Evolutionary Epistemology, Rationality, and the Sociology of Knowledge
Gerard Radnitzky and W. W. Bartley, III
With contributions by Sir Karl Popper, Donald T. Campbell, W. W. Bartley, III, Gunter Wachtershauser, Rosaria Egidi, Gerhard Vollmer, John F. Post, John Watkins, Gerard Radnitzky, Peter Munz, and Antony Flew
Open Court, La Salle, Illinois
Introduction By Gerard Radnitzky and W. W. Bartley, III
PART I: EVOLUTIONARY EPISTEMOLOGY
Chapter I. Philosophy of Biology versus Philosophy of Physics By W. W. Bartley, III
Chapter II. Evolutionary Epistemology By Donald T. Campbell
Chapter III. Blind Variation and Selective Retention in Creative Thought as in Other Knowledge
Processes By Donald T. Campbell
Chapter IV Campbell on the Evolutionary Theory of Knowledge By Sir Karl Popper
Chapter V. Light and Life: On the Nutritional Origins of Sensory Perception By Gunter Wachtershauser
Chapter VI. Natural Selection and the Emergence of Mind By Sir Karl Popper
Chapter VII. Emergence, Reduction, and Evolutionary Epistemology By Rosaria Egidi
Chapter VIII. On Supposed Circularities in an Empirically Oriented Epistemology By Gerhard Vollmer
Chapter IX. Theories of Rationality By W. W. Bartley, III
Chapter X. The Possible Liar By John F. Post
Chapter XI. Paradox in Critical Rationalism and Related Theories By John F. Post
Chapter XII. A Godelian Theorem for Theories of Rationality By John F. Post
Chapter XIII. Comprehensively Critical Rationalism: a Retrospect By John Watkins
Chapter XIV. In Defense of Self-Applicable Critical Rationalism By Gerard Radnitzky
Chapter XV. A Refutation of the Alleged Refutation of Comprehensively Critical Rationalism By W. W. Bartley, III
Chapter XVI. Philosophy and the Mirror of Rorty By Peter Munz
Chapter XVII. Must Naturalism Discredit Naturalism? By Antony Flew
Chapter XVIII. Alienation Alienated: The Economics of Knowledge versus the Psychology and Sociology of Knowledge By W. W. Bartley, III
Light and Life: On the Nutritional Origins of Sensory Perception
By Gunter Wachtershauser
Information does not stream into us from the environment. Rather, it is we whoexplore the environment and suck information from it actively, like food.
Popper's philosophy marks a major achievement: the first and only unified theory of knowledge. One single coherent process of knowledge, a problem-solving process, is seen as stretching from the earliest inklings of life to the latest advances of science and technology.3 We are concerned here with a special aspect of this lifelong process: the evolutionary origins of perception, i.e., the transition from long-term problem solving by biological evolution to short-term problem solving by sensory cognition.
How do we proceed in such an inquiry? Generally speaking, the course of evolution may be unravelled in two directions, forward in time or backward in time. In the first procedure, we follow Popper's method of situational logic 4 We try to spell out a most plausible initial situation and analyze the problems it poses for some organisms. Next, we try to construct a possible solution to this problem situation guided by considerations of physical, chemical, and ecological plausibility. This will lead to a new problem situation for which we try again to construct a solution. In this fashion, we attempt a theoretical reconstruction of the course of evolution in the form of a speculative concatenation of problems and solutions.' In the second procedure, we study the features of modern organisms as they exist today and we try to establish phylogenetic relationships by similarity considerations. The first approach will be taken in section B and the second approach in section C, and we shall see that both approaches lead to compatible results.
Many evolutionary studies focus primarily on the evolution of whole organisms or species. Our study of the origin of perception will focus by contrast on the evolution of features or, more precisely, of biochemical features. The first focus suggests the image of a tree of evolution which grows strictly by a branching process, by bifurcations. It breaks down in cases of symbiosis or endosymbiosis. The focus on biochemical features raises a distinctly different image. It is the image of a flow of evolution where patterns of branching are mixed with patterns of confluence, reaching from the combination of simple chemical compounds to the symbiosis of whole organisms. New features with novel types of functions arise by a combination of previously disconnected features. As an example for such confluence and as a central result of our inquiry, we see the origin of vision as a joint venture of an active locomotion machine with light-receiving pigments derived from the photosynthesis apparatus. Vision began as an active search for nutritive light.
B. The Co-evolution of Photosynthesis and Vision a Speculative Tale
1.The Origin of Active Locomotion
Once upon a time, some four billion years ago, the earth became inhabited by tiny unicellular organisms, protobacteria. They lived in primeval waters containing, as their only source of food, a large variety of abiotically generated organic substances. Wherever this source of food was available, the organisms thrived and multiplied rapidly, which in turn led to food depletion. To some extent these local food shortage problems were alleviated by passive movement resulting from water currents. But eventually an exploratory food searching behavior by active locomotion emerged. This must have occurred by a multitude of small steps, which we shall now attempt to reconstruct.
Our speculation begins with the assumption that in primitive organisms the rate of diffusion of nutrients into the cell and waste products out of the cell was a factor limiting the population. Since diffusion velocity increases with relative movement between the cell and the surrounding fluid, a flow of liquid could facilitate nutrient intake. However, in the absence of any flow turbulence (i.e., in laminar flow) diffusion is severely limited by a stagnating surface layer. If this stagnating layer is broken up by turbulence, the diffusion speed greatly increases. Now it cell wall could induce such turbulence. Therefore, they would have had a strong selective advantage. As a next step, an internally generated active, but still uncoordinated movement of these fibrils might have occurred, further promoting turbulence and thus diffusion. Next, by the coordination of movements of many fibrils (i.e., cilia), and with it a further enhancement of diffusion; an active cell locomotion could have emerged. This cell locomotion must at first have been random. But later, in a most decisive moment in evolution, this random and inefficient locomotion may have turned into controlled locomotion by coupling with
already existent nutrient receptors in the cell wall (chemoreceptors). The identity between the chemical receptors for nutritives and the signal detectors for locomotion control would ensure that only edible nutrients were searched for. If so, this may have been the earliest inkling of a form of perception (see Section E below). It was fully in service of food acquisition and it arose by a joint venture of active locomotion and detection of nutrients. This active and controlled locomotion was one of the two tributaries in the flow of evolution which led to the formation of visual perception. The other tributary we shall turn to now.
