SFW vs NSFW

[jack]

Does "SFW" just mean "not NSFW", or does it mean "actively safe for work"? I guess, they are broad categories, but I'm not sure when to interpret SFW as "typical of the SFW category, eg. societally acceptable prose" and when to interpret it as "right up against the boundary of NSFW or I wouldn't have bothered to specify, but not over it" :)

Comments

[cjwatson]

I usually interpret it as "this channel/feed/whatever sometimes has NSFW links, but in this case I've reviewed it and it's fine"; or sometimes as "safe despite dodgy-looking URL". I agree that it sort of implies some reason why the reader might think otherwise, though.

[gerald_duck]

Firmly the latter, in my case: I'll only say SFW if people might otherwise expect it not to be.

For example, this userpic. Or the Danish word "slutspurt". Or my "NSFW" t-shirt.

[jack]

I think this userpic kind of isn't SFW. I mean, I this is fine for a joke about sfw. But I think if someone is working at a job where seeing a penis on the screen is actually a big deal, I think this is clearly "a penis that turns into a rabbit" not "a drawing of a rabbit that people misinterpret".

[simont]

Yes, what cjwatson said, I think. Grice's maxim of quantity permits you to assume that if someone bothered to say "SFW" then there was some reason why you might have suspected otherwise, and that reason might be that it's actually close to the borderline, but then again it might also be that the content is totally unobjectionable and it's just that some piece of surrounding context (like the title or URL or just the forum in which it appeared) made it not look like a good bet.

(Perhaps even deliberately made it look suspicious, e.g. I have it on good authority that some people still think it's funny to put double entendres about 'pussy' in URLs which turn out to be cat pictures.)

[jack]

Ah, excellent observation. I think that ends up meaning, if someone says "SFW" and it looks nsfw, then it probably looks ok, but if it looks sfw, then it probably isn't :)

[andrewducker]

If we assume that people only apply _necessary_ modifiers, then "SFW" must mean that there was some reason to assume otherwise.

[jack]

Yes, I generally assume that, but I'm not sure if I'm right :) I think people _do_ do that, but I think mathematicians forget that some of the time a modifier seems necessary for a reason that isn't obvious to the person they're talking to, so you can't chain together inferences infinitely and get a reliable answer :)

[andrewducker]

I'm fairly sure I can't do _anything_ infinitely. Let alone with a reliable answer at the "end".

[jack]

I think the case is, if you're dealing with pure logic, then you _can_ chain together as many inferences as you like, and as long as each step is valid, the whole chain and eventually conclusion is valid. This isn't necessarily "a whole chain of inferences, and then a conclusion" which is something mathematical logic talks about but is much more difficult. But "infinite" is a way of saying "true for N inferences, however big N is".

Conversely, in the real world, if a conclusion is only true if each of N necessary steps to deduce it is true, and each of those steps you're 90% certain of, the chance of the conclusion being right is 0.9^N, or, well, not that high if there's more than five steps.

But it seems like a mistake I make an awful lot, and everyone makes often without being aware of, is assuming they know why someone is saying something (eg. using a SFW tag), and reasoning why they would have done that, and from that reasoning what mgiht be behind the cut tag and if it would be ok for their office, etc, etc, and actually, it's easy to have made a small mistake in one of those assumptions.

So "if they used a SFW tag, it's usually because there was a need to", is a good guide, but not a perfect guide.

So, it's something like "if it's not true all the way up to infinity, it's usually not true for five", maybe?

[andrewducker]

Is that statement true up to infinity? Or just to four?

[redbird]

I would agree with that, but sometimes "reason to assume otherwise" would be something like "the person's previous post was NSFW" or the existence of a lot of NSFW material on a given subject, even though the post/story/article in question wasn't in any way sexual.

So if I indicated something was "safe for work" that wouldn't necessarily mean "safe but might be borderline" or "safe unless you work for the government" or even "safe unless your coworkers are going to ask why you're laughing so loudly."