Pragmatics without deception: towards a hermeneutics of speech activities in the Lunyu and Mengzi

The Master said, “To not anticipate deception, to not expect untrustworthiness, and yet to be the first to become aware of such conduct—this is indeed to be a person of substance!”

– Lunyu 14.31

Abstract

The Ruist emphasis upon the role-specificity of appropriate conduct affects the way language is employed in the Lunyu and Mengzi, as well as the hermeneutics we ought to adopt in our interpretation of these texts. In the Lunyu, for instance, reputable teachers, rulers, ministers, sons, and other persons employ language that must seem dishonest or duplicitous if measured by the rubrics of accuracy. Chad Hansen and Steven Geisz have argued, however, that such passages depict language used with an eye to its pragmatic significance, rather than a moral failure on the part of the speaker. If we agree with their assessment, we face the unique hermeneutic challenge of preserving constraints upon our interpretation of the early Ruist texts—a challenge I seek to meet by appealing to the aims, methods, and social roles governing the use of such language.

Introduction

In his account of the preconditions for understanding, Gadamer (1995) insists that we must employ (consciously or otherwise) what he refers to as “the fore-conception of completeness.” ¹We do not encounter the linguistic communication of others in complete naivety but rely upon our own preconceptions and assumptions to make sense of what they say. While these preconceptions and assumptions—or what Gadamer refers to as our “prejudices”—make understanding possible, they can also distort our comprehension of what the other person is saying. The trick is to remain open to the possibility that what others have to say may not be what we expect them to say, that their position on a given subject might be significantly different from our own position. Such openness cannot be accomplished or maintained by force of will, but is, as Gadamer puts it, “the first, last, and constant task” of the practice of understanding (p. 267). To remain open we must find ways to risk our own assumptions so that we might discover the differences between our position and another’s, and thus come to understand what she or he is really saying. This is where the fore-conception of completeness enters the picture. This fore-conception involves two distinct second-order assumptions: first, the assumption that the other person’s claims form a coherent whole, second, that what the person says has a claim to truth. The purpose of these assumptions is to provide constraints upon our interpretations, to prevent misinterpretations due to irrelevant assumptions on our part—since these constraints will, at least initially, challenge any interpretation that renders the other’s claims either inconsistent or false.

If we adopt Gadamer’s account of understanding when we seek to interpret the Lunyu and Mengzi we discover several passages that are particularly challenging. In these passages exemplary Ruists—Kongzi and Mengzi—appear to say things they know to be false, while in other passages they appear to contradict their words by yet other words or by actions incongruous with their own normative claims. Kongzi, for instance, does not appear to condemn lying as such; he appears to practice and applaud it.² There is the story Kongzi tells about Meng Zhifan, a minister in the state of Lu: when his forces were routed he stayed behind to defend the rear, and as he finally entered within the city wall he attributed his delay not to his own courage but to the non-responsiveness of his horse. While it is no doubt a species of modesty that motived Meng Zhifan to explain his conduct in this fashion, what he said was still untrue. Despite that fact, Kongzi appears to praise the minister’s conduct (Lunyu 6.15). Another example occurs when a minster of Chen speaks with Kongzi about Duke Zhao of Lu (7.31). When asked if this Duke understood li 禮, Kongzi replies in the affirmative even though it shortly becomes evident in the same passage that such a claim is ludicrously inaccurate—Duke Zhao, after all, married a woman with the same clan name.³ The Lunyu also records an anecdote of Kongzi feigning illness so as to avoid speaking with a certain Ru Bei (17.20). Another relevant anecdote, this time from the Shiji, concerns an event during Kongzi’s journey from Chen to the state of Wei (Shiji 47.29). Detained in Pu—a place governed by a minster of Wei actively rebelling against his ruler—Kongzi is compelled to promise that he will not continue on his way to Wei; only then do the men of Pu allow him to leave. Yet, after leaving Pu, Kongzi breaks his promise and resumes his journey.

If we can attribute to Kongzi some degree of moral indifference to semantic deception, we seem able to attribute a similar indifference to Mengzi. He appears to utter contradictory claims about his knowledge of Huan of Qi and Wen of Jin—claiming in 1A7 that he has no knowledge of them, yet speaking about them at length in 4B21.⁴ He also appears to endorse speaking for effect rather than accuracy when it comes to matters of propensities beyond our immediate control (ming 命) and spontaneous human dispositions (renxing 人性):

Mengzi said, “The way the mouth is disposed towards tastes, the eye towards colors, the ear towards sounds, the nose towards smells, and the four limbs towards ease is due mainly to spontaneous human dispositions, but also to propensities beyond our control. This is why the junzi 君子does not refer to them as spontaneous human dispositions. The way ren 仁pertains to the relation between father and son, duty to the relation between prince and subject, li to the relation between guest and host, wisdom to the good and wise man, the sage to the way of tian 天, is due mainly to propensities beyond our control, but also to spontaneous human dispositions. That is why the junzi does not refer to them as propensities beyond our control.” (Mengzi 7B24, Lau 2003, modified)

Geisz (2008) has argued that Mengzi frequently engages in this variety of “strategic”—and thus potentially deceptive—language. In his conversation with King Hui of Liang on the value of benefit (li 利), for example, Mengzi appears to not only abandon the Ruist position by endorsing, at least momentarily, the Mohist positive estimation of benefit; his condemnation of benefit, expressed in terms of its mutual exclusion of morality (renyi 仁義), appears to leave him no way to justify morality (Mengzi 1A1). It would seem that in this passage Mengzi declares what he does not endorse; and by providing the King Hui of Liang with an inaccurate picture of his genuine position, Mengzi deceives the king.

As Gadamer’s account of understanding utilizes truth and consistency as the main constraints to prevent misunderstanding, passages in which exemplary Ruists deliberately speak falsely or contradict themselves pose a real interpretive challenge. There are at least three alternative responses in the contemporary scholarship to this hermeneutic challenge. James Legge seeks to preserve the truth and consistency of Kongzi’s normative claims by placing the inconsistency upon Kongzi’s conduct, describing him as incontinent. Chad Hansen and Steven Geisz, on the other hand, accept that the exemplary Ruists will not always speak truthfully and consistently. Hansen would have us think that such passages involve lying or semantic deception only if we read them in light of a semantic-focused theory of language; endorse the classical Chinese view of language, however, and these passages will, at most, constitute cases of accidental or non-deliberate semantic deception. Geisz, on the other hand, suggests that while these passages involve the use of deceptive language, and deliberate or strategic deception at that, such utterances can be justified by their consequences. Despite the differences between the approaches taken by Hansen and Geisz, both scholars effectively abandon Gadamer’s interpretive constraints.

