Use of code-switching in inter-generational and

intra-generational communication in Greek

(Cypriot dialect) and English.

Marilena Onisiforou

May 2017
 Acknowledgments

This work was submitted as a final year Dissertation to the Department of Linguistic and
English Language, University of Manchester, in May 2017. I would like to thank my
dissertation supervisor Professor Yaron Matras for all his guidance and helpful feedback for
which I am so very grateful, without him this dissertation would not have been possible. I
would also like to thank my family and friends for supporting me throughout this process and
to everyone who kindly participated in this study.

2
 Table of Contents

List of tables. ……………………………………………………………………………..…..4

Transcription conventions………………………………………………………………...…5
Abstract………………………………………………………………………………….........6

1. Introduction…………………………………………………………….………………….7
1.1 Aims and motivation for the study………………………………….……..…..…..7
1.2 The community…………………………………………………………................8
2. Literature Review………………………………………………….……………..……....10
2.1 Bilingualism and code-switching……………………………………………...…10
2.1.1 Bilingualism……………………………………………...………...…..10
2.1.2 Code-switching…………………………………...………………...…..11
2.2 Factors influencing language choice………………………………...……….......12
2.2.1 Personal experiences of migration………………………....………......12
2.2.2 Generational patterns………………………………...………..……......14
2.2.3 Language Maintenance………………………………………………....15
2.3 Motivations to code-switch……………………………….……………..….........17
2.3.1 Linguistic gap…………………………………...………………….......17
2.3.2 Accommodation theory…………………………………………….......18
2.4 Conversation analysis…………………………………..………………...……....19
2.4.1 Sequential analysis………………………………………………..........19
3. Methodology………………………………..………………………………………...…..21
3.1 Research objective and hypothesis………………...…………………………......21
3.2 Participants…………………………………………………………………….…22
3.2.1 Participant selection ……………………………..…………...……......22
3.2.2 Participant profiles………………………………….…………...……..23
3.3 Audio-recorded conversations…………………………..……………...………...23
3.4 Transcribing recordings……………………………………………………...…...25
3.5 Interpreting the data……………………………………..……………………......26
4. Results……………………………………….……………………...…………...………..27
4.1 Conversational patterns……………………………………...…………...….…...27
4.2 Accommodation purposes………………………………………....………......…29
4.2.1 Reiteration…………………………………………..……...………......31
4.3 Topic……………………….……………………………………………………..34
4.3.1 Quoting from memories………………………………………...….......34
4.3.2 Topics in local public discourse………………………………...….......39
4.4 Linguistic gap………………..…………………..…………………..………...…41
5. Discussion………………………………..……………………………………………......44
6. Conclusion……………………….……………………………………………………......52

References………………………………………………………………….…………..……55

3
 List of tables

Table 1: Participant profiles

Table 2: Speakers and their conversational patterns

4
 Transcription Conventions

Symbol Meaning
[**] Name of person or specific location
/ Interruption or repair
[laughter] Laughter

5
 Abstract

This study examines the motivations to code-switch in inter-generational and intra-

generational communication between Greek-Cypriots in a North London community. This

study also explores whether the experiences of migration and previously cited generational

language patterns might influence the language choices of these individuals. Three

conversations were recorded where bilinguals discussed their experiences of migration and

living as British Cypriots. A qualitative approach was employed for the analysis of this data

and the conversations were analysed in their entirety, and it was found that the personal

histories of migration and generational patterns were not deterministic of each individual’s

language choices and future studies should consider in greater detail the notion of ‘super-

diversity’. Language maintenance and ‘transnationalism’ factors such as Greek

supplementary schools and frequent contact with Cyprus were found to potentially account

for the continued use of the Greek language across generations.. A conversational and

sequential analysis was undertaken for analysing and interpreting the motivations as well as

the meanings created by code-switching. This analysis finds three main motivations to code-

switch and these were for accommodating purposes, topics of discussion and potentially

filling a lexical gap.

6
1. Introduction

1.1 Aims and motivations for the study

The language-contact phenomenon of code-switching is one of the most common practices

amongst bilinguals. Code-switching can be defined as “the alternation of languages within a

conversation” (Matras 2009:101). I decided to explore this topic of code-switching within the

Greek-Cypriot community of North London because I am a British born Greek-Cypriot and a

member of this community. Growing up bilingual, code-switching has heavily featured in my

everyday language use as well as that of the people around me.

My initial focus for the study was to explore the motivations of bilinguals from the same and

different generations to code-switch in conversation. Christodoulou-Pipis (1991) also

conducted her research on a Greek-Cypriot migrant community based in London. Her

research focused on language maintenance and general features of bilingual speech, including

code-switching and linguistic interference. On delving a bit deeper in my research, I began to

wonder what actually influenced speaker’s language choices in conversations. On reading

Schmid’s (2002) study of German Jews that migrated to Anglophone countries I became

most intrigued by the links she made between speakers’ experiences of migration (such as

trauma, age of migration and language history) and their subsequent linguistic attitudes.

Drawing inspiration from both the above studies, I will aim to investigate, whether a possible

link can be identified between personal biographies and language choices and what possibly

motivates the speakers to code-switch. Further studies involved in the analysis suggest

generalised generational patterns of language shift for various generations of speakers

(Garcia & Diaz, 1992) that may influence the language choices of individuals. The main

literature on language maintenance and code-switching does not often combine these topics

with questions of personal histories and language choices, especially regarding this particular

7
community. Thus, I will attempt to fill this gap by finding connections between these various

topic areas.

I will be taking a qualitative approach for this study. I will be recording and analysing

conversation-based data from members of this particular community and I will be examining

these conversations holistically to identify conversational patterns and language choices, as

well as evaluating the content of their discussions on their personal histories. I will also be

undertaking a conversational and sequential analysis when interpreting the motivations and

meaningfulness of code-switching.

1.2 The community

Currently Greek is considered to be one of the largest linguistic minorities in London and

there are 31,306 individuals residing there who consider Greek to be their main language,

comprising 1.8% of the population living in London (Census, 2011). Many of the first

generation speakers in this community today came to England in two migration waves. The

first wave being around the late 1950s and early 1960s, where migrants came seeking better

economic stability and work opportunities. The second wave being post 1974 after the

Turkish invasion of the island of Cyprus where migrants were forced to leave their homes

and emigrated unwillingly in large quantities as refugees.

According to the Cyprus Educational Mission UK (2017), there are currently over 70 Greek

schools across the United Kingdom, 30 of which are in the Greater London area. Most

Greek-Cypriot children growing up in London attend one of these Greek supplementary

schools usually at least twice a week, in the evenings and at weekends. At these schools,

children learn the Standard Modern Greek language, as well as the culture, history and

8
geography of Greece and Cyprus and they are also able to take formal examinations such as

GCSEs and A-levels in the Modern Greek language. Christodoulou-Pipis (1991) describes

these supplementary schools as forming part of a conscious effort within the Greek-Cypriot

community in London to reduce the effects of a generational gap on their language and

culture.

Christodoulou-Pipis (1991) also explains that within this Greek-Cypriot migrant community

three different languages are used: English, Greek and the Greek-Cypriot dialect. Standard

Modern Greek is the official language that is used as the written standard and in official and

formal situations as in publishing and education (including Greek supplementary schools).

The Greek-Cypriot dialect is used in everyday language and more informal conversations.

The Greek-Cypriot dialect mainly differs from Standard Modern Greek phonologically and

lexically, and to a lesser extent morphologically and syntactically (Papapavlou & Pavlou,

2001). It should be noted that throughout this paper I refer to the language used by members

of this community as Greek and not Greek-Cypriot, despite dialectal features being shown.

9
2. Literature review

In this chapter I explore the relevant definitions of my study and review previous studies that

form the basis of my research.

2.1 Bilingualism and code-switching

2.1.1 Bilingualism

The concept of bilingualism is rather abstract, making it difficult to agree on an exact

definition. There have been various definitions proposed in the literature, each differing on

the importance of proficiency. Li Wei (2000:16) simply describes a bilingual as an individual

that “can function in both languages in conversational interaction”. Clyne (1967:20) is more

specific in his study of post-war migrants from German speaking countries and regards a

bilingual as “a person who understands and speaks two languages, both of which he has

acquired by natural experience rather than by school learning”. It is often cited in the

literature that most bilinguals have a ‘dominant’ and a ‘secondary’ language (Clyne, 1967),

where a dominant language is a language used with greater proficiency and greater

frequency. These types of bilinguals are often referred to as ‘unbalanced’ bilinguals, while

those that are equally competent in using both languages are known as ‘balanced’ bilinguals.

In many studies of bilingual speech communities resulting from circumstances of migration,

there is often an important distinction made between bilinguals. There are those that are born

monolingual and due to circumstances of migration become bilingual, there are children that

are born into these bilingual situations and there are those that are “second language learners”

(Clyne 1967:124). This research paper analyses the language choices and manifestations of

code-switching mainly from the former two of these types, and will be referred to as first

generation speakers and second generation speakers of Greek.

