Repository | Book | Chapter

File clean-up May 24, 2022, 5:21 pm sdvig press ( )

1Published originally in Ladislav Matejka, Peter Steiner & Richard Bailey (eds), The sign, Ann Arbor, Michigan Slavic Publications (1978): 43-67.

The cognitive function of signs

2In his writings on the philosophy of language, Karl-Otto Apel (1973, 271; 353) repeatedly advances a “linguistically oriented” or, in more general and pertinent terms, a “semiotic transformation” of transcendental philosophy. There are no doubt clever outsiders who will assume that philosophy has now joined the ranks of the many younger as well as older sciences already sucked into the wake of the all-pervasive and influential disciplines of linguistics and semiotics. Philosophy, in contrast to most other sciences, can counter this charge on historical grounds. Semiotics, the name included (Locke, 1690, §4.21.4), was child and ward of philosophy up to Peirce and Saussure. However, philosophy is no more subject to history than it is to fashion—apart from the fact that history consists largely of a succession of fashions. An orthodox philosopher is concerned only with the Sachen selbst. The basic subject matter of philosophy is the cognoscibile, the questions, “What can we know?” and “How do we acquire knowledge?”.

3It was this paramount philosophical theme that spurred on the devel|opment of semiotics in the works of modern philosophers from Locke (1632–1704) via Leibniz (1646–1716), Wolff (1679–1754), Lambert (1728–1777), Condillac (1715–1780), and Bolzano (1781–1848) to Peirce (1839–1914) and Husserl (1859–1938).1 The most interesting and exhaustive discussions do not, as might be expected, deal with the long dominant psychological aspect of signs, their emotive function, the expression of inner data, feelings and ideas, nor with the currently prevailing sociological aspect, their communicative function, but rather with what can be called the cognitive function of signs and sign systems. Thus it comes as no surprise that the above-mentioned philosophers ex|plore semiotics largely in connection with mathematics and logic.

4An important function of signs, the finding of philosophical semiotics, consists in making possible and extending consciousness, knowledge and understanding.2 Locke paved the way in his Essay Concerning Human Understanding (1690), not in the third book, “Of Words”, but rather in the second “Of Ideas”, in the sixteenth chapter on numbers. He suggests that we would be unable to produce or clearly distinguish more complex numbers if we had no names for the sums arrived at by the successive addition of units. In support of his thesis, he cites American Indians who, having no words for higher numbers, e. g. 1000, depend on vague figurative means such as pointing to the hair on their heads to indicate a larger sum.

By the repeating [...] the idea of a unit, and joining it to another unit, we make thereof one collective idea, marked by the name two. And whosoever can do this, and proceed on, still adding one more to the last collective idea which he had of any number, and gave a name to it, may count, or have ideas, for several collections of units, distinguished one from another, as far as he hath a series of names for following numbers, and a memory to retain that series, with their several names... For, without such names or marks, we can hardly well make use of numbers in reckoning, especially where the combination is made up of any great multitude of units; which put together, without a name or mark to distinguish that precise collection, will hardly be kept from being a heap in confusion” (Locke 1690, §2.16.5).

5In Book III, Locke attributes to words in general the function of improving knowledge. A predominant function of language3 is to classify. On the one hand, words guarantee the lasting fusion of different modes and relations in one complex idea. Without the word triumphus, we would probably have descriptions but no unified and lasting idea of all that belongs to the essence of this ceremony (ibid. 3.5.10). On the other hand, words provide generalizations of both simple and complex ideas; names can be transferred to similar data (ibid. 3.3). Locke’s theory of language is almost exclusively restricted to the lexical stratum and within this stratum to the class of substantives, i. e. to that part of language whose structure is the least homogeneous, thus yielding abundant support for his atomistic approach. Arbitrarily selected relations of similarity and contiguity between ideas act as the principles of linguistic classification.

6Condillac (1746, §1.4.2.27) radicalizes and generalizes Locke’s discovery of the necessity of signs. Of all the modes of consciousness only sensation and perception can be generated entirely without recourse to signs. Higher modes of consciousness such as awareness and imagination depend for their perfection and autonomy upon signs, upon artificially created signs (signes d’institution). Artificial signs free the human mind from its dependence upon real events, to which accidental and natural signs are bound. Memory and operations of the mind such as reflection, judgment and conclusion are utterly impossible without signs (ibid. 1.2. 4 f.; 1.4.1 f.). Condillac’s concluding attack on the Cartesian thesis of innate ideas is rather timely. In his opinion, this thesis is not only a bias refuted by Locke’s discovery of the necessity of signs for the imagination of numbers, which had served as the model illustration of innate ideas since Augustine; it has also obstructed scientific progress: “Comment soup|çonner la nécessité des signes, lorsqu’on pense, avec Descartes, que les idées sont inées... ?”(ibid. 1.4.2.27).

7Leibniz and Lambert, building upon Locke’s reflections, move in another direction. Leibniz (1765, §2.16.5) observes that a mere accu|mulation of names has only a limited usefulness in the constitution of the natural numerical series. “A certain order and a certain replication”4 are necessary to prevent our memory from being burdened with too many new names. Leibniz advances beyond Locke by surmounting his atomistic approach. Attention is no longer focused on relations of similarity and contiguity among individual ideas but rather, to use Lambert’s terminology, on the figurative or metaphorical, or, to use Peirce’s termi|nology, on the iconic or diagrammatical interpretation and evaluation of systematic relations among ideas.

8Signs can be called scientific “when they not only present concepts or things, but also reveal relations such that the theory of the thing and the theory of its signs can be interchanged” (Lambert, 1774, §2.23).

9One of the most instructive examples of such figurative sign systems is Lambert’s graphic presentation of the modes of propositions and conclusions in Dianoiologie oder Lehre von den Gesetzen des Denkens, the first book of his Neues Organon (ibid. 1.196ff.). It can be viewed as a precursor of the Euler Circles, which have survived in logic and set theory as the superior illustration. For the different propositions, Lambert uses lines of varying length, placed under or next to each other depending on the mode of the proposition. Today, relations between statements and classes are represented by concentric, intersected and adjacent circles. With Lambert’s method, the two propositions “All M are P' and “All S are M' yield figure 1 below. The more figurative Euler Circles yield figure 2.

