Reflexive pronouns are intrinsically related to reflexive verbs. When the subject of an action is also the object of that action, it is said that the action is reflected back onto the subject, thus making the subject the bearer, i.e. the object, of that action. This reflection is expressed thriugh the use of the reflexive pronoun sebe (oneself) or its short, anclitic form se. True reflexiveness is expressed with the use of the enclitic form se in the accusative case, while the long form sebe is used as an emphatic.
Sebe uvek moraš da poštuješ – You must always respect yourself
Smiri se – (You) calm (yourself) down.
Another reflexive pronoun with an emphatic function is the pronoun sam (oneself) which is used with the long or the short form of sebe:
Obećao je samom sebi da neće piti – He promised to himself that he won’t drink.
The reflexive pronoun sebe has no person, gender or number marker, while the reflexive pronoun sam has gender and number. Reflecting the traits of the subject, reflexive pronouns can be used in the following manner:
- As the direct object – in the accusative, both forms are used without a preposition. The reflexive pronoun has to be tracked back to the subject, which in this instance is also object, reflected by the pronoun:
Majka se vratila – Mother has returned (herself)
Čovek treba samog sebe da voli – One needs to love himself
- The reflexive pronoun sebe can also be used in a context of reciprocity when the subject and object have a reciprocal relationship with each other, expresssed through the verb (the English equivalent of ‘each other’ or ‘one another’). In this case the two can be expressed as the subject while the reflexive pronoun se denotes the relationship of reciprocity and reflexivity:
(Džon voli Anku) – (John loves Anka)
Džon i Anka se vole – John and Anka love each other.
Oni se vole – They love each other
- As the indirect object – excluding the accusative case, in the long form:
Čovek treba da se sobom ponosi – One should be proud of oneself
- Following prepositions – all cases, in the long form:
On daje sve od sebe – He is giving all of himself
Izađi na kraj sa sobom – Sort yourself out
Ja imam dovoljno za sebe – I have enough for myself
The reflexive pronoun sabe (self) is not morphologically sensitive to the grammatical person, number or gender of its referent. Browne suggests that despite lack of morphology, these features are present since they occur on the (emphatic) modifier sam. It does not occur in Nominative or Vocative case, but appears in Accusative (sebe), Genitive (sebe), Locative (sebi), Dative (sebi), and Instrumental (sobom) forms. The citation form is Accusative. (The asterisk denotes impossibility of coreference.)
Milan je video sebe u ogledalu
Milan-NOM be-3s saw self-ACC in mirror-LOC
Milan saw himself in the mirror
The clitic form (se) alternates with full form, sebe, in most reflexive constructions:
Milan se (je) video u ogledalu
Milan-NOM self-ACC be-3s saw in mirror-LOC
Milan saw himself in the mirror
The clitic form is not inflected for the person, number, or gender of its referent. It occurs in the Accusative and Dative as se. The clitic form is ungrammatical following prepositions and thus does not occur in oblique cases:
(Ona) govori stalno o sebi/*se
(She-NOM) talk-3s always about self-LOC
She always talks about herself
Particularly when used with inherent reflexive verbs, use of the full form is often contrastive or emphatic:
- Jovan se pere
John-NOM self wash-3s
John is washing himself
- Jovan pere sebe
John-NOM wash-3s self-ACC
John is washing himself (emphatic)
Although sebe is limited to fully reflexive contexts, se also occurs in certain passive and intransitive constructions.
Reflexive Possessive Pronoun
The reflexive possessive pronoun svoj (self’s, one’s own) has no direct counterpart in Modern English, but occurs in Modern Scandinavian languages and Slavic languages. Although Serbian svoj does not agree morphologically with its referent, it is fully inflected as a modifier for gender, number, person, and case.
- Slavko govori o svom konju
Slavko-NOM talks about self’s-masc-s-LOC horse-masc-s-LOC
Slavko talks about his own horse
- Janko daje Marku svoju kniigu
Janko-NOM gives Mark-DAT self’s-masc-s-ACC book-masc-s-ACC
Jankol is giving Mark his own book
Nominative use of the possessive reflexive is limited to idiomatic expressions:
On je svoj čovek
He-NOM be-3s self’s-NOM man-NOM
He is his own man
The reflexive possessive functions syntactically as Specifier of NP; for example,
Ivan je govorio o svojem životu
Ivan-NOM be-3s talked about self’s-LOC life-LOC
Ivan was talking about his own life
Antecedents for the Reflexive Pronoun
Although subject antecedents are preferred, Serbian permits subject and object antecedents for the reflexive pronoun sebe in monoclausal sentences. Even under pragmatic pressure favoring a subject antecedent objects are permitted. However, native speakers exhibit a strong preference for subject antecedents, and object antecedents are only marginally acceptable in some constructions. For example, the object antecedent for the genitive reflexive sebe in sentences such as is marginal to unacceptable. Lexical effects are likely to condition the acceptability of object antecedents in Serbian.