10
It is a widely held view within language contact studies, that the bilingual does not maintain

two separate language systems but one complex communicative repertoire (Matras, 2009).

This complex repertoire consists of linguistic structures that are at the bilingual’s disposal at

all times. A bilingual must select those forms that are context-appropriate (Matras, 2009:4).

In a bilingual context, Grosjean (2001) claims that bilinguals operate in the ‘bilingual mode’.

This differs from a ‘monolingual speech mode’ where bilinguals deactivate one language,

although not fully, and activate only the language of conversation. The ‘bilingual mode’

involves employing a ‘base language’ and calling upon the other language at various times in

the conversation, often being manifested as code-switching.

2.1.2 Code-switching

Traditionally, code-switching, as well as bilingualism, was heavily stigmatised and was

viewed as ‘language corruption’ (Matras, 2009:101). The use of code-switching was believed

to indicate insufficient knowledge in one of the two languages used. However, more recent

studies have supported the antithesis. Further research has described its use as ‘rule governed’

and as requiring a high level of competence in both languages (Poplack, 1980;

Christodoulou-Pipis, 1991) and that code-switching is an additional conversational resource

that bilinguals possess (Gumperz, 1982).

Poplack (1980) also states that there are various types of code-switches that can manifest in

the language use of bilinguals. In her study of bilinguals in a Puerto Rican community she

proposes a distinction between tag-switching, inter-sentential switching and intra-sentential

switching. Tag-switching, which she refers to as ‘emblematic’ switching is the switching of a

word or phrase to an otherwise monolingual sentence. This type of switching is often, but not

always, found to be used by ‘unbalanced’ bilinguals that are less competent in one of the

languages. Intra-sentential code-switching is the alternation of languages within a sentence or

11
clause. The third type she describes is inter-sentential switching and this is the alternation of

languages between sentences. This type of code-switching could occur between two different

speakers in a conversation. For example, if one person uses language A and the other uses

and responds in language B in the same conversational interaction.

Recent literature within this field has claimed that these various code-switches display a

meaningful and “skilled manipulation of overlapping sections of two (or more) grammars”

(Li Wei 2000:17). Interpreting the alternation of two languages ultimately shows that

“language mixing is multilayered and can serve various purposes even in the same

conversation”(Matras, 2009:101).

2.2 Factors influencing language choice

2.2.1 Personal experiences of migration

Monika Schmid’s (2002) study of German Jews that migrated to Anglophone countries

explores first language attrition and maintenance in relation to issues of identity. Language

attrition can be defined as “the gradual loss of a language by an individual” (Schmid,

2002:7). By using a corpus of autobiographical interviews given by individuals who left

between the time the Nazis came to power and the start of World War II, Schmid highlights

the significance of a speaker’s attitude in language attrition and language maintenance by

analysing morphosyntactic features of language, as well as ‘errors’ in free spoken discourse.

She explores how the vicious exclusion that these German Jews experienced from their

society and culture evoked a conflicting sense of identity, which ultimately influenced their

attitudes towards the German language. Schmid goes on to suggest that a correlation exists

between an individual’s attitude and how comfortable they felt using the German language.

She argues that the individual’s perception of themselves and their identity, determines the

effects of first language attrition and language maintenance. She therefore claims that

12
language attrition of the individual is influenced by how they wish to be perceived, but she

simultaneously notes in the case of long-term emigrants the “full range of the repertoire is

still there” and that if the speaker so chooses “proficiency can be reactivated” (Schmid

2002:192). Schmid also highlights the importance of attitudes, not just in availing first

generation speakers of their full linguistic repertoires, but also in second language learning.

Christodoulou-Pipis (1991) also highlights how an attitude towards a language and the group

that uses it determines a speaker’s choice. In supporting this claim, she also cites Grosjean

(1982) who explains that migrant children often elect to not speak their mother tongue if it is

associated with a stigmatised minority, in order to assimilate and fit in with the majority

group. Further to this, in Clyne’s (1967) study of German speaking immigrants in Australia

after World War II, he finds younger migrants trying to ‘assimilate’ and adjust to the majority

language community and suggests that this was perhaps due to a desire not to be viewed as

‘different’ by the community as indeed their parents were. Both Schmid (2002) and Clyne’s

(1967) studies explore speaker attitudes towards a minority language during a time in which

these immigrants spoke a language from a country that most of the world was at war with,

and where being identified as a German speaker would have evoked suspicion and mistrust.

In addition to the above assertions regarding attitudes as being determined by personal

experiences of migration, Christodoulou-Pipis (1991) also highlights the age of arrival as an

important factor when analysing language choices. She suggests that those who arrived

before the age of eleven years old experienced less difficulty in acquiring English, and this

could be due to attending an English school from an early age and feeling the need to adjust

to being English so that they were not viewed as outsiders by their new society. On the other

hand, it was reported that those who arrived when they were much older found learning

13
English more challenging and as suggested by Matras (2009:68) this is due to “the

considerable loss of learning flexibility that sets in with puberty”. It could be suggested that

the difficulty in acquiring English may also impact a speaker’s attitude towards this second

language, as well as their confidence communicating comfortably in it.

2.2.2 Generational patterns

In her study of Greek-Cypriot immigrants, Christodoulou-Pipis (1991) describes that

language loss and language shift often go hand in hand. She describes a typical generational

pattern among immigrants, although she argues that different immigrant patterns may exist

due to language maintenance factors. She argues that the first generation speakers born in

their original country are bilingual with their minority language being the dominant of the

two. Second generation speakers are also bilingual with either language potentially being

dominant. Third generation speakers are bilingual with a dominant majority language. Fourth

generation speakers are monolingual, only speaking the majority language.

Another generational pattern promoted by Garcia & Diaz (1992:14) is known as the “three

generation shift”. In their study of immigrant groups in the United States, Garcia & Diaz also

find that the first generation remain speaking their native language whilst it is the second

generation that begins the shift to the majority language. They state that the third generation

usually completes this shift from the minority language to the majority one. As can be seen,

Christodoulou-Pipis (1991) description of the generational pattern differs slightly as it

suggests that there is one more generation before the language shift is complete. This perhaps

suggests that language maintenance exerts an influence on the transmission of a language

across generations.

14
However, it has been suggested that even within the groupings of individuals into generations

of speakers different patterns manifest. Rumbaut (2004) and Portes and Rivas (2011) claim

that language use of children that migrated from a young age, usually aged five or younger

resembled that of children born in the host country. This is due to the fact that these children

were almost entirely socialised in their host country, and did not receive any formal education

in the language of their original country. Clyne (1967:26) further reinforces this notion by

describing children that migrated before the age of eleven as “linguistically more similar” to

the second-generation speakers of the host country. On the other hand, those who arrived

during their adolescent years are regarded as being closer to adult immigrants than second-

generation speakers born in the host country. When considering language choices in my

study, it might be hypothesised that first generation speakers who migrated during early

childhood may show a preference for English over Greek, and those who migrated as

adolescents or older may have a preference for Greek over English as they “lack the plasticity

of young migrants” (Rumbaut 2004:1167). It appears that differences in language choices can

be found intra-generationally and this suggests that it is important not to overgeneralise

language choices regarding the generations in which speakers belong.

2.2.3 Language maintenance

Language maintenance can be described as the continued use of a language in some or all

domains of a speaker, group of speakers or speech community’s life whilst competing with a

majority language (Pauwels, 2004). Papapavlou and Pavlou (2001:93) describe this as the co-

existence of languages “in a fairly stable relationship”. In Schmid’s (2002) study she

discusses attitudes and experiences that influence the language maintenance of the individual,

however, although not incorporated into her study, language maintenance could also occur

15
with a group of individuals and their transmission of a language possibly from one generation

to the next.

Fishman (1991:113) states that “intergenerational transmission of the mother tongue” and

language maintenance facilitate one another. This can be understood by explaining that

maintenance is the continued use of the language, and without maintenance of a language it is

unlikely that the language will be passed on to future generations. Many researchers also take

this view, including Christodoulou-Pipis (1991), that language maintenance, to some degree,

plays a part in attempting to bridge the language gap between generations.

By adapting a list from Conklin and Lourie (1983), Baker (2011) suggests that there are three

main factors that encourage the maintenance of a language amongst immigrants. The first of

these factors involve political, social and demographic factors, which requires many speakers

living closely together, being able to travel to the original country often and identifying with

the ethnic group rather than the majority language community. The second of these factors

refers to cultural aspects of the community, which includes the attendance of ‘mother-tongue

institutions’ such as schools and community organizations. These two factors are supported

by Christodoulou-Pipis (1991) in her study of Greek-Cypriots in Britain. She states that

Greek supplementary schools and frequent contact with Cyprus, whether it be through travel

or writing letters (this study was conducted before Internet or Wi-Fi access) contributes to

maintaining the Greek language within this community. The third factors Baker (2011)

promotes are of a linguistic nature where he emphasises the importance of the minority

language being standardised as well as having international status.

