The Fiuman dialect is an autochthonous minority Romance language spoken for centuries in the Croatian city of Rijeka and its surroundings, in what is today a Croatian-language dominant environment. There is evidence that it has been uninterruptedly spoken as one of the two indigenous languages of the city of Rijeka (the other being Croatian Chakavian) from the 15th century onwards (Rošić, 2002); it is possible, however, that it was used even earlier, in the Roman period. Because of the historical and political changes in Rijeka in the mid-20th century, today Fiuman is spoken by a relatively small minority group living in several city neighbourhoods (Lukežić, 1993), as a result of which its maintenance has been questioned (Lukežić, 1993; Rošić, 2002). Todayʼs Fiuman speakers are mainly bilingual or multilingual speakers of Fiuman, standard Italian and standard Croatian or one of its substandard varieties (Badurina & Matešić, 2008; Lukežić, 2008). Standard Italian is also a minority language in Croatia, but as the official language of Italian minority nurseries, schools and other institutions, its position is much more stable. Fiuman maintenance thus depends on sustaining the diglossic relationship with standard Italian on the one hand and standard Croatian on the other. This doctoral thesis aims to determine the current degree of vitality of Fiuman and the possibilities for its future maintenance. Ethnolinguistic vitality, introduced by Giles, Bourhis and Taylor (1977), is the ability of a group to maintain its existence as “a collective entity with a distinctive identity and language” (Ehala, 2015, p. 1). In the situations of language contact between two (or more) speech communities of unequal status, it is more probable that minority groups with higher vitality will survive as distincitve entities and maintain their language, while low vitality groups are prone to assimilation with majority groups (Giles et al., 1977). It is the degree of vitality that is a key determinant of language maintenance (LM) or language shift (LS) in favour of the dominant language. The term LM is used to describe a situation in which a minority language persists in some or all domains of life despite the presence of the dominant or majority language. On the other hand, more or less gradual abandonment of the minority language and its substitution with the dominant or majority language leads to LS (Pauwels, 2004, 2016). Based on previous studies, it is difficult to determine the actual position of Fiuman with respect to LM or LS. The studies have mostly dealt with the descripition of Fiuman grammatical properties and vocabulary (Bató, 1993/1999; Berghoffer, 1894/1999; Bidwell, 1967; Blecich & Tamaro, 2015; Bratulić, Đurđulov, Blecich, & Kraš, 2015; A. Depoli, 1913; G. Depoli, 1928/1999; Folena, 1968–1970; Gottardi, 2004, 2007; Lukežić, 1993; Mestrovich, 2001; Pafundi, 2011; Rošić, 2002; Samani, 2007; Spicijarić Paškvan, 2018). Only recently has the interest of linguists shifted to FD vitality, that is, LS and LM (Crnić Novosel & Spicijarić Paškvan, 2014, 2015; Drljača Margić, Kraš, & Smiljanić, 2015; Lukežić, 2008; Spicijarić Paškvan & Crnić Novosel, 2014). The results of the studies into the maintenance of Fiuman have yielded contradictory results by indicating both LS (Lukežić,1993; Rošić, 2002) and LM tendencies (Crnić Novosel & Spicijarić Paškvan, 2014; Spicijarić Paškvan & Crnić Novosel, 2014). According to these studies, Fiuman is a sociolect, with a declining number of speakers, spoken in private domains and informal situations in the institutions of the Italian minority of Rijeka. In the present study, the question of Fiuman vitality has been addressed by investigating two of its aspects. According to Giles et al. (1977), vitality can be objective (OV) and subjective (SV). OV is measured by looking into three groups of factors: 1) demographic factors (size of the group, birth rate, endogamy, exogamy, immigration, emigration, geographical distribution and the proportion of community members), 2) institutional support factors (the level of power in business, industry, administration, education, mass media, culture and sports) and 3) status factors (social prestige, socio-historical status and the prestige of the language and culture). SV, as viewed by Bourhis, Giles and Rosenthal (1981), implies community membersʼ subjective perceptions of OV, and it is considered decisive for group behaviour in LM. The present study aims to investigate both objective and subjective vitality factors. What is crucial for the vitality of a certain community and its language are the continuous intergenerational transmission of the communityʼs language and culture, sustainable demography, active institutions, social cohesion and emotional attachment to the collective identity (Ehala, 2015). However, these factors might be challenging to measure. A fairly good estimation of the group vitality, according to Smith, Ehala, and Giles (2017), can be provided by a language endangerment assessment instrument, such as Expanded Graded Intergenerational Disruption Scale EGIDS (Lewis, 2009; Lewis, Simons, & Fennig, 2016), along with basic demographic data, data on speakersʼ self-reported language use in different domains and their assessment of the vitality. In this study the following vitality factors were taken into consideration: 1) demographic data; 2) intergenerational transmission; 3) the use of Fiuman in different language domains; 4) institutional support to Fiuman; 5) speakers’ (language) identity; 6) speakers’ motivation for knowing Fiuman and 7) attitudes towards Fiuman, standard Croatian and standard Italian, as well as towards Fiuman maintenance. The data were obtained using a mixed-method approach, comprising qualitative and quantitative research methods. A total of 249 Fiuman speakers, recruited by means of snowball sampling, completed the Fiuman Language Questionnaire (Bratulić, Drljača Margić, & Kraš, 2017), comprising closed- and open-ended questions. The questionnaire elicited the participants’ sociodemographic data (age, gender, former and current residence, level of education) as well as information about their language biography, Fiuman proficiency, use of Fiuman in different domains, motivation for knowing Fiuman, (language) identity, and attitudes towards Fiuman, standard Croatian and standard Italian. The data obtained by means of the questionnaire were complemented by the data gained via semistructured interviews with 34 Fiuman speakers. Both the questionnaire and the interviews were conducted in standard Croatian or standard Italian, according to the participants’ preferences. The participants were divided into six age groups (Levinson, 1986) for the purposes of exploring possible intergenerational differences. The results suggest that LS has already started among Fiuman speakers. In spite of speakers’ strong emotional attachment to Fiuman as part of their ancestry, cultural heritage and identity, they increasingly opt for Croatian. Although intergenerational transmission seems to persist, the number of active speakers is decreasing and so is the use of Fiuman in different domains. The qualitative data analysis, conducted mainly on the interview data, shows that the historical changes in the 20th century had an impact on speakers’ attitudes towards Fiuman and their motivation for knowing and using it, and resulted in an increased use of Croatian. Political, ideological, economic and other changes after the Second World War led to a massive emigration of Fiuman speakers from Rijeka and a large-scale immigration of Slavic-speaking population, both of which, together with today’s negative demographic trends, resulted in the reduction of Fiuman use. Nowadays Fiuman is spoken in private domains among family members and friends, and in informal and sometimes formal situations within the Italian national minority institutions. Institutional support is mainly given to standard Italian, while Fiuman seems to lack adequate support. Fiuman is absent from the public use and the mass media, which publish and broadcast in standard Italian. Standard Italian is also taught in nurseries and schools, which admit an increasing number of Croatian-speaking children. As a result, children speak Croatian among themselves in these institutions. One of the most revealing findings, however, is that Fiuman is increasingly used in its written form on the Internet and social networks, which have been recognised as a potential platform for its maintenance. The results also suggest that Fiuman vitality relies more on intergenerational transmission and speakers’ attachment to the collective identity than on the use of Fiuman in different domains, institutional support and positive demographic trends. Fiuman speakers are mainly multilingual and regard Fiuman as their mother tongue. Their use of Fiuman is primarily driven by integrative reasons, such as maintaining close family relations and community cohesion, and not instrumental ones. Their attachment to Fiuman and the Fiuman community is strong and deeper than their attachment to Croatian and Italian; however, younger generations report lower Fiuman proficiency and less frequent use of Fiuman, and attach more importance to being part of the dominant Croatian community. Taking the multilingual character of today’s Fiuman speakers into consideration as well as the fact that multilingualism itself does not necessarily lead to LS, some directions for reversing the LS in the case of Fiuman and boosting Fiuman maintenance in the future have been given. They primarily involve increasing the number of active speakers by sustaining intergenerational transmission and the use of Fiuman in different domains. Furthermore, the importance of stronger instrumental motivation, higher awareness of the benefits of multilingualism and more positive attitudes towards Fiuman has been underlined. Providing greater institutional support to Fiuman maintenance has also been suggested. Both individuals and institutions, as well as the minority and the majority community, are expected to extend efforts to restore and increase Fiuman vitality.