The Message in the Bottle and Lost in the Cosmos

by Walker Percy

  Some further symbolic notation is required to signify the difference. Perhaps we could subdivide Russell’s yellow (A):

  yellow (A) (?) yellow (A) (!)

  * F. Mainx: “…the experience of inorganic and that of organic science join together to give a unitary and consistent picture of the world, derived from the fundamental unity of method.”

  * Although anthropologists differ greatly in their philosophical allegiances, it is my contention that the two main schools, the functionalists and the superorgani-cists, share a common theoretical posture toward their subject matter. A functionalist like Malinowski may understand culture more or less biologically, as “an instrumental reality, an apparatus for the satisfaction of fundamental needs, that is, organic survival, environmental adaptation, and continuity in the biological sense. A superorganicist like Kroeber or White or Sorokin may understand culture as an autonomous reality which is “participated in and produced by organic individuals.”

  In both cases, however, man himself, his personality, is understood as a function of an underlying reality, in the one case, a function of his encounter and response to his environment, in the other, as a function of the culture in which he particpates. In neither case is culture understood as the creation of man which stands over against man as the means by which he can develop the potentialities of his nature— or the means by which he can fall prey to anonymity. For a searching critique of modern anthropology from the point of view of a realistic humanism, see David Bidney’s Theoretical Anthropology (New York, 1953).

  * I use the word “causal” without prejudice. It means whatever the reader would have it mean in the context: either efficient causality or a probability function.

  * Another member of the Viennese circle is highly critical of Carnap’s logicizing of natural laws. Moritz Schlick writes: “It [the natural law] is then not a natural law any more at all; it is not even a proposition, but merely a rule for the manipulation of signs. This whole reinterpretation appears trivial and useless. Any such interpretation which blurs such fundamental distinctions is extremely dangerous.”

  * Heisenberg’s uncertainty relations have been hailed by some enemies of science as proof of the freedom or irrationality or whatnot of the ultimate particles of matter. As Nagel observes, however, “a more sober and prima facie plausible account of the uncertainty relations is that they express relatively large but unaccountable modifications in certain features of subatomic elements, resulting from an interaction between these elements and the instruments of measurement.”

  * It would be possible to develop the same antinomy in other “symbolic forms”— art, history, religion.

  For example, a contrast could be drawn between the pragmatic theories of art as a “play activity” or the behavioristic theory of art as a traffic in emotions, on the one hand, and the seriousness of the artistic enterprise and the revelatory nature of the art experience, on the other.

  In history, a contrast could be drawn between the basically particular and historical character of the scientific observation, on the one hand, and the general character of the ultimate scientific expression, the scientific law, and the inability of the scientific method to grasp singulars except insofar as they exemplify general principles.

  As for religion, although it is listed by Cassirer as a separate “symbolic form,” the very nature of the method used cancels the difference between religion and myth, since it refuses on a priori grounds to grant cognitive content to religion, and so ranks religion as a “higher form” of myth. Once the scientific method is elevated to a supreme all-construing world view, it becomes impossible to consider a more radical science, the science of being. Thus, when Cassirer is confronted with the assertion of pure existence which Moses received from God as His Name, I am Who Am, he is obliged to see it as a piece of semantical magic, a “mythical predication of being.”

  It seemed more expedient, however, to develop the antinomy by using the same three “symbolic forms” used by Cassirer in his major work: myth, language, and science.

  * D. Bidney: “The thesis I am concerned to establish is that the postulate of an ontological human nature is a prerequisite of both individual and social psychology.”

  11

  SEMIOTIC AND A THEORY OF KNOWLEDGE

  A STUDENT OF current philosophies of science must sooner or later become aware of a curious state of affairs. If he is accustomed to the discipline and unity of a particular science, he may reasonably expect that a philosophy of science will in turn confer a larger unity on the elements of the scientific enterprise, not merely the various data of the sciences, but also the conclusions and the activities of the scientists themselves. This is not, however, what he will find. What he is more apt to encounter in the various symposia and encyclopedias of unified science is an inveterate division of subject matters. Some may be written entirely in one language and some entirely in the other; some may be a mixture of both; but neither seems to have much to do with the other. The two approaches are (1) the nomothetic method with which he is familiar, arising from the “inexpugnable belief,” as Whitehead put it, “that every detailed occurrence can be correlated with its antecedents in a perfectly definite manner exemplifying general principles”; and (2) the quite different program which Russell, after completing the Principia Mathematica, staked out for philosophy as its sole concern—the logical analysis of empirical propositions established by perception and science.

