This is a continuation of my discussion from Part IV of some of comments on the essay Learning from Animals: Evolutionary Medicine with a Twist by Dr Harriet Hall and essays on the conviction of Italian geologists.
Science can predict some events with incredible accuracy. If it couldn’t, it would not be possible to fly airplanes, land on the moon, send probes to Mars, or predict the decay of radioisotopes. Other things can’t be predicted as precisely because such predictions are based on stochastic models that have a lot of random noise and factors that we don’t yet understand. Some things, like earthquakes, can’t be predicted very well at all because of high variability and our lack of understanding of key mechanisms by which they occur.
So it appears that David Gorski, MD, PhD, does indeed understand predictive modalities as opposed to predictions made from hypotheses. In light of this admission, note what Gorski stated in comments on the Hall essay, Learning from Animals: Evolutionary Medicine with a Twist:
Indeed. It’s ridiculous to lump all animal models into a “class.” There are too many such animals used for too many purposes. Some animal models are more reliable than others. Such generalizations reveal more about the good Dr. Greek than and his ideology-inspired lack of nuance than they reveal about the actual utility of animal models.
On the contrary, classifying uses allows science to be more precise and science is all about being precise. One breaks down science into physics, chemistry, biology etc even though there are far more topics under each category (and some overlap) than there are ways animals are used in research. Whenever you see a scientist coming out against breaking items down into classes, be suspicious.
As I have stated, there are various ways to classify animal models depending on what you are trying to accomplish with the task. Animal modelers themselves use, and advocate for, such classification. (For example, see (Hau 2003).) Moreover, classifying animal models as “predictive” is not my invention. Hau and others also categorize animal models this way. Having no real defense against classifying animal models, Gorski then resorts to the ad hominem. Finally, whenever someone attempts to discredit his opponent by calling into question why his opponent has this position he is committing the appeal to motive fallacy.
In reality, the use of animals as “predictive models” that Dr. Greek keeps harping on is not nearly as large a slice of the pie when it comes to using animal models as he seems to think it is.
The percentage of animals used as predictive models or the percentage of grant money that goes to such models is immaterial to whether the models are in fact predictive. This is an example of the straw man fallacy in that Gorski is attributing a position to me that I do not hold and a red herring in that it brings a new and irrelevant detail into the discussion.
I’ve pointed this out before elsewhere. . . . Similarly, the various “omics” models are now producing a flood of data that we don’t yet know how to analyze in a sufficiently detailed, robust, and predictive fashion to use them to predict human drug responses, much less human toxicities and responses that aren’t obvious.
This is called a non sequitur in addition to being a straw man. I never claimed omics etc are predictive for humans in the current state of development, merely that animal models are not. I have said that human-based research and testing will be more predictive than animal-based and I stand by that claim. But such a claim is relative not absolute, meaning that even if modality A is better at predicting outcome X than modality B, neither may qualify as a predictive modality. They both may fail.
The non sequitur part is that there is no relationship between the predictive value of animal models and the predictive value of omics etc. One cannot claim animal models are predictive based on the fact that omics etc are not predictive. But personalized medicine, based on genomics etc is currently being used for some diseases and drugs and will be the way medicine is practiced in the future.
More importantly, Dr. Greek appears to be arguing that we should abandon current tools in favor of tools that don’t exist yet, such as computer models cell culture models that predict human response better than animal models and, of course, the various “-omics” of personalized medicine, which have not as yet been validated as reliable predictors of human responses.
Another straw man. Gorski has either not read my writings or is purposefully being duplicitous. I have stated many times that Pharma has few if any predictive technologies. Pharma has also said this. What I am saying is the equivalent of suggesting we abandon homeopathy as an HIV prevention strategy even though we currently have no HIV vaccine.
Why does he do this? Because he’s an animal rights activist and believes that animal research is inherently wrong. (If I’m mistaken about this, then perhaps Dr. Greek would describe for us when animal research, in his opinion, morally justifiable.)
Number 1. Appeal to motive fallacy. Number 2. This is another non sequitur. My position and the position of AFMA concerns the science of animal models, irrespective of the ethics. Contrary to the position of Dr Gorski and numerous animal activists that criticize my blogs, one really can analyze a single aspect of an issue. Number 3. I am no more an animal activist than I am an abortion activist. I have positions on each but do not campaign on either. So while I am an animal rightist philosophically, I do nothing that would qualify one as an activist for my philosophical position.
Number 4. From a science perspective, I have listed many examples of the use of animals in research and science in general that are viable. My philosophical position that sentient animals should not be used in any research that is not of potential direct benefit to them (and or that humans would also be allowed to participate in), obviously does not impinge on my ability to recognize scientific validity.
Number 5. In order to avoid a real analysis of my position that animal models are not predictive modalities for human response to drugs and disease, he must change the subject, or attack character, or use fallacies. What Gorski will not address speaks volumes about his position.
His arguments, particularly his article on personalized medicine, are not persuasive to someone who’s actually doing biomedical research and knows the issues involved in personalized medicine.
Argument from authority and, exactly which aspect of my article is “not persuasive?” Notice that Gorski is surprisingly nonspecific in criticisms where he could be very explicit. As I said, this is why the prediction issue must be addressed before other controversies in animal-based research can be discussed. If this issue cannot be decided, based on the overwhelming evidence and theory available, then one side is simply being disingenuous. For example, Dr Hall commenting on her blog:
Let me see if I understand this correctly. Dr. Greek is willing to discuss whether studying diabetic animals predicted human responses to insulin, but only after we have agreed with him that studying animals doesn’t predict human responses to anything. . . .
