Prisoner's Dilemma citation

July 1, 2009 12:33 PM

How do I cite the originators of the Prisoner's Dilemma (in my thesis) when it looks like none of them actually published anything on it?

The Stanford Encyclopedia of Philosophy says this:

Puzzles with the structure of the prisoner's dilemma were devised and discussed by Merrill Flood and Melvin Dresher in 1950, as part of the Rand Corporation's investigations into game theory (which Rand pursued because of possible applications to global nuclear strategy). The title “prisoner's dilemma” and the version with prison sentences as payoffs are due to Albert Tucker, who wanted to make Flood and Dresher's ideas more accessible to an audience of Stanford psychologists. Although Flood and Dresher didn't themselves rush to publicize their ideas in external journal articles, the puzzle attracted widespread attention in a variety of disciplines. Christian Donninger reports that “more than a thousand articles” about it were published in the sixties and seventies.

I can't find any direct citations, any original articles where this game was formulated/formalized. When I look at research that has used the PD, most don't bother to cite anyone for the formulation of the game. Am I just missing something? Do I just cite (for example) the 1993 Poundstone book where the origination is described?

posted by arcticwoman to education (16 comments total)

I don't think a citation is necessary; it's one of these things that is now common knowledge.

Similarly, when economists use a demand curve in a paper, they don't bother to cite whichever economist first used a demand curve.
posted by matthewr at 12:40 PM on July 1

My supervisor insists, unfortunately.
posted by arcticwoman at 12:50 PM on July 1

“This situation is known as The Prisoner's Dilemma, a puzzle which was first investigated by Merrill Flood and Melvin Drescher, although they do not appear to have published articles on it themselves¹.”
____________________________________________________________________
1. Stanford Encyclopedia of Philosophy: “Although Flood and Dresher didn't themselves rush to publicize their ideas in external journal articles, the puzzle attracted widespread attention in a variety of disciplines.”
posted by koeselitz at 12:50 PM on July 1 [2 favorites has favorites]

In other words: where are you finding out about the theory? Cite that. What tells you they didn't publish? Cite that, and your supervisor will have to be satisfied that you can't cite the men themselves.
posted by koeselitz at 12:52 PM on July 1 [1 favorite has favorites]

Your problem may have something to do with the nature of the game's origin. The RAND Corporation traces its origins to military-funded operations research, specifically, the development of strategies and tactics for nuclear arsenals.

Prior to the twentieth century, when someone came up with a new weapon, its applications were pretty obvious. Bronze swords were a big step up from stone spears, and it didn't take a genius to figure out which end to hit 'em with. Same goes for iron over bronze. Both revolutionized warfare, but really, there wasn't a huge amount of research that went into developing strategies there. Similar things can be said up until about the nineteenth century. Military tactics changed very little during a lot of that time. It was only during the American Civil War that military leaders started to realize that lining your boys up and having them whack away at each other was no longer going to work, but this was still a minority view (Confederate General Longstreet was ahead of his time here). It wasn't until World War I that this was driven home unmistakably, as the machine gun and trench warfare effectively demonstrated that the old ways were no longer useful, and a big, big change needed to happen.

But if the machine gun and the tank changed warfare a lot, this paled in comparison to the difference that nuclear weaponry posed. "Well," the thinking went, "the eggheads gave us these weapons. Maybe they can tell us how to use them too." So places like RAND were tasked with projects such as coming up with the optimal angle of attack for a plane attacking a surfaced submarine, or how many rounds each soldier should carry, or how to station ships in a convoy.

It was in this context that Flood, Dresher, et al came up with the Prisoner's Dilemma game. Nash worked there too, for a time. The game is an attempt to model the interaction between two nuclear-capable states. It's kind of a kluge, but it's from here that the idea of Mutually Assured Destruction originated: because the best move is to always screw your neighbor, and because both sides are presumed to be aware of that, it is in everyone's best interest not to play at all. WarGames is, in fact, based on solid operations research findings.

All of this is background for saying that if you're looking for the paper in which the Prisoner's Dilemma was first described, you aren't going to find it. RAND and its researchers didn't publish the same way that academics normally do, and things like this developed there didn't always get published at all. A lot were classified. Many others simply entered into the broader academic discussion without having been formally published, as a lot of the people at RAND and like places went on to academic postings, making use of their government-sponsored research without bothering to go back and publish what they'd done there (even assuming they'd have been allowed to do so, which is unlikely).

