Deconstructing Flow

What are the neurological underpinnings of the experience of Flow described by Csikszentmihalyi? Do we know enough about the brain to speculate?

Roots

This piece explores the concept of Flow, which I am assuming the reader is already familiar with, since the purpose of this piece is speculative discourse, not instruction. It is not about videogames – when I have finished this exploration, I have some specific points to make in that context, but that will come later. The following pieces may be essential to appreciating my train of thought:

• Paragraph six of this piece on videogame difficulty (the paragraph accompanying the graph of flow) describes Flow in a nutshell.

• The piece on Why You Play Games is also useful, particularly the brain diagram, and an appreciation for the reward system and the role of dopamine.

Disclaimer

This account necessarily takes a reductive view of behaviour. I don't believe that any account of behaviour solely in terms of parts of the brain, neurotransmitters or other reductive elements can ever be complete: one must also take account of the complete experience, and the context of that experience. However, since my express goal is to couple Csikszentmihalyi's psychological observations with a hypothetical neurobiological model, it is necessary to focus on the reductive elements to achieve this goal.

Neurotransmitters and Flow

The Flow channel draws upon three specific concepts: first, there is optimal experience (which is Flow itself) which corresponds with the Flow channel. Second, anxiety, which occurs when the challenges outstrips the skills of the subject. Finally, boredom, which occurs when the skills of the subject outstrip the degree of challenge.

What is the neurology of anxiety? This is easy – the neurotransmitter epinephrine (adrenaline) underlies both excitement and fear, with anxiety occurring when the fear centre (amygdala) perceives threat, loss of control or a similar problem.

What is the neurology of boredom? This is trickier – here we are not looking for the presence of particular neurotransmitters but their absence. Psychologists note that depression and boredom are related, which suggests boredom is related to serotonin, which is involved in maintaining a feeling of well-being. But I rather suspect boredom is simply the absence of epinephrine – which is to say, rather obviously, that boredom is the same as the absence of excitement.

If this is the case, then one dimension of the neurology of Flow is simply epinephrine, which is to say, excitement. This makes sense: if an aspect of Flow is engagement, then feeling a certain degree of excitement in a task is potentially a requirement for engagement. This degree of excitement may be very low, or it could be very high – observationally, we can distinguish between the the Flow of a test pilot or racing car driver, the Flow of a crossword puzzle, or the Flow of meditation – each is progressively less exiting than the other, but all are considered examples of Flow.

There must, therefore, be another element, and unsurprisingly the obvious candidate is the neurotransmitter dopamine, the “goal protein” released from the pleasure centre in the brain (nucleus accumbens), which as I discussed on the ihobo site last week seems to be involved in all enjoyable experience. Linking this with Flow is trivial: the very first condition of Flow that Csikszentmihalyi cites is “clear goals”.

Yet Flow is not goal-orientation, the striving to achieve (although it may include this) since many situations described as Flow are process-oriented – such as being in control of a vehicle, dancing, meditation etc. It seems to me from my observation of players that dopamine is not just released when a goal is achieved, it is also released when a goal is pursued. These smaller hits of dopamine serve to maintain interest in the long-term goal that is being chased. I didn't have any evidence of this until I found this paper by Peter Shizgal and Andreas Arvanitogiannis from the journal Science which describes this exact behaviour in gambling: “a gradual increase in the firing rate of midbrain dopamine neurons in anticipation of uncertain rewards.”

This, then, is my proposed neurological hypothesis for the Flow state: the individual experiences a degree of arousal proportional to the perception of challenge (i.e. excitement). The challenge itself holds out the prospect of reward (i.e. a big hit of dopamine) and this anticipation releases a small amount of dopamine in a continuous flow, helping to maintain interest in the ultimate goal. The achievement of the goal state is essentially anathematic to the Flow state (which is thus inherently process-oriented) it is the pursuit of the goal which produces a steady flow of dopamine and this would then be a candidate hypothesis for the neurology of optimal experience.

