(As two book projects about surrealist games have failed already, I should probably post some of the materials here, just in case I get hit by a car or something.)
The central role of games in surrealism can be understood either through participation in its ludic spirit in actual games, or analytically, for those who lack the experience, or have it and want to interrogate it further and find ways of generalising it. (I was going to say "through participation in actual surrealist games" but added the extra determination "in its ludic spirit" when I came to think of own experiences of people participating in the former without participating in the latter, remaining either unserious or cynical or obstructive or just purely intellectual...)
There has to be circumsciptions. Many historians and sometimes surrealists themselves include various experiments, interrogations, enquiries and creative methods as games (a few even emphasise the ludic element in individual creativity in general). It can be very interesting to investigate the ludicness of all such activities, but the understanding of what this ludicness consists of, must be based on actual games.
Caillois's theory of games
A systematic enquiry into games in surrealism does best to use as a point of departure ex-surrealist Roger Caillois analysis of games from 1958, building on the groundbreaking work of Johan Huizinga but to my knowledge remaining the most clarifying general framework for analysis of games so far.
In Caillois's analysis, the constitution of play as a sphere in itself detached from other social activities, is due to it fulfilling a number of criteria that separates it from useful social activities:
1) free; participation not obligatory
2) separate; circumscribed in space and time, defined and fixed in advance
3) uncertain; course and outcome undetermined
4) unproductive; creating neither goods nor wealth, returning to status quo (except maybe redistributing wealth among participants)
5) governed by rules, or
6) make-believe; accompanied by a special awareness of a separate reality or of a free unreality, as against real life
(where all must appear simultaneously, except only either 5 or 6 which are mutually exclusive)
Surrealist games may stretch some of these criteria a little, and, it seems, thereby, point out some internal problems of Caillois's set of criteria.
But let's continue with the structure of Callois's concept. He introduces a classification of games as focused on one of four basic mechanisms:
A) agon, competition
B) alea, chance
C) mimicry, simulation
D) ilinx, vertigo
(where A and B are mutually exclusive, but C and D can embrace each other or either of A and B)
and for each of these Caillois suggests a bipolar distribution between
I) paidia, improvisation, turbulence and carefree gayiety
II) ludus, binding with arbitrary, imperative, impractical conventions
(this bipolarity exists for A-D though number I appears more connected with C and D and 6 and number II more with A and B and 5)
if focusing on surrealist games we might have chosen some other parameters as main compass directions, but this concerns all games, and surrealist games too can be analysed in terms of the relative importance of each of these four.
Surrealist games in the light of Caillois's theory
Some of Caillois's points are not so crucial for surrealism, or the specific character of the surrealist games might bring their meaning into question.
Free? Sure, for a particular occasion of playing. Yet the commitment to game in surrealism is a more ambitious commitment, and lies at heart at the commitment to the surrealist adventure, the invocation of the poetic transformation of everyday life and the selling of one's soul to poetry and to that cause.
Separate? Yes and no, because since surrealism encourages the game to contribute to reenchanting reality, it is encouraged to grow out of its initial frames and claim its rights in areas we couldn't quite foresee. Though the original establishing of a game is constituted separately, creates its own rules, its own time, its own connections of meaning, and hopefully threatens to turn them back upon our usual world.
Uncertain? Yes, very much. We don't know where it will take us and we don't know what it will make of us.
Unproductive? In the sense of useless: yes very much. Yet indulging in the useless and unguaranteed is exactly the kind of activity that we expect to be productive in terms of experiences. And it keeps producing artifacts, and its artifacts might gain a secondary use as art or new knowledge or whatever. We expect the interesting stuff to come from this source, but we can't calculate with it. In fact, from the surrealist viewpoint, even the most conscious production of art will not be this separated from the mechanisms of play.
Governed by rules, or make-believe? As defined by Caillois, this make-believe is something deeply problematical from the viewpoint of surrealism. Separate reality, yes in that is not governed by utilistic concerns of our usual habits, but that separate reality has a dynamism of not staying separate but instead being able to intervene in quite dynamic ways in all areas of life. However it is difficult to see the mutual exclusivity between such a reality (in fact a "parareality tending towards overreality") and the governing by rules. The experience of surrealist games in fact tells us specifically that ambitious rules are often a fruitful way of invoking such a reality. There is no contradiction. The games of surrealism are clearly governed by rules in most cases but not all, but we are often eager to revise the rules en route, or improvise new sets of rules according to what the previous results suggested, or according to whim, or according to suggestions of lyrical intoxication into a separate reality. If make-believe in the typical cases is a focused temporary complete faith in a particular suggested view of reality, surrealism is usually more about suspending metaphysical judgment, playing along in order for ideas or scenarios to reveal their potential without necessary succumbing to even temporary faith.
Agon? Rarely. Not important.
Alea? Very often.
Mimicry? Oh yes, but often striving towards simulating the unknown to gain insight in it rather that simulating something known.
Ilinx? Sure, but hardly in itself.
Ludus? Very much. But again without seeing a contradiction versus paidia. Most of surrealism has taken place in intimate interplay between improvisation and impractical conventions.
Some important characteristics of surrealist games, or play in surrealism, that has not been accounted for so far in this model:
collectivity: and this collectivity works together with the method (governed by rules), the alea and the ilinx against the individual.
interpretativity; it typically includes a paranoiac-critical moment; assessing the pattern of the artifacts or insights produced either as a conclusion of the game, to be integrated in theoretical speculation, further creativity and further games, or more immediately as a basis for revising the rules and continuing.
and repeating some key points from above:
intervening and reenchanting (transforming life)
interaction of improvisation and rules
using the products yet remaining in the domain of the useless
Again, play, especially as shown by surrealism, is not all that different from art and science. Maybe that is the major weakness of Caillois's analysis, that since it retains a focus on analytic functionality from an academic viewpoint, it is not only about illuminating what games are about but also to classify in social morphology where exactly the border goes between play, art, science and other enterprises. Of course, in reality they may not be all that separate, but the academic will want to keep them separate in order to account for them within that particular analytic framework. While the rest of us may have reason to emphasise those aspects and moments where the boundaries are vague and perhaps even non-existent, where play and art, play and science, play and daily social interactions, are merged or transformed into one another. There is even nothing wrong with practical utility, as long as it is acknowledged as the trivial puzzle-solving it is, that needs to be put into a wider framework of playing and curiousness towards the unknown, and not vice versa.
(I realise that there has been a systematic assessment of Caillois's principles in a surrealist context before, by Martin Stejskal in Analogon #6 (1992), unfortunately largely inaccessible to me because of language...)