More Reflections on "The Blind Spot"

I'm sharing my rough notes on Life, Cognition, and Consciousness after reading Frank, Gleiser, and Thompson's important new book

These are rough notes that capture some of my ongoing reflections after re-reading the core chapters of The Blind Spot on Life, Cognition, and Consciousness. You can read my original review of the book here:

I collected the notes below in preparation for a podcast dialogue tomorrow with biologist Timothy Jackson. Stay tuned for that!

---

Chapter 6, titled “Life,” begins with the authors’ reflection that, from the perspective of physics and chemistry, life is a surprise. There’s nothing in the laws of physics that would allow a physicist to predict the emergence of biological phenomena.

I worry that emphasizing the “surprising” nature of life—all the ways it just doesn’t fit in a world of “mere” physics and chemistry—ends up encouraging us to reify the life/matter distinction. Thompson has argued convincingly for a life-mind continuity. I would follow Whitehead and insist on an even more radical continuity thesis:

“neither physical nature nor life can be understood unless we fuse them together as essential factors in the composition of ‘really real’ things whose interconnections and individual characters constitute the universe” (Modes of Thought, p. 150).

Life can be understood as an amplification of basic powers pregnant in physical processes. Without such an admittedly speculative maneuver (and remember, positing some sort of physical reality beyond or without experience is arguably even more metaphysical!), we're left with a rather large explanatory gap, if not a hard problem: that of understanding how life could come out of matter.

They say that the experience of being alive is a condition for the possibility of knowing life, citing Hans Jonas (Schelling should also be cited, as he says the same thing in his 1797 Ideas for a Philosophy of Nature). The idea is that life can only be known by life. Just as in thermodynamics, where our scientific study of the movement of energy presupposes our feeling of warmth, in biology our study of living organization presupposes our feeling of being alive.

The authors criticize the life-as-machine metaphor. Machines are decomposable because they have externally related parts. Organisms are autonomous wholes and so not decomposable. Their parts produce one another for the sake of the whole. The machine metaphor for life gets the mereology of life all wrong. It ignores life's “needful freedom” (Jonas) and its agency (though most reductionistic approaches end up mobilizing a kind of “molecular animism” anyway by metaphorically attributing agency to molecules, as with “selfish genes” and so on).

The authors argue that life is unlike dissipative structures such as fire, hurricanes, and stars. But they admit stars are an interesting case. Here we might reference the work of astronomer Eric Chaisson on what he calls the “energy rate density” of astrophysical systems, like the metabolism of galaxies and the replication and evolutionary selection of multigenerational stars. The authors ask if we really need to modify physics, or just introduce new principles of biological organization that are irreducible to, but compatible with, existing physics. Whitehead's approach would follow Chaisson in trying to reimagine our understanding of the physical world so as to make life not such a surprise, but rather an intensification of the self-organizing dynamic already present in the physical world. The authors (which include two physicists) appear to prefer the former, more conservative approach.

The authors write: “genes are consequences of organization, not determinants of it.” In other words, genes are always part of a cellular network. The authors reference “closure” and make an analogy between mathematical closure and what Varela called the “operational closure” of self-producing biological systems. Closure refers to any process where the parts of the process recursively depend on each other for their mutual generation as a network, and thereby constitute the system as a unity. The authors also mention Robert Rosen in this context, specifically his attempt at a formal proof that an organism is closed to efficient causation. They argue that in organisms, constraints depend on what they constrain.

All cells are agents. Even the simplest bacteria actively respond to their environment. They even are able to, through gene regulatory networks, have agency over their rate of mutation so as to discover a more adaptable form of behavior, say an enzyme to digest a nutrient in a novel environment. So organisms are not just passive waiting for the environment to select them. There is such a thing as organismic selection as well as natural selection.

The authors distinguish between syntactic and semantic information. Claude Shannon's theory of information deals only with syntactic information. They claim that life pursues values and realizes relevance and so deals with semantic information, which cannot be reduced to statistical correlations. Semantic information carries significance because of its relevance for a system's self-maintenance.