2.Photosynthesis and Vision under a Friendly Sun
Eventually, the prebiotic broth, a consomme of sorts, became globally depleted. The rate at which it was replenished by the abiotic formation of organic substances became the limiting factor of the sustainable biomass. Life was restricted by the scarcity of food. This global food shortage gave rise to what is one of the most important innovations in the history of life: photosynthesis. It enabled its "inventors" to utilize light for the internal biosynthesis of essential organic compounds as a supplement to their ingestion from external sources and it eventually led to a global changeover of the fuel base of life. In this early encounter with life, light functioned solely as food.
The photosynthesis apparatus must at first have been rather inefficient compared with modern green plants, wasting most of the received light. Thus, the amount of light was a limiting factor. Moreover, lighting conditions must have always been locally different and temporally changing. For the organisms this posed the problem of finding and staying in a spot of optimum radiation. It was solved by the emergence of light-searching behavior, a light-controlled locomotion (photo movement). This occurred again by a joint venture, a coupling of the already existent active locomotion machine with light reception by the photosynthesis apparatus. It could have come about comparatively easily, since active locomotion needs a fuel and since photosynthesis provides just this fuel. Later the control chain was shortened by the transduction of control signals from the photosynthesis pigments to the locomotion machine 6 In either case, the primary light-absorbing pigments of bacterial photosynthesis were also used for photomovement. This had an enormous advantage: the type of light searched by photomovement was automatically "edible" light useful for photosynthesis?
Thus, bacterial photomovement, a vision of sorts, is connected with bacterial photosynthesis in a double sense. Biochemically speaking it is a derivative of photosynthesis and functionally speaking it is ancillary to photosynthesis. The search for light began not as a quest for information but as a quest for food.
3. Photosynthesis and Vision under a Hostile Sun
So far, we have only looked at the friendly side of light. But solar radiation also has a hostile, life-destroying aspect due to its ultraviolet components. Let us now unravel that part of the story.
At the beginning of life, our earth was covered by an atmosphere which was utterly different from that of today. It was devoid of any oxygen and as a consequence it did not have a protective ozone layer. Thus, ultraviolet radiation was not absorbed but freely penetrated the atmosphere. Physicists tell us that the sun at that time must have been cooler and hence its radiation less intensive and somewhat shifted away from the ultraviolet range compared to present solar light. But even so, the ultraviolet radiation impinging on the surface of the earth must have been quite intense.
If we now consider that DNA is rapidly destroyed by ultraviolet light, notably near its absorption maximum of 265 nm, we are driven to the question: how could life have ever started under a blazing sun? The answer might be simple: it didn't. According to Hans Kuhn$ the earliest organisms may have dwelt within tiny pores of submersed sand, which were a substitute for cell walls and which fully sheltered the organisms from light. Later, their quest for food drove some of them to dare the open waters. We know they survived, but why? Water absorbs solar radiation, and it does so particularly in the ultraviolet range. Thus, ultraviolet light decreases with increasing depth of water, until it finally vanishes. Therefore, given sufficient depth, it was possible for our ancestral microbes to dare the open yet shun the light.
By doing so, they made an important discovery. Water absorbs unevenly. It changes not only the intensity of light, but also its spectral distribution. The exact nature of these changes depends on the substances that are dissolved or suspended in natural waters. In the turbid waters of Burly Griffin Lake, Australia, the maximum of light transmission is in the red range near 700 nm, while in the clearest waters of the Sargasso Sea it is in the blue range near 470 mm.? Yet, all these waters have one thing in common: as we probe deeper, the toxic ultraviolet radiation decreases faster than nutritive light. For the emergence of photosynthesis, this was decisive. There was always a zone with the necessary protection from ultraviolet rays yet with sufficient exposure to "edible" light. The early photosynthesizers survived in a narrow niche between the devil and the deep blue sea.
Subsequent evolution was marked by a number of important inventions which made bacteria somewhat radiation-resistant and allowed for greater light exposure. The emergence of mechanisms for DNA repair permitted the healing of radiation damage. Protective pigments were incorporated into the cell wall for shielding the sensitive nucleic acids from radiation attacks by converting harmful radiation into harmless heat. It is quite plausible that some simple protective pigments were the precursors of the pigments for photosynthesis and that, to some extent, the evolution of photosynthesis may have been a co-evolution of photoprotective and photoactive pigment functions.
But aside from these somewhat passive strategies, an active protective behavior emerged. We can easily see that a positive photomovement toward ever greater light intensities, while beneficial for photosynthesis, would have been devastating for the organisms. It would have driven an organism to the surface of the water and thus to certain death. Therefore, photomovement came under a dual control enabling not only a search for "nutritive" light but also a flight from "toxic" ultraviolet radiation. With this earliest form of color vision, the quest for light as food became selective.10
4. The Advent of Oxygen or the Blue-green Revolution
We will now turn to the formation of the oxygen atmosphere and its significance for the co-evolution of photosynthesis and vision.
Perhaps the most ancient form of photosynthesis is found in purple bacteria, which produce organic compounds by hydrogenating carbon dioxide (CO2). As hydrogen source they use hydrogen sulfide (H2S), whereby elementary sulfur (S) is formed as a waste product:
2H2S + CO2 by (CH2O); + H2O + 2S.
The requisite life-sustaining hydrogen sulfide, though abundantly contained in the exhalations of the earth, must have been restricted to limited regions, since it was precipitated by metal ions (e.g., Fe++) which acted as H2S scavengers. Thus, the first photosynthesizers were precariously restricted in a double sense: not only to light-exposed yet sufficiently deep zones of water, but also to regions rich in H2 S.
