INTENTIONALITY AND SPEECH-ACTS

Skinner's view is based on the assumption that any statement—oral or written—must be understood at least in part on its own terms. This requires, among other things, understanding an utterer's or author's language as he or she understood that language. If an author's understanding of a word or phrase is different from our understanding, and if we are unaware of or fail to take into account that difference, then we cannot understand fully what the author is trying to say. For Skinner, understanding what the author is trying—intending—to say is an extremely important part of what it means to understand the author's statement.

I have used here such qualifiers as “at least in part” and “understand fully” because Skinner himself does not deny the legitimacy of other modes of textual interpretation, and hence does not insist that his approach is the only useful one. He recognizes that a statement can be fruitfully analyzed in terms of its actual effects on listeners or readers—for example, on the degree to which it changes or reinforces belief—and that such effects may well be unintended. He recognizes, further, that for certain purposes a statement may be explicable in terms of (inter alia) deep-seated social or psychological forces of which the author is or might be unaware, and hence that the statement can be understood or interpreted as itself the effect of some kind of unself conscious causal process (Skinner Reference Skinner2002, 110). But without in any way denying the utility of such interpretive projects, Skinner also wants to recognize—and give a certain pride of place to—the recovery of what might be called intended meaning. Exactly how Skinner understands the relative importance of intended meaning vis-à-vis the other features of a text is a matter of some dispute. There are, so to speak, both stronger and weaker accounts of Skinner's intentionalism, and these are discussed shortly. For the moment, however, it will suffice to say that, in my view, the primary task in interpreting a statement is, according to Skinner, to figure out the author's point in making the statement, and this means neither more nor less than recovering his or her intentions.

In formulating this argument and, at the same time, in adducing (some of) its methodological implications, Skinner relies on the philosophical pragmatics of Austin, Searle, and Grice (Skinner Reference Skinner2002, 2, 106–24; Reference Skinner, Brett and Tully2006, 242).Footnote ⁵ For Austin and Searle, fully understanding an utterance requires understanding that the utterer, in making the utterance, is intentionally doing something—is performing an act (i.e., a speech-act)—and that coming to grips with the utterance requires figuring out just what act is being performed. Sometimes this is indicated directly by the words of the utterance. For example, the speech-act of making a prediction might begin with the phrase “I predict that . . . .” But at other times there is no such “force indicating” phrase, and, in such cases, the nature of the act is indicated, at least in part, by something other than the words themselves (Reference Skinner2002, 106). Thus, if I say “I'm going to call your mother,” this single statement might be a prediction, a warning, a promise, or a threat, depending on a variety of factors largely having to do with the circumstances in which the statement is made. Of course, a prediction, in and of itself, is not a warning, a promise, or a threat. Despite the fact that exactly the same words would be uttered in exactly the same order, these would be different acts embodying different intentions, and if the listener gets it wrong—if, say, he or she thinks the utterance was a warning when it was intended merely as a prediction—then the utterance will have been badly misinterpreted. Its meaning will have been misunderstood.

It is true that Austin and Searle—and Skinner as well—evince a certain uneasiness about the word “meaning,” often preferring to talk of the “force” of an utterance, and it is perhaps at this point that Grice (1989, 213–23) makes a signal and helpful contribution.Footnote ⁶ If I point to the heavens and say, “That dark cloud means rain,” I am using a perfectly intelligible notion of what it means to mean something. But that notion is also different from the notion of meaning that I am using if I say, “When I uttered those words, this is what I meant.” Grice calls the former notion “natural meaning,” presumably because it describes or reduces to the kind of cause–effect processes characteristic of the natural (although also the psychological and sociological) world. He calls the second notion of meaning “non-natural meaning” and analyzes it precisely in terms of a complex structure of intentions on the part of the utterer. Roughly, nonnatural meaning involves the intention of the utterer to induce a belief in the listener in virtue of the listener's recognition (1) that the utterer is intending to do this through the utterance and (2) that the utterer is intending the listener to recognize this intention. It would be plausible to suggest that Austin and Searle, in analyzing the force of a speech-act, are analyzing a certain kind of nonnatural meaning, and indeed, this is a suggestion that Searle (Reference Searle1969, 45–50) himself explicitly endorses. Strawson offers what seems to me a similar view: “The understanding of the force of an utterance in all cases involves recognizing what may be called broadly an audience-directed intention and recognizing it as wholly overt, as intended to be recognized” (Reference Strawson1964, 459).Footnote ⁷

Famously, the effective meaning of an utterance is, according to speech-act theory, not identical to its literal meaning. Again, the utterance “I'm going to call your mother” can mean different things in different circumstances despite the fact that its literal meaning remains perfectly constant. The intended or nonnatural meaning of an utterance—to use the term of art, its “illocutionary force”—is described by the act it is performing, and this, in turn, is determined by the social context in which the act is performed. The latter point is absolutely crucial. Acts—including speech acts—are performed within a structure of established social practices or rules that determine the kinds of things that are appropriate or inappropriate and that determine, more specifically, the significance of words, gestures, tone of voice, body language, and the like (Searle Reference Searle1969, 30). Thus, the rules of our own discursive community determine whether “I'm going to call your mother” is a warning or prediction in virtue of when it is said, how it is said, to whom it is said, what it is said in response to, and so on. Similarly, those rules determine that the utterance “Can you pass the sugar?” when uttered at the dinner table is rarely if ever an inquiry into whether you are able to do something, but, rather, a request that you actually do it. In all such cases, social conventions confer a meaning on the utterance that goes beyond or is otherwise different from its literal meaning.