The Problem of Meditation

This is a great start, but of course there are problems. The principle concern I have is integrating yogic activities with the above model – since meditation in specific (and yoga in general) occur in the absence of excitement. They are, indeed, about reaching profoundly calm states. Why is this a problem? It is because the Flow model inherently requires that when the challenge outstrips the skills, anxiety results, and when the skills outstrip the challenge, boredom results. But in meditation, neither of these happen.

An amateur in meditation or yoga does not experience anxiety when they fail to successfully meditate. They are just unable to quiet their conscious mind and may, as a result, feel bored – the opposite of what Flow theory suggests. Similarly, a master yogi does not feel bored when they fail to reach states of deep meditation – this isn't the nature of the optimal experience of meditation, in which the participant simply reaches as deep a state of mental calm as they are able.

So why should it be that meditation violates the pattern of optimal experience?

A possible answer is that Flow incorrectly conflates different mechanisms into a single mechanism. A paper by Solberg, Holen Ekeberg, Isferud, Halvorsen and Sndvik shows the neurological effects of meditation, which involves two “well-being proteins”, serotonin and melatonin – very different neurology to that I proposed above. However, studies by Andrew Newberg on Tibetan meditators show that the frontal lobe of the brain (which includes the decision centre) is highly active during meditation, while the parietal lobe – which is involved in orienting us in time and space – is largely inactive. The activation of the frontal lobe is associated with concentration, which is a condition of Flow as described by Csikszentmihalyi.

But this isn't the whole story. Kjaer, Bertelsen, Piccini, Brooks, Alving, and Lou published a paper on neurolobiological responses during Yoga Nidra meditation which demonstrated that dopamine was linked to meditation (or at least, to this one kind of meditation): this is a strange result! People who were actively reducing their desire for action (disabling what is termed the executive function of the brain – the brain areas that mediate planning and execution) experienced increased release of dopamine. If this finding applies to all forms of meditation, then perhaps there is a common element to all optimal experiences: dopamine, the “goal protein”.

The Flow Checklist

Now is a prudent time to examine Csikszentmihalyi's list of nine elements associated with the Flow state. He states that not all elements need be present for an optimal experience to occur, but these are in general the hallmarks of the experience as he describes it. Here I associate each element with an aspect of the brain or the neurotransmitter system.

Firstly, activity in the frontal lobe, which links to Csikszentmihalyi's “concentrating and focusing”, “loss of the feeling of self-consciousness”, “distorted sense of time”, “sense of personal control” and “action awareness merging”. According to my hypothesis, all Flow states involve significant activation of the frontal lobe in some way, resulting in the state of intense focus.

Secondly, the dopamine reward system, which links to Csikszentmihalyi's “clear goals”, “direct and immediate feedback”, “balance between ability level and challenge” and “intrinsically rewarding”. Dopamine is the neurotransmitter associated with rewards, and furthermore the anticipation of uncertain reward produces small hits of dopamine. It is these small releases of dopamine which I contend are fundamental to optimal experience. In order for this to occur, the individual must always feel that the goal is achievable (which is how the “feedback” condition links to dopamine) but there also must be sufficient uncertainty that the outcome is not guaranteed (hence the “balance between ability level and challenge”). If the outcome is certain (the challenge is insufficient), boredom results – the anticipatory dopamine hits are cut off.

These two mechanisms I am contending are always involved in experience of Flow, although the relative activation of the frontal cortex versus the degree of dopamine released can be wildly different.

Finally, a third mechanism is required to explain the anxiety that occurs when the challenge outstrips ability, and this is presumably the “flight” of the fight or flight response (which I have discussed before in the context of rushgames), namely adrenalin (epinephrine) and the fear centre (amygdala) which is presumably triggered when the frontal cortex assesses the situation and concludes that the individual is out of their depth with respect to the degree of challenge. This causes the experience of anxiety. Furthermore, the experience of excitement acts to enhance the release of dopamine (which is to say, you experience more reward the more stress you were under prior to achieving the goal) thus creating a range of different Flow states according to the degree of stress associated with them.

