Acupuncture Revisited: Idiot Edition

I haven't been blogging lately (sorry, that whole dissertation thing), but I've been moved to anger over this article "Think acupuncture's a hoax? Think again.":

http://news.cnet.com/8301-27083_3-20006371-247.html

I think what bugs me the most is that the idiot who wrote the article, Elizabeth Armstrong Moore, calls herself "a skeptic". It's like homeopaths calling themselves "practitioners of evidenced-based medicine".

The argument in the article is as follows: some doctors poked mice with needles, and they noticed an upswing in adenosine, and anti-inflammatory, soothing chemical. Therefore acupuncture is not a hoax, it works, and we should all carry around personal acupuncture kits.

You really can't make this shit up. Recall what we learned last time, when we looked at a recent meta-analysis of acupuncture treatments in BMJ:

"A small analgesic effect of acupuncture was found, which seems to lack clinical relevance and cannot be clearly distinguished from bias. Whether needling at acupuncture points, or at any site, reduces pain independently of the psychological impact of the treatment ritual is unclear."

Moore says that there is a "seeming lack of evidence" that acupuncture works. But the case is stronger than that. We have a seeming presence of evidence that acupuncture does not work. It "seems to lack clinical relevance". "Clinical relevance" means "works".

The problem for acupuncture is NOT that there's no plausible physical mechanism by which it might work. Our real skeptic, Orac, said this: "there are needles breaking the skin in acupuncture. It was, at least to me, not entirely implausible that that might have some sort of physiologic effect." The problem is that EVEN IF there's a plausible physical mechanism by which it works and EVEN IF it does in fact cause an increase in adenosine, we still KNOW that it isn't clinically significant. You can't improve people's health by using it, beyond what you could do by giving people placebos. It may "do something". But doing something is not the same as working. No one doubts that sticking needles into someone does something. But why do it if it doesn't relieve their pain?

Furthermore, the journalist (I don't use scare-quotes, because this is about the level of nonsense I expect from journalists) goes on to say that because mice have elevated adenosine when pricked with needles, we should download "an app, called Qpalm Acupuncture, that maps out acupoints and formulae for treating 59 diseases. and another, iLocate-Acupuncture, for finding acupuncturists near you." That's right, needles can fool the body into thinking it's been damaged. Therefore, it's important to stick needles into those exact places that ancient Chinese pseudoscientists thought were points through which the qi flowed, and acupuncture can *cure diseases*. Aaaaaaaaaaaagh! The stupid.

Skeptic my ass.

Pesky Truth-Tables

One of the blog posts on the new Philosopher's Carnival considers how to teach the truth-table for the material conditional to students:

http://www.soulphysics.org/2009/12/where-material-conditional-gets-its.html

I tried to comment, but it didn't work for some reason. Here's how I explain the troublesome cases (where P is false, but (P -> Q) is true).

Suppose Q is true, and P false. All we really need is that Q is true:

1. Q / Q

Monotonicity gets us:

2. P, Q / Q

And then by conditional proof we derive:

3. Q / P -> Q

One final application of monotonicity:

4. ~P, Q / P -> Q

So to reject (P -> Q) when P is false an Q true, the student must reject either conditional proof or monotonicity.

Now suppose P is false and Q is also false. All we really need is that P is false:

1. ~P / ~P

Again, by monotonicity, and conditional proof:

2. ~Q, ~P / ~P
3. ~P / ~Q -> ~P

Here there are several ways to proceed (contraposition being the most obvious). Let's try Modus Tollens:

4. P / P (assumption for CP)
5. P / ~~P (double negation, 4)
6. P, ~P / ~~Q (modus tollens, 3, 5)
7. P, ~P / Q (double negation, 6)
8. ~P / P -> Q (conditional proof, 7)
9. ~P, ~Q / P -> Q (monotonicity, 8)

So to reject (P -> Q) when P and Q are both false, the student must reject conditional proof, double negation, monotonicity, or modus tollens (alternatively, the student must reject conditional proof, monotonicity, or contraposition).

Of course, as pointed out last time, modus tollens, contraposition, disjunctive syllogism, and reductio are all inter-derivable. So my explanation is really of the form: everything follows from a false premise, so get used to it.

Independence of Reductio

The other day I was wondering whether one can have a complete propositional calculus with only the introduction and elimination rules for each connective. So you have modus ponens and conditional proof, but not modus tollens and contraposition; or-introduction and or-elimination, but not disjunctive syllogism; &-intro and &-elim, but not "modus ponendo tollens", i.e. from ~(A & B) and A to infer ~B. In particular, you don't have reductio ad absurdum, from A & ~A to infer B.

Such a system seems simple and intuitive, but I don't see how to do it. At a minimum, we should be able to prove p & ~p -> q from the introduction and elimination rules alone, but I haven't been able to prove it. You can get it assuming any of the principles we've ruled out, e.g. modus tollens, contraposition, or disjunctive syllogism. For example (using / as a turnstile):

1. {p & ~p} / p & ~p (Assumption)
2. {p & ~p} / p (&-elim, 1)
3. {p & ~p} / ~p (&-elim, 1)
4. {~q} / ~q (Assumption)
5. {p & ~p, ~q} / p & ~q (&-intro, 2, 4)
6. {p & ~p, ~q} / p (&-elim, 5)
7. {p & ~p} / ~q -> p (->-intro, 6)
8. {p & ~p} / ~~q (MODUS TOLLENS, 3, 7)
9. {p & ~p} / q (~-elim, 8)
10. / (p & ~p) -> q (->-intro, 9)

The proof assuming contraposition is similar, but instead we show ~p -> ~q (using ~p where we have p in lines 5-7), contrapose, and use p and ->-elim to conclude q (rather than ~p and modus tollens). For disjunctive syllogism, the proof is even simpler:

1. {p & ~p} / p & ~p (Assumption)
2. {p & ~p} / p (&-elim, 1)
3. {p & ~p} / p v q (v-intro, 2)
4. {p & ~p} / ~p (&-elim, 1)
5. {p & ~p} / q (DISJUNCTIVE SYLLOGISM, 3, 4)
6. / (p & ~p) -> q (->-intro, 5)

What's more, we can prove any of these rules (modus tollens, contraposition, disjunctive syllogism) by assuming reductio (it's an easy exercise). So it follows that if we can prove any of these four rules from Assumption and the intro and elimination rules, our system is complete (well, it doesn't follow per se, but I could show you, if asked).

So, here's what I want to know from my readers, few though they may be. Can you prove reductio, modus tollens, contraposition, or disjunctive syllogism from just the intro and elimination rules? And if not, why not? Why does a complete propositional calculus need something other than the intro and elimination rules? I feel like this is something I should've learned in logic, but never did.

(Aside: yes, I know this post is going to make me look foolish when someone presents the answer, whatever it is. So be it.)

Acupuncture: Evidence

So the other day I was at a friend’s house and the topic of acupuncture came up. Now, I’m one of those hard-line skeptics about “alternative” medicine, and I happened to have read several credible sources stating that acupuncture doesn’t work (is no better than a placebo treatment). But not everyone wanted to take me on my word, so I promised to produce the evidence. Here we go.

First, a caveat for those who want to do their own internet research: there are a large number of biased websites out there, on both sides. If you go to the websites of people who practice acupuncture, you’ll find links to many articles which indicate that acupuncture is beneficial. Don’t trust them! For example, on this website:

The Fertile Soul

You’ll find the oft-cited claim that acupuncture helps with in-vitro fertilization. Yet, the study linked is a poor one, and the most systematic review to date on the subject, here:

Cochrane Collective

Concludes: “Acupuncture performed on the day of ET [Embryo Transfer] shows a beneficial effect on the live birth rate; however, with the present evidence this could be attributed to placebo effect and the small number of women included in the trials. Acupuncture should not be offered during the luteal phase in routine clinical practice until further evidence is available from sufficiently powered RCTs [Randomized Clinical Trials].”

Translation: poorly conducted studies are inconclusive. Internet researchers should also be warned about overzealous skeptic sites, which will publicize the negative results that suit them, but not take the time to consider studies which seem at-odds with their views.

The best place to get your scientific evidence concerning medicine is, of course, from recent, systematic meta-analyses of randomized clinical trials published in leading medical journals. Here is one such meta-analysis, published in BMJ. The article is: “Acupuncture treatment for pain: systematic review of randomized clinical trials with acupuncture, placebo acupuncture, and no acupuncture groups” by Matias Vested Madsen, Peter C Gøtzsche and Asbjørn Hróbjartsson and it is available for free here:

Meta Analysis

This article is recent, being published in 2009.

It is also systematic. A systematic meta-analysis looks at all the articles in a large database(s) that meet the inclusion criteria. According to the study’s authors: “We searched the Cochrane Library, Medline, Embase, Biological Abstracts, and PsycLIT. The last search included all trials published before 1 January 2008.”

Furthermore, it is published in a leading medical journal. According to wikipedia: “BMJ is considered to be one of the ‘core’ general medical journals; the others being the New England Journal of Medicine, (N Engl J Med), the JAMA, and The Lancet.”

The authors reached the following conclusions: “A small analgesic effect of acupuncture was found, which seems to lack clinical relevance and cannot be clearly distinguished from bias. Whether needling at acupuncture points, or at any site, reduces pain independently of the psychological impact of the treatment ritual is unclear.”

To sum up: acupuncture treatments for pain are not clinically significant, and they may not even be statistically significant, once bias is accounted for. That seems pretty damning to me.

Still not convinced? I knew you wouldn’t be. So I direct you to one of my favorite science bloggers, Orac, over at Respectful Insolence.

Acupuncture 1

Acupuncture 2

Acupuncture 3

Acupuncture 4

Orac is a medical doctor, and much more knowledgeable than me. In the above blog posts, he dissects a number of studies and meta-analyses. He testifies:

“When I first became interested in ‘alternative medicine’--excuse me, I mean ‘complementary and alternative medicine’ (CAM)--I viewed acupuncture somewhat differently. No, I never bought the traditional explanation that sticking thin needles into the skin somehow alters the flow of qi in order to induce a therapeutic effect. That is no more plausible than reiki or therapeutic touch. However, there are needles breaking the skin in acupuncture. It was, at least to me, not entirely implausible that that might have some sort of physiologic effect. Then I had to go and ruin that lovely kumbaya feeling towards CAM by actually going and looking at the scientific literature on acupuncture… When I actually bothered to do that, I soon realized that the evidence that acupuncture is anything more than a highly elaborate placebo is shockingly thin. More like nonexistent, actually.”

You should read his posts, or the articles he links to, and see if you aren’t equally converted.

Finally, here’s a press article to stew on:

AP Report

It begins: “Ten years ago the [U.S.] government set out to test herbal and other alternative health remedies to find the ones that work. After spending $2.5 billion, the disappointing answer seems to be that almost none of them do.”

Your tax dollars at work!

Accommodationism II

Joshua Rosenau over at Thoughts from Kansas has a new post up about the accommodationist debate.

http://scienceblogs.com/tfk/2009/06/more_on_accomodationism.php

It’s typical of the accommodationist genre: reinterpret Coyne so that he’s saying something he repeatedly says he’s not saying, and then have an irrelevant and one-sided debate. Here’s a taste. First, the old (old, old, getting very old) canard that: look! there are religious scientists, so science and faith are compatible:

“Fr. George Coyne is an astronomer. He was, for many years, the director of the Vatican Observatory, and was a Jesuit priest for even longer. Does he see any conflict between his study of the heavens and his belief in heaven? Not at all”

[Note: this is confusing. JERRY Coyne is the anti-accommodationist; GEORGE Coyne is the catholic astronomer.]

Yes, yes, we get it. One and the same person can say both “I believe in the divinity of Jesus” and “I believe in contemporary evolutionary biology.” That was news, like, never and a half ago. So what’s supposed to be the upshot of Fr. G. Coyne-style accommodationism?

“Theology responds to new scientific discoveries, just as it reacts to cultural shifts. Claims about theology are tested in a different way than scientific claims, indeed cannot be tested as scientific claims, but that does not mean they are invalid. He sees science and religion as connected, as informing one another in certain ways, and as rooted in the same reality, therefore incapable of contradiction. Apparent contradictions must be addressed by further study.”

What does it mean for science and religion to be “incapable of contradiction”? And what is an “apparent contradiction”? I take it the view is this: suppose current science supports theory T1, which entails that p; and current theology supports theory T2, which entails that not-p. Then science and religion are in “apparent contradiction.” But the contradiction is only apparent, because “further study” ultimately resolves the contradiction, by replacing either T1 with another theory that doesn’t entail p; or by replacing T2 with another theory that doesn’t entail not-p; or by replacing both theories in such a way to bring them into consistency.

But here’s where things get interesting. Forget whether science and religion are capable of contradiction, and focus on the question of whether they can be in conflict. I take it that some method of inquiry is in conflict with science if when science supports T1 and the other method T2, where T1 and T2 are inconsistent, it is (always or sometimes) science (that is, T1) that must be abandoned or revised. Surely we cannot tolerate methods of inquiry that conflict with science, so the real question is not whether apparent contradictions can always be resolved, but whether they are always resolved in such a way that science wins, and religion loses.

But this isn’t what we actually see from accommodationists:

“If Jesus lived, he walked the same world I do. If he was divine, the words he spoke struck ears like mine. His miracles were either sleight of hand, embellishments by later storytellers, or suspensions of natural laws. If I could take James Randi back a couple thousand years, I could figure that out, but as it is, all I can do is believe what I believe and let Christians believe what they do.”

Please, for the love of God, can’t one single accommodationist read Jerry Coyne’s New Republic piece? Let me state the argument. Among the basic principles of science is the uniformity principle (UP) as articulated by, among others, Harlequin, Emperor of the Moon: things are always and everywhere exactly as they are here*. That is, if the universal law of gravitation holds on Earth, it holds on the Moon, and the superlunary realm, and a million years ago, and a billion years ago, and a billion years hence. The assumption that nature is uniform is just the assumption that induction works, or that science is possible. And we use UP all the time to argue against creationist nonsense. We say, for instance, that the Earth is more than 6,000 years old, because we know (for instance) that radioisotope carbon-14 decays at a certain rate now and thus must always have done so, and thus some things have been around far longer than 6,000 years. Coyne’s argument, if anyone would bother to read it, is that one must reject UP to hold on to statements like “Jesus came back from the dead,” and “Mary was a virgin when she gave birth.” But if you abandon UP when it suits you, you have no principled grounds for rejecting the young Earth creationist’s claim that carbon-14 used to decay at a much faster rate. So accommodationism is on a par with young Earth creationism, at least so far as it rejects, without principle, fundamental aspects of scientific methodology.

Now, I don’t want to weigh in in favor of Jerry Coyne’s argument (here, at any rate). It certainly excited me when I first read it, and that’s what I hoped to get out of the accommodation debate: whether J. Coyne was right. But it seems that literally no-one has read the argument, or no-one has the intellectual honesty to reply to it. You can’t just baldly assert, in the face of compelling counterarguments, that there’s just nothing one ought to believe in these cases, and that accommodationist Christians are just as rational as atheists. That’s precisely the claim at issue! At least say something about the counterarguments.

Rosenau does try to place some religious doctrines even beyond the bounds of the UP:

“Randi and I could surely figure out what the score is with the loaves and the fishes and the wine. I don't know any way that we could test Jesus' divinity, or whether Mary was born with or without original sin.”

I’m thoroughly unconvinced. It’s like saying we can never know whether there are ionizing particles in a cloud chamber, because we can’t see tiny little particles. Yes, but physical theory predicts a mist around such particles, if present, and we can see the mist. Similarly, divine people can, I don’t know, perform miracles; or their predictions are never false; or they never lie, cheat, or steal. Can’t I observe Jesus to see whether he fits the bill? Does ‘divine’ mean so little nowadays? I’ll readily admit that some claims can’t be tested—e.g. “Mr. X has special property Y, which is in principle undetectable by anyone in anyway”—but most religious claims don’t seem to be of this form.

Alright, final quote:

“[Fr. G. Coyne] believes what he believes, he doesn't impose it on others, and it would be as wrong for others to impose their beliefs about science and religion on him as it would be for him to impose his beliefs on me.”

Arg! Since when did anyone advocate imposing atheism on anyone? Let me just make this very clear: as far as I see the accommadationist debate, there are two fundamental issues: (a) are science and religion in conflict? (b) ought we to say so, if that’s true? No anti-accommodationist has ever asked the question “Ought we to force others to be atheists?” and certainly none has or would answer it affirmatively. So STOP PRETENDING THEY DO.

I leave you with Hume, from the end of On Miracles (in the Enquiries):

“No testimony is sufficient to establish a miracle, unless the testimony be of such a kind, that its falsehood would be more miraculous than the fact which it endeavors to establish… When anyone tells me that he saw a dead man restored to life, I immediately consider with myself whether it be more probable that this person should either deceive or be deceived, or the fact which he relates should really have happened. I weigh one miracle against the other; and according to the superiority which I discover, I pronounce my decision and always reject the greater miracle. If the falsehood of his testimony would be more miraculous than the event which he relates then, and not till then, can he pretend to command my belief or opinion.”

*Check Leibniz, New Essays, for the reference.

Science's Claim to Truth

Here's a common trope: science is not warranted in claiming that its deliverances are true. Science doesn't "get at" the truth; rather, it "gets at" what the evidence most supports. So scientists shouldn't claim that their theories are true; rather, they should claim that their theories are what the evidence most supports.

I see this on blogs now and again (I got on someone's case just the other day for saying it), and I think here's a place where a little simple philosophy can help the non-philosopher out.

1. First objection. If anyone's warranted in asserting certain matters of empirical fact, such as that chiropractic is bogus, scientists are. Some people are warranted in asserting such matters; therefore, scientists are. Anyone warranted in asserting that p is warranted in asserting that "p" is true, because "p" is true when and only when p. Therefore, scientists are warranted in asserting, for example, that it is true that chiropractic is bogus.

2. Second objection. To be warranted in asserting that p, one must know that p. This is why it is infelicitous to say "p, but I don't know whether p" (for example: "it's raining, but I don't know whether it is or not"). So if a scientist is ever warranted in asserting anything, she must know it. But scientists are sometimes warranted in asserting things; therefore scientists sometimes know things. But knowledge is factive: if S knows p, then "p" is true. So to be warranted in asserting something, it must be true; provided scientists know this (and if they didn't before, now they do), they may infer from the fact that their assertions are warranted that what they say is true.

3. Third objection. A standard scientific reasoning pattern is abduction. Thus, we argue from the correlation between a rise in man-made greenhouse gasses and a rise in global temperatures, to the best explanation: the conclusion that humans are causing global warming. But on equally good footing is the inference from "the evidence supports p" to "'p' is true"-- for what better explanation could there be of the evidence supporting p, than "p"s truth?

So there it is, blogospheric soldiers of science: lay your claims to truth.

Accommodationism

The atheist blogs are abuzz with the accommodationist debate: are science and religion compatible, and if we think that they aren’t, ought we to say so? Jerry Coyne has a listing of all the main posts here:

http://whyevolutionistrue.wordpress.com/2009/06/12/the-big-accommodatinism-debate-all-relevant-posts/

Right now I just want to consider this one post by Chris Mooney:

http://blogs.discovermagazine.com/intersection/2009/05/31/civility-and-the-new-atheists/

In it, Mooney outlines 3 reasons he attributes to Barbara Forrest for why we should not criticize the accommodationists:

1. Etiquette. Or as Forrest put it, “be nice.” Religion is a very private matter, and given that liberal religionists support church-state separation, we really have no business questioning their personal way of making meaning of the world. After all, they are not trying to force it on anybody else.

2. Diversity. There are so many religions out there, and so much variation even within particular sects or faiths. So why would we want to criticize liberal Christians, who have not sacrificed scientific accuracy, who are pro-evolution, when there are so many fundamentalists out there attacking science and trying to translate their beliefs into public policy?

3. Humility. Science can’t prove a negative: Saying there is no God is saying more than we can ever really know empirically, or based on data and evidence. So why drive a wedge between religious and non-religious defenders of evolution when it is not even possible to definitively prove the former wrong about metaphysics?

[Note, all three reasons are copied directly, and are not summaries of mine.]

I want to take these reasons in turn.

1. Etiquette. I don’t feel the force of this reason at all. Suppose I believe that accommodationism is false, and I base my belief on a certain array of reasons and evidence. If I write up my position and give my defenses, I will have criticized the accommodationists, because any criticism of accommodationism is ipso facto a criticism of accommodationists. So the “etiquette” principle entails I must shut up. But why should I have to shut up, while the accommodationists run around presenting their reasons and their evidence for the truth of accommodationism? That’s a sorry state for a public debate to be in, where one side is allowed to marshal its defenses and the other has to just be quiet and put up with it.

Maybe the alternative is that I’m allowed to present my case, but I must do it far away from where any accommodationist is, to avoid offense. But why? We’re told that “religion is a very private matter.” I don’t really know what that’s supposed to mean, but I imagine it’s something like: people cherish their religious beliefs, and are very upset when those beliefs are taken to task. If that’s what it means, then I can think of some other things that are very private matters: morality and well-being. But surely no-one thinks that we should just let the anti-abortionists, or the vaccine denialists and conspiracy theorists have the only say on those issues, because morality and well-being are “private matters.” And if religion is so freaking private and Ken Miller is after a “personal way of making meaning in the world” why does he have so many high profile books on accommodationism? This seems like nonsense.

Elsewhere Mooney & Co. argue that we need accommodationists like Miller as allies in the battle against religious fundie anti-science wackaloonery. I agree, and I’m all for accommodationist allies. But I say this: no-one is an ally of mine who is so afraid of any reasons or evidence that runs contrary to their view, that if presented with such would turn tail and run to the other side. That’s not an ally, that’s a passive-aggressive control freak with reality issues. Anybody who actually has respect for science and the scientific method will not be turned off but rather invigorated by critical scrutiny of their positions.

2. Diversity. I find the claims under the “diversity” heading staggeringly silly. If you read Coyne’s New Republic piece, here:

http://www.tnr.com/story_print.html?id=1e3851a3-bdf7-438a-ac2a-a5e381a70472

You’ll find that Coyne’s charge is precisely that the accommodationists do sacrifice scientific accuracy. He presents three specific charges (probably more, I haven’t read the piece for months): Miller-style accommodationism violates (a) the law of biology that says dead people don’t come back to life (b) the law of biology that says virgin births in mammals are impossible and (c) the (admittedly contestable) claim that human-like intelligent creatures are not inevitable products of evolution by natural selection. (a) is presumably necessary for natural selection at all, because death is the method of selection precisely because of its finality; (b) is a precondition of Fischer’s demonstration of the sex ratios; and (c), though it could be false, is certainly worth looking into and it would be absurd to suppress arguments for it on the grounds that otherwise Ken Miller is going to cry.

In fact, I don’t even know what Forrest and Mooney are thinking here. How could there be a scientific critique of accommodationism that wasn’t of the form: accommodationism sacrifices scientific accuracy?

3. Humility. Allow me to me non-humble for a moment, but what Forrest/ Mooney says here is literally stupid. Of course science can prove negatives. Here’s a go: it’s not the case that vaccines cause autism. Or, if negative existentials are your bag: it’s not the case that there exists matter at the top of a mercury barometer. Does Forrest think that it’s impossible to prove the existence of vacuums? Who is she, the Catholic Church circa 1200? What’s more, even in intuitionistic logic, you get the theorem: p → not-not-p, so a proof of anything is a proof of a negative. Suck on that!

I suspect the heart of the issue is that it’s not possible to prove the non-existence of God. But again, I think Forrest/ Mooney is trotting out methodological claims without thinking about them. It’s impossible to prove the non-existence of a deistic God; but one can certainly prove the non-existence of the accommodationist God. The accommodationist God by definition has causal traffickings with the physical world. He’s a watered down Christian God. I mean, Coyne’s whole point, again, is that science (in his opinion) tells against accommodationism. If it tells against accommodationism, it tells against the accommodationist God. Now, Coyne may be wrong in the end, but you don’t get that result for free by saying “humility.”

Just to head off one bit of criticism: yes, I know, I’ve been using “scientific proof” as a standard that delivers less than 100% credence, so it is always conceivable that accommodationism is true. But (a) if Coyne is right about the evidence, then this conceivability is on a par with, say, the conceivable propositions that the Flying Spaghetti Monster or the Invisible Pink Unicorn created the world; and (b) if Forrest/ Mooney is really saying we shouldn’t make scientific cases for claims we can’t establish beyond any doubt whatsoever, then they’re saying we shouldn’t make scientific cases at all.

OK, that’s the end. I’m sure all of these points have been made by posters and commenters elsewhere. But I plan to post more on accommodationism as the debate develops, and I thought a good first start was deconstructing this Forrest/ Mooney crap in detail.

P.S. Sorry I never got around to posting every Friday. This is difficult!