Re: They Found It

1

are meant to make us forget that there's a lot of guesswork in medicine and I have no doubt that in fifty years today's treatments will seem just slightly more sophisticated than bleeding with leeches

Let's not get carried away here.


Posted by: gswift | Link to this comment | 10- 6-07 1:48 AM
horizontal rule
2

Why would that be getting carried away? 40 years ago was the heyday of the transorbital lobotomy, which was just a bit later than the 'Electroshock for everything' fad. And now look! Electroshock is back! Sorta.

max
['And let's not forget that 10-15% of patients acquiring an infection in the hospital thing!']


Posted by: max | Link to this comment | 10- 6-07 2:00 AM
horizontal rule
3

Because shit like leeches really was guesswork, but modern drug therapies really aren't in the same class. Actual chemical targets and such.


Posted by: gswift | Link to this comment | 10- 6-07 2:05 AM
horizontal rule
4

really really really


Posted by: gswift | Link to this comment | 10- 6-07 2:05 AM
horizontal rule
5

Actual chemical targets and such.

Ya know how they find drugs like Prozac and such? They shoot up rats/mice with the stuff, then the carve out their brains, feed the brains into a blender and measure dopamine and serotonin levels, because they think those chemicals may have something to do with depression! And they might even be right, but you know, nobody knows!

max
['Can feel me now? Good!']


Posted by: max | Link to this comment | 10- 6-07 2:16 AM
horizontal rule
6

Gah. Yeah, Eli Lilly took a bunch of chemical analogs to a compound with known antidepressant effects, and tested them to see which which was the most effective serotonin reuptake inhibitor. That, my friends, is just like bleeding someone with leeches.

Oh, wait. No it's not.


Posted by: gswift | Link to this comment | 10- 6-07 2:33 AM
horizontal rule
7

I'd rather have my appendix in front of me than a piece of my brain.


Posted by: eb | Link to this comment | 10- 6-07 4:19 AM
horizontal rule
8

Aren't modern scientists discovering all sorts of useful purposes for applying leeches these days?


Posted by: Jackmormon | Link to this comment | 10- 6-07 6:06 AM
horizontal rule
9

Just to drain blood from healing wounds.

Maggots can be used to clean dead flesh out of wounds also.


Posted by: John Emerson | Link to this comment | 10- 6-07 6:17 AM
horizontal rule
10

The maggot thing sounds so weird: you have to give 24 hours notice that you'll need them, and then you can pick them up from the pharmacy in a little box that, from what I understand, stays on, so that you don't have to watch the maggots doing their thing to your wound. Wouldn't you feel them on there, though? [Shudder.]


Posted by: Blume | Link to this comment | 10- 6-07 7:14 AM
horizontal rule
11

i think that for an extra fee you can watch.


Posted by: John Emerson | Link to this comment | 10- 6-07 7:15 AM
horizontal rule
12

Wouldn't you feel them on there, though?

Yes. Yes you would.

That link is completely work-safe, but do read the comments before following the further link. There are things you can't unsee.


Posted by: apostropher | Link to this comment | 10- 6-07 7:38 AM
horizontal rule
13

Not to derail things from a worthy maggot-on-penis theme, but gswift is reminding me why it is that some people have the negative arrogant-scientist stereotype. I understand what he's reacting against - the derogation of medical science as mere guesswork - but it's coming at the cost of pretending that, for instance, the subject of the original post doesn't exist.

Medical science in 2007 doesn't have all the answers. That's OK.

I think it's safe to say that, in X number of years in the future, the idea of chemotherapy will seem actually worse than leeches. It's incredibly brutal and dangerous and primitive; it just happens to be the best we can do right now.


Posted by: JRoth | Link to this comment | 10- 6-07 7:59 AM
horizontal rule
14

With regard to anti-depressants, there was a thread here or at B's where people talked about the horrendous side-effects that they experiences. Psychopharmacology is still pretty empirical in the "whatever seems to work" sense.


Posted by: John Emerson | Link to this comment | 10- 6-07 8:03 AM
horizontal rule
15

BTW guys, I'm on the way to Fargo today to visit my brother. Ta-ta! Eat your hearts out!


Posted by: John Emerson | Link to this comment | 10- 6-07 8:04 AM
horizontal rule
16

I'm not going to Fargo, alas, but I'll be out all day, so let me add a little gas to the mix before I go:

The can't-you-see-we-know-what-we're-doing-? scientist I'm talking about also comes up in GMO debates/arguments. I really don't want to start one of those here, but the format of them is interesting. Basically, you have a big group of people who have doubts about GMO. Some are, in fact, Luddite radicals who are paranoid to the point of fantasy about what food scientists do. But a lot are just people who recognize that the law of unintended consequences really is a law, and want to move forward under a premise of first do no harm, not "lookit what I can do."

And the response of scientists is absolute disdain that anyone would - could - doubt their intentions and abilities. As if bad science never happens, only bad scientists; and, since these noble scientists only want to feed the world, they are clearly Good, and shouldn't be questioned by hoi polloi. I've seen this firsthand, and I've seen it countless times in threads on GMOs. It's truly astonishing that these people are offended that anyone would question them (I've gotten this from economists too, incidentally).

I don't know if it's underlying insecurity, or if the arrogance goes all the way down, but it always amazes me. I belong to a profession that significantly fucked up in the last century (modern architecture destroying the cores of old cities and, often, promoting sprawl), so I take it as a given that big, beautiful ideas can be radically wrong. I still believe in big ideas and ideals, but I don't imagine to myself that they're foolproof, or that I'm not the fool. I wonder why some other professions don't seem to have learned this lesson.


Posted by: JRoth | Link to this comment | 10- 6-07 8:36 AM
horizontal rule
17

Being the kind soul that I am, I've just sent this link to my sister, whose belly is still swollen from having her appendix removed two days ago.


Posted by: susan | Link to this comment | 10- 6-07 8:58 AM
horizontal rule
18

Please, please, people. There's no need to worry your little heads about what we of the white coat clan are up to. Now, I have to get back to carving out the bleeding edge of human knowledge, so why don't you keep yourselves busy with a videogame or crayons or whatever.


Posted by: cerebrocrat | Link to this comment | 10- 6-07 9:04 AM
horizontal rule
19

There are 'actual chemical targets' but may of the systems aren't well understood. Especially anything psychotropic; the effects are going to be pretty much unknowable unless something is a close analogue to another, known, chemical.


Posted by: yoyo | Link to this comment | 10- 6-07 9:13 AM
horizontal rule
20

Oh, wait. No it's not.

The traditionally held "Father of Medicine" was the Greek physician Hippocrates (460-370 BC), who taught medicine on the island of Cos. One of his major precepts was the rule of harmony, the theory that all body systems were in balance and that disease resulted from an imbalance. Galen (130-201 AD) was the physician to Marcus Aurelius and became the heir to Hippocrates and one of the most influential physicians of all times. He taught the importance of maintaining balance between the four bodily fluids, or "humors" (2): blood, phlegm, yellow bile, and black bile. Each fluid was associated with a specific personality characteristic. Blood was associated with a sanguine personality, that is laughter, music, and a passionate disposition. Someone with a phlegmatic personality was sluggish and dull, while yellow bile represented an individual quick to anger or choleric (cholera meaning yellow as in yellow fever). Lastly, black bile represented a melancholic or depressed personality, melan meaning black. It was the job of the physician to restore harmony in those four humors by the use of emetics, cathartics, purgatives, and by bloodletting. Bleeding was used to reduce excess circulation, to slow the pulse, and to "reduce irritation", all felt to be the cause of inflammation. Dr. Benjamin Rush, a signer of the Declaration of Independence, was a major figure in colonial American medicine and was an important proponent of bleeding, though unfortunately he mistakenly thought that the body held 12 instead of 6 quarts. Shortly before his death, George Washington was bled 4 ‡ quarts in 24 hours for an infected throat and died not long after (3).

max
['{waggling fingers} SCIENCE!']


Posted by: max | Link to this comment | 10- 6-07 9:19 AM
horizontal rule
21

The story I heard about the discovery of the first MAO inhibitor is that it was given to mental patients with colds, under the belief that it was an antihistimine.


Posted by: cerebrocrat | Link to this comment | 10- 6-07 9:41 AM
horizontal rule
22

but gswift is reminding me why it is that some people have the negative arrogant-scientist stereotype. I understand what he's reacting against - the derogation of medical science as mere guesswork - but it's coming at the cost of pretending that, for instance, the subject of the original post doesn't exist.

Neighbor please. We live in a fucking country where something like half the population is still not on board with evolution, homeopathic type remedies are getting popular again, and our president is the member of a political party aggressively pushing "global warming, the jury is still out."

Discussing science with most people is like discussing economics with libertarians. They're just shockingly ignorant of how stuff actually works.

I think it's safe to say that, in X number of years in the future, the idea of chemotherapy will seem actually worse than leeches.

"Chemotherapy" means chemical treatment. You think all chemical intervention will have disappeared? We're going to heal ourselves by flexing out chi or something?

I still believe in big ideas and ideals, but I don't imagine to myself that they're foolproof, or that I'm not the fool. I wonder why some other professions don't seem to have learned this lesson.

Yeah, cause this stuff isn't actually debated in scientific circles. Scientists universally just think everything that can be done should be done.

