Pound of Feathers, Pound of Gold

Which weighs more: a pound of feathers, or a pound of gold?

Close consideration of this riddle – and the conditions under which people tend to get it wrong – is helpful in understanding the limits of human rationality. It is a specific example which leads us to general principles of rationality failure.

These sorts of riddles and similar interpersonal language tricks (such as “Stupid says what?”) are especially popular among children but not among adults. Why is this the case? Partly because adults are more likely to have previously encountered and become familiar with their patterns, but there are other factors – including one very relevant one. Children tend to have less-developed capacities of impulse control.

It takes very little analysis to discover the ‘trick’ in the question; the concepts involved are relatively simple. But we’re confronted with the fact that people do answer it incorrectly, and that by manipulating aspects of the context in which the question is delivered, we can significantly increase the chance people will fall for it. What does this imply? That analysis is not being conducted in the erroneous cases, and that context is a contributing factor to whether people successfully engage in conceptual analysis. Specifically, that context determines whether people will counter their impulses long enough for analysis to be completed.

The key to these sorts of riddles is time pressure. If people feel free to take as much time as they like thinking over the question, they rarely fall for the trick. But if they’re trying to answer rapidly, they’ll screw up. Examples of situations that often result in such behavior include: competing against others to see who can be correct first, trying to demonstrate competence by investing little effort in answering, or encountering the question as part of a limited-duration examination. If several superficially-similar questions whose answer depends on retrieving facts from memory rather than performing logical analysis of the question are asked before the riddle is presented, that also tends to result in a wrong response.

The error occurs because of our weight-related associations with the concepts of ‘feathers’ and ‘gold’, our conditioned assumptions about the sorts of questions people are likely to ask, and a failure to inhibit the first impulses towards response. Feathers are far less dense than gold; any given volume of feathers will weigh far less than the same volume of the metal. Questions about a property rarely contain their own answers in a trivial way – we do not expect the defined quantities in the question to be equivalent relative to the property being asked about. And – this is the most vital aspect – it takes longer for our brains to process the question at a conceptual level than it does to activate our associations.

In the state of nature, organisms are often under intense pressure to produce results quickly. If they take too long, the resource they’re trying to exploit may be taken by a competitor – or worse, they may become exploited resources by a predator. So stimulus-response methods which produce generally-useful reactions tend to be favored over extremely accurate and precisely analysis that takes longer. As a consequence, natural modes of though available to humans favor rapid responses more than rigorous correctness – and in much the same way that the limits of our visual processing systems lead to optical illusions, which can be understood and thus constructed, the limits of our conceptual processing lead to inherent tendencies towards fallacies of reason, which can be exploited to produce riddles and language gags.

Just as other aspects of our behavioral response involve the repression of rudimentary reflexes, our thinking involves the inhibition of associational activation and reflexive reactions. The “more advanced” cognitive functions can take place only because the simpler, less resource-intensive, and faster functions are prevented from initiating responses before them.

In the wrestling match between the modern functions and the ancient ones they try to control, the more subtle and advanced features are at a distinct disadvantage. Which brings us to the next post.

Thoughts on Charles Bonnet Syndrome

Possibly the most important lesson that should be taken from the story of Charles Bonnet Syndrome (CBS) is that correlation implies nothing about causation and even less about etiology.

Patients with general dementia, Alzheimer’s, or strokes were far more likely to communicate seeing things that weren’t there. So if you merely look at reported population statistics, visual hallucinations would seem to be associated very strongly with brain damage.

But the real problem was that such patients were much more likely to be uninhibited in talking about their hallucinations. Older people with loss of sight but properly-working minds feared the consequences of mentioning what they saw to anyone – at least partially because of the existing association in physicians’ minds between visual hallucinations and senility – and so said nothing. And so there was no awareness of the true rate of the phenomenon.

If you look at the actual statistics – the ones collected once the stigma of CBS was reduced, physician awareness increased, and elders gently but insistently questioned – then there’s no particular association between senility or brain disease and the hallucinations.

Remember – it took hundreds of years for the condition to even be mentioned in the English language, despite all of the people who must have experienced vision loss and CBS in that time. Despite all of the physicians who must have aged or had eye damage and suddenly experienced it themselves.

The Absence of Evidence

In 1769, the Swiss naturalist and philosopher Charles Bonnet noticed that his nearly-blind grandfather reported vivid hallucinations of things which he knew to be non-existent. As his own sight worsened with age, Bonnet began to experience similar visions: birds, men and women, strange plants, and architectural patterns came and went. As he was the first person to describe this phenomenon, it was named after him: Charles Bonnet Syndrome.

The syndrome did not enter the English medical literature until 1982. Most physicians and nurses had never heard of it, and there was little literature about it even in non-English journals. Millions of people had seriously-impaired vision without experiencing CBS, so clearly it must be a very rare and unusual condition. If elderly people reported seeing things, it was considered to be overwhelmingly likely to be a sign of dementia or psychosis.

Until doctors actually sat down and questioned patients as to whether they had seen peculiar things, with no suggestions that they were insane or demented. Then it was recognized that more than half of all people with impaired vision, particularly those whose loss of sight occurred suddenly, experienced CBS.

Most of them were afraid they’d be considered mad if they mentioned what they saw to people, only about a third revealed the hallucinations to anyone, roughly a third lived in terror that they were losing their minds, and the vast majority did not tell their physicians. Unless actively encouraged to be honest, and given assurances that seeing things wasn’t a sign of insanity, many people simply denied their experiences if asked.

People who had strokes, were developing Alzheimer’s, and so on were indeed more likely to report having visions. Not necessarily because the visions resulted from neurological impairment, but because they were less likely to possess the contextual awareness of others and their expectations and repress mention of them.

This is a superb example of how important it is to always keep in mind the distinction between reality and our knowledge of it. Absence of evidence is not evidence of absence. Relying upon our collective knowledge base is often useful, but it prevents us from checking that base against reality and correcting it when wrong.

If you assumed that CBS must be rare and unusual, because there was no recognition of it in the literature and doctors knew nothing about it, you’d have make the implicit assumption that physicians’ knowledge was valid and complete. In reality, it wouldn’t be – not through any fault of the doctors, mind you. But you’d be ruling out in advance the possibility that the system itself was lacking, and so wouldn’t have been able to notice the problem and correct it.

Only once physicians actively considered the possibility that they were missing something, and sought out evidence from the patients themselves instead of what the profession thought it knew, was the gap between understanding and reality closed.

The Purpose of Power is Power

You cannot avoid the interplay of politics within an orthodox religion. This power struggle permeates the training, educating, and disciplining of the orthodox community. Because of this pressure, the leaders of such a community inevitably must face that ultimate internal question: to succumb to complete opportunism as the price for maintaining their rule, or risk sacrificing themselves for the sake of the orthodox ethic.

- from “Muad’Dib, The Religious Issues” by the Princess Irulan

- Frank Herbert, “Dune”

I’ve been thinking a lot lately about the nature of power. What it’s good for, what its limits are.

It’s increasingly clear to me that power, used in ways that are not compatible with the maintenance and continued existence of that power, tends to expend itself. Power that persists usually concerns itself with itself and its perpetuation. But there are inherent trade-offs that cannot be avoided.

Wealth is a tool of freedom, but the pursuit of wealth is the path to slavery.

- Frank Herbert

What is power to be used for? What goal is the end to which power is the means? Those who have purposes for their power will sometimes find that purpose in conflict with the perpetuation and increase of the power itself, and so must choose. Those whose purposes include the use of power in a certain way will face even more conflicts and more choices. But those who seek power only for the purpose of possessing and exercising power will not be conflicted, and will be forced to no difficult choices.

Maintaining a democratic system and keeping it functionally in touch with reality is an example of having standards for the use of power. Demagogues and popular tyrants – the most obvious threats to any democracy – are examples of seeking power for power’s sake.

I don’t think this is a conflict we can win.

Watching the Watchers

TGGP asks whether it would be better to have the FDA regulate surgeries instead of drugs.

Even if it were surgeries as well as drugs, I’d say the answer is ‘no’. Why?

Many people do not realize that the FDA doesn’t actually perform any testing of drugs. It merely examines the reported results of tests submitted to it. Who performs the studies of the drugs? Why, the pharmaceutical companies who developed and are hoping to market them, of course.

Does that strike you as a conflict of interest? It should.

The biggest problem with this arrangement is that there’s really nothing preventing the drug makers from only reporting positive results. Yes, there are rules against that, but without keeping tabs on what the companies are doing or granting the authority to discipline violators, they’re paper tigers.

If you go out and look for them, you’ll find that there have been many cases of a drug that appeared to be useful in the experiments submitted to the FDA that turned out to be nigh-useless in actual clinical use. Given a drug that doesn’t appear to be doing better than placebo, you can always just perform lots of tests and then keep the ones in which the drug happens to do better than the control. What about the scientific ethics of the researchers paid to perform the tests? If they get a reputation for being ‘troublesome’, they’re likely to be let go – and find it difficult to obtain employment elsewhere in the pharmaceutical industry. Drug makers can sit on the negative results and never let them see the light of day.

The problem isn’t so much that what the FDA does is bad, but that it’s inadequate and horrifyingly ineffectual – and that, by existing, it helps create the impression that drugs are being properly vetted. People presume that drugs that make it through the process are valuable, that in-all their benefits are substantial and outweigh their costs. That trust is misplaced.

Irreducible Stupidity

By now, there’s a good chance you’ve already heard about the McWhorter-Behe diavlog on Bloggingheads.tv, and its removal for unknown reasons.

So I have very little to say about that.

I would like to point out this post over at BH’s thread regarding the video removal, written by someone with the handle “IRQ Conflict”:

As for entropy. When is the last time you saw an organism gain information and order rather than lose it with time?

Gee, he’s got us there. [takes bite of sandwich] [swallows] I don’t recall any biological organism ever increasing in order and energy in any way… [takes bite] bwff I suppos thair mai bee [swallows] – sorry, shouldn’t type with my mouth full – but I suppose there may be some way in which living creatures might be able to increase their energy. [takes sip of beverage] I wonder what that might be?

I am reminded of the infamous Internet exchange in which a Creationist ridiculed the laws of thermodynamics by noting that they imply that there’s some large source of negentropy pumping energy and order into Earth… thus suggesting that, in his colossal arrogance, he did not permit himself to recall the existence of the Sun.

First, Do No Harm (Part 2)

What does Dr. Vertosick suggest is the cause of such snafus as the EC-IC bypass?

As the EC-IC bypass affair illustrates, experimental operations can jump into the medical mainstream long before anyone establishes their efficacy – or even their safety. Although ego and greed help keep unproved procedures in the operating room, it’s bad scientific judgment that puts them there in the first place.

Before new drugs can be marketed, they undergo three levels of testing. First, volunteers are given the drug to see how toxic it is and how well it’s absorbed and tolerated by the body. (This leads to specifically unexpected yet globally inevitable tragedies every once in a while, most particularly when one of the test subjects happens to be in a minority that has serious reactions to a drug that hadn’t been previously observed.) Then, once the obvious risks of the drug are known, it’s given to ill patients to see if they do better on it than previous, retrospective patients. Finally randomized trials are conducted to compare treated and untreated patients directly. Only when this last stage has been successfully completed is a drug considered for approval.

There is no such regulation on surgical procedures; although the FDA regulates surgical devices, it has no jurisdiction over surgeries. There is no legal obligation for surgeons to test therapies with the third stage of randomized trials. And since they can charge for any surgery, surgeons have no financial reason to put their therapies through expensive and difficult examination.

It is not unusual for surgical procedures to be widely implemented without rigorous testing – it is in fact quite standard. Vertosick offers the example of spinal fusion to treat back pain caused by degenerating disks. The disks can be removed “in a simple, two-hour operation”, or surgeons can remove the disk and implant a steel plate, which in theory helps to stabilize the spine, a procedure which is more involved and incidentally costs two to three times more. As Vertosick points out, “there’s no evidence that it’s any more effective than the simpler procedure”.

The reasons why so little testing is done are legion. It takes lots of money and effort to conduct randomized trials, patients don’t want to be assigned to “nonsurgical” groups – they want to be ‘treated’ when they go to surgeons, and that usually means undergoing surgery – and they go elsewhere for the surgery if they can’t get it in the study. Many doctors are not qualified to evaluate the statistical results produced, and often don’t believe the results when they arrive. Some of the critics of the NINCDS study initially complained that the best candidates for the surgery left and had the surgery elsewhere when they had been assigned to nontreatment, for example. But doctors often simply ignore studies, even when there aren’t potential confounding issues like that one, simply because they want to stick with what they “know” works.

What is known now is that the retrospective studies used by Yasargil did not accurately reflect the rates of stroke at that time. Too little was understood about how people’s health had changed between the time those studies were done and the time EC-IC was first being tested, and our assumptions that the two were comparable turned out to be wrong. As a result, Yasargil reached the wrong conclusion.

He didn’t even do something wrong – or at least, wronger than usual in medicine. He was in fact more careful than most such innovators, and certain more so than those who adopted his ideas without subjecting them to any testing at all. He was merely tragically wrong.

Other, equally tragic mistakes happen not through bad luck, but through incompetence, willful ignorance, and arrogance.

Proving the Negative

There are all sorts of things that people do which annoy me. But what gets on my next-to-last nerve is when someone repeats a cliché that is grossly incompatible with even rudimentary logic. “You can’t prove a negative” is probably the worst offender on the list of such statements.

Every time any statement is disproven, its negation is proven. Negative statements are made and proven constantly – and technically, what constitutes a “negative” isn’t even well-defined. Any statement about any property can be considered a “negative” one from a properly-chosen perspective.

Trying to establish any statement through a process of elimination of its alternatives tends to be tedious and, in practice, unworkable once the space of possible alternatives becomes large enough. Attempting to demonstrate that there is no golden needle in a stack of hay by picking up and examining each and every straw will work. Eventually. It’s hard, certainly, but not impossible. Even if the haystack were infinitely large, the strategy would still work – given infinite time.

When people say “you can’t prove a negative”, it often seems that they’re referring to situations with arbitrarily large haystacks and attempts to rule out a golden needle. Picking through the straws one by one probably isn’t the best available strategy, certainly. But they should just say that instead.

Bertrand’s Example and the Ship of Fools

Bertrand Russel once offered a famous example. Suppose there were a merchant shipper who thought that one of his ships might be less than seaworthy. Instead of investigating and testing the integrity of the vessel, this man instead avoids thinking about the possibility and anything that might remind him of it. He does not look into the ship’s seaworthiness. He does not tell the crew, or suggest to them in any way that the ship might not be safe. He sends it on a voyage, loaded with goods, and frets – not at the chance that he ship could flounder, but with the difficult problem of keeping himself from acknowledging that the voyage might be lost. Every effort is bent towards maintaining his belief and conviction that the ship is seaworthy and that nothing untoward is likely to happen.

Has this man done wrong? Yes, Russell argues. But didn’t he have a sincere belief that the ship was safe? Yes, but that belief was based on nothing – and the man worked to maintain it despite having no valid reason to believe that his conviction was valid. He worked so hard to maintain it precisely because he suspected that conviction was invalid, and wished to avoid the inconvenience and financial loss that being forced to give up that conviction would entail.

If the ship sinks at sea, that in itself does not prove that the ship’s structural integrity was at fault. But let us imagine that a sailor escapes from the wreck with his life and reports that boards gave way halfway through the trip, without a collision or snag. That would mean that the merchant shipper was even more guilty of before, of an additional and even greater crime.

But even if the ship comes home safely, the merchant is guilty. Of recklessness, of endangering the lives and goods of others without their knowledge, of seeking to preserve his belief at the expense of veridical integrity.

Anyone who knew that the merchant behaved this way, and yet chose to sail with him regardless, would be a fool. Anyone who trusted any position of the merchant’s, no matter how sincerely expressed, would be a fool.

It’s easy to agree with Bertrand in the abstract. It’s rather more difficult to disagree with him when facing an issue you don’t want to look into, or on which you have a specific position you don’t wish to give up.

There are many on the Ship of Fools, but few who will admit even to themselves that they sought to sail upon it.

The Accidental Discovery of Ambroise Paré

In the sixteenth century, exciting new advances in technology led to horrifying developments in warfare. For the first time, physicians were forced to confront injuries caused by bullets and shrapnel. Not themselves, of course – surgery and the treatment of wounds were beneath the dignity of physicians, who hired surgeons to deal with such butchery.

In the 1500s it was commonly believed that gunshot wounds were poisoned. We now recognize that high-velocity projectiles often carry foreign matter into a wound and contaminate it, leading to infection and possibly gangrene. Several hundred years before the germ theory of disease was widely adopted, however, everyone was completely ignorant of the nature of infection. Since bullet wounds were ‘poisoned’, they of course had to be cauterized with oil. Burning-hot oil.

During the Siege of Turin, 1537, a young barber-surgeon named Ambroise Paré was overwhelmed by a flood of casualties, so much so that he ran out of oil for cauterization. In desperation, he made a mixture of egg yolks, oil of roses, and turpentine, applied this concoction to the patients, dressed them with clean bandages, then left them for the night. Upon returning next morning, he observed that the patients so treated hadn’t succumbed to sepsis as he feared and were doing well, while the cauterized patients were in agony. Curious, he observed the progression of healing and noted that the non-cauterized patients not only were in significantly less pain but actually healed faster than those scaled by hot oil.

Shocked, he realized that his duty was not simply to provide treatment but to minimize or avoid suffering, and that the standard therapy that everyone used was both tortuous and inferior to simpler methods. His findings were not warmly received by physicians and surgeons at the time, and although his work was well-appreciated by both the military and the monarchy it was rejected for decades by the medical establishment.

Paré went on to challenge many of the surgical dogmas of the time, and although he was often ridiculed for his positions, we recognize them today as obvious. He was the first person to refrain from castrating men who needed surgery for hernias, for example – the belief at the time that eunuchs did not develop hernias was a simple superstition. More generally, he did not resort to surgery unless he felt it was absolutely necessary – a major break with the attitudes of his time. Sadly, the modern era has reverted back to the attitudes of Paré’s contemporaries in some ways.

His most famous saying, which despite being reduced to a cliché through overuse remains in frequent use today, is

I dressed the would, but God healed it.

The theism implicit in the saying is outmoded, but in the real world, humility never goes out of style.