Science Fiction Studies

#2 = Volume 1, Part 2 = Fall 1973

Stanislaw Lem

Remarks Occasioned by Dr. Plank's Essay "Quixote's Mills"

According to our cultural tradition a literary work, is significant in the degree to which it differs in a nontrivial way from all previously published works. This individualistic approach, strongly correlated with the concept of the uniqueness of each literary work, is perhaps dying out in our time, for it is opposed to the trend of mass culture.

A literary work is at the same time something strange and something familiar. Thanks to its familiarity we can understand it well; thanks to its strangeness we can experience through its content a new revelation of what we had thought to be banalized, doomed to silence. Neither the strangeness alone nor the familiarity alone can make a work of art relevant: the fusion of the two is the necessary precondition of creation.

From a bird's-eye view, the whole domain of modern literature is somewhat depressing. Mainstream writing tells us all about practically nothing, and SF tells us practically nothing about all. The mainstream bores us with everyday trivia as seen through a magnificent microscope-- the magnificent system of well-focused lenses created by the self-devouring sophistication of hyperspecialized narrative modes. In SF we look through a badly built apparatus at "all"--but this "all" is dubious if not falsified. We have in SF a counterfeited Cosmos, a nauseating mixture anachronisms, counter-empirical as a whole, and an oversimplified future of the race, perceived only at its black extremes, only in the spectrum of all imaginable holocausts and global suicides; and what stands in SF for human reason is really only a moronized superman version. In SF we do not even see the big problems central to future research (future social structures, the relationship of man to artificial man, mankind endangering itself with the not yet evident consequences of so-called progress) other than through a dimmed glass, darkly: the whole apparatus is out of focus--and intentionally so.

This intentionality must be explained before we can proceed. Our neighbor planets, Mars and Venus, were thought by the old time astronomers to be images of Earth's future (Mars) or of its past (Venus). We know now that neither of these assumptions is true. Venus is not similar to Earth as Earth was some hundreds of millions of years ago, and Mars has nothing in common with the planet imagined by astronomers--a planet on which the once vigorous life forms were dying out, being involved in an heroic but already lost struggle with worsening environmental conditions (the "canals" were thought to be signs of this struggle, with respect to the supply of water). So we see that for the old time astronomers these planets were a kind of Rorschach test: they did not see the planets as they really were; they did not stop at the limit of their knowledge based on observational facts; they instead projected their unconscious, anthropocentric expectations into outer space, building a coherent, intelligible system of relations, a spectrum of celestial bodies in which Earth's position was still central, thus repeating the error of pre-Copernican cosmology. The parallel looks like this: Venus has not YET achieved, and Mars has ALREADY lost, what Earth NOW possesses. This invariant of our perception manifests itself in science only unintentionally, for the sole task of science is to build a model of reality as true as possible. Since the task of science is unambiguous, the attitude of the scientist has little in common with that of the writer, for whereas there is a licentia poetica, there is no such thing as licentia scientifica.

Although the distortion of the world's image is possible in science only as an error, such distortion in literature is often the proper mode of individual expression. But not every kind of intentional distortion and transformation of true facts is of the same value in literature. Highly specialized distortions, or rather their modi in a paradigmatical sense, are selected out of the set of "all possible distortions" by the process of evolution in the various literary genres. This selection is by no means a rational progress, for the writers themselves do not know why this rather than that transformational pattern of the narrative becomes a leading paradigm of creation. Nevertheless Occam's razor is in a way valid in belles lettres. It can be called the minimax tactics of creation: the greatest impact results from a minimum of invested means--a very old maxim, put by J.W. Goethe in the words "In der Beschrinkung zeigt sich erst der Meister." The neglect of this rule results in an inflationary escalation of means and meanings.

While the real universe is a thing too big for our science (and will remain so forever, I think), it is too small and too crowded a place for the expansion of escapist dreams, so SF has invented a lot of other universes. But this remark is only preliminary; the heart of the matter is inflation in art, and specifically in SF. The generally low standard of SF is often defended with an expression now proverbial in SF circles, that "90% of everything is trash." But this defense is not only inappropriate; it misses the point central to any discussion of the genre.