I will argue that Kongzi and Mengzi do not use language to deceive others. The appearance of semantic deception is simply the product of our own interpretive shortcomings. Either we fail to interpret their declarative utterances in light of the relevant cultural assumptions and social cues, or we fail to realize that what they are saying is not declarative in nature and thus incapable of constituting semantic deception. Of course, once we allow non-declarative utterances to play a role in the early Ruist literature we must ourselves go beyond Gadamer’s fore-conception of completeness. With this end in mind, I will conclude this paper with a general sketch of what we might describe as a hermeneutics of role-specific speech activities. I will focus on the role-specific aims and methods of various modes of linguistic behavior in an attempt to provide the interpretive constraints necessary for understanding such non-declarative speech acts.⁵

Incontinence, not deception

James Legge takes the passages of the Lunyu and Shiji, mentioned above, as illustrations of Kongzi’s moral fallibility. In his introduction to his translation of the Lunyu, Legge interprets xin 信as the virtue of truthfulness. This creates a tension between Kongzi’s verbal approbation of xin and the passages in the Lunyu and Shiji where Kongzi appears either to utter or to praise deceptive language—a tension Legge resolves by attributing the inconsistency not to Kongzi’s normative claims, but to a disconnect between the man’s morality and his conduct. Kongzi, it seems, simply lacked the ability to act in the way he knew to be right. As Legge puts it, even though “among the four things which it is said he taught, ‘truthfulness’ is specified, and many sayings might be quoted from him, in which sincerity is celebrated as highly and demanded as stringently as ever it has been by any Christian moralist … he was not altogether the truthful and true man to whom we accord our highest approbation” (1971, p. 100). Not surprisingly Legge, a Congregationalist missionary, attributes Kongzi’s incontinence to his atheism: “natural affections, the feeling of loyalty, and enlightened policy, may do much to build up and preserve a family and a state, but it requires more to maintain the love of truth, and make a lie, spoken or acted, to be shrunk from with shame. It requires in fact the living recognition of a God of truth, and all the sanctions of revealed religion” (1971, p. 101).⁶ Legge assumes that a creator deity, such as one finds in the Christian tradition, not only lays the foundation for a final vocabulary of the way the world really is (and so conditions absolute truth), but such a deity provides a good incentive for being moral. Thus, if being moral includes truthfulness, such a deity provides a good incentive for being truthful. Of course such theological thinking has no place in the interpretation of classical Chinese philosophy; and while Legge’s interpretation is able to salvage the verbal consistency of Kongzi’s assessment of truthfulness, it fails to retain the consistency between Kongzi’s words and deeds—a type of consistency quite important to Kongzi and Mengzi. The inconsistency Legge attributes to Kongzi is, in fact, the kind of inconsistency that is attributed to the village-worthy by both Kongzi and Mengzi.⁷

Accidental manipulation

Rather than take these passages as evidence of moral failure we might do well to follow Chad Hansen’s suggestion that classical Chinese philosophers focused on the regulative use of language rather than its informational use—that the practical impact of what was said had more to do with the success or failure of a philosophy than its accuracy or truth (1985, p. 507). In the early Ruist literature we can clearly observe an awareness of, and dominant interest in, the behavioral implications of language: interest in the identity and character of the speaker and listener, in the specific intended illocutionary action, and in the perlocutionary force of what was said—the “pragmatics” of language, in short.⁸ There are several passages in the Lunyu that illustrate this focus. For example, in 13.3 we read:

When names are not used effectively, language will not be used effectively; when language is not used effectively, matters will not be taken care of; when matters are not taken care of, the observance of ritual propriety and the playing of music will not flourish; when the observance of ritual propriety and the playing of music do not flourish, the application of laws and punishments will not be on the mark; when the application of laws and punishments is not on the mark, the people will not know what to do with themselves. Thus, when the exemplary person puts a name to something, it can certainly be spoken, and when spoken it can certainly be acted upon. There is nothing careless in the attitude of the exemplary person toward what is said. (Ames and Rosemont Jr. 1998, modified)

It is also this pragmatic outlook on language that allows Kongzi to claim that there are sayings that, if they inform a ruler’s conduct, can ruin a state or make a state (Lunyu 13.15). One might object to this pragmatic reading. The sayings in 13.15—“Ruling is difficult, and ministering is not easy either” and “I find little pleasure in ruling, save that no one will take exception to what I say”—might, after all, be understood from a semantic point of view as simply presenting claims that can be evaluated as either true or false. Yet Lunyu 11.22 gives us an insight into the way Kongzi uses language to give advice to others. In this passage Zilu and Ranyou both ask Kongzi the same question: “On learning something should one act upon it?” Kongzi, however, gives these two students different answers. When Gongxi Hua asks Kongzi about this, the Master replies that the two disciples were different: “Ranyou is diffident, and so I urged him on. But Zilu has the energy of two, and so I sought to rein him in” (Ames and Rosemont Jr. 1998). Kongzi is aware of what we are describing as the pragmatics of language—how language affects us or, as Hansen puts it, the way language can function “as a form of social behavior and something that regulates that behavior” (1985, p. 506). In giving advice, or quoting sayings, Kongzi is not aiming at accuracy or a semantically true representation so much as he is seeking to condition efficacious behavior on the part of his listener.

Hansen (1985, 1992) claims that the classical Chinese philosophers “focused” on the pragmatics of language, and thus marginalized matters of semantics.⁹ Yet it is not entirely clear whether Hansen regards this focus as the result of selective attention to pragmatics, or blindness to semantics. At times he describes the classical Chinese philosophers as “emphasizing” pragmatics (1985, p. 507). As any sort of emphasis presupposes an awareness of the competing terms, to speak of emphasis is to suggest that the classical Chinese philosophers were aware of both aspects of linguistic communication and elected to prioritize pragmatics above semantics. Selective attention is not so much a general theory of language as it is a communication ethics, and an emphasis on pragmatics does not amount to a view of language too terribly different from our own. If, however, we are to make sense of Hansen’s general claims about the classical Chinese philosopher’s pragmatic theory of language, we must think of the focus on pragmatics not as a matter of selective attention but as a matter of blindness.¹⁰ Hansen’s thesis must be that the classical Chinese philosophers “focused” on pragmatics because it was all that they saw—either because they could account for all semantic aspects of language from a pragmatic point of view (thus reducing semantics to pragmatics), or because they could justifiably ignore semantics when providing a complete account of language.¹¹

Hansen is aware that an exclusive focus on the pragmatics of language can engender the practical exclusion of the semantic features; that it may produce what he describes as a “manipulative view of the role of language” (1985, p. 503). While concern for semantics or the informational use of language commits one to involving the other person’s conscious awareness in how she or he is affected by what one says, an exclusive focus on pragmatics means that one is merely concerned with the effects of what one says—the perlocutionary force an utterance has upon the listener. A teleologically driven approach to communication naturally gives rise to manipulative language, as one’s end of influencing the other in the desired fashion can often be attained without the other’s conscious involvement.

