Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

Various thinkers have explored versions of the idea that there are ­nonrepresentational and unconscious dispositions that form the background of thought, meaning, and language. In this chapter, I discuss John Searle’s version of this idea. I focus on how he invokes the social background to account for the functioning of social rules. This is a particularly important application of the notion because very many in other respects quite different accounts of social, institutional, and intentional phenomena generally appeal to rules, for example, linguistic rules. But what is the status of such rules? How, in particular, can they play a role in explaining what people do if people normally are not aware of them? The standard answer is that the rules are unconscious and are being followed unconsciously. Searle has challenged this answer on the basis of a thorough critique of the notion of the unconscious mind (e.g., Searle 1992). His alternative explanation appeals to the background. While background dispositions are not representations of rules, they are still “sensitive to the rule structure” (Searle 1995: 145). They produce behavior that is (generally) consonant with the rules, but they are entirely physiological. Rules only enter the picture when we introduce a diachronic dimension. Background dispositions have evolved in response to the rule structure.

It seems to me that Searle’s critique of the notion of unconscious rule following is right on target, and I will take it for granted in what follows.Footnote 1 But I will ­criticize Searle’s positive account and propose an alternative one. In an earlier article (Schmitz 2012), I have argued that we should think of the background as being nonconceptual rather than as being unconscious and nonrepresentational. The main goal of this chapter is to show that this conception of the background will also help us to better understand the social background and the social ­phenomena that give rise to the idea of unconscious rule following. The outline of this chapter is as follows. After an introductory sketch of Searle’s notion of the background, I present his explanation of apparent unconscious rule following in more detail. I then critique it and set out my own explanation. The chapter ­concludes with a suggestion on how to situate the background in the context of a layered view of the mind.

2 The Background According to Searle

Searle characterizes the background through many examples as well as through various theoretical functions that he ascribes to it. Most of his examples try to make plausible the idea that one can take something for granted or be committed to it without believing it. We usually take for granted that the objects around us are solid and will offer resistance to touch and would be very surprised if they just vanished into thin air the moment we make contact with them. But it does not seem right that ordinarily, we entertain beliefs to the effect that objects are solid and offer resistance to touch, though of course we can form such beliefs and sometimes do, for example, when philosophizing. Analogous phenomena can be found in the intersubjective and social domains. It also normally seems inadequate to say that we believe that the people we pass by on the streets are conscious or that a child that is screaming because a car has run over his foot is in pain. As Wittgenstein, whose later work is a main source of inspiration for Searle’s notion of the background, puts it, we are not of the opinion that the other has a soul. Rather, our attitude or stance toward the other is an attitude or stance toward a soul (Wittgenstein 1984, part II, iv). In a similar vein, Searle also often speaks of the sense that we have of others as potential cooperation ­partners (e.g., 1995). We have a sense, for example, that we could approach them on the street and ask them for directions. If it turned out on closer inspection that their eyes were dead, their whole behavior zombielike, and that they were clever automata remote controlled by somebody leaning out of a nearby window, laughing at us, we would be surprised, probably even shocked and terrified before being able to join in the laughter. There are indefinitely many areas in the social and other domains where it seems appropriate to say that we have a sense of something in this way. For example, we also have a sense that it is our turn in a conversation, or that certain things would be socially inappropriate to say, or that they are simply ungrammatical. The reason it seems apposite to prefer these terms to talk of thought, of belief, of intentions, etc., is that these are things that we normally don’t think about and that we normally do not have rules for – unless we are philosophers, psychologists, linguists, or other relevant experts.

In addition to talk of stances and of the sense we have of certain things, Searle uses a host of other terms to refer to elements of the background. Most of these can be divided into two broad groups. The first group contains expressions like “preintentional assumption,” “background presupposition,” and “taking things for granted.” This terminology can be called the “intentional state terminology.” It is most frequently used in Searle’s early writings on the topic (1975, 1980). However, this terminology is rather problematic because of Searle’s insistence that the background lacks representational content. This for him is one of its defining features. But how could an assumption or presupposition lack representational content? Therefore, as Searle is well aware, this terminology has a paradoxical, even “oxymoronic” (1983: 156) ring to it. It is probably for this reason that in his later writings the second group of expressions becomes more prominent, the terminology of dispositions, of capacities, tendencies, skills, habits, routines, and of know-how both in the sense of knowing how things are and of knowing how to do them. However, this way of conceptualizing the background raises the question what the manifestations of these dispositions are. As we shall see, this question is surprisingly difficult to answer in a way that is consistent with the letter and spirit of Searle’s notion of the background.

The idea of a nonrepresentational background is often supported through a regress argument. For example, it is suggested that to avoid a regress of interpretations, understanding must bottom out in a background of nonrepresentational capacities. Even though Searle (1983, 1991) says that he does not employ a regress argument for the background, the background still performs functions in Searle’s theory that correspond to the intent of familiar regress arguments. The most important function of the background in Searle’s theory is that it fixes the application of intentional states: intentional states only determine conditions of satisfaction ­relative to a background. If there is no background, the intentional state does not determine conditions of satisfaction at all; different backgrounds determine different conditions of satisfaction. For example, the statement that a cat is on the mat only determines conditions of satisfaction relative to the background assumption of a gravitational field (Searle 1978). And the verb “cut” and the intentional content that it expresses is applied and interpreted differently against the background of different practices of cutting things. When an order to cut a cake is given, this is usually meant against the background of a practice of cutting cakes with knives. So when in response somebody runs over the cake with a lawnmower, he literally did not do what he was supposed to do. The order has not been satisfied. And this is so even though, according to Searle, “cut” has the same literal meaning whether it is applied to the cutting of cakes with knives or of grass with lawnmowers (Searle 1983, 1992), and intentional content is fully expressed or at least expressible through literal meaning.Footnote 2