Once again, Grice makes an important intervention. His notion of “conversational implicature” describes (among other things) circumstances in which utterers deliberately flout standard practices or conventions (Grice Reference Grice1989, 22–57). I ask you, “How's your love life?” and you respond, “So how about those Mets?” Here, the obvious and intentional violation of a basic rule of appropriateness or relevance—there is, we assume, no discernable connection between your love life and the performance of a baseball team—clearly communicates that you do not want to answer the question, but, again, the clarity of this depends precisely on a shared knowledge of established conventions of communication and a resulting awareness that they are being deliberately flouted.Footnote ⁸

For Skinner, these general lines of argument provide fundamental principles for interpreting historical works of political thought. If we want to understand the arguments of, say, Leviathan, we need to know what Hobbes was trying to do, what act he was intending to perform in writing the words he wrote. This means at least two things. First, it means understanding the particular political circumstances Hobbes was addressing and the particular political controversies he was seeking to resolve. For Skinner, political writing is an activity occasioned by and focused on the real political problems of the world in which that writing occurs. Only with a knowledge of those problems can we come to grips with the author's intentions, the author's point. Second, it means understanding the communicative conventions—the rules or established practices of political discourse—within which Hobbes was working. For Skinner, all political thought is produced within a discursive community that establishes the conceptual and linguistic vocabulary that an author must use in realizing his or her intentions. Only with a knowledge of that vocabulary can we understand the author's words as the author intended them to be understood.Footnote ⁹

In writing Leviathan, then, Hobbes was performing a speech-act or, perhaps more accurately, a series of speech-acts. Because we cannot possibly know what Hobbes was saying—cannot correctly interpret the intended meaning of the text—without a deep understanding of the political problems he was addressing, the serious reader must pay close attention to the controversies and issues that preoccupied mid–seventeenth-century British politics. Likewise, because we cannot possibly know what he was saying without a deep understanding of the linguistic conventions he employed, we also need to develop a systematic knowledge of the full range of political writings—including the voluminous and varied writings of so-called minor authors—that embodied and reflected the prevailing practices of political discourse. Just as “I'm going to call your mother” must be apprehended in light of the pragmatic circumstances in which it is uttered, just as “Can you pass the sugar?” must be apprehended in light of established rules or practices of communication, so, too, must Hobbes' text, like any political text, be interpreted in the light of its political and linguistic contexts. Without doing this, we cannot possibly understand the intentional act that Hobbes performed in writing what he wrote. We are apt to miss the point.

As suggested previously, there is a stronger and a weaker account of Skinner's position. According to the stronger account, understanding and interpreting a statement or a text is primarily or fundamentally a matter of recovering intended meaning. To discover what an utterer or author meant—to get the point—is in itself to understand the text. The intended meaning is or ought to be the focus of any interpretive effort. Although such a view has been standardly attributed to Skinner (2002, 110n), he himself seems to deny that this has ever been his approach. Rather, he endorses a weaker account: a focus on intended meaning is an important and even essential strategy, but by no means the primary or even preferred strategy of interpretation. In his most recent discussion, he says that his theory “does not tell us, nor do I believe, that the intentions of speakers and writers constitute the sole or even best guide to understanding their texts or other utterances” (Reference Skinner2002, 110; see also Skinner Reference Skinner, Brett and Tully2006, 242). He elaborates:

It is an interesting passage. Note that the presence of unintended meaning is something not to embrace but “to concede,” and then only under duress, and that such meaning is to be discovered by readers and interpreters who are not only (we may presume) intelligent and diligent, but also “ingenious” and who examine the text “at every point.” If Skinner rejects the strong intentionalist account, it seems that he does so with some reluctance. Elsewhere, he emphasizes that his principal concern has always been with what the author “meant” and not with the “meaning” of the text (Skinner Reference Skinner2002, 111). His interpretive aims have thus been modest with respect to what a text really means. But surely to ask what the author “meant” is to ask after the author's “meaning,” and Skinner himself identifies one kind of textual meaning—what he calls “meaning₃”—as “equivalent” to the intentions of the author, thereby calling into question any clear distinction between what the author meant and the meaning of the text (Reference Skinner2002, 100). More generally, it is quite evident that, for Skinner, the great culprit is anachronism. Failure to situate a text in its political and linguistic context will lead us to impose on that text our own historically quite different concerns, rules, practices, and linguistic conventions. To the degree that we do this, we are not really confronting the text at all but are inventing, albeit unself consciously, a new text—our own creation—something, moreover, that is related to the original text in ways that we cannot possibly discern because, of course, we do not really know what the original text itself is saying (Schochet Reference Schochet1974, 265–66; also Condren Reference Condren1985, 49). It is, in the end, hard to escape the conclusion that Skinner, while agreeing that any text may have unintended meanings, nonetheless places a strong priority indeed on the discovery and elucidation of intended meaning as the key to any effort at textual interpretation (see Skinner Reference Skinner1974, 286).

PROPOSITIONAL CONTENT AND ASSERTIONS OF FACT

Such a position is enormously attractive. It is also, upon reflection, deeply problematic. Indeed and in particular, I believe that the use of speech-act theory as a model or tool for understanding works of political philosophy has been seriously misconceived and that the implications of philosophical pragmatics for the interpretation of texts are far different from those suggested by Skinner.