16
2.3 Motivations to code-switch

One of the ways in which language choices manifest in conversation is through the use of

code-switching. Once it has been established that a speaker is comfortable using their full

linguistic repertoire in a bilingual context, the motivations to code-switch can be further

explored.

2.3.1 Linguistic gap

According to Christodoulou-Pipis (1991), the reason for code-switching that bilinguals are

most conscious of is the filling of a ‘linguistic gap’. This could include the switching of a

single lexical item or even a phrase. The need to fill a ‘linguistic gap’ arises from either a

lack of memory or a lack of competence in the language of conversation (Auer, 1995). The

use of code-switching to fill a ‘linguistic gap in the repertoires of bilinguals is often preceded

by silence or hesitation showing uncertainty, or in attempting to recall an item from their

linguistic repertoires they are preceded by a pause (Mahsain, 2015).

It has previously been cited that code-switching manifests in the speech of ‘unbalanced’

bilinguals from a lack of competence in one of the two languages. This type of code-

switching serves what is called a ‘referential function’ (Appel & Musken, 1987). It should

also be made clear however, that code-switching also performs a ‘referential function’ when a

bilingual does know an item in both languages but elects to use the language most

appropriate for a given topic.

In a study of Puerto Ricans in the United States, Pedraza (1978) identified three modes of

communication, English-speaking, Spanish-speaking and code-switching. He stated that

many of the motivations for the code-switches in his study were due to individuals lacking a

17
‘full command’ of either English or Spanish. However, there have been many studies to

counter this claim such as Lance’s (1975) study of Mexicans in the United States. He found

that information such as numbers, which were code-switched in conversations were actually

known in both languages and that code-switching in this sense does not indicate a lack of

competence in one language.

2.3.2 Accommodation theory

Accommodation theory, first manifested as Speech Accommodation Theory (SAT), and later

evolved to Communication Accommodation Theory (CAT) (Giles et al., 1991). It proposes

that speakers alter the way in which they communicate both verbally and non-verbally in an

interaction, depending on the situation, content and participants. Speakers may adjust their

speech to converge or diverge from the speech of their interlocutors (Giles & Smith, 1979).

Convergence is the adaptation of an individual’s speech to resemble that of their

interlocutors, thereby suggesting a positive attitude towards their interlocutor as well as a

shared social group, whereas divergence is the distancing of speech from that of an

interlocutor’s.

Schmid’s (2002) findings, to an extent, support this notion proposed by Giles & Smith (1979)

that a speaker’s use of language might be used to converge to others and indicate how they

wish to be perceived. For example, if one wishes to be a part of a particular speech

community, they may accommodate their language to that used by the community, however

if one wishes to disassociate themselves from a community they will do the opposite and

adapt their language to differ from a member of that community. This appears to be the case

in Poplack’s (1980) study of Puerto Ricans living in New York, where intra-sentential

switching was used to signal in-group membership, whilst non-group membership was

18
indicated by emblematic or tag-switching. A further compelling finding of this study showed

that in one of the networks observed, group members, including those that were English-

dominant or bilingual, generally accommodate their language to that of the older, Spanish-

dominant speakers by speaking in Spanish. This encompasses accommodation theory in an

intergenerational setting and may suggest that in my study a younger generation may adapt

their language to that of an older generation speaker.

2.4 Conversation analysis

Conversation analysis is an approach within the study of social interaction, which was

developed by Harvey Sacks, Emmanuel Schegloff and Gail Jefferson. Conversation analysis

investigates spoken discourse; with the intention of understanding how individuals manage

their interactions (Paltridge, 2012). It is fundamentally used to examine the ways in which

spoken discourse is structured and how it develops in conversation. It ultimately analyses

social relations in spoken discourse (Paltridge, 2012).

2.4.1 Sequential analysis

Within the conversation analysis framework, a sequential analysis is often employed to

explore the reasons and implications of code-switching. This type of approach aims to

determine ‘why’ bilinguals code-switch by analysing ‘how’ they code-switch (Li Wei, 1998).

This approach emphasises that the turn that precedes and follows a code-switched utterance

must be analysed in addition to the code-switch itself. Auer (1995:116) coins the

conversational turn that precedes and follows a code-switch as its ‘sequential environment’.

He explains that the utterance that precedes the code-switch may give an insight as to why the

code-switch occurred, and utterances that follow give an indication as to how the code-switch

has been understood. Therefore, when analysing the underlying motivations to code-switch,

19
the turn-taking of participants must be considered, rather than the individual code-switches in

isolation. It aims to investigate the code-switch in the context of the conversation as “the

same cue may receive a different interpretation on different occasions” (Auer, 1995: 123).

Furthering the notion of ‘sequentiality’ is what Auer (1984) refers to as ‘sequential

implicativeness’. This explains that the language chosen by a participant for their turn “exerts

an influence on subsequent language choices by the same or other speakers” (Auer, 1984:5).

Taking all these factors into consideration that in order to interpret the motivations and

implications of code-switching, these alternations must be analysed in relation to their

‘sequential environment’. In my study, I will therefore aim to interpret the motivations

behind code-switching with regards to the context in which they occur.

20
3. Methodology

This chapter explores the method of data collection I found most appropriate and employed

for this study. In this section, I will also describe the methodological considerations during

this process, and how participants were selected for this study.

3.1 Research objective and hypothesis

The aim of this study is to determine whether there is a link between a speaker’s language

choice and their personal histories (such as age of migration, their generation, experiences of

acculturation). I will also consider how language choices are manifested through the use of

code-switching and by undertaking a sequential analysis, I will examine the utterances that

precede and follow, as well as the code-switch itself in determining what motivates a speaker

to code-switch at these particular points in the discourse.

Formulating a hypothesis based on the literature reviewed in the previous chapter, it seems

that people’s conversational behaviour will correlate with their personal histories, perhaps by

following the generational patterns of language use proposed in the previous chapter. It may

also be found that similarly to Schmid’s (2002) study, speakers who migrated reluctantly, and

did not maintain much contact with Cyprus may have a preference for English as opposed to

their first language, Greek. Further to this, their language choices and code-switches, based

on previous studies may also be influenced by who they are speaking to as well as the topic

and content of their conversation.

21
3.2 Participants

3.2.1 Participant selection

Data was collected from a sample that consisted of individuals that were bilingual in both

Greek (Cypriot dialect) and English, all of which were female in order to limit gender as a

confounding factor. To some extent, the sample selected for this study could be described as

a convenience sample (Podesva and Sharma, 2016). All participants were people I already

knew, and they were happy to oblige by making themselves available to meet up for the

recordings. However, when selecting participants from this community I ensured that I had

previously observed each bilingual individual code-switching in their natural speech.

When dividing the participants into pairs there were two main factors I considered. Firstly,

that the participants knew each other prior to the recordings in order for their conversations to

be more natural and less forced. This would have also allowed participants to speak more

openly about their own experiences of migration as these memories can be considered quite

personal and emotional. Secondly, I ensured that there was at least one pairing of same

generation speakers and different generation speakers.

All individuals that were selected for the study received a Participant Information Sheet (see

Appendix I) in advance of being recorded and were given the opportunity to withdraw from

or continue with the study. All participants were reassured that their recordings would remain

anonymous and each signed a consent form (see Appendix II) stating that they were willing

to take part and that they were willing to share their personal biographies of migration.

22
3.2.2 Participant profiles

In Table 1 below, it can be seen that six speakers were recruited for the purposes of this

study. Four of these were first generation speakers of Greek that migrated at different ages to

England and the remaining two were second and third generation speakers.

Table 1: Participant profiles

Speaker Place of Birth Age of migration Generation

A England - 3rd
B England - 2nd
C Cyprus 14 years old 1st
D Cyprus 23 years old 1st
E Cyprus 5 years old 1st
F Cyprus 5 years old 1st

Some of the first generation participants in this study, namely Speakers E and F, migrated

with their families for economic purposes around the early 1960s. On the other hand,

Speakers C and D migrated as refugees reluctantly after the island of Cyprus was invaded by

Turkey in 1974.

3.3 Audio-recorded conversations

In conducting an initial pilot study, I was able to identify any aspects that needed to be

slightly altered in order to yield reliable data for my analysis. As part of the pilot study I

recorded two conversations, each approximately 45 minutes long using the ‘Voice memos’

application on my IPhone. One conversation was between two second-generation speakers of

Greek and the other between a second and third generation speaker of Greek.

It appeared that although I had previously observed these individuals code-switching, these

same individuals did not feel comfortable communicating in Greek whilst being recorded,

and hence did not code-switch. It seemed that to some degree this could be explained by the

23
‘observer’s paradox’. This paradox is the issue arising from trying to “find out how people

talk when they are not being systematically observed; yet we can only obtain this data by

systematic observation” (Labov,1972:209). Being a member of this particular community and

previously knowing all the participants in this study, I had hoped that any effects or influence

of a potential ‘observer’s paradox’ would be minimal. However, participants were aware of

being recorded and made a conscious effort to speak in English. The visibility of recording

equipment perhaps made participants feel as though they were performing for an external

audience and had to consciously modify their language choice to English.

In order to make sure that this issue was overcome in my final data collection I altered the

way in which the recordings were carried out. Firstly, I began recording before participants

commenced their conversation, whilst I was informally explaining the topic of discussion.

This was so that the recording did not have such an overbearing presence when their

conversation began.

Secondly, in order to further ensure that participants felt more comfortable using their full

linguistic repertoires, I licensed the use of Greek by code-switching myself when expressing

the topics I wished participants would discuss during their conversation. After this initial

instruction, as the researcher and observer I was present during the recordings, but I kept a

mainly passive role during the recording of these conversations. This was so that it could be

ensured that the conversations were natural and less ‘interview-like’, and that each participant

was communicating with only one interlocutor. The data would have been less reliable and

would have been more complex to analyse if more than one interlocutor was present as

participants may have felt the need to accommodate their language for multiple speakers.

24
Participants were asked to talk as naturally as possible and to not be conscious of the

language they would be using whilst being recorded. I explained the study being conducted

was about their or their families’ experiences of moving to this country and that each

participant should exchange stories about what they remember of their childhood growing up

as British Cypriots. Participants were told that their conversation should also cover their

experiences in the present day; such as which aspects of the Cypriot culture they felt was

important to pass on to their children and what their current relationship with Cyprus was

like.

By setting the topic of discussion as their personal biographies, I was able to elicit

information on each participant’s personal histories whilst obtaining data on code-switching.

3.4 Transcribing recordings

Each recording was uploaded onto a laptop as an audio file and labelled with the participants’

code names. The audio files were listened on the ‘ITunes’ application and transcribed as a

document in Microsoft Word. I decided against using automatic programs such as ELAN as

Greek and especially Cypriot dialect features would not have been recognised. I transcribed

most of each conversation, mainly segments in which code-switching occurred which I could

apply a conversational and sequential analysis to.

The excerpts in the following chapter are presented adhering to transliteration conventions in

Roman script. This type of transliteration involved the mapping of each letter from the Greek

alphabet into the Latin alphabet making the transcriptions more accessible to read for non-

Greek speakers. It should also be noted that Cypriot dialect features of Greek have not been

standardised and hence been included. In addition to the above, each transcribed utterance

25
was numbered and presented with a ‘soft’ interlinear gloss beneath it, which indicated word

functions. I decided that it was not necessary to provide a full gloss showing all morphemes

as it would not have added to my analysis, and would have been appropriate for a more

grammatical analysis. Each transcribed excerpt was followed by a coherent translation in

English of the meaning of each utterance.

3.5 Interpreting the data

After transcribing each conversation, I adopted a qualitative approach when analysing the

data and attempted to identify conversational patterns. This involved examining each

conversation as a whole and determining what a speaker’s language preference was based on,

and the consistency and frequency of that language within the conversation. By using the

previous literature on personal biographies, I then tried to see if any of the personal histories

of my participants could provide potential explanations for their language choices.

Further to this, when exploring the data more specifically, I selected excerpts that I could

identify recurring patterns of code-switching. The method I used to examine and interpret the

use of code-switching was a combination of a conversational and sequential analysis. This

then allowed me to categorise the different motivations to code-switch into three main areas;

accommodation purposes, the topic of the conversation and filling a lexical gap.

26
4. Results

This chapter explores the language choices and conversational patterns presented in the

recordings. It also demonstrates a conversational and sequential analysis in attempting to

interpret various excerpts of code-switching.

4.1 Conversational patterns

By analysing each of the three conversations holistically (considering the frequency and

consistency of language choices within the entirety of the recordings), two different

conversational patterns can be distinguished:

Type I: Consistent use of code-switching between English and Greek by both individuals

throughout.

Type II: One speaker mainly speaking English and the other mainly speaking Greek.

The type of conversation and its speakers can be seen in the table below:

Table 2: Speakers and their conversational patterns

Conversation Speakers Pattern

1 A&C Type I
2 B&D Type II
3 E&F Type I

The conversational pattern described in Type I appeared in two of the three recordings

collected. As can be seen in the table above, Conversation 1 is between a third generation

(Speaker A) and first generation speaker (Speaker C) of Greek. In the sections that follow, it

will be seen that in Conversation 1 code-switching manifests due to accommodation

purposes.

27
Type I can also be seen in Conversation 3, where both Speakers E and F are first generation

speakers of Greek who migrated to England at the similar age of five years old. An initial

interpretation would be that both speakers seem to feel comfortable using their full linguistic

repertoire, perhaps this is due to similar backgrounds. Code-switching in this conversation

repeatedly manifests through the retelling of direct and indirect memories.

The pattern described in Type II is present in Conversation 2, between a second-generation

speaker of Greek (Speaker B) and a first generation speaker of Greek (Speaker D) who

migrated to England as a young adult. Speaker D appears to mainly speak Greek but

occasionally code-switches into English. On the other hand Speaker B mainly speaks English,

perhaps due to a lack of confidence in their competence in Greek when communicating with

a first generation speaker of Greek. However, Speaker B occasionally code-switches into

Greek for accommodating purposes but then reiterates what she says in English. The way in

which this particular conversation operates is often known as receptive multilingualism.

Receptive multilingualism “refers to the constellation in which interlocutors use their

respective mother tongue while speaking to each other” (Zeevaert and ten Thije 2007:1).

Receptive multilingualism occurs when the conversation functions with speakers using their

respective first languages that differ. In many studies, receptive multilingualism occurs

between languages that are mutually intelligible. However, to some degree the above

conversational pattern could be considered to be receptively multilingual, as both individuals

are able to speak Greek and English, despite the two not being mutually intelligible.

28
4.2 Accommodation purposes

From the data collected, each conversation demonstrated, to a degree, the accommodating

behaviour patterns of speakers in inter-generational and intra-generational communication.

The following example is taken from Conversation 1, where Speaker C is explaining about

the hardships of her first academic year in England:

Excerpt 1:

Speaker C: 1. it was the first year/

Speaker A: 2. oh and you’re an August baby

3. so eisoun i pio michi/ pio micha mes’tin taxi?

so you-was the most small/ most small in-the class

Speaker C: 4. yes če eimoun i pio micha mes’tin taxi mou eimoun

yes and I-was the most small in-the class mine I-was

če pastoua
and skinny

5. anyway um then I became/

Speaker A: 6. really? [laughter]

Speaker C: 7. yeah polla pastoua pou eimoun

yeah very skinny that I-was

8. but I was studying very very hard

Translation of Excerpt 1:

Speaker C: 1. it was the first year/

Speaker A: 2. oh and you’re an August baby

3.so you were the youngest in the class?

Speaker C: 4. yes I was the youngest in my class and I was skinny

5. anyway um then I became/

Speaker A: 6. really? [laughter]

29
Speaker C: 7. yeah I was very skinny

8. but I was studying very very hard

It can be seen that Speaker C sets the default language of the conversation to English when

telling her story. In utterance 3 however, Speaker A seeks confirmation of information in

Greek from Speaker C, and receives it in Greek in utterance 4. By using exactly the same

phrase as Speaker A, but conjugating it differently, Speaker C accommodates her language to

that of Speaker A, thus displaying lexical accommodation. Mahsain (2015:251) explains that

lexical accommodation is the repeating of a word or phrase used by a previous speaker in the

language in which they uttered it. The repeated phrase often contrasts with the language of

the current utterance. In accommodating her language and repeating the phrase used by

Speaker A, Speaker C is expressing an additional confirmation for the inference made in the

previous utterance. This supports the notion of ‘sequential implicativeness’ (Auer, 1984), as

in the case presented above the language choice of Speaker A’s inference has influenced the

language choice of Speaker C in the following utterance. However, she then switches back to

her preference of English in utterance 5. The use of the discourse marker ‘anyway’ and then

continuing in English signals that Speaker C wishes to move the conversation in a different

direction, to a different subject. Speaker A then accommodates her language in utterance 6 to

English when asking a question and when answering Speaker C employs the same alternating

pattern that both speakers employed in the previous utterances. In utterance 8, similarly to

utterance 6 Speaker C uses the English discourse marker ‘but’ to move the conversation in a

different direction.

Gumperz (1982:66) promotes the notion that code-switching is used for meaningful

juxtapositions between “strings formed from the internal rules of two distinct grammatical

systems.” This supports the argument that code-switching is an extra resource that bilinguals

30
have at their disposal in a bilingual context. In the above example, a meaningful juxtaposition

presented in utterances 2 and 3. Utterance 2 can be interpreted as a presupposition, that both

speakers know is true and this is made apparent when Speaker A says ‘oh and’. On the other

hand, utterance 3 is an inference from the information given in utterance 2 and this is

confirmed by the use of ‘so’, linking the two pieces of information. Therefore, contrasting the

languages of the two utterances serves an internal discourse purpose, where a difference can

be seen between information Speaker A knows for sure and information that she wants

Speaker C to confirm.