3 The Background, Consciousness, and the Connection Principle

Searle characterizes the background as both mental and physiological, more ­precisely as neurophysiological, and as unconscious and nonintentional. The first two characterizations might appear to be incompatible, but Searle clarifies that when he says that the background is mental, he merely means that it is internal in the sense of being inside the skull (1983: 153f, 1991: 291f). But why does Searle classify the background as physiological? I believe the answer is that he thinks about background dispositions in terms of their bases, probably even tacitly identifies these dispositions with their bases. He asks: what is the occurrent reality of these dispositions and what makes ascriptions of these dispositions true when they are not exercised? And he assumes the answer must be some physiological ­condition in the brain. What else could it be!

And from this perspective, it then also seems unproblematic that the background is unconscious and nonintentional. However, it is not obvious that a disposition is identical to its base – the occurrent and in this case physiological condition that explains its manifestations. One might rather appeal to Bishop Butler’s dictum that everything is what it is and not another thing and insist that a disposition is one thing and its base quite another. And a question like, for example, what makes it true that I know how to speak English, even when I do not speak it since, say, I am in a dreamless sleep, can be seen to be falsely posed. It is rather like asking what makes it true that somebody is a habitual smoker even when she is not smoking or what makes it the case that she attends class regularly even when she is not there. To ask in this way is just to misunderstand how these expressions work; in particular, it is to misunderstand the temporal resolution at which they represent reality. That she smokes habitually just means something like that she smokes regularly and has not quit, probably also that she will start to crave smoking when she does not get her accustomed dose. (This is not intended to be an “analysis” of this expression; it rather illustrates the level at which it works, which is not that of brain physiology.) In any case, even apart from what the right metaphysical view is of dispositions, their bases, and manifestations, it is important that ordinarily we do not, like Searle, classify dispositions in terms of their bases, but in terms of their manifestations. For example, musical skills are manifest in musical performances, mathematical skills in mathematical performances, and so on. So the crucial question again is the ­following: what are the manifestations of background dispositions? More specifically, what kind of performance is, for example, applying an intentional content? Here, it seems to me, Searle’s account of the background is faced with a dilemma. One horn of the dilemma is that background dispositions are not manifest in ­intentional or conscious performances or events at all. Then it becomes rather ­mysterious what they are manifest in and what they do. Probably the best guess as to what this interpretation might amount to is that background capacities would be manifest in neurophysiological occurrences, which, while not among the immediate neuronal correlates of states of consciousness, would play some sort of supporting role with regard to the latter. But this suggestion is not only rather vague and not very helpful, it also seems to be at odds with various things Searle says. In the few passages where he explicitly talks about what the manifestations of background capacities are, he speaks, for example, of his “intentional behavior” as being a “manifestation of … background capacities” (1992: 185), and he later identifies “seven ways in which my Background abilities manifest themselves in actual ­occurrent forms of intentionality” (1995: 137). This list also leaves no doubt that background dispositions are manifest in consciousness. For example, “the background structures consciousness” (1995: 133, point 3) and motivational background dispositions “condition the structure of our experiences” (1995: 135, point 5). It thus seems clear that Searle grasps the other horn of the dilemma, namely, to hold that background dispositions are manifest in intentional and conscious events.

The problem with this is not merely that it now becomes obvious that Searle is wrong – or at least categorizes dispositions in a nonstandard way – when he describes the background as nonintentional and unconscious. It is intentional and conscious because it is manifest in intentional and conscious performances and events, just like mathematical abilities are manifest in mathematical performances and events. The deeper problem is that under this interpretation, too, the notion of the background is in danger of losing all distinctive content. It is hard to see what its positive content is beyond the trivial point that humans do have capacities to engage in all kinds of intentional events and performances, such as, to take some random examples from the social domain, dancing together, taking turns in a conversation, or forming and executing a joint intention, and the almost equally obvious information that the base of these capacities is neurophysiological. The only interesting point that the notion of the background still makes under this interpretation is a purely negative one, namely, that the bases of these capacities do not consist in occurrent mental events or states. The base of the capacity to ski, for example, is not an inventory of unconscious representations of the rules of skiing that guide the manifestations of this capacity. Rather, this base consists entirely of physiological structures. When activated, these explain the conscious intentional performances of skiing. To understand these performances, it is not necessary to invoke the cognitivist myth of an arsenal of occurrent unconscious representations. Conscious events and performances and the corresponding dispositions are all that is needed.

