I often see a pattern on ELU where someone asks a fairly innocuous question and then many people pile on with answers that seem like just opinions based on what they think, backed up by no real argument or proof, no evidence of usage from the broader corpus, no reference to known, accepted, or respected authorities.

How do you sort the good from the bad?

How can you tell if someone here actually knows something about a subject? It sometimes feels like people just weigh in on whatever takes their fancy, irrespective of their credentials, knowledge, or understanding.

migrated from english.stackexchange.com Sep 27 '17 at 13:07

This question came from our site for linguists, etymologists, and serious English language enthusiasts.

  • 2
    +1 for the question. There are few actual linguists here and as in any discipline even experts can disagree, so expertise is no guarantee of getting a correct answer. So we all just pretend like we are experts, even people here who have no idea what we're talking about. – AmE speaker Sep 27 '17 at 12:39
  • 6
    Give us some examples of things you looked up on the site. As for users, there are linguists, and polyglots, and writers, and know-it-alls, so I'm surprised you haven't found anything informative. – Mari-Lou A Sep 27 '17 at 12:42
  • 6
    If you see anything you believe is "misinformation", you should downvote - and ideally also post a link to whatever you think is an "authority" supporting your position. Which by the way would often be just that - a position. Bear in mind there are a lot of "indeterminate" aspects to language use, since it's largely based on loosely-agreed-upon conventions, not immutable "facts". – FumbleFingers Sep 27 '17 at 12:51
  • 3
    First, my apologies for using the wrong forum. Secondly, I fear I've fallen into the newbie's trap of inadvertent rudeness. I should say that my intention was light-hearted provocation, not trolling. Yes, there's a lot of opinion, but from what I can tell the general rule here is that someone asks a fairly innocuous question and then twenty people pile in with different opinions based on what they think, backed up by no real argument or proof, no evidence of usage from the broader corpus, no reference to known, accepted, or respected authorities. How do you sort the good from the bad? – Which Tyler Sep 27 '17 at 12:56
  • Undoubtedly there are many folks at this site who are guilty of talking through their hat (see phrases.org.uk/meanings/talking-through-your-hat.html). In an open forum such as this is, the folks who DO know what they're talking about are expected to "call" people in the comments sections of Q's & A's. Moderators too are expected to do their part in maintaining scholarly standards, but by and large they do not expect questions and answers which sound as though they're culled from masters' theses and doctoral dissertations. English, after all, is an art form, not an aggregation – rhetorician Sep 27 '17 at 12:58
  • 1
    of rules to which there are no exceptions. Even famous authors (and speakers) "break the rules" from time to time, sometimes consciously and sometimes not. We therefore need to cut people some slack, even if they do not have an arm's length of degrees and have published in the most prestigious of journals. Have you thought of calling people on this site when you spot something which is inaccurate or just plain blather? – rhetorician Sep 27 '17 at 12:58
  • 3
    Were the poor answers you saw posted as comments or answers? Answers tend to be held to a fairly high standard on ELU, so much so that when regulars want to say something unsubstantiated, it's often done as a comment. – Lawrence Sep 27 '17 at 13:15
  • 4
    I'm voting to close this question as off-topic because it does not appear to be about English Language & Usage Stack Exchange or the software that powers the Stack Exchange network within the scope defined in the help center. It is just a rant dressed up as a question which is clearly rhetorical. – David Sep 27 '17 at 13:16
  • 2
    @WhichTyler Wow! British people still say cheerio! That's fantastic. I've never seen it used in the wild before. I feel like I've spotted a rare bird. – Dan Bron Sep 27 '17 at 13:26
  • 2
    @WhichTyler I've cleared up some comments, since the question is now in the right place on Meta. Currently this does look like trolling, though. Please edit in some examples. – Andrew Leach Sep 27 '17 at 13:26
  • 1
    @Mari-Lou A: I wasn't thinking in terms of the OP bolstering his position here with examples of "misinformation". S/he wouldn't need any rep to post an answer - and if that answer convincingly debunked existing answers, I have faith this would be recognised & upvoted. After which the OP would gain enough rep to indulge themselves with downvotes and comments throughout the site. – FumbleFingers Sep 27 '17 at 13:38
  • 2
    I think this is a reasonable question, about the trustworthiness of the answers here in general and in particular. However, it is couched in wording bearing towards the tendentious. Is there a way you can reword this so that it doesn't come across that way? – Mitch Sep 27 '17 at 14:39
  • 1
    @Hank and Mitch, I apologise for my tendentious phrasing. I'm a bit busy with work at the moment, but I'll edit later so that I sound less of an aggressive berk. (I should also add that I don't assume that only grammarians can offer advice - I just used that as an obvious example. With qualifications comes expertise. There are of course other bases for expertise.) Many thanks for all the answers. It certainly clarifies how it works around here. I have more to say, but time is against me today. PS Cheerio, Dan Bron! – Which Tyler Sep 27 '17 at 16:06
  • 2
    I tried to improve the phrasing a bit with some text from the comments (I'm afraid to go too far, but I think I haven't gone far enough and I'm not sure that OP will be back to edit before it's closed). I don't think this should be closed, especially since it'll get messy with the rejected migration. On the subject of the question, I always include some form of evidence in my answers (and it often bothers me that answers don't do this). Also, see this related question. – Laurel Sep 27 '17 at 19:48
  • 1
    @EnglishStudent Rants and unsubstantiated opinion are considered off-topic, main or meta. The wording of the OP is rant-leaning. I suspect that that might be a reason for the close-voters. The OP has been asked to reword. Ranty stuff is for chat. Even then politeness is valued. – Mitch Sep 29 '17 at 13:48