The solution to this problem of local restriction came with the emergence of a novel type of photosynthesis of far-reaching consequences: a modification of the bacterial photosynthesis apparatus in blue-green algae (cyanobacteria) which allowed for a change of the hydrogen source: a substitution of water for hydrogen sulfide (H2S):
H2O + CO2 by (CH2O),z + 02.
Now life was able to colonize the vast expanses of the oceans. For a long time, the waste product of this form of photosynthesis, molecular oxygen, was eliminated by the oxidation of large amounts of reducing metal ions mainly Fe++ Fe+++) But some 1.7 billion years ago this oxygen sink became exhausted and free molecular oxygen started slowly to accumulate, at first in the waters and later in the atmosphere. A global environmental changeover from a reducing to an oxydizing atmosphere ensued. The consequences were enormous. Most forms of anaerobic organisms became extinct by oxygen-poisoning. They were replaced by a multitude of novel oxygen-tolerating or oxygen-dependent (aerobic) organisms.
More importantly, the atmospheric oxygen gave rise to the formation of a layer of ozone high up in the atmosphere. Ozone absorbs ultraviolet light and its absorption maximum happens to coincide with the absorption maximum of DNA. Thus, it absorbs most strongly those portions of the spectrum which cells cannot tolerate. The ozone layer is an effective filter against the most destructive ultraviolet radiation. Now, by virtue of their light-seeking and ultraviolet-fleeing capabilities, the photosynthesizing organisms could rise higher and higher toward ever greater light intensities and eventually to the surface of the water and even onto land. And so it came to pass that life made the earth globally inhabitable for life.
5. The Parasitic Origin of Animal Vision
The earth was now widely colonized by a microbial ecosystem. The food chains were short: a variety of light-searching photosynthesizers (phototrophs) as primary producers, which in turn served as prey for bacteria-devouring (heterotrophic) organisms (phagocytosis). It is widely believed that in this ecosystem endosymbiosis was a frequent occurrence. A photosynthesizing bacterium, after having been ingested, was not digested but rather maintained as an internal cell organelle, a chloroplast. It was kept in a state of slavery, forced to share its photosynthetic products with its master cell. And we may further speculate that it was also forced to lend its light-searching capability for piloting the endosymbiotic organism into areas of optimum radiation. This endosymbiosis, in conjunction with a number of other poorly understood yet thoroughgoing innovations, gave rise to what may be viewed as the eukaryotic (cells with a nucleus) pro-genitors of all higher multicellular forms of life, plant as well as animal. They may be pictured as freely swimming, vision-guided, partly phototrophic and partly heterotrophic organisms. Some of them gave up their animal-like feeding habits and specialized into full-fledged freely swimming unicellular algae, wholly dependent on photosynthesis and light-searching capabilities. Later, when these algae turned into multicellular rooted plants, such vision-guided locomotion of whole organisms was abandoned. But photosynthesis was maintained and perfected.
Others specialized into algae-hunting protozoa by abandoning photosynthesis but maintaining vision-guided locomotion as a way of tracking down algae. These primitive animals did not hunt by directly detecting their prey, but rather by indirectly searching for well- lit areas, using light as a clue for the occurrence of algae. From this earliest form of animal vision, a most simple capacity for discriminating light intensities, all higher forms of animal eyes may be assumed to have arisen. Animal eyes are the offspring of a contraband, stolen from photosynthesizing bacteria through an act of endosymbiosis.
To make a long story short, over billions of years of microbial evolution, the interaction of life and light was marked by a peculiar coevolution of photosynthesis and vision: the photosynthetic machinery served as the biochemical precursor for the formation of a visual apparatus, which in turn served for finding the light for photosynthesis. Early vision bestowed its beneficial effects on the productivity of the very apparatus whence it arose.
C. Confrontation with Some Facts of Biology and Biochemistry
1. The Photochemical Unity of Animal Vision
So far, we have traced the evolution of photosynthesis and vision by moving forward in time, beginning from a fictitious early starting point and reconstructing a sequence of concatenated problems and solutions. We shall now go backward in time, on the basis of a comparative study of extant organisms. We shall see that both accounts are compatible.
Today, photoreceptor organs are found in most (metazoan) animal groups ranging from the simplest light-sensitive cells of hydrozoa to sophisticated eyes in annelida, arthropoda, mollusca, and vertebrata. Admittedly, there are considerable morphological differences, which led L. von Salvini-Plawen and Ernst Mayr to postulate 40 to 65 independent origins of animal eyes (11). But these forms of vision seem to be based on one single identical type of visual pigment (rhodopsin): a vitamin A aldehyde (retinal) connected to a membrane protein (opsin). And they are all based on the same primary photochemical process: a light-induced cis-trans-isomerization of the chromophore retinal:
George Wald, the Nobel-prizewinning pioneer in visual biochemistry, has attributed this astonishing photochemical unity of vision to an extreme case of an evolutionary convergence of biochemicals:
Organisms, under the unremitting pressures of natural selection, have no choice but to rediscover again and again ... the same molecular structure.... If one asks how it comes about that certain molluscs, arthropods and vertebrates agree in possessing vitamin A, the answer is not that these animals are related phylogenetically, but that all of them have eyes .12
This view seems hardly plausible, and most recently it has been clearly refuted by two independent groups,13 who found a considerable degree of homology between the genes coding for the visual opsins of Drosophila melanogaster and several mammals, i.e., two animal groups which have been separated by over 500 million years of evolution. It may well be expected that throughout the animal kingdom there is also a homology in the enzymes for producing retinal.
It should be mentioned that in two younger branches of the animal tree, retinal derivatives figure in place of retinal. 3,4-Dehydroretinal is used by some fishes and amphibians14 and another derivative, 3-hydroxyretinal, is used by diptera (e.g., flies) and lepidoptera (e.g., butterflies).15 Both derivatives emerged as relatively late extensions of the retinal-producing biosynthetic pathway.
Thus, contrary to Wald's opinion, the evolution of the pigments of animal (metazoan) vision appears to be not polyphyletic and convergent, but monophyletic and divergent, with a singular, one-time "invention" of retinal and opsin at its root.
