What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 134 Stack Exchange communities.

Quite often there are debates about definitions or the question being taken literally over the intended question. This completely throws the original question out of track or even extremely discouraging for new users. Immediately anyone who does not fully understand certain definitions and place it in a slightly off context (general idea is there but not to calibre of the complainer), it's immediately downvoted and closed off.

I find that a better solution might be to understand the intended question. For example, someone who is asking about databases may not understand all the definitions such as tables and will ask questions such as how to put text into the database. This should be a good opportunity to guide the user in a less traumatic and frustrating fashion which is often witnessed by people who just take the question literally for exactly that purpose I suspect. Like the type of people, you say sarcastic things and they would simply take the entire meaning literally without understanding the context and intended message behind it.

A good example I found is Creating non-reverse-engineerable Java programs. The user is simply asking an effective way to make decompiling difficult. One of the users responds in a literal fashion in an overly academic and theoretical tone rather than suggest a realistic and pragmatic solution. It's also an assumption that the question asker is oblivious to the same knowledge as the responder. Sort of like when someone has googled and has not been able to find a satisfying answer is told to use Google. Notice that such answers often get high votes for its high usage of cold logic and reasoning, which do not really help much in answering the intended question behind the question.

Sometimes I wonder if this is just a deliberate attempt to be contrarian for the sake of it, or if some people literally cannot take a question without putting it in a literal sense. I find this to be quite often very counter productive as it just introduces more noise with just esoteric jumble. I'm not expecting people to read the question poster's mind but to simply take in everything in its literal format is quite frustrating, much akin to attempt to have a normal social conversation with someone with the Asperger syndrome.

share|improve this question
3  
I've noticed this behavior and it annoys me as well. –  user152743 Nov 27 '11 at 7:00
13  
Can you choose a more recent question as an example? The one given is from 2008. –  Peter O. Nov 27 '11 at 7:13
2  
Peter O, your statement can be a good example for this question. –  Kim Jong Woo Nov 27 '11 at 7:16
16  
@KimJongWoo: Which is part of the problem. Your example is not particularly illustrative of the problem you suggest. I personally do not actually see what you're talking about as a problem. So if you want to convince others that this is a problem, you should find a more legitimate example. –  Nicol Bolas Nov 27 '11 at 7:29
3  
I see this all the time on Super User as well. Users get into 20 comment long threads about minute definitions, instead of focusing on the problem at hand. It's very frustrating, and sometimes ends with a moderator having to lock the post and remove the off-topic comments. I'd provide examples, but we tend to clean them up pretty quickly. –  nhinkle Nov 27 '11 at 7:32
2  
1  
@NicolBolas, I suggest you to simply find examples which matches this question and your understanding of the discussion at hand if the one's I've listed is not to your liking. Nothing stops you from doing this, if you can find a better example, post it. –  Kim Jong Woo Nov 27 '11 at 8:11
17  
@Kim: The fact that someone looked at the question and does not see the problem is evidence that either you haven't stated your point decently, or you're just hypersensitive. Either way, the onus is on you to provide a decent example. It should be enough that we answer the questions for you; we shouldn't have to ask them for you as well. –  cHao Nov 27 '11 at 8:37
2  
'But "glory" doesn't mean "a nice knock-down argument",' Alice objected. 'When I use a word,' Humpty Dumpty said, in rather a scornful tone, 'it means just what I choose it to mean — neither more nor less.' –  AakashM Nov 27 '11 at 8:59
1  
@cHao, that is hardly convincing argument. Again the classic entrapment by quality of example to validate an argument is only self fullfilling prophecy. One can easily downtalk the strength of argument by invalidating the provided example. You need to counter by focusing on the conclusion. It doesn't take a lot of effort to see the problem at hand often on SE, and your statement only enhances my position that people are not able to see beyond the letter but simply put forward their own skewed opinions through entrapment, ad hominem, which are fallacious ways to construct a useful argument. –  Kim Jong Woo Nov 28 '11 at 3:10
8  
@Kim: I don't need to counter at all. You want to say there's a problem, then it's on you to show that there is one. Til you can provide such evidence, i don't even need to care. –  cHao Nov 28 '11 at 5:16
2  
@KimJongWoo: "It doesn't take a lot of effort to see the problem at hand often on SE" Prove it. If you're right, then it should be no problem for you to provide evidence. In an argument, you don't get to declare that you're right and to challenge others to prove you wrong. You must prove your case. And you have not done so, because the one piece of evidence you have provided is flimsy at best. If you can't even prove that the problem exists, why should we believe you when you say it does? –  Nicol Bolas Nov 29 '11 at 3:47
    
@NicolBolas,@cHao you are reading the proof. or shall I say you are all part of the proof you desperately seek. this question, answers and your responses in itself illustrate the problem being discussed. –  Kim Jong Woo Dec 3 '11 at 5:38
1  
@KimJongWoo: So you're saying that your question is poorly phrased, such that a literal interpretation of your argument does not answer your intent (which remains poorly supported, since you offered no support for it), and therefore generates no discussion that is useful to you? How exactly is that the fault of the people answering the question? People have already told you how to improve your question, but you refuse to do so. Instead, you simply assert that any doubt of your question or argument against it on the basis of lack of evidence constitutes the existence of the problem you claim. –  Nicol Bolas Dec 3 '11 at 5:51
3  
"I used this series of comments as the example that conveys the earlier said problem of people treating SO like a court of law to the letter but not the intended context of the questions being asked." And it fails as such an example, because this isn't a question; it's an invitation to discussion (which is allowed on Meta). It's a debate. And a debate is like a court of law. If you want your argument to be taken seriously, you must provide evidence. Thus far, your evidence has been... one question. From 3 years ago. Without evidence, your argument has no merit. –  Nicol Bolas Dec 3 '11 at 6:09

3 Answers 3

Why are people concerned about the letter of the question? This is both simple and obvious: because it is what they asked for.

You can go round and round, guessing at what someone might have been asking for. You can play games to try to deduce what someone is trying to do by infering whatever from their question. But the simple and obvious fact is that the person asked for X. And therefore, answering X is entirely legitimate.

It's not quibbling over semantics. It is giving them what they asked for. That's why it's important to state questions unambiguously. A literal reading of a question is all one can hope for, because the question is 100% of the information we have on the issue. If you're expecting people to try to read between the lines about what someone might want to do, or what they "really mean" when they ask for something like, "a way to deploy a Java program in a format that is not reverse-engineer-able," then that's just not realistic.

You ask a question. Someone answers. It is not the job of the answerer to break out a crystal ball and try to divine your real intent. It is your job to state your intent up-front. If you ask a poorly specified question, there is every reason to expect that the answers you get will not be useful.

I'm not expecting people to read question poster's mind but to simply take in everything in it's literal format is quite frustrating

How else can it be taken? What do you expect? If someone asks for X, then the answers should be about providing X. Anything else is just trying to read around the question to guess the real intent of the question.

For example, take this question about Unicode parsing. The question asks for the ability to "iterate through these codepoints as a series of characters, not a series of codepoints, and determine properties of each individual character, e.g. is a letter, whatever."

Unicode does not define the concept of a character. Before one can legitimately answer this question, it is important to know exactly what the person is actually asking for. Does he want a Unicode grapheme? Yes, that seems likely, but why not ask for clarification first before charging in with an answer that may not be appropriate?

I'm not sure why you find this "quite frustrating". Programmers live by explicit rules, and they die by ambiguity. Everything we do is fundamentally based on a set of well-defined terms and functionality. Asking for clarification when someone asks about something using invented language is far from out-of-bounds.

share|improve this answer
1  
"I got what I asked for, but not what I wanted": surely we can aspire to better? –  user152743 Nov 27 '11 at 8:04
19  
@user152743: The only way to do better is for the person to ask for what they wanted. Then there won't be a problem. –  Nicol Bolas Nov 27 '11 at 8:06
    
You are assuming that what they ask for is literally what they type but this is not always the case and your attitude is exactly of the nature I am addressing. Your assumption about programmers and how they should behave is not a very convincing one, we come in all shapes and sizes. Concepts should be explainable to best answer questions appropriate for the context for someone who lack the same level of expertise without overloading with unnecessary theoretical or academic talk, if you don't have a pragmatic and context appropriate suggestion. –  Kim Jong Woo Nov 27 '11 at 8:09
12  
IMHO the end result - to be at all usable as a reference for future visitors - should be a correct question/answer pair. Not answering to the letter, although perhaps helpful to the current asker, would create "flawed" answers. In my opinion there are two ways out. Either you coach the OP to reformulate the question to what he/she actually intended to ask, or your answer the literal question. Unfortunately the former is more difficult, if at all possible in some cases. –  Bart Nov 27 '11 at 8:15
1  
@Bart, what do you mean by "flawed" answers? Someone who is reading the same question might have the similar intention, and rarely would someone asking the same question themselves would imply it to the letter, I hardly find that is the grounds for an answer being flawed. However, I agree that the OP should be coached to reformulate the original question to best convey the intended question but again this is often met with objection for being "too vague" or "outside of FAQ" or simply closed as off topic, which in itself selfdefeating purpose In other words, SO may not be beginner friendly. –  Kim Jong Woo Nov 27 '11 at 8:21
6  
"Someone who is reading the same question might have the similar intention" IMO that goes way too far. Now we're making assumptions about intent. Makes my head spin. The solution would be either of the two I described before. However, I do agree up to a point with your objection to closing, in that often such questions get closed without any comment. Ideally a comment which explains the situation and coaches the OP to improve his question might be useful. But my opinion remains, the answer should reflect the question asked. –  Bart Nov 27 '11 at 8:35
2  
If the question is genuinely unclear, it should be clarified by e.g. posting comments. But giving a flippant, literal answer to a question that obviously wasn't intended, using an unreasonably literal interpretation you would never take were a coworker to ask you the exact same thing in person at the watercooler, smacks of arrogance and passive-aggression. It does succeed in ridiculing the OP for his ignorance without actually helping him at all, and adds a uselessly literal but correct answer nobody would ever search for to the archives. I've no idea why either is considered a good thing. –  user152743 Nov 27 '11 at 9:40
3  
@user152743 Passive aggression is not a good thing. The FAQ says "be nice" and that should theoretically translate into people not leaving answers just to mock the OP. –  Anna Lear Nov 27 '11 at 23:43
1  
@NicolBolas, If your boss asked you, find ways to prevent piracy, do you simply say you can't and walk off? If a student asked, how do I put stuff into my variables, do you reply "you can't put stuff in there are you crazy woman?". It doesn't hurt to try understand the position the question asker is in providing a useful and relevant answer or a way to guide the user by asking "is this the intended question" or "I think this is the intended question and here's my response to that" rather than to the letter. It's a skill and taking things literally may come across as rude and offensive. –  Kim Jong Woo Nov 28 '11 at 3:44
2  
@KimJongWoo: "If your boss asked you, find ways to prevent piracy, do you simply say you can't and walk off?" I would engage him in a dialog about what exactly he wants done. SO is not a place for engaging in dialog. It's a place where you dump a question and someone else dumps an answer. That's why the comments are so limited compared to the question/answers. That's why talking to someone is so limited compared to talking to the question. It is the responsibility of the person asking the question to do so understanding that this is not a forum. –  Nicol Bolas Nov 28 '11 at 3:47
3  
@KimJongWoo: I don't believe that the person giving the sarcastic, flippant answer is being helpful. But at the same time, they're being unhelpful because the question is wrong. It is always on the person asking the question to do so in a way that will get useful answers. I don't think answering those questions is useful, but I also don't think it's wrong in some moral sense. Garbage in, Garbage out. –  Nicol Bolas Nov 28 '11 at 3:50
1  
I am not disagreeing that a question is badly constructed or incorrect but my concern is that such flippant answers should not be made at all in the first place rather replaced with a useful guide to formulate an increasingly higher quality questions. I do not think that there is a bad question because the yardstick to measure it requires the assumption that the question asker has the same level of understanding of the subject as you. As this is not often possible with beginners, an answer which at least attempts to address pragmatic solutions should be rewarded than sarcasm which is noise. –  Kim Jong Woo Nov 28 '11 at 4:01
    
@NicolBolas, Do you understand the difference between engaging in a long chat vs providing concise pragmatic solutions to a question? The former is completely arbitrary decision, it is possible to engage the question asker without giving any answers, and it's also possible to give a pragmatic solution by understanding the position of the question asker. Nowhere in that example is an invitation for dialog on SO but an example of what is literal response vs a pragmatic one, something which some people on SO obviously understand but withhold from practice on purpose for it's desired effect. –  Kim Jong Woo Nov 28 '11 at 4:12

Asking good questions is a skill.{*}

"Good" encompasses many things but it certainly includes asking the intended question and not some other, vaguely related question.

