Orsborn on Tzohar, 'A Yogācāra Buddhist Theory of Metaphor'

Roy Tzohar
Matthew Orsborn

Roy Tzohar. A Yogācāra Buddhist Theory of Metaphor. Oxford: Oxford University Press, 2018. 296 pp. $99.00 (cloth), ISBN 978-0-19-066439-8.

Reviewed by Matthew Orsborn (Mahidol University) Published on H-Buddhism (February, 2019) Commissioned by Rafal Stepien (Heidelberg University)

Printable Version: http://www.h-net.org/reviews/showpdf.php?id=53031

Roy Tzohar’s A Yogācāra Buddhist Theory of Metaphor is a reworking of his 2011 Columbia University PhD dissertation, “Metaphor (Upacāra) in Early Yogācāra Thought and Its Intellectual Context.”[1] The chapter and subchapter structure of the two are nearly identical, though the monograph under review has been improved in terms of the quality of its writing and clearer focus on the key arguments and conclusions. Not long after receiving my copy for review (thank you H-Buddhism and Oxford University Press), A Yogācāra Buddhist Theory of Metaphor was awarded the 2018 Toshihide Numata Book Award in Buddhism.[2] Tzohar is currently an associate professor in the East and South Asian Studies Department at Tel Aviv University, where he also teaches in the Philosophy Department.

I would like to begin the review with a brief outline and summary of the work, which is composed of an introduction, six main chapters, and a conclusion. I shall use direct quotations where possible, rather than paraphrasing, in order to, as we say, let the author speak for himself.

The introduction begins by pointing out the “paradox” at “the heart of Buddhist philosophical thought,” namely, that “language is ... part of the disease” of human suffering, as well as “a means for representing, describing, or reaching reality.” Tzohar’s diachronic approach “focuses on the ingenious response” to this paradox that the Yogācāra school “proposed through its sweeping claim that all language use is in fact metaphorical (upacāra)” (p. 1). Part 1 (chapters 1 and 2) provides a background of pre-Yogācāra Indian philosophy of language in general and of upacāra and notions of figurative language in particular. Chapter 1, “Metaphor as Absence: The Case of the Early Nyāya and Mīmāṃsā,” covers these two non-Buddhist schools. The Indian concept of upacāra as later found within Sanskrit poetic theory is contrasted against Western post-Aristotelian discourse on metaphor, so that “the emphasis in Indian discourse is not on the changed meaning of a word, but on the difference in the referential relations between a word and its referent” (p. 24). “The Mīmāṃsā’s discussion of the denotation of words is firmly situated within the context of a common adherence to strictly referential theories of meaning” (p. 34). The Nyāya were led to a position wherein nouns refer not to individual referents but to generic properties of things, the former being merely figurative. The notion that nouns for individual referents are figurative is explained as being the same as when an associated term stands metaphorically for a second term.

In chapter 2, “Metaphor as Perceptual Illusion, Figurative Meaning in Bhartṛhari’s Vākyapadīya,” Tzohar examines the 2nd and 3rd ṇḍas of Bhartṛhari’s Treatise on Sentences and Words and its commentaries. This text lays out a theory of language important for virtually all schools of Indian thought, and is “one of the most philosophically innovative treatments of figurative meanings” of the time (p. 42). It is a mark of his thoroughness that Tzohar takes the time to work through the various ways that this difficult text has been read in prior scholarship. In the 2nd ṇḍa, against “Nyāya (realist) semantic externalism and the Buddhist (antirealist) radical conventionalism” as typified by Nāgārjuna, Bhartṛhari’s “mentalist” position “understands meaning to be whatever cognitive content arises by a word’s denotative power, regardless of the ontological status of the referent” (p. 62). This novel approach marks a shift from understanding figurative language as an epistemic, rather than ontological, matter, in its maintenance of the “correspondence relation between language and phenomena” (p. 63). However, it is well noted that this itself threatens the distinctions between primary and secondary meaning in metaphorical and figurative language, where in realistic terms the primary term corresponds to the actual object, against the secondary which does not (and is hence a “perceptual illusion”) (p. 62). In the 3rd ṇḍa, in the face of the assumption of enduring relationships between words and referents, Bhartṛhari indicates the meaninglessness of secondary or figurative expression, due to the lack of an existent object. His proposed solution is that “when something is designated, the things denoted by words have another, Secondary Existence (sattaupacārikī); it shows the individual form of all things in all places” (p. 66, from Vākyapadīya 3.3.39). Tzohar interprets this universalized figurative nature of language by reading the 3rd ṇḍin the light of the 2nd ṇḍa. The “secondary/figurative existence of referents of all words ... allows them to stand in a relation with” referents that do not presently exist, including past or future objects (p. 69). This “radical critique of the realist correspondence theory of meaning ... emphasizes the hermetic nature of language and its inability to truly reach reality” (p. 70). This theory of language has greater implications for philosophy in general, and while not a full-blown idealism, it would prove critical for developments in the subsequent Yogācāra tradition.

