427
Sanda Lucija Udier
Faculty of Humanities and Social Sciences
University of Zagreb
Croatia
sludier@ffzg.hr
PRAGMATIC CHARACTERISTICS OF INDEFINITE PRONOUNS 
IN CROATIAN AS A SECOND LANGUAGE
1 PRAGMATIC COMPETENCE AND ITS IMPORTANCE IN TEACHING 
 CROATIAN AS A SECOND LANGUAGE
Communicative competence consists of linguistic, sociolinguistic and pragmatic compe-
tence (Vijeće Europe 2005: 13), and pragmatic competence includes competence in dis-
course, competence in planning, and functional competence (Nazalević Čučević 2015: 41). 
In Croatian as a second language (CL2) the speaker at the B2 level starts the conversation, 
successfully converses back and forth and concludes the conversation. When describing 
and recounting the conversation, he makes use of examples and details by connecting his 
statements into clear and consistent discourse by various cohesive means. A command of 
pragmatic competence at the B2 level is manifested in the possibility of formulating what 
is to be said in various ways and in the adaptation of the theme, style and tone to match the 
communication setting (Nazalević Čučević 2015: 52). At the B2 level, the ways to express 
opinions and viewpoints are diverse, structurally complex and communicatively function-
al. In order for all this to be possible, it is necessary, in addition to the denotative, to master 
the connotative level of meaning of lexical units, i.e. the level of meaning that is often 
implicit, invisible, difficult to grasp and difficult to explain, and which greatly affects com-
munication. In teaching CL2, one can often hear students recount an experience where “a 
word did not mean what it means” in a communicative situation. For example, if a person 
comments on something with the IP pronoun: “Svašta!” with an exclamatory intonation, it 
may refer to the primary meaning of that pronoun (‘a set of various unnamed objects
1
’), or 
it can also entail a negative attitude towards something that happened. Students of CL2, and 
probably of other languages, find it difficult to master this level of language use, although 
it becomes necessary at higher levels of language competence (from B2 upwards), because 
unfamiliarity with it can make communication difficult and, ultimately, lead the speaker 
into an awkward situation, like any other unawareness of socio-linguistic conventions.
1 Source: rječnik.hr http://xn--rjenik-k2a.hr/search/?q=sva%C5%A1ta accessed November 29, 2020.
Sanda Lucija Udier: PRAGMATIC CHARACTERISTICS OF INDEFINITE PRONOUNS IN CROATIAN...
UDK 811.163.42'243'367.626.6
DOI: 10.4312/vestnik.13.427-444

428 VESTNIK ZA TUJE JEZIKE/JOURNAL FOR FOREIGN LANGUAGES
Precisely because of the need to include such linguistic usages in teaching, i.e. be-
cause of the need to make it easier for students to master the pragmatic function in anoth-
er language, intercultural pragmalinguistics has been developed (Kecskés 2014, 2015), 
which has the task of analysing the pragmatic level of language in an intercultural con-
text, and the results of the research conducted under the auspices of intercultural pragmat-
ics have been successfully applied in the teaching of other languages.
2 LINGUISTIC DESCRIPTION OF INDEFINITE PRONOUNS
In both international and Croatian linguistics, indefinite pronouns (IP) have been covered 
relatively little in comparison with other types of words.
2.1 Indefinite Pronouns in World Linguistics
The formal and semantic features of IP in international linguistics were most systemat-
ically addressed by Martin Haspelmath, who published a comparative study of IP with 
examples from 40 languages (Haspelmath 1997). In his study, the emphasis is on the 
relationship between the formal and functional characteristics of IP. IP typically appear 
in sequences that refer to ontological categories (person, thing, place, time, manner, etc.), 
and so, for example, English has sequences with the following components: some-, no-, 
every- and any-. Russian has sequences with these components: -to, -nibud’ and -libo, 
koe-, ni-, ljuboj- etc. It turned out that very different languages have a similar typology 
of IP, and the differences in meaning and function that IP can express proved interesting. 
The analysis shows that seemingly equivalent IP in different languages cannot be used in 
exactly the same contexts, so it is necessary to make a comparison between languages in 
order to determine where each of them is used, i.e. to see which pronouns in which lan-
guage cover which of the listed categories (Haspelmath 1997: 76). IP are mostly derived 
from interrogative pronouns by adding indefinite markers. Indefinite markers are usually 
particles which come either as prefixes or as suffixes, and they almost never participate in 
inflection. Often interrogative pronouns can also be used as IP. Related languages show 
a very different distribution of IP (Haspelmath 1997: 293). Languages differ not only by 
the forms of their IP, but also by their number and the functions they express. Therefore, 
the comparison of IP in different languages is fruitful. The semantics and pragmatics 
of IP synchronically and diachronically were covered by Guevara, Aloni, Port and Ši-
mik (2011) by analysing them according to Haspelmath’s functions, and IP expressing 
free choice were analysed in particular (Guilar-Gueveara, Aloni, Port, Schulz and Šimik 
2010), where it was concluded that different IP are actually the result of the fossilization 
of different pragmatic effects. This supports the claim that the pragmatic function is es-
sential when it comes to IP.

