New etymologies for some Japanese time-words.

• Author: Unger, J. Marshall

PHASES OF THE NIGHT

The differences between sunset and nightfall and between daybreak and sunrise figure prominently in many cultures. For example,

[T]he solar depression angle corresponding to the beginning of morning twilight (daybreak) and the end of evening twilight (nightfall) ... was frequently treated by Muslim astronomers because three of the live canonical prayer times are determined by these two moments in time. For religious purposes, the Muslim day begins at sunset: evening prayer must be performed in the interval between sunset and nightfall (i.e., during dusk); night prayer in the interval between nightfall and daybreak; and morning prayer in the interval between daybreak and sunrise (i.e., during dawn)" (Chabas and Goldstein 2003: 222). A similar way of partitioning the night is found in the vocabulary of Old and Early Middle Japanese. The times of sunset and sunrise had two names each: OJ yupu and yupubye (

PARTS OF THE DAY PARTS OF THE NIGHT yupu 'evening, dusk' yupubye 'id.' yopi 'evening twilight' yonaka '(mid)night' akatuki 'morning twilight' asa 'morning, dawn' asita 'id.' piru 'noon, daytime' Although one cannot say that yupu(bye), yopi, yonaka, akatuki, and asita ~ asa denoted sunset, nightfall, midnight, daybreak, and sunrise as precise points in time, the overall pattern clearly shows that the beginning and end of nocturnal darkness, as opposed to the setting and rising of the sun, were of special importance to early Japanese.

The order and senses of these terms can easily be verified by reviewing the examples from pre-modern sources cited in standard lexical references such as NKD and JDH. To take just the most striking instance, asita is found in the sense 'morning' from the earliest texts (Man 'yoshu 19: 4209, c. 750) to at least the start of the seventeenth century (s.v. axita in the Portuguese-Japanese dictionary of 1603-4). Apart from one instance in Nihon shoki (720) in which asita might be interpreted in its modern sense 'tomorrow' or 'the next day' (now usually Sino-Japanese yokujitsu), no authoritative dictionary cites an example of such a use before the late 1700s. Indeed, the compilers of both the NKD (the nearest thing to the OED in Japan (1)) and Iwanami dictionary of archaic Japanese (Ono et al. 1974) explicitly comment on the relationships summarized in the foregoing table and call readers' attention to their pattern with cross-references.

Apart from traditional explanations in disparate cultures for why they distinguish sunset from nightfall and sunrise from daybreak, practical reasons are not hard to identify. Comparing the places where the sun crosses the horizon each day with the places where specific stars or constellations cross it provides the most elementary way to reckon the seasons independently of the weather. Rising and setting stars are also helpful in navigation. The farther one is from the Equator, the greater the number of circumpolar stars, which neither rise nor set, but many easily identifiable bodies rise and set during the night, and since they traverse the sky in a plane oblique to the horizon, it is worth being prepared for the first and last opportunities for sighting, and so natural to name those times.

The etymology of OJ yonaka is clear enough, but the etymologies usually given for the other words mentioned so far as well as some other early Japanese time-words are hardly satisfactory. In this note, I reconsider them from the perspective of the importance of the transitions from sunset to nightfall and daybreak to dawn.

'DAY' PREJUDICE

In English and many other languages, today and yesterday are compounds involving the morpheme for 'day', so it is not surprising that most etymologies of the corresponding Japanese words invoke J hi 'sun, day'. The NKD (s.v. kyoo) lists six mutually incompatible theories (some dating from premodern times) for J kyoo 'today'

Given OJ pirn 'daytime, noon', one might be tempted to derive the pit of kyepu and kinopu from an intermediate form *piu. But OJ pint > J hiru, with invariant r, is attested at least as far back as kyepu > kyoo and kinopu > kinoo, and the kye and kino parts of those words still demand explanation. I propose that the common final pu in these words was not related to OJ pi 'day, sun' or piru at all, but was rather a specialized use of a noun denoting the gaps or junctures in a regularly alternating pattern. In particular, I will argue that pu designated the critical time from sunset to nightfall, and show that this proposal opens the way to plausible reconstructions of the pre-OJ forms of the kye and kino parts of the words in question.

EXISTENCE OF THE CRITICAL NOUN

Early MJ pu 'stitch; knot' is used to denote the crossings of strands in roughly woven material, such as a screen, or the orthogonal parts of a fence (Ono et al. 1974: 1115c; NKD s.v. hu; Martin 1987: 416, 418). Compounds containing the form hu include MJ yahugomo 'many-knotted matting' (seventeenth century) and J takehu 'bamboo joint', the term for a frequently occurring formation of unbreakably connected but non-adjacent stones of the same color in the game of go. The non-standard modern word hu 'fork between fingers' (NKD [ibid.] gives examples from Fukushima and Gunma prefectures) may perhaps be related, assuming that hu, denoting here the knuckles at the base of the fingers, came to refer to interdigital skin between...