Introduction

1Public space is the place par excellence where community building takes shape and where different sorts of amenities are provided, and its design, governance and use are of great importance to all who live and work in the city. In this context, the way such space is organised becomes an object of negotiation and conflict, resulting from different visions and preferences among residents and among decision-makers. A particular case in point concerns the controversy about what functions should prevail in urban and peri-urban parks; conceived as places of leisure by some, but as places compatible with motorised traffic by others. If the availability of land makes it easy to develop road infrastructures, motorised traffic limits other forms of access and creates many nuisances for recreational use, far beyond the mere spatial footprint of roads.

2The debate on whether to allow motorised traffic to cross urban and peri-urban parks reflects the contrast between two different priorities – namely urban liveability and geographic accessibility. This controversy is not new and goes back to early phases of mass motorisation, as illustrated by the literature on infrastructure projects of this type. Referring to the US context of the early 1970s, for instance, See speaks of an actual “highways-versus-parks syndrome” [1971: 834], in which many hectares of park land were lost to highways every year. In the late 1970s and the 1980s, the trend changed somewhat, with civic opposition, high maintenance costs and a (slow) change in the urban mobility paradigm, resulting in the abandonment of some highway projects [Gibson, 2011]. In Brussels, for example, the Brussels Agglomeration Sector Plan adopted in 1975 by the national government instituted a moratorium on road construction. Since the early 2000s, it has been redevelopment projects that started to cause controversy worldwide: urban highways have been suppressed and replaced by urban boulevards, public equipment or green spaces in Portland, San Francisco, New York, Lyon and Marseille [Héran, 2020; see also Lecroart, 2012]. In some cases, such as Boston and Madrid, highways have simply been buried underground [Hubert, 2008]. In other cases, such as Prospect Park and Central Park in New York or the Bois de Boulogne in Paris, the interventions have involved the drastic reduction of motorised through traffic [Plitt, 2018; Livois, 2019]. The debate has surfaced again in the context of the outbreak of the COVID-19 pandemic in 2020, as sanitary measures encouraged teleworking, thereby reducing commuting flows and leading to an increase in the use of and demand for green space [among others see da Schio, Phillips et al., 2021]. In this context, many local authorities across the world decided on the – at least temporary – closure of park roads to motorised traffic.

3This trajectory can also be observed in the Bois de la Cambre, a 120-hectare park situated in the south of the Brussels-Capital Region (BCR). This case study is interesting and specific, in that the transit traffic observed there was never the subject of an infrastructure project, but instead the result of the incremental effects of a laissez-faire policy. Although there have never been any concrete projects for the construction of parking areas or a wide highway, conflict between different visions has existed since the park’s creation in the nineteenth century, up to recent years, when it resurfaced due to roads being closed to traffic in the context of the COVID-19 pandemic. In this paper, we start from the hypothesis that prioritising urban liveability or geographic accessibility is strongly connected to citizens’ ways of life, and that those ways of life are themselves infused by past and present dominant urban imaginaries. After providing the context (Section 1), we look at the controversy through a two-pronged approach. In Section 2, we analyse how the tension between leisure and mobility has existed in different forms, but consistently in substance, throughout the 150-year history of the park. Although the presence of a large volume of transit traffic has been normalised by part of the population, it has nevertheless been regularly problematised by some stakeholders since the post-war period, for various reasons, reflecting broader societal changes in terms of dominant urban imaginaries and urban ways of life. Our analysis also shows the similarities between the different spatial solutions proposed over time. In Section 3, we present an explorative survey that was conducted in autumn 2020, at a time when local authorities were testing the partial closure of the park's roadways. The survey aimed in particular to shed light on the relations between respondents’ preferences for the park set-up and their ways of life.

1. Urban liveability, geographic accessibility and ways of life

5Urban liveability can be understood as a way to interpret the concept of quality of life in cities. Chiu [2019: 4] emphasises how urban liveability is a component of the broader concept of urban sustainability, defined as “the ability of cities to reduce the environmental toll of urban activities, while improving liveability and the socio-spatial equity of their inhabitants”. In this article, we are particularly interested in the contribution of urban green space to urban liveability via the so-called ecosystem services [for example, see Stessens et al., 2022]. These include the provision of psychological, cognitive and spiritual benefits obtained from contact with nature; the setting for physical exercise, community-building and social value development; and the mitigation of environmental risks such as air pollution or heat. Not all of these ecosystem services are incompatible with the proximity of motorised traffic, but some of them are degraded by the nuisances generated by it, including air pollution, noise pollution and the risk of accidents.