429
2.2 Indefinite Pronouns in Croatian Linguistics
Just like in international linguistics, IP are covered very little in Croatian linguistics, 
and apart from Croatian grammars (Silić and Pranjković 2005, Barić et al. 1997, Raguž 
1997), only Lučić (2005) has written about them. Grammars mostly focus on their for-
mation in their descriptions, and a description of meaning is given only by Raguž (1997: 
70–72). Apart from Lučić (2005), there is no other research on the meaning and pragmat-
ic characteristics of IP, and thus none on the possibility of expressing a negative attitude 
by using IP. Expressing a negative attitude with IP belongs to the colloquial register of 
language, and colloquial language is otherwise poorly researched in Croatian linguistics.
2.2.1 The Formation of Indefinite Pronouns
According to E. Barić et al. (1997: 206−207), IP are identical to interrogative and rela-
tive pronouns, and more of them are formed by prefixation, or the addition of one of the 
following prefixes, particles or words: ne-, ni-, i-, sva-, koje-, po-, pone-, gdje-, što- and 
-god-. The prefixes sva-, ni-, i-, što-, koje- are usually followed by the form –šta, and 
rarely or never by the form –što, while the other prefixes go with –što. Expressions com-
posed of the interrogative or relative pronoun together with certain words or forms are 
used as IP, some of which are placed only in front of the pronoun, others only behind the 
pronoun, while others may stand either behind or in front of the pronoun. Ma, makar, bud 
and budi go before the pronoun, while the particle bilo comes either before or after the 
pronoun. The words god, mu drago, ti volja, te volja, hoćeš and hoće are placed after the 
pronoun. The numeral jedan can also function as an IP with a meaning corresponding to 
neki, netko, isti, nitko or nikakav, and the ordinal numeral drugi can as well when con-
trasted with demonstrative pronouns and the adjectives ostali and ini.
Silić and Pranjković (2005: 127ff) state that IP appear in simple and complex forms. 
The simple forms are tko, što, čiji, koji and kakav, while the complex forms are those with 
the prefixes ne-, ni-, i-, sav-, koji-, kakav- and the pronominal suffixes -tko, -što, -čiji, 
-koji, -kakav and -kolik. IP can also be (1) prefixes: tko-, što-, čiji-, koji-, kakav-, kolik- 
and the suffix -god, (2) with the particle ma and the pronouns tko, što, čiji, koji, kakav 
and kolik, (3) with the particle bilo and the pronouns tko, što, čiji, koji, kakav and kolik 
as well as (4) with the pronouns tko, što, čiji, koji, kakav, kolik and the particle god, mu 
drago and god mu drago.
2.2.2 The Morphology of Indefinite Pronouns
When it comes to declension, in IP only the forms tko, što, čiji, koji, kakav and kolik are 
declined, both as independent forms and in combination with prefixes and suffixes (Silić 
i Pranjković 2005: 128). Out of all the IP, only the pronoun kolik has both an indefinite 
and definite declension. When prepositions are used with the pronouns nitko, ništa, ničiji, 
Sanda Lucija Udier: PRAGMATIC CHARACTERISTICS OF INDEFINITE PRONOUNS IN CROATIAN...

430 VESTNIK ZA TUJE JEZIKE/JOURNAL FOR FOREIGN LANGUAGES
nikoji, nikakav and nikolik as well as itko, išta, ičiji, ikoji, ikakav and ikolik, they are 
placed between the prepositions ni- and i- and the pronoun. Prepositions are also placed 
between the prefixes (particles) ma and bilo and the pronouns tko, što, čiji, koji, kakav 
and kolik. Therefore, in declension, the prefixes i- and ni- behave as separate words, 
and so they come before the preposition (ičim, ničim, i s čim, ni s čim, i u čemu, ni u 
čemu). Other pronoun prefixes are never separated in that manner (Raguž 1997: 70−72). 
In declension, pronouns functioning as prefixes in compound sentences are not declined 
(koješta, koječega, koječemu).
Based on the linguistic description of IP in the Croatian language (CL), it can be 
concluded that they form a very complex derivational and morphological system. Corpus 
linguistic research should be used to examine how much of this system is really active in 
the language production of native speakers of the CL, and that part should be taught in 
the instruction of CL2. In second language teaching, the language that its native speakers 
produce on a daily basis should be taught, while other segments can be taught as needed 
(Udier and Gulešić Machata 2011a: 349).
2.2.3 Syntax of Indefinite Pronouns
The use of IP in sentences was analysed by Silić and Pranjković (2005: 132−133), and 
they concluded: (1) the pronouns tko, što, čiji, koji and kakav are used only in compound 
sentences (Naiđete li na kakav trag, obavijestite nas. ‘If you come across any clues, let us 
know.’) and in those sentences they can be replaced with the pronouns netko, nešto, neči-
ji, nekoji and nekakav, and tkogod, štogod, čijigod, kojigod i kakavgod. (2) The pronouns 
nitko, ništa, ničiji, nikoji, nikakav and nijedan can be used either in simple or compound 
sentences which must be negative, and they cannot be replaced by any other IP. (3) The 
pronouns itko, išta, ičiji, ikoji, ikakav, ikolik and ijedan are used in interrogative, collo-
quial and affirmative sentences, and they can replace the pronouns bilo tko, bilo što, bilo 
čiji, bilo koji, bilo kakav and bilo kolik (Ako itko o tome može išta znati, to je on. ‘If any-
one knows anything about it, it’s him.’). (4) The pronouns bilo tko... can be used in both 
simple and compound sentences. And (5) the pronouns ma tko, ma što, ma čiji, ma koji 
and ma kakav can be used in both concessive and conditional sentences (Zadovoljit će me 
ma kakav odgovor. ‘Any answer will satisfy me.’), and they can replace the pronouns tko 
god... kakav god... čiji god..., and others.
D. Raguž (1997: 70−72) lists pronouns in different types of hypothetical sentences: 
(1) interrogative (Poznaješ li ijednoga? ‘Do you know any of them?’); (2) conditional 
(Ako ikoga vidiš, prenesi poruku. ‘If you see anyone, pass the message on.’); (3) compar-
ative (Bilo je ljepše nego ikada. ‘It was nicer than ever.’); (4) declarative (Ne vjerujem 
ikomu. ‘I don’t trust anyone.’); and (5) relative (Čovjek koji išta zna ne bi to napravio. 
‘A man who knows anything would not do that.’). Hypothetical sentences also use an IP 
(interrogative-relative) without any prefix (Je li tko dolazio? ‘Did anyone come?’, Daj mi 