10The interpretation of the diagrams reads: “The extension of M is greater than that of S; and that of P greater than that of M. Further, all S are M, and all M are P." The cognitive value of the drawing lies in its representing six propositions at once, namely, in addition to the two illustrated, the following four: 1) Some M are S. 2) Some P are M. 3) Some P are S. 4) All S are P. In fact, even more. Impossible conclusions are revealed as well (e.g. “All P are S”, ibid. 1.201 f.; 3.29). A figurative-iconic presentation is fruitful since it not only encompasses the cognitions that led to its construction but discloses others as well. It has, in Peirce’s terms, a “capa|city of revealing unexpected truth.”

For a great distinguishing property of the icon is that by the direct observation of it other truths concerning its object can be discovered than those which suffice to determine its construction. ... Given a conventional or other general sign of an object, to deduce any other truth than that which it explicitly signifies, it is necessary, in all cases, to replace that sign by an icon.” (Peirce 1931, §2.279).

11Lambert (ibid. 3.34) considered arithmetic number systems more perfect as figurative signs than symbols, emblems, and even Egyptian hiero|glyphics, but it was Husserl (1891, 250ff.)6 who offered the most profound discussion of the semiotic structure of numerical systems as developed by Lambert along the lines of Leibniz stimulated by Locke. The naming of numbers is more than a question of mere nomenclature. There is also more involved than the extension and increase of knowledge, than the extension of the natural numerical series through signs beyond directly and intuitively grasped sums and the disclosure of systematic relations between individual numbers and groups of numbers. Husserl, as Lambert before him, pursued practical interests in addition to the psychological and theoretical cognitive ideal. The calculating methods yielded or at least simplified by arithmetic sign systems are of prime concern in their evaluation. This pragmatic approach explains why Roman numerals gave way to the Indian system of position in which the numerical series is based upon a basic unit (10) with each decimal assigned a certain position (columns, in written form). Thus, in the Indian system the arithmetic operations of addition, subtraction etc. for complex numbers can be separated into single, easily grasped stages in which the operations in question are carried out successively for each column.

Every numerical system grounds mechanisms of calculation appertinent to it, and the best system is surely the one that allows the shortest and easiest ones. From this point of view, those systems are especially advantageous whose base number is divisible by as many other numbers as possible and whose addition and multiplication tables are not all too taxing on our memories. For this reason, mathematicians consider the duodecimal system superior to the now established decimal system. (Husserl, 1891, 267 f.).7

12From Leibniz to Peirce, algebra is considered “the most perfect illustration of the [figurative] characteristica" (Lambert, 1764, § 3.35). The algebraic equation is an icon; it uncovers a multitude of relations among the quantities represented by signs. Geometric and physical phenomena are captured by a system of signs “through whose skilful connection all possible truths may be brought to light” (Wolff, 1719, §324). The choice of algebra again combines cognitive and practical interests. To determine the path of a body in free fall, we can measure each unit of interest to us and set up a table of results. We proceed as if the results were not related to each other. Quite another picture emerges when we focus on the relation between each of the measurements. The relation no longer rests on a simple rule of association, a relation of similarity or contiguity, but instead takes the form of an algebraic equation. This distance covered by a falling object in t seconds is (gt/2)2. The lengthy table with data on each second can be replaced by this short, handy formula.8

13Modern science owes much of its progress to the development of ever new mathematical theories for which a representation or interpretation is subsequently found in nature. Mathematical theories are the most striking illustration for the cognitive or—in the nomenclature normally used in this context—heuristic value of semiotic systems.9

14For Lambert (1764, §3.58) and Peirce (1931, §2,279), only figurative or iconic signs are so structured as to further cognition.10 Comprehension of arbitrary signs necessitates a “theory derived not from a picture but directly from the thing itself” (Lambert, 1764, §3.58). However, figurative or iconic signs also depend upon ordinary language interpretations for their exact determination (ibid. 3.33 and 3.58 f.). Verbal determination of a sign replaces or reinforces contextual determination. We can usually tell from the context whether, for instance, the stylized front view of an automobile refers only to the front view or to the whole car. This drawing in a spare parts’ catalogue would probably call for the former inter|pretation, while on a traffic sign restricting through-traffic, it calls for the latter interpretation. Adequate legal certainty, however, can be obtained only by the verbal commentary of the author. Yet even this linguistically mediated certainty does not satisfy a radical philosopher. The meaning of words and sentences in ordinary language is influenced by experience, perceptions and actions which are themselves not less undefined than the iconic signs whose determinations is being sought.

15To what extent is language itself an iconic sign system and consequently one that furthers cognition? Among the older philosophers, the most positive reply to this question comes from Leibniz (1765, §3.7.6), their most linguistic and structuralistic representative: “Je croye véritablement, que les langues sont le meilleur miroir de l’esprit humain, et qu’une analyse exacte de la signification des mots feroit mieux connoistre que toute autre chose les operations de l’entendement.” This statement concludes a passage in which Leibniz uncovers a general meaning underlying the uses of the English term “but”, for which Locke claims as many different meanings as there are uses. For Leibniz the uses of “but” (“but to say no more”, “I saw but two planets”, etc.) all mark a non plus ultra. According to Peirce (1931, §2.280), the iconic mode of designation on the lexical level has all but disappeared. Logical icons on the other hand are to be found in the syntax of every language (cf. Jakobson 1965d, 350 ff.).

16Today, the interpretation of language as a wellspring of logical and ontological cognitions with which the human mind is equipped prior to all philosophical reflection is advanced above all by ordinary language philosophy and generative semantics. Every child reveals with his language competence that he has unconsciously mastered insights, that have often escaped generations of professional philosophers.

17This can be illustrated by the terms “belief” and “knowledge”.11 Both are modes of consciousness distinguished by an intuition so vague that philosophers have been known to attempt a definition of knowledge as belief. The vagueness of our intuition stands in contrast to the conspi|cuously divergent use of “belief” and “knowledge” in ordinary language, thus defeating every attempt to reduce the one term to the other (cf. Tillman 1967, 38 f.; Ryle 1974, 7 f.).

