Home > Investigating Skeptics > Examining Skeptics > Analysis of Russell

Analysis of Russell

by Ted Dace

Back to...Examining Skeptics

“It often turns out important to the progress of science,” says Bertrand Russell in The Analysis of Mind, “to remember hypotheses which have previously seemed improbable.”1

If only he’d been true to his word. On the brink of uncovering a genuinely scientific account of the mind, Russell cobbled together a straw-man substitute and promptly set it alight. His dismissal of “mnemic causation,” as he called it, was intended to ensure the continued prosperity of his favored belief system, the materialist-mathematical school of thought established by Galileo, Descartes and Newton.

Whether known as the Grand Doctrine, the Mechanical Philosophy, reductionism, materialism or Russell’s own “logical atomism,” the basic idea is that the world consists of simple, discrete entities that behave and combine according to timeless mathematical laws of nature. Reality is particle and law. All else is imaginary, a pointless if amusing dream. In the new intellectual climate, the job of philosophers, if they still have one, is to accept the atomized worldview without protest and investigate issues of human existence in light of it.

A lecture series published in 1921, Russell’s Analysis of Mind was geared around the proposal that ultimately the mind boils down to sense data. “All psychic phenomena are built up out of sensations and images alone,” he says.2 “Beliefs, desires, volitions, and so on” are nothing but “sensations and images variously interrelated.”3 Images may seem more mental than tangible, but according to Russell they “have a causal connection with physical objects, through the fact that they are copies of past sensations.”4 Images reduce to sensations, while sensations reduce to the meeting of the external world with nerve endings. From mind to matter in a few easy steps.

The chief threat to Russell’s scheme came from his arch rival, French philosopher Henri Bergson. In his 1911 book, Matter and Memory, Bergson asks why, if images are faded copies of old sensations, we never confuse the recollection of a loud noise with the sensation of a soft one.5 Unable to answer Bergson’s question, Russell can only observe that we have a “belief-feeling” that a remembered image relates to the past.6 On what basis do we arrive at this belief-feeling? In a world where all images arise from the current of consciousness, where do we get our sense of pastness? Russell cannot say.

The job of the brain, according to Bergson, is to calculate possible actions in response to sensory data.7Inputs are converted in the most efficient possible way to outputs. That’s all there is to it. Within those cerebral folds you will find no representations of the world, no emotions, no thoughts, no desires, no psyche. For Bergson, locating the qualities of mind in the brain amounts to a kind of neural mysticism. Is the brain so special that it can simultaneously be a part of the physical world and yet step outside it to represent it?8

Rather than constructing images of the world, says Bergson, our brains simply facilitate our perception of it. Because the brain does its job, we directly perceive what is around us. But how does Bergson grapple with memory? In this case, the images we perceive are no longer physically given. Surely here we must rely on cerebral storage of images.

Just as he maintains that we actually apprehend what is around us, Bergson argues that in memory we literally perceive the past. Far from merely representing the past, a memory is the resuscitation of a perception.9 To explain how this can be, Bergson must reinvent time itself.

“The past has not ceased to exist; it has only ceased to be useful.”10 Bergsonian time is not a series of discrete moments, each instant replacing the previous one, but an unbroken duration that conveys into our ever-shifting present all that has preceded it. “Our most distant past adheres to our present and constitutes with it a single and identical uninterrupted change.”11 What we call the “past” only appears to be past in the context of the spatial world of matter. As a material object, the brain can only mark the leading edge of the ever-expanding “moment” we call time. By contrast, the mind reflects time as it is, in which past (memory) adheres to presence (consciousness).

Russell regarded Bergson’s memory as a type of long-range causation. Since he saw causation of any kind as a perfectly impenetrable mystery, he conceded that the direct influence of distant past over ongoing present cannot be dismissed out of hand.12 If the present, by some inconceivable mechanism, is influenced by the immediate past, why not by the distant past as well?

