Gentrification of peri-urban spaces in France – the surroundings of Nancy

Abstract The process of gentrification in the peri-urban districts of French cities has scarcely been touched upon in recent research, which has hitherto seen the phenomenon as typically associated with core urban areas. The tendency has been to view the periphery through the lens of the social crisis of the banlieues. In contrast, the present article focuses on gentrification in the metropolitan area of Nancy (Grand Est region) as a development that also plays a role in municipalities around major cities and especially around regional metropolitan centres. Starting with a survey of current research approaches, the article first pinpoints some gaps and methodological imbalances that need to be tackled, before embarking on the case study of peri-urban Nancy. Statistical data and empirical surveys in the form of qualitative interviews indicate how Nancy's peri-urban districts have developed a logic of separation, exclusion and social decoupling – typical features of gentrification – particularly in connection with the construction of new single-family houses as a supplement to existing residential stock. Key questions here concern individual motives for choosing a particular residential location, and the creeping "segregation from above" that accompanies this process. The image of France's peri-urban spaces that arises from this study stands in explicit contrast to the received, markedly negative connotations of the "urban periphery".

1 Périurbain has established itself in French planning and academic research as a designation for the spaces between the banlieues and the rural areas. This usage, rather than the English terms "suburbs" or "suburban", is reflected in the present article. For a detailed discussion see Cusin/Lefebvre/Sigaud (2016) and . of the genesis of urban/rural hybrids (Weber/Kühne 2017;Kühne 2018).
Against this background, this article aims to establish an analytical link between two socio-urban phenomena usually isolated from each other in research: on the one hand, the process of gentrification, designating a movement of displacement of the working and/or lowermiddle by upper-middle and upper social classes 2 , accompanied by changes in housing, or more generally changes of a neighbourhood or even an entire city (Glass 1964;Hamnett 2003;Bridge 2014;Doucet 2014;Lagendijk/van Melik/de Haan et al. 2014); on the other hand, the process of peri-urbanization, which goes hand in hand with urbanization encroaching ever further on rural territories through the paradigmatic residential type of the (detached) single-family house with garden (Jaillet 2004;Bretagnolle 2015;Cusin/Lefebvre/Sigaud 2016;. Our article focuses on the connection between residential choices made by wealthy households wishing to live in single-family houses surrounded by large wooded grounds and the form taken by gentrification in peri-urban areas. Taking the example of the metropolitan area of Nancy (Eastern France), we examine the complex relation between gentrification and residential location choices in peri-urban spaces. In the French context, this approach opens up a new research perspective. Peri-urban areas have at least in some cases become the stamping ground of a specifically bourgeois logic. Nevertheless, the relationship between gentrification and peri-urbanization may at first sight seem paradoxical, since the "urban peripheries" have often been associated with poverty or even in some cases with "relegation zones" (Delarue 1991), and the centre-periphery dualism is, in fact, proving highly resistant, as evidenced by Guilluy (2014Guilluy ( , 2016. The forms of gentrification that we have identified in the peri-urban area of Nancy do not, of course, possess Gentrification of periurban spaces in France -the surroundings of Nancy Gentrifizierung in periurbanen Räumen Frankreichs -am Beispiel von Nancy 259 all the characteristics of the gentrification of more central city areas. In particular, we do not find the social pattern highlighted by Pattison (1977: 168 ff.), who distinguished different populations involved in gentrification and saw these as the result of successive waves of influx by various distinct middle class segments. Initially launched by a small group of intrepid people (the "invaders"), then carried forward by a more risk-conscious clientele ("pioneers/newcomers"), and finally taken over -in some cases after the intervention of public authorities e.g. in labelling "invaded" territories "historic" -by upwardly mobile members of the middle classes ("yuppies"), gentrification in its classic and exemplary form could thus be described as a progressive phenomenon that is at the same time individual, cultural and political.
Peri-urban gentrification is a specific process, if only because in peri-urban spaces gentrification rarely translates into the rehabilitation of the existing built framework , but more often into the construction of new, detached, luxury houses on large plots. A link can possibly be seen here with the construction of housing for the upper-middle classes, particularly in the case of the demolition/reconstruction or regeneration of urban wastelands, which some commentators call "new-build gentrification" (Davidson/Lees 2005;Davidson/Loretta 2010;Rérat 2012). The research issue, then, is to adapt the concept of gentrification to today's peri-urban realities. In other words, it is a question of revisiting the concept in order to investigate its role and that of other, similar developments in the current sociourban evolution of peri-urban spaces.
In pursuit of this goal we first look briefly at the current state of research on gentrification both in innercity areas of France and in the banlieues, taking account of wider peri-urban developments, in particular the role of the single-family house in this context (Section 2). After explaining the empirical methodology to be applied in a case study of the metropolitan area of Nancy (Section 3), we go on to investigate how rural districts become absorbed in processes of gentrification that cause farreaching change in peri-urban social and physical environments (Section 4). A concluding summary (Section 5) highlights the main results of the inquiry.