I believe these three mechanisms are sufficient for an explanation of the neurological underpinnings of Csikszentmihalyi's Flow theory. All that remains is to validate this hypothesis experimentally which I shall not be doing since I am a game designer and not an experimental researcher.

Conclusion

It is my hypothesis that Csikszentmihalyi's Flow theory is explicable in terms of three neurobiological mechanisms: the frontal cortex, which mediates concentration; the pleasure centre (the nucleus accumbens), which releases dopamine in response to both the achievement of goals and the anticipation of such achievement under uncertain conditions; and the fight-or-flight mechanism, specifically the arousal produced by epinephrine (which enhances rewards) and the anxiety that results when the amygdala is activated when an individual feels out of their depth. Between these three mechanisms, all the many and various optimal experiences result.

I welcome discussion in the comments.

Comments

[Matt Mower]

>An amateur in meditation or yoga does not experience
>anxiety when they fail to successfully meditate. They are
>just unable to quiet their conscious mind and may, as a
>result, feel bored – the opposite of what Flow theory
>suggests.

Unless I misunderstand you then I have to disagree. My experiences with attempting (largely unsuccessfully) to meditate do seem to create anxiety, and not boredom.

I don't shrug my shoulders and go do something else more interested instead. I find that I experience a build up of tension that becomes unpleasant and go do something else instead.

[zenBen]

"All that remains is to validate this hypothesis experimentally which I shall not be doing since I am a game designer and not an experimental researcher."

I'll be doing my best :)

"The principle concern I have is integrating yogic activities with the above model"

I believe that the problem lies with the model, because Flow is a distinct form of a more general experience, that is: process engagement. Flow focuses too much on explicit goals to model the experience of meditation, but many of the same principles apply if you take away the requirement to be aiming for something.

[organic io]

More fascinating ideas, Chris.

But, for some reason, reading about such reductive views can get me down. I know you said you don't believe that any 1 specific function of the brain explains behavior.. And I don't believe that either.

But reading about it still bothers me -- if things can be reduced to simple brain chemistry, aren't we all just participating in some sort of odd mental masturbation every time we do any activity we consider enjoyable? Like the biological purpose of pleasure is to prevent us from becoming so depressed that we all commit suicide.

... and what drives us to enjoy one activity over another? Why shouldn't I enjoy knitting or picking up leaves as much as I do video games? If my life had taken a different path, I'm sure I would. Or if video games or computers or even music didn't exist.. I would certainly have different hobbies.

Slightly off topic but still relevant I think:
The addictiveness of video games concerns me, though.. Especially as we go into the future. Certain people are going to have to be very careful, especially those with addictive personalities. In the 70's or 80's, you never read about kids dying at their computer because they played for 3 days straight... But now games are so much more advanced and immersive.. I fear things may get worse. You know like you see on movies where in the future, everyone is wearing VR goggles or something like that.

But, perhaps I shouldn't view it as a negative, but more like an opportunity for myself -- if this kind of thing ever happens, I can find a niche for myself as a virtual-world-addiction therapist. :)

Sorry to ramble. Your blog is always just very thought provoking. :)

-Scott

[Sirc]

"Firstly, activity in the frontal lobe, which links to Csikszentmihalyi's “concentrating and focusing”, “loss of the feeling of self-consciousness”, “distorted sense of time”, “sense of personal control” and “action awareness merging”. According to my hypothesis, all Flow states involve significant activation of the frontal lobe in some way, resulting in the state of intense focus."

I don't think it's necessary to be in a state of "intense focus" to be in the flow channel. When we think of difficulty, we might think of fighting a tough boss in an action game, or passing a tough platforming level in a Mario game, or solving a puzzle. Those will likely put the player in a state of intense focus, but they're not the full extent of the challenges a game could present.

Basically what I'm trying to say is that some games do not require the player to be in a state of intense focus in order to overcome their challenges. And yet they are still challenges, thus there is still difficulty, thus the opportunity to be in a flow state, and there is still a steady release of dopamine because the player is working towards a goal.