In the realm of art we are accustomed to the state of things in which a hierarchical order builds up with the passing of time. The best works find their place at the top of the pyramid, and the hopeless misconstructions find theirs at the bottom. This process of ordering goes on very slowly; as the device filtering out the worst and selecting the best, time is the constantly moving factor in this catalytic cracking. But alas! so much broken glass is given for diamonds in SF that every true diamond must be dimmed, if any at all are to be found in these mountains of waste. The worst eclipses the best; true, but this is only the prima facie notion. In reality the perfection of a work of art is only in its potentiality, its chance of maturing, of expanding in meanings; and there are some necessary preconditions for this kind of growth: the duration of contacts (between a work and its readers), the attention given to each work, and so on. All this can be translated into the technical language of information theory, with its "channel capacity," "noise-signal ratio," etc. The most relevant piece of truth in information theory is the rule that when we have too many signals, their common resultant can only be noise. The principal factor in the killing of creation in our time is the law of large numbers. The hyper-saturation of the market is a suicidal process in our culture: the forces of selection are paralyzed. So what determines the success of a literary work? To an ever greater degree, the determinants are accidental; where there was once self-organization through maturation and discriminative feedback, now there is only chance. If we have not yet reached the breaking point, the point at which the randomization of "natural selection" inculture wipes out all other kinds of selection, we are surely moving, and with acceleration, to this point of no return.

Let me point out some symptoms of this malaise in the artistic genre of SF. Firstly, there is the staggering discrepancy between the literally cosmical aspirations of SF and its realizations. The cosmos is already miniaturized to pocketsize in a vast bulk of SF books. This miniaturization is counteracted only by means of words: by adjectives, say, in statements that the distances between the stars are "immensely great," by the notion of cosmonautics with speeds measured in "kilolights," etc. But a pocketsize universe simply cannot be taken seriously, so we have already a grotesque unintentionally attained in this writing--a dangerous and bad omen.

Secondly, a growing number of books show signs of very strong referential bonds with the whole background of science fiction. These books, when put in the hands of a reader with no experience in SF, remain simply incomprehensible. They look as if they had been mutilated, and truly they have been mutilated: such a book is no closed system, no sovereign independent entity, no self-sustaining organization of meanings but is instead a fragment, a particular and incidental embodiment of the trend now fashionable in SF. In a word, these books have already lost the central. attribute of a work of art: individual, nonrepetitive, unique characteristics. They are very similar each to other, and even if they differ, their differences are dominated by their general resemblance. This is pulp in the primary, material sense of the word.

The disindividualizing trend can eventually become the dominant one in all branches of art; then our present views will perish, as a transitory phase in the evolution of culture. I do not forecast this state of affairs; I consider it only a serious possibility. To claim today that one will spend years in the writing of a novel--as Thomas Mann did, working ten or twelve years on a single book--would be not a sign of spectacular courage but rather of madness. Who can know what the fate of a book planned now would be twelve years from today? There is building up a pressure that necessitates the synchronization of creative plans and practical considerations (marketing) in a writer's production: he is forced to write under conditions more and more resembling those in the typical production of mass goods.

Perhaps all this is so, but what has it to do with Dr. Plank's essay, which introduces into the relation of men and machines such notions as this of the "Don Quixote effect"?

Firstly, one can put the notion "escapist" or "estranged mind" for Dr. Plank's notion "Don Quixote," and "empiricist" for "Sancho Panza." This would already be a somewhat polemical substitution, since I have stated above that I understand a literary work to be a union of these opposites. When you have estrangement only for estrangement's sake, you are escaping the real world. The presence of both Don Quixote and Sancho Panza in Cervantes' novel is no chance accident, neither is the epilogue of this great book.

Secondly, we have in literary criticism the choice between the normative and the descriptive approach. A normative approach must ,necessarily end with the construction of a normative aesthetics. A normativistic critic is forced to throw 98 or 99% of SF to the wolves, and to plead for a "better, future SF"--and so he himself becomes a utopian, since no man can by himself change a big complex trend in the social or in the cultural life of his society. There is simply no substitute for natural selection as a means of preserving prominent works of art and drowning bad ones in oblivion. So even if I am biased here in that I feel a strong attraction toward the normative attitude, I must also sympathize with the descriptive study of literature, for I am a science addict, and the egalitarian mode of research is typical of science.

