Toward a Normative Pragmatics
An ounce of practice is worth a pound of precept.
ENGLISH PROVERB
I. INTRODUCTION
1. Saying 'We'
'We' is said in many ways. We may be thee and me. We may be all that talks or all that moves, all that minds or all that matters. Since these boundaries are elastic, we have a task of demarcation: telling who or what we are, distinguishing ourselves from the other sorts of objects or organisms we find in our world. Saying who we are can contract to an empty exercise in self-congratulation-a ritual rehearsal of the endless, pitiable disabilities of clockworks, carrots, cows, and the clan across the river. Such a mean-spirited version of the demarcational enterprise is not forced on us by the way things are, however.
For what we are is made as much as found, decided as well as discovered.
The sort of thing we are depends, in part, on what we take ourselves to be. One characteristic way we develop and make ourselves into what we are is by expressing, exploring, and clarifying our understanding of what we are. Arbitrary distinctions of biology, geography, culture, or preference can be and have been seized on to enforce and make intelligible the crucial distinction between us and them (or it). But philosophical thought is coeval with the impulse to understand ourselves according to a more principled, less parochial story-and so to be a more principled, less parochial sort of being.
4 Toward a Normative Pragmatics toward a Normative Pragmatics 5
The wider perspective enjoined by principle poses the question, Who are we? in the form: What would have to be true-not only of the quaint folk across the river, but of chimpanzees, dolphins, gaseous extraterrestrials, or digital computers (things in many ways quite different from the rest of us)-for them nonetheless to be correctly counted among us? Putting the issue this way acknowledges an expansive demarcational commitment to avoid, as far as possible, requiring the sharing of adventitious stigmata of origin or material constitution. In understanding ourselves we should look to conditions at once more abstract and more practical, which concern what we are able to do, rather than where we come from or what we are made of. Candidates for recognition as belonging among us should be required to share only the fundamental abilities that make possible participation in those central activities by which we thereby define ourselves. How should we think of these?
The most cosmopolitan approach begins with a pluralistic insight. When we ask, who are we? Or what sort of thing are we? The answers can vary without competing. Each one defines a different way of saying 'we'; each kind of 'we' -saying defines a different community, and we find ourselves in many communities. This thought suggests that we think of ourselves in broadest terms as the ones who say 'we'. It points to the one great Community comprising members of all particular communities-the Community of those who say 'we' with and to someone, whether the members of those different particular communities recognize each other or not.
The reflexive character of the proposal that we use self-demarcation as the criterion by which we demarcate ourselves does not suffice to render it purely formal, however. It does not save us the trouble of contentful self-understanding. For until it has been specified in other terms what one must be able to do in order to count as "saying 'we'," demarcation by appeal to such attitudes remains an aspiration tacked to a slogan-empty, waiting for us to fill it. 'We'-saying of the sort that might be of demarcational interest is not a matter merely of the production of certain vocables-indeed perhaps the relevant kind of attitude is not a linguistic matter at all. Nor again does it consist simply in the engendering of warm mammalian fellow-feeling. Making explicit to ourselves who we are requires a theoretical account of what it is in practice to treat another as one of us.
2. Sapience
What is it we do that is so special? The answer to be explored here-a traditional one, to be sure-is that we are distinguished by capacities that are broadly cognitive. Our transactions with other things, and with each other, in a special and characteristic sense mean something to us, they have a conceptual content for us, we understand them in one way rather than another. It is this demarcational strategy that underlies the classical identification of us as reasonable beings. Reason is as nothing to the beasts of the field. We are the ones on whom reasons are binding, who are subject to the peculiar force of the better reason.
This force is a species of normative force, a rational ‘ought’. Being rational is being bound or constrained by these norms, being subject to the authority of reasons. Saying 'we' in this sense is placing ourselves and each other in the space of reasons, by giving and asking for reasons for our attitudes and performances. Adopting this sort of practical stance is taking or treating ourselves as subjects of cognition and action; for attitudes we adopt in response to environing stimuli count as beliefs just insofar as they can serve as and stand in need of reasons, and the acts we perform count as actions just insofar as it is proper to offer and inquire after reasons for them. Our attitudes and acts exhibit an intelligible content, a content that can be grasped or understood, by being caught up in a web of reasons, by being inferentially articulated. Understanding in this favored sense is a grasp of reasons, mastery of proprieties of theoretical and practical inference. To identify ourselves as rational-as the ones who live and move and have our being in the space of reasons, and so to whom things can be intelligible-is
to seize demarcationally on a capacity that might well be shared by beings quite different from us in provenance and demeanor.
Picking us out by our capacity for reason and understanding expresses a commitment to take sapience, rather than sentience as the constellation of characteristics that distinguishes us. Sentience is what we share with non verbal animals such as cats-the capacity to be aware in the sense of being awake. Sentience, which so far as our understanding yet reaches is an exclusively biological phenomenon, is in turn to be distinguished from the mere reliable differential responsiveness we sentients share with artifacts such as thermostats and land mines. Sapience concerns understanding or intelligence, rather than irritability or arousal. One is treating something as sapient insofar as one explains its behavior by attributing to it intentional states such as belief and desire as constituting reasons for that behavior.
Another familiar route to understanding the sort of sapience being considered here for demarcational duty goes through the concept of truth, rather than that of inference. We are believers, and believing is taking-true. We are agents, and acting is making-true. To be sapient is to have states such as belief, desire, and intention, which are contentful in the sense that the question can appropriately be raised under what circumstances what is believed, desired, or intended would be true. Understanding such content is grasping the conditions necessary and sufficient for its truth.
These two ways of conceiving sapience, in terms of inference and in terms of truth, have as their common explanatory target contents distinguished by their propositional form. What we can offer as a reason, what we can take or make true, has a propositional content: a content of the sort that we express by the use of declarative sentences and ascribe by the use of 'that' clauses.
6. toward a Normative Pragmatics
Toward a normative pragmatics
Propositional contents stand in inferential relations, and they have truth conditions. One of the tasks of this work is to explain what it is to grasp specifically propositional contents, and so to explain who we are as rational or sapient beings. A central subsidiary task is accordingly to offer an account of the relation between the concepts of inference and truth, which complement one another and in some measure compete with one another for explanatory priority in addressing the issue of propositional contentfulness, and so of rationality.
3. Intentionality
The general self-understanding in view so far identifies us by our broadly cognitive capacities: We are makers and takers of reasons, seekers and speakers of truth. The propositional focus of the approach marks this understanding of intelligible contents as discursive. This conception, hallowed by ancient tradition, was challenged during the Enlightenment by a rival approach to cognitive contentfulness that centers on the concept of representation. Descartes’ seminal demarcational story distinguishes us as representers-producers and consumers of representings-from a world of merely represented and representable things. The states and acts characteristic of us are in a special sense of, about, or directed at things. They are representings, which is to say that they have representative content. To have such content is to be liable to assessments of correctness of representation, which is a special way of being answerable or responsible to what is represented.
Another task of this work is accordingly to address the question, How should the relation between representation-the master concept of Enlightenment epistemology-and the discursive concepts of reasons and truth be understood? One of the great strengths representationalist explanatory strategies have developed is the capacity to offer accounts of truth and goodness of inference. There are familiar set-theoretic routes that set out from representational primitives corresponding to sub sentential linguistic expressions such as singular terms and predicates, lead to assignments of truth conditions to sentences compounded out of those expressions, and pass from there to determinations of which inferences are correct. While doubts have been raised, perhaps legitimately, about nearly every phase of this construction, no other semantic approach has been worked out so well.
Yet for all that, the primitives involved have never been well understood.
Descartes notoriously fails to offer an account either of the nature of representational contentsof what the representingness of representings consists inor of what it is to grasp or understand such contents, that is to say, of their intelligibility to the representer. He does not tell us what makes a rabbit-idea an idea of (or purporting to be of) rabbits, or of anything at all, nor what it is for the one whose idea it is to understand or take it as being of or about something. Those things could be represented by and to the mind (have
"objective reality" in it and for it, for the mind to be "as if of" things) is treated as a basic property, an unexplained explainer. But an adequate treatment of the representational dimension of discursive sapience should include an account both of representational purport, and of its uptake.
The topic to be investigated here, then, is intentionality in the sense of the propositional contentfulness of attitudes, not in the sense (if that should turn out to be different) of the directedness of sense. The aim is to understand ourselves as judgers and agents, as concept-users who can reason both theoretically and practically. This is not to say that we should understand our selves exclusively as sapients rather than sentients, in terms of understanding rather than awareness. 'We' is and by rights ought to be said in many ways. The point is just to register and delineate the way that is to be discussed here.
This inquiry is directed at the fanciest sort of intentionality, one that involves expressive capacities that cannot be made sense of apart from participation in linguistic practices. The aim is to offer sufficient conditions for a system of social practices to count as specifically linguistic practices, in the sense of defining an idiom that confers recognizably propositional contents on expressions, performances, and attitudes suitably caught up in those practices. Looking at this sort of high-grade intentionality accordingly risks being beastly to the beasts-not only by emphasizing sapience over sentience, comprehension over consciousness, but also by unfairly ignoring the sorts of beliefs and desires that are appropriately attributed to non- or pre-linguistic animals.
So it is a further criterion of adequacy of this explanatory enterprise that it have something to say about the lower grades of intentionality: not only as to how the lines should be drawn (corresponding to different senses of 'we'), but also as to how the advent of the favored sort of linguistic intentionality can be made less mysterious. How can linguistic abilities arise out of non-linguistic ones? Or to ask a related question, What would sentient creatures have to be able to do in order to count as sapient as well? What is needed is to tell a story about practices that are sufficient to confer propositionally contentful intentional states on those who engage in them, without presupposing such states on the part of the practitioners. The hope is that doing so will offer guidance concerning what would be involved in diagnosing aliens as exhibiting such states, and programming computers or teaching merely sentient animals to exhibit them.
II. FROM INTENTIONAL STATE TO NORMATIVE STATUS
1. Kant: Demarcation by Norms
The demarcational proposal being pursued picks us out as the ones capable of judgment and action. Not only do we respond differentially to environing stimuli, we respond by forming perceptual judgments. Not
8 Toward a Normative Pragmatics
Toward a Normative Pragmatics 9
Not only do we produce behavior, we perform actions. Various ways of talking about this fundamental distinction have been put on the table. It can be made out in terms of truth. In perception what we do is responsively take-true some propositional content that is intelligible to us. In action what we do is responsively make-true some propositional content that is intelligible to us.
