The Vedanta Sutras of Badarayana, Commentary by Sankara (SBE38), tr. by George Thibaut  at sacred-texts.com
24. And as the record of others (viz. the Taittirîyaka) is not such as in the purusha-vidyâ (of the Khândogya), (the two purusha-vidyâs are not to be combined).
In the Rahasya-brâhmana of the Tândins and the Paiṅgins (the Khândogya) there is a vidyâ treating of man, in which man is fancifully identified with the sacrifice, the three periods of his life with the three libations, his hunger and so on, with the dîkshâ, &c. And other particulars also are mentioned there, such as formulas of prayer, use of mantras and so on.--A similar fanciful assimilation of the sacrifice and man the Taittirîyakas exhibit, 'For him who knows thus the Self of the sacrifice is the sacrificer, Faith is the wife of the sacrificer,' and so on (Taitt. Âr. X, 64).--The doubt here arises whether the particulars of the man-sacrifice given in the Khândogya are to be inserted in the Taittirîyaka or not.
Against the view of the pûrvapakshin that they are so to be inserted because in both places we have a purusha-yagña, we maintain that they are not to be inserted because the characteristics of the purusha-yagña of the Khandogas are not recognised in the Taittirîya-text. This the Sûtra-kâra expresses by saying, 'As (the record of the followers of some Sâkhâs, viz. the Tândins and Paiṅgins, is) in the purusha-vidyâ, not such is the record of others,' viz. the Taittirîyakas. For the latter exhibit an identification of man with the sacrifice, in which the wife, the sacrificer, the
[paragraph continues] Veda, the vedi, the sacrificial grass, the post, the butter, the sacrificial animal, the priest, &c., are mentioned in succession; none of which particulars are mentioned in the Khândogya. The use also to which the Taittirîyaka turns the three libations is different from the Khândogya, And the few points in which the two texts agree, such as the identification of the Avabritha-ceremony with death, lose their significance side by side with the greater number of dissimilarities, and are therefore not able to effect the recognition of the vidyâ.--Moreover the Taittirîyaka does not represent man as the sacrifice (as the Khândogya does); for the two genitives ('of him who thus knows' and 'of the sacrifice') are not co-ordinate, and the passage therefore cannot be construed to mean, 'The knowing one who is the sacrifice, of him the Self is,' &c. For it cannot be said that man is the sacrifice, in the literal sense of the word. 1 The two genitives are rather to be taken in that way, that one qualifies the other, 'The sacrifice of him who thus knows, of that sacrifice,' &c. For the connexion of the sacrifice with man (which is expressed by the genitive, 'the sacrifice of him') is really and literally true; and to take a passage in its literal meaning, if possible at all, is always preferable to having recourse to a secondary metaphorical meaning. 2 Moreover the words next following in the Taittirîyaka-passage, 'the Self is the sacrificer,' declare that man (man's Self) is the sacrificer, and this again shows that man's relation to the sacrifice is not that of co-ordination. 3 Moreover as the section beginning with 'Of him who thus knows' forms an anuvâda of something previously established (and as such forms one vâkya to which one sense only must be ascribed), we must not bring about 'a split of the sentence' by interpreting it as
teaching in the first place that man is the sacrifice, and in the second place that the Self and the other beings enumerated are the sacrificer and so on. And as we see that the passage, 'Of him who thus knows,' &c., follows upon some instruction about the knowledge of the Self coupled with samnyâsa, we apprehend that the Taittirîyaka-chapter is not an independent vidyâ but merely supplementary to the instruction previously given. In agreement with this conclusion we observe that the Taittirîyaka promises only one result for both chapters, viz. the one stated in the passage, 'He obtains the greatness of Brahman.'--On the other hand the text embodying the purusha-vidyâ in the Khândogya is an independent text; for we see that an independent result is attached to it, viz. an increase of length of life, 'He who knows this lives on to a hundred and sixteen years.'--Hence the particulars mentioned in the purusha-vidyâ of another Sâkhâ, such as formulas of prayer, mantras and so on, are not to be combined with the Taittirîya-text of the vidyâ.
221:1 And therefore we are not warranted in taking the two genitives as co-ordinate, as otherwise they might be taken.
221:2 Which latter would be the case if we should take the two genitives as co-ordinate and therefore expressing an imaginative identification of the man and the sacrifice.
221:3 If man is the sacrificer he cannot be identified with the sacrifice; he is rather the Lord of the sacrifice.