Now descriptive research is not a univocal term, for we have both synchronical and diachronical modes of description and generalization. The synchronical mode was first worked out by the structuralist school: here the central problem is how a given work has been put together and what its relations are with the proper genological paradigm. The second school, more traditional, treats a work as a link in the chain of sociocultural and philosophical events distributed along the time axis of human history. The structuralist mode of research is, alas, insufficient, since its theoretical apparatus has no necessary resolving power: it cannot distinguish between certain subsets of complex literary texts; e.g., between a dull (and boring) complexity and a brilliant (and fascinating) one. (This is a matter of fact, not openly stated by the structuralists themselves.)

The second school tends to explain away the immanent quality of every literary work, since in this historical perspective a given work is only a point of hybridization of some antecedent current.,; of thought: at the end of the analysis the work in question is reduced to a handful of other works, philosophical viewpoints, and sociological concepts.

Dr . Plank has taken another road, a third road: he has inverted the canonical attitude of considering an artistic work as the ultimate object of study. The SF landscape is for him a projection of the deep currents building up tensions in the base of human minds. He treats SF as an expression of those forces., In The Emotional Significance of Imaginary Beings (1968) he has stated that a bad, primitive SF story is more instructive than an SF masterpiece, since in a primitive story the primary forces working at the base of creation are not camouflaged by the author's, skills, by all the evasive techniques of narration that are typical of the sophisticated and talented writer (pp88-89). In this way Dr. Plank has shown most clearly his own critical attitude, which is by no means a proper one for the "pure" critic of literature.

His is rather the attitude of the scientist and physician, as will appear in the following consideration. For the physician there are of course easy and difficult cases, but there are no "good" or "bad" symptoms of an illness. All symptoms have the same value as signs, and this also applies for the psychologist or psychiatrist, since such a man does not evaluate the enunciations of his patients in the way we would evaluate those same enunciations in everyday life; for him the enunciations are not statements about fact but rather symptoms of the patient's mental state. It is also true that a taxonomist does not distinguish between "good" and "bad" species: cockroaches and gazelles are equals to him, for his function does not include the aesthetic (or moral) evaluation of what he is studying.

This scientistic attitude is alien to literary criticism, and if applied systematically, it nullifies the axiological viewpoint. The analyzed work of art ceases to be a work of art in the proper sense, for it is no longer the ultimate object of study, being used as an apparatus pointing out and magnifying the basic pattern of the emotions active as factors of selection and creation in the author-reader complex.

What survives as organism or species is neither "bad" nor "good" in the eyes of the biologist. There are criteria of "badness" and "goodness" at work in natural evolution, but they are of a "technological" (i.e., pragmatical) character: what survives is evidently "good" and what dies out evidently "bad"--in the sense of "badly built"--and so unfit to survive. Now when we substitute the salability of books for the survival fitness of species, we have before us the proper model of the critical attitude proposed by Dr. Plank. (The set of all readers stands here for the environment of organisms.) Practically all SF is trash; nevertheless not all SF books are selling equally well. Fandom prefers some books and rejects others. Why? Because those others were badly written? No, rather because the privileged and selected out are in tune with certain expectations of the readers. There is apparently a unisonance involved in this process of filtering, a unisonance of all the unconsciousnesses involved, those of the authors and of the readers.

But this is still an interesting mode of approach. It is worth pointing out that Michel Butor has proposed that SF authors work together, in big collectives, to build up a single fictitious world; this proposal caused some indignation on the Olympian heights of SFWA.1 But no one has remarked that Butor's idea is already accomplished to 80 or 90%, and this accomplishment manifests itself in the syncytial character of all produced works that have in common practically the same fictitious universe; that is, falsified in the same way, with "hyperlight travel," with time loops, with parallel worlds, with man-reduplicating machines, etc.

