This was mentioned in another question, so I'm starting a new one to see if we can flesh it out a bit, namely the use of some sort of leveling of questions so that users can filter the site and responders also know how to tailor the question. The idea seems to make sense but we will need a good system in place for it to be successful.
I don't believe the advantage of specifying perspective is enough to justify the otherwise useless meta-tags. The only advantage I can see from such a system is allowing some users to filter out beginner questions, and I'm just not convinced that that's either
a) Something we really want people doing. This isn't like SO where there's a fair chance an advanced learner will be useless on Fortran/Haskell/m4 questions.
b) Worth the overhead of tagging every single question with a level. If you don't tag every single question, a necessarily subjective process, the goal above is not accomplished.
Where does a stylized anime expression, that is fairly basic conversational Japanese, go? One that would be inscrutable to someone only familiar with the "proper" Japanese studied for the JLPT?
Taking your suggested tags in order:
beginner jlpt-n5 Considering that the focus of this site is a language, even very simple questions can have very complex answers (joshi comparisons are a good example of this). By arbitrarily telling people "don't use big words here", you limit the usefulness of the site. Remember that the focus of this site is on the answers, not the specific question.
native-speaker Why? Presumably a native-speaker would be asking very advanced or obscure questions, making this redundant.
advanced-japanese Again... how would this not be obvious? Anyone filtering on this tag would either be missing a number of them, or not successfully filtering out at least half of the actual advanced questions.
One idea that I have is based upon the JLPT levels as most students of the Japanese will encounter it fairly quickly in their studies and should be familiar with it to give a fair idea as to what the test requires. Likewise, it also gives people an idea so to the language skills of the person asking the question so they don't use kanji that the asker may be unfamiliar with. That said, here's my basic idea for the system:
- beginner - A extremely basic beginner who is unfamiliar with the language and likely only understand romaji.
- jlpt-n5 - The asker understands hirajana, katakana, and the kanji required for the level N5 of the JLPT.
- jlpt-n4 through jlpt-n1 - Same as jlpt-n5 conceptually, but increasing in terms of understanding.
- native-speaker - The asker is a native speaker or has a native speaker understanding of the language.
- domain-specific - The terminology being used applies to a specific domain (i.e. medicine).
- unknown-level - Would be a flag for other users on the to categorize the question if they can.
Using jplt is bad though. I didn't feel people with the highest jplt degree had specialized knowledge of the language (I see jplt as a bragging right, though I find most of these people are good at passing exams but forget most of what the studied once passed...) - I'm extremely sorry if that shocks someone, it's just a personal observation over time, not directed at anyone :) Feel free to ignore that part.
The level of a question is relative to one's own knowledge of the language. Those levels are going to be completely subjective.