Tuesday, January 23, 2007


The most forceful (and the only one published, it seems) objection to the concept of a "default mode" of processing in the human brain has been articulated by Alexa Morcom and Paul Fletcher at Cambridge. The brain's "default mode" or "resting state" has been studied most intensively by Marcus Raichle and colleagues (e.g, Gusnard & Raichle, 2001):
So, we propose that there exists a physiological baseline of the human brain, which can be observed when subjects are awake and resting with their eyes closed (an exception might exist for some extrastriate visual areas — see above). Substituting simple visual fixation or passive viewing of stimuli has little effect on this physiological baseline, except to increase blood flow in the visual areas...
Morcom and Fletcher see multiple problems with this proposal:
The case for a default mode comprises three related ideas. The first is that the resting state constitutes an absolute baseline, and is therefore a fixed point relative to which all cognitive and physiological states can and should be considered (Gusnard and Raichle, 2001). Second is the notion that the level of neural activity in this resting state is substantial and therefore functionally important, with changes produced by task demands representing just the “tip of an iceberg” (Raichle et al., 2001). Finally, relative to a wide range of tasks the resting state is said to be associated with higher levels of activity in a consistent set of brain regions. This has led to the idea that, at rest, we return to a ‘default mode’, which plays a critical role in the ‘intrinsic’ functioning of the brain (Shulman et al., 1997 and Gusnard and Raichle, 2001). We believe that these three claims, and their synthesis, should be evaluated critically for theoretical and practical reasons. If they are valid, then the resting state is indeed a context in which to study brain processes that are fundamental and important relative to the small flickers of activity produced by task demands. It would follow that cognitively driven fluctuations cannot be interpreted except in the context of the default system.

We suggest that the case for a default mode does not survive this critical evaluation. We first explain and evaluate the claim that ‘rest’ is a baseline state for the human brain, summarizing the support for this, and the ways in which a baseline might or might not be necessary in functional neuroimaging. We conclude that despite the interesting characteristics of rest as baseline in terms of oxygen balance, these are not relevant to studies that seek to understand how neural activity underpins cognitive processing. Secondly, while we accept that a high level of energy expenditure of the brain at ‘rest’ indicates that the resting state is active, we do not agree that this activity has a special status compared with that in any other task, or that the brain's energy budget is informative about the nature of a ‘default mode’. Thirdly, with respect to the idea that patterns of brain activity found at ‘rest’ are consistent, we point out that the evidence for this is inconclusive. Furthermore, we would question what conclusion could be drawn from such a consistency, if it is ultimately demonstrated. Finally, we note that the idea of a default mode is based not only on three separate claims but upon their synthesis. We scrutinize this synthesis with regard to the support that it gives to rest/default mode as a state that is qualitatively different from any other mental and neural state. We conclude that even if there is empirical consistency in the patterns of activity observed at rest, and a subjective appeal to the notion that when we rest we are in a default state because there is no explicit task to perform, these are insufficient grounds for affording the resting state a privileged status in accounts of human behavior. We further suggest that, in most situations, the aims of cognitive neuroscience are best served by the study of specific task manipulations, rather than of ‘rest’.
It seems to us that there are problems with the assumption that the resting state is a baseline to which other states should be referred, a notion entailed by the call for a distinction between a subtraction and a ‘reverse subtraction’. If, as we believe, such a distinction is irrelevant in cognitive neuroimaging studies, the utility of the term ‘baseline’ in this context must also be called into question. In considering these problems, we need to distinguish between two views: one that the resting task constitutes a processing baseline and one that it serves as a physiological baseline that does not relate directly to processing. In so doing, we highlight the fact that the notion of a ‘default mode’ rests in part on a relationship between the physiological and processing aspects of ‘rest’. We conclude that although rest may be, subjectively, a ‘default’ state (in the sense that it is what we are doing when we are not doing anything else), it is of no utility as a processing baseline. We find that the suggested link between the processing taking place at rest and its physiology is one that can have no direct relevance for neuroimaging.
Most important for the current argument attributing the phenomena of daydreaming or "mind-wandering" to activity in the default mode network is an objection on the grounds of its functional significance. What happens (cognitively speaking) during "rest"? How do you manipulate it and quantify it? Is introspection a valid measure? Certain manipulations can produce global increases and decreases in the network, but is this enough? Shouldn't the component areas be isolated? Why not try to separate activity in disparate regions of the network? Surely superior medial frontal regions subserve a different function than posterior lateral parieto-occipital regions.
Gusnard and Raichle (2001) emphasize the importance of studying the ‘default mode’. It is therefore particularly notable that in the last section of this article, they consider its possible functions by citing studies in which relative activity increases have been demonstrated in the ‘TID’ [task-independent decreases] regions they highlight, in response to cognitive task manipulations. In the case of the precuneus and posterior cingulate, “associated with the highest resting metabolic rates in the human cerebral cortex”, it is pointed out that these regions show specific increases in activity during visuospatial and emotional processing tasks (Vogt et al., 1992 and Maddock, 1999) and thus may play a role in ‘monitoring’ of the environment and in assessing the emotional significance of events (see also Raichle et al., 2001). We believe this to be a telling inference; they are suggesting that although the precuneus/posterior cingulate may indeed frequently be active at ‘rest’, an understanding of the processing such activity subserves depends on the employment of the right tasks and the delineation of the circumstances of activity changes. Comparable observations about posterior lateral cortices, and dorsal and ventral medial PFC are also made (for a similar line of reasoning see Beckmann et al., 2005). It is clear from these observations that the cognitive nature of rest is at present almost entirely a matter of speculation. More importantly, once again, the message must surely be that the resting task seems cognitively interesting, so we should unpack it by devising appropriate tasks.
That concludes our three part series on the concept of a "default mode" for the brain, and whether it ushers in an exciting new era in scientific studies of introspection, adding the technological imprimatur of neuroimaging to that musty old Jamesian field, or whether we should relegate this notion to the dustbin of neuroscience history.

