An Indian Debate on Semantic Internalism and Externalism: Śāntarakṣita’s Tattvasaṃgraha on the Meaning of Words

Abstract

The semantic internalist–externalist debate is one of the central themes in contemporary philosophy. The present paper aims to explore the dispute between Śāntarakṣita and his opponents over the meaning of words (śabdārtha) in the śabdārthaparīkṣā chapter from the perspective of the internalist–externalist debate in semantics. The present paper consists of three main parts. In the first part, I will elucidate the internalist character of Śāntarakṣita’s apoha theory and the externalist characters of his opponents’ theories. The second part will be an examination of an Indian Externalist Argument against Śāntarakṣita’s apoha presented in the chapter. In the third part, I will explore how Śāntarakṣita attempts to refute this argument by resorting to the Dharmakīrtian notion of adhyavasāya (false ascertainment). Additionally, to emphasize a contemporary approach, I will offer and highlight the intercultural interpretations of these Indian externalist and internalist arguments.

1. Introduction

The semantic internalist–externalist debate is one of the central themes in contemporary philosophy. The present paper aims to explore the dispute between Śāntarakṣita and his opponents over the meaning of words (śabdārtha) from the perspective of the internalist–externalist debate in semantics.

In the present paper, the label “internalism/externalism” refers to semantic internalism–externalism. The semantic internalist–externalist debate can be summarized with the question of whether the meaning of words, which determines the reference of words, is solely constituted by the internal factor(s) of an individual. An internalist answer is “yes” while an externalist answer is “no”. Thus, semantic internalism refers to the view that the meaning of words is wholly constituted by something that is internal to the individual. On the other hand, semantic externalism is the denial of such view, namely, that the meaning of words is at least partially constituted by something that is external to the individual.

In Tattvasaṃgraha (TS) Chapter Sixteen, “the Examination of the Meaning of Words” (śabdārthaparīkṣā), Śāntarakṣita (and Kamalaśīla, in the Tattvasaṃgrahapañjikā—TSP) offers a threefold typology of apoha-s,1 and maintains that only the first kind of apoha—the mental apoha which is by nature a reflected mental image (pratibimbaka) or a mental appearance (pratibhāsa)—is what constitutes the meaning of words. On the contrary, his opponents (the Nyāya and Mīmāṃsā philosophers) postulate that the meaning of words is constituted by the combination of individuals (vyakti), configurations (ākṛti), and real universals (jāti)2 (for the Naiyāyikas), or simply by real universals (for the Mīmāṃsakas); and that all these exist independent of the awareness of the speaker and listener. Thus, there is good reason for us to evaluate this ancient Indian debate over śabdārtha in the śabdārthaparīkṣā from the perspective of semantic internalism–externalism.

In this chapter of the TS(P), the opponent launches an argument (the Indian Externalist Argument—IEA) against Śāntarakṣita’s version of the apoha theory which targets its strongly internalist character. The gist of the argument is that the setting up of linguistic conventions (saṃketa), which is generally considered to be a prerequisite of interpersonal communication, would be impossible under Śāntarakṣita’s theory. I will compare this argument to a standard objection against internalism (the Standard Externalist Objection—SEO) offered by contemporary externalists, whose essence is that any internalist conception of meaning would render meaning completely private, and thus interpersonal communication would be impossible. Śāntarakṣita responds to the IEA by resorting to the Dharmakīrtian notion adhyavasāya (judgment/false ascertainment). The present paper will be a philosophical exegesis of these ancient Indian internalist and externalist arguments. Additionally, to emphasize a contemporary approach, I will offer and highlight the intercultural interpretations of these arguments.

2. The Indian Theories of Meaning and Their Internalist–Externalist Characters

2.1. The Nyāya and Mīmāṃsā Theories of Meaning and Their Externalist Characters

The main opponents of Śāntarakṣita (and Kamalaśīla)3 in śabdārthaparīkṣā are not difficult to identify. Considering that many of the arguments against apoha presented in this chapter are quotations from Uddyotakara’s4 Nyāyavārttika5 (NV) or the Apohavāda chapter of Kumārila’s6 Ślokavārttika7 (ŚV), it is safe to say that Śāntarakṣita’s major interlocutors are the Nyāya and Mīmāṃsā philosophers. In this section (Section 2.1), I shall briefly introduce their theories of meaning and explain the sense in which they are externalists.

