Related to: Generalizing from One Example, Connecting Your Beliefs (a call for help), Beware the Unsurprised

The idea of this article is something I've talked about a couple of times in comments. It seems to require more attention.

As a general rule, what is obvious to some people may not be obvious to others. Is this obvious to you? Maybe it was. Maybe it wasn't, and you thought it was because of hindsight bias.

Imagine a substantive Less Wrong comment. It's insightful, polite, easy to understand, and otherwise good. Ideally, you upvote this comment. Now imagine the same comment, only with "obviously" in front. This shouldn't change much, but it does. This word seems to change the comment in multifarious bad ways that I'd rather not try to list.

Uncharitably, I might reduce this whole phenomenon to an example of a mind projection fallacy. The implicit deduction goes like this: "I found <concept> obvious. Thus, <concept> is inherently obvious." The problem is that obviousness, like probability, is in the mind.

The stigma of "obvious" ideas has another problem in preventing things from being said at all. I don't know how common this is, but I've actually been afraid of saying things that I thought were obvious, even though ignoring this fear and just posting has yet to result in a poorly-received comment. (That is, in fact, why I'm writing this.)

Even tautologies, which are always obvious in retrospect, can be hard to spot. How many of us would have explicitly realized the weak anthropic principle without Nick Bostrom's help?

And what about implications of beliefs you already hold? These should be obvious, and sometimes are, but our brains are notoriously bad at putting two and two together. Luke's example was not realizing that an intelligence explosion was imminent until he read the I.J. Good paragraph. I'm glad he provided that example, as it has saved me the trouble of making one.

This is not (to paraphrase Eliezer) a thunderbolt of insight. I bring it up because I propose a few community norms based on the idea:

  • Don't be afraid of saying something because it's "obvious". It's like how your teachers always said there are no stupid questions.
  • Don't burden your awesome ideas with "obvious but it needs to be said".
  • Don't vote down a comment because it says something "obvious" unless you've thought about it for a while. Also, don't shun "obvious" ideas.
  • Don't call an idea obvious as though obviousness were an inherent property of the idea. Framing it as a personally obvious thing can be a more accurate way of saying what you're trying to say, but it's hard to do this without looking arrogant. (I suspect this is actually one of the reasons we implicitly treat obviousness as impersonal.)

I'm not sure if these are good ideas, but I think implementing them would decrease the volume of thoughts we cannot think and things we can't say.


New Comment
47 comments, sorted by Click to highlight new comments since: Today at 2:41 AM

Imagine a substantive Less Wrong comment. It's insightful, polite, easy to understand, and otherwise good. Ideally, you upvote this comment. Now imagine the same comment, only with "obviously" in front. This shouldn't change much, but it does. This word seems to change the comment in multifarious bad ways that I'd rather not try to list.

Uncharitably, I might reduce this whole phenomenon to an example of a mind projection fallacy.

I have a different explanation: this is a status defense mechanism. If you say something that other people find obvious, in a way that suggests that you didn't find it obvious, you lose status-points for not being as smart as them. By adding the word "obviously", you in effect say "please do not infer that I think this remark is a great discovery of mine (and thus that I am ignorant relative to you) from the mere fact that I think it needs to be stated explicitly".

As an added benefit, if the remark turns out not to be obvious to your audience, yet demonstrably true, you gain status for having been smarter than them.

You might think, then, that there is no downside to simply prefacing every statement you think is true with "obviously". Obviously, however ( :-) ), you have to avoid making it transparent what you're doing, and thus restrict your usage of "obvious" to particularly plausible cases. Calibrating this sense of plausibility with your own epistemic powers is one of many mysterious (in the sense of not being spoken about or taught explicitly) techniques of human status negotiation. (And heaven help you if you label "obvious" something that is false...)

More broadly, "obviously" can signal how you expect your audience to react. It can signal, for example, that the reason you're not giving a detailed explanation for your statement is that you take for granted that your audience will agree.

This is a rather important oversight, and because of that this doesn't quite strike me as an article that belongs in "main."

Is this really a contextually relevant oversight? Most terms do have multiple uses, but they depend a lot on the context for their applicability. I might be missing something, but I get the impression that the post's primary purpose is to highlight the problems with using the concept of obviousness here (and could plausibly be extended to do so in other circumstances where you're dealing with an audience to whom you can't immediately measure the inferential distance).

Using the concept of obviousness to signal that you possess or anticipate a certain level of knowledge has its, uh, obvious strengths, but I happily read the post as an explanation of how that usage might include some rather undesirable side-effects.

I'm not really concerned with where the post belongs in a broader sense, so I'm not challenging that statement, just its prior condition.

Good comment. It's a shame I have a policy of never upvoting anything that has a smiley face.

Only one problem: you misinterpreted me, and it's my fault entirely. The proximity of those two paragraphs was not meant to indicate that the latter explains the former. That second paragraph refers to the whole phenomenon of the article, not just what happens to the perception of an otherwise good comment prefaced with "obviously". I actually noticed this mistake before publishing but didn't fix it for some reason.

Is there something we can do about it? Are there community norms we can encourage to lower these status-based barriers to understanding?

If "obvious" things are still being said, prefacing them with "obviously," doesn't seem to be a barrier to understanding.

Signalling doesn't have to be that straightforward. A clever individual (of which we have a few) may choose to be significantly more circumspect, and imply that a piece of knowledge is obvious by omitting it from a statement that presupposes it, or alluding to it off-hand. We do this all the time, but I'm going to say that this probably has more to do with mind projection than anything else. It often simply won't occur to us to modulate a statement to encompass the receivers.

However, I don't know if this is a ploy we can entirely defeat just by making obviousness a bad word. If anything, that just requires people trying to make such a ploy to be circumspect...

Well, my first thought reading this was "look at that, worrying about what people think of you and trying to look cool messes everything up again."

This 'obviously' insertion trick may be rewarded with social pretentiousness brownie points, but as we can see, it also has negative consequences that, I feel, are rather more important. As a remedy, I invite you (and everyone) to join me in working on not caring so much about sounding cool enough.

This is an ongoing project of mine and I'm not nearly at a point yet where social insecurity and pretentiousness don't make any of my decisions for me any more, but at least realising that these are petty and counterproductive things to worry about helps to loosen their grip on your brain a bit.
I'm working on a brand of modesty based on the hypothesis that if you're really good at something, people will often notice it even if you don't signal it, and a need to signal it is just costly nonsense that biases you and gets in the way of your peace of mind, and might even get you stuck in delusions of entitlement to admiration that you haven't earned. And I appease my remaining urge for pretentiousness with the thought that being noticeably great at something without showing it off makes you look all the more badass. Someone with an amazing skill you never would have known they had (and if they've had that hidden in them, who knows what else they can do!) seems a lot cooler to me than someone -- even a more skilled person -- who milks their merits for every last thumbs-up they can get out of them.

Note however that I am not involved with important political matters where my reputation as a Very Smart Person could actually benefit me in more substantial ways than ego boostery.

I'm working on a brand of modesty based on the hypothesis that if you're really good at something, people will often notice it even if you don't signal it, and a need to signal it is just costly nonsense that biases you and gets in the way of your peace of mind

People are also good at ignoring things that are inconvenient for them. Consider an office politics situation where being good at your job may mean that someone else's status gets lowered. You may have to signal that you're good at your job in order to get noticed at all.

There's also the problem that even if it's obvious, obvious+signal is still going to beat out your obvious+no signal. By your reasoning you don't need to walk into a job interview wearing a suit, because your resume should speak for itself. But then the next guy with an equally good resume and a suit comes in and gets hired over you.

More generally: If you've "rationally" deduced that you don't really need to follow pointless social conventions, you're almost certainly wrong and have failed to consider something. Chesterton's Fence applies, at least.

Indeed. Like I mentioned briefly in my footnote, I understand that this is not an approach that you can apply that generally, in any situation. Particularly if you actually somehow depend on other people's impressedness for something that matters to you, actively putting effort into impressing them (if done right) will probably get you more reliable results. If you really need people to think you're amazing, I guess my approach would be a pretty big gamble. The whole point of being subtle is to accept the risk that people won't notice, which works well for art but not for traffic signs.

That's not really my purpose with this, though. The purpose of this idea is mainly to liberate yourself from the urge to impress people at all. Again, you can't always afford to do that -- we all know a job interview is not the moment for modesty -- so the scope would have to be limited to those situations where looking clever really isn't all that important, but I think that still covers a sizeable proportion of them. Including, very much, writing comments on LessWrong that may or may not contain the word 'obviously'.

