LINGUIST List 6.11

Thu 12 Jan 1995

Sum: Names for days

Editor for this issue: <>


  1. Jan Tent, summary of names of days

Message 1: summary of names of days

Date: Wed, 11 Jan 1995 14:11:25 summary of names of days
From: Jan Tent <>
Subject: summary of names of days


In the December 15 issue of LINGUIST I posted a request for
examples of deictic expressions for the names of days either
side of "today". Here is the summary of what I obtained.

The response was overwhelming - a total of 70 LINGUIST
subscribers sent replies. MANY, MANY THANKS TO ALL OF YOU WHO
CONTRIBUTED. Quite a number of respondents supplied
information on two or more languages. There was some
duplication of languages, namely Japanese, Mandarin Chinese,
Danish, French, German, Russian and Hindi. I now have
examples of 48 languages, they are:

 Anejom (Vanuatu), A=FEe A=FEe (Solomon Is), Azerbaijani
 Turkish, Basque, Bauan (Standard Fijian), Brazilian
 Portuguese, Bulgarian, Czech, Danish, Dutch,
 English, Erromagan (Vanuatu), Estonian, Finnish,
 French, German, Hausa, Hawai`ian, Hindustani, Ipili
 (PNG), Italian, Japanese, Kamhmu? (Vietnam), ki-
 Swahili, Kope (PNG), Korean, Lao/Phasa Iisaan,
 Lenakel (Vanuatu), Madarin Chinese, Malay,
 Mauritian/Seychelles Creole, Modern Greek, Modern
 Hebrew, Norwegian, Polish, Quechua, Rumanian,
 Russian, Samoan, Serbo-Croatian, Slovene, Sulka
 (PNG), Swedish, Taiwanese, Thai, Welsh.

My original interest in collecting such deictic expressions
stems from a general interest in symmetry in language and
language systems. Most languages exhibit a symmetry in the
number of deictic expressions for diurnal units either side of
"today", however, some exhibit an asymmetry. C.S. Levinson
(Pragmatics, 1983:75) reports, "The Amerindian language
Chinantec has four named days either side of today; Japanese
has three days back from today, and two ahead; Hindi has the
same word for yesterday and tomorrow." (Levinson gleaned this
information from: Fillmore, C.J. (1975). Santa Cruz Lectures
on Deixis, 1971. Mimeo, Indiana University Linguistics Club.)

I have not personally been able to get a hold of this the
Fillmore reference - living in the academically isolated south
seas does have some drawbacks! If any of you have a copy of
this publication, I would be very interested in obtaining
photocopies of the section on `Time Deixis'. Of course, I
will reimburse you for the cost of postage and photocopying.

Anyway, a symmetrical deictic system of diurnal units/spans
seems to me logical, though I'm not sure I can verbalise the

The reported asymmetry in the Japanese system seemed to me a
bit odd. Especially that the asymmetry was lop-sided towards
the past (see below for reasons). Upon further examination, I
discovered that the Japanese deictic system for diurnal units
does seem to be symmetrical; it has +3 and -3 either side of
the present diurnal span. The expression
`siasatte/yanoasatte' meaning +3 days from "today" is not
known to quite an number of native speakers of Japanese. They
appear to be used infrequently and `yanoasatte and yaneasatte'
are considered to be dialectal (at least in relation to the
Tokyo dialect). Indeed, the lecturer in Japanese in my
department (herself a native speaker of Japanese) was not
aware of these expressions until she looked them up in her

Another interesting consideration in all of this is the
question as to the number of diurnal units before and after
today that languages recognise. What is the maximum/minimum
number of units? Also, what is the average number of units
(or most popular system)? Answers to these questions (based
on my small sample) below.

One respondent asked the following very interesting question:
"If you reach the level of, let's say, 5 (days after present
diurnal span), does this also have the non-specific reading
`at some point in the not-to-far-away future' in those
languages?" The only language in my small corpus that has +5
and -5 diurnal units in its system is Erromangan (Vanuatu). I
will have to ask my source for Erromangan to answer this

There is an inherent problem in determining the number
expressions for diurnal units. This is determining whether
the expressions are lexemes or phrases. Many languages seem
to have "lexicalised" (for want of a better term)
prepositional/nominal phrases.