"But reading about it still bothers me -- if things can be reduced to simple brain chemistry, aren't we all just participating in some sort of odd mental masturbation every time we do any activity we consider enjoyable?"

And what's so wrong about that? :)
Speaking of which, some say philosophy itself is by and large just a form of mental masturbation :D (don't kill me, Chris!)

"The addictiveness of video games concerns me, though.. Especially as we go into the future. Certain people are going to have to be very careful, especially those with addictive personalities. In the 70's or 80's, you never read about kids dying at their computer because they played for 3 days straight..."

Eh, one less idiot to waste our oxygen. If I'm not mistaken, this kind of thing only happens with MMORPGs (Incidentally, probably the videogame genre I loathe the most). I could be out of the loop, but I've never heard of someone dying because they played a single player game for 3 days straight. Which leads me to believe that it might not be the game itself that causes the self destructive behavior, but the presence of other people.

I find it hard to believe that these people played a game for 3 days straight because they were enjoying it so much. I mainly say this because I personally stop enjoying a game if I'm too tired. At which point I stop and go to sleep, otherwise I feel like I'm cheating myself out of a good experience by continuing despite no longer enjoying it.

And it makes sense, right? After all, the only purpose of a game is to entertain you. But in MMOs, something else is going on. Something tells me that these kids might not be thinking "Man, this is so much fun, I can't stop!" But rather "I'm dying, but I have to go on! If I'm the first level 80 Death Knight, then everyone will finally admire and respect me!" or maybe "I have to become high level fast so that the rest of the guild will stop looking down on me!" or perhaps even "I'm dying, but we've been fighting this boss for 10 hours! If I leave now, my friends will hate me!"

Bah, now I'm the one rambling. In conclusion, MMOs suck.

[zenBen]

"I don't think it's necessary to be in a state of "intense focus" to be in the flow channel. When we think of difficulty, we might think of fighting a tough boss in an action game, or passing a tough platforming level in a Mario game, or solving a puzzle. Those will likely put the player in a state of intense focus, but they're not the full extent of the challenges a game could present."

I think that focus is a necessary prerequisite to Flow (it is absolutely central to all the studies), and this is one reason why the Flow experience should be seen as almost like a subset of optimal experience. Or perhaps it is more accurate to say that the Flow model restricts itself (purposely) to focus-dependent experiences.

[Sirc]

I see. So Chris's statement "According to my hypothesis, all Flow states involve significant activation of the frontal lobe in some way, resulting in the state of intense focus." is tautological because the flow model itself chooses to ignore gaming experiences which don't result in intense focus?

I don't see why it should restrict itself though. It seems to me like it could easily apply to other, less focus-dependent experiences.

[zenBen]

"I don't see why it should restrict itself though. It seems to me like it could easily apply to other, less focus-dependent experiences."

Its my view that the Flow model was derived from a study of inherently focus-oriented experiences, and that it shouldn't be stretched too far in trying to explain similar experiences that drop the focus. That's not to say that a model for the latter type of experience wouldn't look very similar to Flow. Like I've been saying, I think you're dealing with variations on a theme. For instance, I believe that some activities that require great focus can lead to experiences where all conscious focus falls away, and all you're left with is a meditative-like state of doing, not thinking...I'd cite my karate practice as an example.

On the other hand, stretching the model could be a shortcut that works, in which case why not?

[Chris]

Wow! Lots more discussion than I thought this would produce. Thanks for getting involved!

Matt: "My experiences with attempting (largely unsuccessfully) to meditate do seem to create anxiety, and not boredom."

Interesting! I may have jumped to a conclusion based on too few cases, then. I shall have to investigate further! :)

zenBen: "I believe that the problem lies with the model, because Flow is a distinct form of a more general experience, that is: process engagement. Flow focuses too much on explicit goals to model the experience of meditation, but many of the same principles apply if you take away the requirement to be aiming for something."

This account makes a lot of sense to me. Do you have some good process engagement papers to throw my way? ;)

Scott: I hope my piece on "you are not your brain" goes part way to addressing your concerns about overly reductive views!

"... and what drives us to enjoy one activity over another? Why shouldn't I enjoy knitting or picking up leaves as much as I do video games? If my life had taken a different path, I'm sure I would. Or if video games or computers or even music didn't exist.. I would certainly have different hobbies."

But you are a product of your circumstances as much as your biology - your love of videogames is a vital part of who you are. Yes, in another life path you could have been a demon knitter, perhaps, but that is not who you are in this life.

"The addictiveness of video games concerns me, though..."

Yes, I am working towards a piece on this. I'm not quite ready, though - I have to be very careful, because if I say the wrong thing it will surely be used against me as a very big stick! :)

"Sorry to ramble. Your blog is always just very thought provoking. :)"

You never, never, ever need to apologise to me of all people for rambling! You've seen the way I work - rambling is the very quintessence of Only a Game, and I love it when I provoke players into rambling too - it means we're all playing the same game, and that's what I love about this blog.

Thanks too, for the kind words!

Sirc: I think zenBen addressed a lot of your issues in his comments.

"So Chris's statement 'According to my hypothesis, all Flow states involve significant activation of the frontal lobe in some way, resulting in the state of intense focus.' is tautological because the flow model itself chooses to ignore gaming experiences which don't result in intense focus?"

That Flow involves intense focus may be tautological, but I don't think anyone else has pointed to the frontal lobe as the brain region, and my hypothesis here is all about tying Flow to neurobiology. I'll admit, it doesn't seem like I'm doing much in this instance because it seems like a given, but science doesn't allow us to jump the connective steps. Hence, it's a hypothesis, not a tautology. :)

"Speaking of which, some say philosophy itself is by and large just a form of mental masturbation :D (don't kill me, Chris!)"

Well, what's wrong with masturbation, exactly? >:)

"Eh, one less idiot to waste our oxygen. If I'm not mistaken, this kind of thing only happens with MMORPGs"

I believe one of the people who died was playing Starcraft, so no, not just MMORPGs. (He was probably playing multiplayer online, though, so your 'presence of other people' may still have a role). I will write about this at some point.

"Something tells me that these kids might not be thinking 'Man, this is so much fun, I can't stop!' But rather 'I'm dying, but I have to go on! If I'm the first level 80 Death Knight, then everyone will finally admire and respect me!' or maybe 'I have to become high level fast so that the rest of the guild will stop looking down on me!' or perhaps even 'I'm dying, but we've been fighting this boss for 10 hours! If I leave now, my friends will hate me!'"

I don't think the people in question know they are dying. But I agree that their motivations for continuing to play may be socially motivated - but they could be as simple as "I really want to beat this guy".

---

And just to complete this run of neurobiology posts, on Tuesday I'll run the piece tying Temperament Theory to the brain. It's a bit speculative, but it should be interesting all the same.

Thanks for the discussion everyone!

[zenBen]

"Do you have some good process engagement papers to throw my way? ;)"

If I do, they don't spring to mind*. All I can remember or forsee is data, data, mining data, forever and ever and ever...I don't remember sunlight, Sam :)

*Having said that, most people to date have seemed to focus on adapting Flow to fit every context, rather than expand it. I suppose you need some good psychological background to expand it, and I mostly read CS/HCI/cogSci type papers.

[organic io [demon knitter]]

Chris, thanks for your insightful responses, and also your post "You are not your brain". I feel honored to have an entire post directed to a question that I raised... Although I still haven't figured out how to respond to it. :)

"Yes, in another life path you could have been a demon knitter, perhaps, but that is not who you are in this life. "

:-D I believe I have finally discovered my character class for only a game!

Can't wait to see how you approach the topic of addictivness in videogames.

And finally...

"I don't think the people in question know they are dying."

This is how I feel about it too. I've read about half a dozen of these news blurbs, but I don't recall them ever stating the actual cause of death. I think heart attack is probably the most likely -- lack of sleep, stress, combined with all the energy drinks that they are consuming... Not a good combination.
A lot of these deaths could probably have been prevented if the person had just relaxed and taken a break 30 minutes before their death.