The Vedanta Sutras of Badarayana, Commentary by Sankara (SBE38), tr. by George Thibaut  at sacred-texts.com
9. And on account of the (omkâra) extending over the whole (Veda), (the view that the term udgîtha expresses a specialisation) is appropriate.
In the passage, 'Let a man meditate on the syllable Om (as) the udgîtha,' the two words 'omkâra' and 'udgîtha' are placed in co-ordination. 1 The question then arises
whether the relation in which the ideas conveyed by these two words stand to each other is the relation of super-imposition (adhyâsa) or sublation (apavâda) or unity (ekatva) or specification (viseshana); for primâ facie each of these relations may present itself to the mind.--Adhyâsa takes place when the idea of one of two things not being dismissed from the mind, the idea of the second thing is superimposed on that of the first thing; so that together with the superimposed idea the former idea remains attached to the thing on which the second idea is superimposed. When e.g. the idea of (the entity) Brahman superimposes itself upon the idea of the name, the latter idea continues in the mind and is not driven out by the former. A similar instance is furnished by the superimposition of the idea of the god Vishnu on a statue of Vishnu. So, in the case under discussion also, the idea of the udgîtha may be superimposed on the omkâra or the idea of the omkâra on the udgîtha.--We, in the second place, have apavâda when an idea previously attached to some object is recognised as false and driven out by the true idea springing up after the false one. So e.g. when the false idea of the body, the senses, and so on being the Self is driven out by the true idea springing up later--and expressed by judgments such as 'Thou art that'--that the idea of the Self is to be attached to the Self only. Or, to quote another example, when a previous mistaken notion as to the direction of the points of the compass is replaced by the true notion. So here also the idea of the udgîtha may drive out the idea of the omkâra or vice versâ.--The relation would, in the third place, be that of 'unity' if the terms 'omkâra' and 'udgîtha' were co-extensive in meaning; just as the terms, 'the Best of the Twice-born,' 'the Brâhmana,' 'the god among men,' all denote an individual of the noblest caste.--The relation will, finally, be that of specification if, there being a possibility of our understanding the omkâra in so far as co-extensive with all the Vedas, the term 'udgîtha' calls up the idea of the sphere of action of the udgâtri. The passage would then mean, 'Let a man meditate on that omkâra which is the udgîtha,' and would
be analogous to an injunction such as 'Let him bring that lotus-flower which is blue.'
All these alterations present themselves to the mind, and as there is no reason for deciding in favour of any one, the question must remain an unsettled one.
To this pûrvapaksha-view the Sûtra replies, 'And on account of extending over the whole, it is appropriate.'
The word 'and' stands here in place of 'but,' and is meant to discard the three other alternatives. Three out of the four alternatives are to be set aside as objectionable; the fourth, against which nothing can be urged, is to be adopted.--The objections lying against the first three alternatives are as follows. In the case of adhyâsa we should have to admit that the word which expresses the idea superimposed is not to be taken in its direct sense, but in an implied sense 1; and we should moreover have to imagine some fruit for a meditation of that kind. 2 Nor can it be said that we need not imagine such a fruit, as scripture itself mentions it in the passage, 'He becomes indeed a fulfiller of desires' (I, 1, 7); for this passage indicates the fruit, not of the ideal superimposition of the udgîtha on the omkâra, but of the meditation in which the omkâra is viewed as the fulfilment of desires.--Against the hypothesis of an apavâda there likewise lies the objection that no fruit is to be seen. The cessation of wrong knowledge can certainly not be alleged as such; for we see no reason why the cessation of the idea that the omkâra is udgîtha and not omkâra or vice versâ should be beneficial to man. Sublation of the one idea by the other is moreover not even possible in our case; for to the omkâra the idea of the omkâra remains always attached, and so to the udgîtha the idea of the udgîtha. The passage, moreover, does not aim at teaching the true
nature of something, but at enjoining a meditation of a certain kind.--The hypothesis of unity again is precluded by the consideration that as in that case one term would suffice to convey the intended meaning, the employment of two terms would be purposeless. And moreover the term 'udgîtha' is never used to denote the omkâra in its connexion with the Rig-veda and Yagur-veda; nor is the word 'omkâra' used to denote that entire second subdivision of a sâman which is denoted by the word 'udgîtha.' Hence it cannot be said that we have to do with different words only denoting one and the same thing.--There thus remains the fourth alternative, 'On account of its comprising all the Vedas.' That means: In order that the omkâra may not be understood here as that one which comprises all the Vedas, it is specified by means of the word 'udgîtha,' in order that that omkâra which constitutes a part of the udgîtha may be apprehended.--But does not this interpretation also involve the admission of implication, as according to it the word 'udgîtha' denotes not the whole udgîtha but only a part of it, viz. the omkâra?--True, but we have to distinguish those cases in which the implied meaning is not far remote from the direct meaning and those in which it is remote. If, in the present case, we embrace the alternative of adhyâsa, we have to assume an altogether remote implication, the idea of one matter being superimposed on the idea of an altogether different matter. If, on the other hand, we adopt the alternative of specification, the implication connected therewith is an easy one, the word which in its direct sense denotes the whole being understood to denote the part. And that words denoting the whole do duty for words denoting the part is a matter of common occurrence; the words 'cloth,' 'village,' and many others are used in this fashion 1.--For all these reasons we declare that the appropriate view of the Khândogya-passage is to take the word 'udgîtha' as specialising the term 'omkâra 2.'
196:1 Sâmânâdhikaranya, i.e. literally, 'the relation of abiding in a common substratum.'--The two words are shown to stand in that relation by their being exhibited in the same case.
198:1 i.e. in the present case we should have to assume that the word udgîtha means, by implication, the omkâra.--Recourse may be had to implied meanings only when the direct meaning is clearly impossible.
198:2 For a special adhyâsa-meditation must be attended with a special result.
199:1 We say, e.g. 'the cloth is burned,' even if only a part of the cloth is burned.
199:2 We therefore, according to, Saṅkara, have to render the passage p. 200 under discussion as follows, 'Let a man meditate on the syllable Om which is (i.e. which is a part of) the udgîtha.'