Adjective reduplication and diagrammatic iconicity

Mario Brdar: Adjective reduplication and diagrammatic iconicity

PROF. DR. SC. MARIO BRDAR Josip Juraj Strossmayer University, Osijek

UDK: 81'373.611=111

Adjective reduplication and diagrammatic iconicity

The present contribution is concerned with some iconic aspects of adjective reduplication expressing intensification in a number of European languages (Germanic, Slavic, Romance and Finno-Ugric) that have so far largely gone unnoticed. First, I consider the role of affixes added to one the two elements in reduplicated constructions, and then the variation in so-called ablaut-motivated reduplications. It is claimed that in both cases we can observe a lengthening of one element, making it morphologi-cally and/or phonologically heavier, which further stresses the idea of intensification.

Key words: reduplication; iconicity; word-formation; intensification; cognitive linguistics.

1. Introduction

1.1. What is reduplication?

In spite of the fact that reduplication intuitively seems to be a relatively simple and widespread phenomenon, it is still surrounded by lot of mystery. A huge body of literature that has accumulated over the last fifty years or so has certainly advanced our knowledge about the phenomenon, but the three basic questions posed by Moravcsik (1978: 302f) about its form, function and distribution within and across languages are still open.

The outcome of reduplication as a morphological process is a single word, i.e. a complex morphological construction consisting of at least two parts. In traditional terminology, one part is considered the base,

489

Brdar, Mario (2013). Adjective reduplication and diagrammatic iconicity. In Liović, Marica (Ed.), Sanjari i znanstvenici. Zbornik u čast 70-godišnjice rođenja Branke Brlenić-Vujić (pp. 489-514). Osijek: Sveučilište Josipa Jurja Strossmayera, Filozofski fakultet.

Zbornik radova u čast Branke Brlenić-Vujić: Sanjari i znanstvenici

i.e. the part that is copied, and the other is considered to be the reduplicant, i.e. a copy of the base. The base is typically an independently existing word, occurring alone, and possibly in combination with other free or bound lexical forms. Consider the following examples from Marshallese (Shetter 2004) in (1) and Zambian English (Crystal 1995) in (2):

(1) wah ‘canoe’ vs. wahwah ‘to go by canoe’ (2) quick-quick ‘very fast’

Typically, there is only one copy of the base, but there may be more than one copy. Some languages exhibit morphological triplication in addition to reduplication, e.g. Thao, an Austronesian language spoken in central Taiwan (3), and Hawaiian (4), as claimed by Blust (2001), or Colloquial Singapore English (5), as demonstrated by Singh and Wee (2002: 519):

(3) a. zay ‘turn’, tau-zay ‘make a turn to left or right’ b. qata-za-zay ‘turn the head from side to side, as when

crossing a street’ c. qata-za-za-zay ‘ceaselessly turn the head from side to side’

(4) a. hiki ‘get to, or reach a place, arrive, come, approach, appear, rise’

b. hi-hiki ‘to come, go’ c. hi-hi-hiki ‘to come or go, of many people’

(5) a. cough-cough ‘keep coughing’ b. cough-cough-cough ‘keep on coughing’

While, as Moravcsik (1978: 308) puts it »[t]here is nothing in the concept of language tacitly assumed here that would put a bound on the number of times a constituent could be reduplicated«, quadruple-cation is rare, quintuplication even more so, but they are nevertheless attested. Cf. some examples for the former from Thao (Blust 2001: 333f), as in (6), and Riau (Gil 2005: 55), as in (7):

(6) a. sas-i ‘deliver it’ b. s-m-a-sas ‘to deliver to someone, as a meal to workers in

the fields’ c. s-m-a-sa-sa-sas ‘deliver on a regular basis, deliver

routinely’

490


(7) Kalau si Pai ambil-ambil-ambil-ambil aja TOP PERS Pai take-REDUP just [Complaining about friend’s behaviour] ‘Pai just takes things all the time’

Gil (2005: 48) also provides an example of vivid demonstrative reduplication which is a case of quintuplication, with four copies of the base:

(8) Ini-ni-ni-ni-ni, dah dia udah boleh, DEM-RED:PROX PFCT 3 PFCT can boleh berga-gabung kami lah, main bola lah can NON.PAT-RED-join 1 CONTR play ball CONTR [Looking at pictures; speaker points to person who is about to join their football team] ‘There, this one here, he can already join us, playing ball’

The absolute record, however, seems to be five copies of the verbal base expressing atelicity in Riau, as reported by Gil (2005: 60):

(9) A, kalau orang itu, sikit EXCL TOP person DEM-DEM-DIST a-little cu-cu-cu-cu-cu-culil, bayar mahal, kalau orang RED-pull pay expensive TOP person perbaiki ini CAUS-good-EP DEM-PROX [About CD-player repairmen] ‘Those people just fiddle a little bit and then you pay a lot, the people who fix things’

1.2. Types of reduplication

According to whether the whole base is copied or not it is possible to distinguish between total, or full, and partial reduplication. Reduplication is total when the entire base is copied, as shown in examples (1-2) above, from Marshallese and Zambian English, respectively. Another example is from Dyirbal (Dixon 1972: 272), in which morphologically simple nominals undergo total reduplication as a way of indicating excessive plurality:

(10) a. midi ‘little one’ b. midi-midi ‘lots of little ones’

491


Partial reduplication copies only a part of the base. The part that is reduplicated can be defined in terms of phonological or segmental units. For example in Agta, plural of nouns can be formed by copying not the whole stem but just its first (consonant-)vowel-consonant sequence (Healey 1960: 7):

(11) a. takki ‘leg’ vs. tak-takki ‘legs’

In French the diminutive of fille ‘girl’ is formed also by partial reduplication: fifille ‘little girl’ (Rainer 1998: 278).

Languages can exhibit both total and partial reduplication. Marshallese, in addition to the total reduplication exemplified in (1) above, also has partial reduplication for verbs expressing the idea of wearing an article of dress of footwear (Moravcsik 1978). It copies the last consonant-vowel-consonant sequence of the base and places it after the base:

(12) a. kagir ‘belt’ vs. kagir-gir ‘wear a belt’ b. takin ‘sock’ vs. takin-kin ‘wear socks’

1.3. The functions of reduplication

When Sapir (1921: 79) notes that

… [n]othing is more natural than the prevalence of reduplication, in other words, the repetition of all or part of the radical element. The process is generally employed, with self-evident symbolism, to indicate such concepts as distribution, plurality, repetition, customary activity, increase of size, added intensity, continuance. (1921: 79)

it appears that he considers it so natural precisely because its iconicity, or »its self-evident symbolism«, as he calls it, is inextricably linked with a number of more or less related concepts it can indicate. In this section we first briefly outline the range of its functions found across the world’s languages, and then consider its iconicity in Part 2.

