solving them a second time: you’ve already had the crucial insight into interpreting them. Also, the indirectness of the keys means that the puzzles aren’t really related to each other in a way that would be complicated by the altered chain. Still, it’s kind of impressive to realize that each puzzle involved had to be designed around two different solutions, with just a word making the difference. Moreover, I find it encouraging to see this kind of playfulness in what was otherwise seeming like a very rigid format. Maybe there’s some real potential in this mini-genre.

In my last post, I hadn’t yet solved the final riddle of Daedalian Depths. I think I have, now. I’ve definitely solved most of it. There’s an overall pattern to the shortest path through the maze, and there are enough hints about that pattern that once you know what you’re looking for, you can find enough minor details confirming the pattern for it to become a certainty. (I took perhaps longer than I should have to discover this pattern: there’s a pretty blatant hint that went over my head until after I figured it out by other means.)

But it really feels like there should be just one more step. The final riddle asks you to find a set of five legendary artifacts “hidden in plain sight” along the shortest path. The picture of the last room has five pillars set up to receive them. This really feels like it’s a setup for a final metapuzzle, building a five-letter word, or perhaps a five-word phrase. That’s how it would work in a puzzle hunt, where answering the last riddle in text is how you confirm that you’ve won. There’s even a really tempting way to extract a word from the treasures: the shortest path is exactly 25 rooms long — 26, if you include the numberless initial illustration. So if you map each room’s position on that path to a letter in the obvious way, and take the letters of the five rooms containing the artifacts, what do you get? Unpronounceable gibberish. Same for a couple of other mappings I’ve tried. I’m starting to really think that I’m carrying things too far, looking for more hidden meanings when I’ve already wrung the thing dry.

This is especially disappointing because it would have been so easy for the author to place the artifacts in rooms where they do decode to something meaningful! With one exception, they’re not particularly bound to the contents of the rooms where they’re found. If I were in the maze for real, I’d totally move them to rooms where they spell something out, so the next guy could have the satisfaction I was denied, if I could find a way to keep the exit gate open while I did so.

Daedalian Depths is a gamebook in the tradition of Chris Manson’s Maze, where “in the tradition of” is a politer way of saying “that blatantly imitates”. Andrew Plotkin has a review of it here; like him, I got a copy when it was released, but I’m in such a puzzle glut that I didn’t get around to going through it until now. It’s by Rami Hansenne, who also created Codex Enigmatum, which is a lot like Journal 29, which is based on web-based riddle chains like notpron. These are all puzzle-hunt-like things where the solutions of puzzles feed into other puzzles. Journal 29 used a web site as an intermediary between its interlinked puzzles; Codex Enigmatum has an online solution checker, but doesn’t absolutely require it; Daedelian Depths doesn’t have an online component at all. It’s meant to be self-confirming, like a cryptic crossword.

But more importantly, it differs from Codex Enigmatum by the Maze format. Everything is placed in illustrations of rooms, with a page of facing text; clues in each room let you know which exit to take, which is to say, which page to turn to next. CE didn’t have any overarching context other than itself as a book. This makes a tremendous difference to the feel of the thing, making it come off as more of a cohesive whole rather than a mishmash of disparate puzzles, even though that’s really what it is. But it still carries a lot of the CE/J29 feel as well, simply due to the cheap paper and fuzzy, indistinct art style. It’s better than Maze in a lot of ways, but production values are not one of them. (CE and J29 at least had the excuse that you were expected to write on them with pencil.)

I’ll reiterate what Zarf said: The most important innovation this book has over Maze is simply that its riddles are reasonably solvable. Maze had a contest associated with it, so it was expected that most people wouldn’t solve it. DD wants you to win, however much it pretends otherwise. Its second most important innovation is redundancy. Every page has multiple clues indicating which door to take. Sometimes I can’t figure them all out — sometimes I don’t even notice them all. But having multiple clues means I don’t need to. Not only that, multiple clues means that individual clues can afford to be sketchy. This is where the self-confirmation factor gets in: multiple sketchy clues that all point at the same thing add up to good certainty. It’s like science that way.

