Michał Głuszkowski
Nicolaus Copernicus University in Toruń  
(Univerza Nikolaja Kopernika v Torunju) 
micglu@umk.pl
Slavistična revija 71/2 (2023): 161–174
UDK 811.162.1'27:81'246.2
DOI 10.57589/srl.v71i2.4109
Tip 1.01
Code-Switching and Code-Mixing on Different Language Levels:  
The Case of a Polish Language Island in Siberia
The aim of the article is to discuss the diverse nature of code-switching and mixing phenom-
ena on different language levels. In the field of phonetics, morphology and lexis, researchers 
most often focus on the interlingual influence in the form of interference, while code-switch-
ing is a phenomenon that occurs at the level of syntax. In this article about linguistic material 
from a Polish language island in a Russian environment (the village of Vershina in Siberia), 
examples of lexical and structural borrowings and various forms of language change during 
speech were indicated. However, this approach only allows for the characterization of selected 
fragments of utterances, while whole texts function in live communication. Therefore, it was 
proposed to include the most complex level of language, i.e., the text as such, in the analyses 
of code-switching and mixing. Selected examples show how this complementation affects the 
perception of phenomena related to bilingualism.
1
Keywords: code-switching, code-mixing, language contact, bilingualism, sociolinguistics
Preklapljanje in mešanje kodov na različnih jezikovnih ravneh: 
primer poljskega jezikovnega otoka v Sibiriji
Cilj prispevka je obravnavati raznolikost pojavov preklapljanja in mešanja kodov na različnih 
jezikovnih ravneh. Na področju fonetike, morfologije in leksike se raziskovalci najpogosteje 
osredotočajo na medjezikovne vplive v obliki interference, medtem ko je kodno preklapljanje 
pojav, ki se pojavlja na ravni skladnje. V članku o jezikovnem gradivu s poljskega jezikovnega 
otoka v ruskem okolju (vas Veršina v Sibiriji) so bili nakazani primeri leksikalnih in strukturnih 
izposojenk ter različne oblike jezikovnih sprememb med govorom. Vendar ta pristop omogoča 
le karakterizacijo izbranih fragmentov izrekov, medtem ko v živem sporazumevanju delujejo 
celotna besedila. Zato je bilo predlagano, da se v analize kodnega preklapljanja in mešanja 
vključi najkompleksnejša jezikovna raven, tj. besedilo, izbrani primeri pa pokažejo, kako to 
dopolnjevanje vpliva na dojemanje pojavov, povezanih z dvojezičnostjo.
Ključne besede: kodno preklapljanje, kodno mešanje, jezikovni stik, dvojezičnost, 
sociolingvistika
1 Introduction 
1.1 The community under study
The issue of code-switching and mixing is one of the most discussed topics in the 
literature on language contacts. The purpose of this article is to analyse the phenomenon 
in question on different language levels in a community that emerged as a result of the 
1
 The publication evoked as part of the project “Code switching in the conditions of Polish-Russian 
bilingualism in the Polish language island in Siberia (the village of Vershina near Irkutsk)”, funded by the 
National Science Centre of the Republic of Poland, granted under decision number DEC-2016/23/B/HS2/01200.

Slavistična revija, letnik 71/2023, št. 2, april–junij 162
migration of Slavic community and their settlement in a foreign, but also Slavic-speaking 
environment. Vershina is a village founded by Polish migrants in Russia in 1910 in 
the Irkutsk oblast. Unlike most Poles in Siberia, they were not exiles, but voluntary 
settlers from various parts of Lesser Poland, who wanted to improve their material 
situation by taking advantage of the land grants and tax reliefs made possible by the 
land reform introduced by the Prime Minister of the Russian Empire Peter Stolypin in 
1906 (Bazylow 1975: 72). The community was bilingual from the beginning, because 
the settlers came from the areas that were part of the Russian partition (Polish lands 
incorporated into Russia in the years 1795-1918). Their bilingualism is connected with 
diglossia, which was initially characterized by the dominance of the Lesser Poland 
dialect in most domains, and currently it is used to a greater extent only in ethnically 
homogeneous Polish families (Głuszkowski 2012: 46-7). 
Vershinian community represents the “folk” and asymmetrical type of bilingual-
ism, which means that one of the languages has been acquired in the course of daily 
communication, and the knowledge and use of both codes is unbalanced – with a 
clear dominance of the majority language, i.e. the official language of the country (cf. 
Bullock, Toribio 2009: 9; Głuszkowski 2015: 56). 
The material under analysis was collected in the course of field research during 4 
expeditions to Vershina and includes almost 100 hours of individual and group inter-
views. A significant part of the material was recorded and allowed for the creation of 
a corpus of 275,000 words.
1.2 The specificity of language contact in language islands
The community in which the linguistic material analysed in this study was obtained 
can be characterized as an insular settlement. According to Claus Hutterer, “language 
islands are internally structured settlements of a linguistic minority on a limited geo-
graphical area in the midst of a linguistically different majority” (cit. and transl. after 
Rosenberg 2005: 221). Thus, the island must be clearly smaller than the sea on which 
it is located (Nowicka 2011), and it is a specific kind of minority. Although this term 
(Ger. ‘Sprachinsel’) was coined to characterize a Slavic community surrounded by a 
German majority in East Prussia over 170 years ago, it is still widely used in various 
subdisciplines of linguistics (Rosenberg 2005: 221). One of the main reasons to study 
language islands was their isolation, because of which the researchers expected to an-
alyse phenomena of language contact as well as preserved archaisms in an undisturbed 
form (cf. Löffler 1987: 387). However, for the same reason the concept of island has 
been criticised, e.g. by Erik Eriksen, who argues that neither a community is entirely 
isolated, nor cultural boundaries are absolute, because “webs of communication and 
exchange tie societies together everywhere, no matter how isolated they may seem at a 
first glance” (Eriksen 1993: 134). This does not mean, however, that this idea should be 
rejected, but treated with appropriate reserve, as an analytical model based on relative, 
not absolute isolation. It was in this spirit that the concept of island communities was 