2. The Algal Connection of Animal Vision
Our story suggests the hypothesis that there must be a direct biochemical connection between animal vision and algal photomovement. The first important evidence of this kind has most recently been produced. It appears that retinal, the universal chromophore of animal vision, is also used as the chromophore of photomovement by the unicellular green alga Chlamydomonas reinhardtii.16 For even stronger evidence we might predict some sequence homology in the genes of animals and chlamydomonads coding for opsin and for the enzyme of retinal biosynthesis. With evidence of this kind coming in, our assumption that animal vision and algal photomovement have been inherited from a common ancestral organism would become compelling.
What could this common ancestral organism have been like? In our story, we adopted the view that it was an endosymbiotic organism consisting of a host cell and a photosynthesizing bacterium as endosymbiont, which later turned into the chloroplasts of the green algae such as Chlamydomonas. The recently discovered phototrophic oxygenic Prochlorophyta17 give considerable support to this view. They have the same photosynthesis pigments as the chloroplasts of green algae and may therefore be considered as closest living relatives of the bacterial endosymbiont.(18) But our account goes beyond this conventional view. It makes the additional claim that the bacterial endosymbiont was also exploited as a photodetector for piloting its host organism and that it is the precursor of algal and animal eyes. That is to say a retinal-based photomovement should have existed already on the level of the phototrophic bacteria prior to endosymbiosis. Perhaps the most recently discovered freely swimming Prochlorophyt (19) might prove to show such a photomovement based on retinal as chromophore. (20)
While such direct evidence for the endosymbiotic origin of vision is not yet available, it should be mentioned that a spectacular analogous case has recently been discovered in Paramecium bursaria. This is a highly derived protozoon, which feeds on algae. It leaves some of its prey (Chlorella) undigested and forces it to share its photosynthetic products with its host and, more importantly, it has been found that the endosymbiotic Chlorella is also employed as a pilot for the lightseeking movement of its host 21 Here a tiny plant organism serves literally as the eye of a unicellular protozoon.
Biologists have widely accepted Euglena as the most likely link between unicellular algae and unicellular animals. If this opinion should prove correct, our speculation on the origin of animal vision would be refuted, since the photomovement of Euglena seems not to be based on retinal, but rather on a compound totally unrelated to retinal, a flavine22 Happily, most recent findings seem to indicate that the chloroplasts of Euglena have triple membranes. This means that they could not have arisen by an early endosymbiosis of a bacterium, which could have only given rise to a double membrane. Rather they must have arisen by a more recent endosymbiosis of a eukaryotic alga within a unicellular protozoon 23 Thus, Euglena can be ruled out as a candidate for the primitive ancestor of animals. It is highly derived rather than primitive, and it must have arisen long after the emergence of unicellular algae and animals.
3.The Biochemical Connection between Pigments of Photosynthesis and Animal Vision
Further, our story suggests a biochemical connection or homology between the pigments of photosynthesis and the pigment of animal vision. However, a quick glance at the formulae of retinal, the chromophore of animal vision, and chlorophyll, the main pigment of photosynthesis, seems to belie any such connection. The structural differences are enormous and the biosynthetic- pathways leading to both compounds are unrelated. But chlorophyll is not the only pigment of photosynthesis. Rather, it becomes increasingly evident that 0-carotene plays an important role not only as a light-harvesting pigment in the antenna complexes but also as a component of the very reaction center of photosynthesis.24 Now it is important to note that animals produce their retinal by oxidative cleavage of 13-carotene.
But animals cannot synthesize the carotene starting material. Rather, they ingest it with their food. Its ultimate sources are plants and photosynthesizing bacteria, which produce it in a lengthy, multistage biochemical pathway. It appears plausible that the eukaryotic ancestor of both animals and algae once had the full capacity for the biosynthesis of 0-carotene and for its conversion to retinal and that in animals, feeding on carotene-rich bacteria or plants, the biosynthesis of carotene became unnecessary and was abandoned.
We can go a little further and look at the conditions for the first emergence of a retinal-based visual system. The multistage biosynthesis of 0-carotene constitutes a heavy investment for the cell. Its product, however, pays full dividends due to its important role in photosynthesis. Now, the biosynthesis of retinal as a terminal extension of the carotene pathway consumes 0-carotene and thus weakens the photosynthesis apparatus. ^ Further, it is plausible that in the earliest stages of the emerging retinal biosynthesis, the function of retinal must have been rather inefficient. Therefore, any selective advantage conferred by such inefficient retinal function could be expected as having been small and certainly offset by the selective disadvantage of the loss of precious 0-carotene. How then could the evolution of a retinal-based photomovement have ever gained momentum? The answer may perhaps lie in an amplifying feedback effect of sorts.
Retinal, right from its first invention, functioned within an apparatus for seeking light. More light meant more photosynthesis and thus higher returns for the investments into the carotene biosynthesis. Thus, the overall effect even of a poor form of early photomovement on photosynthesis was positive. Early retinal-based photomovement could emerge precisely, because it bestowed its beneficial effects on the productivity of the very photosynthesis apparatus whence it arose.
It might be speculated that this dual connection between algal photosynthesis and photomovement is but an example of a more general principle of innovative evolution. In the field of biochemical evolution, this principle of dual connectivity" (chemical and functional) may hold whenever a new biochemical with a novel type of function arises as a chemical derivative of a biosynthetic precursor with a different function, and when the precursor is not abundantly available but is rather itself the precious product of a lengthy biosynthetic pathway. Under these conditions, an overall selective advantage exists only if the novel function is ancillary to the function of the precursor.
4.The Generalized Connection between Photosynthesis and Photocontrol Pigments
So far we have looked at the connection between retinal-based vision and photosynthesis. We will now turn to some other forms of photocontrol. As a consequence of the general principle suggested at the end of the preceding section, it might be expected that any kind of photocontrol pigment may be found to be biochemically related to some kind of photosynthesis pigment. What are the facts?
Let us first consider the case of the halobacteria. They exhibit a light-seeking and ultraviolet-fleeing photomovement25 and also a photosynthesis of sorts (a light-driven proton pump for ATP-synthesis),26 and both are based on pigments consisting of protein-bonded retinal. Thus, the photomovement-photosynthesisconnection is here most evident. The occurrence of retinal in the pigments of halobacteria might suggest even a connection with animal vision. It may be speculated that the halobacteria and the original host cell of the eukaryotes derive from one common ancestor, which was the sole one-time "inventor" of retinal. Some simple facts of biochemistry belie such speculation.
There are three kingdoms of life: eukaryotes, eubacteria, and archaebacteria. Halobacteria belong to the kingdom of archaebacteria. Now, the split between the host cell of the eukaryotes and the archaebacteria must have occurred very early in the evolution of life and certainly long before the change-over to the oxygen atmosphere. This is clearly evidenced by the fact that the ribosomes and other old and most conservative components of the cell are widely different between, yet highly homologous within these two kingdoms of organisms?' On the other hand, the chemical conversion of $-carotene to retinal requires free molecular oxygen.28 Halobacteria are aerobic and they require an external supply of oxygen for producing retinal. Thus, it appears that they could not have acquired a capacity for the biosynthesis of retinal before the advent of an aerobic environment, i.e., before
much more than some 1.7 billion years ago. Therefore, retinal appears to have been acquired twice and independently, once within the archaebacterial line of the halobacteria and once within the eubacterial line of the phototrophic bacterial precursors of the chloroplasts of the eukaryotes. The latter could have "invented" retinal before the advent of an aerobic environment since they produce molecular oxygen internally by their oxygenic photosynthesis (see Section C.2 on Prochlorophyta). This view on the halobacteria is compatible with the lack of sequence homology between mammal and halobacterial opsins.29
Another kind of photomovement is found in Stentor coeruleus, a ciliar protozoon. Its pigment is a protein-bonded quinon, stentorin3° Recently it was announced by Song that the same organism shows light-driven bursts of ATP formation (a photosynthesis of sorts) which seems to be based on the same or a similar pigment as the photomovement 31
As mentioned earlier (section C.2) the photomovement of Euglena32 and some other protists (and also the phototropism of the fungus Phycomyces33) seem to be based on a protein-bonded flavine. Here the biochemical connection with photosynthesis is very weak. Flavoproteins occur merely in the electron transport chain of chlorophyll-based photosynthesis. There is no evidence yet for a form of photosynthesis based on a flavine as a light-absorbing pigment.
Let us sum up the situation: in its long encounter with light, life seems to have "discovered" the nutritional value of light on several independent occasions. Photosynthesis is polyphyletic. And the plural types of photosynthesis pigments gave rise in turn to different types of photocontrol or photomovement pigments. Animal vision is situated on just one of these lines of evolution.
D. The Narrow Band of Visible Light-A Test for Explanatory Power
1. The Window of the Atmosphere-a Case of Make-believe Adaptation
We will now show how our account of the origin of vision can explain the visible light range. We will begin with a criticism of the conventional explanation.
All human and animal vision is limited to a tiny band of visible light. Our eyes do not pick up waves above about 700 nm (infrared waves, microwaves, radiowaves) and below 300 to 400 nm (ultraviolet waves, x-rays, morays). As Bartley put it so succinctly: "We live in an electromagnetic sea, as it were, and nonetheless these waves do not register unassisted on our eyes".34 How can we explain this fact? Let us first consider that the primary reaction of vision must be a photochemical reaction. Photochemistry certainly does not occur throughout the full range of the electromagnetic spectrum, but its range is still as wide as from 100 to 1400 nm.35 Thus, our problem remains, albeit in a narrower form: why is the range of visible light much smaller than the range of photochemistry?
A widely popularized explanation assumes that the atmosphere has a window of light transmission in the visual range, while it absorbs all the other radiation and that the visual pigments simply adapted to the type of radiation transmitted by this window.36 However, the proponents of this "window theory" make a cardinal mistake. Instead of adapting the theory of adaptation to the facts, they seem to adapt the curve of light distribution to their theory of adaptation.
Most spectral curves of light drawn in support of the "window theory" show the radiation energy plotted against the wavelength, and they exhibit a pronounced peak in the visible range. Yet eyes are not light energy meters but rather photon counters 37 Now, if we convert the published energy distribution curves to photon density curves (the energy of a photon decreases with increasing wavelength by the formula: e = he/X) the radiation peak loses its convincing shape. The infrared range becomes more pronounced, and the radiation maximum is shifted somewhat towards longer wavelengths. In fact, many published curves, when properly converted, show about as many photons in the range of 700 to 1000 nm as in the range of 400 to 700 nm38 Thus, the "window theory" is undermined rather than supported by such curves.
But the case of the "window theory" is weaker still. In aqueous habitats, where vision first evolved, the light distribution changes with the type and concentration of impurities and with the depth of water.39 If we probe deeper and deeper into a body of clear ocean water, the light range becomes increasingly restricted to a narrow band of blue light. Thus, it is admittedly always possible to single out a special depth in a special body of water in order to make the available light range coincide with the visible light range and to claim that animal vision must have started there. But the theory of evolutionary adaptation is a theory for explaining evolutionary change by a change in the environment. Now animals, by rising to the surface of the water, by crawling onto land and by spreading into a diversity of habitats with demonstrably different ranges of available light (e.g., dense forests vs. open plains; low vs. high altitudes; overcast vs. clear skies; polar vs. equatorial zones) did greatly change their photic environment, but their range of vision hardly changed and there is no convincing correlation between their photic environments and their range of vision. Thus, an adaptation to the available light, while it might have been of benefit, apparently did not occur to any significant extent 40
2. The Coincidence between the Bands of Vision and Photosynthesis
We will now show how the range of visible light can be explained by the origin of vision in photosynthesis. We will proceed in two stages. First, we will try to give an explanation of the range of photosynthesis and then we will explain the range of vision with reference to this range of photosynthesis.
Photosynthesis is an energy-consuming chemical reaction. It is driven by the absorption of light energy. Light consists of photons, and the energy of a single photon decreases with increasing wavelength. For each elementary photochemical reaction the absorption of a single photon is required 41 If the energy of a photon is high enough, it will satisfy the energy demand of the elementary photochemical reaction and the reaction will take place. On the other hand, if the nergy of a photon is below a certain threshold value, it fails to cause the photochemical reaction.
The oxygen-producing (oxygenic) photosynthesis of bacteria, algae, and plants is based on the abstraction of hydrogen from water. Hence it requires a high amount of energy, and the cutoff point is as low as 700 nm 42 The anoxygenic photosynthesis of purple bacteria is based on the abstraction of hydrogen from hydrogen sulfide. This requires much less energy. Therefore, some purple bacteria are photosynthetically active with wavelengths up to a threshold value of 1100 nm.43 Thus, it is possible in principle to explain the cutoff points of the ranges of photosynthesis toward longer wavelengths strictly by reference to considerations of chemical energy demands .44 There is no need to resort to a process of an adaptation to the photic environment.
The cutoff point of the ranges of photosynthesis toward short wavelengths (350 to 400 nm) is explainable by the toxic effects of the ultraviolet radiation. Thus, photosynthesis is limited to a range of non-toxic and nutritive light. Radiation below this range is poisonous and radiation above this range is not nourishing.
We now turn to the range of vision. There is a peculiar coincidence between the ranges of vision and of oxygenic photosynthesis. Both extend from about 400 to about 700 nm. What does this coincidence mean? According to our story, retinalbased animal vision derived from retinal-based photomovement of algae, which served for finding the light for their (oxygenic) photosynthesis. A photomovement responsive to light above 700 nm or below 400 nm would have been disastrously misleading by maneuvering the organism into non-nourishing or toxic areas. Thus, the restriction of a retinal-based photomovement to the range of 400 to 700 nm had a high selective advantage. Later, when the algal photomovement apparatus was exploited in animal vision for tracking down algae, the responsive range of 400 to 700 nm was still beneficial and simply retained. Thus, the visual range is readily explainable by its. historic origin in photosynthesis and no similarly plausible explanation and certainly not an explanation by adaptation to the photic environment seems to be in sight.
Much later, when animals started to use their sense of vision for many purposes other than algae hunting, adherence to the photosynthesis range of 400 to 700 nm was no longer necessary. Thus, the proponents of the "window theory" should expect an adaptive extension towards the infrared range with its abundance of photons to have taken place. But this did not happen. Instead, in many insects45 and also in some fishes,46 the visual range was extended into the low intensity ultraviolet range and hence in a direction which, in the light of the "window theory", would appear to be counter-adaptive rather than adaptive.
3. The Campbell Coincidence
Donald T. Campbell has discovered a most interesting coincidence: Animal locomotion is blocked by all solid bodies but not by air and water. Similarly the flux of visible light is obstructed by most solid bodies, but again not by air and water. According to Campbell, vision evolved largely by the exploitation of this coincidence as a vicarious (or indirect) substitute for direct exploration by locomotion. In short, vision substitutes for collision?'
We shall discuss two aspects of Campbell's coincidence. First, we shall showthat it cannot explain the range of visible light. Then, we shall discuss the evolutionary connection with a photosynthesis-derived sense of vision.
Ultraviolet radiation as well as near-infrared radiation are transmitted by air and, to a considerable extent, also by water, while they are obstructed by solid bodies such as stones. Thus, the (in)transparency-(im)penetrability coincidence is not exclusive to the band of visible light. It rather extends far into the ultraviolet and infrared ranges. It follows that the restriction of visible light to the narrow band of 400 to 700 nm cannot be explained as an adaptation to the range of Campbell's coincidence. Further, it should be considered that collision prevention requires some capacity for contrast vision, i.e., a higher visual function compared to simple light detection. Thus it is plausible that the range of vision had been established long before eyes for collision prevention emerged.
Our evolutionary story suggests that vision started as a direct process of seeking light for photosynthesis which later turned into an indirect process for the detection of light as a vicarious clue for tracking down algae. How was it possible that, from such simple beginnings, all the sophisticated contrivances of animal eyes could have emerged? Campbell's insight seems to go a long way toward answering this question. The evolution of the higher functions of animal vision might largely be explained as the result of a windfall profit. The range of vision, once established for photosynthesis purposes, happened to fall within the range of radiation in which Campbell's (in)transparency- (im)penetrability coincidence exists, and animal eyes could exploit this coincidence in a long evolution of contrast detection. Thus, our speculation on the origin of vision in photosynthesis and Campbell's coincidence are complementary rather than conflicting.
E. There is More to Vision than Meets the Eye
In the preceding sections, we have inquired into the transition from long-term problem solving by biological evolution to short-term problem solving by sensory cognition. We have located the origin of sensory perception in foraging for food and, more precisely, in a joint venture of a preexisting active locomotion machine with the preexisting receptors for nutritive chemicals or nutritive light. Thus, right from the start, sensory perception is a composite process, arising from two component processes: active locomotion and food absorption. Neither of these two components alone amounts to cognition. But with the conjunction of both comes something radically new: a sensory knowledge acquisition process. In this process, the two contributory components can still be detected, albeit with transformed functions. Active locomotion can now be interpreted as a creative or generative part leading to ever new positional variations, while food reception can be interpreted as a critical or evaluative part for the assessment of each new positional variation. Sensory knowledge acquisition satisfies from the start the Popperian formula41 of an interplay of creation and criticism.
