Book 3 - Pada-kāṇḍa (13): Liṅga-samuddeśa (On Gender)

3.13.1-2 unformalized

स्तनकेशादिसंबन्धो विशिष्टा वा स्तनादयः ।
तदुपव्यञ्जना जातिर्गुणावस्था गुणास्तथा ॥ १ ॥
शब्दोपजनितोऽर्थात्मा शब्दसंस्कार इत्यपि ।
लिङ्गानां लिङ्गतत्त्वज्ञैर्विकल्पाः सप्त दर्शिताः ॥ २ ॥

stanakeśādisaṃbandho viśiṣṭā vā stanādayaḥ |
tadupavyañjanā jātirguṇāvasthā guṇāstathā || 1 ||
śabdopajanito'rthātmā śabdasaṃskāra ityapi |
liṅgānāṃ liṅgatattvajñairvikalpāḥ sapta darśitāḥ || 2 ||

(1) The connection with breasts, hair and the rest; or the breasts and the rest as qualified; the universal manifested by those; the state of the qualities; and likewise the qualities themselves; (2) the meaning-self generated by the word; and 'refinement of the word' as well — seven alternatives concerning gender have been shown by those who know gender's reality.

The samuddesa opens with a doxographic list, and its seven entries map the whole dispute. (i) Stana-kesadi-sambandhah: the *connection with breasts, hair* and the rest — gender as the possession of secondary sexual characters: the crudest account, and the tradition's starting-point. (ii) Visista va stanadayah: or the breasts and the rest *as qualified* — not the connection but the characters themselves, specified. (iii) Tad-upavyanjana jatih: the *universal manifested* by those — femininity as a jati, of which the characters are the manifestors (3.1.29's face manifested by water). (iv) Guna-avastha: the *state of the qualities*. (v) Gunas tatha: and *likewise the qualities* themselves — gender as a quality, not a universal. (vi) Sabda-upajanito 'rtha-atma: the *meaning-self generated by the word* — gender as something the word *makes*: the boldest entry, and the one Bhartrhari's system favours (3.1.6's every word expressing its own universal first). (vii) Sabda-samskara iti api: and *'refinement of the word'* — gender as a purely morphological fact, no meaning at all. Seven, and note the drift: from bodies, through universals and qualities, to the word's own doing. The list's order is an argument.

3.13.3 unformalized

उपादानविकल्पाश्च लिङ्गानां सप्त वर्णिताः ।
विकल्पसंनियोगाभ्यां ये शब्देषु व्यवस्थिताः ॥ ३ ॥

upādānavikalpāśca liṅgānāṃ sapta varṇitāḥ |
vikalpasaṃniyogābhyāṃ ye śabdeṣu vyavasthitāḥ || 3 ||

And seven alternatives of assumption concerning genders have been described, which stand established among words by option and by conjunction.

A second seven, and it is about the words rather than the things. Upadana-vikalpas ca lingaanam sapta varnitah: seven alternatives of *assumption* concerning genders have been *described* — how a word *takes up* its gender, as against what gender is. Vikalpa-samniyogabhyam ye sabdesu vyavasthitah: which stand established among *words* by *option* and by *conjunction* — some words take gender optionally (the tradition's long lists of dvi-linga and tri-linga stems), some by fixed conjunction. So there are two sevens: seven theories of what gender is, and seven ways words get it. And the second seven's existence is already an argument for the second half of the first: if words take gender by option and conjunction, gender is at least partly a fact about words (entries vi and vii).

3.13.4 unformalized

तिस्रो जातय एवैताः केसां चित्समवस्थिताः ।
अविरुद्धा, विरुद्धाभिर्गोमहिष्यादिजातिभिः ॥ ४ ॥

tisro jātaya evaitāḥ kesāṃ citsamavasthitāḥ |
aviruddhā, viruddhābhirgomahiṣyādijātibhiḥ || 4 ||

For some, these are just three universals, standing established — not in conflict with the mutually conflicting universals of cow, buffalo and the rest.

The jati-account (entry iii) stated. Tisro jataya eva etah kesamcit samavasthitah: for *some*, these are just *three universals* — femininity, masculinity, neutrality, each a jati. Aviruddha viruddhabhih go-mahisyadi-jatibhih: *not in conflict* with the *mutually conflicting* universals of *cow*, *buffalo* and the rest. The point is careful: cowhood and buffalohood are mutually exclusive — nothing is both — but femininity is not exclusive of either; a cow is female *and* a cow. So the gender-universals cross-cut the species-universals, which is unusual: most jatis nest or exclude. Hence gender-universals are a special class, pervading the others (3.1.44's brahminhood standing present in all living beings). The account is coherent and its cost is a second, orthogonal system of universals.

