A semiotic approach to mind and culture

The semiotic model and meaning

As defined in the field of semiotics, and in particular the work of Ferdinand de Saussure (1916/1983), a sign is a unit of meaning comprised of a signifier and a signified. The signifier is the form which the sign takes, for example, a word on a page, a trill of music, a gesture. The signified is the mental concept to which the signifier refers in the mind of the beholder, and includes both the denotative (literal, definitional) and connotative (personal, emotional) content. The spoken word “Hi,” for example, is a sign consisting of a sound /hai/ and a concept of greeting, which may include (depending on the tone of voice) a connotation of friendliness, concern, or indifference. The sound signifies—refers to, represents—the semantic content, and the resulting ensemble is a sign or meaning-experience: “Hi.”

A basic tenet of semiotic theory is that the meaning of any signifier is not intrinsic, but depends on a process of interpretation. The same signifier—/hai/—which serves as a greeting in one context, may in another, such as a ski resort or an amusement park, point to the concept of elevation, with connotations of excitement or fear; or in yet another, such as a department store, point to the concept of expense, with connotations of surprise or longing. This effect is even more striking in the context of a different language. In Japanese, for example, the signifier /hai/ may, depending on the situation, represent the affirmative “yes,” a drinking glass, or the lungs. The same signifier stands for a variety of concepts, giving rise to a variety of signs, depending on the social and cultural context of the individual who interprets it. And what is true for words is no less true for images, acts, scents, tastes, textures—anything that refers to something else, and thereby produces meaning for an observer.

Although Saussure (1916/1983) emphasized the arbitrary link between signifier and signified, he recognized that communication would not be possible without a sign system based on a consistent, predictable relation between the two. This relation could not be left to individual whim, then, but was rooted in social convention and consensus: certain signifiers point to certain signifieds because the people of a given linguistic community agree that they do. Subsequent theorists have analyzed a wide range of signifying practices, examining how messages or “texts” in such diverse areas as art, literature, or mass media are shaped by cultural conventions and historically transmitted assumptions about the world (Barthes, 1991). The premise of this research is that the meanings of various texts are not intrinsic to the texts themselves, as if waiting to be noticed and understood, but result from an active construction on the part of the viewer, based on the broader context of signs in which those texts and the viewer are situated. Or, in Saussure’s terms, texts are not signs, but signifiers; and signifiers only become signs—meanings—inasmuch as they point to a signified in someone’s mind.

What remains unanswered, however, is the basic psychological question of how this “pointing to” takes place, and how this process is informed by the surrounding context of signs. Saussure, invoking 19th-century British associationism, described the relation between signifier and signified as an “associative link” (1916/1983, p. 66). The resulting sign, though an indivisible, self-contained dyad, derived part of its meaning—what Saussure termed “value”—from its relation to other signs, and in particular its difference from other signs in the same general category. The meaning of “salty,” for example, could never be understood in isolation from the rest of the English language, or from related words such as “sweet,” “sour,” or “bitter.” Yet this is an observation about signs proper, after they have been constructed, rather than about the construction process itself.

In contrast to Saussure’s dyadic model, the American philosopher Charles Peirce (1978) described a triadic model of signification consisting of the representamen, object, and interpretant. The representamen is the form of the sign, the thing which “stands for” something else, and is comparable to Saussure’s signifier. The object is what the representamen stands for, and in Peirce’s scheme this is an actual referent in the world, the thing itself—material object, event, mental state—rather than a concept of it. The object of the word “zebra,” for example, is the real animal rather than, as in Saussure’s model, a mental concept of the animal. The third element in Peirce’s scheme, the interpretant, is defined as an interpretation or elaboration of the representamen-object relation, for example, the idea that the word “zebra” refers to an animal with black and white stripes.

Peirce viewed the interpretant as a sign in its own right—not something that the representamen stands for, but rather an idea that makes the “standing for” relation of representamen and object possible. Prompted by an encounter with the representamen, the interpretant causes the representamen to signify a particular object in the world. If a knock on the door is the representamen, for example, then the object might be the deliveryman waiting outside with a package, while the interpretant is the idea that a knock on the door means the deliveryman has arrived. For a detective at a crime scene who finds fingerprints on the furniture, the object might be the criminal’s identity (or the criminal himself), while the interpretant is the understanding that fingerprints are unique to each individual.

