1. Classic Theories of Communicative Channels

It is useful to diagram a few classic theories of communicative channels, focusing on their similarities in a relatively schematic fashion.Footnote ¹ In figure 1, we see Claude Shannon’s (Reference Shannon1948) famous model: a message from a source gets turned into a signal, which is then transmitted down a channel. Along the way it may be interfered with by noise, such that only a garbled signal gets to the receiver. Such a signal is then converted back into the original message (with more or less fidelity).

Figure 1. Shannon, mathematical theory of communication

In figure 2, we see Shannon’s (Reference Shannon1946) model of secrecy systems, as it compares with his theory of communicative channels: a private message from a source is encrypted, and the resulting cryptogram is transmitted down a channel to a relatively friendly recipient. The encrypted message may be intercepted by an enemy en route, who will attempt to decrypt it, possibly learning the secret.

Figure 2. Shannon, theory of secrecy systems

In figure 3, we see three key factors in Roman Jakobson’s (Reference Jakobson, Waugh and Monville1990) vision of the speech event as these intersect with Michael Serres’s (Reference Serres2007) understanding of the parasite. In contrast to Shannon, the ends of the channel are slightly simplified (the message is not converted into a signal or cryptogram), while the role of the agent perpendicular to the channel is greatly magnified (and multiplied). Such an agent may interfere with a message (such that it doesn’t reach its original destination), or such an agent may intercept a message (such that it is diverted to a different destination). In effect, such an agent is whatever can knock a message (instrument, action, infrastructure, or organism) “off course.”Footnote ²

Figure 3. Jakobson’s speech event (reinterpreted in light of Shannon and Serres) (channel as much psychological connection and social convention as physical contact).

Finally, in figure 4, we see one way to render Freud’s (Reference Freud1960, Reference Freud1999) understanding of psyche, as an internalization of relatively similar communicative machinery. The id sends messages or ideas to the ego; and, in getting past the superego, they are necessarily rechanneled or recoded.

Figure 4. Freud, diagram of the psyche (reinterpreted in light of Jakobson) (message sent from id to ego must be rechanneled or recoded to get past superego).

Complementing these understandings of communicative channels is Charles Peirce’s (EP 1.5–6) account of semiotic processes, a key modality of “thirdness,” as exemplified in figure 5. Here we have a seemingly simple mode of semiosis known as joint attention—such as a child turning to look at where her parent is pointing. Such a process involves an interpretant (the child’s change in attention), an object (what the parent is pointing toward), and a sign (the parent’s gesture that directs attention).

Figure 5. Peirce, semiotic process

In this example, the channel is not so much a material conduit along which signs and signals are sent (stereotypically composed of something like wires or neurons). Rather, it is minimally a threefold phenomenon that turns on physical contact (e.g., a transparent medium in an illuminated enclosure, with open lines of sight); psychological connection (e.g., the child’s desire to know what the mother desires to make known); and social convention (e.g., who is allowed or encouraged to direct whose attention, in what kinds of contexts, to what kinds of objects). Crucially, if interference and interception are the anti-affordances (or epifunctions) of channels, then to study such a threefold phenomenon is, ultimately, to study all the ways communicating agents may have their contacts, connections, or conventions fail, or be otherwise knocked off course.

2. Freud and Grice

We now turn to a particularly important class of communicative practices that turn on the concatenation of two semiotic processes, or “thirds,” of the Peircean sort. To understand such practices, we will need to reinterpret and generalize some ideas of Grice and Freud. But first, we need to introduce a distinction made by Peirce between dynamic objects (whatever causes a sign) and immediate objects (whatever is caused by a sign, or brought to another’s attention through a sign). For example, when I say, “it’s going to rain,” the dynamic object is my intention to communicate, and the immediate object is the content so communicated. Within such a framing, whenever one explicitly states such an immediate object (what one desires to say) by means of a stereotypic sort of speech act, one implicitly shows such a dynamic object (one’s desire per se). In contrast, symptoms are classically understood as having immediate objects that are more or less overlapping with their dynamic objects. In other words, insofar as the doctor has a particular diagnostic, the symptom (say, a high temperature) brings to the doctor’s attention precisely that which causes the symptom (say, a particular disease or illness).

With this distinction in mind, we may radically rework some key parts of Grice’s (Reference Grice1989) classic proposal regarding “nonnatural meaning” (of the “nonconventional” sort). As diagrammed in figure 6, such forms of communication turn on two interlocking semiotic processes of the Peircean kind, one relatively indexical (ostensive or “concrete”), and the other relatively inferential (elliptic or “abstract”).Footnote ³ For example, suppose we are friends, and I have just arrived at your house, and you are about to invite me in. You look at my muddy shoes with a raised eyebrow. From the standpoint of this framework, such a composite sign has a relatively dynamic object (your intention to communicate), and it has a relatively immediate object (my shoes, which is what you are ostensibly looking at). That is, not only do you point out my muddy shoes (through your eye gaze), but you point out that you are pointing them out (through your raised eyebrow). And so, as an interpreter, I not only attend to my muddy shoes, I also attend to your intention to direct my attention to them. This is the first semiotic process, the relatively concrete one, grounded as it is in indexical contiguities, such that its object is relatively available in the immediate context.

