Report from an Academy 2: Charlatans, Failures, and Frauds

by Paul North

Royal_Academy_of_Music _London_W1With the previous post I began a new series, cognominated "Reports from an Academy." The reports are pure fiction. Imagine, if you would, a middle-aged professor of literature at an elite institution, call it Nevahwen University. Each month, worrying that it will all be over soon, a slipshod, unsystematic professor still in his good years, or so they tell him, pens a bulletin to the outside, without much hope anybody will receive it. Nevahwen is the school's name as well as its motto. Will the conflict between understanding the world and becoming successful in the world finally be decided in favor of understanding? This is the professor's eternal question. He asks: when? Wen? In a cynical mood, his answer is: Nevah. Nevahwen. Postcards from the front lines of a battle over the future—at the clash site of two generations, two or more economic classes, and in the midst of these conflicts, tiptoeing along a schism between the value of understanding and the value of success—postcards from the schism, these reports portray experiences that may be hard for those at home to fathom. Take them as proof the professor is still alive. Take them as recognitions of failure or declarations of hope. Take them as you will.

Charlatans, Failures, and Frauds

These three undesirables have one thing in common, they don't live up to expectations. What seems full of promise turns out to be empty. The product is different than you anticipated, worse than you wanted; promises turn out to be either fast talk, impotence, or lies. Charlatans, failures, and frauds: we want none of them in this Academy. Or, more precisely, we accept a certain amount of failure, but only if it is limited in scope, and it is really only tolerable if failure points the way to success. We must always learn from our mistakes. We must always learn from our mistakes.

Today I report on a point of indistinction. There is a point, hard to reach, even harder to recognize, when the three—charlatan, failure, fraud—become virtually indistinguishable. At this dicey point, someone presents an experiment, a historical thesis, a speculative proposal and neither you nor anyone else can tell whether the hypothesis is trumped up, whether the scientist or researcher has talked themselves into something they will later repudiate and regret, whether the world is simply not as they say and the thesis is flat wrong, or whether they are lying to themselves and as a consequence to others. It is a moment of high dubiousness. Then again, it is also a moment of possibility, where something unexpected could happen. We are not accusing our colleagues in the profession of anything. In fact, they show the utmost in professionalism, thoroughness of research, methodological rigor, and integrity. Yet there used to be a type…

There used to be a type, a type that lived up to the moniker "profess-or." Researchers of this type professed to know while not yet quite knowing for sure and tried then to help the world to conform to their somewhat idle talk. Concentrate on your picture of the "charlatan." We will look at one in particular, who was in some ways inside the Academy, and in other ways very much at its outer limits, a charlatan, if you want to call him that, who still, 100 years later, sits right in that dubious zone, whose results are equal parts highly improbable and full of possibility. Paul Kammerer, the experimental Kammerer2 biologist, from about 1904 to 1926, when he committed suicide, had the intention to prove the Lamarckian theory of evolution, or, from a different perspective, he accidentally discovered that Lamarckianism was true. We can't say which. To this day, Kammerer's story causes academics to condemn him once and for all as a fraud (this happens every ten years or so), to defend him as a simpleminded, misguided failure, or, much more strangely, to revive his peculiar scientific positions on the basis of new theories and new evidence (this also happens about once a decade). Now, it's true, knowledge produced in universities is supposed to be credible. Mostly it is credible because it departs only a very little from known knowledge. It builds little annexes onto others' work, slowly, carefully, making minor revisions, taking two steps forward and one back, adding, after long deliberation and much agreement, a single new brick onto the skyscraper. Even where speculation is allowed, it is never allowed in the empirical and experimental fields. In philosophy or religion someone can make wild claims, "we are all figures in someone else's dream," or "everything is connected," or some such thing. Do this in sociology or chemistry and be drummed out, and rightly. And yet, there was a time…

There was a time, a time when the university (not the universe) was young—the modern university—still arguing over which disciplines of study should be supported, when a newer subdiscipline and method like evolutionary biology, being only about 40 years old around 1900 when Kammerer got his training, was less than thoroughly well-established as a theory and a branch of learning and a university department.

Kammerer was good at breeding amphibians, an Aquarianer, as they called these semi-professional experimentalists. Evolutionary theory to its shame had never been able directly to see natural selection. The process could only be inferred. An Aquarianer like Kammerer, with success breeding finicky animals in the "vivarium," could claim the attention of leading scientists, and this Kammerer did, William Bateson for example, one of the early "synthesizers" of Darwin's theory together with the inheritance theory of Gregor Mendel, whose attention Kammerer got and his disapproval as well. With "newts, lizards and toads" Kammerer was a genius. So Arthur Koestler remarks in the 1971 book that attempted to vindicate the charlatan. So often these amphibians could only be successfully bred for study of heritable traits by Kammerer himself. Koestler made Kammerer his cause célèbre in The Case of the Midwife Toad, which defended the final set of experiments, which caused the most controversy of his controversy-filled career. In the experiments carried out on toads, Kammerer demonstrated conclusively the inheritance of acquired traits—a theory that had been the mark of charlatanary for at least the previous twenty years, although Darwin himself had, early on, accepted parts of it.

