Summary: Terms for 'yesterday' and 'tomorrow'

[Malcolm Ross]

----------------------------Original message----------------------------
In early July I sent the message below to LINGUIST and to HISTLING. I
am very grateful to the many people who sent me information. They
were (I hope I haven't omitted anyone) Ignasi Adiego, Henning
Andersen, Peter Austin, Claire Bowern,Daniel Collins, Ivan A.
Derzhanski, Michele Goyens, Matejka Grgic, Charles Gribble, Jared
Grigg, Joachim Grzega, Alik Guilmy, Christian Kay, Harold Koch,
Konstantin Krasukhin, Joost Kremers, Johanna Laakso, Anthony M.
Lewis, Alexis Manaster-Ramer, Bart Mathias, David Nash, Johanna
Nichols, Jurgis Pakerys, Marc Picard, Robert L. Rankin, Robert R.
Ratcliffe, Nick Reid, Laurent Sagart, Raphael Salkie, Marian Sloboda,
Larry Trask, Shigeru Tsuchida, Miguel Carrasquer Vidal, Rémy Viredaz,
Alexey Vorobyov, Cecil Ward, Andrew Wilcox. I am sorry that I have
not been able to reply to everyone individually.

Below the message is a 'summary': it is a mixture of extracts from
the e-mails I received and, where I received similar information from
several people, my attempts to summarise the information received.

There can be no doubt that there is a crosslinguistic tendency for
terms for 'yesterday' to be derived from 'evening' and for terms for
'tomorrow' to be derived from 'morning'.

A few people asked if I was going to write a paper on this, and some
gave me detailed references to pursue. I am not intending to write a
paper, but if anyone else is interested in doing so and would like
fuller information, I am happy to share it, as well as my own work on
Oceanic languages.
_____________________________________

ORIGINAL MESSAGE

I am interested in the diachronic origins of terms for 'yesterday'
and 'tomorrow'. In Germanic, Slavic and Oceanic Austronesian
languages, at least, we find terms for 'yesterday' derived from
'evening' (Slavic, Oceanic) and for 'tomorrow' from 'morning'
(Germanic, Oceanic). I have three questions:

1) Is there a Slavonic specialist who can tell me how Russian
/vchera/ 'evening' is related to /vecher/ 'evening'? I take it
/vchera/ is a case-marked form of /vecher/, but I haven't managed to
locate the details.

2) Are there similar developments in other language families? I
assume there are, and I would be grateful for examples.

3) Has anyone written anything about these developments? Perhaps I
have been looking in the wrong places, but almost everything I have
found about the linguistics of time is either about aspect and tense
(like Comrie's excellent works) or has a strong philosophical bias.
The development of lexical items seems too mundane to command
attention.
_____________________________________

SUMMARY OF REPLIES

INDO-EUROPEAN LANGUAGES IN GENERAL

1. A number of people drew my attention to Carl Darling Buck, A
Dictionary of Selected Synonyms in the Principal Indo-European
Languages, Chicago/London, The University of Chicago Press, under
"Tomorrow" (14.48, p.999) and "Yesterday" (14.49, p.1000). Buck draws
attention to the facts that most of the words he cites for 'tomorrow'
are derived from words for 'morning' and some of those for
'yesterday' from words for 'evening'. He assumes the semantic
developments 'in the morning' 'on the following morning' 'tomorrow'
and 'in the evening' 'in the past evening' 'yesterday'.

He attributes a 'tomorrow' < 'morning' derivation to almost all the
items he cites: Greek, Italian, French, Spanish, Romanian, Irish,
Welsh, Gothic, Old Norse, Danish, Swedish, English, Dutch, German,
Lithuanian, Latvian, Church Slavonic, Serbocroat, Czech, Polish,
Russian and Sanskrit (but not Latin).

He attributes a 'yesterday' < 'evening' derivation only to the items
he cites from Modern Greek (psés), Lithuanian, Latvian, Church
Slavonic, Serbocroat, Czech, Polish and Russian. Most other words are
attributed to a Proto Indo-European root *^ghes.


2. Rémy Viredaz kindly provided the following additional cases from a
range of Indo-European languages:

Yiddish nekhtn, Bavarian n”chten, Swabian n”hti 'yesterday', Middle High
German nehten 'gestern Abend' (Nibelungen +), an old dative, now widespread
in Northern and Central Germany (L. SainÈan, MÈmoires de la SociÈtÈ de
Linguistique de Paris, 12, 1903, 135).