18Knowledge is forgotten; a belief is lost. It is possible to speak of a loss of knowledge, but this refers not to a known fact but to a potential, or a faculty of knowledge such as a library destroyed by fire or those centers of the brain in which our knowledge is actually founded. Belief is a capacity or an attitude. As such it can only be lost but not forgotten.

19We can of course both stop knowing and stop believing something because we have learned better. But when 1 stop believing that Descartes was a linguist, I can still state retrospectively: “Until the day before yester|day, I believed that Descartes was a linguist.” In the case of knowledge, however, I cannot say: “Until the day before yesterday, I knew that Descartes was a linguist.” The change of knowledge modifies the judgment of my former state of consciousness; the change of belief entails no such consequences. Semioticians are often accused of knowing nothing about pre-Lockian history. There is in fact a wealth of semiotic literature from both Antiquity and the Middle Ages, not to mention non-European, Indian contributions. My thesis is, nevertheless, that Locke represents a caesura. From a structural point of view, ancient and medieval studies may well surpass much found in modern and contemporary studies. From a functional point of view, however, Locke clearly ushers in a new epoch.12

20In Antiquity, from Pythagoras (Proclus 1908, §16) to Augustine (see 1952, §36), cognito rei preceded cognitio verbi.13 Medieval Schoolmen set up a sequitur hierarchy beginning with modes of being of an object, followed by modes of cognition, to which modes of designating and signifying were applied and adapted.14 The words orthotes onomáton and congruitas sum up the treatment of the epistemological problem in ancient and medieval semiotics. The question reads: Does the sign match the thing? Negative findings were attributed to the inadequacy of language or classified as linguistic fictions (nominalism).15

21Since Locke and Leibniz, the relationship between modi intelligendi and modi signandi/significandi has been reversed. First Locke drew attention to the role played by signs in the cognitive grasp of objects, whereupon Leibniz proposed that it was not designations as such but rather an ordered systematic mode of designation that paves the way for cognition. In medieval terminology, modi significandi act as modi intelligendi.

22The reversal of the relation of foundedness between designating and knowing led to a revaluation of verbal and semiotic fictions. Fictions not directly matched by a perceivable reality or an intuitively accessible ideality were no longer shunned; on the contrary, they were thrust into the foreground as constructions furthering cognition. A theoretical grasp of relations between perceivable and concrete data is in many cases possible only with such constructions.

23The mathematician Lambert was the first to succeed in proving the consistency of non-Euclidean geometry. As a semiotician, he was no longer concerned with the immediate evidence of his theses. He was content with a sign system built upon “secure rules”. Society now systematically exploits or reveals this insight into the dependence of consciousness on the so-called external designation. The language of advertising and politics provide the most striking illustrations. In West Germany, not only those who were actually driven out by Soviet occupation but refugees who fled in anticipation of possible banishment are officially registered as expellees. Thus, they profit from the moral justification of compulsory displacement and from the social status and economic privileges of genuine expellees. East Germany, on the other hand, seeks to suppress the negative experiences of the past by underscoring the current status of this group with the designation “new settlers” or “new citizens.”16

24This point is well illustrated by the Jew in Max Frisch’s Andorra (1961). A foundling is thought by the people of Andorra to be a Jew. As the play progresses, the connotations typically associated with this designation in anti-semitic circles are projected onto him, until he finally begins to behave accordingly, displaying a sharp intellect, lack of warmth and trust, obsession with money, broken relationship with native country, etc. After he has been murdered, the people of Andorra learn that he was not a Jew after all. The image one has of a person itself creates what it claims only to reproduce.

25Science treats its linguistic constructions as such and usually considers their existence justified exclusively by the instrumental value of indirectly mediated cognition. Language coiners in the political and business worlds, on the other hand, adhere bona sive male fide to pre-contemporary semiotics according to which signs merely provide names for that which by its nature is accessible through preverbal channels as well.

The semiotic transformation of philosophy

26In view of the pragmatic evaluations of semiotics, the philosopher is confronted with two questions: 1) Is all cognition mediated semiotically or are there regions directly accessible to us, such as perception? 2) What about signs themselves? Are systems and means of designation given to us as a neurological-physiological mechanism, as a habitus—in Ryle’s nomenclature a knowing how or in Husserl’s nomenclature an anonymously functioning intentionality - in any case, as something of which we have no direct, objective knowledge (knowing that)? The extent of the proposed semiotic transformation of philosophy depends on the reply to the first question; its radicality on the reply to the second.

27The development of Husserlian philosophy may serve as a guide in answering these questions. Husserl’s Philosophy of Arithmetic (1891) is perhaps the most comprehensive study on the semiotic conception of the idea of number as launched by Locke and Leibniz. Ten years afterward in his Logical Investigations (1901), Husserl astonished his teacher, Brentano (and probably still astonishes anyone aware of the significance of sign systems for cognition), by advancing a theory of cognition with Platonic undertones focused exclusively on immediate insight into the essence of things and “state of affairs” (Sachverhalte). Toward the end of his life, he wrote a highly stimulating essay entitled—upon posthumous publication in 1939 “On the Origins of Geometry”. Therein he takes up ideas first advanced by Locke (cf. 1690, §4.1.8.) and Leibniz on the inevitability of complementing intuitive knowledge by symbolic knowledge in the pro|gress of science: advances in knowledge, made by such highly developed sciences as mathematics, are intrinsically bound up with a successive semiotization and thus with the sedimentation of antecedent intuitive cognitive steps removed from direct evidence.

28This position in this essay is a development and extension of his original position, but neither here nor in his Philosophy of Arithmetic does he subscribe to an absolute semioticism. The mode of numerical cognition can be divided into three stages. The smallest whole numbers (a maximum of twelve, according to 1891, 250) are accessible to us by direct intuition. The genesis of this direct intuition of numbers, however, rests on the sensory perception of objects. In other words, cognition is possible without the mediation of signs but not without a sensory base as a point of departure. Thus, the thesis of a direct intuition of ideal objects such as numbers is relative although neither the inner structure nor the spatial-temporal order of the sensory substratum plays a constitutive role. Once number concepts have been abstracted, the sensory substratum can be dispensed with. The constitution of higher numbers, as we have said, requires systematic semiotic mediation.

