Study at a glance
- Authors: Kalina Christoff, Caitlin Mills, Jessica R. Andrews-Hanna, Zachary C. Irving, Evan Thompson, Kieran C.R. Fox, Julia W.Y. Kam
- Year: 2018
- Journal: Trends in Cognitive Sciences 22(11): 957–959
- DOI: 10.1016/j.tics.2018.07.004
- Design: Theoretical / conceptual — correspondence Letter (Cell Press REVIEWS), published with a paired reply from Seli et al. (pp. 959–961). No new data.
Key findings
- The field has a definitional problem, and the two camps disagree on what to do about it. ‘Mind-wandering’ is used for phenomena as disparate as attentional lapses, purposeful task-unrelated planning, free-flowing creative ideation and perseverative rumination. Both sides agree this lumping is a problem; they disagree on the remedy.
- Christoff et al.’s position — the dynamic framework (a definitional approach). Mind-wandering has an essential, defining feature: thoughts arise and proceed in a relatively free, unconstrained fashion (from Christoff et al. 2016, Nat Rev Neurosci). A defining feature is needed because ‘for a research field to exist, it needs a definable focus that separates it from other fields’ — without one there is ‘no field of mind-wandering research separable from research on thought in general.’
- The consequence that divides the two views: highly constrained task-unrelated thought. Under the dynamic framework, perseverative rumination and deliberately planning a dinner while in calculus class are not mind-wandering (they are strongly constrained). Under family-resemblances they are, because some researchers have labelled them so. This is the concrete empirical/taxonomic stake, not a semantic quibble.
- Against family-resemblances (Seli et al. 2018, TICS): it dissolves the concept. If ‘there are no specific features that a thought must have’ to join the mind-wandering family, the boundary becomes maximally porous — any thought one researcher calls mind-wandering is admitted — so the term broadens until it is ‘synonymous with thought.’ Christoff et al. read this not as a new approach but as the field’s existing tacit practice, and as the cause of the lumping problem rather than its cure.
- Seli et al.’s reply — The Family-Resemblances Framework Remains Well Clad. (i) It is a false dichotomy (an ignoratio elenchi) that one must either adopt an exclusive definition or proceed unscientifically; the third option is a cluster concept with probabilistic, graded, multidimensional membership and more/less prototypical exemplars. (ii) The dynamic framework fails its own test: ‘relative lack of constraint’ is itself insufficient to distinguish mind-wandering from other thought. (iii) A field can exist without definitional consensus — the existence of mind-wandering research has not in fact been questioned.
- Both frameworks agree membership is graded. The single load-bearing disagreement is whether some features are more defining than others: the dynamic framework privileges unconstrained dynamics as necessary; family-resemblances denies any feature is privileged.
Limitations
- Not an interoception paper, and not an empirical one. This is a two-round theoretical correspondence; it introduces no data, no sample, no method. It earns a wiki page structurally — as the primary statement of a debate the wiki’s Body-wandering construct sits inside — not as evidence about the body.
- The dispute is partly verbal. Both sides accept graded membership and both want researchers to specify which thought type they study; the live residue is whether ‘unconstrained dynamics’ is a necessary feature. Neither Letter adjudicates this empirically, and the exchange turns partly on philosophy-of-science framing (natural kind vs cluster concept vs dynamic definition).
- Scope is definitional, not mechanistic. Nothing here bears on the neural or interoceptive substrate of spontaneous thought; the relevance to this wiki is entirely through what counts as mind-wandering when a study (like Banellis et al.) samples it.
Connections
- Concepts: Mind-wandering, Body-wandering, Default mode network
- Debates: How should mind-wandering be defined?
- Related studies: Banellis et al. (2026), Weng et al. (2020)
A theoretical Letter in Trends in Cognitive Sciences, published with its rebuttal (Seli et al., The Family-Resemblances Framework for Mind-Wandering Remains Well Clad, same issue). The pair is the compact primary statement of a dispute the wiki had been leaning on without ever recording: what, if anything, makes a thought a case of mind-wandering.