Peirce’s model thus obscures Saussure’s distinction between signifier and signified, and fails to explain the “pointing to” relation between the two. For Saussure the signifier can never be known in and of itself, but only as it points to a concept, the signified. Yet in Peirce’s model what the representamen (signifier) points to is something out in the real world, which may or may not be present, and which may or may not be immediately grasped. In the case of the knock on the door, for example, in Peirce’s model the object is not yet known, and the problem is to determine, by means of interpretants—that is, ideas about knocks—what the object might be. This process Peirce called “abduction,” a kind of inference in which rules from the broader sign system are applied to the signifier to find out what it signifies.

Clearly, however, in this case the signifier is not a signifier at all, but a fully fledged sign—a “knock”—as are the interpretations about what it might mean, and the object it refers to (e.g., an as-yet unidentified person). To be comprehended at all a knock must have semantic content; it is not a knock without pointing to a signifying concept in the mind of the person who hears it. The resulting sign may itself point to something in the world, for example to a person behind the door, but this referent is not the same thing as the signifying concept that makes the knock a knock—a sign—in the first place. It is the conception of things, as Langer (1957) noted, not the things themselves, that signifiers directly mean.

Indeed the distinction between representamen and sign is often lost in Peirce’s model, with the terms being used interchangeably, and this confusion is not merely semantic but theoretical. What Peirce’s model ultimately explains is what may be termed “secondary” reference, that is, the reference of one sign to another, rather than “primary” reference, or that between signifier and signified, upon which the sign proper is based. Secondary reference is a more or less deliberate process of inferring what signs stand for, such as the imagery in Edgar Allan Poe’s “The Raven” (1845), or a rise in the unemployment rate. The object can be conceived independently of the sign—the raven represents the poet’s memory, for example; higher unemployment represents an impending recession—because the object itself is a sign, or unit of meaning. Primary reference, by contrast, is the process by which a sign becomes a sign in the first place, a process without which its basic components—signifier and signified—are unknowable. As Myers (1996) argued, Peirce’s model does not explain reference in this primary sense, but rather reference as a “relation between referring signs” (p. 29). A sign refers to an object by referring to the interpretant, which in turn refers to the object, but this rationale takes the signifying function of these various components as a given. It explains how signs can be about other signs, but fails to explain the “aboutness” or “essential intentionality” of the signs themselves (Myers, 1996, p. 29).

This “aboutness,” then, must entail the two components of the sign—signifier and signified—and it must involve something more than what Saussure described as an “associative link” (1916/1983, p. 66). Saussure employed this term from British associationism uncritically, with the result that his model leaves the basic process of signification unexplained. Associationism assumes that the ties of relatedness that make up meaning are extrinsic to the subject, reflecting inherent properties of the object world. Learning is a process of registering discrete, atomic entities from this world, which associate themselves mechanically according to laws of thought such as frequency or temporal contiguity. While the classical exposition of this view is the familiar S-R (stimulus-response) model of orthodox behaviorism, the S-O-R (stimulus-organism-response) or so-called mediational model of cognitive psychology rests on the same objectivist premises. This model too assumes that inputs are representational from the beginning, as are the knowledge structures by which they are manipulated and the outputs they produce. Cognitive processing can be complex, involving multiple operations with low levels of predictability, but the ultimate effect is to replicate ties of meaning intrinsic to the object world.

This conception is clearly at odds with the basic semiotic premise that the signifier has no intrinsic meaning, and that it is up to the subject to produce a “pointing to” relation between signifier and signified. How can the subject do this? For a signifier to point to a signified, it needs to be directed there by a subject’s purpose: by the anticipation of a future outcome or end state to which the “pointing to” relation serves as a means. This is readily illustrated in the case of secondary reference, or that between a sign and an object. When Edgar Allan Poe used the symbol of a raven to suggest memory, he did so with the purpose of conveying a certain idea in the context of his poem. When someone yells “Fire!” in a theater, they do so with the purpose of warning others of an actual fire. In either example the “pointing to” effect of the sign is based on an anticipated state or outcome. For Poe this outcome is some kind of understanding in his reader; for the person in the theater, the evacuation of everyone inside. The sign (raven, “Fire!”) refers to an object (memory, actual fire) as a means to these anticipated outcomes, which can thus be said to determine what the signs point to in the first place.