When the thesis of the background is reformulated in this way, it becomes identical to that of the connection principle. That there are no occurrent unconscious mental states that explain dispositions like the ability to ski or to cooperate with others is just a special case of the general point that there are no occurrent unconscious mental states at all. There are only states of consciousness and dispositions to be in such states. This way of thinking about the background, the connection principle, and their relation can also make sense of the evolution of Searle’s thought. As Searle points out himself (1992: 186ff), when he first employed the notion of the background, he was still under the influence of what he later came to call the “inventory conception” of the mental. This conception embodies the idea denied by the connection principle: that there is an inventory of unconscious occurrent mental states underlying conscious events – the beliefs, intentions, and other states stored away in the dark attic of the mind, waiting for the light of consciousness to shine on them. The notion of the background as introduced in Searle’s early writings rejects this idea for background abilities and presuppositions, but it still remains in place for intentions, beliefs, and other intentional states. Given this picture, the background can also be distinguished from the network of intentional states through its purely dispositional nature – though this isn’t Searle’s official way of distinguishing the two. Conversely, once the idea of occurrent unconscious mental states is abandoned, the critical questions just raised on how to distinguish background and network become even more urgent. If both are dispositional, and if both are manifest in intentional events, how are they distinct at all? Of course, we might conclude that they are not and settle for the purely negative point encapsulated in the connection principle. But then we cannot account for the examples through which the notion of the background was introduced in the first place. How is taking for granted that the earth won’t move and that the people we pass by on the street are conscious, or having a sense of them as potential cooperation partners, different from the corresponding beliefs? And analogously, how are habits and tendencies to act in certain ways, for example, to treat people in certain ways, different from the corresponding intentions? How is having a sense of how to ­continue different from having a corresponding plan? It still seems intuitively plausible that there is a difference here. And from our discussion so far, it is clear that if there is a difference, it must lie in the occurrent mental and thus, according to the connection principle, conscious events in which background dispositions are manifest. Dispositions can only be responsible for the relevant differences if they are manifest in different kinds of events. Furthermore, this difference must pertain to the intentional content of these events, to its structure, or representational format.

It seems to me that a plausible suggestion here is that the difference can be accounted for in terms of nonconceptual intentional content (Schmitz 2012). Background dispositions are those dispositions that become manifest in mental events with nonconceptual intentional contents. While I lack the space here to ­provide a full-blown explication of the notion of nonconceptual content, I will later argue that nonconceptual content has the structural features required to explain the intuitive difference between such states as – to again use our familiar example – the belief that somebody is a potential cooperation partner and the sense that he is. I will begin my argument for understanding the background in terms of nonconceptual content by trying to show that an appeal to nonconceptual content does a better job at solving, or rather dissolving, the problem of rules than invoking the background as conceived by Searle. To do this, however, we first need to get clear about how Searle deploys the notion of the background to respond to that problem.

4 The Background and Rules

The problem for Searle, we recall, is how to make sense of the fact that people ­participate in social institutions, even though in many cases they are not aware of the rules governing these institutions and even though Searle rightly rejects the easy (and popular) way out of this problem by simply asserting that the rules are ­followed unconsciously. The key to solving this problem, Searle says, is to add a diachronic dimension to the explanation of this kind of social behavior:

(…) if you understand the complexity of the causation involved, you can see that often the person who behaves in a skillful way within an institution behaves as if he were following the rules, but not because he is following the rules unconsciously nor because his behavior is caused by an undifferentiated mechanism that happens to look as if it were rule ­structured, but rather because the mechanism has evolved precisely so that it will be sensitive to the rules. The mechanism explains the behavior, and the mechanism is explained by the system of rules, but the mechanism need not itself be a system of rules. I am in short urging the addition of another level, a diachronic level, in the explanation of certain sorts of social behavior. (Searle 1995: 146)

Searle compares his proposed reconceptualization of the role of rules to what he refers to as the “inversion” of traditional intentionalistic or teleological explanations of biological phenomena through Darwinian evolutionary theory. Prior to Darwin, the fact that a plant turns its leaves toward the sun was explained teleologically: it does so in order to survive. Now we explain such facts through “blind,” non-­teleological processes like random mutations and the survival of the fittest. Plants that are disposed to turn their leaves toward the sun through random mutation are more likely to survive and reproduce (Searle 1990). The explanation has been inverted because survival does not appear as the teleological, final, cause of the behavior anymore, but as its effect. This inversion can serve as our model for explaining human functioning in institutional contexts:

A similar inversion should be applied to human background capacities for coping with social phenomena. Instead of saying, the person behaves the way he does because he is ­following the rules of the institution, we should say just, First (the causal level), the person behaves the way he does, because he has a structure that disposes him to behave that way; and second (the functional level), he has come to be disposed to behave that way, because that’s the way that conforms to the rules of the institution.

In other words, he doesn’t need to know the rules of the institution and to follow them in order to conform to the rules; rather, he is just disposed to behave in a certain way, but he has acquired those unconscious dispositions and capacities in a way that is sensitive to the rule structure of the institution. To tie this down to a concrete case, we should not say that the experienced baseball player runs to first base because he wants to follow the rules of baseball, but we should say that because the rules require that he run to first base, he acquires a set of Background habits, skills, dispositions that are such that when he hits the ball, he runs to first base. (Searle 1995: 144)

So on Searle’s picture, the relevant kind of behavior is immediately controlled by the purely physiological mechanisms which underlie background capacities – or, are, on his view, as we have seen, perhaps even identical with them. But if rules don’t immediately control this behavior, what is their role and what is their connection to the background capacities/mechanisms? In the passages I have quoted Searle distinguishes the functional from the causal level of explanation, but he also says such things as that the capacities have evolved so that they will be sensitive to the rule structure and because they produce behavior required by the rules. It seems clear that the kind of explanation Searle has in mind ultimately is also a species of causal explanation. There must be some kind of causal connection between the behavior and the rules. Otherwise the rules could only serve as an external standard for the behavior. But they would not explain it.