You bring up a good point about the reliability of answers and author reputation.

Sometimes I see answers that sound authoritative but offer no supporting evidence, or, as you are deftly not putting into these words, sound like they're pulling stuff out of their ass.

Sometimes I do that. In my (partial) defense, some questions are not covered by dictionaries or style guides or corpus searches. Sometimes all there is is personal introspection (and linguistic science tries to build up beyond that). Other times...now I'm just being defensive.

ELU ain't Stack Overflow. You can't compile a sentence to see if it works. A sentence sounds good or not at all or somewhere in between. Also, as a native speaker, you don't have to have a lot of linguistic or self knowledge to tell if some grammar is weird or to come up with a vocab item but you certainly need both to explain well why the grammar or vocab might be appropriate. That is, non-experts can have something useful to add, they just may not be good at backing it up.

But your real question is about how you can trust what people say, or maybe where you're leaning is how you can trust a particular individual.

The structure of the SE software is intended to capture what the community (all the other visitors to the site) think about the questions and answers. Reputation is a numerical device that is simply the weight of votes for and against each individual's answers (and questions).

Reputation is a proxy for reliability. As a proxy, it is not perfect. People vote for things they like, how they're said, if the match what they already believe (confirmation bias), if they sound fancy (appeal to bombast), if they've said smart things before (appeal to authority, which may be a fallacy or not), etc, etc).

There are lots of visitors to ELU. Most just read. Some stay and ask questions. Some stay and answer. There's no guarantee of authority, just interest. There is a small handful of experts, actual professors of linguistics, people who are writers/editors. But for the most part, it's just regular people.

Sometimes people give their opinion stated as fact. Some people just cut and paste from Wiktionary or some site they googled, both of which have their own difficulties with reliability.

Surely having a high rep is no guarantee of authoritative knowledge. There are a handful of non-native speakers here who have gained a lot of rep points purely because their questions are so interesting. They are self-acknowledged learners of English and don't consider themselves reliable about English patterns. But they have a high rep. (there are also some non-native speakers who I would trust about their intuition on English, well, maybe almost as much as my own).

There are some suggested guidelines for writing answers which can make things more reliable. Citations from authoritative dictionaries (like OED or MW), examples of text found, corpus searches (from COCA or NGrams), all having their own qualifications and difficulties.