But I'll tell you what you can do. I got a lot of my information here from Machine Dreams: Economics Becomes a Cyborg Science, by Philip Mirowski, and from classes I took from him this past year. The book has 50 pages on game theory at RAND. If you're looking for citations, either secondary or direct, this is where you want to look.
posted by valkyryn at 12:56 PM on July 1 [2 favorites has favorites]

What does the OED give as the first recorded use?
posted by ROU_Xenophobe at 12:57 PM on July 1

Can you site a review? Something like "This problem resembles the prisoner's dilemma (reviewed by Smith et al.)."?
posted by mr_roboto at 1:15 PM on July 1

Can you just cite a more recent review about the PD, or if you can find one a review encompassing the initial formulation?
posted by scodger at 1:16 PM on July 1

Thanks, now that I know more about why I can't find anything on it, I feel a little better about my search skills. I'll find a review I like and cite that.
posted by arcticwoman at 1:29 PM on July 1

koeselitz - wouldn't the footnote go outside punctuation?
posted by quodlibet at 3:01 PM on July 1

The Evolution of Cooperation by Robert Axelrod is often cited for The Prisoner's Dilemma if you want to discuss how strategies of cooperation could actually evolve in nature.

In fact, that Wikipedia article references Flood and Dresher through referencing Axelrod.
posted by miasma at 4:13 PM on July 1

Ask the librarians at your school's library.
posted by cestmoi15 at 4:38 PM on July 1

It's not at all true that RAND research is unpublished. Some of it is unpublished (classified), but much of it is available from RAND and much of it was published in academic journals.

In pg. 26 (28 of the PDF) of "A Brief History of Rand's Mathematics Department and Some of Its Accomplishments" we read:

The [game associated with the Prisoner's Dilemma] was first explicitly conceived, and tested by RAND's Merrill Flood and Melvin Dresher in 1949/50."

So I looked at RAND's bibliography on Gaming (Pre-1970) and on pg 20 (21 of the PDF) found "Some Experimental Games. M. M. Flood. June 1952."

That paper is available from RAND in printed form and PDF (direct link).

To confirm that this is the right paper, I looked at Poundstone (via Amazon). This is clearly the paper described in Poundstone pg. 117 (and before). Poundstone notes that the further development as the "Prisoner's Dilemma" is due to Tucker, first in his 1950 Stanford lecture and then in a letter to Dresher. (and see bold below for publication of the Stanford work, alternately described as a lecture or a memo).

A version of Flood's paper was also published in Management Science Vol. 5, No. 1 (Oct., 1958), pp. 5-26 . That's available from JSTOR.

Prior to Poundstone this was discussed in Cook and Levi The Limits of rationality where the attribution of the name to Tucker is cited to a private communication with Flood.

In an on-line timeline of Game Theory, Paul Walker writes that the memo calling it the Prisoner's Dilemma was published: "Publication of Tucker's (1950) memo occurred in 1980 under the title On Jargon: The Prisoner's Dilemma, UMAP Journal 1, 101."

So I guess that's the cite. I can't find that on-line, but it looks like it's a math journal that should be accessible through a university library.

(Hah! And if we'd gone to Wikipedia first we'd have seen a link to Flood's paper there.)

The Stanford Encylopedia is not entirely accurate... Flood at least did publish on it even if he didn't call it that.

Bonus: John Nash's dissertation cited by Flood is available here (and later published in Annals of Mathematics.)
posted by Jahaza at 6:50 PM on July 1

That's really silly to demand specifically citing the PD -- in the social sciences, it's about the most basic thing ever. But whev., I second the recommendation to just cite Axelrod and have done with it.
posted by paultopia at 7:10 PM on July 1

And *bing*, via Rasmusen blog post here's the Stanford handout... and a letter from Tucker suggesting Flood felt he didn't get proper credit!

Turns out the article in UMAP Journal is by Philip Straffin, not Tucker:

Philip Straffin, ``The Prisoner's Dilemma,'' UMAP Journal. 1: 101-103 (1980).

Rasmusen says it was reprinted in Two-Year College Mathematics Journal (1983) 14: 229-230 with an additional interview with Tucker.

It'd still be interesting to see the Straffin article. It's not clear from Tucker's handout alone that he called it the prisoner's dilemma, just that he came up with the prisoner scenario...
posted by Jahaza at 7:23 PM on July 1

Wow, Jahaza- that's way more in depth than I strictly need, but I've spent the last hour reading those links anyway! Interesting stuff - it's amazing how muddy attribution of ideas can get. Thanks for all the work.
posted by arcticwoman at 8:08 PM on July 1