Tsakonian epphÈri 'yesterday', cf. Ancient Greek hespÈra: 'evening' (ibid.).

Latin mane 'in the morning' > e.g. Italian domani, French demain, 'tomorrow'
(but also e.g. Old French main 'morning') (see W. Meyer-L¸bke, Romanisches
Etymologisches W–rterbuch, # 2548 and 5294).

Vedic pra:t·r 'early', 'in the morning', 'the next morning', 'tomorrow'.
Vedic dos.·: 'dark, evening' and 'in the evening': Modern Dardic languages
do:s. (with retroflex s.) 'tomorrow', Nuristani languages e.g. Kati du:s,
'tomorrow' (M. Emeneau, in Ancient Indo-European Dialects, 1966, 138, or R.
L. Turner, A Comparative Dictionary of the Indo-Aryan Languages, 1966, both
after G. Morgenstierne); similarly, Old Iranian daosha(s)-tara- 'Western'
(implying a meaning 'evening' from the basic word): Middle and Modern
Persian do:sh 'the last night, the evening before', Ossetic: Iron dyson /
Digor ÊdosÊ 'yesterday night, yesterday evening' (M. Mayrhofer,
Altindoarisches Etymologisches W–rterbuch, I, 1992, 750).


SLAVONIC

1. I received responses to by question on Russian /vchera/ and
/vecher/ from Henning Andersen, Daniel Collins, Konstantin Krasukhin,
Johanna Nichols and Miguel Carrasquer Vidal. I have attempted to pull
complementary portions out of their responses, which all draw
attention to the same etymology.

Johanna Nichols wrote:
Russian /vecher/ 'evening' and /vchera/ 'yesterday' are related, but
not inflectionally. That is, /vchera/ isn't any case form of
/vecher/, either synchronically (in Russian or any other Slavic
language) or for late Proto-Slavic. The two roots are cognate, but
the vowel alternation is one of (derivational) ablaut. OCS has
/v'chera/ and /vecher"/ (' = front jer, " = back jer; these are
reflexes of short *i and *u respectively).

Henning Andersen wrote:
Common Slavic vIchera is related to CS vecherU as follows:

(i) minor sound change of e to front jer (since lost) conditioned by
the following palatal consonant. Kortlandt has argued that this is a
regular sound change, there are a good handful of examples.

(ii) More importantly, the form is the old instrumental, PIE
*-o-H(1), which yields the regular Li. -uo (final -u (cf. vilku\, but
ger-u/o-j-u. In Common Slavic the o-stem instr. is renewed on the
pattern of u-stems and i-stems (South Slavic -omI//emI) or replaced
by the u-stem ending (West and East Slavic -UmI//-ImI). Since Slavic
merges PIE long *a and *o, the renewal may have been motivated by
this phonetic change. I believe you can find CS vIchera thus
accounted for in Vaillant's Grammaire com. des langues slaves
(Section on instr. of ostems, possibly chapter on adverbs). Stang
argues some place that the stress of the adverb Russ. krugo/m
(contrast regular instr. kru/gom) reflects the accent of the earlier
instr. in CS -a!, which bore the accent in all masc. and neut. nouns
that did not have an accented long (acute) stem vowel.

Miguel Carrasquer Vidal wrote:
Russ. <vc^era>, Pol. <wczoraj> etc. are from C.S. *<vIkera>, according
to Vasmer an archaic instrumental (in accented *-ó:) of *<vekerU> (>
Russ. <vec^er>, Pol. <wieczór> etc.), with reduction of <e> to <I> as
is common in Slavic (a vowel reduction postdating the PIE Nullstufe).

Daniel Collins wrote:
Russian /vchera/ (which has cognates in all the Slavic languages) is
thought by some etymologists to be an archaic o-stem instrumental
singular ending *o: , which was adverbialized prior to the addition
of the morpheme *-mi reflected in the attested o-stem instrumental
case ending (Vasmer, Machek).  Baltic continues to reflect this
archaic form of instrumental; Lithuanian o-stem -u is a reflex of *o:
under acute accent (Endzelins).  The vocalism in the first syllable
(*vikero:) does not have a regular ablaut relation to the other case
forms (nominative singular *vekeros etc.) and seems to reflect
allegro reduction. An alternative explanation is that vchera is a
genitive of time; however, that leaves the accent unexplained.