  To take the most ambitious and interesting example of a “metascience,” semiotic, the science of signs—interesting because, unlike pure symbolic logic, it tries to unite logical analysis with the explanatory enterprise of science, and because, whatever its short-comings, it has at least hit upon the fruitful notion of man as the sign-using animal—here too one encounters the same division of subject matters with no visible means of getting from one to the other, despite the many assurances that semiotic confers unity. If one expects a larger epistemological unity in which the relation of logical analysis to the scientific explanation of natural events is to be made clear, he will be disappointed. He will get logical analysis and he will get scientific theorizing, but he will not learn what one has to do with the other. * There are studies on the biology of sign function, and here one recognizes a basic continuity with the manifold of natural phenomena. When one speaks of animal A responding to buzzer B by salivation in expectation of food F, one is speaking a language familiar to psychologist, physiologist, and physicist alike, the language of spatio-temporal events which lend themselves to causal hypothesis. Stimulus-and-response events occur among natural existents and are mediated by physical structures and a causal nexus which is recognized as valid for organic and inorganic matter.† Thus, whatever the limitations of a biological science of signs in man and animals,‡ one readily recognizes its validity as far as it goes. But then one suddenly finds oneself in the charged atmosphere of the Polish semanticists with their scoldings at the human abuse of signs. At one moment one is studying sign behavior as a natural science, in which “interpreters” behave according to lawful empirical regularities, and in the next moment as a quasiethical science, in which “interpreters” disobey semantical rules and in general behave stupidly and perversely. There will also be articles dealing exclusively with syntactical rules in logic and mathematics, with the arbitrary formation of calculi, with the principles of logical implication. Or one may read statements by the same semioticist that (1) the basic terms of semiotic are all formulable in terms applicable to behavior as it occurs in an environment, and (2) semiotic can be presented as a deductive system with undefined terms and primitive sentences which allow the deduction of other sentences as theorems.

  The fact is that a man engaged in the business of building a logical calculus is doing a very different sort of thing from an animal (or man) responding to a sign, and it is a difference which is not conjured away by ignoring it or by leaping nimbly from res extensa to res cogitans as though there were no epistemological abyss
in between. I cannot say it as well as Professor Crockett: “I do not know whether one should try to describe the universe or whether one should play games with marks arranged according to certain rules; but I do know that one should decide which of these vastly different kinds of activities one is engaged in and inform the reader accordingly.”

  It is not my intention to make a case against either of the two major components of semiotic, symbolic logic, and behavioristics—the shortcomings of each are well known by now. Rather it is my hope to show that a true “semiotic,” far from being the coup de grâce of metaphysics, may prove of immense value, inasmuch as it validates and illumines a classical metaphysical relation—and this at an empirical level.

  I think it will be possible to show (1) that the “unified science” of semiotic is a spurious unity conferred by a deliberate equivocation of the word sign to designate two generically different meaning situations (the sign relation and the symbol relation) and (2) that an open “semiotical” analysis of symbolization—that is, one undertaken without theoretical presuppositions—will encounter and shed light upon two metaphysical relations: the first, the cognitive relation of identity by which a concept, a “formal sign,” comes to contain within itself in alio esse the thing signified; the second, the relation of intersubjectivity, one of the favorite themes of modern existentialists. It may well turn out that the semioticist has good reason to ignore the symbol relation in view of his dictum that sign analysis replaces metaphysics, since an impartial analysis of symbolization can only bring one face to face with the very thing which the semioticist has been at all pains to avoid—a metaphysical issue.

  Let us not be too hasty in surrendering the symbol to the symbolic logician or, as is sometimes done, to the mythist.* It is possible that a purely empirical inquiry into the symbol function, an inquiry free of the dogmatic limitations of positivism, may provide fresh access to a philosophy of being.

  SYNTAX AND SCIENCE

  Semiotic, the science of signs, is an attempt to bring together into the formal unity of a single science three separate disciplines: (1) the semantical rules by which symbols are applied to their designata, (2) the logical analysis of the relations of symbols as they appear in sentences, and (3) the natural science of behavioristics (to use Neurath’s terminology), in which organisms are studied in their relation to the environment as it is mediated by signs. It was soon discovered, as Sellars points out, that the limitation of scientific empiricism to logical syntax is suicidal; and so the semantical and biological study of signs was added under the guidance of C. W. Morris. According to Morris, these three disciplines may be regarded as three “dimensions” of the same science, the semantical dimension of semiotic, the syntactical dimension, and the pragmatic dimension.* This division is held to be analogous to the division of biology into anatomy, ecology, and physiology; a symbolic logician, a semanticist, and a behaviorist are said to be emphasizing different aspects of the same science. Physiology requires anatomy, and ecology requires both; all three conform admirably to the biologist’s conception of organism as a system reacting to its environment according to its needs of maintaining its internal milieu and reproducing itself. Physiology is complemented by anatomy; one flows into the other without a hitch. But how does syntactical analysis flow into behavioristics? One may make a syntactical analysis of the sentences written down by a behaviorist, or one may study the sign responses of a symbolic logician; but in what larger scheme may the two be brought into some kind of order? We find symposia written from either point of view, from the physicalist’s, who starts with matter and its interactions and tries to derive mind therefrom, or from the symbolic logician’s, who conceives the task to be the syntactical investigation of the language of science. Far from the one flowing naturally into the other, the fact is that one has very little use for the other. It takes the encyclopedist to bring them together.

