Introduction

1The urban project has superseded rational and technical planning as the emblematic contemporary organizational form in urban planning (Boutinet 2005, Pinson 2009). It is supposed to provide a comprehensive, specific and local approach by getting as many stakeholders as possible to work together and by breaking with centralized government and a universalist vision of urban planning (Jaquet 2014). In the context of competition between cities, urban projects are tools of differentiation in the hands of urban elites (Pinson 2009), who believe that the adaptation to local specificities could replace the close monitoring of planning and generic urbanism. Urban projects are said to prioritize the context and draw inspiration from the place where they are located, the “already there” city: its history, geography, and identity. In a guide for practitioners, Verdier writes, for example, that every urban project is “unique and local in a globalized world” and represents a far cry from “the interchangeable product city” (2009: 171).

2As part of a neoliberal turn in which supply-side urban policies are becoming widespread (Adam and Comby 2020, Pinson and Morel Journel 2016), urban projects serve as distinguishing features of cities whose local characteristics become a competitive advantage. As tools for differentiation (Pinson 2009), such projects are expected to be characterized by their material diversity and their uniqueness to be widely emphasized in marketing campaigns. At the same time, the dissemination of “good practices”, the continual appraisal and the growing influence of labels and norms bring about a shift towards compliance with generic standards (Adam 2017, Brenner and Theodore 2002). For those walking the streets of recently built projects or browsing urban planning and architecture magazines, the recurrence of identical forms is readily apparent. As conformism seems to take precedence over singularity, the buildings are not the only standardized features; so are their uses and marketing discourses.

3A generalized injunction – reinforced by sustainable development –, which values sensitivity to local singularities and contextualization, paradoxically leads to the material and symbolic standardization of spaces. This article examines this paradox from a particular entry point: the discourses of those who design contemporary cities. I define “designers” the practitioners directly involved in projects (excepting sponsors such as elected officials, financiers, and private investors): developers and urban planners working for local authorities and development companies, project managers for promoters and landlords, (landscape) architects, and engineering consultants (for sustainability, planning, and artistic issues). I conducted semi-structured interviews in 2012 and 2013 with 26 designers involved in two urban projects: Confluence in Lyon and Bottière-Chénaie in Nantes. These are two major projects of several thousand housing units and square meters of offices and shops, built on former wasteland. They are important drivers of attractiveness for their cities, as they became, materially and discursively, the embodiment of the catchwords characterizing contemporary urban spaces: sustainability, innovation, diversity, and participation.

4The interviews reveal the work of designers, the projects they sought to achieve, their successes and failures. Their discourses emphasize the projects’ singularity and its integration of local specificities (history, geography, and sociology). The starting point of this paper is the paradox between the urban projects’ valorization of local singularities and the conformism and standardization of the discourses surrounding them.

5To understand this paradox, I analyze the discourses using mixed methods comprised of a comprehensive qualitative approach and a quantitative approach based on textometry (vocabulary specificity, factorial analysis, divisive hierarchical clustering, similarity analysis). I process the corpus by recursively employing tools of content analysis (interpretation grid) and discourse analysis (textual statistics). This approach offers a comprehensive overview of the raw material, through a denotative analysis, and an interpretative analysis of the actions in which the actors engage. I thus analyze how the designers’ discourse reveal their representations of standardization while also disclosing their own practices of standardization.

6The standardization of space production is the results of market influences, architectural and urbanistic fashions, and the impact of socio-technical standards on production. In a competitive context (Pinson 2009), urban projects, like real estate projects, are designed to meet the expectations of target investors (Adam and Laffont 2019, Delage 2019). Whether they are institutional (especially in the case of real estate for offices) or private (for rental housing), these investors seek above all to limit investment risks by favoring conformist planning or technical solutions in order to rent or resell profitably. Standardization is therefore a safeguarding measure for private and public actors approving projects (Delage 2019). This trend towards conformity is further reinforced as financial actors increasingly influence real estate markets: the office property market in France repurposes buildings into low-risk, partly interchangeable financial assets (Guironnet 2019). Fashion effects have been observed for a long time in the design of urban projects: planning and aesthetic models circulate recurrently, establishing themselves as “good practices” amongst designers and clients. “Good practices” can be understood as “material criteria” which “work as systems of normalization, even standardization” (Devisme et al. 2007: 18). They circulate through project visits, trade fairs and professional media, but also via “good practice” guides drafted by academic authors (e.g. Carmona 2021, Verdier 2009). Additionally, designers favor solutions that comply with technical and socio-technical standards (Adam 2017). These choices are further validated and promoted by systematic evaluation of performance and certified through regulatory monitoring or labeling (Adam 2017, Gaillard and Matthey 2011). Finally, these effects are reinforced by the players’ national or international circulation, who finance, design or implement projects in different cities. Their circulation ultimately contributes to the dissemination of “good practices” and methods.

