For the earliest critical examination of this difficult verse we are indebted to Böhtlingk (1890),1 who was the first to see that a dead man on his way to heaven recites this verse in response to the Moon’s question “Who are you?” Although Deussen’s interpretation2 correctly reproduces the context, in detail it is often wrong. His interpretation, however, next to Böhtlingk’s translation, forms the basis for the later interpretations which can be divided into two groups. The first group, following Böhtlingk, takes erayadhvam and niṣiñcata (or āsiṣikta) as imperatives; the second, following Deussen, takes them as verbs in the past tense.
Let us first follow Böhtlingk’s line of interpretation. Oertel (1898, 117 f.) rejects Deussen’s emendation airayadhvam with the words “the context seems to me to favor imperatives,” without further explaining this assertion. Windisch (1907, 117 ff.) has observed that the imperatives (puṁsi kartari) erayadhvam and (amṛtyava) ābharadhvam form an inconsistency. Therefore he splits the verse into two answers. The first, until āsiṣikta, is pronounced by the deceased who is to be reborn on the earth, the second, sa jāya upajāyamāna … etc., is the discourse of the redeemed. This interpretation is untenable. We expect at least an iti between the two answers, just as the whole discourse is closed with iti. AV 18, 2, 59–60, to which Windisch (1907, 120, n. 1) refers, is a pseudo-parallel: for in a Saṁhitā-text consisting only of mantras, two verses which are used as alternatives in the ritual or elsewhere can follow one another directly. In a prose text, however, which tries to explain the background of the rituals and the application of the mantras, this direct succession is hardly possible without further comment and without iti.
As regards content, Windisch’s artifice is also to be rejected. Let us first look at the context in the three versions. The deceased meets the heavenly gatekeeper i.e. the moon (KauṣU), one of the Ṛtus (JB 1, 49), or all the Ṛtus (JB 1, 18, which is the oldest original version insofar as the vocative can thus be explained), and he should answer the question “Who are you?” with this verse (pratibrūyād, JB 1, 49; KauṣU) or he should introduce himself without being asked (prabruvīta, JB 1, 18). So the role of the verse in the three contexts is indeed more or less the same. How does Windisch’s forced splitting up of this verse now work in practice, i.e. in his German translation? The second answer (or announcement) runs: “This way I am born: a leap month born afterwards …” (Windisch 1907, 22). A really strange answer! The beginning of the discourse is too little “to the point” and too abrupt. In sa jāya upajāyamāna the pronoun sa is probably to be taken anaphorically.
The first answer is just as strange. The deceased commands: “Send me back, do not let me in to your Lord”; for that would roughly be the meaning if imperatives are assumed. Apart from the fact that such an address to a gatekeeper is hardly conceivable, it is nowhere in the three contexts possible to show that the pitṛyāna was the ideal of the authors. Nevertheless, according to Windisch, these authors prescribe for the deceased that he should respond (pratibrūyād) with these commands.
In addition, one does not expect commands, but statements. The single statement in the first answer, “From the moon the seed has been produced,” is only to be understood in the context of an ancient doctrine of water, which was developed into the five-fire-doctrine (pañcāgnividyā) in JB 1, 45, and which, in connection with the doctrine of the two paths, forms the locus classicus of the transmigration of the soul (BĀU 6, 2; ChU 5, 4 ff.). The doctrine of water itself, however, does not yet necessarily imply rebirth; this is evident from ŚB 3, 7, 4, 4.3 Also in the five-fire-doctrine in JB 1, 45, it is only the origin of man that is explained. The water cycle, which Frauwallner4 regards as the starting point of the doctrine of reincarnation, is still absent there. In the verse, the words vicakṣaṇād … reto ābhṛtam, tam mā puṁsi … erayadhvam and mātari māsiṣikta represent three phases from the water doctrine that are difficult to separate. After the past participle ābhṛtam imperatives are therefore excluded, if the ancient water doctrine, which explains only the origin of man, is here dealt with. But if, with Windisch, one reads a water cycle into these words, then the past participle ābhṛtam implies that the second cycle has already begun, and that the deceased has already emerged from the moon. This interpretation, however, is in contradiction with the context.
Besides, it is not plausible that the deceased themselves choose their destiny. The sun or the gatekeepers separate the liberated from the non-liberated. In KauṣU 1, 2, for instance, the moon allows the one who can answer his question (taṁ yaḥ pratyāha) to pass. Whoever cannot answer him (ya enaṁ na pratyāha), he sends down. The answer extends, of course, to the whole verse. That na pratyāha would refer to Windisch’s first answer only, would indeed be hardly possible.
Since, apart from Sivaprasad Bhattacharya (1955), the other representatives of the line of Böhtlingk5 have not substantially modified Windisch’s interpretation, their translations can here be disregarded. Bhattacharya presumes negative imperatives, i.e. injunctives with the negation mā, while all other scholars interpret mā as a pronoun. According to him, mā should be regarded as a negation only in the KauṣU. It is not plausible, however, that an Upaniṣad author would attempt to adapt a transmitted Brāhmaṇa verse to his ideas by means of a grammatical artifice. Moreover, the same arguments that we have used against Windisch, partly still remain in force.