What can this causal connection be? The most straightforward scenario is that the background capacities and mechanisms are acquired on the basis of the subject’s knowledge of the rules. When the player first learns the game, he is conscious of the rules, and this rule consciousness guides his behavior and plays a causal role in the development of his skill. But when the skill has been acquired and the player does not think about the rules anymore, the skill rather than the rules controls the behavior. It is not that he is he now following them unconsciously, as the traditional view has it. On Searle’s view, the rules rather “become progressively irrelevant” and “recede into the background” (1983: 150) – which I take it means they are, so to speak, dissolved and transformed into background capacities/mechanisms. Their role is merely historical. That’s why Searle says his view adds a “diachronic level” to “the explanation of certain sorts of social behavior” (1995: 146).

Searle also countenances what appears to be the interpersonal version of this intrapersonal case. Consider how he develops the baseball example. He imagines “…there were a tribe where children just grew up playing baseball. They never learn the rules as codified rules but are rewarded or criticized for doing the right thing or the wrong thing. For example, if the child has three strikes and he says “Can’t I have another chance?” he is told, “No, now you have to sit down and let someone else come up to bat”” (ibid.: 144). For rules to be involved here at all, we must assume that the adults rewarding or criticizing the children know the rules. So in this case, their awareness of the rules would be a cause of the children’s developing skill and thus a distal cause of the behavior displaying that skill. Note that in spite of Searle’s claim – through his analogy with Darwinian evolutionary explanation – that his proposal inverts the standard explanation of rule-governed behavior, it now seems clear that on his account, the rules are actually just further removed. They are more distal causes in the past and/or act by way of other people’s minds, but apart from this, they still appear in essentially the same causal role as in the standard explanation.

The further removed the rules are, the more doubtful it already becomes intuitively that they are really the source of the relevant normativity. Can the fact that other people’s rule knowledge was a – possibly historically quite distant – cause of the children’s behavior sufficiently account for the applicability of normative notions to it? One may suspect that either the children are just behaving as if they were following a rule in the sense of mere as-if intentionality (Searle 1992) or that there is another form of intentionality present, which – rather than rule knowledge – is the source of the relevant normativity. I will later argue for this second option. However, for Searle, this is not acceptable because he assumes that normativity requires rules:

Somebody might say, “Why do you have these rules at all? Why don’t you just have some kind of behaviorism? These things just happen, people just do these things.” The answer is that where human institutions are concerned, we accept a socially created normative component. We accept that there is something wrong with a person who when the baseball is pitched at him simply eats it; something wrong with the person who doesn’t recognize any reason to do something after he has made a promise to do it; something wrong with the person who goes around spouting ungrammatical ­sentences. And all these cases involve something wrong in a way that is different from the way there is something wrong with the man who stumbles when he walks; that is, there is a socially created normative component in the institutional structure and this is accounted for only by the fact that the institutional ­structure is a structure of rules, and the actual rules that we specify in describing the institution will determine those aspects under which the system is normative. It is precisely because of the rule that making a promise counts as undertaking an obligation that we recognize that certain kinds of behavior within the institution of promising are acceptable and certain other kinds are remiss. (Searle 1995: 146f)

This passage is worth quoting at length because Searle here eloquently expresses an attitude that not only explains why – given his other commitments – he responds to the problem of rules in the way that he does but is also very characteristic for the way that people have thought about these issues at least since the 1940s and 1950s of the past century, at least since Wittgenstein’s very influential discussion of rule following. Indeed, the view that normativity requires the presence of rules is still very widely, probably nearly universally, accepted in contemporary philosophy.

There are several aspects to Searle’s argument that should be distinguished. First, there is the assumption that if we say, in the vein of Wittgenstein, whose philosophy Searle is alluding to at the beginning of the quote above,Footnote 3 that at the bottom of the language game there is action, certain ways of doing things that we just engage in, without further justification, this is already a form of behaviorism in an objectionable sense. Second, there is the assumption that normativity could not be socially created except by creating an institutional rule structure. Third, this assumption is supported by appeal to the claim that only rules could “determine those aspects under which the system is normative.” Let me take these points in turn. First, the claim that at the bottom of the language game there is action does not commit us to behaviorism in any sense. It neither commits us to behaviorist versions of scientific psychology nor to any logical behaviorist reduction of mentality. There is no need to conceive of action behavioristically. Indeed, I think we can and should go along with Searle’s account of action, according to which all action has both a bodily and a mental, intentional component, which at least normally consists in the intentional content of the bodily experience of acting.Footnote 4 This is about as anti-behaviorist as it gets. Second, in the social domain, the relevant kind of action must be joint action. I will argue that an elementary form of normativity is created through the establishment of patterns of joint action. Third, we don’t need rules to specify the “aspects under which the system is normative.” Searle’s argument here is based on his ­general view that all intentional contents specify their objects under aspects – a generalization of Frege’s notion of sense.Footnote 5 But this will only lead to the intended conclusion given a prior commitment to conceptualism. If we assume conceptualism – the view that all intentional content is conceptual content – it seems plausible that the presence of normativity entails the presence of rules. If the subjects can conceptualize what’s right or wrong, what they are socially required to do, their competence can also be expressed in the form of rules about what they should do under certain ­circumstances. But to assume conceptualism would be to beg the question against the account of the background in terms of nonconceptual content to be developed now.