Are we going to ground this in reality at all?


Posted by: gswift | Link to this comment | 10- 6-07 9:42 AM
horizontal rule
23

So gswift, I can't quite figure out what you're arguing. Yes, medical techniques are vastly more sophisticated than they were 50 or 100 years ago. Yes, we know much more than we do. But the history of empirically based medicine is a history of constant realizing our profound ignorance on some topic or other. The anti-depressant example is a really good one: people literally have no good idea why they might work. Are you arguing they'll never figure it out? That there's nothing more to learn? Or that nothing we learn will change the treatment strategy currently in use?


Posted by: Sifu Tweety | Link to this comment | 10- 6-07 9:47 AM
horizontal rule
24

"Chemotherapy" means chemical treatment. You think all chemical intervention will have disappeared? We're going to heal ourselves by flexing out chi or something?

Dude, are you being clueless or just annoying here? "Chemotherapy", in non-technical parlance, refers specifically to cancer treatment rather than drug treatment generally. And while it's the best we've got now, and Yay, modern medicine and all that, a lot of it operates on the "Let's poison you until you almost die, and count on the tumor cells being fucked up enough to die before the rest of you does" principle, if I understand matters correctly.

It's not the fault of current medical science that we don't have anything better yet, they're working as fast as they can, but assuming that fifty years down the road we're going to have better techniques that make our current treatments look pretty darn primitive doesn't seem unreasonable at all.


Posted by: LizardBreath | Link to this comment | 10- 6-07 9:53 AM
horizontal rule
25

What I'm arguing is that we are in fact doing better than "let's bleed this fucker and see what happens."


Posted by: gswift | Link to this comment | 10- 6-07 9:55 AM
horizontal rule
26

I don't think anybody disagrees with you, gswift. The point is that in 50 years we will be doing as much better than we are now as we are now doing better than "let's bleed this fucker and see what happens."


Posted by: Sifu Tweety | Link to this comment | 10- 6-07 9:56 AM
horizontal rule
27

My dad was an MD who referred many patients for cancer treatment, and while he was dying of cancer he said "Oncologists are professional optimists". And sure enough, his oncologist gave him an unreasonably optimistic prognosis.


Posted by: John Emerson | Link to this comment | 10- 6-07 9:57 AM
horizontal rule
28

The fake boobs of today are comparable to George Washington's wooden dentures. In a century fake boobs will be indistinguishable from real boobs, and the boob of the future will be adjustable to fit whichever outfit it's intended to accessorize.


Posted by: John Emerson | Link to this comment | 10- 6-07 10:00 AM
horizontal rule
29

I have a lot of conflicting intuitions on this topic. I've dabbled in enough history of science to think that the attitudes of previous generations weren't irrational. They had theories about how the body works, and their treatments were designed based on their theories. We're better at it now, and we can cure a lot more things, but it's a mistake to think that we're protected from making mistakes just because we now proceed scientifically. It's not just the same as bleeding someone with leeches, because we know a lot more.

(And to the average person, 'Eli Lilly looks at chemical pathways' is about as much of an explanation as 'Eli Lilly does magic to kill the sad demons in your brain.')

But lobotomies and electroshock aren't long enough ago to throw in the pre-scientific method bin. So what explains them?

"Chemotherapy" means chemical treatment. You think all chemical intervention will have disappeared?

Chemotherapy refers to a specific kind of chemical treatment, i.e., marinating the tumor in poision and hoping it kills the tumor before the person. I think I remember reading that new ideas in cancer therapy are more along the lines of personalized genetic therapy, turning cancer into a chronic illness instead of a terminal one. In other words, I sure as hell hope we manage to improve on chemotherapy.


Posted by: Cala | Link to this comment | 10- 6-07 10:01 AM
horizontal rule
30

I think that gswift is correct that, despite our ignorance on many issues, modern science really is qualitatively different , and better, than midieaval medicine. Something happened when we switched to modern medicine and it wasn't just that we aquired more knowledge.

Take the chemotherapy example for instance. It's very unlikely that in the future, we'll find it as backward as leeching. More likely, we'll figure out how to better synthesize chemicals without adverse side effects. It'll be like developing better aim rather than playing a whole new game (pardon the pathetic analogy).


Posted by: WillieStyle | Link to this comment | 10- 6-07 10:02 AM
horizontal rule
31

Discussing science with most people is like discussing economics with libertarians

Awesome.


Posted by: cerebrocrat | Link to this comment | 10- 6-07 10:03 AM
horizontal rule
32

We live in a fucking country where something like half the population is still not on board with evolution

1. Many Americans are not on board with evolution.
2. Many Americans are enthusiastic about the wonders of medical science.

I submit that many Americans are both 1 and 2.


Posted by: Invisible Adjunct | Link to this comment | 10- 6-07 10:06 AM
horizontal rule
33

"Chemotherapy", in non-technical parlance, refers specifically to cancer treatment rather than drug treatment generally.

Except that it's not really clear what the hell he's getting at with the chemotherapy comment. He refers to the "the idea" of it like "the idea" of chemotherapy will have been replaced or something.

Chemotherapy refers to a specific kind of chemical treatment, i.e., marinating the tumor in poision and hoping it kills the tumor before the person. I think I remember reading that new ideas in cancer therapy are more along the lines of personalized genetic therapy, turning cancer into a chronic illness instead of a terminal one.

Chemotherapy as a term is used to differentiate from radiation therapy.

There's a lot being done on more specific targeting of the drugs. Things like using B12 as a Trojan horse to get molecules into turmor cells, then releasing the drug by irradiating those cells with a wavelength of red light, etc.

And that's all still small molecule work. There's other things that are even more in their infancy like engineered T-cells.


Posted by: gswift | Link to this comment | 10- 6-07 10:08 AM
horizontal rule
34

"Chemotherapy" means chemical treatment. You think all chemical intervention will have disappeared?

That's some pretty annoying disingenuousness. Chemo for breast cancer triggered a psychotic episode in a friend of mine, so I have to say I believe some further refinements are probably possible.


Posted by: mcmc | Link to this comment | 10- 6-07 10:10 AM
horizontal rule
35

what WillieStyle said. What gets old are arguments along the lines of "scientists used to thing trash spawned rats, and look how wrong they were about THAT."


Posted by: gswift | Link to this comment | 10- 6-07 10:10 AM
horizontal rule
36

That's some pretty annoying disingenuousness. Chemo for breast cancer triggered a psychotic episode in a friend of mine, so I have to say I believe some further refinements are probably possible.

Christ on a cracker, I didn't say it wasn't going to get better. What I said was the idea of it isn't going anywhere.


Posted by: gswift | Link to this comment | 10- 6-07 10:12 AM
horizontal rule
37

I'm off to a soccer game you primitives.


Posted by: gswift | Link to this comment | 10- 6-07 10:12 AM
horizontal rule
38

But lobotomies and electroshock aren't long enough ago to throw in the pre-scientific method bin. So what explains them?

Doctors aren't necessarily scientists; they're technicians with some scientific training. A lot of clinical progress (or, changes in clinical practice) is made by doctors trying stuff to see if it works on an informal, empirical basis, and passing what works on to other doctors they train. There's much less of that then there was in even the fairly recent past (see, lobotomies), and of course some of it works great. But screwy stuff can show up because someone charismatic thinks like it sounds like a good idea, and teaches it without subjecting it to rigorous scientific analysis.

(I have heard firsthand bitching about this sort of thing still going on from people practicing cutting-edge medicine in prestigious locations. No written speculation about my sources, if people would be so kind.)


Posted by: LizardBreath | Link to this comment | 10- 6-07 10:13 AM
horizontal rule
39

LB you know Heathcliff Huxtable?!?

(whoops!)


Posted by: Sifu Tweety | Link to this comment | 10- 6-07 10:15 AM
horizontal rule
40

The sweaters are even spiffier in person.


Posted by: LizardBreath | Link to this comment | 10- 6-07 10:15 AM
horizontal rule
41

You realize, LB, that moves the doctors slightly closer to the 'let's bleed the fucker and see what happens' camp.


Posted by: Cala | Link to this comment | 10- 6-07 10:16 AM
horizontal rule
42

"scientists used to thing trash spawned rats

That argument would get old pretty fast, certainly.


Posted by: mcmc | Link to this comment | 10- 6-07 10:16 AM
horizontal rule
43

41 to 40.


Posted by: Sifu Tweety | Link to this comment | 10- 6-07 10:17 AM
horizontal rule
44

"Christ on a cracker, I didn't say it wasn't going to get better. What I said was the idea of it isn't going anywhere."

But who is saying nothing will improve? The problem is that many of the treatments aren't nearly as well understood as docs woudl have one believe.


Posted by: yoyo | Link to this comment | 10- 6-07 10:18 AM
horizontal rule
45

For people interested in the (un)scientific nature of psychiatry, I recommend this book:

http://books.google.com/books?id=4T8sKI4cx_wC&dq=&pg=PP1&ots=9Iz2Zy51D1&sig=9kfFvyTrvncB0YlyuRtrV5axNb8&prev=http://www.google.com/search%3Fhl%3Den%26q%3D%2522Mad%2BIn%2BAmerica%2522%2BWhitaker&sa=X&oi=print&ct=title#PPA20,M1

Damn, I really have to learn how to use html code to disguise links...anyone willing to give a tutorial?