Unable to pinpoint why Bergson’s proposal had to be wrong, yet firmly convinced that it was, Russell conjured mnemic causation, not quite what Bergson actually said but close enough that in refuting it, he would seem to have shaken off his nemesis without even mentioning him by name. Utilizing terminology developed by German zoologist Richard Semon (who was inspired by the Greek goddess of memory, Mnemosyne) Russell explains his idea:

Whenever the effect resulting from a stimulus to an organism differs according to the past history of the organism, without our being able actually to detect any relevant difference in its present structure, we will speak of “mnemic causation.”13

A child who has been burned, says Russell, reacts differently to fire than a child with no such experience. If the memory of being burned leaves no trace in the brain, but the child nonetheless reacts to fire in accord with prior experience, this indicates the direct influence of the remote past over the present with no material intermediary.14

By proposing that mnemic causation is demonstrated by the absence of evidence for a neural marker of past events, Russell rigged the results in advance. As we now know, and as Russell anticipated, the brain harbors “memory traces” correlated with past events. The standard materialist assumption is that these synaptic configurations, much like the “memory banks” of a computer hard drive, record the past.

A logician by training, Russell should have realized that mnemic influence would not necessarily imply an absence of “any relevant difference” in brain structure. This is simply the inverse of the fact that the brain’s obvious necessity for memory does not equate to sufficiency. Russell makes this point himself, observing that our dependence on brains for memory does not mean recollection is a strictly neural process or that memories are literally stored in brain tissue.15 Yet Russell failed to appreciate that the action of the distant past on the present, even if necessary to account for memory, in no way negates a role for the brain.

Russell’s plan seems to have been to dispose of Bergson’s past-within-a-present so as to arrive at Semon’s concept of the engram as the only possible explanation of memory. A kind of neural engraving, the engram is the change in the brain’s resting state following an event such as being burned. The engram guarantees that the child who has been burned will be more alert and therefore less likely to be burned again.16 Semon’s explicit denial that engrams could be regarded as “immaterial or metaphysical” must have been music to Russell’s positivistic ears.17 “Already existing engrams are never remoulded,” writes Semon in The Mneme, “but remain as they were first imprinted.”18 The engram’s defining trait, stability over time, not only accounts for memory but helps explain the general stability of the organism in the face of the dominant tendencies of transformation and evolution.19

Yet it’s precisely their stability that makes engrams wholly unlike anything neuroscientists have actually uncovered. Every time an individual remembers an event, the relevant memory trace loses its structure and must be “reconsolidated” from scratch. As John McCrone explains in New Scientist, “Resurrecting a memory trace appears to render it completely fluid, as pliable and unstable as the moment it was first formed, and in need of fixing once again into the brain’s circuitry.”20 If something interferes with reconsolidation, such as a high voltage current or a protein-blocking drug, the memory is lost forever. This finding, which has been consistently replicated, baffles researchers since it means that a memory, once recalled, is lost to the brain and must be re-established on the basis of nothing more than the actual recall, however cloudy, of the past event itself. For a time, just when we’re conscious of it, the memory exists yet is not reducible to information encoded in the brain. How can this be?

No matter how hard scientists try to impress memory into gray matter, it always pops back up, bobbing on the ethereal sea of mind. Rather than storing and retrieving information about the past, the brain seems only to facilitate the mental act of recollection. This accords with common sense, since, after all, the whole point of remembering something is that you don’t have to look it up. To regard the brain as a kind of organic reference library is to banish memory and replace it with mere information storage. That recollection may indeed revitalize past perception is only a problem for the materialistic outlook.

