how to make an abstract for an arxiv paper

December 18, 2010

A couple of days ago, I was venting some pent-up frustration with arXiv and the TR blog which hypes highly speculative and often unsupported papers into the realm of the new scientific holy writ. Well in the comments, a few knowledgeable readers decided to weigh in and posted a link to an arXiv abstract generator a lot like the generators designed to mimic verbose, inflated post-modernist tracts in the wake of the Sokal affair. And having seen my fair share of the stuff that gets submitted to this open repository of physics and math papers, it’s disturbingly similar to the real thing with the only real difference being a slightly exaggerated densities of obscure theoretical buzzwords per abstract. Here’s an example of what the snarXiv composes…

A boundary-dual of a model of dark energy offers the possibility of bounding orientifold black holes at the GUT scale. We use Heterotic strings near noncommutative brane structures, together with fragmentation functions in several holomorphic brane models for instanton liquids to consider the longitudinal XXZ Model. Next, we make contact between the scalar field analytic continuation of the Landau-Ginzburg Model and Z back-reactions to construct instantons on the surface of the sun. Before evaluating unitarity on moduli spaces of RS1 backgrounds of Sp(n) holonomy fibered over m-folds with nonzero spin-structure, we deduce that equivariant S-duality is the final component in solving anomaly matching. We leave the rest for future study.

So let’s evaluate. Introduction mentioning manifolds, black holes, and dark energy? Check. Confusing, wordy, buzzword-laden explanation? Check. Promises of complex and wicked numerology in the actual paper, using barely related concepts as a justification to make a theoretical model of something incredibly vague, obscure, and extremely narrowly focused? Check. Finally, an acknowledgement that no real experimentation has been done and any questions or issues with the paper will be settled by new papers citing this one and made as a least publishable unit to bolster publication count? Check, check, and check. All in all, the typical paper you’re bound to see while cruising arXiv for something interesting or built on solid mathematical foundations. Now, how about a Technology Review post generator? It should begin by rehashing a long-standing physics issue or a discrepancy in astronomical data, explain how a random paper in the archive might have the answer that will settle the issue once and for all, explain how if we only make some unjustified assumption, things would quickly clear up, then end by glossing over the myriad of easily identifiable problems with the work…

Share on FacebookTweet about this on TwitterShare on RedditShare on LinkedInShare on Google+Share on StumbleUpon
  • D.D

    This post had me in stitches. Man that abstract, I feel a black hole has abused my pdf viewer.