This message comments on Marshall Jones' paper. One of the things I like about the paper is its attempt to state its assumptions and assertions directly, making it much easier to spot disagreements.
A starting example would be "pleasure is defined as being passive in nature"--which would entail that food and sex cannot normally give pleasure (force-feeding apart): not a view (or use of language) that I think many would agree with. Still, I agree with the underlying purpose of the argument at that point, which is to do with discussing the issue of how active, how interactive, and how engaged a participant is and the relationship of that to the nature of the experience. This is close to the heart of the matter, and to a critique of flow theory at least in its connection to learning that I would like to try out below. But first a few more basic issues.
Firstly, the paper seems to allow the equating of fun, engagement, flow, immersiveness, and good experience. That may be Csikzentmihalyi's theory (I haven't read enough and carefully enough to be sure myself), but we can question it. I shall mention some tricky examples below. For an independent theory of engagement, Brenda Laurel's work would be the one (Laurel came from a drama background, and applied that perspective to user interface design).
Laurel, B.K., (1986). Interface as mimesis. In D.A. Norman & S.W. Draper (Eds.), User Centered System Design, pp. 67-85. London: Erlbaum.
Laurel, B. (Ed.) (1990) The art of human computer interface design. New York: Addison-Wesley.
Secondly, the issue of what the relationship of fun, etc., is to learning. If we can have both together, 'most everyone would take that option; but the question is: do they go together? For instance, flow seems the opposite of reflection (as Marshall points out in citing Norman's distinction), which seems an argument against. That is, flow is the enemy of reflection, which many think is an essential component of good learning.
Thirdly, computer game playing is still a minority occupation, while learning through education is a majority one. Just because a minority find computer games compelling (as a different minority find heroin compelling) is not only not a reason for adopting it, but not good grounds for assuming that a majority would find it attractive even if it were useful. Urban myth says both are universally irresistible, but there is little evidence of this. You don't like me mentioning heroin in the same sentence as education and computer games, huh? Well, try chess then. This more than most things offers the spectacle of people immersed in a flow experience. But it just doesn't seem a compelling argument to say we should make educational software just like chess, does it? It just does not follow that because some activity is compelling for a minority, that it is a likely recipe for universal absorption.
Marshall's argument, then, has the following structure (I hope I get this right):
I am interested in all these elements (although separately), and I too think they are important. However perhaps the elements fit together in a rather different way.
One of Marshall's axioms (explicit starting points) is "learning can be an immersive activity." I agree. But when I reflect upon this, it immediately (to me) points out how flow theory does NOT describe immersive learning (as opposed to playing) activities. For instance, of the eight properties of flow in his tables, I would attribute property numbers 2, sometimes 5, 6, 7, and 8 to immersive learning experiences; but not properties 1, 3, and 4. Learning is not (1) a task that we can complete: or rather, only shallow learning for a specific known test task but not deep learning which is open-ended and non-terminating. Above all, (3) learning does NOT have clear goals when it is enjoyable but only when it is done for unenjoyable extrinsic reasons (cramming for an exam). And (4) learning does not usually have immediate feedback, at least in the sense meant when game designers talk of immediate feedback (by the machine).
For instance, in writing this message for ITForum, I believe I am learning intensively. I am not watching the clock. But also, I do not have immediate feedback (and if I did, it would distract me from finishing and polishing the message); and I do not have a clear goal: writing causes learning exactly because I only discover what I think as I write. Only non-learning writing is fully planned (i.e., has a clear goal) in advance.
Going back to my first point, I think the really interesting question here is the contrast between kinds of flow, or game flow versus learning immersion.
Marshall's first axiom is "Learning should be fun." I don't think either I or my best students would be happy with that as an exact statement: immersive, motivating yes; but "fun" sounds trivial and seems a word more appropriate to relaxing pastimes than to the most deeply engaging and therefore effortful activities. Of course, perhaps this really just means that Marshall is exploring the design of remedial software for poor students, but would actually prefer something different for his own learning. I find this issue uncomfortable: either I could seem patronizing and write play software only for students who aren't good enough to benefit from the things I myself find best; or else I could be really stupid and act as if all learners were exactly like myself when obviously needs vary enormously. But I think we should try to be explicit about the position we each take on this issue. Marshall does say things about this, but ends up with statements on both sides of this divide (as I see it).
Next we come to my own use of games, including computer games. I regularly spend quite a few minutes most days, but I DON'T choose challenging games: on the contrary, I use games I have used for so long my actions are completely mindless. For instance, at my desk at home, I use a pack of cards to play Patience while I let my mind tick over on some issue (I did this for about 20 minutes before starting to type this message). This fits all the criteria for flow, except for 7b: my sense of self is not stronger afterwards. But it is mindless: the actions cause no learning whatsoever and that is the point. It is also not addictive in the way computer games can be: I may go on rather too long in order to avoid the next burst of effortful activity (work avoidance), but because there is no learning, it is not very compelling, but it is restful (rest through activity). At work I use a very old game on Unix in the same way: I no longer see any new features nor develop any new strategies or tactics. Conversely I very seldom get attracted to a new computer game even when shown it: I have more than enough challenge in the rest of my life, and other software than games I should learn if I have the energy for struggling with yet another user interface.
What might all this mean? Perhaps there is a spectrum of kinds of flow, from the no-learning flow I've just described, through the kind usually described where challenge and learning (about the game) are integral, to immersive learning without games where immersion doesn't follow many of the "flow" characteristics as I described earlier. It seems to me that the flow theory doesn't feature learning at all; while the literature on what makes the most successful computer games does (in terms of challenge, skill acquisition, and learning knowledge about the game), so part of the dichotomy is already implicit in that observation.
Marshall raises the issue about "bells and whistles" (B&W) versus more relevant attempts to engage the user. Here I'd like to suggest that a quite different interpretation might cover what he is getting at. His view is (I think) that we want learning and we want to use engagement for, but only for, promoting this. Now consider, instead of the theory of flow for guiding this, the principle that at all possible times teaching should not only be at the level of concepts but make links to the individual learner's personal experience. (This principle is in the Laurillard model, and is also at the heart of constructivist approaches.)
I have just finished a series of six lectures in an introductory psychology course to our first year students (who in Scotland often arrive at age 17 years). The feedback (both written feedback and manifest shifts in attention levels in class) makes it clear that they put a very high value on the demonstrations I put on, such as mixing color lights to make white light, and a demonstration of after images with complementary colors: many said those were the best features of my lectures. Perhaps because this is a bit hurtful to my vanity, I have a strong tendency to think that these kids just want the B&W of a dramatic show, yet theory actually predicts their importance. But note that it predicts it for a totally different reason from the one focused on by Marshall, i.e., flow theory. Marshall does say that the B&W should be relevant to the content; but perhaps it goes further than that: perhaps flow theory and computer games are totally irrelevant, and all this is just a manifestation of a much more basic educational need to build links between theory and personal (mainly sensori-motor) experience. The big increases in motivation do not come from entertaining the learners, but (as in constructivist and/or problem-based learning approaches) from addressing their (usually starved) learning needs for relevance, meaning, and connection to themselves.