Daydreaming and Thought-Sampling

OK, is there anything new in the daydreaming article in Science?


Fig. 2. Graphs depict regions that exhibited a significant positive relation [with a propensity to daydream], r(14) > 0.50, P < .05 (A) Bilateral mPFC; (B) Bilateral precuneus and posterior cingulate; (C) Right cingulate; (D) Left insula; (E) Right insula. [from Mason et al., 2007]

There's an extensive body of literature on what is called daydreaming or mind-wandering. Raichle and colleagues have argued for a "default mode" or "resting state" of brain function that engages a certain network of brain regions (posterior cingulate and precuneus and medial prefrontal cortex) during "rest." These regions become DEactivated when people are engaged in the typical types of cognitive tasks they're asked to do in a scanner. So it's really only a "resting state" when compared to doing, say, the Stroop task. When asked to rest and stare at a plus sign, you may engage in idle daydreaming or think about what you'll have for dinner or remember your hot date from last night or silently sing.

[You can read more about one of these "default mode" regions, the precuneus, at the links below¹]

So what's up in the new experiment? The idea was to try to manipulate the degree of daydreaming during cognitive tasks by training subjects to become proficient at them. A verbal working memory task involved 4 four-letter sequences (e.g., ZVRT) that had to be recalled either forwards or backwards, while a visuospatial task involved 4 finger-tapping sequences. After 3 days of training with the sequences, and a day of "thought-sampling" (described below), scanning was conducted on day 5. On day 4, participants again practiced the learned sequences but were interrupted periodically and asked

to indicate whether they were having an "irrelevant thought." Consistent with previous studies of this nature, the term "irrelevant thought" was defined to participants as "thoughts that do not facilitate performance and are not immediate reactions to perceptual information gleaned over the course of a trial."

Subjects were more likely to report having irrelevant thoughts during learned sequences compared to novel sequences introduced into the experiment. The entire experiment hinges on accurate introspective reports of daydreaming (stimulus-independent thought, SIT) on the day before the scanning session, since it wasn't assessed during:

On day 4, the proportion of sampled thoughts participants classified as SIT varied by block type (baseline, practiced, or novel)... Participants reported a greater proportion of SIT during the baseline blocks (mean=0.93; SD=0.16) than during both practiced blocks (mean=0.32, SD=0.20), t(17)=9.22, p less than 0.01, and novel blocks (mean=0.22, SD=0.18), t(17)=10.96, p less than 0.01. Participants reported a significantly greater proportion of SIT during the practiced blocks than during the novel blocks, t(17) = 2.11, p less than 0.05...

Scanning took place on day 5, when learned sequences, new sequences, and a baseline condition (no task) were interleaved. Two weeks after the study was over, participants were sent a questionnaire to assess daydream frequency (e.g., On a long bus, train, or airplane ride I lose myself in thought). Scores on the daydream scale were correlated with activations in the default network during practiced stimulus sequences (see illustration above), when presumably, people's minds would wander.

How were the data analyzed? The default network for each subject was determined by comparing baseline vs. working memory task conditions.

This comparison revealed significantly greater recruitment at rest in a distributed network of regions that included aspects of the posterior cingulate and the precuneus [Brodmann areas (BAs) 23 and 31], the posterior lateral cortices (BAs 40 and 39), the insular cortices, the cingulate (BA 24), and aspects of both ventral and dorsal medial prefrontal cortex (mPFC) [BAs 6, premotor and supplementary motor cortex; 8, including frontal eye field; 9, dorsolateral prefrontal cortex; and 10, frontopolar area (most rostral part of superior and middle frontal gyri)] (8, 9).

. . .

The resulting default network contrast was subsequently converted to a binary image and used as an ‘inclusive’ mask in subsequent analyses (at a more lenient threshold of p < .05, k = 10). In effect, this made it possible to identify differences in cortical activity during ‘practiced’ blocks relative to ‘novel’ blocks that occurred within the default network.

However, once the default network was determined, the practiced vs. novel contrast did use a threshold of p<.001. All brain areas showing greater activity for practiced vs. novel that fell outside the default network were discarded as irrelevant for daydreaming. [!!]

So what were the results? Not surprisingly, regions of the default network were more "active" during practiced than novel blocks. The word "active" is in quotes, because really, the finding is that certain default mode regions (e.g., the insula, the posterior cingulate) showed less DEactivation (or no change at all) for practiced than for novel sequences (see figure below). Meaning that daydreams were associated with no brain activity, or a tiny deactivation, in some cases. Hmm.


Regions of the insula (left) and posterior cingulate (right) that exhibited greater activity during the ‘practiced’ blocks (red) relative to the ‘novel’ blocks (blue). [from Mason et al., 2007]

OK, then.

In the final installment, a published critique of the default mode concept (Morcom & Fletcher, 2006) will be discussed.

¹Precuneus References:

Hypnosis and Consciousness

Are You Conscious of Your Precuneus?

The Precuneus and Recovery From a Minimally Conscious State