Hoax paper fools cybernetic boffins
Sokal-style bogosity metrics
An MIT student has had a paper consisting of computer-generated gibberish accepted by technology conference WMSCI. The pretentious gathering bills itself as "an international forum where researchers and practitioners examine Systemics, Cybernetics and Informatics key issues"
Comp sci undergraduate Jeremy Stribling told us that he didn't single out WMSCI because of its subject matter, although it's easy to see how it made a tempting target.
"A Metaphor," the organizers explain. "We are trying to relate the analytic thinking required in focused conference sessions, to the synthetic thinking, required for analogies generation, which calls for multi-focus domain and divergent thinking. We are trying to promote a synergic relation between analytically and synthetically oriented minds, as it is found between left and right brain hemispheres, by means of the corpus callosum." [their emphasis]
But the conference organizers' two minds didn't meet in time to catch the hoax, which fell right through WMSCI's supposedly rigorous review procedures.
Stribling's paper consisted of randomly generated buzzwords munged into complete English sentences by a madlib-like program, so they were grammatically correct but meaningless: much like one of Jonathan Schwartz's weblog entries, or a Cory Doctorow novel.
The paper, entitled Rooter: A Methodology for the Typical Unification of Access Points and Redundancy, and contained such wonderful claims as -
"We can disconfirm that expert systems can be made amphibious, highly-available, and linear-time."
"One must understand our network configuration to grasp the genesis of our results. We ran a deployment on the NSA’s planetary-scale overlay network to disprove the mutually largescale behavior of exhaustive archetypes. First, we halved the effective optical drive space of our mobile telephones to better understand the median latency of our desktop machines. This step flies in the face of conventional wisdom, but is instrumental to our results."
The submission was accepted because the three principal reviewers failed to reject it in time.
"We thought that it might be unfair to refuse a paper that was not refused by any of its three selected reviewers," conference chair Nagib Callaos told Reuters.
The default was set to "accept all".
Not quite. Digging a little deeper, we discover that the organizers are very proud of the quantity of papers accepted. Although they did reject another autogenerated submission.
Stribling told us that he was simply fed up with repeated solicitations from the conference organizers.
"We figured this was a pretty low-grade conference because of the amount of spam they were sending out."
He says the authors have received no negative feedback - but the conference organizers have refunded their fee.
"They don't want us there - we'll try and get there somehow," he hopes.
The most famous academic hoax in recent years was Alan Sokal's paper Transgressing the Boundaries: Toward a Transformative Hermeneutics of Quantum Gravity. The physics professor's parody paper was accepted by cult studies journal Social Text in 1996.
But scientific establishment had equal difficulty with a paper submitted by two French physicists, Grichka Bogdanov and Igor Bogdanov. Their Topological field theory of the initial singularity of spacetime had theoretical physicists arguing for weeks whether it was or a prank, or a genuine breakthrough. Writing to us at the time, Igor Bogdanov denied that it was a hoax, although readers remained suspicious.
Perhaps because of the atomization of the disciplines in both arts and science, the quality of published academic papers appears to be at rock bottom.
And these days, simply being published means you're an authority. The MIT pranks illustrates all it takes to be published, is to submit a paper.
So with one stroke, the democratization of academia is complete! ®
Download the PDF [92kb]
Sponsored: Hyper-scale data management