Jaron’s Blog


The arXiv is a website where scientists post drafts of their research papers, which is said to improve the speed and accessibility of doing science. Peer-approved users can submit their papers anytime, but a light moderation process means that there is a delay between submission and posting: [all papers submitted between 2 PM one day and 2 PM the next appear together at 8 PM][arxiv-times] in the order they were submitted.

While this is arbitrary and uninteresting, research has found that appearing at the very front or back of the order correlates with increased readership, citations, and other standard metrics of influence. The usual causal explanation is that those papers are more visible to people scanning the daily list. Because of this, some scientists arrange to submit just before or after 2 PM, sometimes making their graduate students submit this way as well. That trend lends another, non-causal, explanation: that scientists making the effort to promote their papers this way have also taken the effort to prepare better papers.

No one should ever do this. If the bump is pure coincidence due to self-promotion selecting quality papers, then arranging for a paper to be first doesn’t improve its readership and doing so is a useless for promotional purposes. There is some evidence that being first or last does have a bump independent of whether it was intentional promotion.

It’s hardly necessary to review once again the misrepresentation and elementary fallacies in the work on IQ that we ignore “at our peril,” exposed years ago when the game became popular and now repackaged for today’s purposes — always eliciting much praise for the courage of the authors in lining up with the powerful and “breaking the censorship” imposed by the radical extremists who run the universities and the press. Even if we grant every factual conclusion for which some shred of evidence is claimed, nothing of interest follows, except on assumptions that reflect ideological fanaticism, not science. Specifically, the topic of “meritocracy” is not even addressed unless we grant the tacit assumption that there is something “meritorious” about the particular array of traits, perhaps partially inherited (though nothing relevant is known), that confer power and prestige under particular social arrangements: in some societies, a penchant for torture and murder; in ours, some combination of greed, cynicism, obsequiousness and subordination, lack of curiosity and independence of mind, self-serving disregard for others, and who knows what else, as reflected quite vividly by income distributions and the contributions to society at the high end.

These trivialities aside, one striking finding of the studies reviewed is how poorly IQ correlates with socioeconomic status, possibly a consequence of the fact that it is designed to measure academic success, which is probably correlated only weakly with whatever it is that leads to wealth and privilege under the particular conditions of state capitalist society. If the real factors were studied, possibly better measures could be designed.