Sign and Symbol

The Cratylus is Plato's principal investigation of how names signify. Two positions are set against each other. Hermogenes maintains that names are conventional, so that any sound may be assigned to any thing by agreement among speakers. Cratylus maintains that names are natural, each thing having a correct name that belongs to it by nature. Socrates examines both claims. If names were purely conventional, there would seem to be no ground for preferring one name to another, and the craft of naming would amount to little more than arbitrary stipulation. If names were purely natural, they would have to resemble their objects, a requirement that quickly produces difficulties, since the name "fire" is not itself hot, nor does every sound that names a thing share that thing's properties.

Socrates moves toward a middle account. Names are instruments crafted by a name-maker who attempts to express the nature of a thing in the sounds and syllables of the name. Success is partial. Language reaches toward reality without capturing it completely. In the course of the dialogue Socrates also suggests that a sign ought, so far as possible, to be like the thing it signifies, and that some conventional signs are better than others in this respect. Mental images, which by their nature imitate or resemble their objects, appear on this view superior to spoken words. The dialogue closes with a caution: one should not study names in order to learn the nature of things, but should study the things themselves. Signs are useful instruments of teaching and distinction; they are not transparent windows upon reality. Questions about natural versus conventional naming reappear under the ideas of Language and of Custom and Convention.

Plato thus frames a problem that the later tradition continues to debate: whether the relation of sign to thing is natural, conventional, or some mixture of the two. Aristotle will treat spoken words as conventional symbols of mental experiences that are themselves natural likenesses of things. Augustine will distinguish natural from conventional signs in a theory oriented toward the reading of Scripture. The question whether language mirrors or merely labels what is real remains open.

Aristotle settles the dispute raised in the Cratylus largely on the side of convention, though with an important qualification. Spoken words are symbols of affections or experiences in the soul, and written marks are symbols of spoken words. The sounds and marks themselves are conventional, for different languages use different words for the same things. Nothing is by nature a noun or a name; it is so only when it becomes a symbol by human institution. The mental experiences that words directly symbolize, however, are the same for all human beings, and the things of which those experiences are likenesses are likewise the same. The sign relation thus has three levels: things in the order of nature, thoughts in the mind that naturally resemble those things, and words that conventionally express those thoughts.

Aristotle also limits what counts as a word. Inarticulate sounds, such as those produced by brutes, may be significant, yet none of them constitutes a noun. Such cries, by which one animal calls another or communicates fear or anger, are not strictly conventional signs; they are instinctive modes of expression and therefore natural rather than voluntarily instituted. With respect to natural signs more broadly, Aristotle tends to restrict them to those things which, according to knowledge and experience, permit a necessary or probable inference to something else. That a woman is giving milk he regards as an infallible sign that she has lately borne a child; that a man is breathing fast is only a probable and refutable sign that he has a fever. These distinctions belong as much to the theory of inference as to the theory of meaning.

Within discourse itself Aristotle separates names and verbs, and simple terms from propositions. A name by itself signifies something but makes no claim that can be true or false. Only when terms are combined in a proposition do truth and falsity arise. The analysis thus divides the theory of signs from the theory of truth and assigns each its own domain, a division treated further under the ideas of Truth and of Language.

The middle level of Aristotle's scheme raises further questions. If mental experiences are the same for all and naturally resemble things, then convention encodes a correspondence that is not itself conventional. Hobbes will later deny that there is anything universal except names, and will treat definition as the sole ground of fixed meaning. Locke will return to Aristotle's claim that words signify ideas, while questioning how general names acquire their range of reference. Later discussion often returns to the threefold structure of thing, thought, and word.