This same principle applies to primary reference as well, or that between signifier and signified. Consider the signifier of thunderclouds on the horizon. To someone planning a picnic, this signifier, in addition to the common denotation of a cloud mass foreboding rain, might represent an unwelcome change in plans, with feelings of disappointment. To a farmer growing wheat, by contrast, the same thunderclouds might represent a welcome source of water, with feelings of relief. The signifier (thunderclouds) points to different signifying concepts (disruption of plans; source of water) in the service of different preexisting goals (going on a picnic; watering one’s crops), such that individuals with different goals will experience different thoughts about what they see. Signifier and signified are inseparable, as Saussure (1916/1983) held, and this is because signifier and signified only become what they are in relation to each other and to a particular goal of interpretation. To “mean” means that the signifier is doing something for someone, that it has a function; namely, referring to something for the sake of someone’s purpose or end.

The role of purpose or “telos” in the act of signification is a central issue in semiotic theory (Cornejo, 2010; Hulswit, 2002; Rosa, 2007). Short (1981, 2007), among others, has offered a teleological interpretation of Peirce’s model, arguing that Peirce viewed triadic action as a goal-oriented or “intelligent” action, in which some event (A) causes another event (B) as a means to an outcome (C). This Peirce contrasted with the “brute” action of dyads, in which A produces B, and B produces C, but C has no effect on the likelihood that A will produce B. Yet Peirce himself never explained how such teleological relations could be applied to the triadic model of sign, object, and interpretant (Short, 1981). Although presumably the sign points to the object as a means to some purpose, the third term in Peirce’s model—the interpretant—is not a purpose at all, but a concept or definition of what the sign refers to. Peirce’s semiotic triangle thus fails to capture the goal-oriented nature of a triadic as opposed to dyadic relation.

Toward a new model: The telic triad

For the semiotic model to depict thought itself, then, it must show the teleological nature of primary rather than secondary reference: the reference that makes up a sign, rather than the reference between a sign and an object external to it. It must therefore show how a goal or purpose guides the “pointing to” relation between signifier and signified, in essence transforming Saussure’s dyadic model into a triadic one. Let us call this third term the intendant, and define it broadly as an intention, in the ordinary sense of an aim, goal, or purpose, such that it always entails a future state or outcome in the world, and a subject who in some way anticipates that outcome. The intendant is not a thought in itself, though it is typically based on prior thoughts. The farmer need not be thinking about crops to see thunderclouds as a welcome source of water; nor does the picnicker need to be thinking about the picnic to see those clouds as a hindrance. Indeed the signifier in this case—thunderclouds—is only known inasmuch as it points to one signified or another, which is to say that the effect of the intendant involves no intervening steps of inference. The intendant must therefore be latent and preconscious, akin to a memory, and must in this form guide the relation of signifier and signified in the construction of a sign, or conscious meaning. Each individual possesses an entire field of such intendants, and this forms the basis of a personal point of view from which the world is experienced.

Signifier, signified, and intendant thus make up what may be termed a telic triad (see Figure 1a). “Telic” means “tending toward an end,” and the point of this model is that end-directedness is intrinsic to the sign itself—to the experience of meaning, or primary reference—and not merely to the conscious manipulation of signs, or secondary reference. The signifier, in pointing to the signified, in some way answers to—satisfies, confirms, questions—the intendant, and thereby facilitates the goal or future outcome it entails. The contingent nature of this relation is marked by a dotted line: if the intendant changes, so will the signified. To a paleontologist, an exposed cliff-face appears as a tableau of geological time, while to a rock climber the same cliff-face appears as an ascent with a certain degree of difficulty. The semiotic model is thus “pragmatic” in the sense that a given signifier can assume widely different meanings, as well as different implications for behavior, depending on the purposes of the individual concerned. OPEN IN VIEWER

The end-directedness of the telic triad is what distinguishes it from the computational model (see Figure 1b). For all its advances, this model essentially substitutes the concept of information input and output for the earlier behaviorist concept of stimulus and response. “Information” in this model is not information in the ordinary sense of weather reports or the latest gossip, but information in the technical sense of a correlation between two objects that results from a lawful rather than accidental process, and produces some kind of physical change in one or both objects. Processing of this information, as with the flow of code in a computer program, follows the logic of algorithms, or step-by-step rules and procedures, which manipulate the input efficiently and without the need for conscious intervention. However complex, this is ultimately an objective, third-person process, one which can in principle be broken down into antecedent-consequent dyads. Such dyads, in the absence of a third term or intendant, never amount to a “pointing to” relation, and thus never amount to a sign in the first place.

The cognitive model is not without a concept of goal, and various lines of research have examined the influence of goals and motives on perception (Gollwitzer & Moskowitz, 1996; Kruglanski, 1996). Within a computational framework, however, goals affect the manipulation of precoded information rather than the reference process by which signifiers take on meaning. Goals themselves are understood as knowledge structures or “schemata,” the function of which is to mediate, in sequential algorithmic fashion, between informational input and behavioral output. Once primed by environmental stimuli, goals influence any of the various stages of information processing—encoding, memory storage and retrieval, synthesis—in such a way as to facilitate satisfaction of the goal. Goals may activate other knowledge structures, for example, by which external stimuli most relevant to the goal are encoded and integrated into judgments about how to act (Kruglanski, 1996). As with other forms of information processing, however, these operations are objective, cause-effect events with no special connection to the first-person experience of the subject. Indeed a primary thrust of this research is to show how goals operate “automatically,” unfettered by conscious deliberation (Bargh & Chartrand, 1999).

The cognitive model would nonetheless be hard put to ignore the concept of meaning, and theorists often use the terms “representation,” “meaning,” or “intention” in place of “information.” Yet there is never any suggestion that information processing itself causes meaning; there is no particular step in the encoding-storage-retrieval sequence that causes input to “stand for” output. For there to be meaning in this model, then, it must be intrinsic or “preassigned” to the information. The notion that goals are activated by situational stimuli, for example, and then direct cognitive processes toward fulfillment of the goal—all without conscious awareness—assumes that these various types of information have innate representational power (Deacon, 1997). This assumption in turn stems from the basic associationist premise that the ties of relatedness that make up meaning are intrinsic to the object world, and that units of knowledge—what John Locke (1689/1975) termed “simple ideas”—derive their meaning from a natural correspondence to things in that world. Like “simple ideas,” the units of information in cognitive theory are representational by virtue of their link to veridical properties of the environment, a premise described by Lakoff (1987) as “computational realism.” Cognitive processing in effect manipulates units of information with the ties of relatedness already attached, rather than constructs meanings by building those ties anew.

The telic triad offers a model of this constructive process, showing how goals intervene in the reference process proper. The “pointing to” relation of signifier and signified involves more than a cause-effect dyad: signifier points to signified as a means to the goal, or intendant, and this in turn requires more in the way of goals than objective knowledge structures. The intendant is a “that, for the sake of which” signifier points to signified; not a concrete future event but rather the anticipation of a possible outcome. As an anticipation, moreover, the intendant is inseparable from the first-person perspective of a subject who does the anticipating. When someone makes a plan—to visit Kansas City, to grow strawberries, to take up scuba diving—the goal serves as a standard by which means are chosen and actions taken. This is the sense in which conscious purposes can be understood as “final causes”; they reflect a general possibility or end state “for the sake of which” the subject thinks and acts. The point of the telic triad is that goals and purposes can serve as final causes not merely on a conscious level, in the guidance of thought and behavior, but also on a preconscious level, in the formation of meaning itself.