This brings us to part 2, the Buddhist Yogācāra context. Here, chapter 3, “It’s a Bear ... No, It’s a Man ... No, It’s a Metaphor! Asaṅga on the Proliferation of Figures,” focuses on a key early Yogācāra work by Asaṅga, the Tattvārthapaala in the Bodhisattvabhūmi of the Yogācārabhūmi śāstra, along with corresponding commentarial sections in the Viniścaya. Again, Tzohar takes the trouble to discuss the dating and authorship of the text and to critique earlier scholarship. Interpretations by Janice Willis and Hartmut Beuscher, in particular, are respectively marred by anachronistic applications of later Yogācāra hermeneutic devices and a limited historical approach. “Asaṅga’s engagement with upacāra appears ... in a set of arguments designed to demonstrate through reasoning the inexpressibility of the true, essential nature of reality (svabhāvatā)” (p. 84). Asaṅga critiques the monosemic word/referent relationship “by introducing the possibility of a polysemy of figures—a circumstance in which various metaphors (upacāra) denote the same object” (p. 85). Three arguments found in the Tattvārthapaala are analyzed. The first argument is from polysemy, that is, because multiple words may be used for a single thing, there can be no one-to-one correspondence between word and thing; the second is that “an essential nature is not apprehended or determined by the designation”; and the third is that “an essential nature is not apprehended or determined by the object” (pp. 95, 99, 100). Reading the Tattvārthapaala through the Viniścaya, Tzohar interprets it as “advancing an argument about the limits of language.” Here, “the lack of explanation in the [Tattvārthapaala] for the connection between upacāras and vastu” uncovers “the absence of any essential ties between words and their referents, emphasizing the self-referential nature of language.” While the treatment of upacāras here is “paradigmatic of all designations,” it is not yet the “full-blown, pan-metaphorical claim presented later by Sthiramati in his Triśikābhāya” (p. 124).

Chapter 4, “The Seeds of the Pan-Figurative View: Metaphor in Other Buddhist Sources,” then turns to Vasubandhu’s Abhidharmakośabhāya and Sthiramati’s commentary thereon, the Lakāvatāra Sūtra, and Dignāga’s Pramāasamuccaya. The aim here is “to explore the ways in which these sources possibly influenced Sthiramati’s later full-fledged theory of metaphor, and thus to draw a clearer distinction between the elements of his thought that are grounded in the work of predecessors and those that are innovative” (p. 125). In the Abhidharmakośa and commentaries, “the nonexistence of objectified phenomena, viewed in terms of the referents’ absence from their locus of reference, is taken to imply the pan-figurative nature of all language usage.” However, it is only in Sthiramati’s Triśikābhāsya that the relationship between “upacāras, the nonexistence of objectified phenomena, and their underlying causal reality” are “explicitly tied together theoretically” (p. 136). The Lakāvatāra deals with upacāra mainly in the Sagāthaka, often considered a later addition to the text. Here, as “the primary referents of these expressions [of “self” and “sense-faculties,” etc.] do not exist (as they appear), they are held to refer indirectly to the mental reality that ... brings them about.” This text thus “presents an understanding of upacāra that is broader” than previous accounts. “No longer viewed merely in hermeneutical terms as indicating an implied meaning, or in referential terms as the absence of the primary referent from the locus of experience, upacāra is conjoined with a causal phenomenal description of reality.” It is “a linguistic sign that indicates” both the absence of “ontological existence” and a real referent, and the indirect “presence of a causal deep-structure” (p. 143). The last text for this chapter, Dignāga’s Pramāasamuccaya, largely uses the term “upacāra” as a hermeneutic device to explain away doctrinal inconsistency, and only in one case does it play a philosophical role. While arguing against the idea that “general terms are expressive of instantiations of universals,” Dignāga employs “a particular understanding of figurative use; namely, as the absence of the primary referent from the locus of reference and as requiring qualitative similarity between the primary and secondary referent” (p. 148). Later, Sthiramati would make a similar claim.