Reduplication can have a whole range of functions. To give just a few examples, it can be used to express plurality, collectivity, iteration, continuity, intensification, diminution (but surprisingly, not augmentation, i.e. increase in size, contrary to Sapir above),

492


attenuation, change of word class, etc., as shown in the following set of examples from sundry languages:

(13) a. anak ‘child’ vs. anak-anak ‘children’, gunung ‘mountain’ vs. gunung-gunung ‘mountains (Malay)

b. pingan ‘dish’ vs. pingpingan ‘dishes’ (Ilocano)

(14) a. ren ‘person’ vs. renren ‘people’ (Mandarin Chinese) b. rumah ‘house’ vs. rumah-rumah ‘houses/houses

collectively/various houses’ (Indonesian (Rafferty 2002)

(15) a. guyon ‘jest’ vs. guguyon ‘to jest repeatedly’ (Sundanese) (Regier 1998)

b. -pik- ‘to touch’ vs. -pikpik- ‘touch it lightly repeatedly’ (Tzeltal) (Regier 1998)

(16) a. batok ‘bark’ vs batok-batok ‘bark continuously’, uran ‘rain’ vs. uran-uran ‘rain continuously’ (Bikol)

(17) a. temiz ‘clean’ vs. tertemiz ‘very clean’, dolu ‘full’ vs. dopdolu ‘quite full’ (Turkish)

b. sweet-sweet ‘very sweet’, bitter-bitter ‘very bitter’ (Singapore English)

(18) a. kwák ‘my thing’ vs. kwalakwák ‘my small thing’(Agta) (Healey 1960: 6)

b. djidis ‘tooth’ vs. djidjidis ‘little tooth’ (Comox) (Regier 1998)

(19) a. maji ‘wet’ vs. maji-maji ‘somewhat wet’ (Swahili) (Ashton 1952: 316)

b. mahiya ‘ashamed’ vs. mahiyahiya ‘be a little ashamed’, magwalis ‘sweep’ vs. magwaliswalis ‘sweep a little’(Tagalog) (Schachter and Otanes 1972: 340)

(20) a. bwai ‘boy’ vs. bwayi-bwayi ‘boyish’, mashi ‘smash’ vs. mashi-mashi ‘pulpy’ (Jamaican Creole English)

b. osururu ‘fear’ vs. osoru-osoru ‘carefully’ (Japanese)

On the basis of such a diversity of functions, the question as to whether it is possible to order them somehow and link to each other in a sort of network naturally arises. As Moravcsik (1978: 303) puts it, the first question that needs to be answered is »are there any generalizations to be made about the particular meaning distinctions that are conveyed by quantitative form differences and by its

493


particular subtypes...?« And what are the »conditions that determine the distribution of such constructions within and across languages?«

In spite of the amazing diversity exemplified above their functions can be classified under two most general headings. Reduplication either functions in the same way as the addition of grammatical morphology does, or it functions in the same way as the employment of derivational morphology (in word formation).

The former function normally goes hand in hand with the phenomenon of grammaticalization so that reduplication may be(come) the sole exponent of a given grammatical opposition. It is very often the case that if reduplication is employed to express an opposition within a grammatical category in a language, this language is not found to employ parallel inflectional morphology to express part of oppositions within the same category, e.g. if plural is expressed by reduplication, we do not expect some singular-plural oppositions to be expressed by some sort of plural suffixes in that same language. This does not mean that the meaning in question cannot be expressed in some other, non-grammaticalized way, say periphrastically. Although the tendency is for a given reduplication pattern to become the sole exponent of a given grammatical opposition, in some languages there may exist two rival reduplication patterns for the expression of the same lexical concept. According to Gonda (1949: 171), Malay has two reduplications to refer to ‘climbing perch’, peyu-peyu and pĕpuyu, and two reduplications to refer to ‘pipkin’, labulabu and lĕlubu. However, there are also some exceptions. It is interesting that in a system that exhibits a threefold opposition between singular, dual, and plural (above dual), such as Hopi, we have a mix of plural marking strategies. Unexpectedly, dual is marked by the addition of an inflectional affix, while plural above two is marked by reduplication (Travis 1999: 319).

The second function of reduplication is to indicate the change of the word class (the way that some derivational suffixes are used, or conversion), which may be followed by some concomitant changes in meaning, as exemplified in Jamaican Creole and Japanese examples in (20) above.

However, there are many other residual cases in which reduplication effects some sort of ad hoc lexical enrichment, which may, but need not always, be metonymically or metaphorically motivated. It is difficult to generalize over these because they apply to individual

494


bases. Cf. the following examples from Nahuatl (Tuggy 2003: 113), nakas ‘ear’ can be reduplicated into na-nakas-tli, which refers to ‘(a particular edible) mushroom (shaped somewhat like an ear)’, tiopixki is a priest, while reduplicated form ti-tiopixki is used to refer to a species of grasshopper which has a cross on its back. Further, when pachiwi ‘be covered over (usually by accident)’ is compounded with īx- ‘eye, face’ it comes to mean ‘have your face covered, be blindfolded’, but the reduplication īx-pah-pachiwi, however, means ‘have your vision become blurry.’

However, by identifying these two general functions of reduplication across languages, we have still not come anywhere close to either establishing the most schematic function under which all the rest could be subsumed, or which of them could be posited to be the most basic one (and around which the rest may be somehow organized).

Moravcsik (1978: 324) notes that:

Apart from the meanings of increased quantity, intensity, diminution and attenuation which are concepts capable of pulling together many superficially disparate uses of reduplicative constructions, such constructions also serve to differentiate members of one grammatical category from another. Some of these derivational uses of reduplication do appear to be relatable to one or another of the above-mentioned broad meaning categories in that in some cases the difference in meaning between the reduplicated construction and its unreduplicated counterpart is both a difference in basic grammatical category and also a difference of one of the above types; in other cases, however, no such relation is apparent.

Although she concludes that that there is no common denominator for all cases of reduplication, she nevertheless points out that »[a]ll we may note is that such constructions often express meanings related to increased quantity, intensity, diminutiveness and attenuation« (1978: 325).