2 State of the art and conceptual approaches: staking out the new research field of gentrification and peri-urbanization

Gentrification of cities and banlieues
For more than thirty years certain old city-centre districts of France have witnessed the increasing incursion of members of the middle and upper-middle classes (see Préteceille 2007;Pattaroni/Kaufmann/Thomas 2012). Already observed in an isolated and fragmented way in the 1960s in London, this population and socioeconomic movement became known as "gentrification", a neologism invented by the Marxist sociologist Glass (1964). But gentrification has increasingly spread from run-down central areas to other urban and even rural spaces (see e.g. Phillips 1993;Smith/Phillips 2001;Atkinson/Bridge 2005;Lees/Ley 2008;Stockdale 2010;Smith/Higley 2012). And the gentrification process has at the same time evolved in its forms and concomitant research approaches (see Hackworth/Smith 2001;also e.g. Bridge 2003;Doucet 2014;Lagendijk/van Melik/de Haan et al. 2014). The concept has been extended to other elite-forming processes involving many different actors and including multiple forms of transformation of used space over and above traditional working-class residential areas. From this point of view, gentrification results from a complex "social game" in which sedentary and mobile populations live side by side, where population movements, planning decisions, stakeholder strategies, political aims and specific ways of living and cohabiting between different social groups are intertwined and analysed (Ley 1986;Smith 2002;Bidou-Zachariasen 2003;Uitermark/Duyvendak/Kleinhans 2007).
While gentrification mechanisms are undoubtedly complex -today one would speak analytically of a "broader concept" (Clark 2005;Slater 2006) -many researchers agree on two underlying elements. Gentrification is increasingly seen both as a transformation of the social composition of the residents of a neighbourhood -more precisely the replacement of working classes by salaried upper and middle-income layers (Glass 1964;Pattison 1977;Dansereau 1985; Doucet 2014) -and as a hybrid process combining the rehabilitation and appropriation of (and coterminously investment in) housing across entire working-class neighbourhoods by these new population categories 3 (Hamnett 1991;Hamnett 2003). As Clark (2005: 258) has put it: "It does not matter where, and it does not matter when. Any process of change fitting this description [of population shift] is, to my understanding, gentrification".
In France, the concept of gentrification has so far been used to study transformations of working-class areas of inner cities, and more recently of communes (municipalities) 4 in the banlieues (Préteceille 2007;Pattaroni/Kaufmann/Thomas 2012;Chabrol/Collet/ Giroud et al. 2016). Gentrification can also affect a whole city, for example Paris, which is gradually undergoing gentrification across its entire administrative limits (Clerval 2010;Clerval 2013). And, by a contagious logic, this process has now extended to the first ring of banlieues, starting with the communes of Levallois-Perret, Montreuil and Pantin (Collet 2008;Marchal/ Stébé 2012;Albecker/Fol 2014;Albecker 2015;Marchal/ Stébé/Bertier 2016;Weber/Kühne 2017). In this article we apply the concept of gentrification to peri-urban spaces as instances or pastiche-like complexes of the urban/ rural hybrid (Weber/Kühne 2017;Kühne 2018). This fills a gap, as French research on gentrification has not yet considered the peri-urban dimension, nor (conversely) has research on peri-urbanization incorporated the aspect of gentrification.

The single-family house in France as the motor of peri-urbanization
Social science research tends to see peri-urbanization as rooted among other things in the middle class desire for home ownership (Lambert 2015;Hesse/Siedentop 2018: 101). In France, too, this is a continuing success story. Today 59% of French households own their homes, compared with 35% in 1954 and 50% in the early 1980s; in absolute figures, among France's 34.5 million homes, 19.3 million are owned and lived in by a single family (Marchal/Stébé 2018: 42). Moreover, in 2016 singlefamily houses represented 41% of new construction.
3 "Gentrifiers" in "trendy" cities and suburbs (banlieues) have, in France, often been called bobos (bourgeois-Bohemians) or hipsters by the media. 4 The French term commune has been retained throughout this article: as the smallest autonomous local government body, the commune is managed by a maire (mayor) and a conseil municipal (municipal council), and hence translates roughly as "municipality". It can vary widely in size and is in many cases far smaller than municipalities (or other small local government entities) in Englishspeaking countries.
Thus over the past fifty years many French families have acquired ownership of their own house and garden. Moving ever further away from the cities, single-family houses have spread to rural areas previously spared by urbanization, but which have now, in the course of this development, undergone profound changes (see Jaillet 2004;Bretagnolle 2015;Damon/Marchal/Stébé 2016).