According to the Nyāya school, the meaning of words is constituted collectively by the individual (vyakti), configuration (ākṛti), and real universal (jāti).8 For instance, the meaning of the word “cow” is constituted by the individual cow, the form of a cow, as well as the universal “cowness”. On the other side, for the Mīmāṃsā School, the meaning of words is constituted by real universals, as stated at the beginning of the Ākṛtivāda:

If the meaning (artha) [of words] were different from the real universal9 (ākṛti), then the relation (saṃbandha) [between words and their meanings], as well as its eternality (nityatā), would not be established. Having realized this, it (real universal) is here said to be what is expressed (vācya).10

(Ākṛtivāda 1)

It is said that “ākṛti” simply means “jāti”. Individuals (piṇḍa) are characterized by it (= ākṛti/jāti). And the real universal (sāmānya) of individuals is the basis of the unitary cognition (ekabuddhinibandhana).11

(Ākṛtivāda 3)

Accordingly, ākṛti is just another name for jāti. In fact, for Kumārila, ākṛti, jāti, and sāmānya are synonymous, meaning “real universal”. Mīmāṃsā philosophers maintain that this real universal, which is somehow perceptible by the senses but at the same time also cognition-independent (its existence does not depend on cognition), is the meaning of words— it determines the reference of words. For instance, the meaning of the word “cow” is the real universal “cowness;” and the word “cow” has cows as its referents because “cowness” is inhered in them.

As mentioned before (Section 1), externalism refers to the view that the meaning of words is (at least partially) constituted by something that is external to the individual. I think the sense in which the Nyāya and Mīmāṃsā philosophers are externalists is rather clear. Taking the word “cow” (go) as an example, for the Naiyāyikas, its meaning is constituted by [1] the vyakti, which is the physical body (mūrti) of the cow that is the substratum of the specific qualities like heaviness and firmness; [2] the ākṛti that makes known the jāti, “cowness”, and its marks (liṅga), such as the arrangements of parts; and [3] the jāti that produces the unitary cognition “this is cow”.12 Additionally, for the Mīmāṃsakas, it is constituted by [4] the sāmānya (=ākṛṭi, jāti) that characterizes cows and makes them equally cognized as “a cow” (see Ākṛtivāda 3, translated above). Suffice it to say, all of the above ([1]–[4]) have a separate and independent existence from our minds and cognitions. Therefore, the main opponents in śabdārthaparīkṣā should be regarded as externalists, because according to their theories, the meaning of words is at least partially constituted by an external factor(s). In fact, they should be regarded as extreme externalists (but not moderate externalists) because they maintain that this external and non-mental factor(s) is exactly what constitutes the meaning of words and delimits the word reference.

2.2. Śāntarakṣita’s Theory of Meaning and Its Internalist Character

Unlike the Hindu philosophers, the logico-epistemological tradition of Buddhism (the Pramāṇavāda school) does not accept the existence of real universals. Therefore, Dignāga, the founder of the Pramāṇavāda school, offers the apoha theory to explain how the denotation of things by words can take place, without ontological commitment to real universals. Now, let us set aside Dignāga’s own formulation of the apoha theory and focus on Śāntarakṣita’s (and Kamalaśīla’s) version of apoha.

Dignāga’s apoha theory incurred criticism from Hindu philosophers such as Uddyotakara (in NV) and Kumārila (in ŚV). To respond to these non-Buddhist critics, Śāntarakṣita developed his own version of the apoha theory.13 The salient point of this apoha theory is that there are three different kinds of apoha-s:

In fact, there are two kinds of apoha-s based on [the distinction between] implicative negation (paryudāsa) and non-implicative negation (niṣedha). There are also two kinds of implicative negations based on the distinction (bhedatas) between [the apoha which has] the nature of cognition (buddhyātman) and [the apoha which has] the nature of [external] object (arthātman).14

(TS 1003)