495


2. On inherent diagrammatic iconicity of reduplication

Now, even if we concentrate just on these above-mentioned most common functions of reduplication and try to account how they relate to each other, the task at hand is not easy, as indicated by a general lack of consensus among scholars dealing with the topic. Before we can consider various proposals, it is necessary to broach the issue of iconicity of reduplication, or its lack, as it seems to be the main bone of contention. For one thing, if there is something like the most basic function of reduplication, it stands to reason to expect that it might be the most iconic one. Secondly, there are apparent problems with the iconicity of reduplication expressing two out of the four functions that Moravcsik mentions above.

In general, we can say that there is iconicity if something in the form of a sign reflects something in the world (normally through some sort of mental operation). Applied to language, it this means that something in the form of a linguistic sign reflects (through its meaning) something in its referent (Mayerthaler 1980, 1988). Diagrammatic iconicity is a special term introduced by Bybee (1985a and b) to refer to a phenomenon whereby the relation between certain features of two or more linguistic entities resemble (or is a diagram of) the relation between certain other features of these entities, or of certain features of some other entities.

Reduplication seems to be inherently iconically motivated because »the structure of the language directly reflects some aspects of the structure of reality« (Haiman 1980: 515). Discussing the conduit metaphor, Lakoff and Johnson point out that reduplication exemplifies one of the submetapors (or perhaps one of the mappings) of the conduit metaphor, i.e. MORE OF FORM IS MORE OF CONTENT (1980: 128):

Reduplication applied to noun turns singular to plural or collective. Reduplication applied to verb indicates continuation or completion. Reduplication applied to adjective indicates intensification or increase. Reduplication applied to a word for something small indicates diminution. The generalization is as follows: A noun stands for an object of a certain kind. More of the noun stands for more objects of that kind. A verb stands for an action.

496


More of the verb stands for more of the action (perhaps until completion). An adjective stands for a property. More of the adjective stands for more of the property. A word stands for something small. More of the word stands for something smaller.

A number of authors have argued against the iconicity of reduplication. Travis (1999) thus points out that if reduplication were purely iconic, it is not clear how it could mean both intensive and moderative. Were it truly iconic, we would expect the intensive reduplication to be, in some sense, bigger than the moderative reduplication, but in Tagalog, it is just the opposite: with verbal roots, the domain of the moderative reduplication is a foot, while the domain of the intensive reduplication is a syllable. Cf. also the case of Hopi dual and plural, described above, where reduplication fails to denote ‘two’. In the present context, the biggest problem is how to account for the fact that reduplication can express both intensification and attenuation with adjectives and adverbs. While the semantic difference between Chinese adjectives formed by reduplication and triplication, as pointed out by Zhang (1987: 379) is puzzling, we still have a concomitant difference in form: the former express attenuation, the latter extreme intensity:

(21)ang ‘red’ vs. ang-ang ‘reddish’ vs. ang-ang-ang ‘extremely red’

More difficult are cases of reduplication of adjectives in Jamaican creole, as discussed by Abraham (2005) and Kouwenberg and Charité (2005), where they can express both intensification and attenuation.

However, it must be noted that it is naive to expect it to be iconic all the times and everywhere, given their typical life-cycle, during which they are often endangered by haplology (cf. Anderson 2009), and during which they assume grammatical functions. On a most general level, iconicity is at work in reduplications because they entail the meaning of their bases, while the opposite is not true. Apparently there is no language in which non-reduplicated forms denote intensification, plurality, repetition, etc., reduplications expressing singularity, single occurrence, etc.

Further, it is also naive to question the iconicity in reduplication because the base is not repeated more than once to refer to an increased number of referents between two and indefinite. As an

497


elegant solution to this problem, we point to the metonymy introduced by Van Huyssteen (2004: 280), when he discusses the repetitive function of Afrikaans verbal reduplications such as lek-lek ‘licking repeatedly’. He claims that the metonymy at work there is TWO

PERFECTIVE EVENTS FOR AN INDEFINITE NUMBER OF EVENTS. The metonymy in question can easily be modified to cover the plurality in the case of nouns as well.

What is more, I propose here a slight elaboration that might make Van Huyssteen’s solution even more acceptable and more widely applicable (e.g. to adjective and adverb intensification, and possibly also to adjective attenuation or deintensification). The first problem at hand is that, assuming a scale with an average, neutral degree on one pole, and an extreme degree on the other, reduplicated adjectives expressing intensification tend to go quite a way towards the extreme end of the scale, rather than remain in the vicinity of the average. In the case of plurality, the actual number is unspecified but can be anywhere on the scale, not necessarily in the ‘large’ part of the spectrum.

The second problem is that quantity is a conceptual domain that is not as appropriate for a direct conceptualization of properties and/or states (which are expressed by adjectives) as it is appropriate in the case of replicated objects and events. Nevertheless, we can start from the assumption that the central function of reduplication is to express the quantification concept of replication with nominal referents. There is, I think, a highest possible degree of diagrammatic iconicity here, reflecting changes in states of affairs that are in experiential terms most easily and most directly perceptible.

The bridge getting us to the realm of qualities as expressed by adjectives is the conceptual metaphor QUALITY IS QUANTITY

(Kövecses 2005: 176), possibly via QUALITY IS SIZE (cf. Goatly 2007: 35f). The job of this metaphor is to prepare ground for metonymic mappings. Instead of working with a single metonymy based on number two, we suggest that the metonymy TWO FOR AN INDEFINITE

NUMBER actually first reduces a normal numerical series to a simple numerical set with just two members, i.e. the cardinal number one and a numerical scale. This scale, including anything between two and an indefinite number of occurrences, is then the conceptual source for another metonymic mapping, i.e. THE-LOWER-END-OF-A-SCALE-FORTHE-WHOLE-SCALE (cf. Brdar-Szabó and Brdar 2010), as the reverse of

498


the metonymy THE-UPPER-END-OF-A-SCALE-FOR-WHOLE-SCALE, discussed by Radden and Kövecses (1999: 32). This second metonymy is what makes it possible for reduplicative constructions to express intensification with adjectives and adverbs.

It is, however, also possible that it copies the whole base but that some parts get modified, i.e. replaced by something, and/or that some elements are added to the base in the copy. This may be referred to as inexact total reduplication. Further, just some part of the base may be copied with or without any further modification and subsequent addition. The former may be called exact partial reduplication, the latter inexact partial reduplication. All the four labels have been used in literature; however, their delimitation may be quite problematic, in particular when a segmentally different copy that contains some additional material is involved. Therefore, all the reduplications in which the reduplicant is somewhat different from the base will be treated here as partial ones. They can be grouped in terms of the difference between the base and the copy, i.e.

according to whether there is an element added to the base in the copy o according to what type of element is added to the copy, and

where (i.e. whether the copy precedes the base or the other way round)

according to whether there is an element that is modified within the copy

oaccording to the type of modification according to what was truncated in the reduplication.

In the remaining part of this chapter I consider some additional iconic aspects of adjective reduplication expressing intensification in a number of European languages (Germanic, Slavic, Romance and Finno-Ugric) that have gone unnoticed. First, I consider the role of affixes added to one the two elements in reduplicated constructions, and then the variation in so-called ablaut-motivated reduplications. It will be claimed that in both cases we can observe additional lengthening of one element, making it morphologically and/or phonologically heavier, which further stresses the idea of intensification.

499


3. Adding morphological and/or phonological weight to reduplicative constructions as extra iconicity

Although a more or less strict distinction can be maintained between syntactic and morphological reduplication, the constructions that I adduce are clearly on the morphological end of the pole, or straddling the boundary between the two. This is in part due to the fact that the grammatical (and orthographic) tradition of certain languages may treat something as a single (but complex) word, while its analog in another language may be treated (and written) as a syntactic construction consisting of two words.

It has been stated above that reduplication is total when the entire base is copied and partial if it copies only a part of the base. However, taking a closer look at the issue, we realize that in theory there are at least four possibilities for reduplication considering the degree of identity between the base and the copy. The total reduplication can copy the entire base in an exact manner, i.e. without any modification. This is what we have seen so far, the possibility that might be referred to as exact total reduplication. In several Slavic languages, intensifying reduplication often goes hand in hand with the addition of inflection. Thus, in Russian we find a series of inflected reduplications, where the copy, which precedes the base, contains the adjectival agreement suffix for the instrumental case. These can be used as adverbs or as adjectives:

(22) давным-давно [past-SUFF-past] ‘very long time ago’, белым-бело [white-SUFF-white] ‘very white’, полным полно [full-SUFF-full] ‘completely full’, пьяным-пьяно [drunk-SUFF-drunk] ‘completely drunk’, поздным-поздно [late-SUFF-late] ‘very late’, раным-рано [early-SUFFearly] ‘very early’, светлым-светло [light-SUFF-light] ‘very light’, темным-темно [darl-SUFF-dark] ‘very dark’, черным-черно [black-SUFF-black] ‘very black’

The same pattern is, not surprisingly, found in Ukrainian too, as shown by the following examples:

(23) a. Еще двое давным-давно заслуживают присуждения these two past-SUFF-past deserved awarding докторской степени... doctoral degree

500


‘the two have long time ago deserved to be awarded the PhD degree’

b. … на дворе темным-темно… on yard dark-SUFF-dark ‘it was very dark in the yard’

Another addition to the left item is found in Ukrainian, but this time it is in the form of a derivational prefix по- added to an adverb, as in повік-віки [PREF-century-century] ‘for centuries’.

In what Moravcsik (1978: 327) considers to be an example of total reduplication in Hungarian, i.e. telisteli [full-SUFF-full] ‘completely full,’ we actually have the copy of adjective teli ‘full’ extended by -s, elsewhere used as an adjectivizing suffix. Similarly, in réges-régi [old-SUFF-old] ‘very old’, we assume that the base is preceded by an extended copy. In adverbs örökkön-örökké [eternal-SUFF-eternal-SUFF] ‘for all times, eternally’ and körös-körül [round-SUFF-round-SUFF] ‘all around’, both constituents contain suffixes.

If we now focus on the additions to the item on the right, we again discover a number of patterns in several languages. South Slavic languages have a couple of so-called intensifying suffixes that can be added to the second adjective in a reduplicative construction of the form Adj + Adj-SUFF. They are not productive at all, i.e. they can be added to a just of handful bases, both in reduplicative constructions and elsewhere, but some of these adjectival constructions are used quite frequently. The ending -cat is apparently shared by Croatian, Serbian, and Slovenian. In Croatian and Serbian it is added to several monosyllabic adjectives, whose base exhibits the structure CVC, e.g. pun puncat [full full-SUFF] ‘completely full’, gol golcat [naked naked-SUFF] ‘completely naked’, sam samcat [alone alone-SUFF] ‘completely alone’, nov novcat [new new-SUFF] ‘brand new’, živ živcat [alive alive-SUFF] ‘completely alive’, sav savcat [whole whole-SUFF] ‘absolutely complete’, crn crncat [black black-SUFF] ‘jet black’, etc. Cf. some Croatian illustrations:

(24) a. Svojim her

sjajni shiny

je vlastitim is own ukras… decoration

novcem money

kupila bought

nov new

novcatnew-SUFF

‘She bought the brand new shiny decoration with her own money’

501


b. Mjesec je pun puncat. moon is full full-SUFF ‘It is a full moon’

The Croatian adjectives cijel cjelcat [whole whole-SUFF] ‘absolutely whole’ and bijel bjelcat [white white-SUFF] ‘absolutely white’ do not conform to the phonological template due to the ijekavian reflex of yat, unlike their Serbian counterparts ceo celcat and beo belcat, respectively.

However, there are also some adjectives that apparently exhibit the CCVC structure, e.g. prav(i) pravcat [genuine genuine-SUFF] ‘absolutely genuine’, zdrav zdravcat [healthy healthy-SUFF] ‘absolutely health’, spor sporcat [slow slow-SUFF] ‘extremely slow’, star starcat [old old-SUFF] ‘extremely old’, as well as some clearly disyllabic adjectives like jedan jedincati [one one-SUFF] ‘the only one’, or isti istacati [same same-SUFF] ‘completely same’. In the latter we note that -a- appears between the base and the suffix.

Dictionaries (e.g. Rječnik JAZU, and Skok 1971-1974) also record some further dialectal variants of the above combinations in which other suffixes are found, e.g. -at, as in ceo celat, go golat; -car, as in isti isticar; -it, as in cjel cjelcit or gol golcit; -acit, as in dug dugacit [long long-SUFF] ‘extremely long’; or -it, as in jedan jediti. Cf. also the following parallel forms: pun puncijat, pun puncan, pun puncijan, pun punan.