Yeah, I often see answers that seem like they pulled them out of nowhere or made up as plausible but there is no basis in fact. I find that annoying, and downvote if it is egregious or comment to ask for elaboration. The SE software is set up to deal with this kind of problem. That is, sure, don't trust the first thing you see and question the answer. This ain't as reliable as the Encyclopedia Britannica, but it also isn't as bad as Yahoo Answers (actually you can rely on YA to be not expert advice).

Of course, I usually fail to notice problems if I agree with the answer.

I realize that this is not a very satisfying answer. I'm not really saying either here are the desiderata for a user or a particular answer (because the reason you're asking this question is because rep is not very reliable), or on the other side that there are no good measures and it's the wild west and anything goes. Probably the best is a little experience here, seeing which persons write well, the form of which kinds of answers are trustworthy. Reputation points are a reasonable quick first approximation.

I think the moral is that ELU is an attempt to be as reliable as possible given that it is crowd-sourced by a number of people with varying levels of knowledge.

  • 1
    Ah, a good answer saves the question. I have now retracted my vote to close as "unclear". – NVZ Sep 27 '17 at 15:53
  • 1
    Very good and detailed answer, but I think you forgot to mention one last possible bias, the nativeness one, by those who think their answers are correct just because they are native speakers. – user66974 Sep 27 '17 at 18:22
  • 2
    @Josh there are so many biases I didn't mention. And as with most biases. sometimes they are fallacious and sometimes not. Native speakers re (almost by definition) correct as to what is natural speech, but nativity (?) is independent from conscious knowledge of patterns, history, and explanation. That is, trust a native speaker as to whether an utterance 'sounds' right, but otherwise English speakers may know more about English language facts because most 20th c science about language is written in English and about English. – Mitch Sep 27 '17 at 19:09
  • Reputation is a proxy for longevity. It really has little to do with expertise. And since the voting system is unreliable, so is reputation. – AmE speaker Sep 29 '17 at 14:09
  • 1
    @Clare I agree, with the qualification that the intention is that rep points are a proxy for reputation/reliability, but it turns out to measure both popularity and longevity. I think that calls for tweaking the system of voting or education about the meaning of voting rather than a blanket dismissal. – Mitch Sep 29 '17 at 14:23
  • Reputation has little to do with longevity and popularity, but with activity and consistency. Users upvotes as well as downvotes, but you wouldn’t get reputation points just for being here. – user66974 Sep 29 '17 at 15:49
  • 1
    @Josh Yes, but upvotes count for 10 and downvotes count for -2. So some users who would actually have very low figures (or maybe even negative ones) have very high rep, just by virtue of having posted thousand of answers. So, for example if 4 out of my 5 questions gets a score of minus one and one of them gets plus one, I end up with an overall positive score. So with that kind of hit rate I can get thousands of rep by answering thousands of questions despite the fact that only 20% of my answers are any "good." – Araucaria Oct 8 '17 at 13:17

I would like to raise an issue (often raised in the Academia) of authority versus fact. Not only clashes may occur between those who would "prefer to err with Galen than be right with Harvey" (e.g. sticking with classic grammarians versus modern usage) not to mention that grammar is not a hard science and facts are more difficult to establish.

What disturbs me sometimes in those discussions, is that academic degree, place of birth, etc., seem to matter more than fact and a system of snobism could surreptitiously establish itself, with a ruling class. In that system, people who do not have "credentials" would have to bring more burden of proof than others. In other words, the argument ad personam would still seem to have currency.

In the extreme, consulting the reputation of a person instead of the facts pointed out in an answer (with or without learned references), is still argument ad personam and, from a strict scientific perspective, a fallacy.

The social mechanism of reputation does have its use however, which (as the word clearly states), tells about social acceptance. It is used as a proxy for truth for lack of better, but let's not forget that it is a very imperfect mechanism, just as "peer-review" in Academia. The system has some degree of validity, but we have to live with the risk that some invalid answers may have high rating because they match common opinion, while some good factual ones may be "downgraded" because they don't match that opinion.

