9. Emergentism and Downward Causation
10. Informational Emergentism
11. Mind as an Emergent Phenomenon
Throughout this work, we have noted the conceptual incoherence of reducing intelligence to material representations or to signal processing. Recognizing the shortcomings of a crudely reductionist materialism, many modern thinkers have sought to explain the mind instead as a macrophenomenon that somehow emerges
from the synthesis of neurological microprocesses, as something that is greater than the sum of its parts. Furthermore, they recognize the mind as a real agent that can impose its will on parts of the brain, and by extension the body. The notion of a large-scale or higher-order entity having properties not held by any of its constituents is generally called emergentism,
in contrast with strong reductionism, which would make the larger or higher-order entity nothing more than
a synthesis of its constituents. The idea that the mind or any other supposedly emergent entity can act as a causal agent on its own constituents is known as downward causation.
Emergentism and downward causation are at least nominally opposed to reductionism, but in practice they are often posited in a reductionist or materialist spirit. In such a view, the material constituents are the only causes of the emergent phenomenon, and any property of the emergent entity beyond those of its constituents can be completely accounted for by the relations among the constituents. Thus there is no need to introduce any fundamental principle beyond the material in order to explain
the mind causally. This kind of emergentism covertly reintroduces reductionist materialism, though in a milder form, and becomes subject to the same logical and metaphysical criticisms we have noted previously, such as the confusion of ideas with their representations. There is a further problem with materialist emergentism, insofar as it upholds the thesis of downward causation while still maintaining that the mind is caused solely by its material constituents. This seems to create a causal loop, since the mind is caused by the behavior of its neural constituents, yet somehow it has the power to alter the behavior of these constituents. Clearly, this paradox needs to be resolved if emergentism and downward causation are to be more than hand-waving rationalizations of materialist reductionism.
Before we consider whether the mind may be regarded as an emergent phenomenon, either in a materialistic or non-materialistic sense, we should first establish what kinds of emergent phenomena are possible in nature. In particular, we wish to know whether a large-scale or higher-order phenomenon can emerge that is:
(1) more than a simple composition of its constituents;
(2) has causal agency;
(3) can act causally upon its own material constituents; and
(4) has activity that is not reducible to the local physics of its constituents.
The answer to Item 1 of the question is undoubtedly affirmative, but the remaining parts are more controversial. Item 3, which is downward causation, would result in a causal loop if we denied Item 4, as do materialistic emergentists. We will examine some basic physical phenomena to clarify our thinking on this matter.
It is well established that there are emergent phenomena in nature, if we take emergent
in the most generic sense (Item 1 above) of being more than a composition of its constituents. Common examples are pressure and temperature, which are real physical properties of gases and fluids, yet cannot be defined for any individual molecule in a substance. Since the features of pressure and temperature emerge
only at a large scale, while no individual molecule possesses these properties, they may be considered examples of a gas or fluid being more than the sum of its parts. Nonetheless, the large-scale thermodynamic properties of gases and fluids are fully explainable, at least quantitatively, in terms of the statistics of local interactions among individual molecules, apparently contradicting Item 4 defined above. Further, the application of pressure and temperature changes on a macroscopic level does not cause individual molecules to act locally in a way contrary to the ordinary laws of physics and chemistry, seeming to contradict Item 3.
I note that these contradictions of strong emergentism (Item 3) are only apparent. The laws of thermodynamics mathematically emerge from analysis of statistical mechanics of an ensemble of local interactions, but a mathematical analysis does not necessarily give us all of physical reality. If I knew nothing but statistical mechanics and its proofs of thermodynamic laws, it is highly unlikely, to say the least, that I would grasp anything of what it means for a substance to have pressure or temperature. These are qualitative realities that cannot be expressed in terms of force interactions among point-like particles. I need to grasp a higher-order concept, that of a substance such as a gas or fluid, in order to comprehend such properties. We think we can see
how local physics gives us higher-order properties only because we have learned things in reverse order, from the macroscopic properties down to the subatomic level. We already have well-defined concepts of the macroscopic properties, enabling us to make the requisite conceptual leaps from lower-order physics.
Similarly, the mere fact that a macroscopic phenomenon does not abolish local physics does not necessarily prove that the phenomenon is reducible to local physics (contra Item 4). It is possible, for example, that the phenomenon consists precisely in a global coordination of local interactions. In that case, we should expect to observe no violation of local physical laws, though there is a very real coordination or orchestration of forces. Such a hypothesis is extremely pertinent to a discussion of the mind, for we observe that mind, or more generally soul, consists in the coordination of physical activities to a common end. When we move our limbs, we do not suppose that we are overriding the laws of physics, but rather we are making use of these laws in a directed manner.
Global and local forces in a statistical mechanical ensemble of particles are not two different kinds of forces, but the global force is nonetheless a physical reality to be reckoned. In a coordinated system, we must consider the local forces—i.e., the push and pull between nearest neighbors—in the context of global behavior. Strictly speaking, we cannot use quantitative analysis to disprove that the global behavior is causally or metaphysically fundamental (Item 2), since local and global behavior are simultaneous. We can only prove that the synthesis of local forces is mathematically equivalent to what we observe globally.