29Between the intuitive and the semiotic grasp of numbers, there lies an intermediate mode of cognition which rests on a figurative moment or, in the more familiar terminology of Gestalt psychology, on a Gestalt quality. A specific Gestalt quality enables us to grasp a multiplicity of concrete objetcs without actually counting them, such as a lane of trees, a row of columns, or a flock of birds; likewise, we can estimate a set of numbers on the basis of figural moment. However, according to Husserl, we cannot make accurate judgments beyond a set of five elements. Other authors increase this limit to an average of twenty and, in special cases, e.g. with dice or dominoes, experienced people are said to grasp up to forty units at a glance. Figural moments belong to the psychological aspect of numerical ideas while semiotic apprehension is based on their logical aspect (Husserl, 1891, 244 f., 287ff.).

30Husserl’s rejection of his seemingly modern operationalist conception of mathematical cognition, according to which mathematical entities and operations are founded in an appropriate mode of designation, was apparently motivated by the insight that the mode of designation itself must be systematic and rule-governed, i.e. that it rests on a cognition whose fundament is a direct act of evidence, an intuitive datum.17 Furthermore, the application of Husserl’s analyses of consciousness from mathematics and logic to all fields of cognition reveals that perception, i.e. the fundamental form of consciousness, can be taken neither as a (mere) consciousness of signs or pictures nor as an objective interpreting of several data.

31In a picture, appearance and that which appears are not the same. The red paint on the canvas is not a property, a part of the rose depicted there. When I look at the real rose, however, the color in which it appears to me is a part of it without which it would not be what it is (and not merely a sign for a property not perceived directly. A thing is given (although never completely) in the perception of its appearance and not signified as something that is “in itself” independent of its appearances (Husserl 1931, 99).

32The naive phenomenological concept of perception as an objective interpretation of sense data, i.e. as a semiotic act, is based on the contradictory notion of amorphous sense data. Form is intrinsic to the most primitive sensory data, even if it is only the formal relation of difference between figure and background. It is not the result of an interpretive act. The data of a field of perception refer to each other intentionally through the associative relations of similarity, contiguity, and contrast. On the basis of this network of references, they form manifold phenomenal entities. Associative references are not semiotic relations; they only constitute the basis for potential semiotic relations. A house refers to the village in which it appears, but this does not mean that it functions ipso facto as sign for the village. Nor do disconnected straight lines placed more or less at angles of 45° become a triangle because of an innate idea which enables the percipient to interpret the straight lines as sides of a triangle. Lines refer beyond themselves. The tendency to extend beyond their ends is intrinsic to their perception. Similarly, the Gestalt tendency to become a closed figure is intrinsic to open figures such as loosely disconnected lines. The shaping of perception is never primarily a matter of intellectual interpretation (Holenstein 1972a, 146, 294).

33Husserl’s reasons for maintaining a cautious attitude toward semiotics are supplemented today by a third reason for approaching a semiotic transformation with reserve, a reason deriving, surprisingly enough, from the philosophy of language. Speech-act theory cleared new territory for semantics by questioning the classical conception of meaning. Besides the designative functions of naming and describing, verbal utterances can also order, ask, promise, greet, etc. There are verbal utterances whose primary function consists of constituting a social situation. Nevertheless, even when the foremost intention of a performative utterance such as “I promise to come” is a social commitment, it still involves a designation, albeit an unusual one, namely an auto-reference. The performative utter|ance designates at the same time that which it constitutes. In this respect, it is similar to numerical designations through which higher numbers acquire an intersubjectively accessible existence.

34In his later work, Husserl increasingly reflected upon the gaps in the evidence with which theoretical states of affairs are given to us. He returned to Locke’s insight (although he does not expressly mention him) that a highly developed science such as geometry would be impossible if at every stage of research we wanted to reactivate in full evidence all the preceding stages. Since a higher-leveled problematic not only succeeds earlier phases but is also founded in them, it necessarily shares their deficiencies in evidence (Husserl 1939, 373).

35The entwinement of directly intuitive and semiotically transmitted cognition applies, in my opinion, not only to highly developed sciences but also to perception and motor action, upon which the opponents of a wholly linguistic conception of human consciousness and correspondingly a semiotic-linguistic transformation of philosophy base their arguments. Illustrations of a partial precedence and independence of perception and action over language are not hard to find. The obvious asymmetry between polar terms such as “up/down”, “front/back”, and “right/left” corresponds to an equally obvious asymmetry in the fields of perception and action. Something that is up is generally more visible and more mobile. What lies down below the surface of the earth is out of sight. In terms of visibility and availability, what is in front also has precedence over what is in back. And finally, most people seem more comfortable using the right hand than the left. In a metaphorical context, the spatial terms “up”, “front”, and “right” rather than “down”, “back”, or “left” are used in positive state|ments. A participant in a contest who chooses the right direction may come out in front and on top (up). A person with two “left feet” falls back(wards) and down.

36The dependence of mathematical insights on action or rather acts of coordinated behavior is strikingly illustrated by Piaget’s (1970) example of a mathematician’s childhood discovery. The child had lined up a number of pebbles and was astonished to find that there were always ten pebbles whether he counted them from right to left or from left to right. There|upon he arranged them in a circle and again counted the pebbles in both directions. There were still ten. The number remained constant no matter how he arranged the stones. Thus, through a series of actions, the young mathematician had discovered the law of commutativity, namely that the sum is independent of the order of its elements.

37This process is imbued with semiotics. 1. Probably even a trained intuition can grasp a number of only ten units without a numerical sign, under optimal conditions. The optimal conditions are not likely to be fulfilled when a comparison of several counts is involved rather than the mere grasp of one isolated numerical set. 2. Instead of counting arran|gements of pebbles appearing in the perceptual field simultaneously, successive arrangements of pebbles were compared with each other. A directly given arrangement is confronted with others given indirectly in memory or in planificatory anticipation, i.e. semiotically, as a (mental) scheme (perhaps even more abstractly but nonetheless semiotically, in the form of a rule of spatial arrangement). 3. The conclusion is generalized. The various arrangements are taken to represent x possible arrangements.

38 Without a project or semiotic anticipation, we are capable of only the simplest actions, namely those triggered by the objects of perception through reflexes or association. In their study of children with speech impediments, Luria and Yudovich (1968) demonstrate that children cannot handle behavioral sequences of any length unless they are—in this particular case—linguistically able to formulate a strategy. Action becomes a graduated unfolding of a plan. The individual objects are no longer subjected to random treatment depending on the momentary situation in which they enter the child’s field of vision. Rather, the objects preserve a constant meaning throughout the duration of the action. Their meaning does not depend on the imme|diate phase of action but rather on the entire project. (Blocks, for example, are not put away because the act of throwing them is in itself pleasurable, but in order to make room to set up a train). The progress of the action is constantly checked and evaluated in terms of the project. (“Is the tunnel dark inside?” “No, it isn’t dark enough”, etc.)

39It is more difficult to determine to what extent perceptions are mediated semiotically. The difference in describing a picture—the linguistically underdeveloped child merely names some of the objects in it, while his more skilled counterpart describes the states of affairs that he perceives (ibid.)— does not necessarily indicate progress in perception. It testifies above all to progress in the mastery of descriptive means.

40The benefit of verbal predication as well as of the algebraic equation lies not so much in a faithful record of perception as it does in enabling us to depart from perception by transforming its structure and varying its elements. Instead of saying “Saint Christopher carries the child”, I can say “The child is carried by Saint Christopher”; instead of “across the river”, I can say “across the street”. Semiotic formulation affects perception, as may be assumed, by passively yielding greater receptiveness to unexpected variations and actively stimulating imaginary modifications of the field of perception.

41If we wish to adhere to the insightfulness of our own use of signs, then a transcendental, i.e. cognitive, phenomenological transformation of semiotics is quite as urgent as postulating a semiotic transformation of transcendental philosophy. On the other hand, investigations such as those by Luria and Yudovich demonstrate the need for caution in proclaiming the existence of domains that are “non-verbal”, as Husserl did of perception, Piaget of behavior, and hermeneutics (Ricoeur 1974, 21) of art and play. The borderline in all these fields between a purely intuitive and a semiotic-linguistically mediated performance of constitution still remains to be found. There is much that indicates not only a fluid borderline but also one that is drawn very early.

The planificatory function of signs

42I started out with signs whose primary function lies in constituting cognitive entities such as numbers. In the course of the reflections, brief mention was made of signs such as promises and declarations, whose primary function lies in instituting social relations. Finally, a third function of signs was discussed, the planning of actions. This leads us to the current philosophical tendency to focus less on the theoretical questions posed at the beginning of this paper, “What can we know?” and “How do we acquire knowledge?” and more on the pragmatic interests and consequences of knowledge. Linguistic theory is exploring the implications of the speech act for the addresser and its effects on the addressee of the message. The elucidation of the semiotic implications of praxis seems to me, however, still more fruitful than the elucidation of the pragmatic implications of semiotics. The way for such research has been paved by Soviet Russian psycholinguistics and the cybernetic sciences from automata theory to biology.

43A productive approach to this topic is supplied by a phenomenon neglected by current theories of praxis, namely apraxia. The studies on aphasia, that have proved to be so illuminating for a general theory of language should stimulate increased investigation of apraxia. In motor apraxia, as opposed to paralysis, physical mobility remains intact. Unlike agnostically disturbed behavior, in which the projective representation of the action to be performed is impaired, we can find cases of apraxia in the narrow sense of the word which manifest a perfectly intact “ideational preparation of action.” One of the apraxia specialist H. Liepmann’s (1905) patients was able to execute an action with his left hand that he could not perform with his right. He was also able to formulate and communicate his representation of the action (“the formula of movement”). Liepmann explains this disability as a deficiency of automatic movements normally triggered by intermediate imaginations of a goal. This explanation leads Merleau-Ponty (1945, 161) to criticize classical theories of action for their intellectualism. Not every anticipation of an action need take the shape of a representation before it becomes automatic. According to Merleau-Ponty, Liepmann’s patients suffer from an impaired motor intentionality which he describes as “being in the world”. In the interests of clarity, one should, in my opinion, dispense with the terminology of existential philosophy. Liepmann’s patient suffers from a deficiency of associative references. Every sense datum points beyond itself to its environment and thus motivates our vision or our motor organs to attend to this environment (Holenstein 1972, 116). In fact every movement points beyond itself to a potential continuation.

44The analysis of this case of apraxia and Merleau-Ponty’s charge of intellectualism reveal a distinction vital to semiotics. I would call intel|lectualistic a semiotics that interprets all associative references as a relation between a signum and a signatum (and empiricist a semiotics that reduces all relations to a merely associative relation). Black refers to white by association, without, however, functioning as a sign for white. Associative references found and motivate sign relations, but not every reference is ipso facto a sign as well. A sign relation consists of the intentional represen|tation of one entity by another.18

45Premack’s (1971, 817) now famous chimpanzee Sarah matches up both the plastic chip that she has learned means apple and the apple itself with the pieces of plastic for red (vs. green) and round (vs. square), although the plastic chip for apple is neither round nor red; it is a blue triangle. She has grasped that something with the physical properties of other things is not to be taken as a thing with these properties, but rather as a representation of something else, i.e. as (word)-sign. She seems to have learned that there are categorial distinctions based not on the physical properties of things but on their use—in this case, their highly abstract use as signs.