The first thing to notice is that the pragmatics of Austin, Searle, and Grice focus on spoken conversation rather than formal written communication. Virtually all relevant examples involve the ordinary use of language in everyday spoken discourse. We are dealing here with an analysis of utterances. This is, moreover, no accident. A distinguishing feature of everyday spoken discourse is that it is generally improvised. Of course, not all conversations need be as highfalutin as Oakeshott's “unrehearsed intellectual adventure,” (Oakeshott Reference Oakeshott1962, 198) but the unrehearsed part seems exactly right. In ordinary conversation, we make up things as we go—we proceed to ad lib—and it is precisely for this reason that the established social practices or unwritten rules emphasized by standard speech-act theory are so crucial. Any improvised activity, like modern jazz or soccer football, can be intelligible only if the improvisers operate within a more or less well-recognized structure of conventions or rules that provide a repertoire of moves as well as criteria for making sense of those moves (Steinberger Reference Steinberger2004, 319–21). The speech-act theory of Austin and Searle and the implicature theory of Grice are designed precisely to explain how the making and receiving of utterances functions successfully as an unrehearsed, improvised endeavor governed by communicative rules. That is the goal of their pragmatics.

The written use of language seems different. The author of a written text chooses his or her words with care, edits and revises, decides at comparative leisure whether to share the text with its proposed recipient, and so on. This is not, ordinarily, an improvised activity. It is rather the opposite. Even the most informal missive is apt to be a well-rehearsed and reflectively composed act of communication when compared with ordinary conversational utterances. To the degree, then, that Austin, Searle, and Grice focus on the conditions necessary for successful improvised conversation, it is not at all clear that their work is directly applicable to the production of any written text.

Now there are certainly hard cases that blur the spoken–written distinction. Innumerable spoken (and unwritten) utterances are not improvised at all (e.g., ritualistic utterances such as the bailiff saying, “order in the court,” or the racetrack announcer saying, “and they're off.)”Footnote ¹¹ By the same token, not all written statements are carefully composed. A hastily written note may be nearly as improvised as a conversational utterance. Of course, contemporary forms of electronic communication make the hard and fast distinction between the speaker and the writer seem quaint indeed.

But the existence of hard cases does not obliterate deep distinctions, and with respect to the issue at hand, there is certainly an enormous difference between the act of saying, in ordinary conversation, “I'm going to call your mother,” and the act of composing, over a period of months or years, a lengthy treatise in political philosophy involving pain-staking research, systematic reflection and analysis, the orderly arrangement of ideas, the thoughtful design of a textual structure, the carefully considered use of language. This is to emphasize a difference between the spontaneous use of language (which is often oral and conversational) and the rehearsed use of language (which is often written). To the degree that standard speech-act theory aims to explain the success of improvised conversation—a success based on our socially learned and unself conscious mastery of communicative conventions that enable us to understand the intended meaning or illocutionary force of an utterance—it seems not well suited to the interpretation of philosophical texts. The rules of ordinary, improvised conversation that are the focus of such a theory do not apply—at least not obviously—to the writing and reading of systematic theoretical works; hence, works of this kind would seem to be outside the purview of the pragmatics of Austin, Searle, and Grice.

The Skinnerian will object: if the writing of a theoretical text is not an improvised act, it is an act nonetheless, indeed a linguistic act; and like all such acts, it represents intentions that can be recovered only by situating it clearly within its historical and communicative contexts. The author of a written text, regardless of how elaborate, is dependent on and constrained by established discursive rules, just as much as the casual utterer. Speech-act theory, by sensitizing us to the centrality of rule-based communicative practices, is indispensable in making sense of all linguistic action.

This seems plausible enough. But again, the crucial point is that the particular kinds of rules that govern ordinary, improvised conversational activity and that are necessary for such activity to occur successfully must be quite different from the rules that govern formal writing. At the very least, the burden of proof is on Skinner to show just why this is not so, and, to the best of my knowledge, he never attempts such a proof. More strongly, it is doubtful that a plausible argument along those lines would be possible. To say that ordinary conversation is improvised is to say, again, that it occurs on the fly. Conversational utterances are typically produced under a distinctive set of temporal and social exigencies that encourage and require the development of an elaborate repertoire of verbal shortcuts, feints, allusions, ellipses, vocal intonations, bodily gestures, and the like, all typically underwritten by more or less local conventions. The rules that govern the use of such a repertoire—rules that compose a principal subject matter of standard speech-act theory—are precisely what allow conversationalists to manage successfully the temporal immediacy and physical (including vocal) proximity characteristic of ordinary conversation, and it is not easy to see how those same rules would or could apply to the writing and reading of systematic theoretical works.

Thus, to pursue an analogy mentioned previously, the jazz pianist, when performing with a small group, typically plays a recurrent harmonic pattern—the so-called “form” or “progression” of the piece being played. This pattern is effectively, if not always actually, written down. But in playing the pattern, the pianist must constantly—and instantaneously—make choices (e.g., how to voice a particular chord, exactly where to place the chord temporally within the designated measure). In making such choices, the pianist receives cues from, and also provides cues to, the other musicians. Although this often occurs at great speed and sometimes occurs with great ingenuity, it is, in fact, not rocket science. There are (unwritten) rules for doing it, and one does not have to be a genius to learn them. Nonetheless, learning and applying the rules—the rules for receiving and giving musical cues on the fly—is a large part of what jazz is all about. Those same rules, however, have essentially little if any relevance for the very different activity of composing written music. In composition, one ordinarily does not—and certainly need not—make choices on the fly or exchange instantaneous musical cues with other individuals. The (so to speak) spatio temporal and social conditions that require the development and application of improvisational rules simply do not obtain. Those rules would thus have, in composition, no conceivable function. So, too, I believe, for the difference between ordinary conversation and written (or other kinds of rehearsed) linguistic communication. Again, the circumstances of conversation are profoundly different from those of systematic written communication; hence, the rules necessary for the former are apt to be irrelevant to the latter.