The distinction can be made out in terms of reasons. The judgments that
are our perceptual responses to what is going on around us differ from responses that are not propositionally contentful (and so are not in that sense intelligible) in that they can serve as reasons, as premises from which further conclusions can be drawn. Actions, which alter what is going on around us in response to propositionally contentful intentions, differ from performances that are merely behavior (and so not intelligible in terms of the propositionally contentful intentions that elicit them) in that reasons can be given for them; they can appear as the conclusions of practical inferences.
The distinction can also be made out in terms of the employment of concepts. To be a perceiver rather than just an irritable organism is to be disposed to respond reliably and differentially to the perceptible environment by the application of appropriate concepts. To be an agent rather than just a behaver is to be disposed to respond reliably and differentially to applications of appropriate concepts by altering the accessible environment. Intelligibility in the sense of propositional contentfulness, whether the latter is conceived in terms of truth conditions or capacity to serve as a reason, is a matter of conceptual articulation-in the case of perception and action, that the rely ably elicited response and the reliably eliciting stimulus, respectively, essentially involve the use of concepts.
So sapience, discursive intentionality, is concept-mongering. What is distinctive of specifically conceptual activity? Contemporary thought about the use of concepts owes great debts to Kant. One of his cardinal innovations is his introduction of the idea that conceptually structured activity is distinguished by its normative character. His fundamental insight is that judgements and actions are to be understood to begin with in terms of the special way in which we are responsible for them.
Kant understands concepts as having the form of rules, which is to say
That they specify how something ought (according to the rule) to be done.1
The understanding, the conceptual faculty, is the faculty of grasping rules-of appreciating the distinction between correct and incorrect application they determine. What is distinctive about judging’s and doings-acts that have contents that one can take or make true and for which the demand for reasons is in order is the way they are governed by rules. They are conceptually contentful and so are subject to evaluation according to the rules that express those contents. Being in an intentional state or performing an intentional action accordingly has a normative significance. It counts as undertaking (acquiring) an obligation or commitment; the content of the commitment is determined by the rules that are the concepts in terms of which the act or
State is articulated. Thus Kant's version of the sort of demarcation criterion being considered picks us out as distinctively normative, or rule-governed, creatures.
2. from Cartesian Certainty to Kantian Necessity
This emphasis on the normative significance of attributions of intentionally contentful states marks a decisive difference between Kantian and Cartesian ways of conceiving cognition and action. For Kant the important line is not that separating the mental and the material as two matter-of factually different kinds of stuff. It is rather that separating what is subject to certain kinds of normative assessment and what is not. For Descartes, having a mind (grasping intentional contents) is having representings: states that purport or seem to represent something. Some things in the world exhibit this sort of property; others do not. Where Descartes puts forward a descriptive conception of intentionality, Kant puts forward a normative, or prescriptive, one-what matters is being the subject not of properties of a certain kind but of proprieties of a certain kind. The key to the conceptual is to be found not by investigating a special sort of mental substance that must be manipulated in applying concepts but by investigating the special sort of authority one becomes subject to in applying concepts-the way in which conceptually articulated acts are liable to assessments of correctness and incorrectness according to the concepts they involve.
This approach contrasts sharply with Cartesian demarcations of cognition and action according to the presence of items of a certain matter-of-factual kind. The objection is not to the details of Descartes’ understanding of the descriptive features required for intentionally contentful states and acts: his conception of mental events that are self-intimating cognitions or infallibly performable volitions, takings-true and makings-true that are minimal in that they cannot fail to be successful. It is, more radically, that what sets off the intentional is its liability to assessments of correctness, its being subject to norms (which are understood as codified in rules), rather than any missing feature that it could be described as having or lacking.
Descartes inaugurated a new philosophical era by conceiving of what he took to be the ontological distinction between the mental and the physical in epistemological terms: in terms of accessibility to cognition-in terms, ultimately, of certainty. Kant launched a new philosophical epoch by shifting the center of concern from certainty to necessity. Where Descartes’ descriptive conception of intentionality, centering on certainty, picks out as essential our grip on the concepts employed in cognition and action, Kant's normative conception of intentionality, centering on necessity, treats their grip on us as the heart of the matter. The attempt to understand the source, nature, and significance of the norms implicit in our concepts-both those that govern the theoretical employment of concepts in inquiry and knowledge
10 Toward a Normative Pragmatics toward a Normative Pragmatics 1
edge and those that govern their practical employment in deliberation and action-stands at the very center of Kant's philosophical enterprise. The most urgent question for Kant is how to understand the rulishness of concepts, how to understand their authority, bindingness, or validity. It is this normative character that he calls Notwendigkeit (necessity).
The nature and significance of the sea change from Cartesian certainty to Kantian necessity will be misunderstood unless it is kept in mind that by 'necessary' Kant means ‘in accord with a rule'. It is in this sense that he is entitled to talk about the natural necessity whose recognition is implicit in cognitive or theoretical activity, and the moral necessity whose recognition is implicit in practical activity, as species of one genus. The key concept of each is obligation by a rule. It is tempting, but misleading, to understand Kant’s use of the notion of necessity anachronistically, in terms of contemporary discussions of alethic modality. It is misleading because Kant's concerns are at base normative, in the sense that the fundamental categories are those of deontic modality, of commitment and entitlement, rather than of alethic modality, of necessity and possibility as those terms are used today. Kant's commitment to the primacy of the practical consists in seeing both theoretical and practical consciousness, cognitive and co native activity, in these ultimately normative terms.