- Policajac je ispitivao osumnjičenog o sebi
policeman-N be-3s questioned suspect-A about self-L
The policeman questioned the suspect about himself
- Doktor je pitao pacijenta o sebi
doctor-N be-3s questioned patient-A about self-L
The doctor questioned the patient about himself
- Pacijent je pitao doktora o sebi
patient-N be-3s questioned doctor-A about self-L The patient, questioned the doctor j about himself’/j
- Ivan je poslao Petru odjeéu za sebe/njega
Ivan-N be-3s posted Peter-D clothes-A for self/him-G
Ivan sent Peter clothes for himself//him
In ditransitive sentences, with the reflexive sebe embedded in a prepositional phrase (PP) within a noun phrase, antecedent selection follows the pattern shown. Speakers who accept both local and non-local antecedents indicate a preference for clausal, rather than NP, subjects. Speakers also prefer to disambiguate sentences using pronouns when possible.
- Vera je dala Nini Kristininu knjigu o sebi/njoj
Vera-NOM be-3s gave Nina-DAT [Kristina-GEN book about self/her-LOC]
Vera gave Nina Kristina’s book about herself/her
Antecedents for the Possessive Reflexive
Eligible Antecedents for the reflexive possessive svoj are restricted to clausal subjects in simplex sentences.
- a. Vlado je dao Ivanu svoj/njegov šešir
Vlado-NOM be-3s gave Ivan-DAT self’s/his hat-ACC
Vlado, gave Ivan his own/his hat
- Janko daje Marku svoju knjigu
Janko-NOM gives Mark-DAT self’s book-ACC
Janko is giving Mark his own book
The same coreference pattern occurs when the possessive reflexive modifies an NP complement of a locative PP. As shown in example 13, the reflexive svojoj may only refer to the clausal subject, Ivan. In complementary distribution, the possessive pronoun njen (ACC) shows gender agreement with its NP object antecendent, Nina, as well at agrrment with the NP kući.
- Ivan je poljubio Ninu u svojoj/njenoj kući
Ivan – NOM be – 3s kissed Nina – ACC at self’s/her house – LOC
Ivan kissed Nina at his/her own/her house
Reflexivity in english
Reflexive pronouns are often used when the action described by the verb is directed toward the thing referred to by the subject of the verb. This use of reflexive pronouns is illustrated in the following examples. The reflexive pronouns are underlined.
I washed myself thoroughly before putting on clean clothes
Did you hurt yourself?
Reflexive pronouns can also be used when it is desired to emphasize a personal pronoun. The reflexive pronouns in the following examples are underlined.
I myself saw what happened.
Did he solve the problem himself?
She did the work herself.
In these examples, the reflexive pronouns myself, himself and herself are used to emphasize the personal pronouns I, he and she.
The reflexive personal pronouns are listed below.
|Subjective Case||Reflexive Pronoun|
It can be seen that in the second person, a differentiation is made between yourself, which agrees with singular antecedents, and yourselves, which agrees with plural antecedents.
It should be noted that the first and second person reflexive pronouns are formed from the corresponding possessive adjectives, whereas the third person reflexive pronouns are formed from the corresponding pronouns in the objective case. This is illustrated in the following table.
|Objective Case||Possessive Adjective||Reflexive Pronoun|
Reflexive pronouns are used in three instances in English.
- With Reflexive Verbs
I enjoyed myself last summer.
He’s trying to market himself as a consultant.
Sharon pays herself $5,000 a month.
We encourage ourselves to learn something new every week.
- As an Object of a Preposition Referring to Subject
Tom bought a motorcycle for himself.
They purchased a round trip ticket to New York for themselves.
We made everything in this room by ourselves.
Jackie took a weekend holiday to be by herself.
- To Emphasize Something
No, I want to finish it myself! (I don’t want anyone helping me.)
She insists on talking to the doctor herself. (She didn’t want anyone else talking to the doctor.)
Frank tends to eat everything himself. (He doesn’t let the other dogs get any food.)
