(FWIW various LessWrongers who have studied the issue don’t agree with Eliezer’s bias against arguments from group selection. (Um, me, for example; what finally convinced me was a staggeringly impressive chapter (IIRC ’twas “The Coevolution of Institutions and Preferences”) from Microeconomics: Behavior, Institutions, and Evolution, though I also remember being swayed by various papers published by NECSI.) I’d be very interested in any opinions from well-read biologists or economists.)
The analogies between biological and social evolution are limited. Not only does group selection work in social evolution, but social evolution is Lamarckian in that it retains acquired traits. So you need to be careful when reasoning from one to another; I think that is one reason people keep trying to “justify” group selection in biology.
The “new” group selection (e.g. here and here) works with both organic and cultural evolution.
Dogs pass on fleas they acquired during their lifespan to their offspring—much as humans pass on ideas they acquired during their lifespan to their offspring. Both the fleas and the ideas can mutate inside their hosts—and those changes are passed on as well.
The differences between organic and cultural evolution are thus frequently overstated. Critically, Darwinian evolutionary theory applies to both realms.
except it’s more like viruses than flies: singificant amounts of evolution can hapen within a single host generation, and entirely different species can crospolinate if they end up within the same host.
Depends on yer memes—but sure, often more like viruses.
“Species” is one of the more tricky areas—if there’s much interbreeding, then maybe it’s not two species. It isn’t just memes, though—bacteria and viruses exhibit this too, as you say.
Not only does group selection work in social evolution, but social evolution is Lamarckian in that it retains acquired traits
Isn’t modern opinion that vanilla natural selection is also non-negligibly Lamarckian? (I suppose it’s very possible that the sources I’ve read over-stated the Lamarckian factors.)
The heuristic I generally use is “use parentheses as needed, but rewrite if you find that you’re needing to use square brackets.” Why? Thinking about it, I believe this is because I see parentheses all the time in professional texts, but almost never parentheticals inside parentheticals.
But as I verbalize this heuristic, I suddenly feel like it might lend the writing a certain charm or desirable style to defy convention and double-bag some asides. Hmm.
(A related heuristic for those with little time is to assume that lots of parentheses is correlated with lack of writing ability is correlated with low intelligence is correlated with inability to contribute interesting ideas, thus allowing you to ignore people that (ab-)use lots of parentheses. I admit to using this heuristic sometimes.)
I find that people who use a lot of parentheses tend to be intelligent, and I think this screens off the alleged inference from lots of parentheses to inability to contribute interesting ideas.
I don’t know whether I’m right in thinking there’s a parentheses/intelligence correlation, but if I am there’s a reasonably plausible explanation. Why would someone use lots of parens? Because when they think about something, a bunch of other related things occur to them too and they want to avoid oversimplifying. Of course it’s even better to think of the related things and then find ways to express yourself that don’t depend on overloading your prose with parenthes, but most people who use few parentheses aren’t in that category.
I don’t know whether I’m right in thinking there’s a parentheses/intelligence correlation, but if I am there’s a reasonably plausible explanation. Why would someone use lots of parens?
((Well), that’s (easy).) (((Heavy) users (of (parentheses))) tend to (be ((LISP) weenies))), and ((learning (LISP)) gives ((a) boost (of) ((15) to (30) (IQ) points), (at least))).
(First impression: You’re talking about the 130 vs. 145 distinction whereas I’m talking about the 145 vs. 160 distinction (which you characterize as “even better”). (Can-barely-stand-up drunk (yet again!), opinions may or may not be reflectively endorsed, let alone right.))
Yes, it’s plausible that we’re talking about different distinctions. But even in the range 145-160 I am very, very unconvinced that using fewer parens is a good sign of intelligence. Perhaps you have some actual evidence? Unfortunately, people with an IQ of 160 are scarce enough that it’ll probably be difficult to distinguish a real connection from a spurious one where it just happens that the smartest people are also being careful about writing style.
(Increasingly contemptuous of your too-drunk-to-stand signalling extravaganza; my comments may be distorted in consequence.)
Yes, I think I have evidence—of about 5 people I know of 160+ IQ, none use many parentheses, whereas I know of a greater than 1 in 6 fraction in the immediate predecessor-S.D. that fall into the parenthesis-(ab)using category. Of course, even I myself don’t put much faith in that data.
(Is my drunkenness-signaling (failed) signaling or (failed) counter-signaling (ignoring externalities in the form of diminished credibility)? I can’t tell.)
Is treating “data” as plural rather than singular correlated with difference between high and very high IQs in your experience? :-)
(I wonder whether I’m evidence one way or another here. I’m somewhere around 150, I think, and I used to use an awful lot of parens and have forced myself not to because I think not doing so is better style. But I’m more concerned with writing style than many other people I know who are about as clever as I am.)
((Counter-signalling is a special case of signalling. It isn’t necessarily (failed) just because I don’t like it.))
Is treating “data” as plural rather than singular correlated with difference between high and very high IQs in your experience? :-)
In my experience that seems to correlate a lot more with conscientiousness and caring about writing style after screening off intelligence. (Also: fuck!—I hate when I forget to treat “data” as plural.)
I used to use an awful lot of parens and have forced myself not to because I think not doing so is better style.
Same here, at least when it comes to writing for a truly general audience or for myself.
(Side note: another thing that confuses me is that intelligence doesn’t seem to me to be overwhelmingly correlated with spelling ability. Not quite sure what to make of this; thus far I’ve attributed it to unrelated selection effects on who I’ve encountered. Would be interested in others’ impressions.)
I have found entirely the opposite; it’s very strongly correlated with spelling ability—or so it seems from my necessarily few observations, of course. I know some excellent mathematicians who write very stilted prose, and a few make more grammatical errors than I’d have expected, but they can all at least spell well.
I know many very intelligent good spellers, and several very intelligent mediocre spellers, and one or two very intelligent apparently incorrigibly atrocious spellers. I don’t know any moderate-intelligence good spellers, a few moderate-intelligence atrocious spellers, and quite a few quite a few moderate-intelligence mediocre spellers. I don’t know very many dumb people socially, and mostly don’t know how good their spelling is as they don’t write much. People I met on the Internet don’t really count, as I filter too much on spelling ability to begin with.
(Since you two seem to be mostly using the mentioned IQ scores as a way to indicate relative intelligence, rather than speaking of anything directly related to IQ and IQ tests, this is somewhat tangential; however, Mr. Newsome does mention some actual scores below, and I think it’s always good to be mindful when throwing IQ scores around. So when speaking of IQ specifically, I find it helpful to keep in mind the following.
There are many different tests, which value scores differently. In some tests, scores higher than about 150 are impossible or meaningless; and in all tests, the higher the numbers go the less reliable [more fuzzy] they are. One reason for this, IIRC, is that smaller and smaller differences in performance will impact the result more, on the extreme ends of the curve; so the difference in score between two people with genius IQs could be a bad day that resulted in a poorer performance on a single question. [There is another reason, the same reason that high enough scores can be meaningless; I believe this is due to the scarcity of data/people on those extreme ends, making it difficult or impossible to normalize the test for them, but I’m not certain I have the explanation right. I’m sure someone else here knows more.])
(Hence my use of parentheses: it’s a way of saying, “you would be justified in ignoring this contribution”. Nesov does a similar thing when he’s nitpicking or making a tangential point.)
No, that time passed when you merely had a single parenthetical inside a parenthetical. But when you have a further parenthetical inside the former two, is it then time to break out the curly brackets?
The “new” group selection (e.g. here, here and here) has been demonstrated to be pretty-much equivalent to the standard and uncontroversial inclusive fitness framwork in a raft of papers.
There is widespread agreement that group selection and kin selection — the post-1960s orthodoxy that identifies shared interests with shared genes — are formally equivalent.
That’s not to say that group selection is useless—since it involves different models and accounting methods.
Group selection models, if correctly formulated, can be useful approaches to studying evolution. Moreover, the claim that group selection is kin selection is certainly wrong.
These folk apparently don’t grok the topic too well.