163 Michał Głuszkowski: Code-Switching and Code-Mixing on Different Language Levels
developed in sociolinguistics, especially in relation to languages with a small number 
of speakers:
Островные ситуации возникают вследствие переселения части того или иного этноса 
по тем или иным причинам – социально-политическим, военным, экономическим и 
прочим – в регионы проживания иного этноса (иных этносов) с иным языком (иными 
языками). Оторванные от исходного этно-языкового корня, такие острова постоян-
но ощущают языковой дефицит, особенно что касается его использования в области 
культуры, образования, науки. С одной стороны, их представители должны овладеть 
языком окружающего этноса (что в действительности и наблюдается), с другой – для 
них важно сохранить язык своих предков (Дуличенко 1998: 26).
One has to note, that the inhabitants of Vershina represent a special case of language 
islands – a dialect island. The language used by them is a non-standard variety, and 
exists almost entirely
2
 in the oral form, without a stabilized norm and factors supporting 
the preservation of the language in the minority education system. Such a state favours 
the intensification of foreign influences (Weinreich 1963: 85-6). 
2 Theoretical assumptions 
2.1 Language contact and inter-lingual influence
According to Uriel Weinreich’s definition “languages are in contact if they are 
alternately used by the same persons” (Weinreich 1963: 1). This phenomenon belongs 
to parole, which means that the processes associated with it occur on the “fluency con-
tinuum” between perfect bilingualism and pure monolingualism (Myers-Scotton 2005: 
43). So it is not a state that is achieved once and for all, but a living process and there 
are some disturbances within it. These are primarily the processes of borrowing and 
changing the language within the utterance. What is the basic difference between them? 
The first include both matter (MAT-) borrowings, i.e. “direct replication of morphemes 
and phonological shapes from a source language” and pattern (PAT-) borrowings, i.e. 
“re-shaping of language-internal structures, [in which] the formal substance or matter 
is not imported but is taken from the inherited stock of forms of the recipient or replica 
language (i.e. the language that is undergoing change)” Matras, Sakel 2007: 829-30). 
The latter mean switching between two language systems and allow you to indicate 
the boundaries between them (cf. e.g. Poplack 1988: 220; Sayahi 2014: 81). However, 
it should be borne in mind that an easy and unambiguous indication of the boundaries 
between languages is possible in full bilingualism, but often “speakers change their 
language (A) to approximate what they believe to be the patterns of another language or 
dialect (B)” (Thomason 2001: 142). The example of relations between the Lesser Poland 
dialect with the Russian language in Vershina, belongs to such cases where speakers, 
not having mastered one of the codes fully, make certain approximations by adapting 
foreign elements and uncontrollably changing the language to maintain conversation. 
2
 Although there are few texts written in the Lesser Poland dialect in Vershina, they are limited to the 
lyrics of the folk band Yazhumbek [Less. Pol. Hazel grouse] and have practically no impact on everyday 
communication and code-switching and mixing phenomena (cf. Ananiewa 2013).

Slavistična revija, letnik 71/2023, št. 2, april–junij 164
In the case of a change of language during the speech for the purposes of this article, 
a clarification will be needed, which we will make after Peter Auer, distinguishing 
code-switching as the juxtaposition of codes “perceived and interpreted as a locally 
meaningful event by participants” and mixing as those cases, which “cannot be labelled 
language A or language B”, mainly due to the frequency of switches (Auer 1999: 310-4). 
The diversity of the latter was embraced in the theory of Peter Muysken, who defined 
the following types of code mixing:
a) Insertional (an element from L
1
 language inserted in L
2
 sentence; Muysken 
2000: 3), e.g. tˈeras na koncˈerty xoć stav’ˈajom to
POL
 ńiktˈo
RUS
 ńe xce iść na 
kˈoncert
3
POL
 [Although they are organizing concerts now, no one
RUS
 wants to go 
to the concert]; cˈałe žˈyće
POL
 || dˈa
RUS
, | ja cˈau̯e žˈyće p’isoł, z ʒ́eśuntˈygo rˈoku i 
p’ˈiše
POL
 pr’ėdisłˈov’ė
RUS
 sfˈojej kšˈonšk’i žebˈyśće v’eʒ́ˈel’i
POL
 [Whole life
POL
. Yes
RUS
. 
I was writing whole my life, since I was ten and I am still writing
POL
. The fore-
word
RUS
 to my book, you know?
POL
]. The main language of the utterance is L
1
, 
and the inserted L
2
 element is not subject to integration to L
1
, but is transferred 
in its original form and content. In other words, L
1
 serves as the morphosyntactic 
frame for the clause, and the L
2
 elements are “embedded” in it (cf. Myers-Scotton 
2004: 106-7). 
b) Alternational (full shift from L
1
 to L
2
 in a single speech act), e.g. muj ʒ́ˈadek m’au̯ 
mu̯yn | gdy zmaru̯ | bur’ˈaći bˈarʒo pu̯akˈal’i
POL
 || xαrˈošyj čėłav’ˈek | ńikαgdˈa ńė 
αtkˈazyvəł
RUS
 || xoʒ́‿ʒ́eń | xoʒ́‿noc pšˈyjm’e | tag‿go fspom’inˈal’i
POL 
[My grand-
father had a mill. When he died, Buryats mourned him a lot
POL
. He was such a 
good man, he never refused
RUS
 (to help). It did not matter for him if it was day 
or night. That’s how they remembered him
POL
];  
I1
4
: nˈo jˈo bˈy zˈaros porozmˈov’ou̯ po pˈolsku [Well, I would gladly speak Polish now]
I2: jˈes’l’i v’ˈiće že jˈes’l’i co jo ńe tak no ńe tˈak pˈov’im to [If you notice that I tell something not 
correctly] 
I3: no no [to granddaughter] fs’o | v’ˈižu | bˈol’še ńelz’ˈa
RUS
 [That’s all. I see you do not need 
any more].
I2: bˈo jˈo… [because I…]
I1: nˈe vˈy tˈo dˈobže rozmˈov’oće pˈo pˈo polsku | bˈo tˈak’im jˈag mˈy | ńiktˈužy pšyjižǯˈajum bˈarʒo 
ńˈe ˈiʒ ́ e pojˈuńć tˈak’e swˈova [No, you are speaking Polish very well, just like us. Because 
sometimes those who come here are speaking in the way we do not understand].
According to Muysken, alternation is the only form of juxtaposition of the codes 
during speech that can be defined not as mixing but as code-switching (Muysken 
2000: 4). These types of switches depend on the topic, but also other conditions: place, 
time, participants of communication (Gumperz 1977: 1). In the cited examples, the 
speakers change the code either because they are quoting some else’s words with an 
introducing phrase bur’ˈaći bˈarʒo płakˈal’i [Buryats mourned him a lot] or because they 
3
 All examples have been written in a simplified Slavic transcription, and grammatical descriptions have 
been prepared according to the Leipzig Glossing Rules.
4
 If there were more participants in the conversation, their utterances are numbered.

