Mimamsasutra 3.2

He composes the first topical section of the third quarter: / Regarding dharma, is the tradition of the Aṣṭakā and the like not valid cognition, or is it valid cognition? / It is not a means of valid cognition because it lacks a foundation; if it states the sense of the Veda, it is useless.

Because it is remembered by those who follow the Veda, its Vedic foundation is conceivable; / because it has purpose by condensing dispersed contents, validity belongs [to it].

Tradition statements such as "The Aṣṭakā rites are to be performed" are not a means of valid cognition regarding dharma, because, being human statements, they lack a foundational means of valid cognition, like the statement of a deceiver. Now, to have a foundational means of valid cognition, if it is only the meaning of the Veda, tradition would be useless. Then, since it is merely reiterative, it lacks validity. When this is obtained, we say: Tradition under discussion has the Veda as its foundation, because it is a tradition composed by Vedic sages such as Manu, like the tradition on initiation, study, and so forth. Nor should uselessness be suspected. For the rites to be performed, dispersed across various Vedas directly accessible or inaccessible to persons like us, are condensed in one place. Therefore, this tradition is a means of valid cognition regarding dharma. 1.

In this very [section], he states the prima facie view and conclusion according to another opinion: / The mantra song "Whom the people..." is not valid, because it is an auxiliary to the traditional Aṣṭakā. / Not so: since validity is possible when a foundational Veda for the tradition is inferred.

This mantra, "Whom the people greet with joy...", is an auxiliary to the Aṣṭakā-śrāddha. And that śrāddha is traditional. For we do not find a direct Vedic statement teaching it. Therefore, if it is argued that this mantra statement is not a means of valid cognition regarding dharma, no; because its foundational Veda is inferable. And the inference has been demonstrated. Therefore, that mantra is, regarding dharma,

a means of valid cognition.

(Aphorism in the second topical section on the superiority of direct Vedic statement) / But in the case of conflict, it should be disregarded, for [an inference exists only] when there is no [conflict].

He composes the second topical section: / Is this tradition, "The entire Audumbarī [post] must be wrapped," valid cognition / or not valid cognition? In this doubt, it would be valid cognition, like [that of] the Aṣṭakā and so forth.

Because of the direct Veda: "Touching the Audumbarī [post], he should chant," / on account of conflict, since a foundational Veda is not inferred, it lacks validity.

In the Jyotiṣṭoma, a certain branch of the Udumbara tree is driven into the middle of the hall called the sadas. And wrapping it with cloth is remembered in tradition: "The entire Audumbarī [post] must be wrapped." That tradition is a means of valid cognition, because, as with traditions of the Aṣṭakā and so forth, it is possible to infer a foundational Veda. When this is obtained,

we say: Touching is enjoined by the direct Vedic statement: "Having touched the Audumbarī [post], he should chant the Udgītha." And if it is entirely wrapped, touching is not possible. Therefore, the inference of a foundational Veda is contradicted by direct perception. Therefore, like the statement of a deceiver, the baseless tradition is not a means of valid cognition. 4.

In this very topical section, he states another opinion: / Or rather, because a contradiction is seen between the direct Vedic statement and the inferred one, / when non-validity is suspected, the sublation of the inference is described here.

If a Veda directly perceived by another is the foundation for wrapping here, / let it be so; even so, performance is in accordance with one's own direct perception.

The prima facie view is that, because of mutual conflict between the direct and inferred Vedic statements regarding touching and complete wrapping, neither is a means of valid cognition. / Since the inference is contradicted by direct perception and there is no conflicting direct Vedic statement, the Vedic statement on touching is a means of valid cognition regarding its own meaning. If it is said that a Veda directly perceived by another person is the foundation of the tradition of complete wrapping, then let it not lack validity. Even so, because one's own direct perception is superior to direct perception by another, only touching is to be performed here, not complete wrapping. 6.