Read Part 1 and Part 2.


Beckmann CF, DeLuca M, Devlin JT, Smith S.M. (2005). Investigations into resting-state connectivity using independent components analysis, Philos Trans. R. Soc. London B Biol. Sci. 360: 1001–1013.

Gusnard DA, Raichle ME. (2001). Searching for a baseline: functional imaging and the resting human brain. Nature Rev. Neurosci. 2:685-694.

Maddock RJ (1999). The retrosplenial cortex and emotion: new insights from functional neuroimaging of the human brain. Trends Neurosci. 22 : 310–316.

Morcom AM, Fletcher PC. (2006). Does the brain have a baseline? Why we should be resisting a rest. NeuroImage Oct 16; [Epub ahead of print]

Raichle ME, MacLeod AM, Snyder AZ, Powers WJ, Gusnard DA, Shulman GL. (2001). A default mode of brain function. Proc. Natl. Acad. Sci. 98: 676–682.

Shulman GL, Fiez JA, Corbetta M , Buckner RL, Miezin FM, Raichle ME, Petersen SE (1997). Common blood flow decreases across visual tasks: decreases in cerebral cortex. J. Cog. Neurosci. 9: 648–663.

Vogt BA, Finch DM, Olson CR. (1992). Functional heterogeneity in cingulate cortex: the anterior executive and posterior evaluative regions. Cereb. Cortex 2: 435–443.

Subscribe to Post Comments [Atom]


At January 30, 2007 8:22 AM, Anonymous Anonymous said...

While interesting to see some debate about the default mode hypothesis, I must say that the article (and also the series itself) does fall rather short in offering any decent critique of the default mode hypothesis. The quotes from Fletcher's article amount to "we don't agree with Raichle, nor do we agree with Raichle, oh and we also don't really agree with Raichle much" and "More research should be done to understand these areas" (you don't say! And here I thought science was finished and we could all go home).

Part 2 also failed to make the reader understand what was at issue with the mind wandering study. Having a specific hypothesis and using a targeted ROI approach is exactly the type of analysis that critics of neuroimaging are calling for. It's the whole brain analysis and post-hoc story telling that is generally frowned upon. So why fuss over the fact that they restricted their analysis to ROIs in hypothesized areas?

Also fail to see what the fuss is about over the smaller, or complete lack of deactivation on Rest > practiced blocks. Makes perfect sense given their hypothesis and the previous research.

Generally I agree with your critiques, especially when it comes to the press (or even authors) over-interpreting their data, but in this case I fail to see any reason to go after these sets of papers.


Post a Comment

<< Home

eXTReMe Tracker