Other languages, like for example the English expression `the
day before yesterday', have retained the phrase structure.

One respondent supplied me with a lengthy discussion on the
lexeme/phrase issue. She cites the case of Bulgarian (I hope
she won't mind my including her discussion here):


Bulgarian presents an [interesting] situation with regard to
the phrase/lexeme issue because it doesn't have a real nominal
case paradigm:

 onzi den -2 (lit. "that/yonder day")
 vchera -1
 dnes 0 present diurnal span
 utre +1
 vdrugi den +2 (lit. "other day")

So the same form 'pozavchera' in Bulgarian (with stress on the
penultimate syllable; Russian 'pozavchera' has stress on the
final syllable) perhaps doesn't look as strange because one
wouldn't expect 'po' to require any particular case ending.
However, again we have what looks like a preposition with an
adverb, which should not exist as a phrase. Bulgarian
"tomorrow" is 'utre'. But "the day after tomorrow" is
'vdrugiden' -- i.e., the phrase 'v drugi den' (lit. "on (the)
other day", written without spaces in between the individual
words. Should this be considered a word simply on
orthographic grounds, though? Is it a prepositional phrase or
an adverb? Of course the boundary between prepositional
phrases and adverbs is very fuzzy to begin with. Bulgarian
"the day before yesterday", in contrast, is what looks like a
noun phrase: 'onja den' -- "that day". No preposition 'v'
("on") here. So should this be considered just an NP or a
prepositional phrase? 'Onja den' is used the same way as
'vdrugiden' -- 'Vidjax go onja den' -- "I saw him [on] the day
before yesterday". It seems to me that what's written as a
noun phrase here is really an adverb, like 'vdrugiden' or like
Russian 'pozavchera' and 'poslezavtra', the only difference
being that the other three adverbs are formed from
prepositional phrases while 'onja den' has no preposition. It
would sound strange to call it a prepositional phrase,
because, unlike 'vdrugiden', it contains no preposition. So I
would call it an adverbial phrase. (Time expressions in
Slavic languages are commonly in the accusative, and that may
be the origin of 'onja den', but since modern Bulgarian has
lost its nominal case paradigm for the most part this makes it
harder to determine the status of phrases like this one. (Cf.
Russian 'kazhduju nedelju' ("each-Acc week-Acc"), which means
"every week" in an adverbial sense.)

Bulgarian doesn't have specific words for "the day before the
day before yesterday" -- that would be, I think, 'v denja
predi onja den' ("on the day before the day before
yesterday"), but of course speakers would avoid that and just
would say 'predi dva dena' ("two days ago", lit. "before two
days"). The same would be true for 'vdrugiden' -- 'v denja
sled vdrugiden' is conceivably possible ("on the day after
the-day-after-tomorrow"), but speakers would say 'sled dva
dena' ("in/after two days").

So I think it may be very difficult here to draw a demarcation
line here between the grammatical categories of words vs.


Quite a number of languages (especially Slavonic and Germanic
languages) allow recursion of prepositions/prefixes to "add
on" to the already existing deictic expressions. The extent
to which this recursion can be practiced is limited only by
pragmatic constraints. Here are some nice examples:


 ras-ras-ras-alaltaieri(?) -5
 ras-ras-alaltaieri(?) -4
 ras-alaltaieri -3
 alaltaieri -2
 ieri -1
 azi/astazi 0 present diurnal span
 miine +1
 poimiine +2
 ras-poimiine +3
 ras-ras-poimiine(?) +4
 ras-ras-ras-poimiine(?) +5


 (eer-eergisteren) +3
 eergisteren +2
 gisteren -1
 vandaag 0 present diurnal span
 morgen +1
 overmorgen +2
 (over-overmorgen) +3


 (i for-forgars -3)
 i forgars -2
 i gar -1
 i dag 0 present diurnal span
 i morgen +1
 i overmorgen +2
 (i over-overmorgen +3

The -3 and +3 forms are rather colloquial, but nevertheless
perfectly idiomatic. Further recursions of
prefixes/prepositions are only used in jocular ways, and
predominantly by children.

