Letter from Christian B. Anfinsen to Bernard D. Davis
In this response to Davis, Anfinsen recalled his "biggest error in research" during his career, the conclusions reached
in a 1955 article with William Harrington, Aase Hvidt, Kaj Linderstrom-Lang, Martin Ottensen, and John Schellman.
Thank you for your letter of February 13 regarding the current situation in Congress and, indeed, throughout the public nationally.
Most scientists doing research take for granted that errors in interpretation and judgment are almost always picked up by
the work of others later on and that such errors are, in a sense, automatically corrected. I share with you the worry about
the "mad scientists" image that so many people are beginning to accept.
In connection with your request for "my biggest error in research," I do indeed have one rather amusing example from
my own scientific background. I enclose a photocopy of the paper published back in 1955, and it seems ideal for your purposes
since it contains as authors the names of individuals such as
Linderstrom-Lang, Schellman, Harrington, etc., generally considered to be moderately honest folk. This paper purports to
demonstrate that the activity of ribonuclease is essentially independent of hydrogen bonds since the enzyme digested both
RNA and a synthetic substrate at the same rate in the presence or absence of urea. The summarizing statement at the end of
this short note gives our interpretation at the time. A few years later while Michael Sela was visiting my laboratory at the
NIH, we found that the three-dimensional structure of ribonuclease, although enormously disorganized by 8 mol urea, was snapped
back into essentially native shape by the addition of certain polyanions and, certainly, by RNA itself. Thus the stabilization
of the three-dimensional structure by hydrogen bonds was restored in the presence of the substrate and other polyanionic materials,
and the original conclusion was therefore entirely incorrect. One could term the earlier conclusions "misinterpretation"
or even "professional naivete"; however the name of Kai Linderstrom-Lang in the author list would safely rule out
such appraisal. The same would, I think, apply to the term, "fraudulence." Mistakes are made all the time but most
of them seem to be corrected eventually.
As far as a paragraph for your collection, I would think that the second page of this small reprint beginning with, "All
the data" and ending with the list of authors might be enough to serve Barbara Culliton's purposes.