The sole weakness of Dr. Plank's approach is this: if for the sake of the argument I accept his general assumptions and approve his attitude, I must still disagree with many of his particular statements. But there is no point in opening a polemic, in writing diatribes. What I could say would not in the least change the principal object of disagreement. My arguments, on this level of discourse, would have the same inbuilt weaknesses as those of Dr. Plank. On this level there are simply no methods of proving who is right and who is wrong. Here, in this psychoanalytical approach, we have no simple basal facts--no facts already established beyond any reasonable doubt. Sam Lundwall has written that we have in SF two kinds of artificial beings in accordance with an ambivalence typical and well known in the domain of the unconscious. Androids are sexed but robots are sexually disarmed, because we, the readers, are at one and the same time attracted and repelled by the idea of sexual intimacy with an artificial man or woman.2

Perhaps Lundwall is right, but how could we prove that he is? Breadth of imagination cannot represent the highest instance of appeal in the theory of literature (or of its genres). Of how to evaluate Dr. Plank's hypothesis on the "Don Quixote effect" or "complex," I have no idea. After reading "Quixote's Mills" I have the same feeling as after finishing The Emotional Significance of Imaginary Beings, the feeling that both these texts should start at the point where they come to an abrupt end. I can only hope that this is not a dead end. Perhaps statistical analysis of a truly big set of SF works could be of some help here. Their central leitmotives could be analyzed in search of statistical significance. So we need now some computers, some programs, some men with proper capabilities, and of course a lot of money. This is how literary considerations are ending today.

Perhaps I should add some words explaining that I do not see any serious disagreement between Dr. Plank's position and my own with respect to the "blasphemous nature" of robots in our culture. In my essay on robots, I did not claim that I was enumerating all the factors involved in "anti-robotical" feelings and/or anxiety, but I thought that my argument was reasonable in the form of a selective approximation of the problem. The problem is of course a very complex one, and my essay was a minor thing, written for a limited audience.3


1 See Michel Butor, "Science Fiction: The Crisis of Its Growth," and James Blish, "On Science Fiction Criticism," in Thomas D. Clareson, ed., SF. The Other Side of Realism (1971), ppl57-70.

2 Sam J. Lundwall, Science Fiction: What It's All About (Ace Books 1971), §7.

3 "Robots in Science Fiction," written in German for Quarber Mercur, No. 21 (Nov. 1969); an English translation appears in Clareson pp307-25.



Dr. Plank has inverted the canonical attitude of considering an artistic work as the ultimate object of study. The SF landscape is for him a projection of the deep currents building up tensions in the base of human minds. He reads SF as an expression of those forces. In The Emotional Significance of Imaginary Beings (1968), he stated that a bad, primitive SF story is more instructive than a SF masterpiece, since in a primitive story the primary forces working at the base of creation are not camouflaged by the author’s skill. In this way, Dr. Plank has shown his own critical attitude, which is by no means a proper one for the "pure" critic of literature. His is rather the attitude of the physician and scientist. For the physician there are of course easy and difficult cases, but there are no "good" or "bad" symptoms of an illness. All symptoms have the same value as signs. It is also true that a taxonomist does not distinguish between "good" and "bad" species: cockroaches and gazelles are equals to him, for his function does not include the aesthetic (or moral) evaluation of what he is studying. This "scientistic" attitude is alien to literary criticism. The analyzed work of art ceases to be art in the proper sense, for it is no longer the ultimate object of study. The sole weakness of Dr. Plank’s psychological approach is this: even if I approve his attitude and assumptions, I must still disagree with many of his particular statements. Of how to evaluate Dr. Plank’s hypothesis on the "Don Quixote effect" I have no idea, but after reading "Quixote’s Mills" I have the feeling that the text should start at the point where it comes to an abrupt end.


Responding in a spirit of magnanimity, Stanislaw Lem has widened the area of conceptualization rather than that of argument. Any extended answer by me would merely impair this desirable effect, so I'll be brief.

He is right, of course, that my basic approach is neither normative nor descriptive--in fact, primarily one of psychological study rather than either type of literary criticism. I may well be guilty of not having kept the three approaches as separate as they ought to be.

I can't agree more than with Lem's judgment that my book would be better if it went on (and it will hardly surprise anybody who knows publishing problems to learn that the manuscript was thrice as long originally). I doubt that this would necessarily take us as far as to computers. The science and art of sampling has developed sufficiently, I hope to spare us such a step that I, for one, still feel, without blushing, to be bit "unheimlich."

moonbut.gif (4466 bytes)Back to Home