Gettier Problems in Quiz Bowl

Elaborate on the merits of specific tournaments or have general theoretical discussion here.

Gettier Problems in Quiz Bowl

Postby Muriel Axon » Tue Oct 24, 2017 12:09 am

WARNING: moderately frivolous post

Two or three times as a quizbowl writer, I've written clues where it was conceivable that:

1. A player might buzz in and give the correct answer, and
2. The player arrived at that answer through a train of thought that seems reasonable and justified, but
3. Their buzz nevertheless doesn't seem to represent 'knowledge'

It seems to me that this makes these situations like a Gettier problem -- a situation where someone has a justified, true belief about a situation, fulfilling a classical definition of knowledge from Plato, but where most readers would still intuitively feel like that person doesn't possess that knowledge. If you don't know what a Gettier problem is, read the paper (it's short and easy and will get you points):

Example 1: For the UD/MSU housewrite back in 2012, I wrote a biology question on "birds" that led by mentioning the important (but outdated) Sibley-Ahlquist taxonomy. (Whether that was a poor idea for a clue in a high school tournament, I'll leave to the reader.) I don't care to search for the question, but a similar clue appears in a question from (This) Tournament Is a Crime:

Nadya Medentseva theorized that the evolution of this clade was driven by the loss of thermogenicity of the protein UCP1. The discrepancy between the projected age of fossils of these organisms and their actual discovered age is termed the "temporal paradox." The "cursorial model" of the evolution of one of their characteristic behaviors was modified into the WAIR hypothesis. Based on DNA-DNA hybridization studies, Charles Sibley and Jon Edward Ahlquist put forth a taxonomy of them. Sexual selection led these animals to develop a unique secretion composition in their (*) uropygial gland. David Lack discovered that these organisms adapted to higher latitudes by increasing their clutch size. A scientist who was flummoxed by the tail of one of these organisms studied fifteen species of their subfamily Geospizinae in an early elucidation of adaptive radiation. The genus Archaeopteryx represents an evolutionary transition between dinosaurs and these animals. For 10 points, name these feathered, beaked, egg-laying vertebrates.

ANSWER: birds [or avians; or Aves; prompt to be less specific on peacocks, finches, and other specific birds]

Joe Nutter left a comment on my question to the effect that the name "Sibley" was too early -- anyone with the slightest interest in birds would know that Sibley wrote a famous bird guide. Except that the Sibley of the Sibley-Ahlquist taxonomy, Charles Sibley, is not the same as the Sibley of the bird guides, David Allen Sibley. Wikipedia mentions in the header of the former's article that "Charles Sibley is of no known family relation to renowned bird artist David Sibley," so they weren't both members of some famous ornithological dynasty. (EDIT: The Wikipedia page of David says that Charles did some genealogical research and showed that they could be no closer than fourth cousins.) I mentioned this to Joe and (I think) decided to keep the clue.

Example 2: I ran into a similar situation later where I wrote a question on a particular country, including in the first line a distinctive word related to the country that sounds fairly similar to that country's name, despite not sharing an etymology (as far as I could tell). The editor asked me to remove the clue, and this time I agreed to alter it.

1. If a player were to buzz on either of these clues and give the answer based on knowing the Sibley Guides (Example 1) or noticing a resemblance between the word and the country's name (Example 2), would that be justified? (I would argue that at least in Example 1, it would be at least as justified as the examples in Gettier's paper.)
2. Is it worth scrubbing clues like these from questions? Would it improve game play?

I think the examples provided in the famous Gettier paper are kind of hokey, even though they have the feeling of flashy virtuosity. (No comment on Gettier's character -- I just don't like the paper's style.) Much of the post-Gettier cottage industry in defining knowledge seems to have made more heat than light, too. What interests me about these cases is that it's often taken for granted that Gettier problems are unrealistic and pathological edge cases, but here they seemed to arise naturally -- if you accept my contention that these buzzes are both (1) justified and (2) not knowledge.
Last edited by Muriel Axon on Tue Oct 24, 2017 12:37 am, edited 1 time in total.
Shan Kothari

Plymouth High School '10
Michigan State University '14
University of Minnesota '19 (hopefully!)
User avatar
Muriel Axon
Posts: 616
Joined: Wed Mar 21, 2012 12:19 am

Re: Gettier Problems in Quiz Bowl

Postby vinteuil » Tue Oct 24, 2017 12:21 am

Muriel Axon wrote:unrealistic and pathological edge cases

but isn't this quizbowl too
Jacob Reed
Yale '19
East Chapel Hill '13
"...distant bayings from the musicological mafia"―Denis Stevens
User avatar
Posts: 983
Joined: Sun Oct 23, 2011 12:31 pm

Re: Gettier Problems in Quiz Bowl

Postby everdiso » Tue Oct 24, 2017 2:13 pm

I definitely agree that these cases are justified but are not knowledge. If I ever explain the Gettier problems to anyone in QB from now on, I'll use the Sibley example.
And as far as how we should account for this: I do think it should be avoided. Even though the Sibley clue wasn't actually dropping easy information in power, it created a situation in which someone with less than power-worthy knowledge (just knowing the Sibley taxonomy) would get power. This is just coincidental, but since it's a coincidence that can be predicted, I think we should avoid it. Giving points for something that isn't the knowledge being asked about is generally bad. Same goes for the country case: if someone who actually knows about that country might get beaten to the tossup on it by someone just buzzing from coincidental name recognition, and that buzz without knowledge would be rewarded, then I think that clue should be changed.
Paul Kasiński
University of Toronto, allegedly
User avatar
Posts: 25
Joined: Sun Nov 08, 2015 10:36 pm

Re: Gettier Problems in Quiz Bowl

Postby theMoMA » Tue Oct 24, 2017 3:42 pm

This is a pretty big issue at lower levels, where players are pretty likely to be credulous about clues that are suggestive of a particular answer in a limited answer space, regardless of whether that suggestiveness is coincidental. I wouldn't really expect higher-level players to buzz just because they heard the name "Grombia" and thought it sounded a lot like "Gambia," or whatever, and the higher up you go, the more you can depend on people knowing things like "Charles Sibley wasn't the Sibley of bird guide fame." This issue is still something to be cognizant of at higher levels, however. In cases like the Sibley example above, where it seems likely that higher-level players might not know the difference between the Sibleys (or might know that Charles and David Sibley are different people, but still decide that it's worth buzzing, on the (incorrect) theory that the Sibleys are members of a bird-crazy family), it may be worth considering whether a particular clue can be rephrased to work around a coincidentally suggestive term or name, or should be scrapped altogether.
Andrew Hart
Minnesota alum
User avatar
Forums Staff: Administrator
Posts: 5345
Joined: Mon Oct 23, 2006 2:00 am

Return to Collegiate Discussion

Who is online

Users browsing this forum: No registered users and 3 guests