A model that seems to bear witness to this earliest form of sensory knowledge acquisition is seen in the photomovement of extant purple bacteria 49 Here, an internally programmed active locomotion leads to ever new and spontaneous changes in position, while light reception serves strictly on the critical or evaluative side by the detection and elimination of bad moves (see also section B.2 and footnote 6). Our story suggests that a continuous, albeit lengthy line of evolution stretches from these lowly beginnings to animal cognition and to man's highest cognitive achievements. By extending Crick's principle of continuity,50 we may further speculate that, notwithstanding numerous later sophistications, the two basic components of sensory cognition were always maintained: the active and creative component, stretching from the active mind back to active locomotion, and the critical and evaluative sensory component, stretching back to the pigments of photosynthesis and other food receptors. Hence, in principle, the evolution of such a composite cognitive process can be traced back to the confluence of two non-cognitive processes without violating the principle of continuity.
Our evolutionary account agrees nicely with Popper's claim that all processes of short-term problem solving by sensory knowledge acquisition must necessarily be active processes and that passive knowledge acquisition does not exist 51 Indeed, the alternative notion of knowledge acquisition by a passive flow of bits of information from the environment through the interface of the outer sensoryperiphery into the organism would drive us to an absurd position. We can see this if we attempt to trace the notion of a passive sense data flow back through evolution. In this exercise, more and more sophistications will be seen as falling by the wayside, with the alleged flow of sense data becoming ever simpler and shorter until finally all that is left as a most primitive sense datum is the photoexcitation of a photosynthesis pigment molecule. And so it is that the passive theory
forces us into the absurd position of interpreting the primary photochemical reaction of photosynthesis as a cognitive process. Thus it appears that the theory of passive knowledge acquisition produces severe difficulties if we try to accommodate it within an evolutionary account. No such difficulties in tracing cognitive processes back to non-cognitive origins arise for the notion of a cognitive process as being active and composite from the start.
The cognitive apparatus for performing short-term problem solving through sensory cognition arose by a process of long-term problem solving through biological evolution. Most biologists interpret Darwinian evolution as a process of passive adaptations to the environment. However, in Popper's theory of active Darwinism, the activity of the organisms is seen as the principal driving force of evolutionary change 12
Our account of the nutritional origins of the sensory cognitive apparatus seems to favour Popper's interpretation. In section D, we have seen that one of the main characteristics of visual perception, the restriction of the visible light range, can hardly be understood as a passive adaptation to the environment. It rather seems to be the result of internal biochemical conditions of the organism and its foraging activities within the environment. The situation may be summarized as follows: Perception is not a process of passive acquisition of information from the environment by an apparatus which itself is the result of passive adaptation to this same environment. It is rather a process of active foraging within the environment by means of an apparatus which in its major characteristics is shaped by the organism's own foraging activities. As Popper would have it: organisms by being active seekers are the active makers of their senses.
This essay is a revised and expanded version of a lecture given May 27, 1984, during the 150th National Meeting of the American Association for the Advancement of Science in a Symposium on "New Directions in Evolutionary Epistemology" arranged by Paul Levinson. Copyright © 1984, 1985, Gunter Wachtershauser.
1. The title alludes to a famous lecture by Niels Bohr, delivered at the opening meeting of the
International Congress on Light Therapy, Copenhagen, on August 15, 1932, and printed in Nature 131, 1933, pp. 421-23, 457-59.
2. Karl R. Popper, in "Evolutionary Epistemology", cited below (footnote 3).
3. Karl R. Popper, Objective Knowledge: An Evolutionary Approach (London, Oxford University
Press, 1972); Karl R. Popper, The Open Universe: An Argument for Indeterminism, W. W. Bartley III.
ed. (London: Hutchinson Ltd., 1982), pp 131-62; Karl R. Popper, John C. Eccles, The Self and its
Brain, (Berlin: Springer Verlag, 1977); Karl R. Popper, Unended Quest, (London: Fontana/Collins,
1976); Karl R. Popper, "Evolutionary Epistemology"; in J. W. Pollard ed., Evolutionary Theory: Paths
into the Future, (Chichester: Wiley & Sons, 1984); see also Chapters IV and VI and for further discussions chapters I and II of this volume.
4. Karl R. Popper, Objective Knowledge, pp. 170-90.
5. Excellent examples of this method are found in Hans Kuhn, Jurg Waser, "Selbstorganisation
der Materie and Evolution fruher Formen des Lebens", in W. Hoppe et al., eds., Biophysik (Berlin:
Springer Verlag, 1982), pp. 860 -907, and in Wolfgang F. Gutman, Klaus Bonik, Kritische Evohotiotstheorie (Hildesheim: Gerstenberg Verlag, 1981).
6. Examples may be found in the photokinetic reactions of purple bacteria and cyanobacteria,
and in the phototaxic and photophobic reactions of some cyanobacteria. See Manfred Tevini, Donat-
Peter Hader, Allgemeine Photobiologie (Stuttgart: Georg Thieme Verlag, 1985), pp. 276-78.
7. Jerome A. Schiff, Biosystems 14, 1981, p. 129.
8. Hans Kuhn, Angewandte Chen:i 84, 1972, p. 838.
9. Claus Buschmann, Karl Grumbach, Physiologie der Photosynthese (Berlin: Springer Verlag,
1984), p. 184.
10. An example is found in the photomovement of halobacteria (see section C.4 and footnote 21).
11. L. von Salvini-Plawen, Ernst Mayr, "On the evolution of photoreceptors and eyes", in Evolutionary
Biology 10, 1977, pp. 206-63.
12. George Wald, "Phylogeny and Ontogeny at the Molecular Level", in A. I. Oparin, ed., Evolutionary
Biochemistry (New York: Pergamon Press, 1963), p. 19.
13. Joseph E. O'Tousa, Wolfgang Baehr, Richard L. Martin, Jay Hirsh, William L. Pak and
Meredith L. Applebury, Cell 40, 1985, pp. 839-50; Charles S. Zucker, Alan F. Cowman and Gerald
M. Rubin, Cell 40, 1985, pp. 851-58.
14. George Wald, Nature 139, 1937, p. 1017.
15. K. Vogt, K. Kirschfeld, Naturwissenschaften (1984), pp. 211-13.
16. K. W. Foster, J. Saranak, N. Patel, G. Zarilli, M. Okabe, T. Kline and K. Nakanishi, Nature
311, 1984, pp. 756-59.
