Mar 05 2007
I sense that the skeptical movement has crossed a line in the last few years; we have not only gotten the attention of the woo woo crowd but we’ve shaken them up a bit – at least enough so that they feel they have to go on the attack against skeptics and skepticism. To which I say – bring it on. I’m always up for an intellectual scuffle. It’s a good way to get to the core of the relevant issues and logic.
I was recently sent one article in particular, authored by Winston Wu (a self-proclaimed researcher and explorer of the paranormal), that seeks to systematically dismantle the pillars of scientific skepticism. The aptly named Mr. Wu, however, systematically misunderstands and misrepresents skepticism, but in so doing he exposes many common misconceptions about skeptical principles. So I will use my blog this week to set the record straight.
It isn’t clear from his writing whether or not Wu understands what Occam’s razor really means. He gives both correct and incorrect versions of it without saying what he is advocating. The most common misconception about Occam’s razor is that when there are multiple possible explanations for a phenomenon the simplest one is most likely to be true. This is not quite right.
Wu correctly quotes William of Occam who said, “Entities should not be multiplied unnecessarily.” Put another way, when considering multiple possible explanations the one that introduces the fewest new assumptions (multiplying entities) is to be preferred. This is very different than “simplest.”
Let me give you an example from medicine. If a patient presents with three simultaneous symptoms, A, B, and C, Occam would not favor the hypothesis that the patient contracted three diseases all at the same time. Occam would favor one disease, if it could explain all three symptoms. However, perhaps whatever is causing symptom A also precipitated the cause for symptom B, and symptom C is caused by a common chronic disease that was exacerbated by A. This is a more complex scenario, but it still is only introducing one new disease, and everything flows from that, so it is still complying with the principle of Occam’s razor.
Now let’s look at alleged alien phenomena. Proponents often say that a host of phenomena, including abductions, sightings, cattle mutilations, and crop circles, can be explained with one unifying cause – alien visitation. Therefore, they falsely argue, Occam’s razor would favor this one explanation over a separate explanation for each phenomenon. But this is wrong. If we say that abductions are caused by hypnagogia, sightings by misidentifications, cattle mutilation by scavengers, and crop circles by hoaxes, we are using all established known causes (of course I am oversimplifying for the purpose of argument). But alien visitation is a huge new assumption that has not been independently established, so in this case the more complex explanation has fewer assumptions and is preferred by Occam’s razor.
In the two examples above we also see that the number of new assumptions is not the only factor, but also the “size” of the new assumptions – i.e. how likely or unlikely is the new factor being assumed; how much of a change to our world model would it require? So what Occam is advocating, at least as an intellectual principle, is that one should strive to minimize the overall assumption burden when considering various competing hypotheses.
Wu has four objections to the skeptical use of Occam’s razor: 1) it does not apply to the paranormal; 2) “simpler” is relative and therefore difficult to operationally define; 3) Occam’s razor is only a rule of thumb yet skeptics use it as an absolute; 4) skeptics use Occam’s razor “as an excuse to insert false explanations over paranormal ones.”
The first objection is just silly, and Wu makes no rational defense of it. Logical principles apply to any argument, regardless of the subject matter. What he is advocating for is one set of scientific and logical principles for “natural” phenomena, and a suspension of these rules for the paranormal. This is, in effect, what many paranormal proponents do – they inject wild new assumptions as needed in order to explain phenomena or make their fanciful theories work. So it is no surprise that they do not want to be criticized for this and would therefore argue that Occam’s razor does not apply to them.
It’s possible that what Wu is saying is that Occam’s razor does not apply to the unknown – because it’s unknown, so of course we have to make new assumptions to explain it (or something to that effect). But Occam’s razor does not mean that we cannot make any new assumptions, just that they should be minimized. If an alternative with fewer or smaller assumptions exist, explore that first. Scientists make new assumptions about reality all the time, but they try to take it one step at a time – testing one small new assumption then going onto the next. With each new assumption the probability of being correct goes down, so it doesn’t make sense to waste time and effort researching a hypothesis that requires a string of large assumptions. Scientists don’t like to play the lottery.
The second objection, that “simpler” is hard to define, is not relevant for the reasons I explained above (and is evidence that Wu does not, in fact, understand Occam’s razor). It may be difficult to quantify how much of a new assumption an idea is, but that is irrelevant. That is what scientific judgment is all about – having a sense of what is likely to be true or untrue based upon our current understanding and principles of logic. Saying that ESP is real requires many new assumptions – that there is a medium of information transfer undetectable by modern instruments and not predicted by our physical models, that this medium does not obey any of the usual physical laws we are used to dealing with, etc. It also requires assumptions to explain why ESP is so hard to verify in a lab (the shyness effect, for example). The number of assumptions can be quantified, and the size of the assumptions can be judged from both logic and our current fund of scientific knowledge.
The third objection, that skeptics use Occam’s razor as an absolute, is somewhat of a straw man. It is true that Occam’s razor is often misunderstood and may be abused in arguments. It is further true that Occam’s razor is not an absolute – the hypothesis with the fewest assumptions does not have to be true. Rather, it is about probability. It is possible that my patient may have independently contracted three rare diseases, it’s just damn unlikely.
But Wu’s point is a straw man because what he is arguing against is not the position of skeptics. Wu himself offers no evidence that “skeptics” misuse Occam’s razor in this fashion. Of course, no one can account for every self-proclaimed skeptic, but a quick look at the skeptical literature shows that prominent skeptics, those who write articles and have blogs, generally have a good understanding of Occam’s razor and employ it appropriately. If you Google “Occam’s razor” the first hit is the wikipedia entry, which is an excellent summary. If you Google “Occam’s razor, skeptic” the first hit is for the skeptic’s dictionary entry, which is also quite good and does not commit Wu’s imagined fallacy. We have also written an article about Occam’s razor for the NESS.
Finally, Wu charges that skeptics are simply using Occam’s razor as an excuse to favor “false” explanations. If I try to generously extract anything cogent from his explanation it seems that he is trying to say that skeptics will arbitrarily insist upon a non-paranormal alternative explanation for an alleged paranormal phenomenon by declaring it “simpler,” without justification. His caricature of a typical skeptical thought process, however, betrays just another straw man, and the fact that he clearly does not understand skeptical thinking.
Wu writes: “For example, if someone had an amazing psychic reading at a psychic fair (not prearranged) where they were told something very specific that couldn’t have been guessed by cold reading, skeptics would start inventing false accusations such as: “Someone who knew you must have tipped off the psychic in advance”, “A spy in the room must have overheard you mention the specific detail before the reading”, “You must have something in your appearance that reveals the detail”, “You must have remembered it wrong since memory is fallible”, etc. Even if none of these accusations are true, skeptics will still insist on it simply because it‚s the simpler explanation to them.”
First, he is making a huge assumption in his premise that “they were told something very specific that couldn’t have been guessed by cold reading.” As far as I know, no alleged psychic has ever given a performance that could not be duplicated by admitted mentalists using cold reading techniques – so Wu is simply assuming his conclusion in this statement.
Second, his use of the word “must” in characterizing the skeptical response betrays a misunderstanding of science and skepticism. It is not a subtle distinction to say, rather, that the skeptical approach is to consider all alternate hypothesis – that the appearance of a psychic reading can also be produced by a “hot” reading (where the reader had prior knowledge of the target or used tricks to obtain knowledge), or that the reading was due to standard cold-reading techniques, including visual or other cues, or that the subject did not accurately recall the details of the reading.
Now, if we properly apply Occam’s razor to the above, we can say that it is well established that con-artists in the past have secretly obtained knowledge about their marks in order to fake supernatural knowledge, and doing so would not be difficult or require any new assumptions about reality. Further, it is well established that skilled cold-readers can seem to obtain very accurate information about their subjects – no new assumptions required here. In fact, the best readings I have ever seen by far were by mentalists, not alleged psychics. It is also well established that human memory is fallible, and in fact comparisons have been made between a subject’s memory of a reading and a transcript of the reading documenting extreme flaws in the subject’s memory. So no new assumptions are required for any of the non-paranormal hypotheses. The paranormal hypothesis requires great new assumptions about the very nature of reality. So Occam’s razor, properly applied, strongly favors the non-paranormal hypotheses (and it is therefore no wonder that paranormal proponents wish to subvert Occam’s razor with misdirection and logical fallacies).
But it is true, as stated above, that this does not mean that one or more of the non-paranormal explanations must be true, or that psychic ability cannot exist. It just means that they are vastly more likely. But the implications of this for proper scientific inquiry should be clear – these alternate non-paranormal possibilities should be carefully eliminated before accepting the paranormal explanation as probably true. This is the follow up to Occam’s razor where we do actual experiments or investigation. It turns out that no one has been able to reproduce a psychic reading under controlled conditions that rule out trickery, that cannot be explained by cold reading, and that do not rely upon subjective memory for evaluation.
If the paranormal explanation were true against all odds, it should be possible to eliminate the alternate hypotheses favored by skeptics, but believers have not been able to do so. Some proponents, unwilling to make the obvious conclusions dictated by science and logic, commit further atrocities against Occam’s razor by adding still more assumptions to explain away the inability to demonstrate psi under controlled conditions – such as the ad hoc nonsense that the belief of the experimenter affects the functioning of psi.
The rest of Wu’s arguments against skeptics are similarly flawed and fallacious. I will cover several more of them this week – discussing good skeptical habits by way of exposing the straw men of our detractors.
Leave a Reply
You must be logged in to post a comment.