In linguistics, reduplication is a morphological process in which the root or stem of a word (or part of it) or even the whole word is repeated exactly or with a slight change.

The classic observation on the semantics of reduplication is Edward Sapir's: "generally employed, with self-evident symbolism, to indicate such concepts as distribution, plurality, repetition, customary activity, increase of size, added intensity, continuance."[1] Reduplication is used in inflections to convey a grammatical function, such as plurality, intensification, etc., and in lexical derivation to create new words. It is often used when a speaker adopts a tone more "expressive" or figurative than ordinary speech and is also often, but not exclusively, iconic in meaning. Reduplication is found in a wide range of languages and language groups, though its level of linguistic productivity varies. Reduplication is found in a wide variety of languages, as exemplified below. Examples of it can be found at least as far back as Sumerian, where it was used in forming some color terms, e.g. babbar "white", kukku "black".[2]

Reduplication is the standard term for this phenomenon in the linguistics literature. Other terms that are occasionally used include cloning, doubling, duplication, repetition, and tautonym when it is used in biological taxonomies, such as Bison bison.

Reduplication is often described phonologically in one of two ways: either (1) as reduplicated segments (sequences of consonants/vowels) or (2) as reduplicated prosodic units (syllables or moras). In addition to phonological description, reduplication often needs to be described morphologically as a reduplication of linguistic constituents (i.e. words, stems, roots). As a result, reduplication is interesting theoretically as it involves the interface between phonology and morphology.

The base is the word (or part of the word) that is to be copied. The reduplicated element is called the reduplicant, often abbreviated as RED or sometimes just R.

In reduplication, the reduplicant is most often repeated only once. However, in some languages, reduplication can occur more than once, resulting in a tripled form, and not a duple as in most reduplication. Triplication is the term for this phenomenon of copying two times.[3] Pingelapese has both forms:[4]

Triplication occurs in other languages, e.g. Ewe, Shipibo, Twi, Mokilese, Min Nan (Hokkien), Stau.[3]

Sometimes gemination (i.e. the doubling of consonants or vowels) is considered to be a form of reduplication. The term dupleme has been used (after morpheme) to refer to different types of reduplication that have the same meaning.

Full reduplication involves a reduplication of the entire word. For example, Kham derives reciprocal forms from reflexive forms by total reduplication:

Another example is from Musqueam Halkomelem "dispositional" aspect formation:

Partial reduplication involves a reduplication of only part of the word. For example, Marshallese forms words meaning 'to wear X' by reduplicating the last consonant-vowel-consonant (CVC) sequence of a base, i.e. base+CVC:

Many languages often use both full and partial reduplication, as in the Motu example below:

Reduplication may be initial (i.e. prefixal), final (i.e. suffixal), or internal (i.e. infixal), e.g.

Internal reduplication is much less common than the initial and final types.

A reduplicant can copy from either the left edge of a word (left-to-right copying) or from the right edge (right-to-left copying). There is a tendency for prefixing reduplicants to copy left-to-right and for suffixing reduplicants to copy right-to-left:

Initial L → R copying in Oykangand Kunjen (a Pama–Nyungan language of Australia):

Internal reduplication can also involve copying the beginning or end of the base. In Quileute, the first consonant of the base is copied and inserted after the first vowel of the base.

In Temiar, the last consonant of the root is copied and inserted before the medial consonant of the root.

Internal R → L copying in Temiar (an Austroasiatic language of Malaysia):

A rare type of reduplication is found in Semai (an Austroasiatic language of Malaysia). "Expressive minor reduplication" is formed with an initial reduplicant that copies the first and last segment of the base:

All of the examples above consist of only reduplication. However, reduplication often occurs with other phonological and morphological process, such as vowel alternation,[6] deletion, affixation of non-reduplicating material, etc.

For instance, in Tz'utujil a new '-ish' adjective form is derived from other words by suffixing the reduplicated first consonant of the base followed by the segment [oχ]. This can be written succinctly as -Coχ. Below are some examples:

Somali has a similar suffix that is used in forming the plural of some nouns: -aC (where C is the last consonant of the base):

This combination of reduplication and affixation is commonly referred to as fixed-segment reduplication.