6Geographic accessibility, in turn, is a concept that can be useful to understand mobility both from the perspective of infrastructure supply and of the demand for it. There are many definitions, including very broad ones such as that by Hansen [1959], in which accessibility is defined as “the potential of opportunities for interaction”, or more specific ones, where accessibility is “the extent to which land-use and transport systems enable (groups of) individuals to reach activities or destinations by means of a (combination of) transport mode(s)” [Geurs and van Wee, 2004: 128]. In general, most definitions include a land-use dimension (in which people and activities are located) and a transport dimension (how they are connected to each other). A smooth interplay between these two dimensions is a critical asset for a city and its residents, because it implies that opportunities for interactions are high, while resources spent on travel are low. In a context of car-centred urbanity, the fluidity of individual, motorised traffic is a dominant component of geographic accessibility. It ought to be noted that under these conditions, the relationship between accessibility and liveability becomes ambiguous: though a high level of geographic accessibility may well be considered as a component of urban liveability, the nuisances caused by excessive motorised traffic give rise to a trade-off between the two, at least to a certain extent [see for instance da Schio, Boussauw and Sansen, 2019].

7The tension between urban liveability and geographic accessibility is particularly pronounced in the case of Brussels, inasmuch as its mobility agenda has long been influenced by divergent visions of the city, articulating the issues of metropolitan accessibility and local liveability in different ways [among others: see Vanhellemont, 2016; Pelgrims, 2019b]. The evolution of the mobility infrastructure in Brussels can be understood as the result of three major dynamics that reflect diverging urban narratives or visions for the city to be [Pelgrims, 2019b].

8The first dynamic that emerged in the post-war period comprised the construction of a new urban landscape, separating the transport modes to allow cars to travel faster. Relevant policies were intended to increase the attractiveness of the region – geographic accessibility being part of urban liveability – but instead reinforced existing trends of an urban exodus. This shift in policy-making and urban planning can be accurately depicted by the French expression of tout automobile (everything for the car, or the car as king), referring to an approach to mobility in which priority is given to individual car use over all other modes of transport, in terms of individual behaviour, public space and the utilisation of human, technical and financial resources. This was expected to facilitate access to various areas where active transport modes were still favoured [Leloutre and Pelgrims, 2017; Pelgrims, 2019a]. Other modes could indeed exist, and even develop, provided that they did not hinder the place of cars [Héran, 2020]. Historic city centres, the first pedestrian areas and public parks were thought of as “pockets of persistence” of sustainable mobility [Schipper, 2020] – while the dominance of cars throughout the city is fully normalised, in these singular spaces, the mobility practices that are seen as “normal” are leisure-related active mobilities, and the car is merely tolerated.

9The second dynamic, emerging in the context of the post-1968 urban struggles, questioned the modernist conceptions of the city and car dominance: urban liveability was seen as being threatened by car accessibility. We observe an ambition to return to more “traditional” urban forms, in which slower motorised traffic would make it possible to share the roads and redevelop civic life. Nevertheless, the policies were focussed on Brussels’ central areas and barely challenged the dominance of the car elsewhere.

10Since the turn of the twenty-first century, in a context of urban population growth, the third dynamic invests more widely in the hitherto neglected peripheral areas. This approach relies on the acceleration of active mobility and public transport, to the disadvantage of automobile mobility. The aim is to deliver geographic accessibility without jeopardising urban liveability, among other things by upscaling mobility practices still persisting in green spaces. The overall objective is to create a more attractive and calmer urban environment. Notwithstanding the incipient effects of the second and third dynamics, the Brussels region and its outskirts can still be considered an example of car-centred urbanity. This approach to urban planning and policy also has a counterpart in terms of mobility practices, illustrated by the notion of car dependency [Dupuy, 1999]. Research has shown how car dependency results from land use characteristics (e.g. suburbanisation or amenities located far from public transport hubs) and transport systems that make it impossible to take part in certain activities within an acceptable travel time, effort or financial cost if people do not have a car. Car dependency also results from situations in which people simply find it hard to imagine life without a car, even when other modes of transport are easily accessible [Van Eenoo, Fransen and Boussauw, 2022]. Although the three dynamics are presented chronologically, there are no clear ruptures between them; on the contrary, they are combined over time, and the older ones still have great evocative power within certain social groups [Pelgrims, 2019b].