Against the Kantian objection that this view of language is likely to raise, Hansen offers something towards its defense. He suggests that such a view of language need not be morally suspect if we were to adopt the dominant conceptions of agency and moral responsibility employed by the classical Chinese—conceptions that do not assign agency or responsibility to individuals. His claim is quite similar to something François Jullien says:

Chinese thought … had no qualms about conceiving of manipulation [of humans] upstream, in an ongoing process. That is because, from its own strategic point of view, it never drew a distinction between the world and consciousness (or nature and the internal life of a human being, physical laws and moral laws, and so on). So it never subsequently had to bring the two orders together … For Chinese thought, everything constituted a process—everything, including human behavior. Manipulation could thus be imperceptible. (2004, p. 137)

“And unobjectionable,” we might add.

If Hansen is right that the classical Chinese philosophers were effectively blind to semantic considerations, then we must conclude that Kongzi and Mengzi could not willfully deceive others. To speak of these passages in the Lunyu, Shiji, and Mengzi as involving intentional falsehood or deception is to evaluate them in terms of a “semantic” or “semantic-pragmatic” theory of language—a theory of language that does not exclude semantic considerations. It is, in other words, to involve an alien theory of language when interpreting these texts. Such deception, falsehood, or manipulation as we find in the texts is a product of our interpretation. If we persist in claiming that Kongzi and Mengzi engage in deceptive language and praise deception, we must at least acknowledge that—from the perspective of the pragmatic theory of language—such deception is not deliberate, but wholly accidental. If such “deceptive” language on their part is deceptive only from our perspective and our interest in semantic values, such language, from a Ruist point of view, is not—contrary to what Legge suggests—a moral failure on the part of Kongzi or Mengzi, nor is it evidence of moral incontinence. Not only that, but on Hansen’s interpretation there is no textual inconsistency either. This is because a pragmatic theory of language precludes the second general assumption of Gadamer’s fore-conception of completeness: that the person is saying something true about the subject under discussion. For truth is, at least in part, a semantic notion.¹²

There are, however, several linguistic practices mentioned in the Ruists literature that cannot be accounted for on a purely pragmatic theory of language, viz., making good on one’s word (xin 信) and reporting (gao 告).¹³ Hansen says that xin corresponds to the combined requirement of truth-telling and promise-keeping, which one might translate as ‘trustworthiness.’ Yet Hansen maintains that truth-telling and promise-keeping have little to do with a concept of semantic truth. Truth-telling, he says, “does not require that we utter only semantic truths; it does require that our speech honestly reflects what we believe” (1985, p. 515). Presumably, what we believe need not actually be true. Yet from a first-person perspective it seems impossible to have a belief (at least in the propositional sense of the term) without assuming it to be true, and thus having a concept of semantic truth. There are also several passages in the early Ruist literature that involve the use of language to make a report (gao): reporting on one’s surroundings, on one’s destination, on political events, on one’s intention to marry, on what another person said, on the results of a trial, who was capped or married, and so on.¹⁴ While there are certainly pragmatic dimensions to reporting, and even the possibility of using a report for the sake of its results, it is rather untenable to claim that such language does not require at least some awareness of the semantic features of language, such as truth and accuracy.¹⁵

These linguistic practices suggest that we cannot attribute a pragmatic theory of language to the Ruists. As a result their deceptive utterances can no longer be described as accidentally deceptive. When semantics is no longer marginalized or erased by a pragmatic theory, we can no longer say that the Ruists were oblivious to semantics; instead, we must conclude that these passages involve Kongzi and Mengzi deliberately deceiving others—that their language is not merely “manipulative” (in Hansen’s sense of the term of speaking for effect and thus occasionally engaging in non-deliberate acts of deception), but strategic. This is how Steven Geisz interprets the relevant passages in the Mengzi.

Strategic deception

Geisz adopts what he refers to as a strategic–pragmatic reading of the Mengzi, built on the claim that Mengzi uses language strategically to promote the “Confucian dao.” The basic assumption of Geisz’s hermeneutics is that “whenever truth-telling and full descriptive accuracy seem to Mengzi to conflict with promoting the Confucian dao, Mengzi himself would jettison truth-telling and descriptive accuracy in favor of doing with language what seems most likely to achieve his Confucian dao-oriented ends” (2008, p. 193). This is not the blindness to semantics that Hansen speaks of, but a selective focus on Mengzi’s part on the pragmatics of language. If we can attribute such strategic language to Mengzi it would follow that when he appears to make a straightforward claim about his position on any given topic he might in fact be speaking for effect, and thus intentionally dissembling—“he might be more concerned with what effects the uttering of that statement will have on his audience’s behavior than with doing what we (as outside observers) would call ‘stating the truth’ or achieving ‘full descriptive accuracy’ ” (p. 192). Hence, as Geisz puts it, “Mengzi’s real views and his stated position may occupy two separate layers; what he actually says and does may or may not, in any given recorded situation, directly reveal what he really thinks” (p. 193).

Mengzi 1A1 is one of the passages Geisz uses to justify this hermeneutic approach.¹⁶ This passage records a meeting between Mengzi and King Hui of Liang. The King asks Mengzi how the latter might “benefit” (li 利) the King’s state. Mengzi replies by apparently problematizing the King’s endorsement of this norm. In doing so Geisz assumes that Mengzi is offering a critique of the Mohist talk of “benefit.”¹⁷ Yet what is interesting, according to Geisz, is that Mengzi rejects talk of benefit because it is unbeneficial in the Mohist sense of the term: ‘if your lordship speaks of benefitting the state, the ministers will speak of benefitting their family, knights and commoners will speak of benefitting their persons—in short, everyone will take a competitive view of the good and will undo the order of the state.’ By providing this counsel Mengzi seems guilty of two inconsistencies: first, he must assume the merit of benefit in order to reject its value as unbeneficial; second, Mengzi is committed to emphasizing morality (renyi) and ignoring benefit—but how is morality to be justified if not by appeal to benefit? Geisz would have us preserve these inconsistencies but justify them by an appeal to their results. The idea is that Mengzi may be permitted these inconsistencies because he is speaking for effect, not for accuracy or consistency. He is saying what needs to be said to promote the Ruist agenda.

