Dimensions of Linguistic Space: Variation – Multilingualism Conceptualisations Dimensionen des sprachlichen Raums: Variation – Mehrsprachigkeit – Konzeptualisierung

The article deals with the use of different lexical variants on both sides of the AustroGerman border. In particular, it addresses the question of whether crossborder commuters adapt to the language usage of the neighbouring country in terms of lexical variants. Along with this, the question is prompted to what extent this state border establishes a language border. The collected data show that a sense of national inclusion proves to be the most important influencing factor. Synchronisation processes are currently emerging among crossborder commuters and younger informants from Austria. They use significantly more variants, which are typically assigned to the neighbouring country, than the nonbordercrossers and older informants from Austria. Accordingly, it can be assumed that the AustroGerman border also forms to a certain extent a linguistic border on the lexical level.


Introduction *
Political borders which pass through interlinked language areas can be seen as one of a range of factors that affect language variation and change on both sides of the border (cf. Auer 2004;Auer et al. 2015;Harnisch/Reinhold/Schnabel 2008;Niebaum 1990;Scheuringer 1990;Smits 2011). Political borders cause the dialects of a 'former' dialect continuum to become associated as dialects of a certain language. An often-cited example of this are the dialects along the German-Dutch border. People associate some dialects along the German-Dutch border as dialects of Dutch while others are being considered dialects of German. The German dialects are heteronomous with respect to Standard German, and the Dutch dialects to Standard Dutch (cf. Chambers/Trudgill 2009: 9). As a consequence, divergence processes can be identified along the political border. Smits (cf. 2011) shows that the dialects of the 'former' West Germanic dialect continuum in the German-Dutch-border region are drifting apart due to the different standard languages. The people adjust their dialects to their respective standard language. Generally spoken, there are convergence and advergence processes taking place within one * The essay is largely based on Bülow/Kleene (in press). We would like to thank Stephan Elspaß, Katharina Jakob and Rüdiger Harnisch for their support and valuable information. country so that the language development depends on the linguistic situation in the particular country (cf. Auer 2004: 167-177).
According to traditional dialectology, the root of the different language behaviour relates to the fact that political borders -similar to geographical obstacles such as rivers or mountain ranges -occur as a hindrance to communication (cf. Goossens 1977: 56-77;Pickl 2013: 28-31). Thus, accommodation processes, or rather synchronisation processes, are inhibited. 2 Auer (cf. 2004) and Auer et al. (cf. 2015) take a rather different approach, stating that political boundaries, unlike the aforementioned physical obstructions, no longer seem to be effective obstacles to language contact as well as accommodation and synchronisation processes. Instead, mental boundaries seem to be much more relevant to processes of divergence. There are many indications that effects of linguistic division by political boundaries have never been as marked as they are nowadays, when their actual significance for both traffic obstruction and the obstruction of speaker contact, respectively, is virtually non-existent (cf. Auer et al. 2015: 324). This can only mean that changes to conditions with regard to synchronisation of speakers (cf. Schmidt/Herrgen 2011) would therefore not play a part in linguistic variation and changes across political borders.
Es sind also nicht die faktischen Verkehrsgrenzen, sondern der Raum als mentales Konstrukt, der die Wahrnehmung sprachlicher Variabilität steuert und gegebenenfalls auch in der sprachlichen Produktion sprachliche Grenzen (Isoglossen) bewahrt oder sogar aufbaut. Allenfalls können natürliche oder politische Grenzen für diese mentalen Raumkonzepte auslösend sein, nicht aber für die sprachlichen Divergenzen im Raum selbst. Auch die immer wieder beobachtete Tatsache, dass Konfessionsgrenzen die sprachliche Raumwahrnehmung (auch noch heute) beeinflussen, ist mit Simmels Idee der ‚seelischen Begrenzungsprozesse' eher zu erklären als mit dem Akkommodationsmodell. (Auer 2004: 162) 3 2 Hereafter, we refer to the synchronisation process by Schmidt/Herrgen (cf. 2011) without questioning the accommodation theory. Regard the concept of language accommodation cf. Lanwer (2015: 44-48) among others. 3 'It can thus be said that it is space as a mental construct, rather than actual borders, which determines the perception of linguistic variability and perhaps even maintains or creates linguistic boundaries (isoglosses) in linguistic production. Physical or political boundaries can at best act as a trigger for these spatial concepts, but not for linguistic divergences within an area. The often observed fact that confessional boundaries exert an influence (even today) on linguistic spatial conception can also better be explained by Simmels' idea of spiritual limitation processes than the accommodation model. ' (our translation) Auer (cf. 2004: 177) stresses the importance of his deliberations with respect to the linguistic conditions in the Austro-German border region. Several different divergence processes on different levels can be ascertained (cf. chapter 2 as well as Kleene 2017;Mihm 2000;Scheuringer 1990;Zehetner 1985). Kleene's (cf. 2017) and Fischer's (cf. 2016) studies contribute attitudinal-perceptional findings for the Austro-German border situation in the area of Passau. It can be shown from the results of both studies that the national border does indeed constitute a mental border for the majority of the participants. However, the role of synchronisation in divergence processes cannot be marginalised based on these findings. The role of synchronisation processes could be better assessed using data on the language behaviour of cross-border commuters. The question is whether crossborder travel can influence language variation and language change in the border region. Moreover, the question arises whether the mental boundaries could be strong enough as to prevent synchronisation processes of even cross-border commuters altogether. This article explicitly considers the question of how relevant speaker contact in the Austro-German border region is for the usage of lexical variants. To this end, cross-border commuters (n = 18) and non-border crossers (n = 136) from seven localities close to the border were interviewed using an onomasiological survey. Whilst this research design does not allow a collection of mental constructs and borders in people's minds, it should enable to better empirically assess the impact of cross-border mobility on language use. This contribution clarifies the role which theoretical synchronisation actually plays when explaining language variation and language change in border regions.
The following chapter deals with the concept of synchronisation (2.1) as well as the general language development in the Austro-German border region (2.2). This is followed by a deduction of the hypotheses and research questions (2.3) and a description of the methodology (3). Subsequently, a presentation of the results (4) and a related discussion in the framework of theoretical concepts (5) will be offered. Finally, there will be a short conclusion at the end of the contribution (6).