46The criterion for determining that something is acting as a sign is not individual introspection but an intersubjectively observable substitution. What is to guarantee, however, that Premack’s chimpanzee has not simply been conditioned to produce the same response to the sign for apple as to the apple itself, just as Pavlov’s dog learned to react to a light signal in the same way as to food? Unequivocal proof of genuine understanding rather than mere associative transferral can only be supplied by the use of compound signs, which can be exchanged for other, differently structured signs according to certain laws of transformation. The structure of signs is not significant in the substitution of one simple sign for another (“autumn”—“fall”). For compound signs, however, structure is the decisive factor in determining the admissibility of a substitution. A child cannot grasp that a cube measuring 2 x 4 x 6 cm has the same volume as one measuring 2 x 3 x 8 cm, until he has mastered and comprehended the rules governing the relations between the properties of the cube—its length, width, height—and that the shortening of one dimension is compensated by the lengthening of another. Similarly, he demonstrates his understanding of a sentence by admitting only those modifications that preserve its meaning (“a is larger than b”—“b is smaller than a”—“a surpasses b in size”, etc.). The modification of one element implies rule-governed modifications of other elements. Compound signs are understood when the rules governing their translation into each other are mastered and when the relations governing the individual elements of a sign system are grasped.

47Understanding must not be misconstrued as a static grasp of relations. A relation is not grasped in isolation but only as part of a more comprehensive relation, as congruent with an identical relation, or in contrast to a different relation. Thus even the introduction of simple signs involves acts of translation. The referent is usually entwined with something else that could be intended with the same pointing of one’s finger or the same iconic representation. Indeterminacy can be reduced by positive and negative definitions and circumscriptions: “That is red. ‘Red’ does not designate a spatial extension. ‘Red’ is a color.”—“The photograph in the hairdresser’s window depicts the hairdo in fashion and not the person modelling it.”

48Intralingual transformations are as well as interlingual translations usually involve a slight shift in meaning. Correspondingly, the mere mastery of all possible transformations does not suffice for the complete understanding of a sentence. Total understanding is demonstrated by the ability to comment upon the extent of the semantic shift in such transformations or by the situationally appropriate use of transformations.

49By substituting “translatable” as the determination of the signatum for the classical determination “intelligible”, as suggested by Jakobson (1965d, 345) following Peirce (1931, §4.127), a semiotic conception of understanding replaces the mentalist conception, which has been explicated psychologically as empathy and biologically as assimilation. Understanding thus becomes an intersubjectively accessible operation.

50Significantly, not only the transcendental-hermeneutic concept of understanding is explicated as translation but another fundamental concept as well, the concept of life,19 neglected by philosophy since Bergson and Driesch.

51Molecular biology has extricated two processes basic to all life. According to the one process, a nucleotide sequence which specifies a gene is replicated when two strands of the sequence along which the nucleotides are arranged in a rigorously complementary order split, whereupon the unit complementary to each nucleotide is reconstituted. This replication, a tautological repetition, underlies the stability of the species, which often spans millions of years. According to the other process, genetic infor|mation consisting of the arrangement of nucleotides in groups (words) of three units each (letters) is translated into a sequence of amino acids, the building blocks of proteins, by pairing each triplet (word) of the nucleic sequence with a specific protein unit. These so-called codons, consisting of three nucleotide bases, obey “syntactic” laws of distribution in forming more complex units, cistrons and operons, whose start and end are marked by three special codons with no corresponding amino acid, comparable to the delimitative devices well known in phonology since Trubetzkoy (Monod 1970, 119f.; Jakobson 1974, 51).

52The build-up of the organism as a whole is ultimately regulated by the genetic information contained in the composition of the amino acid sequences in proteins. In contrast to a machine whose design depends upon the planning of an independent mind and the action of external forces, a living being is created by an autonomous, self-regulating morpho|genesis.

53In addition to regulation through protein structure, upon which all is founded, there are constantly new self-regulating homeostatic processes on the macroscopic levels of biology such as physiological homeostasis in the maintenance of a constant body temperature or, on a still higher level, ecological homeostasis in the oscillation of the population of predatory animals and their victims. These homeostatic processes can be divided into two subprocesses, i.e. into two distinct effects. The primary process results in the production of something, warmth for instance. The secondary process, seen in isolation, can also be described as production. In a homeostatically regulated heating element for example, it can be described as the thermodynamic expansion of a chemical substance, which at a certain point of expansion triggers a mechanical process that in turn influences the main process by stopping or accelerating it. Instead of de|scribing this performance as thermodynamic and mechanical production, it can from a holistic and teleonomic point of view be described as “control”, as a process that regulates the main process according to a functionally more or less meaningful program.

54In contrast to conscious behavior, biological processes of a homeostatic nature make no distinction between the signs in which the information for executing a process is constituted and the process itself. The information is already translated in the arrangement itself, i.e. in the course of a process, as in a paternal slap in which communication of what the son deserves and the execution of the punishment coincide. The son can deduce the “message” transmitted by his father only after the fact from the act itself. Likewise, in the case of physiological homeostasis, science can derive the “information” transmitted to the heating agent by the biological process that functions as a “thermostat” only after the fact from the effect of the thermostat on the heat source.

55Processes of understanding and life processes can both be explicated as translation. The differences, however, must not be obscured.

56Understanding is an intentional process. The translation, as which it can be defined, is reversible, unlimited and non-arbitrary. “A is larger than b” can be translated into “b is smaller than a” and conversely. There is no limit to the possible circumscriptions of these sentences. On the other hand, translation is not random; it is bound to the semantic content of the sentence and its components.

57The process defined as translation in molecular biology is purely mechanical. On the whole, biologists hold that translation is irreversible, finite and arbitrary. There is no process in which a certain arrangement of proteins can ever act in reverse upon the arrangement of nucleic units. Corresponding to each nucleic unit there is only one single protein unit and not a chain of possible alternative translations. There is no chemical reason in the form of a steric affinity for pairing a nucleic unit with one particular protein unit and no other. Thus, phonological theories of word formation and word assignment are a more suitable model for the process of translation in molecular biology than the processes explicated on the linguistic level as translation and transformation. The formation and assignment of words is basically arbitrary. The fact that an oak tree is called an oak and not a maple is not based on any affinity between the designated tree and the phonemic elements that constitute the word “oak”.

