I've noticed an exponential increase in NYTimes articles being massively upvoted here lately with such garbage articles as this, "A New Luxury Retreat Caters to Elderly Workers in Tech (Ages 30 and Up)" and "NSA has ended phone monitoring based on an anonymous aide."
NYT is a quality journalism outlet, obviously not every article is amazing, but if enough people upvote it, it gets to the front page. It's not some big conspiracy.
I was thinking about that one, and while it can mean destroying one tenth as the Romans said it, it also seems like it could mean destroying all but one tenth, linguistically speaking.
The NYT is right; 'exponentially' is widely misused.
The font-rendering article on today's HN front page [1] is guilty of it, for example: "There is a pixel or less of extra space between kerning pairs and the problem grows exponentially when letter-spacing values are fiddled with". No, it doesn't. There is no exponential growth in the size of the extra space.
I don't see a problem with metaphorical overstatement (e.g. "he was bleeding all over the floor"), especially in casual usage. The real problem would be if they insisted on claiming it's "literally exponential" ... except, of course, everyone insists on breaking that API too, because they're too lazy to come up with a good intensifier.
As for the example you've cited, I think you're strawmanning it -- it's claiming that the problem grows exponentially, which could be true, or close enough, if the missing pixel required you to add custom behavior for all possible combinations of a parameter's value that you didn't even have to worry about before.
>The NYT is right; 'exponentially' is widely misused.
Not really, since both the intended and conceived meaning is "increasingly more", "a hella lot" -- and has nothing to do with exponents.
So it's a successful use of language, Wittgenstein-style. Of the dictionary definition, one is the standard mathematical one, but another pertains to the casual use: "rising or expanding at a steady and usually rapid rate". And there's an even more casual element of actual use of the word, that the dictionary is a few years behind in reflecting.
Interesting case where, on HN, I used "exponential", intending its technical meaning as a critical part of my point, and was accused of using the lay meaning (see last response):
Sorry, didn't read because of the paywall, but the reckless use of this term has bothered me for years, and some things to keep in mind when deciding what is exponential: (1) When something is increasing exponentially, the rate of increase increases in direct proportion to how much of it there is. So unless there is some plausible explanation of growth driven by size, 'exponential' is unlikely to be apt. (2) A true exponential curve increasing over time started infinitely long ago with an infinitesimal magnitude, has grown by the same constant percentage each millennium since, and will continue thus growing forever to eventually exceed every finite limit. (3) The actual rapid growth of anything significant we observe will apparently coincide with such a curve for only a much shorter time. Thus, if you call attention to rapid exponential growth of anything, you are almost certainly and most importantly describing a process that must reach some limit sooner or later, stabilize or crash.
Think of accepting misused exponentiallys, decimateds, literallys, etc., as obedience to Postel's Law: Be liberal in what you accept, and conservative in what you send. Stretch your brain to comprehend around malapropisms while striving to avoid them yourself.
Of course if one of the things that you comprehend is that the speaker is not well spoken there's no need to discard that datum.
Mmm, I think someone could misuse a word like “exponential” while still making a good point or saying something worth thinking about. And vice versa. There are more important things than being “well-spoken”.
All these pundits don't understand that language is its use (as any linguist could tell, and dictionaries try to reflect, e.g. covering "literally" as meaning "virtually", "nearly" too).
Language it's not the etymology of the words, or some canonical meaning set in stone.
You can still make an error: it's when your use of a word is not a widely understood use (and thus you fail to communicate what you meant) -- not when your use doesn't match the etymology.
27 comments
[ 3.0 ms ] story [ 47.1 ms ] threadMaybe it's just me, but this seems like garbage.
NYT is a quality journalism outlet, obviously not every article is amazing, but if enough people upvote it, it gets to the front page. It's not some big conspiracy.
The font-rendering article on today's HN front page [1] is guilty of it, for example: "There is a pixel or less of extra space between kerning pairs and the problem grows exponentially when letter-spacing values are fiddled with". No, it doesn't. There is no exponential growth in the size of the extra space.
1: https://news.ycombinator.com/item?id=19312404
As for the example you've cited, I think you're strawmanning it -- it's claiming that the problem grows exponentially, which could be true, or close enough, if the missing pixel required you to add custom behavior for all possible combinations of a parameter's value that you didn't even have to worry about before.
"in effect; in substance; very nearly; virtually: I literally died when she walked out on stage in that costume."
Not really, since both the intended and conceived meaning is "increasingly more", "a hella lot" -- and has nothing to do with exponents.
So it's a successful use of language, Wittgenstein-style. Of the dictionary definition, one is the standard mathematical one, but another pertains to the casual use: "rising or expanding at a steady and usually rapid rate". And there's an even more casual element of actual use of the word, that the dictionary is a few years behind in reflecting.
https://news.ycombinator.com/item?id=13869092
https://blog.codesolvent.com/2016/10/hype-on-hacknews-and-si...
Of course if one of the things that you comprehend is that the speaker is not well spoken there's no need to discard that datum.
[But yes, it bugs me too.]
(that's another one that grates on me). I don't know who suffers more: physicists, mathematicians, or classicists.
Language it's not the etymology of the words, or some canonical meaning set in stone.
You can still make an error: it's when your use of a word is not a widely understood use (and thus you fail to communicate what you meant) -- not when your use doesn't match the etymology.
http://existentialcomics.com/comic/268