This paper is forthcoming in a special issue of the Journal of Chinese Philosophy devoted to Mohist logic. In lieu of an abstract, I’ve reproduced the Introduction here.
Truth in Mohist Dialectics
Chris Fraser
University of Hong Kong
The Mozi famously proposes three “standards” (biao 表) or “models” (fa 法) as criteria for evaluating teachings, claims, or policies. A longstanding controversy in the interpretation of Mohist thought concerns exactly what the three standards are criteria of. Are they intended to evaluate whether a teaching is true, morally right, pragmatically useful, or something else? A seemingly natural interpretation, motivated partly by Western philosophical assumptions, is that the models are criteria for judging the truth of an assertion or theory. Watson, for instance, interprets them as three tests of the “validity” of a “theory.” Schwartz and Wong both take them to be three tests for “verifying a proposition.” Graham calls them three tests of “assertion” and contends that they concern issues that are “purely factual.” In a more recent discussion, Van Norden suggests that they are “indicators of truth.” Against these interpretations, Hansen has contended that the best explanation of the first and third models is that the Mohists are concerned not with truth, but with “appropriate word or language usage” or pragmatic “assertibility.” He suggests that the Mohists are not treating the semantic issue of how to determine whether a sentence is true, but the pragmatic one of how to determine whether the use of words is appropriate.
The general approach of evaluating statements, actions, and policies by distinguishing whether they are relevantly similar to a standard figures prominently both in the core books of the Mozi (books 8–37) and in the Mohist dialectical texts—the six books that form the so-called “later Mohist” texts or Mohist “Dialectics.” Both use the same terminology for such criteria, referring to them as fa (model, standard). Unlike the core books, however, passages in the dialectical texts explicitly treat semantic issues, such as the grounds by which to distinguish whether things fall under the same general term and the status of utterances disputants might make in a debate over which of two terms fits an object. If the three standards are not criteria of truth, are these later Mohist texts also evaluating utterances in terms of some pragmatic status, rather than truth? Does a concept akin to truth have any role in Mohist dialectics, whether in the core books or the dialectical texts? Hansen argues boldly that “Chinese philosophy has no concept of truth” and that later Mohist thought instead applies purely pragmatic, not semantic, terms of evaluation. Utterances are evaluated as to whether they are “admissible” or “assertible” by practical standards, not by whether they are correct in a specifically semantic sense. He offers three main arguments for this interpretation. The first is that early Chinese theories of language had a pragmatic, not semantic, orientation, and thus there was no role for a concept of truth. The second is that early Chinese thinkers did not theorize about the status of sentences, the units of language that admit of evaluation as true or false. The third is that Mohist dialecticians evaluated the status of utterances not in terms of a concept corresponding to truth, but in terms of whether they were ke 可 (“permissible”), a concept with a pragmatic connotation.
This essay will review the case for the claims that the Mohists’ three standards are something other than standards of truth and that even later Mohist dialectics employs no term of semantic evaluation corresponding to “true.” I will argue that Hansen is correct that the three standards are not criteria of truth, specifically, but of a more general notion of the correct dao 道 (way). However, they do not preclude a concern with truth, and their scope probably covers questions of truth. Later Mohist dialectics likewise does not focus specifically on truth or employ a concept that aligns exactly with “true.” Nevertheless, I will argue, the texts do employ terms that play the same expressive role as “…is true.” Thus, contra Hansen’s thesis, these texts can justifiably be said to have a concept of semantic truth.
To download the full paper, use this link.