So for Kant, concepts are to be understood by the theorist in terms of the rules that make them explicit, rules that specify how the concepts are properly or correctly applied and otherwise employed. Kant’s appreciation of this normative significance of concept use is one of the lenses through which he views his relationship to his rationalist and empiricist predecessors. With the wisdom of hindsight, Kant can see a normative strand of concern with responsibility as fundamental to the Enlightenment.
2 Thus the Meditations is to be read as motivated by the demand that the meditator take personal responsibility for every claim officially endorsed-be prepared to answer for it demonstrate entitlement to that commitment by justifying it. This theme remained merely implicit in Descartes’ theorizing about us (as opposed to his motivation and methodology), for his explicit theory remains naturalistic (though not, of course, physicalistic).
Leibniz insists, against the empiricists, that inferential transitions between representations ought not to be assimilated to matter-of-factual, habitually acquired causal dispositions. He understands them rather as applications of general principles that must accordingly be available prior to any knowledge of empirical matters of fact. Kant takes over from his reading of Leibniz the general idea of rules as what underwrite cognitive assessments of inferences and judgments. He understands such a priori principles, however, not as very general statements of fact (even metaphysical fact), but as rules of reasoning. They are conceived not as descriptive but as prescriptive as (in Sellars' phrase) “fraught with ought."
This lesson dovetails neatly with the moral he draws from Hume’s thought. On Kant's reading, Hume's contribution is to see that ordinary empirical discourse involves commitments that reach beyond the sequences of representations, however regular, in which the concepts deployed in that discourse are taken to originate. Kant’s Hume recognizes that cognitive experience crucially involves the application and assessment of the correctness of the application of rules. For Kant, Hume's inquiry after the nature of the authority for this inferential extension takes the form of a quest for the nature of the necessity, understood as normative bindingness, exhibited by the rules implicit in empirical concepts. It is under this conception that Kant can assimilate Hume's point about the distinction between saying what happens (describing a regularity) and saying what is causally necessary (pre scribing a rule) to his point about the distinction between saying what is and saying what ought to be. One need not buy the metaphysics that Kant uses to ground and explain his norms, nor accept his answer to Hume, in order to appreciate the transformation of perspective made possible by his emphasis on the normativeness of the conceptual, and hence of cognition and action the latter distinguished in the first instance as what we are responsible for.
3. Frege: Justification versus Causation
Kant's lesson is taken over as a central theme by Frege, whose campaign against psychologism relies on respecting and enforcing the distinction between the normative significance of applying concepts and the causal consequences of doing so. For Frege, it is possible to investigate in a naturalistic way acts of judging or thinking (even thinking conceived in a dualistic way), but such an investigation inevitably overlooks the normative dimension that is essential to understanding the propositional contents that are judged or thought. Sometimes this point is put in terms of reasons, invoking inferential relations among judgeable contents, as when he com plains that psychologism "loses the distinction between the grounds that justify a conviction and the causes that actually produce it,"3 or again when he argues that "the laws in accordance with which we actually draw inferences are not to be identified with the laws of correct richtigen inference; otherwise we could never draw a wrong inference.?" Sometimes the point is put in terms of truth, as when he says, "It is not the holding something to be true that concerns us, but the laws of truth. We can also think of these as prescriptions for making judgments; we must comply with them if our judgments are not to fail of the truth.t''' Put either way, the point is that concern with the propositional contents that are thought or judged is inseparable from the possibility of assessments of correctness. Besides empirical regularities, there are also proprieties governing inferring and holding-true. Besides questions of which judgeable contents are held true, and under what circum stances, there is the question of which ones ought to be, and when. Besides the question of what consequences holding-true or making a judgment with
12 Toward a Normative Pragmatics
Toward a Normative Pragmatics 13
a certain content in fact leads to, there is the question of what those consequences ought or must rationally be. Psychology can study the matter-of-factual properties of contentful acts of judging and inferring, but not the semantically determined proprieties that govern them, the norms according to which assessments of truth and rationality are to be made.
Psychologism misunderstands the pragmatic significance of semantic contents. It cannot make intelligible the applicability of norms governing the acts that exhibit them. The force of those acts is a prescriptive rather than a descriptive affair; apart from their liability to assessments of judgments as true and inferences as correct, there is no such thing as judgment or inference. To try to analyze the conceptual contents of judgments in terms of habits or dispositions governing sequences of brain states or mentalistically conceived ideas is to settle on the wrong sort of modality, on causal necessitation rather than rational or cognitive right. Such natural processes "are no more true than false; they are simply processes, as an eddy in the water is a process. And if we are to speak of a right, it can only be the right of things to happen as they do happen. One phantasm contradicts another no more than one eddy in water contradicts another .r''' Contradiction, correct inference, correct judgment are all normative notions, not natural ones.
The laws of nature do not forbid the making of contradictory judgments.