The point, however, is not to claim that the insights of speech-act theory per se have no relevance for the analysis of theoretical texts. Rather, it is to suggest that the implications of philosophical pragmatics, broadly construed, for the interpretation of such texts are, in fact, sharply different from what Skinner supposes.

Specifically, explaining the illocutionary force of an utterance, Searle helpfully distinguishes it from the utterance's “propositional content” (Searle Reference Searle1969, 30). Even though “I'm going to call your mother” might have different intended meanings depending on conventions and circumstances, in all of its uses it will have the same propositional content (i.e., the same reference and predication). This is simply the (actual or possible) fact of me calling your mother. Whether the utterance is a prediction, a warning, a promise, or a threat, it will always express that content. It will refer to two people, me and your mother, and it will predicate me calling your mother. Of course, we must bear in mind that propositional content is not the same thing as literal meaning. The literal meaning of “Can you pass me the sugar?” is different from its intended meaning, but is also different from its propositional content, which is simply the actual or possible act of me passing the sugar to you. More generally, propositional content is not the same as any kind of statement or a speech-act. It is, to be sure, a constitutive feature of many statements or speech-acts. But for a collection of words to be such a statement or speech-act (i.e., a statement or speech-act that has propositional content), an utterer must actually do something with—must add something to—the propositional content. By itself, then, propositional content is not much. Nonetheless, a focus on propositional content and its relationship to the speech-act is absolutely crucial because, among other things, it encourages us to consider a kind of speech-act different from those we have considered thus far, namely, a speech-act that is intended not to predict, warn, threaten, or promise but, rather, to state and defend the actual or possible fact of a proposed content. The act performed by such an episode of speech is simply the act of asserting a fact to be a fact. It is the act of saying what is. In performing such an act, the focus is on the actuality or possibility of the propositional content itself, and to assert a propositional content is to assert that the content proposed is or may be. In such an act, the assertion of the fact of the propositional content is the point.

For Searle, acts of this kind compose one of the five principal categories of speech-acts—specifically, the category that he calls “representative” (Reference Searle and Gunderson1975, 354–55)—and there can be no question that such acts are performed all the time, both in everyday, improvised conversation and in formal written discourse. Of course, the realization that asserting the fact of a proposed content is itself an act—every bit as much as predicting, warning, etc.—is in large part the reason that Austin himself abandoned his original notion that constatives (assertions of fact) and performatives (speech-acts) were mutually exclusive categories, and concluded instead that all statements, including constatives, have a locutionary dimension (roughly, literal meaning), an illocutionary dimension (intended meaning), and a perlocutionary dimension (the actual effect on the hearer) (Reference Austin1975, 133–47). What is interesting, however, is that, in performing acts that assert facts, the utterer or author often takes special pains precisely to abjure, or to avoid relying on, the kinds of unwritten rules, practices, or conventions that govern much oral communication. That is, the utterer or author often tries to make things explicit. This seems to reflect, at least in part, a distinguishing feature of the act of asserting the actual or possible fact of a content. If my act is a warning, it cannot be refuted; if it is a promise, it might or might not be kept, but it remains a promise nonetheless; if it is a threat, either it works or it does not; if it is a prediction, it might not come through. In Austin's terminology, such acts might be “happy” or “unhappy”—“felicitous” or “infelicitous”—depending on whether they effectively express their intended meaning (Reference Austin1975, 148). But the assertion of a fact (which, of course, might include the assertion as to whether or not a warning, promise, or threat actually was made and, if so, what it was) is different in that it is, here and now, either “true” or “false”—again, Austin's words. Such an assertion can be (in principle) disproved, and what this suggests, I believe, is that the inherent legitimacy, so to speak, of the act is at risk in a way that distinguishes it from many other kinds of speech-acts and that establishes a special incentive to be precise.

Again, it is plain that many spoken utterances are intended to assert the fact of a proposed content, and hence seek a kind of precision that is often unnecessary or unhelpful in other kinds of ordinary conversational acts. Thus, if I want simply to assert the fact of me calling your mother, and if I anticipate any skepticism, I may well be inclined to provide detailed evidence (e.g., a demonstration that I know how to reach your mother, proof that I have called your mother before in similar circumstances), but, more important, I will often attempt to present my claim and my evidence in language—involving, say, the use of force indicating devices—that minimizes or eliminates the difference between the literal meaning of my words and their intended meaning. If, that is, my point—my intention—is to assert the actual or possible fact of a proposed content, I may well choose to avoid to the degree possible the kind of nuance, indirection, allusion, implication, insinuation, metaphor, flouting, and so on that invoke and depend on the established conventions or practices of ordinary discourse. At least in certain circumstances, I will strive mightily to avoid any possible source of ambiguity. I will lay things out.

But certainly what is true of many conversational acts that assert the fact of a proposed content is, if anything, even more true of systematic theoretical writing, that is, linguistic acts that purport to make propositional claims of unusual complexity and importance and that, moreover, leave enduring traces of themselves—texts—that invite the close, detailed scrutiny of, in principle, a multiplicity of critics.Footnote ¹² To my knowledge, the point is largely ignored by Austin, Searle, and Grice. Again, their focus is on improvised conversation. The point is certainly not ignored by Skinner, but here, it seems to me, he has it exactly wrong. As indicated previously, he recognizes that in conversation we sometimes append force indicating phrases to our utterances (e.g., “I predict that. . .”) so as to make explicit what we are doing and thereby avoid misunderstanding, and he recognizes further that “the problem of interpretation arises in part because we do not generally trouble” to do this (Skinner Reference Skinner2002, 106–107). In other words, interpretation is a challenge precisely in those circumstances where the speaker has failed clearly to specify the force of what has been said. But then Skinner unaccountably asserts that this problem—the absence of force indicating devices—is especially characteristic of complex written texts (i.e., “acts of communication [that] normally attract the attention of literary critics and intellectual historians”) (Reference Skinner2002, 107). The apparent and blithe assimilation here of literary texts—novels? poems?—to theoretical or philosophical texts is troubling enough, but the notion that theoretical texts pose the same kind of interpretive problems as underspecified conversational utterances is peculiar indeed. Surely, the truth of the matter is rather the reverse. Of all linguistic acts, the composition of a written theoretical treatise is in fact among the most likely to rely heavily on precisely the kinds of devices that maximize clarity and minimize the possibility of miscommunication and hence minimize the particular kinds of interpretive problems with which philosophical pragmatics are primarily concerned.