There is no reason an action like this can't have a compound cause. I would guess that, in the hypothetical, the person is not actually thinking "Okay, I'll preface this with 'obviously' so that I look good." However, it is likely that, since saying "obviously" is high status, they wouldn't think too hard about whether the thing is in fact obvious - certainly not as hard as if they were about to say something low status.

I might propose another way someone might use "Obviously": As a codeword for arguing by definition. There are obviously several ways in which you can use that word. (See what I did there?)

I wish I could remember where I read this (or even in what academic field). But some academic once wrote that his most acclaimed, most cited papers were always the ones he thought of as mere summaries of existing knowledge. This made a strong impression on me. In most cases when dealing with high-level ideas, very good restatements of previous research are not only valuable, but likely to make those ideas click for some non-trivial number of readers. A few other thoughts:

Perhaps you read it here: Explainers Shoot High. Aim Low!:

A few years ago, an eminent scientist once told me how he'd written an explanation of his field aimed at a much lower technical level than usual. He had thought it would be useful to academics outside the field, or even reporters. This ended up being one of his most popular papers within his field, cited more often than anything else he'd written.

Addendum: With his gracious permission: The eminent scientist was Ralph Merkle.

That's it, thanks. I should have known it was on Less Wrong!


Mathematician Gian-Carlo Rota also made a similar comment in his 10 Lessons I Wish I Had Been Taught, giving some examples of mathematicians better known for their expository work.

There is plenty of low-hanging fruit in merely doing your research and saying the obvious.

Darn, I wish I'd come up with that line myself!


The lukeprog era of Less Wrong: "It's amazing what you can learn when you look shit up!"

I can completely believe that these papers were successful (as measured by citations for example), but that does not necessarily mean they were the most useful papers or that people got the most out of them.

In a typical paper, somewhere in the introduction it will be necessary to say some basic "establishing the field" statements. Academics want to support these statements with references. A reference that says some basic thing, in plain words with very little technical hedging, is much easier to find and cite than a series of more targeted and precise points that add up to the same thing. At least in my field the papers that get the most citations are exactly these introduction citation ones.

A good example of an arguably "obvious" result making big waves is DiVincenzo's criteria. I don't think that many people really would have questioned that a "useful quantum computer" needed to be able to [I have changed the order of the 5 criteria]  (2) write data, (4) do logic gates. (5) read data. While also being  (1) big enough to be useful and (3) quantum. Its not a million miles from a tautology, with (1) and (3) translating to "useful" and "quantum" and (2,4,5) being pre-requisites for a thing to be called a computer. But, if I am writing a paper introduction saying "X satisfies the Divincenzo criteria [1]" sounds so much cooler and more considered than "X is a possibly a good platform for quantum computing".

[I am sure that spelling out the criteria was useful. But suspect that the level of attention, as measured by citation, is probably outsized relative to the usefulness.]

This is not (to paraphrase Eliezer) a thunderbolt of insight. [...]

This sentence seems exactly the same to me as saying, "This was obvious, but, [...]".

Sometimes, people assert obviousness as a self-deprecating maneuver or to preempt criticism, rather than because they believe that everyone would consider the statement in question obvious.

There a running joke in mathematics that saying that a statement is "obvious" means:

I know this "has to" be true, but don't feel like figuring out how to prove it.

When lecturing, saying the word "obvious" is a signal for the students to begin panicking and self-doubting. I wonder if some instructors do this intentionally.

Saying that something is 'obvious' can provide useful information to the listener of the form "If you think about this for a few minutes you'll see why this is true; this stands in contrast with some of the things that I'm talking about today." Or even "though you may not understand why this is true, for experts who are deeply immersed in this theory this part appears to be straightforward."

I personal wish that textbooks more often highlighted the essential points over those theorems that follow from a standard method that the reader is probably familiar with.

But here I really have in mind graduate / research level math where there's widespread understanding that a high percentage of the time people are unable to follow someone who believes his or her work to be intelligible and so who have a prior against such remarks being intended as a slight. It seems like a bad communication strategy for communicating with people who are not in such a niche.

"Yes, gentlemen, it IS obvious!"

I always imagine that line being delivered by John Cleese (as in "he IS the Messiah!").