Geisz’s interpretive approach to the Mengzi might be applied to the Lunyu. It is obvious that Kongzi lied about his health in order to avoid meeting Ru Bei, and it is equally obvious that Meng Zhifan, whom Kongzi praises for his lie, did so to play down his conduct.¹⁸ While it is not entirely clear why Kongzi misrepresented Duke Zhao of Lu’s understanding of ritual propriety (li) to the minister of Chen, Slingerland (2003) suggests that doing so was perhaps the only viable response given the situation. The minster of Chen, Slingerland says, “fails to see that asking Confucius to criticize a former lord of his home estate—especially in the presence [of] an official of a rival state—is itself a grave violation of ritual, and that Confucius’s praise of Duke Zhao was the only ritually proper response” (p. 75). These passages appear to reflect as much strategy as what Geisz finds in the Mengzi.

Rather than appeal to alternative theories of agency and moral responsibility to justify his interpretative approach, as Hansen does, Geisz is able to justify the strategic use of language by appealing to nothing stranger than consequentialist reasoning. So long as such language promotes the Ruist agenda or “Confucian dao”—as Geisz puts it—any deception or deliberate manipulation can apparently be sanctioned. The pragmatic–strategic reading of the literature effectively attributes something akin to the Buddhist practice of upaya kausalya, or “skillful means,” to the Ruists.¹⁹ The Buddhist use of language for effect is normally governed by the moral requirements of proper motivation and proper effect, and we might expect a similar set of constraints to govern the Ruist use of strategic language. Geisz, for instance, suggests that the key focus of Mengzi’s philosophical engagements is not the preservation of truth, but the preservation of “pragmatic force”—or what J. L. Austin describes as perlocutionary force (Geisz 2008, p. 207).²⁰ Of course, were King Hui of Liang to realize that Mengzi spoke deceptively and for the sake of effect, that realization might alter and perhaps even undermine the perlocutionary force of Mengzi’s future utterances to the King. At the very least the King may no longer trust Mengzi. To be able to justify strategic language we must be able to identify those relationships in which its utilization will not, ultimately, prove fatal to the relationship and contrary to the desired effect.²¹ Yet, just as deceit can sour a relationship between persons, so too can deceit undermine our ability to properly interpret a text.

Geisz is well aware of the hermeneutic risk taken by his strategic–pragmatic reading of the Mengzi: “If the position we attribute to Mengzi is different from the positions he seems to espouse, then can we not just attribute to Mengzi whatever position we wish him to hold?” (2008, p. 193). The challenge is to find hermeneutic constraints other than truth and consistency that can prevent distortion or misunderstanding. Geisz proposes two constraints. First, we must assume that the means Mengzi employs to bring about the desired modification in the other person’s behavior are morally consistent with the overall Ruist agenda he is promoting.²² Second, “even if some of what Mengzi says is strategic, it seems plausible for us still to assume that much of what he says is intended to be true and that we should look for alternatives only insofar as there is some clue that an alternative is plausible”—in other words, we ought to employ a strategic reading only when there is some indication that this is appropriate and otherwise assume that he is speaking truthfully. Since presumably in many situations the best way to use language to shape an interlocutor’s behavior will be to tell the truth and to accurately describe the relevant state of affairs, perhaps the only indication we will have that Mengzi is speaking strategically is when truth-telling and full descriptive accuracy are incompatible with promoting the Ruist agenda (pp. 193–194). Unfortunately, these constrains are insufficient. Each of them, with their appeal to the content of the Ruist agenda, is circular. After all, any passage we rely upon to inform our grasp of the Ruist agenda may very well be strategic in nature.

The insufficiency of these constraints is demonstrated by Geisz’s misreading of Mengzi 1A1. To begin with, it is not obvious that this passage concerns the Mohist notion of “benefit.” Li 利 is also discussed in the Lunyu, where it is contrasted with “appropriateness” (yi 義) and means something akin to seeking personal advantage. So the question is whether Mengzi is discussing li in the sense of Mohist “benefit” or in the Ruist sense of “personal advantage.” It is very likely that Mengzi is talking about the Ruist conception of li since talk of seeking li for one’s state, one’s clan, one’s person is necessarily an exclusive and individuating notion of li. Hence, Mengzi’s argument is that thinking in terms of exclusive, “personal” advantage (my state, my clan, my person) undermines itself. Mohist “benefit,” on the other hand, does not lead to such individuated thinking if it is properly understood. According to the Mozi, “inclusive care” (jianai 兼愛) is the best way to achieve benefits (li), and jianai excludes personal advantage—in fact, in one of the three versions of the “Inclusive Care” chapter of the Mozi, jianai is defined by a lack of ownership or “mine-ness” that characterizes the Ruist concept of li. Furthermore, ren 仁—part of the binomial in Mengzi 1A1 that is translated as “morality”—is best realized, according to the Mozi, by adopting a policy of jianai. The tension Mengzi sees between ren and li is untenable if he is dealing with the Mohist notion of li. In short, if this passage is a critique of Mohism, it relies upon a strawman fallacy; if we are going to be charitable to Mengzi and not charge him with this fallacy, we must assume that he is working with the Ruist rather than the Mohist notion of li. That Mengzi is dealing with the Ruist notion is further substantiated by the King’s initial salvo: ‘To come from so far away—you must have a way of benefiting this state, my state (wu guo 吾國), and not other states.’²³ It is an exclusive benefit the King has in mind—the very definition of li in the early Ruist literature.

But let us imagine that Mengzi is offering a critique of the Mohists; even still, he is not necessarily being inconsistent in the manner Geisz suggests. The first kind of inconsistency Geisz attributes to Mengzi—endorsing the Mohist value of benefit only to prove that this value is itself unbeneficial—is best understood as nothing more than a perfectly legitimate reductio ad absurdum. He is trying to convince the King that endorsing the value of benefit will undermine itself. Of course, persons who perform an RAA of an opponent’s position need not take any position themselves; by objecting with an RAA one need not endorse the terms that lead to the contradiction.²⁴ The second kind of inconsistency Geisz mentions (that is, ignoring benefit for the sake of morality and thus dismissing the ultimate justification of morality) requires that we think of the early Ruists as endorsing a consequentialistic moral theory. Geisz claims that “implicit in Mengzi’s Confucian project is an assumption that a focus on rites, benevolence, and rightness is incompatible with a focus on profit/benefit and the Confucian concerns properly trumps any concern with profit/benefit” (2008, p. 196). Everything hangs on the meaning of ‘focus’ here and unfortunately Geisz equivocates between focus in the sense of ‘one’s attention to any degree’ and focus in the second sense of ‘one’s primary, fundamental, or highest concern.’ It is true that the early Ruists claim that “profit/benefit” may not be pursued if by immoral means, and so morality remains fundamental (focus in the second sense); yet this does not prevent the Ruist from giving some thought to benefit (focus in the first sense)—especially in considering how to be moral (specifically, how one ought to engage in gift-giving, diplomacy, remonstrance, teaching, and the like). While we might suggest that Xunzi’s defense of Ruist practice comes down to “benefit” (in the sense of staying alive, among other things), this seems an indefensible way of reading the Lunyu or the Mengzi.²⁵ Besides, Mozi and Kongzi have contrary notions of good consequences; Mengzi’s rejection of Mohist “benefit” need not exclude the normative relevance of good consequences discussed by the Ruist, such as “social harmony” (he 和).²⁶