3.13.5 unformalized

हस्तिन्यां वडवायां च स्त्रीति बुद्धेः समन्वयः ।
अतस्तां जातिमिच्छन्ति द्रव्यादिसमवायिनीम् ॥ ५ ॥

hastinyāṃ vaḍavāyāṃ ca strīti buddheḥ samanvayaḥ |
atastāṃ jātimicchanti dravyādisamavāyinīm || 5 ||

In a cow-elephant and in a mare, there is a concomitance of the cognition 'female'; therefore they desire that universal, inhering in substance and the rest.

The argument for 3.13.4. Hastinyam vadavayam ca stri iti buddheh samanvayah: in a *cow-elephant* and in a *mare*, there is a *concomitance* of the *cognition 'female'* — the same cognition arises across species: one recognizes femaleness in an elephant and in a horse, and the recognition is one. Samanvaya: concomitance, the running-together that is a universal's mark (3.1.14's anuvrtti, recurrence). Atas tam jatim icchanti dravyadi-samavayinim: *therefore* they desire that *universal*, *inhering in substance* and the rest. So the recurrent cognition evidences a recurrent universal, by the standard inference. The argument is the tradition's best for gender-realism, and its evidence is genuine: femaleness *is* recognized across kinds, and that is what universals explain.

3.13.6 unformalized

परतन्त्रस्य यल् लिङ्गमपोद्धारे विवक्षिते ।
तत्रासौ शब्दसंस्कारः शब्दैरेव व्यपाश्रितः ॥ ६ ॥

paratantrasya yal liṅgamapoddhāre vivakṣite |
tatrāsau śabdasaṃskāraḥ śabdaireva vyapāśritaḥ || 6 ||

That gender which belongs to what is dependent, when abstraction is intended — there that 'refinement of the word' rests upon words alone.

The seventh entry (sabda-samskara) placed. Paratantrasya yal lingam apoddhare vivaksite: that *gender* which belongs to what is *dependent* — the adjective's gender, which is not its own but taken from its noun — *when abstraction is intended*: when the analyst extracts the word (3.1.1's apoddhara). Tatra asau sabda-samskarah sabdaih eva vyapasritah: *there* that *'refinement of the word'* *rests upon words alone*. So the dependent word's gender is purely morphological — 'sukla' is feminine in 'sukla gauh' because 'gauh' is, and nothing about whiteness is female. The seventh account is thus certainly right for *some* cases: adjectival gender is agreement, and agreement is a fact about words. The question is how far the concession runs — and 3.13.7 extends it.

3.13.7 unformalized

बुद्ध्या कल्पितरूपेषु लिङ्गेष्वपि च संभवः ।
स्त्रीत्वादीनां व्यवस्था हि सा लिङ्गैर्व्यपदिश्यते ॥ ७ ॥

buddhyā kalpitarūpeṣu liṅgeṣvapi ca saṃbhavaḥ |
strītvādīnāṃ vyavasthā hi sā liṅgairvyapadiśyate || 7 ||

And there is a possibility even in genders whose form is constructed by the intellect; for that settled arrangement of femininity and the rest is designated by the genders.

The concession extended. Buddhya kalpita-rupesu lingesv api ca sambhavah: there is a *possibility* even in *genders whose form is constructed by the intellect* — genders assigned to what has no sex: 'tata' (the shore, masculine), 'nadi' (the river, feminine), 'kula' (the bank, neuter) — the same water's edge in three genders. Stritvadinam vyavastha hi sa lingair vyapadisyate: for that *settled arrangement* of femininity and the rest *is designated by the genders*. So where there is no sex, the gender is buddhi-constructed and the endings designate the construction. And since most Sanskrit nouns name sexless things, the constructed case is the *normal* one. This is the samuddesa's decisive turn: 3.13.5's cross-species femaleness is real and rare; gender's ordinary employment has nothing to do with it.

3.13.8-9 unformalized

यथा सलिलनिर्भासा मृगतृष्णासु जायते ।
जलोपलब्ध्यनुगुणाद्बीजाद्बुद्धिर्जलेऽसति ॥ ८ ॥
तथैवाव्यपदेश्येभ्यो हेतुभ्यस्तारकादिषु ।
मुख्येभ्य इव लिङ्गेभ्यो भेदा लोके व्यवस्थिताः ॥ ९ ॥

yathā salilanirbhāsā mṛgatṛṣṇāsu jāyate |
jalopalabdhyanuguṇādbījādbuddhirjale'sati || 8 ||
tathaivāvyapadeśyebhyo hetubhyastārakādiṣu |
mukhyebhya iva liṅgebhyo bhedā loke vyavasthitāḥ || 9 ||

(8) As an appearance of water arises in mirages — a cognition arising, where there is no water, from a seed conducive to the apprehension of water; (9) just so, from indescribable causes, differences among stars and the like stand established in the world, as though from primary genders.

