एष उ एव ब्रह्मणस्पतिः; वाग्वै ब्रह्म, तस्या एष पतिः, तस्मादु ब्रह्मणस्पतिः ॥ २१ ॥
eṣa u eva brahmaṇaspatiḥ; vāgvai brahma, tasyā eṣa patiḥ, tasmādu brahmaṇaspatiḥ || 21 ||
21. This alone is also Brahmaṇaspati (lord of the Yajus.) Speech is indeed Brahman (Yajus), and this is its lord. Therefore this is also Brahmaṇaspati.
Similarly the self of the Yajuses. How? This alone is also Brahmaṇaspati. Speech is Brahman or Yajus, which is a kind of speech. And this is its lord, the lord of that Yajus. Therefore this is indeed Brah-manaspati, as before.
How is it known that the words ‘Bṛhatī’ and ‘Brahman’ mean the Ṛc and the Yajus respectively, and nothing else? Because at the end (of this topic, in the next paragraph) the word ‘speech’ is used as co-ordinate with ‘Sāman,’ ‘Speech is indeed Sāman.’ Similarly in the sentences, ‘Speech is indeed Bṛhatī’ and ‘Speech is indeed Brahman,’ the words ‘Bṛhatī,’ and ‘Brahman’, which are co-ordinate with ‘speech’, ought to mean the Ṛc and the Yajus respectively. On the principle of the residuum also this is correct. When the Sāman is mentioned, the Ṛc and the Yajus alone remain. Another reason is that they are both forms of speech. The Ṛc and the Yajus are particular kinds of speech. Hence they can well be co-ordinated with speech. Moreover, unless they are taken in that sense, there will be no difference between the two terms of each sentence. (In the next two paragraphs) ‘Sāman’ and ‘Udgītha’ clearly denote specific objects. Similarly the words ‘Bṛhatī’ and ‘Brahman’ ought to denote specific objects. Otherwise, not conveying any specific object, they would be useless, and if that specific object be mere speech, both sentences would be tautological. And lastly, the words Ṛc, Yajus, Sāman and Udgītha occur in the Vedas in the order here indicated.