The basic problem facing Geisz’s approach is that he has unwittingly endorsed what Gadamer calls, following Paul Ricoeur, a “hermeneutics of suspicion” (1984). It is a hermeneutical approach that anticipates “pretext.” In his essay, “Text and Interpretation,” Gadamer distinguishes between the hermeneutical concepts of text and countertext. Despite the fact that everything written may appear to be a text, only texts that are “textualized”—that is, “that fix the original announcement in such a way that its sense is unequivocally understandable”, or that make “the effort… to avoid strife, to exclude misunderstanding and misuse, and to make univocal understanding possible”—are texts in the primary sense of the term (Gadamer 1989, pp. 35–36). Textualization is accomplished by successfully minimizing any reliance upon the details of the original communicative situation or any unarticulated cultural assumptions. Gadamer suggests that judicial codification and judicial hermeneutics well illustrate the process of textualization. It is significant that in his example Gadamer attributes the process of textualization to both the act of writing and the act of interpreting because, in textualization, the “task of writer corresponds to that of the reader, addressee, interpreter: that is, to achieve such an understanding and to let the printed text speak once again” (p. 35). “Countertexts,” on the other hand, are texts in a loose sense of the term “that offer resistance and opposition to textualization” (p. 37). They are those “many forms of linguistic communicative behavior”—such as speaking ironically, joking, lying, engaging in indirect communication, and so on—“that cannot possibly be subjected to this kind of finality” (p. 37).

One form of countertext is “antitext”: a text or form of discourse the proper interpretation of which relies upon the spoken word (the tone, timing, and so on), context, and cultural pre-understanding.²⁷ Antitexts “resist textualization because in them the dominate factor is the situation of interactive speaking in which they take place” (Gadamer 1989, p. 37). Jokes and irony are examples of antitext: they appear to state one thing, but are not understood properly if we take them to mean what they state; they are properly understood only when we go beyond the “text.”²⁸ With a joke what is said is not seriously meant, while with irony what is said is often the opposite of what is meant. We can tell someone is joking not by what is said so much as by how it is said—who said it, under what circumstance it was said, the manner in which the words were spoken, and other such situational signals. A joke is designated not textually but contextually. Even if we were to attempt to represent these situational or contextual signals in a detailed narrative of the joke there is a sense in which “a joking remark clearly belongs to the moment and thus really cannot be repeated” (p. 37).²⁹ Repeating the words of a joke often proves insufficient; in such cases one must recreate or perform the joke for it to come off. Beyond their reliance on situational or non-textual cues, jokes also resist textualization insofar as they require un-verbalized cultural assumptions—assumptions that remain un-verbalized since to verbalize them is often to destroy the joke’s capacity to be humorous. As for irony, while it may not rely upon situational clues to the degree jokes do, a shared pre-understanding still plays a significant role in its proper detection—and those who speak ironically, by definition, want their irony to be detected and their hidden meaning worked out. Were one to attempt to textualize ironic communication by replacing irony with straightforward formulations the result would not fully capture the communicative significance of irony.

Any communicative behavior that cannot be reduced to text because it relies upon contextual cues of cultural assumptions is very likely an example of antitext. Yet there is an additional aspect that sets antitext apart from other forms of communication: every case of antitext presupposes “solidarity.” Such solidarity is rather complex as it involves interlocutionary, prejudicial, and topical forms of solidarity. In every case of antitext the speaker or writer intends what he or she means to be understood; even if that meaning is cloaked by humor, irony, or indirection, it cannot count as genuine antitext if the speaker or writer does not at least intend for the listener or reader to grasp what is really meant. Secondly, as we have already discussed, antitext resists textualization because it requires that the speaker and listener share certain cultural assumptions—a prejudicial form of solidarity. Finally, antitext involves a shared focus, between speaker and listener or writer and reader, on a common topic. The meaning of what is said by antitext concerns some subject matter (die Sache) both speaker and listener are focused upon, and not merely the opinion of the speaker or his or her true motives. In other words, when we encounter antitext there is no need to abandon the fore-conception of completeness—particularly the assumption that what the other persons says may be the complete truth of a given subject matter.

Another form of countertext Gadamer discusses, “pretext,” resists textualization not simply because it relies upon situational signals or cultural assumptions but because what is truly meant is neither said nor left unsaid, but concealed; and such concealment is incompatible with its textualization (Gadamer 1976, p. 90). While the meaning of what is said is hidden in both antitext and pretext, one consideration that sets these two types of countertext apart concerns the intention of the speaker. With antitext the ironic or humorous meaning is left unsaid and thus hidden, but the speaker of antitext intends the hidden meaning to be discovered by the listener (regardless of whether she or he is successful in this task). This is not the case with pretext; here, speakers use language to conceal their own motives. What is said is merely an excuse, a pretense—pretext—to conceal the true motives of the speaker. Propaganda, lying, and dreams (from the perspective of depth psychology) are all examples of pretext (Gadamer 1976, pp. 90–91; 1989, p. 39). “The complicated interweaving of interpersonal relationships encountered in lies”, Gadamer explains, “has in itself no primarily semantic character. He who lies like a book does so without stuttering and without showing embarrassment, that is, he even conceals the concealment that his speaking in fact is” (1976, p. 90). The concealment of concealment is also encountered in ideologically slanted texts. Such texts use the spread of information as an excuse or pretext to conceal their true objective of shaping public opinion. Hence, the interlocutionary solidarity presupposed by antitexts is absent in pretexts as the speaker of the latter seeks to hide their true intentions from the listener.

