>>> Posting number 312, dated 12 Jul 1996 18:02:46 Date: Fri, 12 Jul 1996 18:02:46 -0400 Reply-To: Discussion of Fraud in Science Sender: Discussion of Fraud in Science From: "W. R. Gibbons" Subject: Re: anyone could win Comments: To: "Paul R. Gross" In-Reply-To: <199607121920.PAA131444@faraday.clas.Virginia.EDU> On Fri, 12 Jul 1996, Paul R. Gross wrote: > Ray Gibbons disagrees, "emphatically," with Lipps to the effect > that it somehow comes right in the end. I would ask him, then, > what fraction of the standard texts in his own field, which I > take to be physiology or biophysics, is -- to his reliable > knowledge -- wrong or at least suspect. I mean, textBOOKS, not > monographs or compendia of recent "research" (which as one of the > contributors to this list has just noted, with justice, is not > the same thing as "science"). > > PRG > Perhaps I should have said that I disagree emphatically with the view that we *need not be concerned* with dishonest, misrepresented, etc. science because it will all come right in the end. We should be concerned with it for a great many reasons, even if we thought it would all eventually work out and truth would finally emerge, maybe after you and I are dead. If nothing else bothers you, consider that people may be harmed by misinformation in medical science before the truth fights its way inevitably to the fore. The recent fraud in clinical trials comparing breast cancer treatments may not have altered the essential conclusions about appropriate treatments, and so patients may not have been harmed, but that is just pure good luck and not evidence that bad science is harmless. I haven't the faintest idea what exact percentage of textbooks is wrong. You may think you've scored some sort of debating point, but the question is irrelevant to my original comments and pretty silly to boot. If the textbooks were wrong because the authors read bad or fraudulent science without detecting it, how would I know? I read the same science. But the fact that a question is preposterous and unanswerable should never stop a professor. It is an interesting question, even if irrelevant, so I will take a stab at it. In those few areas I am qualified to judge, physiology textbooks often say things that are wrong. Sometimes they are wrong because nobody knows better. Very likely they are wrong sometimes because the authors have read bad, fraudulent, or careless research and believed it, but of course we cannot detect that. Most often, I think they are wrong because the authors haven't read much science at all. I have a very old Physiology textbook which discusses the theory that the semen contains homunculi, which "grow to man's estate" in the womb. It denies that this can be true, because it is obvious to the author that the semen is the product of putrifaction and decay. I can't really blame the author; he felt he had to say something and because he was an MD, he could hardly admit he did not know anything. I own an early 20th century Physiology textbook whose original owner was Joseph Erlanger (Nobelist in physiology). The margins contain notes to the effect of, "I cannot confirm this." This text may have been based on reading of the primary scientific literature; at the least, what it contained was of current interest to a prominent scientist. Certainly, Erlanger thought the book contained errors. I have not opened it for many years, but I am certain much of what it contains would now be considered laughable. Is that an example of the truth coming out? Only if you are sure that what we have now is perfect understanding of the truth. We may simply have moved to a different level of wrongness. Today, my guess is that, with a few rare and welcome exceptions, textbooks are based on other textbooks. I cannot imagine that text authors struggle through all the primary literature of physiology and assess it critically; I cannot keep up with it in my very narrow field. Once errors are in textbooks, they have a long life and propagate from book to book. Even if science occasionally corrects itself, it may not find its way into the textbook. For example, one of my pet peeves is that many medical physiology textbooks present as fact that cardiac muscle cannot be tetanized because it has a long electrical refractory period. The first paper I can recall refuting this was published about 40 years ago. Why do they say this? Not because of any fraudulent research. I guess because other textbooks said it. I am currently reading a textbook of electrocardiography that is very popular with our students. I am only a few pages into it, but I find egregiously wrong blanket statements about the innervation of the vasculature. That's about 4% of the book right there. Not only has this author not read original research, he does not appear to have read other textbooks. I cannot recommend one of the most popular texts of medical physiology to our students (unless they have a table with one short leg) because the author has, as far as I can tell, made the book read well by making up plausible (to him) explanations for things nobody knows, which he presents as established fact. So I don't know what percentage is wrong, but it is significantly above zero and probably irrelevant to whether one should be sanguine about sloppy, dishonest science. Bad science cannot help the textbooks, that is for sure. WRG Ray Gibbons Dept. of Molecular Physiology & Biophysics Univ. of Vermont College of Medicine, Burlington, VT gibbons@REDACTED.med.uvm.edu (802) 656-8910 [. . .] >>> Posting number 330, dated 15 Jul 1996 09:50:44 Date: Mon, 15 Jul 1996 09:50:44 -0600 Reply-To: Discussion of Fraud in Science Sender: Discussion of Fraud in Science From: "John C. Bailar III" Subject: Re: anyone could win As I have said here before, we never, never know as much as we think we do -- even after allowing for all of the uncertainties we can think of. This is relevent to the discussion of textbooks and their errors. But the important question is perhaps not whether there are errors, but whether the authors have made a serious effort to understand what thery are writing, and put in all of the caveats. Then the reader must apply the extra (large) dollop of uncertainty. Unfortunately, this is not taught in an organized wayt anywhere in our schools, from kindergarten through post doc. We tend to cogive students the impression that if something is in the secondary or tertiary literature it has been thoroughly checked out and is very likely to be true. But all this uncertainty is no reason for doubting the general efficacy and effectiveness of science. The fruits of the method are all around us, and I will not bother to enumerate some of them. It is rather a call to do the very best we can (hence the gravity of misconduct in science) and remember both the uncertainty about each part of the whole enterprise as well as its general overall success. We aren't always wrong enough to matter when it comes to important applications. Newtonian physics is still pretty useful to almost everyone with a real problem to solve. John C. Bailar III Chair, Department of Health Studies University of Chicago MC-2007 5841 S. Maryland Ave. Chicago, IL 60637 Phone 312-702-2453 Fax 312-702-1295