Skip to main content

Prediction Versus Everything Else (Part II)

Continuing the prediction theme from yesterday (or maybe earlier today, it is getting late) . . .

Prediction has been, in large part, ignored by scientists except for those in Pharma and by skeptics fighting pseudoscience like astrology, psychics, and complimentary and alternative medicine. Douglas:

It has become canonical among philosophers of science that explanation is a central goal of science. Many philosophers of science also recognize prediction as a central goal of science, and often they are paired together as two potentially competing goals of the scientific enterprise. This has been apparent in various works over the past half century . . . Despite the fact that most philosophers acknowledge the general importance of prediction for science, the vast majority of the intellectual focus between the two goals rests on explanation. Prediction is rarely a topic in its own right, appearing mainly in discussions of confirmation, realism, and other topics. It has been this way for over 40 years. (Douglas 2009)

Granted, explanation is important. Albert Hofstadter stated in 1951: “Prediction and explanation are the two main functions of scientific knowledge” [(Hofstadter 1951) p339]. However, at times at least, prediction is more important. Further, if the topic under discussion is prediction it is not fair to claim animal models are predictive when in fact what the vivisection activist is using them for, or the new topic he is bringing into the discussion, is something other than prediction.

So, the most common discussions about prediction today come from skeptics battling pseudoscience and Pharma seeking to predict human response to drugs. (These are not the only sources but they are the most common, in my experience.) Pharma knows what predict means because they lose money when their models fail to predict human response. Nothing makes a person acknowledge reality better than losing lots of money when his version of reality differs from the real deal. That is why Pharma scientists say things like the following.

The executive director for cancer research at Merck Research Laboratory 1997:

The fundamental problem in drug discovery for cancer is that the [animal] model systems are not predictive at all (Gura 1997).

An editorial in Nature Reviews Drug Discovery 2005:

Clearly, one part of the problem [of drug research] is poorly predictive animal models, particularly for some disease areas and drug classes with a novel mechanism of action, a topic we continue to cover in our ongoing 'Model Organisms' series. But arguably the best 'models' for drug discovery are human subjects and as the need to have proof of concept or mechanism for a drug before moving on to larger, more costly clinical trials has never been greater, more big-pharma companies are now embarking on programmes in experimental or translational medicine. (Editorial 2005)

Chabner and Roberts:

Fewer than 10% of new drugs entering clinical trials in the period from 1970 to 1990 achieved FDA approval for marketing, and animal models seemed unreliable in predicting clinical success . . . (Chabner and Roberts 2005)

The above examples can be easily multiplied.

Skeptics clearly understand the definition and use of prediction. (There are numerous books about this topic, see almost any skeptic-related website for a selection.) They routinely use the scientific definition of prediction to debunk astrology, people claiming to be psychics, and other pseudoscience. Astrology, psychics, and various alternative medicine modalities have been tested for their predictive and or curative powers and found lacking. Astrology is not a predictive practice even though it occasionally forecasts the future neither does an occasional correct guess make someone a psychic. Despite what vivisection activists’ claim, one right answer does not make a practice predictive or even useful. When judging claims by vivisection activists, substitute the phrase “animal model” or “animal research” or whatever similar phrase is being used, with astrology and see how well the claim holds up. For example, in response to my blog One Example is Not Evidence to the Contrary, Don Earl stated:

It proves you're a bald faced liar

One example is sufficient to do that, even in the absence of the well established track record of doing exactly that on many occasions.

And Dr Ringach stated:

One example is enough

Yes, a general statement can be falsified with a single example.

So if you say "All odd numbers are prime", then all I have to do is to bring up 21 to falsify it.

When you say "The results of research from mice, basset hounds, and bonobos can't be extrapolated to people.", all I have to do is bring one example to the contrary.

I hope this is clear.

One example can indeed disprove a position or even a paradigm. But judging whether a practice or modality is predictive is not a “general statement.” Judging predictability is based on many examples or even better a consecutive series of cases in question. Like drugs or medical interventions, for example. Keeping that line of reasoning, there have been many vaccines that have prevented infection with SIV or SHIV in monkeys but none have been successful in humans. If such a vaccine were found to be effective tomorrow, that would be great news but it would not make monkeys vaccinated against SIV predictive for humans with HIV. One example does not falsify the statistical examination of many examples or a series. One aberration does not counter a hundred failures. In basic research, one success in a hundred may be considered above average, even excellent. But basic science research makes no claims to prediction and if someone is foolish enough to make such a claims, the claim can be easily falsified. (See “Is the use of sentient animals in basic research justifiable?”) Changing the subject from prediction to examples of single successes in basic research, then claiming that such successes prove animal models are predictive is a common practice for the vivisection activist. Totally dishonest, but very common.

Lets substitute astrology in the above and see how it looks.

Let’s make Don Earl’s statement read: “One example from astrology is sufficient to do that, even in the absence of the well established track record of doing exactly that on many occasions.” Do you buy that one example of success from an astrologer falsifies the claim that astrology is not predictive?

Now for Dr Ringach’s claim. "The results of astrology can't be . . . ” lets substitute predictive for people for “extrapolated to people”. I think that maintains the spirit of the statement (see below for more on this). Dr Ringach continues with “all I have to do is bring one example to the contrary.” Let’s put that in context: “All I have to do is bring one example to the contrary, where astrology did accurately forecast a person’s future.” Hmm. I am not buying that a single example of a fortuneteller or astrologist forecasting an event is enough to prove that astrology or fortune telling is a predictive practice. Nope—not buying it. Neither do skeptics who write about and criticize astrology and other pseudoscientific endeavors. In fact, anyone with any knowledge whatsoever about philosophy of science would not make such a claim.

Unless of course money were involved. Then a person might make any claim. Why, he might even claim that there is no evidence that smoking causes cancer (a claim based on research with animals (Greenstein 1954) (Northrup 1957) (Janofsky 1993) (Utidjian 1988) (Clemmensen and Hjalgrim-Jensen 1980)). Or that Enron was financially viable. Or that all the oil is gone from Gulf oil spill. Yes, money and or threats to his power can cause a man to make strange claims indeed.

When Don Earl and Dr Ringach claim a single example can falsify the position that animal models are not predictive, they are either inadequate in their knowledge about philosophy of science, statistics and so forth, or they spinning. We are all familiar with spin. Politicians live by spinning their version of events especially when the real version is at odds with their best interest.

Dr Ringach may claim that the blogger he was quoting, referred to in his statement above, said extrapolated instead of predict. But since I have stated more times than I can count that my issue is with prediction, denying that extrapolate in this context meant predict, would be a very hollow sounding defense. Especially when my blog, the one he was responding to (Animal Research: Similar is Still Not Close Enough), was all about prediction and when the following blog, One Example is Not Evidence to the Contrary, that generated the above quotes, was also about prediction. Antivivisectionists have frequently used the word extrapolate to mean predict. That is how they write and even what they scream and chant, as Dr Ringach can attest having had many protesters screaming similar things at his residence. I think it would be very disingenuous for anyone to claim that he thought extrapolate meant something other than predict. But money makes a man claim curious things.

I doubt the above will convince vivisection activists. Max Planck famously said:

A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it.

In light of all the young scientists being trained as vivisectors, I question whether the next generation will be any more receptive to truth than the present generation. In light of all the money involved, I am confident they will not be.


Chabner, B. A., and T. G. Roberts, Jr. 2005. Timeline: Chemotherapy and the war on cancer. Nat Rev Cancer 5 (1):65-72.

Clemmensen, J, and S Hjalgrim-Jensen. 1980. On the absence of carcinogenicity to man of phenobarbital. In Human Epidemiology and Animal Laboratory Correlations in Chemical Carcinogenesis, edited by F. Coulston and S. Shubick: Alex Pub.

Douglas, Heather E. 2009. Reintroducing Prediction to Explanation. Philosophy of Science 76 (October):444-463.

Editorial. 2005. The time is now. Nat Rev Drug Discov 4 (8):613.

Greenstein, J. P. 1954. Biochemistry of Cancer, 2nd ed. New York: Academic Press.

Gura, T. 1997. Cancer Models: Systems for identifying new drugs are often faulty. Science 278 (5340):1041-2.

Hofstadter, Albert. 1951. Explanation and Necessity. Philosophical and Phenomenological Research 11:339-347.

Janofsky, Michael. 1993. On Cigarettes, Health and Lawyers. New York Times, December 6.

Northrup, E. 1957. Science looks at smoking: A new inquiry into the effects of smoking on your health. . New York: Coward-McCann.

Utidjian, M. 1988. The interaction between epidemiology and animal studies in industrial toxicology. In Perspectives in Basic and Applied Toxicology, edited by B. Ballantyne. London: Butterworth-Heinemann.


Popular Video