Figure 6. Grice, indexical-inferential communication

Crucially, my interpretant is not just to look at, or attend to, my own shoes (immediate object), and attend to your attention to them (dynamic object), it is also to consequently take my muddy shoes as a sign that points to something else in a less immediate, or more abstract, inferential context. In particular, knowing what you pointed out, and knowing that you pointed it out on purpose, may lead me to a hypothesis or abduction (itself a kind of logical interpretant): my friend wants me to take off my shoes, or to wipe them before I come in, or to note that the neighbor’s dog has done it again (such that his complaining last night about his neighbor’s laxness, or the absence of leash laws, was warranted). This interpretant, then, is relatively abstract or inferential—a hypothesis, grounded in my assumptions about minds as much as worlds, as well as the ongoing dynamics of our interaction and its immediate context.Footnote ⁴ It may be more or less proximal (take off your shoes) or distal (forgive my complaining).Footnote ⁵

It is worth pausing a moment to appreciate this insight. People are always going on about the generativity of language, with an emphasis on that key affordance of grammar: finite means (say, a relatively bounded number of words and rules) enable infinite ends (say, a relatively unbounded number of possible sentences). Note, instead, how “generative” a single sign can be (such as the pointing gesture). The emphasis here is not the usual one, that a pointer is a shifter (a conventional means of recruiting any contextually available entity into the role of object). Nor is the emphasis that something like metalanguage is operative: there are signs that take as their objects other signs. The emphasis rather is that any otherwise nonsemiotic entity or event in the world can be turned into a sign (of something else entirely) just by pointing to it; and one and the same entity or event can be a sign of an infinite number of different objects, depending on the context and culture in which it is pointed out.

Such a reformulation of complexly cascading semiotic processes is not only meant to capture classic Gricean processes (while reimagining them in a way that is less dependent on certain commitments of Grice that are easy to critique). In particular, Freudian processes arguably turn on a similar logic (see fig. 7). As this story goes, my dreams not only have a manifest content (what they point to concretely, or conventionally, or iconically). They also have a latent content (what they point to abstractly, or elliptically). And to recover the latent content (i.e., the immediate object of the second semiotic process) from the manifest content (i.e., the immediate object of the first semiotic process), an analyst has to make reference to the dynamic object of the first semiotic process.Footnote ⁶

Figure 7. Freud, dream interpretation

Crucially, such dynamic objects are not communicative intentions, but rather repressed wishes. It is only by knowing (or positing) that a dream was the product (or effect) of a censored desire, itself due to the superego’s parasitic interference with, or interception of, the id’s wishfulness, that an analyst can figure out what the latent content of the dream actually is (recall fig. 4).

Needless to say, such semiotic processes always depend on semiotic grounds: the sensibilities and assumptions agents have regarding the qualities, causalities, and conventions that seem to govern a particular environs, or organize a particular mind, society, or world. (Recall our doctor’s diagnostic.) It is only by reference to such grounds that such agents can signify and interpret in particular ways, insofar as such grounds enable such agents to draw iconic, indexical, and symbolic connections among otherwise disparate entities and events. Such grounds are not necessarily, or even usually, backgrounded, but may themselves be the figures and figurations of semiotic processes. For example, Freud’s own writings, through their wide dissemination and institutional uptake (however erroneous or inadequate), became an important part of many groups’ common ground: not just practicing psychoanalysts, and their (ever internalizing) patients, but also layfolks influenced by pop psychology, Hollywood films, and critical theory.

It should be emphasized that there are three kinds of circularity, or self-justification, in both Gricean and Freudian communication. First, one cannot get to the abstract object without first getting to the concrete object; but, as Willard Quine (Reference Quine1960) argued, it is not all that easy getting to the concrete object in the first place (the so-called inscrutability of reference). So it is usually the very fact that you arrive at an interesting abstract object (as a kind of destination) that assures you that you settled on the correct concrete object (as a kind of origin). Second, in one sense the dynamic object is the initial cause (which leads to the concrete object, which leads to the abstract object); and so, if you assume its existence, you can go very far (inferentially speaking). But, in another sense, it is the very fact that it took you so far that legitimizes your original assumption (as to its existence). And, finally, there is circularity of a more obvious Austinian sort: the very grounds an agent may presume in order to undertake a semiotic process may be produced by the very performance of that semiotic process (Austin Reference Austin1962).

Crucially, at least in the case of psychoanalysis, it is not just an interpretive leap one makes, it is a kind of inferential struggle, whereby one moves more slowly than surely from one inference to another as one tries to figure out what is being abstractly pointed to. Indeed, as seen in the diagram, Freud considered interpretation a kind of work (as do anthropologists, or else we couldn’t justify our salaries, just like psychoanalysts couldn’t justify their fees). His model of “work” was closely aligned with the way the term is used in physics: a summation over forces multiplied by distances. In particular, the forces are resistances, and the distance is the inferential movement one makes against such resistances—from signs to interpretants, from premises to conclusions, and thus from the concrete object to the abstract object. Interestingly, Grice’s theory of communication is also often rendered in terms of an economy, especially when focusing on his maxim of “quantity.” For example, speakers are understood to minimize encoding “effort” while maximizing communicative “effect.” They are often posited as offering the simplest sort of sign (no. 1), qua means, to produce the most complicated sort of object (no. 2), qua ends. From the standpoint of such a hermeneutic, an imaginary is coupled to an instrumentality, the latter to enlist simple indices, the former to license complex inferences.

A key factor in both kinds of communication is prosocial motivation and common ground, albeit in very different ways. For Grice, the addressee can only make such an inference by assuming that both speaker and addressee share a lot of assumptions, and by assuming that the communicative event is a joint action that both parties are—to some degree—committed to and accountable for. Similarly, in the case of Freud, much of the work of the analyst is devoted to establishing common grounds of a more narrow sort (only you know about your childhood, let me know as well, or else I can’t help you), and battling relatively über prosocial forces (because the superego—that key source of resistance—is precisely society as it sets itself up in psyche).