It is here for one reason. The wiki’s Body-wandering construct — the embodied slice of the resting stream of thought, from Banellis et al. (2026) — is a kind of mind-wandering, and its own concept page notes that “the wiki had no page for mind-wandering or spontaneous thought at all” when it was written. Whether body-wandering counts as mind-wandering at all depends on which side of this debate one stands on, because body-wandering can be either free-drifting or the ruminative, constrained fixation on a bodily symptom that the two frameworks classify oppositely.
The disagreement, stripped down
Everyone in the exchange agrees the term has been overstretched, and agrees that membership in the category is graded rather than all-or-none. The entire fight is over one question:
Is there a feature a thought must have to be mind-wandering — or not?
Christoff et al. (dynamic framework): yes. The defining feature is relatively unconstrained dynamics — thoughts that arise and move freely, not held on course by either deliberate goals or automatic (e.g. affective, perseverative) constraints. This is a definitional approach: it draws a boundary and accepts that some things previously called mind-wandering fall outside it.
Seli et al. (family-resemblances): no. Mind-wandering is a cluster concept — a family whose members share overlapping resemblances with no single common thread, membership probabilistic and multidimensional, some exemplars more prototypical than others. No feature is privileged over any other.
The test case that makes it concrete
The two views are distinguishable exactly where they classify the same thoughts differently. Christoff et al.’s own table:
| type of thought | dynamic framework | family-resemblances |
|---|---|---|
| perseverative task-unrelated thoughts (rumination) | not MW | is MW |
| purposeful thoughts about holiday activities (task-unrelated) | not MW | is MW |
| deliberately planning a dinner date in calculus class | not MW | is MW |
| letting one’s mind wander by a lake | is MW | is MW |
The dynamic framework excludes the top three because they are constrained — by perseveration or by deliberate goals. This is the stake for a wiki that cares about the body: ruminating on a symptom and idly noticing the breath are both “thought about the body,” but only the second is mind-wandering under the dynamic view. That distinction is the same one the wiki keeps meeting from the clinical side — elaborative vs sensory reactivity, Decentering, Cognitive reactivity (to mood) — where what harms is not the bodily content but whether thought is locked-on and elaborating.
Each side’s strongest move
Christoff et al.’s sharpest argument is the dissolution worry: family-resemblances makes the boundary so porous that “as long as at least one researcher refers to a type of thought as ‘mind-wandering’, this type of thought would become part of the family,” until the concept is synonymous with thought and the field loses any focus that separates it from cognition in general — “the mind-wandering emperor really has no clothes.”
Seli et al.’s sharpest reply is the false-dichotomy charge: the choice is not “exclusive definition or nothing.” A cluster concept is a third, scientific option with real structure (graded, probabilistic, prototype-based), and — turning the knife — the dynamic framework’s own criterion, “relative lack of constraint,” is itself vague enough that it fails to cleanly separate mind-wandering from other thought. And empirically, the field has flourished without consensus, so the existential threat is overstated.
What it does to the wiki
It supplies the definition the Body-wandering page was missing, and it supplies it as a live dispute rather than a settled fact — which is the honest state of the field. The wiki now records mind-wandering as a concept and the definitional split as a debate.
It contradicts nothing. No wiki page had committed to a definition of mind-wandering; Banellis et al. and Weng et al.’s EMBODY both sample or decode mind-wandering operationally without adjudicating what it essentially is. This Letter names the assumption underneath both.
It plants one flag worth keeping. If the dynamic framework is right, then the empirical question “is body-wandering protective or harmful?” (Is more interoceptive awareness better?) may be mis-posed whenever body-wandering blends free drift toward the body with constrained fixation on it — two states this debate says should not share a label. Banellis et al.’s single-probe MDES cannot separate them; the debate says they must be separated.
Provenance
Two Letters in the Correspondence section of Trends in Cognitive Sciences 22(11), November 2018. Christoff (UBC Psychology / Centre for Brain Health) is the senior voice of the dynamic-framework camp; the reply is led by Paul Seli (then Harvard) with Kane, Metzinger, Smallwood, Schacter, Maillet, Schooler and Smilek — Jonathan Smallwood among them, the same MDES architect whose framework Banellis et al. later extend to the body, which places the wiki’s body-wandering instrument on the family-resemblances side of its own author lineage. The file also contains, on its first two pages, an unrelated Continuous Flash Suppression correspondence (Sklar, Deouell & Hassin) that is not ingested.