431
koju. ‘Give me some.’, Ako te tko pita, reci da ne znaš. ‘If anyone asks you, say you don’t 
know.’, Ma što odabrao, dobro je. ‘Whatever you choose, it’s good.’). It could be argued 
whether IP with the prefix i- are indeed in explicit sentences, and whether they are attest-
ed in the modern CL. This claim should also be verified by corpus linguistic research.
Radovan Lučić conducted an analysis of IP with constituents bilo and god based on 
the possible positions they occupy in the sentence (Lučić 2005). He showed that the po-
sition of these IP in a sentence depends on their meaning. IP beginning with the prefixes 
sva- can be in the starting position, but they cannot be in the initial position for pronouns 
in i-. When bilo means i-, it cannot stand in the initial position (Bilo tko čini grijeh, čini 
i bezakonje. ‘Whoever commits sin also commits lawlessness.’), but when it means sva-, 
then it can (Bilo što da zatreba, ja sam ovdje. ‘Whatever he needs, I am here.’). In such 
use, the conjunction da is added. When it comes to the final position, it is “reserved” for 
pronouns with the component bilo which has the meaning ‘whichever’, and can have both 
ironic and pejorative meanings (Govore kako pravo na grb ima bilo tko. ‘They say that 
anyone has the right to a coat of arms.’). In colloquial language, it is not uncommon to use 
pronouns with the component god as well (On priča što god! ‘He says whatever.’). Lučić 
also showed the syntagmatic limitations of the use of the particles bilo and god as parts of 
IP: the conjunctions ako, makar and nego can stand in front of compound pronouns with 
bilo, but not in front of complex pronouns with god (Lučić 2005: 443).
2.2.4 The Meaning of Indefinite Pronouns
There are very few texts examining the meaning of IP in the CL, and corpus-linguistic 
research on this topic does not currently exist. Grammars of the CL tend to deal less with 
the meaning of IP and more with their formation and syntax. Barić et al. stated that each 
of these prefixes that participate in the formation of IP, in addition to the meaning of 
indefiniteness, also add additional connotations to the meaning of IP: (a) ne- gives the 
meaning of uncertainty when something is not known or is something one does not want 
to say; (b) ni- gives the meaning of negation and is used in negative sentences in contrast 
with ne- in the affirmative; (c) i- introduces a permissive meaning and entails existence 
in its lowest quantities, measures or values and is therefore the opposite of ni-; (d) sva- 
means wholeness or comprehensiveness; (e) gdje- means occasionalness, sporadicness, 
scattering in time and space, a meaning the prefix po- also has when alongside pronouns 
beginning with the prefix ne-; (f) što- and koje- mean diversity of kind and quality, often 
not the best; and (g) god- means non-selection, indeterminacy and indifference to choice 
(1997: 206−207). It is said, therefore, that the prefixes što- and koje- can have a pejorative 
meaning, but this is not elaborated on in more detail.
Silić and Pranjković only state in principle that IP are used alongside indefinite ob-
jects, beings, traits, concepts and phenomena. Nevertheless, they concretise this assertion 
by listing the particles god, ma (makar) and bilo (mu, vam) drago which, along with the 
Sanda Lucija Udier: PRAGMATIC CHARACTERISTICS OF INDEFINITE PRONOUNS IN CROATIAN...

432 VESTNIK ZA TUJE JEZIKE/JOURNAL FOR FOREIGN LANGUAGES
meaning of emphasis, also have the additional connotation of giving permission (2005: 
127ff). They come with relational words, and the combination of relativity and conces-
sion results in a habitual meaning (Silić and Pranjković 2005: 254), i.e. the meaning of 
repeating something. Thus, the relative pronoun becomes the indefinite koji god = svaki 
= bilo kakav = svakakav (Tko god dođe, neka je dobro došao. ‘Whoever comes, let him 
be welcome.’). Pragmatics is not mentioned by these authors.
D. Raguž states that the number of pronoun words increases by adding prefixes as 
modifiers of meaning, and these modifiers are shared amongst pronouns and adverbs 
(1997: 70−72). Modifiers only come with a basic IP. The prefix ne- does not indicate ne-
gation, but rather indefiniteness, while the prefix ni- means negation (completely), while 
the prefix i- gives the pronouns a permissive and hypothetical meaning to the smallest, 
and not just any, degree. The prefix sva- has the meaning of comprehensiveness, and the 
prefix po- appears before the prefix ne- and gives the meaning of random selection or 
distributivity. The prefixes koje- and što- add a concessive connotation to the quality and 
choice, thus permitting even the worst quality and choice (and so these words often have 
a derogatory meaning). The prefix koje- is associated with all pronoun words, and even 
with pronominal adverbs, while the prefix gdje- has a limited number and limited quality 
of meanings.
The following can come as independent modifiers or suffixes in IP: (1) the suffix 
-god which has the meaning of hypothetical indefiniteness (kakogod, kogagod), and (2) 
the independent modifier god (always after the pronoun) which has a concessive meaning 
with pronouns (Koga god da pitaš, reći će ti to isto. ‘Whoever you ask will tell you the 
same thing.’). The following modifiers are separate: ma, makar, bilo mu (ti, vam...) drago 
– they have a concessive meaning.  Bilo can come before or after the pronominal word 
(tko bilo, bilo tko), as well as either before or after a preposition (s bilo kim, bilo s kim). 
Syntagmata with mu (joj, vam...) drago come only after the pronominal word (komu ti 
drago). Makar and ma come only before the pronominal word (makar što, ma što). These 
modifiers (bilo, što mu drago) can be found in independent sentences as well (Može 
učiniti što mu je drago. ‘He can do what he pleases.’). The modifiers ma, makar and god 
are more rarely found in independent sentences, especially in the colloquial language 
(Može makar što. ‘Sure, anything.’).
In the article Kaj god! R. Lučić (2005) deals with the intensifiers bilo and god and 
with their influence on the meaning and syntax of the complex IP in whose formation 
they play a part. The function of the intensifiers god and bilo is to extend the concessive 
meaning to indefiniteness, arbitrariness, and unlimited possibility of choices. They can 
have an ironic and pejorative meaning (Govore kako pravo na grb ima bilo tko. ‘They 
say that anyone has the right to a coat of arms.’). According to Lučić (2005: 340) the 
pronouns itko and išta mean the existence of at least one being or object of least value, 
quality or measure. The pronouns netko and nešto have the meaning of an indefinite per-
son, an unknown object or phenomenon, they can mean that the noun is not known or not 

433
wanted to be pronounced, and it can also have the meaning of uncertainty. The pronouns 
svatko and sve mean completeness, comprehensiveness regardless of quality and number, 
and indifference to choice.
We can conclude that Croatian grammars (Barić et al. 1997, Silić and Pranjković 
2005) generally do not provide information on pragmatics, so they do not answer the 
question of how the usable context contributes to the meaning of IP. Valuable data on 
the pragmatic value of IP are provided by Lučić (2005: 342) and Raguž (1997: 70–72), 
and their conclusions can be considered the beginning of the research on the pragmatic 
properties of IP in the CL that needs to be conducted.
3 INDEFINITE PRONOUNS IN THE CONTEXT OF TEACHING 
 CROATIAN AS A SECOND LANGUAGE
Because some of them (netko, nešto, sve, svašta, nitko, ništa...) are among the most com-
monly used words in the CL, IP are taught right from the start when learning CL2, most 
often through a lexical approach (Bergovec 2007), while structured teaching is conducted 
from the B2 level onwards (Čilaš Mikulić, Gulešić Machata and Udier 2015: 88−89, 
Udier and Gulešić Machata 2017: 194−196). The formation, syntax and morphology of 
the compound IP deriving from the pronouns tko, što, čiji, koji, kakav and kolik in com-
bination with the prefixes ne-, i-, ni- and sva- are taught systematically, while the IP with 
the prefixes koje-, gdje-, with the suffix -god and with the particles ma and bilo are only 
mentioned with a note that they belong to a higher variety of the Croatian standard lan-
guage and are predominantly found in the written language.
However, there are very few occurrences found in the corpus for part of the IP sys-
tem that is taught, and that shows a discrepancy between the system and usage, which 
leads to the question of which forms and uses should be taught – all those that are in the 
system or those that are closer to the actual usage, i.e. the conclusion that in teaching one 
should pay attention to the sometimes asymmetrical relationship between the system and 
usage, where preference should be given to usage (Udier and Gulešić Machata 2011a: 
340, Udier and Gulešić Machata 2011b: 68). It follows from all of the above that IP are 
a major challenge for CL2 language instructors given their complex semantics, morphol-
ogy and syntax.
4 THE PRAGMATIC POTENTIAL OF INDEFINITE PRONOUNS 
 IN THE CROATIAN LANGUAGE
The pragmatic properties of pronouns in general, and consequently of IP, have not yet 
been dealt with in more detail in Croatian linguistics. The reason is probably the fact that 
Sanda Lucija Udier: PRAGMATIC CHARACTERISTICS OF INDEFINITE PRONOUNS IN CROATIAN...

434 VESTNIK ZA TUJE JEZIKE/JOURNAL FOR FOREIGN LANGUAGES
it is difficult to analyse the meaning of pronouns because it is always highly dependent 
on the context and the meaning of the word they replace. Kuna (2007) noticed the prag-
malinguistic potential of IP, writing about pronominalization as one of the ways of form-
ing euphemisms. Pronominalization refers to the alternation of a content word (i.e. the 
autosemantic word) with a pronoun that acquires lexical meaning in a speech situation 
or linguistic context (Kuna 2007: 105). In his examples he cites demonstrative pronouns 
(vidjeti onu stvar ‘to see that thing – to see one’s sexual organ’, raditi one stvari ‘to 
do those things – to have sexual relations’) and the reflexive-possessive pronoun (imati 
svoje dane ‘to have one’s days – to menstruate’), and of the IP he mentions only nešto 
and sve (učiniti nešto za njega ‘to do something for someone – to do something illegal 
for someone’ and bilo je tu svega ‘here there was everything – a lot of unpleasant things 
happened’). 
Other linguists have also noticed the pragmalinguistic potential of IP (Raguž 1997: 
70–72 and Lučić 2005: 340) in expressing implicit negative attitudes, and their potential 
for such use is also revealed by some humorous non-scientific texts, for example the en-
try for Makaršta on the online dictionary: Ausvajs bite
2
. The pragmalinguistic potential 
of IP as a challenge in acquiring CL2 was also a motivator in conducting this research.
5 RESEARCH
By researching the command of pragmatic properties of IP among CL2 students
3
, specif-
ically in using them to express implicit negative attitudes, an answer was sought to the 
question to what extent CL2 students at the B2 and C1 levels of language competence 
have acquired a command of the meaning of IP and their use in expressing a negative 
attitude.
2 Makaršta − A word which universally negates value. It is a favourite of grandparents when they want to express 
a focused critique of an entity in its entirety, but also in detail, from 1969 onwards. And not only that, this word also 
describes extraordinarily well any and all aspirations, dreams, hopes, efforts and intimate and social values   of young 
people now and onwards. Examples: Ma, on studira nešto neke kompjutere, ma makaršta. “He studies something 
with computers or whatever.”, Internet? Makaršta! “Internet? Whatever that is!”, Ta njegova ženskica.… Makaršta. 
“That little lady of his...whatever”, Deda, vidi, fotografije s Marsa! Deda:  Makaršta… “Grampa, look, photographs 
from Mars! Grampa: Whatever.” It is perhaps interesting to examine the thought demon which this word lets run 
amok. This word is a linguistic perpetual motion machine in a vicious cycle of creation and destruction of existence. 
As a particle it is permanently locked on to the “event horizon”. Makaršta constantly indicates the possibility of the 
existence of some value (makar-) and subsequently demolishes and destroys it with the question word -šta, as if it 
didn’t even exist. So, if we were to seek out the antithesis of the notion of existence and being, we would easily agree 
that “the non-existent” or “what is not” describes one such concept relatively well, relative to “what is”. That is, we 
would easily agree with such a dichotomy only if we are not familiar with “makaršta” – because this word is not a 
soft and stale concept of non-being, not at all! This word shares nothing with anything you know, it stands between 
being and non-being and kills God in both. Example: Sendvič!? Šta jedeš to makaršta! (A sandwich!? Why are you 
eating that rubbish!) https://leksikon.thinking-garment.com/makarsta/ (Accessed November 9, 2020)
3 The term CL2 student is used as a generic term and means “person who learns Croatian as a second language” 
regardless of that person’s age and in what way and in what context they learn CL2.

435
5.1 Research Hypotheses
The research question is concretized by the following hypotheses:
H1: CL2 students at levels B2.1, B2.2 and C1 have mastered the meaning of IP.
H2: CL2 students at levels B2.1, B2.2 and C1 are not as well versed in the pragmatic 
nature of IP (the ability to express a negative attitude) as in their meaning.
H3: With the increase of the language competence of CL2 students, the mastery of the 
pragmatic properties of IP (the ability to express a negative attitude) also increases.
H4: CL2 students at levels B2.1, B2.2 and C1 whose first language is one of the Slavic 
languages have a better command of the pragmatic dimension of IP than students 
whose first language is not Slavic.
H5: CL2 students at levels B2.1, B2.2 and C1 are less proficient in the pragmatic dimen-
sion of IP in production than in comprehension.
5.2 Sample in Research
The research was conducted on a suitable sample of 60 CL2 students who attended se-
mester classes at Croaticum in the 2019/2020 academic year. There were 20 participants 
in the study at the lower level of B2 (B2.1), 22 at the higher level of B2 (B2.2), and 18 
at the level of C1. The first languages of the respondents expressed in percentages are: 
Polish: 22%, Spanish: 15%, English: 10%, German: 8%, Bulgarian: 8%, Italian: 5%, 
Hungarian: 5%, others: 27%. The total number of speakers of Slavic languages was 39%, 
and non-Slavic was 61%.
5.3 Ethical Assumptions of Research
The study Ovladanost pragmatičkim svojstvima neodređenih zamjenica u hrvatskome 
kao inom jeziku (‘The Command of the Pragmatic Properties of IP in Croatian as a 
Second Language’) was approved by the Ethics Committee for Scientific Research at 
the Faculty of Humanities and Social Sciences, University of Zagreb by a decision on 
September 10th, 2019. Respondents signed an informed consent form and were familiar 
with the purpose of the survey and how to conduct it. It was explained to them that the 
results of the research would be calculated in bulk and that their personal data would be 
kept strictly confidential, i.e. available only to the researcher.
5.4 Research Method and Research Instrument
The research was conducted by the quantitative method, and the research instrument was 
a questionnaire intended to determine to what extent CL2 students know the meaning 
Sanda Lucija Udier: PRAGMATIC CHARACTERISTICS OF INDEFINITE PRONOUNS IN CROATIAN...