7Standardization contradicts a priori the trend to develop policies and projects that take local particularities into account (Kaufmann and Arnold 2018). Pinson (2009) highlights the standardized nature of urban projects and notes the paradox between this conformism and the fact that urban projects are markers of the differentiation processes of urban policies and governance; differentiation that would be a way for urban elites to address the pressures of globalization and the recomposition of states. Competition between cities drives urban elites to adopt stereotyped strategies, but also to exploit “hidden or under-exploited” resources (Pinson 2009: 61) to produce a “territorial offer” (ibid.: 76) that maximizes the comparative advantages of every city: heritage, industry, real estate, scenery, etc.

9Unlike institutional discourses such as urban plans (Buhler and Lethier 2020) or press articles (Comby et al. 2019), semi-structured interviews cannot be considered a priori as standardized discourses. The advantage of semi-structured interviews resides in their ability to explore the subjective narrative of designers, rather than official or written discourses. They allow more nuanced accounts to emerge, and in particular provide the ground for a critique of one’s professional field and emerging trends. The interviewees’ different professions and levels of involvement in the project contributed to a diversity of views and discourses. Finally, textometry is used to complement the qualitative discourse analysis offering a comprehensive insight into the variations among discourses.

10This article is divided into two sections. In the first part, I advance the theoretical framework and the hypotheses, introducing the research sites and the statistical methods and tools deployed to test the hypotheses. The second part presents the two main findings. The first suggests that discourses are very little influenced by territorial context and are very similar in Lyon and Nantes. The second argues that promoting local singularities appears to be a conformist dimension of the urban planning discourse.

Discourses as material for critical urban studies

Theoretical framework and hypotheses

13Borrowing from social psychology (Author PhD, Gamby-Mas et al. 2012, Jodelet 2003 for a synthesis), I define representations as practical knowledge – both psychological and social, individual and collective – enabling individuals to understand reality and to act on it. Representations play three roles for individuals and social groups who build and use them: informative (interpreting reality), regulating social relationships (unifying and differentiating groups) and, deriving from the first two, operative (enabling and guiding action). Representations are a bridge between individual and social (Moscovici 2003), because they are the fruit of interactions between individuals and between groups, but also because they are constitutive of the existence of the group itself, and in particular of its members’ adherence to a common vision of reality. The tendency of urban projects towards conformity and the common representations shared amongst the different categories of our interviewees are an illustration of how representations function. Finally, to connect this theoretical framework to that of the production of space and to that of critical discourse analysis, a last point needs to be added: ideology is defined as a finite, hierarchical system of values that can generate an infinite number of representations to legitimize a form of power organization (Adam, 2016). This allows us to consider how the values of the neoliberal ideology that infuses the urban project (Adam and Comby 2020, Pinson 2009) influence the representations of the actors we interview, and how these representations can be revealed in the opinions and attitudes they express during the interviews.

14A critical discourse analysis reveals the social representations and relations, and in particular power relations, since discourse contributes to the maintenance or change of power structures and ideologies (Fairclough, 2010). The ideological dimension is of particular relevance since the dominant ideology (Fairclough’s aim is to discuss the role of language in maintaining or contesting its hegemony) influences representations. Fairclough insists on the need to link the language analysis (which we do here with textometry) to that of discursive practices (we focus on interviews as a discourse register) and of the social context in which discourse takes place (the professional world of urban planning). While geography and urban studies have mobilized Fairclough’s approach (Comby et al. 2019; Buhler and Lethier 2020), a link has never been made explicitly with Lefebvre’s work. This link is a logical one nevertheless, insofar as both authors are concerned with unveiling ideological mechanisms and social contestation. In the case of space production, this requires a detailed understanding of how the various urban professionals play their roles, as well as an analysis of shared representations and the potential contradictions or even conflicts they may reveal. This is made possible here using semi-structured interviews, which can grasp the way in which designers view their actions; whereas written discourse (in planning plans, technical or marketing documents) is uncritical and calibrated to ensure an apologetic reception. Interviews also allow a focus on the process rather than the product while shedding light on the actors’ reflexivity. Overall, through semi-structured interviews we can unpack the nuances and contradictions lying at the heart of their actions.

15I present below three hypotheses that document and analyze the identified paradox:

1. Critical Discourses Analysis reveals the urban planning actors’ representations as well as the ideology underpinning them (the conceived space);

2. These representations, as regulators of social relations, help both to unify the group of designers and to differentiate it from other groups (that of the inhabitants here taken as an external reference). This happens at the level of what I call designers, but also for each profession (urban planner, architect, project management assistant, landscape architect, developer). It is expected that, since they share the same profession, designers will have relatively similar discourses, and that differences related to specific professions will also emerge. Conversely, it is expected that residents will have more varied discourses;

3. As guides to action, these representations are reflections of the economic and cultural constraints these actors are facing. This is likely to explain how a paradox can persist despite its identification or even denunciation.

Bottière-Chénaie and Confluence, two projects reflecting the Zeitgeist

16Bottière-Chénaie and Confluence can thus be considered as two projects reflecting the Zeitgeist in urban design. The term Zeitgeist (spirit of the age) emphasizes the notion that the dominant paradigms can vary and are embedded in materiality. As such, the Zeitgeist can impose itself upon individuals, whether they are conscious of it or not (Durkheim 1895). It highlights the influence of ideology, as well as that of fashions in architecture, urban planning and city governance.

17Bottière-Chénaie and Confluence won awards in the 2009 national EcoNeighborhood competition and play an important role in the “sustainable” communication strategy of their respective local governments. Bottière-Chénaie was a central component of Nantes’s successful bid to become “European Green Capital” in 2013. In Lyon, Confluence is the driving force behind urban sustainability policies, which are reflected in the development of a local eco-neighborhood charter, a municipal “Fair and Sustainable City” label, and actions in partnership with the WWF. Both contracting authorities are publicly owned development companies, meaning that the local government controls the major planning decisions, in a context where partnership between public and private interests is the prevalent rule in Nantes and Lyon.

18As products of the same era, these projects share many common points that allow us to grasp both the global issues of urban production and their local variations. The two projects are new neighborhoods built on former wasteland in the industrial port area in Lyon and the market-garden area in Nantes. They are multi-functional, even if Bottière-Chénaie is primarily residential, while Confluence hosts numerous shops, office buildings, and large public institutions (regional council hall, museum). The catchwords used to publicize both projects are similar: social diversity, “urban” character, quality of life, innovation, etc. Their urban and architectural forms are consistent with current trends, inspired by “new urbanism” (Ellis 2002, Grant 2005) and “supermodernism” (Ibelings 2002).

19In total, 26 interviews were conducted, 11 with the designers of Bottière-Chénaie, 14 with those of Confluence, and 1 with an architect who worked on both projects. I met the urban planning directors and landscape architects of both projects, urban planners from each city, project managers from the development companies (Nantes Métropole Aménagement, SPLA Lyon Confluence), 11 architects, the project managers of a property developer from Nantes and a social housing landlord from Lyon, and 3 engineering consultants. Although these designers share a common professional culture, the fact that they occupy different functions, have different educational backgrounds (urban planning, architecture, landscaping, law, engineering, etc.), and are of different generations (our respondents were between 30 and 70 years old) does not necessarily mean that their discourse would be standardized. However, this gradually transpired during the interviews.

20Every corpus of discourse is shaped by the researcher’s hypotheses, analysis and interpretations (Comby 2016). This is particularly crucial in the context of semi-structured interviews: as an interviewer and a transcriber, I helped co-producing the discourses I analyzed. The main pitfalls of analyzing the discourse of urban professionals to assess urban policies are the heavy use of rhetoric and the a posteriori rationalization of actions (Buhler et al. 2018). Yet, the effects of rhetoric and an a posteriori reconstruction are not smokescreens to be dispelled, but part of what is observed, considering that such screens are part of actors’ representations and reflect the prevailing ideology (Fairclough, 2010). Although not entirely possible to verify, the assumption is that the discourses collected in interviews are more personal, less polished or smoothed out than those in official documents. However, as a co-producer of the collected discourses, to what extent I also co-produced their homogeneity through my questions? The orientation of the questions is a bias limited by two factors: a) the interviews were conducted by a single interviewer, which potentially reproduces the same biases from one interviewee to the other; b) the interviews were not intended to focus on the singularity and conformism of discourse, and were therefore not a priori designed with this purpose in mind. It was instead the fieldwork that generated this research question, as the discourses appeared to be more similar to one another than the projects themselves. Transcription was carried out by a single person, without software assistance. To prevent the homogenization of the discourse, the transcription was done word by word, following as closely as possible what was actually said, i.e. preserving not only the vocabulary but also the marks of orality: unfinished sentences, repetitions, hesitations. In this way, I tried to limit the influence of transcription on the analysis.