It was discovered evidence had been faked. Ink had been injected into the pads of the toads, making it look as though an acquired trait had been inherited. A lab assistant did it, Kammerer insisted, but his estrangement from the warm center of the discipline, which did not begin with his suicide, ended with it. Koestler describes the life of the researcher. True, there were celebrity elements. Alma Mahler became Kammerer's assistant for a time, studying the molting habits of the praying mantis. On the norm, however, the work was long and isolating. Fifteen years Kammerer devoted day in and out to experiments like these, burning into his consciousness the one after another of mating, egg laying, larval stage, and maturation into adulthood. What Kammerer discovered, or so he thought and then aimed to demonstrate, or, as some think, what he already believed and then set out to prove, twisting the data to support it, was that traits acquired during a lifetime could be passed down to offspring. This is the Lamarckian thesis, which was at the time and is still today in great disrepute. Much more interesting than the results of his experiments and the theory they supposedly support is, for us, the desires and habits that developed along this type of life. The life pattern that Kammerer more or less fell into was very unlike the life patterns of his Viennese cultural consorts like Alma Mahler, the conductor Bruno Walter, and others. The pattern of attention burned into him, by perception giving itself again and again, year by year, in the closest of reviews, to minuscule divergences in essentially similar, tiny bodies, such habits of mind drove him over the edge of evolution's own logic.

What a semi-professional like Kammerer does in his day job may be worth critiquing. What he does at night, in private, calls for another reaction altogether. What scientist doesn't dream? Kammerer dreamed, in private, that the types of regularities he saw in the vivarium could, with the proper method, be found across all regions of being, as much in salamanders as in the Vienna Philharmonic. List after list he made in multiple notebooks, gathering mundane events with the scarcest similarity to each other; he categorized them, drew them out into longer and longer kinship series. Here is one notebook entry under the category "numbers": "On June 15, 1912 I went to a concert at Beethoven Hall in Vienna and had a seat in the 18th row on the inside; on the very next day I sat in the great hall of the Musical Association, once again: the 18th row." Yellow dots on black salamanders correlated at the vivarium; words, Salamanders names, gestures, numbers, personal correspondence, people, dreams, memories, accidents and deaths and other things noted in the newspapers, correspondences without end, crowded his notebooks. You may call them coincidences, but, without the slightest bent toward mysticism, Kammerer calls them "serial experiences." Right after the war he writes a book about them, The Law of Series. He writes of the epiphany he wanted for readers, "serial experiences… appeared before now to [the reader] as meaningless accidents." He continues: "From now on however [the reader] will learn to pay careful attention to them as the expression of a secret lawfulness within them."

Crackpots at the edges tell us a lot about a discipline, about its limits and where they fall, and also about the secret desires and dreams that motivate even sane practitioners. Natural science aims at a theory of everything. Practitioners are usually humble enough to conclude that it is not for them in their lifetimes. Not for being totally mad are charlatans pushed out, but rather, at least at times, because they reflect in gross caricature assumptions that those in the center half-consciously intuit and sometimes repress. For sure, crackpots exaggerate the basic assumptions of a science; yet, as a result, they make them visible. Kammerer's belief in a "secret lawfulness" of nature is not even really marginal. Since Newton, finding or defining "laws" is a chief preoccupation of many sciences. That there are such laws is a central assumption, and because laws themselves are not visible, they make up the corpus mysticum of the science, around which the practitioners carry out their rituals. Kammerer took extreme positions of course—less extreme in his time than they would seem now, but still well out of the mainstream. He believed, or came to believe in inheritance of acquired traits. Tired of being flabby? Work out for a few years and buff up. By this theory, your biceps can now be inherited by your offspring. Kammerer made sightless salamanders see (see "Dunkeltiere im Licht und Lichttiere im Dunkel," 1920), water breeding salamanders breed on land and land breeders breed in the water. Or so he claimed. He also claimed these traits were then inherited.

These were more than observations, they were makings, more than makings, they were markers of a deep antipathy to mechanical explanations of nature. They were the marks of a strong anti-natural inclination, a desire to free matter and life from natural kinds, as well as from the cold, willless Darwinian constraints of accident and environment. Kammerer's amphibians were less experiments in breeding than living canvases, skin painted with spots by laboratory techniques, dark skin made light, light made dark, eyes drawn out where they were dormant; and, most unusual, the alterations would then enter into the hereditary line and contrive new species. Or that was the claim…

That was the claim, and Kammerer did not shy away from the associated revolutionary hopes. If acquired traits were heritable, education was equivalent to revolution. Not only good for one generation—a lesson given now might be learned by all subsequent generations. He wrote in 1923: "It must not be believed that the processes described above are limited only to salamanders and lower organisms: similar facts underlie the life processes of other classes of the animal kingdom, of man himself" ("Inheriting Acquired Characteristics"). In a speech that same year he told a popular audience: "Frobel, Pestalozzi [educational reformers much in vogue at the time] and their schools relied on the potential dispositions which the child inherits from its ancestors, hereditary dispositions which the educator hoped to enrich; but he could not hope to bestow on the children a permanent heirloom in which their children's children would be able to share-only a gift for the fleeting duration of an individual existence." However, "on the hypothesis of the inheritability of acquired characters, which seems to be closer to the truth, the individual's efforts are not wasted; they are not limited by his own lifespan, but enter into the life-sap of generations" (quoted by Koestler).

