Complexity and Complementarity

The topic of the Default Mode Network (DMN) has been raised in our discussions before (I believe by Jake Mahon), in the context of how it might correlate with the function of the right hemisphere. The DMN was recently also discussed during an interview with Dr. Ev Fedorenko on the Many Minds podcast:

[paste relevant portion from the transcript once it becomes available]

Has this subject been discussed by McGilchrist? Tangentially. Paraphrasing an earlier interview with Scott Barry Kaufman:

Kaufman: “Our work is actually really consistent. We should coauthor an article together.”

McGilchrist: “That would be great.”

Kaufman: “I’m obsessed with the default mode brain network because I think that network offers us the core of human experience of what it means to be really human. And I think you’re obsessed with the right hemisphere for the same reason. So I looked deeper into our Nature paper and I noticed that there were preferential right hemisphere activations in the default mode network. Now, we didn’t originally set out to test that specifically, so don’t quote me on that as a statistically significant effect, because we didn’t have the methodology, but I noticed a trend…

It’s mapping the ‘network approach’ to the ‘hemisphere approach’. You’ve got me thinking about this in a deeper way than I ever have before, how these are probably both saying the same thing. They’re just different levels of analysis. You know, I’ve been arguing against the tyranny of the executive attention network, where you’ve been arguing against the tyranny of the left hemisphere. But conceptually, I feel like we’re both bothered by the same thing.”

McGilchrist: “I think that that’s right. There’s a lot of overlap.”

I would love to see such a paper, if it were to be written. 

Any discussion of the DMN and the anti-correlated Task Positive Network (aka "executive attention network") might cause one to wonder if McGilchirst is like that proverbial "man who was found searching for his keys, not where he had dropped them, but under the lamplight, because that was where he had enough light to search." The location of damaged brain regions, whether on the left or right, has long been noted in lesion-deficit studies, so it would obviously be easier to articulate a hypothesis about the hemispheric lateralization of our modes of attention. Those same studies do not necessarily "shine a light" on whether portions of the DMN or TPN are damaged or not. Consequently this lack of information prevents the likelihood of a possible "Network Hypothesis" from being put forward as a competing theory to the "Hemisphere Hypothesis". After all, the DMN was only described in 2001, with the majority of papers published after 2007, so there wasn't a whole lot of research at the time of the publication of The Master and His Emissary in 2009.

Now, if information about these brain networks should prove to be more explanatory than the hemisphere hypothesis, would this invalidate the sort of arguments that McGilchrist is putting forward in his work? Not necessarily. It's conceivable that it would merely switch the attribution of his observations to a different neurological mapping, that is, from hemispheres to networks.

Kaufman and McGilchrist are engaged in complimentary research. And as he indicated, the most likely possibility is that one level of analysis is at a higher integrative level [1] than the other, one of these can only be fully developed by seeing it in light of the other, located at a higher integrative level. If they were to write a paper comparing the network and hemisphere approaches, I suspect they might find just this sort of hierarchical relationship, such that one is "necessary but not sufficient," one of these underwrites critical aspects of the other, but does not explain the full observations. But which is it? 

Is it more accurate to say that each hemisphere has access to bilaterally distributed networks with anti-correlated roles? Or is it more accurate to say that these networks cut across hemispheric divisions to access lateralized hemispheres with anti-correlated modes of attention? Whichever it is, this is the "what."

Following that, "how" these neurological networks are engaged by either hemisphere in the deployment of asymmetric modes of attention is the purview of the hemisphere hypothesis. Or conversely, "how" these hemispheres are engaged by either network in the deployment of asymmetric roles is the purview of the network approach. So is the lateralization aspect overlain upon the network approach, or vice versa?

In the conclusion of The Master and His Emissary, McGilchrist writes that the "divided nature of mental experience... might have some literal truth." If we assent this much, then does it really matter which "chunks" of the brain "mirror the dichotomies that are being pointed to"? In principle, I don't think it should. Both analyses involve opponent process theory (anti-correlated networks or hemispheres) with asymmetric roles. The 'coincidence of complementary opposites' is the primary insight. In order to be 'antifragile' and adaptable the brain operates near a critical threshold at the edge of chaos, between order and disorder (or so it's been speculated), which is the most resilient place to be. Exactly where these coinciding oppositions are instantiated is arguably of secondary importance.

A point in favor of the network approach is that it does more easily satisfy the "bridge law objection" as articulated by Spezio (and referenced by Thompson) because it appears to better identify a reductive "locus" for various features. It also better comports with a postmodern cultural context, in part because it is more conceptually abstract, and because an eliminativist perspective (which is where the network approach tends to lead) means one can avoid addressing the question of how science and the humanities intersect, a preferable outcome since it is feared that any proposed answer to that question might support discriminatory practices and revisit the experience of historical traumas. 

McGilchrist noted that Marcel Kinsbourne identified three main oppositional pairings within the brain. The network approach described by Kaufman here articulates still others in addition to these, or perhaps merely recombines and reinterprets them in different formulations, subdivided at different scales of resolution. And this is to be expected. We should not be surprised to find a multitude of opponent pairings across many scales and within a wide variety of biological systems. 

A point in favor of the hemisphere hypothesis is that, and here McGilchrist cites Joseph Bogen, "hemispheres can sustain the activity of two separate spheres of consciousness." The network approach cannot claim this as easily. In part this is due to the inherent complexity of the networked neurological structures being described and their morphological dissimilarity to each other. I think we are now able to postulate a few general rules: 

• Once structural complexity has reached an apex or point of diminishing returns, mutually entailed opponent processing must take over to do what elaborate complexity cannot achieve alone, for example, in order to be able to address problems such as those raised by the "no free lunch" theorem. McGilchrist shares the Heraclitean view: "We need resistance. We need opposition."
• The highest integrative level or scale of opponent processes will provide the greatest explanatory power, as these are the "highest leverage points" (Donella Meadows) in the system. These large scale dynamics exert downward constraint that entrains and thereby overrides lower scale differences (Stanley Salthe). This also suggests an apparent telos or convergent drive acting behind the system to sustain opponent dynamics. 
• In general, the higher the level of opposition, the more evenly matched 'the agonist and the antagonist' may appear to be. This is needed in order to sustain the tension between them. Lower level analyses however tend to focus on similarities among the parts instead of the higher level opposed gestalten. (Thus to a trained eye, the apparent redundancies in the bisected brain should suggest that a higher level of integrated opponent processing may be occurring, with corresponding phenomenological specialization.)

And indeed, these implications appear to follow from the underlying argument in The Matter with Things, which leads from neurology to "naturalized metaphysics" in a single vision, re-uniting the sciences and the humanities. 

Notes: 

[1] Joe Scott described 'top down' causation in a recent video: "There's not just one consciousness going on inside our minds. There's two. But it actually gets crazier. Turns out there's actually multiple modules inside each of the hemispheres that can display a kind of consciousness, and they have their own things to say. In a lot of ways, our brains work pretty much like a hierarchical brain system. Consciousness is a result of both 'bottom up' and 'top down' causation. We are simultaneously making decisions on a subconscious level but also shaping those decisions at a conscious level. In Gazzaniga's words, "Action is made up of complementary components arising from within and without... What is going on is the match between our ever present multiple mental states and the impinging contextual forces within which it functions."