A Gaseous Vertebrate

Interestingly enough, once we stop identifying finalist activity with conscious intelligence, some of our objections to Ruyer's position can begin to evaporate naturally.  Because if the human brain is merely a tool of a larger organic activity, if it merely reveals, but does not create finality, then we no longer need to see this ongoing organic finality as a tool of an even larger brain.  This may not answer our question of where finalist activity begins, or what it is the ultimate finality.  But it at least defuses the threat of a regress of conscious intelligence that would remake this finality in the image of man -- God is not a "gaseous vertebrate" who designed human intelligence as a tool of his divine plan.  Once we leave behind the anthropocentric assumption that all intelligence must look like our own, once we see the workings of our own brain as merely a peculiar case study in how finalist activity can be materialized, we can appreciate how the root of Ruyer's activity is much stranger than we might have thought.  Ruyer's explanation of this is so clear that I'm just going to quote the whole thing.

Let us translate this thesis into our schema (Figure 19). In the finalist act in external circuit, the brain is an indispensible link (the cook uses his brain to make a dish), whereas in internal circuit, the stomach, for example, operates like a mixer or a furnace regulated by its own nervous centers. If we then admit organic finality by equating the stomach itself to an invented tool, it seems that a second external—"supernatural"— circuit is needed to explain this organic tool, a circuit controlled by a consciousness and even a supernatural brain. Such a duplication of humans or animals by a fabricating God—by a "gaseous Vertebrate," as Huxley said—is not very seductive to biologists.

 

But the error is clear. Finalist action in external circuit, which presupposes a good condition of the brain, is only a complication of finalist activity in internal circuit. It is thus logical to consider the brain as the instrument of this complication, but not as the instrument of finalist action in general. The central nervous system, extended by the eye and the hand, allows the organism to project its finalist activity into the external world; it enables it to structure and organize a vast domain, beyond its integuments and internal organs. The brain expands the field of organic finality; it allows finality to spill over onto the world, to discover materials within it, to construct tools and machines that are organic by their form and not by their matter. But "to transport" or "to expand" is not synonymous with "to create" or "to bring into existence." This is not a curious, paradoxical, or even personal thesis; it is not even, strictly speaking, a thesis. It is the pure and simple statement of obvious facts, of one fact above all, which no one can contest: the organism forges its nervous systems before using it. The brain is thus an "organ of transport" of finalist activity; it is evidently not its "organ" tout court.

While simple and maybe even obvious, I feel like this realization that human intelligence is produced is so profound that we should pause here to appreciate some of its other consequences.  Because we are constantly back-projecting the endpoint of this production as something that was there essentially from the start, then covering over the fact that this is what we have done.  Consider our use of the machine as an analogy for causal determinism.  To say that something operates 'like a machine' or 'mechanistically' is to say that it is purely a meaningless series of feed forward causes and effects, just a chain of billiard balls bumping into one another.  And indeed, this is how our machines operate ... because we built them that way.  We built them as tools for accomplishing goals that come from outside them.  Yet we go on to apply this same notion of what a machine is to natural organisms, and even to nature as a whole, when we claim that everything operates as chains of purely mechanical cause and effect.  Then we go on to complain about how foolish people are for taking a gaseous vertebrate to be the designer of the machine of nature even though the analogy between mechanism and nature relies entirely on the human designer who makes the machine what it is.  That is, we apply the analogy, but erase half of it in a way that reminds me of Einstein's famous joke about wireless telegraphy.  In short, we define the whole notion of mechanistic cause and effect as a chain that can't aim itself.  But this very definition depends on the notion that there is something which can aim itself.  Chalk this up as another of Ruyer's "paradoxes of antifinalism".  

Something similar happens when we define intelligence as how the normal adult human brain operates.  Then we forget that this intelligence requires an incredibly complex organic and cultural production to come into existence.  It is anything but simple or fundamental or 'normal'.  These reflections may seem to be at odds with a neo-finalist philosophy which embraces the existence of overarching ends that organize step-by-step means.  But this tension evaporates when we realize that Ruyer is simply cautioning us not mistake a particular means for the end.  The end is not a particular point but an activity, a process of laying out a space and establishing goals and directions. So the brain is not the goal.  It is a means of reaching the goal that we might simply call 'thinking'.  It's not that the brain thinks, but that thinking may, at some point, employ a brain.