165 Michał Głuszkowski: Code-Switching and Code-Mixing on Different Language Levels
are addressing a person who does not speak Polish (a granddaughter). Such changes 
include a full switch in terms of the language of the matrix as well as the elements 
filling it from L
1
 to L
2
.
c) Congruent lexicalization (largely or completely shared syntax structure, lexicalized 
by elements of either language), e.g. 
to śe stroˈiu̯o
MAT
 tam i tam to to fstr’ˈeća bywˈa fs’ˈex
RUS
 | kto co skunt pšy… popšyjˈiž:ou̯ i 
ugošˈeńije
RUS
 bˈyu̯o fs’o
RUS
 bˈyu̯o dˈobže | da
RUS
 | no a pˈotym jo ńe v’ėm | coś ńe stˈal’i tudˈa 
zaxˈoʒ ́ ić krˈom’e
RUS
 kośćˈou̯u
POL
 ńikudˈa
RUS
 | a cˈyje to to ńe znaj…
RUS
 ńe v’im [It was being 
built
MAT
 there, and there was the meeting for everyone
RUS
 | who came here from various 
places, there was the treat
RUS
, everything
RUS
 was good. Yes
RUS
. And after that, I do not 
know. For some reason we stopped going there, except
RUS
 of the church
POL
, nowhere
RUS
. 
And whose it is? I do not kn…
RUS
 I do not know].
This is a special type of code mixing in which some elements of the other two can be 
find, and words from both languages are “inserted more or less randomly” (cf. Muysken 
2000: 3-5). In the case of the above excerpt, the intention of the speaker was to formu-
late a statement in Polish, but there were cases of inserting single words (ugošˈeńije 
[the treat], fs’o [everything], da [yes]) and component sentences (fstr’ˈeća bywˈa fs’ˈex 
[there was the meeting for everyone]; ńe stˈal’i tudˈa zaxˈoʒ́ić krˈom’e
RUS
 kośćˈowu
POL
 
ńikudˈa
RUS
 [we stopped going there, except
RUS
 of the church
POL
, nowhere
RUS
]), which, 
according to the interpretation of many researchers, should be treated as alternations 
(cf. Auer 1999: 309-10; Myusken 2000: 3-4; Deuchar 2020: 1-2). The situation is ad-
ditionally complicated by the fact that in the second sentence there is a Polish element 
of kośćˈowu [church
GEN.SG
], used in accordance with the dialectal paradigm of inflection, 
so it cannot be treated as a Polish borrowing in a Russian utterance, but as an insertion. 
There is also a switch, which resulted from self-correction: ńe znaj…
RUS
 ńe v’im
POL
 
[I do not kn…
RUS
 I do not know
POL
]. The informant started sentence in Russian, but 
did not finish it and quickly came back to the main language of the utterance. Russian 
influence is also visible in a MAT-borrowing śe stroˈiu ̯ o [it was being built] (cf. Rus. 
строилось), which has been integrated in the target language. Moreover, due to the 
structural closeness of two Slavic language in contact, the utterance contains words that 
could be characterized both as Polish and Russian, e.g. to [it], tam [there], kto [who] 
and other, underlined in the analysed passage. 
In order to analyse such cases, after Svyatlana Tesch and Gerd Hentschel, two levels 
of articulation will be distinguished. The first embraces the lexis and syntactic structures 
as well as those elements of the expression plan that determine their morphological 
form in the mind, i.e. they relate to the content plan (signifié). The second level is a 
phonetic-phonological surface, i.e. the plan of expression (signifiant). Thus, there are 
three possible convergence relations, visible primarily in cognate languages: a) both 
at the deep-morphonological and phonetic-phonological level, e.g. tam
POL
 and tam
RUS
 
[there]; b) only at the deep level, while the phonetic-phonological form is different, 
e.g. bˈyła
POL
 and bˈyła
RUS
 [was
PST.F.3SG
]; c) the units are different on both levels, e.g. 
fšystko
POL
 and fs’o
RUS
 [everything] (Теш, Хенчель 2009: 210). Of course, if congru-
ently lexicalized words with the same morpho- logical or morpho- phonological and 

Slavistična revija, letnik 71/2023, št. 2, april–junij 166
phonetic structure occur in the course of a longer utterance in one of the languages, we 
treat them as an element of a given code, e.g. the negation ńe [not
NEG
] in the sentence 
no a pˈotym jo ńe v’ėm [and then I do not know], despite to the similarity to the second 
language. However, if they occur at the junction of fragments in L
1
 and L
2
, they can be 
considered part of each of them, e.g. bˈyu̯o [was
PST.N.3SG
] in ugošˈeńije
RUS
 bˈyu̯o fs’o
RUS
 
bˈyu̯o dˈobže
POL
 [there was the treat, everything was good]. Since it is impossible to 
unambiguously attach this word to any of the codes, it is a prototypical example of 
congruent lexicalisation, and the entire excerpt, due to the mixture of elements both 
form L
1
 and L
2
 can be characterized as such. 
2.2 Different language levels and the phenomena of language contact
An attempt to break down congruently lexicalized speech into prime factors is to 
relate it to the analytical levels of language distinguished in structuralism: phonet-
ics and phonology, morphology, lexis, syntax (Haas 1960: 267). In the traditional 
structuralist approach, foreign language influences are analysed at different levels of 
language (cf. e.g. phonetic, grammatical and lexical interference in Weinreich 1963). 
In considering the nature of these influences, it should be taken into account that all 
residents of Vershina speak Russian at the level of monolingual native speakers, and the 
knowledge of the Lesser Poland dialect is varied and ranges from the communicative 
minimum among the younger generations in mixed families to a relatively high level 
among older people living in homogeneously Polish families. The Russian language 
is mastered both in everyday communication and in the course of school education, 
which results in a strong orientation to the norm and limits the scope of interference 
to a minimum. This results in an uneven nature of the linguistic influence: changes 
under the influence of contact are observed practically only in the heritage language, 
i.e. the Lesser Poland dialect. The switches that take place between the two languages 
take place as part of utterances intentionally formulated in Polish, in which Russian 
elements are intertwined, but also in the matrix language itself, the Russian influence is 
also present, although not to varying degrees on all levels. As Svetlana Mitrenga-Ulitina 
notes, the phonetic system of the Vershinian dialect is the most resistant to Russian 
influence (Mitrenga-Ulitina 2015: 145), which is also confirmed by our analyses. The 
characteristic features of the Vershinian dialect, such as mazuration (or mazurism), i.e. 
the pronunciation of the alveolar consonants š, č, ž, ǯ as dental s, c, z, ʒ (see Sawicka 
2020), have not only been preserved in the original Polish lexicon, e.g. cˈysty [clean, 
pure] cf. stand. Pol. czysty and Rus. чистый, but also appear in adapted borrowings 
from the Russian language, e.g. polˈacka [a Polish woman
COLL
], cf. Rus. полячка.
The Russian influence at the morphological level is largely connected with the 
lexical one. In the Vershinian Lesser Poland dialect, borrowed morephemes are stems, 
not prefixes or suffixes, and therefore, according to the most widespread point of view 
in the study of language contacts (cf. e.g. Poplack, Sankoff 1984: 104; Thomason, 
Kaufmann 1988: 37), they should be treated as a form of adaptation of borrowings 
through the exchange of affixes, and not interference at the level of morphemes. The 
borrowed stems take a phonological form and are equipped with affixes typical of the 

167 Michał Głuszkowski: Code-Switching and Code-Mixing on Different Language Levels
recipient language, such as -stroj- [-build-] in its basic meaning ‘to build’: o te dˈumy 
v’ˈincėj o tak strojˈune bˈywy [Oh, these houses were being built in this manner]; dˈum 
tˈak’i v’ˈe:lg’i postrojˈiwy [They built such a big house], as well as in the metaphorical 
colloquial meaning ‘to find a place, to be settled’: ˈone f posˈolstvo kajś pšystrojˈune 
[They have been settled somewhere in the embassy]. The combination of lexical and 
morphological influence is also visible in newly coined words reproducing Russian 
models in the form of word-formation calques (i.e. PAT-borrowings) morpheme-by-mor-
pheme, e.g. pšʼev ʾėść na nacjͻnˈalny jˈiῖ̯zyk [to translate into the national language], cf. 
Rus. перевести [to translate] – the word has been created in the course of replacement 
of Russian morphemes with the Polish ones: p’er’e- : pše-, -v’est’i- : -v’eść-. 
However, there are cases that can be described as half-calculus or linguistic hybrids, 
e.g. pˈotem zapšecˈone bˈyu̯o fśo ‘after that everything was forbidden’ – the new word 
zapšˈecać
INF
 was created under the influence of Rus. запрещать [to forbid]. The mor-
pheme – pr’e- was replaced with its Pol. equivalent -pše-, and the second part has been 
phonetically adapted in the course of sound substitution š:’/šč’ : c. It is a similar case 
to Polish bawełna, or Czech bavlna [cotton] reproducing German Baumwolle, where 
the word is partially “translated” (morpeheme substitution) and partially phonetically 
adapted (sound substitution) (cf. Weinreich 1963: 57-63). The new word has Polish 
conjugation and can be considered a form of PAT-borrowing.
The most common and most noticeable, and thus the best described manifestation 
of foreign influence at the lexical level are MAT-borrowings, i.e. single words and 
complex lexical units, which appear in the target language because of the need to 
describe a fragment of reality that requires the use of vocabulary present only in the 
source language. The process of borrowing in this respect has a lot in common with 
the insertional type of code-mixing. The main cause of the use of an integrated (MAT-
borrowing) or not integrated (insertion) foreign element is code repair (cf. Gafaranga 
2012: 509-10). Since these processes are especially vivid in insular communities expe-
riencing deficit of their lexical resources (cf. Дуличенко 1998: 26), they are frequent 
in the Lesser Poland dialect in Vershina, too, e.g. bestau̯kovyj [dumb, fool], cf. Rus. 
бестолковый: ńˈe suxˈajćė tˈak’ėgo bestau ̯ kovˈėgo [Do not listen to this fool]; izˈucać 
[to study
INF
, to learn
INF
], cf. Rus. изучать: grˈupa pͻjexˈaa v:akˈac
y
ji jˈiῖ̯zyk | izˈucać 
pˈolsk’i jˈiῖ̯zyk [the group went for holidays to study Polish]; dvojńˈašk’i [twins
PL
], cf. 
Rus. двойняшки: ʒ́ˈou̯xy u mńˈe dvojńˈašk’i [My daughters are twins]. The borrowed 
items have been adapted to Polish inflectional system, and izˈucać also reflects the 
phenomenon of mazuration (č > c).
A type of interference characteristic at the syntax level are PAT-borrowings including 
reproduction of syntactic structures (syntactic calques), e.g. constructions expressing 
time relations, such as location of events in hourly time: v + X goʒ́iny/goʒ́in
NOM
 following 
the Russian scheme в Х часа/часов [at X o’clock], cf. Pol. o X godzinie: f śtˈyry goʒ
́ ˈiny 
śˈe fstavˈawͻ | i do v’ėcˈora do pˈuźna śˈe robˈiwͻˈ [We were getting up at 4 o’clock and 
worked until the evening]. Another example is the expression of an event taking place 
before a specified time: do + X
GEN
 [before X], cf. Pol. przed X
INS
: ješčˈo do vojny
GEN
 | 

Slavistična revija, letnik 71/2023, št. 2, april–junij 168
a ˈadek ve vˈojne ˈumaru̯ [It was before the war, and the grandfather died during the 
war]. Because PAT-borrowings are not foreign elements, but a reproduction of foreign 
models using the resources of the recipient language, they are more difficult to notice, 
and they are often treated as native by speakers of a given language. 
At the syntax level, there are also switches described in paragraph 2.1, but the example 
of congruent lexicalization in section (c) shows that the distinction of insertions and 
alternations, as well as manifestations of phonetic, morphological and lexical, as well 
as syntactic interference is possible in short excerpts – single phrases and sentences. In 
longer utterances, they overlap and mix, causing, for example, that the alternation is made 
not to pure L
2
 code, but to L
2
 containing PAT- and MAT-borrowings and insertions. 
Therefore, we decided that in addition to the traditionally mentioned analytical levels 
of language: phonetics, morphology, lexis and syntax, the analysis of code-switching 
and mixing should also take into account the most complex level – text. Such approach 
is in line with Francois Grosjean’s  postulate of a “wholistic” view on bilingualism, 
according to which there are no isolated instances of juxtaposition of codes, because 
“the bilingual is an integrated whole which cannot easily be decomposed into two 
separate parts”, and the texts he or she produces are a “unique and specific linguistic 
configuration” (Grosjean 1992: 54-5). Having made this addition, we can summarise 
the phenomena occurring at successive analytical levels of the language in terms of 
both borrowings and code-switching/mixing in the table below.
Table 1: Forms of interlingual influence on different language levels.
Language level
Code-switching and 
code-mixing
Borrowing processes
phonetics and phonology – 
phonemes
n/a phonetic interference
morphology – morphemes n/a
morphological interference, 
language hybrids
lexis – words
insertions
PAT-borrowings (word-for-
mation and morphological 
calques), MAT-borrowings 
(„classical” borrowings)
syntax – phrases and sen-
tences 
alternation, insertions
PAT-borrowings (syntactic 
calques)
text all listed above all listed above

169 Michał Głuszkowski: Code-Switching and Code-Mixing on Different Language Levels
Since only at the textual level are individual phenomena visible in a broader context 
showing their mutual connections and dependencies, we will adopt this perspective 
when discussing examples of this phenomenon in the Vershinians’ speech.
3 Discussion
Treating not only the processes of code-switching and mixing, but the entire 
Vershinian bilingualism as texts, rather than separate sentences, allows us to illustrate 
the most complete approach to this phenomenon. The first excerpt is an example of a 
text maintained relatively consistently in the Lesser Poland dialect. The informant – 
an elderly woman, is answering of the researchers, and is aware of the lexical deficit 
in the heritage language, but nevertheless she is able to control the main language of 
the utterance:
zadˈėjće m’i jˈak’ė te vaprˈosy | po rosˈyjsku [śmiech] vaprˈosy | bo jo | bˈabuške
MAT
 po pˈolsku 
zvˈal’i
MAT
 [first name] a po rˈusku [first name] bˈėńʒ ́ ė dva lˈata | jˈagžė śe to nazˈyvo | śˈerp’iń 
| vžˈeśiń | bˈėʒ ́ ė u̯ośėmnastˈėgo vžˈeśńa dva lˈata jag’ juž ńė v’ˈiʒe ńˈic | fcale [Please ask me 
some questions. In Russian (laughter) – questions. Because I… My grandmother’s
MAT
 na-
me
MAT
 in Polish was (first name) and in Russian (first name). It will have been… How was 
it? August? September? By September 18, it will have been two years since I do not see 
antyhing. At all].
- Wcale? [At all?]
i gu̯ˈuxom | na jˈed… na to ˈuxo jo juš to no dvajˈeśća 
u̯
ˈośėm lot ńė su̯ˈyše | a na to žė tak trˈoške 
su̯ˈyšė | dˈobžė vrˈoʒ ́ ė žė rozmˈov’o alˈe co | ńic ńė ńė pojmˈujė [And I am deaf. Out of one… 
I have not been able to hear out of one ear for 28 years now. And with this other ear I can 
only hear a little: someone seems to be
PRT
 talking something, but what? I do not understand 
anything].  
- A do doktora pani jeździła? [And did you go to the doctor?]
no mńe bˈyu̯o v: ˈośėmʒ ́ ˈeśunt tˈyśuũ̯c ʒ ́ ev’ˈińcėt ośˈėmʒ ́ eśˈuũ̯tym rˈoku d’v’e opėrˈacjė na gu̯ˈove 
| 
u̯
ot | i m’e zˈaro sparal’ižovˈau̯o na strˈone | jak m’i ˈino zėrvˈal’i tėn nu | po rosˈyjsku trˈojńič-
nyj ńerf nap’isˈal’i | nu pojmujˈećė mńe [Well, in 1988 I had two head operations. Yes. And it 
immediately paralysed my side. As soon as they ruptured this… in Russian it is the trigemi-
nal nerve. As they wrote. So do you understand me?]
There are only two unambiguous cases of a MAT-borrowing: bˈabuška [grand-
mother], which has been adapted to Polish declension (cf. Pol. bˈabuške
ACC
 and Rus. 
бабушку
ACC
), and zvˈal’i [they called
PST.3PL
] (cf. Pol. nazˈyvać  and Rus. звать). The 
other instances of possible MAT- and PAT-borrowings or code switching and mixing 
has been underlined, because their interpretation is not obvious:
- The form of vaprˈosy [questions] in NOM.PL would be the same both in the case 
of Russian original and its Polish adaptation. Thus, on the basis of the given excerpt 
is impossible to answer this question and the reference to the entire corpora is needed 
to check the frequency and other forms. There are only 5 instances of the use of this 
word and all of them in NOM.SG and NOM.PL, which do not show any differences 
in inflection. However, the Polish equivalent pytanie is much more frequent and 

Slavistična revija, letnik 71/2023, št. 2, april–junij 170
productive. The words built with a help of the stem -pyt-, nouns and verbs, e.g. pytać, 
spytać, zapytać [to ask
IMP/PERF
] appear 23 times, which suggests that vaprˈos has not 
been fully established in the Vershinian dialect yet and should be considered either an 
insertions or a nonce borrowing (cf. Halmari 1997: 17).
- The first part of the last underlined phrase trˈojńičnyj ńerf [trigeminal nerve] could 
be treated as a prototypical example of insertional code-mixing: due to lexical deficit 
an item from L
2
 is embedded in the L
1
 statement. However, the speaker first hesitated 
looking for the right word, and then announced the change of language, therefore it 
was conscious, as in the case of alternation. The decisive factor in such cases is the 
continuation of the utterance: if there is a return to the original language of the utterance 
(L
1
), it is an insertion, and if the utterance is continued in L
2
, one should classify it as 
an alternational switch. The continuation of this statement, however, is not unambig-
uous, because the form nap’isˈal’i [they wrote
PST.3PL
] is identical in L
1
 and L
2
, both at 
the morphological and surface-articulation level (cf. Теш, Хенчель 2009: 210). Since 
the rest of the utterance is in Polish, nap’isˈal’i is a congruently lexicalized element in 
the common syntactic structure, and may fulfill the function a switch point between L
1
 
and L
2
 in this fragment of utterance: zėrvˈal’i tėn nu | po rosˈyjsku
POL
 trˈojńičnyj ńerf
RUS
 
nap’isˈal’i
RUS-POL
 | nu pojmujˈećė mńe
POL
 [they ruptured this… in Russian it is
POL
 the 
trigeminal nerve
RUS
. As they wrote
RUS-POL
. So do you understand me?
POL
].
The analysed example shows that although interpretations of text excerpts may lead 
to unambiguous classifications of individual words, word combinations and phrases, 
in a broader context their ambiguity is revealed. The interpretations in the next frag-
ment are also ambiguous, where, however, the frequency of clear L
1
–L
2
 switches and 
Russian influences is higher.
jak jˈuš stˈal’in ˈumar to to | i fs’ˈo pˈošu̯o
PAT
 | d’emokrˈat’ijaRUS drˈugo stˈau̯a
PAT
 | a tˈak 
demokrˈac’ji ńe bˈyu̯o | f kou̯xˈoźe | f proizvˈoctfax | na fˈabr’ikax | rukovod’ˈit’el’i bˈyl’i
RUS 
| ˈuny m’ˈau̯y t’ˈekst | tˈyle pov’iń:ˈiśće zarˈob’ać | tˈyleśće dˈou̯žńi
MAT
 zapu̯ˈaćić lˈuʒ́um | tˈyle 
dˈou̯žno
MAT
 iś gosudˈarstfu
MAT
 | ukˈazane
MAT
 | a tˈeras ńiktˈo
RUS
 ńikˈomu ńe ukˈazuje
MAT
 | kˈaždy 
targˈuje | jak pˈošu̯a ta | sfobˈodno targˈovl’a
MAT
 | to | fs’ˈo pošu̯ˈo po inˈakšymu
PAT
 | no co 
zrˈob’iće | no | naćˈalstfo jˈest naćˈalstfo
RUS
 | v’ˈiʒ́i na očax
PAT
 že k’ˈepsko rˈob’i | jˈego sprˈava 
| i k’ˈepsko | jˈesl’i s šerˈegu śe odˈezv’eš | to ty jˈuš vrˈak narˈoda
RUS
 | pšysfˈajival’i tˈak
[Once Stalin died, everything went on
PAT
. Another democracy
RUS
 began
PAT
. And before that, 
there was no democracy in kolkhoz, in production, in factories
RUS
. There were directors. 
And they had an instruction: you are supposed to earn this much, this much you have to
MAT 
pay the people, this much you have to
MAT
 pay to the state
MAT
. (It was) specified
MAT
. And 
now no one
RUS
 is giving orders
MAT
 to anyone. Everyone is trading. When the free market 
economy
MAT
 began, everything went on differently
PAT
. What can you do? The directors are 
the directors
RUS
. They see with their own eyes
PAT
, that someone is working badly. It is his 
business. And it is not good when you make any comment – you are already the enemy of 
the people
RUS
. (Such a man) was described this way].
There are several instances of possible insertions, but only two of them are not com-
bined with another type of code-mixing: ńiktˈo [no one] cf. Rus. никто and vrˈak narˈoda 

171 Michał Głuszkowski: Code-Switching and Code-Mixing on Different Language Levels
[the enemy of the people]. There is also d’emokrˈat’ija
NOM.SG.RUS 
and demokrˈac’ji
5
GEN.SG.POL
, 
but the interpretation is not clear, because a) first instance is Russian and the second one 
is Polish, b) they appear in a longer passage of PAT-borrowings (fs’ˈo pˈošu̯o [everything 
went on] cf. Rus. всё пошло and drˈugo stˈau̯a [became different] cf. Rus. другая стала) 
combined with insertion and is switched to a fragment of utterance that can be interpreted 
both as Russian and as MAT-borrowings (f kou̯xˈoźe | f proizvˈoctfax [in kolkhoz, in produc-
tion]). The part i fs’ˈo pˈošu̯o
PAT
 | d’emokrˈat’ija
RUS
 drˈugo stˈau̯a
PAT
 | a tˈak demokrˈac’ji ńe 
bˈyu̯o | f kou̯xˈoźe | f proizvˈoctfax | na fˈabr’ikax | rukovod’ˈit’el’i bˈyl’i
RUS
 is predominantly 
Russian, but because the presence of elements integrated to Lesser Poland dialect and 
shared lexis
6
 determines the excerpt in whole as congruent lexicalisation. The speaker’s 
intention was to speak Polish, but he was not aware of the switch points. 
The excerpt naćˈalstfo jˈest naćˈalstfo
RUS
 [The directors are the directors] is a separate 
sentence, which could be treated as a full switch from L
1
 to L
2
, i.e. alternation, but as 
a fixed phrasematic structure with limited possibilities of the changes in the scheme 
X есть X [x is x], has a lot in common with insertions. Muysken pointed out that the 
types of code mixing distinguished by him are not independent states, but that real 
phenomena are located on continuums between the three ideal types. The presented 
example shows just such an intermediate state between insertion and alternation.
There are frequent MAT-borrowings. Those related to the domain of economics or 
politics can be characterised as cultural borrowings, i.e. names for objects and process-
es acquired in the course of socio-cultural contacts, for which the recipient language 
had no names before (cf. Sayahi 2014: 89; cf. Weinreich 1963: 53-4): gosudˈarstfo 
[state] Rus. государство, sfobˈodno targˈovl’a [free trade, free market economy] Rus. 
свободная торговля. However, the others are replacing words that already existed 
in the recipient language (Myers-Scotton 2005: 215): dˈou̯žno, dˈou̯žńi [should] Rus. 
должно, должны, ukˈazane [specified] Rus. указано. 
The prepositional phrases of phrasematic character are clear examples of PAT-
borrowings: po inˈʼakšymu [in a different way, differently] reproduces Rus. по-другому, 
and na očax [with someone’s own eyes] – Rus. на глазах. The last highlighted item 
– pšysfˈajivać was used under the influence of Rus. присваивать [assign
INF
] and may 
be treated both as a PAT-borrowing (word-formation semi-calque in which the prefix 
is replaced with its Polish equivalent and the stem is borrowed), and as an adopted 
borrowing (MAT-borrowing), in which the replacement of prefix is a form of morpho-
nological adaptation (cf. Grek-Pabisowa 1999: 225-6).
Despite the intentions of the speaker, who was an elderly man with a good command 
of both languages (with Russian as the language better known and used more frequently 
5
 Polish demokracja and Russian демократия are an example of words identical only on the morphono-
logical level, but the differences on the phonetic level refer to d : d’ and c’ja : t’ija. Such similarities favour 
mixing (congruent lexicalisation) (cf. Muysken 2000: 1-5).
6
 D’emokrac’ija, kou ̯ xoz, proizvoctfo, rukovod’it’el are of Russian origin, but appear in the corpus as 
adopted borrowings.

Slavistična revija, letnik 71/2023, št. 2, april–junij 172
in recent years), it was not possible to produce the entire utterance in the Lesser Poland 
dialect. In his efforts, our interlocutor focuses on a version of the heritage language 
that he considers native, i.e. containing numerous lexical and syntactic influences of 
the Russian language (cf. Mitrenga-Ulitina 2015: 130, 146-7). 
4 Conclusion
In the fragments discussed, the transition between different types of code switching 
and mixing as well as forms of borrowing often takes place depending on the adopted 
perspective: partial or “wholistic”. The analysis of the Vershinians’ bilingual statements 
on the text level shows that many phenomena, as in the case of congruent lexicalization 
in dysfluent speech defined by John Lipski, result from: incomplete fluency in one of 
the languages coupled with the intention to maintain the utterance in this language, as 
well as from the lack of social consequences for involuntary mixing (Lipski 2009: 33). 
Attempts to complete or fix the code by our informants are too visible to talk about a 
homogeneous L
1
 language that has so many features of a mixed code (or a fused lect, 
cf. Auer 1999: 309-10). Among the types of code-mixing, it is congruent lexicalization, 
and especially its variant referring to dysfluent speech, that is the most adequate to de-
scribe the processes taking place as part of the contact of cognate languages. However, 
it is only in the analysis at the text level that all dimensions of the interrelationships of 
various types of juxtaposition of the codes and borrowing processes are fully visible.
RefeRences Natalia An Aniew A, 2013: Teksty Polaków pisane grażdanką w syberyjskiej wsi Wierszyna. 
Acta Baltico-Slavica 37. 287-98.
Peter AueR , 1999: From code-switching via language mixing to fused lects. Toward a 
dynamic typology of bilingual speech. The International Journal of Bilingualism 
3. 309-32.
Ludwik BAzylow , 1975: Syberia. Warszawa: Wiedza Powszechna. 
Barbara Bullock , Almeida ToRiBio , 2009: Themes in the study of code switching. The 
Cambridge handbook of linguistic code-switching. Eds. Barbara Bullock, Almeida 
Toribio. Cambridge: Cambridge University Press. 1-17.
Margaret Deuch AR, 2020: Code-Switching in Linguistics: A Position Paper. Languages 
5/22. 1-19. Also online.
Thomas e Riksen , 1993: In which sense do cultural islands exist? Social Anthropology 
1/1B. 133-47. Also online.
Joseph Gafaranga, 2012: Language alternation and conversational repair in bilingual 
conversation. International Journal of Bilingualism 16/4. 501-27. Also online.
Michał Głuszkowski . 2012. Bilingwizm z dyglosją w sytuacji wyspowej. Polska wieś 
Wierszyna na Syberii. Z polskich studiów slawistycznych. Seria 12. Językoznawstwo. 
Prace na XV Międzynarodowy Kongres Slawistów w Mińsku 2013. Ed. Małgorzata 
Korytkowska et al. Warszawa: Polska Akademia Nauk. 45-51.

173 Michał Głuszkowski: Code-Switching and Code-Mixing on Different Language Levels
Michał G ł u s z k o w s k i, 2015: Creolization and balkanization as a result of language 
(dialect) contact. Is the origin of mixed languages universal? Slavia Meridionalis 
15. 53-66. Also online.
Iryda GRek -PABisow A, 1999 (1983): Typy zapożyczeń i sposoby przyswajania wyrazów 
polskich. Staroobrzędowcy. Szkice z historii, języka, obyczajów. Warszawa: SOW. 
211-27.
François GRosjeAn , 1992: Another view on bilingualism. Advances in Psychology 83. 
51-62. Also online.
John Gum Pe Rz , 1977: The Sociolinguistic Significance of Conversational Code-Switching. 
RELC Journal 8/2. 1-34.
William h AAs , 1960: Linguistic Structures. Word 16/2. 251-76.
Helena h Alm ARi , 1997: Government and codeswitching: Explaining American Finnish. 
Amsterdam, Philadelphia: John Benjamins Publishing Company.
John l iPski , 2009: “Fluent dysfluency” as Congruent Lexicalization: A Special Case 
of Radical Code-Mixing. Journal of Language Contact 2/2. 1-39. Online.
Heinrich l öffle R, 1987: Sprache und Gesellschaft in der Geschichte der vorstrukturali-
stischen Sprachwissenschaft / Language and Society in the History of Prestructuralist 
Linguistics). Sociolinguistics. An International Handbook of the Science of Language 
and Society. Soziolinguistik. Ein internationales Handbuch zur Wissenschaft von 
Sprache und Gesellschaft. Vol. 1. Eds. Ulrich Ammon, Norbert Dittmar, Klaus 
Mattheier. Berlin, New York: Walter de Gruyter. 379-89.
Yaron m ATRAs , Jeanette s Akel , 2007: Investigating the mechanisms of pattern repli-
cation in language convergence. Studies in Language 31/4. 829-65. Also online.
Swietłana m i TRen GA-u li Tin A, 2015: Język polski mieszkańców wsi Wierszyna na Syberii. 
Lublin: Wydawnictwo Uniwersytetu Marii Curie-Skłodowskiej.
Peter m uysken , 2000: Bilingual speech. A typology of code mixing. Cambridge: 
Cambridge University Press.
Carol m yeRs -s coTTon , 2004: Precision Tuning of the Matrix Language Frame (MLF) 
Model of Codeswitching. Sociolinguistica 18. 106-17.
Carol m ye Rs -s co TTon , 2005: Multiple voices. An introduction to bilingualism. Malden: 
Blackwell.
Ewa n owick A, 2011: Wierszyna, czyli z morza na ocean. Paper at the conference 
“Słowiańskie wyspy językowe i kulturowe”. Nicolaus Copernicus Univesity in 
Toruń, September 16-17th 2011.
Shana Po Pl Ack , 1988: Contrasting patterns of code-switching in two communities. 
Codeswitching: anthropological and sociolinguistic perspectives. Ed. Monica 
Heller. Berlin-New York, Amsterdam: Mouton de Gruyter. 215-44.
Shana Po Pl Ack , David s Ankoff , 1984: Borrowing: the synchrony of integration. 
Linguistics 22. 99-135.
Paul Rosen Be RG, 2005: Dialect convergence in the German language islands (Sprachinseln). 
Dialect change: convergence and divergence in European languages. Peter Auer, 
Frans Hinskens, Paul Kerswill. Cambridge: Cambridge University Press. 221-35.
Lofti s A y Ahi , 2014: Diglossia and Language Contact. Language Variation and Change 
in North Africa. New York: Cambridge University Press.

Slavistična revija, letnik 71/2023, št. 2, april–junij 174
Irena s A wick A, 2020: Merger of Alveolar and Dental Fricatives and Affricates (Mazurism). 
Encyclopedia of Slavic Languages and Linguistics Online. Ed. Marc Greenberg. 
Leiden: Brill. Online. 
Sarah ThomAson , 2001: Language contact. Washington: Georgetown Univerity Press.
Sarah Thom Ason , Terrence k Aufm Ann , 1988: Language Contact, Creolization, and 
Genetic Linguistics. Berkeley-Los Angeles-Oxford: University of California Press.
Uriel w ein Reich , 1963: Languages in contact. Findings and problems. The Hague: 
Mouton.
Александр Дуличенко, 1998: Языки малых этнических групп: статус, развитие, 
проблемы выживания. Языки малые и большие... In memoriam acad. Nikita 
L. Tolstoi. Ed. Александр Дуличенко. Slavica Tartuensia IV. Tartu: Universitas 
Tartuensis. 26-36.
[Aleksandr Duličenko, 1998: Jazyki malyh etničeskih grupp: status, razvitie, pro -
blemy vyživanija. Jazyki malye i bolʼšie... In memoriam acad. Nikita L. Tolstoi. Ed. 
Aleksandr Duličenko. Slavica Tartuensia IV . Tartu: Universitas Tartuensis. 26-36].
Святлана Т еш, Герд Х енчель, 2009: Переключение кодов в трасянке (некоторые 
количественные наблюдения). Славянские языки. Аспекты исследования. Eds. 
Наталья Ивашина, Елена Руденко. Минск: БГУ. 209-15.
[Svjatlana Te š , Gerd h en čel ʼ, 2009: Pereključenie kodov v trasjanke (nekotorye koli-
čestvennye nabljudenija). Slavjanskie jazyki. Aspekty issledovanija. Eds. Natal’ja 
Ivašina, Elena Rudenko. Minsk: BGU. 209-15].
PovzeTek Vas Veršina v regiji Irkutsk v Sibiriji so v začetku 20. stoletja ustanovili prostovoljni nase -
ljenci iz južne Poljske. Od začetka je bila dvojezična skupnost, vendar je zaradi družbenih in 
političnih sprememb ruščina postopoma nadomestila malopoljsko narečje v številnih funkcijah. 
Dvojezičnost potomcev poljskih naseljencev poleg diglosije spremlja tudi pojav jezikovnih 
sprememb med govorom. Namen tega članka je obravnavati raznolikost pojavov kodnega 
preklapljanja in mešanja na različnih jezikovnih ravneh. Na področjih fonetike, morfologije in 
leksike se raziskovalci najpogosteje osredotočajo na medjezikovne vplive v obliki interference, 
medtem ko je kodno preklapljanje pojav, ki se pojavlja na ravni skladnje.
Članek temelji na jezikovnem gradivu, zbranem med terenskimi odpravami v Veršini. 
Opredeljuje primere leksikalnih in strukturnih izposojenk ter različne oblike jezikovnega 
preklapljanja med govorom. Te so v literaturi tradicionalno analizirane kot: insercije, alternacije 
in kongruentna leksikalizacija. Vendar pa tak pristop omogoča opisovanje le izbranih delov 
izreka, medtem ko v živem sporazumevanju delujejo celotna besedila. Zato je bilo predlagano, 
da se obravnava najkompleksnejša raven jezika, tj. besedilo kot tako. Če se analizira celotno 
besedilo, se v njem prepletajo različne vrste kodnega preklapljanja in interference, in tisto, kar 
bi bilo na ravni posameznega stavka opredeljeno kot izposojanje ali alternacija, se v resnici 
izkaže kot kongruentna leksikalizacija. Izbrani primeri kažejo, kako to dopolnjevanje vpliva 
na dojemanje pojavov dvojezičnosti.