11In this paper, we hypothesise that the priority given to urban liveability or geographic accessibility is strongly connected to different ways of life, which in turn are supported by different urban imaginaries. In line with Anthony Giddens [1991], we refer to ways of life as a set of routine practices in and through space adopted by an individual in response to essential needs and to a certain expression of self. Ways of life evolve slowly with individual identities, while remaining structured by the social environment. Perceived as the result of personal motivations, tastes and values, lifestyles also remain structured by peer influence, the presence of role models and the socioeconomic context. They must therefore be thought of both as social and as individual constructs. The concept of urban imaginaries refers to abstract visions of the city that are at the same time descriptive and performative, drivers of action and objects of controversy [Genard, Berger and Vanhellemont, 2017]. Both the concepts of ways of life and of urban imaginaries are particularly useful to better comprehend the Bois de la Cambre controversy and the dynamics underlying it.

2. A history of diverging visions

2.1. The early years of the Bois de la Cambre and the rising tension between mobility and leisure

12As a first step to shed light on the controversy, we conducted historical research into the development of the Bois de la Cambre park and how it resonated with the broader evolution of imaginaries about the city, reviewing a corpus of archival material and press articles. Our analysis shows how the organisation of mobility within the park has been the subject of much controversy since its creation.

13Both the Bois de la Cambre and the Avenue Louise, which was designed in the second half of the nineteenth century to provide access to the park, were places where new urban forms of traffic separation were tested. The Avenue Louise in particular, was divided into seven corridors for each mode of transport, operating segregation not only between leisure and utility mobility, but also between the different modes of leisure mobility. The park (Figure 1) was in turn designed to offer a triple network of promenades separating pedestrians, horse riders and horse-drawn carriages. Designed by Edouard Keilig in 1862, in a part of the Sonian wood acquired by the City of Brussels, the park was laid out in the “English style” and evokes a forest atmosphere, particularly appreciated by Brussels’ elites of the time. This choice, maintaining some parts of the forest, also allowed Keilig to reduce the costs of the park’s development. Later in the century, the bicycle arrived in the city, mainly as a leisure activity of the elites. To provide space for it, a bicycle lane was added in the late nineteenth century in the Avenue Louise, between the central part and the parallel road. In the Bois though, the rigidity of the modal segregation system made it impossible to add a fourth network, and thus the park's traffic regulations were adapted to accommodate this new mode of leisure mobility. With the advent of the automobile in the early 1900s, car and motorcycle traffic was limited to a part of the carriage network [Duquenne, 1989].

14The interwar period was a critical moment in the transformation of the park’s use, as the neighbourhoods around it were developed and the speed of cars increased. Although car use was originally only linked to the recreational practices of the elites, the urbanisation of that part of the region led to transit traffic through the park. After World War II, the spatial integration of “modern roads” within existing or newly created green spaces became a major concern for landscape designers [Danneels, Notteboom and De Block, 2017]. Additionally, infrastructure projects in the city were intended to modernise the urban fabric by bringing in more daylight, space and greenery. Urban highway projects were thus accompanied by greening initiatives, although these were very rarely realised [Leloutre and Pelgrims, 2017]. Yet in the specific case of the Bois de la Cambre, the increase of transit car traffic was more the result of a laissez-faire attitude than a deliberate attempt to transform it into an entrance to the city. There were no projects to transform the roads of the Bois into arterials for automobile traffic, even during the period of fast-paced construction of urban highways in the Belgian capital (see Section 1). The roads inside the park were not even drawn on the plans for the road network to be developed in view of the 1958 World Fair. Nevertheless, the park progressively became a major gateway to Brussels from the south of the metropolitan area, as the car became more and more important in the suburban lifestyle.

2.2. Different proposals to reconcile accessibility and leisure

15It was not until 1966 that the City of Brussels set up a committee of experts to care for the urban park and restore the planting. Ten years later, the park was classified and protected as a historical site. Since then, car traffic has been prohibited in most of the Bois on summer weekends, in order to respect the historical function of the park – mainly taking a stroll. Car traffic is allowed on all other days [Duquenne, 1989].

16Motorised traffic flow, nonetheless, has been regularly questioned and several different set-ups have been proposed over time. The early proposals illustrate a traditional approach to transport engineering, giving priority to traffic fluidity. One example worth mentioning is the SOBEMAP proposal of 1982, consisting of a tunnel for the tram between the Avenue Churchill and the Etoile roundabout. This could have eased the west-east connection for car traffic and improved the speed of public transport, yet the project was abandoned because of the associated high costs and because a consensus was reached to preserve the design of the park as drawn in the nineteenth century.

17In 1985, two researchers from the Université libre de Bruxelles proposed a reorganisation of the park's road networks to better accommodate car traffic, public transport and leisure needs [Frenay and Frankinet, 1985]. In their view, the function of the Bois as a transit area for car traffic coming from the south-east of the agglomeration could not be a long-term solution, both because it limited the recreational functions of the park and because it led to an extremely chaotic situation when the Bois was closed due to strong winds. They proposed a reorganisation of the traffic (Figure 2) consisting of diverting the existing car flows via the Chaussée de La Hulpe and the Avenue Franklin Roosevelt, the construction of a tunnel in the north part of the park and a two-way access on the west side of the northern loop.

18At the turn of the twenty-first century, while remaining important, the imperatives of traffic fluidity and accessibility lost support over time in favour of arguments relating to the quality of life in the city and efforts to realise a more attractive and calmer urban environment [Pelgrims, 2019b]. This was visible in the 2012 draft of the City of Brussels Mobility Plan, which proposed to close two thirds of the park to car traffic and to implement a circulation plan very similar to that of Frenay and Frankinet. Although there was insufficient consensus to approve the plan, since then the park has been partially closed to car traffic, not only on summer weekends, but also on weekends and public holidays throughout the year.

19The most recent debate regarding the set-up of the Bois de la Cambre started during the first COVID-19-related lockdown, in spring 2020. This in fact resulted in a drastic reduction of traffic flows, due to widespread teleworking and a substantial increase in visits to parks and green spaces. In that context, the City of Brussels decided to temporarily redirect motorised traffic away from the park to increase the availability and quality of public open spaces for leisure activities. The debate became heated when the authorities declared their intention to extend the new set-up beyond the COVID-19 crisis: the park would be closed to vehicular traffic on all weekends of the year, and every day in July and August. Although the park is located on the territory of the City of Brussels, it also borders three other municipalities (Uccle, Ixelles and Watermael-Boitsfort) that asked to have their say in the matter. In May 2020, the municipalities and the regional authorities agreed to reopen a few access roads for motorised traffic through the northern part of the park (Figures 3) and to work out a new mobility plan to enable the public to take full advantage of the Bois de La Cambre and at the same time guarantee a north-south and a west-east connection for car traffic, though not necessarily through the park. Different options were proposed by the City of Brussels, leaning towards a reduction of motorised traffic through the Bois and the Municipality of Uccle, which was concerned with keeping transit traffic outside of the surrounding neighbourhoods. Between 15 September and 15 November 2020, one of the options proposed by the City of Brussels was tested – opening the western portion of the northern loop in both directions for motorised traffic, as shown in Map 3d. This resulted in a minor increase in travel times [Espaces-Mobilités, n.d.]. After a legal action initiated by the municipality of Uccle, the Avenue de Boitsfort from the southern loop was reopened on weekdays (Figures 4), which was still the position at the time of writing.

Legend 
— Motorised traffic 
— Active mobility 
--- Parking areas 
— Public transport, school bus, emergency vehicles 

Legend 
— Motorised traffic 
— Active mobility 
--- Parking areas 
— Public transport, school bus, emergency vehicles 

20The organisation of different mobilities within the park has thus been the subject of much controversy over time. Motorised traffic flow has never been the result of an explicit plan, but the consequence of a laissez-faire approach. Nevertheless, multiple alternative setups, including limitations to traffic and diversions, have been proposed since the late 1960s. These alternatives reflected societal evolutions and differing urban imaginaries, encompassing evolving levels of attention paid to ecology, heritage, car traffic fluidity and, most recently, more attention paid to the recreational use of the park. The next section will analyse the extent to which support for or opposition to motorised traffic reflects the ways of life of today's users of the park.