Finally, and only somewhat paradoxically, note the radical pleasure often afforded by such semiotic labor, such hermeneutic work. More generally, note the pleasure offered by many inferential struggles: a mechanic troubleshooting an engine; a programmer debugging an algorithm; a physicist explaining a phenomenon; a detective sussing out an assailant; an anthropologist interpreting a taboo or ritual. Each such domain is (hypothetically) governed by a complex ensemble of causal (qualitative and conventional) relations, and each such agent has a distinctive competence in connecting otherwise disparate events by means of drawing out, and on, such causal (qualitative and conventional) relations. Needless to say, there is often a distinct pleasure in exercising, through praxis, a hard-earned competence, be that competence enminded or embodied, be that praxis inferentially abstract or indexically concrete.

The real pleasure discovered by Freud, thus, is not the wanton lusts of the id, nor the sadistic gratifications of the superego.Footnote ⁷ It is the fact that he offered a productively constrained space to engage in inference, a causal (qualitative and conventional) ground organizing that space, and a promise that such inferences would lead to some significant disclosure (a remedy, a secret, a weapon, a key). Indeed, it is really our tarrying in that space that affords so much guilty pleasure. And this is, arguably, the real guilty pleasure, not so much discovered by Freud as inaugurated by him: look at all the cool critical theory arguments I can make! (But, alas, look at how half-baked they so often are.)Footnote ⁸

3. Angels and Devils (a Brief Aside)

Michael Tomasello (Reference Tomasello2008) and his colleagues at the Max Planck Institute for Evolutionary Anthropology have shown that apes point for imperative ends. In their experiments, for example, apes will gesture to out-of-reach food in order to get humans to give it to them. And apes will even point to a tool that a human would need in order to secure such food for them. (Though they wouldn’t make such a gesture if the same humans were looking for food to feed themselves.) Apes, then, will engage in semiotic processes with humans that resemble a relatively truncated version of figure 5. Tomasello argues that apes never use such pointing gestures with other apes because there is no prosocial motivation to help each other. (In particular, the other ape would just eat the fruit himself rather than give it to whichever ape pointed it out.) And he argues that if the social environment of apes became more cooperative, they would point imperatively to request help from other apes with no additional cognitive machinery needed.

Let me draw out just one amazing entailment of these claims: apes have an incredible ability that remained latent until the moment they were locked up by humans and needed to enlist their help to eat (or to find release, more generally, from the agonies of their experimental subjugation, itself a key means to delimit the boundaries of their semiotic subjectification). This makes me think that there truly are angels and devils in this world. Only they don’t lie at the edge of our human world (popping in and out, as it were, like hyperobjects to Flatlanders). They lie at the edge of the apes’ world—indeed, they reside just outside the confines of their jungles and cages. And they are not something more or less than human (say, winged babies without belly buttons, or red-hot goteed men). They are, rather, humans of the “all too” variety: popping in and out of the apes’ world, and either locking them up or handing out bananas.

4. From Wish Hunting to Witchful Thinking

Let us turn now to a famous hermeneutic, studied by the anthropologist E. E. Evans-Pritchard (Reference Evans-Pritchard1976) in his monograph Witchcraft, Oracles, and Magic among the Azande (an ethnic group in southern Sudan). Consider the often-discussed case of the granary that falls down on a party. Something bad happens to someone. In particular, there has been an unfortunate coincidence of an efficient cause and a telic cause. Regarding the efficient cause, termites ate through the posts of the granary, and so it came crashing down; regarding the telic cause, a host and his guests were sitting there in order to escape the scorching noonday sun. That is, the party had a goal that got thwarted. To return to figure 3, they had a trajectory, a path toward a destination, that got intercepted or deflected.

Crucially, both of these events—by themselves—are easy enough to explain by reference to such everyday motivations and basic causations. What is difficult to explain is why both events happened to coincide in space-time: the granary fell just when the party was under it. And so a form of abduction, or hypothesis, takes place. Paraphrasing Peirce’s (EP 1.196–98, 2.230–32) well-known description of such inferences, the process might be understood to go something like this: (1) an unlikely coincidence has occurred (which leads to an unfortunate event); (2) if the coincidence were caused by a witch, its occurrence would be “a matter of course”; (3) so there is reason to believe there was a witch. Unfortunately, there is a plethora of possible witches to choose from; and so the only remaining question is which?

Evans-Pritchard said such a hermeneutic was locally construed in terms of the “second spear”: the telic and efficient causes were the first spear, but the fact of their coincidence was the second spear. Just as two spears are needed to bring a hunt to its conclusion (the one that weakens the animal, and the one that kills the animal), two causal spears must be posited in order to inaugurate an inferential search—the hunting of a witch.

Crucially, this second spear wasn’t just thrown by a witch per se; it was thrown by a witch with a good reason for being jealous of the victim, and thus a desire to hurt them. A particular affect (jealousy, itself leading to a malicious desire, intention, or wish) caused the witch to act as such. But what caused this affect? If we can figure out who would be experiencing such an affect, we would know who threw “the second spear,” and thus who we can hold responsible for the misfortune.

In this regard, Evans-Pritchard offered a few key criteria, which are precisely ways of productively constraining an otherwise way-too-wide inferential space, ways of grounding inferences. First, the suspect should have contact with the victim—living nearby, or otherwise rubbing elbows with them. Second, the suspect should be relatively marked—not just marginal, but someone least in conformity with social norms (and/or the victim’s own values). And, third, the suspect should be similar in status to the victim, and someone who recently got less than their share of some valuable good (in comparison to the victim).

Note, then, that we have a very similar situation as described above in our discussion of Freud. First, posit that some event, such as a dream or misfortune, is not senseless or contingent. Second, posit a repressed wish or a jealous witch as the dynamic object that caused that event. And, third, offer a productively constrained space for grounding inferences; for example, all the other ideas of psychoanalysis, plus the remains of the day. Or, all the other conventions and convictions in Azande culture, plus the victim’s and perpetrator’s day-to-day.