Anyway, sciences are frequently in the position where systems are not well understood and so a lot a lot of theories are compatible with the observed phenomena, and experiments cannot easily distinguish between the theories. It's pretty obvious that we understand the body way better than they did in the 16th century, but still not very well at all. Especially in the case of the brain. Hard to see a big payoff in manufacturing a glass half full / glass half empty fight over that.

The more interesting thing to wonder about is whether there are systems that are so complex or self-recursive or ill-defined that science can never understand them. In the case of some disciplines that try to use the scientific method to examine culture, like economics, I think there's a decent argument that it will never work. Anyone want to make the same argument for the body or (more likely) the brain?


Posted by: marcus | Link to this comment | 10- 6-07 10:20 AM
horizontal rule
46

34: Chemo for breast cancer triggered a psychotic episode in a friend of mine,

This happened to someone I knew. I wonder if this is common.


Posted by: politicalfootball | Link to this comment | 10- 6-07 10:22 AM
horizontal rule
47

Treatment of atrial fibrillation is another very good example. If you read the minutes of afib conferences from even five years ago until now, you'll find people completely reversing themselves on the use of various techniques. It's amazing that they can cure some people, but there's no denying that they're doing a guess-and-test burn off of parts of the heart, and when they've done it to enough people, they'll have a pretty good idea of what works. Again, it's amazing, but also very crude. My point is just that a lot of medicine is like that.


Posted by: ogged | Link to this comment | 10- 6-07 10:23 AM
horizontal rule
48

My humours are disordered. It's like melancholy melancholy melancholy CHOLER!!! melancholy.


Posted by: mcmc | Link to this comment | 10- 6-07 10:23 AM
horizontal rule
49

Well, yeah. There's doubting science (SCIENCE!) and doubting medicine. Lots (most) of medicine is great and science-based and very well supported and will look sane even in retrospect, but clinical practice isn't, on a social level, something that operates with the same sorts of controls and practices that scientific research does; there are areas where the science isn't there yet, so practice is driven by ad hoc empiricism, and there are areas where actual practice diverges from what research would suggest as the most rational course of conduct for social reasons.

I'm not trying to put down doctors, just saying that your doctor is most likely not a research biologist -- she's more like an a auto mechanic with a lot of education in the relevant biology -- and there are some circumstances where that's going to lead to the application of 'clinical judgment' without a lot of research behind it.


Posted by: LizardBreath | Link to this comment | 10- 6-07 10:23 AM
horizontal rule
50

<a href="http://www.yourlink.com">Your linked text</a>


Posted by: apostropher | Link to this comment | 10- 6-07 10:24 AM
horizontal rule
51

55: <a href="PASTE THE LINK HERE">PASTE WHATEVER YOU WANT TO SHOW UP AS THE LINKED TEXT HERE</a>

Like that.


Posted by: LizardBreath | Link to this comment | 10- 6-07 10:26 AM
horizontal rule
52

she's more like an a auto mechanic with a lot of education in the relevant biology

Exactly (one doctor friend says "plumber," but the idea's the same).


Posted by: ogged | Link to this comment | 10- 6-07 10:26 AM
horizontal rule
53

My humours are disordered. It's like melancholy melancholy melancholy CHOLER!!! melancholy.


Posted by: ben w-lfs-n | Link to this comment | 10- 6-07 10:26 AM
horizontal rule
54

49 hits the nail on the head. Especially the "ad hoc empiricism" comment. One can also point to what seems to be massive geographic variation in practice norms with no real justification.


Posted by: marcus | Link to this comment | 10- 6-07 10:27 AM
horizontal rule
55

it probably doesn't help that medical school selects really strongly for people who are huge pricks.


Posted by: yoyo | Link to this comment | 10- 6-07 10:28 AM
horizontal rule
56

Your Mama


Posted by: John Emerson | Link to this comment | 10- 6-07 10:28 AM
horizontal rule
57

whereas most scientists i've met are pretty cool


Posted by: yoyo | Link to this comment | 10- 6-07 10:28 AM
horizontal rule
58

34, 46: A friend of my father's had a horrible reaction to chemo. She had breast cancer that had spread to her brain, and she and the doctor decided to stop treatment, the doctor commenting 'we won't know if this stuff works for another 50 years anyway.'

The tumors shrunk. She lived another ten years, with recurrent episodes of tumors popping up and then disappearing, before it got her. This doesn't tell me that chemotherapy isn't effective for most people, but it does tell me that we've got a lot more to learn.


Posted by: Cala | Link to this comment | 10- 6-07 10:29 AM
horizontal rule
59

Where I worked the PhD candidates taught basic science to the med students, for whom they felt an amused contempt. Some med students would take an extra year to get a PhD too, IIRC.

One way or another, med students tend to be very, very practical. OHSU got a fair proportion of hipster med students, though. I don't know if that's a national trend or not.


Posted by: John Emerson | Link to this comment | 10- 6-07 10:32 AM
horizontal rule
60

Lizardbreath, do you have any sourcing for your claim about the "ad hoc" nature of medical research? All applied science relies a great deal on empricism but, but "ad hoc" empricism doesn't sound at all like anything I've seen practiced in any lab anywhere.


Posted by: WillieStyle | Link to this comment | 10- 6-07 10:35 AM
horizontal rule
61

yeah, i've brought up some of the conflicting research on saftey of one of the drugs my psychitrist was thinking of prescribing me, and the conversation indicated that her basis for deciding how to prescibe stuff was mostly based on common practices in the profession, not on the research itself. Not that their aren't people looking at the research, but there is a tenuous connection between when evidence indicates something, and doctors start to use it.


Posted by: yoyo | Link to this comment | 10- 6-07 10:36 AM
horizontal rule
62

Ok a clear distinction has arisen in this thread between "scientists" and medical doctors. Not being a doctor, I can't speak to actual medical practice. So, given my ignorance, I'll shut up now.


Posted by: WillieStyle | Link to this comment | 10- 6-07 10:38 AM
horizontal rule
63

I'm not talking about labs, I'm talking about hospitals (as would be suggested by the use of the words 'clinical practice' and 'clinical judgment'). Treatment decisions aren't uniformly research driven (they are largely research driven -- I'm really not intending to imply that doctors are incompetent or irresponsible). Where research doesn't uniquely determine a set of best practices, and even sometimes when it should but isn't being followed, it's my understanding that treatment decisions are at least sometimes based on an ad hoc let's-throw-things-at-the-wall-and-see-what-sticks method, the results of which get informally passed around through training.

That's not evil or irresponsible -- there are some areas where the research isn't really there yet and practicing doctors have to do something. But it's not what we normally think of as science.


Posted by: LizardBreath | Link to this comment | 10- 6-07 10:40 AM
horizontal rule
64

63 crossed with 62.


Posted by: LizardBreath | Link to this comment | 10- 6-07 10:40 AM
horizontal rule
65

If you read the minutes of afib conferences from even five years ago until now, you will come to see how surfing the internet on a Saturday afternoon can constitute a dynamic lifestyle.

medical school selects really strongly for people who are huge pricks.

And he means "huge" literally. I always wonder how these poor, not-enough-time-to-sleep, med students and residents manage to spend so much time at the gym. 'Cause a whole lot of them look pretty good in their hospital blues.


Posted by: cerebrocrat | Link to this comment | 10- 6-07 10:43 AM
horizontal rule
66

What I'm arguing is that we are in fact doing better than "let's bleed this fucker and see what happens."

See, the thing is that if the leeches folk had been operating under this mindset they would indeed have been operating scientifically (if you also throw in adjusting behavior based on "what happens" over time and possibly also sharing such findings with others). Instead they were operating under a set of presumptions regarding bodily humors in which any results (good or bad) got fed into said assumptions. This does not prevent the accumulation of knowledge, it simply makes it more difficult. Chinese medicine often gets results, despite a rather convoluted understanding of human physiology, precisely because it has paid attention to what has worked. And, while it makes a lovely magic wand for allopathic medicine to dismiss the effectiveness of any other medical system, I don't think that it is all explained by the placebo effect.


Posted by: JPool | Link to this comment | 10- 6-07 10:52 AM
horizontal rule
67

The PhD candidates say that med students cheat a lot.


Posted by: John Emerson | Link to this comment | 10- 6-07 10:57 AM
horizontal rule
68

more like an auto mechanic

Physicians say this about surgeons.


Posted by: Jesus McQueen | Link to this comment | 10- 6-07 11:00 AM
horizontal rule
69

a clear distinction has arisen in this thread between "scientists" and medical doctors

As it should be. I always find it kind of... hard to work with that biological science and medicine overlap so much in people's minds. There's an obvious connection between the two, but really, they're different professions with different cultures practiced by different kinds of people, and what's true about one is rarely true about the other. And do not get me started on the subject of M.D.-Ph.D.'s.

I don't know (or to be honest, care) a whole lot about the philosophy of science, so this isn't an argument I'm advancing so much as a personal preference I'm expressing: I think of science as just a methodology for asking how things work that allows you to hack your own confirmation bias. With that understanding, a lot of the more embarrassing scientific greatest hits don't actually count as science - I don't think science loses points for humours and animal spirits and phlogiston, etc, because these ideas were not the product of this method, ie, "modern science." Hypothesis generation on its own isn't science.