Still, we have some reason to believe that the local forces among constituents are more fundamental, for these forces can be used to account for all systems, simple or complex, while the putative global forces only apply to sufficiently complex systems, and never act in a way that contradicts local force laws. Nowhere in the corporeal world do we find evidence of agency that cannot be explained in terms of the fundamental forces of physics (gravitational, electromagnetic, weak and strong nuclear forces). Attempts to demonstrate empirically the presence of an elan vital or other overtly global causal agent have all failed, which is what led to the abandonment of vitalism in the first place. In particular, the behavior of individual neurons is explainable in terms of ordinary forces, so no new force ought to be needed to explain an ensemble of neurons. Nonetheless, there remains the possibility of global cooperation or coordination among neurons, even without the introduction of a new kind of force.
Modern scientists and philosophers of science generally prefer local theories of causation over holistic or global theories. This is not a rational necessity, but a cultural preference. There is nothing intrinsically absurd about action at a distance,
for it does not entail any logical contradiction. Nonetheless, there is a centuries-old prejudice against the concept, so that scientific thinkers will avoid it even when it is suggested by the facts. Isaac Newton famously refused to accept that massive bodies could exert gravitational force upon each other across a void without any intermediary, even though he could not explain his force law in terms of local interactions. His understanding of a rational physical cause involved one body physically touching another body. To act at a distance would seem to require the mediation of intellect, since one particle would have to effectively know,
either by its own power or the power of another being, the state of a remote object. This was too magical or mystical for Newton (who was not averse to mysticism in other parts of his philosophy), as he insisted that space must be the sole medium of physical interaction. Modern scientists have adopted this position even more strenuously, motivated either by outright atheism or by an epistemic bias against intelligent causation in the natural world. These philosophical positions are cultural developments, not the product of any logical or empirical necessity.
As another example, electromagnetic interactions could be most naturally interpreted in terms of action at a distance, but physicists have chosen to postulate substantive fields
that mediate these forces locally. There is no empirical proof that electromagnetic fields are real, substantive causal agents, rather than mere mathematical conveniences for calculating force strengths. Nonetheless, most physicists prefer to think of electromagnetic interactions in terms of local interaction with fields permeating empty space, rather than particles acting on each other from a distance. The belief in electromagnetic fields as substantive entities persists, even though the existence of an electromagnetic medium or "aether" has been disproved. This shows that the belief in locality is motivated more by philosophical preference than by what the facts suggest. (See Causality and Physical Law, II for a further discussion of electromagnetic fields.)
This bias towards causal locality is even more evident in the physicist’s treatment of quantum entanglement. This is a phenomenon where the state of one particle is necessarily the complement of its remote partner, so that if one measurement forces a particle to resolve itself into a determinate state, then its distant partner must be resolved into the complementary state. This seems to provide weighty evidence against a strictly local, direct contact model of physical interactions. Still, most physicists will deny that there can be any causative link between two remote entangled particles, or at least insist that there can be no communication of the state of one particle to the other. Indeed, the strained rationalizations they employ to deny action at a distance only strengthens our contention that the principle of locality is more of a philosophical bias than a straightforward analysis of empirical results.
A cultural aversion to any kind of holistic action at a distance also prevails in biology. Biologists like to pretend that all the actions of an organism can be explained entirely in terms of local chemical interactions. Yet the existence of a mind that is a real causal agent and is not simply the slave of its brain chemistry flies in the face of such a philosophy. If my will is free, it cannot be reduced to local chemical processes, which are all deterministic. The neurons must be viewed as agents for manifesting and implementing my will, and they might even influence the will to some degree, but if we allow that we are not complete automata, we must acknowledge a role for non-local causation, since conscious states manifest themselves as large-scale phenomena across large regions of the brain. This non-locality need not be action at a distance between individual neurons (such as the field postulated by Libet), but might be a holistic, coordinated activity among the neural signals. I should note that there is no fact of experimental neuroscience that positively contradicts this supposition. Rather, like their physicist counterparts, neuroscientists choose to interpret all observed facts without this supposition, out of a similar philosophical prejudice. However, if we were to accept a materialist principle of locality as a dogma, we would be forced to admit a strong determinism of the mind, which we have seen to be a self-stultifying position, not to mention utterly contrary to our most fundamental psychological experience, our sense of self, upon which all other interpreted observations, including scientific observations, are dependent.
There is, of course, a reason for this philosophical aversion to holism. To admit any quarter to non-locality would be to allow the soul to enter biology. The very function of a vitalistic soul is to coordinate diverse parts to a common end, which is exactly the sort of purpose-driven mysticism modern science prides itself in expunging from reality. Yet, as we have noted, there is nothing properly irrational about holism. In fact, if we were truly to deny holism in biology, we would have to say that ‘organism’ is just a convenient label for a collection of molecules, and that there is no more agency to the organism than that which occurs in local interactions. If this is really the case, then even lower animal consciousness is an inexplicably superfluous phenomenon, and locomotion is but an illusion, for the supposedly self-moving animal is merely pushed and pulled along by forces just as inexorably as dust in the wind.
We see that this nineteenth-century style determinism leads to views about animals as misguided as those of Descartes, who regarded brute beasts as automata. Modern materialists compound the paradox by including humans among the animal-machines, as if all biology could be reduced to chemistry. Biology without some form of holism or vitalism is essentially a specialization of chemistry; it denies its subject, for life
does not really exist as a physical agent, but is just a subclass of deterministic chemical processes. Although modern physics allows for non-deterministic activity at the subatomic level, this washes out at the scale on which biochemistry occurs, so that all biology is effectively explainable in deterministic terms, if we accept that local chemical interactions are the sole driving force of an organism.