Oddly enough, Semon proposed another concept, known as mnemic homophony, that accounts for memory far better than the engram. Russell praises Semon for this ingenious idea, not for its application to memory and habit but its explanation of how the richness of experience is rendered into abstractions, a process that befuddled philosophers ranging from Berkeley to Hume.21

Semon compares the emergence of abstraction to the process of superimposing similar photographs. So long as they are close enough in form, a series of mental images generates a somewhat fuzzy general image when layered over each other. Each time you see a tree, for instance, it calls to mind all the other times you’ve seen one, and this new image is superimposed over the rest, producing a composite picture you think of as “tree.”22

Neither Russell nor Semon saw the contradiction between mnemic homophony and the engram. As material objects, engrams (or memory traces) cannot simply blend into one another to form a generalized engram. The fact that successive perceptions may coexist in a single, generalized perception does not give the brain the power to construct generalized bits of matter.

Semon’s unwillingness to accept the irretrievably immaterial nature of mnemic homophony is ironic given that physics, the foundation of natural science, had ceased to be materialistic by the early 20th century. What draws a pair of magnets to each other is not a pair of invisible hands but a magnetic field. The moon remains in the sky because the earth generates a gravitational field that warps the space-time through which the moon travels. In both cases, a force of nature allows objects to act at a distance on other objects. Perhaps organic events can be influenced, at a distance over time, by sufficiently similar previous events.

Mnemic homophony gives us memory and habit without the need for permanent neural engravings. Semon always thinks of Capri when he smells a particular cooking oil because this oil happened to be wafting over from a nearby restaurant as he gazed upon the island across the Bay of Naples.23 No information storage is required, only the principle that any given mental state is influenced by a similar previous state. Rather than recording the past, memory traces simply enable us to recall it, much as radio antennae facilitate the long-range influence of electromagnetic waves.

As to that special type of memory known as habit, Semon illustrates the role of mnemic homophony with a game of fetch. Each time you cock your arm, the dog understands that you’re going to throw the stick. Even if you don’t actually toss it but only pretend to do so, the dog will chase the chimera because your gesture has awakened its memory of when you actually did throw it. Of course, this works only so many times. Before long, the dog refuses to run until it has perfect homophony between the new stimulus and the old stimulus, that is, when it actually sees the stick emerge from your hand. Habits are activated by mnemic homophony, whether rough or perfect, between current and past circumstances.24

This line of reasoning would have served Semon well in his discussion of the inheritance of acquired characteristics. Like Darwin before him, Semon regarded evolution as unworkable without the ability of organisms to pass on newly developed instincts and bodily forms.25 Otherwise, adaptation to changing environments would play no role in evolution, leaving us with a lifeless mechanism of chance mutation and natural selection to explain the world’s dazzling array of exquisitely adapted creatures. The evidence produced by Darwin’s successors, though never refuted, was dismissed and forgotten. Why look at evidence for a phenomenon which, from a genetic standpoint, is impossible?

Yet the evidence is extensive. Semon reports on salamanders coaxed into either holding their young in utero longer than usual or releasing them early, in both cases their young carrying on the newly-altered behavior.26 He reports on trees transplanted from temperate to tropical regions and vice versa, either way their new adaptations continuing to crop up even in progeny restored to their normal habitat.27 Echoing Darwin’s observations on farm animals, he reports on praying mantis populations becoming more tame with each generation in captivity despite the complete absence of selection for this trait.28

Theorist August Weismann tried to refute claims such as these - - and there were many, many more - - by cutting off the tails of 1500 mice and observing that all their progeny still grew tails as usual.29 Yet experiments demonstrating inheritance of acquired traits succeeded precisely because researchers induced organisms to make the changes themselves, just as the environment prods organisms into adapting rather than mechanically imposing new behaviors. This distinction was lost on Weismann, who saw organisms as nothing more than automatons to be reconfigured like watches or water pumps.

Semon believed that engrams migrate to germ cells in order to propagate newly developed habits and forms. If it’s true, however, that such traits cannot influence genes, then we must consider the possibility of an immaterial mechanism of transmission. If mnemic homophony can connect past and present within a single organism’s lifespan, why not across many lifespans?