Interweaving content analysis and discourse analysis to reveal uniformity

21The gradual improvement of software tools for textual analysis (Jacobs 2006) and the emergence of the “linguistic turn” (Mondada 2003) have given the analysis of discursive material an increasingly important role within social sciences. It is “considered a via regia for the study of the psychosocial dynamics of representations” (Kalampalikis 2003: 148). Social psychologists have progressively selected or developed statistical tools and software adapted to their needs in order to deal with diversified and voluminous corpora. Iramuteq, the software I use, was developed by Pierre Ratinaud and his team of psychosociologists specialized in the analysis of political discourse at LERASS (Toulouse 3). Widely used in France, ALCESTE was also developed by a psychosociologist, Max Reinert (2001). I share with them the belief that textual statistics do not lead to purely quantitative analysis but also supports qualitative analysis. I use textual statistics not to produce a linguistic interpretation but rather to identify meanings based on denotation. In doing so, the analysis goes beyond the simple enumeration of forms to highlight the more complex interplay of proximities and oppositions of forms in which individuals rely. This approach privileges a focus on the structure and context of the utterance, by highlighting the similarities and differences between speakers and categories of speakers. As a complement to connotative interpretation, discourse analysis opens up new avenues for reflection on content analysis and reinforces its results. In line with Comby’s (2016: 132) proposal, my work aims “to rethink the dichotomy between quantitative and qualitative approaches to data, by intertwining these two logics”.

23I try to avoid these pitfalls by acknowledging the theoretical assumptions on which they are built. Drawing on the theory of representations (Kalampalikis 2003) and on literature on spatial policy discourses (Hastings 1999, Segaud 2012), we can argue that:

• discourses are not neutral vectors of information but contribute to the actors’ reality construction;

• actors are competent discourse producers, i.e. they are capable of translating into discourse the meaning they give to reality;

• it is possible to infer representations from a corpus of semi-structured interviews.

24The content analysis of interviews is done according to systematic criteria defined by identification and qualification keys, based on an analysis grid applied systematically. The textual data analysis (Lebart et al. 2019) relies on textual statistics tools. A statistical approach “can in no way replace the researcher’s intuition, nor the hermeneutic finesse of classical content analysis, nor the interpretive work necessary for any reflection in the social sciences” (Kalampalikis 2003: 151). Discourse analysis offers nevertheless additional insights to those afforded by content analysis, particularly on voluminous corpora, as in our case (810,000 words): “textual statistics allow for the objectification and synthesis of this qualitative information in order to bring out a common and diverse representation” (Garnier and Guérin-Pace 2010: 9).

25The joint use of content and discourse analysis to interpret the data relies on the principle of recursiveness. Through successive sieving, this recursive combination of quantitative and qualitative tools reveals the nature and content of representations shared by the social groups studied. Differently put, the qualitative approach initiated the hypothesis, the quantitative method followed to test the hypothesis of homogeneity, and then the qualitative process was drawn again to better highlight the standardization and delineate the singularity traits.

26By combining the qualitative approach with discourse analysis, we can link the text to discursive practices and to a spatial and temporal context (Fairclough 2010). This combination is therefore particularly suited to a multifaceted geographic approach (Comby et al. 2016): a) textual statistics can be used to analyze new corpora (press, interviews) to shed light on spatial issues; b) they are particularly well-suited for comparative approaches, since they can highlight both common features and divergences between study sites; c) their use sheds light on the representation that speakers have of lived, perceived and conceived spaces (and on their professional practice, in the case of urban planning actors).

Statistical tools and corpus

27I used four complementary statistical tools included in the Iramuteq software (Rocha de Souza et al. 2018). These mathematical tools combine commonalities and contrasts to highlight both the unity and multiplicity of discourses. Their main features are summarized below.

• The lexical specificity score highlights the words typically used by a population, shedding light on the differences in vocabulary between groups.

• Factor analysis (FCA) structures the words in the corpus according to their distribution in text segments (context units) and in relation to variables chosen for analysis. FCA structures the vocabulary according to the linguistic context of use and the characteristics of the speakers. It shows the co-occurrences of terms and associates them with variables.

• Top-down or divisive hierarchical clustering (DHC) starts from the raw discourse (rather than the variables associated with the speaker(s)) to identify its logic, organization, and structure. The objective is not “the calculation of meaning, but the topical organization of discourse” (Kalampalikis 2003: 151) by revealing the semiotic traces inscribed in the discourse independently of any interpretation. The value of the classes formed by DHC lies both in the meaning that can be given to their autonomy and in the meaning that can be associated with their links/imbrication.

• • 5 The analysis of similarities is based on the calculation and representation of co-occurrence trees. (...)

  The analysis of similarities has a comparable objective to that of the DHC. Seeking to highlight the proximities and oppositions between lexicons and themes, it focuses on the correspondences and regularities, considered the weak point of specificity calculations, FCA, and classification, which underline the oppositions5.

28These tools are used to analyze four separate but related corpora. The general corpus of the designers’ discourse highlights the elements structuring their discourse and the difference in discourse between actors and across sites. The separate corpora of the Bottière-Chénaie and Confluence designers’ discourses are set side by side to test their specificities and differences. The corpus of discourses belonging to designers and inhabitants (the total corpus) is used to highlight the territorial differences and the greater uniformity across the designers’ speeches. Two main variables are used in the analysis: the project studied (two categories: Bottière-Chénaie or Lyon Confluence) and the profession/function of the designers (five categories: urban planner, architect, project management assistant, landscape architect, and developer (or social landlord, who plays the same role as the project manager here)). The analysis was carried out on discourses produced entirely in the French language.

The Zeitgeist pervades uniform and delocalized discourses

Local Context has little influence on discourses

29Discourse analysis tools are useful to analyze the form (vocabulary) and structure of interviews in a logic of recursive reduction of the corpora, from the corpus that gathers all the interviews to those that contain only the discourse of the designers of Bottière-Chénaie or Confluence. The vocabulary gives an indication of the diversity of the discourse, while the structure provides information on the themes and issues identified by the designers in the production of the city. In a first step, we decipher the total corpus.

Figure 1: Dendrogram (result of DHC) and associated FCA on the total corpus

30DHC performed on the total corpus distinguishes four classes (each represented on the factorial plan by a distinct color) deriving from two branches (see dendrogram in the top right). FCA and the associated lexical specificity score are used to interpret these classes. The two main branches distinguish unambiguously the designers’ discourses (classes 1 and 2) from those of inhabitants (classes 3 and 4).

31The main finding concerns the homogeneity of the discourses, which is much important among designers than among inhabitants. Classes 1 and 2 can be defined as “product” (practical issues of the project) and “process” (way of producing the city, design work, and reflections on it). On the factorial plan, they are totally layered. Class 3 corresponds to the “practice of space”, while class 4 represents “the inhabitants’ reflection on their relationship to the space” (judgment and narrative). Classes 1 and 2 are strongly intertwined and, above all, much more homogeneous than classes 3 and 4, which are characterized by their compactness on the factorial plan. These classes correspond to the three elements of Lefebvre’s triplicity of space (1974): conceived space (1 and 2), lived space (3), and perceived space (4).

Designers of Confluence	Saône (infinite), Lyon (infinite), Confluence (infinite), darse (13.9), SPLA (12.5), autoroute (12.2), loisir (11.7), Name of the urban planning director (anonymized) (10.9), Perrache (10.3), port (10.1), quai (9.8), SEM (9.3), Lyon Confluence (8.9), très (8.6), Presqu’Île (8.4), parc (8.4), CL2 (7.9), Rhône (7.8), niveau (7.6), Ville de Lyon (7.6), Collomb (7.6), premier (7.3), Part-Dieu (7.1)
Designers of Bottière-Chénaie	Nantes (infinite), maison (infinite), Bottière (infinite), Deux (15.9), Bottière-Chénaie (14.3), ruisseau (13.6), pièce (13.5), quoi (12.2), mec (11.2), jardin (10.7), collectif (10.2), gens (10), abordable (8.7), bagnole (8.5), Name of the urban planning director (anonymized) (8.4), loggia (8), médiathèque (7.3)

Table 1. Terms with a lexical specificity score greater than 7 in the “all designers” corpus

32The analysis of smaller corpora, which includes the discourse of all the designers, refines this first observation. The first calculation is the lexical specificity score, with the site as a discriminating variable. It shows that most of the terms that are highly specific to each terrain are elements of pure description, with toponymy and proper names leading the list (cf. Table 1). The vocabulary focusing on the problematization of local issues differentiates the designers’ expressions to a lesser extent. This is an initial confirmation of homogeneity in the designers’ discourse.

33With regard to Bottière-Chénaie, some terms reflect the range of vocabulary or verbal tics of some designers (quoi, mec, gens, bagnole (like, guy, folk, motor)) and, sporadically, specific issues. For example, abordable (affordable) refers to the policy of aiding home ownership, maison (house) qualifies the bordering urban forms. At Confluence, terms referring to local issues are rarer. Niveau (level) refers both to the standing of the project and to the different floors of Lyon’s buildings, très (very) can be interpreted as reflecting the superlative character of a project with a demonstrative vocation. In both cases, toponymic elements (Lyon, quai, ruisseau (quay, brook)) or programming elements (parc, deux, loggia (park, two, loggia)) dominate.

Figure 2. FCA of the classification on the “all designers” corpus, superposition of shapes and variables (profession / study site)

34DHC calculation on this same corpus further reinforces the idea of a homogeneous discourse that varies little across the two projects. The lexical worlds that constitute the five proposed classes are thus close-knit but the “site” variable is discriminating only weakly (cf. Figure 2). On the other hand, the classes are strongly correlated with the respondents’ profession. The DHC indicates that urban planners, together with the engineering consultants, are at the center of a factorial plane that distinguishes the lexical worlds characteristic of architects, promoters and landlords, and landscape architects. The structuring of the productive model – landscape architects expressing themselves more on public space and biodiversity, for example, and architects on the built environment – and the professional referents shape discourses more than the geographical context.