There are many reasons why the French are attracted to the single-family house. For half a century, many studies have examined the motivations put forward for attachment to the idea of individual home ownership As early as the 1960s, the Institute of Urban Sociology highlighted the extent to which the pavillon de banlieue (single-family suburban house) was explicable in terms of a cultural habitus rather than of individualist ideologies (Raymond/Dezès/Haumont et al. 1966). Later, in the early 2000s, Jaillet (2004) argued that the peri-urban single-family house is like a "mosaic space" in which the different socio-professional strata that comprise the intermediate layers of society settle in order to benefit from attractive land prices combined with a tranquil living environment and visible social success. In the same vein, Rifkin (2000) underlined the extent to which the purchase of a single-family house in the suburbs is an "entry ticket" to a residential club, a group that benefits from a social universe whose features range from a specific physical landscape to the promise of security this entails. Based on research conducted in France, Charmes (2005Charmes ( , 2011 has similarly shown that living in a private housing development means above all inhabiting a social space with discreet neighbours and enhanced well-being and security. In this context, he uses the metaphor of the "closed club". More recently, on the basis of perspectives inherited from the study of social stratification, Cartier, Coutant, Masclet et al. (2008) have highlighted how access to a peri-urban single-family house, in the eyes of people rooted in the working class, substantiates a distance from their modest social origins and at the same time allows them the opportunity to live "like everyone else". For Donzelot (2009) this indicates the extent to which the single-family house supports self-assertion and social distinction. Among other things, Donzelot (2009) notes that the owners of single-family houses, anxious to defend their tranquillity, their neighbourhood and, by extension, their quality of life, keep their distance from other residents -especially those from impoverished social housing estates -who they fear may invade and disturb their "little happiness". Other authors (e.g. Maumi 2008) have shown that the choice of a house in the peri-urban ring is also based on the proximity of nature and the countryside (on the significance of "landscape" see Kühne 2018; Kühne/Weber 2018). The ideal sought in many cases corresponds to that of a city set in the countryside -or at least a residential environment, if not in harmony, then as close as possible to nature. A fundamental point in this context is that although the socio-economic profiles of inhabitants of peri-urban areas have diversified (Keil 2018), the "home of one's own" can be regarded as a common anchor.
The research carried out over the past fifty years has provided a number of interesting insights into gentrification processes, as well as into the global dynamics of peri-urbanization (for post-suburbia in Europe see also Bontje/Burdack 2011, for urban peripheries Harris/Vorms 2017), but it has not yet (in the French context) established links between residential choices in peri-urban space and the logic of gentrification (see Damon/Marchal/Stébé 2016). In fact, on closer examination, only the rare authors who worked during the 1960s and 1970s on processes of "rurbanization" highlighted the displacement of rural populations by new middle-class homeowners (Chamboredon 1985). But this is not, strictly speaking, a matter of gentrification, since the urbanization of the countryside at that time was part of a reconfiguration of French society that saw the middle classes and part of the working class settling in villages located on the outskirts of cities (Kayser 1981).

Methodological choices: the case study of Nancy
For the analysis of gentrification processes in the periurban spaces of France, we will focus here on the metropolitan area of Nancy. Situated in Eastern France, the city of Nancy, with its suburbs, has 220,000 inhabitants. Its aire urbaine -an INSEE 5 category comprising the city centre, the banlieues and the peri-urban ring -has 445,000 inhabitants. 50% of the population of Nancy's aire urbaine lives in a peri-urban area, which makes it an exemplary French city in terms of peri-urbanization. By explicitly not focusing on the Greater Paris area, we also avoid the challenge of the French special case and the Our corpus consists of 32 peri-urban communes 6 (first and second ring of peri-urbanization) located beyond the banlieues7 but whose centres are not more than 11 km from the city centre of Nancy ("périurbain proche") -a firmly applied selection criterion tailored to the situation under analysis (see Figure 2) 8 . The largest of these communes, Essey-lès-Nancy, has just over 8,500 inhabitants and the smallest, Dommartin-sous-Amance, 250.
Our methodology is organized around two central components. The first, based on INSEE statistics 9 relating to the 32 communes, consists of isolating two potential indicators of gentrification over the period 1999-2014 10 : (1) the evolution of socio-professional categories (INSEE PCS); (2) development in the number of higher education graduates. The second component inquires into the residential aspirations and justifications of upper and upper-middle class residents who have chosen to settle in one of the communes where the gentrification process is more advanced. Nine such communes in the process of advanced gentrification have been identified in the course of our statistical analysis: Bouxières-aux-Chênes, Cerville, Chaligny, Eulmont, Fléville-devant-Nancy, Lenoncourt, Lupcourt, Pulnoy, and Velaine-en-Haye (see Figure 2). In order to grasp the (gentrifying) motives of their owner-occupiers within the overall context 6 The 32 communes in our corpus are (in alphabetical order): Agincourt, Amance, Art-sur-Meurthe, Bouxières-aux-Chênes, Bouxières-aux-Dames, Cerville, Chaligny, Champigneulles, Chavigny, Dommartemont (already for a long time a richer and gentrified commune), Dommartin-sous-Amance, Essey-lès-Nancy, Eulmont, Fléville-devant-Nancy, Frouard, Heillecourt, Houdemont, Laîtresous-Amance, Laneuvelotte, Laneuveville-devant-Nancy, Lay-Saint-Christophe, Lenoncourt, Liverdun, Ludres, Lupcourt, Maron, Pompey, Pulnoy, Saulxures-lès-Nancy, Seichamps, Velaine-en-Haye, and Velaine-sous-Amance. 7 We refer to the banlieues of Nancy as defined by INSEE: all communes directly adjacent to the city centre of Nancy. 8 We chose this perimeter in order to conduct our research in the first peri-urban ring, which is also the oldest, and which is now subject to strong land pressure. 9 Our reference statistical population does not include persons under 15 years of age, or students in secondary or tertiary education. Our interest is in people in gainful employment likely to settle in peri-urban Nancy. However, in order to understand the overall dynamics of these peri-urban communes we have included retired people in our sample, as they form an important population segment (for a comprehensive study of this segment in France see Berger/Rougé/Thomann et al. 2010). 10 Our focus is on working-class districts engaging in or on the way to gentrification, which explains our decision not to go back beyond 20 years. of residential and life trajectories (Authier/Bonvalet/Lévy 2010), we conducted eighteen semi-structured one-anda-half hour interviews at the household homes, i.e. two interviews per commune. The households were chosen using first contacts in the snowball principle, supported by the local administrations, and were characterised by having settled in the peri-urbain area near Nancy after 1999.

