SMALL CLAUSES IN ENGLISH AND BOSNIAN, CROATIAN, AND SERBIAN

Small Clauses in English represent one of the basic concepts of modern generative grammar and they are characterized as minimal units of non-verbal predication, where the predication relation between the subject and the predicate is established in the absence of a finite verb form. In traditional grammars, they are generally referred to as “secondary predication” constructions. As the concept of a “small clause” is not familiar in the context of traditional descriptive grammars of Bosnian, Croatian, and Serbian, the aim of this paper is to make a preliminary contrastive analysis of how these or similar constructions are realized in Bosnian, Croatian, and Serbian through translation equivalents and how they are treated in traditional grammars. Additionally, we have applied several tests typically used in English literature to prove the constituency of small clauses in order to establish whether or not these constructions could form a separate constituent at the sentence level in Bosnian, Croatian, and Serbian as well.


Introduction
Small Clauses (henceforth referred to as SCs) 1 appear in the constructions of the type [NP 1 V NP 2 XP] 2 , generally referred to as "secondary predicative constructions", where the string [NP 2 XP] is taken to be a separate constituent in the form of a clause. Thus, the string [NP2 XP] represents a small clause and X can be realized as a noun (N), an adjective (A), a preposition (P), or a non-finite verb form (V), which we illustrate with the examples in (1a-d).
(1) a. We consider [ SC John a genius]. ( Semantically, SCs are assumed to be minimal units of non-verbal predication, corresponding to a proposition, where the predication relation between the SC subject NP 2 and the SC predicate XP is established in the absence of a finite verb form. The SC predicate XP can be categorially realized as NP, AP, PP or VP (with a non-finite verb: infinitive, -ing form or past participle) and according to the categorial type of the predicate XP there are four different types of SCs -nominal, 1 In this paper, we will use the terms "clause" and "sentence", hence we have to explain what we actually mean by them. The term "sentence" usually escapes precise definition, nevertheless it is commonly understood that the sentence is the most complex language unit to which the syntactic rules apply, it is used independently and it is bounded by the initial capital letter at the beginning and the punctuation mark of full stop at the end. We regard the "clause" as an element that is hierarchically subordinated to the sentence, but superior to the syntagm (Crystal 1987: 95). A sentence can consist of multiple clauses with a verb in the finite or non-finite form. In this paper, the clause bears the epithet "small" because it is either verbless or it contains a non-finite verb. The clausal status is assigned on the basis of "tacit", or unexplained predication, which extends the traditional understanding of predication usually based on the verb. 2 We use the following abbreviated English forms: NP -Noun Phrase; V -Verb; XP stands for the predicative phrase, where X can be realized as N -Noun, A -Adjective, P -Preposition, or V -Verb. adjectival, prepositional and verbal. 3 Another important feature of this construction is its ability to express at least three different semantic meanings, as illustrated by the examples in (2): ( In this paper, we will use the term qualifying small clause for the construction in the example (2a), the depictive small clause for the construction in the example (2b), and the resultative small clause for the construction in the example (2c).
Small Clauses emerged as one of the basic concepts of modern generative grammar, relying heavily on some of the most fundamental principles of Government and Binding (GB) Theory, such as the Theta Criterion, the Projection Principle and the Binary-branching requirement (Chomsky 1981;1986). Many questions about them have been addressed by numerous modern linguistics scholars, but without achieving any consensus in finding the appropriate and final solutions. Some of the most controversial issues about SCs are related to their constituency at the sentence level, the empirical range of the possible constructions to be analyzed as SCs, as well as their internal structure and categorial status.
that the term "small clause" is not used in contemporary BCS grammars, we have attempted to offer a preliminary contrastive analysis of small clauses in English and BCS through translation equivalents and based on the most common and simple types of small clauses in the English language. This paper also offers a brief insight into the literature related to the BCS linguistic context in order to examine whether and how the same or similar constructions to small clauses in English are treated by native linguists in the Bosnian, Croatian, and Serbian language. Furthermore, we have applied several standard English constituency tests to determine whether there is a possibility that the constructions referred to as small clauses in English could also be realized as separate sentence constituents in BCS.
The main goal of this brief contrastive analysis of small clauses in English and BCS is to identify some possible similarities and differences between the two language systems when it comes to the use of such and similar constructions in both languages, and also to provide some new insights and different perspectives to the description of the linguistic phenomena in both English and BCS.
There are four major lines of analysis which have attempted to clarify the problems posed by different semantic relationships established between the constituents within the syntactic structure of the type [NP 1 V NP 2 XP]. Traditionally, descriptive grammars analyze the pattern [NP 1 V NP 2 XP] as a "complex-transitive" complementation structure, where NP 2 is the direct object and XP is taken to be a "predicative complement". From the generative perspective, there are three conceptually different theoretical approaches to these constructions -The Small Clause Theory, The Predication Theory and The Complex-Predicate Analysis. The predicative relationship between NP 2 and XP is the basis of the Small Clause Theory analysis, which regards NP 2 as the subject and XP as the predicate of a Small Clause which is selected by the matrix verb (Stowell 1983). The Predication Theory treats NP 2 as a direct object of the main verb, and the XP as a secondary predicate that is licensed by the argument complex of the primary predicate (Bresnan et al. 1978;Schein 1995;Williams 1983). The Complex-Predicate Analysis treats NP 2 and XP as two separate constituents and assumes that the predicative relationship obtaining between these two phrases is attributed to the fact that the main verb and XP form a complex predicate, which selects NP 2 as its direct-object argument (Chomsky 1957;Rapoport 1993;Contreras 1995).
In the framework of GB theory, many different proposals have been suggested related to the categorial status and internal structure of SCs, ranging from those which assume that SCs are lexical projections (LP) of their predicate, via those which include functional projections in their structure (IP, AgrP, PrP, TopP etc.), and finally ending up with those which totally assimilate the internal structure of small and full clauses (CP) 5 .