And so there is just one last question. What was the original point? What was the concrete semiotic process (that causes some head to turn) which leads to the abstract semiotic process (that causes some mind to search)? Our stereotype of pointing is a gesture that directs another’s line of sight to some object. Crucially, though, another way to point is simply to put an object directly in another’s line of sight. And so, in the case of misfortunes, nobody has to direct your attention to them, because they precisely cross your path (recall our parasite). Indeed, in many cases, they literally run you down or jump in your way.

5. Desire: The Prose and Lacan

Let us now turn to a village of Q’eqchi’ (Maya) speakers in the cloud forests of Guatemala. A woman squats by the hearth fire in a one-room, thatch-roofed home. Making tortillas for dinner, she repeatedly pats cornmeal into hand-sized circles, which she then lays down on a hot griddle. Intermittently, she pulls cooked tortillas off the griddle, and places them in a basket so that they may retain their heat and be passed around to her hungry family. In attempting to pull one such tortilla off the griddle, it tears in half. “Ay dios,” she says, looking up at her husband.

In short, a minor mishap or mistake occurs while a woman is working (one of her actions has gone “off course”), and her husband’s attention is drawn to the mistake through the use of an interjection (“ay dios”). In terms of the foregoing framework, a vocalized sign concretely stands for an object (in this case, an event) and calls another’s attention to that object (recall fig. 5). In some sense, then, we are not so much interested in the action of pointing as the point of such a “reaction.”

In this village there is also a local system of taboo (awas), whereby certain kinds of events (actions, experiences, mistakes) are understood to be causally connected to other kinds of events, usually because there is putative resemblance between them. In particular, if the woman’s husband were to eat such a torn tortilla, he might cut himself with his machete while working. And if the woman herself were to eat the tortilla while pregnant, her child might be born with a cleft palate. Note, then, that insofar as two events are understood to be related to each other as cause and effect, by pointing to the cause (qua proximal event), one is also pointing to the effect (qua distal event, or potential repercussion). To be sure, the inferential process often works the other way: that is, upon experiencing an event that may be framed as the effect of such a process, one goes looking for the event that must have been its cause. Recall our example of misfortune among the Azande, and the so often circular nature of Freudian and Gricean hermeneutics.

Figure 8A reframes this event in terms of such grounds. First, the woman points to something (concrete object) which points to something else (abstract object), insofar as the former relates to the latter as cause to effect (in a local, relatively shared, causal ground). The interjection “ay dios” constitutes this initial point. Its dynamic object (i.e., the reason for, or cause of, its expression) is—arguably—the woman’s knowledge of such causal processes (as a kind of path) and, in particular, her fear of such caused effects (as a kind of destination). More generally, one points to such events (like the torn tortilla) not so much because they are significant in themselves, but because they are potential causes of significant effects (insofar as such effects are positively or negatively valued). In this framing, then, a key dynamic object is one’s fear of the effect of the cause one is pointing to. That is, the dynamic object (which causes the first sign) is not so much a communicative intention, but rather negative affect in light of a possible causal repercussion.

Figure 8. A, Interjecting “Ay dios!” (first framing). B, Interjecting “Ay dios!” (second framing, itself built upon first).

But there is much more to this event than that. By calling attention to the torn tortilla, with its potential to harm a man or woman if eaten, the woman also effectively calls attention to the fact that she is remedying a risky situation. In particular, when her husband turns to look he doesn’t just see the torn tortilla, he also sees that his wife is setting it aside so that it won’t be inadvertently eaten (by either her or him). In effect, she seems to be (unintentionally?) calling attention to the fact that she is making sure a potentially dire causal process cannot proceed.

Before continuing it should be emphasized that interjections are funny little signs. Philosophers, layfolks, and linguists often imagine them to lie on the edge of language (Kockelman Reference Kockelman2010b). Indeed, they are usually defined as conventional linguistic forms (or words) that enter into minimal syntactic relations with the rest of language. For example, they prototypically undergo few inflectional or derivational processes (e.g., one cannot pluralize or nominalize them), and they don’t richly combine with other words to make up larger sentences. In some sense, they minimally partake of langue, and maximally partake of parole. Moreover, as intimated above, speakers are often accorded relatively little agency over their expression: they seem to be “speech reacts” rather than “speech acts.” Finally, the objects they index seem to be relatively limited in number, and hopelessly bound to the immediate context. In a tradition that goes back to Aristotle’s Politics, such signs seem to belong to that peripheral part of language where human voice most closely resembles animal sound.

Among speakers of Q’eqchi’ who live in this village, the interjection “ay dios” is typically used in the following situations. First, it may be uttered in the midst of an experience that is markedly intense, or graded. For example, upon struggling to lift an unexpectedly heavy bag, a boy might say, “ay dios, mas aal,” or “goodness, how very heavy.” Such a usage typically functions as an attention getter, and allows the speaker to take the floor. Second, one may use this interjection to comment on the relative intensity of another’s recounted experience. This use of “ay dios” functions as a back-channel cue (saying, in effect, “wow, what you’re saying is intense, keep talking, I’m listening intently”). Third, this interjection often prefaces marked responses to questions or offers. That is, if someone asks you a question and you don’t know the answer, or if someone makes you an offer that you cannot accept, you can preface your normatively unexpected response with this interjection: “Ay dios, toje’ xinwa’ak,” or “goodness, I just ate” (and so cannot accept the bowl of food you are offering). Fourth, a parent whose patience is running low can use this interjection with a child when they misbehave or make a potentially injurious mistake. Such a usage often functions as a sign of exasperation as much as admonishment. Finally, an older child can use this interjection when a younger sibling misbehaves in order to call a parent’s attention to their actions—acting, in effect, like the displaced eyes and ears of their otherwise disengaged parents.

In the case of the torn tortilla, uttering “ay dios” is probably multifunctional, aligning with several of the uses just described. It indicates a quantitatively marked event or experience (in particular, it is not just a little tear in the tortilla, which would usually go unremarked). And it functions to secure another’s attention and take the floor. It also, arguably, positions the woman as an older sibling pointing out the misbehavior of a younger sibling to a parent. As George Herbert Mead might see it, the woman as I inhabits the role of “older sibling” (an overseer, sounding an alarm), the woman as Me inhabits the role of “younger sibling” (someone who has made a mistake or misbehaved), and the husband inhabits the role of “parent” (someone who can hold the Me accountable for the mistake or mishap, as well as hold the I accountable for drawing attention to it, or not). We have then a very interesting mode of doubled self-interpellation through other-enrollment.

Crucially, around the time of this event, two kinds of state (and religious) moral injunctions were particularly salient. In particular, there was a well-organized and heavily signaged campaign to stamp out two alleged problems, or sins (maak), of Q’eqchi’ speakers: kok’alib’ (having too many children) and kalaak (drunkenness). If awas was a Q’eqchi’-specific form of taboo, this was a state-level variety.

Such injunctions were particularly salient in this young woman’s life. In particular, she and her husband had just had their third unplanned child. And a question she often turned to in conversations was birth control: what kind to get, where to get it, and how to pay for it with their meager income. Just as telling, and quite chilling, her aunt had recently died after a long illness, and when they prepared her body for burial they discovered she had been beaten badly; it was probably by her husband, who was well known to be an angel when sober and a devil when drunk.

Pulling all these details together, as troubling as they are, it is not at all surprising that a woman would point out such a mistake (and her remedying of it) to her husband. In part, such a sign could function to let him know that she might be pregnant, and that he is not doing enough to take care of her (by getting contraceptives, or being more careful, so they don’t have another child). In part, such a sign could function to let him know that she is taking care not to do anything that might get him hurt down the line (which could lead to him getting angry at her, with potentially harmful repercussions).

Finally, if it is the case that many speakers understand interjections as being sounded rather than said, such that there isn’t a full-fledged communicative intention behind them, such that they don’t involve the full agency of a stereotypic speech act (Kockelman Reference Kockelman2010b), then the woman can engage in a behavior that points to her carefulness (and her husband’s carelessness) without calling attention to any communicative intention, so that such context- and culture-grounded inferences (or more distal meanings) are relatively defeasible. This makes it difficult to hold the woman accountable for intentionally pointing out her own lack of accountability.

Figure 8B reviews the foregoing steps, which themselves build on the analysis undertaken in figure 8A. First, the interjection points not just to the mistake, but also to the woman’s remedying of the (conventionally) causal repercussions of that mistake. The mistake itself points to its possible effect (given certain assumptions among villagers): a man cutting himself, a woman’s child being born with a birth defect. The remedying of the mistake, in pointing to the cancellation of such a causal process, points to the woman’s carefulness (in making sure not to endanger her husband) and her husband’s carelessness (in not getting contraception). In light of recent events in this woman’s life, and recent state-level policies (against having too many kids and drunkenness), there is good reason for this woman to point out her own carefulness and her husband’s carelessness. Using an interjection, the woman sounds out these implications without spelling them out: perhaps she is unconsciously doing this; perhaps she is simply not intentionally doing it; perhaps she is strategically and self-consciously doing it in a way that can be passed off as unconsciously sounded. In any case, insofar as the meanings are seemingly implicit, the woman herself is—for those meanings—seemingly nonaccountable or noncomplicit.

In effect, there seems to exist a doubled dynamic object: not just her desire to point out that she is being careful while her husband is being careless, but also, arguably, her desire to keep this informative desire implicit (by use of a communicative resource that is ostensibly sounded rather than said), such that she cannot be held accountable for having pointed it out. (And all this on top of the other dynamic object we just discussed.)

If Freud (Reference Freud1990) sought to analyze the meaning of mistakes in his Psychopathology of Everyday Life, we are here analyzing similar kinds of events through a framework that might best be called “The Semiopathology of Everyday Strife.”Footnote ⁹

6. Meeting the Universe Two-Thirds of the Way

Suppose there is a statue, or some other large-scale “material object,” that is shrouded in fog, as illustrated in figure 9. We throw basketballs into the fog and catch them when they come out, noting their positions and velocities when thrown and caught. We imagine some kind of material the statue might be made of (stone, steel, and so forth), as well as some kind of form (curvature, texture, density, shape, size) that material might take. And we hypothesize an interaction between our balls and the statue (as well as a relative lack of interaction between the fog and our balls). The unseen statue, then, is that which deflects the trajectories of our balls in a particular and, hopefully, predictable way. A form of matter throws the balls off course in a way that can be made “a matter of course.”

Figure 9. Big O objects rendered through their interactions with little o objects

For example, as shown in the insert of figure 9, we might presume that collisions between the balls and the statue are relatively elastic (so that a ball’s speed after a collision is the same as the ball’s speed before the collision). And we might presume that the angle of a ball just after it hits the statue is the same as the angle of the ball just before it hits the statue (relative to a line perpendicular to the surface of the statue at the point of collision).

By measuring the trajectories of all the balls (when thrown and caught), and by theorizing the details of ball-statue interactions, we infer the shape of the statue, as a kind of abduction or hypothesis. Such an inferential process might go something like this: A strange event has occurred (our balls were thrown “off course”). If those balls had interacted (in a certain way) with a statue (of a certain sort) such deflections would be “a matter of course.” Thus, there is good reason to believe that the statue has a material form of precisely that sort. From a certain perspective, then, rendering res extensa (the dotted line in fig. 9) is not much different from rendering res cogitans (such as a communicative intention, repressed desire, jealous witch, negative affect, or muted desire).

Figure 10 diagrams the foregoing ideas, showing two-“thirds” of the Peircean kind. There is a set of signs (data, so to speak) that represent, in a relatively concrete fashion, the input and output trajectories of all the balls: ingoing and outgoing velocity and position of ball 1, ingoing and outgoing velocity and position of ball 2, and so on. Such changes in trajectories of balls (that such signs point to) are themselves signs, in a relatively abstract fashion, of the shape (material, properties, position, etc.) of whatever changed the trajectories of those balls. In particular, we can get from the concrete object to the abstract object in light of an imagined dynamic object: an obstacle, obstruction, or parasite, which interferes with the balls in particular ways. It is only by postulating the features of this obstacle, in relation to the complementary features of the balls we throw and catch, that we may render the obstacle. Note, then, that such parasites don’t so much stand perpendicular to our paths; they are, rather, the very destinations of those paths—the very objects we are aiming at.

Figure 10. Rendering of “objects” reconsidered

Such ideas are readily generalized. At issue is not just a statue enshrouded by fog and impinged on by basketballs. Such an analysis also applies to a brain tumor enshrouded by a skull, as impinged on by x-rays. It applies to gold atoms, enshrouded by their incredibly small scale, as impinged on by alpha particles. It applies to an enemy submarine enshrouded by the ocean as impinged on by sound waves. It applies to a Mayan temple, enshrouded by dirt, as impinged on by radio waves. And, of course, it applies to a bat’s rendering of its nocturnal surrounds via echolocation. (For the calculations in question can be engenomed as much as enminded; and the throwing and catching media are just as often embodied in instincts and habits as they are embedded in instruments and infrastructure.)

Such an analysis applies to a rock, enshrouded by the atmosphere, as impinged on by visible light. (For we can catch with the eye what is thrown by the sun, candle, or lamp.) It applies to any entity that “bends light,” such as a prism or pane of glass. (For refraction is as important a process as reflection.) It applies to any entity that “breaks up light,” such as an aperture or grating. (For diffraction is as important a process as refraction and reflection.) And it applies to a rock, enshrouded by darkness, as impinged on by a hand. (For, in some sense, we sense and instigate in myriad ways simply by throwing and catching our bodies, and their various parts.)

Such an analysis applies to Huck Finn as enshrouded in a dress and bonnet, Southern politeness, and social norms, as impinged on by the tossed pin cushions and unthreaded needles of Ms. Loftus. Indeed, it also applies to a “neurotic,” enshrouded in fin-de-siècle culture and Victorian privacy, as impinged on by a psychoanalyst. And it applies to a “native,” enshrouded in local customs, as impinged on by an anthropologist. (For questions, and inquiries more generally, not to mention commands and insults, are just as often “deflected” as balls.)

In short, the properties of what we “throw” and “catch” (little o), of what there is to “hit” (big O), and how such entities might interact with, inform, and transform each other may be understood and imagined in any number of ways. Stereotypic physical processes of the wave/particle kind (such as reflection, refraction, or defraction) barely scratch the surface.Footnote ¹⁰

As should now be apparent, any such “experiment” or “experience” requires an agent (big A), itself a collectivity of smaller agents (little a). Such agents can throw and catch little o objects (such as particles, waves, fingers, and questions). They can measure the input and output features of such objects. They can imagine some big O object that such little o objects might be interacting with (such as an atom, a tumor, or a submarine; an identity, an illness, or an intentionality). They can theorize the details of such interactions. They can make inferences accordingly. And they can plan their instigations and update their theories in light of those inferences. (As well as do much else besides.)

As the history and philosophy of science has shown us (Hacking Reference Hacking1983, inter alia), for such agents to learn about big O (the object in question), they must already know a lot about little o (all the entities thrown and caught). And they must not only know that such little o objects have certain properties (that govern their interactions with big O), they must also know how to “throw” and “catch” them. To represent (big O) they must first intervene (little o), and their interventions turn on their residence in the world as much as their representations of the world. Moreover, just as the little o’s of today were probably the big O’s of yesterday, the big O’s of today may become the little o’s of tomorrow. Just as past inferences lead to present interventions, present interventions lead to future inferences. Such knowledge is not just embodied and enminded, of course; it is most often embedded—in all the technologies, for example, that are used to throw and catch. Such is the realm of the “real” as experimentally—or experientially—rendered.

As intimated above, big O is itself just as often an agent, and so cannot only catch what we throw, but also throw it right back. Indeed, many such agents can also duck, flee, shatter upon impact, cower in a corner, confiscate our balls, or simply whack them out of the park. Perhaps more often than not, they refuse to be rendered, such that they never become a matter of course (in regard to the ways and reasons they throw our matter off course). To return to nonhuman primates, as a boy I would often visit the San Francisco zoo, where the apes used to throw their feces whenever they caught you pointing.

Figure 11 returns to our original scenario. As may be seen, depending on which kind of ball is thrown and caught (and, in particular, how that ball interacts with the statue enshrouded by fog), one and the same statue may be rendered in very different ways. For example, if parts of the statue have a curvature whose radius is smaller than the radius of the balls, then those parts of the statue cannot be resolved. Similarly, if the statue is really a mobile, whose movements are faster than the rate we can throw and catch balls, then all we may be able to resolve is a blur. Or, if the statue is around the same size as the ball, then the ball’s momentum may be transferred onto it, such that it too will be “thrown” (and we will have to measure how much). And so forth.

Figure 11. Multiple renderings of the “same” big O object

Very similar observations apply to each of the other scenarios mentioned above—submarines and temples, rocks and tumors, illnesses and identities. More generally, depending on what kind of little o’s are thrown and caught, the properties of big O may vary greatly—not just its shape and size, but also its internal contents and inherent qualities, its position and momentum, its color and texture, its energy and temporality, its use value and exchange value, its frequency and amplitude, its beliefs and values, its affect and attitude, its spin and charge, its very “nature.” One and the same physical reality, so to speak, may be rendered through experimentally repeatable and epistemologically objective techniques, in a potentially infinite variety of ways. And so there is no way such agents will ever see some real object in all its glory, so to speak. For this reason, the big O object is sometimes best understood as that which stands at the intersection, or virtual center, of such an infinity of possible renderings, each made possible by the throwing and catching of different little o’s.

That said, there may always be the hope that big O may be reduced to an ensemble of seemingly natural kinds (or a similar sort of sort), so that such an infinity of possible renderings is itself renderable into a single image, essence, icon, idea, habit, form of life, worldview, or equation. Indeed, a key prize in many imaginaries will always be such an ultimate interpretant of the abstract object, especially insofar as such an ultimate interpretant renders the abstract object relatively simple, portable, perspicacious, commensurate, generative, or citable (at least to those interpreters who have been trained to perceive, and produce, its implications). Such are some of the perceived stakes of nature’s putative secrets.Footnote ¹¹

Crucially, for a wide range of big O objects, the “fog” is integral to them: there’s no blowing it away, or waiting for it to dissipate, or going around it, or even reaching inside it. Indeed, in some sense, everything the agents require to find and create indices, as well as make inferences and undertake actions, is the fog. Their very ontologies and instruments, their mentalities and media, help constitute not just that which is to be illuminated, but also that which is doing the obfuscating. Just as each agent is often its own worst enemy, it is also the foremost source of its own noise.

As should also be clear, big A can be more or less overlapping with, or even identical to, abstract O. That is to say, big A can itself be the target of its own sensations and instigations, its own throwings and catchings. One can put oneself through trials to find out what one is “really made of.” (For tests of mettle are no different from tests of metal.) And so such an analysis also applies to the self (qua Me), as enshrouded by “unconscious drives” or “social structures” or “global capital” (and other well-known monsters, however real or imagined), as impinged on by the self (qua I) in its reflexive attempts to become more conscious of itself. In other words, such an analysis also applies to many self-reflexive projects, be they collective (who are we) or individual (who am I). And similarly for other projects: while the answers to questions like who am I, what is that, and who are they (not to mention why oh why and what the fuck) are different, the techniques used to answer them are often more or less the same.

From the foregoing examples, it should be clear that any such big O object may itself be understood as a sign of something else (so far as the interpreting agent has a particular causal imaginary, or ground more generally). For example, archaeologists may use ultrasound (or ordinary shovels) to render a causeway; and that causeway, or any of its features (shape, size, composition, age, location, and so forth), may be understood as a sign of the society that built it—and thus provide evidence of that society’s beliefs, desires, knowledge, know-how, social relations, causal grounds, cosmology, and so forth. A submarine rendered by ultrasound may constitute evidence of the immanence and size of an enemy attack. Physical characteristics, as rendered by ultrasound, may constitute evidence for the sex of an unborn child. More generally, insofar as the form, material, properties, and position of any object are themselves signs of the transformative processes that brought that object into being, one learns many other facts just by rendering such features of that object (so far as one has knowledge of such transformative processes).

While we are not here taking up such issues, it should be understood that it is often precisely these other features that constitute the reason for our pitching and catching activities in the first place. That is, we render a big O object precisely because it constitutes a sign (in our causal imaginary) of some other object. Moreover, as a function of such a rendered identity, we don’t just infer certain other properties (like immanent attacks and gendered bodies), we instigate further actions—where to place our ships, what to name a child, where to dig next, how best to debug, and so forth. Phrased another way, the rendering of a big O object is usually a means to further ends, as much as a sign of other objects.

Just as the final abstract object may itself be a sign that points further, the initial concrete index may itself be an entire semiotic process (and so on, indefinitely). For example, one can point in front of a child to see if they can follow your point. And their interpretant of your sign becomes a key index of their identity (to a doctor or therapist trained in a certain diagnostic tradition, or habituated to a certain ground). In particular, their failure to look (if they are nine months or older) is sometimes taken as a symptom of autism. So just as we may push back, in developmental time, the gendering (animating, enumerating, segmenting, and it-ing) of a child to a point prior to birth (via ultrasound), we may push back the disabling of a child to a point in time that is allegedly prior to language (via joint attention). In the first case, it often feels oppressive (the disciplining of a child’s gender starts earlier and earlier); in the second case, it is often thought restorative (the earlier the therapeutic intervention, the more successful the outcome). As brilliantly developed in the writing of scholars like Goffman (Reference Goffman1959), Butler (Reference Butler1993), Mol (Reference Mol2002), Taylor (Reference Taylor2004), and Barad (Reference Barad2007), such ontological baptisms are as important to critical theory, ethics, and ethnography as they are to clinical interventions and scientific understanding.