Figure 4. Similarity tree of the Bottière-Chénaie designers’ corpus

35As the size of the corpus was further reduced, we calculated the similarity trees of the corpora gathering the words of the Confluence and Bottière-Chénaie designers (cf. figures 3 and 4). An analysis of the similarity trees reveals the main clusters around which the designers’ discourses are organized as well as the links between these clusters. The graphs show a similar organization around very similar branch nodes. These clusters reflect the major issues related of the design of the space, in particular the vocabulary related to the project, reflections on the design and the completion scale.

36The similarity of the patterns making up these clusters also confirms the weak differentiation amongst the discourses produced around each project. However, the similarity trees also highlight the differences between these two sites.

37At Confluence (Figure 3), the issue of the project [projet, yellow cluster] and its design are central [penser, question, light green]. The term “project” is associated with both the general question of the production of the urban space and the specific urban project. These two subjects are mentioned together, while Confluence is a starting point for discourses that develop to include contemporary urban issues. Three major secondary nodes correspond to the project’s scales: the architecture [chose, green] (focused on visual aspects and housing), the public space [grand, espace, purple] (concerned with the qualitative treatment of spaces), and the new neighborhood and its integration in the city [mettre, pink]. The issues of housing [logement, red] and use [gens, quartier, blue] are still significant but more peripheral.

38In Bottière-Chénaie (figure 4), the project is an important cluster, but it is less central. The discourses are organized around lexicons related to housing [logement, red], uses [gens, purple], and the production of the city [penser, light green]. This last issue, which is not specifically connected with the Bottière-Chénaie project, reflects a looser entanglement of specific and general discourses on the urban question in Nantes than is the case in Lyon. Rather than different representations, these different nodes correspond to what distinguishes the material and organizational characteristics of the two projects, as well as the composition of respondents’ panel. The larger number of actors involved and the more complex set-up in Lyon (numerous engineering consultants, grouping of five architects per block on the first phase of the project, etc.) mean that the question of the project occupies a more central place in the designers’ discourse than in Bottière-Chénaie [projet, light green]. Conversely, the issue of producing housing for a target population of first-time buyers may explain why housing and uses are more prominent in the discourse on the Nantes project.

39Finally, the forms associated with sustainable development, participation, or social diversity do not appear as focal points of the designers’ discourse. Although widely discussed during the interviews, these issues appear to be transversal and not specific to one or other of the themes addressed. This underlines the ubiquity of the catchwords used in the production of contemporary urban spaces and their normalization; they are not (or no longer) transposed into a specific theme.

Promoting singularities, a conformist aspect of urban planning discourse

40The homogeneity of discourses – and therefore their delocalized character – highlighted by the lexicographic tools reinforces the impression felt during the interviews and their transcription. At first glance, this homogeneity seems to contradict the claim running through the collected discourses that close consideration is given to the local character.

41In both Lyon and Nantes, the designers working at the urban scale show their desire to address contemporary issues by paying close attention to the local context in its temporal, spatial, and social dimensions. They balance these dimensions with a national and international perspective embraced by the two local authorities. We find similar comments made by the Confluence consultant on the project’s use and occupancy (“We are not just transposing marketing ideas, we are really linking it to the local fabric, the idea of the sustainable city, and the idea of continuity with respect to the history of the neighborhood”), by the urban planning director of Bottière-Chénaie, for whom “the project reinterprets the site and the history of the site”, or by the project manager of a real estate developer in Nantes, for whom the project “is part of a kind of familiar history and geography which proposes an improved quality of urban form and architecture and uses”.

42Architects argue that they rely on local particularities for inspiration and to mark both materially and symbolically the integration – through some form of allusion (the terms most often used are “echo” and “resonance”), reinterpretation or even rectification – of their productions in what they perceive as the identity or the specificities of the pre-existing space. They also proclaim the importance of considering the multiplicity and the imbrication of scales (building, neighborhood, city, globality of the production of space).