5 The Background as Nonconceptual

One important clue that the background is nonconceptual lies in the fact that it is natural to talk about it in terms of having a sense for certain things as well as in terms of certain kinds of feelings and emotional responses, for example, feelings of familiarity – but also of surprise when events deviate from familiar patterns, that is, in cases of “background breakdown” (Searle 1983). As we have seen, Searle also often feels compelled to describe the background in this way. The crucial point now is that having a sense that somebody is a potential cooperation partner or that something is right or wrong, familiar or unfamiliar, is clearly different from having the corresponding concepts or beliefs. For example, children have a sense that certain things are familiar before they develop a concept of familiarity and start wondering whether something is familiar or not and forming beliefs about what is familiar and what is not. Likewise, children will have a sense of what is right and wrong, what they are supposed to do, and what not before they develop concepts of right and wrong, of morality, of rights and obligations, and so on. For example, a child may have a sense that she ought to do what she has promised without having a concept of obligation – or of promising, for that matter. Even dogs (and most likely other animals as well) often appear to have a sense of which kinds of behavior are ­permissible and which are not and display corresponding emotions such as shame. Moreover, we often have a sense of what is unfamiliar, wrong, or inappropriate, without being able to conceptualize what it is that is unfamiliar, wrong, or ­inappropriate, and in which way and why. For example, we often have a sense that there is something unfamiliar about a person’s appearance or demeanor without being able to pinpoint, to form a conception of, what it is that is unfamiliar. The sense or ­feeling of unfamiliarity is holistic, gestaltlike, and undifferentiated – hallmarks of nonconceptual content. Similarly, we often experience speech patterns as unfamiliar, sometimes as jarring, without being able to conceptualize the ways in which these patterns deviate from the patterns we are familiar with. Linguists try to develop theories of grammar partly based on these kinds of experiences, but they exist prior to and independently of a concept of grammaticality, or of concepts such as verb, noun, adjective, or case.Footnote 6 Finally, children, but also adults, often operate on the basis of a sense of what would be morally wrong or right, or even illegal or legal, in the absence of knowledge of the relevant rules or laws and without being able to ­conceptualize what would be wrong about the relevant course of action. They have a feeling of entering forbidden territory without having a conception of the ­boundaries of that territory.

How can these nonconceptual forms of intentionality and normativity, these senses and feelings of what is familiar or unfamiliar, appropriate or inappropriate, right or wrong, have a social dimension such that they may partially embody the identity of a group and its institutions, communal practices, and ways of living? The answer to this is straightforward: they can express the shared background of a group because – and to the extent that – the background skills, tendencies, and habits that they display have been introduced and established in the joint interactions of the group. These background dispositions are reinforced through elementary forms of normativity – negative and positive reactions in the form of actions and emotional responses.

Consider how children jointly develop a game, as they often do, by interacting with one another, picking up on patterns in their interactions and developing shared patterns by responding to one another through action and emotion. Suppose a child is kicking a ball around and repeatedly shoots it through the opening between two trees. Other children come by and after observing this for a while, start to join in the fun. They will try to show that they have got the point of the game, shooting through what is now being established as the goal for this emerging game and trying to prevent the other child from what could almost already be called “scoring.” Certain patterns of going about this will be established as familiar or are already familiar from other games. (New games are often made from old parts.) For example, ­kicking the ball may have been established as the familiar and thus acceptable way of ­moving it. What if one of the players deviates from these patterns, for example, by picking up the ball and running it through the goal? This could be simply rejected, even met with outrage as a blatant violation of the spirit of the game, thus reinforcing the pattern that the ball is kicked rather than carried. But it could also be embraced as a fun, even genius extension of the game. That might be more likely if it is done under special circumstances. For example, perhaps carrying the ball is acceptable after it has first been kicked and then caught in the air. The other player(s) might copy this move, such that it will be established as an acceptable move in the game. The crucial point is that all this can happen in the interaction, without rules being formulated in any way. Certain patterns of interaction are established as familiar and thus acceptable. Deviations from these patterns are sanctioned ­negatively or positively through actions and emotional responses, leading to ­reinforcement or modification of these patterns and thus of the game that is being jointly played. This is a basic kind of normativity, and it does not depend on the presence of rules. It is not essential that adults who know the rules give the feedback as in Searle’s baseball example. It is sufficient that players react normatively to one another. Their emotional reactions are primitive forms of directives and evaluations. In this way, common (shared, collective) background dispositions, common skills, habits, and tendencies are established. Because this happens at the nonconceptual level of action, perception, and emotion, we can also think of these dispositions as sensory-motor-emotional schemata, thus extending the established concept of sensory-­motor schemata by adding the crucial emotional dimension. This dimension is essential both because the dispositions are created partially through ­emotional reactions and because they are themselves displayed in emotional responses, for example, in the surprise at deviations from the patterns established and thus ­perceived as familiar.