- Problem Areas
Many languages besides Serbian, such as Italian, French, Spanish, German, and Russian often use verb forms which employ reflexive pronouns. Here are some examples:
alzarsi – Italian / get up
cambiarsi – Italian / change clothes
sich anziehen – German / get dressed
sich erholen – German / get better
se baigner – French / to bathe, swim
se doucher – French / to shower
In English, reflexive verbs are much less common. Sometimes students make the mistake of translating directly from their native language and adding a reflexive pronoun when not necessary.
I get myself up, shower myself and have breakfast before I leave for work. SHOULD BE I get up, shower and have breakfast before I leave for work.
She becomes herself angry when she doesn’t get her way. SHOULD BE She becomes angry when she doesn’t get her way.
Reflexivity in English, on the other hand, is canonically represented by pure reflexives, verbs followed by the reflexive pronoun, which are, however, to be found relatively rarely in Modern English. Reflexiva tantum are now to be found mostly in literary discourse. These are verbs such as bethink, comport, perjure, pique, bemean, bestir, betake, etc. It is important to notice, though, that they are all semantically intransitive.
There is also a very strong tendency to omit the reflexive pronoun or to replace it with the personal pronoun or other non-standard forms, which is more commonly found in informal styles of communication. For example:
I’ve bought me a new car.
Had a pint after work to cheer self up.
 HAMMOND, L, Serbian: An essential Grammar, London: Routledge, 2005.
 BROWNE, W. Serbo-Croatian. In B. Comrie and G. Corbett, eds. The slavonic languages. London: Routledge, 1993.
 HAWKESVORTH, C. Colloguial Serbo-Croatian. London: Routledge & Kegan Paul, 1986, p. 92.
 Bidwell, C. “The reflexive construction in’i,’SerboCroatian”. Studies in Linguistics 18, 37-47, 1965.
 Browne, 1993.
 MIHALJEVIĆ, M. 1990. “Upotreba povratnoposvojne zamjenice svo; u hrvatskom ili srpskom jeziku”. In G. Holzer, ed. Croatica. Slavica. Indoeuropaea: Wiener slavistiches jahrbuch. Ergàn zungsband VIII (pp.145-156), 1990.
 HAWKESVORTH, C. Colloguial Serbo-Croatian. London: Routledge & Kegan Paul, 1986, p. 144.
 MIHALJEVIĆ, p. 145.
The overall pattern of results indicates that reflexive binding in a second language is constrained by Universal Grammar. In addition, the results show that the anaphor type found in the native language may crucially affect the way in which reflexives are interpreted in the target language. A significant number of L2 learners in this study showed evidence of transfer of the LI X° anaphor type to their interlanguage grammar. The effect of the consequent misclassification of English XP reflexives as XC anaphors is shown in high levels of acceptance of long-distance antecedents for reflexives in sentences that lack an X-bar compatible subject (i.e., AGR) in the local domain. Finally, it appears that subjects do not simply surface transfer lexical elements or whole constructions, but are able to access a UG-constrained deductive system to establish binding domains for the target language.
Results of the two sentence comprehension tasks lead to the following set of conclusions and implications:
(1) Response patterns on the two tasks suggest these L2 learners have a +AGR parameter setting but that the morphological complexity of the English reflexive has not been recognized by a substantial number of these L2 learners.
This conclusion is supported by results on the object position CPNP Type 1 sentences and Type 3 tensed biclausal sentences. LD responses on Type 1 sentences primarily provide information about the anaphor type assumed by these L2 learners, while local responses to Type 3 sentencesestablish the +AGR parameter setting operating in the grammars of these learners.
CPNP sentences occur in both Serbian and English. However, the governing category for reflexives in CPNP constructions in these languages differs. For the X° reflexive in Serbian, the [NP,NP] Specifier subject of the complex NP is not an eligible subject, since only clausal AGR, another X° element, can set the local binding domain. Thus, CPNP constructions provide syntactic environments in which the effect of anaphor type is apparent in languages that have morphologically overt AGR.
The pattern of LD binding by L2 learners on object position CPNP sentences on both tasks suggests that the X°/ +AGR binding configuration is present in the interlanguage grammar of a significant number of these learners. This result supports Hypotheses A and B.
Results on tensed biclausal Type 3 sentences across task type strongly indicate that these L2 learners have grammars with overt morphological AGR. Although the source of the AGR parameter setting (i.e., UG or transfer) cannot be unambiguously established when the L1 and L2 grammars have identical values, it is clear that these L2 learners have not initially adopted a Chinese-type -AGR parameter setting. This finding is also supported by results on biclausal Type lA (CPNP) sentences on the MCC task.
