LINGUIST List 6.52

Mon 16 Jan 1995

Sum: Reduplication

Editor for this issue: <>


  1. Jan K Lindstrom, 200

Message 1: 200

Date: Mon, 16 Jan 95 15:31:24 +0200
From: Jan K Lindstrom <jklindstwaltari.Helsinki.FI>
Subject: 200

Summary on reduplication

A month before Christmas (in LINGUIST Vol-5-1354) I sent out a
query about *reduplication* and how this strategy is
grammaticalized in languages to convey some facet of
*intensification*. This is a summary based on the 45 replies I
received from LINGUIST netters. Allow me to THANK YOU all
collectively now, since I will not be referring to single
respondents in the following discussion. The topic of
reduplication interests me because I am writing on my PhD on
repetition with respect to intensification and iconicity in Swedish.

The point of departure in the original query were the claims
found in Lakoff & Johnson (Metaphors we live by; 1980)
according to which reduplication may indicate:

 - plural or collective
 - intensification or increase
 - continuation or completion
 - diminution

Thus, 'more' of physical linguistic 'form' corresponds to
'more' in the 'content' of the reduplicated expression as
opposed to a non-reduplicated form -- as an instance of
diagrammatic iconicity (or as a realization of the *conduit* metaphor
in the terms of Lakoff & Johnson).

There are reflections of the outlined principles in colloquial
English in expressions like 'an old old man', 'the music got
louder and louder', 'she talks talks talks', 'the ball bounced
and bounced', 'there's forms forms forms forms', 'it was
sinking down down', etc. I originally mentioned, however, that
these structures were *marked* in a language like English. The
choice of words was perhaps not so lucky. What I meant by
"marked" is that even if we may think that such expressions are
frequent in casual speech, they are rare in more constrained
contexts. A less marked way of intensifying would probably be
to use a specific intensifier or quantifier, e.g. 'a very old
man', 'talks a lot'. I hope no one will be offended by this
simplistic analysis, it is only leading us a bit beside the
point. (For a comprehensive account on "Repetition in English",
see Persson 1974, Univ. of Uppsala, Sw.)

I was in my query more after pointers to and examples of
languages where reduplication as a means of *intensification*
is an integrated, so to speak, a standard characteristic of the
grammar. This was also, more or less, what I got. I will
summarize the outcome in the following; I am aware of that in
most instances the "facts" are probably subject to
qualification, but for the sake of clarity & brevity I will lay
the examples ahead in a rather list-like manner. Furthermore,
the examples certainly represent only a sample of the
languages in the world that have incorporated reduplication in
their system. Note that I have left out -- in order to make it all
more illustrative -- possible tone marks and diacritics of the

1. Plural or collective

* In Malay/Bahasa (Indonesia), full word reduplication turns
singular to plural:
 anak 'child'
 anak-anak 'children'

* In Nahuatl (or "Aztec"), prefixed reduplication with nouns
conveys a plural meaning:
 cih-tli 'hare' (where 'cih-' is the stem)
 ci:cih-tli pl.

 -- Moreover, there are no structurally distinct adjectives in
Nahuatl but certain "adjectival" suffixes that can take on a
plural notion via reduplication:
 -pi:l diminutive suffix
 -pipi:l plural

* Hausa is cited to have reduplication in the formation of
plurals of a limited set of nouns.

* In Japanese, reduplication may turn a noun to a collective,
but the scope of this strategy is said to be marginal:
 hito 'person'
 hitobito 'people'
 kami 'god'
 kamigami 'gods'

* In Mandarin (Chinese), reduplication turns singular to
collective (but this is said to be rare):
 ren 'person'
 renren 'people'

-- Moreover, both in Mandarin and Cantonese *classifier* items
may be reduplicated to convey "universal quantification" in
contrast to more unique reference. This fits readily in the
notion of intensification (or augmentation):
 ge ren 'a person' (Mandarin)
 ge ge ren 'every person'
 douh 'place, there' (Cantonese)
 douh-douh 'everywhere'

* Dakotan patterns plurals of stative intransitive verbs with
reduplication. This is said only to apply to instances with
inanimate subjects though. North American languages in general
may be of interest when considering reduplication but I am short
of examples at present. Following languages were mentioned in the
replies, though: Klamath, Nez Perce, Sahaptin.

2. Intensification or increase

* Mandarin Chinese intensifies adjectives with reduplication:
 xiao 'small'
 xiaoxiao 'very small'
 gaoxing 'happy'
 gaogaoxingxing 'very happy'

As you notice, with disyllabic adjectives the reduplication
pattern is AABB.
-- Cantonese also uses reduplication for augmentation or

* Turkish is, apparently, a point in case. Reduplication of
adjectives indicates intensification or increase in the
following way:
 temiz 'clean'
 tertemiz 'very clean'
 dolu 'full'
 dopdolu 'very full'
 bos 'empty'
 bosbos 'completely empty'

Here we have prefixed reduplication accompanied with a binding
consonant. Interestingly, there do not seem to be "rules" for
which binding consonant should be inserted in a given case.