The way out of this deterministic morass, which would force us into a view of humans and other animals that is profoundly contrary to our experience (and as anti-empirical as Parmenides’ denial of motion), is to scrutinize the philosophical error of reductionism. By reductionism, I mean the view that the smallest constituents are the most real
things, and all larger scale phenomena are merely artifacts or effects of the activities of the constituents. A clock, for example, is nothing more than
the synthesis of its parts, which in turn are nothing more than
the synthesis of their constituent substances, all the way down to fundamental particles. Thus it is less proper to speak of a clock as an objective physical entity than of an electron. The term ‘clock’ is just an artificial construct we impose on a group of particles acting together in a certain way.
The philosopher Mortimer Adler proposed, on the contrary, that the macroscopic whole can be considered more real
than the constituents. If I may take his argument to the atomic level, when atoms merge to form a molecule, they no longer have the same properties in the bound state as they did as free particles, nor is the combined wavefunction equal to the sum of constituent wavefunctions. From now on, the atom can act only as part of the molecule. It is only virtually
an atom, since it contributes its attributes to the whole, and has the potential of becoming a free atom again, but it is really
a part of the molecule. Now the molecule is the thing that can act as a physical agent, not the atoms as atoms. Similarly, when a molecule is merged in a polymer, that plastic acts as a single substance, and the constituent molecules are now molecules only in virtu. Lastly, a machine consisting of many parts bound together now acts as a single machine, and the parts have no choice but to act as part of the machine, not as free substances. Thus, the higher order macroscopic objects are the most real.
This interpretation of reality makes perfect sense, and would account for why we have been able to give a rational explanation of causality among macroscopic objects long before we discovered atomic physics. The large-scale interactions are not just accidents or artifacts of particle physics, but real physical forces acting on bodies as a whole. Atomic and molecular physics can explain how particles combine to form unified bodies that each act as a whole, but they do not thereby abolish the reality of the whole. The fact that large-scale physical laws can be mathematically derived
from small-scale interactions proves only that the universe is quantitatively consistent, as indeed it could not be otherwise, regardless of the direction of causality or metaphysical priority.
It is one thing to make the metaphysical judgment that the whole is more real,
or fully actual, than the constituents, which exist as distinct entities only virtually, but it is another to say that the activity of the whole can cause local activity in the parts. (Item 3) The local and global activities are simultaneous in time, so, on empirical grounds, we could argue for causality in either direction. It is obvious that a ball cannot move unless its constituent atoms move, but it is also the case that, as long as they are in the bound state, the constituent atoms cannot move far unless the ball as a whole moves. The bound state is a physically real state of affairs that denies the atoms their primitive freedom, so they are as much slaves of the whole as the whole is constrained by the properties of its constituents.
When I set a ball rolling down a hill, I impart motion to it on a macroscopic scale, touching countless atoms at once. It does not really matter much which parts of the ball I touch, but only the magnitude and direction of the macroscopic forces imparted (i.e., straight pushing, spin, torsion) with respect to its center of mass. As it is rolling down the hill, I can model its motion with perfect predictive power—a common criterion of causal agency—without recourse to atomic physics. In fact I can predict the positions of every microscopic part of the ball based on my knowledge of the motion of the whole. Intuitively, it makes sense to say that the ball as a rigid body is what is moving, while the individual atoms are just along for the ride. It is arguably an abuse of language to say that the atoms cause the ball to roll, as this seems to suggest that they all just happened to move at the same time in a highly coordinated way. In fact, they have no choice but to move with the whole. Once they are bound as a rigid body, the constituent atoms are mechanically subservient to the collective. This is not the only possible physical account of a rolling ball, but it is fully intelligible, valid, and physically intuitive.
Nonetheless, we must recognize that the global constraints enabling the ball to act as a unity are mediated or implemented by local field interactions among the constituents, so we do not need to invoke a new kind of force beyond the local forces. Perhaps, then, we could say that the global force of the ball is nothing more than
a synthesis of many local forces acting collectively and cohesively upon a particular constituent. Such an account gives due regard for both the reality of global agency and its dependence on its constituents.
Yet in biology, we are concerned with a different kind of downward causation. When we consider animals and their minds as real causal agents, we are affirming the existence of global forces
that are not reducible to the local physics of constituents. (Item 4) This is because in animals we perceive a kind of self-agency that is not strictly determined by extrinsic conditions. If this agency were reducible to local forces, which are all deterministic chemical interactions, it would lose this character of self-determination.
It should be clear, then, that a strictly materialist or mechanistic whole-to-part
downward causation (affirming Item 3 while denying Item 4) cannot account for mental agency, without stripping it of its distinctive character. Emergentist
arguments in this vein are really reductionism in disguise. Those who wish to perpetrate the deception most completely will even try to pretend that they have thereby explained
free will, when they have actually explained it away. If the global agency of the mind is entirely reducible to the local chemistry of neurons, with the latter being strongly deterministic, it follows with inexorable logic that the global agent must also be strongly deterministic. The fact that the local interactions are so complex and numerous that we cannot reliably predict a future state does lessen the objective determinism of the system in the slightest. Trying to generate free will
in this manner is either getting something for nothing, or perpetrating a fraud by calling such a will free.
This reality might be more clearly perceived if we recognize that there is nothing special about neural networks, and that we could effectively replicate them with other much less mysterious systems. Instead of neurons firing signals at each other when certain potentials are reached, we could have a system of pipes and valves opening and closing when certain pressure is reached, so the pressure effectively acts as a signal like voltage. Few would believe that any such system of pipes and valves would become conscious and self-determining if they were made sufficiently complex, yet this is precisely what many believe about an analogous neural network. A trillion pipes and valves might be configured to make a powerful calculating tool, but all will recognize that its behavior is strictly deterministic. Going further, we might conceive of a network with billions of people sending signals to each other by semaphore. Shall we consider the collective network to be conscious, above and beyond the consciousness of individual people? For an ideology that is committed to avoiding mysticism, this kind of emergentism has the potential to create many new mystical entities.
In economics, it is well understood that aggregates or macroscopic variables should not be viewed as economic agents. It is wrongheaded to say, for example, that inflation reduces unemployment, for both inflation and unemployment are large-scale aggregate results of innumerable microeconomic decisions. It is folly to seek agency in the aggregate, though many try to do this with the stock market, insofar as they use technical analysis
(i.e., trying to predict market behavior based on the shape of the curve) These predictions are only as strong as the observed statistical correlations, but go wrong when they suppose that there is some kind of causality going on at the macro level. Philosophers of science make precisely this error when they try to make the mind just an aggregation of neural signals without holistic coordination.
We should repeat that if there is, on the contrary, holistic coordination among neurons or other cells, present scientific techniques are incapable of determining this. Since the question cannot be decided empirically, we may at least inquire whether the supposition of such coordination is necessary in order to have truly self-determining mental agency.
Before addressing the question of whether holistic coordination is necessary to account for mind, there remains another kind of emergentism to be considered, which Paul Davies calls level entanglement.
As Davies admits, this can be a confusing choice of terminology, as it has nothing to do with the phenomenon of quantum entanglement.
Instead, level entanglement is when some higher order phenomenon controls a lower order phenomenon in the same system. Putative examples of level entanglement include: investor behavior being affected by macroeconomic aggregate variables; software controlling hardware; quantum wavefunctions affecting particle behavior; and mind controlling the body. Let us consider each of these in turn.
We have already noted that macroeconomic variables are just aggregations of microeconomic results, and are therefore powerless to exert causation on some macroeconomic order, though it may sometimes be convenient to speak of them in this way, saying, Inflation lowers unemployment,
for example. There is, however, a way in which macroeconomic variables can affect microeconomic activity. Investors are familiar with the computed values of macro variables such as stock market indices, unemployment statistics, and GDP growth. They may, and often do, choose to change their investment decisions based on knowledge of the macroeconomic data. In the case of the stock market, individual investor behavior may cause the average market prices to be volatile. The investors in turn become aware of this volatility, and alter their investing behavior. It is not that the aggregate variable as such has the power to alter investor behavior, but rather knowledge or information about aggregate activities can cause economic constituents to behave differently. This example is different from putative whole-part causation only if we can construct a non-materialist theory of information, so that information is something other than the material constituents of a system. Again, we will see that materialists try to have it both ways, by invoking the concept of information
to explain non-materialist phenomena, and at the same time treating information as if it were reducible to its material representations.
This comes out a bit more clearly in the hardware-software example. We commonly speak of software
as commanding or controlling hardware,
but what is software
really? Is it not just the same computing system considered on a higher order: the order of information?
The hardware is simple enough to understand, for it is just a mechanical or electrical system that slavishly obeys the laws of physics for the local interactions of its material constituents. A software program is basically a configuration of physical states—for a bit is ultimately the physical state of an electrical component—which causes the hardware to alter its state in accordance with an algorithm defined by that configuration of states. Software
acts on
hardware not by imparting knowledge to the unconscious machine, but by informing
it in a more generic sense, namely, imposing its form onto the hardware. That is, the physical configuration of states that is the software imposes a corresponding change of state in the hardware, following the usual laws of physical interaction. The intelligence
of software is in the men who, knowing how the hardware is wired, know what configuration of states to put in their software in order to achieve the desired result. (This assumes they code in assembly language. Higher-order languages are just higher-order representations of states, so the programmer does not even need to know anything about hardware.) There is no need here to invoke a concept of information
that is anything more than a configuration of physical states.
If the information
of computers is just a set of physical states, then this cannot suffice as an explanation of human knowledge, which is more than mere physical representations. A computer does not need to have any understanding of what it is doing, either on the software or hardware levels. Still, the broader concept of level entanglement
may prove useful in trying to account for the mind. In the case of hardware and software, hardware is the physical stuff of a system, while software is a configuration of physical states in a system. Clearly, changing the physical state of the system changes the stuff of the system. There is nothing here that is contrary to ordinary physics. In fact, Davies regards this kind of emergentism as only so much word play, since no new physical force is introduced. We simply alternate between conceptualizing in terms of stuff
and information
(i.e., physical states).
In The Ghost in the Atom (1986), Davies proposed a clear example of hardware-software feedback where a robotic arm is programmed to rewire itself, which effectively changes its programming for future self-modifications. Here, the software
might be thought of us as the mind,
while the robotic arm and its circuitry is the brain. Still, as Davies points out, this system can be accounted for completely in terms of hardware, so there is no strong reason to believe that software
is anything more than a conceptual convenience.
Could the mind be a kind of software
that re-wires the brain? We do in fact observe the brain constantly re-wiring itself by extending new dendritic spines. On a local level, these protrusions seem to grope about haphazardly, until making a connection with the first axon they happen to intercept. This hardly seems like a promising way to set up an efficient neural network, unless these connections are coordinated on a large scale. Yet if the mind is software
only in the sense of computer software, being nothing more than a configuration of physical states of neurons, we are stuck with the same problem that plagues all materialist accounts of the mind: there is not even the rudiment of subjective mental experience (intension) in the mere corporeality (extension) of the mathematically defined physical states of neurons. To say that the mind is just the brain reconceptualized is to make an unjustified, and indeed impossible, reduction of intensionality to extensionality. We cannot square the circle by simply defining the circle to be a reconceptualized square.
Perhaps, then, the mind is software
in a higher sense than the software of a computer, being a real ghost in the machine
that is genuinely causally distinct from the neural hardware, not just a reconceptualization of that same hardware in terms of its physical states. This reality is suggested by the brain’s remarkable plasticity, as it is constantly re-wired, and more importantly, the mind is able to reinterpret the significance of neural signals. One astounding example is given in Norman Doidge’s book The Brain That Changes Itself (2007), where several monkeys had the nerves in their hands rewired, so that each input came from the wrong part of the hand. Within minutes, after a brief adjustment period, the monkeys’ spatial map of the hand was nearly normal again, as it somehow figured out
which signals were really coming from which part of the hand. This proves, at a minimum, that the brain is not a slave of the signals it receives, like a computer. Yet how can this be, if the brain is just a signal processor? Clearly, there is some interpretation going on, if only at the unconscious level. Someone or something is effectively saying, Wait, this can’t be right. The signals should be interpreted this way instead.
Before we get carried away, we should note that the normal recovery time for monkeys in other rewiring experiments is around 20 months. What we have here is not a simple matter of a sentient mind understanding what is wrong with the inputs and then correcting it. Plasticity occurs at a neuronal level, and is limited by real physiological constraints, which is why stroke victims often take months or years to re-learn processes they already know
intellectually. This praxis knowledge (learned executable processes) is somehow encoded neurologically, and its re-acquisition depends on neurological means. Still, it is not clear how purely mechanistic signal processing could know
that it should initiate a re-mapping, unless perhaps it gets visual feedback by looking at the hand, and seeing that the wrong
finger is moving. This kind of re-learning should take a lot of time, so it is hardly a plausible account for the rapid re-mapping in the monkeys discussed above. Even in animals re-learning slowly through visual feedback, it would seem that this feedback is necessarily mediated by the mind. There is an important role, then, for the mind, rather than the brain, in the phenomenon of neuroplasticity.
Although most neuroscientists plainly acknowledge that psychological experience
is able to cause a re-shaping of the brain (Items 2 and 3 from the previous chapter), those of a materialist bent tend to conceive of neuroplasticity as the brain rewiring itself.
This is a strange way of speaking, as if the brain (or any entity) could be abstracted from itself, and the same thing is both cause and effect. There are two problems with this model, an ontological problem and a causal problem.
The ontological problem is that this model treats psychological experience as if it were a brain state and nothing more. This makes no sense if the brain is conceived in purely materialistic terms. It is profoundly self-evident that psychological experience qua experience is immaterial. ‘Experience is material,’ is a grammatically valid sentence, but does not correspond to an ontologically coherent judgment that any person can actually comprehend, since it is contrary to what we know in our very mode of consciousness. My experience is eo ipso experienced subjectively, so it is pointless to say that my experience is objectively a brain state, but only subjectively immaterial. The subjectivity is the experience. The materialist must either deny the existence of subjectivity, or affirm that our experience is subjectively material. Both of these options are instantly refuted by simply contemplating one’s own subjective experience for a moment. Thus, in humans at least, and probably in other higher animals as well, there is a role for an immaterial mind in neuroplasticity.
We still acknowledge the likelihood that much of neuroplasticity can be accounted for by processes not involving subjectively experienced feedback. Although the ontological problem may not apply here, there is still a causal problem that needs to be sorted out. If the thing that rewires itself is nothing more than the brain states, we must make sure that we are not guilty of creating any causal loops or paradoxes.
Taking the hypothetical example of Davies’ self-modifying robotic arm, we can avoid causal paradoxes by using sequential, discrete steps. First, the existing program causes the arm to perform an action of removing or replacing a circuit. Supposing that circuit itself contains the instructions for its removal, we would have an obvious problem if the program needs to be running simultaneously with the arm’s execution. There must be some delay between the delivery of signals from the circuit and the execution of the code by the arm. This way the act of removing the circuit will not prevent the arm from moving, since it has already received its instructions. Now, when a new circuit is placed, a new set of instructions is defined, and the arm will execute these in the future until they are altered.
Note that it is critical for there to be some delay between these sequential, discrete steps in order for there to be a coherent direction of causality. We may express this schematically below:
If, however, we made the execution of instructions dependent on their simultaneous presence in the circuitry, we would have either a causal paradox or a complete failure to execute. What about something like the brain, which runs many processes at the same time? We might conceive of this as many robotic arms as rewiring their own or each other’s circuits. Here there can be no guarantee that the rewiring process will not interfere with an arm’s execution, unless everything is superbly timed. There is no reason to believe this is the case in the fast-paced environment of the brain, where there is constant re-wiring and initiation of new processes simultaneously and continuously. While it may be impossible for a single process to alter itself in real time, various processes might re-program each other, without creating any genuine paradox, though the feedback will soon become insanely complex.
There is room, then, for some cases of the brain rewiring itself,
though to avoid paradox we most postulate that certain processes alter other processes, or that a process can alter itself only after some iterative delay. This last supposition seems incongruous with the continuity of neuroplastic behavior, so it is more promising to propose that various processes are able to alter each other. This type of interaction cannot give a complete account of neuroplasticity, however, for those animals that mediate feedback through subjective experience. Here we cannot use the analogy of multiple arms, since subjective experience is unified and continuous, at least in humans, and probably in other animals as well. If it is admitted that the mind (in the psychological order) in higher animals can rewire the brain (in the neurological order), then the mind cannot just be a reconceptualization of the brain, for that would make its activity simultaneous with, and indeed identical with, neurological activity. Yet the conscious psyche is experienced as a unity, so it is analogous to the single robotic arm, which cannot alter its own programming if execution is simultaneous with the presence of circuitry. If my mind has the power to re-wire the brain, it must do so in a way that does not eo ipso alter my subjective experience, except perhaps at some time in the future. If we admit that there must be some temporal delay between neurological states and psychological states in order to avoid causal paradoxes, then we have effectively admitted that psychological states are not just neurological states reconceptualized.
I should remind the reader, who may be tempted to reach into the vagaries of quantum mechanics for a more flexible concept of causality, that neurological activity occurs on a scale in which classical mechanical and electromagnetic principles apply. (Besides, the causal paradoxes of quantum mechanics are artifacts of the faulty Copenhagen interpretation, as I discuss in another work.) As for the less than linear causality of complex systems, I have given due consideration to this possibility by allowing for unpredictably complex feedback among multiple processes, but not in a way that positively contradicts the postulate of causality. Those who wish to abandon causality for the expedient of advocating a materialist mind (though the proof would remain incomplete, due to the ontological problem) should accept the full consequences of this denial. Without causality, there is no reason to demand scientific explanations for everything, and entire worlds of paranormal phenomena will now have their doors reopened.
Lastly, I should reiterate that I do not intend to deny that neurological processes play a necessary role in neuroplasticity. I only admonish against assuming that correlation equals causality or ontology. It has been proven, for example, that mRNA distribution is needed in order to generate proteins that trigger the extension of dendritic spines. Failure to distribute this mRNA results in mental retardation, due to fewer neuronal connections, and inhibited ability to learn. This does not mean that we learn by distributing mRNA and creating neuronal connections. Rather, this interconnected neural architecture is a necessary apparatus for learning, which may occur within this network by other means. To take a hardware-software analogy; I need to build an adequate circuit in order to run complex software. This does not mean that the building of the circuit is itself the execution of the software. I still have to develop and install the software, then run it.
Davies uses quantum mechanics as another putative example of level entanglement
or informational emergentism. This model depends on a peculiar interpretation of quantum mechanics where the particle aspect of a physical system is stuff
and the wave aspect is what we know
about the system. It is by no means obvious that this interpretation is correct, though at least, unlike many physicists, Davies recognizes that the wavefunction is not stuff.
If we assume, for the sake of argument, that this model is correct, then the wavefunction—what we know
—affects the behavior of particles or stuff.
Thus there are two levels, information and stuff, interacting causally.
Such analysis plays fast and loose with the concept of information,
for it is not clear if we are talking about a mere set of physical states (or possible physical states), or an actual human observer interpreting the system, or if non-human physical systems are capable of interpreting information.
This conflation of concepts becomes evident when Davies suggests that information
is the basis of statistical entropy, and that genes are repositories of information. This is to ascribe semantic intention to unintelligent objects, when in fact none of these things serve as information
in the sense of knowledge, except insofar as human beings impose interpretations on these things, making them serve as representations.
A nucleic acid does not contain information in the same way that our words convey meaning. To assert the contrary is to adopt a purely functionalist, input-output definition of knowledge. The nucleotides in a DNA strand act as coded information
only in the sense that one thing effectively acts as a surrogate for another thing or process that it causes to come into being. A sequence of nucleotides acts as code
in this limited sense for a certain protein, in very crude analogy with the way the utterance ‘dog’ represents a certain animal. Yet this is where the similarity ends. It is needless anthropomorphism to speak of a system interpreting
this code; rather, physical objects interact with each other out of blind physical necessity, making any semantic interpretation superfluous. As far as we know—though this is not certain—the ribosomes interpret
RNA through sheer chemical necessity.
It is meaningful to ask to what extent, if any, is the genetic code arbitrary,
meaning, could the correspondence between nucleotides and amino acids have been completely different? We do observe some limited variation in the mitochondrial genetic code, but it is not clear if this is simply related to a corresponding variation in organelle structure, which would make the so-called code
just as blindly deterministic, without any need for the superfluous, though intuitively convenient, concept of information.
As far as we know, cellular and subcellular biological components do not interpret
information in the sense of actually understanding anything or ascribing meaning to anything. Any attempt to reduce human or higher animal mental activity to the informational
processes of cells and molecules is nothing but a clumsily fallacious argument by metaphorical analogy. Since those who oppose a holistic or vitalistic concept of the mind are almost invariably materialists, they are hardly willing to ascribe mystically soul-like
understanding to our microscopic constituents. Rather, they would like to wave their hands and say that our functions of knowledge and attribution of meaning are nothing more than
the pseudo-informational processes found in physics and chemistry. They wish to diminish the former rather than exalt the latter, so that there is no soul anywhere.