Linguistic dynamics in the Austro-German border region
This paper focuses on the lexical variation of everyday language (Alltagssprache), defined in Möller/Elspaß (cf. 2014: 122) as those forms of speech that speakers of German use in everyday communication, i.e. in the social and functional realm of the private 'language of immediacy', spontaneous conversation among friends, relatives or acquaintances, or even in informal exchanges among acquaintances from the same place / town / village, e.g. at the local grocery store.
Despite a number of cross-border dictionary and glossary projects (cf. Ammon/Bickel/Lenz 2016; Eichhoff 1977Eichhoff -2000Elspaß/Möller 2003 ff.;Pickl et al. in this volume), little empirical research has been carried out with respect to the particular situation along the Austro-German border to date, except for the work by Scheuringer (1990Scheuringer ( , 1995. However, the development of lexis can reveal a good deal about language change processes (cf. Pickl 2013), affecting the part of language usage most likely to be influenced by external changes such as demarcations of political borders. As Scheuringer (1990: 273) explains: Mehr als die anderen sprachlichen Teilbereiche ist der Wortschatz ein sehr variables und flexibles Teilsystem, das früher als jene auf außersprachliche Veränderungen reagiert und darum diese Veränderungen am frühesten widerspiegelt. 4 It can be stated that one of the empirical challenges may be that for vocabulary and some other systemic levels of language (cf. Scheutz 1985 and Kleiner 2011 ff. on the phonetic level) a number of complex speech level-specific patterns of variation have to be assumed (cf. Scheuringer 1995).
The theoretical part of this paper, therefore, examines the issue of varieties and speech levels more closely (cf. 2.2). In the first instance though, the role of speaker mobility and speaker contact will be addressed, based on the synchronisation approach used in Schmidt/Herrgen (cf. 2011) and Schmidt (cf. 2010).

Synchronisation approach
According to Schmidt (2010: 211), linguistic dynamics is defined as "the study of the influences acting on the constantly shifting complex of language and the resultant modifying and stabilizing processes". Schmidt/Herrgen (2011: 32) claim to be able to explain linguistic divergence and convergence using a unified concept called "synchronisation" (cf. also Schmidt 2010: 212). This concept is based mainly on the critical review of the methodological separation of synchronic and diachronic language considerations. A key point of origin in the theory of linguistic dynamics is the observation that every natural language in its spatial and temporal dimension is always dynamic, variable and heterogeneous (cf. Kehrein 2012: 33-34).
Ultimately, the concept of synchronisation works as tuning or alignment of differences in competency of actual speech, resulting in a "stabilization and/or modification of the active and passive competencies involved" (Schmidt 2010: 212;Schmidt/Herrgen 2011: 28). This stabilisation and/or modification of individual language competency is a non-linear process, continually accomplished with the respective communication partners (cf. Bülow 2017: 172) whereby strategies of speakers as well as listeners have to be taken into account. 5 Schmidt/Herrgen (cf. 2011: 29-34) differentiate between micro-, meso-and macrosynchronisation, where meso-and macrosynchronisation stem from microsynchronisation. Whilst microsynchronisation is confined to individual interactions, mesosynchronisation depicts the result of a "series of parallel acts of synchronization, performed by individuals in personal contact situations, which lead to the establishment of common context-dependent linguistic knowledge" (Schmidt 2010: 213). It can be said that mesosynchronisation results in "the establishment of group and contextdependent linguistic conventions and thus, in the final instance, [it is responsible] for the formation of varieties" (Schmidt 2010: 213).
Therefore, mesosynchronisation could have diverging effects on the process of language change in the border region. For speaker mobility, however, it is crucial that common communication strategies can be established through frequent speaker contact over time as well as through a high degree of emotional integration. As a result of mesosynchronisation, a high level of speaker contact can eventually lead to the development of regional languages (cf. Schmidt/Herrgen 2011: 70;Lanwer 2015: 41). The question arises whether mesosynchronisation can also operate across political borders which simultaneously represent mental boundaries.
Meanwhile, macrosynchronisation is oriented towards the overall linguistic system and refers to acts of synchronisation "via which the members of a linguistic community orient themselves to a common norm" (Schmidt 2010: 214), without the need of personal contact. They also define "the boundaries of the dynamic system of a particular language" (Schmidt 2010: 214;Schmidt/Herrgen 2011: 32). Since the 1980s, a large influence can be ascribed to forms of mass-media orality in which large groups (e.g. Federal German, Austrian and Swiss German variants of the standard language) take part, however primarily on a receptive level (cf. Schmidt/Herrgen 2011: 32).
The different processes of synchronisation can, however, hardly be viewed in isolation. Interactions and interferences result in a complex network of divided linguistic bases of knowledge which can be defined as varieties and language levels (Sprechlagen) (cf. Schmidt/Herrgen 2011: 51-52; Kehrein 2012: 36).

