The Vedanta Sutras of Badarayana, Commentary by Sankara (SBE38), tr. by George Thibaut  at sacred-texts.com
9. But the same (soul returns from Brahman); on account of work, remembrance, text, and precept.
Here we have to enquire whether the soul when awaking from the union with Brahman is the same which entered into union with Brahman, or another one.--The pûrvapakshin maintains that there is no fixed rule on that point. For just as a drop of water, when poured into a large quantity of water, becomes one with the latter, so that when we again take out a drop it would be hard to manage that it should be the very same drop: thus the sleeping soul, when it has become united with Brahman, is merged in bliss and not able again to rise from it the same. Hence what actually awakes is either the Lord or some other soul.--To this we reply that the same soul which in the state of sleep entered into bliss again arises from it, not any other. We assert this on the ground of work, remembrance, sacred text, and precept; which four reasons we will treat separately. In the first place the person who wakes from sleep must be the same, because it is seen to finish work left unfinished before. Men finish in the morning what they had left incomplete on the day before. Now it is not possible that one man should proceed to complete work half done by another man, because this would imply too much. 1
[paragraph continues] Hence we conclude that it is one and the same man who finishes on the latter day the work begun on the former.--In the second place the person rising from sleep is the same who went to sleep, for the reason that otherwise he could not remember what he had seen, &c., on the day before; for what one man sees another cannot remember. And if another Self rose from sleep, the consciousness of personal identity (âtmânusmarana) expressed in the words, 'I am the same I was before,' would not be possible.--In the third place we understand from Vedic texts that the same person rises again, 'He hastens back again as he came, to the place from which he started, to be awake' (Bri. Up. IV, 3, 16); 'All these creatures go day after day into the Brahma-world and yet do not discover it' (Kh. Up. VIII, 3, 2); 'Whatever these creatures are here, whether a lion, or a wolf, or a boar, or a worm, or a midge, or a gnat, or a musquito, that they become again and again' (Kh. Up. VI, 10, 2). These and similar passages met with in the chapters treating of sleeping and waking have a proper sense only if the same soul rises again.--In the fourth place we arrive at the same conclusion on the ground of the injunctions of works and knowledge, which, on a different theory, would be meaningless. For if another person did rise, it would follow that a person might obtain final release by sleep merely, and what then, we ask, would be the use of all those works which bear fruit at a later period, and of knowledge?--Moreover on the hypothesis of another person rising from sleep, that other person would either be a soul which had up to that time carried on its phenomenal life in another body; in that case it would follow that the practical existence carried on by means of that body would be cut short. If it be said that the soul which went to sleep may, in its turn, rise in that other body (so that B would rise in A's body and A in B's body), we reply that that would be an altogether useless hypothesis; for what advantage do we derive from assuming that each soul rises from sleep not in the same body in which it had gone to sleep, but that it goes to sleep in one body and rises in another?--Or else the soul rising (in A's body) would be
one which had obtained final release, and that would imply that final release can have an end. But it is impossible that a soul which has once freed itself from Nescience should again rise (enter into phenomenal life). Hereby it is also shown that the soul which rises cannot be the Lord, who is everlastingly free from Nescience.--Further, on the hypothesis of another soul rising, it would be difficult to escape the conclusion that souls reap the fruits of deeds not their own, and, on the other hand, are not requited for what they have done.--From all this it follows that the person rising from sleep is the same that went to sleep.--Nor is it difficult to refute the analogical reasoning that the soul, if once united with Brahman, can no more emerge from it than a drop of water can again be taken out from the mass of water into which it had been poured. We admit the impossibility of taking out the same drop of water, because there is no means of distinguishing it from all the other drops. In the case of the soul, however, there are reasons of distinction, viz. the work and the knowledge (of each individual soul). Hence the two cases are not analogous.--Further, we point out that the flamingo, e.g. is able to distinguish and separate milk and water when mixed, things which we men are altogether incapable of distinguishing.--Moreover, what is called individual soul is not really different from the highest Self, so that it might be distinguished from the latter in the same way as a drop of water from the mass of water; but as we have explained repeatedly, Brahman itself is on account of its connexion with limiting adjuncts metaphorically called individual soul. Hence the phenomenal existence of one soul lasts as long as it continues to be bound by one set of adjuncts, and the phenomenal existence of another soul again lasts as long as it continues to be bound by another set of adjuncts. Each set of adjuncts continues through the states of sleep as well as of waking; in the former it is like a seed, in the latter like the fully developed plant. Hence the proper inference is that the same soul awakes from sleep.
147:1 There would follow from it, e.g. that in the case of sacrifices occupying more than one day, there would be several sacrificers, and that consequently it would be doubtful to whom the fruit of the sacrifice, as promised by the Veda, belongs. And this would imply a stultification of the sacred text.