A brief overview of the literature in BCS
In BCS descriptive grammars, the term "small clause" is not mentioned anywhere, but the same or similar constructions are referred to by different names and treated as such in a different way. The most commonly used terms are: secondary predication, predicate extension, nominal or adjectival predicate, attributive predicate, etc. These specific terms indicate that constructions such as small clauses in English are mainly discussed in the chapters dealing with predicates and establishing a predicative relationship within a sentence in BCS descriptive grammars.
When we refer to the Bosnian language, the authors Jahić, Halilović, and Palić (2000: 367-368) use the term semi-copulative predicate, which consists of a semi-copulative verb in a personal form and with a lexical nucleus. Semi-copulative verbs, unlike copulative verbs, have a lexical meaning, but it is not complete. Therefore, they must have a complement in the form of a lexical nucleus, which will characterize the fundamental lexical meaning of the predicate, i.e. what is attributed. The lexical core is composed of substantive or adjectival words. Semi-copulative verbs can be intransitive (e.g. BCS: postati, ostati, ispasti, izgledati, praviti se, činiti se, zvati se, prozivati se, etc.) and transitive (e.g. BCS zvati, nazvati, imenovati, smatrati, držati, proglasiti, etc.). When the intransitive semi-copulative verbs are used, the lexical content of the predicate is attributed to the subject, but when the transitive ones are used, it is attributed to the object. With the intransitive semi-copulative verbs, the lexical core appears in the form of the nominative or instrumental case, e.g.
Oni su već postali nestrpljivi; I ocu uvijek ja ispadnem kriv; Osjeća se izigranim, prevarenim, zaboravljenim; Ona mnogo toga zna, ali se čini nevještom; etc. (Jahić et al. 2000: 367). On the other hand, with the transitive semi-copulative verbs, the lexical core is in the instrumental case, e.g. Zvali su one što imaju pametnim, a one što nemaju -budalastim; Neki su je držali ljepoticom; etc. (Jahić et al. 2000: 368). Besides, these authors also mention that, with some of the transitive semi-copulative verbs, the lexical core can be in the If we make a comparison with the English language and what is meant by a "small clause" in that language system, it would be possible to relate such constructions in the Bosnian language to English small clauses with verbs like consider, find, regard (as), view, deem, counts, appointments, etc. The main difference is that Jahić et al. (2000) treat the postverbal NP, e.g. ga in the sentence Smatrali su ga za ludaka as the direct object of the main verb, 6 and the sequence za ludaka as a predicative complement, whereas according to the small clause model, the entire sequence ga za ludaka should be observed as a separate constituent in the form of a small clause which represents the complement of the main verb smatrati.
In the section on sentence condensation, Jahić et al. (2000: 450-451) distinguish a special type of sentence condensation by means of the so-called predicative appositive. It occurs in clauses with copulative nominal predicates whose lexical core is filled by an adjectival word. The complex sentence in which such condensation takes place must be either "mono-subject", i.e. both clauses being part of it have the same subject, or the subject of the condensed clause must be the same as the principal clause object. In the process of condensation, the copulative verb is dropped, and the clause is reduced to the lexical nucleus of the nominal predicate, i.e. an adjectival word, which may also be accompanied by the corresponding dependent member of the adjectival phrase, e.g. Zato što je bila sposobna, lako je našla posao > Sposobna, lako je našla posao; Zato što je bio napušten od svih, bio je tužan > Napušten od svih, bio je tužan (Jahić et al. 2000: 450).
Jahić et al. claim that this type of condensation with the predicative appositive mostly occurs in (i) independent clauses with accompanying situational meaning, e.g. Išla je uza zid, zamišljena i spora, po 6 The main issue here is that if "ga" is taken as a direct object in this sentence, then we would also have to be able to make sentences like e.g. *Smatrali su ga, or *Smatrali su njega, or *Smatrali su Mustafu, which is obviously not the case, because these sentences are ungrammatical. mention that the predicative appositive in the grammatical structure of the sentence expresses a double syntactic dependency, i.e. on the one hand, it is dependent on the substantive word which performs the function of the subject or the object it agrees with, and on the other hand, it is dependent on the predicate, while referring to it as an adverbial expression, since it functions as a condensing clause with the "adverbial" meaning. Thus, the predicative appositive is a unit with dual function, as an attribute and an adverbial. Jahić et al. (2000: 451) assume that the adverbial function of the predicative appositive, i.e. its connection with the predicate, is more important syntactically and semantically. 7 These constructions could also be related to English small clauses in the form of descriptive, subject-oriented and object-oriented constructions that function as adjuncts. Thus, for example goli in the sentence Trčali su goli, and dobro opremljen in the sentence Dobro opremljen, mogao bi stići daleko, could be taken as descriptive constructions with a PRO subject, which imply the control structure in generative grammar, and according to the fact that they belong to the type of subject-oriented structures. On the other hand, for example uhvaćen in the sentence Uhvaćen, odmah je predat policiji, would also be a depictive small clause with PRO subject, but this time as the object-oriented control structure. 7 The predicative appositive is mainly separated with a comma or different intonation from the rest of the sentence in order to emphasize its adverbial function and its difference in relation to the congruent attribute, e.g. Vruće, gvožđe se lako kuje (= Kad / zato što je vruće, gvožđe se lako kuje) vs. Vruće gvožđe se lako kuje (= Gvožđe koje je vruće lako se kuje) (Jahić et al. 2000: 451).