This raises the question of how, if meaning depends on a subjective experience of reference, a goal could operate outside of awareness. Numerous contemporary philosophers who support the essential link between meaning-making and consciousness have appealed to the “connection principle” as a way to incorporate preconscious intentions in the explanation of behavior (Ludwig, 1996; Searle, 1992). The key assumption of this principle is that any mental state, to count as a mental state, must in some way be connected to the first-person, experiential aspect of mind, either as a conscious thought or as a preconscious state with the causal power to produce such first-person experiences. At any given moment most of one’s memories, wishes, or purposes are not present to mind, yet they remain accessible and capable of influencing current thoughts. This is not to say that such preconscious intentions are fully formed, autonomous meanings lurking below awareness, like fish below the surface of the water, because in that condition they lack the experiential, first-person perspective essential to any referential relation. Preconscious intentions are understood rather as latent intentions with a “dispositional causal capacity” to bring about subjective experiences (Ludwig, 1996; Searle, 1992). Even outside of awareness, goals, wishes, and expectations retain what Searle (1992) termed an “aspectual shape”—a particular content as grounded in the subject’s point of view—and this aspectual shape persists as a disposition of the brain with the power, should the right conditions occur, to produce conscious thoughts with a similar or related aspectual shape.

All of this fits with the concept of intendant as a preconscious goal with the potential to guide the “pointing to” relation between signifier and signified. By the present view there is no meaning apart from a triadic relation among signifier, signified, and intendant, such that these three elements are what they are only in relation to each other. In particular the “pointing to” relation of signifier to signified would not be possible without a purpose or end state in the service of which the “pointing to” relation serves as a means. As a preconscious intention, then, the intendant is best understood not merely as a disposition to think or act in a certain way, but as a disposition toward an end. This follows because, though not a fully fledged meaning itself, the intendant is based on prior conscious meanings—purposes, expectations, motives, desires—all of which involve an object or outcome for the sake of which the subject acts.

A detective arrives at a crime scene, for example, with the conscious goal of solving the crime, and various behaviors—searching for physical evidence, taking pictures, interviewing witnesses—all serve this goal. He then gets a call from headquarters and begins talking about a different case, when he suddenly sees a stray fiber on the counter as a possible clue to the criminal’s identity. Though not immediately present to consciousness, the goal of solving the crime nevertheless operates as a disposition toward an end—an end in the service of which the signifier (stray fiber) points to a signified (clue). Consistent with the connection principle, the latent intention—to solve the crime—has a “dispositional causal capacity” to produce conscious thoughts. It has an ongoing, ever-ready potential to become manifest in consciousness in the form of meanings conducive to its realization.

A further premise of the connection principle is that preconscious intentions produce meanings grounded in the subject’s perspective, and the intendant—by guiding the relation of signifier and signified—does precisely that. The intendant, after all, entails not only a future outcome in the world, but also a subject who anticipates that outcome. When the detective sees the stray fiber as a clue, the latent disposition by which the signifier-signified relation is framed—to find the criminal’s identity—is a disposition not toward an abstract, impersonal end, but toward an end conceived by and involving the subject. The resulting meaning—“clue”—though not about the detective himself, is nonetheless experienced by him as “something to me”—experienced from an inward, first-person perspective.

The telic triad thus helps explain one of the perennial mysteries of subjectivity: how thoughts feel personal and “owned” even when those thoughts are not about the self at all. In contrast to introspection, in which the self is a direct object of awareness, subjectivity is the immediate, pre-reflective “acquaintance” with one’s own mental processes, the unobserved yet ineliminable sense of “I-ness” accompanying all conscious states. Meanings come in many forms—the familiarity of an old song, the cold taste of ice cream, the clarity of a new idea—but all have a subjective or phenomenal character, a sense of “what it’s like” to experience them from within. This is what Smith (1989) termed the “egocentric structure of experience,” an immediate awareness of self as subject “built right into the experience itself” (p. 88). The experience of “this song” is always “I hear this song” even though “I hear” is not part of the object of the thought.

In semiotic terms, then, the question of subjectivity is how the signifier-signified relation conveys an immediate awareness of self even when the resulting sign is unrelated to that self. The answer lies in the fact that the signifier-signified relation is always defined by a third term, or intendant. We have an immediate familiarity, or direct acquaintance, with our own point of view, because all signifiers, in referring to a signified, refer to that very point of view—to a field of goals, expectations, or purposes derived from prior experience. All thoughts are indeed “owned” by us in the sense that the signifiers on which thoughts are based only become signifiers by virtue of pointing to, and thus being assimilated by, this dispositional field. Where intendants are widely shared, signifiers tend to have similar meanings, and this is the basis of conventional understandings and behavior. When intendants are unusual or idiosyncratic, however, the result may be an entirely new way of understanding a given signifier, as in the case of creative discovery.