“We provide an architecture... In fact, each time we are dealing with situations in which we find singularities – whether geographical, political, urban, economic, and so –, our work is to make a synthesis of these various components and landscape representations to come up with an architectural proposal. So, each time we hope our proposals are singular and unprecedented.” (Architect, Lyon)

“In fact, the city already exists and it is the result of a long historical process of stratification. We have to try to understand it, to observe it, to see it, and once we have been able to do so, we try to build intelligently with it. [...] The building that we did, I think, partly extends this idea. The molded walls are evocations of the stone walls that existed before in market gardens.” (Architect, Nantes)

“I think that the soul of the project comes obviously from our capacity to read a site and to use everything that works, everything that is beautiful, and to modestly improve to make everything that is not going well a little better, or better altogether. So, there is a reading of the site which for me is a fundamental element.” (Architect, Lyon)

“How we design our projects? In fact, we conceive them at the same time with the program and in resonance with the site and often with the history of the site. [...] And then all the rereading of the past, not agricultural but of market gardening of the site, because it is an old market-gardening site. We took this into account by including large vertical greenhouses, an educational greenhouse on the school roof, the drawing of the vegetated facades and then the whole refurbishing of this large facade on the mall. Then the red and white tiles and the gray slate, which is a kind of abstraction, if we can put it that way, of the small Chénaie neighborhoods next door, made up of small houses from the 1950s, fitted with mechanical tiles and slate roofs.” (Architect, Nantes)

43It may seem ironic that one of the most explicit examples of the delocalized uniformity of discourse is precisely the praise of local particularities (architectural, historical, and geographical). Above all, this reflects the strong integration of this component in the shared representation of what a desirable urban space is. The recurrent reference to both elements of innovation (or contemporaneity) shared on a global scale and of local identity or history reveals the double universe of designers’ representations. We see here a strong integration of one of the postmodernist principles, i.e. the promotion of innovation and the adaptation to particularities – historical, geographical, cultural – of the concerned area (Ibelings 2002, Jameson 1992). This idea, reinforced by the promotion of sustainable development (Adam 2017), reflects the desire to reverse the universalist vision of modernism. These shared representations are rooted both in the context of spatial design, where project-based urbanism replaces plan-based urbanism, and in the education received by architects and urban planners. Indeed, in response to questions related to their professional career, they share visions shared by rather anti-modern scholars. The prevalence of such representations amongst the designers we interviewed are matched by common negative views of modern architecture and urbanism, which they share unanimously (e.g., that of suburbanization). This is more broadly congruent with the importance of urban “identity” in political and media discourses (Castells 2009, Davis and Duren 2011).

44In addition to their homogeneity, these discourses reflect the extent to which the designers’ representations are infused by the Zeitgeist. The idea that the way the space is conceived is very much a sign of times is explicitly evoked on numerous occasions during the interviews.

“I think that there is a question of time, we are in a time of great diversity of cultures, tastes, etc., there is no reason why architecture should escape it. We are not in a royal, imperial era, of completely homogenized culture. So, it reflects the diversity of the times [...] There are effects related to construction methods, which are the same, which are repetitive, and there are also effects of fashion which makes it all less cacophonous. Yes. What do you want me to say? It’s production, it’s what we know how to do today.” (Urban planning director, Lyon)

“It is something that is part of a social dynamic and therefore architecture is undoubtedly the implementation of inspiration of our time. We can have a critical, or distanced, or regionally activist eye with regards to this. But in any case, we are above all situated in relation to a cultural environment of the time. I believe that when I speak about context, perhaps the founding element, even more than geography, is time, it is the epoch, it is what our society produces as its way of thinking.” (Architect, Lyon)

45This common awareness is evident in the designers’ discourses. For example, the calculation of the specificity indices of the vocabulary of the total corpus (cf. Table 2) reveals that the word “today” is very strongly representative of the designers’ discourses. This term, which is not a professional term, is used by the designers to describe the frameworks of the city’s production, as shown by the specificity indicators for this word.

	Designers Lyon	Designers Nantes	Inhabitants Lyon	Inhabitants Nantes	Number of occurrences
today	infinite	infinite	-22.3	-25.5	406

Table 2. Lexical specificity scores of the word “today” in the total corpus

46If the Zeitgeist permeates the designers’ representations, their attitude towards what is considered inescapable varies considerably. Total adhesion to the Zeitgeist is rare, even if some designers make it an objective of their work in order to update the urban to their epoch.

“We are in the Presqu’île, so we are in the middle of Lyon but we are not in the traditional Lyon, that’s clear, we are in our century. It’s 21st-century Lyon, that’s all, it’s as simple as that.” (Engineering consultant, Confluence)

47Others see the Zeitgeist as inevitable and positive as soon as it meets the resident’s expectations and is likely to become a success.

“It’s a bit fashionable, a bit trendy, but it’s not a big deal, it’s not embarrassing, it’s quite nice. The proof is that it works well, there are many people walking around, showing it to visitors, wishing to work there. So, from that point of view it’s a success.” (Architect, Confluence)

48Others offer a contrasting view and refer critically to “fashion”. This involves a criticism of the substantial weight and ephemeral content of the catchwords characterizing urban design.