A problem science still has, and always will, is that questions often ripen before there are any good tools available to ask them with, and there's a lot of floundering around in the mean time. Brain science is a great example of this; before electrophysiology and imaging, all you could really do was carve pieces of brain out and see what your subject couldn't do any more. I'd argue that areas of inquiry today that don't quite seem like science (economics, social psychology) are perfectly legitimate and important sciences that are still working the kinks out of their particular methodologies. Behavior is a phenomenon that can be studied like any other, once you've got a high enough signal:noise ratio in your data.


Posted by: cerebrocrat | Link to this comment | 10- 6-07 11:04 AM
horizontal rule
70

Physicians say this about surgeons.

Have you ever seen a video of major surgery? Given the tools they're using, "auto mechanic" sounds about right to me.


Posted by: Josh | Link to this comment | 10- 6-07 11:05 AM
horizontal rule
71

68: You don't what to know what surgeons say about physicians.


Posted by: LizardBreath | Link to this comment | 10- 6-07 11:05 AM
horizontal rule
72

what s/b want


Posted by: LizardBreath | Link to this comment | 10- 6-07 11:06 AM
horizontal rule
73

I'd argue that areas of inquiry today that don't quite seem like science (economics, social psychology) are perfectly legitimate and important sciences that are still working the kinks out of their particular methodologies. Behavior is a phenomenon that can be studied like any other, once you've got a high enough signal:noise ratio in your data.

Yeah. We've had a lot of discussion of 'is economics really science' around here in the past, and I think your characterization is perfectly accurate. There's a layperson's problem though, in that the hard sciences have absolutely incredible prestige, which is fully deserved, as a method of producing reliable new knowledge about the world, because they've developed very powerful, reliable methodologies, and theoretical frameworks that make it possible exploiting the data they have to a very great extent.

When you get into the softer sciences, like all the sciences of human behavior including economics, they aren't yet at a point in their development where they have the power of the hard sciences, but practitioners nonetheless seem to often expect to be granted the prestige of those sciences, and to have their results relied on to the same extent.


Posted by: LizardBreath | Link to this comment | 10- 6-07 11:15 AM
horizontal rule
74

This thread is really annoying. I just want to drive by to say that I hope I can contribute to evidence-based medicine, because every condition has so many treatments and doctors have to use guesswork most of the time to decide what to prescribe for a given patient, based on a minimal amount of population-level data or no comparative data at all. That's unacceptable.

The research I'm doing now is only clinically relevant to the notion of identifying vaguely-defined categories of a small subset of patients who are only susceptible to this particular disease in the first place because they have been (usually artificially) immunosuppressed. I wish I was doing something relevant to a larger number of patients.


Posted by: Cryptic Ned | Link to this comment | 10- 6-07 11:24 AM
horizontal rule
75

i definatly agree that there is some stuff that has weird evidence. check out these abstracts:

Arch Intern Med. 1999 Oct 25;159(19):2273-8.
A randomized, controlled trial of the effects of remote, intercessory prayer on outcomes in patients admitted to the coronary care unit.

Harris WS, Gowda M, Kolb JW, Strychacz CP, Vacek JL, Jones PG, Forker A, O'Keefe JH, McCallister BD.

Mid America Heart Institute, Saint Luke's Hospital, Kansas City, MO, USA.

CONTEXT: Intercessory prayer (praying for others) has been a common response to sickness for millennia, but it has received little scientific attention. The positive findings of a previous controlled trial of intercessory prayer have yet to be replicated. OBJECTIVE: To determine whether remote, intercessory prayer for hospitalized, cardiac patients will reduce overall adverse events and length of stay. DESIGN: Randomized, controlled, double-blind, prospective, parallel-group trial. SETTING: Private, university-associated hospital. PATIENTS: Nine hundred ninety consecutive patients who were newly admitted to the coronary care unit (CCU). INTERVENTION: At the time of admission, patients were randomized to receive remote, intercessory prayer (prayer group) or not (usual care group). The first names of patients in the prayer group were given to a team of outside intercessors who prayed for them daily for 4 weeks. Patients were unaware that they were being prayed for, and the intercessors did not know and never met the patients. MAIN OUTCOME MEASURES: The medical course from CCU admission to hospital discharge was summarized in a CCU course score derived from blinded, retrospective chart review. RESULTS: Compared with the usual care group (n = 524), the prayer group (n = 466) had lower mean +/- SEM weighted (6.35 +/- 0.26 vs 7.13 +/- 0.27; P=.04) and unweighted (2.7 +/- 0.1 vs 3.0 +/- 0.1; P=.04) CCU course scores. Lengths of CCU and hospital stays were not different. CONCLUSIONS: Remote, intercessory prayer was associated with lower CCU course scores. This result suggests that prayer may be an effective adjunct to standard medical care.

and it works even if its done afterward:

BMJ. 2001 Dec 22-29;323(7327):1450-1.
Effects of remote, retroactive intercessory prayer on outcomes in patients with bloodstream infection: randomised controlled trial.

Leibovici L.

Department of Medicine, Beilinson Campus, Rabin Medical Center, Petah-Tiqva 49100, Israel. leibovic@post.tau.ac.il

OBJECTIVE: To determine whether remote, retroactive intercessory prayer, said for a group of patients with a bloodstream infection, has an effect on outcomes. DESIGN: Double blind, parallel group, randomised controlled trial of a retroactive intervention. SETTING: University hospital. SUBJECTS: All 3393 adult patients whose bloodstream infection was detected at the hospital in 1990-6. INTERVENTION: In July 2000 patients were randomised to a control group and an intervention group. A remote, retroactive intercessory prayer was said for the well being and full recovery of the intervention group. MAIN OUTCOME MEASURES: Mortality in hospital, length of stay in hospital, and duration of fever. RESULTS: Mortality was 28.1% (475/1691) in the intervention group and 30.2% (514/1702) in the control group (P for difference=0.4). Length of stay in hospital and duration of fever were significantly shorter in the intervention group than in the control group (P=0.01 and P=0.04, respectively). CONCLUSION: Remote, retroactive intercessory prayer said for a group is associated with a shorter stay in hospital and shorter duration of fever in patients with a bloodstream infection and should be considered for use in clinical practice.

There is similar evidence for homeopathy too

its such a huge wtf


Posted by: yoyo | Link to this comment | 10- 6-07 11:26 AM
horizontal rule
76

The intercessory prayer stuff was debunked. Science at work, baby.


Posted by: ogged | Link to this comment | 10- 6-07 11:33 AM
horizontal rule
77

one study doesn't 'debunk' anything.


Posted by: yoyo | Link to this comment | 10- 6-07 11:38 AM
horizontal rule
78

Sure, but neither does one study establish anything. For a WTF!?!?!!! result that would require all sorts of exciting new theory to explain it, like 'intercessionary prayer works', you'd really want to see it replicated a bunch before hanging anything on it.


Posted by: LizardBreath | Link to this comment | 10- 6-07 11:40 AM
horizontal rule
79

I love Bob's hypothesis that prayer works better in areas where there's less of it going on.


Posted by: ogged | Link to this comment | 10- 6-07 11:44 AM
horizontal rule
80

That is a truly great theory. Whatever happened to Bob, anyway? And why am I in the office on a weekend again? Who are all you people, anyway?


Posted by: LizardBreath | Link to this comment | 10- 6-07 11:50 AM
horizontal rule
81

Bob went off and had a baby (not all by himself) and might have started a non-profit.

I can't answer your other questions.


Posted by: ogged | Link to this comment | 10- 6-07 11:53 AM
horizontal rule
82

Who are all you people, anyway?

I be the first even Asian astronaut.


Posted by: Cryptic Ned | Link to this comment | 10- 6-07 11:53 AM
horizontal rule
83

75: I think there've been some recent studies (in late '06 or early '07) that actually found that patients who were being prayed for did worse than the controls, which suggests that prayer has no real effect one way or another.

About lobotomies and similarly dubious 'therapies'--I think two big reasons why those caught on the way they did was that they happened in an era when a)Clinical testing and evidence-based medicine played a much smaller role than they do now and b) Bioethical standards were a lot looser than at present (the modern IRB system dates from either the late 1960s or the early 1970s). I don't think that our current research system is perfect, and the uncertainty over atrial fibrillation therapies is a really good example of that, but things are a lot better than they used to be.


Posted by: the Other Paul | Link to this comment | 10- 6-07 11:54 AM
horizontal rule
84

One failure to replicate doesn't debunk anything, but one study can undermine confidence in a methodology, and thereby weaken the collective significance of all the studies that rely on it. And like LB says, if you're arguing for a result that violates the relevant theoretical structure, you need a lot of good evidence. Unfortunately, I can't check what the case is here because fucking PubMed is crapping its pants, AGAIN. Any other frequent PubMed users been having trouble this week?


Posted by: cerebrocrat | Link to this comment | 10- 6-07 11:55 AM
horizontal rule
85

hey LB: what is document review, and is easily learnable on the job? specifically privilege review on summation legal for what i think is an accounting firm


Posted by: yoyo | Link to this comment | 10- 6-07 11:55 AM
horizontal rule
86

Article about the prayer study here: http://www.nytimes.com/2006/03/31/health/31pray.html?_r=1&oref=slogin


Posted by: the Other Paul | Link to this comment | 10- 6-07 11:56 AM
horizontal rule
87

In any case, public health education continues to advance by leaps and bounds.


Posted by: LizardBreath | Link to this comment | 10- 6-07 11:57 AM
horizontal rule
88

I don't think that our current research system is perfect, and the uncertainty over atrial fibrillation therapies is a really good example of that, but things are a lot better than they used to be.

I think so too.

It's hard for me to imagine that all the consent forms and privacy regulations are really necessary, because I can't imagine someone not caring one whit about their research subjects and patients, but I guess they are.


Posted by: Cryptic Ned | Link to this comment | 10- 6-07 11:58 AM
horizontal rule
89

I be the first even Asian astronaut.

As long as we're asking questions, am I the only one who frequently reads things here that are funny even though I don't get the reference?


Posted by: cerebrocrat | Link to this comment | 10- 6-07 11:58 AM
horizontal rule
90

i think the most recent meta-review don't show any difference, but that wasn't the case recently. i'm not arguing for IP or homeophaty, just that research is messy and changes and tha probably contributes to the way the practice of medicine isn't all that scientific


Posted by: yoyo | Link to this comment | 10- 6-07 11:59 AM
horizontal rule
91

89: Dammit, it should be "first ever", not "first even".


Posted by: Cryptic Ned | Link to this comment | 10- 6-07 11:59 AM
horizontal rule
92

85: It's totally monkey work -- you can learn what to do in fifteen minutes.

A document is privileged if it either (a) contains a communication between a lawyer and her client that wasn't made public to a third party, or (b) contains a lawyer's work in preparation for litigation. What this comes down to is that if there's a lawyer's name on a piece of paper, either as the writer or the recipient, and no evidence that it was sent out to a third party, there's a good shot it's privileged.

If you're being hired to do this (you're not a lawyer, right?), you're doing a first cut. You'll be looking through a whole bunch of documents for the appearance of the names of the lawyers involved in the case, or any other indication that there's legal advice in the document (like the words "as your attorney, Bob, you should do X" even if you don't know the name of the person writing it.) You'll flag the relevant documents, and then a lawyer will go through the documents you flag to figure out which ones are really privileged -- your job is to be over rather than under inclusive.

Mindnumbingly dull, but not difficult.


Posted by: LizardBreath | Link to this comment | 10- 6-07 12:04 PM
horizontal rule
93

OT: Did any of you NYers show up at home last night with my jacket or my friend's sweater?

Also, I promised to send someone a song or set of songs, and now I forget who or what it was. Email me if you have any idea what I'm talking about.


Posted by: A White Bear | Link to this comment | 10- 6-07 12:07 PM
horizontal rule
94

For a wonderful, fleeting moment, I believed the link in 87.


Posted by: Penny | Link to this comment | 10- 6-07 12:17 PM
horizontal rule
95

quick detour: I dare anyone to look at this and not laugh.


Posted by: Cryptic Ned | Link to this comment | 10- 6-07 12:17 PM
horizontal rule
96

85: It's totally monkey work -- you can learn what to do in fifteen minutes.

If you're being hired to do this (you're not a lawyer, right?), you're doing a first cut

Right, it's common for lawyers doing it to be asked to classify it much more finely, but they can't ask you to do that.


Posted by: I don't pay | Link to this comment | 10- 6-07 12:19 PM
horizontal rule
97

Mindnumbingly dull, but not difficult.

And yet, better than a lot of other paralegal work. At least when I was doing document reviews I could sit in an office and listen to music and be left alone.

One of my friends works for a company that does automated review (i.e., you give them a bunch of documents in electronic format, they tell you what's in them). My reaction when he told me he was considering taking the job was that a) it sounded like an awesome idea but b) I couldn't see the legal profession ever actually trusting something like that.


Posted by: Josh | Link to this comment | 10- 6-07 12:21 PM
horizontal rule
98

They can't mean automated in the sense of non-human -- they must just mean 'outsourced to the service's lawyers and paralegals'. At which point it's not all that untrustworthy, just depends on the quality of the service. But I don't believe anyone would rely on something really automated.

Document review on paper, IME, while mindless, is often a good time -- it tends to be a conference room full of junior lawyers and paralegals shooting the shit as they work. Doing it electronically is more along the lines of zoning out as you listen to music.


Posted by: LizardBreath | Link to this comment | 10- 6-07 12:25 PM
horizontal rule
99

They can't mean automated in the sense of non-human -- they must just mean 'outsourced to the service's lawyers and paralegals'. At which point it's not all that untrustworthy, just depends on the quality of the service. But I don't believe anyone would rely on something really automated.

No, I mean fully automated. And your reaction is exactly what I expected.


Posted by: Josh | Link to this comment | 10- 6-07 12:31 PM
horizontal rule
100

Well, for one thing it would have to be limited to documents in an electronic format -- scanned paper documents have handwritten notes on them, often, that OCR software isn't up to handling. For another, it totally ignores the problem that the list of relevant lawyers you go in with is always incomplete. I've never done a privilege review where there weren't a couple of phone calls to the client as follows: "Hey, Richard Roe's name keeps on showing up in a context where he seems to be dealing with legal stuff, but he's not on the list of lawyers you gave me. Who is he, and is he an attorney?"

The idea of a really automated review is just goofy.


Posted by: LizardBreath | Link to this comment | 10- 6-07 12:36 PM
horizontal rule
101

So you OCR the documents, index them by the names that appear in them along with instance of the words "lawyer" and "legal advice", and flag them for manual if they appear to contain anything other than machine readable text. Doesn't seem wildly useless.


Posted by: Sifu Tweety | Link to this comment | 10- 6-07 12:38 PM
horizontal rule
102

But "name X appears a lot in legal-ish contexts, maybe we should check if X is an attorney" is the sort of thing you could program in.


Posted by: Unfoggetarian: "Pause endlessly, then go in" (9) | Link to this comment | 10- 6-07 12:39 PM
horizontal rule
103

Handwriting doesn't work for OCR, and "legalish contexts" sounds like a really hard AI problem to reliably detect. While lawyers should always do this sort of thing (law firm emails often generate an automatic notice along these lines), they don't say: "as your lawyer, this is privileged legal advice" every time they give advice. You pick it up by reading for context, not key words.

I'm not saying it's theoretically impossible; presumably handwriting OCR even of scrawled marginalia will at some point approach human abilities, and 'legalish contexts' is conceivably automatable with enough keywords. But I've never heard of an actually existing system I'd trust with either.


Posted by: LizardBreath | Link to this comment | 10- 6-07 12:44 PM
horizontal rule
104

The idea, I'm assuming, would be too do a first cut of the first cut; if a document is fully machine-readable, has no lawyer's names on it, and doesn't contain any terms broadly relating to legal issues, you can exclude it from the human first pass. Not wildly useless, no?


Posted by: Sifu Tweety | Link to this comment | 10- 6-07 12:48 PM
horizontal rule
105

But I've never heard of an actually existing system I'd trust with either.

Yes, but this miserable present is but a necessary stage on the path to the One True System. Do not waver, comrade!


Posted by: Josh | Link to this comment | 10- 6-07 12:49 PM
horizontal rule
106

One of my friends is a linguist at a company that does automated legal document searching of some sort, and they seem to get plenty of business. So they must be able to do something well enough to be hired. I'm not sure if this is precisely the stuff they're working on, but they have a lot of computer scientists and linguists, so they at least have a chance at solving at least somewhat hard AI problems.


Posted by: Unfoggetarian: "Pause endlessly, then go in" (9) | Link to this comment | 10- 6-07 12:51 PM
horizontal rule
107

104: People do that now -- in a review, documents will be machine-sorted into folders with search-term hits and those without, and it does speed things up. You can motor through the no hit documents at high speed. But you still need human eyes on the documents with no hits, even if just for a quick glance, because there's no such thing as an exhaustive list of terms broadly relating to legal issues. I could stand here and give you legal advice for an hour without using a technical term.

Computers help, but there's enough necessary human work that calling the process 'automated' is either overstating the degree to which humans are still doing it, or using the service would be malpractice.


Posted by: LizardBreath | Link to this comment | 10- 6-07 12:53 PM
horizontal rule
108

1) I actually would love to know what surgeons say about physicians.

2) I know two MD-PhD(candidates). They're both really smart, and really nice. But they came to grad school after med school and seem to have forgotten... um... rather a lot.


Posted by: TJ | Link to this comment | 10- 6-07 12:55 PM
horizontal rule
109

108.1: That they aren't precisely brain surgeons, or rocket scientists, IYKWIM. (I am merely repeating, not endorsing, these slanders.)


Posted by: LizardBreath | Link to this comment | 10- 6-07 1:02 PM
horizontal rule
110

oh, ok.

Yeah, i didn't take the bar exam this summer. I just got a call about this and wasn't quite suire what it might involve. I probably won't do this actual assignment, since its 3 hours away, but i'm guessing i'll get similar ones in the future.


Posted by: yoyo | Link to this comment | 10- 6-07 1:24 PM
horizontal rule
111

Should have guessed.