A imprecise questions is a bad question.

Good questions are good for the site, and bad questions are bad for the site.

By giving answers to the questions asked instead of some guess at the question intended, we teach this valuable skill, and encourage them to contribute good content later on.


{*} A skill that is valuable in life every bit as much as on the internet; even more so on Stack Overflow and the other technical sites, and most of all for people in (or thinking of entering) technical professions.

share|improve this answer
1  
I agree but giving a good relevant answer is also a skill. To be literal with obviously out of context answers are not at all helpful in real life and SO. I understand if a question is blatantly lacking any foresight but punishing a question asker by being literal for it's ulterior motives is pathetic. Again due to the anonymity and lack of emotional feedback mechanisms but I still find that even people who read to the letter and cannot provide a relevant answer is often rewarded instead of being discouraged by making non-sequitar or cleverly passive aggressive statements. –  Kim Jong Woo Nov 28 '11 at 3:31
    
@Kim: I'm going to take these one at a time. "I agree but giving a good relevant answer is also a skill." I stand by my record of three years and a thousand answer on this. –  dmckee Nov 28 '11 at 5:12
1  
@Kim "To be literal with obviously out of context answers are not at all helpful in real life and SO." Among my real like tasks I am and have been a teacher, and I know how the inability to frame of good question can cripple a student. Often the best help I can provide them is to show them how bad the question is. –  dmckee Nov 28 '11 at 5:13
3  
@Kim "punishing a question asker by being literal" They are not being punished. No one is taking anything from them. Instead we're using our time and effort in a attempt to help them. –  dmckee Nov 28 '11 at 5:15
    
@Kim "it's ulterior motives is pathetic" I will thank you not to assume you know my motive and again to cast no aspersion. –  dmckee Nov 28 '11 at 5:16
1  
@Kim: "Again due to the anonymity and lack of emotional feedback mechanisms" I'm not anonymous here. Anyone who cares enough can find out who am I (though woe betide him who contacts me out of band without a damn good reason!), nor am I emotionally isolated from this community. Again you are assuming that you can deduce my emotional state over this same thin channel that you accuse of being so unresponsive. Don't. –  dmckee Nov 28 '11 at 5:19
    
@Kim: "or cleverly passive aggressive statements" More of the same. –  dmckee Nov 28 '11 at 5:19

The ability to ask the right question requires a solid understanding of the foundations. Therefore, getting the question right is almost a mandatory requirement for the answer to be useful to the asker. Hence the asker should be grateful for suggestions for improvement of the question. In some cases, thinking about what the question really wants can eliminate the need for the question altogether.

That said, what's the best way to respond to a poor question? It All Depends.

If the asker is clearly on track and has a vague grasp of the matter, then a comment to ask for clarification and to suggest improvements is surely the best course of action. If the asker is responsive and interested in dialogue, and willing to revise the question in light of new information, then that's certainly the ideal outcome.

But sometimes a poster is insistent on wanting it his way. This may even come prefaced with "I know the standard way, but I want to do X wihout Z" or something like that. In that case, a to-the-letter answer (perhaps with a comment about the ill-advisedness of the solution) should be an acceptable course of action.

Another situation is where the original poster doesn't engage with the community at all (post-and-run). In that case, it might be worth waiting a day or two, and if the question is still unclear or unreasonable, you might just vote to have it closed.

post scriptum. A bit of experience and gut feeling goes a long way. For example, in C++ I can usually tell when someone proposes a solution that's littered with char * variables even though string processing is called for whether the person might simply not know any better and should be recommended a tidy std::string solution, or whether the question warrants being taken literal.

The more unusual a question, the more well-crafted and researched it should be in order to be taken at face-value without confirmation.

share|improve this answer
    
Just as you say the problem is beginners may often not have the foresight to consider the existing foundations to properly formulate an acceptable level of quality of those established experts. So knowing this, I think it's better to guide the user to reformulate the question by pointing them at the correct direction (as many already do on SO) than simply post responses that are hardly constructive by being literal and clog up the question. –  Kim Jong Woo Nov 28 '11 at 3:48

You must log in to answer this question.

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