As physicist Feynman said:

For a successful technology, reality must take precedence over public relations, for Nature cannot be fooled.

For my part, I will lend my full unreserved respect to the answer of any "maverick" born in a non-English speaking town, even with no degree or reputation at all, who can bring about facts in an answer that appeal to my curiosity and which I can endeavour to verify. Otherwise, how could anyone acquire reputation?

But even beyond this, this tells a lot about the attitude of modern society toward facts. Facts are what is observable, not what authorities, commissions, boards and even academics and teachers tell us to observe.

Of course, it is not always easy to discern facts from opinion and many people are asking for guidance.

But that's where the problem is. The value that we should all cultivate on this forum, is that defined by Immanuel Kant as Enlightnement, which is:

Enlightenment is man's emergence from his self-imposed nonage. Nonage is the inability to use one's own understanding without another's guidance. This nonage is self-imposed if its cause lies not in lack of understanding but in indecision and lack of courage to use one's own mind without another's guidance. Dare to know! (Sapere aude.) "Have the courage to use your own understanding," is therefore the motto of the enlightenment.

And that is my answer to the question:

How do you sort the good from the bad?

Have a look at reputation, citations and whatnot.

But dare to know! Use your own observation and understanding! Liberate yourself from the self-imposed nonage of social acceptance and intellectual authority!

Stack Exchange is a fairly workable mechanism for a forum. It is far from perfect and I would like to commend those who are behind it because they are fully aware of that and permanently trying to improve it.

But to give credence to the myth that truth will somehow emerge fully formed from a system of social metrics such as reputation and degrees, etc., would reduce these forums (as Feynman would put it) into cargo cult science: a series of rituals devoid of substance.

Oh well, I am sure that my words are going to raise discomfort and anger and rebukes as "who do you think you are to think you can think"? But isn't that the lot of any person who wants to trust fact rather than opinion, even authoritative opinion?


Yes, many users here are expert linguists or professors of English. I would recommend looking at the bios of some of our top users. However, as with every stack exchange site on the network, there is no rule prohibiting any person from creating an account and participating by asking questions or answering questions, and potentially accruing high reputation if their answers are consistently upvoted.

The main two ways to determine whether an answer has veracity, besides using your own judgment, are to observe the reputation of the user who posted the answer and to note the number of up- or down-votes they receive.

There will undoubtedly be cases where a non-expert or poorly-researched answer gets many upvotes. There are also many excellent expert answers that get little attention and few up-votes. This is an unavoidable aspect of the Stack Exchange model, and other sites on the network are no different in that regard.

If you believe an answer is not useful, lacks research effort, or is unclear, you can help our community by downvoting those answers and upvoting answers that you believe are high-quality.

  • I hope we didn't go against your wishes too much by undeleting this answer. I thought it was pretty good and needs to be said. – Mitch Sep 28 '17 at 14:43
  • @Mitch not at all, if people find it useful I have no problem with it being undeleted. – RaceYouAnytime Sep 28 '17 at 14:54
  • Why did this Q receives 4 close votes as off-topic? It is very much on-topic as it inquires about how users are qualified to write credible answers on ELU. Why vote to close? Simply write answers telling newbie OP how we are required to support our arguments with credible references (also how reputation and downvotes work) -- so I really appreciate the answers of Mitch and @RaceYouAnytime! – English Student Sep 29 '17 at 12:39

Wow, what happened? I write responses to questions when I feel like it. I'm always on point. My responses are casual and targeted. I respond differently if the person posting is an ESL user or a native speaker trying to write a job cover letter. I've noticed some people who write responses here are pedantic and ostentatious for no reason.

I say, go with your gut feeling. If a response seems condescending, then screw it. I received negative votes for pointing out that "waker" is not a noun (as in a person), but if said out loud sounds like "wanker" - British slang for jerkin' off. You can't please everyone.

You must log in to answer this question.

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