Posted by: TJ | Link to this comment | 10- 6-07 1:24 PM
horizontal rule
112

Document review teams of lawyers are often, but not always comprised of recent grads. The project I just came off of had a crew of surprising depth and maturity; there wasn't a recent grad in the bunch. We must have been selected on that basis, and in fact the categories and different projects and transactions they wanted us to keep track of and classify were quite complex. It just varies.

The "tentatively privilaged" classification was clearly automatic, on the names of the lawyers. And as LB says, the names list kept growing longer by the hour. And we were supposed to eliminate the non-responsive, which buries things much deeper than "privilaged," which may be challenged, does. If something could be both, it was better for it to be non-responsive.


Posted by: I don't pay | Link to this comment | 10- 6-07 1:25 PM
horizontal rule
113

I'd argue that areas of inquiry today that don't quite seem like science (economics, social psychology) are perfectly legitimate and important sciences that are still working the kinks out of their particular methodologies. Behavior is a phenomenon that can be studied like any other, once you've got a high enough signal:noise ratio in your data.

I don't buy this. Not that you can't learn anything from studying society, of course you can learn a whole lot. And you can learn a lot using quasi-scientific methods. But it isn't and will never be "science", because the phenomena you're studying never stay still long enough to get replicable scientific results. There is no important human social behavior that is separable from the larger cultural context in which it occurs, and that cultural context is always changing. Part of the description of a social behavior is the full cultural meaning of that behavior in the minds of all the participants, and that description is incredibly hard to get in any reliable and replicable way.

There are of course plenty of cross-cultural commonalities and consistencies in human nature, but not as many as you'd need. Contrast the ability to isolate a single phenomena, like motion or gravity, in physics. Biology differs from physics in part because the body is a massive, complex, interdependent system like a society. And I think most biologists would say that we are a very, very long way from understanding the full complexity of the interdependencies in the body. Yet the body is a much simpler system than society, and of course unlike society it is not a moving target, since it is consistent across time and people.


Posted by: marcus | Link to this comment | 10- 6-07 3:21 PM
horizontal rule
114

112:Think of how many additional categories my spelling alone creates!


Posted by: I don't pay | Link to this comment | 10- 6-07 3:22 PM
horizontal rule
115

113: and you can't do intervention studies on lots of things, or if you can, its unethical


Posted by: yoyo | Link to this comment | 10- 6-07 3:50 PM
horizontal rule
116

114: I wasn't going to say anything. I, myself, find that an extraneous 'd' tends to creep into the word in question.


Posted by: LizardBreath | Link to this comment | 10- 6-07 3:52 PM
horizontal rule
117

113: Boy. Well, I don't know what you mean by "quasi-scientific methods," but by your criteria, much of biology and all of ethology aren't science. Also, climatology. You don't need an object of study that doesn't vary over time to develop a science around it, and you don't need completely isolable variables. You need to be able to collect quantifiable data, make predictions, and test them. That's pretty much it.

Just because a body of data is large, noisy, and multivariate, doesn't mean you can't meaningfully apply scientific method to it. Embryology is much like you're describing human social behavior - there's a complicated and dynamic interaction between genotype, prenatal environment, and the parent's ecosystem, including behavior of conspecifics in social species. Yet, most people consider embryology a science.

And this: the body is a much simpler system than society, and of course unlike society it is not a moving target, since it is consistent across time and people is a statement that needs reconsidering. The body is consistent across time and people? Wha?


Posted by: cerebrocrat | Link to this comment | 10- 6-07 4:18 PM
horizontal rule
118

I like the part in Star Trek IV where Bones, back from the future, sees a present-day woman on a dialysis machine. Shocked at the archaic (to him) treatment, he gives her a pill instead. Later, he sees her again and she's feeling just fine.


Posted by: Gaijin Biker | Link to this comment | 10- 6-07 7:08 PM
horizontal rule
119

84: PubMed has been working splendidly for me all week and is working fine now. I believe I have experienced periodic sluggishness throughout the past few months, however.


Posted by: Otto von Bisquick | Link to this comment | 10- 6-07 7:41 PM
horizontal rule
120

117: "wha" yourself. The same knowledge you use to treat sick bodies today could have been used in Ancient Greece to do the very same thing. Hearts do the same thing in roughly the same way in everybody, even if they vary in effectiveness and so forth.

I already analogized biology to social sciences in my first post...you could do the same with climatology. But what biology and climatology have over social sciences is that they are rooted in natural laws of physics and chemistry, and these laws are invariant, can be understood and mathematicized with almost complete precision, and then used to construct models. Societies are not governed by natural laws, but by the understanding and interactions of intelligent, self-conscious agents. These are constrained by the brute facts of existence, but they don't follow laws the way matter follows the laws of physics.

Just because a body of data is large, noisy, and multivariate, doesn't mean you can't meaningfully apply scientific method to it.

Societies are not bodies of data. Human behavior generates tons of data, and you can perfectly well make up stories about society and then look for correlations in the data that are consistent with your story. An interesting thing to do, but it doesn't start with replicable, consistent laws that have explanatory power and there's no particular reason to think it will get you to them. As an occupation, it has some vague similarities with science but it's probably closer to doing history or journalism than running the lab experiments or building the natural-law grounded models that are the bread and butter of science. (As Yoyo points out lab experiments are very difficult to run in the social sciences...and where they can be run, the way that social meanings of actions vary across environments means that they will have quite limited applicability outside of the lab).


Posted by: marcus | Link to this comment | 10- 6-07 10:33 PM
horizontal rule
121

And I'm not saying the social sciences aren't useful. Take evaluation research as just one example. Done right, it can give good information about the effects of some particular intervention. That's very useful, but it doesn't necessarily tell you much about the effects of the very same intervention ten years later, or in a different region.

Some of the very same issues come up for medical interventions based in biology, but they are much more limited and more possible to handle (e.g. by increasing sample size and looking for the effects in subpopulations).


Posted by: marcus | Link to this comment | 10- 6-07 10:39 PM
horizontal rule
122

I don't know (or to be honest, care) a whole lot about the philosophy of science

See, the problem is, scientists often say this sort of thing and then go on to express opinions about science that would be a hell of a let more interesting and better informed if they did know some philosophy of science.

There's a vast difference between doing science on the one hand, and having informed opinions about science and the latter 'meta' discussions are ones that research scientists often think themselves qualified to have (simply by virtue of being scientists). Crudely, and somewhat snarkily, this is roughly akin with your man in the street thinking he has a lot of interesting things to say about anthropology simply by virtue of being human.

Your stereotypical scientist of this type will dismiss the philosophy of science and then come out with their own 'philosophy' of science, which 9 times out of 10 turns out to be some crude Popperianism that represents roughly the state of the philosophical art from about 1934.

Of course, a lot of research scientists have really interesting and intelligent things to say about science and may have things to say that philosophers of science really ought to listen to, but those research scientists tend, mostly, to be the ones who've actually read a bit a of philosophy of science, and have a sense of the dialogue they are entering into, rather than just dismissing it.


Posted by: nattarGcM ttaM | Link to this comment | 10- 7-07 2:56 AM
horizontal rule
123

Marcus, my "wha" was in response to what seemed, without further explanation, like a remarkable statement; bodies of course change a lot over the time course of an individual's life and over evolutionary time. Now I see what you meant, but it doesn't really help: the same knowledge you use to predict human behavior today could have been used in Ancient Greece to do the very same thing.

Your argument sounds like the teleological case against evolution: human behavior is too complicated to study, so there must be no natural laws governing it. The underlying assumption of behavioral science (as any) is that behavior, like everything else under the sun, is rooted in invariant natural laws; identifying those laws is kind of the point. When you say that societies are not governed by natural laws, you're also making an assumption, and I can't see what justifies it.

Societies, not to mention individual people, are certainly complicated things, which is in part why economics and social psychology are still not very mature sciences. But you can still collect data, make predictions, control variables, and test your predictions. That's pretty much all you need to be in the science business. Erlenmeyer flasks are not required (although a lot of social psych experiments are done in labs). If you like math with your science, I suggest network theory as a complement to your social sciences, or information theory if your taste runs to something a little drier.


Posted by: cerebrocrat | Link to this comment | 10- 7-07 3:10 AM
horizontal rule
124

122: huh, I don't think I dismissed philosophy of science (didn't mean to anyway) so much as indicate that I wasn't that interested in it, by which I mean, of the many things that compete for my attention, philosophy of science isn't high on my list. I'm certainly receptive to being schooled if there's something you think I should know.


Posted by: cerebrocrat | Link to this comment | 10- 7-07 3:18 AM
horizontal rule
125

re: 124

Perhaps I misread you. I have no problem with people not being interested in the philosophy of science [as you say, competing interests, etc.] and I'm fairly sure that it's not required to be a good and thoughtful scientific practitioner.

I do have issues with people with strong 'meta' opinions on science who simultaneously dismiss the work of the people who spend their entire lives thinking about precisely those 'meta' questions. Happy to accept that you're not espousing that viewpoint, though.


Posted by: nattarGcM ttaM | Link to this comment | 10- 7-07 4:21 AM
horizontal rule
126

Yeah, definitely not. My "don't know/don't care" qualifier was precisely because I was expressing a meta opinion that I knew was not particularly informed.


Posted by: cerebrocrat | Link to this comment | 10- 7-07 9:16 AM
horizontal rule
127

OK, really was gone all day, and haven't read the whole thread, because I can't get past gswift's thoughtful rejoinder at 22.