Let me give a concrete example of what I’m talking about. In one room, there’s a portrait of Beethoven on the wall, showing him standing in front of a large full moon. It’s not the focus of the room or anything, it’s just a detail in the background. But it’s rendered in enough detail to seem important. The juxtaposition of Beethoven and moon suggests the Moonlight Sonata, aka Piano Sonata #14. And indeed the room contains a door labeled “14”. This connection is tenuous enough that it might not convince you on its own that door 14 is the right one, but it’s a strong confirmation.

At any rate, I’m mainly posting about it here as a way to get eyes on my notes. Here they are! If you have the book, you can use this as a source of hints, but what I really want you to do is comment and add to it. Even though I’ve found the correct path through the maze, there are redundant clues that I do not understand, and I want to understand them. To that end, I tried to find an existing forum or wiki, but the results were disappointing, particularly for rooms off the main path; even the author’s own message board had very few comments. So I’m trying to fill that gap.

I recently solved my way through a couple of unrelated puzzle books, Journal 29 and The Librarian’s Almanaq — although perhaps they’re better described as metapuzzle books. These are not merely books that contain puzzles, but books that are puzzles: multi-stage puzzles that take advantage of the physical properties of books. And yet they approach this in such different ways! To put it briefly, Almanaq aims at being a puzzlehunt¹I’d never seen “puzzle hunt” contracted into one word before this book, but I embrace it. in book form, while Journal is modeled on web-based riddle games like notpron, where the solution to each puzzle yields the URL of the next.

Structurally, this means that while Almanaq is kind of freeform and unpredictable, with lots of paging around through the book to look for things, Journal 29 is designed as a sequence of discrete puzzles, each two-page spread acting like a single web page. Journal isn’t quite as linear as its inspirations, though. Since you can’t actually lock access to pages of a book, it settles for solution dependencies: solving each puzzle yields a “key”, a word or other short text, which later puzzles can reference by number. A puzzle might reference multiple keys, and finding them all is generally necessary — often the keys are the only indication of a puzzle’s goal.

One weird point about this: The puzzles don’t yield the keys directly. Instead, there’s an online component. When you solve a puzzle, you go to the journal29.com website and type in the solution, and that gives you the key. A couple of the puzzles rely on other resources from the website, too, combining keys into URLs for you to visit. I really don’t care for this. I’ve complained about such things in the IF Comp before, but it’s a bigger concern there, where archival is half the point, than here, where the pages are designed to be marked up with pencil over the course of solving puzzles, making the whole thing ephemeral and unreplayable. Still, I find the reliance unaesthetic. I suppose it has one virtue, that it makes it impossible to work backwards, solving puzzles from the keys they produce. But I suspect the main motivation is metrics. The online component makes it possible for the author to track which puzzles go unsolved. I hope they’re at least finding this instructive.

Once you have a mandatory online component, you have to ask: Why the book? Is this in fact just an online riddle-game that puts its graphics in the form of print instead of some digital format? But no, some of the puzzles use the medium of print in clever ways. One puzzle requires you to fold over some pages to complete a picture, another uses the uninked parts of a page as a grille for obtaining letters from the next page, and so forth. It’s a small fraction of the puzzles that pull such tricks — by my count, just eight of the 63 puzzles really need the book, although sometimes this is a judgment call. But that ratio means it can come off as a surprise every time it happens.

The reason it can come as a surprise is that the puzzles here are scant on instruction. The basics of how to look up keys on the website are covered in the book’s introduction, but other than that, you’re generally just given some pictures and have to figure out what to do on your own, just like in those web-based riddle games. There’s a fictional premise that the book was recovered from a lost archeological expedition, and some of the graphics lean into this with pictures of ruins or rubbings or sketches of alien skulls or whatever. The puzzles aren’t really designed to make sense with this fiction, but they use it as mood-setting, as a way to invest the act of poring over a mysterious book with a little meaning, however tangentially. It reminds me a little of how hidden-object games tend to adopt mystery and detection themes as a reason for why you’re scouring scenes for random objects, however little sense that makes.

