When you think of a psychological experiment involving the use of functional magnetic resonance imaging (fMRI) you probably imagine a person laying down inside the scanner and performing one or several tasks. While this is true in many cases, there are also some studies that do not require a task at all. These studies examine brain activation during so called resting-state, looking at what your brain does while you sit back and relax. The goal in resting-state fMRI experiments is to identify brain networks. These networks are formed by regions that activate or de-activate simultaneously, and thus show a certain similarity in their function.

A default state of the brain

Thanks to these types of experiments, the default-mode network (DMN) was discovered, which shows more activity when people rest compared to when they are performing a task. This network is composed of regions that are located at maximal distance from sensory and motor areas of the brain and are involved in abstract information processing. The DMN was given its name due to the idea that it represents the default state of the brain.

But what does the default state of the brain refer to? The mental state associated with the DMN has also been called a “conscious resting state” and has been linked to a wide range of higher cognitive functions. Among the suggested functions are retrieval of memories from one’s past and knowledge about the world, mobilization of memories in order to solve problems and to develop future plans, and imagining the perspective of others. It was also hypothesized to be related to self-referential thought, self-awareness, free thought, and even creativity.

 

Challenges for the DMN: Consciousness for everyone

There is a serious problem to these interpretations: Activation of the DMN is not specific to simply being at rest and performing whatever humans think without anything to do. A certain degree of DMN activity has been found during sedation, a pharmacologically induced state of reduced consciousness familiar to anyone who has ever been anesthetized before on operation. Furthermore, DMN-like networks have been identified in non-human primates, dogs, and rodents. The monkey-counterpart has even been shown to be more active during rest than during task performance, analogous to findings in humans. Although the cognitive abilities of non-human animals shall not be debated here, it should be safe to say that they certainly do not reach the same degree of conscious thinking as humans.

It seems that the existence of the DMN is an evolutionary preserved feature of (mammalian) brains that entails an advantage for brain organization. This does, however, not mean that the story of the DMN is told and no higher-order functions can be assigned to it; rather, one has to be cautious in defining the role of a multi-faceted entity.

 

The solution: A wandering mind

A fresh perspective on the mental operations underlying DMN activation that has received considerable attention among DMN researchers is mind-wandering. From its first conceptualization as “task-unrelated thought”, referring to thoughts not concerned with the current operation or task, mind-wandering has evolved into a more differentiated concept. For instance, Poerio and colleagues argue that during mind-wandering, attention first has to be diverted away from perceptual input. In a second step, autobiographical and semantic knowledge can be retrieved from memory to form the content of thought.

Other emerging frameworks converge on the notion that mind-wandering can occur either deliberately or spontaneously. Deliberate mind-wandering relates to a controlled state in which you allow yourself to let your thoughts and ideas flow freely, while spontaneous mind-wandering features thoughts that intrude your focused state of mind rather unexpectedly and hinder concentration on ongoing tasks.

Christoff and colleagues additionally point out that defining mind-wandering in terms of its content fails to capture its key feature – the unstable nature of our thoughts that jump from one idea to the other, often without an obvious connection. They frame the above mentioned distinction in terms of deliberate and automatic constraints on thought that can change dynamically in the amount to which they are exerted. Deliberate constraints refer to cognitive control, which is exerted when you force yourself to remain mentally focused on a topic, such as when reading a book. Automatic constraints take over when you experience a very noticeable stimulus or thought that you simply cannot ignore, such as your favorite song being played on the radio while you are trying to read your book. Mind-wandering occurs when both kinds of constraints are low and you are not limited in your thinking and in the way you link your ideas.

 

Dynamic minds – dynamic brains

If the wandering mind is so heterogeneous, how can the same consistently found brain regions that constitute the DMN subserve it? The answer lies in the dynamics of brain networks. The DMN itself is composed of several sub-networks that are functionally different. The core DMN subsystem serves as a gateway for information flowing through the overall system. Another subsystem is centered on the medial temporal lobe (MTL), a region responsible for all sorts of memory functions, and the structures it projects to. Apart from memory, it is also involved in the construction of ideas from past experiences.

As mentioned above, brain networks are not static over time, but can exhibit changes in their connectivity, i.e. the amount of co-activation that the regions which constitute the network express. Importantly, networks are not only able to change in their connectivity, but can couple and de-couple with other networks. Coupled networks tend to activate together and deactivate together as well, essentially forming a super-network.

The resolution to the DMN-mind-wandering problem lies, as Christoff and others suggest, in that “fluctuations in level and type of constraints on thought correspond to changing interactions between large-scale brain networks”. The MTL-subsystem of the DMN, for instance, is recruited to a higher degree when people are unaware of their task-unrelated thoughts. The researchers thus suggest that it plays a role in spontaneity of mind-wandering. However, this subsystem does not differentiate well between task-related and task-unrelated thought. When the MTL-subsystem and the DMN-core couple, changes in the amount of this co-activation correlate with fluctuations in mind-wandering. This means that there is a link between changes in the coupling of the subsystems and changes in mind-wandering.

DMN networks are not restrained to coupling among each other, but can also partner up with attentional and executive control networks. While you think about the last movie you saw and how the countryside reminded you of your summer vacation, the MTL-subsystem will exhibit connectivity with the DMN-core. As you shift from this unconstrained mind-wandering to worrying about your knee that you injured while hiking, this cognitive shift is likely to be accompanied by an increased coupling of the MTL-subsystem with attentional networks. States of salient stimuli that constitute automatic constraints on thought are thus associated with a strong coupling between the DMN and attentional networks.

During states of high deliberate constraints, the executive control network can regulate the influence of distracting stimuli. This regulation is implemented via interference of the executive control network with the coupling between attentional networks and the DMN. When you start to plan how to prevent injuring your knee further, you will be less able to ruminate about the pain it causes you, and the executive control network will have diminished the influence of your attentional networks on the DMN.

It appears that we have to take our analysis of brain organization one step further, from brain regions in isolation, over networks formed by brain regions in conjunction, to the dynamic interaction of these brain networks over time.