In utterances 2 and 3, Speaker A signals an understanding of the context of the story that

Speaker C is describing by already interpreting the setting and making conclusions. When

Speaker A receives confirmation from Speaker C it can be seen that she understood the

implicature of utterance 2 (Grice, 1975). In making this inference Speaker A is essentially

helping Speaker C with her narration of the story, abiding by Grice’s (1975) ‘cooperative

principle’. The inferential procedure is carried out in English as can be seen by the use of ‘oh

and’, ‘so’ and also ‘yes’ in utterance 4, perhaps because the default language of the

conversation has been set to English. Greek on the other hand is only used for the inference in

utterance 3, showing that Greek is used as a sort of window into the content of what is being

said, whilst English is used for managing the interaction. Thus, this intra-sentential switching,

which seemingly appears to not have neat boundaries between the two languages in fact does

have neat boundaries.

4.2.1 Reiteration

The use of code-switching to repeat a word, phrase or expression, could be used for

accommodating purposes as well as for emphasis in the discourse.

31
An example of this type of accommodation can be seen below in an excerpt taken from

Conversation 2:

Excerpt 2:

Speaker B: 1. pote irthes Agglia [**]?

when you-come England [**]?

2. when did you come to England?

Speaker D: 3. eh ego irtha to ogthonta

eh I I-came the eighty

Speaker B: 4. ah ok meta pou to polemo

ah ok after from the war

5. it wasn’t when the war/

Speaker D: 6. ne
yes

Speaker B: 7. ah it was after

Speaker D: 8. o polemos itan prin

the war was before

Translation of Excerpt 2:

Speaker B: 1. when did you come to England [**]?

2. when did you come to England?

Speaker D: 3. um I came in nineteen eighty

Speaker B: 4. ah ok after the war

5. it wasn’t when the war/

Speaker D: 6. yes

Speaker B: 7. ah it was after

32
Speaker D: 8. the war was before then

It is apparent that Speaker D’s default language throughout Excerpt 2 and indeed for the

majority of the conversation is Greek. As this excerpt was taken from the start of the

recording, it seems that Speaker B anticipates Speaker D’s default language, as in utterance 1

Speaker B ‘pre-accommodates’ her language to Greek when asking Speaker D a question. As

there is no previous evidence of Speaker D’s language preference, it can therefore be said that

Speaker B is pre-empting the situation and converges. This type of accommodation is

different from the type found in Excerpt 1 as it has a different evidence base. In Excerpt 1,

Speaker C accommodates her language to that of Speaker A after she speaks Greek.

However, there is no evidence yet of Speaker D’s language preference and so this

accommodation is based on Speaker B’s expectation.

The question is why does Speaker B then repeat the utterance in English after she has already

accommodated her language to that of Speaker D? It may be the case that Speaker B wishes

to reinforce or emphasise what she has asked or perhaps it is because the question itself is

rather unusual. Another explanation might be that Speaker B’s choice of Greek conveys an

unusual effort and the question might seem unclear and therefore needs to be reinforced in

English. This reiteration therefore not only emphasises the question, but is also a way of

hedging and neutralising the discourse in order to make the interlocutor feel at ease. It

ultimately alleviates the discomfort of asking the question in the first place but might bring a

new discomfort by sounding insistent.

In utterance 1 Speaker B seems to be making a gesture, but in utterance 2 she finds her

comfort zone by speaking English. In utterance 4, Speaker B also accommodates her

33
language and a negotiation can be seen between the two languages. Speaker D is certain of

her language and Speaker B is the one accommodating, but once it is established that she can

use both languages she is more comfortable using English in utterance 5. In utterance 5 a

contrast is presented, as Speaker B changes the direction of conversation, which is marked by

a language switch with Speaker B returning to her comfort zone. In utterance 7 Speaker B

can be seen to continue using English, and this choice is not negotiated at all. After the

utterances in the above excerpt, the conversation does indeed continue with Speaker D

speaking Greek and Speaker B preferring to communicate in English with some negotiations

in Greek.

4.3 Topic

4.3.1 Quoting from memories

When a speaker is in the bilingual mode, they have two options when quoting someone from

a memory. They can either quote in the original language of the quote or they can translate it

into the dominant language of the current conversation. Sebba and Wootton (1998) claim that

code-switching “is a frequent correlate of reported speech in conversation”. The most

common motivation to code-switch in all three conversations seems to be when individuals

quote from direct and indirect memories. Quoting from direct memories is the repetition of

the exact words in their original language from a conversation in which the participant is

usually present. Quoting from indirect memories consists of expressing the meaning of what

was said rather than quoting the exact words spoken from a memory that participants were

not present in but told about perhaps by a parent.

The following excerpt is an example taken from Conversation 3 where participants are

describing conversations they had from direct memories:

34
Excerpt 3:

Speaker E: 1. I told my dad once when I got home

2. apofasisa na gino ithopoios [laughter]

I- decided that I-become actress [laughter]

3. you know what he said to me?

4. he said to me nambou na yineis?

He said to me what that you-become

5. ithopios dad
actress dad

6. ithopoios? če na se fila o enas če o allos?

actress? and that you-will kiss the one and the other

lalli mou
he-says me

Speaker F: 7. I told my mum I wanted to be an air hostess and she goes to me

8. mono oi poutanes kamnousin etsi douleiés

only the prostitutes they-do those jobs

Translation of Excerpt 3:

Speaker E: 1. I told my dad once when I got home

2. I have decided to become an actress [laughter]

3. you know what he said to me?

4. he said to me what are you going to become?

5. an actress dad

6. an actress? And you’ll be kissed by every Tom Dick and Harry?

he says to me

35
Speaker F: 7. I told my mum I wanted to be an air hostess and she goes to me

8. only prostitutes do those sort of jobs

From the onset it can be seen that the default language of this excerpt is English. This is made

conspicuous as the first utterance, which ultimately sets the language for the segments that

follow, is in English. It can be seen that throughout the excerpt both speakers adopt the

technique of introducing a quote in English and consistently ‘authenticate’ these original

quotes in Greek (Gumperz, 1977). By adopting this technique of alternating between the two

languages, speakers are able to establish two things; firstly, that the original setting of the

memory was in Greek and secondly, that they’re retelling the story in the present and the

language of the present is English, but quotes are re-enacted in Greek.

This technique is used until utterance 5, by which time it is an established technique and an

introduction is not needed for the quote that follows. From listening to the recording, it seems

that intonation is used at this point to introduce the quote. This displays how inventive

speakers are in spontaneous speech, as the conversation follows a particular pattern which is

used to facilitate the interlocutor’s understanding of what is being said, but once this pattern

has been established, it is no longer required and this does not leave any gaps in the

interlocutor’s understanding.

In utterance 6 however, Speaker E then departs from this pattern, by following a quote in

Greek with a quotative expression in English. This highlights that the two languages have

different functions within this conversation, which causes this alternation and makes this

excerpt of Type I.

36
In the above example, code-switching could be interpreted as being used as an identity

construction. The two participants share common languages and function in the bilingual

mode making use of their full linguistic repertoire (Matras 2009). This portrays a sort of

solidarity element as both speakers can go beyond and manage an interaction in both

languages. This established solidarity is furthered when Speaker F reciprocates by replicating

the pattern used previously by Speaker E. This shows that Speaker F not only understands

what Speaker E is saying, but the interaction is mutual.

Most of the memories described in the data took place during the participants’ childhoods

when the dominant language of the home was Greek (as in Excerpt 3). However, there are

exceptions presented in the only Type II conversation where the first generation speaker of

Greek recounts of a more recent conversation that took place in an English setting. This can

be seen in Excerpt 4 below, where Speaker D is explaining that she does not tolerate the

shortening and Anglicizing of her son’s Greek name and retells of a phone call that she had

with one of her son’s friends.

Excerpt 4:

Speaker D: 1. mia fora tilefona enas mikros

one time it-called one small-boy

2. can I speak to [**]

3. lego no nobody is called [**] in this house

I-say no nobody is called [**] in this house

Speaker B: [laughter]

Speaker D: 4. lei mou can I speak to [**]

he-say me can I speak to [**]

5. lalo no [**] is here

I-say no [**] is here

37
 6. lali mou ok thank you
he-say me ok thank you

Translation of Excerpt 4:

Speaker D: 1. once a young boy phoned

2. can I speak to [**]?

3. I said no nobody is called [**] in this house

Speaker B: [laughter]

Speaker D: 4. he said to me can I speak to [**]?

5. I said no [**] is here

6. he says to me ok thank you

As can be seen from the above excerpt, the same pattern of alternation that was used in

Excerpt 3 is also present in Excerpt 4. However, in this case it can be seen from the start that

Speaker D sets the default language of the interaction to Greek, which is the language for the

introduction of each quote. Whilst Greek is used for the quotative expressions of these

utterances, English as the original language of the conversation she is recalling, is used for

each quote. This alternation pattern can be seen in utterances 3, 4, 5 and 6. However, in

utterance 2, the English quote does not appear to have an explicit introduction like the

utterances that follows. Intonation could be the technique used when quoting her son’s friend,

or utterance 1 which precedes it could be interpreted as an indirect way of introducing the

quote that follows. Alternatively, the use of English and the code-switch itself could present a

pattern of language specialisation, where it is understood that utterance 2 is a quote as it is in

English, and an explicit introduction is not needed.

38
It can therefore be seen in the above examples, that bilinguals are able to establish an

interesting alternating pattern when retelling of a memory and quoting speech.

4.3.2 Topics in local public discourse

A further use for code-switching appears to be when speakers are discussing topics and using

lexical items that would typically appear in local public discourse.

The following excerpt is taken from Conversation 2, which is of Type II and presents two

lexical items that are code-switched.

Excerpt 5:

Speaker D: 1. yiatí niothoume oti eimaste community

because we-feel that we-are community

Speaker B: 2. and we feel that we have to have our/

Speaker D: 3. ne
yes

4. community simeni na eheis ta idia kaina

community means to you-have the same common

charaktiristika
characteristics

5. eh society eine oles oi files

eh society is all the nationalities

6. i kinonia etho einai diaforetiki poliethniki

the society here is different multicultural

Translation of Excerpt 5:

Speaker D: 1. Because we feel that we are a community

Speaker B: 2. and we feel that we have to have our/

39
Speaker D: 3. yes

4. community means that you have the same characteristics

5. And society is all nationalities

6. The society here is different its multicultural

It is apparent from the start of this excerpt that Speaker D’s preferred language is Greek,

whilst Speaker B’s is English. Code-switching manifests when Speaker D manages the

conversation in Greek but says ‘community’ in English in utterances 1, and 4, and when she

says ‘society’ in utterance 5. It could be argued that the fact that ‘community’ is consistently

expressed in English might be an indication of a lexical gap in the speaker’s communicative

repertoire. However, it seems more likely that this is a choice, where making use of the

English form is simply a replication of public discourse. The lexical item ‘community’ is

most probably not a commonly used word for the speaker and most possibly would only be

heard or spoken in an English-speaking environment.

An alternative theory, which is perhaps less probable, is that Speaker D is accommodating to

Speaker B’s preferred language of English by switching this lexical item. This could be

justified by perceiving Speaker D as having a lack of confidence that Speaker B will

understand the word in Greek. By code-switching she effectively ensures that the message

she is trying to get across is understood.

In utterances 4 and 5 it becomes apparent that there is a sort of juxtaposition between the

terms that are code-switched and the rest of utterances 4 and 5. The lexical items

‘community’ and ‘society’ are expressed in English, whilst their descriptions are

communicated in Greek. This pattern of code-switching creates a sort of symmetry between

40
the utterances in 4 and 5. In utterance 5, similarly to ‘community’, the choice of English for

‘society’ reflects its contextualisation in local public discourse.

However, in utterance 6 it can be seen that Speaker D departs from this pattern.

It seems that there is a contrast that is marked between utterance 6 and the utterances that

precede it. The consistent use of Greek in utterance 6 reflects a more conscious effort to make

a coherent statement. Speaker D is specifically talking about the society in London and does

this by expressing both the term ‘society’ and the rest of the utterance in Greek. Speaker D

could potentially have marked this distinction by expressing the entire utterance in English,

however this is not her preferred language and settles on subtly switching ‘society’ to

‘kinonia’. Therefore, the previous utterances can be interpreted as a build up to the statement

in utterance 6, which is the message Speaker D wants to get across.

4.4 Linguistic gap

In the literature, ‘filling a lexical gap’ can be used to describe an occasion where a code-

switch occurs due to a lack of competence in a language or a lack of memory. In the

following excerpt, taken from Conversation 2, a first generation speaker of Greek discusses

her experiences of integrating into her present community. It can be seen that she code-

switches a single lexical item into English and it seems likely that this is due to a difficulty in

recalling the Greek form at the time of the utterance, with the English equivalent being more

readily available in her linguistic repertoire.

Excerpt 6:

Speaker B: 1. esthanese o kosmos oksa I ídia esthanese/

you-feel the people or the yourself you-feel

2. do the people treat you like you’re a foreigner still you think?

41
 3. or is it you yourself because you grew up in Cyprus?

Speaker D: 4. ne nomizo ehoume diafores

yes I-think we-have differences

Speaker B: 5. yeah

Speaker D: 6. pou en kataligoun pouthena vasika

where they-don’t result anywhere basically

niotho oti
I-feel that

7. eimai diaforetiki, ipoxreotika na yino

I-am different, by-force to I-become

adjust [laughter]
adjust [laughter]

Translation of Excerpt 6:

Speaker B: 1. do you feel it’s the people or do you feel its you?

2. do the people treat you like you’re a foreigner still you think?

3. or is it just you yourself because you grew up in Cyprus?

Speaker D: 4. yes I think we have differences

Speaker B: 5. yeah

Speaker D: 6. there is no end to it basically I feel that

7. I am different and forced to adjust

As in previous examples from Conversation 2, Speaker D’s default language is Greek.

Although Speaker B appears to mainly speak English, there seems to be a negotiation

between the two languages in the first three utterances. Similarly to Excerpt 2, it can be seen

42
that Speaker B ‘pre-accommodates’ to Speaker D’s language preference in utterance 1,

followed by a reiteration of the question in English in utterances 2 and 3. This negotiation,

similarly to Excerpt 2 eventually ends with Speaker B returning to her comfort zone and

communicating in English. The rest of the conversation continues with both individuals

speaking in their preferred language.

Speaker D maintains speaking Greek for the majority of the recorded conversation, however

in utterance 7 the lexical item ‘adjust’ is code-switched into English. The fact that Greek is

her preferred language throughout this excerpt as well as throughout the entire recording

indicates that this code-switch is due to a difficulty in recalling this particular lexical item

rather than not knowing the word in Greek . The English word for ‘adjust’ seems to be more

readily available and hence more easily retrievable from the speaker’s complex

communicative repertoire (Matras, 2009). Another potential explanation could be that this is

an example of lexical accommodation where perhaps Speaker D accommodates to the

preferred language of Speaker B due to a lack of confidence in Speaker B’s competence, but

this seems unlikely.

The code-switching of this single lexical item seems to further add to the meaning of what

Speaker D is actually saying. If it is the case that Speaker D is accommodating her language

to that of Speaker B, her need to ‘adjust’ that she describes in the excerpt is what she is

currently doing in the conversation and this is mirrored by the code-switch itself. This

reflection of content in the switch itself further reinforces the message she is expressing.

43
5. Discussion

The conversational patterns and language preferences of individuals were analysed and

interpreted on a holistic level, by analysing each conversation in its entirety. Two distinctive

types of conversations manifested in the recordings: Type I which shows speakers

consistently alternating between the two languages and Type II which involves one speaker

speaking Greek and the other mainly English, with occasional code-switches.

Two of the three recordings used in the analysis were of Type I. Conversation 1 was of this

type and was between Speaker A (a third generation speaker) and Speaker C (a first

generation speaker). It seems that Speaker A’s language choices do not appear to fall in line

with either of the patterns described by Christodoulou-Pipis (1991) and Garcia & Diaz

(1992). Speaker A describes her language history of growing up learning both languages

simultaneously, with one parent speaking Greek and the other in English. She describes that

attending a Greek supplementary school in addition to travelling to Cyprus every year and

communicating with family members living there reinforced her acquisition and maintenance

of the Greek language. This forms part of Baker’s (2011) theory of social, demographic and

culture factors where attending a ‘mother tongue institution’, like the Greek supplementary

school, and frequent travel appear to encourage the language maintenance of this speaker.

Therefore, this speaker’s language choices where she appears comfortable and confident

alternating between Greek and English consistently does not reflect the previously suggested

patterns of the generation of speakers in which she belongs.

An analysis of Speaker C is most compelling because she appears to digress the most from

the typical generational patterns described in section 2.2.2. Being a first generation speaker

of Greek and migrating at the age of fourteen, Speaker C displays no signs of having Greek

44
as her dominant language, which does not support the generational pattern previously

described in the literature. As can be seen in Excerpt 1, which is similar to other parts in the

entire conversation, Speaker C sets the default language of the conversation to English when

telling of the hardships she faced at school when first moving to London. Further to this,

Speaker C was very open in discussing what she described as experiences of racism that she

encountered when first migrating to London, which she felt was due to her inability to speak

English fluently at the time. Similarly to the first generation speakers investigated in

Schmid’s (2002) study, Speaker C speaks of a trauma, although less brutal, that she felt was

the result of being different and an ‘immigrant’. Using the ideas from Schmid’s study the

case of this individual might support the claim that the way in which an individual wishes to

be perceived and their attitude towards a particular language, especially if it is influenced by

a desire to ‘fit in’ with their new society might influence their language choice. However, this

might be a likely explanation but it cannot be known for sure that the experience that Speaker

C describes triggered a particular linguistic attitude, but it could perhaps be considered as a

possible contributing factor.

As can be seen in Tables 1 and 2, the participants in the third conversation (Speakers E and

F) were both first generation speakers of Greek that migrated at the age of five years old with

their families in the early 1960s. Their acquisition of English was described as a ‘sink or

swim’ approach being totally immersed in a new language and attending formal education in

a primary school much like native speakers would. However, unlike the predictions cited in

section 2.2.2, both speakers did not appear to have a preference and a dominant language of

English. They in fact displayed no predisposition to either language as they consistently and

frequently alternated between the two. This shows both speakers to be very comfortable

communicating in the ‘bilingual mode’ and exploiting their full communicative repertoires. A

45
possible explanation for this seemingly confident use of both languages could perhaps be

attributed to the language maintenance within the Greek-Cypriot community in which they

grew up and still live in. With both sets of parents being unable to speak any English, they

attended Greek school and acted almost as translators or ‘mediators’ (Clyne, 1967) for their

parents. Further to this, both individuals also frequently visited Cyprus and still have family

residing there. All these factors can be considered to have effectively contributed to the

retention of Greek in their communicative repertoires.

It is important to note here that during the recordings all participants spoke of travelling to

Cyprus at least once a year and maintaining frequent communication with people living there,

as well as having some kind of experience with a Greek supplementary school. These aspects

may be referred to as contributing factors of ‘transnationalism’. Transnationalism can be

defined “as the process by which immigrants build social fields that link together their

country of origin and their country of settlement” (Glick Schiller et al. 1992:1). These ‘social

fields’, which include increased communication through “new technology…with increasing

speed and efficiency” (Vertovec 1999: 447) and the Greek supplementary school allow for

the members of this community to maintain the Greek language. This ‘transnationalism’ can

be described as culpable for the fact that participants in this study, such as Speakers A, E and

F explored above, do not follow the previously proposed generational patterns of language

use (Christodoulou-Pipis,1991; Garcia & Diaz, 1992). Unlike the studies previously cited

such as Schmid’s (2002) research on German Jews, ‘transnationalism’ appears to be

prevalent in the lives of these individuals. The cultural demarcations that may have existed

previously when migrants moved from their homeland to their host society do not appear to

exist within this community and the language and culture of Cyprus transcends beyond

geographical boundaries and is maintained across the generations.

46
The only conversation of Type II was between a second-generation speaker (Speaker B) and

a first generation speaker of Greek (Speaker D) that migrated as an adult. The entire

conversation showed Speaker D’s dominant language to be Greek which in this case does

align with the generational patterns described in section 2.2.2 (Christodoulou-Pipis, 1991;

Garcia and Diaz, 1992; Rumbaut, 2004 and Portes and Rivas, 2011). This could to a degree

be explained by the fact that Speaker D migrated at a much older age and second language

acquisition is more difficult after puberty as cited by Matras (2009). Further analysis of the

content of her autobiographical conversation shows Speaker D explaining that she identifies

more with Greeks in Cyprus than the Greeks within this London community. This is made

evident in Excerpt 4, where Speaker D appears to be proud of her Greek heritage and refuses

to allow the Anglicizing of her son’s Greek name. This could perhaps suggest that similarly

to the literature correlating language and self-perception (Schmid, 2002), Speaker D wants to

maintain her Greek identity and shows this through her language choices. However, in

Excerpt 6 it can be seen that Speaker D expresses that she does feel people view her

differently and perhaps like a foreigner and says that she must ‘adjust’ more, however this is

not made apparent by her language choices.

From this same conversation, Speaker B, a second-generation speaker of Greek, is shown to

have a preference of speaking English, although she does code-switch into Greek for

accommodating purposes, often through reiteration. This shows that Speaker B does have a

competence in both languages, but does have a preference of English as predicted by the

generational patterns in section 2.2.2, her dominant language is English. Her experiences of

growing up as a British Cypriot showed her first language to be Greek, with this being the

dominant language of her parents. Further to this, she explains that she attended Greek school

47
as well as frequently visiting Cyprus with her parents, who now in fact live there. Initially,

from analysing the content of her conversation, one might predict that her language

preferences might be similar to Speaker A who as a third generation speaker of Greek and

showed no indication of a preferred language. However, unlike Speaker A, Speaker B stated

that she identifies as an “English Cypriot” rather than as a Greek Cypriot. This statement may

perhaps give an insight to this speaker’s attitude towards the Greek language and how she

wishes to be perceived. Perhaps this indicates that a combination of attitudinal factors as well

as language maintenance need to be taken into account when analysing language preferences.

By analysing the content of the conversations as well as the language preferences of each

individual, it appears that the personal histories and the proposed generational patterns of

language shift do not seem to be deterministic of all speakers. Although the experiences of

migration of some speakers such as Speaker C might offer an explanation for an apparent

language shift of the individual, others such as Speakers E and F do not. These speakers do

not appear to be less confident speaking in Greek and this could be attributed to the strong

language maintenance within this Greek-Cypriot community. This community is an example

of ‘transnationalism’ in their attempts to try and maintain their sense of identity with Cyprus.

This is made conspicuous as all participants stated they travel at least once a year to Cyprus

on holiday or to visit family. The well attended Greek supplementary schools by the children

of the speakers as well as themselves in the past, is testament to the ongoing efforts to hang

on to their ethnicity by way of language. These factors can also be considered to be what

Baker (2011) and Conklin & Lourie (1983) describe as some of the ‘cultural factors’ that

encourage language maintenance.

48
A further exploration of these language maintenance or ‘transnationalism’ factors highlights

the influence of ‘super-diversity’ (Vertovec, 2007) on this community. ‘Super-diversity’ can

be defined as a ‘diversification of diversity’ and can be described as “a kind of complexity

surpassing anything...previously experienced” and as “a dynamic interplay of variables” such

as country of origin, migration channel, legal status and access to employment (Vertovec

2007:1024). It is through the ‘super-diverse’ society that is London, where various and

diverse immigrants reside, that members of the Greek-Cypriot community feel confident in

using and retaining the Greek language and culture, and hence are able to maintain multiple

identities. The Greek-Cypriot and British identities within this community appear to co-exist

concurrently in this ‘super-diverse’ city. Therefore, this ‘super-diverse’ society in which they

function, where language maintenance features greatly, states that old models of assimilation

and language attrition across the generations no longer apply.

One of the ways in which this confident use of both languages manifests is through the use of

code-switching. Further exploration of the code-switches in the excerpts of the previous

chapter, show that all speakers exploited their full communicative repertoires at some point in

the discourse. A conversational and sequential analysis of these excerpts shows three main

motivations to code-switch: for accommodation purposes, the topic of conversation and

filling a linguistic gap.

As can be seen in sections 4.2 and 4.2.1, code-switching is used for accommodating

purposes. In both Excerpts 1 and 2, speakers are seen to be converging to the language

preference of their interlocutors. There appears to be two ways identified that speakers show

this convergence, through lexical accommodation and reiteration. In Excerpt 1 it can be seen

that Speaker B utilises the two languages to convey a meaningful juxtaposition, by

49
contrasting a presupposition and an inference. Speaker C lexically accommodates (Mahsain,

2015) her language to Greek, which can be interpreted as an additional confirmation to the

inference. It is most compelling however, that in Poplack’s (1980) study of Spanish-English

bilinguals the younger English-dominant generation of speakers generally accommodate their

language to that of the Spanish-dominant older generation in intergenerational

communication. This does not appear to be the case in this particular excerpt. Speaker C who

migrated as an adolescent does not display any signs of being Greek-dominant, and is

accommodating to the use of Greek of a third generation speaker of Greek.

However, Excerpt 2 does display a similar pattern of accommodation to that described by

Poplack (1980). It can be seen that through reiteration Speaker B pre-accommodates her

language to Greek when asking Speaker D a question, and later repeats this in English, which

ultimately appears to be the language she feels more comfortable communicating in. On the

other hand, Speaker D does not appear to accommodate her language to English. Perhaps, it

could be argued that instead of converging, Speaker D is diverging from the language

preferred by Speaker B.

The topic of conversation also seems to be an additional motivation to alternate between the

two languages. In section 4.3.1 it can be seen in Excerpts 3 and 4 that quoting from memories

seems to elicit an interesting pattern. The default language of the conversation is used for

introducing quotes, whilst the original language of the quotes is used when repeating them.

Topics in local public discourse also seem to determine language alternations. It can be seen

in Excerpt 5, that code-switching manifests in the utterances of Speaker D by repeating the

lexical item ‘community’ that perhaps is heard and spoken of more often in local public

50
discourse which is English in this setting. This English form could then perhaps be described

as being more readily available in their communicative repertoire.

The use of code-switching also appears to be used to ‘fill a lexical gap’ which arises from a

difficulty in recalling lexical items, not a lack of competence in one of the two languages. In

section 4.4, it can be seen in Excerpt 6 that the single lexical item ‘adjust’ appears to be the

only word code-switched in a set of utterances that are all in Greek. This might suggest that

the English form is more readily available in Speaker D’s communicative repertoire than the

equivalent Greek form. However, in alternating between the two languages in this way,

Speaker D cleverly reflects the content of what she is saying. In switching to English when

describing that she needs to ‘adjust’ to the community here in London, she is essentially

showing that in her choice of language. This may present an alternative theory for this

particular code-switch, that Speaker D is not experiencing difficulty recalling a lexical item

but is additionally conveying her message through her language choice.

51
6. Conclusion

In this research paper, I aimed to identify potential motivations to code-switch when

bilinguals availed themselves of their full linguistic repertoires in inter-generational and intra-

generational communication. Further to this, I also investigated whether a potential

connection existed between participants’ personal experiences of migration and their

language choices. However, I incorporated an exploration of language maintenance and an

aspect of ‘transnationalism’ within this community into my analysis of language choices.

Schmid (2002) ultimately suggests that experiences of migration may determine a linguistic

attitude, which influences how confident individuals are speaking their first language. In

addition to this, Christodoulou-Pipis (1991) and Garcia & Diaz (1992) propose two rather

similar generational patterns that typically arise in immigrant communities, resulting in

language shift and language loss for future generations. However, by employing a qualitative

approach and analysing conversation-based data holistically it seems that the personal

histories of these individuals and their pre-determined generational patterns do not always

appear to determine their linguistic choices. Although it is possible that their experiences

might evoke a need to assimilate into their new society in order to be perceived as less

different, ultimately these conclusions cannot be drawn. It may be the case that language

choices may be a part of a more complex, multifaceted phenomenon, that of ‘super-diversity’

(Vertovec, 2007). It seems that language maintenance and factors of ‘transnationalism’ such

as the Greek supplementary school and frequent contact with Cyprus account for some of the

different language choices of these members of the Greek-Cypriot community in North

London.

52
Further to this, I was able to undertake a conversational and sequential analysis of the three

conversations, which allowed me to interpret potential motivations for code-switching in a

bilingual context. Code-switching was found to be a compelling communicative tool in which

contrasts and additional meanings were given by alternating between the two languages. The

three main motivations that I identified in my analysis included aspects of accommodation

theory, topics of discussion and potentially filling a lexical gap. There were two main types

of accommodation that had different evidence bases that were found in the study: lexical

accommodation and reiteration. With regards to topics, quoting from memories created a well

established alternating pattern in the discourse where one language was used to introduce

quotes and another for the quotes themselves. Topics in local public discourse also appeared

to motivate the switching of a single lexical item. Filling a lexical gap when there is difficulty

recalling a lexical item was the third motivation identified in this study.

By combining the different approaches to analysing language choices cited in the previous

literature, I have filled a potential ‘gap’ in previous research with regards to language choices

and code-switching. The language preferences of this small sample of the Greek-Cypriot

community in North London can only begin to be explained by an interplay of various factors

such as personal experiences of migrations, generational patterns and language maintenance.

A most compelling finding is the strong sense of language maintenance and

‘transnationalism’ existing within this Greek community which has resulted in the continued

use of the Greek language across the various generations of speakers within this study.

Ultimately, this study finds that the personal histories of migration and generational patterns

alone are not deterministic of language choices, and members of present day communities

differ in their multilingual repertoire maintenance due to ‘transnationalism’ factors such as

53
supplementary schools and travel, and future studies should explore individuals in a ‘super-

diversity’ context on a case by case basis.

54
References:

Appel, R. & P. Muysken. 1987. Language Contact And Bilingualism. London: E. Arnold.

Auer, P. 1984. Bilingual conversation. John Benjamins, Amsterdam.

Auer, P. 1995. The Pragmatics of Code-switching: A Sequential Approach. In: L. Milroy &
P.Muysken, (eds.). One Speaker, Two Languages: Cross Disciplinary Perspectives on Code
switching, 115-35. Cambridge: Cambridge University Press.

Baker, C. 2011. Foundations Of Bilingual Education And Bilingualism. Buffalo, N.Y.:

Multilingual Matters.

Census 2011: Census Snapshot: Main Language (2013).

<https://files.datapress.com/london/dataset/2011-census-diversity/2015-11-
03T10:50:22/CIS2013-01%20Main%20Language.pdf > [accessed 27 April 2017]

Christodoulou-Pipis, I. 1991. Language use by Greek-Cypriots in Britain. Nicosia: Diaspora

Books.

Conklin, N. & M. Lourie. 1983. A Host of Tongues. New York: The Free Press.

Clyne, M. G. 1967. Transference and Triggering. The Hague: Nijhoff.

Cyprus Educational Mission UK (K.E.A). 2017.

< http://kea.schools.ac.cy/index.php?id=whoarewe> [accessed 27 April 2017]

Fishman, J. A. 1997. Reversing Language Shift. 1st ed. Clevedon, Eng.: Multilingual Matters.

Garcia, R, & Diaz, C. 1992. The status and use of Spanish and English among Hispanic youth
in Dade County (Miami) Florida: A sociolinguistic study. Language and Education 6, 13-32.

Giles, H., J. Coupland, & N. Coupland. 1991. Contexts Of Accommodation. Cambridge:

Cambridge University Press.

Giles, H. & P. Smith. 1979. Accommodation Theory: Optimal Levels of Convergence. In: H.
Giles & R. St. Clair, (eds.). Language and Social Psychology. 45-65. Oxford: Blackwell.

Glick Schiller, N., L. Basch, & C. Blanc-Szanton. 1992. Towards a transnational perspective
on migration. New York: New York Academy of Sciences.

Grice, H. P. 1975. Logic and Conversation. In: P. Cole & J. Morgan, (eds.), Syntax and
semantics 3: Speech Acts. New York: Academic Press.

Grosjean, F.1982. Life with Two Languages: An Introduction to Bilingualism. Cambridge,

Mass: Harvard University Press.

55
Grosjean, F. 2001. The bilingual’s language modes. One mind, two languages: Bilingual
language processing 7.11, 1-22.

Gumperz, J. 1977. Sociocultural Knowledge in Conversational Inference. In: M. Saville-

Troike (ed.), 28th Annual Round Table Monograph Series on Languages and Linguistics.
Washington, DC: Georgetown University Press.

Gumperz, J. 1982. Discourse Strategies. Cambridge: Cambridge University Press.

Labov, W. 1972. Sociolinguistic Patterns. Oxford: Blackwell

Lance, D. M. 1975. Spanish-English code-switching. In El lenguaje de los chicanos, E.

Hernandez-Chavez & A Beltramo (eds.). Arlington, VA: Center for Applied Linguistics.

Li Wei. 1998. The ‘why’ and ‘how’ questions in the analysis of conversational code-
switching. In P. Auer (ed.), Code-switching in conversation: Language, interaction and
identity, 156-176. Routledge.

Li Wei. 2000. The Bilingualism Reader. London: Routledge.

Mahsain, F. 2014. Motivations Behind Code-switching Among Kuwaiti Bilingual Schools’

Students. [Thesis]. Manchester, UK: The University of Manchester.

Matras, Y. 2009. Language Contact. Cambridge: Cambridge University Press.

Paltridge, B. 2012. Discourse Analysis. 2nd ed. London: Bloomsbury

Papapavlou, A. & and P. Pavlou. 2001. The Interplay Of Language Use And Language
Maintenance And The Cultural Identity Of Greek Cypriots In The UK. International Journal
of Applied Linguistics 11.1, 92-113.

Pauwels, A. 2004. Language Maintenance. In A. Davis & C. Elder (eds.). The handbook of
applied linguistics, 719-738. Cornwall: Blackwell Publishing.

Pedraza, P. 1978. Ethnographic observations of language use in El B'arrio. (Unpublished

master’s thesis).

Podesva, R. J., & D. Sharma. 2016. Research Methods In Linguistics. 1st ed. Cambridge:
Cambridge University Press

Poplack, S. 1980. Sometimes I’ll start a sentence in Spansih y termino en español: Towards a
Typology of Code-switching. Linguistics, 18, 581-618.

Portes, A. & A. Rivas. 2011. The Adaptation of Migrant Children. The Future of Children,
21(1), 219-246.

56
Rumbaut, R. G. 2004. Ages, life Stages, and Generational Cohorts: Decomposing the
Immigrant First and Second Generations in the United States. International Migration Review
38, 1160-12.

Sebba, M. & T. Wootton. 1998. We, they and identity: Sequential versus identity-related
explanation in code-switching. In P. Auer (ed.), Code-switching in conversation: Language,
interaction and identity, 262-286. Routledge.

Schmid, M. S. 2002. First Language Attrition, Use And Maintenance. 1st ed. Amsterdam: J.
Benjamins Pub.

Vertovec, S. 1999. Conceiving and researching transnationalism. Ethnic and Racial Studies
22(2), 447-462.

Vertovec, S. 2007. Super-diversity and its implications. Ethnic and Racial Studies 30(6),
1024-1054.

Zeevaert, L. & J.D. Thije,.ten. 2007. Introduction. In J.D. ten Thije & L Zeevaert (eds.),
Receptive Multilingualism. Linguistic analyses, language policies and didactic concepts. 1-
25. Amsterdam: Benjamins.

!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!

57