Beyond Arbitrariness: Kūkai’s Theory of Languages and Scripts

Nicholas Morrow Williams, University of Hong Kong


Kūkai, theory of language, deconstruction, Shingon Buddhism, calligraphy


Shingon 真言 patriarch Kūkai 空海 (774–835) had an original understanding of the role and significance of language, to which modern scholars have already drawn considerable attention. This essay challenges, however, the contemporary view that Kūkai’s perspective on language parallels that of deconstruction in skepticism about the stability of meaning; to the contrary, Kūkai vigorously asserts the truth and reliability of the Sanskrit mantras and even the seed syllables. More generally, modern scholarship tends to follow the assumption of the “arbitrariness of the sign,” in which languages are structured by systems of differentiation without any inherent significance. But again, Kūkai would have rejected this view, since he considers Sanskrit superior to other languages as a vehicle for the truth of the dharma. Finally, rather than considering writing as a secondary representation for natural, spoken language, Kūkai sees both Siddham and Chinese calligraphy as having a profound truth of their own. Kūkai and Derrida happen to be in agreement on one point, though, as they reject the priority of spoken language.


Kūkai 空海 (774–835), also known as Kōbō daishi 弘法大師 or the Great Dharma-Promulgating Teacher, is known as the founder of Shingon 真言 or Esoteric Buddhism in Japan, and as a polymath with cultural achievements ranging from his treatise of Chinese poetic criticism, Bunkyō hifu ron 文鏡秘府論, to his influential calligraphy and much else besides. Apart from his voluminous writings, Kūkai led an eventful life, traveling to Chang’an in the years 804–806, returning to become a close associate of the Emperor Saga 嵯峨 (r. 809–823), and finally establishing the vast Shingon sanctuary at Mount Kōya 高野, which now hosts his mausoleum. Kūkai has attracted a vast scholarship in Japanese and also substantial interest in the West. One point of continuing interest has been his theory of language, integrating elements from Mahāyāna Buddhism in general, from Esoteric Buddhism, from pre-Buddhist Chinese thought, and—intangibly but unmistakably—from his own homeland of Japan as well.

In my view, though, Kūkai has no theory of “language” per se, because he would not have accepted the assumption of modern linguistics that there is such a fixed category encompassing Chinese, Japanese, Sanskrit, etc. Instead, Kūkai has a theory of “languages,” assigning fundamentally distinct status to the different languages he knew.2 To be more precise, one should say that he has a theory of “languages and scripts,” as reflected in the title of his famous essay, Shō-ji-jissō gi 聲字実相義, “On the Meanings of Sound [i.e. language], Letter [or graph, character, etc.], and Reality.”3 A key element of this theory is that Sanskrit mantras and the letters of which they are composed have a unique and privileged status as keys to reality. For instance, one of Kūkai’s essays is devoted entirely to the Sanskrit syllable hūṃ.4 Kūkai’s view of the power of mantras is grounded in Indian tradition tracing far beyond Tantric Buddhism to the Upanishads.5 To this conception, though, Kūkai attaches a traditional Chinese conceptions of the characters as reflecting natural patterns so that they possess a privileged, though secondary, status.6

Inscribed GravestoneFigure 1 Gravestone inscribed with Siddham letter “A” at the Okunoin cemetery, final resting place of Kūkai, at his sanctuary complex on Mount Kōya. Photo by author, January 23, 2020. For Kūkai, who does not subscribe to the modern linguistic dogma of the “arbitrariness of the sign,” neither Sanskrit nor Chinese writing has an arbitrary, socially constructed relation to reality. This modern view of language originates with the great linguist Ferdinand de Saussure and his posthumously published Cours de linguistique génerale, in which he suggests that the elements of language possess meaning only by virtue of their differentiation from other signs, and lack any intrinsic significance.7 Though Saussure’s own thought balanced contradictory impulses from Romanticism and modern science––one reason he was unable to complete any publication on general linguistics in his own lifetime––his legacy to modern thought is clear.8 It is reflected in the tendency of Kūkai’s contemporary interpreters to present him as a having a theory of an abstraction called “language.” Even Takagi Shingen 高木訷元 and Thomas Eijō 瑩浄 Dreitlein, who together have authored a scrupulously accurate translation of several key texts, mislead the reader slightly in the title of their volume, Kūkai on the Philosophy of Language.

In fact, Kūkai’s view of languages is grounded in a sense of the mutual entanglements of culture, language, and reality. Classical Chinese (kanbun 漢文) is not one sign-system parallel to many others but is instead the basis of a cultural system all its own. Kūkai’s extant writings are all composed in classical Chinese, and he was one of the greatest masters of the Chinese rhetorical and poetic tradition in his day. His view of languages, like those of other topics, are expressed in a singularly dense style embedded with countless allusions to classical Chinese literature. Just as with languages, nor did Kūkai regard different forms of discourse within one language equally, but privileged certain forms of poetic and allusive writing. All this is particularly relevant for Sanskrit (and its written form in the Siddham script), which for Kūkai is not a language parallel with Chinese, let alone Japanese, but is the privileged means of expression for the Dharma-body itself. The fundamental sounds and letters of Sanskrit have a relationship to reality that is not symbolic but productive and iconic. For instance, the letter a “is the mother of all letters, the essence of all sounds, and the fountainhead of every aspect of reality” 即是一切字之母,一切聲之體.一切實相之源.9

In this paper I examine the hierarchical conception of language fundamental to Kūkai’s thought. Though in this paper we may only have space to examine the most basic features of Kūkai’s view, the schematic arrangement is clear. Spoken utterances and written texts can be arranged in two parallel hierarchies, proceeding in one case from ordinary spoken language, such as spoken Japanese or Chinese dialects; to more patterned and elegant speech, such as the language of the Sages or Boddhisattvas, among which Sankrit is already privileged; and culminating in the Mantras of enlightenment. Written texts can be arranged in a parallel scheme, but here Japanese has no place at all, lacking an independent writing system, so the hierarchy consists simply of Chinese/kanbun and Siddham, both aspiring to truth and elegance, but Siddham privileged far above Chinese, being the direct written representation of the Dharma.