But the topical solidarity of antitext is also missing from pretext, and this is a second consideration that separates pretext from antitext. Compare, for instance, educational or pedagogic texts with those that are ideologically slanted. Pedagogic texts, like ideological ones, seek to influence their audience. It is also possible that the pedagogic texts of yore (such as historical accounts of the causes of the American Civil War or the valorization of Christopher Columbus) may be seen as ideologically slanted texts today. Yet there is a difference to be found between these two sorts of texts, and it has to do with the means by which a pedagogic text transforms into a propagandist text—from text to pretext. It occurs only when our cultural assumptions shift and we no longer take these earlier pedagogic texts as true of their subject matter but, rather, as symptomatic of the cloaked assumptions and implicit motivations of their authors. Pretext is discovered only when we experience a loss of topical solidarity with the speaker or author, either because we encounter some sort of practical contradiction when acting on what the other says, or a disruption of possible agreement in understanding or consensus. As we become suspicious that what the other has to say is not the truth we may eventually abandon the fore-conception of completeness. When this occurs the nature and aim of understanding is altered; we engage in what Gadamer describes as a “hermeneutics of suspicion” (1984). Rather than allow what is said to influence our understanding of the topic at hand (die Sache), we focus on understanding the psychology or historicity of the speaker (1995, p. 294). We go beyond what is said and its intended meaning, and instead seek to uncover the concealed motive of the speaker. Hence, pretext differs from antitext not simply in terms of a speaker’s intention to conceal her or his true motives, but also in terms of how interpreters receive what is said—whether they meet the text with an attitude of suspicion.³⁰

An overriding interest in the effects of what is said, such as Geisz attributes to Mengzi, will occasionally motivate one to engage in pretext. This is apparently the case in Mengzi 1A1. There Mengzi is said to conceal his own Ruist agenda behind a conversation about Mohist values. Were his true intentions to come to the surface, not only would he fail in his aim but very likely he would lose the trust of the King. While there is no indication that the King receives what Mengzi says as pretext, Geisz recommends that scholars only properly understand this passage when they grow suspicious and focus not on the truth of what Mengzi says, but upon Mengzi’s true intentions. This is, of course, to engage in a hermeneutics of suspicion.

The trouble is that from the moment we attribute pretext to another we cannot be confident that anything the person says is not also pretext. Once we begin to suspect a text of a concealed meaning it becomes almost impossible to stop. Gadamer’s interpretive constraints of assuming that the other’s claims are consistent and true involve both respect and modesty: we must respect the other enough to stop ourselves from willfully misconstruing what they say, and we must have enough modesty to admit that they might know more than we do and thus might be able to correct us. These characteristics lead Robert J. Dostal to describe Gadamer’s hermeneutics as one of trust (1987). A hermeneutics of trust will hold a high standard when it comes to the effort we put into rendering a text consistent with itself; a hermeneutics of suspicion, on the other hand, lowers that standard. It releases the pressure to find an interpretation that renders the text consistent by allowing us to assume that the author or speaker might simply be dissembling. Ultimately, by lowering the standards of a successful interpretation, these interpretations become unimpeachable and unfalsifiable.

Such is arguably the case with the hermeneutics of suspicion Sigmund Freud employs in his interpretation of dreams. In an attempt to defend his thesis that the meaning of every dream is wish-fulfillment, despite the apparent counterexamples provided by painful dreams, Freud distinguishes between manifest and latent dream-content. With dreams that are easily recognized as cases of wish-fulfillment the manifest and latent dream-content coincide; but this is not the case with dreams that appear to be painful. With these dreams the manifest dream-content is distorted and deformed. Freud claims that such distortion is intentional. He posits two psychical forces to dream formation. The first is the unconscious that “forms the wish uttered by the dream” (1999, p. 113). The second force acts in a fashion analogous to a political censor. The first force seeks to communicate the fulfillment of a wish, while the other seeks—when such a wish-fulfillment would disturb our sleep—to distort the dream-wish beyond recognition. There is even a dialectic involved: knowing when the manifest fulfillment of a particular wish will be censored, the unconscious may moderate and willfully distort the dream-content so that it might go uncensored. To use Gadamer’s vocabulary, we might say that the unconscious engages in a bit of antitext—“the stricter the censorship, the more far-reaching the disguise and often the cleverer the devices which nevertheless put the reader” or conscious mind “on the track of what is really meant” (p. 113). In short, the unconscious seeks to be understood, and may engage in the willful, strategic distortion of the wish “uttered” by the dream for the sake of communicating the latent content or real meaning of the dream to our conscious mind. The psychical force that censors our dreams, however, has a very different intention—it engages in “distortion and deformation … as a means of pretense” (p. 112). The problem is that, in defending his general thesis that all dreams are wish-fulfillment, Freud employs his thesis as an ideology—as an interpretive lens. He does not approach the matter empirically and articulate his thesis in a falsifiable manner. In short, a hermeneutics of suspicion sabotages one’s own ability to learn from the other.

Another consequence of a hermeneutics of suspicion is that we become hypersensitive to inconsistencies, seeing them in passages where they may not be occurring at all. This approach to the text assumes, as Dostal puts it, “too high a standard of logic and control of the text by the author. Every time there is a contradiction or a difficulty in the text does not mean that the author is dissembling” (2008, p. 259). To think otherwise is to assume that the author is godlike in their control of the text and hidden meanings. It is, to quote Gadamer, “Talmud in the wrong place.”³¹ Thus Geisz’s pragmatic–strategic reading of the Mengzi, as a hermeneutics of suspicion, too easily sanctions the imputation of strategic language. We cast about for inconsistencies, taking them as symptoms of strategy; and rather than attempt to resolve these inconsistencies by appealing to what is said in the text and its context we hunt for ways in which such inconsistencies might serve the Ruist agenda.

Besides these common objections to a hermeneutics of suspicion, there are specific reasons for rejecting its use in the interpretation of the early Ruist literature. First, these passages in the Mengzi and, by extension, the Lunyu do not involve pretext. How can one determine whether a passage involves antitext or pretext? The fluid transition from pedagogical text to propagandist pretext that we discussed above illustrates the point that designating a “text” as either antitext or pretext cannot be resolved simply in terms of what is said. It is partly a question of whether or not the listener enjoys or does not enjoy solidarity with the speaker—whether they share a horizon of pre-understanding. Yet someone might lie to us and we might not catch it because it seems to be within our horizon of assumptions. So, solidarity or its lack is an insufficient criterion. Let us then add reference to the speaker’s intentions: whether we hunt for the psychological causes of their utterances or text hinges upon our classification of the text as either antitext or pretext. Generally speaking, however, we can say that pretext is text that is caused by a motive to conceal rather than communicate by what is said (“text”) or even by what is unsaid (“antitext”). In fact intention seems to take precedence over reception—as someone might take irony as deception, and yet the speaker can sensibly refute that reception, saying that it was merely a failed ironic statement, just as one might try to disown one’s racist ideology by asserting one’s claim was ‘only a joke.’ Looking to both the speaker’s intention and the listener’s reception there are then two questions we can ask of any utterance or text to test whether it is in fact pretext. First, is there a loss of solidarity or trust; does the listener search for the true motives and ignore the truth-claim of what is said? If so, this means the listener receives the text as pretext. This is suggestive of pretext, but not itself conclusive. Second, is the true motive of the speaker necessarily concealed? That is, does the discovery of the true motive of the speaker undermine the efficacy or performance or success of what is said? Does the speaker, in other words, intend to manipulate the listener? If so, what is said is assuredly pretext.