The main reason for the scant approval of Deussen’s interpretation among later translators probably lies in the fact that the forms erayadhvam and āsiṣikta are not verbs in the past tense, but imperatives. In āsiṣikta one may presume an augment, but for erayadhvam Deussen saw himself compelled to propose the emendation airayadhvam. Keith (1908, 17 f.), however, notes that the augment is not required, and Geldner (19282, 142) translates as a past tense, without comment. Unfortunately, a detailed exegesis of the text and a discussion with the followers of Böhtlingk is missing in the work of these translators, who thus accept an optional use of the augment. On the other hand, Fürst (1915, 22, n. 2) has defended the assumption of verbs in the past tense, and he rightly observes “that the whole discourse that follows the question ‘Who are you?’, is only a lengthy answer in mystic terms to this question, and does not contain any request or appeal.” He considers (1915, 62) erayadhvam and some other forms as exceptionally preserved old forms from a time when the augment could be left out, when the meaning of past tense was evident from the context.
Since the publication of Hoffmann’s Der Injunktiv im Veda (1967), however, this view can no more be accepted just like that. Hoffmann (160 ff.) rejects the optional use of the augment and concludes (110) that the non-prohibitive injunctive is no longer used in Vedic prose. The few seeming injunctives which exist, are “either formally deviating subjunctives or forms in the past tense which lost their augment secondarily.” With regard to erayadhvam in our verse, he observes that the lack of augment can be explained “from the endeavor, to distinguish the form of the preposition-less imperfect airayadhvam” (108, n. 6). But it is questionable whether the assumption of forms in the past tense matches Hoffmann’s own theories. The succession of imperfect (erayadhvam) and aorist (āsiṣikta), at which Böhtlingk6 already took offence, is in contradiction with Hoffmann’s observation (270), “In the ‘reporting narrative’ of facts of the distant past no aspect difference is indicated, there is always the imperfect.” But is erayadhvam really an imperfect? Would a Brāhmaṇa author let two forms in the past tense follow each other without a recognizable augment, just to keep the preposition ā? In prose texts injunctives are not excluded when they occur in archaizing mantras, as Hoffmann (107, n. 1) observes. On the basis of the contents, it is easy to conceive that our verse is archaizing.7 Admittedly, the second person plural in the injunctive is avoided, since it coincides with the imperative,8 but such a form is not impossible, especially since we are dealing with archaïsms. The difference in function of the imperfect and the injunctive must therefore bring the decision.
Hoffmann (163) sets the “mentioning description” of the injunctive against the “reporting narration” of the imperfect. In a report the speaker tells a fact which he assumes to be unknown to the listener (160). On the other hand, e.g. in a dialogue, shared experiences are not “reported” as news, but are only “mentioned” (199). One could also call the injunctive “memorative.”
Let us now look at the verse. The deceased says: “From the moon the seed has been produced. You, seasons have led me into a man and have poured me into a mother by means of this man. So I am then born, produced by the year as an intercalary month. I know that for sure. Lead me therefore to immortality.” This can hardly be considered as a “reporting narration.” Nothing new is being reported to the gatekeepers . Every deceased recites the same verse. He “mentions” a general truth, describes some important singular facts without giving an ordered narrative. Moreover, the gatekeepers are not interested in details about the deceased, e.g., how he was born at a certain time in a certain place as the son of a certain father. Whoever answers in JB 1, 18 to the sun’s question “Who are you?” with his name or gotra, is sent back. In fact the question is not “Who are you?” but rather “What are you?”
In this entrance examination the deceased shows by means of his answer, i.e. by means of this “memorative, mentioning description” of his (and everyone’s) immortal origin, that he possesses the liberating knowledge. Thus, he concludes his discourse with saṁ tad vide ’ham prati tad vide ’ham “That I know thoroughly, that I know certainly.” The repetition of vid- with different preverbs does not imply a twofold knowledge as one has assumed, but is merely stylistic9 and expresses emphasis, among other things. The fact that saṁ … vide and prati … vide form a unit and are derived from vid- “to know,” is evident from AĀ 2, 3, 1; 4; 6, where one reads the phrase yo ha vai … veda … sa samprativid. I therefore do not believe that Thieme (1951–1952, 26 f.) was correct in separating sam from pratividaḥ in KauṣU 1, 4 and in deriving vide from vid- “to find” in our verse.
In this “mention” before saṁ tad vide the tenses and moods must match. The participle ābhṛtam is probably to be conceived as a statement. The injunctive present (erayadhvam) and injunctive aorist (āsiṣikta) may stand side by side (Hoffmann 1967, 171). The alternation is evoked by an aspectual difference (271 ff.): The production-within-the-man of the seed occurs gradually and is expressed as a progressive action. The pouring-into-the-mother is to be conceived as purely punctual. The present indicative (jāye) may stand in place of the injunctive (165). Here, an event is “mentioned” which is basically free of duration. The context defines the past time period.
If we leave out mā and replace the first person (jāye) with a third, a timeless process is described. Every man has the year for his father; the seed from which every man comes forth is the water of immortality from heaven. See the five-fire-doctrine in JB 1, 45, where amṛtam and āpaḥ are the first sacrifice. The earthly father is only a kartṛ-, a handler of the year (or the seasons), who performs what the real father lets him do. It is the year as the totality of the Ṛtus, who have led the human beings from heaven to earth, who is this real father. The year is equated with Prajāpati and the sun, represents the totality of time and is the prototype of the imperishable.10 The earthly seed is only a phase of the heavenly waters, that is led by the sun as rain to the earth.11 The man, who emerges from it, is as a year’s child equal to the other year’s child, the upamāsa-, in the classifying way of thinking in the Brāhmaṇas. It is interesting to note that in these texts the upamāsa- is considered identical with the year.12 Man is thus identical with his father, the year (= the sun, Prajāpati, Brahman). Whoever is aware of this, is liberated.