Denyse O’Leary uses Bill Dem­b­ski’s blog (and a dozen oth­er ID blogs) to report a com­ment from a friend about the mis­sion state­ment for Nature. The mis­sion state­ment reads:

First, to serve sci­en­tists through prompt pub­li­ca­tion of sig­nif­i­cant advances in any branch of sci­ence, and to pro­vide a forum for the report­ing and dis­cus­sion of news and issues con­cern­ing sci­ence. Sec­ond, to ensure that the results of sci­ence are rapid­ly dis­sem­i­nat­ed to the pub­lic through­out the world, in a fash­ion that con­veys their sig­nif­i­cance for knowl­edge, cul­ture and dai­ly life.

Her friend replies:

To report advances and serve sci­en­tists means not to report set­backs, or the expo­sure of fal­lac­i­es in wide­ly-held the­o­ries that would tend to put main­stream sci­ence in a bad light.

The press and pub­lic oper­ate under the impres­sion that Nature and Sci­ence mag­a­zines report any sig­nif­i­cant devel­op­ments in sci­ence, whether pos­i­tive or neg­a­tive, and that both serve the pub­lic; but both pub­li­ca­tions are very up-front that they only report advances and suc­cess­es because they exist to serve scientists.

This reflects a dra­mat­ic mis­un­der­stand­ing of what an advance in sci­ence looks like. A neg­a­tive result is not a “set­back,” it is an advance, because it improves our state of knowl­edge. A set­back is when your fly colony devel­ops par­a­sites and dies off, or when you real­ize that you put fer­til­iz­er on your con­trol plot, and don’t even know how much. That means you have to start over, so the time invest­ed in that exper­i­ment was wasted.

But any exper­i­ment that goes through to com­ple­tion is an advance, even espe­cial­ly when it over­turns some exist­ing set of ideas. 

To choose an exam­ple O’Leary is prob­a­bly aware of, con­sid­er the reac­tion to a paper Nature pub­lished in 2005, show­ing that Ara­bidop­sis plants seemed to be able to restore a gene to the state found in their grand­par­ents but not their par­ents. Nicholas Wade’s New York Times report on the find­ing described it as “star­tling,” poten­tial­ly “an unprece­dent­ed excep­tion to the laws of inher­i­tance,” and “sur­pris­ing.” One of the dis­cov­er­ers described the result as “a com­plete shock,” and an edi­tor at Nature explained that the author’s hypoth­e­sis about an RNA back­up copy of the genome as “the least mad hypoth­e­sis for now.” If it were true, it would set our knowl­edge of genet­ics back to the time of Mendel, or at least the redis­cov­ery of his work in the ear­ly 20th century.

How did sci­en­tists react to this poten­tial set­back? One inde­pen­dent researcher described it as “mar­velous,” anoth­er told Wade it was “fas­ci­nat­ing” and “unprece­dent­ed.” Oth­ers observed that it could force a recon­sid­er­a­tion of the evo­lu­tion of sex as well as the field of genetics.

Did Nature sit on the result? Hard­ly, indeed they may have been too rushed to report it. The review process which nor­mal­ly takes months was trun­cat­ed to six weeks, and sub­se­quent research sug­gests that the result is not a result of non-Mendelian inher­i­tance. The plants in the exper­i­ment typ­i­cal­ly self-fer­til­ize, but the muta­tions they intro­duced have a side effect of increas­ing out­cross­ing. That may explain how some non-mutat­ed genes got intro­duced into the mix. Oth­er researchers observed that DNA from parental chro­mo­somes has been shown to cross into devel­op­ing seeds, pro­vid­ing a more like­ly locus of the grand­par­en­t’s gene than a pre­vi­ous­ly unseen RNA back­up. Anoth­er researcher found ver­sions of the cor­rect (unmu­tat­ed) DNA sequence in oth­er parts of the genome, sug­gest­ing that inter­nal error-check­ing may allow cor­rec­tion of muta­tions. All of those reac­tions were pub­lished in Nature. The orig­i­nal authors respond­ed to some of these crit­i­cisms in Nature. And of course, Pan­da’s Thumb blog­ger Reed Cartwright turned a blog response into a pub­lished reac­tion in Plant Cell, propos­ing that nat­ur­al selec­tion on pollen com­bines with dele­te­ri­ous effects of the mutat­ed gene to pro­duce the appear­ance of non-Mendelian inheritance.

In each phase of that dis­course, peo­ple were dealt per­son­al set­backs. The orig­i­nal paper sug­gest­ed that some­thing unprece­dent­ed was going on in plant genomes, and pos­si­bly in oth­er species. The respons­es sug­gest­ed fail­ures of tech­nique or the­o­ry by the orig­i­nal group. At each stage, the sci­en­tif­ic jour­nals accept­ed these crit­i­cal com­ments. They did so because each paper was an advance for sci­ence, even if it set cer­tain indi­vid­u­als back. 

I won’t claim that ego and per­son­al bias­es play no role in sci­ence; they do. But the struc­tures of sci­ence – peer review, repro­ducibil­i­ty, testable hypothe­ses – exist to over­come those human fail­ings. Unlike IDol­a­tors, who con­duct “research” in order to pro­mote a philo­soph­i­cal view­point, real sci­en­tists research and pub­lish their results in order to advance our state of knowl­edge, even if those results under­mine some­thing those same sci­en­tists pre­vi­ous­ly claimed. Since every result (bar­ring those pro­duced by erro­neous meth­ods) advances our knowl­edge, every exper­i­ment pro­duces some sort of advance, even if it sets some­one back.