(Aphorism in the third topical section on the non-validity of tradition based on visible motives) / And because a visible motive is seen.

He composes the third topical section: / Is the tradition of taking the cloth belonging to the Vaisarjana offering / valid cognition or not? Since it is not sublated by a direct Vedic statement, it would be valid cognition, like [that of] the Aṣṭakā and so forth.

Since no direct Vedic statement is postulated when a foundation solely in visible greed is possible, / this [tradition], even though free of sublation unlike the complete wrapping, is indeed not valid cognition.

In the Jyotiṣṭoma, when the framework of the animal sacrifice to Agni-Soma has commenced, the Vaisarjana offering is enjoined. There, having covered the sacrificer, wife, sons, and brothers with an unwashed cloth, and tying the ladle handle to the edge of the cloth, he makes the oblation. In tradition, regarding that cloth, it is remembered: "The Adhvaryu priest takes the cloth belonging to the Vaisarjana offering." This tradition is not sublated by a direct Vedic statement like the tradition of complete wrapping. Therefore, it is a means of valid cognition, like the traditions of the Aṣṭakā and the like. When this is obtained,

we say: At some time a certain Adhvaryu priest took that cloth out of greed. The assumption is also possible that this tradition has just that as its foundation. And this assumption follows what is seen, because greed is observed in priests hired by the sacrificial fee. That being so, when this tradition can be accounted for otherwise, a foundational direct Vedic statement cannot be postulated as with the Vedic texts for the Aṣṭakā and so forth. Therefore, even though there is no sublation, this tradition is not a means of valid cognition, because there is no different foundation. / [End of topic].

(Aphorisms 5-7 in the fourth topical section on the superiority of the principal matter) / If it is said to conflict with the non-violation of what is taught [by tradition];

no, because [the sacrifice] is measured by the scripture;

or rather, when a cause is not perceived, those [words] would be understood as employed [in that meaning].

He composes the fourth topical section according to the opinion of the commentary: / Is this tradition, "[it should be done] by one who has sipped water," valid or not valid? It must not be valid, / because of conflict with the direct Vedic sequence, "Having made the veda..."

Sipping water and so on is a substantive rite here; sequence is a property residing in substantive rites. / Because a property depends on the property-bearer, validity exists, since there is no sublation.

Sipping water enjoined by "Having sneezed, he should sip water" serves the purpose of the person. But when an occasion such as sneezing occurs in the middle of a sacrifice, then occasional sipping of water as a subsidiary element to the sacrifice is enjoined by tradition: "It must be done by one who has sipped water." This tradition is not a means of valid cognition. Why? Because it is in conflict. In the direct Vedic statement, "Having made the veda, he prepares the vedi," sequence is understood by the suffix -tvā, which expresses an earlier time. A veda is an instrument of sweeping made of darbha grass. The vedi is the ground between the Āhavanīya and Gārhapatya fires, dug four fingers deep. If in between them he should perform sipping of water occasioned by sneezing and the like, then the uninterrupted sequence stated by the direct Vedic statement would be contradicted. Therefore, like the tradition of wrapping, the tradition of sipping water is not a means of valid cognition. When this is obtained—

We reply: Just like the things to be performed mentioned in direct Vedic statements such as the knowledge of the Veda, sipping water and the like mentioned in tradition are things to be performed. But sequence is a property residing in things to be performed, and it depends upon those things. Therefore, through conflict with that upon which it depends, sequence alone is sublated, but sipping water is not sublated by sequence. Therefore, this tradition is a means of valid cognition.

In this very [topic], the author of the Vārtika formulated two inquiries in a different manner. There, he presents the first inquiry:

Is non-injury stated by the Śākya dharma or not dharma, being found in revelation?
It is not dharma; for cow's milk held in a dog's skin would not be pure.