Augustine offers one of the most complete early accounts of the sign. A sign is a thing which, over and above the impression it makes on the senses, causes something else to come into the mind as a consequence of itself. He first suggests a threefold division of things: some are simply things and not signs at all; some are things that are also signs of other things, as the ram Abraham offered in place of his son; and some, such as words, are never employed except as signs. Every sign, he adds, is also a thing, for what is not a thing is nothing at all. Within this division falls the further distinction between natural and conventional signs. Natural signs lead to knowledge of something else without any intention of being used as signs: smoke indicates fire, and the footprint of an animal that has passed by belongs to the same class. Conventional signs are those which living beings mutually exchange in order to show the feelings of their minds, their perceptions, or their thoughts. Of these, words hold the chief place.

Augustine's practical interest is the interpretation of Scripture. He distinguishes literal signs, which signify what they directly name, from figurative signs, which signify something beyond their ordinary meaning. Much of Scripture is figurative: the rock in the phrase "God is my rock" signifies not a stone but divine steadfastness. Correct interpretation of figurative language requires knowledge of the original languages, acquaintance with the natural world from which metaphors are drawn, and regard for the teaching of the faith, with charity as the interpretive norm. When a word can bear more than one meaning, the reader must determine from context which sense the author intended. Augustine also notes that some words signify kinds of words rather than kinds of things, as "noun" and "adjective" do, and that the same word may signify itself or something other than itself according to its use in a sentence. These observations anticipate later distinctions between first and second imposition, treated more fully under Language and under Universal and Particular.

Augustine's rules for reading Scripture leave open the question how one knows when a sign is figurative and what it figures. His appeal to charity and to the faith of the Church provides a theological check, though it does not settle every philosophical difficulty about ambiguous signs. Aquinas will develop a more systematic account of the senses of Scripture and of the names applied to God. Later disputes about figurative interpretation, including those of the Reformation, continue to turn on distinctions Augustine first made central.

Aquinas inherits Aristotle's framework and develops it with care. Words are signs of ideas, and ideas are similitudes of things; words therefore function in the signification of things through the conceptions of the intellect. Ideas are the immediate or proximate objects that words signify, and through them words ultimately signify the real things that are themselves the objects of ideas. One can give a name to anything only insofar as one can understand it, so that words cannot signify things without the mediation of concepts. Aquinas is also among the writers most explicit in treating images and ideas as, in their very nature, meanings or intentions of the mind. He calls them "mental words." The exterior vocal sound is called a word because it signifies the interior concept; first and chiefly, the interior concept itself is called a word. Angels may make their concepts known to one another immediately, but men require the medium of external speech.

The most distinctive part of his account of naming concerns the names of God. Can human language, drawn from experience of creatures, say anything true of God? If a word such as "good" applied to God and to creatures univocally, with the same sense in both cases, God would be treated as if he were a creature. If it applied purely equivocally, with wholly diverse meanings, nothing could be known or demonstrated about God from creatures. Between these extremes Aquinas places analogical signification: the meaning is not one and the same as in univocals, yet it is not totally diverse as in pure equivocals. Words such as "good," "wise," and "being" are, on this view, intrinsic denominations when said of both God and creatures, whereas names such as "angry" or "jealous" belong to God only metaphorically. The pattern of meaning resembles that of "being" as said of heterogeneous things more than that of a merely attributive name such as "healthy" said of animal and of medicine. The problem of likeness between infinite and finite being is treated more fully under the ideas of Same and Other and of God.

Aquinas also extends Augustine's concern with Scripture. The signification whereby words signify things belongs to the historical or literal sense. That whereby things signified by words themselves have a signification is the spiritual sense, which is based on the literal and presupposes it. The spiritual sense he divides into the allegorical, the moral, and the anagogical. Divine things are metaphorically described by means of sensible things, as in the sacraments physical things serve as outward signs of inward grace. Metaphors in sacred doctrine he regards as both necessary and useful, whereas in poetry he thinks them employed chiefly for pleasure. Philosophers and scientists often take the opposite view, restricting metaphor to poetry and excluding it from demonstration.

The doctrine of analogy becomes a central resource in Catholic theology and in later philosophy of language whenever speech is asked to describe what exceeds ordinary experience. Hobbes will reject universal natures and treat names as imposed marks. Locke will agree that words immediately signify ideas, yet without adopting the scholastic account of mental words or analogical predication of the divine. The triadic relation of word, concept, and thing continues to structure debate long after Aquinas.

Hobbes treats words primarily as instruments for the recording and communication of thought. They serve as marks by which an individual recalls his own conceptions, and as signs by which those conceptions are made known to others. There is nothing natural in the connection between a word and what it is used to signify. Names are imposed by decision, and their meaning is fixed by definition. Truth consists in the right ordering of names in affirmations. True and false are attributes of speech, not of things; where speech is not, there is neither truth nor falsehood. Without language there would be no truth and no falsehood, and without general names there would be no reasoning of the sort Hobbes describes. These claims place his account at the extreme of conventionalism within the tradition, and connect the theory of signs closely with the ideas of Truth and of Reasoning.

On universal names Hobbes is a thoroughgoing nominalist. There being nothing in the world universal but names, a name is universal when it is imposed on many things for their similitude in some quality or accident. A proper name brings to mind one thing only; a universal recalls any one of those many. Universal names such as "man" or "horse" do not correspond to universal things or to universal ideas existing apart from speech. He also distinguishes kinds of names with care: some are names of the things conceived; others are names of the imaginations or ideas of those things; still others are names of names, as "universal," "plural," and "singular." The abuse of words, their use without fixed definitions or with contradictory meanings, he regards as a principal source of philosophical error, and he is severe toward metaphysical discourse that he takes to consist in meaningless speech.

Hobbes is likewise hostile to metaphor in science. One main cause of absurdity, he holds, is the use of metaphors, tropes, and other rhetorical figures instead of words proper. What is lawful in common speech is not to be admitted in reckoning and the search for truth. In this he stands with those who would purge demonstration of figurative language, against the view, found in Augustine and Aquinas, that metaphor is sometimes necessary for teaching spiritual truths, and against the later observation of Freud that even scientific psychology cannot escape metaphorical expression.

If words mean only what convention assigns, the definitions that ground reasoning are themselves products of imposition. Locke will attempt to tether definitions to ideas derived from experience, hoping thereby to connect language with something beyond mere agreement among speakers. Berkeley will attack the abstract general ideas to which Locke appeals. Hobbes's nominalism continues to oppose any theory that treats general names as signs of natures or of abstract concepts in the mind.

Locke devotes Book III of the Essay Concerning Human Understanding to language, its imperfections, and its abuses. Men, he holds, needed to make articulate sounds and to use those sounds as signs of internal conceptions, making them stand as marks for the ideas within their own minds, whereby those ideas might be made known to others. Words came to be used as the signs of ideas not by any natural connection between particular articulate sounds and certain ideas (for then there would be but one language among all men), but by a voluntary imposition, whereby such a word is made arbitrarily the mark of such an idea. The proper and immediate signification of words is therefore the ideas they stand for. Locke goes further and maintains that words can be signs of nothing else, though men often suppose their words to stand also for the reality of things, not wishing to be thought to talk only of their own imaginations. Obscurity and confusion enter the signification of words whenever we make them stand for anything but the ideas we have in our own minds.

He distinguishes kinds of names according to the ideas they signify. Names of simple ideas are least liable to confusion, because the idea is determined by experience. Names of mixed modes are most arbitrary, because the mind constructs the complex idea. Names of substances are intermediate, guided by experience but not fully determined by it. General words work by abstraction. Ideas become general by separating them from the circumstances of time and place and from any other ideas that may determine them to this or that particular existence. Common nouns such as "man" or "cat" become general by being made the signs of general ideas. What general words signify is a sort of things, and each does that by being a sign of an abstract idea in the mind. The problem of how general names acquire meaning, and of what universality belongs to, is discussed more fully under the idea of Universal and Particular.

Locke is acutely aware that the imprecision of language is a major source of philosophical dispute. Communication succeeds when the hearer has associated with a word an idea similar to that which the speaker intends; it fails when speakers attach different ideas to the same sign, which happens more often than is commonly recognized. Careful definition is the remedy he recommends. Like Hobbes, he often dismisses theories as consisting of meaningless words rather than merely as false. Interest in freeing discourse from the tyranny of unclear signs, which later centuries call semantics, is already present in this part of the tradition.

Locke's ideational theory of meaning becomes the usual account in the British empiricist line. Berkeley attacks the doctrine of abstract general ideas on which the account of common names depends. Hume refines the analysis of particular impressions and ideas. The question whether words primarily signify ideas rather than things continues to generate debate, and later writers on Language and on Idea return repeatedly to the distinction between the voluntary imposition of words and the natural character of ideas.

Berkeley's Introduction to the Principles of Human Knowledge criticizes Locke's doctrine of abstract ideas at its foundation. Locke had held that general words get their meaning from abstract general ideas, produced by stripping particularities from particular ideas. Berkeley finds this unintelligible. He asks whether one can form the idea of a triangle that is neither oblique nor rectangle, neither equilateral, equicrural, nor scalene, as Locke's description requires. Every idea in the mind, Berkeley maintains, is particular and determinate. A word becomes general not by standing for an abstract idea but by being made the sign of several particular ideas, any one of which it indifferently suggests to the mind. An idea which, considered in itself, is particular becomes general by being made to represent or stand for all other particular ideas of the same sort. Generality is thus a matter of representative use, not of an abstract content stripped of all determination.

On this view the attempt to account for the meaning of general names is itself a principal cause of the acceptance of abstract ideas. If there had been no such thing as speech or universal signs, Berkeley writes, there never would have been any thought of abstraction. Men mistakenly suppose that every name has, or ought to have, one only precise and settled signification, and therefore invent abstract, determinate ideas as the true and only immediate signification of each general name. In truth, he concludes, there is no such thing as one precise and definite signification annexed to any general name. Where Locke would say that a common name gets its general meaning by signifying one idea that itself has general significance, Berkeley holds that a general name gets its meaning from a great number of particular ideas, all of which it signifies indifferently. The related issue of what universality belongs to things, words, or ideas is treated under Universal and Particular.

Berkeley's critique has consequences beyond the theory of general terms. If there are no abstract ideas, many philosophical puzzles are said to dissolve, because they arise from treating words as if they named general entities that exist only as abstractions. Terms such as "matter," "substance," and "real essence" are, on this showing, abstract terms without corresponding ideas. Elsewhere Berkeley also tends to substitute the relation of sign and thing signified for the relation of cause and effect among natural things: the fire he sees is not the cause of the pain he suffers on approaching it, but the mark that forewarns him; the noise he hears is not the effect of a collision of bodies, but the sign thereof. Nature itself becomes a language of marks, a theme that belongs as much to his theory of knowledge as to his theory of signs.

Berkeley's attack on abstract ideas reshapes the empiricist theory of signs. Hume takes up the point and grounds his philosophy on particular impressions and ideas. The question what general words mean, if not abstract entities, remains open long afterward, and reappears whenever the theory of meaning is asked to explain how a finite stock of particular experiences supports speech about sorts and kinds.

Kant distinguishes two modes of rendering concepts intuitable: schematism and symbolism. A schema provides a direct sensible illustration of a concept, as a drawn triangle does for the concept of triangularity. A symbol provides an indirect illustration through analogy. When the state is called a machine, or a monarchy a body, the point is not that the state is literally a machine, but that the relation of parts to whole in a machine is analogous to the relation of citizens to the state. The symbol works by transferring a pattern of relations from a familiar domain to one that cannot be exhibited directly. In this respect Kant's account continues the older analysis of metaphor based on proportion, such as Aristotle's example of calling a king the father of his people, while placing that analysis within a critical theory of how concepts are given sensible content.

All our language about God, Kant argues, is symbolic in this sense. We speak of God as a father or as governing the world not because these terms apply literally, but because the relations they describe are analogous to divine attributes as we must think them. Knowledge of God is therefore merely symbolic. The claim recalls the theological problem of the names of God treated by Aquinas under the heading of analogical predication, though Kant's solution belongs to the critical philosophy rather than to scholastic metaphysics. He seeks a middle path between literal application of creaturely predicates to the divine and a pure equivocation that would empty theological speech of cognitive use. Related questions about the limits of discourse concerning God appear under the ideas of God and of Theology.

In the Critique of Judgment Kant also discusses aesthetic ideas: rich representations of the imagination in art and poetry that occasion much thought without any definite concept being adequate to them. A beautiful image may present moral ideas in a way that discursive language cannot. Rather than labeling what is already clear, the symbol makes partially available what no definite concept can fully capture. The productive power of symbolic presentation thus extends from theology into the theory of art and beauty, and connects the analysis of signs with the chapters on Art and on Beauty.

Kant's theory of the symbol offers a path beyond both purely literal and purely arbitrary signification. Later writers develop related distinctions in the theory of art, including accounts of symbolic, classical, and romantic forms. Freud transforms the notion of symbolism again when he treats dreams, slips, and symptoms as meaningful signs of unconscious wishes. The critical insight that some signs present their objects only indirectly, by analogy rather than by direct illustration, remains available for both theological and aesthetic use.

Freud extends the theory of signs into the analysis of dreams, symptoms, and ordinary slips of speech and action. From Augustine's definition of the sign as a thing that causes something else to come into mind, through Locke's and Berkeley's concern with the meaning of words, the tradition had treated signification largely as a matter of conscious discourse and natural inference. Freud treats the relation of significance, intention, and tendency as operating also below the threshold of awareness. A dream is not random noise but a meaningful production. Its manifest content, what the dreamer remembers, is a symbolic representation of latent content: repressed wishes and thoughts that the dream-work transforms through condensation, displacement, and the requirements of visual representability. A single image may combine several trains of thought; emotional charge may shift from one idea to another; abstract thoughts may appear only as scenes and figures.

Freud resists a fixed dictionary of dream symbols. The same figure may mean different things for different dreamers; context and free association determine meaning. He extends the same method to parapraxes, jokes, and neurotic symptoms. In each case a surface sign carries a hidden meaning produced by unconscious processes. Interpretation is the art of reading these signs, tracing them back through the patient's associations and history. Symbolism, on this view, is not confined to poetry, theology, or deliberately instituted conventional marks; it pervades the works of man, normal and neurotic, dreaming and awake. Questions about memory images as signs of absent things, already raised by Aristotle, reappear here in a psychological form and connect with the ideas of Memory and Imagination and of Mind.

Freud is also aware of how thoroughly metaphorical scientific language itself remains. The difficulty of picturing certain psychological processes, he observes, comes from being obliged to operate with scientific terms that are metaphorical expressions peculiar to psychology. Without them one could neither describe the corresponding processes nor even have remarked them. Physiological or chemical terms would themselves constitute a metaphorical language, only one longer familiar and perhaps simpler. In this he parts company with those, such as Hobbes, who would ban metaphor from the search for truth, and stands nearer to those who accept figurative speech as sometimes unavoidable in the exposition of what cannot be stated in wholly proper terms.

Freud enlarges the domain of the sign from logic, rhetoric, and theology into the interior life of the individual. Whatever the later scientific standing of particular interpretations, the claim that symbolic meaning operates outside conscious intention remains influential in psychology, literary criticism, and cultural theory. The older problems of natural and conventional signs, of literal and figurative speech, and of the relation between surface mark and hidden sense receive, in his work, a new field of application.