Such judgments are forbidden in a normative sense. It is incorrect to endorse incompatible contents: rationally incorrect, incorrect according to rules of reason, prescriptions governing what is proper in the way of inferring and
Judging. The ‘must’ of justification or good inference is not the ‘must’ of causal compulsion. But the possibility of expressing each in terms of rules or laws, so central to Kant's enterprise, misleads if these two different sorts of laws are not kept distinct, as they are not by psychologism and association ism. "What makes us so prone to embrace such erroneous views is that we define the task of logic as the investigation of the laws of thought, whilst understanding by this expression something on the same footing as the laws of nature ... So if we call them laws of thought, or, better, laws of judgment, we must not forget we are concerned here with laws which, like the principles of morals or the laws of the state, prescribe how we are to act, and do not, like the laws of nature, define the actual course of events
Frege expresses his views about the normative character of judgeable contents, which he understands as having truth conditions, and so about the application of concepts, which he understands as functions whose values are truth-values, by talking about the nature of logic, which he understands as the study of the laws of truth.
Logic, like ethics, can also be called a normative science.
8 The property 'good' has a significance for the latter analogous to that which the property 'true' has for the former. Although our actions and endeavours are all causally conditioned and explicable in psychological terms, they do not all deserve to be called good.9 Discussion of just how these remarks about the normative or prescriptive character of logic relate to a commitment to the normative or prescriptive significance of the exhibition of conceptual content by judgments must await more detailed consideration of Frege's Begriffsschrift theory of logical vocabulary as expressive of conceptual contents, in Chapter 2.
4. Wittgenstein on the Normative Significance of Intentional Content
Frege emphasizes that concern with the contents of concepts and judgments is inseparable from concern with the possibility of the concepts being correctly or incorrectly applied, the judgments correctly or incorrectly made, whether this correctness is conceived in terms of truth or of the goodness of inference. Understanding this point requires distinguishing normative from causal modalities. Beyond enforcing this distinction, however, Frege has little to say about the nature of the norms that matter for the study of conceptual contents, and so for logic-though of course he has a great deal to say about the structure of such contents (some of which will be rehearsed in subsequent chapters). His concerns are at base semantic rather than pragmatic. In the twentieth century, the great proponent of the thesis that intentionally contentful states and acts have an essentially normative pragmatic significance is the later Wittgenstein.
The starting point of his investigations is the insight that our ordinary understanding of states and acts of meaning, understanding, intending, or believing something is an understanding of them as states and acts that commit or oblige us to act and think in various ways. To perform its traditional role, the meaning of a linguistic expression must determine how it would be correct to use it in various contexts. To understand or grasp such a meaning is to be able to distinguish correct from incorrect uses. The view is not restricted to meaning and understanding but extends as well to such intentionally contentful states as believing and intending. This is one way of developing and extending Kant's point that to take what we do as judging and acting is to treat it as subject to certain kinds of assessments as to its correctness: truth (corresponding to the world) and success (corresponding to the intention). A particular belief may actually relate in various ways to how things are, but its content determines how it is appropriate for it to be related, according to the belief-namely that the content of the taking-true should be true. A particular intention may or may not settle how one will act, but its content determines how it is appropriate to act, according to the intention-namely by making-true that content. To say this is in no way to deny that occurrences of intentional states of meaning, understanding, in tending, and believing have causal significances. It is simply to point out that
14 Toward a Normative Pragmatics toward a Normative Pragmatics
15 understanding them as contentful involve understanding them as also having normative significances. L''
The issue constantly before us in Wittgenstein's later works is how to understand these normative significances of intentional contents-the way in which they incorporate standards for assessments of correctness. Many of his most characteristic lines of thought are explorations of the inaptness of thinking of the normative 'force', which determines how it would be appropriate to act, on the model of a special kind of causal 'force'. The sense in which understanding or grasping a meaning is the source of the correct use is quite different from the sense in which it is the source of what one in fact goes on to do.11 Enforcing the Kantian and Freeman distinction between grounds in the order of justification and grounds in the order of causation is what is behind talk of the "hardness of the logical 'must"'12 and the picture of the dominion or compulsion intentional states exercise over what counts as correct performance as a machine whose "super-rigid" construction precludes any sort of malfunction. "The machine as symbolizing its action ... We talk as if these parts could only move in this way, as if they could not do anything else. How is this-do we forget the possibility of their bending, breaking off, melting, and so on? Yes; in many cases we don't think of that at all ... And it is quite true: the movement of the machine-as-symbol is predetermined in a different sense from that in which the movement of any
Is the expression 'in a queer way’? The rest is all right; and the sentence only seems queer when one imagines a different language game for it from the one in which we actually use it"15-the different game, namely, of attributing natural states and properties, rather than normative statuses such as commitments. What is determined is not how one will act but how one ought to, given the sense or content grasped, or the rule one has endorsed. "'How am I able to obey a rule?'-if this is not a question about causes, then it is about the justification for my following the rule in the way I do.1116 That is, it is a question about what actions accord with the rule, are obliged or permitted by it, rather than with what my grasp of it actually makes me do.
5. Norms and Intentional Explanation
Although Wittgenstein often uses specifically linguistic examples, and some commentators have focused exclusively on these cases, the normative phenomena he highlights are part and parcel of intentional attribution generally, whether or not language is in the picture. Ceteris paribus, one who believes that it is raining, and that moving under the tree is the only way to stay dry, and who desires to stay dry, ought to move under the tree. The intentional states make the action appropriate. Indeed, the concept of rationality achieves its paradigmatic application in just such circumstances, as given actual machime i. s predetermm.
e d. 1113
conduct warranted by the attributed intentional states is characterized as The relation between the content of an intention and the performances that would fulfill that intention does not leave any room for misfire, corresponding to the melting or bending of the parts of a mechanism, for it is already a normative relation. The state is to settle what ought to be done, what must be done if it is to be realized. What actually does or would happen is another matter. The images of superrigidity-of being guided by rails that one cannot fall away from-are what one gets if one assimilates normative compulsion to causal compulsion, ignoring the Kantian distinction. That is, if the normative 'must' were a kind of causal 'must', it would have to be a puzzling, superrigid sort-but the point is not to start with this sort of naturalistic prejudice.