17. R. A. Lewin, Nature, 261, 1976, pp. 697-8.
18. Prochlorophyta contain chlorophyll a and chlorophyll b just like the chloroplasts of green algae.
19. T. Burger-Wiersma, M. Veenhuis, J. J. Korthals, C. C. M. Van de Wiel, L. R. Mur, Nature,
320, 1986, pp. 262-4.
20. Cyanobacteria, which owing to their lack of chlorophyll b are not the likely ancestors of the
chloroplasts of green algae, utilize for their photomovement photosynthesis pigments (phycobilins
and carotenoids) (see the literature reference in footnote 6).
21. D. Neiss, W. Reisser and W. Wiesser, Planta 152, 1981, pp. 268-71.
22. Hader Tevini, Allgemeine Photobiologie, p. 277.
23. L. W. Wilcox, G. J. Wedemayer, Science 227, 1985, pp. 192-94.
24. R. J. Cogdell and J. Valentine, Photochemistry and Photobiology, 38, 1983, pp. 769-72.
25. L. Spudich and R. A. Bogomolni, Nature 312, 1984, p. 509.
26. D. Oesterhelt, W. Stoeckenius, Proceedings of the National Academy of Science, US 70, 1970,
27. C. R. Woese, L. J. Magrum, G. E. Fox, Journal of Molecular Evolution 11, 1978, pp. 245-52.
28. J. A. Olson and O. Hayaistis, Proceedings of the National Academy of Science US 54, 1965,
p. 1364; D. S. Goodman, H. S. Huang, M. Kanai and T. Shiratori, JBC, 242, 1967, p. 3543.
29. Yu. A. Ovchinnikov, FEBS Letters 148, 1982, 179-91.
30. P. S. Song, E. B. Walker and M. J. Yoon, in F. Lenci and G. Colombetti, eds., Photoreception
and Sensory Transduction in Aneural Organisms (New York: Plenum Press, 1980), pp. 241-52.
31. Announced by P. S. Song in the Meeting of the American Society for Photobiology, New
Orleans, June 1985, and reported in The Tinies-Picayune/The States -Item, New Orleans, June 27, 1985, .
32. G. Colombetti and F. Lenci, eds., Photoreception and Sensory Transduction in Aneural Organisms
(New York: Plenum Press, 1980), pp. 172-88.
33. M. K. Otter, M. Jayaram, R. M. Hamilton and M. Delbruck, Proceedings of the National
Academy of Science US 78, 1981, pp. 266-69.
34. W. W. Bartley, III, Chapter I of this volume.
35. Ultraviolet radiation below 100 nm leads to destructive ionization. Infrared and microwaves
above 1400 nm generate merely heat through vibrational and rotational excitations. The range of 100
to 1400 nm is the range of electron excitations, which lead to highly specific chemical reactions. This
is the range of photochemistry.
36. G. Vollmer, Evolutiondre Erkenntnistheorie (Stuttgart: S. Hirzel, 1981), p. 98; H. von Ditfurth,
Im Anfang war der Wasserstoff (Hamburg: Hoffman and Campe, 1972), p. 100.
37. Barbara Sakitt, journal of Physiology 223, 1972, pp. 131-50.
38. H. H. Seliger, Environmental Photobiology, in K. C. Smith, ed., The Science of Photobiology
(New York: Plenum, 1977), pp. 143-73. U. Kull, Evolution (Stuttgart: J. B. Metzlersche Verlagsbuchhandlung,
1977), p. 76, shows a photon density curve with substantially more photons in the range
of 700 to 1000 nm than in the range of 400 to 700 nm. Strangely, the same author has adopted the
39. See section B.3 and footnote 9.
40. 3-Dehydroretinal (see section C.1 and footnote 14) has an absorption maximum shifted by
some 20 nm toward longer wavelengths. But the pattern of its distribution seems to be so erratic that
no convincing correlation with the photic environment can be found. (F. Crescitelli, "The vertebrate
visual pigments", in H. Gutfreund, ed., Biochemical Evolution (Cambridge: Cambridge University Press,
1981), p. 346.)
41. Albert Einstein, "Uber einen die Erzepgung and Verwandlung des Lichts betreffenden
heuristischen Gesichtspunkt", Annalen der Physik 4, Folge 17, 1905, pp. 132-48.
42. The energy demand of this reaction is so high that it cannot be met by one photon alone.
It requires two photons in a series connection of two elementary photochemical reactions. Anoxygenic
photosynthesis requires excitation by only one photon.
43. E.g., Rhodopseudomonas viridis (N. Pfennig, Annual Review of Microbiology 21, 1967, p. 285).
44. See also E. Strasburger, Lehrbuch der Botanik (Stuttgart: Gustav Fisher Verlag, 1983), p. 250,
and J. Schiff, Biosysrems, p. 141.
45. See section C.1 and footnote 15.
46. Chr. Neumeyer, "An ultraviolet receptor as a fourth receptor type in Goldfish color vision",
Naturwissenschaften 72, 1985, pp. 162-63.
47. See Chapters II and III of this volume.
48. Karl R. Popper, Die beiden Grundprobleme der Erkenntnistheorie, (Tubingen: J. C. B. Mohr
(Paul Siebeck), 1979), based on manuscripts from 1930 to 1933, see notably pp. 19-32; Karl R. Popper, Logik der Forschung, (Vienna: Julius Springer, 1934); Karl R. Popper, Conjectures and Refutations, (London: Routledge & Kegan Paul, 1963); see also the references in footnote 3.
49. E. Strasburger, Lehrbuch der Botanik, p. 448.
50. F. H. C. Crick, 1. Mol. Biol., 38, 1968, p. 367.
51. See the references in footnote 3.
52. Karl R. Popper, Auf der Suche nach einer besseren Welt, (Munich: Piper, 1984), pp. 11-40,
Karl R. Popper, "The Place of Mind in Nature", in Richard Q. Elvee ed., Mind in Nature, (San Francisco: Harper & Row, 1982), pp. 31-59; and Objective Knowledge, pp. 256-84.