The only problem with a denial of the soul or psyche’s existence is that is demonstrably wrong. Indeed, it is more certain that the soul exists than that matter exists, for we perceives psychic operations directly, while all our knowledge of matter comes, as it were, at second-hand through the soul. Our knowledge of corporeality, texture, color—considered qualitatively or quantitatively—indeed, even our knowledge of neuroscience and empirical psychology, all comes through the lens of conceptual understanding. We must know something interpretively, or not at all. This is the diametric opposite of the supposedly informational
processes found in nature, where it is not the least bit necessary for the decoding
system to conceptually interpret anything. It suffices to act like an unthinking drone, just following the orders of the laws of physics and immediate local conditions, without worrying about the big picture, or what the language
being parsed really means. I have proven, in my analysis of the one-bit adder, that the same computational
process can be given diverse interpretations or meanings. The only thing an electronic or chemical component knows
is what is in its immediate vicinity. Since it reacts to this knowledge with deterministic regularity (at least on the non-quantum scale with which we are concerned), the supposition that it truly knows even this miniscule amount is superfluous.
Let us suppose that the mind really is a higher-order phenomenon acting downward on its material constituents. If the mind’s activity is entirely reducible to the local physics of its constituents, it would indeed be completely analogous to the pseudo-informational processes discussed above, and so the act of interpreting and understanding conceptually would be just as superfluous. The intelligent mind would just be a helpless bystander that enjoys the delusion of being in control, when in fact it is just an observer of the outcomes of inexorably deterministic processes completely beyond its power, both prior and posterior to consciousness. If this is reality, then it is, to say the least, a most cunningly executed deception. We would remain absurdly at a loss to explain why conceptual consciousness is necessary, if it is powerless to break the cycle of determinism. No evolutionary explanation would make sense if consciousness does not affect biological outcomes. Much less could we explain why an elaborate system of deception would be biologically necessary, since even a mind aware of its impotence would be unable to change anything. Needless to say, the supposition that the mind is powerless to overcome biochemical determinism is an absurdity that only an ideologically driven intellectual could believe. The greater mass of humanity is helpfully inured from this error by a dose of earthy common sense.
As a matter of fact, the mind does have the power of its own conscious acts, and these are regularly translated into physical actions. It is true, as we have discussed, that unconscious neural activity plays a greater part in shaping our perceptions and executing our decisions than simple intuition would lead us to believe, but there is nothing that abolishes the sovereignty of the conscious willing agent.
There is no logical or scientific obstacle to believing that local biochemical forces can be subjected to global rules. None of our current techniques would be able to discern whether or not this is the case, due to the staggering complexity of even the simplest organisms. Of course, the physics and chemistry of inanimate objects can be explained in terms of local forces—though we have noted that they could also be explained non-locally—but this is not strong grounds for denying that there could be non-local harnessing of forces in biological organisms. Living things, at least from the amoeba onward, are markedly different from other types of physical objects in that they coordinate various diverse parts to a common end. This orchestration to a common end, which was traditionally called soul,
may vary in kind from one class of organism to the next. In the lowest forms of life, it may be little more than a set of physical constraints eliminating certain degrees of freedom.
In higher forms, there may be some active sharing of resources for the good of the whole rather than of a particular cell. When we move to animals with nervous systems, soul
activities are coordinated through the phenomenon of sensation, the subjective experience of which is still a mystery to us, as far as other animals are concerned. Increasingly complex animals may coordinate sensations, and at some point they might be able to have a mind of sorts that can prefer one thing over another subjectively, not just out of mechanical necessity. We see evidence of such appetition in frogs, which can choose to jump one way rather than another. In humans, we find that the various appetites can be subjected to rational or discursive understanding, which may follow or override these impulses as it sees fit.
Even if it is admitted that global rules of these varieties must exist in biological organisms, the question remains unanswered as to how they arise. Do the higher forms of soul arise simply out of increasing complexity? Clearly, at some point between proteins and humans, there must arise something genuinely new in order to break the deterministic loop. Ascertaining what this point must be, and whether there are other similar points of rupture, depends on our ability to accurately assess the psychological abilities of other animals. Alfred Russel Wallace, co-discoverer of the theory of evolution, had enough philosophical sense to recognize that there must still have been at least three special acts of creation. The first would have created life from non-life; the second would have created subjectively conscious animals; and the third would have created rational humans from appetitive brutes. This insight recognized that the coordinated behavior of biological organisms has no true analogue in the physics or chemistry of inorganic matter; that subjective consciousness is altogether categorically distinct from anything else in physics or chemistry, which indeed have no real way of describing it, being qualitatively barren; and that human conceptual understanding is quite irreducible to a mind that deals only in perceptibles. It is not necessary, however, to assume, that there were complete ruptures in evolutionary history, with entire organisms being created out of whole cloth. Rather, it is necessary only to affirm that a new physical principle had to be introduced at various points in natural history, or at a minimum, that a principle that had existed from the beginning only now was able to be employed, when the right material conditions were met.
The exact solution to the problem of how the mind emerged is necessarily speculative, but it is unfaithful to the facts of existence to pretend that higher biological faculties are reducible to mechanical principles that do not contain even their rudiments. Emergentism is not a panacea that can be applied willy-nilly to explain any complex or higher-order thing in terms of anything simpler. We must consider the features of the proposed constituents and local forces to determine if they are adequate to the task. The forces and objects of basic physics, as they are commonly known, are not sufficient to account for radically incorporeal phenomena such as the subjective experiences of sensation, appetition and volition. No appeal to complexity alone will make a deterministic thing become non-deterministic, nor will it make a world of unthinking corporeal objects a world of subjective sensation, feeling and thought. It makes no sense to deny the soul as magical while proposing a natural alchemy that is far more absurd. The objections against materialism are not simply that I, a fallible individual, cannot conceive how soul could arise from matter,
but rather that, when the principles and objects of materialism are clearly articulated, it is clear that the scope of their operation does not extend to the qualities of subjective sensation, perception, understanding, and volition, just as it is clear that no combination of vectors in a plane, however complex, would ever lead to a vector pointing outside the plane. It is not necessary to exhaustively specify the possibilities to apprehend this as clear as day.