In fact, by contrast, "The laws of inference can be said to compel us; in the same sense, that is to say, as other laws make human society. They determine, in a sense yet to be specified, what one ought to do. Being compelled in this sense is entirely compatible with failing to act as one ‘must’. Indeed, the physical or causal possibility of making a mistake, or doing what one is obliged, by what one means, intends, believes, and desires, not to do, is essential to the conception of such states and shows the essentially normative nature of their significance. '"But I don't mean that what I do now (in grasping a sense) determines the future use causally and as a matter of experience, but that in a queer way, the use itself is already present'.-But of course it is, 'in some sense'! Really the only thing wrong with what you say rationally appropriate. The qualification marks the option being reserved to deny that the conduct is, say, morally, politically, or aesthetically appropriate. (It is a further question whether this explanatory role of rationality justifies conceiving of what is rationally appropriate as reducible to what is prudentially or instrumentally appropriate.) Taking the category of rational it to be essentially involved in intentional explanation, as Dennett and Davidson for instance do, is one way of recognizing the normative dimension of intentionality.
It is important to keep this acknowledgment distinct from further theses one may then want to endorse concerning that normative dimension. If one keeps one’s eye resolutely on the causal dimension of intentional explanation, the normative aspect can be masked. For instance, Dennett conjoins his recognition of the constitutive role of rationality in intentional explanation with the claim that such explanation involves a substantive "rationality assumption," the assumption in effect that the system will by and large in fact act as it ought rationally to act. There is nothing wrong with considering explanations of this sort as intentional explanations, but it is important to distinguish normative intentional explanation from causal intentional explanation. The former explains only what the subject of the intentional states ought or is obliged or committed (rationally) to do in virtue of its exhibition of the attributed states. The latter makes the substantive rationality assumption and goes on to explain what in fact happens. Normative intentional
16 Toward a Normative Pragmatics toward a Normative Pragmatics 17
Explanations are more fundamental; they are presupposed and built upon by causal ones. The same normative considerations arise if one approaches intentional states from the direction of their functional roles in mediating perception and action. Where Wittgenstein, Sellars, and Dummett, for instance, look at meaning, understanding, believing, or intending something in terms of mastery of the public proprieties governing the use of linguistic expressions, others would see intentional states as already definable by their role m accounting for the conduct of rational agents, whether linguistically adept or not. Both views are functionalist, in a broad sense. They differ over how to draw the boundaries around the functional systems within which alone something can have the significance of an intentional state. The issue of the extent to which mastery of linguistic social practice is a prerequisite for possession of intentional states of various sorts is of course an important one. But one need not have settled on one or the other of these explanatory approaches in order to appreciate that intentional states belong on the normative side of the Kantian divide. For talk of functional roles is itself already
Normative talk.
Specifying the functional role of some state in a system is specifying how
It ought to behave and interact with other states. It is with reference to such a role that one makes sense of the notion of a malfunctioning component, something that is not behaving as it is supposed to. One may go on to offer various stories about the source of the correctnesses involved in functional roles: invoking the intentions of the designer, the purposes of the user, or the way the system must function if it is to realize the evolutionary good of survival, or even the cognitive good of accurately representing its environment. The point is that all of these are accounts of the source of the norms of proper functioning that are an integral part of functional explanations. The job of a designer's drawing of a machine is to specify how the machine is supposed to work, how it ought to work, according to the intentions of the designer. It is for this reason that “we forget the possibility of the pieces bending, breaking off, melting, and so on." Wittgenstein is of course concerned to understand how it is possible to understand such normative roles. But the current point is just that the roles one seeks to specify, in explaining the significance of intentional states, must, to begin with, be understood in normative terms of proper or correct functioning. Once again, this is not to deny that the fact that some component or system ought (functionally) _to behave in a certain way may under many circumstances have a causal significance regarding how it will in fact behave. The issues are in principle distinct, however, and causal functional accounts presuppose normative Functional ones.
The recognition that the consequences of attributing intentionally contentful states must be specified in normative terms may be summed up in the slogan, "Attributing an intentional state is attributing a normative status." This is one of the leading ideas to be pursued in the present investigation. Intentional states and acts have contents in virtue of which they are essentially liable to evaluations of the "force of the better reason." It is this mysterious "force evidently the core of the social practices of giving and asking for reasons-that Greek philosophy investigated and appealed to in demarcating us from the nonrational background of items that we can think and find out about but that cannot themselves think or find out about other things. This “force of the better reason" is a normative force. It concerns what further beliefs one is committed to acknowledge, what one ought to conclude, what one is committed or entitled to say or do. Talk of what is a reason for what has to do in the first instance not with how people do or would act but with how they should act, what they should acknowledge. The sophist may not in fact respond to this "force," but even the sophist ought to. To understand rationality and states whose contents are articulated according to their role in reasoning, one must understand the force of such “ought's”. The relevance of reasons to the attributing and undertaking of intentional states and acts is prima facie reason to employ a normative meta language in analyzing such activity.