When we look back at the passages in the early Ruist literature that gave us pause, and make use of these two questions, we can see that many of these passages are more justifiably interpreted as antitext than pretext. What at first appears to be dissemblance becomes, upon closer inspection, something else. Take Kongzi’s rebuff of Ru Bei. Kongzi may falsely report his state of health, but there is no intention on his part to hide the truth—as is plain by his playing the lute and singing right after communicating this falsehood. This passage does not contain pretext, but antitext. In this case, Kongzi is quite possibly communicating a hidden insult much like someone might communicate a hidden meaning in a bit of irony. As for Kongzi’s misrepresentation of the propriety of the Duke Zhao of Lu in Lunyu 7.31, it is undeniable that the minister of Chen takes this misrepresentation as willful deceit, which he thinks is motivated by partiality. Yet how a listener receives what is said is only indicative of pretext; essentially, pretext is a matter of intention. Yet when Kongzi’s student, Wuma Qi, reports the minister’s assessment to Kongzi, the Master makes no attempt to vindicate his earlier claim that the Duke Zhao of Lu understood ritual. Were it indeed partiality that motivated Kongzi, and thus made his misrepresentation intentional, we would expect Kongzi to attempt to defend the Duke’s reputation, if only to his own student. While we might read Kongzi’s final remark—“I am indeed fortunate; if I have committed a transgression, others are certain to notice it”—as ironic, as Slingerland suggests, it is just as plausible—and, indeed, consistent with the Master’s love of learning—that he is accepting responsibility for misrepresenting Duke Zhao of Lu’s understanding of ritual. Falsehood or misrepresentation, however, is insufficient for pretext.³²

A second reason to avoid implementing a hermeneutics of suspicion when interpreting the early Ruist literature is that while Kongzi and Mengzi are both fully aware that others might try to deceive us, both are committed to the value of assuming that others are trustworthy. As Kongzi expresses it, “To not anticipate deception, to not expect untrustworthiness, and yet to be the first to become aware of such conduct—this is indeed to be a person of substance!” (Lunyu 14.31). To anticipate dissemblance on the part of the other is to sabotage the chance of a socially productive relationship, just as trusting someone even before they prove themselves worthy of being trusted can cultivate a socially productive relationship. The way that Yao trusted Shun with official positions and thus cultivated Shun’s trustworthiness is a case in point. Another classical example is recorded in Mengzi 5A2. After once again dodging his family’s attempt on his life, Shun returns to his home only to find his step-brother, Xiang, about to make off with his belongings. Shun, extending trust even to one who just attempted to kill him, mistakes his step-brother’s presence in his home and the concern expressed on his step-brother’s face as genuine concern for his welfare. This perpetual trust on the part of Shun has much to do with his ability to eventually cultivate his family members. We might also suggest a contemporary example, such as loaning the family car to the child who is learning to drive. This is an act of trust on the part of the parent and comes before the child has demonstrated her or his trustworthiness; and yet such an act can afford the child an opportunity to earn that trust. In this way trusting another can be transformative.

Yet how can we simultaneously avoid mistrusting others and still be the first to detect dissemblance on their part? Even if it is psychologically possible to do so, is Kongzi not contradicting what he says he learns from Zaiwo: that you cannot trust that people will necessarily do what they say they will, but you must watch what they do? (Lunyu 5.10). Perhaps the solution is to distinguish between being trustful and being gullible. We see this kind of distinction in Lunyu 6.26:

Zaiwo asked, “With consummate persons (renzhe仁者)—if they were informed that there was another consummate person down a well, would they go in after him?”

Kongzi replied, “How could this be? The junzi might save him but they would not fall into the well themselves; the junzi may be cheated but not deceived.”

When exemplary persons (junzi) trust the words of others, they do not allow themselves to be significantly compromised by this trust. Theirs is a critical form of trust—one that looks to the person’s conduct and does not simply trust a person’s words. When Kongzi remarks that Zaiwo’s desultoriness taught him to watch what a person does and not simply trust that they will do what they say they will, it is important to point out that Kongzi is not now suspicious of others. Kongzi is not abandoning trust and embracing mistrust; rather, he is expanding the conditions of continued trust, rendering his trust informed or inquisitive. In other words, Kongzi can maintain his trustful optimism when he hears what another promises to do, but that does not stop Kongzi from also attending to what the other person actually does—and it is this perpetual awareness of the other that takes Kongzi beyond blind trust or gullibility. A hermeneutics of suspicion appears, therefore, to be inconsistent with the hermeneutic approach that Kongzi enjoins when it comes to the evaluation of others.

A hermeneutics of speech activities

We have shown that interpreting passages in the Lunyu and Mengzi that appear to involve semantic deception as cases of moral incontinence, manipulative language, or strategic language produce unsatisfactory results. All three interpretive approaches share a similar assumption about these passages—that they are descriptive or constative utterances. However, not every form of linguistic behavior is descriptive or best evaluated in terms of accuracy. Mourning, oath-taking, encouraging, joking, storytelling, cursing, blessing, as well as the enactment of courtesies are all examples of communicative behavior, or speech activities—to borrow Michel Foucault’s expression—that are not declarative in nature.³³ As a properly grasped non-declarative speech activity is incapable of semantic deception, it would be inaccurate to say that a person speaking strictly courteously or humorously is either accidentally (as with manipulative language) or intentionally (as with strategic language) deceiving others.³⁴ To assure the other that what we just said was merely a joke is to effectively disarm our utterance of any declarative significance, robbing it of any truth-value.³⁵ We can see Kongzi doing just this in Lunyu 17.4:

The Master was at Wucheng when he heard the sound of stringed instruments and singing. Smiling he said, “Why use an ox-clever to slaughter a chicken?”