Let me now argue for this account by way of comparing it with Searle’s. Perhaps the strongest argument in favor of the present account derives from the fact that in the baseball example, it can localize the relevant kinds of intentionality and normativity in the minds of the children themselves rather than just in the minds of the observing adults. Why should the fact that there is rule intentionality and ­normativity in the adult’s mind be sufficient to confer such normativity on the children’s ­playing? Conversely, why should the children’s practice stand in need of such outside supervision? Again, there seem to be many cases where such practices evolve without such outside involvement. Moreover, I now want to argue: there is a dilemma for Searle’s position if we think about how the connection between the adults’ rule-based intentionality and the children’s actions would have to look like in order for the former to be normatively relevant. It seems we must ascribe to the children some kind of understanding of the rule-based normative responses of the adults observing the game in order to forge a relevant link between the rules and the children’s behavior. It wouldn’t be sufficient for them to react only to the merely physical properties of the relevant symbols, gestures, and so on. Now this link-­forging understanding is either conceptual or it isn’t. If it is conceptual, the conceptual level is doing the work. To fully understand what the adults mean, the children would need to understand the rules reflected in their responses. So if this horn of the dilemma is grasped, conceptual level rule intentionality rather than the background is doing the work. If on the other hand the children’s understanding in some sense falls short of the adults’ conceptual level rule intentionality – and that of course is the entire point of the thought experiment – then it can’t be essential either that the conditioning role should be played by rule intentionality. What is beyond the children’s understanding can’t be essential to the intentionality inherent in their practice. And that of course was only to be expected since a practice of this kind can exist without being guided by rule intentionality. In other words, if the children’s understanding is at a nonconceptual level, so can the normative reactions that they are understanding. The ­conditioning role can equally well be played by a form of intentionality that is on the same level as the children’s response and that means they can also simply respond to one another.

Note further, for what it is worth, that the proposed account rather than Searle’s opens up the possibility of inverting the traditional explanation. As was argued, on Searle’s account, the rules are just further removed and act by way of other people, but they are still in their customary role as causes of the relevant behavior. By ­contrast, on the present view, the patterns established in joint action and perception can be the basis for, and also (part of) the cause of, social rules. This will be the case when, for example, the kids from our example will negotiate and establish rules to codify the patterns they have established, to legislate contentious cases and ­conflicts, to make their game more easily accessible to others, or for any number of other reasons. This building of a level of conceptual level rule intentionality on top of the nonconceptual level of collective sensory-motor-emotional schemata will soon be discussed further.

Let me now address two objections to the present account coming from ­opposite directions. The first objection questions whether this account is really so different from traditional accounts as I make it out to be. Aren’t the negative and positive emotional reactions I have been talking about rather like the reactions of assent and dissent that have often been invoked in the history of ideas, for example, in ­interpretationist – and, one might add, more often than not also behaviorist – ­philosophy of language? Now, I don’t want to deny any similarity, but there are at least two ­fundamental differences. First, the responses I have in mind are primarily responses to actions – rather than to sentences or propositions – and, again, there is a mental component to these actions, and it is nonpropositional and nonconceptual. Second, even though the responses normally have a negative or positive valence – but sometimes they will be “mixed” – they are not binary items displaying as yes/no polarity. There is rather a rich repertoire of emotions: excitement, joy, exhilaration, and pride, but also annoyance, outrage, disappointment, shame, and many more, and all these come in various degrees. So the picture I have sketched is sufficiently different from one of people merely assenting to or dissenting from propositions.

But is it too different in order for the notion of normativity to be applicable? This is the suspicion underlying the second objection. “Normative,” the objector points out, “is derived from “norm,” and what are norms if not rules, laws, and the like?” As a purely terminological comment, this remark is well taken. It may indeed be more appropriate to use a term such as “proto-normative” instead, and I will do so from now on. However, the crucial point for present purposes is that, whatever we choose to call these responses, their causal role is similar to that of rules, laws, directives, and of evaluative statements and attitudes usually also discussed under the heading of “normativity.” They shape people’s behavior – and often much more effectively than rules or other forms of discursive, conceptual level instruction. Moreover, as I have argued already, if we look at elementary forms of joint action, we find that collective action patterns are established through sensory-motor-emotional responses rather than by way of rules.

The upshot is that we don’t need to choose between behaviorism on the one hand and the conceptual level intentionality of rules on the other, as Searle and an entire tradition of thinking about these issues suppose. There is something in between, the level of joint nonconceptual sensory-motor-emotional schemata and capacities I have described. And on that level, we also find something between brute behavior patterns and the full-blown normativity of conceptual level rules and laws: the proto-normativity inherent in shared action patterns established in joint interactions through emotional and actional responses.

6 The Social Background and Layers of Collective Intentionality

There are still a number of questions that I need to say more about, and I want to conclude this chapter by giving at least a rough sketch of how I think these questions ought to be answered. First, more needs to be said about how nonconceptual representations are different from conceptual level ones. So far I have used a number of examples to distinguish them and the contrast between having a sense of something and having a corresponding belief, but we need more theoretically advanced criteria. Fortunately, we can largely rely on well-known criteria from the literature on nonconceptual content here. Even though this literature (e.g., Gunther 2003; Bermúdez and Cahen 2012) is mostly about perception, the criteria discussed there carry over to action (Pacherie 2011) and also to the social domain, including joint action and perception. So in the present context, I will be content to mention some of these criteria and to illustrate how they apply to the social domain. Second, we should get clear about if and how the background, when reconceptualized as nonconceptual, can still fulfill some or all of the functions ascribed to it in Searle’s account. Third, building on the first two points, I want to say at least a little bit about how nonconceptual and conceptual forms of collective intentionality are related in a layered picture of the social mind, and finally, I will give a glimpse of how this picture might be extended to include the institutional world of states and organizations of various kinds.