The normative dimension of intentional attributions is equally apparent if the propositional contents of the states and acts that are attributed, exhibited, or performed are conceived, not in terms of their accessibility to reasons, but in terms of there being circumstances under which they would be true. Assessments of truth, no less than assessments of rationality, are normative assessments. Truth and rationality are both forms of correctness. To ask whether a belief is true is to ask whether it is in some sense proper, just as to ask whether there are good reasons for it is to ask whether it is proper in a different sense. The business of truth talk is to evaluate the extent to which a state or act has fulfilled a certain kind of responsibility. This normative aspect of concern with truth can be masked by offering a descriptive, matter-of-factual account of what truth consists in. But doing so should be understood as offering a theory about this variety of semantic correctness, not as a denial that correctness is what is at issue. Thus Dummett argues that one does not understand the concept of truth when one has only a method for determining when it correctly applies to a claim or belief-a practical mastery of its circumstances of application. One must also know the point of applying it, must understand that truth is the proper goal of assertion and belief, that the language game of assertion and belief implicitly but essentially involves the injunction that one ought to speak and believe the truth. That is what one is supposed to be trying to do. Without an appreciation of this normative significance of application of the concept truth, one does not understand that concept.
Raising the question of what a belief or claim represents or is about can be understood as treating it as in a special way answerable for its correctness to what is represented, what it is about.17 Thus the claim that semantically
18 Toward a Normative Pragmatics toward a Normative Pragmatics 19
or intentionally contentful states and acts have, as such, pragmatic significances that should be specified in normative terms does not depend upon what particular model (for instance, reasons, truth conditions, or representation) is employed in understanding such contents. The theoretical task of the intentional content of a state or act is to determine, in context, the normative significance of acquiring that state or performing that act: when it is appropriate or correct to do so and what the appropriate consequences of doing so are. The content is to determine proprieties of use, employment, or performance for states, acts, and expressions that exhibit or express such contents. The content must (in context) settle when it is correct to apply a concept in judging, believing, or claiming, and what correctly follows from such an application. Correctnesses of application are discussed under the general headings of assessments of truth or representation; correctnesses of inference are discussed under the general heading of assessments of rationality.
19 To pick out intentional states and acts as ones to which any of these sorts of assessments-truth, accuracy of representation, or reasonability-are in principle appropriate is to treat their normative articulation as essential to them. For this point, it does not matter which sort of assessment is treated as fundamental, whether the goodness of claiming of the sort concepts of truth try to capture, the goodness of representation that concepts of correspondence try to capture, or the goodness of reasoning of the sort concepts of rationality try to capture. All are prima facie normative or evaluative notions.
III. FROM NORMS EXPLICIT IN RULES TO NORMS IMPLICIT IN PRACTICES
1. Regulism: Norms as Explicit Rules or Principles
The first commitment being attributed to Wittgenstein, then, is to taking the significance of attributing intentional states to be normative, a matter of the difference it makes to the correctness or justification of possible performances (including the adoption of other intentional states). The second commitment he undertakes concerns how to understand the normative stat uses of correct and incorrect, justified and not justified, which this approach to intentionality concentrates on. The question of how the normative significances of intentional states are to be taken to be related to the matter-of factual consequences of those states, which would be one way into this issue, can be put to one side for the moment. It is a question Wittgenstein is much interested in, but it ought to be seen as a rising at a different point in the argument. For an account of the normative pole of the Kantian dualism need not take the form of a specification of how the normative is related to the non-normative. Instead, Wittgenstein considers, and rejects, a particular model of correctness and incorrectness, roughly Kant's, in which what makes
A performance correct or not is its relation to some explicit rule. To understand his argument and the lesson he draws from it, it is necessary to see what this model of the normative is, and for what sort of explanatory role he claims it is unsuitable.
According to this more specific Kantian view,19 norms just are rules of conduct. Normative assessments of performances are understood as always having the form of assessments of the extent to which those performances accord with some rule. Reference to proprieties of performance is taken as indirect reference to rules, which determine what is proper by explicitly saying what is proper. On this account, acts are liable to normative assessments insofar as they are governed by propositionally explicit prescriptions, prohibitions, and permissions. These may be conceived as rules, or alternatively as principles, laws, commands, contracts, or conventions. Each of these determines what one may or must do by saying what one may or must do. For a performance to be correct is, on this model, for the rules to permit or require it, for it to be in accord with principle, for the law to allow or demand it, for it to be commanded or contracted. It is because Kant is someone for whom the normative always appears in the explicit form of rules, laws, and commandments that he could see the rationalists' insistence on the essential role of principles in cognition and action as a dark appreciation of the fundamentally normative character of those faculties. It is for this reason that when Kant wants to say that we are creatures distinguished from others by the normative dimension of our conduct (both cognitive and practical), he puts this in terms of our being bound by rules.
On an approach according to which normative assessment of conduct
whether prospectively, in deliberation, or retrospectively, in appraisal always begins with the question of what rule is followed in producing the performances in question, norms are likened to laws in the sense of statutes. For conduct is legally appropriate or inappropriate just insofar as it is governed by some explicit law that says it is. Assessments of legal praise and blame must at least implicitly appeal to the relation of the performance in question to some law. In this way, the model appeals to a familiar institutional context, in which the norms most in evidence clearly take the form of explicit principles, commands, and the like.
The influence of the jurisprudential analogy is evident in Kant's conception of the normative aspect of cognition and action in terms of following rules. Kant inherits the Enlightenment tradition, handed down from Grotius and Pufendorf, which first studied the normative in the form of positive and natural laws, conceived as the explicit commandments of sovereigns or superiors of one sort or another. As a result, Kant takes it for granted that it is appropriate to call a 'rule' or a 'law' whatever it is that determines the propriety or impropriety of some judgment or performance. For him, as for most philosophers before this century, explicit rules and principles are not simply one form among others that the normative might assume. Rules are
20 Toward a Normative Pragmatics
The form of the norm as such. This view, that proprieties of practice are always and everywhere to be conceived as expressions of the bindingness of underlying principles may be called regulism about norms.
2 According to this intellectualist, Platonist conception of norms, common to Kant and Frege, to assess correctness is always to make at least implicit reference to a rule or principle that determines what is correct by explicitly saying so. In the best-known portion of his discussion of rule-following in the Philosophical Investigations, Wittgenstein argues that proprieties of performance that are governed by explicit rules do not form an autonomous stratum of normative statuses, one that could exist though no other did. Rather, proprieties governed by explicit rules rest on proprieties governed by practice. Norms that are explicit in the form of rules presuppose norms implicit in practices.
20. Wittgenstein's Regress Argument
Norms explicit as rules presuppose norms implicit in practices because a rule specifying how something is correctly done (how a word ought to be used, how a piano ought to be tuned) must be applied to particular, circumstances, and applying a rule in particular circumstances is itself essentially something that can be done correctly or incorrectly. A rule, principle or command has normative significance for performances only in the context of practices determining how it is correctly applied. For any particular performance and any rule, there will be ways of applying the rule so as to force the performance, and ways of applying it so as to permit or require it. The rule determines proprieties of performance only when correctly applied.
If correctnesses of performance are determined by rules only against the background of correctnesses of application of the rule, how are these latter correctnesses to be understood? If the regulist understanding of all norms as rules is right, then applications of a rule should themselves be understood as correct insofar as they accord with some further rule. Only if this is so car the rule-conception play the explanatory role of being the model for understanding all norms. A rule for applying a rule Wittgenstein calls an “interpretation" (Deutung}. "There is an inclination to say: every action according to the rule is an interpretation. But we ought to restrict the term 'interpretation' to the substitution of one expression of the rule for another."21 The question of the autonomy of the intellectualist conception of norms, presupposed by the claim that rules are the form of the normative, is the question of whether the normative can be understood as "rules all the way down," or whether rulish proprieties depend on some more primitive sort of practical propriety. Wittgenstein argues that the latter is the case. Rules do not apply themselves; they determine correctnesses of performance only in the context of practices of distinguishing correct from incorrect applications of the rules. To conceive these practical proprieties of application as themselves rule-governed is to embark on a regress. Sooner or later the theorist will have to
Toward a Normative Pragmatics
21 Acknowledge the existence of practical distinctions between what is appropriate and what not, admitting appropriateness’s according to practice as well as according to rules or explicit principles.
This regress argument shows that the Platonist conception of norms as rules is not an autonomous one, and so does not describe the fundamental form of norm. "What does a game look like that is everywhere bounded by rules? Whose rules never let a doubt creep in, but stop up all the cracks where it might?-Can't we imagine a rule determining the application of a rule, and a doubt which it removes-and so on?
22 In each case the doubt is the possibility of a mistake, of going wrong, of acting incorrectly, for instance in applying a rule. The point is to be that a rule can remove such a doubt, settle what is correct to do, only insofar as it is itself correctly applied. '"But how can a rule show me what I have to do at this point? Whatever I do is, on some interpretation, in accord with the rule.'-That is not what we ought to say, but rather: any interpretation Deutung still hangs in the air along with what it interprets dem Gedeuteten, and cannot give it any support. Interpretations by themselves do not determine meaning.
23 No sequence of interpretations can eliminate the need to apply the final rules, and this is always itself subject to normative assessment. Applied incorrectly, any interpretation misleads. The rule says how to do one thing correctly only on the assumption that one can do something else correctly, namely apply the rule.
This was our paradox: no course of action could be determined by a rule, because every course of action can be made out to accord with the rule. The answer was: if everything can be made out to accord with the rule, then it can also be made out to conflict with it. And so there would be neither accord nor conflict here. It can be seen that there is a misunderstanding here from the mere fact that in the course of our argument we give one interpretation after another; as if each one contented us for at least a moment, until we thought of yet another standing behind it. What this shows is that there is a way of grasping a rule eine Auff assung einer Regel which is not an interpretation, but which is exhibited in what we call "obeying the rule" and "going against it" in actual cases.24
Absent such a practical way of grasping norms, no sense can be made of the distinction between correct and incorrect performance-of the difference between acting according to the norm and acting against it. Norms would then be unintelligible.
3. Wittgenstein’s Pragmatism about Norms
The conclusion of the regress argument is that there is a need for a pragmatist conception of norms-a notion of primitive correctnesses of performance implicit in practice that precede and are presuppose