Gutenberg’s invention of the printing press, as described by Koestler (1989), is a good example. In late medieval Europe, books were still written by hand or made by carving text on a block of wood and rubbing paper over it. Gutenberg wanted a faster method, and he found it one day while watching a wine press, a traditional device used for pressing grapes by turning a large screw over a flat tray. In the moment of insight Gutenberg saw this device in a different light—as a means to press paper onto a bed of inked type. For everyone else who shared the goal of making wine, the wine press was just a wine press, too ordinary to cause much notice. But for Gutenberg, whose goal was entirely different, the wine press stood for something new. The signifier—the physical apparatus of the wine press—referred to a signified—printing press—as a means to the goal of finding a better mode of printing. At the crucial moment this goal was a preconscious intention—a latent disposition toward an end—but it was Gutenberg’s intention, based on an outcome envisioned by him. The resulting insight was, for Gutenberg, “something to me,” an inward creative glimmering, and this sense of ownership allowed him to pursue his idea and turn it into an invention that revolutionized printing in Europe and the rest of the world.

The telic triad and intentionality

As with many creative acts, Gutenberg’s insight concerned an innovation that did not yet exist. But this in no way prevented his insight from being “about” something—from being a bona fide meaning based on a referential relation between signifier and signified. The telic triad thus clarifies certain long-standing questions about the nature of intentionality, questions stemming from a definition of intentionality as the ability of mental states to represent actual objects or states of affairs in the world. The thought we have of Mount Kilimanjaro, for example, represents the real mountain, while the thought we have of banana bread represents the real food. By this logic, however, some mental states do not seem to represent or be “about” anything, such as pains, itches, or moods. And some thoughts, such as those of mythical lands or fantastical beasts, don’t have anything in the real world to be about. How can a thought point to unicorns, the question goes, if there are no unicorns in the world to point to? How could Gutenberg’s idea be about a printing press, if printing presses did not yet exist? In light of these problems, it is commonly assumed that intentionality describes some thoughts, but not all.

Another approach to this question, based on the work of the German psychologist and philosopher Franz Brentano, defines intentionality more broadly as a mental “directedness” on objects (Crane, 1998). Brentano (1874/1995) argued that intentionality is the mind’s ability to “refer” or be directed to objects—images, ideas, sensations—in the mind. All thoughts are intentional, Brentano held, because all thoughts are about something: “aboutness” is a matter not of referring to something beyond the mental act, but of referring to something within it. Just as in perception something is perceived, so in emotion something is felt, and in sensation something is sensed. A similar logic applies to signs having no external referents. There may be no unicorns in the real world, but this doesn’t keep a thought from being directed on unicorns. Thoughts may not point beyond themselves, in other words, but they still point.

Thus defined, Brentano’s thesis hinges on a semantic distinction. Describing intentionality broadly as “directedness” suggests that all thoughts are intentional, while describing intentionality more narrowly as “representation” seems to rule out thoughts lacking an external object. Another approach, known as “representationalism,” argues that even experiences like pain or emotion are representational (Lycan, 1996; Tye, 1995). Pain is usually associated with specific parts of the body, and by this view it represents harm or disturbance at those sites. Yet critics have argued that the experience of pain does not necessarily include the idea of damage to the body. Headaches, for example, typically don’t mean damage to the head, so viewing headaches as intentional on this basis is unwarranted. The representationalist approach also has difficulty explaining how nonexistent entities like unicorns can be the referent of a thought (Siewert, 2003).

The telic triad helps explain Brentano’s notion that every thought “includes something as object within itself” (1874/1995, p. 88), and moreover shows how all thoughts are representational. It does this by distinguishing between signifier and sign, and between primary and secondary reference. The question of intentionality has traditionally been framed in terms of secondary reference, or that between a thought—a sign—and something outside of it—another sign. But this is to neglect the primary reference—that between a signifier and signified—on which the thought itself is based. A thought is “about” pumpkin pie, for example, not because there happens to be a referent in the real world, but because the signifier points to a signified within the thought itself, as directed by an intendant. By the time we experience the signifier of pumpkin pie—an image in our minds, for example, or a word on a restaurant menu—it already does represent something—a perfect snack, 320 calories, something to indulge in or avoid.