(FWIW various LessWrongers who have studied the issue don’t agree with Eliezer’s bias against arguments from group selection. (Um, me, for example; what finally convinced me was a staggeringly impressive chapter (IIRC ’twas “The Coevolution of Institutions and Preferences”) from Microeconomics: Behavior, Institutions, and Evolution, though I also remember being swayed by various papers published by NECSI.) I’d be very interested in any opinions from well-read biologists or economists.)
The analogies between biological and social evolution are limited. Not only does group selection work in social evolution, but social evolution is Lamarckian in that it retains acquired traits. So you need to be careful when reasoning from one to another; I think that is one reason people keep trying to “justify” group selection in biology.
The “new” group selection (e.g. here and here) works with both organic and cultural evolution.
Dogs pass on fleas they acquired during their lifespan to their offspring—much as humans pass on ideas they acquired during their lifespan to their offspring. Both the fleas and the ideas can mutate inside their hosts—and those changes are passed on as well.
The differences between organic and cultural evolution are thus frequently overstated. Critically, Darwinian evolutionary theory applies to both realms.
except it’s more like viruses than flies: singificant amounts of evolution can hapen within a single host generation, and entirely different species can crospolinate if they end up within the same host.
Depends on yer memes—but sure, often more like viruses.
“Species” is one of the more tricky areas—if there’s much interbreeding, then maybe it’s not two species. It isn’t just memes, though—bacteria and viruses exhibit this too, as you say.
Yea, I oversimplified a bit.
Isn’t modern opinion that vanilla natural selection is also non-negligibly Lamarckian? (I suppose it’s very possible that the sources I’ve read over-stated the Lamarckian factors.)
When you have a parenthetical inside a parenthetical inside a parenthetical, is it time to break out the square brackets?
No, it’s time to take out some of the round ones.
I find that even the trivial heuristic “delete all parentheses” usually improves what I write.
(But it’s no fun if you can’t construct [all kinds of {silly }] elaborate nested parentheses [in your comments {in case that wasn’t clear}])
You forgot a period.
Well, that’s the danger with using parentheses.
The heuristic I generally use is “use parentheses as needed, but rewrite if you find that you’re needing to use square brackets.” Why? Thinking about it, I believe this is because I see parentheses all the time in professional texts, but almost never parentheticals inside parentheticals.
But as I verbalize this heuristic, I suddenly feel like it might lend the writing a certain charm or desirable style to defy convention and double-bag some asides. Hmm.
(A related heuristic for those with little time is to assume that lots of parentheses is correlated with lack of writing ability is correlated with low intelligence is correlated with inability to contribute interesting ideas, thus allowing you to ignore people that (ab-)use lots of parentheses. I admit to using this heuristic sometimes.)
I find that people who use a lot of parentheses tend to be intelligent, and I think this screens off the alleged inference from lots of parentheses to inability to contribute interesting ideas.
I don’t know whether I’m right in thinking there’s a parentheses/intelligence correlation, but if I am there’s a reasonably plausible explanation. Why would someone use lots of parens? Because when they think about something, a bunch of other related things occur to them too and they want to avoid oversimplifying. Of course it’s even better to think of the related things and then find ways to express yourself that don’t depend on overloading your prose with parenthes, but most people who use few parentheses aren’t in that category.
((Well), that’s (easy).) (((Heavy) users (of (parentheses))) tend to (be ((LISP) weenies))), and ((learning (LISP)) gives ((a) boost (of) ((15) to (30) (IQ) points), (at least))).
(First impression: You’re talking about the 130 vs. 145 distinction whereas I’m talking about the 145 vs. 160 distinction (which you characterize as “even better”). (Can-barely-stand-up drunk (yet again!), opinions may or may not be reflectively endorsed, let alone right.))
Yes, it’s plausible that we’re talking about different distinctions. But even in the range 145-160 I am very, very unconvinced that using fewer parens is a good sign of intelligence. Perhaps you have some actual evidence? Unfortunately, people with an IQ of 160 are scarce enough that it’ll probably be difficult to distinguish a real connection from a spurious one where it just happens that the smartest people are also being careful about writing style.
(Increasingly contemptuous of your too-drunk-to-stand signalling extravaganza; my comments may be distorted in consequence.)
