俱舍論疏 5.243

5.243

論「云何令了」，經部問也。論「謂語發名至乃能令了」，有部答也。《正理論》云「此中經主作如是言：豈不此三語為性故用聲為體色自性攝，如何乃說為心不相應行？此責非理。所以者何？由教及理知別有故。教謂經云『語力、文力』，若文即語，別說何為？又說『應持正法文句』，又言『依義不依於文』。又說伽他因，謂闡陀文字。闡陀謂造頌分量語為體。又契經言『知法知義』，法謂名等，義謂所詮。又契經言『文義巧妙』，又言『應以善說文句讀誦正法，惡說文句讀誦正法義即難解』。又說『如來獲得希有名、句、文身』，又說『彼彼勝解文句甚為希有』。由此等教證，知別有能詮諸義名、句、文身。猶加語聲，實而非假。理謂現見有時得聲而不得字，有時得字而不得聲，故知體別。有時得聲不得字者，謂雖聞聲而不了義。現見有人粗聞他語，而復審問汝何所言。此聞語聲不了義者，都由未達所發文故。如何乃執文不異聲？有時得字不得聲者，謂不聞聲而得了義。現見有人不聞他語，覩唇等動知其所說。此不聞聲得了義者，都由了達所發文故。由斯理證，文必異聲。又見世間隱聲誦呪，故知呪字異於呪聲。又見世間有二論者，言音相似一負一勝，此勝負因必異聲有。又法與詞二無礙解，境界別故，知字離聲。是故聲者但是言音相無差別，其中屈曲必依迦、遮、吒、多、波等。要由語聲發起諸字，諸字前後和合生名，此名既生即能顯義，由此展轉而作是言。語能發名，名能顯義，故名、聲異，其理極成。應知此中聲是能說，文是所說，義謂非二，如是則為無亂建立。」

On the Śāstra's statement: "How do they make it understood?": The Sautrāntika asks. On the Śāstra's statement: "Namely, speech produces a name [...] then it can make it understood": The Sarvāstivādin replies. The Nyāyānusāra says: "Here the author of the Śāstra [Vasubandhu] states: 'Are these three not of the nature of speech, using sound as their substance, and therefore subsumed under the nature of form? How then can they be stated to be conditioning factors disassociated from thought?' This reproof is unreasonable. Why so? Because by both scripture and reasoning, we know they exist distinctively. Scripture: as a sūtra says, 'the power of speech and the power of syllables (vyañjana)'. If vyañjana were simply speech, why state them separately? It also says, 'One should uphold the words and phrases (pada) of the true Dharma.' It also says, 'Rely on the meaning, not on the letter (vyañjana).' It also speaks of the cause of a gāthā, which is meter (chandas) and syllables (akṣara). Chandas refers to speech as substance measuring the composition of a verse. A sūtra also states, 'Knowing the Dharma and knowing the meaning'—Dharma refers to name and so forth, while meaning refers to what is expressed. A sūtra also states, 'The letters and meanings are subtle and wonderful,' and also, 'One should recite the true Dharma with well-spoken words and phrases; if one recites the true Dharma with ill-spoken words and phrases, the meaning is difficult to understand.' It also says, 'The Tathāgata has attained rare collections of names, phrases, and syllables,' and also, 'Their respective resolute faith in the words and phrases is exceedingly rare.' From such scriptural proofs, we know that collections of names, phrases, and syllables that express meanings exist distinctively. Like vocal sound, they are real and not nominal. Reasoning: because we empirically observe that sometimes one apprehends the sound but does not apprehend the phonemes, and sometimes one apprehends the phonemes but does not apprehend the sound; hence we know their substances are distinct. 'Sometimes apprehending the sound but not the phonemes' means: although hearing the sound, one does not understand the meaning. We empirically observe someone who cursorily hears another's speech and then asks closely, 'What did you say?' This hearing of speech-sound without understanding the meaning is entirely due to not having arrived at the phonemes produced. How could one grasp that phonemes are not different from sound? 'Sometimes apprehending the phonemes without apprehending the sound' means: not hearing the sound, yet understanding the meaning. We empirically observe someone who does not hear another's speech, but by watching the movement of the lips and so on, knows what was said. This understanding of meaning without hearing the sound is entirely due to understanding the phonemes produced. By this logical proof, phonemes must be different from sound. Moreover, we observe in the world that people recite mantras with suppressed voice, so we know the mantra-phonemes are distinct from the mantra-sound. Moreover, we observe in the world two disputants whose voices and sounds are similar, yet one loses and one wins; this cause of victory and defeat must exist distinct from the sound. Furthermore, because the spheres of the two unobstructed knowledges (pratisaṃvid) of Dharma and Nirukti are distinct, we know that phonemes are separate from sound. Therefore, sound is merely vocal tone without differentiation in characteristic; its modulations necessarily rely on ka, ca, ṭa, ta, pa, and so forth. Only through vocal sound are the phonemes manifested; the phonemes, aggregated in sequence, produce a name; once this name is produced, it can manifest the meaning. By this succession, it is stated thus. Speech can produce a name, and a name can manifest a meaning; therefore name and sound are distinct, and this principle is solidly established. It should be understood that here sound is the expresser, the syllable is that which is expressed, and the meaning is neither of the two; thus the establishment is without confusion."