Autopoiesis is not explained by evolution, according to the authors. It's not explained by natural selection, because the latter presupposes self-producing organisms that can also reproduce and vary. They briefly discuss the origin of life, claiming it likely occurred as a result of the emergence of autopoetic protocells that were minimally autonomous. Evolution would then kick in to complexify autopoiesis further, leading to greater forms of autonomy. “The point is conceptual: organizational concepts aren't reducible to genealogical ones.” Strangely, they only spend one sentence on this monumental transition from the abiotic to the biotic world. The sentence includes a citation to a great paper by the co-author Gleiser, as well as Sarah Imari Walker, that discusses lipid or peptide-first theories of the origin of life, and that also cites biochemist David Deamer and the wet-dry cycling thesis he developed with my friend and colleague Bruce Damer (see my 2022 chapter with Damer, “The Cosmological Context of the Origin of Life: Process Philosophy and the Hot Spring Hypothesis”). But this question of the origin of life deserves more care, as given the embodied phenomenological or what could be called “biopsychist” orientation defended by the authors, the emergence of living organization signals a kind of rupture in being, a rift opening up in the very nature of reality so as to make room for genuine “agency” and “subjective concern” that's seemingly no less severe than the old rift between matter and mind.

The authors argue that life is inherently a planetary phenomenon. In Mind in Life (2007), Thompson had coined the term ecopoesis to refer to the Earth system, or Gaia, as itself a self-producing living entity.

They close the chapter on life by putting the idea coming out of physics of “phase space” in its place. They argue that while all trajectories in phase space are logically determined by laws and by initial and boundary conditions, in reality we can never know the initial conditions with the degree of precision necessary to make certain predictions. Phase spaces are logical constructions. They are not real places out there in nature. The authors write that “[phase spaces] are our remarkable and very effective invention in order to make physical phenomena intelligible.” They claim that unlike idealized physical systems, life comprises unprestatable phase spaces.

Chapter 7 focuses on cognition. The authors begin by talking about classical symbolic processing, or good old-fashioned AI, as the original form of computationalism, contrasting it with the newer form based on neural network architecture and called “connectionism.” This latter approach is related to the machine learning explosion that's driving the AI revolution at present. Connectionism doesn't view cognition as a kind of symbolic processing or logical deduction, but rather sees cognition in essentially numerical terms, as having to do with sub-symbolic activation patterns in weighted networks. They discuss John Vervaeke’s idea of relevance realization and mount a defense of the non-computability of human intuition. Human intuition is not simply brain activity. They claim that machines don't really understand anything. They worry that we're letting our so-called artificially intelligent machines remake our life-world for the worse. We're allowing our devices to force us to conform to their processing limitations.

The authors are against representationalism in cognitive science, but at least in this text they rely on an unexamined conception of “sensory perception,” which they parody as providing “the territory,” while the representation would be “the map.” They quote Di Paolo et al.: “cognition is not about transposing a world of predefined significance to the inside of an agent. It is about agents moving within the world and singly or collectively changing it in ways that are significant according to the forms of life they enact.” There may not be any predefined domain of biological significance, but to insist that organisms are entirely “operationally closed” to prehensions of an outside environment seems to me to flirt awfully closely with solipsism (even if only at the species level). If the authors had engaged more deeply with Whitehead, they might have been able to distinguish between the presentational immediacy of sensory perception and the causal efficacy of bodily reception, the latter being strikingly similar to Merleau-Ponty's insistence that every sense percept has a bodily reference. They talk about how all human intelligence presupposes a meaning-providing background of pre-conceptual sensemaking and culturally configured and socially situated bodies. They draw a sharp distinction between calculation and reasoning in lieu of this.

They mention Grossberg and Carpenter's “adaptive resonance theory,” which might allow for truly self-generating models of brain activity via unsupervised neural network learning. So it would appear that they are not totally rejecting the idea that truly unsupervised machine learning might achieve something approaching artificial sensemaking.

Chapter 8 is on consciousness. They defend the horizonal structure of consciousness, which is to say that you can't step outside of it and measure it against something else. As you proceed towards what initially seems to be its edges, the edges just continually recede, like the horizon. Consciousness is thus not just another natural phenomenon. It is a precondition of world-disclosure, including the disclosure of nature for scientific investigation. Consciousness thus has existential and cognitive as well as transcendental primacy. It is the a priori condition of possibility for knowledge of anything. The authors also affirm the primacy of embodiment.

They argue that our awareness of pure experience, or of non-dual consciousness without an object, doesn't necessarily prove that consciousness is the very ground of being. They're arguing for phenomenological primacy, not metaphysical or ontological primacy.

They argue that consciousness is intersubjectively grounded. Our sense of self is sourced in other selves. In other words, developmentally speaking, we come to be a self because we internalize the perspectives of others. And we always draw on a shared language and set of concepts with which to think about ourselves. We're always already bound together in community. From a phenomenological point of view, metacognition or self-consciousness is a special case of social cognition or intersubjectivity.