3. A survey of preferences for the contemporary set-up of the park

3.1. Data collection

22This exploratory survey has two important limitations. First, although the sample was large, it was not representative of the population of the Brussels’ region, the metropolitan area or the neighbouring communes (see Table A1). Both the very conflictual topic and the way the survey was distributed lead us to presume that more extreme opinions on the matter are overrepresented. Second, the survey took a “snapshot” of the controversy at a particular time – one at which the public debate was particularly heated and polarised, due to the changes in the park's set-up. Within a different timeframe, the same survey might have produced a more nuanced picture. As a consequence, this article is limited to the analysis and comparison of the profiles of different groups, against the background of the liveability versus accessibility controversy, and as a complement to the historical analysis. A more extensive presentation of the methods, data and results of the descriptive analyses was published as a separate report [da Schio, Pelgrims et al., 2021].

3.2. In favour of or against motorised traffic

23One of the central questions in the survey concerned the extent to which people wanted motorised traffic to be allowed in the park. The options presented were: a) full access, b) access limited in time or c) space, and d) no access at all for motorised traffic. Respondents were asked to rank these options in order of preference. Based on this question, three groups of respondents were identified with similar preferences concerning the set-up of the park.

24The first group (n=2779) labelled “Allow Traffic”, represents the respondents who strongly believe the park should be open to (transit) motorised traffic: they all picked the unlimited use by motorised traffic as their first choice (100 per cent) and the banning of motorised traffic as their least favoured (100 per cent). In the wording of the question, motorised traffic did not refer to parking in and around the park, but to traffic passing through it.

25The second group (n=2346), consists of respondents who prefer a partial restriction of traffic in the park, either geographically or temporally. The area-bound restriction is the most popular measure among this group (the favourite option for 48,9 per cent of respondents) and 43,4 per cent chose the two forms of partial restriction as their number one and two options (implying that the preferences of this group are not that homogeneous). We labelled this group “Middle Ground”.

26The third and final group (n=2128), comprises respondents who represent the mirror image of the first group. They strongly believe that motorised traffic should be banned from the park: everyone in this group chose the banning of traffic as their first option (100 per cent) and the admission of traffic as their least favourable choice (100 per cent). Hence, we labelled this group “Ban Traffic”.

27When enquiring about good reasons to maintain/reduce car traffic in the park, the Allow Traffic group strongly agrees with the different reasons related to fluidity of traffic both to and from Brussels and – at a more local scale – avoiding an increase of car-related nuisances in surrounding neighbourhoods. Respondents were more divided concerning the idea that driving in the park is enjoyable. In the Ban Traffic group, reasons related to the recreational function of the park are stressed somewhat more than reasons related to mobility or nature protection.

3.3. Linking preferences to personal background and habits

28To examine whether different groups have different profiles in terms of background characteristics and habits, a multinomial logistic regression was performed with the Middle Ground group as the reference category. The regression parameters of the model (see Table 1) indicate the extent to which respondents in the Ban/Allow Traffic group are more (positive sign), less (negative sign) or equally (non-significant) likely to demonstrate a particular background characteristic (e.g. age, owning a garden) compared with the Middle Ground group. When the Ban traffic and Allow traffic groups present opposite (and significant) parameter signs, they resolutely differ on that characteristic, with the Middle Ground group taking an intermediary position. Equal signs indicate that the Ban Traffic and Allow Traffic groups differ in the same sense from the Middle Ground group.