In the direct Vedic statement, ‘One should scrupulously observe student-celibacy, non-injury, non-possession, and truth,’ non-injury and the like are stated as dharma. That very dharma has also been stated by the Śākya. Therefore, tradition of the Śākya is a means of valid cognition with respect to dharma—if this is argued: / No. Even though in its own nature it is dharma, when there is connection with the Śākya, it would become non-dharma by the principle of the cow's milk. Non-injury and the rest are not to be understood through his text. Therefore, that tradition is not a means of valid cognition regarding dharma.

He shows another inquiry:
Is the conduct of the virtuous an invalid source or not? It is invalid for lack of a root [in the Veda];
It is a means of valid cognition because, like the Aṣṭakā rites and the like, it has a root [in the Veda].

If it is argued that the conduct of the virtuous, such as the Holāka festival and the like, lacks validity because of the absence of a root Veda—no; for, because it has been accepted by orthodox learned people, it has its root in the Veda like the Aṣṭakā rites and the rest. For this very reason, even though not taught in a specific form by Manu and others for fear of text elongation, the conduct of the virtuous has been taught in a general form, through the citation of means of valid cognition regarding dharma in the manner of: ‘Direct Vedic statement, tradition, and the conduct of the virtuous.’ Therefore, the conduct of the learned is a means of valid cognition.

(In the fifth topical section on the validity of words established in authoritative treatises [in the section on the Ārya and Mleccha], aphorisms 8–9)

Because no contradiction is observed among them, there would be equal dispute;

Or rather, [the meaning] found in authoritative treatises [is accepted], because it is the cause thereof.

He sets forth the fifth topical section:
Do words such as ‘yava’ have two meanings or not, through the equality of the Ārya and the Mleccha?
Both meanings, such as long-bearded barley and panic seed, are optional alternatives.

‘Where other [plants]...’—the usage found in authoritative treatises is stronger;
Therefore, in ritual dharma based on authoritative treatises, panic seed and the rest are not taken here.

It is heard [in scripture]: "There is a caru made of yava" and "He puts on two shoes made of varāha." There, the Āryas use the word 'yava' for long-awned barley and the word 'varāha' for a pig. The Mlecchas, however, use the word 'yava' for panic seeds, and the word 'varāha' for a black bird. When that is so, since the meanings of words are to be determined by common worldly usage, and because the usages of the Āryas and the Mlecchas are of equal strength, both kinds of meanings should be accepted optionally. In response to this prima facie view—

We reply: In the comprehension of ritual duty taught in scripture, scriptural usage is stronger, because it is more proximate and comes down through an unbroken lineage of tradition. In the scripture, the corroborative statement to the injunction concerning yava is heard thus: "When other plants wither, then these stand up, rejoicing as it were." Growth at the time of the destruction of other plants is seen in long-awned barley, but not in panic seeds; because they ripen before the ripening of other plants. And the corroborative statement to the injunction concerning shoes of varāha is this: "Cattle run after a varāha." Running after by cattle is possible in the case of a pig, but not in that of a black bird. Therefore, long-awned barley and so on are the meanings of the words 'yava' and so on. / On this point, the author of the Vārttika gave the example of the word 'pīlu'. And the Mlecchas use that for an elephant, but the Āryas for a tree. There, since uncorrupted usage is found among the Āryas, the tree alone is the meaning of the word 'pīlu'. / In this very topical section, he states the view of the Guru: / If it is objected that because the meaning of 'yava' and so on is undetermined, that sentence is not a means of valid cognition; / [The reply is:] No, because scriptural text is stronger, the meaning is determined by its usage.

The meaning is clear. / In this very topical section, he frames another interpretation according to the view of the author of the Vārttika: / Is that conduct of the cultured in marrying a maternal uncle's daughter and the like valid, or not? / It has validity like other conduct, or it lacks validity because of sublation by tradition.

For everywhere the conduct of the cultured is rooted in tradition; therefore, here too / Tradition is to be inferred, but that [inferred tradition] is sublated by directly perceived tradition.

In certain southern regions, cultured people marry the daughter of a maternal uncle. If it is said that this conduct is an authoritative means of valid cognition, because it is the conduct of the cultured, just like the conduct regarding the Holākā festival and the like;

No, because, being opposed to tradition, it is vitiated by the fallacy of contradiction [by a stronger authoritative source]. And likewise there is the traditional text:

"Having married the daughter of a maternal uncle, and likewise one belonging to the mother's clan,

And one of the same lineage-sages, having abandoned her, he should perform the cāndrāyaṇa penance." Thus.

And one should not suspect that tradition and conduct are of equal force on the ground that they are equal in inferring the Vedic source. For righteous conduct like the Holākā festival and the like cannot infer a Vedic source in the manner of the tradition of Manu and others. For cultured persons of the present time cannot, like Manu and others, directly perceive with divine knowledge the Veda that is remote in space and time, by which conduct of the cultured could infer a Vedic source. And any cultured person anywhere, in a specific country and at a specific time, is able to see some traditional text that is the source of conduct like the Holākā and the like. Therefore, by the conduct of the cultured only a traditional text can be inferred, not a direct Vedic statement. And an inferred traditional text is sublated by a conflicting, directly perceived traditional text. Hence they say:

"From conduct tradition is inferred, and tradition assumes a direct Vedic statement; / Thereby their authority, being separated by two [steps], is far removed." Thus. / Therefore, the non-authoritativeness of such conduct must be accepted. || 17

He frames another interpretation in this very section: / Is the meaning of 'trivṛt' and so forth the worldly one or that found in the sentence, because they are equal? / Both [are possible]; but let the latter be [accepted] here, because of the unity of sentence between injunction and corroborative statement.

In the direct Vedic statement "The Bahiṣpavamāna is threefold (trivṛt)", threefoldness is the worldly meaning of the word 'trivṛt'. From the supplementary sentence, in the three

hymns consisting of triplets of verses that are capable of producing the chant known as the Bahiṣpavamāna, namely "Sing to him, o men!" and so forth, the meaning is a set of nine verses. If it is said that there, although the Veda is stronger in determining dharma, the worldly and Vedic [usages] are of equal strength in determining words and word-meanings, and therefore both meanings are to be accepted optionally; it is not so. If the worldly meaning is accepted, in the injunctive sentence the meaning would be threefoldness, and in the corroborative sentence the meaning would be a set of nine verses used for chanting; thus there would not be unity of sentence, because the injunction and corroborative statement would refer to different subjects. Therefore, for the sake of unity of sentence, only a set of nine verses used for chanting is the determined meaning in the injunctive sentence.

(In the sixth topical section on the authoritativeness of meanings established among Mlecchas, there is the aphorism:) / But that which is enjoined should be understood by a means of valid cognition, because there is no contradiction.

He frames the sixth topical section: / Is the meaning of words like 'pika' to be postulated, or accepted from the usage of the Mlecchas? / It is to be postulated, because it is not established among the Āryas and because the non-Āryas are disregarded.

The usage of the Mlecchas, however, is to be accepted, since no contradiction is seen; / Therefore, words like 'pika' and 'nema' have the meanings of the cuckoo and the like.

The Āryas do not use the word 'pika' and the like for any meaning, and the Mlecchas are not authoritative. Therefore, the meaning of words like 'pika' and 'nema' is to be postulated through Vedic glossaries, etymology, and grammar—if this is urged; / It is not so. Since no contradiction with Āryan usage is seen, in such a doubtful matter even Mleccha usage is to be respected. Conventional Mleccha usage is better than an unstable meaning that is merely postulated. Where even such conventional usage is absent, there etymology and the like serve their purpose. Therefore, through non-Āryan usage, 'pika' is a cuckoo, the word 'nema' signifies half, and the word 'tāmarasa' signifies a lotus; thus it should be understood. || 20

In this very topical section, he states the view of the Guru: / The injunction to immolate a pika is unauthoritative, because the meaning is not understood; / It is not so, because that [meaning] is understood even from Mleccha usage, since it is not contradicted.