The same is true for pains and moods, as well as nonexistent entities like unicorns. To ask “What does a headache represent?” or “What does a unicorn represent?” is to miss the point that, by the time they occur as thoughts, the signifiers of headaches and unicorns already do represent something. A headache may be interpreted as a sign of stress, for example, or the symptom of illness. The pain itself may be felt as sharp or dull, or focused in a certain part of the head, but in any case the headache is known only as it connotes something to the subject. The signifier of unicorn is likewise known only as it stands for something, for example mythical creature, object of beauty, or nonexistent entity frequently invoked in philosophical discussions. The point is that the question of intentionality is not whether the thought represents something beyond itself, but whether the signifier of a thought does. And the answer to that question is always yes, because signifiers cannot be known otherwise.

Indeed the structure of this intrinsic representation is such that the signifier does point to something in the external world, even in the absence of obvious external referents. This follows from the nature of the intendant, understood as a goal by which the “pointing to” relation between signifier and signified is defined. Such goals are by definition interactional; all involve a motive, wish, or expectation based on the individual’s relation to the world, and in particular to a future state or outcome in the world. Every signifier, in pointing to a signified as a means to a goal, points to the goal as well, and thus to a future state or outcome in which the signifier is entailed.

For humans, moreover, this world is invariably mediated by symbols, as are the goals and purposes needed to survive in this world. In directing the signifier toward a signified, then, these purposes direct the signifier toward the broader symbolic context in which the signified—and its connection to the signifier—is grounded. To someone who sees a slice of pumpkin pie as 320 calories, for example, this purpose might be “to go on a diet.” To someone who sees the same signifier as a perfect snack, this purpose might be “to live life to the fullest.” Note that the very notions of “dieting” or “living life to the fullest” are entirely observer-relative; they exist by virtue of cultural consensus, and indeed in some cultures these notions don’t exist at all. Precisely the same is true for the signifieds “320 calories” and “snack”: they have no natural, innate reality apart from conventional understandings. Yet this feature does not prevent “320 calories” or “snack” from being part of the external world. The symbolically mediated world too counts as the external world to which thoughts can be directed, for indeed this is the only world humans can know.

By this logic even signifiers like headaches or unicorns point to something in the external world. In the case of headache, for example, connotations such as “sharp,” “throbbing,” or “dull” depend on a linguistic context that delineates such distinctions as well as makes it possible to think of pain as a “headache”—a discrete, localized entity—in the first place. Depending on the culture headaches may be attributed to witchcraft, chocolate consumption, changes in weather, or fatigue (Tan, 1997). The same signifier points to entirely different signifieds depending on the stipulative effect of different purposes and goals, but all of these goals are grounded in a preexisting conceptual framework, and all direct the signifier to a signified that belongs to this framework and only makes sense in terms of it. The signifier of “unicorn” likewise connotes different things in different cultures—for example, an omen of good times, a symbol of purity, an antidote to illness—but which of these applies depends on the culturally defined goals of the individual who is thinking about unicorns. There is no philosophical quandary about reference to “nonexistent objects” here; all of these concepts exist as part of the collective imagination of their respective cultures, and the signifiers of headache or unicorn could not be signifiers without them.

Intentionality and mutual constitution

The importance of intentionality in explaining the constructive and social nature of mind was not lost on the founders of psychology. Mental representations are not atomistic entities built up from passively registered associations, Wilhelm Wundt argued, but reflect an active process of representational “synthesis” and the capacity for inward attention (Kim, 2008). Wundt was interested in the evolution of human culture, and he saw this process as a way to explain how the mind both shapes and is shaped by changing historical contexts. For William James (1890/1981) as well, the stream of consciousness is never a passive succession of ideas and feelings, but involves an “active element… which seems to go out to meet” those ideas and feelings (p. 285; emphasis in original). The process of “selective interest” is shaped by experience, James argued, but at the same time shapes that experience “much as a sculptor works on his block of stone” (p. 277).