I note in passing that Kūkai would not distinguish, like modern scholars, between written Chinese and its composition by Japanese scholars (kanbun); nor does he find that the Siddham script is the sole possession of Indian culture. In this regard Kūkai is more hierarchical in his thinking but also more cosmopolitan in his overall outlook than we are today.

Difference and Distinction

In approaching this topic we start out with our own assumptions about the use of language. Does language represent reality, and if so how? One of the fascinations of Kūkai, like other Buddhist thinkers of both South and East Asia, is his recognition of the limitations of language, which seems to anticipate similar queries of 20th-century philosophy. In particular, the scholar of Japanese and comparative literature, Thomas Hare, has attempted to read Kūkai from the perspective of deconstruction.10 Hare argues that Kūkai uses various analytical tools to draw out the ambiguities or contradictions in language, while also admitting that he combines this skeptical activity with a faith in esoteric knowledge that may be immune to such skepticism. In his conclusion, Hare offers the inspiring suggestion that Kūkai may offer a path for resolving the contradictions of modern thought.

Much of Hare’s analysis is focused on a single text by Kūkai, “Explaining the Title of the Diamond Sutra” 金剛般若波羅蜜経開題 (on the Vajracchedikā prajn͂ā pāramitā sūtra).11 The Diamond Sutra is of course one of the classic scriptures of Mahāyāna Buddhism, and an important reference point for Kūkai, but not one of the Esoteric scriptures central to his thought, so while it is potentially a valuable source, this approach risks confusing typical Buddhist arguments with those unique to Kūkai. At one key point, Hare discusses a passage where Kūkai is presenting the views of the Indian philosopher Nāgārjuna, borrowing mainly from the Shi Moheyan lun 釋摩訶衍論.12 Kūkai explains that the numberless Buddhas may differ drastically in their individual qualities and level of insight, while still being identical with respect to the dharma itself:

The virtues of each of these Buddhas surpass in number the sands of the Ganges and exceed the number of grains in an entire world reduced to dust, but each and every one maintains separate distinction in terms of root and branch, principal and auxiliary. All the same, they remain equal and, being equal, one; hence, the name “unconditioned dharma.”13:


Hare concludes that Kūkai uses the passage “to explain the existence of distinction and difference on a plane of perfect enlightenment, a kind of difference written under erasure, ~différance~, which allows the unbounded meaning of the profound and recondite interpretive strategy yet maintains a possibility of cognitive differentiation.”14

Though it is true that this kind of dialectical move from difference to identity to identity-in-difference is an important ingredient of Kūkai’s thought, however, it is a general feature of Buddhist philosophy rather than a particularity of Kūkai’s. There is no doubt that Kūkai’s thought is in large part built on general Mahāyāna assumptions, and also that he employs some of the traditional Mahāyāna arguments about the relativity of truth and the conditionedness of language.15 But elsewhere, Kūkai attempts to move beyond the paradoxes of Nāgārjuna’s philosophy by means of Esoteric absolutes.16 For instance, just a little later in the same text, Kūkai introduces the four samādhi he has identified in the Diamond Sutra, corresponding to the “four immeasurable states of mind” (Skt. catvāri-apramānāṇi, si wuliang xin 四無量心). These are actually Esoteric concepts which he is applying to the Diamond sutra rather than ideas already found within it. The third is the mind of great joy:

The doctrines of adaptation and transformation according to a person’s faculties, explain many terms and sentences in shallow and summary form. If one were to explain the deepest secrets of the mantras for someone with a high degree of native ability, so the text says that in order to cultivate the Greater Vehicle, the interpreters should refer to exoteric teachings. But in order to cultivate the Supreme Vehicle the interpreters refer to the mantras to explain the deepest and most profound meaning. This is why the Vajraśekhara tantra and other sutras are named the supreme vehicle.

According to the Mahā prajn͂āpāramitā śāstra, truth (prajn͂ā) has many levels of depth. For those of the earlier levels truth should be explained in its shallow form, and for those at the advanced levels can be explained in depth. At the higher levels the depth will also vary accordingly. The bodhisattvas corresponding to each separate level may also differ in knowledge. The bodhisattva of the first level does not know the truth of the bodhisattva at the second level. The bodhisattva of the ninth level does not understand what the bodhisattva of the tenth level does. The bodhisattva of the tenth level cannot understand what the Tathāgata Buddha does. Thus knowledge has depths without limit, without end, knowledge has breadth that is without limit, without end.17


This later passage adds context to the one analyzed by Hare. The point that bodhisattvas can be enlightened while having different levels of attainment cannot properly be related or even compared to deconstructive theories. In fact, all that Kūkai is saying here is that Buddhist practice is incremental; conceptually it is the same as the observation that under a graduated income tax, two people who make $50,000 per annum and $100,000 per annum, respectively, will both pay the same rate on the first $50,000, but the latter person will pay a higher rate on the second half of his income.

Even more importantly, Kūkai does not end here but goes on to cite a sutra dear to his heart, the Mahāvairocana tantra (Dari jing 大日經), which is fundamental to Esoteric teachings:

According to the Mahāvairocana tantra:18 ordinary men only understand the shallow and summary meaning of many terms and sentences of the Two Vehicles. But the Buddhas and Bodhisattvas can understand the infinite meaning of just the individual terms and sentences. It is the same here for the depth and breadth both of the title and of the contents of the sutra. Thus among the countless terms, sentences, and phrases, explained in the body of the sutras, in each case what is explained is the meaning of what has no attachments and is unchangeable: all of them are simply referring to the samādhi of great joy belonging to Mañjuśrī, bodhisattva of great wisdom.19


According to the editors of the Kōbō daishi chosaku zenshū, the source for this idea within the sutra to which Kūkai attributes it is unclear. Presumably the general point is the way that insight and enlightenment can be found through the recitation of mantras and are even embodied in particular Siddham syllables. This idea is even hinted at by Kūkai’s striking use of “in the body of” 身中 where he seems to mean simply “in the midst of.” The critical issue here, though, is that when Hare suggests that Kūkai has two complementary interpretive strategies, he is overlooking their intimate relationship in Kūkai’s eyes. For Kūkai, the profound and esoteric meaning of the text is not opposed to but is an extension of the difference apparent on the surface. Moreover, the especially potent words and symbols, namely the fundamental Sanskrit sounds, Siddham symbols, and Esoteric mantras, in themselves contain the profound implications of the Dharma.