So let us begin with the criteria for distinguishing nonconceptual from conceptual content. One fundamental feature of nonconceptual states is that they are ­independent from thought and that also means from beliefs in the theoretical and intentions in the practical domain. The best-known example for this is the belief-­independence of perceptual illusions (Evans 1982). An example from the social domain is the independence of joint action patterns from collective intentions. For example, the members of a football team may have evolved a pattern of getting ­careless or of playing too defensively after taking a lead that proves recalcitrant to contrary joint intentions, just like perceptual illusions prove resist contrary beliefs. Paradigmatic social emotions like trust can also be rather recalcitrant to disillusioning experiences recorded in thought. One may still have feelings of trust toward a person in spite of knowing that this person has betrayed that trust many times.Footnote 7

The nonconceptual content of perceptual and actional experiences is also presentational, whereas thought is representational in the sense that it is repeatable and may represent a state of affair that has already been present to the subject. That is, while perceptual and actional experiences, including those of joint action and perception, always concern what is present and thus their intentional content cannot be repeated, there can be several occurrences of the same, practical, or theoretical, thought. I will still, following Searle (1983), sometimes use “representational” as a cover term for all kinds of intentional content, presentational as well as representational.

Conceptual thought is also less context dependent than nonconceptual perception and action. For example, perceptual experiences in all their richness are usually only possible in the presence of the objects of those experiences. Likewise, joint skills may only be accessible in the presence of other group members. For example, the movements of a complex dance routine may only be available in all their intricacy to the members of a dance group in the context of actually performing the dance. The representations used in thought are independent from that context, but they are also not as rich and more abstract. So the dancers can jointly think and reason about how to improve their dance quite apart from the dance context, but the representations they employ when doing so will be impoverished relative to the sensory-motor dance representations.

Sensory-motor-emotional imagery of joint action may of course also be employed in such a context, notably when jointly imagining performing a joint action. Joint imagination is nonconceptual in nature, but intermediate between joint action/­perception and joint thought in terms of both context dependence and richness/concreteness of the relevant representations. That is, while it does not require the context of joint dancing, it does require some kind of face-to-face context for coordination – though this may sometimes be, for example, in the form of a video call. And while the content of joint imaginations is more concrete than that of joint thought, it still falls short of the richness of the experiences in joint action and perception.

Conceptual thought also differs from nonconceptual forms of representation through representational role differentiation. In thought, there is a structure ­determined through singular terms, concepts, and other elements that have distinct representational roles within thoughts and their linguistic expressions. There is no equivalent structure in actional and perceptual experience, including the experience of joint action and perception, though conceptual and linguistic structure may be prefigured in and emerge from figure/gestalt structures in those experiences (Langacker 1987/1991). This point is closely related to the observation that nonconceptual experience and representation is continuous, whereas thought comes in ­discrete, discontinuous units – individual thoughts.

Another closely related point is that the nonconceptual level is characterized through the absence of logical operators. There is no “and,” “or,” and “all” on the level of actional and perceptual experience, including their joint varieties. This observation in turn is related to one regarding certain epistemological features of the nonconceptual level. Though we sometimes of course act hesitantly, like when I waveringly pass to my teammate in a football match, full-blown doubt belongs on the level of thought rather than on the sensory-motor-emotional one. Only in thought do we doubt our strategy, do we wonder whether it was the right one, and do we deliberate, consider, and weigh reasons for and against it, or for and against our ­collective belief. We are always operating in a logical space that also contains the negation of our practical or theoretical attitude. By contrast, on the nonconceptual level, we just take in how the world is and act accordingly.

Nonconceptual representation is also dense and gestaltlike relative to conceptual representation. Conceptually we differentiate features that at the nonconceptual level we only experience as gestalts or packages. For example, while we conceptually separate color and shape, we perceptually experience them as a single package when perceiving ordinary objects. Similarly, when we have a sense of somebody as a potential cooperation partner, this experience is gestaltlike and not differentiated with regard to whether, for example, he or she would make a good business, bridge, or bowling partner.

Let me now summarize the account of the background in terms of nonconceptual content developed so far by giving a more extensive analysis of this last example that has accompanied us for the entire text. To have a sense of somebody as a potential cooperation partner means to possess certain sensory-motor-emotional dispositions and experiences: to perceive that being in a certain way; to be ready for certain kinds of interactions, for example, to smile at it and to approach it in a friendly manner to point at something and assume a position of joint attention vis-à-vis some object; to be disposed to certain emotional reactions, for example, to be surprised, even shocked, and probably angry, if the putative cooperation partner responds with an angry growl or not at all; and by contrast, to experience a friendly and cooperative response as familiar and reassuring. While we may say, if we want to, that in virtue of being habituated to certain patterns of interaction we are expecting certain events, but not others, it is important not to confuse expectations in this sense with beliefs and the sense of the other as a potential cooperation partner with an application of the corresponding concept.Footnote 8 To have the concept of a cooperation partner would enable its possessor to think that somebody is a cooperation partner apart from any perceptual context; to doubt and wonder whether he will really cooperate; to consider reasons for and against this and settle on a corresponding belief; to consider reasons for and against certain courses of actions and settle on corresponding plans; and to wonder whether he will be equally cooperative in business, bridge and bowling, and so on. And it seems plausible that a sense of somebody as a cooperation partner can exist in the absence of these conceptual capacities. For example, small children and even animals may have such a sense without being able to think the thought that somebody is a cooperation partner. Moreover, one may have the sense that somebody is a cooperation partner in spite of a contrary belief and conversely.

It may be objected that the conceptual capacities listed are themselves hierarchically structured such that, for example, one might be able to think the thought that somebody is a cooperation partner without being able to wonder or doubt whether he will really cooperate. This point is well taken; one should think in terms of such a hierarchy of conceptual capacities, and I also want to be clear that the boundary between conceptual and nonconceptual capacities is likely not very sharp either. But this does not invalidate the distinction or make it less important.

