I'm DM-ing an upcoming campaign and am considering banning or nerfing spells like Comprehend Languages because they can be used to instantly solve potentially interesting problems like reading an ancient text.
I'd prefer to avoid banning spells and would like some advice on techniques for working around (better yet, working with) these kinds of spells that could make otherwise interesting puzzles and challenges boring.
I've gathered that concealing the information in something other than the literal meaning of words (e.g. ciphers, steganography, Thieves' Cant) would defeat the effect of the spell. This either makes understanding the language essential to solving the problem or it renders the spell utterly useless, so it seems suboptimal to me.
How might I design a puzzle such that Comprehend Languages (or a similar spell such as Tongues) would provide an advantage without outright solving it?
The puzzle that inspired this question was from a D&D podcast where an ancient text was on a pedestal and the text itself basically gave away the solution to the puzzle once translated. The DM forgot that this particular spell existed, so the Wizard used the spell and totally thwarted the puzzle. Perhaps this puzzle was intended to be solved with a Rosetta Stone or some other form of problem-solving and inferences. Perhaps it was an alternate script of English text intended to be inferred through frequency analysis and other clues. The intended solution could have been something along those lines, but since I'm not the one who thought up the puzzle and I don't remember any other details, I don't know what the DM had in mind.
I'm hoping to be able to create a similar puzzle without either nullifying the spell or requiring its use. I'm not totally sure if I will use this in my campaign, but I want to keep it in my DM toolbox.
Side note: I am working in D&D 5e, though this question could apply to other systems with spells in the same vein.
Best Answer
This is a gating problem.
In the design of games occasionally we like to put bits of the world behind gates, be they literal or metaphorical. The upper floors of the tower are behind a locked door... and now we look for the key. A dungeon-area is flooded and inaccessible... until we learn to breathe water or redirect the river. Clues are hidden in the foreign-language journal... until we comprehend languages.
You've noticed, though, that certain spells can shortcut the gates we put into games.
That's fine. In fact that's good. Barriers in adventure design should be used to pace narrative, to keep squishy parties from too-hard regions, to tantalize upcoming adventures! But once they're surpassed, their remaining would just introduce a new chore into the world. If the key to the tower's upper floors is a one-use key, requiring a new fetch-quest each time we want to go there... we simply won't go there.
Work with your gates. Before your party gets there.
For every gate you put into your adventures, you've got to look at your party composition and resources (friendly NPCs? Spell scrolls?) and figure out the lowest level at which the party could get past that gate. Flight? Toxic air? Going to another plane? Getting past a guardian-monster?
Then evaluate whether that's when you want your party to get past there. If not, you need a tougher gate. (Unfortunately, 5e makes it very difficult to come up with mid-tier gates.)
In your case, it means that "ancient text" is not a significant gate. But (as Robert Pain VanZant's answer describes) "text is literally hidden" or "meaning is coded" or another method may achieve the same gate. Or maybe you need to look to another gating method, based on your party's composition.
(Don't get rid of all the gates your party has a solution to, though: it's good to feel character choices being organically rewarded by making some things easier. Just be aware, when you design, that you're not actually gating for that character.)
See also AngryGM's Megadungeon Monday: Gating which probably is the source for half of my examples up there ^^ =D