In the context of these hopes, his private notebook jottings become legible. Alongside decades-long work on physical changes in organisms, against the backdrop of revolutionary plans to re-fertilize natural history with "the good" and de-fertilize "the bad," not forgetting the eugenical implications of these impulses, alongside and we could say underlying all of this work, in the private notebooks lay Kammerer's big dreams. In the notebooks and then the book The Law of the Series (Das Gesetz der Serie, which has not been translated into English) he produced a typology of series the likes of which had not and has not been seen. Engineering the births of countless salamanders, writing dozens of books and more than a hundred articles over 25 years, he also mothered an alt-biology and alt-cultural studies—a new science of everything: serialogy—into existence.

A series is a lawful association of events that do not have a shared cause. Sitting in the 18th row in two different theaters on two consecutive days is an example of one kind of series. Wind blowing off your hat on one corner and then it happening again on the next corner is an example of a different kind of series. The hat scenario is especially exciting to Kammerer because, he notes, there are much windier days when you even put your hat on loosely and it doesn't blow off. Why just this time? And then again? The law of the series. Another way to see this is the following. There are so many causes and potential causes for the twice or thrice blown hat, they cannot be exhaustively described, and so "cause" is not an appropriate explanation. "Accident," he insists, is also bad. Cause overdetermines the blowing. Accident underdetermines it. What determines the series of lost hats to perfection is a general law of series that says simply: events resembling one another gather up in space and in time. If true, this would explain in the strongest manner the coherence and orderliness of all sorts of phenomena. It sounds mad, or like a matter of faith, but the evidence mounts, and much of it seems plausible—alluring even.

There are, Kammerer shows, endless rows of things and families of events in space as well as countless periodic events in time of different frequencies. Series have a tendency to proliferate. One series gives tise to other series, series are contained in higher series. If, Kammerer argues, if there is something like a law of the series then a single series doesn't really exist at all. All series are parts of others, and the whole of nature and the whole of culture fall into nested and intersecting arrays of like things. Existence is a stream of serial happenings. Obviously any branch of the tree of life is an evolutionary series, species stretching from the proto-ancestor to your pet cat. Likewise the history of the world is a series of likenesses, where French Revolutionaries wear Roman togas, or sovereignty continues to take medieval forms. All series fall into shapes; some are cyclical: AA BB AA BB. Cyclical series have shorter or longer periods: XXX Y AAA BB XX YY / XXX Y AAA… Sometimes you find a penumbra of lightly diverging resemblances, like a family of faces:

Kennedy family
The Kennedy Family, 1931

In the image above, JFK would be the first on the left in the back row. That face is like his face. Similarly, we can show how a series can be made up of pictures of JFK over his lifetime.

Jfk7
Jfk7
Jfk7

Jfk7Jfk7

A less adventurous evolutionary biologist, one in the center of the field, would hold that these likenesses have a material base and respond to a mechanical impulse. Regularities traced to heredity, to genetic constraints and processes. We should understand them as expressions of an underlying causal order. Kammerer's great leap into absurdity and wild invention was to envision a law of likeness—things of the most divergent material sorts gather according to likenesses. No one has to perceive these likenesses. It is not a form of idealism. Nature and culture—on their own—series up.

Another series, or the same one, would include the actor who reportedly will play JFK in a movie,

Lookalike
Caspar Phillipson

as well as all the photographs and paintings of the nation's favorite son, which, although we take them to be "JFK," are obviously only like him in some respects. The list could include likenesses in other respects, such as Alfred E. Neumann, a sometime JFK lookalike, when he ran for president in 1960:

Neuman

And of course beyond caricatures, other series, larger, more inclusive—meta-series after meta-series—that would include or vector off from this JFK set would be things like the series of bipeds, the series of suit-wearers, the series of the super-wealthy centrist democrats, embodied beings. Members of each of these series are unlike on another in some respects and alike in some respects. Let them overlap and trace out the implications of each series, of every series that branches off or includes others. From here a picture of the all and the whole of everything could be built up from the arrangement of its elements, under the thesis that everything is like everything in some respect.

And who can tell whether this is all a glorious lie, faulty methodology, or fast talk with no hope of substantiation? Today I report to you, members of the public, on this point of indistinction. It would be too irresponsible of me to suggest that if we disregard the standards of probity and take Kammerer's crackpot theory as containing some truth, it would open for us the doors of perception.

Like what you're reading? Don't keep it to yourself!
Share on Facebook
Facebook
Tweet about this on Twitter
Twitter
Share on Reddit
Reddit
Share on LinkedIn
Linkedin
Email this to someone
email