As should be clear from this last example, the pitching and catching of little o objects is just one way to imagine our interactions with big O (see fig. 12). Phenomenologically construed, we instigate certain actions and we sense the results of those instigations. We pull something and, concomitantly, feel how it stretches (or rips or screeches).Footnote ¹² We signify something and, concomitantly, perceive how it is interpreted (or even if it is interpreted). And it is precisely such an instigation-sensation relation, or an ensemble of such relations, that constitutes the key indices we use to infer an object’s identity. Indeed, what is important is never the qualia per se (some experience of a sensation), but rather the quantia (and often quantities) of such sensory-instigatory qualia: the ways intensities in sensation (along various dimensions) correlate with intensities in instigation (along various dimensions). Note, then, how far we are from the “bundle of qualities” imaginaries so presciently ridiculed by Locke.

Figure 12. Indices as relations between relations

Such relations between relations may be understood technically instead of experientially: any input-output relation may constitute an index of the internal identity (e.g., circuity or algorithm) of some complicated technology. Indeed, it may index a “malady” afflicting that internal identity: a bug in some software, a disconnect in a machine, an enemy or noise on the line or in the works. (Indeed, the “machines” in question are just as often some little a in big A’s ensemble that is failing to function during the course of an experimental endeavor, as they are some big O that big A is attempting to elucidate through such an endeavor.) Finally, and in line with the points above, a key interpretant of such a pitch-catch, instigate-sense, or input-output relation is usually another such move: what we will pitch (and catch), or instigate (and sense), or input (and output) next. That is, we use the relation between our action and its result to determine our next action (with an eye toward its result).

As such examples clearly show, the interaction of little o with big O is only sometimes governed by stereotypic physical processes (reflection, refraction, defraction). Most actual interactions are far more complicated: diagnosing an illness, reckoning a terrain, assaying a substance, interpreting a gesture, explicating a social relation, understanding a taboo, or measuring a forcefield (not to mention explaining phenomena like global warming, the origins of the galaxy, the French Revolution, or what my three-year-old wants for lunch). This is especially true insofar as big O “internalizes” little o, and transforms accordingly. For example, in measuring the temperature of a gas, one may change the temperature of the gas. In trying to understand the beliefs and values of a person, one may change the beliefs and values of that person.

To be sure, many such effects are more or less easily handled (say, by making sure a thermometer has far fewer degrees of freedom than the object whose temperature is being measured). But other such effects are more pernicious. Indeed, many such effects are fundamental to our understanding of matter and meaning. For example, in measuring the polarization of a light beam, one changes its polarization. In distinguishing between ham and spam, one changes the attributes of spam (and even ham). More profoundly, any population subject to natural selection internalizes its environment by transforming its genotype (and thereby, often, changes its environment). Finally, as well known to anthropologists, humans very often internalize others’ interpretations of their indices, and so come to transform their indices (including their throwing and catching behavior, as well as their signing and interpreting behavior) in order to “throw off” the inferences of such interpreting agents.Footnote ¹³

Perturbing Hacking’s (Reference Hacking1995) famous phrasing, we don’t so much “come to act under a description” as we come to counter act a representation (or any interpretation, more generally), or simply come to act under the radar of an intervention (e.g., any attempt to make us interact with some little o).Footnote ¹⁴

7. Capture and Enrapture

In our discussion of Freud and Grice we described the pleasure of residing in productively constrained inferential spaces. Of interest, then, is not just the enclosure of an object, or the production of a phenomenon; it is also the capture of an agent. Gell (Reference Gell1998) famously described the capture of one particular kind of agent: someone caught up in the design patterns of an art object, trying to make sense of some ensemble of sensual features, trying to find coherence in some unfolding of events.

Crucial to this capture, it should be argued, is the agent’s assumption that it can and should make sense of the features or sequencing: not just that there is a pattern to be found, but also that the pattern somehow pertains to that agent, and that the agent has the means to find that pattern should it work hard enough. From whence comes this promise—not just that there is a secret, but that the secret is particularly relevant to you, and that you have in yourself the means to reveal that secret (providing you don’t dillydally).Footnote ¹⁵ This, arguably, is the key dynamic object of most forms of imaginative and experimental inquiry. The question is not why do I turn. It is, rather, why do I pitch and catch, hoping to intuit what caused it to turn.

(Note, then, that just as big O is very often a parasite, big A is very often an enemy. It doesn’t so much intercept secrets as render them, listening in on the world’s conversations with itself, through channels of its own providing.)

But Gell’s notion of capture is not enough for our purposes in still further ways. For the agents captured are not simply trying to infer patterns off the indices of found objects. As we just saw, they are intent on producing the indices through their sensations and instigations. Indeed, they are dedicated to creating big O through their throwing and catching of little o. Such indexical-inferential and sensory-instigatory interrelating, in all its potentially productive pleasure, is perhaps best understood as enrapture.Footnote ¹⁶

A key question, in the tradition of scholars from Erving Goffman and Michel Foucault to Ian Hacking and Lorraine Daston, is the history or genealogy of these indexical-inferential and sensory-instigatory worlds. Who can reckon in them? To what ends? How did they get stabilized, or closed? What kinds of events lead to their destabilizing, or their reopening? For our purposes, a key issue in such histories is the (often inadvertent) production and (usually resistant) detection of semiotic strain. Building on Mary Douglas’s (Reference Douglas1966) notion of anomalies, such strain might be understood as barely evident evidence that some indexical-inferential imaginary, some ground, hermeneutic, or ontology, is “out of touch with,” or “insensitive to,” some world. In particular, such strain consists of all the evidence one might gather (in light of a more inclusive, incipient, or counter, ontology) that the kinds of causal interactions we imagine are incorrect, that the indices we produce and interpret are inadequate, that the individuals we aim our instigations and sensations at are nonexistent, that the distributed agencies we incorporate are incoherent, that our modes of inference are unsound, that our treatment of subjects is unethical (if not diabolical).Footnote ¹⁷ At issue, then, is not just how to (elude) capture, but also how to rupture (or sustain) enrapture.