2. Several people also pointed out that the Slavonic terms for
'tomorrow' are derived from 'morning'.

Daniel Collins wrote: Old Church Slavonic and Old East Slavic zautra
= za 'during' + utra genitive singular 'morning'.

Charles Gribble wrote: 'Tomorrow' is related to 'morning' also in
Slavic: Russ. {zavtra} is *za utra, Serbo-Croatian {sutra} is *s
utra, Bulgarian {utre} is from *utro, etc.

Marian Sloboda wrote: Slovak has /zajtra/ and for "morning" there is
an etymologically different stem (the word is /rano/ connected with
"early"). Czech also /zi:tra/--/rano/, but has also a word /jitro/
(which is not used much, but can be seen in older texts (from 19th,
first half of the 20th century).

Matejka Grgic wrote: Slovene /vc^eraj/ = yesterday and /vec^er/ =
evening, /jutri/ tomorrow and /jutro/ morning.


BALTIC

Jurgis Pakerys wrote:
Lithuanian:
1) vakar "yesterday", cf. vakaras "evening"
2) rytoj "tomorrow", cf. rytas "morning"

Latvian:
1) vakar "yesterday", cf. vakars "evening"
2) rit "tomorrow", cf. rits "morning" (long /i/ in both words)

Johanna Nichols wrote: Lithuanian /vakar/ 'yesterday' and /vakaras/
'evening' have the same vowel grade and different morphology for the
adverb (no ending). ... judging from the facts here there was a
single Balto-Slavic root from which adverb and noun were both derived.


CELTIC

Cecil Ward wrote:
The modern Celtic words for "tomorrow" all do seem to come from a
root meaning "morning". Possibly "in (the) morning".

As for "yesterday", the modern words all come for the old IE root
ghdhyes- but have been influenced by the word for "today". They have
all become combined with the definite article.

Larry Trask wrote: Welsh <yfory> 'tomorrow' is etymologically a
prepositional phrase meaning 'in the morning', from <bore> 'morning',
with mutation.  Irish <i mbaireach> 'tomorrow' has a similar origin.


GREEK

Andrew Wilcox wrote: As for Mod. Greek, the Tegolpoulos-Fytrakis
dictionary (standard serious "family dictionary") gives etymology of
'avrio (=tomorrow) as Ancient Grk. 'avos (=morning). No etymology for
chthes (=yesterday) is given except to note that it is unchanged
since Anc. Greek.

Konstantin Krasukhin  wrote: Greek AURION "tomorrow" can be connected with
Proto-Indo-European *AUS-R; cf. Greek EOS < *AUSOS, Latin AURORA < *AUSOSA etc.


ROMANCE

This is my summary of responses from Michele Goyens, Laurent Sagart,
Larry Trask, Marc Picard and Anthony M. Lewis.

Late or low Latin /de mane/ 'starting from the morning', lit. 'from
morning' is the source of French /demain/,  Provencal /deman/,
Italian /domani/. Latin /maane/ > Romanian /miine/.

Spanish /mañana/ and Portuguese /manha/ each mean both 'morning' and
'tomorrow'.

On the other hand, French /hier/ "yesterday" (and its cognates in
other Romance languages) derives from the Latin adverb /heri/
"yesterday", from IE *ghes-, and thus cognate with /yesterday/ and
German /gestern/.


FINNIC

Johanna Laakso wrote:
In the Finnic languages, the word for 'tomorrow' is obviously related
with 'morning', although the relationship is not quite transparent
for modern speakers. Finnish _huomenna_ is an essive form (old
locative case, typically appearing in crystallized idioms like this)
of _huomen_, which is by now obsolete; the word _huomen_ 'morning'
only appears in the greeting _hyvää huomenta_ 'good morning!' and in
derivatives like _huominen_ 'of tomorrow' (adj.), while the normal
word for 'morning' is _aamu_. In other Finnic languages, other
derivatives of the same stem may appear, as in Estonian, where
_homme_ 'tomorrow' is an opaque form (perhaps, originally illative?)
of the same stem, while another derivative, _hommik_ (< Proto-Finnic
*hoomenikko), is used for 'morning'.

However, the word for 'yesterday', Fi. _eilen_ ~ Est. _eile_ (<
Proto-Finnic ?*eklen) is a Proto-Finnic innovation of an obscure
origin; the attempts by some etymologists to connect it with the word
for 'former, situated before' are not very convincing. There is a
word for 'evening' (Proto-Finnic *ekta-kV-), which is phonologically
not very far away, but the relationship between these two words,
however, remains obscure.


AUSTRALIAN

Nick Reid wrote: In Ngan'gityemerri (non-Pama-Nyungan, NT, Australia)
/kultyi/ 'evening' and /kultyinimbi/ 'yesterday', where -nimbi is a
nominal case suffix marking 'direction from' and 'cause'.

The lexeme for 'tomorrow' is also a root plus agentive/instrumental suffix
/ngunyine-ninggi/ though in this case the root is obscure (to me anyway).


Peter Austin has written an article which shows the correspondence between day
terms and day period terms (tomorrow=morning, yesterday=evening) in
Australian Aboriginal languages.

Harold Koch provided evidence form his database of the
'evening'/'yesterday' and 'morning'/'tomorrow' connections in Arandic
languages and of the 'evening'/'yesterday' connection in Western
Desert languages.


CHINESE

Larent Sagart wrote:

Then there is Classical Chinese xi1 'evening; last night', discussed
in my book of 1999 "The roots of Old Chinese" (Amsterdam: John
Benjamins), on p. 160. I assumed the development 'evening' >
'yesterday evening' for this word because it appears to contain a
root meaning 'night'. In the same book, I also supposed that the
classical and modern Chinese word for "yesterday" (Mandarin zuo2)
also contains the root for 'night' and also came to mean 'yesterday'
out of a basic meaning 'evening', through 'yesterday evening' in both
cases.

The Mandarin word for 'to-morrow'  ming2 tian1, and its classical Chinese
precursor ming2 ri4 also have a good chance of being from an original
meaning 'morning, dawn', since ming2 means 'bright' ri4 means 'sun, day',
and tian1 means 'sky, day'. I do not know of any published references on
this, though.


SIOUAN

Bob Rankin wrote:
In the Siouan language, Kansa, (and other, related languages)
'tomorrow' is /gasi~/. 'Tomorrow' in the sense of 'on tomorrow' or
'during tomorrow' is /gasi~da~/. 'Morning' is /gasi~xci/, literally
'real morning', where /-xci/ is a common intensifier, often
translated 'real' or 'very'.

In Lakota (Teton Dakota) 'evening' has the root /xta/ and 'yesterday'
is /xtaleha~/. The second part, /-leha~/ is a locative deictic 'now,
thus far, at this place'. /hi~ha~na/ 'tomorrow' is paralleled by
/hi~ha~na-xci/ 'forenoon', literally 'real morning' with the /xci/
intensifier noted in Kansa, above.

In my transcription /~/ nasalizes the preceding vowel and /c/ is a
voiceless, unaspirated palato-alveolar affricate = c-hachek; /j/ is the
voiced counterpart of /c/ = j-hachek.


TURKIC

Alik Guilmy wrote:
for example, in Tatar language (Turkic group):

'kich' (evening) - 'kiche' (yesterday)
'irte' (morning) - 'irtege' (tomorrow)

Larry Trask wrote:
Turkic languages also show some of this, though I am no Turkicist,
and I can't readily interpret the data I have here:

Kurtulus Oztopcu et al. (1996), Dictionary of the Turkic Languages, London:
Routledge.

I see here that words for 'tomorrow' and 'morning' are typically
closely related.  Some examples:

        'morning'       'tomorrow'

Kazakh  tangyerteng     yerteng
Kyrgyz  tang erteng     erteng
Tatar   irt”  irt”g”
Turkmen ertir   ertir
Uyghur  ”tig”n      ”t”

Turkish has none of this, but Turkish <sabah> 'morning' is the same word as
Azerbaijani <sabah> 'tomorrow'.  However, this item, with its final
/h/, cannot be native Turkic, and must be borrowed, probably from
Persian.

We also have links with 'evening' and 'yesterday':

        'evening'       'yesterday'

Kazakh  kesh    keshe
Kyrgyz  kech    kechee
Tatar   kich    kich”

The words for 'yesterday' appear to be case-suffixed forms of the
word for 'evening'. Again, Turkish shows none of this.


BASQUE

Larry Trask wrote: I'm afraid Basque <atzo> 'yesterday' and <bihar>
'tomorrow' are morphologically unanalyzable and etymologically
opaque: they cannot be related to <arrats> 'evening' or to <goiz>
'morning' or indeed to anything else at all.


BANTU

Raphael Salkie wrote: In Shona (Bantu, the majority language of
Zimbabwe) the word for TOMORROW is "mangwana", and the word for
MORNING is "mangwanani". I've always assumed that the similarity with
Spanish is just a coincidence...


JAPANESE

Bart Mathias and Shigeru Tsuchida note the following words:

/asa/ 'morning', /asita/ 'tomorrow' ('morning' in earlier Japanese),
/asu/ 'tomorrow' (a little bookish), /asatte/ 'the day after
tomorrow'.

Shigeru Tsuchida wrote further:

ONO Susumu, a specialist in Japanese, suggests that the root as-,
which appears common in the above three words, might have had a
meaning of `dawn' (under the item asu in his dictionary
``Kogo-Jiten'' [Old (Japanese) Dictionary] (Tokyo: Iwanami Publishing
Co. 1974 first ed.)(Old Japanese around 8 to 9 centuries.)

According to the above dictionary, there were two series of words
indicating time in Old Japanese: daytime-oriented, and nighttime-
oriented. (In the following, F indicates a voiceless bilabial
fricative, which has become h in initial position and has been lost
in medial position in Modern Japanese.)

Day-centric: asa --> Firu --> yuFu
            morning  daytime   evening
Night-centric:              yuFumbe --> yoFi --> yonaka --> akatuki --> asita
                             evening     night    midnight   dawn      morning

Asa (Dc) and asita (Nc) both meant substantially the same `morning',
but asita (Nc) meant, by implication, the morning after having passed
a night when something happened, and thus in the middle age the
emphasis began to shift to `tomorrow morning', and finally `tomorrow'
in modern Japanese. (Likewise, yuFumbe (Nc) `evening' now means `last
night' in modern Japanese.)

One reason why there are more words in the night-centric series may
be in the fact that in those days we had a matrilocal society in
Japan, a man visiting a girl at night to copulate. Thus yoFi was the
time when it became dark and a man visited a girl; yonaka was the
time when the man stayed with his girl-friend; akatuki was the time
when it was still dark, but the man had to leave for his house.


SEMITIC

Robert Ratcliffe wrote:
In Classical Arabic the words for 'yesterday'
('amsu) and 'tomorrow' (ghadan) have related words meaning 'evening'
('umsiiya(t)) and 'morning' (ghadaa(t)) respectively. I am not sure how
these forms are related derivationally or historically. 'Yesterday' and
'tomorrow' are adverbial formations, which I suppose are based on noun
forms no longer in use, and the noun forms for evening and morning are
apparently secondary nominal formations which  (I suppose) have replaced
the earlier words under the pressure of ambiguity. (By the way these are
not the normal (= taught in the textbook) words for evening and morning
which are  masaa' and SabaaH.)

In modern spoken Arabic, these forms are largely replaced by words
originally meaning 'early' (tomorrow) (Egyptian bukra), and 'past'
(yesterday) (Egyptian imbaariH).

Ivan A Derzhanski also noted that the nouns seem to be derived from
the adverbs in Arabic, not vice versa.

Joost Kremers wrote:
.... the word for "tomorrow" in Egyptian Colloquial Arabic (ECA) is
"bukra", which in Classical Arabic (from which ECA derives) means
"early morning".


Rémy Viredaz wrote:
Kurdish s at be (with schwa) 'tomorrow', from Arabic s.aba:h. 'morning' (with
emphatic s. and pharyngeal h.), Kurdish s at ba 'morning' (H. Adjarian,
MÈmoires de la SociÈtÈ de Linguistique de Paris, 16, 1910-11, 365).
--
_____________________________________
Dr Malcolm D. Ross
Senior Fellow
Department of Linguistics
Research School of Pacific and Asian Studies
Australian National University
CANBERRA ACT 0200