The meaning is clear.

(In the seventh topical section on the Kalpasūtras' lack of intrinsic validity, there are aphorisms 11–14:) / If it is said that it is a treatise on ritual performance;

No, because there is no restriction;

And because there is no supplementary sentence;

And because of application everywhere, and because of the treatise on juxtaposition.

He frames the seventh topical section: / Are the Kalpasūtras and the like authorless, or artificial? They are not artificial, / Because, since awareness of dharma arises from direct Vedic statement and tradition, intrinsic validity is equal [in both].

They are of human origin because they are designated by the names of men, not being equal to the Kāṭhaka and others; / For some people remembered their authorship by Āpastamba and others.

The Kalpasūtra texts marked with the names of Baudhāyana, Āpastamba, Āśvalāyana, Kātyāyana, and others; the six auxiliary limb texts such as the Nigama and Nirukta; and the smṛtis of Manu and others are not of human authorship, because they generate the cognition of dharma, like the Veda. Nor should one suspect that they are unlike the Veda by virtue of depending on a root means of valid cognition; for, once cognition has arisen, it is independent due to the acceptance of intrinsic validity.
Not so. Because the stated inference has been set aside by time [i.e., is contradicted by perception]. Those texts are designated with human names, as 'the aphorism of Baudhāyana' and 'the aphorism of Āpastamba.' Nor is it appropriate that it be on account of exposition, like the designations Kāṭhaka and so on; for at the time of the composition of those texts, it was perceived by people of that time.
And that [perception] continues through an unbroken succession. Therefore, they are of human authorship, like the texts of Kālidāsa and others. Even so, they are a means of valid cognition because they are rooted in the Veda. || 23

On this very point, he states the Guru's [Prabhākara's] view:
Because the performance of the new-moon sacrifice on all lunar days is stated in the Kalpa, direct Vedic statement is not a means of valid cognition.
Not so; because its status as Veda in the Kalpa has been abandoned, as it is to be established, and because it is weaker.

If one objects: The author of the Kalpasūtra states that the new-moon sacrifice is to be performed on all lunar days: 'The new-moon sacrifice is to be performed on all lunar days.' But direct Vedic statement declares that it is to be performed only on the new-moon day. Therefore, direct Vedic statement is not a means of valid cognition, because it is contradicted by the Veda in the form of the Kalpasūtra—
Not so. The status of the Kalpa as Veda has not yet been established; rather, it is to be established with effort. And it cannot be established; for its human authorship has been established through its designation and through the perception of its author. Therefore, because the Kalpasūtra is weaker, direct Vedic statement does not lack validity.

(In the eighth topical section, on Holākā, aphorisms 14–23)

Because of the restriction of inference, [only] that which is connected with it would be a means of valid cognition.

Or rather, it would be a duty for all, because an injunction has that principle.

The application would be due to [its] being seen.

And because of the absence of an indicatory mark for a permanent [duty].

For the name is due to connection with a region.

If it be argued that it would not exist in other regions—

It could be, for the name is etymological, as in the case of a person from Mathurā.

Or it is the duty of one [region], like the slope.

But it is equal to the duty of an agent.

He sets forth the eighth topical section:
Is there a restriction of Holākā and the rest, or are they common to all? The former,
because they are observed according to different regions; but they are equal, because their root is equal.