It should not be a matter of controversy that the principles of materialism do not extend to the higher faculties of the soul, since materialists themselves are constantly trying to deny the higher faculties, explaining them as illusory or secretly deterministic, while deriding the subjective as imaginary or insubstantial. The fact remains that, even if my soul were an illusion, it is a clearly perceived illusion, and materialism would not be able to account for illusion, being qualitatively barren. As a practical matter, I would recommend treating the materialist precisely as the automaton he claims to be, and do not take his mindless utterances seriously or treat him with human or even animal dignity, since he is a mere collection of objects, while ‘human’ and ‘animal’ are just convenient labels for certain groups of molecules. Where reason is impotent, perhaps self-interest will persuade him to reconsider his foolish position.
Apart from the questions of the existence and origin of the soul, we might also inquire how the soul is able to actually operate. In the lowest organisms, the soul may be little more than a set of physical constraints. In organisms with enough of a central nervous system, there may be actual subjective perception and decision-making, the rudiments of an I,
though perhaps not a self-aware one. The nature of the interface between a mental soul and body is not clear, since we are not knowledgeable enough about how the brain works. It is not as if one neural signal mechanically triggers another signal in an adjacent neuron, and so on, indefinitely, in a deterministic serial circuit. There is massively parallel processing going on, and we cannot tell why some processes continue while others terminate, and others just spontaneously appear out of nowhere. How does each region know
when to start or stop working? This is similar to the equally mysterious problem of how a ribosome knows
when to read which part of an RNA strand. I will not presume to say that these processes are holistic simply because they are currently unknown. However, the temporal and spatial discontinuity of these processes strongly suggests that the mind does not interface with the brain at a single point, but permeates it, much as the soul was traditionally conceived as filling the entire body. This does not mean that the mind is stuff
in some other dimension, but rather it is the organizing principle that coordinates various processes in the brain, consciously or unconsciously. It is beyond our present scope to consider if there is more than one such principle in a given brain. In another work, I will postulate that only in human beings is the mind a substantive entity, not just an organizing principle or form of the material brain and its processes. This supposition will require metaphysical justification.
I am not sure that the techniques and epistemology of contemporary science are well equipped to determine the presence and activity of a holistic soul or mind as I have outlined it, even if such a thing truly exists. Empirical science restricts itself to what can be observed by the senses and measured. As long as we can mathematically analyze the quantitative phenomena, there is little interest in a metaphysical explanation of the thing-in-itself or noumenon. As long as there are no local violations of the laws of physics, there is little incentive to probe deeper.
Even if scientists were motivated to look for holism in psychology, it is doubtful that they could empirically demonstrate its existence or non-existence. The staggering number and rapidity of changes in the brain makes it quantitatively formidable, to say the least, to determine whether such activity is coordinated, if we could even define a basis for making such a judgment. Our best evidence for holism remains in subjective experience. I can speak and feel pain at the same time, though disparate regions of the brain are responsible for these faculties, and there is no single experience center.
Scientific investigations of psychology may make quantitative correlations between psychological states and neurological states, but correlation does not prove causality, and in large, complex systems with multiple overlapping processes, the directions of causality are practically impossible to discern.
Some aspects of psychology cannot be addressed by empirical neuroscience even in principle. When dealing with properly psychological phenomena on their own terms, we must turn to abstract ratiocination. You can only talk about the neural correlates of psychological phenomena for so long. At some point you must face the psyche itself, especially since we ourselves are psychological beings. The psyche qua psyche is not observable by any scientific instrument, which makes it properly outside the domain of empirical science. This is why pure psychology (i.e., psychotherapy and other non-neurological aspects of the science) has not had the empirical rigor of physical science, but has frequently involved the subjective rational judgments of conflicting schools of thought. We can observe
the psyche indirectly of course, by understanding what people mean when they say or do certain things. This requires interpretation, not mere sensory detection and quantification, which is the characteristic procedure of modern physical science, leading to regular and reliable results.
Although pure psychology loses the mathematical reliability of neurology, it is a necessary endeavor if we are to respect the fact that the thoughts of the psyche contain an internal causal logic of their own. I choose to do things based on what I experience psychologically. Neuroscience is useful in helping to understand how the objects of psychological experience (i.e., perceptibles) are created behind the scenes. Yet I am responsive to neurological phenomena only indirectly; I respond directly to what is experienced psychologically. If my brain is malfunctioning, and creating the wrong perceptibles, medical treatment may be appropriate. However, even if the brain is functioning normally, I may act unwisely in my response to certain psychic conditions or perceptibles. This is the domain of ethics and psychology proper. In a more purely rational treatment, we may consider the faculties of the soul and the proper use of these faculties, abstracted from determinate neurological conditions. Obviously, the presence of neurological maladies has some bearing on ethics, insofar as they can lessen culpability by preventing or limiting the free exercise of the will. Yet, even without neurology, I can generally tell the difference between whether I have willed something or if it is a spontaneous imagining. While neuroscience teaches us to at least be wary of trusting our introspective analysis, pure psychology offers us a domain in which to analyze our mental life as it is really experienced. Psychology, after all, is concerned with the experience of a subject. Any attempt to reduce the subject to an object would not be an authentic psychology, but a denial of the psyche.
© 2010 Daniel J. Castellano. All rights reserved. http://www.arcaneknowledge.org
| Home | Top |