  It is well known that logical empiricism is without a theory of knowledge since it restricts itself to an abstract theory of the logic of language. It is equally well known—and perhaps one is a consequence of the other—that the history of logical empiricism is the history of wide fluctuations on the mind-body axis. Examples of the extremes are the solipsism of Mach, Wittgenstein, and the early Carnap of Der logische Aufbau and the physicalism of the American behaviorists and the later Carnap. But even in the more modern attempts at unity, one is aware of the tendency to construe the field exclusively from the logical or the physicalist point of view—and indeed, how can it be otherwise when the problem of knowing is ruled out of court? A semioticist can easily take the position that the only genuine problem, as Carnap claimed, is one of logical analysis; that is, the question of the formal relations among the concepts that describe the data of first-person experience, the concepts of physics, and those of behaviorist psychology. Or one can begin at the other end with the causal relations between signs and interpreters and derive mind and consciousness with never a thought for syntactical analysis. Anatomy is indispensable to physiology, but syntax can get along very well without neurology. Neither symbolic logicians nor behaviorists are constrained to make contact with each other, and it is perhaps proper that they do not. But it is the semioticists who have brought them willy-nilly together to form the new organon. It is not unreasonable, therefore, to expect that this metascience will provide a larger order. Perhaps, then, it is the semanticists who fill the gap. For semantics professes to deal with both the words of the logicians and the natural objects of the scientist.

  We are destined to disappointment. Semantics, it turns out, abstracts from the user of language and analyzes only the expressions and their designata. Like syntax it operates from the logical pole in that it is chiefly concerned with formation of “rules” for the application of symbols to things. Korzybski, we discover, is not interested in how it is that words get applied to things, in the extraordinary act of naming, but only in our perverse tendency to use words incorrectly, and in making a “structural differential” so that one may use words with the full knowledge of the level of abstraction to which they apply. Or if we turn to Tarski’s classic paper on the semantic conception of truth with high hopes that at last we have come to the heart of the matter, we will find as the thesis of the article the following criterion of “material adequacy”: X is true if, and only if, p is true, which when interpreted yields: “Snow is white” is true if, and only if, snow is white. I do not wish to deny the usefulness of Tarski’s criterion within the limits he has set; I only wish to point out that Tarski by his own emphatic asseveration is not concerned with the problem of knowing.*

  If, in order to bring the twain together by the semiotic method, we strain forward to the furthest limits of behaviorism and backward to the earliest take-off point of semantics, we will find that the gap between them is narrow but exceeding deep. Logical syntax begins with the “protocol statement,” the simplest naming sentence; semantics is exclusively concerned with its rules of designation. In regard to the logical syntax of the language of science, Carnap wrote: “Science is a system of statements based on direct experience and controlled by experimental verification….Verification is based on protocol statements.” Protocol statements are “statements needing no justification and serving as the foundations for all the remaining statements of science.”

  Behavioristics, even taken at its own estimation, brings us to a point considerably short of the relation of denotation and the protocol sentence. It deals with the sign behavior of animals and man according to the method of natural science—that of discerning empirical regularities and later attributing them to a causal function, a =f (b). An organism’s response to a stimulus is resolvable into a sequential series of commotions mediated by structures, beginning with an air vibration and ending with an efferent nerve discharge into a glandular end-organ.†

  An object-science of behavior can only make sense of language by trying to derive it from some refinement of sign response. As Susanne Langer has pointed out, when
the naming act is construed in these terms, when the situation in which you give something a name and it is the same for you as it is for me, when this peculiar relation of denotation is construed in terms of stimulus response, one has the feeling that it leaves out the most important thing of all. What is left out, what an object-science cannot get hold of by an intrinsic limitation of method is nothing less than the relation of denotation—a name above all denotes something. If you say “James” to a dog whose master bears that name, the dog will interpret the sound as a sign and look for James. Say it to a person who knows someone called thus, and he will ask, “What about James?” That simple question is forever beyond the dog; signification is the only meaning a name can have for him.

  The upshot is, even if we go no further than Mrs. Langer, who is otherwise in sympathy with the positivism of the semioticists, that in semiotic symbol analysis and the science of sign behavior are brought willy-nilly together into a unity which has no other justification than that both have something to do with “sign.” No larger sanction can be forthcoming because of the dictum that sign analysis replaces metaphysics. To say to a semioticist that he is confusing the logical with the real is unacceptable to him because of the “metaphysical” presuppositions involved. One might nevertheless expect that, within the limits of the semiotical method, some attempt might be made to achieve the continuity so highly prized by semioticists since the time of Peirce.* Failing this, one cannot help wondering whether to do so, to explore the gap between pragmaticsmatics and symbol analysis, will not run squarely into an “extrasemiotical” relation—not as a “metaphysical presupposition” or a “naive realism” but as an issue which is precisely arrived at by the semiotical method itself.

  SIGN AND SYMBOL