In Tohono O'odham initial reduplication also involves gemination of the first consonant in the distributive plural and in repetitive verbs:

Sometimes gemination can be analyzed as a type of reduplication.[citation needed]

In the Malayo-Polynesian family, reduplication is used to form plurals (among many other functions):

In pre-1972 Indonesian and Malaysian orthography, 2 was shorthand for the reduplication that forms plurals: orang "person", orang-orang or orang2 "people".[8] This orthography has resurfaced widely in text messaging and other forms of electronic communication.

The Nama language uses reduplication to increase the force of a verb: go, "look;", go-go "examine with attention".

Chinese also uses reduplication: rén for "person", 人人 rénrén for "everybody". Japanese does it too: toki "time", tokidoki 時々 "sometimes, from time to time". Both languages can use a special written iteration mark to indicate reduplication, although in Chinese the iteration mark is no longer used in standard writing and is often found only in calligraphy.

Indo-European languages formerly used reduplication to form a number of verb forms, especially in the preterite or perfect. In the older Indo-European languages, many such verbs survive:

Those forms do not survive in Modern English but existed in its parent Germanic languages. Many verbs in the Indo-European languages exhibit reduplication in the present stem, rather than the perfect stem, often with a different vowel from that used for the perfect: Latin gigno, genui ("I beget, I begat") and Greek τίθημι, ἔθηκα, τέθηκα (I place, I placed, I have placed). Other Indo-European verbs used reduplication as a derivational process: compare Latin sto ("I stand") and sisto ("I remain"). All of those Indo-European inherited reduplicating forms are subject to reduction by other phonological laws.

Reduplication can be used to refer to the most prototypical instance of a word's meaning. In such a case, it is called contrastive focus reduplication. Finnish colloquial speech uses the process; nouns can be reduplicated to indicate genuinity, completeness, originality and being uncomplicated, as opposed to being fake, incomplete, complicated or fussy. It can be thought as compound word formation. For example, Söin jäätelöä ja karkkia, sekä tietysti ruokaruokaa. "I ate ice cream and candy, and of course food-food". Here, "food-food" is contrasted to "junk-food". One may say, "En ollut eilen koulussa, koska olin kipeä. Siis kipeäkipeä" ("I wasn't at school yesterday because I was sick. Sick-sick, that is"); that means that one was actually suffering from an illness instead of making up excuses, as usual.

Words can be reduplicated with their case morphemes, as in lomalla lomalla, where the adessive morpheme (--lla) appears twice. While reduplication is intelligible to most Finns, its usage is confined mostly to subgroups of young women and children (and possibly fathers of young children when they talk to their children). However, most young women and children do not use reduplication. Reduplication has a somewhat childish connotation and may be perceived as annoying.

In Swiss German, the verbs gah or goh "go", cho "come", la or lo "let" and aafa or aafo "begin" reduplicate when they are combined with other verbs.

In some Salishan languages, reduplication can mark both diminution and plurality, with one process being applied to each end of the word, as in the following example from Shuswap. Note that the transcription is not comparable to the IPA, but the reduplication of both initial and final portions of the root is clear: ṣōk!Emē'’n 'knife' reduplicated as ṣuk!ṣuk!Emen'’me’n 'plural small knives' (Haeberlin 1918:159). Reduplication has been found to be a major part of Salish languages.[9]

At 25–50 weeks after birth, typically developing infants go through a stage of reduplicated or canonical babbling (Stark 198, Oller, 1980). Canonical babbling is characterized by repetition of identical or nearly identical consonant-vowel combinations, such as nanana or idididi. It appears as a progression of language development as infants experiment with their vocal apparatus and hone in on the sounds used in their native language. Canonical/reduplicated babbling also appears at a time when general rhythmic behavior, such as rhythmic hand movements and rhythmic kicking, appear. Canonical babbling is distinguished from earlier syllabic and vocal play, which has less structure.