Ruyer goes on to explain in detail why the brain cannot have a monopoly over the factors we associate with thinking -- memory, habit, invention, signifying activity, consciousness -- for the simple reason that the organism that produces the brain already exhibits all these traits.  Obviously, in the case of consciousness, this is a controversial statement, and leads Ruyer to distinguish between a primary organic consciousness and the secondary consciousness of objects that we usually associate with the brain.  The latter is a differentiation or complexification of the former in the same way that the fully developed organism is a differentiation of the embryo.  In both cases, there is a primary activity that is carried further, but not created, by the organs that serve as its tools of realization.  This point is actually closely related to Varela's insistence that the brain is there to do things, not represent things -- it is already embodied.  We will fail to understand it if we treat it as a purely computational device designed to represent pre-given objects because the significance and even coherence of these objects is defined at the level of the whole organism that uses the brain.  At the risk of adding terminology, we might clarify things if we say that dualistic consciousness is merely a small subset of organic awareness.  Awareness doesn't begin with machine constructing adult humans -- these are merely steps in its amplification or differentiation.

Of course, we still don't really know where this organic awareness begins nor much about what it is like (except by analogy with our own sense of self-existence).  So far, Ruyer's goal has been simply to convince us to admit its existence as a dimension of reality alongside the material world.  The main idea of Chapter 6, which compares the functioning of two 'loci' of awareness or finalist activity-- the brain and the embryo -- reiterates this same dualistic thesis by arguing that the remarkable capacities of these tissues cannot be localized in space and time.  That is, the capacity of the brain to produce a thought or the embryo to produce an organism cannot be explained as an outcome of step-by-step causal interactions of genes or neurons or any other physical components.  These structures are merely the objective expression of an awareness that is outside of space and time, an expression which only proceeds through these tissues as a whole.  So, again, the main point is to maintain a dualism between the "domain of sense" and the "domain of causality" (NF, 44).  But this long and complex chapter also gives us the first hint of a structure of finalist activity itself.  It proceeds from what we would usually call the abstract to the particular, from the whole to the part, just as the embryo proceeds from a definition of axes (anterior-posterior, medial-lateral, superior-inferior) to a formation of appendages to a specification of digits, all through a series of cascading differentiations.  Finalist activity works from the top down, not the bottom up.

Ruyer calls the crucial feature that the brain and the embryo share "equipotentiality".  This means that they operate as a whole and in terms of unities, and not step-by-step in terms of parts.  This isn't because they are special chunks of matter though.  Ruyer is quick to dismiss the idea that its something unique about genes or neural networks as physical phenomena that somehow magically turn them into unities.  While it's tempting to think of Ruyer's idea as a form of emergence, and consider his finalist activity as a form of feedback loop, I think this would be inaccurate.  For Ruyer, nothing emerges from matter.  As a dualist, there's no 'hard problem' for Ruyer because awareness is there from the beginning (though we still might wonder why it seems especially related to this particular bit of matter).  These tissues are equipotential because they already have finalist activity dissolved, as it were, or expressed, in each and every part.  The brain and the embryo are 'holographic' in this sense -- the whole can be reconstructed from just a part.  This whole is not the sum of the parts, but stands outside them, organizes them, and tries to express its ends through them in the same way we try to express our desire to build a house by using hammers and nails. 

This organic awareness outside of space and time may all still sound kinda vague and mysterious.  Fortunately Ruyer's discussion of the embryo gives us a concrete reference to guide us.  Obviously, an embryo is in some sense holographic.  The whole tree is there in the seed, or at least the seed seems to aim at the tree as a whole.  And the tree is also, in a manner of speaking, equally distributed throughout the seed.  In fact, we can often cut a seed in half and still come out with a whole tree.  That is to say, something about the seed reflects a unity that doesn't have to do with the extensive parts of seeds or trees.  Normally, of course, we see embryology in exactly the opposite sense, as a demonstration of the mechanical step-by-step construction of an organism.  We think of the DNA as a blueprint for how to build a machine organism.  But if the final organism is a machine, it sure is a peculiar one.  It is not assembled as a synthesis of independent parts but rather as a diffferentiation of a whole.  Somehow this whole seems to be given right from the start, which is in fact precisely why we frequently resort to the metaphor of a blueprint.  If the organism were a step-by-step construction following a blueprint, we would see trees with branches only on the left side, or that made ninety degree turns in the middle.  That is, if a genome were really like a blueprint, then its mode of failure would also be like a blueprint's.  We would see pieces randomly shuffled from their correct position, or the whole structure collapse because the foundation wasn't connected correctly.  This is how our human machines fail -- catastrophically.  One missing piece in the causal chain results in a heap of junk.  But this is not at all how an embryo fails, as Ruyer's brief tour of the study of monstrosity indicates.  People have modified embryos and later genomes in every kind of way.  And of course, you can make such large alterations that nothing grows at all.  But what you can't do is simply treat the organism like a heap of parts that can be reassembled in any order.  You can't just draw up a new blueprint that puts the living room in an arbitrary location.  The modifications all have to follow and influence the overall trajectory of development that's normal for that organism, no matter how surprising the results may look to our eyes (Armand Marie Leroi's Mutants provides a great look at the sort of variations that are possible).  Experimental embryology has uncovered a lot of how this mechanism works.  The organism is structured as a series of progressive differentiations, each level of which seems to have a well defined end or guiding theme. 