Gentrification of urban peripheries: Nancy's first and second peri-urban rings
What connections can be established between gentrification, peri-urbanization and the spread of owner-occupied single-family houses, and what characterizes local environments that are undergoing such a development? Beginning with a quantitative survey, we will go on to present a qualitative analysis of the motivations underlying processes of peri-urban gentrification. With the greater Nancy area, we are looking at an area in the Lorraine strata in a wide valley basin along the Rivers Meurthe and Moselle, as well as the Rhine-Marne Canal -and thus a landscape that is usually regarded as particularly scenic.

Gentrification processes traced by statistical data (INSEE): growth in density of management staff and higher education qualifications
If we look at the evolution of the population of executives and higher professionals (cadres et professions intellectuelles supérieures) within our corpus, we note that 26 of the 32 analysed communes recorded an increase in this official INSEE category between 1999 and 2014. 11 Some communes have seen their professional and managerial populations double or even quadruple during this period. In Lay-Saint-Christophe, for example, the ratio of senior managers rose from 8.1% in 1999 to more than 14% in 2014, Laître-sous-Amance grew from 9.3% to nearly 19%, Lupcourt recorded an increase from 4.9% to 9.2% over the same period. But 11 All the statistical data cited in the text comes from the French population censuses (Recensements généraux de la population, RGP) conducted in 1999 and 2014.
the commune with by far the strongest increase in its population of executives and higher professionals was Lenoncourt, where this group grew from 5.1% to 19.8% over the same period. By way of comparison, it should be recalled that during this period the increase in the professional/executive category within both the former Lorraine region 12 and metropolitan France was only 4 points, rising from 10% to 14% and from 13.1% to 17.1% respectively (Berrard 2013;INSEE 2016).
Turning to the statistical category of mid-range professions (professions intermediaries), we note that nine communes of the 32 recorded a consistent decrease in this category. This was due to two phenomena: first of all, a strong increase of executives and higher professions in communes where land and property prices were increasing (as in Laître-sous-Amance and Dommartemont); secondly a marked aging of the population, as in Dommartin-sous-Amance and even more so in Fléville-devant-Nancy. This was reflected in a marked increase in the number of pensioners who still occupied their own homes and thus did not allow other population categories, particularly the middle classes, to settle in higher numbers. The statistics are revealing: in 2014 Dommartin-sous-Amance recorded 35% retirees and Fléville-devant-Nancy nearly 40%. During the period 1999-2014, when these two communes saw their number of retired people strongly increase, they also experienced a fall in their economically active middleclass populations (Dommartin-sous-Amance 11% to 7% and Fléville-devant-Nancy 17% to 14.5%).
At the same time, other communes such as Agincourt, Cerville, Chaligny, Eulmont, Lupcourt and Velaine-en-Haye continued their process of gentrification with a different dynamic. These six communes experienced a simultaneous increase in the categories "mid-range" and "executive and higher" professions (see Figure 1). Compared with the statistics for Lorraine (+1.5 points: 23% to 24.5%) and France as a whole (+2.5 points: 23.1% to 25.6%) over the same period, the consistent increase in the "mid-range professions" category in periurban Nancy is remarkable.
If we observe the evolution of the category of semiskilled employees (employés) the results of our statistical analyses again reveal a process of gentrification in many communes situated in Nancy's first and second periurban rings. Generally speaking, 23 communes have seen a decline in this category, six have stabilized, and three have seen a slight increase. Among the communes with the largest decreases are Lupcourt, which recorded a drop of 9 points (from 19.7% to 10.7%), Cerville (down 8.8 points from 27.9% to 19.1%), Bouxières-aux-Chênes (down 8.5 points from 21% to 12.5%), and Fléville-devant-Nancy (down 8 points from 20.7% to 12.7%). These statistical trends are all the more interesting in terms of the problem of gentrification in Nancy's peri-urban areas, since during the same period Lorraine experienced a 1.5 point increase in the semi-skilled employee category (from 23% to 24.5%), while in France as a whole the general trend was towards stability.
Finally, the development of the statistical category of unskilled employees (ouvriers) between 1999 and 2014 indicates the occurrence of a process of gentrification in most of the 32 communes in our corpus, 19 of which experienced a fall in this category, without any recording an increase. In other words, although we are dealing with "urban peripheries", there is no working class consolidation and, by extension, no sign of a precariat developing in the communes in Nancy's first and second peri-urban rings. Among the communes most affected by the downward trend in the unskilled worker category are (again) Cerville (13.5% to 5.2%), Amance (13.3% to 6.1%), Dommartin-sous-Amance (11.2% to 5%), Chaligny (19% to 13%), and Frouard (17.4% to 12.4%). These figures accord with regional and national trends where, during the same period, the decline in unskilled workers was 6 and 5 points respectively.
Looking more systematically at this data, we can distinguish two types of commune that have undergone significant change: A first group of older, upscale communes have consolidated their bourgeois evolution over the past twenty years. As we have already seen, this is the case with Dommartemont, historically well-off and "clubby", which continues to welcome an ever growing number of executives. In line with this development, the number of unskilled workers has stagnated at around 3.5%, although in the centre of this village clinging to the hillsides surrounding Nancy there are a few modest rental dwellings where low-income households live. Within this same group, we find two other communes, Laître-sous-Amance and Lay-Saint-Christophe, which have also experienced growth in the number of executives (18.7% and 16.7% respectively). Historically, these two communes have also had relatively small numbers of unskilled workers (no more than 8%), and those numbers are continuing to drop. As in the case of Dommartemont, we cannot speak here of a "classical" form of gentrification, as the working classes have not been predominantly displaced by upper and uppermiddle social categories. The initial weakness of the unskilled categories in all three communes rules out such a process. Other communes, however, have undergone more typical processes of change.
The second group has recently gentrified and, as such, records both a growth in the executive, higher professional and mid-range professional classes and an "mid-range" professions "executive and higher" professions equally marked decline in the semi-skilled and unskilled worker classes. The two communes of Chaligny and Lenoncourt can be taken to exemplify the trend. For Chaligny, between 1999 and 2014, while the executive/ higher professional and mid-range professional categories grew by 3.9 points (4.0% to 7.9%) and 6.2 points (11.3% to 17.5%) respectively, the semi-skilled and unskilled categories decreased by 5.5 points (22% to 16.5%) and 6 points (19% to 13%) respectively. As far as Lenoncourt is concerned, the statistics indicate a more salient gentrification process: between 1999 and 2014 the executive/higher and mid-range professions increased by 14.7 points (5.1% to 19.8%) and 9 points respectively, while the semi-skilled and unskilled categories declined by 5.7 points (19.2% to 13.5%) and 3.6 points (14.1% to 10.5%) respectively. These two communes appear, then, to exemplify par excellence a cross movement of gentrification, where a decline in modest socio-economic categories goes hand in hand with a growth in wealthier categories.
Looking now at the development of higher education qualifications in peri-urban Nancy, the statistical data from INSEE reveals a strong increase in levels in line with the gentrification process already noted in specific communes. Between 1999 and 2014, all the communes surveyed experienced a marked increase in educational levels. In some communes, the increase even reached 22 points: in Lupcourt the graduate population surged from 16.2% to 38.4%. Other communes recorded an increase of over 15 points, starting with Lenoncourt (14.5% to 31.1%), Velaine-en-Haye (25% to 41.2%), and Eulmont (24% to 39.6%). As for the three traditionally well-off communes of Dommartemont, Laître-sous-Amance and Lay-Saint-Christophe, all three also experienced strong growth in terms of people with tertiary education qualifications, Dommartemont recording 50% and the other two communes almost 40% -figures that again confirm their status as socially and economically privileged locations. This being said, it should be noted that communes such as Eulmont, Lupcourt and Velaineen-Haye, which originally had a lower rate of university graduates, have now caught up with historically well-off communes such as Laître-sous-Amance and Lay-Saint-Christophe. In the same period Lorraine's graduate population grew by only 5 points, reaching a quota of 22% in 2014, when France as a whole had 30% in this category. In other words, a number of communes in Nancy's first and second peri-urban rings have seen a significant increase in their graduate populations over the past twenty years, and one that is well above the regional average: 28 of the 32 communes in our sample have a university graduate rate above the regional average (22%); 19 communes have a graduate rate above the national average (30%).
The overall picture arising from this data is as follows. Apart from the three historically posh "residential-club" type communes (Dommartemont, Laître-sous-Amance and Lay-Saint-Christophe) classical gentrification, with strong growth of the higher professional and managerial classes combined with an increase in the proportion of university graduates and a matching decrease in the incidence of less qualified categories, is marked in nine communes in our sample (see Figure 2 13 ). Our statistical analysis, then, shows recognizable trends. However, without some evidence of what these figures actually mean for the people who live there, they remain abstract indices. In order to explore the experience of gentrification more concretely, we have, therefore, conducted semi-structured interviews in the nine communes in peri-urban Nancy most deeply involved in a recent process of gentrification ( Figure 2). These are: Bouxières-aux-Chênes, Cerville, Chaligny, Eulmont, Fléville-devant-Nancy, Lenoncourt, Lupcourt, Pulnoy 14 and Velaine-en-Haye.

Motives for selection of residential location: social exclusiveness, rising property prices
The inhabitants we interviewed all belonged to the upper and upper-middle social classes. Analysis of the interviews revealed a number of converging factors, all of which imply some sort of demarcation, regarding respondents' aspirations and rationales for their choice of residential location. The households interviewed cited above all the village, neighbourhood or specific development where they lived, highlighting its amenities (e.g. woods, public park, square, golf course, riding school, river). Then they mentioned the large plots of land available for the construction of single-family houses (between 700 and 1,500 sq. m.). Finally, they cited the importance of neighbours who shared their way 13 The basis for inclusion in the category of communes undergoing gentrification is an increase or decrease in the selected categories that is above the national average. 14 Although this commune of 4,500 inhabitants located 11 km east of Nancy has on the whole only undergone a weak process of gentrification, it is known within the urban/rural hybrid of Nancy for its landscape quality, due in particular to a large golf course adjoining several housing developments.
Gentrification of periurban spaces in France -the surroundings of Nancy Gentrifizierung in periurbanen Räumen Frankreichs -am Beispiel von Nancy 265 of living and thinking. These factors will now be analysed in greater detail as aspects of peri-urban gentrification.

A decisive and secure site logic
The vast majority of the households we interviewed emphasised the privileged quality or "site logic" of their residential location, with its topographical and environmental amenities, as well as its attractive heritage and its functional urban planning. The small town of Chaligny (under 3,000 inhabitants, 12 km from downtown Nancy) is prototypical in this respect -in particular the neighbourhood located in a green setting along a hillside facing due south and enjoying an unobstructed view over the valley of the Moselle. Here Madame A 15 (45 years old, nurse, married, two children) welcomed us spontaneously on the doorstep with the words: "Look at this magnificent 15 All respondents' names are anonymized for data protection reasons.
view of the Moselle! What a nice place we have here, surrounded by nature! My husband and I wouldn't leave here for anything!" 16 Similarly Monsieur and Madame B (retired couple, former public service executives), residing in a cul-de-sac development in Pulnoy (4,500 inhabitants, 11 km east of Nancy), persistently invited us to look through the large bay windows of their singlefamily house at the perfectly maintained fairways and greens of the golf course with its reservoir and, in the background, the hill marking the beginning of the national forest: "Oh yes, it's true that this view of the golf course was decisive in our choice to settle here" says Monsieur B and adds: "We knew that every morning we would see this landscape, and we know how lucky we are [...]. What a pleasure to eat on the terrace with this well-maintained park before us [...], we enjoy it every day and we are not 16 Original: "Regardez donc cette vue magnifique sur la Moselle! Qu'est-ce qu'on est bien ici avec toute cette nature! Pour rien au monde, mon mari et moi, on ne partirait d'ici!" the ones who have to look after it." 17 It is evident here that to be able to afford such a location is one of the main driving factors in the move to a peri-urban location, where the stereotypically beautiful landscape assumes the additional quality of a "home environment", with all the emotional and demarcational connotations of that concept (see Kühne 2018).
One of the couples we met in Eulmont (1,000 inhabitants, 10 km from downtown Nancy) focused more on the historical heritage (patrimoine) of their adopted village. Monsieur C (47 years old, engineer, married, one child) was full of praise for the 18th century château whose entrance could be seen from his large residence in the centre of the commune, a house that was itself perfectly in keeping with the landscape. The young couple who had moved into a small development at Lupcourt (442 inhabitants, 12 km from downtown Nancy) three years ago also justified their move to this commune by emphasizing its heritage (church with Romanesque tower, château, remains of an 18th century priory): "Here, you can really envisage having children, because wherever you walk, you find nice views [...] there's the château and our little church. And frankly, pushing a buggy here is cool!" 18 (Madame D, 30 years old, veterinary surgeon, one small child). In Fléville-devant-Nancy (2,300 inhabitants, located 12 km south of downtown Nancy), alongside the historical heritage (notably a 16th century château), the well-kept public spaces and the cleanliness of the village, as well as the proximity of cultural and commercial amenities (multiplex cinema, hypermarket with shopping malls), were cited by two respondent households as an asset and a determining element in their residential choice. Monsieur E (36 years old, general practitioner, married, two children) was explicit on these points: "We were charmed by the village. It's clean, and there are lots of facilities to accommodate tourists visiting the château and to make the locals feel good." 19 He explained how the hypermarket, located 4 km away, made their 17 Original: "Ah oui, c'est sûr que cette vue sur le golf a été décisive dans le choix de s'installer ici […] On savait que tous les matins, on allait avoir ce paysage devant les yeux, et on sait notre chance […] Quel plaisir de manger sur la terrasse avec ce parc toujours entretenu […], nous on en profite tous les jours et ce n'est pas nous qui l'entretenons." 18 Original: "Ici, on peut envisager d'avoir des enfants car partout où on peut se balader, on trouve des vues sympa […] il y a le château et notre petite église. Et franchement prendre la poussette ici, c'est cool!" 19 Original: "Nous, c'est clair, on a été séduit par le village. C'est propre, et il y a beaucoup d'aménagements pour recevoir les touristes du château et pour que les habitants s'y sentent bien." lives easier for shopping and for the children's leisure activities. It is evident here that the middle/professional class choice of a peri-urban living location -and with it an accelerated gentrification process -is also motivated by the presence of thoroughly urban amenities.
The inhabitants of these communes, then, are aware that they benefit from a sort of "frozen" nature -nature as in a painting or museum, without any of the disadvantages of reality -a nature of châteaux, landscapes, golf courses and woods whose maintenance is not their responsibility, since it is maintained either by private owners or by public authorities. In this respect, it is important to stress that it is not Nancy's first and second peri-urban rings as such that are being gentrified, but rather specific communes which, through their heritage of architectural and environmental assets have established themselves as peri-urban niches of gentrification. And one of the key qualities of such niches is a "beautiful" landscape.
Generally speaking, the interviews reveal the extent to which site logic operates here in the residential choices of the wealthier households: the communes most deeply involved in the process of gentrification are characterized by architectural, patrimonial and landscape qualities which, in the eyes of the gentrifiers, guarantee both a pleasant living environment and a sound long-term financial investment. In contrast with the situation in popular old town centres, peri-urban gentrifiers run little or no financial risk. The initiators of peri-urban gentrification in the first and second periurban rings around Nancy -which is in any case a recent (and sometimes merely inchoate) phenomenon -is not the work of daring initiators with a taste for risk. Here we are far removed from the concept of "invaders" starting an improbable movement in neglected and stigmatized neighbourhoods in which no one previously wanted to invest. Peri-urban gentrification of the kind described in these paragraphs is informed by widely accepted positive associations of environmental beauty and tranquillity which carry a real estate price tag that not everyone can afford -a connection with (undoubtedly conscious) implications in terms of social exclusiveness.

A sought-after and valued community of neighbours
Alongside this site logic, interviewees stressed the importance of the social environment. This is above all concerned with the search for neighbours who one knows in advance will tacitly share certain values and points of view on the education of children and respect for others, on the maintenance of gardens and the cleanliness of public spaces. It ensures a social environment that is reassuring, soothing and, on occasion, accommodating or even vigilant. Madame F (39 years old, human resources manager in industry, divorced, one child), for example, absolutely wanted to keep her house after her divorce, for fear of finding herself in a different social environment where she would know no one. It must be said that she had established over the last ten years many relations, if not friendly then at least cordial, in her neighbourhood, located at the edge of Velaine-en-Haye (1,800 inhabitants, 14 km from downtown Nancy): "You know when I got divorced, I was able to count on two of my neighbours who looked after my child, and whom I completely trust. They took time to listen to me when things went wrong. So I wanted to keep my house at all costs, because here we form a small world with the neighbours. Even the ones I don't know much about, we say hello and everything goes well." 20 The familiar, supportive environment of "home" is a value to be unconditionally preserved.
For other households what is important is involvement in local and citizen initiatives: joining the neighbours for a meal, preparing the annual garage sale (vide-grenier), or getting involved in the renovation of historical heritage -for example the restoration of a 17th century wash-house in the village of Maron (856 inhabitants, 12 km from Nancy). Monsieur G (34 years old, pharmacist at the hospital, single) justified his move to Bouxières-aux-Chênes (1,400 inhabitants, 15 km from downtown Nancy) by highlighting a village atmosphere that met his aspirations: "I'll tell you, I came to live in Bouxières because my friends, whom I often came to see here, have always praised the good atmosphere of the neighbourhood. And that was the most important thing in my choice." 21 And he went on to say how much he enjoyed working together with his neighbours, whom he knew well, in organizing the local garage sale. In its transcendence of anonymity the motivation here is distinctly non-urban: indeed it is reminiscent of the village working community of an earlier age. 20 Original: "Vous savez lorsque j'ai divorcé, j'ai pu compter sur deux de mes voisins qui ont gardé mon enfant, et en qui j'ai complètement confiance. Ils ont pris le temps de m'écouter lorsque ça allait mal. Du coup, j'ai voulu à tout prix garder ma maison parce qu'ici on forme avec les voisins un petit monde. Même ceux que je ne connais pas beaucoup, on se dit bonjour et tout se passe bien." 21 Original: "Moi, je vais vous dire, je suis venu habiter à Bouxières parce que mes amis que je venais voir souvent ici m'ont toujours vanté la bonne ambiance du quartier. Et c'est ça qui a compté le plus dans mon choix."

A large garden for yourself, and your children out of sight
Finally, the interviews revealed a third determining element in the upper-middle and upper class residential aspirations and rationales of peri-urban Nancy. It is a question here of having a large protected space surrounding the house which allows you to benefit from privacy and freedom of movement without being disturbed by your neighbours. The size of your plot becomes an indicator of income and wealth levelsa distinguishing social feature and by extension an indicator of gentrification. Monsieur and Madame H (lawyers, 48 and 52 years old, married, three children), for example spent a long time looking in Nancy's first suburban rings for a building plot of more than 1,200 sq. m for a single-family house: "What we wanted was a big enough piece of land. I didn't see myself with the neighbours on our backs. I feel like breathing when I get home from work. That's why we came to Bouxières-aux-Chênes to build. So for my wife's kids, it was more fun to play with their friends, and we arranged the plot of land in our own way." 22 High-end land in the inner city cannot compete on this score. So it is scarcely surprising that the gentrifiers in our survey were radically opposed to any idea of redrawing subdivisions in order to increase the density of housing. If peri-urban gentrifiers have chosen to live in developments with large plots, it is certainly not so that they can see their neighbours arriving right next to their home.

The impact of rising real estate and land prices on exclusiveness
The nine peri-urban communes of Nancy in our qualitative corpus that are part of a gentrification process -Bouxières-aux-Chênes, Cerville, Chaligny, Eulmont, Fléville-devant-Nancy, Lenoncourt, Lupcourt, Pulnoy and Velaine-en-Haye -are undergoing new social and territorial differentiation. Nancy's first and second periurban rings are experiencing new forms of top-down segregation (see for France Pinçon/Pinçon-Charlot 2007;in general Musterd/Ostendorf 1998 and Friedrichs/ 22 Original: "Nous, ce qu'on voulait, c'était de la surface. Je ne me voyais pas avec les voisins sur notre dos. J'ai envie de respirer lorsque je rentre de mon travail. C'est pour cela que nous sommes venus faire construire à Bouxières-aux-Chênes. Du coup pour les enfants de mon épouse, c'était plus sympathique pour jouer avec leurs copains, et nous, on a arrangé le terrain à notre façon." 268 Hervé Marchal, Jean-Marc Stébé, Florian Weber Triemer 2009) that go hand in hand with the formation of residential neighbourhoods and an increasingly assertive distance from more modest social income groups, which are excluded from the emerging gentrification niches by increasing land and property prices.
The notion of "being among one's own kind", with its connotations of exclusiveness and social differentiation, is applicable here insofar as the inhabitants of these niches of peri-urban gentrification belong to the same socio-professional categories, have attended the same schools and universities, and go to the same associations and leisure clubs. This gentrification process certainly heralds new forms of housing decline -to be analysed in detail -based on the desire to preserve the value of one's own property (see in general Schwartz/Seabrooke 2009;Doling/Ronald 2010;van Gent 2010), in order to guarantee a homogeneous social environment and to ensure the landscape quality of one's living space. The question therefore arises as to whether these peri-urban gentrification niches are, in fact, undergoing a process of "clubbization" (Charmes 2011).
Over and above the communes that are currently most advanced in recent gentrification processes, there are others, particularly Dommartemont, Laître-sous-Amance and Lay-Saint-Christophe, which have for several decades been engaged in a process of social differentiation that today takes radically exclusive forms. These communes end up embodying urban fragments that homogenize socially and economically from above. Although we are not speaking here of strictly gated communities, the fact remains that we are confronted with territorial entities in which the exclusiveness of the wealthier households has become increasingly apparent.

Conclusion: assessment and outlook
Gentrification has in recent decades -and even more so in recent years -become a field of research that is by no means restricted to the inner-city areas of Western Europe. So far as France is concerned, detailed studies have been published on inner-city gentrification (see Préteceille 2007;Donzelot 2009;Clerval 2013), as well as on the radical changes taking place in those Parisian banlieues that are within close reach of the city centre (see Marchal/Stébé 2012;Albecker/Fol 2014;Weber/ Kühne 2017); but -leaving aside recent work on the specific issue of urban/rural hybrids (Kühne 2018) -little attention has been paid to the more distant ring at the interface of the urban and rural. Yet what in France has earned a designation of its own, the "peri-urban ring", reveals a specific pattern of upheaval, with changes in social structures and concomitant displacement of population groups that justify its inclusion within the concept of gentrification. It is these developments that have lain at the centre of our analysis, and an outcome of this analysis is the realization that the rural communities around -but not in direct proximity to -Nancy (as a prototypical medium-sized French city) can by no means be classified either as "urban periphery" or as a terra incognita to be entered at one's (social) peril.
After a brief glance at the current state of research, we presented quantitative and qualitative data on the metropolitan area of Nancy in Eastern France as an exemplary case-study for our analysis of periurban gentrification. A growing number of peri-urban communes in this area have experienced an influx of higher professionals (with correspondingly higher educational qualifications) and a simultaneous decrease in the number of less skilled and unskilled employees. About a third of the communes included in our survey can accordingly be considered as fulfilling the criteria of ongoing gentrification.
In order to shed light on people's motives for moving into the peri-urban ring and for selecting their specific house or plot of land, the quantitative data was complemented with a qualitative survey consisting of 18 in-depth interviews. These indicated the importance for new home-seekers of a "beautiful environment" in terms of landscape. In contrast to speculative innercity gentrification, there is no danger in the peri-urban area of making a risky investment in a downmarket neighbourhood in the hope that substantial numbers of like-minded purchasers will follow. In other words, in contrast to the classical urban model, peri-urban areas do not need "high risk" pioneers to precede the gentrifiers: these come directly and invest in safe, but already expensive, land. Their financial resources are sufficient to demarcate them from those unable to afford such privileged locations. And privilege is to a considerable extent what motivates them to seek -and by the same token increasingly to establish -a homogeneous neighbourhood in Nancy's peri-urban communes. Other motives apparent from the survey are the (no less privileged) tranquillity and sense of security provided by their chosen location. On the other hand, the proximity of like-minded neighbours should not be exaggerated: large building plots, unattainable in an urban context, ensure that no one comes too close for comfort. Finally, an aspect related to a typically postmodern "both-and" hybridity (Bhaba 1994; Weber/Kühne 2017) is the combination available here of urban infrastructures (cinemas, shopping centres, leisure facilities etc.) and a lifestyle that lacks the anonymity of the city.
Against the background of these results, further research is called for into the ongoing pluralization of lifestyles in peri-urban space as an aspect of gentrification. Other issues not even touched upon here are, on the one hand, what happens to those who are displaced (or move voluntarily) away from centres of peri-urban gentrification and, on the other hand, the political strategies and impact of communes in the gentrification context (see in general Smith 2002;Uitermark/Duyvendak/Kleinhans 2007). Finally, the emphasis on landscape and its qualities evident from the interviews as a motive for residential location choices suggests that a promising and as yet unexploited link might be made between the research fields of landscape and peri-urban gentrification. In the meantime our quantitative and qualitative surveys and the analysis presented in this article represent an initial attempt to describe gentrification processes in France more fully.