Part 3 (chapters 5 and 6) and the conclusion combine together the various threads from the previous chapters to examine Sthiramati’s pan-figurative position. Chapter 5, “What It All Comes Down To: Sthiramati’s Pan-Metaphorical Claim and Its Implications,” focuses on Sthiramati’s presentation of metaphor in his commentary on the Triśikā. Tzohar first provides Sthiramati’s definition in his commentary: “The metaphor (upacāra) of ‘things’ (dharmā) and of ‘self’ in its various forms, which is set in motion, that is to say in the world and in treatises, that is [with reference to] the transformation of consciousness” (p. 157, insertion original). With “things” and “self” covering all phenomena, “they are figurative because their primary referents ... are absent from their locus of reference” (p. 160). By defining metaphor (upacāra) purely in terms of “its underlying referential mechanism and without regard to any of the discursive and pragmatic conditions,” Tzohar astutely notes that this would seem “to undermine the ordinary-language distinction between the literal and figurative use of words” (p. 161). It “does not necessarily entail the complete absence of word-referent correspondence,” however (p. 166). While this “radical change in the status of language” based on the transformation of consciousness thus refers to “imagined entities and is hence ... incapable of truly representing reality,” “it changes nothing in our understanding of language as function of its use.” Therefore, “ordinary-language metaphors become something like second-order metaphors, the visible tip of a more fundamental “submerged” metaphorical structure.” This view “amounts to a figurative theory of sense more than of reference” (p. 167). This Yogācāra position provides two benefits: first, “it creates a kind of philosophical meta-language that ‘cleanses’ ordinary discourse of its essentialist and reifying quality by casting it in terms of mental descriptions—but without changing its basic vocabulary”; and second, “this framework allows the Yogācāra to distinguish between different levels of discourse and posit a hierarchy of meaning within the conventional realm of language” (p. 169).

Chapter 6, “Conversing with a Buddha: The Yogācāra Conception of Meaning as a Means for Overcoming Incommensurability,” explores broader epistemic ramifications. Incommensurability is “the potentially troubling scenario in which there is simply no common measure or standard of comparison across conceptual schemes because the meaning (and possibly also the reference of terms) differs radically between schemes” (p. 178). An example is how a Buddha freed from conceptualization could communicate with an unenlightened common person. It is the “unique polysemic quality” of the Buddha’s voice that overcomes incommensurability, because, “despite its diverse interpretations, [it] is still the object of a shared experience,” which “also contains certain private, unshared elements or discrepancies” (p. 190). This leads us to a discussion of “deep structural affinities between the Yogācāra understanding of linguistic meaning and its understanding of experience, particularly intersubjective experiences of the external world” (p. 180). A key element of this is that in Buddhism in general, and particularly in Yogācāra, “the notion of a ‘world’ designates more than a mere physical locus,” as the various types of living beings “occupy entirely different realms of existence” (p. 196). Between shared and private experiences, “whatever causal activity is shared at any given moment by our respective mind-streams will appear as intersubjective, and whatever ... is not shared will be experienced privately.” Both types of experiences are still “pervaded by conceptual categories” (p. 198). Underlying such conceptions are “impressions of speech,” which are causally efficacious (p. 199). “Intersubjectivity for the Yogācāra is not about how disparate subjects share knowledge of the world so much as it is about the incoherence of any egological view of consciousness; it is intersubjectivity that precludes any given notion of a subject” (p. 201). The two matters of intersubjectivity of experience and incommensurability of interpretation are thus both founded on a similar “deep structure,” based on causal terms and the functioning of the ālayavijñāna (p. 204).