Russell never mentions mnemic homophony by name. Nor does he mention Semon’s insistence that evolution is unworkable without inheritance of acquired traits. You would never suspect, reading Russell, that Semon denies the reduction of memory to a machine-like process. While Russell claims that an organism is no more purposive than a river flowing to the sea, Semon notes, in his “law of ecphory,” that unlike machinery, which requires a complete input to produce a complete output, a memory can be fully realized even when the trigger, such as the smell of cooking oil, contains only a hint of the original event.30 Semon seems to have resembled Bergson more than he was willing to admit. But Russell was too committed to establishing Semon’s materialist credentials to notice where he and Bergson overlapped.

A half century after Russell’s investigation, the task of synthesizing Semon and Bergson fell to a young biologist-in-training at Cambridge University, a theoretical nonconformist who took a year off from his laboratory work to study philosophy at Harvard. Unlike Russell, whose reading of Bergson was colored by professional rivalry, Rupert Sheldrake was captivated by Bergson’s radical take on time and its implications for memory. By coupling Bergson’s time-as-duration with Semon’s mnemic homophony, Sheldrake obtained the basis for a scientific theory of mind, the very thing Russell had sought with his Analysis of Mind.

Designed to explain organic development from egg to maturity, Sheldrake’s theory of morphic resonance is based on his Bergsonian reading of Semon. Where “mnemic” emphasizes the emergence of organic form as a memory-based process, Sheldrake’s use of “morphic” turns it the other way round, highlighting the proposition that organic memory operates on the basis of form. The more similar - - or homophonous or resonant - - a current organic form is to a previous form, the more likely it will be influenced by that form. It was Sheldrake who realized that if the mnemic principle applies within a single lifespan, it ought to apply between current and past organisms as well. Thus human embryos develop along the same lines as previous human embryos, while chimpanzee embryos develop the same way as previous chimpanzees, and so on. Sheldrake also realized that this principle, if correct, would apply not just to the organism as a whole but to all levels of structure comprising it. Every organ, every tissue, every cell is busy reproducing the actions it undertook in previous, similar situations.

Morphic resonance is most clearly revealed wherever successive generations of a given species become more adept at a given task without receiving any guidance from their parents. The best-documented spontaneous case of this kind concerns birds that learned to pop open milk bottles in Western Europe. The technique was first observed in 1921 in Southampton, England among blue tits and spread primarily through simple imitation. Since blue tits rarely travel more than a few miles, it’s unlikely that imitation could account for the appearance of this habit in Sweden, Denmark and Holland. “The Dutch records are particularly interesting,” writes Sheldrake.

Milk bottles practically disappeared during the war, and became reasonably common again only in 1947 or 1948. Few if any tits that had learned the habit before the war could have survived to this date, but nevertheless attacks on bottles began again rapidly.31

Of course, it’s always possible that postwar birds simply learned the process again from scratch. For a really compelling demonstration of the direct influence of past behavior over current behavior, we must turn to controlled laboratory conditions. The best known example is William McDougall’s experiment on learning in rats. One of many scientists around the turn of the 20th century to have demonstrated the inheritance of acquired traits, McDougall placed rats in a water maze and found that each generation solved the maze more quickly than the previous generation. Like Semon, he assumed that the animals’ genes were somehow incorporating and transmitting the acquired ability. But when the experiment was replicated, first in England and then Australia with rats unrelated to McDougall’s, the tendency for improvement continued as before, an outcome totally inexplicable except in light of species memory via morphic resonance.32

A few tests on human subjects have also revealed long-range memory. In one such test, non-Japanese speakers were better able to memorize authentic Japanese nursery rhymes than rearranged, nonsensical versions.33 According to Sheldrake, this result follows from the fact that untold millions of people have already learned the rhymes, and anyone trying to memorize the correct versions is influenced by their cumulative experience. When subjects of another experiment were shown Persian words for ten seconds, some real and some only Persian-like fakes, and then asked to reproduce the words from memory, they fared far better at reproducing the real words.34