Because someone who's done grad-level work in evolutionary biology clearly should be lumped with superstitious primitives who think Darwin was the Devil. And by "chemotherapy" I obviously meant, like ibuprofen.

I don't know where I got the idea that scientists can be arrogant pricks who can't acknowledge any shortcoming. What was I thinking there?

Christ.


Posted by: JRoth | Link to this comment | 10- 7-07 9:54 AM
horizontal rule
128

Speaking of appendices, I love Max's bracketed comments after his name.

[these. I think they're hilarious.]


Posted by: heebie-geebie | Link to this comment | 10- 7-07 10:07 AM
horizontal rule
129

Appendectomy scars are sexy.


Posted by: apostropher | Link to this comment | 10- 7-07 10:20 AM
horizontal rule
130

I don't know where I got the idea that scientists can be arrogant pricks who can't acknowledge any shortcoming. What was I thinking there?

SILENCE! The Primitives will speak only when queried. Some of us have thinking to do.

Christ.

Now that's more like it. Yes?


Posted by: cerebrocrat | Link to this comment | 10- 7-07 10:23 AM
horizontal rule
131

Appendectomy scars are sexy; performing an appendectomy is sexier.


Posted by: Jesus McQueen | Link to this comment | 10- 7-07 10:24 AM
horizontal rule
132

the same knowledge you use to predict human behavior today could have been used in Ancient Greece to do the very same thing.

I don't really believe this, except in the crudest and roughest sense of "predict" (they have families! there is trade and exchange! They fight wars!). And of course, with the proviso that part of our knowledge of human behavior today is the historical knowledge of what happened in ancient Greece.

you can still collect data, make predictions, control variables, and test your predictions.

any ability to control variables in the social sciences is very, very limited. I mean, under certain circumstances you occasionally can (though less than some people think...often what exactly is being held constant is not well understood at all). But the external validity of what you're doing is extremely doubtful. And if you believe regression analysis truly constitutes "controlling variables" in anything like the experimental sense, then there's a bridge I'd like to sell you.

Bottom line, I'm very familiar with the social sciences, and that's the source of my skepticism. I understand quite well that the social sciences try hard to use the scientific method. I just don't think the phenomena under study are very amenable to the method. Not just because it's complex, because of the way in which it's complex.

On the other hand, the very act of gathering lots of data about people and society is a very useful contribution of the social sciences. Lots of data is better than little or no data.


Posted by: marcus | Link to this comment | 10- 7-07 11:03 AM
horizontal rule
133

cerebrocrat has gotten it exactly right twice now.

On a less divisive note, thanks to this posting, I believe that I was part of history last night. My daughter awoke asking what an appendix is (it's in a Madeleine book), and I was actually able to tell her - among the first parents in human history with a correct answer (we think)!


Posted by: JRoth | Link to this comment | 10- 7-07 11:24 AM
horizontal rule
134

133: And she was all like, "Daddy? This section of maps and graphs at the end of my book is a repository for bacteria? Yucky!"


Posted by: A White Bear | Link to this comment | 10- 7-07 11:29 AM
horizontal rule
135

Okay, so I guess my bottom line would be that in my eyes, if you're using scientific method, you're doing science. I think of things like economics as being something like electrophysiology - before you could ask a neuron or groups of neurons what they were doing, you had to have a way to reach them (microelectrodes), a way to amplify their signals (amplification), tell one signal from the other (isolation), filter out noise (filtering algorithms), and mathmatical tools adapted for the job of making sense of what you found (stochastics, information theory). All of those technological steps had to be in place before the answers could start to get satisfying. In this view, the economists are still working out their amplifier circuits.

As for the ancient Greece thing, I think we're talking apples and oranges. If you mean predicting what a society will be like from one time point to another, then yes, chance events and an impoverished data set make that practically impossible. Yet historians still do identify cyclic patterns in societies over time, even with their comparatively poor data set. I was thinking more of predicting behavior Khanman & Tverksy-style; given some information, people's decisions are predictable to a certain precision. Probably we'd have to figure out specifically what behavior it is we want to predict to get any further with this.

To understand your position, I guess I'd need some elaboration on this: Not just because it's complex, because of the way in which it's complex. Where we part ways seems to be on the issue of whether the phenomena under study are qualitatively different in a way that makes them resistant to scientific inquiry, rather than just quantitatively different enough to make it really hard.

Here's a question: would you consider ethology a bona fide science, or one of the kind-of-but-not-really variety?


Posted by: cerebrocrat | Link to this comment | 10- 7-07 11:54 AM
horizontal rule
136

133: I'm afraid to know which two...

I'm dying to blather lengthily on the subject of the arrogance of scientists but good god, have I got to get back to work. Speaking of which, now see, god dammit, with all this arguing, now I AM kind of interested in philosophy of science, which I totally don't have time for. But nattarG, if you were to submit a reading list, it'd be welcome. You bastard.


Posted by: cerebrocrat | Link to this comment | 10- 7-07 12:01 PM
horizontal rule
137

re: 136

I'll do that tomorrow. There's a couple of introductory text books and some collections of readings that would take you a long way without huge amounts of work.


Posted by: nattarGcM ttaM | Link to this comment | 10- 7-07 1:09 PM
horizontal rule
138

137: Could you please post it here or send it to me as well? My first priority right now is making my experiments work, but it would be nice to have such a list for the long term.


Posted by: Otto von Bisquick | Link to this comment | 10- 7-07 1:33 PM
horizontal rule
139

re: 138

Will do.


Posted by: nattarGcM ttaM | Link to this comment | 10- 7-07 1:35 PM
horizontal rule
140

If anyone here knows SPSS very well and wants to contribute to a truly useless social scientific endeavor that the (better $*&@ing second) author of this paper believes in not at all, and he/she would like to email me this afternoon, I'm not sure if tia@unfogged still works (I think so), but if not no/tickling@gmail.com would do, and she would be endebted to you not forever, but a good long time. Only true love lasts forever.


Posted by: Tia | Link to this comment | 10- 7-07 2:53 PM
horizontal rule
141

Damn, I'm decades out of date on SPSS. But there must be someone around.


Posted by: LizardBreath | Link to this comment | 10- 7-07 2:55 PM
horizontal rule
142

mrh is already on the case, although if anyone else shows up he doesn't consider himself totally expert, and I might wind up having other questions too. you know, when I signed on to this they told me I was going to do a lit review and write the introduction. Contemplating the fact that I (I!) am basically the data manager and running all the analyses on this paper is like touring a sausage factory.


Posted by: Tia | Link to this comment | 10- 7-07 3:04 PM
horizontal rule
143

I don't really believe this, except in the crudest and roughest sense of "predict" (they have families! there is trade and exchange! They fight wars!).

On the other hand, a person can read the extant plays of Euripides and get the sense that damn little about how people feel and act has changed in 25 centuries.


Posted by: CharleyCarp | Link to this comment | 10- 7-07 3:17 PM
horizontal rule
144

Heh. Have I told the story of my only academic publication ever? I was in high school, and got hired as a RA/programming monkey by an education researcher who was herself entirely uncomfortable with mathematics of any sort. I had basically no relevant skills, and am horrified that the statistical results I came up with were published in any sort of respectable journal. But I did get to be second co-author (anyone who knows my name, google FN SecondhalfofmycurrentLN IQ to admire the dullness).

But the experience of it put me right off the social sciences generally. If that sort of thing could happen (publication of research done in important part by an untrained and incompetent 15-year-old, working with a researcher without the skills to check what was being done on her behalf) without repurcussions, I wanted no part of it.


Posted by: LizardBreath | Link to this comment | 10- 7-07 3:17 PM
horizontal rule
145

But it's on gifted kids. They had to use a gifted kid to make it work.


Posted by: A White Bear | Link to this comment | 10- 7-07 3:30 PM
horizontal rule
146

Two of us, in fact, and the prof. was an alumna of the same school. Doesn't change the fact that I lost half the Terman data set somehow, and the stats are run on a comparison of our data to the remaining half of the Terman data.


Posted by: LizardBreath | Link to this comment | 10- 7-07 3:37 PM
horizontal rule
147

Ha!


Posted by: A White Bear | Link to this comment | 10- 7-07 3:38 PM
horizontal rule
148

Yet historians still do identify cyclic patterns in societies over time, even with their comparatively poor data set.

Not the good ones.


Posted by: eb | Link to this comment | 10- 7-07 3:52 PM
horizontal rule
149

Yeah, a strong takeaway from that whole experience was that people who work with 'gifted children' are wildly overimpressed by 'gifted children'. Clever a child as I may have been, hiring a teenager with no relevant academic knowledge to do the math for your research? Not responsible.


Posted by: LizardBreath | Link to this comment | 10- 7-07 3:52 PM
horizontal rule
150

Yeah, it's like the way my mom thought about me when I was 13. "If my appendix burst, I'd want you to take it out! You're smarter than any doctor!" That, my friends, is a direct quote. My mom is pretty smart in the areas of a lot of things, but she assumes there's no such thing as education or training.


Posted by: A White Bear | Link to this comment | 10- 7-07 3:58 PM
horizontal rule
151

So, Tia's research is comparatively infinitely respectable and un-sausagelike.


Posted by: LizardBreath | Link to this comment | 10- 7-07 4:17 PM
horizontal rule
152

Yet historians still do identify cyclic patterns in societies over time, even with their comparatively poor data set.

I once spent quite awhile trying to make sense of the cycles in Eurasian history reported by Andre Gunder Frank, and as far as I could tell they were an garbled mess of arbitrary quantities.


Posted by: John Emerson | Link to this comment | 10- 7-07 4:36 PM
horizontal rule
153

Next thing you people will be telling me that Spengler was wrong.


Posted by: LizardBreath | Link to this comment | 10- 7-07 4:36 PM
horizontal rule
154

We know how you feel about Spengler, LB, so no. I would instead say that, in certain respects, Spengler may not be entirely fucked up at all.


Posted by: John Emerson | Link to this comment | 10- 7-07 4:38 PM
horizontal rule
155

Okay, I see the history point. Maybe historians identifying constants over time works better? Or maybe I should quit while I'm ahead?


Posted by: cerebrocrat | Link to this comment | 10- 7-07 4:44 PM
horizontal rule
156

135: thanks for the interesting reply.

On Greece and cyclic patterns in history: my question would be, does a "social scientific" approach *add* to our ability to understand society as compared to a literary/observational/journalistic approach? My general answer: except for formalizing the gathering of mass statistics, not really. If I was a ruler in ancient Greece, I would learn a lot more reading Thucydides than I would by reading a modern intro econ textbook. On the other hand, a suitably translated modern medical text would be much better than Hippocrates, in fact so much better that it would revolutionize life there overnight.

Where we part ways seems to be on the issue of whether the phenomena under study are qualitatively different in a way that makes them resistant to scientific inquiry

yes, qualitatively different, although one can trace a path of increasing complexity from physics to biology to ethology to human society. But the qualitative difference is that human beings are self-conscious agents. One doesn't understand the behavior simply by observing it, one must understand what the behavior *means* to the agents. This isn't true for an atom or a cell or a neuron. And meanings are always changing as cultures shift. Economics is probably better understood as an ideology that corresponds to a particular shift in the social meaning of markets than as a set of scientific laws about how people respond to money.

I would add too that another issue with people being self-conscious is that any scientific theories about society that gain wide acceptance will be incorporated into the society and therefore shift the nature of that society.

Ethology is getting up there toward some of the problems that the social sciences have, but I'd be much more optimistic about its ability to be a very precise predictive science. Good to bring that up as an example, because the question of how successful social sciences can ever be in prediction is closely related to just how "animal" we are. Gary Becker wrote some stuff in the 70s about the need to root economics in sociobiology. Thus creating the perfect bogeyman discipline for Unfogged commenting.


Posted by: marcus | Link to this comment | 10- 7-07 5:29 PM
horizontal rule
157

Okay, so now I can see the root of our disagreement, and why it's not likely resolvable.

One doesn't understand the behavior simply by observing it, one must understand what the behavior *means* to the agents.

I'd say that the most profound cumulative lesson of human behavioral and cognitive science is that our undertanding of what motivates us does not correspond closely to the actual determinants of our behavior. Self-consciousness and its complement, theory of (others') mind, are significant mainly as sources of noise*. Similarly, to this

the question of how successful social sciences can ever be in prediction is closely related to just how "animal" we are

I'd have to say I see no sense in which we are anything but animal. I asked about ethology because, again, I see it as different in degree but not kind from the human behavioral sciences. If you see humans and our societies as having special status because of our particular kind of consciousness, then I can see why you'd hold little hope for social science. I don't think I could even try to do behavioral science if I believed that.

*when trying to study behavior, I mean. They're perfectly nice things to have in general.


Posted by: cerebrocrat | Link to this comment | 10- 7-07 6:21 PM
horizontal rule
158

noise*

Feedback and/or nonlinearity, I'd argue.


Posted by: Sifu Tweety | Link to this comment | 10- 7-07 6:22 PM
horizontal rule
159

So, the reason this paper is so useless is that none of the original data collected in this survey is interesting, nor do any of them have any interesting relationship to each other. It's a bunch of measures of what teachers do, so all I can do is run a bunch of meaningless statistical tests to show that teachers who do one thing also do other things, whereas teachers who don't do one thing don't do other things either. It's going to be insanely boring, and who even knows if it measures anything besides how people take surveys; maybe some people just like to answer no to all the questions, and other people just like to answer yes. But! There was one question that had a chance to show something that was a little bit interesting, because just one question asked about a teacher-reported student outcome. So maybe, I just thought to myself, if I run some chi-squareds on the student outcome versus the measures of teacher behavior, I will find something, which might actually be able to gesture at a statement about how teacher behavior helps or hurts students. My p's: .09 and .077. I piss on hypothesis testing.


Posted by: Tia | Link to this comment | 10- 7-07 6:35 PM
horizontal rule
160

158: I'd argue
What's stopping you?

159: My p's: .09 and .077
Sooo... is the silver lining here that you don't have to finish writing up the study?

And now, since this the Science! thread, a tip for all of our fine young and aspiring scientists: When measuring out powdered chemicals at your lab's balance, best safety practice demands that you only play shitty music on your iPod. If you carelessly let something groovy come up on the playlist, your inability to suppress the urge to whistle along can seriously compromise the measurements for your solution. Particularly problematic if the stuff was were toxic.


Posted by: cerebrocrat | Link to this comment | 10- 7-07 7:32 PM
horizontal rule
161

Because someone who's done grad-level work in evolutionary biology clearly should be lumped with superstitious primitives who think Darwin was the Devil. And by "chemotherapy" I obviously meant, like ibuprofen.

I don't know where I got the idea that scientists can be arrogant pricks who can't acknowledge any shortcoming. What was I thinking there?

Christ, how many of you philistines have taken me to task for actually using the fucking word "chemotherapy" properly?

I was talking about cancer treatment, not ibuprofen. For fucks sakes, "chemotherapy", even in the context of cancer treatments, means treatment by chemicals, whether small molecule, proteins, or what have you. So when you write "in X number of years in the future, the idea of chemotherapy will seem actually worse than leeches", you are in fact saying that in the future the idea of treating cancer with chemical intervention will be seen the same way as leeching people.

Maybe that's not what you meant, but that's what you wrote.


Posted by: gswift | Link to this comment | 10- 7-07 7:53 PM
horizontal rule
162

160.1: what, you want more than a hedged assertion? I call undergrad!

Seriously it seems strange to me to say that self-conscious, reflective conscious doesn't effect behavior; I just think that it's likely to be almost impossible to model predictively because it's a feedback system. And shit.


Posted by: Sifu Tweety | Link to this comment | 10- 7-07 7:59 PM
horizontal rule
163

consciousness would be more appropriate in the second instance above.


Posted by: Sifu Tweety | Link to this comment | 10- 7-07 8:01 PM
horizontal rule
164

An average undergrad electrical engineer would laugh heartily at the idea that feedback makes a system impossible to model. Modelling - and hell, designing - feedback systems is an entire subfield.


Posted by: Nathan Williams | Link to this comment | 10- 7-07 8:38 PM
horizontal rule
165

164: predictive modelling of nonlinear systems is complex when possible at all, no?


Posted by: Sifu Tweety | Link to this comment | 10- 7-07 8:41 PM
horizontal rule
166

Especially multi-dimensional nonlinear systems with no particularly applicable differential equations.


Posted by: Sifu Tweety | Link to this comment | 10- 7-07 8:44 PM
horizontal rule
167

I'm not saying that it's not a complicated system we're talking about, or that we're going to understand it all tomorrow, but invoking "feedback" doesn't create any kind of magic jump into un-analyzable territory.


Posted by: Nathan Williams | Link to this comment | 10- 7-07 8:46 PM
horizontal rule
168

No, I know that. I was using somewhat inept shorthand, and actually it has so far proven pretty complicated to create useful error and update functions in neural networks that incorporate any kind of complex feedback.


Posted by: Sifu Tweety | Link to this comment | 10- 7-07 8:49 PM
horizontal rule
169

Maybe that's not what you meant, but that's what you wrote.

You told them! Another blow struck in the fight for truth and accuracy!


Posted by: mrh | Link to this comment | 10- 7-07 8:54 PM
horizontal rule
170

multi-dimensional nonlinear systems with no particularly applicable differential equations

Aieee! Run away!

It seemed obvious at first, but now that I think about it, I'm not sure I get how consciousness would function as a feedback system. And of course, speaking of consciousness,

Aieee! Run away!


Posted by: cerebrocrat | Link to this comment | 10- 7-07 9:12 PM
horizontal rule
171

170: well, it would (really, really approximately (and just for instance (getting pretty far from things I actually understand, here))) get input from somatosensory areas, create a plausible narrative to explain them, and given that narrative feed back into executive motor control areas. And shit.


Posted by: Sifu Tweety | Link to this comment | 10- 7-07 9:15 PM
horizontal rule
172

And shit.

You are Dan Dennett. I claim my five pounds.


Posted by: nattarGcM ttaM | Link to this comment | 10- 8-07 8:47 AM
horizontal rule
173

I claim my five pounds.

Tthat reminds me so much of usenet (it spread from the Brits to all the regulars on the rec.arts.sf hierarchy).


Posted by: emir | Link to this comment | 10- 9-07 7:56 AM
horizontal rule