The Wandering Mind: Neuroimaging of the Default Mode

The default network constitutes a collection of brain regions associated with and active during mind wandering. Although studies on the cognitive foundations of mind wandering abound, this week’s readings are the first (so far in class) to address the brain activity associated with mind wandering in an in-depth manner. I find this very exciting because it allows us to pair evidence from cognitive studies and with neurobiological data to generate a more complete image of mind wandering. Interestingly, despite the similarity in the neuroimaging methods used across all of this week’s studies, their results are not entirely consistent and the authors sometimes reach rather distinct conclusions based on the pattern of brain activation shown in a resting state.

Summaries and Comparative Analysis

Fox et. al. (2005)’s study was motivated by the previously established finding attention-demanding tasks routinely lead to the activation of some brain regions (task positive regions such as frontal eye field and middle temporal region) and the deactivation of others (task negative such as medial prefrontal cortex and posterior cingulate/prenucleus). The authors were interested in investigating the degree to which this “task related dichotomy” is present intrinsically in the resting human brain. To test this theory, Fox et. al. collected fMRI data from ten normal subjects and analyzed the correlations and anti-correlations in the spontaneous BOLD (blood oxygen level dependent signal) associated with three task-positive brain regions and three task-negative brain regions. From this analysis, the authors found the existence of two “diametrically opposed” anti-correlated networks in the resting brain.

The task-positive network is composed of regions activated in goal-directed tasks whereas the task-negative network consists of regions with suppressed/decreased activity during task response. The latter is usually referred to as the “default system.” The authors argue that this data supports the concept of a resting-state functionality in the brain.

Mason et. al. (2007) expanded on Fox et. al.’s study by investigating the role of the default network in mind wandering or the formation of stimulus-independent thoughts. Specifically, he uses thought sampling and brain imaging studies to demonstrate that mind wandering is associated with activity in the default network, which constitutes brain regions active during rest. fMRI studies were used to demonstrate a correlation between high-incidence mind wandering periods and an increase in default network activation and subject self-reports on individual tendency to engage in mind wandering were also related to their default network activation patterns. Furthermore, no default network area exhibited significant activity during low-incidence mind wandering period. Taken together, these results support the theory that the tonic activity present in the default network during resting states is associated with mind wandering.

The authors offer two possible explanations for the functional roles of mind wandering. For one, it could aid in the maintenance of an optimal arousal level that enhances performance on “mundane tasks.” Another possibility could be that SITs help make sense of past, present, and future experiences. These results largely agree with those of Fox et. al in that they also support a resting state functionality. However, the results and interpretations also differ from those of Fox et. al. in an important respect. Whereas the latter proposes that the resting brain is characterized by representations of two anti-correlated networks (the default network and the task-positive network), Mason et. al. proposes that resting-state mind wandering is attributable only to the default network.

Andrews-Hanna et. al. (2010) was interested in distinguishing between functional contributions of the default network to external attention shifts and internal mentation. To this end, the authors essentially “decoupled” these two processes by manipulating factors that specifically promote spontaneous cognition. Specifically, the authors conducted a neuroimaging study in which subjects were asked to detect a signal present under three fixation conditions: broad attention, focal attention, and passive attention. The authors found that broad attention had different effects than the default network. Furthermore, stimuli and responses were held constant and only expectations differed across these conditions, thus lending to the separation of spontaneous cognition from factors that affect the scope of the external environment. Results indicate that the default network is involved in spontaneous cognition, but not on the level of broad attention. Thus, they argue that the default network is not responsible for broad external attention and thus is involved in internal mentation rather than strictly mind wandering. These results and interpretations clearly differ from those of Mason et. al. and to some respect, also Fox et. al.

Stawarczyk et. al. (2011) expands upon the previous studies’ results by providing a multi-dimensional scheme of consciousness as a function of task-relatedness and stimulus-dependency. Specifically, the authors introduce four classes of conscious experiences: complete task-directed focus, districted focus by irrelevant sensations, interfering thought related to the task, and mind wandering.

Subjects were probed for each of the four conditions to investigate the brain activation patterns associated with each class of conscious experience. Like previous studies, they found specific brain regions highly conducive to mind wandering. The interesting part of this study lies in the fact that they addressed the concern voiced in Andrews-Hanna’s 2010 study, namely the tangling mind wandering and other processes. Specifically, the multi-dimensional scheme lends to the possibility that the neurobiological mechanisms for mind wandering and external shifts in attention may overlap and indeed coexist. Furthermore, the consideration of both task-relatedness and stimulus dependency also introduces a variable of complexity (that accounts for the complexity of the neural networks) not present in the previous studies. This incorporation proved essential, as evidenced by the authors’ finding of an additive effect along the midline default mode network regions.

Questions and Comments

Although I find this week’s readings particularly interesting, I still have some questions about the methods used and result interpretation of some studies. For instance, virtually all of the authors interpreted results from neuroimaging studies without rigorous analysis of possible confounding factors. Fox et. al. and Mason et. al.’s studies provides cases in point. Brain regions activated during rest are classified as “task negative.” However, it is possible that these brain regions are in fact engaged in essential tasks, and thus the term “task negative” seems like an inappropriate description of these brain areas. Furthermore, given the complex and interconnected nature of the brain, the search for an engram or even a collection of engrams for mind wandering seems to ignore the essential role of virtually every brain region in maintaining coherent thought and understanding.

Furthermore, I also took issue with some of the methods and experimental design in some of these studies. For instance, Fox et. al. only tested ten subjects in his fMRI study, and while the results were definitely interesting and holds a lot of potential for future investigation, the low number of subjects increase the probability that the results presented could just be an artifact of the sample population chosen. In addition, the self-reporting technique utilized by both Mason et. al. and Andrews-Hanna et. al. present the same problem discussed in early posts: inconsistency across subjects’ reports stemming from the absence of a rigorous definition of stimulus-independent thought and precise mechanism of evaluating its presence.

On the brighter note, I found the dimensions of conscious experiences presented by Stawarczyk et. al. (2011) very enlightening and insightful. Although, like previous authors, Stawarczyk and his team also found certain brain regions highly conducive to mind wandering, I think this multi-dimensional scheme on consciousness achieves what the three previous studies failed to achieve: an integration of mind wandering with other cognitive experiences to provide an overall image of conscious experiences and thought. Furthermore, this method provides data on the relative activations of brain regions in mind wandering compared to other conscious thought, and thus provides a more physiological description of conscious thought and mind wandering. Clearly, I find this study the most compelling out of the four studies we read for this week.

Links to Neuroscience

Since this week’s reports were primarily neuroimaging-based, the links to neuroscience are intrinsic and already present. However, I think it would be interesting to expand on some of the observations and interpretations offered in these studies. For example, I would like to further the possibility offered by Mason et. al. that mind wandering contributes to the generation of a coherency between past, present, and future experiences, almost like the glue that links the pieces of experiences across the temporal space to create meaning out of these experiences. This is extremely similar to memory and a key function of memory. As neurobiologist Eric Kandel once said, memory is glue that holds our lives together. Without memory, we are capable of living only in a given moment in time, without any connection to our past experiences and future hopes. I feel like mind wandering may have a similar function. Indeed, it would be interesting to explore the role of mind wandering in memory consolidation, which is the topic of my research proposal. I am very excited about this topic, and I look forward to expanding and refining this idea.

Word Count: 1490, not including references