Contribution
A survey of how flow is measured, and an argument about what it is. Across 42 PsycINFO studies from July 2014 to July 2019, flow was operationalized in 24 distinct ways (Table 1). Three inconsistencies are traced to their sources: continuous against discrete, enjoyable or not, and the state against its conditions. The paper’s proposal is to measure flow only as Csikszentmihalyi conceived it and to use task involvement and intrinsic motivation for everything below it.
Key insight
Much of what is called flow is absorption. Absorption, as task involvement, is a purely cognitive matter of attention and is not enjoyable in itself; flow, as Csikszentmihalyi defined it, is absorption that is also highly enjoyable and rare. Whether “flow” is positive therefore depends first on which of the two a study measured.
Assumptions
- Csikszentmihalyi’s conception is the only valid one (Summary and Conclusion): “flow” is a proper noun with his conceptualization “pre-installed”, so it is “the default conceptualization, and this is true regardless of its merits”.
- Appraisal theory (Issue 2, Source #1): appraisals are usually automatic, need little attention, and become more automatic with practice, so emotion can arise during absorbed, skilled activity.
- Emotion does not need self-awareness, only its recognition and labelling does (Issue 2, Source #2), citing infants, non-human mammals and children without a cortex.
Key results
- The review (Table 1; Appendix). 111 PsycINFO hits under the heading “Flow (consciousness state)”, 70 with an operationalization, 42 after exclusions (trait flow, collective flow, unclear or multiple operationalizations, validation studies, non-English). 24 distinct operationalizations; most continuous; 17 of 42 without enjoyment or autotelic experience; most common scales mix experience items with condition items.
- Issue 1: discrete or continuous. Csikszentmihalyi described flow as rare and optimal (1975 preface, quoted: “On the rare occasions that it happens, we feel a sense of exhilaration, a deep sense of enjoyment that is long cherished”). Continuous scales were adopted for statistical power and turn “low flow” into “a contradiction in terms”. They also lose discriminant validity against task involvement.
- Issue 2: is flow enjoyable? Csikszentmihalyi “conceptualized flow as
an enjoyable experience” and has kept to it (1975 preface, quoted: “the
experience itself is so enjoyable they will do it even at great cost, for
the sheer sake of doing it”). Three sources of the view that it is not:
- Seligman (Authentic Happiness, Flourish), quoted: “it is the absence of emotion, of any kind of consciousness, that is at the heart of flow”, and “the concentrated attention that flow requires uses up all the cognitive and emotional resources that make up thought and feeling.” Abuhamdeh answers from appraisal theory: appraisals are automatic and need not use attention.
- Absence of self-awareness read as absence of emotion (Kyriazos et al. 2018). Abuhamdeh: “self-awareness is not a precondition for the experience of emotions, only the recognition and labeling of them.”
- Csikszentmihalyi’s own word “pleasure”. He quotes Flow (1990): “None of these [flow] experiences may be particularly pleasurable at the time they are taking place, but afterward we think back on them and say, ‘That really was fun’ and wish they would happen again.” He answers that Csikszentmihalyi reserves “pleasure” for experiences that satisfy biological needs, “restorative homeostatic experiences”, and that an artist working all night without food or rest had no “pleasure” in that sense, “But this should not be misinterpreted as implying that the artist did not enjoy him/herself.”
- Issue 3: state against conditions. Clear goals, immediate feedback and challenge–skill balance are conditions, not the state; scales that average them in conflate cause and effect. The 10-item Core Flow Scale is the exception and “may be the best option”.
- Where to draw the boundary (Question 1). No sharp threshold is likely; the quadrant model labels a quarter or more of daily experience as flow; self-identified flow episodes (the Flow Questionnaire) come closest to the concept but were used by one study in 42. Taxometric methods are suggested.
- Below flow (Question 2). Task involvement, “the degree to which an individual concentrates on and becomes absorbed in an activity”, is “a purely cognitive phenomenon representing the degree of attentional involvement in an activity; it is not inherently enjoyable and motivating in concept, though it often predicts both.” Intrinsic motivation, measured by enjoyment and interest, captures flow’s emotional and motivational side.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Flow is operationalized inconsistently: 24 ways in 42 recent studies | strong as a description | the review, Table 1; non-exhaustive by its own criteria |
| C2 | On Csikszentmihalyi’s conception flow is inherently enjoyable | strong as exegesis | quotations from 1975 and later work |
| C3 | Flow is enjoyable during the episode, not only afterwards | weak | argument from appraisal theory and from a reading of Csikszentmihalyi’s “pleasure”; no data |
| C4 | Absorption (task involvement) is not inherently enjoyable | moderate | cited prior work (Abuhamdeh & Csikszentmihalyi 2012a), not reread here |
| C5 | Flow should be measured as a discrete, enjoyable state distinct from its conditions | normative | argument |
Concepts
- Flow — on Csikszentmihalyi’s conception, a rare, discrete, “optimal” state of consciousness that is highly enjoyable and autotelic.
- Autotelic — intrinsically rewarding; the paper treats it as synonymous with intrinsic motivation.
- Pleasure against enjoyment (Csikszentmihalyi, as reported) — pleasure is the satisfaction of biological needs, “restorative homeostatic experiences”; enjoyment is the reward of absorbed, challenging activity, which may conflict with those needs.
- Task involvement — degree of concentration on and absorption in an activity; cognitive, continuous, not inherently enjoyable.
- Conditions of flow — clear goals, immediate feedback, balance of challenge and skill (Figure 1, from Nakamura and Csikszentmihalyi 2002).
Connections
It rests on Csikszentmihalyi (1975, 1990) and on Nakamura and Csikszentmihalyi (2002, LIT-914 in the record, unread), from which its Figure 1 is taken. Its foil is Seligman (2002, 2011). It draws on appraisal theory (Moors, Scherer, Lazarus) and on the task-involvement literature (Harackiewicz; Elliot). Swann et al. (2018) is cited for a Kuhnian diagnosis of the field.
Bearing on the record
- CLAIM-441: partly. The objection waits on whether flow is reported as positive during the episode, and names exit 3, that “people in flow are absorbed rather than happy, and that the enjoyment is reported afterwards”. This reading confirms that the literature is divided on exactly that point: Seligman says flow is emotionless, and Csikszentmihalyi (1990), as quoted, says flow experiences may not be “particularly pleasurable at the time” and are called “fun” afterwards. Abuhamdeh argues that flow is enjoyed during the episode, but by definition and by appraisal theory, not by data, and the scales he reviews are filled in after the activity. So exit 3 is neither shown nor excluded, and the objection’s defeater (“flow is not reported as positive during the episode, only after it”) is not met. Two things the reading adds. First, flow is enjoyable by Csikszentmihalyi’s definition, so “flow is positive” is partly analytic: an absorbed episode that is not enjoyed is, on this view, task involvement, which is “not inherently enjoyable”. Second, as Abuhamdeh reports it, Csikszentmihalyi separates pleasure, the restoration of homeostatic needs, from enjoyment, which can run against them. That bears on exit 1: flow’s positivity is placed on the enjoyment side and not on the side of an ordering of the organism’s states by their bearing on its maintenance. That is half of exit 1; it does not show that the enjoyment is intrinsic pleasantness.
- TERM-076: partly. Its nonvacuity condition makes valence undefined where nothing is contested, and flow is the case pressed against it. The reading does not decide whether flow is positive in the episode. It does supply a distinction the record can use: absorption alone (task involvement, as in CASE-051’s absorbed writer) is not inherently enjoyable, so undefined valence there is no anomaly; and if flow’s enjoyment is not homeostatic “pleasure”, it takes flow’s positivity off the maintenance ordering, half of exit 1, without showing it is intrinsic pleasantness; exit 1 in full would leave the definition standing. Nothing here bears on exit 4.
- CLAIM-440: partly, by way of CLAIM-441. The same: the flow case against its first consequence is narrowed, not answered.
- CLAIM-162 and CLAIM-433. CLAIM-162 reads flow as nothing overridden; this paper does not discuss conflict among motives. CLAIM-433’s restriction to goal-conduciveness valence is the distinction exit 1 relies on, and the pleasure/enjoyment split is evidence for it within flow research, at second hand.
- LIT-914. It stays unread. This paper uses it only for Figure 1, so it is not a substitute for reading it.
- No instruction for machine-learning practice; nothing for the anthology.
Limitations
- No data on experience. The review counts operationalizations; the claims about how flow feels rest on quotations and argument.
- Second-hand quotations. Csikszentmihalyi and Seligman are quoted, not read here; the “pleasure” passage is interpreted by Abuhamdeh.
- Definition by stipulation. That Csikszentmihalyi’s conception is the only valid one “regardless of its merits” makes the enjoyment of flow partly a matter of naming.
- The review is not exhaustive (Appendix): five years, one database, one subject heading.
Open questions
- Is flow positive while it happens? Experience sampling during absorbed, skilled activity, with valence probed at the moment and again later, would settle exit 3; the reading names no such study.
- Does enjoyment in flow dissociate from goal-conduciveness, as the pleasure/enjoyment split suggests? That would decide between exits 1 and 4.