Finally, if you have read this far, you are about to be
rewarded with a summary of the data I have at this point.

Number of diurnal units either side of "today":


 -1 0 +1

 -2 -1 0 +1 +2

 -3 -2 -1 0 +1 +2 +3

 -4 -3 -2 -1 0 +1 +2 +3 +4

 -5 -4 -3 -2 -1 0 +1 +2 +3 +4 +5
 1 EXAMPLE - Erromangan (Vanuatu)


 -2 -1 0 +1 +2 +3

 -2 -1 0 +1 +2 +3 +4
 1 EXAMPLE - Malay

 -2 -1 0 +1 +2 +3 +4 +5
 1 EXAMPLE - Hausa

 -3 -2 -1 0 +1 +2
 Colloquial Spanish (Southern Spain)

It's interesting to note that no language in the sample
employs the -1 0 +1 system. And the majority of asymmetrical
systems are lop-sided towards the + side of the present
diurnal span. Why this should be so, I don't know. However,
this sample is very small and perhaps a larger sample would
show no such lop-sidedness towards the + "today" side.

Are the languages that show a lop-sidedness in a state of
change? Most languages have a symmetrical vowel system e.g.
the same number of front vowels as back vowels. However,
there are quite a number of languages with asymmetrical vowel
systems. What lies behind this asymmetry?

The answer to this question may lie in the following scenario
[From: Crowley, T. (1992) An Introduction to Historical
Linguistics, OUP, pp. 200-201]

Suppose that in a language with a nice symmetrical five vowel

 i u

 e o


the vowel /e/ beocomes raised towards /i/ and ultimately
merges merges with /i/. The resulting asymmetrical system:

 i u



will put structural pressure on the system, and, in time, it
would not be surprising to see /o/ merging with /u/.

If a system becomes asymmetrical (or uneven) so as to create
some sort of `gap', then a change is likely to take place as a
way of plugging that `gap', thereby re-establishing an even
(or symmetrical) system.

If this is indeed a natural process in phonological systems,
perhaps analogous processes occur in other systems, like
temporal deictic systems. A language with an asymmetrical
temporal deictic system may be in the process of historical
change and the `gap' plugged in due time. Japanese may be a
language that is in the process of losing its +3 deictic term.
If this is indeed the case, it would be interesting to see
whether, in due course, this "imbalance" is "balanced" again
by the loss of its -3 deictic term. I don't know how valid
any of this is, I'm just speculating.

Anyway, back to symmetry.

The most symmetrical system of all is that found in Hindi:

 tarso -4
 narso -3
 parso -2
 kal -1
 aaj 0 present diurnal span
 kal (bihaan) +1
 parso +2
 narso +3
 tarso +4

where there are not only an equal number expressions for
diurnal units either side of "today", but the expressions on
either side are the same. Distinction between "before today"
and "after today" is made through use of past or future tense
of the verb.

In looking at a language's temporal deictic system I guess we
must not fail to take account of a number of important
linguistic/cultural aspects. Among these would be:

* The culture's view of time. Is it a linear view or a
 cyclical view?

* The language's tense/aspect systems.

* The historical development of its deictic expressions

In short, we can't study deictics in isolation. We must look
at how it meshes grammatically with the language.

I would be very interested in any of your views on the reasons
for symmetry and asymmetry we find in this area of deixis.

Also, if you are interested in adding to my list of
languages, I'd be very grateful to receive your contribution.

Thanks again to all those generous people who sent me so
much interesting data and opinions.

Yours today, tomorrow and ...

Jan Tent
Department of Literature and Language
School of Humanities
The University of the South Pacific
P.O. Box 1168

TEL: (679) 313900 Ext. 2263
FAX: (679) 305053
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue