Skip to content Skip to navigation

Personal Deixis: The Early Stages


1. Acquiring self- and other-reference: pronouns vs. verb inflexions


The linguistic means for expressing the pragmatic roles of speaker, addressee and non-participant in a communicative situation are a part of a complex deictic system – that of personal deixis. The personal and possessive pronouns represent the core of this system. In pro-drop languages, however, verb inflexions for person are often the only markers of personal deixis. This structural difference between the linguistic systems is also reflected in the acquisition of these systems. Thus, many children acquiring pro-drop languages start marking the notions of speaker, addressee and non-partocipant with verb inflexions (cf. Clark 1986 for Spanish and Italian; Smoczyńska 1992 for Polish; Georgov 19051 and Stojanova-Trajkova for Bulgarian). This possibility is not discussed for children acquiring non-pro-drop languages. However, examples illustrating self- and other-reference with inflected verbs without (own) name are to be found at least in languages such as German2 and Russian, as in these languages, unlike English, the category of person is regularly expressed in some of the verb paradigms, although partially or completely neutralized in others3.

1.1. Order of acquisition: personal and possessive pronouns

Most of the publications devoted to the acquisition of personal deixis try to shed light on the highly disputable question regarding the order of acquisition of the notions of speaker, addressee and other, and of the different linguistic means for marking these notions.

Numerous studies state that children acquiring English start with first person pronouns I, my, mine and the 3rd person inanimate it, followed by the 2nd person you (Brown 1973; Huxley 1970; Clark 1978; Chiat 1986, etc.), while the order of appearance of the remaining pronominal forms is not well established. Chiat (1986) emphasizes the existence of significant individual variations in the appearance of the 2nd person you and the 3rd person animate forms s/he: while one of the children in her naturalistic sample introduced these pronouns almost simultaneously, others showed a gap between these pronouns lasting up to 6 months.

For German, the precedence of 1st over 2nd person pronominal forms is confirmed in naturalistic, as well as in experimental studies (cf. Stern & Stern 1928; Deutsch & Pechmann 1978; Mills 1986).

The same order of 1st and 2nd person pronouns is reported for French (Clark 1986), where the 3rd person singular masculine il appears more or less simultaneously with the 2nd person tu; next to emerge are the 2nd and 3rd person plural. A somewhat delayed appearance of the 1st person plural nous is explained with the (specific for French) replacement of this pronoun with on ‘one’.

1.2. Is the pre-pronominal stage a pre-personal one as well?

Publications concerning acquisition of pronouns are usually not interested in the earliest phase of self- and other-reference which can be characterized as pre-pronominal. Numerous authors mention, however, that before mastering pronominal forms, many children use names for referring to self and the addressee. In languages such as English and German, children’s sentences in this period lack any person marker, while in pro-drop languages, children already use verbs marked for 3rd person indicative. Clark (1986) emphasizes the fact that “the order of [of acquisition] of pronouns in French is similar to the order of emergence for verb forms” in Spanish and Italian, which, being pro-drop languages, mark person “directly in the verb”. If the verb inflexions for person are concerned, the author suggests, Italian and Spanish children start with 2nd person singular imperatives, followed by the 3rd person indicative; later on emerges the 1st person indicative and, in several months, the 1st person plural (Clark 1986).

Thus, the earliest linguistic forms for self-reference in Spanish and Italian are the 3rd person verb forms, with or without (own) name.

Being pro-drop languages, Polish and Bulgarian show developmental patterns closely related to those reported about Spanish and Italian. According to Smoczyńska’s (1992) detailed study on the acquisition of personal deixis, Polish children between 1;6 and 1;8 years of age start referring to self and the addressee with 3rd person verb forms. A similar stage of early development is reported also for Bulgarian (cf. Georgov 1905; Stojanova-Trajkova1986).

As the above mentioned data show, children acquiring pro-drop languages pass through a pre-pronominal stage during which 3rd person verb forms are the only markers of personal deixis. Sporadically used 2nd person imperative forms, reported by some authors (Smoczyńska 1992 for Polish; Clark 1986 for Spanish and Italian; Mills 1986 for German, Georgov 1905 for Bulgarian, etc.), seem to be rather formulaic expressions. This means that, during a couple of months, the 3rd person verb forms are not contrasted to any other grammatical means for marking "person". Therefore, children in this stage do not seem to have in their disposal much knowledge about the formal distinction between speaker, addressee, and other. Isn't it then reasonable to claim that the pre-pronominal stage is a pre-personal one as well? The answer is positive if only the personal deixis is taken into account (cf. Stojanova-Trajkova 1986). Some languages, however, such as Bulgarian and Polish, possess an additional grammatical marker - a vocative inflexion in the noun – for differentiating the pragmatic role of addressee from the roles of speaker and non-participant. Data about Bulgarian prove children's ability to mark addressee with vocative nouns a couple of months before the contrast between 3rd person and 1st (2nd) person forms emerge, since vocative endings are among the earliest grammatical morphemes used by children (cf. Stojanova 2005). So, in the pre-pronominal stage, the 3rd person verb forms used undifferentiated for referring to speaker, addressee, and non-participant, can be combined with a noun in vocative to construct the initial opposition between speaker/non-participant, on one hand, and addressee, on the other.

The pre-pronominal stage is typical of early speakers (Stojanova-Trajkova 1986; Smoczyńska 1992) and is one of the characteristic features of the referential style of acquisition (cf. Bates, Dale and Thal 1995).

2. The Bulgarian data: “Pragmatic-dominant” vs. “formal-dominant” strategy

The Bulgarian sample includes 5 longitudinally studied subjects as well as 2 children who were tape-recorded in single 60-min. sessions (cf. Table 1).

Table 1: Data about the Bulgarian children

Child’s name

and sex

Source of material

Child’s age



Georgov’s sample

Stojanova’s sample


Vlado – male



0;7 – 3;0

Ženja – male



0;6 – 2;9

L. –female


Diary & tape-rec

0;8 – 2;6

S. – male



1;7 – 2;9

O. - female



1;10 – 2;9

K. – male




D. – female





Unlike the Polish subjects, reported by Smoczyňska, Bulgarian children are divided into two groups according to their strategies for acquiring self- and other-reference. The two groups differ significantly in their ways of acquisition of the linguistic means for expressing personal deixis: they follow two opposite strategies, called in Stojanova-Trajkova (1986) with the terms “formal-dominant” vs. “pragmatic-dominant”. Although the strategies correlate with the individual styles of acquisition: referential (nominal, analytic) vs. expressive (pronominal, holistic), not all tendencies characteristic of the styles relate to the strategies as well.

The term "formal-dominant" emphasizes children's greater fluency and their concern with the formal rather than pragmatic aspect of their speech: They use less amount of rote units and acquire grammatical contrasts relatively early, including the formal contrasts for marking person within the verbal and pronominal paradigms. These latter contrasts, however, are first related to wrong pragmatic meanings: 3rd person is ovegeneralized to refer to speaker and addressee, 2nd person is used instead of 1st person for self-reference. Five of the Bulgarian children listed in Table 1 followed the "formal-dominant" strategy in the acquisition of the deictic category of person.

The term "pragmatic-dominant" describes an opposite strategy of acquisition. "Pragmatic-dominant" children rely much more on rote forms in their early developmental stages. These children, however, use verbs and pronominal forms for marking person in their correct pragmatic meaning from the very beginning. Overgeneralizations of 3rd and 2nd person do not appear in their speech, or occur only incidentally. Two of the Bulgarian subjects, Ženja, and O., acquired self- and other reference according to the "pragmatic-dominant" strategy.

Since the most important difference between the two groups concerns the acquisition of self-reference, the tables and figures bellow aim at giving detailed illustration of the development of that deictic category.


2.1. A "pragmatic-dominant" child: Marking speech roles is the most important task in speaking

Table 2 and Figure 1 show O.'s data as representative for the "pragmatic-dominant" strategy4. Table 2 displays a clear tendency: The child starts referring to self exclusively with 1st person forms. At the age of 2;3 these forms are mostly 1st person singular verbs; in the next two months the percentage of 1st person pronominal forms grows, and at 3; 5 O. refers to self with 1st person verb and pronoun forms in almost 100 percent of the cases. Self-reference with 3rd person verbs is

Table 2: Development of linguistic means (scored in percentages of tokens) used by O. for self-reference.


3rd V

3rd V & Name


1st V

1st Pr

1st Pr &1st V

1st Pr &3rd V

1st Pl

1st Poss



































Figure 1: Development of self-reference in O.'s speech: percentages of 3rd person verb forms and/or name scored against 1st person verbs and/or pronominal forms.

incidental. Nevertheless, O. referred to self with her own name or with 3rd person verb forms in 28.57% of the cases when she was 2;3 years old, and in 11.54% of the cases when she was 2;4. Most of these utterances in O.'s earliest sample, however, were pronounces either in a situation where she was describing a picture of herself taken several months ago, or in a context where she was answering adult's prompting questions in 3rd person forms which are typical of Baby talk. Example 2 shows that even in such contexts O. tended to prefer the adult-like self-reference with 1st person forms.

(2) Mo: Oni kakvo pravi?

Oni what do-3rd-P-SG-PRES

(What is Oni doing?)

V: Pipam [r]adioto.

Touch-1st-P-SG-PRES the radio.

(I) am touching the radio.

In (2), O.'s mother addressed her with a 3rd person verb form plus child's name "Oni": Oni kakvo pravi? (What is Oni doing?), as it is usual within the Baby talk register. O., instead, not accepting the "Baby talk proposal", referred to herself with a 1st person verb: Pipam [r]adioto, [I] am touching the radio.

It is worth noting that O's correct self- and other-reference contrasted with her underdeveloped morpho-syntactic competence. At the age of 2;5, when she reached an almost adult-like level regarding the pragmatic means for expressing person, she had not still mastered some important components of the simple syntax: grammatical morphemes, such as prepositions, prefixes, particles, etc., were sometimes omitted; morphological paradigms were still not complete. She used only few types of complex sentences and her utterances left the impression of a fragmentary and inconsistent grammatical competence.

2.2. A "formal-dominant" child: Marking speech roles can wait until more important grammatical markers have been mastered

Although S.'s linguistic development is not as precocious as that of other children following the "formal-dominant" strategy, it is taken as representative of that strategy. Table 3 and Figure 2 show the development of S.’s linguistic means for self-reference.

He starts referring to self and the addressee before the age of two with 3rd person forms, but single examples of 1st person verbs and pronouns, as well as of pronominal reversal are registered, too. Since S.'s verbal production until 2;0 years of age is insufficient for scoring, his earliest data are not included in the Table below.

Table 3: Linguistic means used by S. for self-reference (percentages of tokens)


3rd V

3rd V & Name


1st V

1st Pr

1st Pr &1st V

1st Pr &3rd V

1st Pl

1st Poss















































*2/1 = 2nd person forms used instead of 1rs person ones.

Figure 2: Development of self-reference in S.'s speech: percentage of 3rd person verb forms and/or own name scored against other means (1st person verbs and/or pronominal forms, cases of pronominal reversal, and agrammatical combinations of 1st person pronouns with 3rd person verbs).

As Table 3 shows, S.'s means for self-reference at the age of 2;0 are quite variable. Third person verbs plus child's own name predominate, but 1st person verbs with or without 1st person pronominal forms are also frequent. Examples of pronominal reversal are registered, too, but they are marginal in S.'s speech. What is typical of his development is the occurrence of 3rd person verb forms in combination with 1st person pronouns. Within the following three months, the child overcomes this violation of the rule of morpho-syntactic congruence. In the Bulgarian sample, S. is the only child5 who uses a considerable amount of that type of incongruent constructions (cf. example 3).

(3) E: A na plaža kakvo praveše?

And what were you doing on the beach?

S: (I)g(r)ai-3/1 az-1/1 tam pačka [= pjasǎka]

Play-3rd-P-SG-Pres I there on the beach.

(I am playing on the beach).

E: Igraeše na pjasǎka?

You were playing on the sand?

Mo: Kakvo goniš tam?

What are you running after?

S: (I)g(r)ai-3/1 az-1/1.

Play-3rd-P-SG-Pres I.

(I am playing).

E: Igraeš na pjasǎka?

You are playing on the sand?

S: Pei-3/1 az-1/1.

Sing-3rd-P-SG-PRES I.

(I am singing)

E: Da peeš?

(You want) to sing?

S: N'ama pei. N'ama pei-3/1 az-1/1!

Not sing-3rd-P-SG-PRES. Not sing-3rd-P-SG-PRES I.

(Won't sing. Won't sing I).

As Table 3 shows, examples of 1st person plural verbs in S.'s speech appear simultaneously with 1st person singular forms. Initially, the 1st person plurals are more frequent – 12,5% at the age of 2;0, but 2,5% three months later. This fact could be possibly explained with the overgeneralizations of the 1st person inflexion –m neutralizing the difference between singular and plural forms. After overcoming the overgeneralized forms, the child uses 1st person plural verbs error-free, although less frequent.

3. The case of L.: 2nd person before 1st person, and plural before singular

L. is the earliest speaker among the Bulgarian children developing personal deixis according to the "formal-dominant" strategy. Compared to the other 4 subjects in the "formal-dominant" group, she shows developmental models typical of that strategy as well as unique characteristics that are worth noting.

L. begins coding the communicative roles of speaker, hearerand otherat the age of 1;5. Like all "formal-dominant" children, L. starts with a pre-pronominal stage of self- and other-reference, additionally marking the difference between addressee/non-addressee trough the opposition between vocative and the general noun form. However, L.'s pre-pronominal stage extends over 6 months, gradually turning out to be a "method" of avoiding personal deixis. At the same time, the child makes a significant progress in her grammatical development: About the age of 1;11, she produces seven of the nine Bulgarian tense forms, including , pluperfect, and future in the past in its counter-factive meaning; morphemes, such as definite article, prepositions, prefixes, reflexives,particles and conjunctions areregularly used, as well as pronominal clitics in their function of direct or indirect objects and of object doubling;different types of clause and sentence appear, too; substantial examples of complex sentences relative, object, temporal, subordinate clauses are registered as well.

At the same time, the child continues referring to herself mostly with 3rd person verbs plus verbal predicate. As far as 's own name is used for self-reference, it is in and object position, the being more frequent, e. g.:

(5) Keksče da napravi mama na Ijto.

Cake-DIM to make-3rd-P-SG- Mommy to Lili-Dim.

(Let Mommy make a cake to Lili)

The child uses her name in possessive construction as well, of a1st person possessive pronoun:

(6) Kǎde e lǎžičkata na Lilito?

Where is the spoon-DIM of Lili?

Where is Lili’s spoon?

Provided that the "formal-dominant" strategy is characterized with a preferential mastering of grammatical over pragmatic aspect of language, L.'s development illustrates this tendency in an evident and most pronounced manner through clear discrepancies between her formally precocious, but pragmatically underdeveloped speech. Her development, however, is of special interest due to some peculiarities in the process of acquisition of self- and other-reference that are discussed in sections below.


3. 1. First person plural forms before first person singular ones.

As Table 4 and Figure 3 make it clear, between the age of 1;5;13 and 1;9 L. used mostly 3rd person verb forms and/or names/kinship terms to refer to self, the addresseeand non-participant. A comparison between L.'s (Table 4) and S.'s (Table 3) early development shows an interesting difference: The frequency of 1st person forms in S.'s speech grows so fast that only in a month, between 2;0 and 2;1, these forms overpass the 3rd person forms used for self-reference6.  L.'s 1st person singular verbs and/or pronouns do not show a real growth between 1;5 and 1;10 years of age and, what is even more amazing,

they almost never occur in child's speech spontaneously, but only after adult's prompting utterances of the type "Say…[1st person verb and/or 1st person pronoun]!". Prompting utterances are used also the purpose of eliciting 2nd personforms for referring to the addressee.

The expected growth of 1st person forms is replaced in L.'s speech with a kind of 2nd person expansion. Self-reference with 2nd person forms (pronominal reversal), is the only alternative to 3rd person self-reference in this child's speech production for a period of 5 months, between 1;5;13 and 1;11. The percentage of 2nd person forms used instead of 1st person ones is especially high until the age of 1;9 – between 32% and 26%.

Adults' prompting utterances for eliciting 1st person forms appear only in cases of pronominal reversal, i.e. after an utterance where the child refers to self with a 2nd person form. Instead, overgeneralization of 3rd person forms for self- and other-reference are accepted by child's environment and even reinforced by the mother who often uses analogous examples of 3rd person forms when referring to self- and the child as addressee, since 3rd person overgeneralization is a salient characteristic of her CDS7.

At the beginning of the above mentioned 5-month-period, the child was only expected to repeat the right forms supplied by adults' prompting utterances.

(7) Mo: Tova ne sa kǎnki, a avtomobilče. .

These aren't skates, but an automobile-DIM. A car.

L(1;6): [Da] ka(r)aš-2/1!


[You to] drive! = Let me drive!

Mo: Kaži: "Iskam da karam kolata"!

Tell: "[I] want to drive the car"!

L: Mamo, i(s)kam (ko)lata ka(r)am!

Mommy-VOC, want-1st-P-SG-PRES the car [to] drive-1st-P-SG-PRES.

(Mommy, I want to drive the car)!

Later on, however, adult's promptings are transformed into mere reminders which do not include the form to be usedby the child. In such cases, L.'s adequate linguistic reaction proves she has no difficulties with the formal aspect of the adult-like elf- and other-reference, e. g:

(8) Mo: Kakvo iskaš? Da te svalja li?

What do you want? (You want) me to put you down?

L(1;9) : Da.


M: Kak šte kažeš?

How will you say (it)?

L: Mamo, svali me!

(Mommy-VOC, put-2nd-P-SG-IMP me-1st-P-PRON-ACC-CL down!)

Mommy, put me down!

(9) L (1;9): Da te-2/1 vzeme-3/2 mama!

Take-2nd-P-SG-EXHORT you-2nd-P-OBJ-CLIT Mommy!

Let Mommy take you up!

Mo: Kak šte kažeš?Mamo

How will you say (it)? Mommy…

L: Mamo, vzemi me!

Mommy-VOC, take-2nd-P-SG-IMP me-1nd-P-SG-OBJ-CL.

(Mommy, take me up!)

Table 4: Linguistic means used by L. for self-reference (percentages of tokens)


3rd person

1st sg

1st pl

2nd spontan





1;5;13 –








1;7 – 1;8







1;8 –1;9







1;9 – 1;10







1;10 – 1;11















Figure 3: Development of self-reference in L.'s speech: percentage of 3rd person verb forms and/or own name scored against 1st person forms and cases of pronominal reversal. (The child's age is given in months).


While first person singular verb and pronoun forms appear only after adults’ prompting, 1 person pluralverbs, contrary to the expectations, are in a varieof examples (,4% of all utterances for self-referencegiven between 1;8 and 1;9).During that period, L. has shown no difficulties both with the form and the meaning of 1st person plural forms. With these forms, the child referred to self as an integral part of the "alliance" speaker – addressee:

(10) Mo: Zašskǎsa knižkata?

Why did you tear the book?

L (1;8): Šte ja zalepime!

Stick-1st-P-Pl-FUT it-3rd-P-FEM-OBJ-CL together.

(We will stick it together).

(11) L(1;8;18): Da risuvame tuka. Da risuva Ijto s himikalčeto edno patence.

Draw-1st-P-PL-EXHORT. Draw-3rd-P-SG-EXHORT

Lili-DIM with the ball pen a duck-DIM.

(Tet us draw here. Let Lili drow here a duck-DIM).

(12) L (1;8;25): P(r)i Mečo! Da ideme p(r)i Mečo!

To Teddy-Bear! Go-1st-P-PL-EXHORT to Teddy-Bear!

(To Teddy-Bear. Let us go to Teddy-Bear!)

Mo: Može i pri Mečo da idem.

We can go to Teddy-Bear, too.

L: … kato dojde tate.

(…as soon as Daddy comes).


3.2. Second person instead of 1st person: pronominal reversal

Within a five-months-period (between 1;5 and 1.10) L. used 2nd person forms of verbs and pronouns for self-reference a relative high amount (. Table 4 and Fig3). the age of 1;9 almost one third of L.'s self-references were with 2nd person verb and/or pronoun, while between 1;9 and 1;10 the percentage was reduced to 8,3%.

In the psycholinguistic literature, cases of pronominal reversal are usually regarded as marginal. The majority of authors illustrate this phenomenon with single examples, that is why it is difficult to make quantitative comparisons on that basis. Nevertheless, some information about the quantitative aspect of pronominal reversal is available.


3.2.1. Referring to self with 2nd person: is it marginal?

Morgenstern and Brigaudiot (2005) found that 12% of the utterances produced by their French subject Guillaum between the age of 2;2 and 2;8, contained pronominal reversal, “along with adequate usage”.

Smoczyńska (1992) reports that for one of her Polish children, Kasia, 48 cases of 2nd person forms for referring to self were registered between the age of 1;7 and 1;8, as opposed to 320 other cases of self-reference during the same period. These scores are interpreted by the author as proving the marginality of the pronominal reversal phenomenon. However, the frequency of reversed 2nd person forms calculated in percentage is not as low as one could conclude on the bases of Smoczyńska's comments: It amounts 15%. Hence, although Kasia's utterawith pronominal reversal are not as abundant as Lilia's, they do not seem to be marginal either.

Compared to another child of the Bulgarian sample, K., L.'s percentage of utterances containing 2nd person self-reference is not unusually high: At the age of 2;0, K. referred to self with 2nd person forms in 40,63% of the cases, as Table 5 shows.

Table 5: Linguistic means used by K. for self-reference (percentages of tokens)


3rd pVerb and/or Name

1st sg


1st pl


2nd (spontan)










1st P (total)

2nd P (total)






Unfortunately, K. was not observed longitudinally, so it is not possible to reveal the dynamic of his acquisition of self- and other-reference.

In summary, it is out of doubt, that 3 (L., K. and V.) of the 5 Bulgarian children acquiring personal deixis according to the "formal-dominant" strategy, passed through a stage of pronominal reversal, i.e., they used 2nd person verbs and/or pronouns for self-reference not occasionally, although the frequencies of this usage varied considerably.


3.2.2. What does self-reference with 2nd person mean?

Different explanations of the pronominal reversal phenomenon have been given in the psycholinguistic literature (Clark 1978; Charney 1980; Chiat 1986; Deutsch et al. 2001, etc.). Most of them are not discussed here because of their irrelevance for the topic of this study. Children talk to themselves: Echo-repetitions and quotations of adults' utterances.

In their analysis of two French children, Brigaudiot & Morgenstern (1999) take an unusual perspective over the pronominal reversal: the authors offer a non-cognitive and non-linguistic explanation, connected with the development of children's personality within the family mirco-socium.

Brigaudiot & Morgenstern (1999) emphasize the psychoanalytic aspect of what they call “comments about themselves”. The believe that children can discernadults’ utterances describing them as “non-ordinary”, . e. “good” or “bad”children, on the basis of prosodic features or “mimics expressing astonishment, wonder or anger”. adults’emotionally markedutterances of the type, children’s echo-like responses, which results in pronominal reversal or in self-reference with a 3rd person pronoun. Later on, children reproduce the memorized echo-utterances as a kind of quotations, but only in comments about themselves and in autobiographic narratives. Brigaudiot & Morgenster illustrattheir claim with excerptions of two b , between 2;2 and 2;8years of age,sometimes referred to self with tu, il own name, il, although they have already mastered the self-reference with 1st person pronouns.

In a more recent publication, Morgenstern and Brigaudiot (2005) add to their psychoanalytic interpretation a cognitive one and test it against the speech production of one French child. The authors accept Chiat's (1986) hypothesis8 that the reversal errors are a kind of a shift of mental perspective, but they do not agree that this shift is "deliberate" as far as reversals are found in particular, well-defined contexts: "We can see that the child is producing an utterance out of a fixed scenario. He uses his auditory memoty with a situation associated to a sort of “quotation”. […] The script exists, the child does not create an utterance, he uses it because it applies to the present situation.” (Morgenstern & Brigadiot 2005).

The examples given by Morgenstern and Brigaudiot include child’s utterances meaning “congratulations” (Bravo tu marches!) or “reproaches” (T’as avalè encore!). Freud's conception of "ego-ideal", the authorssuggest that the "variability of the linguistic means for self-reference are to be explained with the diversity of self-conceptions and self-images, especially the good and bad selves, the hoped-for-selves, the ideal selves, etc." (Brigadiot & Morgenstern 1999).

The idea that children refer to self with 2nd person forms in situations where they anticipate, on the basis of their previous experience, what adults would say in such a situation, is not new. It was initially set forth by Ivan Georgov as early as at the beginning of the 20th century (cf. Georgov 1905). Georgov comments on his first son's self-reference: "Sometimes he speaks of self in second person; however, I don't think he doesn't know the difference in usage between 1st and 2nd person, rather he uses this way of expression either because he merely repeats the utterances, or because in these utterances he speaks to himself; one can come to this conclusion due to the fact that at the same time he speaks of self in 1st person…"9 Georgov illustrates his claim with examples proving the regulating function of adults' internalized utterances. Although it seems difficult to reduce them to "congratulations" or "reproaches", as Brigaudiot and Morgenstern suggest on the basis of their French excerptions, Georgov's examples (13 –14) similarly exibit internalized instructions, such as advice, warnings, negotiations, etc.

(13) Vlado (2;1): Papa šte (v)zeme (V)lado, ako bideš-2/1 miren.

Daddy take-3rd-P-SG-FUT Vlado, if be-2nd-SG quiet.

(Daddy will take Vlado, if you are quiet)

(14) Vlado (2;3): Šte se kači-3/1 (V)lado, ama da ne padneš-2/1.

Climb-3rd-P-SG-FUT-REFL Vlado, but not to fall-2nd-P-SG-EXHORT down.

(Vlado will climb, but you-2/1 should be careful not to fall down!)

While Georgov (1905) and Brigaudiot and Morgenstern (1999; 2005) are convinced that all cases of pronominal reversal in children's speech can be interpreted as "comments about themselves", Smoczyńska (1992) concedes that not all examples of Kasia's 2nd person self-reference should be interpreted as "anticipations of a possible utterance" addressed to the child by somebody in her surroundings. Besides, Smoczyňska reports data about another child, Agnieszka, who used 2nd person forms for self-reference only as immediate repetitions of adults' utterances. Hence, for some children, echo-repetitions could not only be regarded pre-conditions for "comments about themselves", but also, during a short initial period, as more or less independent means of self-reference. Internalized parents' representations in L.'s sample: Echo-repetitions and "super-ego" comments about herself

In order to compare L.'s 2nd person forms used for self-reference with those described in the above mentioned studies, these forms were classified into three types (cf. Table 4): "spontaneously" used, "repetitions", and "super-ego" ones. The term "super-ego" describes the regulative function of what Morgenstern & Brigadiot understand under "internalized parents' representations".

As the scorings in Table 4 show, L.'s "super-ego" usage of 2nd person forms is only active during a relative short period of time, between 1;7 and 1;9, and its percentage is rather low (8,51 and 3,1 respectively). Lilia's "super-ego" utterances, like those given by Georgov's son Vlado (cf. examples 13 – 14), are to be interpreted mostly as warnings and instructions (examples 15 - 16).

In some of these utterances, the child uses her name in a vocative form (cf. example 21), which reinforces the impression that she takes the perspective of a "meaningful other": She addresses herself in a way this "meaningful other" would have probably done, and reminds herself, on the behalf of the "meaningful other", what consequences her action could have (15), or what she is not allowed to do (16).

(15) L (1;7) is touching the glas-frame of the book-case and tells herself:

(Šte) po(r)ežeš (r)ǎčičkata tuk!

Cut-2nd-P-Sg-FUTUR hand-SG-DIM-ART here.

(You will cut your hand here)

(16) L(1;9): Da ne pipaš, Lilinke-Milinke, tuka kopčetata!

Touch-2nd-P-SG-, Lili-DIM-VOC-Mili-DIM-VOC, button-PL-DET here.

(Don’t touch, Lili-Mili, the buttons here!)

While in L.'s sample only few examples were registered where she addressed herself with her name in vocative, in the speech of Smoczyňska's Polish subject, Kasia, for the period between 1;6 – 1;8, they represented 11,25% of all utterance with self-reference and 75% of all utterances with self-reference in 2nd person forms.

In their "comments about themselves", Brigadiot and Morgenstern's (1999) French subject use not only 2nd person forms, but also the 3rd pronoun il plus own name. L.'s sample includes a single "super-ego" utterance where she speaks of self not as an "addressee", but as a "non-participant", i.e. using a 3rd person verb plus her own name (example 27). This utterance expresses what Brigadiot and Morgenstern (1999) call "congratulations": the child praises herself as a "hero", thus adopting the parents' role of a "judge" over her own behavior.

(17) Lilia (1; 7): Bravo na Lilito! Ig(r)ae si s balončeto.

Bravo to Lili-DIM. Play-3rd-P-SG-PRES-REFL with the balloon-DIM.

(Bravo to Lili! She is playing with the baloon!)

As it was already mentioned, the "super-ego" function of 2nd person forms for self-reference is not the only one exhibited in children's speech. If we take into account L.'s and K.'s data summarized in Table 4 and Table 5, as well as Agnieszka's case reported by Smoczyńska (1992), we cannot overlook the fact, that an important part of 2nd person forms for referring to self consists merely of echo-repetitions of adult's utterances. Since Agnieszka's 2nd person overgeneralizations are not scored, we will analyze only L.'s and K.'s data.

In K.'s speech (cf. Table 5), repetitions predominate, accounting for 30,62% of all cases of self-reference. For the period between 1;5;13 and 1;7 years of age, L.'s repeated 2nd person forms for self-reference amount 10,2%, but after that they completely disappear.

Both K. and L. used 2nd person verbs and pronouns in referring to self, which cannot be interpreted as repeating or echoing adults' 2nd person forms. Such cases are scored under the label "2nd spontaneous" in Table 4 and Table 5. While this usage accounts for only 10,2% of K.'s means of self-reference, L. used them much more frequently, in 22,4, 23,4 and 26,4% respectively, during three-and a half-months-period between 1;5;13 and 1;9. Before overcoming the 2nd person overgeneralizations at the age of 1;10, L. still referred to self spontaneously in 2nd person forms in 8,3% of the cases.

In summary, the greatest part of the 2nd person forms for self-reference in L.'s data belong to the "spontaneous" category. This "spontaneous" usage characterizes her speech production for several months (between 1;6 and 1;9), whereas the "super-ego" function of these forms as well as their "echo"-function are more restricted in time and quantity. Taking her collocutor's perspective – a general tendency in L.'s speech

Before discussing some plausible explanations of the high percentage of L.'s 2nd person overgeneralizations, let us mention an important characteristic of her communicative style: a general tendencyto take her collocutor's perspective. A large number of examplesillustrate that tendency.

Thus, in (18) and , the child finishes her collocutor's sentences, adding what the other would have possibly said. In (19), she cites her father's utterance, incorporating it in an adult-like way in her speech and demonstrating a precocious knowledge about the difference between direct and indirect speech. In (20) and (21), imaginary situations are verbalized, where the child pretends to be "talked to" by the kitten and by the doll. What is "redressed" as a quotation, is a playful construction of a "dramatized" dialogue. In (21), L.'s special sensitivity regarding her collocutor's perspective causes a precocious mastering10 of some Baby talk features, which, however, is observed mostly in fictitious dialogues with toys.

() Mo (to Lilia, 1;9): Haide, ljagaj, štom tolkova iskaš…

(Come on, go to bed, if you are so eager…

Lilia: … da nankaš.

… to sleep-BT-2p-sg-pres).

(… if you touch it!)

(19) Lilia (1:7): Tate kazal: "Nedej (da) pipaš, Iji!"

Daddy say-3rd-P-SG-PERF: not touch-2nd-P-SG-IMPP, Lili!

(Daddy said: Don´t touch, Lili!)

Mo: A Lilito kakvo?

And what is Lili doing?

Lilia: Pipa.


(She is touching)

(20) Lilia (1;10): Izleze kotenceto ot prozorčeto. "Mjau-mjau! Lili, kǎde si?"

Go-3rd-P-SG-AORIST out kitten-DIM-DEF of the window-DIM-DEF: Miau-miau, Lili, where be-2nd-P-SG-PRES?

(The kitten went out of the window: "Miau-Miau, Lili, where are you?")

(21) Lilia (1;10): Mahna Lilito kraka (na kuklata) i sega še plače.

Take-3rd-P-SG-AORIST away Lili-DIM the (doll’s) leg and now cry-3rd-P-SG-FUT.

(Lili took away the (doll’s) leg and now she will cry.

Še kaže: "Koj mi mahna kraka?" – še kaže. – “Lilito”.

Say-3rd-P-SG-FUT: “Who take-3rd-P-SG-AORIST away

me-1st-P- PRON-DAT-CLIT the leg?” - Say-3rd-P-SG-FUT. – Lili-DIM.

(She will say: “Who took my leg away?”- she will say. - “Lili”. Referring to the addressee: transition from 3rd person to 2nd person forms.

Although in pro-drop languages such as Italian, Spanish, Polish, Bulgarian, etc., 2nd person singular imperative verbs are among the first grammatically marked forms to be produced by children, these early 2nd person units appear merely as frozen forms and do not play any significant role in the acquisition. Indeed, the majority of studies on the development of personal deixis mention the fact that children start using 2nd person forms for referring to the addressee only after having acquired the 1st person forms for self-reference (cf. Georgov 1905; Deutsch & Pechmann 1978; Chiat 1986; Deutsch et al. 2001; Clark 1986; Smoczyńska 1992, etc.). That is why many children initially refer to the addressee with 3rd person forms plus names. This way of other-reference is characteristic of CDS as well, which reinforces its usage by the children.

What happens, however, when a child uses 2nd person forms in a self-referring function? Is it possible for these forms to play a "double-bind" role changing from speaker to the addressee according to the context?

The development of L.'s personal deixis proves that it is possible.

As it was already shown, during the 5 month-period between 1;5;13 and 1;10 years of age, L. used 2nd person forms with the function of self-reference. At that time, the addressee in her speech was usually marked with 3rd person verbs plus nouns in vocative. However, the 2nd person forms appeared in their proper function as well.

This happened mostly in contexts where the addressee was a kiof a fictitious collocutor, that is, non-present adult (cf. examples 22 and 24) toy(cf. example 23). As Table 6 shows, this function accounts for 80 to 90% of all 2nd person forms produced until the age of 1;8. In the following two months, the 2nd person verbs and pronouns start fulfilling their proper pragmatic function, that is, they begin referring to a real addressee in 33 to 34% of the cases, but the child still addresses fictitious communicative partners more often than real ones. At the same time, self-corrections, replacing the previous repetitions, mark the transition from overgeneralized 3rd person forms to the adult-like usage of 2nd person (cf. example 25). Self-corrections demonstrate how an adult-child interaction on a metalinguistic level can help the acquisition.

(22) L. (1;8), looking through the window at two men who can not hear her:

Ej, čičkovci! Bjagajte! Bavničko bjagajte!

(Hey, uncle-PL! Run-2nd-P-PL-IMPER! Slowly-DIM Run-2nd-P-PL-IMPER!

(Hey, uncles! Run! Run slowly-DIM!)

(23) L. (1;8), playing with her dolls Mimi and Dida:

Papkaj, Mimi! Pij kafence, Mimi!

Eat-2nd-P-SG-IMPER, Mimi! Drink-2nd-P-SG-IMPER coffee, Mimi!

(Eat, Mimi! Drink some coffee, Mimi!)

Zašto padna, Dido?

Why fall-2nd-P-SG-AORIS down, Dida-VOC?

(Why did you fall down, Dida?)

(24) L. (1;9), "addressing" her absent grandparents:

Zašto ne idvate, babo Mimi i djado Vanjo?

Why not come-2nd-P-PL-PRES, grandma Mimi and grandpa Vanjo?

Why don't you come, grandma Mimi and grandpa Vanjo?

(25) L. (1;9): Mamo!


Mo: Kakvo iskaš? Da te svalja li?

What do you want? (You want) me to put you down?

L.: Da.


M: Kak šte kažeš?

How will you say (it)?

L: Mamo, svali me!

(Mommy-VOC, put-2nd-P-SG-IMP me-1st-P-PRON-ACC-CL down!)

Mommy, put me down!

At the age of 1;10 L. refers with 2nd person forms to real addressees in 70%, but continues talking to fictitious ones in 30% of the cases. This is the age at which the great shift in her system of personal deixis takes place: she completely abandons the overgeneralized 2nd person form for self-reference, so that these forms continue appearing only in their proper function.

Table 6: Contexts of appearance of 2nd person forms for referring to the addressee in L.'s speech.




referring to fictitious collocutors

referring to a real addressee






1;5;13 - 1;7






1;7 – 1;8






1;8 – 1;9




6, 67


1;9 – 1;10






1;10 – 1;11







3.3. Baby talk register, pretended communicative roles, and the mastering of personal deixis

In playing scenarie with dolls or animals, L. demonstrates her growing knowledge of Baby talk and her ability to use it11 taking the adult part in that fictive, playful communication.

(26) L. (1;9): Mimi, ela na masata da te oblečem!

Mimie, comend-P-SG-IMP to the table (so that) to dressst-P-PL-PRES yound-P-Pron-OBJ-Cl.

(Mimi, come to the table so that we can dress you)

Ne možeš-2/1 [da] oblečeš-2/1 tova. Mama [da] obleče-3/2! Da

zakopčee-3/1 Lilito!

(You) cannd-P-SG-PRES not dressnd-P-SG-PRES it. Mommy (to) drend-P-SG-PRES (it). Lili buttonnd-P-SG-PRES this up.

(You/I cannot dress it. Let Mommy put-3/2 it on. Lilie cannot-3/1 put it on).

Mo: Njama kopčenca, milo!

(There are no buttons, darling).

L.: Ne može-3/1 Lilito da zavǎrže-3/1 tova. Čakaj2/2, Mimince, da ti složi 3/1 tova kaka Lili!

Lili-DEF cannotrd-P-SG-PRES tierd-P-SG-PRES it together. Waitnd-P-IMP, Mimie-DIM, older-sister Lili putrd-P-SG-PRES it to yound-P-PRON-CL!

(Lili can-3/1 not tie this together. Wait, Mimi, so that the older sister Lili put-3/1 this on you)

(27) L. (1;10): Stiga si-2/2 spala, Didko, tam na mama na legloto!

Stop sleep-2nd-P-SG-PERFECT, Dida-HYPOC-VOC, there on Mommy´s bed.

(Stop sleeping, Dida, there on Mommy´s bed)

Hajde glavata, Mimi! Da sǎ bleče-3/1 mama Mimkata! (she takes doll's shirt off)

Come on, the head, Mimi! Undress-3rd-P-SG-EXHORT Mommy Mimi

(Come on, the head, Mimi! Let Mommy undress Mimi!)

(28) L. (1;10): Zašto plačeš-2/2, Mečo? Nedej-2/2 plaka!

Why cry-2nd-P-SG-IMP, Teddy-Bear? Do-2nd-P-SG-NEG cry-INF

(Why are you crying, Teddy-Bear? Don’t cry!)

Kaka Lili e-3/1 tuka.

Older sister Lili is-3rd-P-SG-PRES here.

(Older sister Lili is here)

Several Baby talk markers are to be found in these examples. , in referring to the toys as addressees, the child uses:

  1. Only 2nd person forms (in contrast to the 3rd person for referring the adults)

hypochoristic derivation „Mimince“ of doll’s name Mimi and the diminutive “Didka” of the doll’s name Dida.

Secondly, in referring to self, she uses:

    1. Nouns expressing the pretended adult role: “mama” (Mommy) or “kaka” (older sister) plus 3rd person verb forms. Although before the age of 1;10 she uses reversed 2nd person forms for self-reference, these forms never appear when she talks to toys!

1st person plural forms in a kind of a patronizing, “nursery” talk: “Mimi, ela na masata da te oblečem!“ (Mimi, come to the table so that we can dress you).

Even more interesting is the dialogue demonstrating L.´s attempt at playful reversal of the adult-child social relationship. Addressing her mother, the child says: “Čakaj, maminko, da te izrešem -1st P-PL malko“. (Wait-2nd-P-SG-IMP, Mommy-DIM-VOC, to you-2nd---CL comb-1st-P-PL-PRES a little! - Wait, Mommy, for us to comb you a little!) the adult-like role in regard to her mother, the child (1;9) the proper 2nd person imperative and pronominal form for the addressee at the time she uses 2nd person forms for other-reference only when addressing fictitious collocutors, plus the ‘nursery’ 1st person plural for self-reference. An additional marker of this protective, patronizing style is the diminutive-vocative form of Mommy: Maminko!

4. Conclusions

4.1. The need for creating a shared communicative perspective

Let us summarize some specific features marking the process of acquisition of personal deixis in L.´s verbal development between the age of 1;5;13 and 1;11:

(1) During the whole period, she uses (with continuously diminishing frequency) the 3rd person verb forms/pronouns and/or own name for self-reference;

(2) 2nd person forms appear in her speech production simultaneously with the 3rd person forms, fulfilling two functions:

  1. Between the age of 1;5; 13 and 1;10, these forms refer to self (pronominal reversal)

Between the age of 1;5 and 1;8, 2nd person forms are used in their proper function, too, but only when the child addresses fictitious collocutors, that is, non-present adults, toys or animals.

After 1;8 years of age, the percentage of reference to real addressees increases, whereas the cases of pronominal reversal decrease, and at the age of 1;11, 2nd person is used only in its proper deictic function.

(3) The first person forms are initially quite infrequent and appear only

  1. as repetitions, after adult´s promptings,

and as quotations;

during the same period, the child uses 1st person plural forms for referring to herself as a part of the alliance “me and my collocutor”;

L. starts referring systematically to self with 1st person forms only after the age of 1;10, when the 2nd person overgeneralization have been overcome.

This short summary of L.’s developing personal deixis system makes it clear, that the child has no difficulties with the formal aspect of the deictic paradigms: starting from 1;6 years of age, she is able to produce 1st, 2ndand 3rdverb forms both in singular and plural; in a short time, corresponding personal pronouns appear, so that at the end of the period under analysis, at about 1;11 years of age, L. has filled all slots of the personal deixis paradigm, with the exception of the possessive pronouns12. This development, typical of the formal-dominant strategy, exhibits this child’s greater concern about the form than about the function of the personal deixis.

The most interesting phenomenon in L.’s individual strategy concerning the acquisition of self- and other-reference is the way she treats the 2ndperson forms. During 5 months (between the age of 1;5 and 1;10) L. refers to self with these forms, using them in

different contexts: initially as repetitions, then spontaneously. Between the age of 1;7 and 1;9, the child uses a small part of these spontaneously produced 2nd person forms (8, 51 and 3;1% respectively) which can be described as “super-ego” utterances. The term "super-ego" describes the regulative function of what Morgenstern and Brigadiot (2005) understand under internalized parents' representations. L.’s “super-ego” utterances clearly exhibit a “shifting perspective” (Chiat 1986): they express instructions, praises, advice, reproaches which the child supposes would have been pronounced by her “meaningful others”. This kind of a metalinguistic play (Morgenstern and Brigadiot 2005), however, can not explain the residual of over 20% of cases where the child refers to self with 2nd person forms without the purpose of relating her point of view to the supposed or memorized one of the adults in her surroundings. Similar examples are found also in the speech of Georgov’s son Vlado and of Smoczyńska’s Polish girl Kasia. Indeed, it still seems plausible to explain such spontaneous 2nd person reversals with Chiat’s (1986) hypothesis of “perspective shifting”, but, as the author herself admits, this explanation is rather speculative and needs additional investigations.

Let us now consider the second function of the 2nd person forms in L.’s speech during the analyzed period, the one of referring to the addressee. This function starts developing simultaneously with the one of self-reference, but for the period between 1;5;13 and 1;8 years of age the child addresses with 2nd person forms only fictive collocutors: toys or non-present adults.

However unusual this usage seems to be, it is not unique for L.’s development. In a kind of an elicitation procedure, other two of the “formal-dominant” children showed the clear tendency to produce 2nd person forms when addressing toys, and 3rd person forms when addressing adults (cf. the discussion in Stojanova-Trajkova 1986).

How could this phenomenon be explained?

Compared to real collocutors, fictitious ones represent the addressee’s role par excellence: While communicating with a real collocutor demands a permanent shift between the pragmatic roles of “speaker” and “hearer”, which leads to shifting pronominal reference, a fictitious collocutor is steadily addressed with the 2nd person. This saves the child the trouble to decide over and over which deictic form should be currently used.

Besides, talking to a fictitious collocutor gives the child the opportunity to create a shared perspective with the adult: The playful “talking” to toys and non-present persons, indeed, consolidates caregiver and child, thus setting them together against “the other”.

For L.’s development this kind of sharing perspective proves to be quite important. It explains the early emergence of 1st person plural forms in her speech at the background of 1st person forms as mere citations; the prolonged usage of 2nd person forms for self-reference along with the initial use of these forms when addressing fictitous collocutors, and, finally, the precocious (starting with the age of 1;9) mastering of Baby talk features, especially regarding the forms of personal deixis.


4.2. Could dolls and Teddy Bears play the role of missing siblings?

Recent studies suggest that various kinds of social upbringing, which shape the nature of input, can influence the acquisition in different ways. This is especially true of the domain of personal deixis.

In their study on the acquisition of personal and possessive pronouns, Deutsch et al. (2001) come to the conclusion that, as far as the development of pronominal reference is concerned, siblings are in a better position than singletons: “… singletons are disadvantaged in that they do not experience dyadic speech from the outside (as spectators) as often as siblings do. Such (observer) situations can provide a model of shifting reference in personal deixis, which could promote the use of pronouns in personal reference.”As singletons have fewer opportunities to observe “how other participants refer to themselves as speakers and others”, their input could be regarded as impoverished (Deutsch et al.,2001; cf. also Oshima-Takane 1988, etc.). Deutsch et al. (2001) point out that another explanation of the sibling effect in the domain of personal deixis could be possibly found in the interaction style called “confrontational”, as opposed to the “accomodational” style (Demuth 1992). The former style is more common in interactions between peers than in the adult-child dyads, and is typically used, for example, in conflict situations about alienable possession which provoke older children to use pronominal forms of the type: “Mine, mine, mine”. Such emphatic usage of pronouns “direct the attention of the younger sibling to both the form andthe function of personal deixis” (Deutsch et al. 2001).

Let us consider the children in the Bulgarian sample with regard of their family constellation. Only two of the seven children had older siblings: Georgov’s son Ženja and S. Presumably, Ženja profited from the sibling effect better than Stefan: while Ženja, acquiring the pronominal reference according to the “pragmatic-dominant” strategy, did not refer to self with 3rd or 2nd person forms, S. used 3rd person verbs and pronouns for several months, before the adult-like personal deixis has been mastered. Even more interesting is the development of Smoczyńska’s Polish subject Kasia, who had an older brother, but nevertheless passed through a stage of both 3rd person and 2nd person self-reference (cf. Smoczyńska 1992). On the other hand, O.13,  who followed the “pragmatic-dominant”, error-free acquisition of the personal deixis, is a singleton. It is clear that, as far as individual differences are concerned, we should speak only of tendencies, which could not predict the development of each concrete child.

Nevertheless, it seems plausible to hypothesize that in cases the child experiences a kind of an impoverished linguistic input with regard to the pronominal reference, s/he would try to compensate this by searching for opportunities to expand his/her experience over imaginary situations. L.’s development of personal deixis speaks in favor of such a hypothesis. She creates a shared communicative context with the adults when talking to fictitious collocutors such as dolls, Teddy Bears, and non-present adults. She memorizes what the others say or imagines their utterances and reproduces or construes them in a kind of “dramatized” turn-taking which incorporates direct speech into indirect one. She acquires Baby talk precociously and uses it in fictitious dialogues with toys in order to “rehears” the communicative role played with her by the adults.

Therefore, we could conclude that L.’s efforts are unconsciously directed to create a richer social environment and thus compensate the impoverished linguistic input typical of a singleton. This hypothesis seems to be supported through elicited responses received by K. and S. as well, who showed a clear tendency to address toys with 2nd person forms more often than adults. In order to prove this claim, additional and more exhaustive cross-linguistic analyses are needed.


The author would like to acknowledge Ina Vishogradska for her help with the English version of this paper.




  • Bates, E., Dale, Ph., Thal, D. 1995. Individual differences and their Implications for Theories of Language Development. In Fletcher,P., MacWhinney, B. (eds.), The Handbook of Child Language. Oxford: Blackwell. 96 – 151.
  • Bloom, L., Lightbown, P., Hood, L. 1975. Structure and variation in child language. In Monographs of the Society for Research in Child Language Development 40 (Serial N0 160).
  • Brigaudiot, M., Morgenstern, A. 1999. The Self as Other. Own name, other name, second and thirs person pronouns as designators for self by French speaking children. Poster presented at the VIIIth International congress for the study of Child Language, San Sebastian, 12 – 16 July, 1999.
  • Brown, R. 1973. A First Language: The Early Stages. Cambridge, Mass.: Harvard University Press.
  • Charney, R. 1980. Speech roles and the development of personal pronouns. Journal of Child Language 7. 509 - 528.
  • Chiat, S. 1986. Personal pronouns. In P. Fletcher & M. Garman (eds.), Language acquisition: Studies in first language development. Second Edition. Cambridge: Cambridge University Press. 339 – 355.
  • Clark, E. V. 1978. From gesture to word: on the natural history of deixis in language acquisition. In J. S. Bruner, A. Garton (eds.), Human growth and development. London: Oxford University Press. 85 - 120.
  • Clark, E. V. 1986. The Acquisition of Romance, with Special Reference to French. In Slobin, D. I. (ed.), The Crosslinguistic Study of Language Acquisition. Volume 1: The Data. Hillsdale: Lawrence Erlbaum. 687 – 782.
  • Deutsch, Werner, Pechmann, Thomas. 1978. Ihr, dir, or mir? On the acquisition of pronouns in German children. Cognition 6. 155 – 168.
  • Deutsch, W., Wagner, A., Burchardt, R.,. Schultz, N., Nakath, J. 2001. Person in singletons, siblings, and twins. In M. Bowerman & S. C. Levinson (eds.), Language acquisition and conceptual development. Cambridge: Cabridge University Press. 284 – 315
  • Georgov, I. A. 1905. Die ersten Anfänge des sprachlichen Ausdrucks für das Selbstbewusstsein bei Kindern. Archive für gesamte Psychologie 5. 329- 404.
  • Halliday, М. A. K. 1975. Learning How to Mean. In Lenneberg, E. H.Lenneberg, E. (eds), Foundations of Language Development: A Multidisciplinary Approach. New York, Academic Press. 239 - 265.
  • Horgan, D. 1980. Nouns: love ‘em or leave ‘em. In Teller, V., White, S. (eds.), Studies in child language and multilingualism. Annals of the New York Academy of Sciences 345.
  • Huxley, R. 1970. The development of the correct use of subject personal pronouns in two children. In Flores d’Arcais, G. B., Levelt, W. J. M. (eds.), Advances in psycholinguistics. Amsterdam North-Holland. 141 - 165.
  • McNeil, D. 1963. The psychology of you and I: a case history of a small language system. Paper presented at the APA Symposium on Child Language: Structural Aspects. August, 1963.
  • Mills, A. 1985. The Acquisition of German. In D. I. Slobin (ed.), The Crosslinguistic Study of Language Acquisition, Volume 1: The Data. Hillsdale: Lawrence Erlbaum. 141 – 254.
  • Morgenstern, A., Brigaudiot, M. 2005. “Bravo, tu marches!” A French child’s use of second person pronoun. Poster presented at the Xth International Congress for the Study of Child Language, Berlin, July 25 – 29, 2005.
  • Nelson, K. 1975. The nominal shift in semantic-syntactic development, Cognitive Psychology 7. 461 – 479.
  • Oshima-Takane, Y. l988. Children learn from speech not addressed to them: The case of personal pronouns. Journal of Child Language, l5. 94-l08.
  • Oshima-Takane, Y. l992. Analysis of pronominal errors: A case study. Journal of Child Language 19. 111-131.
  • Peters, A. 1977. Language learning strategies: does the whole equal the sum of the parts? Language 53. 560 – 573.
  • Savić , . 1972. Usvajanje lične zamenice prvog lica jednine u dečjem govoru. Psihologija (1). 59 – 67.
  • Smoczyńska, M. 1992. Kategoria osoby w języku dziecka. In B. Bokus& M. Haman (eds), Z badań nad kompetenciją komunikacijną dzieci. Warszawa: Energia. 203 – 228.
  • Stern, C., Stern, W. 1928 [1907]. Die Kindersprache. Leipzig: Barth.
  • Stojanova-Trajkova, J. 1986. Young children's acquisition of formal and pragmatic aspects of the category of person. In I. Kurcz, G. W. Shugar, J. H. Danks (eds.), Knowledge and Language. North-Holland, Elsevier Science Publishers. 415 - 430.
  • Stojanova, J. 2005. Samonazovavane i personalen deiksis. (To appear in a book in honor of Ruselina Nicolova, 2005).
  • Vasić, Vera. 1983. Govor sa sestrom i bratom. Novi Sad: Forum.
  • Wills, D. D. 1977. Participant deixis in English and baby talk. In C. Snow, Ch. A. Ferguson (eds), Talking to children: Language Input and Acquisition. Cambridge. 271 – 296.
  • 1. Ivan Georgov is one of the first to discuss how this difference affects the acquisition of the category of person.
  • 2. Mills (1986) cites an example of self-reference with a verb in the homonymous 1/3 person, produced by a 1;10-old child: Mag nicht esse, interpreted as “I don’t want to eat”. She comments the unclear character of the e-ending for 1st person present which can be interpreted as a variation of the infinitive –en suffix with the omission of –e as well.
  • 3. In German, 1st and 3rd person singular are homonymous for all preterit verbs and for all present and past modal verbs. In Russian, the present tense paradigm of the verb byt’ (be) ceases to exist (in the 13th century); as a result of this ø –morphemic realization, present tense copula predicates and all past tense verbs are not inflected for person.
  • 4. Veronika is a singleton and a late speaker, while Ženja has an older sibling and is about 6 months in advantage regarding Veronika. Unfortunately, Georgov's study (1905) on the development of self- and other-reference did not include quantitative data, only the first occurrences and examples illustrating the developing forms are given. Ženja started using the 1st person pronoun with the age of 1;7;16, while the 1st person verbs appeared a month later (with the age of 1;8;20). Second person pronouns and verb forms were two months postponed in regard to the first person forms (1;9;13 and 1;10;8 respectively).
  • 5. In her description of the transitional stage between self-reference with 3rd person and self-reference with 1st person forms Smoczyńska registered this transitional stage in most of her Polish subjects.
  • 6. Authors describing a pre-pronominal stage in the acquisition of personal deixis (cf. Georgov, 1905, Smoczyńska 1992, etc.), emphasize, too, that in this stage children use, at least sporadically, 1st person verbs for self-reference which in a month or two significantly grow in number, become more frequent and successfully compete with the 3rd person forms in the same function.
  • 7. In her study on the acquisition of personal dexis by Polish children, Smoczyńska (1992) discusses some interesting interdependencies between the use of 3rd person forms for self- and other-reference in children’s and their mothers’ speech.
  • 8. “The child says what he expects to hear or what “should” be said. If this is the case, we might not consider the use of the pronoun you as a reversal and as being the “incorrect” form, but think of the whole sentence as being uttered by the “wrong” speaker".
  • 9. The English translation of Georgov’s citations is made by the author of this paper according to the Bulgarian edition of his study published simultaneously in German and in Bulgarian: Georgov, I. Părvite načala na ezikovija izraz za samosăznanieto u decata. V: Periodičesko spisanie na bălgarskoto knižovno družestvo, 66 (1-2), 1905, 31 – 94.
  • 10. Usually, children do not start using some characteristics of Baby talk in their speech to younger siblings before the age of 3 (cf. for example the study of Vasić, 1983 on children acquiring Serbo-Croatian, a South-Slavic language like Bulgarian).
  • 11. Early usage of Baby talk components by children has been reported (cf. Vasić, 1983), but they are registered in the speech of children at the end of their third year.
  • 12. An important difference between Bulgarian and Polish, on one side, and English, German and French, on the other, consists in the acquisition of possessive pronouns. It has been considered for the first time by Georgov (1905), who mentions that in his Bulgarian material, unlike German, English and French, possessive pronouns do not appear before personal ones, and that his sons never replaced personal with possessive pronoun forms. This is true also of the other children in the Bulgarian sample (cf. Stojanova-Trajkova 1986) as well as of Smoczyńska’s (1992) Polish subjects.
  • 13. There are other factors, too, which could explain Veronika’s error-free acquisition of the category of person as contrasted to Kasia’s development: it is the difference between a late (Veronika) and an early (Kasia) speaker. Probably, after the age of 2;0, children pay more attention to the communicative roles and their linguistic marking than before that age (Stojanova-Trajkova 1986).