Accordingly, there are altogether three kinds of apoha-s, namely, [1] the apoha which has the nature of cognition, [2] the apoha which has the nature of object, and [3] the apoha which is an absolute negation.15 For the sake of brevity, let us call them the mental apoha, the object apoha, and the absolute apoha. In this paper, I shall mainly focus on the mental apoha, because it is described by Śāntarakṣita and Kamalaśīla as the only meaning of words (śabdārtha).16 In TS 1004–1005, Śāntarakṣita explains what a mental apoha is17:

Similar objects such as an Abhayā,18 which are distinct from the others by their very nature (prakṛtyā eva), have been previously said [in the Sāmānyaparīkṣā] to be the causes of the unitary judgment (ekapratyavamarśasya). Based on them (the similar objects), a reflected mental image of the object (arthapratibimbakam) appears in a conceptual awareness. Even though [the reflected mental image] does not have the nature of the [external] object, it is [mistakenly] judged just as the [external] object.19

(TS 1004–1005)

Following Dharmakīrti’s line of argument in Pramāṇavārttika (PV) 1.73–7420, Śāntarakṣita explains in the Sāmānyaparīkṣā of the TS that even though things are different from one another, still, they can have the same potency (śakti). For instance, some plants, despite being different from one another, can have the same potency of removing diseases.21 Śāntarakṣita here maintains that likewise, based on the similar but different objects which produce unitary judgment, a reflected mental image of the object, namely a mental apoha, appears in the verbal awareness. More importantly, even though the reflected mental image (the mental apoha) itself is not an external object,22 it is falsely ascertained by ordinary people as an external object. Kamalaśīla elaborates as follows:

“Based on them” means based on the similar objects such as an Abhayā, namely, [these things] having been made into a cause, due to the power of experiencing them, a verbal awareness is produced. There is a reflected mental image of the object (arthapratibimbakam)—viz., an appearance of the object (arthābhāsaḥ), being manifested in that [verbal awareness] as identical with the object.23

(TSP 1004–1005)

Accordingly, arthapratibimbaka and arthābhāsa are synonymous24, both referring to the mental apoha. This arthapratibimbaka or arthābhāsa is internal to the awareness of the individual, but it manifests as external—viz., as identical with the external object. I will explain more about this mechanism when we come to discuss the notion adhyavasāya (false ascertainment) in Section 4.2.

A few śloka-s later, Śāntarakṣita proclaims that the first kind of apoha, namely, the mental apoha (arthapratibimbaka/arthābhāsa) is what is conveyed by words.25 Kamalaśīla explains as follows:

“The first” refers to [the apoha whose] nature is a reflected mental image of an [external] object mentioned [in TS 1004–1005]. [To indicate] the reason with respect to it (it = the first exclusion), [namely, why it is said to be what is conveyed by words,] [Śāntarakṣita] explains beginning with “bāhyārthādhyavasāyinyāḥ”. To explain, only (eva) that which manifests in a verbal awareness can be the meaning of words (śabdārthaḥ). But there is no false ascertainment (adhyavasāyaḥ) of non-implicative negation in that [verbal awareness], and there is no appearance of unique particular (svalakṣaṇapratibhāsaḥ) [in that verbal awareness] as there is in sensory awareness. But rather, only a verbal awareness, which consists in falsely ascertaining [a reflected mental image] as an external object, arises. Therefore, only (eva) the reflected mental image of the object which appears directly (sākṣāt) in verbal awareness and appears as identical with it (the object), and nothing else (nānyaḥ), can be the meaning of words. This is the intended meaning.26

(TSP 1010)

There are two points in this commentary. First, the mental apoha is the internal factor of an individual—it is internal to awareness but falsely ascertained as something external. Second, the meaning of words is constituted by the mental apoha and nothing else. In TSP 1013–1014, Kamalaśīla emphasizes this point by asserting that the meaning of words is not even constituted by the other two kinds of apoha-s—they are only metaphorically said to be (upacaryate) the śabdārtha.27 Now, we can summarize the following two features of Śāntarakṣita’s apoha theory:

• A mental apoha is internal to awareness.
• Only the mental apoha is the meaning of words.