The mirage, and it is the samuddesa's central image. (8): salila-nirbhasa mrga-trsnasu jayate — an *appearance of water* arises in *mirages*; jala-upalabdhy-anugunad bijad buddhir jale asati: a *cognition* arising, *where there is no water*, from a *seed* conducive to the apprehension of water. Bija: the seed, the trace (1.84's ahita-bija; 3.9.39's deposits). So the mirage's water comes from a disposition in the seer, not from the desert. (9) applies it: tatha eva avyapadesyebhyo hetubhyah taraka-adisu — just so, from *indescribable causes*, among *stars* and the like — bheda loke vyavasthitah: *differences* stand established *in the world* — mukhyebhya iva lingebhyah: *as though* from *primary* genders. So 'taraka' is feminine and 'naksatra' neuter, and the stars have no sex: the gender-differences arise from indescribable causes as the mirage's water does, and present *as though* primary. The iva is the doctrine. Gender, in its ordinary employment, is a mirage — real as appearance, seeded in the speaker, absent from the thing.

3.13.10 unformalized

व्यक्तेषु व्यक्तरूपाणां स्तनादीनां तु दर्शनात् ।
अव्यक्तव्यञ्जनाव्यक्तेर्जातिर्न परिकल्प्यते ॥ १० ॥

vyakteṣu vyaktarūpāṇāṃ stanādīnāṃ tu darśanāt |
avyaktavyañjanāvyakterjātirna parikalpyate || 10 ||

But from the seeing of breasts and the rest, whose forms are manifest, among manifest things — from the non-manifestation of the unmanifest manifestor, a universal is not constructed.

An argument against the jati-account. Vyaktesu vyakta-rupanam stanadinam tu darsanat: from the *seeing* of *breasts* and the rest, *whose forms are manifest*, among *manifest* things — where there are sexual characters, they are plainly visible. Avyakta-vyanjana-avyakteh jatir na parikalpyate: from the *non-manifestation* of the *unmanifest manifestor*, a *universal* is *not constructed*. The argument: a jati needs manifestors (3.1.29's face in water); where the characters are absent — in stars, rivers, shores — nothing manifests femininity, so no universal can be posited there. And most nouns are of that kind. So the jati-account works exactly where the characters are and fails everywhere else — which makes it an account of sex, not of gender. The samuddesa is separating the two, and 3.13.5's cross-species recognition is being confined to its proper domain.

3.13.11 unformalized

अस्तित्वं च प्रतिज्ञाय सदादर्शनमिच्छतः ।
अत्यन्तादर्शने न स्यादसत्त्वं प्रति निश्चयः ॥ ११ ॥

astitvaṃ ca pratijñāya sadādarśanamicchataḥ |
atyantādarśane na syādasattvaṃ prati niścayaḥ || 11 ||

And for one who, having asserted existence, desires that it is always unseen — in utter non-seeing there would be no ascertainment as to non-existence.

The realist's escape blocked. The jati-realist might say: femininity is present in the river too, merely unmanifest. Astitvam ca pratijnaya sada adarsanam icchatah: for one who, having *asserted existence*, *desires* that it is *always unseen* — the universal is there and never shows. Atyanta-adarsane na syad asattvam prati niscayah: in *utter non-seeing* there would be *no ascertainment* as to *non-existence*. The reply is double-edged and Bhartrhari lets both edges show: if the universal is never seen, one cannot establish that it is absent — and equally, one cannot establish that it is present. An always-unseen posit is unfalsifiable and unconfirmable, which is not a defence but a diagnosis. The escape saves the thesis by making it idle.

3.13.12 unformalized

न चालमनुमानाय शब्दोऽदर्शनपूर्वकः ।
सिद्धे हि दर्शने किं स्यादनुमानप्रयोजनम् ॥ १२ ॥

na cālamanumānāya śabdo'darśanapūrvakaḥ |
siddhe hi darśane kiṃ syādanumānaprayojanam || 12 ||

And a word preceded by non-seeing is not sufficient for an inference; for when the seeing is established, what would be the purpose of an inference?

The realist's second escape blocked. He might infer the unseen universal from the word: 'nadi' is feminine, therefore femininity is there. Na ca alam anumanaya sabdo 'darsana-purvakah: a *word* preceded by *non-seeing* is *not sufficient* for an *inference* — a word cannot evidence what nobody has ever seen; inference runs from a mark to what has been observed with it (the vyapti requires prior co-observation). Siddhe hi darsane kim syad anumana-prayojanam: for when the *seeing is established*, what would be the *purpose* of an inference? Where the characters show, one sees and does not infer; where they do not, no inference is available. So the word cannot bridge to the universal, and the realist has neither perception nor inference for gender in the sexless case. This closes the argument begun at 3.13.10: gender-as-jati has no evidence outside the bodies.

3.13.13 unformalized

आविर्भावस्तिरोभावः स्थितिश्चेत्यनपायिनः ।
धर्मा मूर्तिषु सर्वासु लिङ्गत्वेनानुदर्शिताः ॥ १३ ॥

āvirbhāvastirobhāvaḥ sthitiścetyanapāyinaḥ |
dharmā mūrtiṣu sarvāsu liṅgatvenānudarśitāḥ || 13 ||

Manifestation, disappearance, and standing — these unfailing properties in all corporeal things have been exhibited as gender.

The samuddesa's own account begins, and it is startling. Avirbhavas tirobhavah sthitis ca iti anapayinah: *manifestation*, *disappearance*, and *standing* — the three that 3.8.25-26 gave as the vikaras' reduction, and 3.9.11's moments of a thing's career. Dharma murtisu sarvasu linga-tvena anudarsitah: these *unfailing properties* in *all corporeal things* have been *exhibited as gender*. So the three genders are these three: feminine is manifestation (the coming-forth), masculine is standing, neuter is disappearance — or some such mapping; the verse does not assign them, and the tradition's assignment varies. What matters is the move: gender is not sex but *phase* — the universal three-beat of arising, abiding, ceasing, which every corporeal thing exhibits. This is why every noun has a gender: everything has the three, and the ending marks which is salient.

3.13.14 unformalized

सर्वमूर्त्यात्मभूतानां शब्दादिनां गुणे गुणे ।
त्रयः सत्त्वादिधर्मास्ते सर्वत्र समवस्थिताः ॥ १४ ॥

sarvamūrtyātmabhūtānāṃ śabdādināṃ guṇe guṇe |
trayaḥ sattvādidharmāste sarvatra samavasthitāḥ || 14 ||

Of sound and the rest, which are the very selves of all corporeal things — in each quality, those three properties beginning with 'sattva' stand established everywhere.

The three grounded in the Samkhya gunas. Sarva-murty-atma-bhutanam sabdadinam gune gune: of *sound* and the rest — the five sense-qualities, which are the *very selves* of all corporeal things (nothing is but its qualities: 3.11.7's para-upakara-tattvanam) — *in each quality*. Trayah sattvadi-dharmas te sarvatra samavasthitah: those *three* properties beginning with *'sattva'* stand established *everywhere* — sattva, rajas, tamas: the three gunas of Samkhya, present in every quality of everything. So 3.13.13's three phases are the three gunas: manifestation is sattva's, standing is rajas', disappearance is tamas'. And since the gunas are in everything, the three genders are in everything — which explains why every noun has one, and why the assignment can seem arbitrary: all three are always present, and the ending marks whichever the word fixes on.
contested: the mapping of the three genders onto the three gunas is implied rather than stated, and the tradition's assignments differ; we set out the structure the verses build and decline to fix the correspondence the text leaves open

3.13.15 unformalized

रूपस्य चात्ममात्रानां शुक्लादिनां प्रतिक्षणम् ।
का चित्प्रलीयते का चित्कथं चिदभिवर्धते ॥ १५ ॥

rūpasya cātmamātrānāṃ śuklādināṃ pratikṣaṇam |
kā citpralīyate kā citkathaṃ cidabhivardhate || 15 ||

And of the mere selves of form — white and the rest — at every moment, some one dissolves away and some one somehow increases.

The three phases at their finest grain. Rupasya ca atma-matranam suklaadinam prati-ksanam: of the *mere selves of form* — *white* and the rest — *at every moment*. Ka cit praliyate ka cit katham cid abhivardhate: *some one* *dissolves away* and *some one* *somehow increases*. So even a colour is, at every instant, partly waning and partly waxing — the three phases running continuously in the smallest quality. Katham cit — *somehow*: the shrug (2.171, 3.7.34, 3.8.62's kathamcit), marking that the mechanism gives out. And prati-ksanam, at every moment: this is momentariness in a grammarian's treatise, and it is offered not as Buddhist doctrine but as what the guna-analysis requires. If the three gunas are in every quality and they are phases, then every quality is always arising and ceasing — and gender marks which phase a word fixes on.

3.13.16 unformalized

क्वथितोदकवच्चैषामनवस्थितवृत्तिता ।
अजस्रं सर्वभावानां भाष्य एवोपवर्णिता ॥ १६ ॥

kvathitodakavaccaiṣāmanavasthitavṛttitā |
ajasraṃ sarvabhāvānāṃ bhāṣya evopavarṇitā || 16 ||

And their unsettled operation, like boiling water — ceaselessly, of all entities — has been described in the Bhasya itself.

The momentariness of 3.13.15 given its image and its authority. Kvathita-udakavat ca esam anavasthita-vrttita: their *unsettled operation*, *like boiling water* — the pot at a rolling boil, where nothing holds still and the whole is a churn of arising and subsiding. Ajasram sarva-bhavanam: *ceaselessly*, of *all entities* — no thing is exempt, and no moment. Bhasya eva upavarnita: has been described *in the Bhasya itself*. So Patanjali said it, and Bhartrhari is drawing it out rather than importing it (2.475's bhasye eva upavarnitah; 3.9.103's Bhasya-citation). The image is chosen well: boiling water is unmistakably one body of water and unmistakably never the same twice — which is exactly the standing-and-not-standing that gender's three phases (3.13.13) mark.

3.13.17 unformalized

प्रवृत्तेरेकरूपत्वं साम्यं वा स्थितिरुच्यते ।
अविर्भावतिरोभावप्रवृत्त्या वावतिष्ठते ॥ १७ ॥

pravṛtterekarūpatvaṃ sāmyaṃ vā sthitirucyate |
avirbhāvatirobhāvapravṛttyā vāvatiṣṭhate || 17 ||

Uniformity of operation, or evenness, is called 'standing'; or it stands established by an operation of manifestation-and-disappearance.

What sthiti — standing — amounts to on the boiling-water picture. Pravrtter eka-rupatvam samyam va sthitir ucyate: *uniformity of operation*, or *evenness*, is called *'standing'* — a thing 'stands' when its arising and ceasing are in balance: not because nothing happens but because the churn is steady. Avirbhava-tirobhava-pravrttya va avatisthate: *or* it stands established by an *operation of manifestation-and-disappearance*. So standing is not a third thing beside manifesting and disappearing; it is those two in equilibrium (3.8.26's janma eva asrita-sarupyam sthitih iti abhidhiyate: birth alone, having resorted to a likeness of form, is spoken of as standing). The middle of the three phases dissolves into the other two, and what looked like a triad is a pair with a balance-point.

3.13.18 unformalized

गुणा इत्येव बुद्धेर्वा निमित्तत्वं स्थितिर्मता ।
स्थितेश्च सर्वलिङ्गानां सर्वनामत्वमुच्यते ॥ १८ ॥

guṇā ityeva buddhervā nimittatvaṃ sthitirmatā |
sthiteśca sarvaliṅgānāṃ sarvanāmatvamucyate || 18 ||

Or 'standing' is held to be the state of being the occasioning ground for the cognition 'qualities' merely; and from standing, all genders are said to have the state of being pronouns.

A second account of standing, and a startling consequence. Guna iti eva buddher va nimittatvam sthitir mata: 'standing' is held to be the state of being the *occasioning ground* for the cognition '*qualities*' merely — a thing stands just insofar as it occasions the thought 'here are qualities'. And then: sthiteh ca sarva-lingaanam sarva-namatvam ucyate — from *standing*, *all genders* are said to have the state of being *pronouns*. Sarvanaman: the pronoun, Panini's own category — and the pronouns are precisely the words that take all three genders (sa, sa, tat). So gender's ground in standing explains why the pronouns are gender-indifferent: what merely stands, occasioning the thought of qualities, is not yet any gender, and the pronoun names exactly that (3.4.3's dravya defined by pronominal referability: whatever you can point at with 'this').

3.13.19 unformalized

स्थितेषु सर्वलिङ्गेषु विवक्षानियमाश्रयः ।
कस्य चिच्छब्दसंस्कारे व्यापारः क्व चिदिष्यते ॥ १९ ॥

sthiteṣu sarvaliṅgeṣu vivakṣāniyamāśrayaḥ |
kasya cicchabdasaṃskāre vyāpāraḥ kva cidiṣyate || 19 ||

All genders standing established, the operation of some one — resting on the restriction of intention — is somewhere accepted in the refinement of the word.

The samuddesa's central mechanism. Sthitesu sarva-lingesu: *all* genders *standing established* — every thing has all three, since all three gunas are in everything (3.13.14). Vivaksa-niyamasrayah kasya cid vyaparah: the *operation* of *some one* — resting on the *restriction of intention*. Kva cid isyate sabda-samskare: is *somewhere accepted* in the *refinement of the word*. So: all three genders are present, vivaksa restricts to one, and the restriction shows as the word's morphological form. Sabda-samskara — the seventh entry of 3.13.1-2's list, now adopted as the account. Gender is where the speaker's selection among ever-present phases gets into the word — which is 2.400's directed eye (the word, when intended, becomes the expressor of a meaning) applied to what the endings mark.

3.13.20 unformalized

संनिधाने निमित्तानां किं चिदेव प्रवर्तकम् ।
यथा तक्षादिशब्दानां लिन्गेषु नियमस्तथा ॥ २० ॥

saṃnidhāne nimittānāṃ kiṃ cideva pravartakam |
yathā takṣādiśabdānāṃ lingeṣu niyamastathā || 20 ||

When the occasioning grounds are in proximity, some one alone is the impeller — as with the words 'taksan' and the rest, so is the restriction in respect of genders.

The selection of 3.13.19 exemplified. Sannidhane nimittanam kim cid eva pravartakam: when the *occasioning grounds* are in *proximity*, *some one alone* is the *impeller* — many conditions are present and one operates (3.7.119's sa-vyaparatara: one property understood as more operative; 3.8.39's kascid pradhanyena pratiyate). Yatha taksadi-sabdanam lingesu niyamas tatha: as with the words *'taksan'* (carpenter) and the rest, so is the *restriction* in respect of *genders*. The taksadi class is Panini's list of stems with fixed gender-behaviour — the same word, one gender, no option, and no reason in the thing. So gender-selection works as every selection in this kanda works: a crowd of available conditions, one operative, and the operating not derivable from the crowd (3.12.14's niyatah sabda-saktayah: the capacities of words are fixed).

3.13.21 unformalized

भावतत्त्वदृशः शिष्टाः शब्दार्थेषु व्यवस्थिताः ।
यद्यद्धर्मेऽङ्गतामेति लिङ्गं तत्तत्प्रचक्षते ॥ २१ ॥

bhāvatattvadṛśaḥ śiṣṭāḥ śabdārtheṣu vyavasthitāḥ |
yadyaddharme'ṅgatāmeti liṅgaṃ tattatpracakṣate || 21 ||

The cultivated, who see the reality of entities, standing in the meanings of words — whatever gender comes to be a limb in respect of whatever property, that they declare.

The authority for the assignments. Bhava-tattva-drsah sistah: the *cultivated*, who *see the reality of entities* — the sistas, whose usage is the criterion of correctness (1.27's sadhutva established from the cultivated; 1.141's aviccheda). Sabdarthesu vyavasthitah: *standing in the meanings of words*. Yad yad dharme 'ngatam eti lingam tat tat pracaksate: *whatever* gender comes to be a *limb* in respect of *whatever* property, *that* they *declare*. Angata: the Mimamsa relation of subsidiary to principal (2.81-82). So the sistas assign the gender that serves the property in view — and their assignment is the fact. There is no appeal past it, because gender is a sabda-samskara (3.13.19) and the cultivated are who refine words. The verse locates gender's authority exactly where Kanda I located all linguistic authority.

3.13.22 unformalized

स्वरभेदाद्यथा शब्दाः साधवो विषयान्तरे ।
लिङ्गभेदात्तथा सिद्धात्साधुत्वमनुगम्यते ॥ २२ ॥

svarabhedādyathā śabdāḥ sādhavo viṣayāntare |
liṅgabhedāttathā siddhātsādhutvamanugamyate || 22 ||

As words are correct in another domain by a difference of accent — so, from an established difference of gender, correctness is followed along.

The parallel that settles gender's status. Svara-bhedad yatha sabdah sadhavo visayantare: as *words* are *correct* in *another domain* by a *difference of accent* — the same syllables, differently accented, are correct in different senses (and the Indra-satru of 3.10.5 is the tradition's monument to what accent decides). Linga-bhedat tatha siddhat sadhutvam anugamyate: so, from an *established difference of gender*, *correctness* is *followed along*. So gender is like accent: a formal feature that fixes correctness domain by domain, learned rather than derived, and carrying no meaning of its own beyond the fixing. This is the samuddesa's cleanest placement of its subject — and the accent-parallel is exact, since nobody supposes an accent means anything.

3.13.23 unformalized

प्रयोगो विप्रयोगश्च लोके यत्रोपलभ्यते ।
शास्त्रमारभ्यते तत्र न प्रयोगाविपर्यये ॥ २३ ॥

prayogo viprayogaśca loke yatropalabhyate |
śāstramārabhyate tatra na prayogāviparyaye || 23 ||

Where employment and non-employment are found in the world, there the science is undertaken — not where there is no inversion of employment.

Why the grammar legislates gender at all. Prayogo viprayogas ca loke yatra upalabhyate: where *employment* and *non-employment* are *found in the world* — where usage varies, some saying one thing and some another. Sastram arabhyate tatra: *there* the *science is undertaken*. Na prayoga-aviparyaye: *not* where there is *no inversion* of employment — where usage is uniform, no rule is needed. So the grammar's gender-rules exist precisely where the sistas' usage wavers, and are absent where it does not. This is a fine account of what a rule-system is for: it is a record of the contested cases (2.343's vyavahara constructed by division for the sake of the operation), and its silence marks agreement rather than oversight — which is why 3.12.24's argument from the Bhasya's silence had to be answered.

3.13.24 unformalized

उपाधिभेदादर्थेषु गुणधर्मस्य कस्य चित् ।
निमित्तभावः साधुत्वे विवक्षा च व्यवस्थिता ॥ २४ ॥

upādhibhedādartheṣu guṇadharmasya kasya cit |
nimittabhāvaḥ sādhutve vivakṣā ca vyavasthitā || 24 ||

From a difference of adjunct among meanings, the state of being the occasioning ground, in correctness, belongs to some quality-property; and intention stands established.

The two variables that fix gender. Upadhi-bhedad arthesu: from a *difference of adjunct* among *meanings* — upadhi: the limiting condition (3.2.2's asatya-upadhi; 3.9.36's kriya-upadhi). Guna-dharmasya kasya cin nimitta-bhavah sadhutve: the state of being the *occasioning ground*, in *correctness*, belongs to *some quality-property*. Vivaksa ca vyavasthita: and *intention* stands established. So gender-correctness has two conditions: which quality-property is in view (fixed by the adjunct) and what the speaker intends. Neither is a fact about the thing's sex. And 3.13.25 will show both working on a single pair of words.

3.13.25 unformalized

हिमारण्ये महत्त्वेन युक्ते स्त्रीत्वमवस्थितम् ।
ह्रस्वोपाधिविशिष्टायाः कुट्याः प्रसवयोगिता ॥ २५ ॥

himāraṇye mahattvena yukte strītvamavasthitam |
hrasvopādhiviśiṣṭāyāḥ kuṭyāḥ prasavayogitā || 25 ||

In 'hima' and 'aranya', femininity stands established when joined with greatness; and of a hut, distinguished by the adjunct 'small', there is the fitness for the productive suffix.

The two variables of 3.13.24 exhibited. Himaranye mahattvena yukte stritvam avasthitam: in *'hima'* (snow) and *'aranya'* (forest), *femininity* stands established when *joined with greatness* — himani, aranyani: the feminine forms mean the *great* snow, the *great* forest (Panini 4.1.49's himaranyayor mahattve). So femininity here marks magnitude, and has nothing to do with sex. Hrasva-upadhi-visistayah kutyah prasava-yogita: and of a *hut*, distinguished by the *adjunct 'small'*, there is the *fitness for the productive suffix* — kuti, the little hut, taking the feminine for smallness. So the *same* feminine marks greatness in one word and smallness in another, and both are correct. Nothing could show more plainly that gender is a sabda-samskara: one ending, opposite significations, and usage settling each.

3.13.26 unformalized

शब्दान्तरानां भिन्नेऽर्थ उपायाः प्रतिपत्तये ।
एकतामिव निश्चित्य लघ्वर्थमुपदर्शिताः ॥ २६ ॥

śabdāntarānāṃ bhinne'rtha upāyāḥ pratipattaye |
ekatāmiva niścitya laghvarthamupadarśitāḥ || 26 ||

The means, for comprehension, of other words in a different meaning — having determined as though a oneness — have been shown for the sake of brevity.

A methodological note on the grammar's gender-classes. Sabdantaranam bhinne 'rtha upayah pratipattaye: the *means*, for *comprehension*, of *other words* in a *different meaning*. Ekatam iva niscitya: having *determined as though a oneness* — the grammar groups words of different senses into one gender-class as though they were one kind. Laghv-artham upadarsitah: have been shown *for the sake of brevity* — laghava, the Paninian virtue (2.343's laghu; 2.176's instruction by the general for brevity's sake). So the gender-classes are economy-devices: a list of stems treated as one because listing them severally would be endless, and the iva marks that their oneness is the list's and not theirs. This is 3.7.78's method (one karaka exhibited by way of difference) inverted: many things exhibited as one, for the same reason.

3.13.27 unformalized

उत्पत्तिः प्रसवोऽन्येषां नाशः संस्त्यानमित्यपि ।
आत्मरूपं तु भावानां स्थितिरित्यपदिश्यते ॥ २७ ॥

utpattiḥ prasavo'nyeṣāṃ nāśaḥ saṃstyānamityapi |
ātmarūpaṃ tu bhāvānāṃ sthitirityapadiśyate || 27 ||

For others, 'origination' is production, 'destruction' is congealing; but the own-form of entities is designated as 'standing'.

A rival naming of the three phases. Utpattih prasavo 'nyesam: for *others*, *'origination'* is *production* — prasava: bringing-forth, the word 3.6.10 used for what direction-words lack (no engendering). Nasah samstyanam iti api: *'destruction'* is *congealing* — samstyana: thickening, coagulating; destruction as a hardening rather than a vanishing (3.7.118's samurcchita iva: the meaning's self as though congealed). Atma-rupam tu bhavanam sthitir ity apadisyate: but the *own-form* of entities is designated as *'standing'*. So the three are renamed: production, congealing, and the thing's own form. And the middle term is now the thing itself — standing is not a phase alongside the others but what the entity *is*, with arising and ceasing as its edges. The renaming shifts the doctrine's weight: gender marks the thing's own form and its two boundaries.

3.13.28 unformalized

दृष्टं निमित्तं केसां चिज्जात्यादिवदवस्थितम् ।
दृष्टवच्छब्दसंस्कारमात्रं तु परिकल्पितम् ॥ २८ ॥

dṛṣṭaṃ nimittaṃ kesāṃ cijjātyādivadavasthitam |
dṛṣṭavacchabdasaṃskāramātraṃ tu parikalpitam || 28 ||

For some, the occasioning ground is seen as standing established, like a universal and the rest; but a mere refinement of the word, as though seen, is constructed.

The two camps set side by side, and the samuddesa's verdict is in the second clause. Drstam nimittam kesam cij jaty-adivad avasthitam: for *some*, the *occasioning ground* is *seen* as *standing established*, *like a universal* and the rest — the realists (3.13.4-5's jati-account): femininity is there and shows. Drstavat sabda-samskara-matram tu parikalpitam: but a *mere refinement of the word*, *as though seen*, is *constructed*. Drstavat: *as though* seen — the sabda-samskara is a construction that wears the look of an observation. So the two accounts are not two readings of one datum: on the second, the datum itself is manufactured, and the realists are reading a construction as a finding. Parikalpitam: constructed (3.13.9's mirage). And matram — *mere*: nothing but a refinement of the word.

3.13.29 unformalized

यथा प्रसिद्धेऽप्येकत्वे नानात्वाभिनिवेशिनः ।
नानात्वं जनयन्तीव शब्दा लिङ्गेऽपि स क्रमः ॥ २९ ॥

yathā prasiddhe'pyekatve nānātvābhiniveśinaḥ |
nānātvaṃ janayantīva śabdā liṅge'pi sa kramaḥ || 29 ||

As, even where oneness is well established, words — for those addicted to manifoldness — as it were generate manifoldness; that same procedure is in gender too.

The mechanism named. Yatha prasiddhe 'py ekatve nanatva-abhinivesinah: as, even where *oneness* is *well established*, for those *addicted to manifoldness* — abhinivesin: clinging, obsessed (2.266's gauna-artha-abhinivesinam, settling upon a secondary meaning). Nanatvam janayanti iva sabdah: *words* — *as it were* — *generate manifoldness*. Linge 'pi sa kramah: *that same procedure* is in *gender* too. So words manufacture plurality where there is one thing, for minds disposed to find it — and gender is that manufacture. The iva keeps it honest: words do not really generate; they present. And nanatva-abhinivesinah is a diagnosis of the hearer, not of the language: the manifoldness is generated *for those addicted to it*, which is everyone in vyavahara (3.6.18's dik-kala-parikalpana standing like consciousness itself).

3.13.30 unformalized

इदं वेयमयं वेति शब्दसंस्कारमात्रकम् ।
निमित्तदर्शनादर्थे कैश्चित्सर्वत्र वर्ण्यते ॥ ३० ॥

idaṃ veyamayaṃ veti śabdasaṃskāramātrakam |
nimittadarśanādarthe kaiścitsarvatra varṇyate || 30 ||

'This' or 'she' or 'he' — a mere refinement of the word: from the seeing of an occasioning ground in the meaning, it is described by some as everywhere so.

The doctrine at its most general. Idam va iyam ayam va iti: '*this*' or '*she*' or '*he*' — the three pronominal genders, and note that the neuter comes first. Sabda-samskara-matrakam: a *mere refinement of the word* — matraka: nothing but. Nimitta-darsanad arthe kaiscit sarvatra varnyate: *from the seeing of an occasioning ground in the meaning*, it is described by *some* as *everywhere* so. So: some hold that gender is a sabda-samskara *everywhere* — not merely in the sexless cases (3.13.7's rivers and shores) but in all cases, including the bodies. Even 'she' said of a woman is a refinement of the word, occasioned by something seen in the meaning and not naming it. This is the seventh entry of 3.13.1-2 taken to its limit, and it is reported (kaiscit) with evident sympathy: the samuddesa's whole argument has been driving here.

3.13.31 unformalized

नावश्यं विषयत्वेन निमित्तं व्यवतिष्ठते ।
इन्द्रियादि यथादृष्टं भेदहेतुस्तदिष्यते ॥ ३१ ॥

nāvaśyaṃ viṣayatvena nimittaṃ vyavatiṣṭhate |
indriyādi yathādṛṣṭaṃ bhedahetustadiṣyate || 31 ||

The occasioning ground does not necessarily stand established as an object; as the sense-faculty and the rest, being unseen, is accepted as the cause of difference.

The samuddesa's last defence of 3.13.30. The objection: if gender is a refinement of the word occasioned by something in the meaning, that something must be an object one could point to — and in rivers there is none. The reply: na avasyam visayatvena nimittam vyavatisthate — the *occasioning ground* does *not necessarily* stand established *as an object*. And the precedent: indriyadi yatha adrstam bheda-hetus tad isyate — as the *sense-faculty* and the rest, being *unseen*, is accepted as the *cause of difference*. This is 3.1.30-31's argument exactly: the eye's condition differs and is never itself perceived, and it makes my seeing differ from yours. Ungrasped, it conditions. So gender's occasioning ground may be real and never an object — like the eye, working by not appearing. The realist demanded a visible femininity in the river; the reply is that causes need not be objects, and the finest ones never are.