They discuss “the strange loop,” the way that the world is in us and that we are in the world. They relate this to Merleau-Ponty's idea of the intertwining flesh of the world. They write: “the bodily subject is a project of the world, a way the world locally self-organizes and self-individuates to constitute a living being.” We could compare this to what Whitehead says in Process and Reality about how the subject emerges from the world as a superject, a view Whitehead contrasts with Kant’s transcendental idealism, where it is thought that a merely apparent world of objects emerges from the structure and activity of the subject. This contrast is quite important because one might worry that the phenomenological/transcendental perspective that the authors of The Blind Spot are articulating ultimately collapses into the kind of Kantian position that Whitehead criticizes.

The authors raise the problem of ancestrality and try to move beyond it by talking about a kind of life-world/universe strange loop. They then say that the very idea of a nebula is just our human way of sorting astronomical phenomena, which makes me worry that they're succumbing to a kind of nominalism, and again, I much prefer Whitehead's organic realism. It may be true that science and/or common sense sometimes carves up the world in arbitrary ways, but that doesn't mean that there aren't real structures in nature that we could come to know through some form of conformal relationship or resonance.

They claim that the hard problem of consciousness is not solved by dualism, by substance-property forms of panpsychism, or by illusionism, because all assume a blind spot metaphysics.

They criticize predictive processing models. They might be useful for studying brain function, but such models offer no explanation for consciousness. They are critical of Anil Seth, as many times his claims (eg, our conscious perceptions are just “controlled hallucinations”) are just metaphysics dressed in a white lab coat that encourage us to adopt a solipsistic view of our own minds. The authors point out that the predictive processing model ends up swallowing up its own physical basis. This is because, according to the thesis itself, the scientists can only have a kind of controlled hallucination of the brains they are trying to study. There is no outside-the-model physical brain that they might access in order to test the theory. I think here of Whitehead’s quip in Science and the Modern World that “some people express themselves as though sense organs, nerves and brains were the only real things in an entirely imaginary world.”

On the other hand, in an earlier part of the book, the authors themselves positively explore and even recommend the QBism interpretation of quantum mechanics, which would suggest that this predictive processing or Bayesian inference process does in fact go all the way down to quantum events in the physical world. So again here, one has to wonder if we're needing to change our very understanding of the physical world, or just come up with new principles of biological organization that are compatible with but beyond the laws determining that physical world. It's this latter approach that the authors often seem to take, but they leave the door open in certain especially curious moments for the alternative (more Whiteheadian) approach. The authors write that “the brain is an organ of perception, not the perceiver. The perceiver is the whole person, or animal, geared into its world.” I agree, but find the choice of mechanistic metaphor (“geared into”) unfortunate!

The authors argue that the version of panpsychism developed by philosophers like Galen Strawson and Philip Goff only inserts experience into nature without acknowledging that experience is a condition of possibility for the intelligibility of nature. They say that this kind of panpsychism is like sprinkling consciousness into the physical world as though it were a kind of pixie dust. This makes me wonder what they think about Whitehead's panexperientialism. He certainly cannot be accused of neglecting the transcendental status of experience (see this article by James Bradley wherein he defends Whitehead’s “transcendental cosmology”); but he also affirms that the universe is suffused with experients and that outside of the experience of subjects there is nothing, nothing, nothing.

The authors helpfully criticize the abstract idea of qualia, which is important to analytic panpsychists. Qualia are part of the analytic panpsychist's intrinsic nature argument for the fundamental nature of consciousness, but as the authors argue, qualia are really relations, they're not intrinsic properties. Here they again could have linked directly with Whitehead's idea first introduced in Science in the Modern World and then systematically elaborated in Process and Reality: that of prehension. Prehensions are feeling-vectors. They're how feeling is transmitted from one occasion of experience to another. Prehensions are the process-relational panexperientialist’s alternative to the substance-property panpsychist’s qualia.

They close the consciousness chapter with the idea of neurophenomenology, which is a novel approach to the study of consciousness proposed by Varela that is alternative to the reductive materialist approach. A neurophenomenological approach would seek the mutual constraints between cultivated/trained experience and brain measurements, always attending closely to the actual texture of experience and not allowing the model to be surreptitiously substituted for that experience.

The authors end with a crucial reframe: “the problem for neuroscience can no longer be stated as how the brain generates consciousness. Instead, the problem is how the brain as a perceptual object within consciousness relates to the brain as a part of the embodied conditions for consciousness.”