Zi You replied, “Formerly, Master, I have heard you say, ‘if a junzi studies the way he is able to care for others; if a xiaoren studies the way he is easily employed by others.’ ”

The Master said, “Students—what Zi You said is correct; my previous comment was merely in jest (xi 戲).”

While Kongzi’s initial comment may be nothing more than a joke, Zi You seems to think that Kongzi is using the semblance of a joke to indirectly attack some perceived form of overkill—either the ruler’s attending to music before he has seen to the basic needs of the people, or someone of Zi You’s ability serving in the small city Wucheng. Reassuring Zi You that what he said was merely a joke, however, Kongzi seeks to strip his initial comment of any declarative significance, hidden or otherwise.

The non-declarative nature of certain speech activities is particularly salient to an interpretation of the Lunyu and Mengzi. In fact, in the texts we have not yet addressed—Lunyu 6.15, Mengzi 7B24, and Shiji 47.29—the crucial point in each text, the very utterance that would suggest deception, is in fact non-declarative in nature.

It is to substantiate his claim that Meng Zhifan did not boast about his meritorious accomplishments that Kongzi recounts the story of the minister’s staying at the rear during retreat. Meng Zhifan’s utterance—“It is not that I stayed behind out of bravery but that my horse was unresponsive”—is to be understood as a humble or modest utterance. Qidiao Kai, one of Kongzi’s students, performs a similar utterance in Lunyu 5.6. Kongzi announces that he regards Qidiao Kai ready to assume office. Yet when Kongzi hears the student reply, “I am not yet to be trusted with an undertaking such as that”, the Master is pleased. By such expressions of humility, as we have in these two passages, one not only demonstrates a commitment to the inherent value of one’s accomplishments, one also preserves that commitment. It is the difference between studying to develop oneself and studying to impress others (Lunyu 14.24). Surely an expression of humility need not be accurate to be valuable and effective.

The worst we can say about Mengzi 7B24 is that, in it, Mengzi is advocating the practice of emphasis. While emphasis may mislead, it is not a case of semantic deception. Furthermore, the rationale behind the emphasis he suggests is that in each case it serves to encourage moral development. To attribute our sensual desires to propensities beyond our control encourages us to ignore their cultivation, while to attribute moral conduct to our spontaneous human dispositions is to emphasize the part we can play in developing them. Mengzi is not describing; he is using emphasis to encourage. He appears to be up to something similar when he says that the sage is only human, or that we all possess the four sprouts of moral excellence just as we possess the four limbs of our body (2A2, 2A6).

Finally, in Shiji 47.29 Kongzi justifies his apparent breach of trust. He explains that he broke no promise because one of the conditions for the successful performance of promising—namely, that one is not compelled to make the promise—did not apply to his situation. A promise under duress is, he intimates, no promise at all.³⁶

How are we to properly interpret speech activities, such as the expression of humility or encouragement? Initially, at least, we must have a way of properly identifying the speech activity that is involved in a given text. As Xiao (2005) argues, the pragmatics of any given utterance may sometimes be textualized in the early Chinese literature by means of specific grammatical particles, or so-called “empty words” (xuzi 虛字) (p. 2; cf. Pulleyblank 1995, p. 12). But, as Xiao also argues in the same essay, such textualized indicators are neither necessary nor always sufficient: “the grammatical … features of linguistic expression cannot determine how they can be used pragmatically, even though they may provide useful clues. This suggests,” he continues, “that it is not enough to take the logical, grammatical, and semantic approaches when we study Classical Chinese texts. Following Classical Chinese scholars, we should … focus directly on the communicative practice” (p. 19). The “communicative practice” in the early Ruist literature will certainly include “concrete contexts,” as Xiao suggests. But it will also include the identities and reputations of the speakers, their roles vis-à-vis those listening, and—significantly—the speech activities that are relevant to these roles. Once we have a sense of the speaker’s social character we can use our understanding of role-specific speech activities to properly discern the speech activities involved.

Developing the resources to recognize speech activities is certainly important, yet it does not address the question of how we ought to interpret non-declarative speech activities. The fact is that the interpretive constraints of Gadamer’s fore-conception of completeness are uniquely suited to deal with declarative utterances. We might agree with Gadamer that declarative utterances, especially those that are fully textualized, “are the preferred objects of hermeneutics” but for the very real possibility that these utterances may be the only objects his hermeneutics is equipped to handle (1976, p. 90). Unfortunately, this preference for declarative utterances can become a distorting lens when it leads us to define certain speech activities as antitext or pretext when they ought rather to be understood independently of the concepts of text and countertext. Jokes, for instance, may very well indicate; but it is a stretch to say that they report. A racist joke neither reports its teller’s bigotry nor does it directly state any matter of fact about race, even if telling it or laughing at it indicates a person’s bigotry. While irony may very well communicate, in a cloaked fashion, some declarative, jokes are not essentially connected to such content. As such, while irony may be an example of antitext, as Gadamer suggests, it is difficult to see how jokes are antitext as well. Then there is Gadamer’s account of what he refers to as “Oriental forms of courtesy” (p. 90). Interpreting such communicative customs in terms of the rubric of accuracy, Gadamer concludes that these courtesies are themselves a form of lying, pretext. But to suggest that the proper response to many of the verbal courtesies of ritual propriety (li) is to grow suspicious—this is going too far. In their account of the pragmatic interests of the Ruists Hansen and Geisz at least freed us from a strict focus on declarative utterances and Gadamerian “text.” Yet if we are to improve upon their accounts, we must not only recognize when a speech activity is or is not descriptive in nature, but also appeal to the interpretive constraints relevant to the particular speech activity—constraints that may not always fall within the fore-conception of completeness.

One such interpretive constraint—and one that can apply to declarative as well as non-declarative speech activities alike—must be the fore-conception that the speech activity is successful. When we are dealing with a declarative speech activity, success is a matter of truth; when we are dealing with a string of declarative utterances, truth and consistency are relevant interpretive constraints. The nature of success will, however, vary widely among the different types of non-declarative speech activities.³⁷ For instance, we assess the success of remonstration, oath-taking, and encouragement in very different ways. Yet if we are able to not only identify the relevant speech activity, but also have a sense of its aims and methods, we will have a good sense of what success looks like.

Understandably, not every speech activity will be successful. Yet if we stand a chance of understanding what is said and thus learning from the speaker, we must begin with the assumption that it works, looking for a social or psychological explanation of failure only if our initial assumption cannot be vindicated. This is to preserve a hermeneutics of trust, and is the only way to sustain our ability to learn from what the early Ruists have to say.