If one objects: The validity of the conduct of the virtuous such as Holākā, and of specific smṛtis such as that of Hārīta, is restricted according to the difference of the performing persons. Why? Because they are observed in a specific region. Holākā and the like are performed only by easterners; Holākā is the spring festival. Āhnīnaibuka and the like [are performed] by southerners; the worship of stationary deities such as the Karañja tree and the Sun, passed down in one's own family, is designated by the word 'āhnīnaibuka.' The Udvṛṣabhayajña and the like [are performed] by northerners; having worshipped bulls on the full-moon day of the month of Jyeṣṭha, they make them run: this is the Udvṛṣabhayajña. In this way, a certain smṛti, like that of Hārīta, is seen somewhere in a specific region. Therefore, their validity is restricted—
Not so. Because the Veda inferred as their root is common to all, they too are common to all.
On this very point, he states the Guru's [Prabhākara's] view:
If it be argued that when the direct Vedic statement containing words such as 'easterners' is inferred, it lacks validity because the meaning of the word is not understood,
that is not so, because the inference is general.

If one objects: When Holākā and the like are performed by easterners and others in a restricted manner, the direct Vedic statement that is their root must also be inferred as containing words such as 'easterners.' There, the meaning of a word such as 'easterner' is not understood. The very men who live in the east for some time are observed living in the west at another time. There, because the meaning of words such as 'easterner' in direct Vedic statement is not understood, the direct Vedic statement is not a means of valid cognition—
Not so. Because the generality of performance is what leads to the presumption of the root direct Vedic statement. Therefore, since there is comprehension of meaning when words such as 'easterner' are absent, the inferred direct Vedic statement is a means of valid cognition.

(In the ninth topical section, on the use of correct words, aphorisms 24–29)

Because usage and origination are not [governed by] scripture, there would be a restriction regarding words.

Because a word is produced by effort, it is subject to error.

And multiplicity of words is contrary to principle.

In that case, the correct state would be due to the expertise of experts.

And their inability [is inferred] from resemblance.

And because of being a part, [the meaning] would obtain even when there is an interchange of case-endings.

He sets forth the ninth topical section:
Among 'gauḥ', 'gāvī', and the like, is there any restriction regarding correctness or usage,
or is there none? There is none, because grammar lacks a root [in the Veda].

One should use only correct words; only words like 'gauḥ' are correct:
there is such a restriction, because it is rooted in preceding and preceding grammars.

The word 'gauḥ' is used by experts in grammar to denote an object possessing a dewlap and other features. But by those ignorant of it, following their own regional languages, words such as 'gāvī', 'goṇī', and 'gopotalikā' are used. There, in the sense that 'only such a word is correct, not such [another],' there is no regulator. Likewise, in usage too, there is no [rule] that 'only such a word should be used, not such [another].' First, the usage of elders is not the regulator, because it is the same for all words. Nor is the smṛti of grammar the regulator, because it lacks validity as it lacks a root [in the Veda]. For the usage of experts cannot be its root, because that would entail circular reasoning: once the validity of the smṛti of grammar is established, the expertise of the users according to it is established; and once that is established, the validity of grammar is established by having their usage as its root. Therefore, there is no restriction regarding correctness and usage—such being the prima facie view—

We reply: there is indeed a restrictive rule that "one should employ only correct words, and not corruptions." For Vedic passages declaring respectively fault and merit are heard in both regards: "One single word, properly understood and well employed, becomes a wish-granting cow in the heavenly world" is the statement of merit; "Therefore a Brahmin must not speak like a barbarian nor utter corrupt words; for that which is an incorrect word is indeed barbarous" is the statement of fault. Just as there is a restriction regarding usage, so a restriction must also be acknowledged regarding correctness: "Only 'gauḥ' and the like are correct words, not 'gāvī' and the like," because grammar is the regulator in this matter. Nor is it baseless, since preceding grammars serve as its basis. Intending this very point, it is said: "There, as in the topical section on the sacrificial post, no fault can be found, because the succession of grammatical works is beginningless." The authority of grammar has been directly presented by the Veda itself. Thus the followers of the Atharvaveda recite: "'Two sciences are to be known'—so indeed say the knowers of Brahman—'the higher and the lower.' There, the lower is the Ṛgveda, the Yajurveda, the Sāmaveda, the Atharvaveda, phonetics, ritual procedure, grammar, etymology, metrics, and astronomy." In the Taittirīya Brāhmaṇa also, the acceptability of grammar is heard: "Speech indeed spoke formerly unanalyzed. The gods said to Indra: 'Analyze this speech for us.' Having entered into its midst, Indra analyzed it. Therefore this analyzed speech is spoken." And the grammarians—Pāṇini, Kātyāyana, and Patañjali—are on a par with Manu and others. Therefore, the traditional text of grammar is an authoritative means of valid cognition regarding the employment of correct words alone, and regarding the correctness of 'gauḥ' and the like alone. || 28