It seems that in writing Leviathan, Hobbes was engaged in the activity of presenting, and defending at length, a series of assertions about the fact of some proposed content and that, in doing so, he sought largely to employ a language that did not require, indeed would not allow, the discovery of an illocutionary force different from the literal meaning. As is discussed later in this article, this hardly means that the text itself is innocent of rhetorical technique. Far from it. Nonetheless, the constant effort by Hobbes to define terms, to clarify the ways in which words are being used, to make explicit distinctions—all of this suggests a dissatisfaction with, an unwillingness to rely on, hence an effort to break free from particular or local conventions of communication in favor of a more abstract and universal mode of expression. In emphasizing the “apt imposing of names” and the rigor of strict definition and “ratiocination,” Hobbes' intention was plainly to produce a text that would transcend the local community of discourse (see Condren Reference Condren1985, 219).

Just as plainly, his intention was to address substantive issues of consequence—issues of political theory—that were hardly confined to the particular controversies of his day. Ordinary, everyday human activity occurs locally. I am doing this here and now, and doing it in response to this particular situation—I'm going to this restaurant, I'm painting this room, I'm campaigning for this political office. That is the character of most human performance, and what is true of ordinary, everyday activity in general is true as well of ordinary, everyday conversational activity. Most speech-acts are performed in response to other speech-acts and have, as their focus, some matter of more or less immediate concern. But, again, the production of systematic theoretical treatises seems to be a different kind of endeavor. Now, surely Skinner is correct in suggesting that political writing is usually or always occasioned by active and urgent political issues. But here, it seems to me, he is in danger of ignoring a distinction that he himself emphasizes, namely, between motive and intention. Hobbes may have been motivated—moved, inspired, compelled—by the crises of the mid-seventeenth century. These crises may have caused him to write about politics and may even have caused him to assert the particular propositions that he did. But this would also be perfectly consistent with an intention to produce a political philosophy that would address issues characteristic not only of seventeenth-century British politics, but also of politics in general, and that would refute not only, say, a partisan of parliamentary government, but also a defender of vain philosophy such as Aristotle (Minogue Reference Minogue1981, 537–38; Parekh and Berki Reference Parekh and Berki1973, 171–75; Schochet Reference Schochet1974, 270–71). In seeking a systematic, self-conscious, and highly rationalized mode of argumentation and expression involving careful definition and logical explication, he was seeking a mode of argumentation and expression appropriate to issues that far transcended the here and now.

Hobbes was thus engaged in performing speech-acts of a particular kind: the carefully planned, painstakingly executed, utterly nonimprovised activity of asserting the actual or possible fact of a proposed content, the actuality or possibility of which was understood to be of general, rather than local, relevance.

But the contextualist will persist: is it not possible that Hobbes adopted the language of logic and definition as a literary tool authorized, as such, by the communicative conventions of his day? Is it not equally possible that his abstract or universalized assertions of proposed content were merely coded or disguised or even unconscious assertions about local problems? The answer, of course, is yes and yes; however, the answer, further, is that none of this would give us any reason not to focus our attention primarily on the propositional content of Hobbes' text on its own terms. For the assertion of the actual or possible fact of proposed content is clearly and plainly what Hobbes, for whatever reason, intended the content of his work to be. It is, as such, the expression of Hobbes' political philosophy, and if we want to understand that philosophy rather than the motives for its production, this is what we must examine above all else.

And what of the claim that Hobbes was addressing issues of general, rather than merely local, relevance. Does this not uncritically presuppose the existence of general—or “perennial”—issues of politics? Isn't such a presupposition unacceptable in the absence of strong supporting evidence? In fact, however, I am making no such presupposition. My argument is simply that Hobbes was intending to address issues of general relevance. This was a constitutive and defining part of the act he was performing, and to understand the act does indeed require recognizing the intent. It is, I believe, impossible to read Leviathan as not embodying such an intent. But the question of whether Hobbes really was addressing general issues, as opposed to merely believing that he was, is an entirely separate question, as is the question of whether such issues really exist. I am not here addressing those questions—although presumably doing so would benefit greatly from the effort rationally to reconstruct the arguments that Hobbes, and many others, actually make. We cannot discover if a set of arguments is addressing perennial questions without first knowing what those arguments are.

Of course, what is true of Hobbes is true of anyone—including Skinner himself—who performs the act of writing a text that seeks, through explicit definition, careful analysis, and patient step-by-step exposition, to assert the actual or possible fact of some proposed content of, per hypothesis, quite general application. It is, thus, far from folly to approach such a text by focusing not on considerations of political or linguistic circumstance but, rather, on (to borrow Grice's phrase) “what is said”—on the structure of argumentation considered on its own account, independent of historical context. Indeed, this is precisely the kind of analysis that such texts seem to call for.