* Hausa is cited to use reduplication in the intensification of

*Celtic languages have full word reduplication to indicate
intensification. The method is reminiscent of the case of
English, but it may be more "integrated" in the grammar of the
Celtic (are there any opinions about this?). Could the Celtic
model have influenced Germanic languages so that we still today
have rather similar reduplication in the colloquial registers
(just my own modest idea...)? Here are some examples provided
by netters:

-- Welsh:
 ty bach bach 'a very small house'
 oglau cryf cryf 'a very strong smell'

-- Gaelic:
 fada fada 'very long'
 trua trua 'very pity'

-- Breton:
 braz-braz 'very tall'
 This is said to be common especially in negative sentences.

* In colloquial Russian reduplication / repetition of adj's /
adv's has likewise an intensifying function. The use is
probably stylistically similar to repetition in colloquial
English, but it is said to be limited to *predicate*
 belyj-belyj 'very white'
 tixo-tixo 'very quietly'

* It was also pointed that Finnish has prefixed reduplication
in some intensified adjectival forms. This is true, indeed, but
-- what it seems -- wholly lexicalized and non-productive. In
any case, here are some examples I and my colleague enjoyed
coming up with:
 taysi 'full' (umlaut a)
 tapo-taysi 'completely full' (uml. a's & o)
 tyhja 'empty' (uml. a)
 typo-tyhja 'compl. empty' (uml. a's & o)
 puhdas 'clean'
 puti-puhdas 'compl. clean'
 uusi 'new'
 upo-uusi 'brand new'
 pinta 'surface'
 piri-pinta 'right on the surface'
 suomalainen 'Finnish'
 supi-suomalainen 'purely Finnish'
 yksin 'alone'
 ypo-yksin 'compl. alone' (uml. o)
 tiessaan 'lost' (uml. a's)
 tipo-tiessaan 'compl. lost' (uml. a's & o)

The prefixed items 'tapo', 'typo', 'tipo', 'supi', 'puti',
'piri', 'upo', 'ypo' do not seem to mean anything, at least for
the speakers today. It seems that the first vowel or the pair
first consonant & vowel of the stem are reduplicated with a
binding syllable that most often has a 'p' and a vowel, e.g.:
 ta-po-taysi, ty-po-tyhja, u-po-uusi, y-po-yksin

Could this have been productive in some earlier stage of the

* Finally, in Classical (only?) Greek a small number of *verbs*
may have reduplication to communicate some sort of
intensifying, expressive or affective notions:
 pam-phain-ei 'it shines brightly' ((*phan-phan-j-ei)

3. Continuation or completion

Reduplication is, evidently, often used in the formation of
present, progressive or perfective. I take here the view that
present or progressive forms combine with *continuation*,
whereas perfective forms communicate *completion*.

It was nice to discover that "classical" linguistic tradition
can offer fitting data here. It feels appropriate to begin with
these examples.

* In Sanskrit, the perfect stem is formed by reduplication:
 budh- 'know'
 bubodh- perf.
 jan- 'born'
 jajan- perf.

Aorist ('true perfect') roots may also be formed by a kind of
 jan- ajijana- (aor.)

In addition, present stems may be formed by reduplication:
 bhii- 'drink'
 bibhii- pres.

* Classical Greek uses partial reduplication, i.e. prefixation
of the initial consonant of a verb plus the vowel 'e', in stems
for perfect tenses:
 le-lu-k-a 'I have freed'
 pe-poie:-k-a 'I have made'
 ge-grap-tai 'it has been written'

If the root begins with a vowel, the vowel is augmented.

Furthermore, some common verbs take reduplication in their
*present* tense; the initial consonant is reduplicated with the
vowel 'i':
 di-do:-mi 'I give/am giving'
 gi-gno-mai 'I am becoming'
 ti-the:-mi 'I am placing' ((*thi-the:-mi)

* In Latin, perfect forms may involve reduplication:
 curro 'run'
 cucurri perf.

Let us then go over to more living languages.

* In Hausa, reduplication applied to verbs in two different
ways expresses completion or continuity respectively. Modified
suffixed reduplication gives an idea of *completed* action:
 cika 'fill'
 cikakke compl.
 jefa 'throw'
 jefaffe compl.

Prefixed reduplication communicates "something like
continuity", as expressed by the respondent:
 buga 'beat'
 bubbuga 'keep on beating'
 kira 'call'
 kikkira 'call various people'

* In Tagalog (Philippines), reduplication distinguishes
*imperfective* actions from perfective:
 bili root of 'buy'
 bibili irrealis imperfective
 upo root of 'sit'
 uupo irr. imp.
 kuha root of 'get'
 kukuha irr. imp.

Some scholars call irrealis perfective *completed* aspect and
irrealis imperfective *contemplated* aspect. This makes sense,
since irrealis imperfective 'bibili' seems to translate to
'will buy'.

* In Wailevu/Fijian (Austronesia), reduplication is used in so
called "object defocusing" that involves repetition of the
action and a progressive marker:
 Au xau-ta na agone 'I carry the child'
 Au xau-xau jixo 'I'm carrying now'
 Au dola-va na xaatuba 'I open the window'
 Au dola-dola jixo 'I'm opening now'

Moreover, in this language adjectives may be derived from verbs
through reduplication. Even if noted to be "not productive",
the process is fascinating:
 sava-ta 'wash-TR'
 sava-sava-a 'clean'

The adjective could be understood to represent *completed*
action, the result, in a way very similar to how the perfect
participle works in, say, Germanic:
 She has washed the cloth. --) A washed cloth

* In Nahuatl ("Aztec"), reduplication may affect the meaning of
a verb in several ways.
 -- an action carried out in a "systematic" (progressive?) way:
 tequi base for 'to slice' or 'to hack'
 te:tequi 'slice / carve something'
 cho:ca 'weep'
 cho:cho:ca 'weep continuously'

 -- of the same base verbs, an action carried out in a
"random" (involving repetition?) way:
 tehtequi 'hack something up'
 chohcho:ca 'to sob'

 -- Furthermore, reduplication may be recursive (e.g. of 'weep
continuously' & 'sob' respectively):
 cho:cho:cho:ca chohchohcho:ca

 -- Verbs with certain morphologic characteristic may take on
reduplication to indicate *repeated action*:
 tzili:ni 'make a metallic sound'
 tzttzilica 'jingle'
 tzitzilitza 'make something jingle'

* In Afrikaans, reduplication applied to some verbs seems give
a progressive idea:
 Hy loop eet-eet 'He eats continuously while walking'

4. Diminution

There are not many examples of *diminution* in the total of
replies. In a sense, the process of "making something smaller"
could possibly go under the more general strategy
of *intensification or increase* (although it is literally
about "decrease"). Nevertheless, there were a couple of
possibly fitting cases:

* Cantonese may use reduplication to convey *diminution* or

* It was pointed that English (!) has some "diminutive"
reduplication in formations like 'itsy-bitsy', 'eensie-
weensie', 'teeny-tiny'. These carry the content 'very small',
but they belong merely to the child-speak register. In any
case, the examples are diagnostic of the diminutive case.

5. Other

There are uses of reduplication that go beyond the suggested
four categories above. With some restraint, some of these could
perhaps be seen as slight offsprings of the more general
principle of marking *plurality*. I do not go into detailed
examples but mention some most common semantic categories:

 -- distributive reduplication: presupposes two or
more referents that are taken separately (type: 'one by one')

 -- reflexive notions go naturally hand in hand with
distributive (type: 'one another')

 -- the so called *delimitative aspect*, i.e. 'doing
something for a while', in Mandarin and Cantonese does not at
first seem to fit in *continuity* or *completion*. However, the
semantics of the delimitative aspect is said to encompass
"implied repetition":
 tai-tai 'take a look at' (Cant.)
 chang-chang 'have a taste' (Man.)
 shi-shi 'have a go' (Man.)

What would seem logical to me is that a casual, temporally
short action like 'take a look' reaches its termination quicker
than standard 'looking'. I don't know if this kind of an
implicit sense of "rapid completion" could be the motivation of
this reduplication?

Then there are numerous instances of reduplication that
communicate *contrastive emphasis*. These do not necessarily
combine with the above functional categories at all. The
motivation of such reduplication is merely to *make a point*
clearer. The (American) English "double" is a seemingly nice
example of this:
 Shall we rent a CAR car, or would you rather have a jeep?

Of course, emphasis is a sort of *intensification* and, thus,
it touches the theme of our discussion.

 * * * *

Now I think it's time to close this rather lengthy summary.
Thanks for reading it. If you would like to add something to
the discussion -- like comment on my suggestions --, feel free
to e-mail (or snailmail) the "stuff" directly to me. (For
literature, see below)

All the best for 1995, - Jan.

Jan Lindstrom
Dept. of Scandinavian Languages
PB 4
00014 Helsinki University


In the end, I will give some bibliographic hints that were
pointed to me (sorry, the order is random):

Martin, Samuel E. 1988. A reference grammar of Japanese.

Tai, James. 1993. Iconicity: motivations in Chinese grammar. In
Principles and prediction: the analysis of natural language,
ed. by Mushira Eid & Gregory Iverson. Benjamins. 153-174.

Whitney, William Dwight. 1896. A Sanskrit grammar. (e.g.) Harvard
University Press.

Carnochan, J.C. 1957. Gemination in Hausa. In studies in
Linguistic analysis, special vol. of the Philological Society.

Schachter & Otanes. 1972. Grammar of Tagalog. University of

Moravcsik, Edith A. 1978. Reduplicative constructions. In
Universals of human language, vol. 3. Word Structure. ed. by
J.H. Greenberg et al. Stanford University Press. 297-334.

Davis, Stuart. 1988. On the nature of internal reduplication.
In Hammond & Noonan (eds.), Theoretical morphology: approaches
in modern linguistics. Academic Press. 305-323.

Li & Thompson. Mandarin Chinese: A functional reference
grammar. (eg. pp. 28-36)

Newman, Paul. 1989. Reduplicated nouns in Hausa. Journal of
African languages and linguistics vol 8 nr 2 (Oct.). 115-132.

Matthews & Yip. 1994. Cantonese: A comprehensive grammar.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue