Knowledge of language in action

Abstract

Knowledge of a language is a kind of knowledge, the possession of which enables a speaker to understand and perform a variety of linguistic actions in that language. In this paper, I pursue an agency-oriented approach to knowledge of language. I begin by examining two major agency-oriented models of knowledge of language: Michael Dummett's Implicit Knowledge Model and Jennifer Hornsby's Practical Knowledge Model. I argue that each of these models is inadequate for different reasons. I present an Acquaintance Knowledge Model, in which a speaker's knowledge of a language is a combination of the speaker's first-order linguistic ability and second-order acquaintance with his ability and actions.

Keywords:

• agency
• acquaintance
• practical knowledge
• Michael Dummett
• Jennifer Hornsby

Acknowledgements

Earlier versions of this paper were presented at the Body, Cognition & Meaning (BC&M) Conference held at Academia Sinica in 2009, at the Faculty Research Colloquium held at Soochow University in 2010, at the 2010 Annual Conference of Taiwan Philosophical Association (TPA), and at the Knowledge and Logic Conference held at National Taiwan University in 2012. I am grateful to the participants and especially to Kuang-ming Cheng (my commentator at the BC&M conference), Wan-chuan Fang, Chien-hsing Ho (my commentator at the TPA conference), Jih-Ching Ho, Richard Hou, Daisy Ku, Cheng-hung Lin, and Norman Teng for helpful comments and questions. I am also grateful to Barry C. Smith and two anonymous referees for their valuable comments and constructive criticisms on earlier versions of this paper. This work was supported by National Science Council (NSC 99-2410-H-031-009-MY3).

Notes on contributor

Cheng-hung Tsai is Associate Professor in the Department of Philosophy at Soochow University, Taiwan. His research interests are in epistemology, philosophy of language, and philosophy of mind. He has published articles in journals such as Philosophia, Theoria, The Southern Journal of Philosophy, and Phenomenology and the Cognitive Sciences.

Notes

1. Cf. “Knowledge of a language supports both the active ability to speak in that language and the passive ability to understand what others say in it” (Dummett 2007, 562). See also Dummett (2006, 49–50).

2. The assumption that what constitutes being a hearer is receptivity is widespread but not unchallengeable. Some scholars might assert that listening, like speaking, is something we do, not something that happens to us. For example, Crowther says, “It is a familiar thought that in perception we are passive and at the mercy of our immediate environment. But perceptual goings-on like listening and watching are also active” (2009, 173). In this paper, I do not dwell on the issue of whether listening is active because even if it is, this fact would not affect the study of this paper; on the contrary, this fact would support the agency-oriented approach. Nonetheless, the hearer-oriented approach, once it is integrated into the agency-oriented approach, must reveal something more than what it originally said, i.e. it must convey more than how it would accommodate the agential character of listening into its original account of knowledge of language.

3. According to Dummett, “the rationality of using language is sufficiently recognized … if knowledge of a language is treated as of the intermediate type” (2007, 563); the conditional indicates that “[k]nowing a language is a species of knowledge intermediate between pure practical knowledge and pure theoretical knowledge” (1993b, x).

4. For a critical overview of the debate, see Smith (1992).

5. Here, I assume that the anti-intellectualist view of knowledge is correct. Generally, intellectualism in epistemology claims that (i) propositional knowledge is the only genuine knowledge, and (ii) other kinds of alleged knowledge, such as practical knowledge and acquaintance knowledge, are either not genuine knowledge or they can be genuine knowledge only when they are defined in terms of propositional knowledge. Anti-intellectualism in epistemology contests these claims. Much of the recent debate between intellectualism and anti-intellectualism focuses on the nature of practical knowledge or knowledge-how: the intellectualist position claims that knowledge-how is a species of knowledge-that (Stanley and Williamson 2001; Stanley 2011), whereas the anti-intellectualist position denies this claim (for a weak version of this position, see Ryle 1949, and for a strong version, see Hetherington 2011, which argues that knowledge-that is a species of knowledge-how). My own defence of the anti-intellectualist view of knowledge-how can be found in Tsai (2011a, 2011b, 2014).

6. In proposing this constraint, Dummett aims to avoid behaviourism in the theories of meaning, such as Quine's linguistic behaviourism (1960, 1975).

7. The second-order “perspective” is posited to explain intentional action and can fulfil its role only if understood as functioning at the personal – not the sub-personal – level.

8. Although the notion of intentional action characterized here is commonsensical, it also enjoys theoretical and experimental support. See particularly Malle and Knobe's “The Folk Concept of Intentionality” (1997), in which they devise experiments to develop a model of the folk concept of intentional action. According to this model,

performing an action intentionally requires the presence of five components: a desire for an outcome, beliefs about an action that leads to that outcome; an intention to perform the action; skill to perform the action; and awareness of fulfilling the intention while performing the action. (Malle and Knobe 1997, 111, emphasis mine; see also Malle 2004)

The concept of intentional action characterized in the present paper refers to three components: (the first-order) ability, (the second-order) knowledge of one's own ability, and (the second-order) knowledge of one's own action. These three components roughly correspond to the skill component, the belief component, and the awareness component in Malle and Knobe's model.

9. The two distinctive conditions for intentional action – that an agent must have a perspective on either his action (i.e. knowing what he is doing) or on his ability (i.e. knowing how he does it) – are intimate in the case of linguistic action:

There is no gap between knowing what it is to speak Spanish and knowing how to do so … : you do not first learn what speaking Spanish is and then learn a means by which this feat can be executed. (Dummett 1978, 95)

10. See also Dummett (1981) for his more detailed objections to the (Chomskyan) unconscious conception of knowledge of language.

11. This version of the argument is primarily provided and adopted by Harman (1967, 76) and Devitt (2006, 92), respectively. I have elaborated this argument in greater detail in Tsai (2011a).

12. Ernest Sosa is the most prominent philosopher to advocate a two-level model of knowledge, which is now known as virtue perspectivism (see esp. Sosa 2007, 2009, 2011). My characterization of the two-level model of knowledge of language in its most basic form is inspired and influenced by Sosa's works, although with some important differences. For example, Sosa construes an epistemic agent's second-order perspective on first-order ability as propositional knowledge, whereas I leave this issue open. In particular, I demonstrate in Section 3 that this particular construal should not be supported with respect to knowledge of language.

13. The notion of “knowledge without observation” is understood by Hornsby as follows:

if [an agent] were to attend to, or to reflect upon, what she is doing, then it is something she could find herself doing, and in finding herself doing it, she would not need to make observations of the sort that a spectator might make. (2005, 121)

Applying this notion to the case of knowledge of language, Hornsby says that

[i]f the speaker did exercise procedural knowledge of how to voice her thoughts, then, even if the procedure were something of which she was not explicitly aware, she should be in a position to know of it “without observation” as she spoke. (2005, 121)

14. In fact, Stanley also knows about these caveats:

I am conscious, however, that there are many caveats and complexities in Hornsby's paper that I may have missed. For example, … [t]he subtle distinction [i.e., one must not confuse the idea that we are able to simply voice our thoughts with the idea that voicing our thoughts is something that we simply do] Hornsby has in mind here is one that has no doubt escaped me, and perhaps some of the considerations I have adduced against her position are idle once this distinction is made perspicuous. (2005, 143)

Stanley's criticisms nonetheless merit attention because they help explain why Hornsby offers these caveats for her notion of knowledge of language.

15. I suggest that knowledge of language is acquaintance knowledge initially (but only briefly) in Tsai (2010).

16. According to Hayner (1969, 426–428), prior to Russell, philosophers such as Plato, Aristotle, and Spinoza regard acquaintance as a distinctive kind of knowledge. In addition to Russell, contemporary acquaintance theorists include Fumerton (1995), Fales (1996), BonJour (2001), Chalmers (2003), and Gertler (2011, Chap. 4).

17. In this paper, I use the terms “acquaintance” and “knowledge by acquaintance (or acquaintance knowledge)” interchangeably. Some scholars might question whether acquaintance (direct consciousness or direct awareness) can be treated as a distinctive kind of knowledge (for a discussion of this issue, see Hicks et al. 1919; Hart, Hughes, and Findlay 1949). I offer two responses to this question. First, this question arises because some critics assume that propositional knowledge is the only kind of genuine knowledge. When this assumption is dropped, acquaintance can be treated as knowledge, which is consistent with the treatment of acquaintance by many great philosophers throughout history (cf. Hayner 1969). Second, Tye provides two possible answers that may be more constructive:

[W]hy should consciousness of something … yield knowledge of that thing? One partial answer is that such consciousness is undeniably epistemically enabling; via consciousness of a thing, one is put in a position to know facts about the thing. A more direct answer is that it is simply incoherent to suppose that one might be genuinely (non-inferentially) conscious of an entity and yet not know it at all. (2009, 98)

The second answer is what I intend in treating acquaintance as knowledge.

18. For his full treatment of acquaintance, see Russell (1984); for a comprehensive summary of Russell's account of acquaintance, see Miah (2006, Chap. 2). Like certain philosophers, I do not accept Russell's account of acquaintance unreservedly; regarding this point, see Tye's (2009, Section 5.1) modification of Russell's account of acquaintance.

19. Cf.:

Russell's idea of logical form as opposed to grammatical form is strikingly like Chomsky's idea of depth grammar as opposed to surface grammar. Of course Russell's forms, which we now call first-order predicate logic, look not at all like Chomsky's kernel. But it is not monstrous to propose that first-order predicate logic is the core of a deep grammar of English … A foremost advocate of such a programme is Donald Davidson. (Hacking 1975, 91)

20. Cf.: “A grammar … should … be a model of whatever it is that enables people to recognize new grammatical sentences. … [G]rammar becomes part of a theory about how we understand what is said” (Hacking 1975, 90).

21. In this paragraph, I do not intend to offer an exegesis of the notion of logical form in Russell's writings, but to identify a link between Russellian logical form and linguistic rules. Russell's concern of logical form is not on linguistics, but this absence of linguistic concern does not mean that his notion of logical form does not advance the study of language. For more on the notion of logical form (Russellian and others) in both linguistics and philosophy, see the anthology edited by Preyer and Peter (2002).