SMALL CLAUSES IN ENGLISH AND BOSNIAN, CROATIAN, SERBIAN
In Croatian descriptive grammars we also find a different approach in treating the same type of constructions that are considered to be small clauses in English. Silić & Pranjković (2005: 290-291) apply the same analysis model as in Jahić et al. (2000), with the same terminology, so they also offer a chapter on the semi-copulative predicate, that is composed of a semi-copulative verb and a complementary nominal element, e.g. Postali su nerazumni; Nazivao se predsjednikom (with a nominal complement in the nominative or instrumental case with intransitive semi-copulative verbs), and e.g. Smatraju ga varalicom; Njega su prikazali poštenim (with a nominal complement in the instrumental case with transitive semi-copulative verbs). Silić & Pranjković (2005: 292) consider this nominal complement as a nominal predicate extension, which usually appears as an adjective, in the indefinite form, and which can agree either with the subject, e.g. Ana je ležala smirena, or with the object, e.g. Našli smo ga iscrpljena. These authors provide an additional term, the so-called predicative attribute, which is introduced by the predicate and becomes its part, while agreeing with the subject or the object. Silić & Pranjković (2005: 292) also claim that the nominal predicate extension can appear in the form of a noun, so they mention several possibilities for its realization: (i) with the subject agreement, e.g. Dugo se mučila jadnica, or the object agreement, e.g. Našli su ga bijednika u drvenjari; (ii) in the instrumental case (the marked usage in the standard Croatian language), e.g. Djevojkom je bila šutljiva; (iii) as the genitive case without a preposition and with an obligatory attribute, e.g. Ležala je na plaži otvorenih usta; (iv) as an expression which includes preposition and case, e.g. Našli su ga u ranama; (v) in the construction with kao, e.g. Vratio se iz rata kao bogataš. Barić et al. (2005) also use the term predicate extension, as an obligatory complement of verbs with incomplete meaning, and which commonly occurs as a nominal word in the nominative, instrumental, or accusative case, and as an expression with a preposition, e.g. There is a special chapter on the nominal extension, where they take into consideration the nominal predicate extensions which usually appear when an adjective or a noun is used as a predicate and the predicate of one sentence is incorporated along with the verbal predicate into the structure of another sentence, which is called predicate extension (2005: 570). If the subject is the same in both resulting sentences, then the adjective performing the role of a predicate extension remains in the nominative case and is marked by grammatical indefiniteness, e.g. the sentence Ja sam ležao miran originates from two initial sentences, i.e. Ja sam ležao and Ja sam miran (2005: 570-571). Barić et al. claim that adjectives in these sentences are not attributes, because they do not appear next to nouns, but next to verbal predicates as the complements to their meaning. If the subject of the initial sentence with the adjectival word in the predicate is different from the subject of the sentence with the verb predicate, but the same as some other noun or pronoun in it, then the predicative adjective agrees with that other noun or pronoun, e.g. the sentence Sutradan nađoše djevojku mrtvu u šljiviku originates from two initial sentences Sutradan nađoše djevojku u šljiviku and Djevojka je mrtva (2005: 571). Barić et al. also provide the examples of the predicate extension in the form of a noun in the nominative case if the subjects in both initial sentences are the same, which is usually considered as a nonstandard and poetic usage, e.g. Prognanik u pučkoj rulji sam se skrio, from the initial sentences U pučkoj rulji sam se skrio and Bio sam kao prognanik (2005: 252). When the subjects in the initial sentences differ, the predicative noun that is introduced as the predicate extension agrees with the noun or pronoun it refers to, e.g. Našao sam ga siromaha originating from the initial sentences Našao sam ga and On je siromah (2005: 572). Barić et al. also mention the example where the predicate extension is used independently, as a "special element in the sentence sequence", e.g. Stadoh, začuđen;

SMALL CLAUSES IN ENGLISH AND BOSNIAN, CROATIAN, SERBIAN
Ustao Mika još uvijek svečan, važan i ozbiljan; etc. (2005: 573). Again, the above mentioned constructions considered by Barić et al. (2005) could be related to the English descriptive constructions that imply the subject or object control structures.
Among the descriptive grammars for the Croatian language, the most detailed description of constructions with the so-called secondary predication including many examples from literary works is provided by Katičić (2002). Like Barić et al. (2005), this author also uses the terms predicate extension and predicative extending (2002: 481-489). In addition to grammars, there are special studies on secondary predicative constructions in the Croatian language. In his monograph The Predicate Extension, Peti (1979) deals with this topic in detail, and in particular with the syntactic position of what he assumes as the "predicate extension", which are syntactic categories that include what is termed "secondary predication" in contemporary linguistic literature, but also those constructions that are considered to be mandatory predicative complements. Šarić (2008) also deals with secondary predication in the Croatian language, with special reference to descriptive constructions, their semantic and grammatical characteristics, their relation to adverbials, and their semantic typology. Marković (2009) offers a study on the resultative secondary predicate in Croatian. Although these constructions are rarely encountered in standard Croatian, Marković states that they are not impossible, and examines the reasons why the resultative meaning of the Croatian secondary predicate is rarely realized.
When it comes to describing secondary predicative constructions in the traditional descriptive grammars for the Serbian language, there are still no small clauses mentioned. Piper et al. (2005: 314) offer a chapter on copulative and non-copulative predicates, where the secondary predicative constructions, which are actually related to English small clauses, are considered as non-copulative predicates that may occur with transitive or intransitive verbs, e.g. Petar je postao direktor and Ministarstvo je postavilo Petra za direktora. What is new in this grammar is that these authors use modern linguistic terms when referring to different types of these predicates, thus distinguishing: (i) subject-oriented predicates, and (ii) object-oriented predicates. The difference is established on the basis of whether the content of the predicate relates only to the subject of the sentence, or it partly relates to the subject and partly to the object (with transitive verbs), e.g. On se proglasio za vladara and Narod je proglasio stranca za vladara. Piper et al. (2005: 334) also distinguish the attributive predicative in a non-copulative predicate, which they call the predicative attribute, e.g. for vesela in the sentence Ana ulazi vesela. At the same time, they treat the non-copulative predicate used with the predicative attribute as a complex predicate.
In the literature referring to the Serbian language system that was available to us for this paper, we only found the term "small clause" used by Ivić (2005: 8), who defines them as "jezičke datosti koje inače, s obzirom na formu u kojoj se ostvaruju, nemaju apsolutno nikakvih uslova za to da budu svrstane u kategoriju rečeničnih struktura, ali koje, ukoliko se uklope u sastav neke proste rečenice, tu prostu rečenicu, po iskaznom informacijskom sadržaju, preobraćaju u složenu". Ivić relates them to similar syntactic constructions in Serbian, which are formally significantly different from small clauses, but functionally the same, i.e. they have the same informative effect -it relates to the specific extension of the informative potential of a simple sentence, that this author also calls the semantic "loadedness". The author considers the use of different constructions in Serbian that are supposed to provide the same informative effect as small clauses do in English, e.g. the declarative sentence Vidjeli smo ga can be extended additionally with živog i zdravog.
Ivić claims that such a complement contains the same information which could otherwise be expressed with the sentence construction e.g. Vidjeli smo ga: on je živ i zdrav ( In addition, Ivić (2005: 9) shows two specific uses of the adjective lexeme for the purpose of enriching the informative content of a sentence, wherein the informative content introduced by such an adjective may be identified with a corresponding (dependent) temporal sentence, e.g. On je umro mlad (= On je umro dok je još bio mlad) or Ona se sasvim mlada udala (= Ona se udala dok je još bila sasvim mlada); or with a different conjunction as in e.g. Kupus se jede ukiseljen (= Kupus se jede onda kad je ukiseljen) or Kafa se pije vruća (= Kafa se pije onda kad je vruća). In English, these constructions are considered to be depictive small clauses.