Flabbergasted by Sheldrake’s audacious proposal, neuroscientist Steven Rose designed an experiment that would surely dispose of it once and for all. The experiment involved day-old chicks divided into two groups. Test chicks could peck at yellow diodes, while control chicks could peck at chrome beads. After pecking, the test chicks were injected with lithium chloride, a toxic substance that made them mildly nauseous, while control chicks were injected with a harmless saline solution. The same procedure was followed for 37 days with a new batch of chicks each day. The data indicated that successive batches of test chicks became gradually more hesitant to peck relative to control chicks. While this finding indicated that test chicks were influenced by previous test chicks, the most clear-cut result concerned control chicks that were allowed to peck at either the yellow diodes or the chrome beads three hours following their injection of saline solution. Over the course of the experiment, successive batches of control chicks became increasingly reluctant to peck at the yellow diodes, indicating that they were influenced by the cumulative experience of chicks that had pecked at the yellow diodes and then been injected with lithium chloride. After stalling for months, Rose reneged on his agreement to write up the results with Sheldrake for publication.35

Needless to say, a handful of anecdotes and mostly unrepeated experiments falls far short of proof. While interesting, Sheldrake’s theory remains largely untested. The important thing, however, is not that it hasn’t been properly and repeatedly put to the test but that the theory is indeed testable and falsifiable. The same cannot be said of the quaint notion that DNA is a kind of blueprint or program of the developing organism.

Around the time he was mutilating mice in a misguided effort to disprove inheritance of living adaptations, August Weismann proposed that organisms develop from the egg on the basis of information conveyed from parents in the form of “determinants” (now known as genes). But the very gains in molecular biology that subsequently fleshed out this theory would ultimately abolish it.

A theory is scientific insofar as it reduces a complex phenomenon, such as the organization of a living body, to something simple, such as the information stored in its DNA. At the core of Weismann’s proposal was the assumption that genes are relatively simple, static units of information. In particular, the genes associated with development must vary from species to species, reflecting changes in form over evolution. Neither he nor any of his intellectual descendants expected that developmental genes, known as homeobox or “hox” genes, would be virtually identical in species ranging from insects to people. What changes in the course of evolution is not the genes themselves but the regulatory DNA that switches them on and off to ensure that development is species-appropriate.

Usually adjacent to the hox genes they regulate, epigenetic tags or “switches” operate at blinding speed. According to molecular biologist Sean Carroll, typical developmental processes involve “tens of thousands of switches being thrown in sequence and in parallel.”36 The operation of switches is so complex that they can be analyzed only with combinatorial logic. “Because the combination of inputs determines the output of a switch, and the potential combinations of inputs increase exponentially with each additional input, the potential outputs of switches are virtually endless.”37 Every switch position and associated pattern of protein production is but a snapshot, a single frame in “one hell of a movie with nonstop action.”38

Imagine a forest overflowing with lightning bugs, except this forest is actually produced by the incomprehensibly complex and ever changing patterns of lightning bug flashes. Altering this pattern alters the shape of the forest. This, according to molecular biology, is basically how our bodies develop.

Whether we’re looking at the organization of proteins in a cell or neurons in the brain, the number of possible configurations is virtually infinite, preventing successful physical analysis. Genes were supposed to be the exception. Genes were something we could bring within our orbit of comprehension. Now we find that the computation of genetic activity also escalates to infinite complexity, leaving us with the absurdity of reducing one complexity to another. That certain genes can be correlated with certain traits and illnesses indicates only that genes play a role in development, not that development is reducible to genetic information.

Weismann’s proposal may have captivated generations of biologists - - and it may even be true - - but it’s not a scientific theory. Morphic resonance, on the other hand, reduces the body’s infinite complexity to the perfect simplicity of Semon’s principle that “like causes produce like effects.”39 With the loss of the DNA-based theory of development, morphic resonance is the only game in town.