Finally, in the conclusion, Tzohar draws out features and themes touched on above, and extends their possible applications. Sthiramati’s position “according to which all language is metaphorical” ties together multiple earlier threads in the face of the dichotomy of language being on one hand necessary for teaching but on the other an obstacle to the realization of truth (p. 205). Bringing together Yogācāra doctrines concerning both linguistic and perceptual meaning, Tzohar states that Sthiramati engages in a criticism of the correspondence theory of language and joins it to a “positivistic account of the causal and mental underpinnings of language in terms of the activity of consciousness.” Tzohar points out how Sthiramati’s use of polysemy (to counter monosemy) “reinforces the need to approach Indian philosophy diachronically, as a series of central debates and themes rather than through the prism of ‘schools’” (p. 206). The traditions differed in their understandings of metaphor as an absence of referent, from ontological nonexistence in Bhartṛhari, Asaṅga, the Lakāvatāra, and Sthiramati, to epistemic absence in the Mīmāṃsā and Abhidharmakośabhāya, or nonactuality in the Nyāya. The “alterity of figurative meaning,” which marks “the demise of the correspondence theory of meaning,” is that “secondary denotation is invariably viewed as both reliant on direct denotation and at the same time fundamentally different from it” (p. 208). For Sthiramati, this alterity “suggests the breach between ordinary language and reality, while also establishing a figurative causal theory of sense that salvages meaning discourse” (p. 209). Finally, Tzohar returns to examples of how classic Buddhist metaphors are employed in śāstra and commentary. While they do not fully reveal reality, “the clustering of figures highlights not only their referential exchangeability, but also the ways in which association governs description and the ultimate inability of description to penetrate reality” (pp. 218-219). This is an “unending process of association in which signs are stacked one on top of the other and which is incapable of exhausting its subject matter.” In terms of their study, “approaching Buddhist metaphors, therefore, requires us to consider the broad linguistic worldview in which they are embedded and the philosophical and poetical stakes involved in their use, as manifested in the case of Yogācāra discourse” (p. 219).

The monograph features two appendices. Appendix A is “A Translation and Exposition of the Vākyapadīya 2.250-256.” The translation includes the translated and original Sanskrit text itself, plus the īkā and Vtti. “Exposition” refers to Tzohar’s own running commentary. Appendix B is “A Running Translation of the Vākyapadīya 2.285-2.297.” This again includes the īkā and Vtti, though without any of Tzohar’s own discussion. There are nineteen pages of references, and the index is both extensive and well arranged.

In total, A Yogācāra Buddhist Theory of Metaphor is an outstanding study. It is solidly grounded in its historical context and demonstrates the need for a diachronic approach to the questions it raises. This can be seen in that this study goes beyond a single text, school, or even Buddhism alone, but extends out to provide knowledge about a period in Indian religious philosophy in general. There is appropriate sensitivity to earlier text-historical and interpretative scholarship on each of the texts it deals with, and awareness of how much such scholarship can influence later studies, but without merely rehashing old arguments and debates. In terms of its field, while the title referring to metaphor may suggest it is linguistically focused, its philosophical scope is far broader, extending to causality and theory of mind and perception. In this way, I feel it will soon become essential reading for all scholars and graduate students who work on Yogācāra/Vijñaptimātra and the other traditions upon which it touches (including but not limited to Nyāya, Mīmāṃsā, and even, obliquely, Madhymaka).

For this review, however, I would be derelict in my duties if I were not to indicate some possible avenues of improvement. At the very least it may hint at directions where future studies can complement and further flesh out a few incomplete corners of research. In particular, I would like to mention metaphor theory and sources beyond Sanskrit (and Tibetan).

The title of the monograph refers to “metaphor,” and it is worth pointing out that the original dissertation included “upacāra,” the Sanskrit term that it translates. Tzohar states that while other Buddhological studies on this theme have tended to “appeal to contemporary philosophical and literary theories of metaphors,” the present study will “attempt to reconstruct a body of theory on metaphor as formulated by Buddhist thinkers (i.e., using their own terms)” (p. 3). The problem pointed out was and still remains a critical issue for Buddhist thinkers, and is certainly deserving of study. I also appreciate the matter of what kinds of approaches we apply to our study of the material, largely modern Western academic disciplines, and the question of whether we are ultimately trying to understand Buddhism or reinterpret its formulations in forms conducive to certain other cultures (or even whether understanding can take place without reinterpretation). Despite this, we still have a problem here, which largely centers on the words “upacāra” and “metaphor” but extends into other phrases, such as “figurative language.” 

Yogācāra Buddhist thinkers did not use the word “metaphor,” or any English term, and when we make the decision to interpret the languages of our texts into any other language, we are then required to use those terms appropriately. There are many occasions in this work where “upacāra” does roughly equate to English “metaphor,” though about halfway through the work, I would argue that it takes on a different sense, significant enough to warrant that it no longer refers to metaphor (in the usual sense of the English word). We thus cannot avoid looking at Western thought to understand the term “metaphor,” which is defined in the Oxford English Dictionary as “a figure of speech in which a word or phrase is applied to an object or action to which it is not literally applicable.” A distinction between primary and secondary meaning would thus appear to be necessary, a prerequisite for the idea, and so without “literal” correspondence between word/phrase and thing, does the term “metaphor” mean anything at all? Or, to put it another way, can metaphor be universal, that is, does it abolish the distinction between primary and secondary meaning? I believe that to deal with this issue, some examination of contemporary metaphor theory is required, and quite probably the adoption of other terminology. Perhaps a generic structuralist approach would help, working with signs composed of signifier and signified, along with the common usage of “tenor” and “vehicle.” This may in turn run into problems once the signified loses ontological status but may help to work around the matter of “primary” and “secondary” meaning. Another approach could be to just use the Sanskrit term “upacāra,” and thus rather than translate it and effectively preface the study with a definition, let the entire study reveal to the reader its meaning and usage. This of course may lead to the problem of scholarly works overrun with technical jargon, particularly as we try to extend our audiences beyond our immediate Buddhological peers. I feel that Tzohar is quite aware of this problem of rendering the term into English and wonder if the choice of using the term “metaphor” in the title (at least) may have come from the editorial side. There is no easy solution to this, but I think it deserves further consideration.

The second matter is one of source texts and their languages. This monograph largely uses Sanskrit and Tibetan, as well as references to some Pāli sources. I feel that the latter two are a little problematic for this period of thought. The Tibetan sources, while usually literally accurate, are still often relatively late. The Pāli sources simply do not lead into the Sanskrit Abhidharma or Yogācāra traditions. What is missing is an examination of the Chinese sources. Granted, these are translations (though so are the Tibetan), but there is a wealth of Chinese Sarvāstivādin/Sautrāntika Abhidharma and early Yogācāra literature that includes both early and alternate sources for this period. An example is the Mahāvibhāsya of the Sarvāstivādin Vaibhāika tradition, an extremely influential text, which influenced the Yogācāra either through direct continuation or dialectically. These translations, due to being translations, can also provide insights into how such terms as “upacāra” were understood by other early exegetes, such as Paramārtha, Xüánzàng, their students, and commentators, approximate contemporaries of Sthiramati. Granted, there is a challenge of ascertaining the original Sanskrit term behind a given translation idiom. However, given Xüánzàng’s general consistency of idiom, we can trace terms from the Sanskrit Abhidharmakośabhāya or parts of the Yogācārabhūmi to his Chinese, and from this Chinese back to the Sanskrit of other texts for which we have Chinese translations (and sometimes no extant Sanskrit). To choose a handy example, we may look to the Kośaor Tattvārthapaala. In the former, we have 假說 and 假立 for upacāra. In the latter, where the original has vādopacāra and prajñapti vādopacāra, Xüánzàng consistently uses 假說 and 假說詮表, indicating a kind of non-ultimate designation or expression akin to prajñapti alone. We can also find these terms in the MahāvibhāyaNyāyānusāraVijñaptimātratāsiddhi, and other relevant textual sources. There may be a challenge of confirming without doubt that these other uses of the term are indeed upacāra, but if the focus is on the general sources of the Yogācāra theory of referential language, then it should not be limited to a single word, and these textual sources can provide still further valuable material for tracing the progression of the idea.


[1]. Roy Tzohar, “Metaphor (Upacāra) in Early Yogācāra Thought and Its Intellectual Context” (PhD diss., Columbia University, 2011), https://academiccommons.columbia.edu/doi/10.7916/D80017ZV/download.

[2]. See “Announcements,” Buddhist Studies, University of California, Berkeley (October 2018), http://buddhiststudies.berkeley.edu/announcements/ToshihideNumataBookPrize_2018_winner.html.

Citation: Matthew Orsborn. Review of Tzohar, Roy, A Yogācāra Buddhist Theory of Metaphor. H-Buddhism, H-Net Reviews. February, 2019. URL: http://www.h-net.org/reviews/showrev.php?id=53031

This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.