In other words, rather than having a skeptical view of language, Kūkai has an incrementalist one. Ordinary people using ordinary language will not be able to express meaning fully, but Buddhas and Boddhisattvas are a different story. It is worth noting two further points. First, the source of this citation is not actually clear. Moreover, upon reflection, Kūkai’s view of language cannot be the same as that of the original Indic sutra because he is viewing the sutras from the outside, as a higher body of knowledge.20 But to locate the ground of language for Kūkai, we need to consider another key text.

Language and Script

The Shō-ji-jissō gi 聲字実相義 is one of the most-discussed of Kūkai’s writings precisely because it explicitly addresses issues of language and meaning, and so appears to be a useful guide for interpreting his other works. As we have seen from the essay on the Diamond Sutra, many of Kūkai’s Buddhist writings are responses or introductions to other texts. One might say they are “dialogic” in constantly alternating between Kūkai’s voice and that of other texts, or sometimes of imaginary interlocutors. A significant part of Shō-ji-jissō gi consists of two poems by Kūkai with his own extensive commentary. It is not an argumentative essay with a single thesis, but the theme is indicated by the title, which places the three concepts of sound, letter, and reality in parallel. Kūkai is arguing that within his own Buddhist cosmology, spoken language, written letters, and the reality all around them are three parallel strata, constantly in correspondence, shaping and forming one another.

The most detailed study of Kūkai in English is Ryūichi Abé’s The Weaving of Mantra: Kūkai and the Construction of Esoteric Buddhist Discourse. Abé’s detailed study of Kūkai’s life and works makes a number of important points, notably: that many of the disparate elements of Shingon Buddhism had already been transmitted to Japan before his time, so that Kūkai’s innovations lay in more radical, fundamental transformations; that Kūkai’s esoteric thought was not just a sectarian Shingon project, but influenced Japanese Buddhism and culture much more broadly; and that Kūkai’s establishment of Shingon Buddhism was in part a political project resisting the ritsuryō 律令 (essentially Confucian) organization of society. For the purposes of my discussion here, the most relevant section of Abé’s book is its seventh chapter, “Semiology of the Dharma; or, the Somaticity of the Text.” In this chapter, Abé emphasizes that Kūkai’s mature theory of text and language is founded in his esoteric thought, and in esoteric scriptures like the Mahāvairocana tantra. Moreover, Abé sees the “somaticity of the text” as a key element in Kūkai’s thought: Kūkai repeatedly emphasizes the way that human use of language, whether expressing words as sound or inscribing letters physically, is a physical, ritualistic act that relates the embodied subject to the dharma-body of the universe.

While this is indeed a critical aspect of Kūkai’s thought, the “somaticity of the text” is ambiguous with regard to the reliability or truth of bodies and/or texts. In spite of Abé’s comprehensive study of Kūkai’s works and life, he too, like Hare, misrepresents Kūkai’s view of language, in Abé’s case presenting Kūkai as a structuralist or semioticist who sees only a universe of signs. For instance, consider the critical second quatrain from Shō-ji-jissō gi, on which Kūkai presents extensive commentary. I quote directly from the authoritative recent translation by Takagi Shingen 高木訷元 and Thomas Eijō 瑩浄 Dreitlein:21

Form (rūpa) has such qualities as color, shape, and movement.
Both the sentient and insentient––beings and their worlds––have these.
There is the spontaneous, and there is conditioned arising.
They can delude or then enlighten.

The whole verse is describing a complex set of relations between form and reality, the spontaneous Dharma and conditioned beings. Abé, by contrast, renders the third line “As [the Dharmakāya’s] spontaneous play and as their consequences, [these letters].”22 This curiously unliteral translation identifies “letters” with the concept of “play,” Abé’s rendering of unconditioned spontaneity, hōnen 法然 (also hōni 法爾). But the original poem is not merely indicating the diverse relations among forms (such as speech and writing), living beings, and reality (the conditioned and the unconditioned). By translating hōnen 法然 as “spontaneous play,” Abé’s translation gives the impression that Kūkai sees everything outside of the Dharma itself as “play” or performance without fixed meaning.

In fact, Kūkai often distinguishes at least one stratum of “true words,” mantras which are grounded in the teachings of supreme Buddhas, as he declares elsewhere in Shō-ji-jissō gi: “Though the Mantras possess infinite variations, when you investigate their fundamental origin, it is no other than the King-Mantra of the Ocean-Seal Samādhi of the Lord Mahāvairocana” 雖云真言無量差別,極彼根源,不出大日尊海印三昧王真言.23 Abé finds that for Kūkai “mantras are not intrinsically distinct from ordinary language.” But his argument for this equivalence seems to rely on a misreading of Kūkai’s text, the final chapter of the Himitsu mandara jūjūshinron 秘密曼荼羅十住心論.24 Abé translates “Because the letters of worldly languages are already capable of expressing what reality is, Tathāgatas empower [some of] them and present them as mantras. If one takes the position that the letters of worldly languages are external to the nature of the Dharma, that is nothing but the false view of a delusory mind.”25 According to this translation, Kūkai would be treating all forms of language as fundamentally equivalent in “expressing what reality is.”

My own translation of the full passage, however, differs considerably. Here is the same passage, including the two sentences which follow Abé’s quotation:

“O Lord of Secrecy, do you know what the path of the mantra for the Thusly-Come Buddhas is? That is, by protecting and nurturing these written scripts,”26 according to the actual meaning of worldly script and language. Therefore the Thusly-Come receive and enact them according to the actual meaning of the mantras. As for other worldly writings that depart from the nature of the Dharma, these must be the erroneous views of deluded minds. Everything that has no real body to be sought, but is only protected and nurtured by the Buddhas employing their divine power, would be following an inverted course. These are not mantras (true words).”27

祕密主,云何如來真言道?謂加持此書寫文字 ,以世間文字語言實義。是故如來即以真言實義而加持之。若出法性外,別有世間文字者,即是妄心謬見。都無實體可求,而佛以神力加持之,是則隨於顛倒,非真言也。

But both this commentary and the original passage of the sutra on which it comments are focusing on the nature of mantras rather than of ordinary language.28 Our discussion in the previous section has already prepared us to understand what this means. The commentary is simply clarifying that not all uses of language represent the truth. At one extreme there are the mantras, and at the other lies and deceptions.29

Finally, Abé concludes his discussion of Kūkai on writing with this statement:30

As he repeatedly states, “All sorts of letters are differentiations” (KZ 1:530). That is, for Kūkai, there is no identity among diverse signs of the text, except that they are all differentiating from one another manifesting emptiness ingrained in the text, whether it is book, a royal palace, or the universe.

Abé is trying to present Kūkai as a Heian forerunner of Saussure, with the “diverse signs” of the universe having no inherent meaning of themselves, but only manifesting the same emptiness. This is an appealing vision, but it may not be exactly what Kūkai had in mind. The assertion that “All sorts of letters are differentiations” comes once again from the Shō-ji-jissō gi, where after enumerating how the various impressions of the senses are understood by the mind, Kūkai concludes: “These kinds of differentiations are themselves letters of the text” 如是差別即是文字也.31 But this assertion just means that the differentiations in sense impressions themselves constitute a kind of writing, not that they are equivalent to writing (“A is B” can mean A⸦B as well as A=B). For, later in the same text, Kūkai says that “Paintings of various sentient and insentient beings are also the letters of form. Silk brocade, embroidered cloth, patterned cloth, silk gauze, and so on are also the letters of form.” 又彩畫種種有情非情亦名色文字,錦繡綾羅等亦是色文字也.32 Here he is simply saying that decorative patterns are comparable to written language, the formal equivalent of script.33

The difference is fundamental. If there is no identity but only infinite differentiation, how can Kūkai’s mantras be a reliable key to enlightenment? It is true that with the advantage of historical perspective, it is easy to view Tantric Buddhism in general as a kind of marginal innovation in Buddhist practice whose great contribution lies in certain prominent distinctions from earlier Mahāyāna Buddhism. But this can hardly have been Kūkai’s point of view. Instead he seems to be willing, at least at moments within his arguments, to view the sensory universe as an intricate inscription (by the Mahāvairocana Buddha, perhaps), of which the Siddham letters form a kind of cryptographic key. But the origin of all these signs is not underlying emptiness but the omnipresent Dharma-body.

I would not want to exaggerate the importance of these specific textual issues; Kūkai's texts are difficult, mingling translated passages from sutras and commentaries with his own innovative (or eccentric) usages. But overall, Abé’s understanding of Kūkai is open to the same objection as Hare’s. Both scholars correctly point out Kūkai’s presentation of language as a ubiquitous phenomenon underlying and perhaps also distorting our perception of the world. In different contexts this argumentation might appropriately be compared to structuralist or poststructuralist views. But Hare and Abé neglect to note that for Kūkai, only certain forms of language are conditioned. Recall that Kūkai does not write that language as a whole is just a system of differentiations; to the contrary, he says instead that any system of differentiations in physical form, such as visual or auditory sense data, can be understood as a kind of writing; but only one kind among several, because there are forms of writing which are unchanging and unconditioned.34

Though these are only quite limited examples from Kūkai’s vast corpus, they suggest that the contemporary tendency to present him as engaging in deconstruction is misguided and reads into his thought a “spontaneous play” that is not actually stated by Kūkai. Instead, Kūkai isolates a special mode of language: the mantras which must be recited in their original Sanskrit and inscribed in Siddham calligraphy.

The Ladder of Meaning

One of the fundamental themes of Kūkai’s thought is that the route to enlightenment is incremental and can be divided into stages of awareness, as in the ten stages of his magnum opus, the Himitsu mandara jūjūshinron. A simpler version of this is his essential distinction between “Exoteric” and “Esoteric” teachings. Although these terms existed before Kūkai, he thought through the distinction more systematically and applied it more generally to Buddhist doctrine.35 It is relevant to many of his extant writings and to his claim as a prelate to be the leader of the Shingon sect superior to the preexisting Buddhist sects. Kūkai expounded the distinction in the “Disquisition Distinguishing the Exoteric and Esoteric Teachings” 辨顯密二教論.36 The essay opens:37

The Buddha has three separate bodies, but only two kinds of teachings. The explanations of reception and transformation are called exoteric. Their words are obvious, summary, and adapted to the occasion. When the dharma-buddha speaks, it is called the secret treasure. Its words are secret, profound, and substantive teachings.

Here Kūkai rejects the notion that words in general can represent truth directly. Instead, he believes that the words of the “exoteric” sutras were only partial and inaccurate representations. At the same time, the esoteric teachings are the actual words of the dharma-buddha and are “substantive” 實, conveying the truth in and of themselves.

This distinction leaves open various questions, of course, such as why the Buddha would need to preach in the Exoteric form, and how Kūkai obtained access to the Esoteric teachings. Thus he comments on these issues in a dialogue with an anonymous questioner:

Question: The response body and the transformation body preach the dharma, as confirmed by all the different schools. As for the dharma body, it has no appearance, no form. It is cut off from the reach of human language, when the mind approaches it vanishes, it cannot be explained or demonstrated. The various sutras all teach this lesson, and the various treatises also discuss it in the same way. So how can you now discuss the dharma-body preaching the dharma? Where is the proof of this?


Response: In the various sutras and treatises this idea is sometimes expressed. Although the text may be hidden from those of stubborn views, its meaning appears in accordance with one’s ability. It is just like how gods and demons see different views, or how men and birds perceive light and dark differently [according to their distinctive natures].38


The final sentence here uses two Buddhist allusions to present very concisely the theory behind graduated levels of doctrine. The first is the parable in which four different classes of beings look differently upon the same pool of water. The gods see it as a pond encrusted with jewels; the mortals see plain water, the hungry ghosts see blood and pus, and fish see their own abode.39 Similarly, the nighttime looks dark to men but not to the kalaviṅka birds which have night vision.40

At this level of abstraction, Kūkai’s teaching might seem similar to the conventional Buddhist one of upāya, expedient means, as if the exoteric version of Buddhism is simply tailored for less percipient minds. But actually there is an important distinction in his conception of the Esoteric which both these parables help to clarify. The theory of the Esoteric is not just about production but about reception. The Esoteric teachings have existed before, but only a skilled interpreter can recognize them, like the kalaviṅka birds which look into the dark of night as if staring into midday sun.