Table 1. Logistic regression parameters estimating the likelihood of Ban Traffic respondents or Allow Traffic respondent to manifest a certain characteristic compared to the Middle Ground group (reference category) 

	Ban Traffic vs. Middle Ground				Allow Traffic vs. Middle Ground
	b		SE		b		SE
Intercept	-,44		,33		,60	**	,22
Age
-25	-,73	**	,25		-,05		,29
25-34	-,22	*	,10		-,13		,11
45-54	-,35	***	,10		-,19	*	,09
65+	-,43	***	,13		-,07		,09
Professional status
Self-employed	-,40	***	,10		-,06		,08
Retired	-1,05	***	,17		-,32	**	,12
Unemployed	,09		,25		-,35		,32
Student	-,09		,26		,11		,32
Other	-,13		,18		-,05		,15
Living < 800m La Cambre
Yes	-,39	***	,09		,33	***	,08
Visit green spaces
Very often	,50	*	,23		-,23		,12
Regularly	,22		,24		-,06		,13
Residence
Pentagon and first belt	,67	***	,19		-,39	**	,13
Second belt	,07		,18		-,23	*	,11
Other	,03		,29		,08		,16
Place of work
Pentagon and first belt	-,23		,15		-,26	*	,11
Second belt	-,13		,16		-,13		,12
Other	-,30		,30		-,08		,24
Physical difficulties
Yes	-,13		,16		,28	**	,10
Regular transportation usage
Car	-,86	***	,08		1,08	***	,10
Walking	-,19		,15		-,03		,10
Bike	,56	***	,09		-,58	***	,08
Public transport	,00		,08		-,31	***	,08
Use of La Cambre
Through-fare by car	-1,31	***	,12		,47	***	,07
Through-fare by bike	,49	***	,08		-,71	***	,10
Visit week	,65	***	,09		-,87	***	,10
Visit weekend	,21	*	,09		-,45	***	,08
Visit < 1/month	,24	*	,12		-,36	***	,10
Station car	-,87	*	,37		,19		,19
Access to greenery
See trees	,00		,09		,18	*	,08
Private garden	,10		,08		-,20	**	,08
Friend's garden	-,15		,11		,10		,10
Public space with trees	-,03		,09		-,06		,08
Park or public garden	-,04		,08		-,11		,08
Other	-,40	*	,20		-,13		,17
No access	,11		,16		-,20		,18

Note : ^* p < .05, ^** p < .01, ^*** p < .001. n = 6392, 𝛘²(72) = 4727, p < .001, R² (Cox&Snell) = .50

30Respondents were also asked about their frequency of use of different modes of transportation over the preceding 12 months. A comparison with the BELDAM mobility survey [Cornelis et al., 2012, data for 2010] shows that the mobility practices of our sample differ greatly from the Brussels population as a whole. This is unsurprising, as already mentioned. Some 90 per cent of the Allow Traffic group acknowledged using individual motorised modes of transport at least one day a week (and 67 per cent, more than five times a week). On the opposite side, 75 per cent of respondents from the Ban Traffic group reported using a bicycle (or equivalent) at least one day a week (and nearly half of them at least five times a week). By contrast, in the BELDAM survey only 14 per cent of the respondents declared travelling by car between one and four times per week, and 22 per cent more than five times per week; 8 per cent declared traveling by bicycle between one and four times per week, and 5 per cent more than five times per week.

31All things being equal, the modal choice is a distinctive characteristic of the different groups. In comparison with the Middle Ground group, the Ban Traffic group used a bicycle more often (b = 0,56) and a car less often (b = -0,86), whereas the opposite is true for the Allow Traffic group – using the car more often (b = 1,08) and the bicycle less (b = -0,58). This demonstrates that modal choices cannot be fully explained by residential choices: Ban Traffic respondents tended to use a bicycle more often than respondents from the other two groups, regardless of their place of residence.

32The survey also collected data on respondents' relation to green spaces: the type of green spaces respondents usually have access to and their habits in relation to them. The (overall small) proportion of people lacking access to green spaces is three times larger in the Ban Traffic group than in the Allow Traffic group, with the Middle Ground group taking an intermediary value. These results are mirrored in the question asking whether the respondents could see trees from their window: 65 per cent of Ban Traffic respondents gave a “yes” response, versus 75 per cent of Allow Traffic respondents. A similar picture appears when looking at the availability of a private garden: 38 per cent of the Ban Traffic group has a private garden and 10 per cent can access the garden of a friend; for the Allow Traffic group, these percentages are 59 and 15, respectively; the Middle Ground group again takes a position in between. However, the regression analysis shows that these differences between the groups can be largely attributed to background characteristics (municipality, age, etc.).