Almanaq, by contrast, has lots of instructions. The premise here is that if you follow the book’s instructions precisely, you will attain enlightenment — and that’s all the premise you get. Indeed, there’s one sub-puzzle that consists entirely of instructions, but instructions that are complicated enough, with enough conditional exceptions, that they’re difficult to follow correctly. (It took me three tries.) But the presence of instructions means that the average page can be far less informative and more enigmatic than the average page in Journal. Flipping though the book, you’ll see a pattern of hexagons containing letters here, a crossword grid without clues there, and then a page with just some thick lines running through it, or a page full of Shakespearean dialogue in iambic pentameter. Some of these things will be used in puzzles. Others are decoys. There are entire types of puzzle, entire genres, that are only seen as decoys. Only by faithfully following the instructions will you know which is which.

The very first thing that the initial instructions tell you to do is to rip the initial instructions out of the book. Ironic, for a book with “Librarian’s” in its name! This goes against everything I had learned about how to treat books so much that I was hesitant to do it at first, but I convinced myself that the pursuit of enlightenment would necessarily involve overcoming my ingrained habits. This is, I think, important. To tear out that page is to commit yourself, to declare that you trust the instructions and will follow them wherever they lead.

The second thing the instructions tell you to do is find certain other pages and rip them out as well, for use as tiles in an assembly puzzle. The instructions give advice on how to rip neatly, but ripping is never neat, and I think that’s part of the point. It’s still humbling you, to make you receptive to enlightenment, or at least to ensure your obedience. There’s some verbiage about how you “don’t need” to use a knife or scissors, which is phrased as encouragement or kindness but acts more like a taunt or criticism. I got better at the ripping with practice, but the more neatly-ripped pages still shared the puzzle with my earlier, rougher efforts.

That initial puzzle gives you the key to finding the pages for the midgame, consisting of several parallel puzzles of different sorts, all of which are ripped out, and in some cases manipulated further. (This time, scissors are sometimes necessary.) In keeping with its puzzehunt ambitions, the book suggests playing with a team, in which case each midgame puzzle can be physically handed to a different team member. This all funnels into a final metapuzzle, and the final step of that final puzzle wowed me. I won’t say why here, in part because I don’t think the impact could be adequately communicated by a mere description. Viewed dispassionately, without being primed by the rest of the game, it may be barely better than the “Aha!” moments from Journal. But at the time, it felt like the book was making good on its promise of enlightenment. It wasn’t, of course, but it did a good job of making me feel like it was.

Now, I enjoyed both of these books. But I enjoyed Almanaq a great deal more. It was purer in its bookness, and took advantage of it more thoroughly, turning the entire book into an arts-and-crafts project. Its less-linear structure made it less stuckable. And of course there’s that ending. Journal, like certain other riddle-games, uses familiar ancient-astronaut imagery to suggest earth-shattering revelations without actually delivering anything other than vagueness. Somehow, Almanaq‘s approach of not even having anything to be vague about works better for me. Aliens just make me more aware that you’re pretending.

And then… I’m fairly sure that I’ve exhausted Journal, but Almanaq leaves some possibility that there’s more to be discovered. I’ve looked at the unused pages a bit, and found one page with a fairly obvious hidden message on it. Is there more like that? Are any of the mysterious decoy puzzles actually things that can be solved? A particularly devious designer could exploit ambiguity in the introductory puzzle to create multiple paths through the rest of the book. I really don’t think anything like that is going on here, but just imagining the possibility is a little exciting.

Apparently the author of The Librarian’s Almanaq is currently preparing a sort-of-sequel called The Conjuror’s Almanaq, this time modeling it after escape rooms. I’m looking forward to that. It’ll be interesting to see how it differs.