Recall that internalism refers to the view that the meaning of words is constituted only by something that is internal to the individual. Thus, the internalist character of Śāntarakṣita’s apoha theory is well reflected by the above two features (A and B): the meaning of words (śabdārtha) is only constituted by the mental apoha that is internal to the awareness of the individual (but is falsely ascertained as something external). More important is Kamalaśīla’s explicit assertion that what makes the mental apoha qualified to be the meaning of words is that only it manifests in verbal awareness. The same principle is also articulated in TS 879ab: which does not manifest itself when the verbal awareness produced by a word takes place, that cannot be the meaning of that word.28 From these, it is evident that Śāntarakṣita’s and Kamalaśīla’s version of the apoha theory should be counted as a semantic internalist theory of meaning.

3. The Indian and Contemporary Critiques of Internalism

3.1. The Indian Externalist Argument in the Śabdārthaparīkṣā

In the śabdārthaparīkṣā of the TS, the opponent launches an argument against Śāntarakṣita’s apoha theory which targets its internalist characters (features A and B). In this argument, the opponent argues that the setting up of linguistic conventions (saṃketa) would be impossible under Śāntarakṣita’s theory:

How can [the setting up of] a linguistic convention be possible under [your] apoha theory? And how can it (linguistic convention) be fruitful (sāphalyam)?29 To explain, it (=a mental apoha) is not established [as the same] for both the speaker and the listener, because awareness (jñānam) is not made known to one another (parasparam).30

(TS 1207–1208ab)

As explored in Section 2.2, for Śāntarakṣita, the meaning of words is constituted only by the mental apoha (the reflected mental image). The opponent here argues that such a theory would render linguistic conventions impossible. Kamalaśīla elaborates in the commentary:

Why would linguistic conventions be impossible under it (Śāntarakṣita’s apoha theory)? To explain, [the opponent] said “na dvayoḥ sa siddhyati.” The word “hi” indicates the reason, namely, because the apoha which is by nature a reflected mental image (pratibimbātmako ’pohaḥ) cannot be established as an object of linguistic convention (saṃketaviṣayayoḥ) that is the same for both the speaker and the listener.

Why? To explain, [the opponent] said “na hi jñānam” and so forth. Ordinary people (arvāgdarśanānām) can only perceive their own awareness. This is because ordinary people (aparadarśanaḥ) cannot know the awareness of the others; and ultimately, the apoha which is by nature a reflected mental image is not different from the awareness (jñānād avyatiriktaḥ). And therefore, since there is no establishment of an object of linguistic convention that is the same for both the speaker and the listener, how could a linguistic convention be set up or apprehended?31

(TSP 1207–1208)

I will offer an intercultural interpretation of this argument in the next section (Section 3.2). Now, let us break it down and see how it assumes the two abovementioned features of Śāntarakṣita’s apoha theory:

• Premise 1: A mental apoha is not different from awareness.
• Premise 2: Only the mental apoha is the meaning of words.
• Premise 3: Ordinary people cannot know the awareness of other people.
• Premise 4: The speaker’s meaning of words cannot be known to the listener.
  (From Premises 1, 2, and 3)
• Premise 5: If the meaning of words is not the same for both the speaker and the listener, then the setting up of linguistic conventions would be impossible.
• Conclusion: The setting up of linguistic conventions would be impossible (under Śāntarakṣita’s apoha theory) (from Premises 4 and 5)

One can readily see that Premises 1 and 2 correspond to features A and B of Śāntarakṣita’s theory. Thus, suffice it to say that this argument targets the strongly internalist character of Śāntarakṣita’s apoha theory. Henceforth, let us refer to this argument as “the Indian Externalist Argument” (IEA) and the opponent as “the Indian externalist”.

