Book 3 - Pada-kāṇḍa (4): Dravyalakṣaṇa-samuddeśa (On Substance)

3.4.1-2 unformalized

संसर्गरूपात् सम्भूताः संविद्रूपादपोद्धृताः ।
शास्त्रे विभक्ता वाक्यार्थात् प्रकृतिप्रत्ययार्थवत् ॥ १ ॥
निमित्तभूताः साधुत्वे शास्त्रादनुमितात्मकाः ।
केचित् पदार्था वक्ष्यन्ते संक्षेपेण यथागमम् ॥ २ ॥

saṃsargarūpāt sambhūtāḥ saṃvidrūpādapoddhṛtāḥ |
śāstre vibhaktā vākyārthāt prakṛtipratyayārthavat || 1 ||
nimittabhūtāḥ sādhutve śāstrādanumitātmakāḥ |
kecit padārthā vakṣyante saṃkṣepeṇa yathāgamam || 2 ||

(1) Arisen from the form of commingling, abstracted from the form of awareness, divided in the science out of the sentence-meaning — like the meanings of stem and suffix; (2) being the occasioning grounds of correctness, their selves inferred from the science — certain word-meanings will now be stated, in brief, according to tradition.

The Padakanda's methodological charter, and it is placed at the head of the samuddesa that names the kanda's central category. Four descriptions of what a padartha is, each subtracting something. Samsarga-rupat sambhutah — *arisen from the form of commingling*: word-meanings exist only where things have been related (3.1.20's prapanca arising only among things commingled; 3.3.57's turbidity born of interpenetration). Samvid-rupad apoddhrtah — *abstracted from the form of awareness*: they are extracted from cognition, not found in the world (2.132's buddhi-visaya; 3.3.41's division which, having made the object external, is itself an abstraction of capacities). Sastre vibhaktah vakyarthat — *divided in the science out of the sentence-meaning*, prakrti-pratyaya-arthavat, *like the meanings of stem and suffix*: 3.1.1's opening move repeated verbatim, and 2.10's doctrine — the pada is to the vakya what the suffix is to the pada. (2) adds their office and their epistemic status: nimitta-bhutah sadhutve — the *occasioning grounds of correctness* (they exist so that the grammar can say which forms are right: 1.27's sadhutva, the dharma-bearing property); sastrad anumita-atmakah — their *selves inferred from the science*: we know them only because the science needs them, which is as clean a statement of theoretical posits as antiquity produced. And the closing modesty: kecit padarthah vaksyante samksepena yatha-agamam — *certain* word-meanings will be stated, *in brief*, *according to tradition*. Kecit: some, not all. Yatha-agamam: as received. The fourteen samuddesas are announced as a selective, traditional, abbreviated account of items that are commingling-born, awareness-abstracted, science-divided and inference-known.

3.4.3 unformalized

वस्तूपलक्षणं यत्र सर्वनाम प्रयुज्यते ।
द्रव्यामित्युच्यते सोऽर्थो भेद्यत्वेन विवक्षितः ॥ ३ ॥

vastūpalakṣaṇaṃ yatra sarvanāma prayujyate |
dravyāmityucyate so'rtho bhedyatvena vivakṣitaḥ || 3 ||

Where a pronoun is employed as the indicator of a thing — that meaning is called 'substance', being intended as what is to be differentiated.

The definition of substance, and it is startlingly formal. Not: substance is what bears qualities; not: substance is what persists. Rather: vastu-upalaksanam yatra sarvanama prayujyate — where a *pronoun* is employed as the *indicator of a thing* — dravyam iti ucyate so 'rthah: *that* meaning is called *substance*. The test is grammatical: whatever you can point at with 'this' or 'that' is a dravya. And the second criterion: bhedyatvena vivaksitah — *intended as what is to be differentiated*: substance is whatever the speaker means to qualify. So dravya is defined by two features of *usage* — pronominal referability and intended qualifiability — and by nothing about what it is. This is exactly what 3.4.1's charter promised: word-meanings are science-divided posits inferred from the grammar's needs, and the grammar needs a category of things-that-get-qualified. Compare 3.2.1's list, where dravya's synonyms ran from atman and tattva to the grammarians' ipsita: here the metaphysical end of that list has dropped away entirely, and what remains is the pronoun's target. The two samuddesas on substance thus approach it from opposite ends — 3.2 from the eternal reality, 3.4 from the grammar's slot — and the Padakanda's method is to let both stand (2.179's licensed double description).
contested: the relation between this samuddesa and the Dravya-samuddesa (3.2) is a standing question: whether they treat two senses of dravya (the metaphysical and the grammatical) or one sense from two angles. The tradition's own titling — this one Dravyalaksana, 'the definition of substance' — suggests the second, and we read it so