58Homeostatic processes of feedback are also mechanical processes that are finite inasmuch as closed, merely repetitive operations are involved, and arbitrary (according to prevalent opinion) inasmuch as they owe their genesis to chance and merely their continued existence to the functional advantage they have to offer.

59Modern philosophy was motivated to explore semiotics by the dis|covery of the cognitive function of signs, their ability to promote, extend, and pinpoint knowledge. In current scientific theory, a different function of signs has come to dominate: the planning and control of actions.

60In modern times, the natural sciences have prevailed in the philosophy of science imposing their models on the humanities.

61With the planning and control function of signs, a semiotic discipline has become the first of the humanities to reverse this relation of foundedness and to supply a humanist model as the base for several natural sciences and technical disciplines. This radical change in turn had a liberating effect on the humanities themselves. The computer sciences had no small influence on the supersession of behaviorism by cognitive psychology (Neisser 1967, 8f.). Their highly differentiated, non-mechanical notion of the machine proved a fruitful model for the operation of the human mind, of which it was originally a rough copy. Computers are physical systems whose operation leads not only to processes which, being physical achievements, can be described in physical language, but they also lead to processes which can be described in humanist language as elaborations of information and as control of behavior. As in phonology with its distinction between phonetics, which investigates the physical and physiological properties of sounds, and phonemics, whose target is their linguistic function, the computer sciences can distinguish between an etic discipline aimed at the physico-mechanical aspects of the computer and an emic discipline aimed at its computing performance.

62We must not forget that the transmission of information in self|regulating systems is achieved by totally mechanical means. Viewed in isolation, the informational process is mechano-causal; only the holistic, teleonomic concept of its function is revolutionary. Furthermore, as already mentioned, current biological theory holds that teleonomic processes, inasmuch as they can be found in nature and do not have their source in human production, like automatic machines, are of completely and utterly fortuitous origin. Nevertheless, the humanist who believes in the primacy of Gestalt (form) and meaning (function) over mechanical causality may be consoled—and encouraged—by the fact that apparently even in the realm of nature, “creations” distinguished by perfection of form and high sensefulness are superior to other formations which make no contribution to increased order in the cosmos. The criterion for selection in a fortuitously initiated evolution is functionality.

Appendix

63What still remains to be clarified is the role of the communicative function. The communicative function appears as an independent function when signs are posited for no other purpose than to establish or prolong contact between addresser and adressee (Malinowski’s phatic function). However, it can also be of service to other functions. Semiotic communication offers considerable advantages on both the cognitive and planificatory-cybernetic planes.

64On the cognitive plane, it has been shown that higher-level cognitions require the semiotic presentation of the cognitions upon which they are founded. Through intersubjective written and oral communication, scientists have access to semiotic versions of sciences, upon which they can build directly without having to find and follow the long path of scientific development themselves. When a scholar wishes to check the path already cleared by science, the semiotically transmitted tradition provides him with one clue after the other on how to “bundle” the infinite multiplicities encountered in perception and thought to the best advantage.

65On the planificatory-cybernetic plane, a more primal form of commu|nication, sexuality; can demonstrate the role of semiotic-verbal communi|cation. Before sexuality appears in evolution, every genetic program is a faithful copy of one single program, from which it has split off and according to which it regenerates itself. The possibility of variation is restricted to mutations determined by the accumulation of quanta-like disturbances. With obligatory sexual reproduction, variation becomes integrated into programming. Every new program springs from the combination of two older programs. With each new generation, new combinations are not only possible, but compulsory (Jacob 1970, 330 f.). Similarly, linguistic communication can yield a wealth of new programs, which would be unthinkable without it. In contrast to sexual communication, its realisation is voluntary; and not every realisation represents a gain. In contrast to nature, the mind recoils from disastrous realisations.

66For the philosopher, the subordination of the communicative function to the cognitive and planificatory-cybernetic functions cannot be the final word. The ultimate aim of all human activity is the human subject or rather the communicating society of human beings. This hierarchy explains the repugnance towards realizable but disastrous programs of action. With the subordination of the other two functions, communi|cation becomes meaningful and is no longer mere contact for the sake of contact as in the exclusively phatic function of language.

67On the genetic plane, the communicative function also holds first plane. The primary motive for learning language is its suitability as a means of contact and communication. But very early, already in the so-called egocentric language of the child, the cognitive and planificatory function becomes evident as an increasingly independent aspect of language. It has been demonstrated that the child uses language more and more for the cognitive and planificatory mastery of a task when difficulties emerge demanding conscious insight and deliberation. Egocentric language is still closely connected with communicative language, of which it is an offshoot. It resembles communicative language as far as vocalization and grammatical structure are concerned and usually occurs only in the presence of other people who, it is assumed, understand the utterances. Egocentric language is then gradually internalized and ultimately replaced by inner language at about the age of seven. Compared with external language, inner language displays an extremely elliptical structure on the phonological and grammatical levels (Vygotsky 1934).