To summarize the account given so far, there are background skills, habits, ­tendencies, and other dispositions, but the background as conceived here is still conscious and representational because it is manifest in experiences with presentational contents. These manifestations of the background can be straightforwardly distinguished from other intentional states through their nonconceptual nature. In this way, the notion of the background does not only make the purely negative point that there are no (unconscious) rules behind the exercise of certain elementary skills, but we can also give a clear positive account of Searle’s examples. To mention once more our favorite ones, we now have the tools to more precisely characterize the difference between having a sense of something – say, as being a cooperation partner or a linguistic error – and the corresponding beliefs. And as I argued at some length, we can also give an account of the elementary normativity or proto-­normativity inherent in group practices and jointly established patterns of collective action that is more satisfactory than Searle’s and overcomes the strict dichotomy between rule-governed and “brute,” nonintentional behavior that has characterized much thinking in the last 70 years or more.

This account of the background is in some sense “enactivist” or “interactionist” because it thinks of basic forms of normativity and intentionality as being embodied in perception, action, and patterns of social interaction and joint action. But it parts company with popular versions of enactivism and interactionism when such views show anti-representationalist, anti-intentionalist, and even behaviorist tendencies. To emphasize the onto- and phylogenetic primacy of action, perception and interaction over thought, rules, and deliberation should not mean to diminish the all importance of consciousness, intentionality, and representation. Opposing tendencies generally reveal a thought-, rule-, and deliberation-centric view of consciousness, representation, and intentionality. An anti-intentionalist and behaviorist interactionism about elementary forms of normativity or proto-normativity is like a mirror image of Searle’s rule-centric account. Both accept a dualism of rules, intentionality, and consciousness on one side and “mere” patterns of action and social interaction on the other, while I believe we should try to overcome this dualism by giving an account of perception and action, including social interaction and joint action, in terms of perceptual and actional experiences with nonconceptual intentional content. In this chapter, I have given a rough outline for such an account.

Where does this leave us with regard to some of the functions Searle ascribes to the background, such as that intentional states only determine their conditions of satisfaction relative to the background and that the background “fixes their application”? I think we get a similar picture provided we are willing to replace the reference to intentional states with a reference to conceptual level intentional states and to accept that fixing the application of these states may involve going beyond their intentional contents, filling in their blanks as it were. For example, our joint intention to execute a pass play only determines conditions of satisfaction relative to our joint background skills of playing and executing pass plays. If there was no joint practice of executing pass plays, there would be no clear sense of what intending to execute a pass play would amount to and what might satisfy such an intention. This is not to deny that we might sometimes devise a game or other practice in a top-­down way, at the green table as it were, and only then realize it in joint action. But in so establishing patterns of joint action, we would also be giving a clearer meaning to the plans and rules we had devised, and the whole enterprise would only be possible if we could build on other existing practices, say of playing other games. The conceptual level need not be tied to the background at each point, but some relevant capacities must be available for conceptual level thought to be able to determine conditions of satisfaction. And when we actualize our skill to execute our intention, we will need to fill in the blanks that the more abstract, schematic ­conceptual level representation had to leave out, but which our more concrete, richer, sensory-motor-emotional experiences of executing the play will fill in.

From this vantage point, we can now also see how the notion of the background can be naturally extended in the context of a layered conception of the mind in general and the social mind and the social world in particular. So far we have used the notion of the background to refer to the lowest, nonconceptual layer or level of the social mind, and I have just argued that the conceptual level only functions against the background in this sense. But it seems plausible to extend this picture and distinguish further layers – with the understanding that these distinctions are not necessarily very sharp and just for purposes of orientation. In particular, I’m drawn toward a tripartite division. Something important seems to happen when written language and other forms of documentation are introduced into social practices, and so I believe we should recognize a corresponding documental level.Footnote 9 We can illustrate this tripartite distinction of layers by means of our football example. At the bottom, nonconceptual level, acceptable patterns of jointly kicking a ball are established through sensory-motor-emotional interactions. At the conceptual level, corresponding concepts such as “goal” and “offside” are introduced, and rules employing these concepts are negotiated and jointly formulated. The next level is reached when these rules are written down. With documentation, a qualitative jump with regard to such features as the degree of stability of these representations and their context independence occurs. The game or other social practice now becomes much more independent of the context of face-to-face sensory-motor and linguistic interaction and can reach much larger groups of people in a standardized form. Institutions such as clubs and international organizations such as FIFA – which are unthinkable without documentation – are made possible. But the documental level is just as much dependent on the conceptual level – and thus ultimately on the nonconceptual level – as the former is on the latter. Accordingly, we can say now say generally that higher level function against the background of the lower level ones. Rulebooks and institutions like FIFA can only function against the background of people who know the rules and the statutes of the institutions and have beliefs, intentions, and other conceptual level attitudes with regard to them. At the level of political organization and the state, constitutions can only determine conditions of satisfaction and be applied against the background of a public understanding of their provisions and the concepts that they involve. For example, Werner Binder (2013) argues convincingly that the US Supreme Court adapted its interpretation of the US constitution to shifts in the collective background understanding of certain key ­concepts induced by the events of 9/11 and their effect on collective consciousness and the public sphere. So there is much potential for the notion of the background to help us understand the different layers of the social mind and world and their relation, but to explore this further must be left for another occasion.