The Proto-Indo-European language used partial reduplication of a consonant and e in many stative aspect verb forms. The perfect or preterite (past) tense of some Ancient Greek,[10] Gothic, Latin, Sanskrit, Old Irish, and Old Norse verbs preserve this reduplication:

Proto-Indo-European also used reduplication for the imperfective aspect. Ancient Greek preserves this reduplication in the present tense of some verbs. Usually, but not always, this is reduplication of a consonant and i, and contrasts with e-reduplication in the perfect:[11]

Reduplication in nouns was rare, the best example being Proto-Indo-European *kʷé-kʷl-os 'wheel' (cf. Lithuanian kãklas 'neck', Sanskrit cakrá 'wheel', Greek κύκλος (kýklos) 'circle'), which doubled *kʷel-o- (cf. Old Prussian kelan 'wheel', Welsh pêl 'ball'),[12] itself likely a deverbative of *kʷelh₁- 'to turn'.

English has several types of reduplication, ranging from informal expressive vocabulary (the first four forms below) to grammatically meaningful forms (the last two below). See also the for cases like flip-flop, dribs and drabs, etc.

Of the above types, only shm-reduplication is productive, meaning that examples of the first three are fixed forms and new forms are not easily accepted.

The double copula is in some cases a type of reduplication, which may be regarded as non-standard or incorrect.

More can be learned about English reduplication in Thun (1963), Cooper & Ross (1975), and Nevins & Vaux (2003).

While not common in Dutch, reduplication does exist. Most, but not all (e.g., pipi, blauwblauw (laten), taaitaai (gingerbread)) reduplications in Dutch are loanwords (e.g., koeskoes, bonbon, (ik hoorde het) via via) or imitative (e.g., tamtam, tomtom).[15] Another example is a former safe sex campaign slogan in Flanders: Eerst bla-bla, dan boem-boem (First talk, then have sex; lit. First blah-blah, then boom-boom). In Dutch the verb "gaan" (to go) can be used as an auxiliary verb, which can lead to a triplication: we gaan (eens) gaan gaan (we are going to get going). The use of gaan as an auxiliary verb with itself is considered incorrect, but is commonly used in Flanders.[16] Numerous examples of reduplication in Dutch (and other languages) are discussed by Daniëls (2000).

Afrikaans makes use of reduplication to emphasize the meaning of the word repeated and to denote a plural or event happening in more than one place. For example, krap means "to scratch one's self," while krap-krap-krap means "to scratch one's self vigorously",[17] whereas "dit het plek-plek gereën" means "it rained here and there".[18] Reduplication in Afrikaans has been described extensively in the literature – see for example Botha (1988), Van Huyssteen (2004) and Van Huyssteen & Wissing (2007). Further examples of this include: "koes" (to dodge) being reduplicated in the sentence "Piet hardloop koes-koes weg" (Piet is running away while constantly dodging / cringing); "sukkel" (to struggle) becoming "sukkel-sukkel" (making slow progress; struggling on); and "kierang" (to cheat) becoming "kierang-kierang" to indicate being cheated on repeatedly.[19]

In Italian reduplication was used both to create new words or word associations (tran-tran, via via, leccalecca) and to intensify the meaning (piano piano "very softly").[citation needed]

Common in Lingua Franca, particularly but not exclusively for onomatopoeic action descriptions: "" ("The Spaniards came, cannonaded, and left. The English came, cannonaded heavily, and left. The French came, trumpeted on bugles, and captured it.")[20]

Spagnoli venir...boum boum...andar; Inglis venir...boum boum bezef...andar; Francés venir...tru tru tru...chapar.

Common uses for reduplication in French are the creation of hypocoristics for names, whereby Louise becomes Loulou, and Zinedine Zidane becomes Zizou; and in many nursery words, like dada 'horsie' (vs. cheval 'horse'), tati 'auntie' (vs. tante 'aunt'), or tonton 'unkie' (vs. oncle 'uncle').

In Romanian and Catalan, reduplication is not uncommon and it has been used for both the creation of new words (including many from onomatopoeia) and expressions, for example,

In colloquial Mexican Spanish it is common to use reduplicated adverbs such as luego luego (after after) meaning "immediately", or casi casi (almost almost) which intensifies the meaning of 'almost'.

The reduplication in the Russian language serves for various kinds of intensifying of the meaning and exists in several forms: a hyphenated or repeated word (either exact or inflected reduplication), and forms similar to shm-reduplication.[21]

Reduplication is a common feature of Irish and includes the examples rírá, ruaille buaille both meaning 'commotion' and fite fuaite meaning 'intertwined'.[22]

Typically all Indo-Aryan languages, like Hindi, Punjabi, Gujarati and Bengali use partial or echoic reduplication in some form or the other. It is usually used to sound casual, or in a suggestive manner. It is often used to mean etcetera. For example, in Hindi, chai-shai (chai means tea, while this phrase means tea or any other supplementary drink or tea along with snacks). Quite common in casual conversations are a few more examples like shopping-wopping, khana-wana. South Asian languages are also rich in other forms of reduplication: morphological (expressives), lexical (distributives), and phrasal (aspectual).

Reduplication also occurs in the 3rd gaṇa (verb class) of the Sanskrit language: bibheti "he fears", bibharti "he bears", juhoti "he offers", dadāti, "he gives". Even though the general idea is to reduplicate the verb root as a prefix, several sandhi rules change the outcome.

There are a number of constructions in Hindi and Urdu that are constructed by reduplication. Nouns, adjectives, verbs, adverbs, pronouns, all have possibility of reduplications.[24][25][26]

In Armenian, reduplication follows the same classification as in Turkish (see below), namely emphatic reduplication, echo reduplication,[27] and doubling. Many appear as lexical entries in Armenian lexicographical sources.

Reduplication is also used in Dravidian languages like Telugu for the same purpose.[33]

Reduplication is a common phenomenon in Bantu languages and is usually used to form a frequentive verb or for emphasis.[35][36]

Semitic languages frequently reduplicate consonants, though often not the vowels that appear next to the consonants in some verb form.[37] This can take the shape of reduplicating the antepenultimate consonant (usually the second of three),[clarification needed] the last of two consonants, or the last two consonants.[38]

In Hebrew, reduplication is used in nouns, adjectives, adverbs and verbs for various reasons:

In Amharic, verb roots can be reduplicated three different ways. These can result in verbs, nouns, or adjectives (which are often derived from verbs).

From the root sbr 'break', antepenultimate reduplication produces täsäbabbärä 'it was shattered'[40] and biconsonantal reduplication produces täsbäräbbärä 'it was shattered repeatedly' and səbərbari 'a shard, a shattered piece'.[41]

From the root kHb 'pile stones into a wall', since the second radical is not fully specified, what some call "hollow", the antepenultimate reduplication process reduplicates the k inserting the vowel a along with the consonant as a place holder for the hollow consonant, which is by some criteria antepenultimate, and produces akakabä 'pile stones repeatedly'.[42][43]

A small number of native Japanese nouns have collective forms produced by reduplication (possibly with rendaku), such as 人々 hitobito "people" (hb is rendaku) – these are written with the iteration mark "々" to indicate duplication. This formation is not productive and is limited to a small set of nouns. Similarly to Standard Chinese, the meaning is not that of a true plural, but collectives that refer to a large, given set of the same object; for example, the formal English equivalent of 人々 would be "people" (collective), rather than "persons" (plural individuals).

Japanese also contains a large number of mimetic words formed by reduplication of a syllable. These words include not only onomatopoeia, but also words intended to invoke non-auditory senses or psychological states, such as きらきら kirakira (sparkling or shining). By one count, approximately 43% of Japanese mimetic words are formed by full reduplication,[44][45] and many others are formed by partial reduplication, as in がささ〜 ga-sa-sa- (rustling)[46] – compare English "a-ha-ha-ha".

The wide use of reduplication is certainly one of the most prominent grammatical features of Austronesian languages.[47]

In the Malay language, reduplication is a very productive process. It is used for expression of various grammatical functions (such as verbal aspect) and it is part in a number of complex morphological models. Simple reduplication of nouns and pronouns can express at least three meanings:

Notice that in the first case, the nasalisation of the initial consonant (whereby /p/ becomes /m/) is repeated, while in the second case, it only applies in the repeated word.

The Māori language (New Zealand) uses reduplication in a number of ways.[48]

Reduplication can convey a simple plural meaning, for instance wahine "woman", waahine "women", tangata "person", taangata "people". Biggs calls this "infixed reduplication". It occurs in a small subset of "people" words in most Polynesian languages.

Reduplication can convey emphasis or repetition, for example mate "die", matemate "die in numbers"; and de-emphasis, for example wera "hot" and werawera "warm".

Reduplication can also extend the meaning of a word; for instance paki "pat" becomes papaki "slap or clap once" and pakipaki "applaud"; kimo "blink" becomes kikimo "close eyes firmly"

The Mortlockese language is a Micronesian language spoken primarily on the Mortlock Islands. In the Mortlockese language, reduplication is used to show a habitual or imperfective aspect. For example, /jææjæ/ means "to use something" while the word /jæjjææjæ/ means "to use something habitually or repeatedly".[49] Reduplication is also used in the Mortlockese Language to show extremity or extreme measures. One example of this can be seen in /ŋiimw alɛɛtɛj/ which means "hate him, her, or it". To mean "really hate him, her, or it," the phrase changes to /ŋii~mw al~mw alɛɛtɛj/.[49]

Pingelapese is a Micronesian language spoken on the Pingelap atoll and on two of the eastern Caroline Islands, called the high island of Pohnpei. Pingelapese utilizes both duplication and triplication of a verb or part of a verb to express that something is happening for certain duration of time. No reduplication means that something happens. A reduplicated verb means that something IS happening, and a triplication means that something is STILL happening. For example, saeng means 'to cry' in Pingelapese. When reduplicated and triplicated, the duration of this verb is changed:

Few languages employ triplication in their language. In Micronesia, Pingelapese is one of only two languages that uses triplication, the other being Mokilese. Reduplication and triplication are not to be confused with tense however. In order to make a phrase past, present, or future tense, a temporal phrase must be used.[50]

Rapa is the French Polynesian language of the island of Rapa Iti.[51] In terms of reduplication, the indigenous language known as Old Rapa uses reduplication consistent to other Polynesian languages. Reduplication of Old Rapa occurs in four ways: full, rightward, leftward, and medial. Full and rightward are generally more frequently used as opposed to the leftward and medial. Leftward and medial only occur as CV reduplication and partial leftward and medial usually denote emphasis.[51]

For the Rapa Language the implementation of reduplication has specific implications. The most evident of these are known as iterative, intensification, specification, diminutive, metaphorical, nominalizing, and adjectival.[51]

Metaphorical (typically comparing an animal action with a human action):[51]

Philippine languages are characterized as having the most productive use of reduplication, especially in Tagalog (the basis of the Filipino language). Reduplication in Tagalog is complex. It can be roughly divided into six types:[52][53][54]

They can further be divided into "non-significant" (where its significance is not apparent) and "significant" reduplication. 1, 2, and 3 are always non-significant; while 5 and 6 are always significant. 4 can be non-significant when used for nouns (e.g. lalaki, "man").[52][53][54]

Full or partial reduplication among nouns and pronouns can indicate emphasis, intensity, plurality, or causation; as well as a diminutive, superlative, iterative, restrictive, or distributive force.[52][53][54]

Adjectives and adverbs employ morphological reduplication for many different reasons such as number agreement when the adjective modifies a plural noun, intensification of the adjective or adverb, and sometimes because the prefix forces the adjective to have a reduplicated stem".[55]

Number agreement for adjectives is entirely optional in Tagalog (e.g., a plural noun does not have to have a plural article marking it):[55]

The entire adjective is repeated for intensification of adjectives or adverbs:

In verbs, reduplication of the root, prefix or infix is employed to convey different grammatical aspects. In "Mag- verbs" reduplication of the root after the prefix "mag-" or "nag-" changes the verb from the infinitive form, or perfective aspect, respectively, to the contemplated or imperfective aspect.[55] Thus:

For Ergative verbs (frequently referred to as "object focus" verbs) reduplication of part the infix and the stem occur:

The complete superlative prefix pagka- demands reduplication of the first syllable of the adjective's stem:

Reduplication is not a productive noun derivation process in Wuvulu-Aua as it is in other Austronesian languages. Some nouns exhibit reduplication, though they are considered to be fossilized.[56]

Verb roots can undergo whole or partial reduplication to mark aspect. Actions that are continuous are indicated by a reduplicated initial syllable. A whole reduplication can also be used to indicate imperfective aspect.[57]

The onomatopoeia in Wuvulu language also uses reduplication to describe the sound. These onomatopoeic words can be used as alienable nouns.