Saturday, June 27, 2009

What epistemology?

Learned of the following via the Center for Inquiry. It's an article by physicist Lawrence Krauss in yesteday's Wall Street Journal.

Science is only truly consistent with an atheistic worldview with regards to the claimed miracles of the gods of Judaism, Christianity and Islam. Moreover, the true believers in each of these faiths are atheists regarding the specific sacred tenets of all other faiths. Christianity rejects the proposition that the Quran contains the infallible words of the creator of the universe. Muslims and Jews reject the divinity of Jesus.

So while scientific rationality does not require atheism, it is by no means irrational to use it as the basis for arguing against the existence of God, and thus to conclude that claimed miracles like the virgin birth are incompatible with our scientific understanding of nature.

Finally, it is worth pointing out that these issues are not purely academic. The current crisis in Iran has laid bare the striking inconsistency between a world built on reason and a world built on religious dogma.

Perhaps the most important contribution an honest assessment of the incompatibility between science and religious doctrine can provide is to make it starkly clear that in human affairs -- as well as in the rest of the physical world -- reason is the better guide.

I really don't understand why anyone bothers defending religion. It claims to have a valid epistemology. Well, ok, show us. What supernaturalistic claim does any religion have that's been verified to be true? We've been waiting for several millennia now. Face it. Religion has tons of supernatural claims. None have been found or is known to be true It has tons of empirical claims too (and it even claims to have a real epistemology) and a good number have already been disconfirmed (eg. the earth is not--as some sacred texts claim--flat, the age of the earth is way more than ten millennia, there is just no evidence for any worldwide deluge, evolution is the origin of species not some mythic anthropomorphic militaristic macho prick as claimed by a tribe in the Bronze Age, ....)

Provide us a means by which we can find out whether religious epistemology actually works. For instance offer a procedure by which revelation can be shown to be a valid way of knowing. And show that revelation (whereby a "truth" or idea is placed in someone's head by a supposed supernatural entity) actually is a revelation from a supernatural entity and not just natural firing of neurons. Show that whatever "truth" is supposedly uttered by this prophet is not true simply due to coincidence or was arrived at by other ways of knowing.

In contrast, science has demonstrated again and again and again over hundreds of years that its epistemology does work and does provide us reliable knowledge. If you've ever used a computer and know of the things called airplane and antibiotic and seen those stunning Hubble deep space photographs then you would be a freaking idiot to claim that science doesn't work.

Epistemic progess is of course a barometer of the validity of the epistemology that any discipline flaunts. So which domain has had epistemic progress? Yes that's a no-brainer. Which religion has had any epistemic progress vis-a-vis supposed supernatural "truths"? If this were a contest (and I think it is!) religion should've been booed out of the stadium eons ago.

Now tell me which of the two is arrogant: Science which lets reality be the ultimate arbiter as to which explanations, hypotheses, theories are true, or religion which rides roughshod over reality whenever any of its doctrines/dogmas are contradicted by reality?

Yes, I am so uber frustrated over the fact that even educated adults keep getting duped into clinging to superstitions in sheep's clothing.

Thursday, June 11, 2009

How not to find cause of and cure for cancer

A friend brought to my attention (via a forwarded email that has been circulating for some time it seems) a certain Professor Jane Plant who claims in a book of hers that she's discovered a causal factor of breast cancer and that she was cured of her own affliction by eliminating milk products in her diet. The book may be old hat to some of you but I just got wind of it.

Just to be sure that this wasn't another urban legend, I turned to Well, Jane Plant and her book Your Life in Your Hand (the US edition is entitled The No-Dairy Breast Cancer Prevention Program: How One Scientist's Discovery Helped Her Defeat Her Cancer) checks out. So much the worse for Plant as you'll see. Go read the widely available excerpt from her book (that particular website goes to show how governments are not necessarily keepers of the light, enlightenment that is).

Allow me to put cart before the horse and provide you my conclusion: Geologist and professor Jane Plant is incompetent. I say that because she's supposedly a scientist and yet she commits errors in thinking and analysis that would earn science undergrads a failing grade. How she could, in the same breath, remind us she's a scientist and write as a woowoo is jaw dropping.

Now for my arguments.

Plant tells us that it dawned upon her that in China practically no one drinks cow's milk and that dairy products including cheese are not part of the diet. She also says that statistics show that only 1 in 10,000 women in China die from breast cancer, while the figures for Western countries is around 1 in 10. (Let's at the moment just take for granted that she has her numbers right, although those would need to be checked too of course--I'm suspicious of the 1 in 10 stat). So Plant puts two and two together and comes up with the hypothesis that dairy product consumption might be a or the culprit.

Can we jump to the conclusion that milk is dangerous to women's health? Most certainly not. There's a truism in statistics and science: Correlation does not necessarily imply causation. Correlation is the phenomenon wherein two or more variables/events are associated with one another. For example, the temperature of the ground is correlated with the time of day--the closer it is to noontime the warmer the ground is. And of course this is because the sun heats the earth up. However, just because two variables are correlated does not mean one causes the other. The clock we used in recording the time of day obviously does not cause the ground to heat up. Yet another example. Over two decades ago researchers in Taiwan found that there is a strong correlation between the number of electrical appliances (including toasters) that a household owns and the frequency of use of birth control methods. Does this mean then that buying more appliances causes people to resort to contraceptives more often? Or does it mean that higher frequency of employing contraceptives makes Taiwanese buy more appliances? One of these would have to be our conclusion if correlation were equivalent to causation. The truth of the matter, however, is that the above variables are both correlated with yet other variables, namely, income and educational attainment. And it is these two latter factors that cause the increase in both number of appliances owned and contraceptive use. Income and education are both correlated with the former two variables. They also are causal factors.

In summary, if X is correlated with Y then either X causes Y, or Y causes X, or neither X nor Y is a cause of the other. On the other hand, if Q is the known cause of P, then Q and P will by necessity be correlated with one another. So while correlation does not necessarily imply causation, causation necessarily implies correlation.

In discussing Plant's hypothesis, another friend reminded me of how the Chinese consume a lot of soybean in its various forms--tofu, soybean milk, soy sauce, salted soy beans, etc. That in itself would correlate significantly with breast cancer mortality since Westerners consume less soybeans than Orientals. Just as with dairy products we could also say something like, It might be that the high consumption of soybean products guards against the occurrence of breast cancer. And these surely are not the only variables that correlate with breast cancer incidence and mortality. You could scour the world for various factors and find correlations, both positive and negative.

Plant goes on to tell us that based on this correlation and hypothesis of hers, she stopped taking any product that contained milk. She narrates what happened soon thereafter:
About two weeks after my second chemotherapy session and one week after giving up dairy produce, the lump in my neck started to itch. Then it began to soften and to reduce in size. The line on the graph, which had shown no change, was now pointing downwards as the tumour got smaller and smaller.
She goes on to conclude that based on her experience she was right in identifying milk as the cause.
It was difficult for me, as it may be for you, to accept that a substance as ‘natural’ as milk might have such ominous health implications. But I an i living proof that it works....
Well, unfortunately for Plant, her reasoning is flawed. She commits the post hoc ergo propter hoc fallacy (after this, therefore, because of this). This is a causal attribution error whereby just because event B comes after A, we conclude that A caused B. Thus, just because the sun rose after the rooster crows obviously does not imply that the chicken made the sun rise. But this is precisely what Plant is saying. She tells us that she stopped taking any dairy product after which she noticed that her lump started waning until it totally disappeared. She then attributes this to her diet change. This is a textbook case of causal attribution error.

But note how she herself tells us that she was simultaneously undergoing chemotherapy, a treatment modality that is known to work against cancer. Why does she not attribute the remission to chemo? So what actually caused her lump to up and disappear? Well, we don't know for sure. It may have been the chemo, the diet change, both of these, spontaneous remission, or something else. To jump to the conclusion that it was diet change is to commit the post hoc fallacy.

Furthermore, Plant has the gall to jump to a conclusion based on her own personal experience. One anecdote is not evidence. And even if a thousand other women shared a similar story, it would still not be evidence. Thus, the aphorism goes: The plural of "anecdote" is not data, it's "anecdotes." And even if this had been a clinical trial, it is impossible for it to be a randomized controlled study for the simple reason we'd need to have at the very least two participants in the study--with one serving as the control. A clinical study with a sample size of two is in itself laughable--the margins of error would be so huge as to make its results practically useless. Hence, a sample size of one is just absurd.

Linda Bily, a reviewer on of Plant's book, likewise apprehends the lack of critical thinking that Plant manifests and has this to say:
The premise that since Oriental women don't consume a lot of dairy products and have less incidence of breast cancer is plausible, but unproven. I shudder to think of the thousands of women who will change their diets based on this book. I am most concerned that the high intake of estrogens and phytoestrogens, especially in the soy products recommended, could be detrimental to some women. There is still controversy in the medical community about the use of soy. If you read this book as an interesting scientific, but unproven, premise, you will be fine. If you take this book to heart, without consulting your medical specialist, you could be opening a can of worms. Dr. Plant is a respected scientist in her field. As a breast cancer survivor and advocate, I question some of her findings. The studies she cites to validate her ideas are older, some of obscure practice and are not widely confirmed. I also take issue with her description of her own breast cancer diagnosis. It returned 5 times according to the author and yet she states that it was an early stage at diagnosis. The tumor on her neck disappeared during chemo and she credits only her non-dairy diet for this shrinkage. She says that it spread to her lymphatic system, but her lymph nodes were clear. The book is interesting reading, but while I do not doubt her personal beliefs or her expertise as an earth-based scientist, I do hesitate to recommend this book to anyone. I am afraid that too many women, looking for a quick fix, will adapt her lifestyle without question. There still is no known cause or cure for breast cancer. Feel free to search alternative options and methods, but please, discuss any changes in your treatment, diet or life with your medical team and make an informed decision.

So, is Jane Plant's hypothesis that dairy products are a casual factor in breast cancer wrong? Certainly not. Let me repeat that in case you think that's a typo. It may be that regular consumption of dairy products are (partly) responsible for breast cancer. My critique above does not imply that Plant's hypothesis is totally off the mark. What is utterly awry are the methods/reasoning by which she reaches her conclusion, which means it is nowhere close to being conclusive. Keep in mind that an argument may have false premises but true conclusions. When the argument contains various fallacies then the conclusion cannot be known to be true. However, if the argument is sound (i.e, the premises are known to be true and the argument contains no fallacies) then the conclusion must by necessity be true. Because Plant does not follow scientific protocol (ie., objective, unbiased methods of testing hypotheses) we cannot have any confidence in her conclusion.

As a scientist, what Plant could have done is applied for a grant and performed randomized, double-blind, controlled clinical trials (RCT) employing at least several dozens of participants. Barring this (for ethical or whatever reason), she could've performed an epidemiological study (just as was done with tobacco use and lung cancer decades ago), although such studies hardly provide the degree of certitude of RCTs. (But since she has no degree nor expertise in medicine I doubt she would've been awarded research money in the first place; thus she should've left testing of this hypothesis to the experts).

It is said that when a layperson makes a mistake in matters of, say, rocket science then that mistake is out of ignorance. But when a rocket scientist commits an error involving rocket science, that's stupidity. Prof. Jane Plant claims to be a scientist. But she made elementary mistakes about hypothesis testing and induction. Now that's utter stupidity.

As we've seen above Plant says of her cancer treatment: "I am living proof that it works." No, Prof. Plant. You're living proof that you failed to learn the essentials of Scientific Method 101. For shame!

Sunday, June 07, 2009

Speak univocally, not equivocally

What's wrong with the following argument?
Mad men should be put in mental asylums. Annie's dad is mad--he just berated her for one full hour for taking money from his wallet without asking permission. Therefore, Annie's dad should be carted away in a straitjacket and locked up in a mental institution. (example is adapted from Introduction to Logic and Critical Thinking, 3ed., by Merrilee H. Salmon, Harcourt Brace, 1995, p.47)
If that made you chuckle, then you hit the nail on the head intuitively. In the first sentence (statement) "mad" is used in the sense of "insane" / "mentally disturbed" / "demented". However, in the second statement "mad" is used in the sense of "angry." Because Annie's dad was angry and not insane, the conclusion (the last statement) simply does not follow. If you say that the conclusion does follow from the preceding statements then you're mad! (no, not angry, but nuts and off your rockers).

The use of the same term but with different meanings within the argument is known as the fallacy of equivocation. Professors of logic Copi and Cohen tell us that equivocation is an:
informal fallacy in which two or more meanings of the same word or phrase have been confused. If used with one of its meanings in one of the propositions of the argument but with a different meaning in another proposition of the argument, a word is said to have been used equivocally (p.688)

Equivocal arguments are always fallacious. (p.192)

[Irving M. Copi & Carl Cohen, Introduction to Logic, 10th ed., Prentice-Hall, 1998]

As we've seen from the (facetious) example above the premises (1st and 2nd statements) have nothing to do with one another given that the same word "mad" was used to mean quite different things. The Philosophy Pages tells us that: "The inferential relationship between the propositions included in a single argument will be sure to hold only if we are careful to employ exactly the same meaning in each of them," in other words, if the terms are used univocally--with only one meaning--and not equivocally--several meanings.

As with other logical errors, equivocation is sometimes used in humor. Lewis Carroll, for instance, employs it in Through the Looking Glass:
"Who did you pass on the road?" the King went on, holding his hand out to the messenger for some hay.
"Nobody," said the messenger.
"Quite right," said the King; "this young lady saw him too. So of course Nobody walks slower than you."

(quoted in Copi & Cohen, p.192)
Moving on to more serious examples.
It's the duty of the press to publish news that's in the public interest. There is great public interest in UFOs. Therefore the press fails in its duty if it does not publish articles on UFOs. (Theodore Schick, Jr. & Lewis Vaughn, How to Think About Weird Things: Critical Thinking for a New Age, 2ed., Mayfield, 1999, p.286)
Did you catch how "public interest" shifted in meaning? From meaning "welfare of the public" in the first statement it changed to "what the public wants to read about" in the second.

Here's one that's requires some prior scientific understanding. It's also an example that has an ethical side to it:
[A] sugar advertisement ... argued for increased consumption of sugar on the grounds that "Sugar is an essential component of the body ... a key material in all sorts of metabolic processes." (Howard Kahane & Paul Tidman, Logic and Philosophy: A Modern Introduction, Wadsworth, 1995, p.311)
It is true that sugar is an essential component of the body. But this "sugar" is glucose. On the other hand, the "sugar" which the advertisement is promoting is sucrose--table sugar. While both glucose and sucrose are examples of sugars (i.e., saccharides), using "sugar" to mean glucose in one part of the advert and "sucrose" in another part (even if just implicitly) is a blatant commission of equivocation. Given that this is an ad, it is almost certain that the ad makers were fully aware of what they were doing and intentionally took advantage of the equivocal meaning of "sugar" in an attempt to mislead and dupe the consumer (perhaps to counter the prevailing notion that table sugar in one's diet should be reduced to a minimum).

Now that wily, deceitful ad appropriately segues to an example of equivocation by a member of the sect/cult Jehovah's Witnesses (JW) who had the temerity of locking horns with an atheist using illogic and absurd claims. Like the above ad, his argument misleads readers by resorting to the equivocal use of a pivotal term.

This JW made the following argument:
The gods in the religion of the atheists are the atheists themselves. The atheists deny the existence of all gods. But the atheists believe that they are gods. So they idiotically deny their existence.
If you're a bit nonplussed as to how and why he can make the claim that atheists are gods, he offers the following idiosyncratic definition of "god":
A god can be the true God, can be any powerful being, any person with power in high position or anyone can be a god over a group of people under him, anyone makes himself a god if he denies the true God, a god can be also a thing like money, sex, idol, etc.
Going back to his argument about atheists, let's number the sentences/statements therein:

1. The gods in the religion of the atheists are the atheists themselves.
2. The atheists deny the existence of all gods.
3. But the atheists believe that they are gods.
4. So they idiotically deny their existence.

In #1 since atheists are human beings, natural (not supernatural) phenomena, we know that he uses "gods" in the sense he has defined it.

In #2 however, "gods" can only pertain to supernatural entities since that is what atheists don't believe in, that is what "gods" mean when atheists declare "I/We do not believe in gods". It cannot be in the sense that this JW has defined it since needless to say atheists believe in the existence of powerful persons in high positions, in the existence of other humans beings, in the existence of sex, money, and idols (presumably he means that sex, money and other objects in the world can be idolized, i.e., inordinately valued by some people to the point of obsession, reverence, etc.). This JW cannot of course claim that "gods" in #2 refer to a subset of the "gods" as he has defined it since he tells us that "atheists deny the existence of all gods" (emphasis added). Insisting that "gods" here is the same as in #1 would mean that premise #2 is false, pretty obviously so, thus pulling the rug from his argument.

In #3,ostensibly, he uses "gods" in the sense as he does in #1.

Given that this JW uses "gods" in two different senses, his argument commits the fallacy of equivocation.

#4 is the conclusion. But as we've seen and learned above this conclusion cannot legitimately follow from the premises because the word "gods" has been used equivocally.

In order to further see more clearly how the above argument in fact commits the fallacy of equivocation here is an example that uses "God" (capitalized) equivocally:
Some religious arguments can also include equivocations, for example:
It is not possible for the universe to exist without a cause, therefore there must have been a First Cause, which we can reasonably call "God." I already believe in the God of the Bible, and now you have no excuse for not doing so as well.
[W]e can see that God is being used in two entirely different ways. In the first sense, God is simply being used as a convenient term to describe a First Cause of the universe, with no particular attributes beyond that which is necessary to cause a universe. But in the second sense, the term God is used for something much more specific and with many more attributes: a traditional Christian conception of God.

(Fallacies of Ambiguity: Equivocation)

In logic, a deductive argument is said to be valid if it contains no fallacies. In a valid argument the conclusion logically and necessarily follows from the premises. An argument is said to be sound if the argument is valid and all the premises are known to be or have been shown to be true. Given a sound argument the conclusion therein must necessarily be true.
It is important to stress that the premises of an argument do not have actually to be true in order for the argument to be valid. An argument is valid if the premises and conclusion are related to each other in the right way so that if the premises were true, then the conclusion would have to be true as well. We can recognize in the above case that even if one of the premises is actually false, that if they had been true the conclusion would have been true as well. Consider, then an argument such as the following:
All toasters are items made of gold.
All items made of gold are time-travel devices.
Therefore, all toasters are time-travel devices.
Obviously, the premises in this argument are not true. It may be hard to imagine these premises being true, but it is not hard to see that if they were true, their truth would logically guarantee the conclusion's truth. It is easy to see that the previous example is not an example of a completely good argument. A valid argument may still have a false conclusion. When we construct our arguments, we must aim to construct one that is not only valid, but sound. A sound argument is one that is not only valid, but begins with premises that are actually true.

(Validity and Soundness)

Given this primer on validity and soundness, is the argument by the JW a valid argument? No, since it contains at least one logical fallacy.

Is it a sound argument? No, since it is invalid. And we don't even need to ask if the premises are true (a necessary condition for an argument to be valid), because an invalid argument can never be sound (validity is also a necessary condition for an argument to be sound).

The lesson: Make sure you use terms consistently. You can preclude equivocation by defining your terms precisely and in detail at the very beginning and double checking that every instance of the term is consistent with how it has been defined. Remember: the existence of equivocation renders an argument invalid. And sometimes, if flagrant, it may make it rather silly too.