ANALYSIS AND INTERPRETATION

None of this is to deny the extraordinary value of the detailed investigations undertaken by Skinner. But it is to suggest that their value is, indeed, historical rather than philosophical. At least on certain influential accounts, the principal task of the historian is to discover a kind of intelligibility in the events or performances of the past (Oakeshott Reference Oakeshott1962, 137–67). This is accomplished by, for example, examining those events—political, military, economic, sociological, aesthetical, ideological—explicitly in the context of one another and, more specifically, by considering how the one informs, conditions, contradicts, or otherwise sheds light on the other, so that an examination of the range of events, taken together, allows the historian to construct a coherent picture of a (not necessarily coherent) social whole, which, in turn, helps us understand the function, role, or historical significance of each of the parts themselves.

A focus on what is said—on the act of asserting the actual or possible fact of a proposed content—has a different goal requiring a different method. Generally, the approach is one of paraphrase, reconstruction, inference, and extrapolation. It is the effort to restate what has been said with a view to achieving maximum clarity and comprehensiveness, and this means eliminating, to the degree possible, all sources of ambiguity or imprecision. It is, I believe, the method of analysis.

In saying this, I am, in a sense, saying nothing startling. Commentators frequently use the word “analytic” to refer to scholarship in the history of political thought that focuses on the logic of argumentation and that pays comparatively little attention to historical (or literary/rhetorical) considerations (e.g., Baumgold Reference Baumgold1981, 929). But the positive sense of the term—the very idea of what it means to focus on the logic of argumentation in political theory—is often underspecified, and one consequence is that the real nature of the challenge that the analytic approach poses for Skinner's approach has remained all too obscure. To be sure, Skinner's own work has sometimes been identified with the analytic tradition, and not without reason. In relying on Austin, Searle, and Grice, he is certainly relying on the work of analytic philosophers (Richter Reference Richter1990, 61). But his use of their work is not analytic in the way that I intend. Properly understood, the approach to the interpretation of texts that I am proposing is “analytic” in the specific sense that this term has acquired through the influence of a handful of works of early twentieth-century thought, of which Russell's “On Denoting” is arguably the most important and which are especially germane to the question of how to treat speech-acts that assert the actual or possible fact of some proposed content.

Russell's (1905) immediate goal in “On Denoting” was to show how it is possible to talk or write intelligibly about negative existentials, and to show this by rationally reconstructing—perspicuously restating—what has been said in a way that reveals its underlying logic. Russell demonstrates the intelligibility of the statement “The King of France is bald” not by examining any particular sociohistorical contexts or conversational conventions but by discovering the logical structure of the assertion of proposed content. When one says that “The King of France is bald,” one is really saying that “There is one and only one King of France, and he is bald.” The apparent unintelligibility of the first assertion—because there is no King of France, how can one refer to a King of France?—dissolves when the assertion is shown really to express (in part) the perfectly intelligible, although also false, assertion that there is a King of France.

Of course, a great deal of subsequent philosophical analysis—both before and after the emergence of so-called ordinary language philosophy—proceeds more or less along these lines. Indeed, “philosophers of all persuasions are continually recasting sentences in what they regard as a better form” (Urmson Reference Urmson1956, 25). Thus, to pick just one especially convenient example, consider Grice's essay on meaning, mentioned previously. Again, Grice asserts the difference between assertions of natural meaning and assertions of nonnatural meaning. He does this by rationally reconstructing what is actually being asserted in particular claims about meaning. To say about a dark cloud that “That means rain” is really to say something about natural cause and effect, whereas to say about some words I uttered that “This is what I meant” is really to say something about my intentions. It is important to emphasize that the difference is not at all dependent on politics, history, or the nuanced, socially located practices of conversation. It is, rather, an abstract conceptual difference that reflects distinct, possibly universal structures of thought. Grice's essay thus exemplifies the larger analytic project: to rearticulate the claims we make about the world—whether on the street corner or in the laboratory, whether in ordinary language or in high-brow discourse, whether in the popular press or in elaborate, scholarly treatises—in such a way as to specify with precision the concepts that are being used and the relationships that are being proposed. Such a project—a project of rational reconstruction—reflects Wittgenstein's conviction that “(a)ll philosophy is ‘critique of language’” and that “Russell's merit is to have shown that the apparent logical form of the proposition [Satz] need not be its actual form” (Wittgenstein Reference Wittgenstein1955 [1922], 62–63).

It is one thing to investigate, as Skinner does, the particular materials of a discursive community—of a “language”—for the purpose of determining how particular terms are used and understood by the members of that community. It is quite another to analyze a particular set of propositions with a view toward discovering and explicating their underlying argumentative structure. The one is, roughly, a lexicographical project, the other philosophical. Of course, to call a project lexicographical is hardly to demean it; the potential value of such a project for historical study is obvious. But the critical analysis of the “real form” of a proposition—whether for the purpose of developing an ideal language or for the purpose of pursuing philosophical puzzles or confusions that are posed by ordinary language—is quite another matter.Footnote ¹³