“We can’t chase trends forever, because then we are in the spirit of the times, there is no durability. Yes, it’s pretty, but it’s fashionable, it’s only fashion. And then there will be something else […] So you have to try to be part of something else. But even then, you are ultimately influenced by the Zeitgeist, let’s not kid ourselves.” (Architect, Nantes)

49The temporalities of the production of the city (Simonsen 2017, Tomas 2003) are also matters of concern for designers. Even when regarded as just another stage within a temporal succession, the present still influences considerably the general representation of the production of urban space. Beyond the discourse homogeneity and the conformist representations and catchwords, these phenomena are acknowledged by designers, who do not shy away from criticizing them.

Conclusion

50This paper highlights the complementarity of content analysis and textual statistics. Lexicometry and lexicography allow to go beyond the researcher’s intuition and impressions to show that discourses are highly standardized. This standardization reflects the importance of fashion as well as the circulation of representations, values and vocabulary amongst the designers involved in contemporary urban projects. It also reveals the frames of reference specific to each profession (the designers’ professional training structures their discourse much more than the geographical context). The qualitative analysis foregrounds the role of local specificities in discourse standardization and reveals the fine balance designers seek to strike between the two.

51Our analysis of urban designers’ discourses shows that, just like their material form, their linguistic dimension is highly standardized and both influences and is influenced by the urbanistic Zeitgeist. My investigation also shows that the temporality of the production of the city is a preoccupation for designers. Whether they adopt a critical or positive attitude on the role of fashion, the present influences to a great extent the designers’ general representation of the production of urban space. The present is considered as just one stage in the series of trends. Despite the discourse homogeneity, these phenomena are identified by the designers who often criticize them.

52It is impossible to say that nothing changes: the trends renewal modifies the projects and the discourse of those conceiving them. At the same time, it is difficult to assume that project-based urbanism has renounced an universalist and context-insensitive approach (Jaquet 2014, Verdier 2009). Our observation and analysis of specific situations (we generally agree here with the conclusions of researchers working on different geographical areas (Adam 2017, Delage 2019, Lin 2007)) raise the following question: What if the injunction to contextualize projects and to integrate local specificities renewed urbanism more in discourse than in practice?

53Pinson observes that, with the growing importance of communication, urban projects are responding to imposed frameworks, while at the same time creating differentiated urban identities by enhancing the “already there city” (2009, 139). This is another way of formulating the tension between conformity and singularity. Starting from a similar line of inquiry, my work shows that stereotypes take precedence over differentiation, and that the valorization of local resources is one of the ways in which standardization occurs. The race for the singularity inevitably leads to conformity. The conclusions drawn in this article echo those of studies on the standardization of tourism and heritage experiences (Ponzini et al. 2016) or those focusing on the conformism of territorial marketing and promotional strategies, whether to produce a critique (Adam 2020) or to invite strategy designers to pay attention (Anttiroiko 2014). When projects and territories become products to be sold, even what is “already there” is selected to correspond to market standards. Local specificities become conformist because they are selected based on formalist criteria.

54While one of the challenges of contemporary urban projects is to include local particularities to distinguish themselves in interurban competitions (Kaufmann and Arnold 2018), this is often a purely discursive practice, which has no purchase in real life aside from. The density of discourses on the specificity and uniqueness of projects cannot conceal their similarity – the projects of the 2000s or 2010s are as similar to one another as were the large-scale projects of the 1970s – since they rely on identical and generalized procedures or “good practices” (Carmona 2021, Devisme et al. 2007). The integration of contextual elements (heritage, architecture, landscape) and the discourses around them are also largely standardized; they work as narrative devices rather than observable features. The designers' discourses are thus just one more proof that the race for the singularity in urban planning projects is run for the sake of conformism.

55A final question may be raised. Why does this narrative paradox of pointing out contextual singularities in standardized discourses perpetuate even when the designers themselves acknowledge it ? Firstly, it is precisely a narrative paradox, not a contradiction. The paradox is a matter of discourse. The contradiction is resolved in material production, which is standardized for the reasons mentioned above. The weight of the real estate industry and urbanistic or architectural fashions impose themselves on the stakeholders. These effects are strong and reflect massive economic stakes in an industry with significant inertia and where standards renewal, from the modern city to sustainable neighborhoods, has historically been the norm (Gaillard and Matthey 2011). The actors interviewed are not powerful enough to change the rules of the game. As a result, they find themselves acting in a conformist manner while deploring the heavy weight of standardization. Added to this are technical constraints - linked in particular to various construction standards - which also push towards technical and socio-technical standardization. Economic and technical normativity are, moreover, additional justifications for the standardization designers decry (Adam 2017). Finally, as pointed out earlier, the apology of local specificities has itself become a prerequisite of architectural and urban planning discourse, i.e. a dimension of conformism itself. In fact, this tends to reinforce the paradox, turning it into a vicious circle that is hard to break and highlighting once more capitalism’s ability to reinforce itself by integrating its contradictions (Author PhD, Boltanski and Chiapello 2007).