On this very topic, he states the view of the Guru [Prabhākara]:
Because of doubt between a dental and a palatal [sibilant] in the scriptural text concerning the seizure of a horse,
If it lacks authority, the determination of the non-dental [palatal] follows from the power of the grammar declared by authoritative persons.

In "He should seize a horse [aśvam]," if the sound lying between the letter 'a' and the letter 'va' is dental [asvam], the meaning becomes: "'Sva' means wealth; he should seize a destitute person lacking that."

If it is palatal [aśvam], it denotes a horse. If it is objected that, on account of doubt resulting from that, it lacks authority—
No. According to grammar, when the Uṇādi suffix 'va' occurs after the verbal root "aśū [meaning] pervading," the determination of its being palatal is made.
Therefore the scriptural text concerning the seizure of a horse is an authoritative means of valid cognition.

(In the tenth topical section, on the identity of words between everyday life and the Veda, or on the generic form, aphorisms 30–35):
There is identity of meaning because there is no injunction to employ [new words] and because of non-division.

Because they are not words denoting substance.

And because another is seen.

Rather, the generic form [is the meaning], because of being meant for ritual action.

If it is objected that there would be no ritual action, [we reply] that if there is an injunction with reference to a different thing, it is not the substance.

Because it has that for its meaning, usage is undivided.

In the tenth topical section, he sets forth the first sub-inquiry:
Are the word and word-meaning that exist in everyday life not those in the Veda, or are they present here?
A word is different due to difference of form; the distinction of supine [cattle] and the like is evident.

Identity of a word exists from the identity of its phonemes; difference of form is occasional;
Through the general identity of words, the identity of their meanings is of the same kind.

The Vedic word and word-meaning are different from the everyday word and word-meaning. Why? Because of difference of form. To begin with, difference of form is seen in the word. The word 'ātman'
is fixed in everyday usage as having the initial vowel 'ā' and so forth; but in the Veda it is sometimes read without the initial vowel 'ā': "the resolute man by himself [tmanā]." In everyday usage it is "brāhmaṇāḥ"; but in the Veda it is read otherwise: "The Brahmin fathers [brāhmaṇāsaḥ], worthy of Soma." Difference of meaning is also evident. "Supine indeed do the divine cattle convey," and it is
heard: "To the gods, O tree, [carry] the oblations, O golden-leaved one, to your ancient goal." Therefore, when it is prima facie maintained that "the word and word-meaning in everyday life and the Veda are different"—

We reply: those very meanings of words that are everyday are Vedic as well. To explain: to begin with, the eternality of phonemes was established in the first quarter. Just as, despite the difference among the persons employing them, and despite the difference in many utterances made even by a single person, there is identity of phonemes because of the recognition: "These are those very same phonemes"—likewise, one must accept identity of words by virtue of the uncontradicted recognition: "Those very words such as 'gauḥ' in everyday life are the ones being recited in the Veda." As for the difference of words such as 'tmanā' and 'devāsaḥ', it is merely occasional; not by that much can the identity of words, grasped through abundant recognition, be set aside. Otherwise, even the identity of Devadatta recognized as "This is that Devadatta" could be set aside merely by a difference of place and the like. And when there is identity of words, identity of meaning is inevitable; otherwise, because of the absence of a separate grasping of convention in the Veda, the consequence would be its lack of expressive power. Intending this very point, it is said: