Diana Kwon (Scientific American, March 2018, p 29): … recently, roboticist Angelo Cangelosi of the University of Plymouth in England and Linda B. Smith, a developmental psychologist at Indiana University Bloomington, have demonstrated how crucial the body is for procuring knowledge. “The shape of the [robot’s] body, and the kinds of things it can do, influences the experiences it has and what it can learn from,” Smith says.
Having just read Kwon’s “Self-Taught Robots” article in this month’s Scientific American, I’ve decided to put aside my resolution not to deal with the phenomenology component of Darwin, Dogen and the Extremophile Choice in this blog. (In fact, I’ve been advised to take the “Zen stuff” out of the book altogether!) This new research confirming the importance of a body, even for AI learning, emboldens me. And in fact I’ve always felt that Dogen’s body-mind orientation is important to the Extremophile Choice argument, because it allows us to better understand our animal natures, and ultimately the relationship between technology and the Natural world. Also, in my mind at least, the ecological and sociological components of the Extremophile Choice message can’t be fully appreciated without this phenomenological component. I feel Kwon’s article supports my position, especially as she gets more specific with the following observations:
In a 2015 study, Cangelosi, Smith and their colleagues endowed an iCub with a neural network that gave it the ability to learn simple associations and found that it acquired new words more easily when objects’ names were consistently linked with specific body positions. The experimenters repeatedly placed either a ball or a cup to the left or the right of an android, so that it would associate the object with the movements required to look at it, such as tilting its head. Then they paired this action with the items’ names. The robot was better able to learn these basic words if the corresponding objects appeared in one specific location rather than in multiple spots. [new paragraph] Interestingly, when the investigators repeated the experiment with 16-month-old toddlers, they found similar results: relating objects to particular postures helped small children learn word associations. Cangelosi’s laboratory is developing this technique to teach robots more abstract words such as “this” or “that,” which are not linked to specific things.
So here are two excerpts (segues added) from Darwin, Dogen and the Extremophile Choice that are based on my own “meditation research” over the last 30 years:
From Essay 31:
It may very well be that in our conscious inner lives the interplay among the senses is what constitutes the sense of touch. Perhaps touch is not just skin contact with things, but the very life of things in the mind? —McLuhan1
That even Marshall McLuhan, author of Understanding Media (1964), saw touch as the product of sensory interplay, rather than as the sensory framework, or primary means, by which we think, is evidence of the deep resistance our thinking minds have to admitting the co-dependent and transitory nature of the ‘self’. Even though it was suggested many years ago (e.g. Wilder Penfield and Oliver Sacks—writing after McLuhan) that a physiological ‘sense of self’ is generated by various kinaesthetic organs that tell us how our joints, tendons, and voluntary muscles are moving (a wide array of other proprioceptors generates internal sensation elsewhere in the body); and while it’s also obvious that only our skeletal muscles can reshape, and conform to the rest of the world, the quite separate light touch system at the body’s intervening surface (the skin has different receptors for soft-contact, pain, heat, and cold); yet our thinking minds remain concertedly blind to the role of this deep sixth sense when we construct our phenomenologies. One reason is surely that deep touch is the only sense that’s accompanied by efferent impulses—i.e. movement is voluntary. In fact, efferent and afferent are so indistinguishable here that I will continue to use the one word, kinaesthesia, to mean both.
Speaking from experience, I am convinced that the best carpenters make practical ‘sense’ of the messages coming from their eyes—as they scan a drawing or a built framework for instance—only because they use their bodies to under-stand connectivity itself. And yet my reading on the subject reveals only this: that ever since literary men first began to write down their philosophies, the objects of perception have typically been de-scribed in primarily visual terms. …
… But can passive vision really extract ‘meaning’ from an animal’s, or even an animated machine’s, environment? In boldly dualistic terms, how else, from the moment our infant eyes are opened, can knowledge about ‘there’ be registered as meaningful, if not by responses of the body ‘here’? If only by the voluntary or involuntary contraction, or the resisted displacement, of a single muscle fibre attached to a vertebra, a wrist bone, an eyeball or its lens? We call even those ‘motionless’ shadows that seem to map out our fainter impressions, feelings; so how can the geography of cognitive tensions, or our brain maps of projected in-tension, be other than the body itself? I find it hard to accept that any of my thoughts or imaginings can have meaning before they are registered by some impulse in my brain aiming, if only tentatively, at the fibres that relate my body to its environment. So let’s agree for the moment (allowing for dreams, and for Steven Hawking’s profound immobility) that my use of the term, motor program, need not imply actual movement. Indeed, until some threshold of a brain’s efferent population of impulses is crossed, their effect on muscle fibres might not be afferently, let alone act-ually, ‘felt’.
… I’ve been looking for Western body-mind literature for forty years now, and except for Merleau-Ponty and some later behaviourists (McGuigan, Essay 44), I keep bumping into this ‘out of body’ gray-matter fetish that over-complicates things. Since Aristotle first proposed that knowledge must ultimately derive from the five senses—thus providing a foundation for modern science4—it seems Western philosophers have felt the need to trot out his teacher’s Idealist ‘profundity’ to complete this shallow phenomenalist wisdom. But surely the wanted depth can be more easily recovered by showing a proper respect for an older pre-literate intelligence, and for the body’s deep-touch sense?
When you take into account the full subtlety of behaviour (see my attempt at an anatomy in the next essay, 32), then abstract forms like Kant’s pure intuition (Anschauung),5 and Descartes’ soul, or mental substance, become sensible as the body’s covert (thus ‘abstracted’) activity. The empiricist Locke, who deemed the primary qualities—“bulk, figure, number, situation, and motion or rest”—to be sensible as touch,6 implicitly disembodied his “inner sense” of “mental operations”. But of course even if these ‘solid’ truths, and our thoughts about them, are seen to derive from motor-sense, this doesn’t necessarily make them secondary fabrications. That even optical illusions, hard-wired in the eyes and brain, must be registered (in some cases explained) by a body with its own logistical limitations, just means our inner lives are more tangible than the thinking mind wants to admit. And in fact we do find phenomenologies that get around the dualistic five senses plus a soul fixation: all the while Western philosophers had their suspicions (and talked around them), and long before modern physiology located all the bodily senses, farther to the East, in meditative stillness and silence, the mind itself was easily recognized as the body’s “sixth sense”.
From Essay 38:
… I personally think the curve of human technological evolution will play out something like the Cambrian explosion, which was also a new kind of evolution. But then, what carrying capacity, or what ‘technological barrel’, will this latest evolutionary diversification reach or fill? I’ll speculate more about this later because it has everything to do with humans getting to know who we are: with our finding and securing, like the supreme extremophile rather than the supreme opportunist, survival strategies that are progressively ‘contained’ with respect to authentic ecosystems. But for now, what about our hopes for artificial intelligence of the “I am alive” kind? Frankly, I don’t know. But perhaps, just perhaps, when our technology starts to level out a bit towards the top of its sigmoid curve, we might see that we don’t really need this kind of intelligence, or even want it, from our tools. And, returning to my earlier thought, since it feels to me like this being alive, or this being self-aware, is directly related to our capacity for joy, or at least to a memory and a hope of joy; and since our joy in life—a joy we do share with other species—is the product of three billion real-time years of good luck accruing to our personal germ-lines; then such true and equal fellowship with our less ancient, our less fortunate, technology could have some way to go yet.
So, just for fun, let’s keep our technological slaves working on their artificial neural nets. Perhaps we can even allow them to ‘feel’ the consequences of their actions in the cosmos somehow? We have nothing to lose as long as we allow ourselves to feel this difference too. Whatever forms intelligence might take in the future, they can never be wholly strange to us once we see that good will is at the root of evolving awareness. The intelligence of ecosystems, LAST Niche primates, and nanotech space bugs, even if tied up sometimes in self-centred knots, can never be complete without touching this common root, and in the touching, this is us. I wonder: if the primate strikes the right attitude to the ecosystem, will his own success convince him that respect for life is the mature state of all intelligence—including that scary future space bug?