Notes

• 1 For a survey, see Coseriu 1968ff.; Jakobson 1975. For an older, diversified, systematic and also historical presentation of semiotic problems, see Vol. II of Gomperz’s Weltanschauungslehre, entitled Noologie (1908). Gomperz began a correspondence with Husserl upon sending him Vol. I (1905). Vol. II is sometimes (Jakobson 1965, 345) considered a possible link between Saussure and older philosophical semiotics.
• 2 “The third branch of science may be called semeiotike, or the doctrine of signs... the business whereof is to consider the nature of signs, the mind makes use of for the understanding of things, or conveying its knowledge to others” (Locke 1690, §4.21.4). (The final italics are the author’s).
• 3 Locke usually places communication second (§3.3.20; 4.21.1 and 4). An exception is §3.5.7, where communication is called “the chief end of language.”
• 4 “La mémoire seroit trop chargée, s’il falloit retenir un nom tout à fait nouveau pour chaque addition d’une nouvelle unité. C’est pourquoy il faut un certain ordre et une certaine replication dans ces noms, en recommençant suivant une certaine progression.”
• 5 See 1.201 and 209ff. The dotted line indicates that the extension is undefined.
• 6 Husserl’s most interesting contribution to semiotics is not to be found in his 1890 manuscript “On the Logic of Signs (Semiotics)”, first published in 1970, but rather in the last three chapters of the Philosophy of Arithmetic (1891). In his manuscript headed “Semiotics”, Husserl develops a classification of signs based on a similar classification undertaken by Bolzano. In his Wissenschaftslehre (1837), Bolzano, on the other hand, does not present his classification in the chapter entitled “Semiotics” in the table of contents and “Theory of Signs” in the body of the text (§4.637ff.). Here he deals broadly with questions of applied semiotics such as what new signs introduced by logic should be like. His classification of signs appears in a paragraph (ibid. 3.285) devoted to the “Designation of our Ideas”.
• 7 Not all sign systems are equally practical for both human and mechanical calculators. Thus, the dyadic number system is obviously less suited to the human mind and memory than to the electronic computer.
• 8 This example stems from G. Miller and is taken from a paper by Jerome S. Brunner (1964, 59) significantly entitled, “Going beyond the information given.”
• 9 According to Locke (1690, §3.3.11), ideas are signs no less than words. The crux of the matter would be grossly neglected if this were taken to indicate a naive mentalism. Expressions such as representation, imagination, perception, thought, etc. all have a subjective and an objective sense, a distinction strictly and consistently maintained in philosophy only since Husserl’s critique of psychologism. Subjectively, these expressions mean the act of representation, imagination, perception, thought etc.: objectively, they mean the intentional object of these acts, the represented, the imagined, the perceived, the thought, etc. as such, i.e. the state of affairs which is represented, imagined, or thought. Such a state of affairs can also be a mathematical theory. What may be considered untenable in Locke is not so much his mentalism as his “imaginativism”, his pictorial view of abstract ideas, which gives rise to Berkeley’s inquiry about measuring the angles of the abstract idea, triangle.
• 10 Wolff (1729, §952; cf. 1733, §97) also discusses the cognitive value of the natural signs that Peirce was to classify as indices. “In specie Signum demonstrativum dicitur, cujus signatum praesens: Signum prognosticum, cujus signatum futurum est: Signum denique rememorativum, vel Memoriale cujus signatum praeteritum est.” The terminology suggests that an older tradition is involved in this classification.
• 11 These two terms will be useful to elucidating Augustine’s view of the function of signs. See fn. 3.
• 12 The opposite thesis is maintained by — among others — Coseriu (1968, 136): “Thus it can also happen that, at the end of the seventeenth century, John Locke seems to be declaring something entirely new when he speaks of the necessity of semiotics. In reality, he hardly says anything novel by pointing out the necessity of semiotics — as if it had never existed before and was yet to be established.” With my thesis, I do not wish to exclude completely the possibility of isolated steps taken in Antiquity and the Middle Ages in the same direction as Locke’s interpretation of signs. With a bit of historical diligence, forerunners can be found for any idea. But the question remains: What value is assigned to such an idea; is its effect on the author himself or on his followers heuristic and systematizing, paradigmatic in the sense of Thomas S. Kuhn’s conception of the history of science, or does it fall by the wayside, unexploited?
• 13 “Cum enim mihi signum datur, si nescientem me invenit cujus rei signum sit, docere me nihil potest: si vero scientem, quid disco per signum?... Itaque magis signum re cognita, quam signo dato ipsa res discitur” (ibid. 33). — Augustine’s argumentation rests on a thoroughly atomistic doctrine of signs. The fact that compound signs (complex numbers) and combinations of signs (premises, conclusions) can yield new cognitions is either overlooked due to the fixation on the correspondence between thing and sign or repressed by interpreting what is gleaned from statements of others as mere belief and not knowledge. Thus, according to Augustine (ibid. 37), the story of the three youths in the oven does not add to my knowledge. It only provides me with something to believe. What happens, however, when I hear another version of the deliverance of the three youths, which for reasons internal or external seems more convincing, or when I learn from a biblical scholar that the story is not historical fact but didactic legend? I lose my belief in the factuality of the events in the story. The state of affairs constituted by the story still remains a datum of knowledge as a state of affairs in itself. In Augustine, the relation between knowledge (scire), evidence (intellegere), and belief (credere) is not satisfactorily elucidated. Every belief implies an affair-complex and knowledge that can be true or false in reference to a particular real or possible world. The question is whether the human mind would be capable, without the support of signs, of comprehending and remembering a complex story like that of the three youths in the oven (quite apart from the linguistic semiosis which the story itself contains in the form of reported speech) as a coherent whole.
• 14 “Modi essendi seu proprietates rerum seu entium praecedunt modum intelligendi, sicut causa effectum... Modum autem intelligendi sequitur modus seu ratio signandi, quia prius intelligitur res et etiam concipitur antequam per vocem signetur quia voces sunt signa passionum... Modum autem signandi sequitur modus significandi sicut rem sequitur modus rei” (Sigerus de Cortraco [Siger von Kortrijk], see 1913, 93 f.).
• 15 Locke retains this orientation in Book III of his Essay. The direction semiotics has taken in modern times, however, does not spring from Book III but rather from the chapter on the idea of numbers in Book II.
• 16 This example is taken from Lübbe 1975, 141ff.
• 17 This line of thought is clearly expressed in a critical letter from Brentano to Husserl (9.1.1905): “Is not the discovery of differential calculus also and especially the discovery of a methodological procedure, so that even the positive determination of a certain mode of designation by Leibniz, superior to that of Newton, proved a great step forward? Even the invention of the computers could be included. — True, one does not simply learn to add, one also learns individual laws of the equation of several addends with one sum, e.g. 2 + 5 = 7...” Husserl’s transcendence of Brentano’s semiotic operationalism could be compared with Carnap’s later transcendence of Bridgman’s metric operationalism. The designation and measurement of data are preceded by a “theoretical” conception of data which yields the framework for possible operationalizations.
• 18 “A l’intention d’une Psyché s’est substituée la traduction d’un message” (Jacob 1970, 10).
• 19 For some additional aspects of a semiotic restructuration of philosophy, see Holenstein (1976, 134ff.).