Often, it seems, analysis is not necessary. If I assert the fact of me calling your mother, there may well be nothing more that need be said. The assertion may be clear enough as it stands. It may be, similarly, that Grice's account of the meaning of meaning—his assertion(s) of content about what we assert about content when we speak and write about meaning—is the last word, not simply in terms of being (exhaustively) correct, but in terms of being expressed with a kind of lucid accessibility that makes any restatement superfluous. Perhaps, in other words, it speaks for itself. But certainly a great many assertions of the actual or possible fact of a proposed content are not of this nature. Indeed, even the most carefully written, rigorously argued, and painstakingly constructed assertion of proposed content is apt to present problems of unclarity, opacity, incompleteness, and error that have nothing to do with the conventionally mandated elisions, nuances, and codes of ordinary conversation and everything to do with the intricacies of logical reasoning, broadly construed. Surely this is true of most of the texts with which political theorists have been primarily concerned. These are texts that propose assertions of content of substantial complexity, assertions that are themselves constituted by or dependent, explicitly or otherwise, on numerous other assertions, the minor assertions serving as premises or qualifiers of the major assertions. Of course, the specific relations among these various assertions are also being asserted, although again not always as explicitly as one might like. And this results in an elaborate, intricate, many layered structure that may or may not hold together and that may be presented, moreover, in any number of ways — sometimes with close attention to transparency and rigor, as with a Hobbes or a Rawls; sometimes with an interest less in laying things out than in deploying a kind of literary rhetoric designed to persuade rather than prove, as with a Rousseau or an Arendt; sometimes with an effort to formulate a new linguistic apparatus uniquely suited to the content being asserted, as with Hegel; and so on. In all such cases, the analytical interpreter seeks to pursue what is said and what is implied, not with a view toward discovering the rich and varied sociohistorical world of which the text is a part, but with a view toward reconstructing the complex structure of propositional content whose actual or possible fact the author intends to present and defend. It should now be clear that the features of theoretical texts that make them ill suited to the kind of conversational analysis pursued by standard speech-act theory—their nonimprovised nature, their focus on assertions of fact, their permanence, the inherent value placed on eliminating all sources of ambiguity, the consequent attempt to discover an abstract and universal mode of expression free from the imprecision (although also the nuance) of local idioms—are precisely those features that call for a systematic effort to focus closely on what is said, to piece together the structure of argumentation, and to restate assertions of content with ever more accuracy and clarity.

The differences between this kind of approach and the approach recommended by Skinner—along with their respective strengths and weaknesses—can be well illustrated by looking briefly at some aspects of Skinner's (1996) own important study of reason and rhetoric in Hobbes. In this book, Skinner makes four central and interrelated claims: Hobbes' intellectual formation was heavily influenced by a predominately humanist tradition in which the study of classical rhetoric played a central role. Despite this, he emphatically rejected that tradition in works such as the Elements of Law and De Cive, which pursued an austerely scientific and resolutely antirhetorical mode of argumentation; however, in his somewhat later writing—including and especially Leviathan—he sharply reversed fields and embraced a style of presentation rich in materials of a highly rhetorical nature. This change, moreover, reflected Hobbes' newfound conviction that, at least in the real-world politics of his own day, rational argumentation was insufficient by itself and needed to be supplemented by the rhetorical arts of persuasion. Skinner's defense of these claims seems to me spectacularly successful. After providing a comprehensive and authoritative account of the elements of classical rhetoric as formulated primarily by Cicero and Quintilian and as adapted and purveyed by Renaissance humanists, he traces in copious detail and with great resourcefulness the literary-rhetorical dimensions of Leviathan. He demonstrates unequivocally the ubiquity, in that work, of all manner of rhetorical tropes and figures that reflected the linguistic/literary conventions of the time, and he carefully and convincingly categorizes each of them according to the curriculum of classical rhetoric—from diasyrmus to meiosis, from epanodos to epanaphoria, from synchoresis to litotes, and so on. Skinner's presentation is exhaustive, dare I say ingenious, and entirely persuasive. The result is nothing less than a new Leviathan. Familiar passages must now be read as never before (e.g., as pieces in an elaborate literary jigsaw puzzle or as parts of an ornate, even baroque tapestry comprising a vast repertoire of rhetorical devices). By clearly situating Leviathan within the prevailing humanist and linguistic practices of sixteenth- and seventeenth-century intellectual life, and also vis-à-vis Hobbes' own earlier writing, Skinner substantially deepens and enriches our understanding of the book as an historical document. For the first time, it seems, the historical character of the work comes into clear focus, and does so precisely because we are able to observe it, via Skinner, in its natural setting. At the same time, of course, the account of Leviathan deepens and enriches our understanding of the setting itself (i.e., the very world out of which it emerged). This is marvelous history—complex, encyclopedic, utterly persuasive, and deeply edifying.

However, Skinner also suggests (although not without some equivocation) that such a reading may be helpful in eliminating doctrinal misinterpretations—erroneous accounts of Hobbes' philosophy—attributable to “literal-minded interpreters who have concentrated on what Hobbes says. . . without paying any attention to his tone and manner of saying it” (Reference Skinner1996, 13–14). Here, it seems to me, the results are much less encouraging. Consider, to begin with, Skinner's claim that Hobbes, in the dedicatory letter to Leviathan, unmistakably invokes—and in rather elevated terms—the classical rhetorical doctrine of inventio, and does so specifically by asserting, at the outset, the importance of his book, the novelty of its arguments, and the neutrality or impartiality of its author (i.e., himself) (Reference Skinner1996, 376–77). The evidence that Skinner provides in support of this claim seems to me rock-solid. But it is not at all obvious how such an account helps us understand Hobbes' assertion itself because the language of the dedication seems clear on its own. It plainly says that the work being dedicated is, indeed, important, novel, and the product of an impartial commentator. The fact that this assertion is of Ciceronian inspiration, that it reflects a common rhetorical practice of the day, that it also reflects Hobbes' intention to emphasize the importance of the persuasive arts in any practical political endeavor, and that it is made, as Skinner says, “in tones of calculated moderation”—all of this seems, in doctrinal terms, neither here nor there. The discovery and elaboration of the rhetorical dimension adds nothing to our understanding of the assertion qua assertion.

Now, in this particular case, Skinner does not claim to have changed or improved our understanding of doctrine. But the independence that we see here of asserted proposition on the one hand and rhetorical form on the other is, I would suggest, typical of a great many of the passages that Skinner examines, including those where he does claim that a focus on the tone and manner of expression can sharpen, revise, deepen, or clarify our understanding of what is actually being said. It seems to me, moreover, that such a focus, far from making a contribution to our philosophical understanding, may actually be a distraction.

Thus, for example, Skinner notes that Leviathan, in describing the state, invokes, as did The Elements of Law and De Cive, the age-old image of an organism. The state is a body-politic. Just as the human body is a natural entity composed of a variety of naturally occurring organs, so, too, is the state. Skinner notes, however, that Leviathan also invokes a new and quite different set of images in describing the state, according to which the state is not a body but a kind of machine (e.g., “what is the Heart, but a Spring; and the Nerves, but so many Strings; and the Joynts but so many Wheels . . . .”) (Hobbes Reference Hobbes1980 [1651], 81). Skinner places great emphasis on the doctrinal importance of these mechanistic images. Specifically, they “underpin his [Hobbes’] claim that commonwealths can in no sense be regarded as God-given creations or natural occurences. They are wholly man-made contrivances, mechanisms we construct with the sole aim of furthering our own purposes” (Skinner Reference Skinner1996, 387). There can be no doubt, of course, that such mechanistic images do occur in and are an important part of Leviathan, and that Skinner is right to examine them. But their role in helping us understand Hobbesian doctrine is, I would suggest, uncertain. To begin with, it is not entirely clear just what it means to say that such images “underpin” a claim. Is this underpinning supposed to have some kind of effect on the reader and, if so, what is the literary-psychological theory that would account for such an effect? The point may seem uncharitable, but it gains considerable force when one considers—as Skinner certainly acknowledges—that organic, hence, nonmechanistic or, indeed, antimechanistic imagery is never abandoned in Leviathan but, in fact, recurs throughout. Mechanistic imagery does not replace but is super-added to the organic. What does this mean in terms of the rhetorical strategy of Leviathan? What underpins what? Why doesn't the organic imagery—imagery suggesting that the state is natural and not humanly constructed—undermine the mechanistic? What can Hobbes—understood as a penetrating student and virtuosic practitioner of the arts of classical rhetoric—have had in mind when employing these “contrasting” and “partly mixed” metaphors (Reference Skinner1996, 386)? Where exactly is the philosophical payoff? Certainly, none of this is to deny that Hobbes did indeed assert the artificiality of the state. But that assertion is the result of an elaborate argument based on conceptual, empirical, and psychological claims about the natural condition of humankind—a condition of anarchy—and presented in the language of abstract philosophical discourse. The recovery, explication, and evaluation of that argument is and can only be a matter of philosophical, rather than literary/historical, analysis. From a philosophical perspective, the rhetorical materials seem at best unhelpful, at worst counterproductive.

Or, to pick a different example, consider Skinner's account of Hobbes' attack on the scholastic doctrine according to which God “infuses” grace into his subjects. Hobbes writes as follows: “the proper use of the word infused, in speaking of the graces of God, is an abuse of it; for those graces are Vertues, not Bodies to be carryed hither and thither, and to be powred into men, as into barrels” (Reference Hobbes1980 [1651], 441). In Skinner's view, we have here a model case of the conventionally recognized figure of tapinosis, whereby the rhetor uses “deliberately inappropriate and undignified terminology to belittle what we are talking about. . .” (Reference Skinner1996, 422). The suggestion is that the infusion theory is repudiated via a rhetorical strategy of ridicule and hence is dismissed out of hand. Once again, Skinner's literary account is entirely compelling; Hobbes' language does indeed seem to belittle the doctrine of infusion and in the way that Skinner describes. But a focus on the rhetoric, however interesting and important in historical terms, is apt to turn our attention away from the fact that Hobbes' criticism is actually based on, and is the culmination of, a lengthy and elaborate discussion—at once philosphical and theological—of the meaning and coherence of the concepts “body” and “spirit,” and of related concepts such as “substance” (Reference Hobbes1980 [1651], 428–42). That discussion is serious, detailed, and systematic. It is presented in language that clearly aims to be precise, abstract, and general. This is a linguistic act the unmistakable goal of which is to prove through rational argument that the nature of spiritual things is such that they cannot be infused in the way that (certain) bodies can. The scholastic view is demonstrably in error; it is subjected to, and cannot survive, close philosophical scrutiny. The fact that Hobbes' conclusion is, in the end, couched in what contemporaries would have recognized as belittling language says nothing about, and makes no contribution to understanding, the actual argument that we find in Leviathan. Again, interpreting the text requires explicating the argument, and explicating the argument is, I would suggest, a matter of philosophical analysis involving rational reconstruction and logical evaluation.

The point is not to deny that Skinner, in his writings on Hobbes and elsewhere, ignores propositional content. Even under what I have previously called the “stronger version,” his work admits that illocutionary force is not the only aspect of a text worth examining. My primary intention, however, is to examine not the particular ways in which Skinner himself applies his methodology but, rather, the logic of the methodology itself. That logic would suggest, among other things, that we cannot actually know what propositions an author is defending without attending to the historical and linguistic contexts that make those propositions what they are. In some sense, illocutionary force determines propositional content. This seems to me incorrect. Again, I believe that the implications of philosophical pragmatics for the analysis of written theoretical works are quite different from what Skinner proposes and that those implications suggest a quite different and philosophically more satisfying approach to the history of political thought.Footnote ¹⁴