33Focussing on the actual use of green spaces, descriptive analyses suggest that the ambition to enhance the leisure and ecological functions of the Bois de La Cambre goes hand in hand with the frequency with which people visit green spaces. While a large majority of respondents in all groups (85 per cent or more) indicate visiting green spaces at least several times per month, this proportion is the largest in the Ban Traffic group (96 %), followed by the Middle Ground group (92 %) and the Allow Traffic group (85 %).

34With regard to the use of the Bois (Figure 5), the Allow Traffic respondents, who were used to passing through by car or motorcycle (92 per cent) in the period prior to the partial closure, still tended to perceive the park more as a car mobility axis for transit (b = 0,47) and seemed less likely to visit it during the week (b = -0,87), or even at the weekend. The partial closure of the park even seems to have reduced their tendency to visit the park at weekends (from 45 to 32 per cent). The Middle Ground respondents were also affected by the partial closure, in terms of how often they went through the park by car or motorcycle (from 54 to 32 per cent). Whether they visited the park at the weekend seems unaffected by the partial closure, but they made use of the opportunities the traffic ban entailed: the likelihood of visiting the park during weekdays seemed to have increased – from 30 per cent prior to the partial closure, to 38 per cent after. Lastly, the reorganisation seems to have encouraged Ban Traffic respondents’ use of active modes of transport to go through the park (49 per cent prior to partial closure, 64 per cent after). In comparison with the Middle Ground group, they were less likely to cross by car (b = -1,31) and more likely to cross by bicycle (b = 0,49). The partial closure also led the Ban Traffic respondents to visit the park even more than before: from 69 to 78 per cent at weekends and from 41 to 64 per cent during weekdays. Taking other variables into account, they were more likely than both other groups to visit the park during the week (b = 0,65).

35The profile of the Ban Traffic group is also weakly marked by the professional status of the respondents, more often being employees. In terms of age structure, it emerges that respondents in the Allow Traffic group are older (average of 50 years; with the mode being: 55-59 years), followed by the Middle Ground group (average of 46 years; mode: 45-59 years) and then the Ban Traffic Group (average of 41 years; mode: 30-34 years). The regression analysis shows that – taking other differences into account – the ages 35 to 44 are more strongly represented in the Ban Traffic group. The Middle Ground group is also distinguished from the other groups in terms of age, as the former are more likely to be between the ages of 25 and 34. The profile of the Allow Traffic group is somewhat marked by their physical health (more likely to have physical difficulties because of permanent handicap or age, using certain means of transportation).

Conclusions

36The case of the Bois de la Cambre illustrates the tensions that can operate in (peri)urban green spaces between geographical accessibility and liveability. While conflicting visions for the park emerged again in 2020, they were not a new phenomenon and have characterised the history of the Bois de la Cambre. Every redevelopment proposal made since 1966 has induced strong reactions, either positive from those who see their opportunities expanded, or negative from those who feel them to be more constrained. The last reorganisation of car traffic in the park, proposed in 2020, is in line with the evolution of recent decades, even if the way in which motorised traffic has been problematised over time has changed. The archival analysis suggests that the adaptations that were made over time to the layout of the park reflect changing leisure and mobility practises and the evolution of broader ways of life, rather than being a consequence of an explicit project for the park. However, this laissez-faire attitude resulted in an incremental transformation of the main parkways into major roads, perceived by some as a socio-environmental problem and by others as a necessity vis-à-vis the lack of alternative access routes to the city from the southern periphery. The different claims that are made today regarding the use of the park are strongly connected to residential location, habitual modes of transport and the use of and access to green spaces. These practices reflect divergent lifestyles that are inspired by dominant imaginaries of mobility and the city, and that evolve slowly. In that sense, our analysis stresses the need to work both on the elaboration of concrete alternatives – concerning more sustainable transport options, the affordability of urban residential locations, and/or the geographic relation between places of residence and places of work – and on the construction of new urban and mobility imaginaries. Rather than leading to concrete recommendations, our research highlights the eminently political nature of the controversy and of its possible solutions, whatever they may be. In this context, further research could explore in greater depth the social practices of those who live nearby, visit and cross the area, in order to co-construct a scenario for a sustainable future for the Bois de la Cambre.