Contrastive analysis of small clauses in English and BCS based on translation equivalents
Assuming the fact that a comprehensive and thorough contrastive analysis of small clauses in English and BCS is far beyond the scope of this paper, we have decided to do at least a small part of it, i.e. by means of a contrastive analysis of some of the typical English constructions that are treated as small clauses and often cited as such in the literature, through their translation equivalents in BCS. Generally, the aim has been to of examine how these English constructions are realized in BCS, whether with the same or at least similar constructions such as small clauses in English, or in some completely different ways. In the analysis, we used examples for qualifying, resultative and depictive small clauses, including examples of nominal, adjectival, prepositional, and verbal small clauses.
The example (1) illustrates how the qualifying nominal small clause [ SC him a fool] in the English sentence They considered him a fool is realized through translation equivalents in BCS. In the BCS translation equivalent there is a clitical form of the pronoun in the accusative case ga, whereas a fool corresponds to the noun in the instrumental case. Additionally, in BCS translation variants it is possible to use the prepositional constructions with za or kao, in which case the noun is realized in the accusative case.
(1) E: They consider him a fool.

BCS: Smatraju ga budalom / Smatraju ga za budalu / Smatraju ga kao budalu.
In the example (2), there is another qualifying small clause in English, this time an adjectival one, [ SC him intelligent] in the sentence They consider him intelligent, which is realized in BCS with a translation equivalent that includes a pronoun in the accusative case ga and the corresponding adjective in the instrumental case. The prepositional constructions with za or kao and the accusative case become acceptable only when we use a noun in the accusative case along with the adjective, e.g. čovjeka.

BCS: *Vidio sam ga ulaziti u biblioteku / Vidio sam ga da/kako ulazi u biblioteku / Vidio sam ga da je ušao u biblioteku.
In the example (5), we have a resultative small clause in English [ SC his son an honest man] in the sentence He raised his son an honest man. In the BCS translation equivalent we have the noun in the accusative case sina along with the given adjective, or the noun in the instrumental case poštenim (poštenjakom). Again, there is a possibility to use the alternative construction in which the preposition za or kao is followed by a suitable adjective or noun in the accusative case, e.g. za poštena čovjeka or kao poštenog čovjeka.
(5) E: He raised his son an honest man.

BCS: Odgojio je sina poštenim (poštenjakom) / Odgojio je sina za poštena čovjeka / Odgojio je sina kao poštenog čovjeka.
The example (6) illustrates a typical resultative small clause in English [ SC the metal flat] in the sentence They hammered the metal flat. In the BCS translation equivalent we have the instrumental ravnim again, but alternatively, we could also use the adverb ravno.
Besides, it is also possible to use a finite clause introduced with tako da, which actually clarifies the resultative character of this construction.
(6) E: They hammered the metal flat.

BCS: Iskucali su (čekićem) metal ravnim (~ravno) / Čekićem su iskucali metal tako da postane/ bude/ je ravan.
The example (7) also introduces a resultative small clause in English [ SC the table clean] in the sentence The waitress wiped the table clean, but which cannot be realized as a construction with the instrumental case in the BCS translation equivalent. Instead, it is possible to use only a finite clause with tako da, which explains the resultativeness expressed by the given construction in English.
(7) E: The waitress wiped the table clean.

BCS: *Konobarica je obrisala stol čistim / Konobarica je obrisala stol tako da postane/ bude/ je čist.
In (8) we have another example of a resultative small clause in English [ SC the pavement thin] in the sentence They ran the pavement thin, which could also not be realized as a construction with the instrumental case in the BCS translation, but rather as finite clauses that clearly explain the resultative character of the whole construction.

BCS: *(Is)trčali su pločnik tankim / Trčali su i pločnik se stanjio / Trčeći su stanjili pločnik.
The example (9a) shows an object-oriented depictive small clause in English [ SC him drunk] in They found him drunk, which can be realized in the BCS translation equivalent as the pronoun in the accusative case + the adjective in the accussative, but not as the adjective in the instrumental case *pijanim. The example (9b), with the same construction, illustrates a subject-oriented small clause in English which contains the subject element PRO that is not phonetically realized [ SC PRO drunk], and which is coindexed with the subject of the main clause within the sentence structure I i found

SMALL CLAUSES IN ENGLISH AND BOSNIAN, CROATIAN, SERBIAN
him [ SC PRO i drunk]. Thus, the past participle form drunk does not refer to the object him, but to the subject I. The comma that is used as a punctuation sign emphasizes the fact that we deal with an adjunct structure here. In the BCS translation equivalent we have the adjective in the nominative case, that agrees with the subject, so it becomes clear what it actually refers to in the whole sentence structure. Additionally, it is also possible to use the adjective at the beginning of the sentence.
The example (10) illustrates another object-oriented depictive small clause in English [ SC his coffee cold] in the sentence He drank his coffee cold. In the BCS translation equivalent, we can have the adjective used attributively instead of the resultative construction, e.g. hladnu kafu, although the resultative construction kafu hladnu is also acceptable. Due to the fact that the word order is not so strict in BCS as it is in English, in BCS it is also possible to use the construction in which the adjective is moved at the beginning of the sentence for putting more emphasis to it, as in Hladnu je popio kafu.

BCS: Popio je hladnu kafu / Popio je kafu hladnu (dok je bila hladna) / Hladnu je popio kafu.
This brief contrastive analysis through translation equivalents has shown that constructions representing small clauses in English are often realized by different type of constructions in BCS, although in some cases we have also observed some coincidence. In BCS, the subject NP from the English small clause is most often realized as a noun in the accusative case, while the predicative phrase XP is often realized in the instrumental or accusative case, with a preposition za or kao. Most discrepancies were observed when it came to English prepositional and verbal small clauses, which are most often realized in BCS translation equivalents as finite clauses. Furthermore, even the resultative structures in English cannot be adequately realized in BCS translation equivalents, so once more, finite clauses appear to explain the resultative meaning of the given constructions.

Constituency tests applied on BSC examples
For the purpose of contrastive analysis of English and BCS with respect to constructions of the type [NP1 V NP2 XP], in which the English string [NP2 XP] is treated as a small clause, and for the purpose of determining the possible clausal status of the same string in BCS, we have applied some of the tests commonly used to determine constituency of the sentence elements. Although these tests are applied specifically in English, and thus some of them are not appropriate for the BCS language system, we thought that at least some of these tests could still be applied to BCS examples, and that in this way we could come up with some new insights and interesting data.
There are 5 commonly used constituency tests that are usually cited in the English literature on small clauses: (1) coordination test; (2) independent use of the string [NP 2 XP]; (3) predicative string [NP 2 XP] as a subject; (4) adverbial modification of the predicative string [NP 2 XP] and (5) derived nominals. In the following part of this section, we will examine the possibilities for applying each of these tests, using the BCS examples, to determine whether we can use the string [NP2 XP] as a separate constituent in BCS as well.
When it comes to the coordination test, the examples in BCS show that it is possible to coordinate these structures in BCS. For illustration, we used one qualifying (1), one depictive (2), and one resultative construction (3): ( In order to understand this BCS example in the right way, i.e. so that it represents a resultative construction with the actual meaning of Obojili su zid tako da je postao crven, it is necessary to emphasize in pronunciation the adjective crvenim at the end of the sentence. In that way, we would differentiate this construction from another one with a different meaning, e.g. Obojio je zid crvenom (bojom). The examples in (7) illustrate that the sentence in (7a) is more acceptable both in BCS and E, i.e. if we use a qualifying construction with kao (E: as) in the subject position of the copulative construction, because the absence of kao in BCS would definitely result with an ungrammatical sentence. 9 In (8) we provide a similar example to the one that is widely used in literature in order to prove the constituency of small clauses in English (Safir 1983: 732), where such an exemplifying sentence is acceptable in English, but not in BCS. 10 9 A similar phenomenon was observed in English (Aarts 1992: 44-45). 10 The original sentence used as an example in Safir (1983: 732) is: Workers angry about the pay is just the sort of situation that the ad campaign was designed to avoid. After having compared the examples in BCS and the English ones provided by Kayne (1984: 151-152), and assuming the fact that it is possible to have derived nominals with transitive verbs, but not intransitive ones as e.g. smatrati, which cannot provide a suitable deverbative noun *smatranje, we have decided to use another verb instead, e.g. the verb tretirati in (12), which could have a derived nominal tretiranje. Here we also include the examples with a depictive (13) and a resultative construction (14). Since we have transitive verbs here, it is possible to form derived nominals. Although Kayne (1984) based this test with derived nominals exclusively on English and some specific rules, we can see in the example (12)  In that view, we quote a conclusion provided by Aarts (1997: 271), referring to the constituency of these constructions: "The conclusion we can draw from our deliberations must again be a provisional one: while not all the tests point to the bracketing in (101) [here in the examples (1a), (2a) and (3a), the author's remark], the semantic and syntactic evidence taken together does seem to do so. Further research is needed to explain why some of the tests fail."

Conclusion
In this paper we briefly discuss the possibility of introducing the concept of "small clauses" in BCS, when it comes to analyzing the same or similar constructions that are considered as small clauses in English, but that are traditionally analyzed in a different way in BCS traditional grammars. A contrastive analysis of some typical forms of small clauses in the English language and corresponding constructions in BCS through translation equivalents has shown that in some cases a coincidence is possible, while in others it is not. The possibility of matching was most evident in qualitative and depictive constructions, and the least or almost not evident at all in the case of resultative constructions, as well as in prepositional and verbal small clauses, in which the finite clauses are most commonly used in BCS translation equivalents. The subject NP in all English small clauses is most commonly realized as a noun or pronoun in the accusative case, while the predicative NP or AP in English nominal or adjectival small clauses is most commonly realized as a noun or adjective in the instrumental case, or with a preposition za and in the accusative case, or in the construction with kao.
Generally, the structures that are considered to be small clauses in English are treated in a completely different way in the BCS traditional descriptive grammars. General terms used for these structures are secondary predication or secondary predicative constructions. The subject NP in English small clauses is traditionally treated as a direct object in referential BCS grammars, whereas the predicative XP is most commonly referred to as the so-called predicate extension, nominal or adjectival predicate, attributive predicate etc.
After having applied five standard constituency tests in BCS, which are otherwise commonly used in English to prove that the string [NP2 XP] in the structure [NP1 V NP2 XP] makes a small clause, we have found out that there is a possibility for the predicative string [NP2 XP] to be analyzed as a separate constituent in the form of a small clause in BCS as well. However, this possibility needs to be further explored, with the use of additional tests and the inclusion of a far larger number of examples. We consider that this smallscale contrastive analysis of small clauses in English and BCS could contribute to the cross-linguistic study of these structures, and open some new opportunities for further research.
Given that the linguistic phenomenon of small clauses has only recently attracted the attention of linguists, it has not been systematically described yet in most recent and referent descriptive grammars, but the specific aspects of their syntactic and semantic properties are mainly discussed in scientific papers and monographs. Therefore, a modest contribution of this paper would be to consolidate the current knowledge and insights from different sources, and to systematize and classify them. Although small clauses may still be one of the controversial issues in modern linguistics, their existence is evident, and they play a significant role when it comes to a syntactic description of not only the English language system but also of other languages based on the increasing number of cross-linguistic research.