Number 1. Her statement that I claim: “studying animals doesn’t predict human responses to anything,” is not fact-based. I have explained many times that at levels of organization that are lower than where disease and drug actions occur, the reactions of animals can be identical to humans. So can yeast in many cases. Number 2. As I said above, unless an obvious reality, like the failure of animal models be qualify as predictive modalities for human response to drugs and disease, can be agreed upon, discussing less clear issues is unlikely to be enlightening. Number 3. Dr Hall apparently really does not understand the difference between predictions from hypotheses and whether a modality (i.e. minor earthquakes) can qualify as a predictive modality (for a major earthquake). As she is an MD, this is unfortunate. Finally, this is an example of why the meaning of predict must be nailed down. Single instances of correlation do make a modality predictive. If it did, astrology and minor earthquake activity would be predictive modalities.
Dr Hall continues:
You are not entirely wrong. That’s the pity. You have some good points to make, but you undermine them by going too far. I would be more disposed to take you seriously if you didn’t say so categorically that animal models are “never” predictive of human responses. That is demonstrably not true.
Still a straw man. Throw a frog out of an airplane and you will see the same gross disruption of anatomy that you see if a human is thrown out. That does not mean frogs react to disease the way human react.
Dr Gorski again changes the subject from animals as predictive modalities:
Nor is anyone going to really learn about research, basic or translational, without actually doing basic or translational research. There is no way anyone is going to learn how to do research via blogs or just reading about it in textbooks or journals.
Straw man. First, one does have to be a surgeon in order to evaluate surgical complication rates. If surgeons A, B, C, D, and E have an infection rate of 3% for routine scheduled mastectomies and surgeon F has an infection of 33%, ceteris paribus, surgeon F is doing something wrong. One does not need to complete a 6-year surgery residency to figure this out.
Second, my position is not about “how to do research.” I have stated many times that the methodology employed by animals-based researchers is usually sound. (I will address this more completely in a future blog, as there are some problems even with methodology.) I am addressing the premises on which the research is based. As this is under the category philosophy of science and as most scientists are ignorant in this area, most do not understand the difference between criticizing methodology and criticizing premises.
Third, I have also been critical of explaining science via blogs and have suggested issues like this be debated in the scientific literature or in public venues like universities. Gorski et al have refused to participate. This is also informative.
Indeed. Dr. Greek boxes himself in with his steadfast refusal to admit to the existence of even one good predictive animal model, as if doing so would completely invalidate his overall argument.
Actually, I have stated that there must be some aspect of drug development where animal models are useful if not predictive just based on the great number, and diversity, of situations where animal models are used. In addition, I have also stated that animal models can be predictive but not for drug and disease response as, at that level of organization of a complex system, small difference can have profound effects. But, having discussed this many times, I am suspicious that Gorski is again pretending there is no difference between hypothesis-generated predictions and predictive modalities. “(E)ven one good predictive animal model” would have to meet the criteria of having a high PPV and NPV and I do not know of any. Moreover, given current knowledge of evolved complex systems such a model is very unlikely in terms of outcomes at higher levels of organization (such as response to drugs and disease).
Dr Hall stated in her essay Learning from Animals: Evolutionary Medicine with aTwist: “There is only one medicine and one biology. Terms like “alternative medicine” or “veterinary medicine” create false dichotomies. There is much to be gained from a regular collaboration between people doctors and animal doctors.” I addressed Dr Hall’s inspiration for this topic in my review of Zoobiquity: What Animals Can Teach Us About Health and the Science of Healing, as this attitude is representative of the problem. In pre-Darwin days, this attitude was somewhat justified: the Creator used a common theme in creating life so studying dogs could inform one about human disease. The father of modern-day animal-based research was a creationist and this influenced his position. (Bernard 1957; Elliot 1987; LaFollette and Shanks 1994) Note that I did not say that it invalidated his position. Only an analysis of the position can determine whether it is viable. Drs Hall, Gorski etc have assumed my position is wrong based on the fact that I am philosophically an animal rightist. Were they to actually prove that my position was wrong they could then speculate as to why I was wrong, just as I have pointed out that Bernard’s position as a creationist unduly influenced his philosophy of science. But as Gorski et al have not addressed my position much less falsified it, such speculation is merely an ad hominem.
Skeptic and feminist Rebecca Watson authored an article in Slate, It Stands to Reason, Skeptics Can Be Sexist Too, where she stated: “I also believe that old line about sunlight being the best disinfectant. Ignoring bullies does not make them go away.” I have stated that vivisection activists are bullies because they misrepresent science in order to fund their research or stroke their egos. They are bullies because they threaten society with: “Your dog or your child,” even though this is a false dichotomy and they justify their behavior with fallacies and lies. As is the case with most bullies, they will not fight fair. Call a bully out and he will probably find a way to avoid the fight, as most are cowards. Likewise, Gorski et al avoid exposing their claims to the sunlight of a debate, be it in the scientific literature or a university.
I encourage you to read Learning from Animals: Evolutionary Medicine with a Twist along with all the comments on the essay.
Bernard, Claude. 1957. An Introduction to the Study of Experimental Medicine. 1865. Translated by Henry Copely Greene. New York: Dover.
Elliot, P. 1987. "Vivisection and the Emergence of Experimental Medicine in Nineteenth Century France." In Vivisection in Historical Perspective, edited by N Rupke, 48-77. New York: Croom Helm.
Hau, Jann. 2003. "Animal Models." In Handbook of Laboratory Animal Science. Second Edition. Animal Models, edited by J Hau and G.K. van Hoosier Jr, 1-9. Boca Rotan: CRC Press.
LaFollette, Hugh, and Niall Shanks. 1994. "Animal Experimentation: The Legacy of Claude Bernard." International Studies in the Philosophy of Science no. 8 (3):195-210.