Cultural psychology, by making the interdependence of mind and culture its signature principle, seems to be moving toward a notion of intentionality as defined by these early thinkers. But with basic processes tied to the notion of a central processing mechanism, many theorists have turned to the culture concept itself to explain mutual constitution—to some special domain of knowledge that gives mind its transformative, constitutive power. “Cultural” knowledge is defined variously as knowledge that is tacit, historically rooted, intersubjectively shared, and codified in language. It is sometimes held to be restricted to arbitrary or “nonrational” propositions, known to philosophers as “synthetic a prioris,” or statements about the world whose validity is established without appeal to experience. Other debates focus on how this special domain of knowledge exerts its constitutive influence, and whether such influence occurs “inside” or “outside” the head.

All this confusion is understandable given what the beleaguered culture concept is being asked to do, namely, convey the mutual constitution of mind and society in terms of the information-processing model. Despite this confusion—or perhaps because of it—the term “culture” has assumed an almost magical status as a catchword of psychic diversity and as a talisman against claims in general psychology that basic categories are innate and universal. Basic categories vary because of “culture,” so the argument goes, but this implies that a system of symbols—however delineated—can operate as a disembodied causal force. This in turn implies that symbols are meaningful in and of themselves, that what signs refer to is intrinsic to them rather than dependent on a process of reference (Linger, 1994). If cultural symbols are intrinsically meaningful, however, it is not easy to see how such symbols are both constituted by—and constitutive of—mind.

The telic triad, as a model of intentionality, offers a way to explain mutual constitution without relying on any particular concept of culture. Intentionality cannot be reduced to a sequence of dyadic relations, as found in the computational model, because to do so is to eliminate its goal-directed, and thus its social nature: signifier points to signified as a means to the subject’s goals and purposes, as grounded in the subject’s experience in a particular cultural context. This “pointing to” process does not depend on any particular domain of content; it’s the way the mind works. Norms, narratives, metaphors, classificatory schemes—all may be internalized as part of the intendants that guide this process. Cultural context “situates” intentional states, as Bruner (1990) wrote, and it does this by defining the intendant, which in turn directs the “pointing to” relation between signifier and signified.

Again, however, cultural context is not an abstract, impersonal entity that exists “out there.” What stars mean to an astrophysicist is different from what stars mean to a medieval astrologer or to an ancient Polynesian navigator, and without this subjective process of meaning-making, stars mean nothing at all. This is not to say that stars have no empirical reality, only that stars can only be known as they are interpreted from a particular point of view. And what is true of stars is true of tables, televisions, and skyscrapers, not to mention less tangible things like coffee breaks and online shopping. That signifiers mean more or less the same thing in a given culture is not because their meaning is intrinsic, but because the people in that culture share the same goals by which those signifiers take on meaning. Changes or idiosyncratic variations in this interpretive framework open the way for novel relationships between signifier and signified, such as when words acquire new meanings in response to changing needs of communication. Human minds and the sociocultural environment are interdependent, then, because the process of reference is inseparable from a particular point of view, and thus from a shared symbol system, but at the same time, this symbol system is dependent on and continually shaped by this process of reference, as experienced by a community of individuals.

Understood this way, psychic diversity is by no means incompatible with psychic unity. Consciousness, as James (1902/1994) observed, consists of both an objective part—the content of what we are thinking at any given moment—and a subjective part—the “inner ‘state’ in which the thinking comes to pass” (p. 542). Psychic diversity corresponds to the content side of consciousness, to what thoughts are “about”: a feeling, a memory, a choice to be made. Such contents vary widely by culture, but ultimately exist only as experienced, only inasmuch as certain signifiers point to certain signifieds to individuals in those cultures. This “pointing to” relation, then, corresponds to the “process” side of consciousness, a universal process underlying psychic diversity in all its depth and nuance. To argue for intentional worlds—worlds not “given” but existing only intersubjectively, by virtue of shared understandings—is to argue for the individual capacity to be directed at things, to have thoughts that refer to or represent things from a personal perspective. This capacity is a prerequisite for adapting to such intentional worlds, regardless of time or place.