3.2. Intercultural Interpretation of the Indian Externalist Argument

To emphasize a contemporary approach to the IEA, let us compare it to an objection against internalism offered by contemporary externalists. Riemer (2016, p. 37) discusses a common objection (which I call “the Standard Externalist Objection”—SEO) to an internalist conception of meaning. According to the SEO, interpersonal communication is considered as the transfer of something (e.g., thoughts or concepts) that is the same for the speaker and the listener, from the speaker to the listener. Thus, interpersonal communication has a prerequisite, namely, the speaker and the listener have a joint grasp of content. However, any internalist conceptions of meaning would render meaning completely private and hence make the shared content among the speaker and the listener impossible. Therefore, any internalist theories of meaning would preclude the possibility of interpersonal communication.

To compare the IEA and the SEO, I shall introduce Pagin’s (2000, p. 164) thesis of basic publicness (BP), which is stated as follows:

What a speaker means by his words can be known by others.

(BP) (Emphasis in original)

As mentioned by Pagin (Ibid.), what BP denies is the possibility that an individual’s meaning of words is something private and inaccessible to another individual(s).32

In what follows, I will compare the IEA to the SEO. The aim is to offer an intercultural interpretation of the IEA and examine it from the perspective of contemporary philosophy. Before that, it is worth noting that while the SEO is a general objection against any internalist conceptions of meaning, the IEA targets a specific version of the internalist theory, namely, Śāntarakṣita’s and Kamalaśīla’s apoha theory, according to which the meaning of words is solely constituted by reflected mental images. However, this difference may not be as significant as it seems because, as already suggested by Riemer (2016, p. 37), the internalists (the targets of the SEO) usually consider mental entities as the meaning of words. This being the case, it is safe to say that both the IEA and the SEO are attempts to undermine the very idea that the meaning of words is constituted by something mental in nature.

The first comparison is that both the SEO and the IEA attempt to critique internalist theories by showing that any internalist conceptions of meaning would lead to the violation of BP. To appreciate how this might work, let us explore the understanding of the mind or mental content in both traditions. In Western philosophy, the mind is often understood as something private and its contents are only accessible to oneself. This understanding of mind is perhaps derived from Descartes, and Dennett (1991) termed it as “the Cartesian theater”. Accordingly, the mind is a private and inner stage for an audience of one, and only one—every individual can only access what is going on in his/her own mind, but not that in others’ minds. In other words, one may guess or infer (e.g., based on observation) the contents of other’s minds, but can never directly and authoritatively know them in the way (s)he knows the contents of his/her own mind (Kim 2018, pp. 63–64). It appears that in the SEO, the contemporary externalist assumes this Cartesian understanding of the mind and thereby argues that an internalist conception of meaning—according to which the meaning of words is constituted by mental entities—would lead to the violation of BP. Indian philosophers have a similar understanding of the mind. As described in the Abhidharmakośabhāṣya (AKBh), the cognition of other minds (paracittajñāna or paracittavida/cetoparyāyābhijñāna) is one of the six supranormal powers or abilities (abhijñā) that ordinary people do not possess; and, as suggested by its name, it has the minds and mental states of other people as objects (paracittacaittaviṣayatva).33 In other words, one does not have any knowledge of other’s minds, not until one acquires the supranormal power of paracittajñāna. Therefore, in the IEA, the Indian externalist asserts that “ordinary people (arvāgdarśana/aparadarśana) cannot know the awareness of another people” (Premise 3). With this premise, (s)he then shows that Śāntarakṣita’s theory, which has features A and B (Premises 1 and 2), would entail Premise 4 (namely, the speaker’s meaning of words cannot be known to the listener), and this is exactly what is denied by BP.

Second, both the SEO and the IEA argue that if the meaning of words were something private and inaccessible to others (=~BP), then communication between the speaker and the listener would be impossible. According to the SEO, this is because interpersonal communication has the shared content among the speaker and the listener as its prerequisite. For contemporary externalists, this may not require further justification or elaboration because it represents the received wisdom among Western philosophers, including towering figures such as Wittgenstein and Davidson, that linguistic meaning must be public. On the other hand, from the perspective of the ancient Indian philosophers, this may require further explanation. Thus, in the IEA, the Indian externalist attempts to explain this by resorting to the notion of saṃketa (linguistic convention). Saṃketa, a synonym of samaya, refers to that which establishes the connection (sambandha) between words and their meanings. The setting up of saṃketa-s is considered to be a prerequisite for communication between the speaker and the listener because, as stated in the Vaiśeṣikasūtra (VS), it (saṃketa/samaya) is what allows the listener’s understanding of meaning from the speaker’s word.34 In the IEA, the Indian externalist argues that if the meaning of words were completely private as implied by Śāntarakṣita’s apoha theory, then the setting up of saṃketa-s would be impossible because the meaning, i.e., the viṣaya of saṃketa, as being something private, cannot be the same (eka) for both the speaker and the listener; and without the setting up of this very saṃketa, the listener’s understanding of meaning (arthapratyaya) from the speaker’s word would be impossible.

4. Internalist Responses to the Critiques

4.1. Contemporary and Indian Internalist Arguments: From Type-Identity and Adhyavasāya

As introduced by Riemer (2016, p. 37), contemporary internalists refute the SEO by resorting to the notion of “type-identity”. Type-identity is a kind of identity that is between two instances of the same category. For example, the kidney of Individual_A is type-identical with that of Individual_B, in virtue of the fact that both are instances of the same category, namely, “KIDNEY”. According to Riemer (Ibid.), the SEO is only a pseudo-critique because contemporary internalists can easily dismiss it by holding that there can be a type-identity between the mental contents of the speaker and the listener, which (= mental contents) is proclaimed by the internalists to be the meaning of words. Thus, an internalist conception of meaning does not imply that the meaning of the speaker’s words is unknown to the listener because the possibility that there is a type-identity between the speaker’s and the listener’s mental contents (viz., the meaning of words) cannot be eliminated. Henceforth, let us call this argument the argument from type-identity.

Recall that in the SEO, the externalist attempts to show that internalism would lead to the violation of BP. Thus, here in the argument from type-identity, the internalist’s strategy is to demonstrate that an internalist conception of meaning does not imply the violation of BP: even though the mental contents of the speaker are internal factors, if they are type-identical to those of the listener, then the listener can still somehow know them. In such a way, using Riemer’s wording (Ibid.), there is the possibility that meanings are “private” but at the same time also “shared”—knowable to one another.

On the Indian side, Śāntarakṣita replies to the IEA in TS 1209–1210:

Even though one [only] cognizes one’s own appearance (svasya svasyāvabhāsasya), still, that [linguistic convention] can take place, because the false ascertainment (adhyavasāye) [of the mental appearance] as an external object is common for both [the speaker and the listener]. To explain, just as the one whose eyes are afflicted with an eye disease (timira) says “there are two moons” to someone like him/her, all verbal usages are considered likewise.35

(TS 1209–1210)

Śāntarakṣita here attempts to refute the IEA by employing the example of eye disease. Accordingly, two people with cataracts can communicate with each other about the second moon despite that it is just “the beetle in the box”: (i) the second moon only appears to one’s own and (ii) it is possible for the speaker and the listener to have different second moons in their own awareness (“boxes”). Kamalaśīla elaborates in the commentary as follows:

Even though ultimately, the speaker and the listener [can only] perceive (saṃvedanam) their own appearances (svasya svayaivāvabhāsasya), but still, just as two people with an eye disease, because the source of error (bhrāntibījasya) is common to both, the speaker and the listener can equally falsely ascertain [their own mental apoha-s/reflected mental images] as an external object. The speaker would have the erroneous conception (abhimānaḥ) “that person (viz., the listener) understands the object which I understand”. This is also the case for the listener.

[Objection:] How do the speaker and the listener mutually apprehend that they have the ascertainment about the same object?

[Reply:] Even though it is not ultimately apprehended,36 but still, because the source of error is common [to both the speaker and the listener], this very erroneous verbal expression—which ultimately conforms to their own mental appearances—takes place. This is our presented view. Therefore, on account of falsely ascertaining [their own appearances as] the same object, the setting up of linguistic conventions (saṃketakaraṇam) is in fact possible.37

(TSP 1209)

Again, an intercultural interpretation of this argument (argument from adhyavasāya) will be offered in next section (Section 4.2). Here, it is worth noting that while Śāntarakṣita simply employs the case of timira as a simile, Kamalaśīla explicitly mentions that it is the source (bīja) of the erroneous cognition of two moons. In other words, Kamalaśīla interprets this as an account of the source of the error, namely, the ground or basis of people’s falsely ascertaining of something internal as external.

4.2. An Intercultural Interpretation of the Argument from Adhyavasāya

As mentioned in Section 4.1, the argument from type-identity refutes the SEO by showing that an internalist conception of meaning does not imply the violation of BP—it does not entail that the meaning of the speaker’s words would be unknown to the listener because one cannot eliminate the very possibility that their mental contents are type-identical. Here, in the argument from adhyavasāya, the Indian internalists (Śāntarakṣita and Kamalaśīla) adopt the same approach: even though people can only perceive their own mental appearances (svasya svasyaivāvabhāsasya … saṃvedanam), still, our apoha theory—according to which the meaning of words is constituted only by mental apoha-s (reflected mental images)—does not imply that the speaker’s meaning of words would not be known to the listener (Premise 4 of the IEA). Unlike the contemporary internalists, however, Indian internalists cannot explain this in terms of type-identity, because, as I will illustrate shortly, this would violate the ontological doctrine of the Pramāṇavādins. So instead, they explain this by resorting to the Dharmakīrtian notion of “adhyavasāya” (false ascertainment). To emphasize a contemporary approach, in this section (Section 4.2), I will (i) introduce the main usage of the notion “adhyavasāya” in the Pramāṇavāda tradition and explain how it may help to undermine the IEA; (ii) discuss why the argument from adhyavasāya might not be a satisfactory response to the IEA, considered from the perspective of contemporary philosophy; and (iii) elucidate why the Indian internalists could not appeal to the notion of “type-identity” as contemporary internalists do.

There are two main usages of the notion adhyavasāya in Dharmakīrti’s Svavṛtti (PVSV). Let us see two of its occurrences in PVSV Chapter 1:

In that [conceptual awareness], the image of an object manifests (i) as external, (ii) as unitary, and (iii) as accomplishing that [effect], despite that it does not accomplish any effects. This is because [it] is falsely ascertained to be so by language users.38

(PVSV 1.75d)

And according to the apoha theorists, verbal awareness does not grasp what is real: under delusion, it operates [by] falsely ascertaining something as causally efficacious even though that thing is causally inefficacious.39

(PVSV 1.97)

First, in both sources, adhyavasāya refers to an awareness’s (or a cognition’s) disposition to falsely judge something without causal efficacy (arthakriyā) as causally efficacious. Apart from this usage, in the source PVSV 1.75d, it refers to the disposition to falsely judge something internal (e.g., something mental in nature) as external. This second usage of adhyavasāya is emphasized by Śāntarakṣita and Kamalaśīla and repeatedly occurs in the śabdārthaparīkṣā. Here are two examples:

The verbal cognition, by which an external reality (bāhyatattva) is not cognized, falsely ascertains its own image as something external due to natural delusion (prakṛtyopaplavāt).40

(TS 1065)

The exclusion that is by character a reflected mental image is ascertained as an external individual (bāhyavyaktirūpatvena) by the mistaken people. Because of [this] false ascertainment, there would be the connection [of the apoha] with gender, number and so forth through the individual.41

(TSP 1142)

In the argument from adhyavasāya (TS(P) 1209–1210), the Indian internalists aim to refute the IEA by showing that their apoha theory does not imply that the speaker’s meaning of words would be unknown to the lister. The argument is as follows: it is true that (i) the meaning of words is only constituted by an individual’s mental contents; and that (ii) the speaker and the listener cannot directly perceive each other’s mental contents, and this is exactly where the notion adhyavasāya comes into play—on account of sharing the same source of error, the speaker and the listener falsely ascertain their own mental contents (e.g., the mental images of a cow) as the same external object (e.g., an individual cow), just as the eye disease cataract allows two people to falsely ascertain their own images of the “second moon” as the same external second moon. This false ascertainment thus allows the speaker and the listener to someone know each other’s mental contents, because both mental contents, despite being different, are falsely represented as the same external object. In other words, even though it is true that the “beetle” in the speaker’s “box” is not directly perceivable by the listener and vice versa, in virtue of falsely ascertaining the different beetles in boxes as the same beetle that is out of a box, the speaker and the listener can somehow access and know the beetle in each other’s box.

