From 2018... Self-generated thought: The medial temporal lobe is a key site of thought-generation; the posterior cingulate cortex & inferior parietal lobule have little say

From 2018: Neural origins of self-generated thought: Insights from intracranial electrical stimulation and recordings in humans. Kieran C.R. Fox. Chapter in The Oxford Handbook of Spontaneous Thought: Mind-Wandering, Creativity, and Dreaming, May 2018, DOI: 10.1093/oxfordhb/9780190464745.001.0001

Abstract: Functional magnetic resonance imaging (fMRI) has begun to narrow down the neural correlates of self-generated forms of thought, with current evidence pointing toward central roles for the default, frontoparietal, and visual networks. Recent work has linked the arising of thoughts more specifically to default network activity, but the limited temporal resolution of fMRI has precluded more detailed conclusions about where in the brain self-created mental content is generated and how this is achieved. Here I argue that the unparalleled spatiotemporal resolution of intracranial electrophysiology (iEEG) in human epilepsy patients can begin to provide answers to questions about the specific neural origins of self-generated thought. I review the extensive body of literature from iEEG studies over the past few decades and show that many studies involving passive recording or direct electrical stimulation throughout the brain all point to the medial temporal lobe as a key site of thought-generation. At the same time, null effects from other brain regions suggest that various other default network hubs, such as the posterior cingulate cortex and inferior parietal lobule, might have only a marginal role (if any) in the self-generation or initiation of mental content like dreams, visual imagery, memories, and prospective simulations. Ultimately, combining a variety of neuroscientific methods that compensate for each other's weaknesses and complement each other's strengths may prove to be the most effective way to understand the brain's remarkable ability to decouple from the immediate environment and generate its own experiences.

Introduction
An enormous amount of scientific interest has recently begun to focus on spontaneous and self-generated forms of thought (Andrews-Hanna, Smallwood, & Spreng, 2014; Christoff, 2012; Christoff, Irving, Fox, Spreng, & Andrews-Hanna, 2016; Seli, Risko, Smilek, & Schacter, 2016). As experience sampling studies (see Stawarczyk, this volume) and questionnaires develop an understanding of the associated subjective content (Delamillieure et al., 2010; Diaz et al., 2013; Fox, Nijeboer, Solomonova, Domhoff, & Christoff, 2013; Fox, Thompson, Andrews-Hanna, & Christoff, 2014; Stawarczyk, 2017; Stawarczyk, Majerus, Maj, Van der Linden, & D'Argembeau, 2011), functional neuroimaging research over the past two decades has delineated a rough but increasingly refined picture of general brain recruitment associated with these self-generated forms of thought (Fox, Spreng, Ellamil, Andrews-Hanna, & Christoff, 2015). Throughout this chapter, by ‘selfgenerated thought’ I will simply mean mental content that is relatively independent of and unrelated to the immediate sensory environment (Andrews-Hanna et al., 2014; Fox, Andrews-Hanna, & Christoff, 2016); taken broadly, self-generated thought includes mental processes such as stimulus-independent thought (McGuire, Paulesu, Frackowiak, & Frith, 1996), task-unrelated thought (Dumontheil, Gilbert, Frith, & Burgess, 2010), spontaneous thought (Spiers & Maguire, 2006), mind-wandering (Christoff, Gordon, Smallwood, Smith, & Schooler, 2009), creative thinking and insight (Ellamil, Dobson, Beeman, & Christoff, 2012), and dreaming (Fox et al., 2013).

Now that a general picture of the subjective content and neural correlates of selfgenerated thought is emerging, deeper and more subtle questions are being posed: for instance, whether specific neural correlates are associated with specific self-generated content (Gorgolewski et al., 2014; Tusche, Smallwood, Bernhardt, & Singer, 2014); whether differences in brain morphology are associated with individual tendencies toward certain types of self-generated thinking (Bernhardt et al., 2014; Golchert et al., 2017); what the relationship of self-generated thought is to various psychiatric and neurodegenerative conditions (Christoff et al., 2016); and whether specific neural origin sites of self-generated thought can be identified (Fox et al., 2016). It is this final question that I will focus on throughout this chapter: what brain structures are the primary initiators, drivers, and creators when the brain decouples from its sensory environment and self-generates its own experiences? Is this question even valid? Can there be a specific answer?

First, I will very briefly review what is known about the broad neural correlates of self -generated thought from functional neuroimaging investigations. These studies point to the primacy of the default network in initiating self-generated thought, but cannot seem to go beyond this level of specificity and offer a more detailed answer due to their inherently poor temporal resolution. Next, I will delve into the relatively smaller (but fast-growing) human intracranial electrophysiology literature to explore the neural origins of selfgenerated thought in more detail. Starting with the broad set of regions identified by functional neuroimaging as being involved in self-generating thought, I synthesize data from human electrophysiology to hone in on the most likely origin/initiation sites and to tentatively exclude other areas from a primary regenerative role.