How much would it go against the spirit of SE to allow conlanging flavoured challenges here? Some possibilities (not a complete list) I could see working with the format of SE include:

  1. Reconstruction games. In these, one is given a set of supposed cognates, either as a table or on a map (potentially fictional) and has to try and reconstruct the original word(s). This thread has a bunch of them, each more elaborate than the previous.

  2. Deciphering challenges. Given a short text (sentence or small paragraph) in a conlang and some additional information, either in the form of a direct grammar/lexicon dump or other sentences, figure out a translation for the text.

  3. Construction challenges. Within a limited timespan, construct a conlang sketch and translate a small text (1-2 sentences).

I’m not sure 3) would work well here, but I could definitely see the other two being a fun addition to the site. The question is merely… do we want fun? I do hear that there is a puzzling SE, but the knowledge required to work these challenges might be too specific there.


A simple-ish answer is that it would indeed go against the spirit of Stack Exchange. We do have two sites - Puzzling and Programming Puzzles & Code Gofl - dedicated to solving puzzles, but there's one key feature: Those sites are explicitly designed for puzzles. The other Stack Exchange sites are not.

The official, tongue-in-cheek position of Stack Exchange is that We Hate Fun. Obviously, that's not totally true; we do have fun things like Winter Bash semi-regularly, and chat rooms are more laid-back. But on the whole, we keep things serious. It's not really that we hate fun; it's just that it distracts from the serious, real-life problems that people have.

This sort of thing is especially true on small beta sites. I've got no idea what sort of activity Constructed Languages will see in the future, but it won't be huge, maybe five questions a day in the coming months (if we get past private beta). This means that if we get one puzzle question every day or two - and we would get that many, if not more - then 10-20% of the site would be for fun. That ratio's higher than I think most people would like.

That said, we could maybe do this sort of thing in chat, maybe having a weekly game session, where it would be tucked away and wouldn't distract from the main site (although there's plenty of serious business in chat). Maybe a custom room could be made, so the site's main chat room can remain open for more serious stuff. But I wouldn't recommend having this sort of contest on the site itself.

  • 1
    I think that a separate room in the chat is a great idea. I don't see anything wrong in such challenges and I will really like to see them here, at least on the chat.
    – RedClover
    Feb 10 '18 at 11:26

By way of comparison, I am on another Stack Exchange site that has such a challenge type: Chemistry.SE. Within organic chemistry, suggesting syntheses to various organic compounds using known reactions is considered an interesting puzzle. Often, the goal is to find elegant pathways that please the organic chemist’s eye and with as few reactions as possible. There’s a reason why chemists speak of the art (and not science) of total synthesis.

The way the challenges — which are subsumed under the umbrella tag and which are explained in this chem.meta.SE question for those interested — work is that they are posted once every month, users have the chance to submit their answers and finally the accepted answer is awarded to the route with the lowest number of steps (assuming the community agrees by voting that the route will likely work).

This would immediately set an outline of when and how this would be possible.

  • Chemistry receives some thirty questions per day yet the challenges are only once a month. The ratio of normal questions to questions is about 900 to 1. (Note, however, that Chemistry has a somewhat unhealthy closing rate.)

  • While the challenge in itself seems simple, the main question posters take special care to ensure that the challenge is neither too hard (= no answers) nor too easy. They have misjudged a few times which lead to a daunting drop in answers.

  • A challenge needs an unambiguous way to award the checkmark; i.e. a way to measure the goodness of answers. I consider choosing the first answer would be a very bad design principle. Choosing the highest-voted answer would work partially, but chances are that it is just the first answer that rises highest. I’m not sure how one would rate the answers since any constraint I am imagining right now seems very abusable when one has full control over the output.

    (That does not mean that a good method does not exist, it means I can’t think of one off the top of my head.)

    Of course, if the asker has designed the challenge in a way that leaves only one correct answer, then FGITW is a valid checkmark selecting method.

I think it would be a good idea to implement at some point in the future but not before a steady stream of ‘normal’ questions come in. When exactly a good time to start would be and how often to post a challenge would have to be discussed at a later date.


I agree they shouldn't be a regular thing.

One possibility is though to borrow from Mi Yodeya (Judasim) where each year they have one month where Purim Torah questions are allowed, for humorous and satirical misinterpretations of their scriptures. Questions must be tagged, and once the period is over they are all immediately closed.

In the same way, we could have one month a year where these language puzzles would be allowed, perhaps timed to coincide with International Linguistics Olympiad.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .