The objection stated in two words, and it functions as a heading for the whole unit. Somebody claims there are words that are never used. The source marks it as prose, not as a vārttika, but v1 will quote it exactly, which is how you can tell it is a fixed formula.

The objector says it again at full length. Note the shape of the sentence, because the whole of the first reply will turn on it: an existence-verb and a participle of non-use, in one breath.

The four stock examples. They are quoted, not spoken, and they will be repeated at s32–s35 and again at s56–s58, so they are worth fixing in memory here.

The other side asks what the objection is supposed to prove. Granting the fact, why should a grammarian care? The next two sentences give the answer, so this question is asked in order to have the argument made explicit.

The premise the objection needs, and it is quoted back from the grammarian himself. You said usage is the authority. That was the answer lokataḥ two blocks back, and the objector is holding him to it.

The conclusion of the objection. On your own criterion, the words your rules generate but nobody says come out incorrect. The optative is doing careful work — the objector is deriving a consequence from the grammarian's premise, not asserting it on his own account.

The counter-attack, and it goes at the wording rather than the fact. The claim is said to cancel itself. Note that nothing has yet been said about whether the four examples are used; the charge is that the sentence cannot be true as it stands.

The first horn. To exist, on this reading, is to be in use — so what exists cannot be unused. Notice how much weight is being put on santi; the objector's defence at s14 will be to give it a different sense.

The second horn, and it is the first one turned round. Either way you lose one of the two words. The pair of sentences is a small dilemma, complete in six words.

The dilemma summed up. Assert both halves at once and you assert nothing. The charge is now complete, and it has been made without examining a single one of the four examples.

A second and sharper charge. You said ūṣa, tera, cakra and peca aloud in order to claim they are never said. The objection is now performative, not merely logical, and it is a stronger form of the same complaint.

The charge pressed to its limit. If anyone's usage counts as usage, yours does. So your saying the words is itself their being used. Watch sādhu change its application — everywhere else in this unit it qualifies a word-form; here it qualifies a person.

The defence opens with a flat denial, and the ground follows in the next two sentences. The way out will be to show that "exist" and "unused" are said in different respects.

The first half of the distinction. "They exist" means the rules generate them — grammarians can produce and account for these forms. Existence here is existence within the system, not currency in speech.

The second half, and one word does all the work: loke. The two predicates were never about the same field. They exist in the śāstra; they are unused in the world. So the sentence was elliptical, not self-cancelling.

The second charge is now taken up for answer, quoted in full. The bhāṣya answers objections one at a time and in order, and it signals which one it is on by quoting it.

The answer to the performative charge. The claim was never that we do not say these words. Adding asmābhiḥ shows exactly which agent was never intended.

The standing follow-up: you have said what you do not mean, now say what you do. The formula appears throughout the bhāṣya and is always answered immediately.

The answer in three words. Not unused by us; unused in the world. The two locatives, or rather the locative against the instrumental, are the whole of the defence.

The rejoinder, and it is neat. You escaped by saying "in the world". But you are in the world. So if you say them, they are used in the world after all.

The answer, and it is the last word of this exchange. Granted; but being part of the world is not being the world. What "unused in the world" denies is general currency, and one man's citing a form is not that. Whether this really disposes of the charge, the bhāṣya does not argue further — it simply moves to the vārttikas.

The first of Kātyāyana's vārttikas, and it is built on the standing three-part pattern: quote the position, refuse it, give the reason in a single ablative. This is a different voice from Patañjali's prose — terser, and with no dialogue in it. The reason it gives is that words go with meanings, and these four have meanings.

The gloss begins by adding one word to the vārttika, tat, so that the na has something to deny. Compare the treatment of the vārttika in the previous block, where kriyate was the one word supplied.

The question is put in order to have the vārttika's ablative read out as an answer. This is how the bhāṣya shows which part of an aphorism answers which question.

The vārttika's third member, now standing as the answer to a question. Nothing has been added; the aphorism has simply been laid out as a dialogue.

The compound is unpacked into a sentence. That is the whole content of this segment, and it makes the next one possible.

The argument completed. The four forms are not meaningless noises; each has a sense, and having a sense is what makes a word a word that gets used. So they belong to the class of used words even if you have not heard one lately.

The second vārttika, in two words. It concedes the fact and explains it away. These forms are indeed not heard, and the reason is that people say something else in their place. Kātyāyana leaves out both what the non-use belongs to and what "otherness of usage" means; Patañjali supplies each in turn.

The gloss supplies both what was missing and a verdict. Not merely that these forms are unused, but that their being unused is exactly right. The concession is turned into part of the answer.

The same device as s23. Question inserted so that the aphorism's reason can be given as an answer.

The vārttika's second word, standing as the answer. What it means is still not clear, and the next sentence explains it.

The compound explained. "Otherness of usage" means the meaning is expressed, but by a different form. The slot is filled; it is just filled by something else. That is a very different thing from the meaning going unexpressed.

The illustration is announced. Four examples follow, one for each of the four forms.

The first replacement shown. Instead of the perfect ūṣa people say uṣitāḥ with a pronoun. This is where you can see that ūṣa is a second person plural: its replacement carries yūyam.

The second replacement, same shape as the first. Note that the frame is shortened as the series goes on — etasya śabdasya becomes asya.

The third replacement. Here the participle is the active -vat form rather than the passive, because the sense wanted is "you did", not "was done by you".

The fourth and last. The demonstration is complete — every one of the four has a living replacement, so the meaning was never left unexpressed.

The third vārttika, and the most compressed so far — a locative and a comparison, and nothing else. It answers a different worry from v1 and v2. Even if a form really is never said, the grammar must still provide for it, and there is a precedent for providing rules for what nobody performs.

The vārttika unfolded, and the supplement is large — a concessive, an adverb, an instrumental and a gerundive. Compare v1, where only tat had to be added. The claim is now clear: the grammar's obligation to derive a form does not wait on anybody using it.

The precedent is announced.

The precedent named. Ritual sessions lasting a hundred or a thousand years. The lengths are what make them unperformable, and that is exactly why they are the right parallel.

The fact that makes the parallel work. These are rites nobody does, exactly as the four forms are words nobody says. Note the honesty of the example — it is chosen because it concedes the objector's point as fully as possible.

The analogy completed, and its mapping is exact. The ritualists are the grammarians; the unperformed sattras are the unspoken words; providing for them by śāstra is what both disciplines do. What the parallel proves is that a śāstra's coverage need not track current practice. Where it stops: the ritualists' reason is that the rite is dharma, and the bhāṣya does not here say what the corresponding reason is for words.

The fourth vārttika, and the shortest — two words. It offers a fourth defence, independent of the other three. The forms are in use; you have just not been where they are used.

The vārttika unfolded with a noun and a verb. The claim is a factual one about the geography of Sanskrit, and it will be defended with real observations at s49–s52.

The obvious reply. Show me. Nobody has actually met these forms anywhere. The objection is about evidence, and the answer to it is about where you have looked.

The answer is an imperative, not an argument. Then look harder. Absence of observation is only evidence if you have searched the field, and the next sentences will describe how large that field is.

The ground for the imperative. The field of Sanskrit usage is enormous, so a failure to find a form proves very little. The next sentence measures the field.

The measurement of the field, and it is meant to be overwhelming. Land, worlds, the Vedas and all their branches, and then the non-Vedic literature as well. The list is not glossed in the bhāṣya, and the individual items are not what matters — the length of the list is the argument.

The verdict on the objection, and note what it is and is not. It does not say the claim is false. It says it was made without doing the work. On the evidence available to anyone, the claim is not something you are entitled to assert.

The general fact behind the fourth vārttika. Words have territories. Within the whole vast field, particular forms are current in particular places and not elsewhere. Note the doubled pronouns — they are the idiom for a one-to-one distribution, this word in this place, that word in that.

The regional evidence is announced. What follows is a small piece of dialect geography.

The first regional fact, and it is a real observation about a real people. The verb is alive in one region only. That is the pattern the vārttika needed.

The other half of the observation. Elsewhere the root is not used as a verb but shows up inside a noun. The connection between the two halves is not stated and I leave it as the Sanskrit leaves it.

Three regions, three verbs for going. Each community has its own, and none of them is wrong. This is the clearest statement of the point — a form absent from your speech is current in somebody else's.

A last regional pair, and the same shape as s49–s50 — verb in one region, derived noun in another. The evidence for the vārttika is now in.

The argument now comes back to the four forms it started from. Not merely words in general: these four, the ones you named. Their use can be shown.

One word, demanding the evidence. The bhāṣya's habit of pausing for a challenge before producing its proof.

The answer in one word, and it changes the terms of the whole dispute. The forms are used, and where they are used is the Veda. So "the world" in which words are current includes the recited texts, not only conversation.

A Vedic line quoted without a source, and I can only construe it roughly. What matters for the argument is not its sense but its last word — ūṣa, the very form that was said to be unused, occurring in the Veda. That much is clear even though the line is not.

A second Vedic line, and this one construes more easily. The point again is the last word — cakra, second person plural perfect, standing in the text where the objector said it never stands. My rendering of śrutyam brahma is uncertain.

The last of the three citations, closing both the run and the unit. cakra again, in a second Vedic passage. The unit ends on evidence rather than on a formula — the forms the objector called unused have been shown in use, and the objection is answered without ever having conceded the criterion it attacked.