At this point we need to return to the fundamental centrality of Sanskrit in Kūkai’s conception of Buddhism, language, and reality. While Kūkai does not seem to have had the opportunity to learn Sanskrit grammar or actually to read or write in Sanskrit, he became devoted to the Siddham script, fondly writing encomiums to individual Siddham letters even without being able to compose them into sentences.41 Though Sanskrit letters and mantras are ubiquitous in his writings, there are also important documents where he discusses their particular significance for him. For instance, in a memorial to the Emperor Saga on August 30, 814, he presents a number of valuable Siddham and Chinese texts. The opening paragraph recapitulates the classic Chinese account of the origins of writing, based on the prognostications of the trigrams invented by Fu Xi and employed for moral teaching, to transmit the lessons of the sages throughout the realm and into the future; but then continues with Siddham, before celebrating both forms of script together from the point of view of Kūkai’s love of writing:42

Memorial Presenting Siddham Characters and Various Texts 獻梵字并雜文表

I, seeker Kūkai,43 say: I have heard that if the way of the emperor moves Heaven itself, then the secret register will surely be revealed; if the imperial winds [of influence and control] stir the earth, then numinous writings will triumph as well. Thus, it was only possible for the dragon trigrams and the tortoise inscriptions44 to be marked out properly once the Yellow Emperor and Fu Xi ruled; the phoenix writings and the tiger characters had to wait for Lord Bai [sage-emperor Shaohao] and Lord Ji [King Wen of Zhou] to reveal their forms. Then discarding the old way of notched ropes, the texts of the Three Mounds flourished.45 The wood-carving method was forgotten, and the Five Canons all thrived gloriously.46 The Resplendent Emperor promulgated his teaching and influence, encouraging moral transformation, and the common people all looked up and recognized it, coming to observe. Without even leaving the gate of their courtyard, he seemed to gaze ten thousand leagues away; without sagely wisdom, he mastered all the methods of the heaven, earth, and man. Examining antiquity and refreshing past knowledge, revealing the proper model by his own person, without writing how is all this possible?

As for the marvelous writings of Siddham, the letters of Sanskrit writing: in form they condense the former Buddhas, in their principles they contain the wisdom of multitudes. Thus the fully Awakened ones of past, present and future have been honored as teachers; the Boddhisattvas of all ten realms were revered beyond life itself. Treasures filling the entire universe barely equal the wisdom of half a gāthā; the obstructions accumulated through the kalpas can all be swept away in a single thought. Truly the sense and function of writing are mighty and far-reaching indeed!

I bow in reverence before your majesty the Emperor, declaring the title of the “three joined by a line” [to make the character 王 “king”], establishing primary authority as the “five with one line deleted” [again making the character 王]. Your path proceeds properly according to compass and square, your radiance is equal to that of Sun-crow or Moon-hare. Your benevolent dew descends on all under Heaven, and within all six corners you govern by Non-Action. The breezes stir your zither, and by means of your solitary person all is at peace. Regulating governance by jade and candle, reflecting in the bronze mirror, today we can all observe the motion of the auspicious Wheel.

I, Kūkai, am just a shard or pebble of a man, but ever regard in reverence the Golden Immortal [the Buddha]; in my rectitude I bow in awe before Chao or Xu,47 but still I have long reclined in the clouds of Yao and the Yellow Emperor. In my spare time from contemplation in my grotto, I have made a study of the writing of India. After sitting some time with my boiled tea, I peruse for a while the books of China.48 Whenever I look at the ancient seal script of Scribe Cang Jie, or the modern clerk script of Wang Xizhi,49 or the force of the grass script by Master Du,50 I always feel a surge of energy and forget my sorrows, a passion from the mountains full of laughter. A proverb says: “What is sweet for the servant is sweet for the master too.” For a long time I have wanted to present these writings according to this principle. But since they are disarrayed and defiled, I feared to brush with dust your sagely vision. I only hope that my slight sincerity may confidentially penetrate and be heard by heaven.

Humbly bearing the oral command of the sea of Fuse,51 leaping in delight, I have bound and prepared the Encomia to Scripts of Past and Present, Wang Xizhi’s Lanting Inscription, and also Sanskrit writings and Siddham, for eleven fascicles in all, which I dare to present now to your Majesty. I only beg that Heaven in its mercy not reject a single drop, but peruse in entirety these scattered traces. I bow in prayer that when your Majesty unfolds these Sanskrit words, the guardianship of Indian gods will be manifold and mighty; that when you further examine these divine books, the protection of deities and men will be numerous and vast. May the farthest shores of the Anavatapta River swiftly enter your own domain; may the lofty ridges of Mount Song come receive your inaugurating calendar.52 The permanently abiding letters protect and nurture your imperishable body; may the people as in remote antiquity continue to play their wood-tossing game and plough the fields in tranquility up to the present day.53 You appoint officials according to dragon auguries, who dally with you at Mount Guye; with phoenix-like auspices you designate the offices, and revel in relaxation at the Golden Tower.54 I have ventured to besmirch momentarily your dangling jade and the axe-patterned screen; bowing deeply, I tremble in fear before you. I, the seeker Kūkai, in sincerest awe and reverence, make these humble statements.

Sanskrit Words and Siddham Letters, with Meanings Explained, one fascicle.
Encomia to Writings of Past and Present, three fascicles.
Forms of Seal and Clerk Script of Past and Present, one fascicle.
Emperor Wu of Liang’s Evaluations of Grass Script, one fascicle.
Wang Xizhi’s Lanting Inscription, one fascicle.
Vinaya Master Tanyi’s Stone Inscriptions, one fascicle.55

Presented by seeker Kūkai on August 30, 814.

After effusive encomia to the power of writing, the final section praises the emperor’s power in even more extravagant and deferential terms. The key point for my purposes, though, is how, even within the context of Kūkai’s devotion to calligraphy and his gift to the emperor, the text distinguishes precisely between Chinese and Indic writing. The first paragraph praises the political function of Chinese characters, mentioning its legendary inventors, but the second paragraph discusses how Siddham script allows one to sweep away doubt and achieve enlightenment. Thus Kūkai, while seeing Chinese characters and Siddham writing as comparable in form, also distinguishes them radically in terms of value. Moreover, Kūkai has no conception of “language” or “writing” in itself. Instead, he sees both Chinese and Siddham writing as representing millennia—or even kalpas—of historical development, recalling and even reviving the culture-heroes and Boddhisattvas of the past. In other words, he has no notion of—or would reject—value-neutral “writing” intended solely as a transcription of spoken language.

I suspect that Kūkai was particularly well-placed to appreciate the special potency of Sanskrit mantras in Esoteric Buddhism by his experience as a student of classical Chinese in Japan. To conquer the difficulties of Chinese expression, he memorized and recited hundreds of Chinese texts, and was a incomparable master of Chinese style. His attention to the form of the characters is reflected in his famous calligraphy. Yet when he encountered Sanskrit, he perceived it as occupying a higher level yet. Kūkai’s theory of languages and scripts places the languages and scripts that he knew in a hierarchy of propinquity to the Dharma. The key difference between Shingon thought and earlier Mahāyāna philosophy is that language is decidedly not a tool to be discarded when enlightenment is achieved. By means of language we can approach the truth, but only if we hone our own linguistic abilities, and employ language as effectively and elegantly as we can, and learn the different languages we need to access higher insights.

Universal and Particular

More broadly, the trouble with many previous discussions of Kūkai’s view of language is that they suggest he had one generalized view of letters, or words, or languages. We are accustomed to seeing the essential unity of the world’s many languages. Since Saussure, this approach has flowered in the field of linguistics with Chomsky’s Universal Grammar and related initiatives. Though modern linguistics provides insights into the actual operation of spoken languages, however, its axioms are not necessarily the sole or most effective ones in dealing with the intracultural significance of language.

In spite of the huge differences between them, in fact, on this particular issue Derrida does offer a valid approach to Kūkai in his critique of Western logocentrism. One of Derrida’s most influential works, De la Grammatologie, is devoted to a critique of the Saussurean assumption of spoken language as primary and writing as secondary. Indeed, in that work Derrida takes pains to discuss how the varied modes of writing, including not just Chinese characters but also cuneiform and hieroglyphics, fit into a theory of writing’s priority.56 Kūkai never makes the error of treating writing merely a substitute for the authenticity of speech, which Derrida finds ubiquitous in Western philosophy, and at least in some of his writings Kūkai even seems to identify the written letters of Siddham as the basis of meaning. This point of contact between the two thinkers is deceptive, needless to say, since Derrida goes on to question the stability of Saussurian difference altogether, while Kūkai sees both the sound and the inscription of mantra as reliable sources of truth. And yet from a specifically East Asian point of view, Kūkai’s vision of Esoteric truths only properly expressible in Sanskrit might potentially be understood as a radical destabilizing of the Sinographocentric view. Finally, there is scope for a deconstructive reading of Kūkai, beginning with how his concept of Esoteric truth is founded in complementary relation with superficial and transient uses of language, but that would exceed the goals of this essay.

I would like to suggest simply that the assumptions of equality among languages and of the priority of spoken language to writing have distorted modern interpretations of Kūkai’s thought. By reducing his views of languages and scripts to a theory of “language” abstracted from place and culture, scholars have neglected some of his central claims. Kūkai does not see language as a vast system of symbols to be manipulated, but rather, in keeping with his soteriology and epistemology, as a ladder of understanding, in which we first employ those letters and words we have already learned, and only painstakingly advance to higher levels of knowledge, accessing higher mantras and deeper levels of meaning.

On the other hand, I would agree with Hare that Kūkai has insights which may remain relevant for us today, reflecting on the limitations of a global civilization that is simultaneously homogenized and fractured. From this point of view one aspect of Kūkai’s thought that may have enduring appeal is his loving attention to the particularity of words, languages, and scripts. For Kūkai, different sounds and letters have their own iconic features and a physical, somatic power, well described by Abé, that is not simply semiotic in nature. In that sense, there is an affinity between Kūkai’s view of language and that of unorthodox modern linguists espousing “linguistic relativism,” a research program that “moves away from a rigid dichotomization of structure and practice, focusing instead on their complex interactions.”57 In his quest to understand the power of Buddhist texts and practice ranging from his own Heian environment to China and India, Kūkai too emphasized the particularity of linguistic and written forms in relation to religious practices. That is to say that languages are constructions of the human imagination embedded in distinct cultures, with world-building, mythopoeic functions. Finally, without an accurate perception of how historical thinkers have understood the capabilities of their own languages, we have little hope of appreciating how our own perceptions of the world are structured by our own languages and media.

  1. This paper is a fortuitous outcome of my ongoing investigation into Kūkai’s unique view of the relation between literary art and religious insight. This project has been generously supported by a Japanese Research Fellowship from the Hakuhodo Foundation in 2019–2020, and General Research Fund grant #17615818 from the University Grants Council of the HKSAR. I am particularly grateful for the hospitality of Professor Midorikawa Hideki 緑川英樹 and the Graduate School of Letters at Kyoto University while I was gathering materials for this project.

  2. A valuable survey is Matsunaga Yūkei 松長有慶, “Kōbō daishi no gengokan” 弘法大師の言語観, in Kūkai no shisō to bunka: Onozuka Kichō hakase koki kinen ronbunshū 空海の思想と文化:小野塚幾澄古稀記念論文集 (Tokyo: Nonburusha, 2004), 1:35–52. Matsunaga emphasizes the multiplicity of Kūkai’s conception of language, with roots in both India and China, and distinguishing among words, script, etc., though he does not directly discuss the relative significance of Sanskrit and Chinese.

  3. The definitive edition of Kūkai’s works is Teihon Kōbō daishi zenshū 定本弘法大師全集 (Kōyasan, Wakayama-ken: Mikkyō bunka kenkyūjo, 1992–1997). This piece is in Teihon Kōbō daishi zenshū, 3:33–49. The notes and commentary in Kōbō daishi chosaku zenshū 弘法大師著作全集, ed. Katsumata Shungyō 勝又俊教 (Tokyo: Sankibō busshorin, 1968, 1994) are also very helpful and I will refer to them when necessary.

  4. . Teihon Kōbō daishi zenshū, 3:51–72. Both this and the essay mentioned in the previous note are conveniently accessible in the bilingual edition by Takagi Shingen 高木訷元 and Thomas Eijō 瑩浄 Dreitlein, Kūkai on the Philosophy of Language (Tokyo: Keio University Press, 2010).

  5. On this point see Matsunaga Yūkei, “Kōbō daishi no gengokan,” 45. Note that there was debate within the Brahmanic tradition over how and even whether mantras express meaning: see Johannes Bronkhorst, A Śabda Reader: Language in Classical Indian Thought (New York: Columbia University Press, 2019), 10.

  6. A good account of this topic, and how Kūkai’s view of Chinese characters was imbued with the patterns of nature in the Chinese Daoist tradition, is in Kamitsuka Yoshiko 神塚淑子, “Kūkai no mojikan” 空海の文字観, in Rikuchō Dōkyō shisō no kenkyū 道教思想の研究 (Tokyo: Sōbunsha, 1999), 415–39. However, see also Timothy Michael O’Neill, Ideography and Chinese Language Theory: A History (Berlin: Walter de Gruyter, 2016), presenting Confucian language theory in another light altogether.

  7. Saussure, Cours de linguistique générale (Paris: Payot, 1960), 166: “. . . dans la langue il n’y a que des différences sans termes positifs. Qu’on prenne le signifié ou le signifiant, la langue ne comporte ni des idées ni des sons qui préexisteraient au système linguistique, mais seulement des différences phoniques issues de ce système. . . . Un système linguistique est une série de différences de sons combinées avec une série de différences d’idées. . .”

  8. For a thoughtful reconsideration of the development of Saussure’s own ideas, see Boris Gasparov, Beyond Pure Reason: Ferdinand de Saussure’s Philosophy of Language and Its Early Romantic Antecedents (New York: Columbia University Press, 2013).

  9. . Unji gi 吽字義 (The Meanings of the Letter Hūṃ), in Teihon Kōbō daishi zenshū, 3:53; translation in Kūkai on the Philosophy of Language, 130–31.

  10. Thomas Hare, “Reading Writing Cooking: Kūkai’s Interpretive Strategies,” Journal of Asian Studies 49.2 (1990): 253–73.

  11. . Teihon Kōbō daishi zenshū, 4:255–68.

  12. . T 1668. This text was of great importance to Kūkai but probably not actually written by Nāgārjuna.

  13. Hare’s translation from “Reading Writing Cooking,” 270–71; original text from Teihon Kōbō daishi zenshū, 4:258.

  14. “Reading Writing Cooking,” 271.

  15. This passage is an odd choice as a representative of Kūkai’s thought, since it is one where he is seeking only to relay the relevant views of the principal philosopher of Mahāyāna ontology, Nāgārjuna. Moreover, the passage Hare translates to illustrate this point is dealing with a very specific question, the relation of enlightened beings to one another and how they can share one enlightenment while also differing in various ways. That is to say, it is more a point about the organization of the pantheon than about language or even belief per se.

  16. Cf. Ian W. Mabbett, “Nāgārjuna and Deconstruction,” Philosophy East and West 45 (1995): 203–25; David R. Loy, “Language against Its Own Mystifications: Deconstruction in Nāgārjuna and Dōgen,” Philosophy East and West 49 (1999): 245–60.

  17. . Teihon Kōbō daishi zenshū, 4:260–61. My translation.

  18. Original source unclear; Katsumata suggests Kūkai is summarizing his impression of the sutra rather than quoting it directly. See Kōbō daishi chosaku zenshū, 2:436, n. 2.

  19. . Teihon Kōbō daishi zenshū, 4:261. My translation, with emphasis added.

  20. The sutra itself does explain that the mantras transcend past, present, and future, being “formed through inconceivable and profound interdependent arising.” See Stephen Hodge, tr., The Mahā-vairocana-abhisaṃbodhi tantra (Abingdon, Oxon: RoutledgeCurzon, 2003), 172. See also Kūkai’s Shō-ji-jissō gi: “The deep, esoteric interpretation is that each syllable, each word, and each phrase contains infinite meanings and truths” 若作秘密釋者一一言一一名一一成立各能具無邊義理 (Kūkai on the Philosophy of Language, 91).

  21. . Kūkai on the Philosophy of Language, 104–5. See also Teihon Kōbō daishi zenshū, 3:41.

  22. . The Weaving of Mantra, 284. I have reformatted slightly and inserted “the verse reads” since Abé translates the preceding sentence and verse separately.

  23. Translation slightly modified from Kūkai on the Philosophy of Language, 100. See also Teihon Kōbō daishi zenshū, 3:40.

  24. . Weaving of Mantra, 295.

  25. Ibid.

  26. The first two clauses are from the Mahā-vairocana tantra itself (T 848: 18.10a). Kūkai defines kachi 加持 in Hizōki 秘藏記 (Kōbō daishi chosaku zenshū, 2: 652; Teihon Kōbō daishi zenshū, 5:149–50): “Ka is the protection and consideration of the various Buddhas, chi is our own independent action. Kachi together is like when the father brings the seed inside the mother, and the mother’s womb guards it well and nurtures the seed.”

  27. . Kōbō daishi chosaku zenshū, 1:568. Incidentally, Abé misidentifies as passages from Kūkai which are actually quotations from Yixing’s一行 (683–727) Dari jing shu 大日經疏, T 1796: 39.650c (clearly identified as such in the 1968 annotated edition by Katsumata, cited by Hare but not Abé).

  28. For reference, the relevant passage goes: “In what way do Tathāgatas transform the Mantra Method into letters? [This is the first part of the quotation, but Hodge adds a note suggesting that the Chinese version is different from the Tibetan.] Lord of the Secret Ones! They are the words of truth, the Four Noble Truths, the Four Recollections, the Four Bases of Supernatural Power, the Ten Strengths of a Tathāgata, the Seven Precious Limbs of Enlightenment, the Four Divine States and the Eighteen Uncommon Qualities of a Buddha, which were accomplished by the Tathāgatas over many billions of eons.” See Hodge, Mahā-vairocana-abhisaṃbodhi tantra, 132.

  29. Similarly, when Abé writes, “. . . Kūkai points out, ‘the infinite merit inherent in every one of the letters of worldly language is already equal to that of mantras.’” he is again discussing a passage from the Yixing commentary requoted by Kūkai, not Kūkai’s own views; and in any case, the passage under consideration says nothing about “worldly language” at all. See Kōbō daishi chosaku zenshū, 1:569.

  30. . Weaving of Mantra, 304.

  31. . Kūkai on the Philosophy of Language, 108–9 (translation slightly modified); Teihon Kōbō daishi zenshū, 3:43.

  32. . Kūkai on the Philosophy of Language, 114. Teihon Kōbō daishi zenshū, 3:45.

  33. Note discussion in Matsunaga, “Kōbō daishi no gengokan,” 42.

  34. Fujii Jun 藤井淳 has also clarified this point by discussing how Kūkai was following in the tradition of Six Dynasties philosophical debate on the question of whether “words exhaust meaning” 言盡意. Kūkai offered a new answer to this question. See Kūkai no shisōteki tenkai no kenkyū 空海の思想的展開の研究 (Tokyo: Transview, 2008), 376–91.

  35. Fujii Jun discusses the relationship of Kūkai’s distinction to that of his contemporary Gen’ei玄叡 as well as contemporary Esoteric thought in China (Kūkai no shisōteki tenkai no kenkyū, 147–87).

  36. Abé points out that the main purpose of the text, pace its title, is to find textual justification within the Exoteric literature for Kūkai’s conception of the Esoteric (Weaving of Mantra, 213–19 and 267–70). For my purposes, though, only the introduction, which explains Kūkai’s sense of this fundamental distinction, is pertinent.

  37. . Teihon Kōbō daishi zenshū, 3:75.

  38. . Teihon Kōbō daishi zenshū, 3:75–76.

  39. The story is told in full in the She dasheng lun shi 攝大乘論釋. See T 1598: 31.402c.

  40. The source for this point is the Shi moheyan lun, which was dear to Kūkai’s heart. See T 1668: 32.623c.

  41. Secondary scholarship does not always explain this point clearly, but the uneven trajectory of Sanskrit in premodern East Asia is traced in full by Robert Hans van Gulik, Siddham: An Essay on the History of Sanskrit Studies in China and Japan (Nagpur: International Academy of Indian Culture, 1956). Hare also discusses Kūkai’s incomplete understanding of the six types of Sanskrit compounds.

  42. The memorial is included in the fourth fascicle of Kūkai’s literary collection, the Seireishū 性靈集, in Teihon Kōbō daishi zenshū, 7:62–64; Kōbō daishi chosakushū, 3:195–200; Sangō shiiki; Seireishū 三教指歸.性靈集 (Tokyo: Iwanami shoten, 1965; Nihon koten bungaku taikei 71), 241–44. Cf. Kūkai’s earlier memorial reporting on sutras he has brought with him immediately on his return from China, translated in full in Yoshito Hakeda, Kūkai: Major Works (New York: Columbia University Press, 1972), 140–44.

  43. I render the Sanskrit śramaṇa, translated into Chinese as shamen 沙門 (J. shamon), as “seeker.” In the East Asian context it implies a Buddhist monk.

  44. The editors of the Nihon koten bungaku taikei edition identify parallels with the Ming compilation Shushi huiyao 書史會要 (Siku quanshu), 1.3b.

  45. The writings associated with the three sages Fu Xi 伏羲, Shen Nong 神農, and the Yellow Emperor (Huangdi 黃帝).

  46. The canons established (in writing) by Shaohao 少皞, Zhuanxu 顓頊, Di Ku 帝嚳, Yao 堯, and Shun 舜.

  47. The ancient recluses from the time of sage-emperor Yao, Chao Wen 巢文 and Xu You 許由.

  48. Kūkai uses the Chinese transcription Zhendan 震旦of the Indic name for China, Cīnaṣṭhāna.

  49. Wang Xizhi 王羲之 (303–361), whose “Preface to the Collection of the Lanting Gathering” 蘭亭集序 remains one of the most cherished pieces of Chinese calligraphy.

  50. Du Bodu 杜伯度 of the Latter Han dynasty.

  51. The modern Jūnichōgata十二町潟 lake in Toyama prefecture, which is mentioned frequently in the Man’yōshū. But it is unclear why it appears here.

  52. Mount Song is the central one of the five sacred mountains or “marchmounts” of China. In these lines Kūkai expresses the fanciful wish that by means of Chinese and Sanskrit writing, Emperor Saga’s benevolent rule will be sustained by both Indian deities and Japanese kami, and extend to encompass the territories of the continent as well.

  53. Under the sage-kings of antiquity the peasants lived in such contentment that they were said to devote much time to the game of jirang 擊壤, which involved tossing a wooden tablet to strike another one imbedded in the earth.

  54. Both Mount Guye and the Golden Tower are elegant references to the imperial palace. Mount Guye is a legendary home of immortals.

  55. Tanyi 曇一 is the style name of the monk Jueyin 覺胤 (692–771).

  56. . De la Grammatologie (Paris: Édition de Minuit, 1972), 131–42.

  57. See Jane H. Hill and Bruce Mannheim, “Language and World View,” Annual Review of Anthropology 21 (1992): 400. I am also inspired here by the work of the interdisciplinary linguist Paul Friedrich, including “The Symbol and Its Relative Non-arbitrariness,” in Language, Context, and the Imagination (Palo Alto: Stanford University Press, 1979), 1–61; and The Language Parallax: Linguistic Relativism and Poetic Indeterminacy (Austin: University of Texas Press, 1986).