A somewhat different picture is presented by the results on Type lC sentences. Low profiency learners produce a significantly higher number of LD responses on sentences with reflexives occurring in subject internal position. LD binding outside tensed clauses is only permitted in
languages with a X° reflexives -AGR binding configuration. In light of the robust result on Type 3 sentences the presence of -AGR in the grammars of these learners seems improbable.
It is likely that other, perhaps pragmatic, factors may account for LD binding across the indicative clausal barrier in Type lC sentences by low proficiency learners. There is some indication from Ll acquisition research that sentences of this type are difficult to process.
Overall, L2 learners who permitted LD binding in their interpretations of English reflexives restricted antecedents to the minimal finite clause. This is consistent with Serbian, Russian, and Scandinavian languages which have X° reflexives and morphologically overt AGR.
(2) The high level of local binding of reflexives achieved by most of the L2 learners indicates morphological complexity of anaphoric elements is a learnable feature of language. Over 50% of the L2 learners in this study show evidence of consistent local reflexive binding in the individual subject analysis.
(3) When a binding domain resulting from the interaction of an XO anaphor type and an AGR parameter setting is not instantiated in the native language, L2 learners must use their grammar to set the domain when relevant constructions are encountered in the target language. Computation of binding domains by L2 learners requires operation of a UG-constrained deductive system.
This claim is supported by the results on Type 2B object control infinitival sentences. Object control structures provide the crucial test structure for isolating aspects of the interlanguage grammar that may be attributed to surface transfer (i.e., X° anaphor type, +AGR parameter setting) from those that may reflect access to Universal Grammar.
Evidence of LD binding by L2 learners lacking L1 knowledge of the domain restrictions on anaphor-antecedent coreference in these constructions suggests that L2 learners do not simply transfer whole constructions, but that they resort to a UG-constrained deductive system to define binding domains in the target language.
LD binding of reflexives in object control sentences by low profiency L2 learners suggests that the predicted pattern of binding found in languages like Russian, which has the +AGR/ X° reflexive configuration, may occur in the L2 despite lack of L1 instantiation. Since object control sentences provide the only null subject environment for testing LD binding across infinitival [NP,IP] subjects, this is an important result. It suggests that computation of binding domains is not limited to L1 acquisition and that the interaction of the transferred XO reflexive and the +AGR setting results in the predicted domain extension. Results on Type 2B sentences support Hypothesis C.
The contrast between results on ECK (Type 2A) and object control (Type 2B) infinitival sentences raises a number of methodological and theoretical questions. In contrast to the signficant level of LD binding on object control sentences by low proficiency and adolescent L2 learners, interpretations of reflexives in ECK infinitivals show no evidence of LD binding. This result on ECK sentences on the PI task fails to support Hypothesis C. Differences in response behavior on object control and ECM infinitival sentences suggest that reflexive binding may be sensitive to the type of subject present in the embedded clause. In English infinitival sentences, the subject in control sentences is null (PRO), while in ECM sentences, an overt lexical subject is present. Since ECM structures are not present in Serbian, the binding domain for reflexives in English ECM structures must be established in L2 acquisition. For learners with an X°/+AGR configuration in their interlanguage grammar, it appears that the presence of a lexical NP in embedded subject position blocks coreference outside the infinitival clause. The split between ECM and object control sentences may also reflect task differences since these sentence types occur on different tasks. Although selection of a local antecedent is a legitimate option in languages with X° reflexives, the rejection of LD binding on ECM sentences in the PI task needs to be further investigated.
 BENNETT, S. “Second language acquisition of reflexive binding by native speakers of Serbo-Croatian”, McGill University, Montreal, 1993.
- BENNETT, S. Second language acquisition of reflexive binding by native speakers of Serbo-Croatian, McGill University, Montreal, 1993.
- BIDWELL, C. “The reflexive construction in SerboCroatian”. Studies in Linguistics 18, 37-47, 1965.
- BROWNE, W. Serbo-Croatian. In B. Comrie and G. Corbett, eds. The slavonic languages. London: Routledge, 1993.
- CATFORD, J. C. “Contrastive analysis and language teaching”. In James E Alatis (ed) Contrastive linguistics and its pedagogical implications. Report of the 19th Annual Round Table Meeting on linguistics and language studies. Washington DC: Georgetown University Press, 159-173. 1968.
- CHESTERMAN, A. Contrastive functional analysis. Amsterdam: John Benjamins. 1998.