Thus embryonic equipotentiality, like cerebral equipotentiality, is bound up with the thematic character of development. The cascade of determinations has a thematic character, because determination precedes differentiation and differentiation proceeds in its turn through themes that can only be designated with abstract terms: a limb bud is determined as a "foot" (as "foot" in general) before it is determined as right foot or left foot. It is also tied to its finalist character (in the strictly etymological sense of the word) because, in all the extraordinary regulations allowed by equipotentiality, the normal end is reached despite the operative disruption of conditions, materials, and means. (NF, 50)

While we've discussed the embryo because it is a particularly clear example, Ruyer argues that the brain works the same way -- as a whole with a goal or theme.  This is why localized lesions to the rat brain don't seem to have much specific effect, and it takes massive damage to the cortex to make it forget a maze.  It also accounts for why the rat never forgets just part of the maze, or why it can reconstruct the overall trajectory even when its previous means of accomplishing the goal is blocked (NF, 63).  The end of reaching the cheese can be expressed in the brain as a whole because this end doesn't exist at a particular point in space and time.  It's as if this theme is infused throughout the brain equally, present in every point, just as the whole tree is present in the seed.  Since the end is everywhere within the tissue, any part of it can serve as a whole, which accounts for the remarkable fecundity of equipotentiality.

I want to pause for a minute here at the conclusion to address an objection that continues to pop up in my mind even though Ruyer has actually already addressed it.  Somehow, this idea that the embryo has a goal of producing a 'normal' organism still seems hard to swallow.  I'm so accustomed to thinking of it as a purely casual chemical process.  Sure, it doesn't work like a normal machine because of its top down differentiation.  But perhaps this merely indicates that our concept of a machine is needlessly narrow.  Does the difference really indicate that embryogenesis cannot be explained as a step-by-step causal process?  Certainly, I'm sympathetic to the argument that this causal process is not limited to the action of genes, and that the genes therefore cannot be thought of as a blueprint.  The seed needs the soil and the water and the light, and the hundreds of genetically identical seeds from the same tree will develop into surprisingly different looking organisms.  So the causal process cannot be laid at the doorstep of genes alone.  But does this license us to say that no causal process can explain embryogenesis?  The more I've thought about it though, the more I've seen that this objection misses the point because so much depends on the word "explains".  Ruyer isn't denying the existence of a causal mechanism that produces a tree, he's simply insisting that this process cannot be explained by a law-governed but meaningless or pointless unfolding, but is somehow guided by its endpoint right from the beginning.  This explanation fails for factual reasons -- if the embryo expressed no goal it would not degrade in the systematic way when its normal development is thwarted.  So it sure seems to act as if it 'wanted' to make a tree.  But it also fails for logical reasons -- strictly speaking, the causal explanation doesn't try to explain the tree at all because it doesn't consider it a real entity.  In this mechanistic explanation, the form we call "tree" is simply a convenient shorthand for certain statistical regularities in the behavior of a set of chemical reactions.  Strictly speaking, there is no tree, and therefore nothing to be explained.  And if we ask why just these regularities exist, we are pushed back to an evolutionary explanation that claims the 'tree-form' just happens to be the set of DNA that best propagates itself in these conditions.  In other words, we are told that the end that apparently guides the development of the seed is ... chance.  Evolutionary explanations never actually explain any thing as anything more than chance, which is why they end up not explaining anything at all.  But as we saw with the axiological cogito, this view of the universe as a purely chance causal process works only up to the point it is articulated, at which moment it becomes self-negating.  But if we admit the meaningfulness of our own thinking, we have admitted at least one thing in the universe falls outside its purely causal mechanics and yet certainly appears to effects those mechanics (judging, at least, by the frequency with which I actually articulate what I mean when I open my mouth, as opposed to simply a series of grunts and hoots -- this frequency is certainly not 100%, but I submit to you that it is higher than chance).  From there we enter this slippery slope where we are forced to reintroduce agents and their meaningful activity back into the world as guides for causal mechanisms.  This gives these mechanisms an end or meaning.