Yes, I think I have evidence—of about 5 people I know of 160+ IQ, none use many parentheses, whereas I know of a greater than 1 in 6 fraction in the immediate predecessor-S.D. that fall into the parenthesis-(ab)using category. Of course, even I myself don’t put much faith in that data.
(Is my drunkenness-signaling (failed) signaling or (failed) counter-signaling (ignoring externalities in the form of diminished credibility)? I can’t tell.)
Is treating “data” as plural rather than singular correlated with difference between high and very high IQs in your experience? :-)
(I wonder whether I’m evidence one way or another here. I’m somewhere around 150, I think, and I used to use an awful lot of parens and have forced myself not to because I think not doing so is better style. But I’m more concerned with writing style than many other people I know who are about as clever as I am.)
((Counter-signalling is a special case of signalling. It isn’t necessarily (failed) just because I don’t like it.))
((()))
In my experience that seems to correlate a lot more with conscientiousness and caring about writing style after screening off intelligence. (Also: fuck!—I hate when I forget to treat “data” as plural.)
Same here, at least when it comes to writing for a truly general audience or for myself.
(Side note: another thing that confuses me is that intelligence doesn’t seem to me to be overwhelmingly correlated with spelling ability. Not quite sure what to make of this; thus far I’ve attributed it to unrelated selection effects on who I’ve encountered. Would be interested in others’ impressions.)
I have found entirely the opposite; it’s very strongly correlated with spelling ability—or so it seems from my necessarily few observations, of course. I know some excellent mathematicians who write very stilted prose, and a few make more grammatical errors than I’d have expected, but they can all at least spell well.
I have the opposite impression, but now that I have that correlation it’s hard to make further unbiased observations.
I know many very intelligent good spellers, and several very intelligent mediocre spellers, and one or two very intelligent apparently incorrigibly atrocious spellers. I don’t know any moderate-intelligence good spellers, a few moderate-intelligence atrocious spellers, and quite a few quite a few moderate-intelligence mediocre spellers. I don’t know very many dumb people socially, and mostly don’t know how good their spelling is as they don’t write much. People I met on the Internet don’t really count, as I filter too much on spelling ability to begin with.
(Since you two seem to be mostly using the mentioned IQ scores as a way to indicate relative intelligence, rather than speaking of anything directly related to IQ and IQ tests, this is somewhat tangential; however, Mr. Newsome does mention some actual scores below, and I think it’s always good to be mindful when throwing IQ scores around. So when speaking of IQ specifically, I find it helpful to keep in mind the following.
There are many different tests, which value scores differently. In some tests, scores higher than about 150 are impossible or meaningless; and in all tests, the higher the numbers go the less reliable [more fuzzy] they are. One reason for this, IIRC, is that smaller and smaller differences in performance will impact the result more, on the extreme ends of the curve; so the difference in score between two people with genius IQs could be a bad day that resulted in a poorer performance on a single question. [There is another reason, the same reason that high enough scores can be meaningless; I believe this is due to the scarcity of data/people on those extreme ends, making it difficult or impossible to normalize the test for them, but I’m not certain I have the explanation right. I’m sure someone else here knows more.])
(Hence my use of parentheses: it’s a way of saying, “you would be justified in ignoring this contribution”. Nesov does a similar thing when he’s nitpicking or making a tangential point.)
That would ruin the aesthetic.
No, that time passed when you merely had a single parenthetical inside a parenthetical. But when you have a further parenthetical inside the former two, is it then time to break out the curly brackets?
The “new” group selection (e.g. here, here and here) has been demonstrated to be pretty-much equivalent to the standard and uncontroversial inclusive fitness framwork in a raft of papers.
Here’s Marek Kohn writing in 2008:
That’s not to say that group selection is useless—since it involves different models and accounting methods.
There are still a few dissenters. E.g. Nowak, Tarnita and Wilson (2010) apparently disagree—saying:
These folk apparently don’t grok the topic too well.
For a more modern and knowledgeable group selection critique, see:
Social semantics: how useful has group selection been?, West, Griffin and Gardner (2009).
Here’s Stuart West on video, covering much the same topic.