Given that we don’t seem to be machines, it’s odd that Russell had such faith in the reduction of organism to mechanized assemblage of atoms. The most compelling data in opposition to this belief are generated daily by that ongoing half-baked experiment we call life. Unlike materialism, the mnemic theory makes room for the mind as a thing-in-itself, the seat of our self-existence and autonomy as living beings. We appear to be thinking, feeling, freely acting people - - and not genetically programmed organic robots - - because we are in fact people leading meaningful human lives. Atoms, by contrast, do not lead meaningful lives, no matter how they’re assembled.

Weismann, Semon and Russell clung to materialism like children to their mother. Bergson and Sheldrake, on the other hand, recognized that it’s precisely against matter that memory is defined. With memory freed from the smothering embrace of matter, mind is at last made sensible.

So long as it’s restricted to the brain, the mind can be dismissed as mere shadow play. But extended throughout the body, it finds its home. By reproducing prior behavior on the basis of similar context, the brain is no different from any other organ. Mentality is associated with every organ in the body, guiding its development and maintaining its form via resonance with similar, past form.

“Mind and body” is more phrase than reality. We have two words for the same thing because we see body-mind from two perspectives, one in terms of space and the other in terms of time. As body is the spatialized surface of mind, the mind is the temporal depths of the body. Accordingly, death is where the body loses its mind, where matter and memory cease to be united.

What the ancients called soul or spirit has been translated in the modern mind as the immaterial element of life. But we don’t have to define organic memory in the negative, any more than body-mind must be defined as the unconscious. The immaterial element is simply the influence of the deep past on the present. Those actions undertaken in situations most resembling the current situation are the ones most likely to materialize.

Blinded by his need to verify the reduction of the world to tangible matter and timeless law, Russell missed the message of the mind, which is neither one nor the other. In the end, he got it wrong because he just had to be right.

1. Russell, Bertrand, The Analysis of Mind, London: George Allen & Unwin, 1921, p. 92
2. Ibid, p. 279
3. Ibid, p. 300
4. Ibid, p. 110
5. Bergson, Henri, Matter and Memory, London: Swan Sonnenschein, 1911, pp. 318-319
6. Russell, p. 159
7. Bergson, p. 20
8. Ibid, p. 11
9. Ibid, p. 179
10. Ibid, p. 193
11. Bergson, Henri, The Creative Mind, New York: Philosophical Library, 1946, pp. 180-181
12. Russell, p. 89
13. Ibid, p. 86
14. Ibid, p. 77
15. Ibid, p. 91
16. Ibid, pp. 79-83
17. Semon, Richard, The Mneme, London: George Allen & Unwin, 1921, p. 275
18. Ibid, p. 240
19. Ibid, p. 14
20. McCrone, John, 'Not-so total recall', New Scientist, May 3, 2003, p. 27
21. Russell, pp. 218-219
22. Semon, p. 164
23. Ibid, p. 92
24. Ibid, p. 156
25. Ibid, p. 290
26. Ibid, pp. 58-60
27. Ibid, p. 64
28. Ibid, p. 133
29. http://en.wikipedia.org/wiki/August_Weismann
30. Semon, pp. 273, 124
31. Sheldrake, Rupert, The Presence of the Past, New York: Times Books, 1988, p. 178
32. Ibid, p. 175
33. Ibid, p. 189-190
34. Ibid, p. 192
35. Sheldrake, Rupert, 'An Experimental Test of the Hypothesis of Formative Causation', Rivista di Biologia - Biology Forum 86 (3/4), 1992, 431-44 http://www.sheldrake.org/Articles&Papers/papers/morphic/formative.html
36. Carroll, Sean, Endless Forms Most Beautiful, New York: W.W. Norton and Company, 2005, p. 114
37. Ibid, p. 124
38. Ibid, p. 128
39. Semon, p. 173
Top of Page

Back to...Examining Skeptics



If you have any comments or suggestions on this website please email... morphlist@aol.com
Copyright © The Association for Skeptical Investigations