436 VESTNIK ZA TUJE JEZIKE/JOURNAL FOR FOREIGN LANGUAGES
of IP and how much they recognise the attitude expressed by them. The questionnaire 
consisted of two parts that the respondents answered separately so that the part in which 
they had to indicate the IP would not affect the part in which they are not explicitly men-
tioned. The parts were later linked together using a code that the respondents typed in 
independently.
In the first part of the questionnaire the respondents had to complete six sentences 
with words that corresponded in meaning. The sentences were taken from the hrWaC 2.0 
corpus (Ljubešić and Erjavec 2011),with the IP that was originally in them omitted, and 
words that matched the meaning had to be filled in (with an unlimited number of possi-
ble options). The purpose was to see how many of the words entered would be IP. The 
sentences used were: Hodala bih pognute glave jer bi vojnici na prozorima dobacivali 
(...) Većina takvih radnika je nakon osnovne škole išla na (...) tečajeve upitne obrazovne 
vrijednosti. Vidiš, postoje (...) ljudi, oni koji traže osvetu ili jednostavno žele nekome 
nanijeti zlo. Zadnjih par dana Maja nije dobro spavala i budila se po noći tako da je bila 
sva (...) Nastavi li se situacija u državi kotrljati nizbrdo, bit će (...) iza Nove godine. Lako 
je predvidjeti reakciju normalnog čovjeka, ali od budale možeš (...) očekivati.
The second part of the questionnaire consisted of 12 test items that checked the re-
spondents’ command of the meaning of IP and the recognition of the negative or neutral 
meanings that they had (here they are listed in the case forms in which they appeared 
in the questionnaire): koješta, kojekakve, svakakve, nikakva, svašta, koječemu, nikakav, 
sve, svašta, kojekakvim, svakakvi and svega). Thus, knowledge of the meaning of each 
IP was checked on two test items, in one of which the pronoun had a basic meaning 
(that is, a specific type of indefiniteness), and in the other where it implicitly expressed 
a negative attitude. Examples were chosen in which it was not obvious from the context 
of the sentence what its pragmatic function was. For example sentences like Ljudi su 
poslije toga postali ludi i vjeruju u koješta. (‘People went crazy after that and believe in 
whatever.’) were not included, because from of the adjective “ludi” (crazy) it could easily 
be concluded that the attitude towards what the IP refers to is negative. Examples were 
chosen so that they did not contain vulgar words, political views or politically incorrect 
expressions that could offend someone. Efforts were made to choose short and simple 
sentences so that the respondents’ attention would not be diverted. Sentences used in 
the questionnaire were: Mi bismo vam u Splitu o tome mogli koješta reći. Treba kupiti i 
okititi božićno drvce i još kojekakve druge sitnice. Imao sam diplomu u džepu i morao 
sam raditi svakakve poslove. Mačići su posve ovisni o majčinu mlijeku i ne treba im 
nikakva druga hrana. O njoj se danas piše svašta.  Pričali smo o koječemu, sve onako 
bez reda. To je za mene prije svega nikakav čovjek. Perfekcionisti sve uspiju napraviti 
dobro. Naravno da možete još svašta iskombinirati u ovom receptu. Cilj je udruge da se 
suprotstavi kojekakvim nevladinim udrugama. Danas na tržištu postoje svakakvi modeli, 
od šarenih i veselih do onih vrlo sofisticiranih i modernih. Kad sam vidio tko sudi, znao 
sam da će biti svega. 

437
Since there is no firm criterion by which it could be established with certainty that an 
IP expresses a negative or neutral attitude in a certain context, three independent evalua-
tors were hired, all CL2 experts (M. Čilaš Mikulić, M. Gulešić Machata and D. Matovac), 
and the questionnaire included sentences which all assessors agreed expressed an appro-
priate attitude (negative or neutral, depending on the test item).
5.5 Results of the Research
The results of the research, which relate to knowledge of the meaning of the IP koješta, 
kojekakav, svakakav, nikakav, sve and svašta, as well as to the recognition of the negative 
or neutral attitudes they express, are shown in Table 1. The results show that a very high 
percentage of CL2 students at the B2 and C1 levels of language competence mastered 
the meaning of all IP which were studied (mastery is an acquisition ranging between 
80% and 88%), while their recognition of neutral and, particularly, the negative attitude 
expressed by IP was much weaker (on average 52.8%). They showed a particularly poor 
recognition of a negative attitude expressed by the pronouns koješta and kojekakav (21% 
and 30%, respectively).
Table 1. Knowledge of the meaning of the IP koješta, kojekakav, svakakav, nikakav, sve and 
svašta and recognition of the attitudes which they express
Indefinite 
pronoun
Knowledge  
of meaning
Recognition  
of a neutral attitude
Recognition  
of a negative attitude
koješta 80% 58% 21%
kojekakav 81% 75% 30%
svakakav 85% 54% 63%
nikakav 88% 61% 73%
sve 86% 59% 66%
svašta 86% 49% 64%
When it comes to the results obtained from the analysis of the second part of the 
questionnaire in which the respondents had to write in words that match according to 
their meaning, and which express a negative attitude (Table 2), respondents at the B2.1 
level wrote an average of 5.70 words, of which 0.95 or 16.67% were IP, respondents at 
the B2.2 level wrote an average of 6.14 words, of which 1.00 word were IP (16.29%), 
while respondents at the C1 level entered an average of 8.50 words, of which 1.50 words 
were IP (17.65%).
Sanda Lucija Udier: PRAGMATIC CHARACTERISTICS OF INDEFINITE PRONOUNS IN CROATIAN...