All the above adjectival bases belong to the most frequent adjectives in Croatian. They denote some basic adjectival concepts such as age, colour, value, and physical properties (Marković 2007: 151, Marković 2011).

The construction must be considered to be a case of syntactic reduplication not only because the two words are pronounced and written as two separate words, but also because both adjectives carry the declension, i.e. take inflection for case and gender, e.g. punu puncatu [full-FEM.ACC full-SUFF-FEM.ACC], but there is at least one such reduplication in Croatian that is written with a hyphen:

(25) pun-puncat [full-full-SUFF] ‘completely full’

Slovenian also has a number of constructions in which the suffix – cat is added to the second adjective:

502


(26) ... kot je povedal, prav pravcat šok… as has said genuine genuine-SUFF shock

‘… as he said before, a real shock…’

It is interesting that the construction sam samcat is actually a false friend: one might think that that given the same form (order of the base and the suffixed adjective) and the intensifying function, and given that the languages are closely related, the construction should mean the same in all the three languages. Surprisingly, while in Croatian and Serbian its meaning is ‘completely alone’, in Slovenian it means ‘just a single’. The Slovenian construction that means the same as the Croatian or Serbian sam samcat actually exhibits the reverse order, the first adjective is suffixed:

(27) Slovenci prekratki za sam samcat gol Slovenians too-short for one one-SUFF goal ‘Slovenians remain short by a single goal’

(28) … ostal sam samcat sam remained am alone-SUFF alone

‘I remained all alone

Slovenian has another suffix, -(ot)en, added to the second adjective in the construction, as in en enoten [one and one-SUFF] ‘one and the only’, lahko lahkotno [easily easily-SUFF] ‘extremely easily’, or as in the following example:

(29) … cel celoten direktoral personala… whole whole-SUFF directory staff-GEN ‘absolutely the whole staff directory’

We should also mention the Croatian and Serbian adverb reduplicated following the same general pattern, the second item taking the suffix -ice:

(30) Jedva jedvice smo došli. barely barely-SUFF are arrived ‘We were able to arrive only with great difficulties’

The same pattern seems to be at work in the formation of the adverb dan-danas [day-day-SUFF] ‘still to this day’, where the second item carries the suffix -(a)s which derives adverbs from nouns. However, it is better to treat this as a subtractive partial reduplication in which the base is not the noun dan ‘day’ but the adverb danas ‘today’, and the

503


copy is just a part of this base. Cf. the German counterpart with the same pattern of reduplication: tagtälich [day-daily]. This sort of analysis may also be appropriate for reduplications like danodnevni [day-o-today] ‘applying to every day’, from dan ‘day’ + danas ‘today’; raznorazni [various-o-various] ‘various, all kinds’, from razni ‘various’, and vjekovječni [age-o-eternal] ‘lasting for ages’, from vijek ‘age’ + vječni ‘eternal’. It is true that the right-hand constituents are in most cases derived from the left-hand items, but the reduplications as wholes function as intensifications of right-hand constituents.

Bulgarian exhibits reduplicated adjectives in which the second item is extended by the suffix -ичък. The two adjectives are sometimes hyphenated, and sometimes written as two words. Cf. the following examples:

(31) a. Сам-самичък ли си? alone-alone-SUFF Q are ‘Are you completely alone?’

b. хубав хубавичък сайт pretty pretty-SUFF site

‘a very pretty site’

We have seen above in syntactic reduplications in Slavic languages that the second constituent may be preceded by the superlative prefix. The Russian equivalents are spelled with a hyphen, and are considered here to be genuine morphological reduplications, e.g. белыйпребелый [white-SUP-white] ‘very white’. Prefixes раз- and на- also found in the same environment and with the same function, e.g. веселый-развеселый [merry-PREF-merry] ‘in an extremely merry mood’ or большой-пребольшой [big-PREF-big] ‘very big’; крепконакрепко [strongly-PREF-strongly] ‘very strongly’, мелко-намелко [finely-PRE-finely] ‘very finely’, прямо-напрямо [frankly-PREFfrankly] ‘very frankly’, строго-настрого [severely-PREF-severely] ‘very severely, strictly’. A diminutive suffix complex consisting of ень- and -к- can also be attached to the second adjective, as in рано раненько [early-early-DIM-DIM]. This can be further elaborated into even more complex diminutive clusters, e.g. in рано-ранехонько [early-early-DIM-DIM-DIM], рано-ранешенько [early-early-DIMDIM-DIM-DIM], жив-живёхонек [alive-alive-DIM-DIM] ‘very much alive’. What is more we have also come across constructions

504


where both constituents carry suffixations, e.g. холодным холоднёхонек [cold-SUFF-cold-SUFF-DIM-DIM] ‘very cold’.

Ukrainian also has this pattern with the superlative prefix on the second item: сонна-пресонна [sleepy-PREF-sleepy] ‘very sleepy’. It is interesting that the Ukrainian adverb тихо-тихо [silently-silently] can also be modified by diminutive suffixes, but it is the first item that gets extended: тихенько-тихо [silently-DIM-DIM-silently].

Apparently an isolated case of prefixation of the second constiutent in German seems to be treugetreu [faithful-PREF-faithful], cited by Fleischer and Barz (1995: 232).

In Romance languages the base form of an adjective can be followed by a diminutive of the same adjective, which is very similar to what we have seen above in Slavic languages in terms of formally fleshing out the second element in order to reinforce the idea expressed by the adjective. Cf. the following examples from Spanish (32) and Italian (33) (Bartens and Sandström 2006: 356):

(32) Trenzas, peinados a lo rasta, pero lo que no se la da nada mal es cortar el pelito de los bebés. Suave suavito. Con maquinilla. ‘Plaits, rasta hairstyles, but what she is really good at is cutting the hair of babies. Softly, very softly-DIM. With the shaver’

(33) C’era una gallina seccha secchina. ‘Once upon a time, there was an extremelly skinny hen’

In Northern Italian dialects the second element may be reinforced by the suffix -ent, e.g. novo novente ‘very new’.

Hungarian has a handful of adjectives that can appear in well-entrenched reduplicative constructions in which the first adjective is augmented by two suffixes, which is just the opposite of what we have seen in Slavic languages in terms of the ordering. The adjective is first nominalized by the addition of the suffix –ság/ség, which is then followed by –os/-es, which turns nouns into adjectives. The effect of this is the intensification of the property denoted by the adjective: nagyságos nagy [great-SUFF-SUFF great] ‘greatful/respectful’, vénséges vén [old-SUFF-SUFF old] ‘very old’, szépséges szép [beautiful-SUFF-SUFF beautiful] ‘very beautiful’, jóságos jó [good-SUFF-SUFF good] ‘very good’, and szentséges szent [saint-SUFFSUFF saint] ‘most holy’.

505


In addition to these some reduplicative constructions are attested that are nonce-formations. The first adjective is modified by suffix – án which turns adjectives into adverbs. Cf. some recent examples:

(34) a. Simán sima ordinary-SUFF ordinary ‘To wash with absolutely ordina

vízzel water-with

ry water’

megmosakodni... wash

b. Tisztán tiszta fehér liliomból… pure-SUFF pure white lily-from ‘from absolutely pure lilies…’

c. … azután sűrűn sűrű then frequent-SUFF frequent

öntözgetve splashing

d. nem not

ritkán rare-SUFF

ritka rare

madárfajták bird-types

In one of the examples, it is the second element, and not the first that is longer, but it appears that the first element has been truncated, i.e. stripped of its adjectival suffix, so that it looks like a now:

(35) A szin színes zenei programkínálatával... the colour colourful musical offer-of-programmes-with ‘With a really colourful choice of musical programmes…’

In view of the fact that the addition of the suffix makes the second adjective even heavier, i.e. adds more form, and that some of the suffixed adjectives can express intensification in their own right, we can say that we have a sort of double iconicity here: it is not only that the properties in question are expressed by doubling the form, i.e. reduplicating the base, but also by making the second element heavier.

Now we turn to cases in which there is an element that is modified within the copy. These come in two main types. The first type is characterized by onset alternation such that an initial segment of the base (not including any vowel) is replaced by a different segment in the copy (which again does not contain any vowel), which results in phonological overlap between the rest of the constituents, hence their being labelled as rhyming(-motivated) reduplications. In the second type we observe vowel antiphony, i.e. a vowel alternation pattern similar to ablaut, hence the label ablaut-motivated reduplications.

What we have said above applies to English as well, which is quite rich in reduplication with onset alternation. There are at least a couple

506


of such reduplicated adjectives expressing intensity. Cf. some examples:

(36) a. teeny-weeny, teency-wency, eentsy-weentsy ‘very small’ b. easy-peasy ‘very easy’ c. lovely-jubbly ‘excellent’ d. rolly-polly ‘short, round and fat’ e. super-duper ‘excellent’

In the first example, either constituent may be argued for as the base. Teeny is an early 19th century variant of tiny, but weeny has also been around since the late 18th century, formed from wee on the pattern of tiny. Its synomym peewee is a reduplication from wee. Lovely-jubbly, used to refer to good news, was originally part of an advertisement for a frozen orange drink, and later popularized by the sitcom Only Fools and Horses. Rolly-polly is included here as it might imply some sort of intensity or exaggeration in view of its possible metonymic motivation by the link to the verb roll (rolly as an EFFECT-FOR-CAUSE

metonymy in the sense that an entity that is extremely short, round (and fat) is quite likely to be rolly, i.e. roll around. In Australian English slang we find the adjective ridgy-didge ‘true, genuine, dinkum’, considered to be a reduplication of an earlier form ridge, from thieves’ slang word ridge meaning gold or gold coinage. Herkyjerky ‘spasmodic, irregular and unpredictable’ is a reduplication from jerky ‘making sudden starts and stops’. Loosey-goosey is ‘visible loose or relaxed, not tense’.

Pell-mell, not motivated in English, is a late 16th borrowing from French. The present-day French adverb pêle-mêle, goes back to an earlier form, pesle mesle, mesle pesle, reduplication from mesler ‘to mix’.

In Romanian we find rhyme-motivated reduplicated adverbs such as tanda-manda ‘confused, disorganized’, or hurduz-burduz from burduz ’rumble and roll’. Cf. also Hungarian tarka-barka ‘very colourful, from tarka ‘colourful’.

Schicki-micki ‘very stylish, superficial, very chic’, as in That’s a schicki-micki bar, is a borrowing from German. The model was the noun Schicki-micki ‘someone who cares a lot about fashion’, but it is nowadays also used as an adjective. The word is a reduplication of schicklich ‘appropriate’, truncated and extended by suffix -i:

507


(37) Also ich denke nicht, dass das schickimicki ist… well I think not that that fashionable is ‘Well, I do not think that it is fashionable’

A German adverb that is apparently the result of reduplication and expresses intensity is e.g. holterdiepolter ‘helter-skelter’, based on the verb poltern ‘rumble’.

Generally, in ablaut reduplications the vowel of the first constituent is almost always a high vowel while the second constituent contains a low vowel. There is also a tendency for the former to be front and for the latter to be back one. Among the ablaut-motivated reduplicated adjectives and adverbs expressing intensity in English we find:

(38) a. teeny-tiny ‘very small’ b. tip-top ‘excellent’

Ticky-tacky is used both as a noun (‘cheap building material’) and as an adjective. It is a reduplication from adjective tacky ‘cheap, tawdry, badly made and/or lacking in taste’. Teetotaller ‘pledged to total abstinence from intoxicating drink’ may also be mentoned here as it is motivated by ablaut reduplication. According to some sources, it was possibly formed from total with a reduplication of the initial letter T for emphasis (T-totally, i.e. totally totally). The way it is nowadays spelled suggests vowel alternation following abrreviation rather than total reduplication.

Ablaut-motivated reduplicated adjectives expressing intensity can also be found in Hungarian. In some cases both constituents end in suffix -(V)s, which is elsewhere used to derive adjectives from nouns. The pattern seems to be still productive. Cf. some examples:

(39) a. rissz-rossz ‘very bad’, from rossz ‘bad’ b. fidres-fodros ‘very much curled, curly’, from fodor

‘wave, curl’ c. gidres-gödrös ‘full of pits and holes’, from gödör ‘pit’ d. girbegörbe, girbegurba ‘full of curves’ from gödör

‘curves’ e. dimbes-dombos ‘hilly, have a wavelike appearance’,

from domb ‘hill’ f. gizgazos or gizes-gazos ‘very weedy, overgrown with

wee’, from gaz ‘weed’

508


We now finally move to partial reduplications in which there are clear signs of truncation in the copy. Let us first take a look at the so-called reduced reduplication in Estonian. It is a rare type of reduplication exemplified in:

(40) Taevas oli sini-sinine sky was blue blue ‘The sky was blue-blue’

According to Erelt (2008), in Standard Estonian, reduced asyndetic reduplication may occur optionally with adjectives carrying the suffixes -ke(ne) or -ne, especially if they are colour adjectives, such as sinine ‘blue’ above. This suffix can be omitted in the first constituent, as in rohe-roheline ‘green-green’, puna-punane ‘red-red’; őhu-őhuke ‘thin-thin’, pisi-pisike ‘tiny-tiny’ etc. Mäger (1966) describes some more possibilities in various Estonian dialects, e.g. ilu-ilus ‘beautifulbeautiful’, julk-julge ‘brave-brave’, kank-kanged ‘stiff-stiff’, kőverkőveride ‘curvy-curvy’, kerk-kergede ‘light-light’, väiku-väikuke ‘tiny-tiny’ etc.

The few partial reduplications in French often seem to involve truncation. Thus bête ‘dumb, silly’ is reduplicated by repeating the first syllable in front of the base, bébête ‘dull, silly’. Fofolle ‘mad, foolish’ from folle ‘mad’is produced in a similar fashion.

An isolated example of partial reduplication pattern in Serbian (more precisely in a South Serbian dialect spoken in Kosovo) is described in Boissin (1957: 34): beo ‘white’ is reduplicated as bembeo, apparently due to close contact with Turkish (cf. bem beyaz ‘snow white’ from beyaz ‘white’, or ceo ‘whole’ reduplicated as cip ceo and cipanceo.

Lindström (1995) reports on a peculiar reduplication pattern found with Finnish adjectives where the copy (which precedes the base) contains only the initial (C)V segment of the base, followed by a syllable with the structure CV, where C is /p/, /t/ or /r/, while the choice of V is determined by the rules of vowel harmony. Cf. some examples:

(41) a. täysi ‘full’ vs. täpö-täysi ‘completely full’ b. tyhjä ‘empty’ vs. typö-tyhjä ‘completely empty’ c. puhdas ‘clean’ vs. puti-puhdas ‘completely clean’ d. uusi ‘new’ vs. upo-uusi ‘brand new’ e. pinta ‘surface’ vs. piri-pinta ‘right on the surface’

509


f. suomalainen ‘Finnish’ vs. supi-suomalainen ‘purely Finnish’

g. yksin ‘alone’ vs. ypö-yksin ‘completely alone’

4. Concluding remarks

On the basis of an inventory of reduplicative constructions involving adjectives and adverbs in Germanic, Romance, Slavic and Finno-Ugric languages and a comparison of their functions, we can put forward a number of tentative conclusions, which, needless to say, require further testing against data found elsewhere. This overview has by necessity been rather selective, as not all the languages in the four groups can be described in the same detail.

It appears that morphological reduplication seems to be more important and more widespread, as far as the four groups of European languages are concerned, than indicated in Rubino (2005). It appears that the distribution of the reduplicative constructions (from the nonoccurrence to the inventory of types) is on the one hand dictated by the structural givens of the linguistic system, i.e. whether their morphology is predominantly flective, analytic, or agglutinating, whether they rely in word formation more on compounding or affixation, etc.

We have seen that the addition of a suffix occurs, seen from a crosslinguistical point of view, more often in the right-hand constituent than in the left-hand one. This clearly makes the second adjective even heavier, i.e. adds more form, and that some of the suffixed adjectives can express intensification in their own right, we can say that we have a sort of double iconicity here: it is not only that the properties in question are expressed by doubling the form, i.e. reduplicating the base, but also by making the second element heavier.

In cases in which there is an element that is modified within the copy we have observed similar tendencies. These constructions come in two main types. The first type is characterized by onset alternation such that an initial segment of the base (not including any vowel) is replaced by a different segment in the copy (which again does not contain any vowel), which results in phonological overlap between the rest of the constituents, hence their being labelled as rhyming(motivated) reduplications. However, even here we could observe the

510


addition of certain elements such that the extra phoneme is present only in the right-hand element. Alternatively, in some cases the wovel quality of the right-hand element changes, it is very often a longer vowel, or a diphthong instead of a monophtong in the left-hand element.

In the second type we observe vowel antiphony, i.e. a vowel alternation pattern similar to ablaut. Generally, in these reduplications the vowel of the first constituent is almost always a high vowel while the second constituent contains a low vowel. There is also a tendency for the former to be front and for the latter to be back one, i.e. a heavier one.

References

1. Andersen, Henning (2009): »Reduplication in Slavic and Baltic: loss and renewal«, in: Morphology, 19, 2, 113-134.

2. Ashton, Eric O. (1952). Swahili Grammar (Including Intonation). London, New York: Longmans, Green and Co.

3. Bartens, Angela & Niclas Sandström (2006): »Towards a description of Spanish and Italian diminutives within the NSM framework«, in Peeters, Bert & Anna Wierzbicka (eds.): Semantic Primes and Universal Grammar: Empirical Evidence from the Romance Languages. Amsterdam, Philadelphia: Benjamins, 331360.

4. Blust, Robert (2001): »Thao triplication«, in: Oceanic Linguistics, 40, 2, 324-335.

5. Boissin, Henri (1957): »Quelques procédés de renforcement nominal en serbo-croate«, in: Revue des études slaves, 34, 1-4, 3236.

6. Brajdić, Katarina (2010): »Temeljita studija o pridjevu. Ivan Marković: Uvod u pridjev,« Fluminensia 22, 1, 176-181.

7. Brdar-Szabó, Rita & Mario Brdar (2010): »’Mummy, I love you like a thousand ladybirds’: Reflections on the emergence of hyperbolic effects and the truth of hyperboles«, in Burkhardt, Armin & Brigitte Nerlich (eds.): Tropical Truth(s). The Epistemology of Metaphor and Other Tropes. Berlin, New York: Walter de Gruyter, 383-427.

8. Bybee, Joan (1985a): Morphology. A Study of the Relation Between Meaning and Form. Amsterdam, Philadelphia: John Benjamins.

511


9. Bybee, Joan (1985b): »Diagrammatic iconicity in stem-inflection relations«, in Haiman, John (ed.): Iconicity in Syntax. Amsterdam, Philadelphia: John Benjamins, 11-47.

10. Crystal, David (1995): The Cambridge Encyclopaedia of Language. Cambridge: Cambridge University Press.

11. Dixon, R.M.W. (1972): The Dyirbal Language of North Queensland. Cambridge: Cambridge University Press.

12. Erelt, Mati (2008): »Intensifying reduplication in Estonian«, in: Linguistica Uralica, 44, 4, 268-277.

13. Fleischer, Wolfgang & Irmhild Barz (1995): Wortbildung der deutschen Gegenwartssprache. Tübingen: Max Niemeyer.

14. Gil, David (2005): »From repetition to reduplication in Riau Indonesian«, in Hurch, Bernhard (ed.): Studies on Reduplication. Berlin, New York: Mouton de Gruyter, 31-64.

15. Goatly, Andrew. (2007). Washing the Brain: Metaphor and Hidden Ideology. Amsterdam, Philadelphia: John Benjamins.

16. Gonda, Jan (1949): »The functions of word duplication in Indonesian languages«, Lingua, 2, 170-197.

17. Haiman, John (1980): »The iconicity of grammar: Isomorphism and motivation«, in: Language, 56, 515-540.

18. Healey, Phyllis M. (1960): An Agta Grammar. Manila: Bureau of Printing.

19. Kövecses, Zoltán (2005): Metaphor in Culture. Universality and Variation. Cambridge: Cambridge University Press.

20. Lakoff, George & Mark Johnson (1980). Metaphors We Live by. Chicago, London: University of Chicago Press.

21. Mäger, M. (1966): »Intensiivistavad reduplikatiivsőnad eesti murretes«, in: ESA 12, 91-107.

22. Marantz, Alec (1982): »Re reduplication«, in: Linguistic Inquiry, 13, 435-482.

23. Marković, Bojana (2011): »Pridjevske sintagme tipa ‘gol golcat’ u jezičnim priručnicima i rječnicima hrvatskoga standardnog jezika i čakavskoga,« Fluminensia, 23, 1, 23-38.

24. Marković, Ivan (2007): »Repeticija i reduplikacija u hrvatskome«, Suvremena lingvistika, 33, 2, 141-157.

25. Mayerthaler, Willi (1980): »Ikonismus in der Morphologie«, in: Zeitschrift für Semiotik, 2, 19-37.

26. Mayerthaler, Willi (1988): Morphological Naturalness. Ann Arbor: Karoma Publishers.

512


27. Moravcsik, Edith A. (1978): »Reduplicative constructions«, in: Greenberg, Joseph H. (ed.): Universals of Human Language, Vol. 3. Word Structure. Stanford: Stanford University Press, 297-334.

28. Radden, Günter & Zoltán Kövecses (1999): »Towards a theory of metonymy«, in Panther, Klaus-Uwe & Günter Radden (eds.): Metonymy in Language and Thought. Amsterdam, Philadelphia: John Benjamins, 17-59.

29. Rafferty, Ellen (2002): »Reduplication of nouns and adjectives in Indonesian«, in Macken, Marlys (ed.): Papers from the Tenth Annual Meeting of the Southeast Asian Linguistics Society. Tempe: Arizona State University, Program for Southeast Asian Studies, 317-332.

30. Rainer, Franz (1998): »La reduplication francaise du type fifille d’un point de vue diachronique«, In Ruffino, Giovanni (ed.): Atti del XXI Congresso Internazionale di Linguistica e Filologia Romanza, Palermo. Niemeyer: Tübingen, 279-289.

31. Regier, Terry (1998): »Reduplication and the arbitrariness of the sign«, in Gernsbacher, Morton A & Sharon A. Derry (eds.): Proceedings of the Twentieth Annual Conference of the Cognitive Science Society. Lawrence Erlbaum Associates, Mahwah, New Jersey, 887-892.

32. Rubino, Carl (2005): »Reduplication«, in Haspelmath, Martin, Bernard Comrie, Davil Gil & Matthew Dryer (eds.): World Atlas of Linguistic Structures. Oxford: Oxford University Press, 114117.

33. Sapir, Edward (1921): Language. An Introduction to the Study of Speech. New York: Harcourt, Brace.

34. Schachter, Paul & Fe T. Otanes (1972): Grammar of Tagalog. Berkeley: University of California Press.

35. Shetter, William Z. (2004): »No hocus-pocus here! Repeating a word or a syllable«, http://mypage.iu.edu/~shetter/miniatures/reduplication.htm.

36. Singh, Rajendra & Lionel Wee (2002): »On so-called triplication in Colloquial Singapore English and Thao: A response to Blust: In memory of Stan Starosta«, in: Oceanic Linguistics 42, 2, 514-522.

37. Travis, Lisa (1999): »A syntactician’s view of reduplication«, in: Toronto Working Papers in Linguistics 16, 313-331.

38. Tuggy, David (2003): »Reduplication in Nahuatl: Iconicities and paradoxes«, in Casad, Eugene H. & Gary B. Palmer (eds.): Cognitive Linguistics and Non-Indo-European. Berlin, New York: Mouton de Gruyter, 91-133.

513

http://mypage.iu.edu/~shetter/miniatures/reduplication.htm


39. Van Huyssteen, Gerhard B. (2004): »Motivating the composition of Afrikaans reduplications: a cognitive grammar analysis«, in Radden, Günter & Klaus-Uwe Panther (eds.): Studies in Linguistic Motivation. Berlin, New York: Mouton de Gruyter, 269-292.

40. Zhang, Zheng-Sheng (1987): »Reduplication as a type of operation«, in: Papers from the Annual Regional Meeting of the Chicago Linguistic Society, 23, 1, 376-388.

Sažetak

Reduplikacija pridjeva idijagramatička ikoničnost

U prilogu se razmatraju neki dosad zanemareni aspekti ikoničnosti pridjevskih reduplikacija s pojačajnom funkcijom u više europskih jezika (germanskim, slavenskim, romanskim te ugrofinskim). Nakon analize uloge sufiksa koji se dodaju jednom od dva elementa u reduplikacijama, prikazuje se i fonološka varijacija (razlika u odnosu na samoglasnike ili suglasnike u bazi). Tvrdi se da u oba slučaja dolazi do produljavanja jednog elementa (tipično onog drugog), što ga čini morfološki i/ili fonološki težim te se tako dodatno izražava pojačajnost.

Ključne riječi: reduplikacija; ikoničnost; tvorba riječi; pojačajnost; kognitivna lingvistika

514

Adjective reduplication and diagrammatic iconicity

Documents