Varieties and speech levels
Communication in the Austro-German border area takes place using several different varieties and speech levels. Wiesinger (2014: 76-84), for example, differentiates between four distinct manners of speech for the Bavarian speaking part of Austria (Basisdialekt 'base dialect', Verkehrsdialekt 'common language', (regionale) Umgangssprache 'colloquial language' and Standardsprache 'standard language'). Schmidt/Herrgen (cf. 2011: 86) assume two or more varieties within Germany, which corresponds to what is a typical estimate for conditions there. 6 It may in fact be the case that one may assume regionally marked speech levels between dialects and standard languages, 7 because the levels between base dialect and standard language in particular are difficult to define based on language structure alone. Scheutz (1999: 107), therefore, rightfully criticises: [S]ogar bei einer Beschränkung auf den phonetisch-phonologischen Bereich ist nicht ersichtlich, in welcher Weise die Zuordenbarkeit einzelner Formen zu den vier Schichten in eindeutiger Weise geregelt sein könnte. 8 Any classification can always be treated as an idealisation and shortening by the researchers. Scheutz (1999: 106) summarises the definition of language layers as an attempt Indications of different varieties or speech levels can also be provided by examinations on a subjective level. Therefore, two to four speech levels 10 can be ascertained for the border region around the towns of Passau and Schärding (cf. Kleene 2017). The speech level defined as 'dialect' (Dialekt) by informants in the Bavarian language area mostly serves as a means of everyday communication within a family setting. In contrast, the mode of speaking described as 'standard language' (Hochdeutsch) is restricted to official occasions as well as to communication with senior or unknown people or those of higher authority. Intermediate speech levels (Umgangssprachen) are applied on occasions when dialect would be too informal and standard language too formal. Kleene (cf. 2017) detects the intermediate speech level as one which is knowingly used by informants if they are in a more mobile setting with more frequent exposure to people from other regions. In addition, Wiesinger (cf. 2014: 81) finds that the majority of young Austrians now see everyday or colloquial language in much the same way as an intermediate speech level. This development can also be applied to conditions in Bavarian border communities, although with linguistic innovations emanating from Munich or the southern German standard rather than Vienna.
One has to bear in mind though that -on an objective as well as a subjective level -a uniform ordinary language does not exist: "Es gibt also sozusagen 'Register' der Umgangssprache […] -sowohl in Richtung Standard als auch in Richtung Dialekt. " (Ammon/Bickel/Lenz 2016: XLV). 11 However, the corresponding ordinary language on both sides of the border can ultimately be regarded as a result of balancing intra-national processes of convergence and advergence between dialect and standard language (cf. Wiesinger 2014: 82;Scheutz 1999). The crucial point is that the national border has ultimately led to advergence and convergence processes, further resulting in processes of divergence on all language levels.
This observation then leads to the question to what extent the Austro-German national border can be treated as a language border on the different levels. In the framework of linguistic dynamics, Schmidt/Herrgen (cf. 2011: 62) demonstrate that macrosynchronisation is restricted to members of a language community which in turn has led to distinct oral norms for Germany, Austria and Switzerstandard and dialect, emanating from his linguistic competence and intuition. ' (our translation) 10 Speech level is understood here as a subjective device and is not to be confounded with the definition of speech level by Schmidt/Herrgen (2011). 11 ' As such, there are different kinds of everyday language […] both tending towards standard as well as towards dialect. ' (our translation) land. A crucial factor had to be the mass media (cf. de Cillia 2016: 336-339) with public broadcast in particular. However, Bavarian/German as well as Austrian media (radio and television broadcasts) have been available on both sides of the border most of the time. 12 Moreover, news, advertisements, films, series etc. are widely available on the internet nowadays which undoubtedly has an effect on language also. The oral norms are now partly de-nationalised (cf. Herrgen 2015: 157). Based on a listener judgement test and an analysis of further subjective datasets of people from Austria and Bavaria, Kleene (cf. 2017) demonstrates that Austrians do in fact differ between two oral norms (see also Herrgen 2015). The Austrian standard variety can be regarded as standard norm, whereas the German standard variety has to be interpreted as a cognitive standard norm for Austrians. Thus, a predominantly mental boundary can be determined on the level of standard language as shown by Scheuringer (cf. 1990: 300) who declares that the national inclusion does influence language development more than ever before. This primarily affects the lexical level, and more specifically the administrative and merchandise trade vocabulary (cf. Scheuringer 1990: 374; Ammon/ Bickel/Lenz 2016). For instance, the high school diploma or A-level, for example, is called Matura in Austria, while German students do Abitur (cf. Ammon/ Bickel/Lenz 2016). Divergence processes take place not only on the dialectal and standard level but also on intermediate speech levels like everyday language. Zehetner (1985: 62) observes that the 'common languages' (Verkehrssprachen) or 'urban ordinary languages' (städtische Umgangssprachen) in Bavaria and Austria have increasingly diverged. People in Bavaria would increasingly be oriented towards language use in Munich, whilst conformity with the Viennese language use is prevalent in Austria. Mihm (cf. 2000Mihm (cf. : 2119 adds that there are marked differences in sets of variants and situational areas of application between the ordinary languages on both sides of the border. The pair of variants Meerrettich/Kren ('horseradish') can be named as an example: Scheuringer (cf. 1995: 49) shows that Kren used to be fairly widespread in the Bavarian language area and beyond, but is progressively superseded by the standard language form Meerrettich, whereby the national border essentially becomes, or already is, a variety border. This shows that processes of language change attributable to the border do have an impact on the dialectal level.

Research questions and hypotheses
The following research questions and hypotheses are based on theoretical concepts as described above, but also touch on other premises and assumptions which can only briefly be outlined in the following section.
The first two hypotheses have the underlying assumption that the informants have primarily been socialised in a specific linguistic context. It is expected that based on corresponding processes of meso-and macrosynchronsiation, Austrian informants are more likely to describe pictures using a significantly higher number of variants categorised as typical Austrian than German. Likewise, German informants use more German/Bavarian variants during the onomasiologic survey.
H1.1: Austrian informants use significantly more characteristically Austrian variants than German informants. H1.2: In contrast, German informants use significantly more characteristically German/Bavarian variants. 13 Two factors are critical for the following two hypotheses (H2). Firstly, the increasing influence of mass media on language behaviour has to be taken into account. On various occasions, it has been assumed that the German media landscape exerts a significant influence on the everyday and hence ordinary language in Austria (cf. de Cillia 2016: 336-339; Ammon/Bickel/Lenz 2016). This chiefly affects younger speakers that are growing up with this influence. It also has to be assumed that younger speakers use the colloquial everyday language more frequently than older speakers, e.g. when they leave their hometown to study (cf. Kleene 2017). Secondly, because higher-level variants accepted as German have a higher (overt) prestige, it can be assumed that younger informants from Austria tend to use more typical German/Bavarian vocabulary than older Austrian informants in our study. No significant difference is supposed to be present between the German informants' age cohorts. H2.1: The younger Austrian informants use significantly more typical German/Bavarian variants compared with the older informants. H2.2: The younger German informants use the same amount of typical Austrian variants as the older informants.
As was shown in the depiction of the problem, the influence of processes of micro-and mesosynchronisation on the use of lexis will also be examined. Crossborder commuters working or going to school on the other side of the border on a regular basis should be expected to be more familiar with the language use of their neighbours due to the language contact and consistent processes of microsynchronisation compared with people not commuting across the border (non-border crossers). Thirdly, it can therefore be assumed that cross-border commuters use significantly more typical variants of the neighbouring countries than non-border crossers.
H3.1: Austrian informants working (or attending school) in Germany use significantly more typical German/Bavarian variants than Austrian informants working (or attending school) in Austria. H3.2: German informants working (or attending school) in Austria use significantly more typical Austrian variants than German informants working (or attending school) in Germany.
In sociolinguistic studies and studies on the use of dialects, it has often been proven that gender is a relevant factor (cf. Chambers/Trudgill 2009). Therefore, it will also be considered whether gender has an influence on the usage of lexical variants in the border region.
RQ1: Is gender a relevant factor with regard to the use of lexical variants in the Austro-German border region?

Methodology
Using an onomasiologic research design, it will be examined which lexical variants are being used as part of the everyday language on either side of the Austro-German border, with an emphasis on the influence of the independent variables as described above. The focus will be put on classic sociodemographic (such as age, gender and place of residence) as well as border region-specific factors (such as primary socialisation as well as commuting behaviour).

Survey procedure
The data shown here have been collected using an onomasiological survey, with one of the main aspects being the gathering of lexical variants on the level of ordinary and everyday language. The applied research design enables a targeted survey of lexis previously classified as relevant, without influencing the informants' response behaviour through linguistic stimuli. The survey was conducted within the framework of a "face-to-face-interview", generally within a private setting of the particular informant in order to test which variants dominate in the oral and everyday language use. The survey was conducted using exclusively image stimuli which were introduced with the following questions: "What is this? How would you call this?" In some cases, introductory questions were left out in order to prevent the interruption of the informants' "flow" of replies. Informants were also clearly made aware that the survey was not designed to find right or wrong answers, but to focus on the everyday language use as it would be common within families and among friends or acquaintances. The warming-up phase comprised four items. For the analyses conducted here, the first variant mentioned by informants was of particular importance as it has to be assumed that this one is predominantly retrieved in the mental lexicon and thus is used most often in everyday language use as well. In general, it should be assumed that speakers apply different variants depending on factors such as communication partner, the specific situation and other parameters. This is why it was explicitly noted when informants replied that they would use different variants in various situations.
Social and mobility data were collected from all informants in addition to those stemming from the onomasiological survey. Moreover, semi-structured interviews were conducted at selected locations (Austria: Wernstein am Inn, Oberkappel, Neufelden; Germany: Neuburg am Inn, Wegscheid).

Informants
For this study, data from a total of 154 informants from three German and four Austrian locations were analysed (see Fig. 1). The selection of the informants took place primarily according to the criteria of age and gender as will be outlined below. With respect to their residential location, we ensured that the younger informants had been living there since birth and the elderly since the start of their professional life. Based on dialectological and sociolinguistic apparent-time studies, the survey was initially designed for a younger (15 to 25 years) and an older (55 to 70 years) age cohort in order to find possible generation differences. In the following, the term age cohort refers to the distinction between the younger (15 to 25 years) and the older (55 to 70 years) informants. The term age group refers to another subdivision into 15-to 19-year-olds, 20-to 25-year-olds and 55-to 70-year-olds.
The first age cohort basically covers speakers who can be described as young. These informants are in a special transition phase from childhood to adulthood which is important for their identity formation (cf. Zimmermann 2012). This cohort was further subdivided during the survey into two age groups. The 15-to 19-year-old informants were usually pupils or students; most of the 20-to 25-yearolds had already engaged in employment. Both age groups have in common that they essentially grew up with the current open-border conditions in the region. Hence, they have never experienced border controls or had to change currency at the German-Austrian border.
The informants from the second age cohort had known the political border as a limitation on their mobility at a time when they already were adults and before the border was opened as a result of the implementation of the Schengen Agree-ment in 1997. These informants can be considered to have social security. Their everyday language usage can be classified as stable, whereas younger informants show more fluctuation in their language behaviour.
The aim was to collect data from at least five female and five male persons from each age cohort for each location. For Wegscheid (DE), Oberkappel (AT) and Neufelden (AT), however, no data are available for the older age cohort (55 to 70 years). Table 1 gives an overview of the informants' distribution by location, gender and age. One subgroup within the 154 informants are the 18 cross-border commuters. Table 2 gives an overview of their distribution by location, gender and age.

Items and stimuli
The selected items and stimuli can be assigned to the semantic fields of occupation, food and miscellaneous. The images used in the onomasiological survey are stimuli that correspond to the items that relate to the concepts. The images should evoke the different variants. When selecting the items, care was taken to ensure that at least one variant was considered to be typical for Germany/Bavaria and that one variant was typical for Austria on the level of everyday language (see section 2.1). The selection of items, however, represents a methodological problem. What is to be classified as a typical German/Bavarian or Austrian variant is actually an empirical question. This question can be answered for example by objective-(through onomasiological surveys) and/or subjective-linguistic (through language perception surveys) studies. However, such field work has not yet been done systematically and empirically. The study presented here can actually be seen as a step in answering the question. The methodological problem is that this investigation presupposes a certain typical distribution of the variants.
The specification of what is understood here as a typical German/Bavarian or Austrian variant was obtained from multiple sources. We classified, for example, so-called Austriazismen like Spritzer, Jause or Bankomat as typical Austrian variants (see Table 3). The following sources provided evidence for the timeliness and relevance of the assignment: a) research literature (cf. e.g . Scheuringer 1990;Scheuringer 1995: 49-53;Wiesinger 2014;Ammon 1995Ammon , 1997Ebner 1988;Zeman 2009), b) relevant dictionaries and word atlases (cf. e.g. Ammon/Bickel/Lenz 2016;ÖWB 2012;Ebner 2009;Eichhoff 1977Eichhoff -2000, c) pre-tests performed with students during the seminar "Language Contact" at the University of Passau in the summer term 2014, d) Internet search (e.g., http://www.oesterreichisch.net/, last accessed on 15.11.2017) and e) personal experiences of the explorers.
The following 23 items and variants were taken into account for the analysis carried out. Often, more items than listed below were elicited by the explorers. However, for our research only those were considered which were queried in all locations with all informants. To explain some problems, the item 'chimney sweep' will be briefly discussed as an example. For this item, the Austro-German state border does nowadays form a word boundary on the level of everyday language (cf. Scheuringer 1995: 55). The variant Kaminkehrer is dominant in Bavaria whereas Rauchfangkehrer is dominant in Upper Austria. At the level of the base dialect, the variant Rauchfangkehrer might also be used by older informants in Bavaria (cf. Scheuringer 1995: 54) whereas younger dialect speakers dominantly use Kaminkehrer in their base dialect. In the standard language, another variant that has become increasingly important due to the influence of the media is Schornsteinfeger. Although Kaminkehrer is primarily used in Southern Germany both in standard and in everyday language (cf. Scheuringer 2011: 204; Ammon/Bickel/Lenz 2016: 363), Schornsteinfeger is not completely unknown among the younger speakers. Thus, Kaminkehrer -as well as Metzger and Schreiner -are used in the standard variety in most parts of Southern Germany. In Upper Austria, the variant Rauchfangkehrer applies for all speech levels (base dialect, everyday language, and standard language). However, Kaminkehrer is also known in some regions in Western Austria such as Vorarlberg or Tyrol (cf. Wiesinger 2014: 150-151;Scheuringer 1995: 55). Scheuringer (1995: 52) calls the distinction of Kaminkehrer vs. Rauchfangkehrer at the German-Austrian border for the higher speech levels in the dialect-standard constellation (everyday and standard language) a "sprachliche Bereinigung der Staatsgrenze" which means that the political border is at the same time an isogloss for the variants. The use of the variant Rauchfangkehrer in Bavaria could, therefore, indicate language contact on the level of everyday language between Bavaria and Upper Austria and between dialect and standard, respectively. The use of the variant Kaminkehrer in Austria (except for Vorarlberg, Tyrol and parts of Salzburg) would probably have to be explained primarily by speaker contact on the level of microor mesosynchronisation.

Results
The data analysis was carried out with the statistics software SPSS. On the one hand, the analysis shows that almost all informants on both sides of the border name typical variants of the neighbouring country. On the other hand, multifactorial analyses of variance (Two-Way ANOVA) and unpaired t-tests reveal significant differences in some of the factors investigated here. Tukey and GT2 after Hochberg were used as post-hoc tests in contexts of the ANOVAs. For the data analysis, only the initial answers were considered. socialisation (hypotheses 1.1 and 1.2) Due to the primary linguistic socialisation associated with meso-and macrosynchronisation processes, it was assumed that the informants on both sides of the border name significantly more typical variants of their home country. An ANOVA shows significant differences that correlate strongly with the nationality of the informants. The Austrian informants (n = 83) use significantly more typical Austrian variants (M = 89.7 %, SD = 7.4 %) than the German informants (n = 71, M = 12.6 %, SD = 9.2 %). Accordingly, the German informants use significantly more typical German/Bavarian variants (M = 87.6 %, SD = 9.2 %) than the Austrian informants (M = 10.3 %, SD = 7.4 %). The specific location is also a significant factor in this context, F (5) = 12.4, p < 0.00.

Linguistic
For the German locations, a post-hoc test shows that only the difference between Neuhaus am Inn and Neuburg am Inn (MD = 5.5) is significant, p < 0.01. With regard to the use of typical German/Bavarian variants in the Austrian locations, the post-hoc test shows that informants from the locations Schärding vs. Oberkappel (MD = 13.8) and Oberkappel vs. Wernstein am Inn (MD = 9.6) perform significantly different from each other, p < 0.05. The hypotheses 1.1 and 1.2 can, therefore, be considered confirmed. However, it is noticeable that every German informant name at least one typical Austrian variant, whereas seven Austrian informants (six of them from Schärding and one from Wernstein) do not use any typical variant of the neighbouring country.

Age groups
Based on the assumption that younger informants are influenced more strongly by the (Federal) German media language and that (Federal) German variants have more prestige in the higher varieties in the dialect-standard constellation, we expected that the younger Austrian informants tend to use more typical Ger-man/Bavarian vocabulary than the older informants from Austria. No difference is expected between the two German age cohorts.
With regard to the three age groups, the ANOVA initially shows no significant difference for the use of typical German/Bavarian variants of Austrian informants, F (1) = 1.1, p = 0.29. However, the post-hoc test clearly shows that both the difference between the 15-to 19-year-olds and the 55-to 70-year-olds (MD = 6.5) is significant, p < 0.01, as well as the difference between the 20-to 25-year-olds and the 55-to 70-year-olds (MD = 5.2), p < 0.01. The difference between the younger age groups (MD = 1.3) is clearly not significant, p = 0.69. The informants from the two younger age groups use about twice as many German/Bavarian variants (M = 12.2 %, SD = 7.5 %) as the older Austrian informants (M = 6.4 %, SD = 5 %). The result is also validated by a t-test between the two age cohorts, t (81) = 3.5, p < 0.01. This also confirms hypothesis 2.1 for the data set.
For the German informants, the ANOVA also shows no significant difference between the three age groups, F (1) = 1.3, p = 0.26. However, this time the posthoc test reveals a significant difference between the two younger age groups (MD = 4.9), p < 0.01. There is also a significant difference between the 20-to 25-yearolds and the 55-to 70-year-olds (MD = 4.5), p <0.01. In contrast, the difference between the 15-to 19-year-olds and the 55-to 70-year-olds (MD = 0.4) is clearly not significant, p = 0.96. A t-test confirms that there is no significant difference between the age cohorts 'young' and 'old', t (69) = 1.01, p = 0.32. This also confirms hypothesis 2.2. The younger German informants do not conspicuously use more typical Austrian variants (M = 13.3 %, SD = 10.1 %) than the older informants (M = 10.8 %, SD = 6.9 %).

Gender differences
To answer the question whether gender has an influence on the use of lexical variants in the border region, the Austrian and German informants are compared separately. Both an ANOVA and a t-test show no significant difference between male (n = 41) and female (n = 42) Austrian informants, F (1) = 0.08, p = 0.78. This also applies for the German informants. There is no significant difference regarding the factor gender, F (1) = 0.44, p = 0.51. The answer to research question 1 is that men and women behave in the same way when confronted with the stimuli. Men and women use typical variants of the neighbouring country in the same way.

Discussion and prospects
Of the 154 informants, 147 (95.5 %) use at least one typical variant of the neighbouring country (the use of one variant corresponds to 4.3 %). The rare use of typical German/Bavarian variants by Austrian informants (M = 10.3 %, SD = 7.4 %) and typical Austrian variants by German informants (M = 12.6 %, SD = 9.2 %) is surprising given the profound political changes in the last 20 years. For example, Scheuringer (cf. 1995: 46) predicted major linguistic changes in the border region regarding the imminent economic and political transformations in Europe. From today's perspective, the Monetary Union of 2000, the Schengen Agreement of 1997 and various joint EUREGIO projects between border communities should have led to a transformation of the language situation in the border region (cf. Bülow/Schifferer/Dicklberger 2015). This is why some linguists also expected a Europe of the regions in terms of language development in the border regions. Linguistically speaking, a Europe of the regions does not seem to be developing so quickly. National identities continue to be projected consciously and unconsciously into linguistic identities. These findings underline how much the mental borders affect language usage and language attitudes. The results of this study support Auer's (cf. 2004: 177-178) résumé: A Europe of the regions across borders has no linguistic reality; it remains segmented into the scope of national standard varieties. Also the results presented by Kleene (cf. 2017) confirm the findings shown here: The majority of her informants from Passau and Schärding stated that they perceived the state border as a language border. Furthermore, Kleene's informants could easily distinguish speech samples from both sides of the border in a listener judgment test (cf. Kleene 2017).
It should not be concealed that the data presented do allow for other interpretations as well. The younger informants from Austria, for example, use significantly more German/Bavarian variants than the older informants from Austria. This has to do with the cross-border media consumption as well as the tendency for their everyday language to be nested in a higher variety of the dialect/standardconstellation. Their everyday language tends to be closer to the standard pole.
Furthermore, the data indicate that speaker contact across the border is an essential factor. The cross-border commuters (whether from Germany or Austria) use significantly more typical variants of the neighbouring country than nonborder crossers. People who cross the state border very regularly (to work or to go to school) synchronise themselves linguistically. The state border is not a border that prevents synchronisation processes because it is not a barrier for mobility, but because the regions on both sides of the border have developed in a complex situation. Apart from the speakers' attitudes, the most important factors certainly are the respective national economic, administrative and educational characteristics. These characteristics will keep the state border a linguistic border as well. Only when micro-, meso-and macrosynchronisation processes intermesh on a crossborder level, will convergence processes across the state border be conceivable.
However, one has to keep in mind that the items we examined represent a particularly salient part of the everyday vocabulary. The results that Pickl et al.
However, homogeneity and continuity in using lexical variants could not be found in this study.

Conclusion
The present study shows that clear differences in using specific lexical variants exist on both sides of the border. It plays a crucial role whether the informants were linguistically socialised in Austria or in Bavaria. The factor age is relevant at least for the Austrian informants. The younger informants from Austria use significantly more typical German/Bavarian variants than the older generation. For the two German age cohorts, no age differences can be found. Mobility has also proved to be a decisive factor. Informants working or going to school in the country across the border use significantly more typical variants of the neighbouring country. Where there is intensive cross-border speaker contact, micro-and/ or mesosynchronisation processes have an effect on the language usage. These specific synchronisation effects should not be underestimated for the explanation of language change in border regions.
According to the results of the study presented, it is obvious that the linguistic situation at the Austro-German border around Passau also functions like other German borders: as a mental border between two countries in which German varieties are spoken.