438 VESTNIK ZA TUJE JEZIKE/JOURNAL FOR FOREIGN LANGUAGES
Table 2. Proportion of IP among words that match in meaning and context  
(expression of a negative attitude) according to levels of language competence
Level of  
language competence
B2.1 B2.2 C1
Proportion of  
indefinite pronouns
16.29% 16.67% 17.65%
When comparing the results achieved by the respondents who are native speakers 
of one of the Slavic languages with the results of the respondents who are speakers of 
non-Slavic languages (Table 3), it can be seen that the speakers of Slavic languages 
wrote an average of 6.96 words, of which 1.22 (17.53%) were IP, while non-Slavic na-
tive-speakers wrote an average of 6.56 words, of which 1.08 (16.46%) were IP.
Table 3. Proportion of IP among words that match in meaning and context  
(expression of a negative attitude) according to type of first language
Type of first language Production of indefinite pronouns
Slavic first languages 17.53%
Non-Slavic first languages 16.46%
6 DISCUSSION
The first hypothesis, that CL2 students at the B2.1, B2.2 and C1 levels have acquired a 
command of the meaning of IP, is confirmed. This was to be expected given the fact that 
the structure of the CL is already completely acquired by the B2 level (Udier 2015: 249), 
but this hypothesis had to be confirmed by the research because a command of the mean-
ing is a prerequisite for acquiring a command of the pragmatic features.
The second hypothesis, that CL2 students at the B2.1, B2.2 and C1 levels have a 
weaker command of the pragmatic property of IP (the possibility of expressing a negative 
attitude) in comparison with their meaning, is confirmed only in part, i.e. it is confirmed 
only in the IP koješta and kojekakav, while it is not confirmed for the pronouns svakakav, 
nikakav, sve and svašta. This difference in acquisition can be partly explained by the fre-
quency of pronouns (the pronoun koješta in the hrWaC corpus has 4,789 tokens, and the 
pronoun kojekakav has 18,467, while the pronoun svakakav has 18,731 tokens, nikakav 
331,572 tokens, sve 786,886 tokens and svašta 55,313 tokens
4
), especially for the pro-
noun koješta, which is significantly less represented in the hrWaC corpus than other IP 
4 All data about the number of tokens is taken from the hrWaC corpus at the following link: https://www.clarin.si/
noske/all.cgi/first_form?corpname=hrwac;align= (accessed December 2, 2020)

439
(with only 4,789 tokens). Greater acquisition can be attributed to some extent to a higher 
frequency in the language which, since it is a sign of greater repetition and use, leads 
to more effective cognitive implantation (Langacker 1988, 2000). However, frequency 
alone cannot explain the fact that the respondents acquired a command of the pragmatic 
function of the pronouns kojekakav (18,467 tokens, acquisition of 30%) and svakakav 
(18,731 tokens, acquisition of 63%), which are very close, in terms of representation, 
in the corpus. The reasons for the uneven acquisition of pronouns of similar frequency 
should be sought elsewhere, most likely in the way students were taught the pronouns, 
which cannot be investigated in this study.
That the increase in the language competence of CL2 students also increases their 
command of the pragmatic properties of IP (the ability to express a negative attitude), 
which was claimed in the third hypothesis, was also confirmed. Although explicit in-
struction of the pragmatic properties of IP is not contained within the CL2 textbooks that 
have been analysed (Čilaš Mikulić, Gulešić Machata and Udier 2015, Udier and Gulešić 
Machata 2017, Korom 2005), CL2 speakers at high levels of language proficiency still 
acquire it. The percentages that show the representation of IP among the words that they 
correspond to by context are not large (ranging from 16.29% to 17.65%), and the in-
crease in the use of IP in the appropriate context is proportional to the level of language 
competence (B2.1 16.29%, B2.2 16.67% and C1 17.65%), although it is not large (the 
difference between the level of B2.1 and C1 is only 1.36%). This fact leads to the con-
clusion that there is still room for improvement in teaching, and it can be assumed that 
systematic and explicit teaching of the pragmatic properties of IP would result in greater 
and pragmatically more appropriate usage. This hypothesis should be tested by further 
research preceding the development of a model for teaching IP that would include their 
pragmatic properties.
The hypothesis that CL2 students at the B2.1, B2.2 and C1 levels whose first lan-
guage is one of the Slavic languages better master the pragmatic dimension of IP than 
students whose first language is not Slavic was also confirmed, but with very small 
difference in values – the representation of IP in Slavic speakers was 17.53%, and in 
non-Slavic ones 16.46%, so the difference is only 1.07%. The results show that CL2 
students who speak related Slavic languages generally have a slightly richer vocabulary 
with a slightly higher share of IP compared to CL2 students of non-Slavic origin, on 
the basis of which it can be assumed that there was a positive transfer from the other 
Slavic languages into Croatian. However, caution should be exercised when relying on 
positive transmission from related languages, because analyses have shown that related 
languages can have significantly different IP systems (Haspelmath 1997: 293), and that 
seemingly equivalent IP in different languages, even related ones, cannot be used in all 
the same contexts. It would be useful to conduct a comparative analysis of the meaning 
and pragmatic potential of IP in different Slavic languages, as well as other languages, 
in order to illustrate and explain similarities and differences in relation to Croatian, and 
Sanda Lucija Udier: PRAGMATIC CHARACTERISTICS OF INDEFINITE PRONOUNS IN CROATIAN...

440 VESTNIK ZA TUJE JEZIKE/JOURNAL FOR FOREIGN LANGUAGES
to determine in which contexts and meanings seemingly equivalent IP are used or not 
(Haspelmath 1997: 293).
From the results of this research, specifically from the small difference in the pro-
duction from Slavic and non-Slavic respondents (only 1.07%), it follows that the differ-
ence between Slavic speakers and non-Slavic speakers learning CL2 is not so large as to 
require a different approach when teaching. The research also confirmed the fifth hypoth-
esis that CL2 students at the B2.1, B2.2 and C1 levels are less proficient in the pragmatic 
dimension of IP in production than they are in comprehension.
Some of the results of this research were expected – respondents at higher levels 
of language competence in CL2 listed more words which match in meaning in compar-
ison with students from lower levels of language competence. The amount of IP with a 
certain pragmatic function used appropriately increases with the amount of words, and 
CL2 students whose first language is one of the Slavic languages have a slightly richer 
vocabulary and use slightly more IP.
However, the second part of the results of this research was less expected and 
shows that the proportion of IP in all words used by respondents is relatively small 
(between 16.67% and 17.65%), which leads to the conclusion that respondents use 
relatively few IP that match the meaning in a particular context, that is, that their rec -
ognition of a pragmatic function predominates over its production in specific commu-
nication settings. Although it is universally known that among learners of a second 
language comprehension is dominant over production (they are much more able to un-
derstand when they listen or read than they are able to say or write), still, the proportion 
of IP in production is very minor and points to the fact that with awareness of them and 
better teaching things could be improved, which would allow CL2 students to better 
realise the specific pragmatic function of IP that was discussed in this research (that of 
expressing a negative attitude).
The research also opened many other questions, for example the question of how 
much of the complex system of IP is actually used in everyday communication in the 
CL, since CL2 teaching also has the purpose of teaching only that which is actually 
used. Frequency analysis would show us which IP have a higher frequency in the CL, 
and these pronouns should then be taught chronologically earlier and quantitatively to a 
greater degree. But out of everything, the most important factor regarding the teaching of 
CL2 in lessons would be an exhaustive linguistic description of the pragmatic function 
of those IP that are actively and often used in the CL in order to base their explicit teach-
ing on that description. It would also be necessary to explore the meanings and feelings 
expressed by IP in idiomatic expressions such as: nitko i ništa (‘nobody and nothing, a 
person with no value’), sve i svašta (‘anything and everything’), svakakav (‘all sorts, a 
euphemism for something bad’), onakav (‘not much of a person/thing’), etc., because 
their understanding is necessary for quality communication in the CL at higher levels of 
language competence.

441
7 CONCLUSION
The results of the research show that the pragmatic level needs to be included more 
frequently and more thoroughly in the teaching of CL2 because its mastery is an es -
sential component of language competence, especially at higher levels. The results of 
the research indicate the conclusion that – in addition to everything else already taught 
– the pragmatic properties of IP should also be taught, and additionally idiomatic ex-
pressions with IP should be learned lexically (for example ima nas svakakvih ‘there are 
all kinds of us’, on je nitko i ništa ‘he is nobody and nothing’, svakakva osoba ‘shady 
person’). 
The teaching of IP in CL2 should be based on the results of corpus research on rep-
resentation and frequency, on research on the acquisition of the system of IP and, first 
and foremost, on a scientific description of the meaning of IP that includes the pragmatic 
level. It is necessary to not only teach from an explicitly grammatical approach, but also 
to teach lexically in such a way that includes the pragmatic level, because IP (as well as 
other types of pronouns) have great pragmatic potential.
 In the future, it will be necessary to make a more complete linguistic description 
of the pragmatics of IP that could then be applied in CL2 teaching, as well as to make a 
better general description of the pragmatic level of the CL, which is still relatively unde-
scribed, and whose description will be a great challenge for experts given the fact that this 
level cannot be described by formal linguistic instruments.
REFERENCES
AGUILAR-GUEV ARA, Ana/Maria ALONI/Angelika PORT/Katrin SCHULZ/Šimik 
RADEK (2010) Free choice items as fossils. Workshop on Indefiniteness Cross-
linguistically (DGfS) Berlin, February 25−26, part one and part two. https://www.
semanticscholar.org/paper/Free-choice-items-as-fossils-Guevara-Aloni/e8c16bfdb-
621ff49a03ffb0fe4af03320fc80527#paper-header (Accessed March 2, 2021)
AGUILAR-GUEV ARA, Ana/Maria ALONI/Angelika PORT/Šimik RADEK  (2011) Se-
mantics and Pragmatics of Indefinites: Methodology for a Synchronic and Diachron-
ic Corpus Study. Proceedings of the Workshop Beyond Semantics: corpus-based 
Investigations of Pragmatica and Discourse Phenomena, 1−16. Goettingen, Germa-
ny, 23−25 February 2011, Bochumer Lingvistische Areitsberichte (3) https://www.
academia.edu/9963031/2011._Semantics_and_pragmatics_of_indefinites_meth-
odology_for_a_synchronic_and_diachronic_corpus_study._In_Proceedings_of_
the_workshop_Beyond_Semantics_--_Corpus-based_investigations_of_pragmat-
ic_and_discourse_phenomena._Bochum_Bochumer_Linguistische_Arbeiten (Ac-
cessed March 2, 2021)
Sanda Lucija Udier: PRAGMATIC CHARACTERISTICS OF INDEFINITE PRONOUNS IN CROATIAN...

442 VESTNIK ZA TUJE JEZIKE/JOURNAL FOR FOREIGN LANGUAGES
BARIĆ, Eugenija et al. ²(1997) Hrvatska gramatika. Zagreb: Školska knjiga.
BERGOVEC, Marina (2007) Leksički pristup nastavi stranih jezika s posebnim osvrtom 
na hrvatski. Lahor: časopis za hrvatski kao materinski, drugi i strani, vol 1, br. 3, 
53–66.
BRALA VUKANOVIĆ, Marija/Mihaela MATEŠIĆ (2015) Croatian ‘pointing words’: 
Where body, cognition, language, context and  culture meet. Branimir Belaj (ed.) 
Dimenzije značenja, Zbornik Zagrebačke slavističke škole. 31−62.
ČILAŠ MIKULIĆ, Marica/Milvia GULEŠIĆ MACHATA/Sanda Lucija UDIER (2015) 
Razgovarajte s nama! Udžbenik hrvatskoga kao drugoga i stranoga jezika za razinu 
B2. Zagreb: FF press.
HASPEMATH, Martin (1997) Indefinite pronouns. Oxford: Oxford University Press.
KECSKÉS, Istvan (2014) Intercultural Pragmatics. Oxford: Oxford University Press. 
KECSKÉS, Istvan (2015) Intracultural Communication and Intercultural Communica-
tion: Are They Different? International review of pragmatics 7 (2015), 171−194.
KOROM, Marija (2005) Kroatisch für die Mittlestufe: Lese- und Übungstexte. München: 
Otto Sagner Verlag. 
KUNA, Branko (2007) Identifikacija eufemizama i njihova tvorba u hrvatskom jeziku. 
Fluminensia 19, 1, 95−113.
LANGACKER, Ronald W. (1988) A usage-based model. Brygida Rudzka-Ostyn 
(ed.) Topics in Cognitive Linguistics. Amsterdam, Philadelphia: John Benjamins, 
127−164. 
LANGACKER, Ronald W. (2000) A Dynamic Usage-Based Model. Michael Barlow, 
Suzanne Kemmer (ed.) Usage-Based Models of Language,  Stanford: CSLI Publi-
cations, 1−64.
LJUBEŠIĆ, Nikola/Tomaž ERJA VEC (2011) ‘hrWaC and slWaC: Compiling Web Cor-
pora for Croatian and Slovene’. In TSD’11: Proceedings of the 14th international 
conference on Text, Speech and Dialogue, Ivan Habernal and Václav Matoušek (ed.) 
395–402. Berlin, Heidelberg: Springer.
LUČIĆ, Radovan (2005) Kaj god! Jagoda Granić (ed.) Semantika prirodnog jezika i me-
tajezik semantike, zbornik Hrvatskoga društva za primijenjenu lingvistiku, Zagreb i 
Split: HDPL, 437−445.
NAZALEVIĆ ČUČEVIĆ, Iva (2015) Pragmalingvistična kompetencija na razini B2. 
Milvia Gulešić Machata and Ana Grgić (ed.) Hrvatski B2, Opisni okvir referentne 
razine B2 u hrvatskom jeziku, Zagreb: FF press, 41–54.
RAGUŽ, Dragutin (1997) Praktična hrvatska gramatika. Zagreb: Medicinska knjiga.
SILIĆ, Josip/Ivo PRANJKOVIĆ (2005) Gramatika hrvatskoga jezika za gimnazije i vi-
soka učilišta. Zagreb: Školska knjiga.
UDIER, Sanda Lucija/Milvia GULEŠIĆ MACHATA (2011a) Kakvom jeziku poučavati 
neizvorne govornike? Viši i niži varijetet u poučavanju hrvatskoga kao inog jezika. 
Nova Croatica, V, 5, 329−349.

443
UDIER, Sanda Lucija/Milvia GULEŠIĆ MACHATA (2011b) Registri i hrvatski kao ini 
jezik. Pavel Krejčí, Elena Krejčová and Michal Przybylski (ed.) Výuka jihoslovan-
ských jazyků v dnešní Evropě, Porta Balkanica, 64−71.
UDIER, Sanda Lucija (2015) Gramatička kopetencija. Milvia Gulešić Machata and Ana 
Grgić (ed.) (2015) Hrvatski B2: Opisni okvir referentne razine B2. 247–272.
UDIER, Sanda Lucija/Milvia GULEŠIĆ MACHATA (2017) Razgovarajte s nama! Gra-
matika i pravopis hrvatskoga jezika s vježbama za razine B2-C1, drugo izdanje. 
Zagreb: FF press. 
VIJEĆE EUROPE (2005) Zajednički europski referentni okvir za jezike. Zagreb: Školska 
knjiga i Vijeće Europe.
POVZETEK 
PRAGMATIČNE ZNAČILNOSTI NEDOLOČNIH ZAIMKOV V HRV AŠČINI KOT DRU-
GEM JEZIKU
Prispevek obravnava nekatere pragmatične značilnosti hrvaških nedoločnih zaimkov, zlasti mo-
žnost njihove rabe za izražanje negativnega odnosa oziroma stališča. V prispevku predstavljamo 
rezultate kvantitativne raziskave, katere namen je bil ugotoviti, v kolikšni meri učenci hrvaščine 
kot drugega jezika obvladajo rabo nedoločnih zaimkov za izražanje negativnega odnosa. V vpra-
šalniku, ki je bil sestavljen iz dveh vsebinskih sklopov, smo preučevali raven razumevanja in 
dejanskega izkoriščanja pragmatičnih danosti hrvaških nedoločnih zaimkov. V vprašalnik smo 
vključili primere, ki smo jih pridobili iz korpusa hrvaških spletnih besedil hrWaC 2.0 (Ljubešić in 
Erjavec 2011). Raziskavo smo izvedli med učenci hrvaščine kot drugega jezika, ki so se v študij-
skem letu 2019/20 udeležili jezikovnih tečajev Centra za hrvaščino kot drugi in tuji jezik Croati-
cum na Univerzi v Zagrebu. Raven znanja hrvaščine anketirancev je ustrezala stopnjama B2 in C1 
Skupnega evropskega referenčnega okvira (N=60). Rezultate naše raziskave bo mogoče s pridom 
uporabiti pri razvijanju učnih gradiv, namenjenih obravnavi semantičnih in pragmatičnih lastnosti 
nedoločnih zaimkov pri pouku hrvaščine kot drugega jezika. Raziskava je pokazala, da se je pri po-
uku hrvaščine kot drugega jezika v prihodnje treba bolj posvetiti obravnavi pragmatičnih vidikov 
rabe nedoločnih zaimkov, saj je ravno obvladovanje pragmatične ravni jezika zlasti pri govorcih 
na višji jezikovni stopnji eden ključnih pokazateljev jezikovnih kompetenc. Poleg pragmatičnih 
lastnosti nedoločnih zaimkov bi bilo treba pri pouku obravnavati tudi idiomatične izraze, v katerih 
se pojavljajo nedoločni zaimki. Zaradi številnih pragmatičnih danosti teh zaimkov je nujno, da jih 
pri pouku ne obravnavamo le s strogo slovničnega, temveč tudi z leksikalnega vidika, ki omogoča 
ustrezno obravnavo pragmatične ravni jezika.
Ključne besede: nedoločni zaimki, hrvaščina kot drugi jezik, pragmatične strukture, izražanje 
negativnega odnosa 
Sanda Lucija Udier: PRAGMATIC CHARACTERISTICS OF INDEFINITE PRONOUNS IN CROATIAN...

444 VESTNIK ZA TUJE JEZIKE/JOURNAL FOR FOREIGN LANGUAGES
ABSTRACT
PRAGMATIC CHARACTERISTICS OF INDEFINITE PRONOUNS IN CROATIAN AS 
A SECOND LANGUAGE
This paper deals with some of the pragmatic characteristics of indefinite pronouns, specifically 
about the potential for expressing a negative attitude or stance by using them. The results of a 
quantitative survey will be presented exploring to what extent the potential to express a negative 
attitude using indefinite pronouns has been mastered. The research questionnaire consisted of two 
parts and examined the respondents’ command in terms of both comprehension and production of 
the pragmatic properties of indefinite pronouns in Croatian as a second language (CL2), and the 
examples used in the questionnaire were taken from the hrWaC 2.0 corpus (Ljubešić and Erjavec 
2011). The research was conducted among students of CL2 at the levels B2 and C1 (N=60) who at-
tended the programs of Croaticum – Centre for Croatian as Foreign and Second Language, Univer-
sity of Zagreb in the 2019/2020 academic year. The results obtained in this research can be applied 
in the development of learning materials intended for teaching the semantic and pragmatic features 
of indefinite pronouns in the instruction of CL2. They show that the pragmatic level needs to be 
included more thoroughly in the teaching of CL2 because its mastery is an essential component of 
language competence, especially at higher levels. The pragmatic properties of IP should also be 
taught, as should idiomatic expressions with IP. It is necessary to not only teach from an explicitly 
grammatical approach, but also to teach lexically in such a way that includes the pragmatic level, 
because IP have great pragmatic potential.
Keywords: indefinite pronouns, Croatian as a second language, pragmatic structures, expressing 
a negative attitude