One possible objection to this argument is that the Indian internalists are obliged to offer an adequate account of how different people (the speaker and the listener) with different mental contents can equally falsely ascertain their mental contents as one and the same external object. Appealing to their sharing of the same source of error only defers the problem but does not really solve it. For, one can further question how different people can come to have this very source of error. While this objection seems tenable, it might not be applicable in the context of the Indian Buddhist tradition. As suggested in TS 1065, for Śāntarakṣita and Kamalaśīla, the source of error which brings about the false ascertainment is natural delusion, that is, ignorance (avidyā). Since ignorance is understood to be beginningless (anādi) in the Buddhist tradition, the Indian internalists can comfortably reply that the origin of this source of error (namely, ignorance) requires no explanation or justification.42 Thus, despite the fact that resorting to the sharing of the same source of error may not be considered a satisfactory explanation from the perspective of philosophy, nevertheless, it could be an adequate explanation in the context of Indian Buddhism.

Finally, I will briefly explain why Śāntarakṣita and Kamalaśīla could not resort to notions like “type-identity” to ward off the externalist argument, as contemporary internalists do. In TS 1006a,43 in explaining why reflected mental image (mental apoha) is called “apoha”, Śāntarakṣita points out that every single one is different from the others. In other words, reflected mental images are by nature unique particular, and this very point is later emphasized by Dharmottara to argue why they cannot be the meaning of words.44 The Sanskrit equivalent of “identity” is “ekatva”, which is usually translated as “sameness” or “oneness”. According to the Pramāṇavāda ontology, there cannot be any genuine sameness (ekatva) between any two unique particulars, as every unique particular is excluded (vyāvṛtta) from both its similars (sajātīya) as well as its dissimilars (vijātīya); all “sameness” among two unique particulars is in fact only forms that are erroneously superimposed (āropita) on the unique particulars. Thus, the Indian internalists could not accept the possibility that the reflected mental images of the speaker are type-identical with those of the listener, as this ontological commitment to identity among unique particulars would violate the radical nominalist view of the Pramāṇavāda School.

5. Conclusions

Let me summarize the main points of this paper as follows:

(1)

Śāntarakṣita’s (and Kamalaśīla’s) version of the apoha theory should be counted as a semantic internalist theory of meaning. The theories of his interlocutors (Uddyotakara and Kumārila), on the contrary, should be regarded as semantic externalist theories.

(2)

The arguments in TS(P) 1207–1208 and TS(P) 1209–1210 represent an externalist–internalist debate on the meaning of words in the history of Indian philosophy.

(3)

Both the (ancient) Indian and contemporary externalists attempt to undermine internalist theories by showing that any internalist conceptions of meaning would lead to the absurd consequence that the speaker’s meaning of words cannot be known to the listener. Behind this line of argument is the conception of mind that is shared by both traditions, namely, that the mind is a private stage that “no one else is permitted a look”. (Kim 2018, p. 63)

(4)

By resorting to the notions “adhyavasāya” and “type-identity”, respectively, the Indian and contemporary internalists attempt to ward off the externalist critiques by showing that an internalist conception of meaning does not necessarily imply that the speaker’s meaning of words cannot be known to the listener.

In contemporary scholarship, Buddhist–Hindu debates on apoha are generally philosophically explored from the perspective of antirealism/realism about universals.45 I have demonstrated in this paper that it may also be fruitful to evaluate these debates from the perspective of semantic internalism–externalism. By emphasizing and highlighting the intercultural interpretation of the Indian externalist and internalist arguments, I illustrated how ancient Indian notions and arguments can be understood from the perspective of contemporary philosophy. A complete evaluation of the soundness of these arguments is, however, beyond the scope of this paper. Additionally, it would be worth pursuing an examination of other versions of the apoha theory in the post-Dharmakīrtian period, such as those of Dharmottara and Karṇakagomin, from the perspective of semantic internalism.