There's an excellent definition of how "obvious" should be used in mathematics: something is obvious if and only if "a proof immediately springs to mind".


The problem remains that it's not particularly helpful to know that a proof immediately springs to the lecturer's or textbook author's mind. And so, operating under the assumption that people are trying to communicate only relevant information, whenever I see 'obvious' or 'easily seen' in a mathematical text, I can't help but read it as an obnoxious 'you should know this already -- unless you're dumb or something'. I think that the best norm for using the word 'obvious' and its variations would be to not use it at all.

This might be a defensive mechanism. Being explicit and formal is very important in mathematics but when you choose which proofs to omit, you're making a judgement about which kind of presentation will lead to the greatest level of understanding among the intended audience. This judgement is more about psychology rather than mathematics and is necessarily based on fuzzy intuitions. Maybe mathematicians are uncomfortable with that and call such omitted proofs 'obvious' to make criticism costly in terms of status.

No! How is this different than saying:
There's an excellent definition of how "obvious" should be used: something is obvious if and only if "an explanation immediately springs to mind".

This should be in Main.

This is kind of funny. I had removed a sentence from a draft of this post about how I considered putting it in Main, but it obviously didn't belong there.



Has anyone else had the experience of being afraid to post "obvious" things?

I write up comments and delete them because I think they're obvious or meaningless more often than I actually post.

I upvoted this for being an interesting admission, but I'd encourage you to go a head and post more of your comments.

Excellent points, all of them. Well, researched, too. Belongs in Main?

I suppose if one really wants to communicate to the reader that a certain point does not appear very deep to the author, they can start with "It may already be obvious to you that..."

I can't really find any objections to this, and agree that maybe it should be a top level post. I like the comparison with probability. It suggests that we should probably talk more about obvious things, because those are the things we understand well enough to actually talk about.

I like this post, and I would support putting it in Main if you added some material on the signaling implications of calling things "obvious" and integrated them with the rest of the post. (I would be fine with putting it in Main anyway, but it would be better with the added content.)

I'm not sure if these are good ideas, but I think implementing them would decrease the volume of thoughts we cannot think and things we can't say.

Could you explain how you think that would work?

The first follows from the second. (In that light, it's obvious now that I put them in the wrong order.) If the pressure (real or imagined) to not say "obvious" things is lifted, that allows people to say things they otherwise could not. As a consequence, people who never would have thought of the supposed "obvious" thoughts are now figuratively allowed to think them by virtue of having seen them in the first place.

That is not what Eliezer and Graham mean by "cannot think" and "can't say" in the essays you linked.

Sometimes, people hesitate to state the obvious because they falsely think the social norms or social consensus would disapprove. Cf. "The Emperor has no clothes."

I once worked at a job where the decision-makers were notably conservative but the line workers were not. The line workers had a tendency to spin reports conservatively, even in situations when I think the decision-makers would not have naturally been as conservative.

However, merely lifting a norm against saying obvious things is not going to solve that problem.


Then let me get something obvious off my chest. I could paraphrase this in terms of UDT to get upvoted, or I could do it the clarity way to craft the kind of LW I want. Here it is:

This is how you change:

Change a thing, change a process or reintrepret.

That's inspired by the of-for approach to policy analysis

I agree with most of this, though occasionally when

(someone says something is obvious) AND (it's not so to me) => I have a knowledge gap I should be looking at.

but yeah, if we have to decide on a norm I'd go with the non-obvious option

I think a better approach than doing away with the notion that obviousness is bad (because, to be honest, if something really is obvious to you, getting a detailed explanation of it can be very annoying), might simply be to explain concepts like inferential distances and mind projection to posters who don't seem to understand them. If people understand those problems of communication and others like them implicitly, they can more easily allow themselves to say something that might be obvious. At least it works that way for me. I won't explain seemingly obvious preconditions of a discussion unless called upon to do so, but I do my best not to assume that everything that might be obvious, is. There are usually plenty of clues. Even if it sometimes requires someone eventually saying "Uh, what does that mean?"

Maybe I'm being terribly optimistic. In my example of one, however, knowing that I have knowledge others might not share is usually enough to make me check if they understand me instead of making the supposition that they do.


According to a professor I once had, in mathematics, "obvious" means something like "something I think is true but don't have a convenient proof for."

[This comment is no longer endorsed by its author]Reply