Rethinking the place brand: the interactive formation of place brands and the role of participatory place branding

This article attempts to ‘rethink’ place brands after examining in detail how people form them in their minds. The article starts with a very brief account of the place branding literature to provide the necessary background and goes on to identify what we see as a shortcoming in current understanding of the place brand: the dominant idea that brands are formed as sums of mental associations. The article attempts to take current understanding of place brands further by going beyond associations and adding a missing element: the interactions between those associations. We propose a rethinking of place brands based on two pillars: first we incorporate more geographical understanding into place branding and, second, we outline a process that allows place elements and placebased associations to combine and form the place brand. The place brand formation process starts when people use place-making elements (materiality, practices, institutions and representations) to form mental associations with the place. These associations are not static but evolve and change over time as they interact with each other on several dimensions. These interactions constitute the way in which the place brand is formed. The argumentation leads to a novel conceptualization of the role of place branding in the above processes. The practical applicability and implications of the proposed rethinking of place brands suggested here are explored in detail through the examination of the branding process followed recently in Bogotá, Colombia, where our approach to place branding has found practical application.

city-state slogans (eg, 'Uniquely Singapore') and city slogans (eg, 'Only Vegas', 'One Leicester'). A few years ago the Spanish region of Andalucía also used the word 'one' in its campaign but with a significant difference. The slogan used was: 'there is only one Andalucía: yours!' The word 'yours' added after the claim makes it a very different proposition and it is, in essence, this difference that is explored here. The place branding claim of Andalucía was not that the uniqueness of the place lies in the place itself but that it can only be found in the interaction and relationship between the (unique by definition) place and the (also unique) individual. Such an uncommon understanding, which is largely missing from place branding practices, is the heart of the argument developed here, leading to a reassessment of the nature of place brands and of the role of place branding.
The main argument developed in the paper is that there is a need to rethink the way in which place brands are conceptualized, based on a more refined understanding of place and how this is constructed. The paper integrates different understandings of places and place brands in order to arrive at a refined view of how place brands are formed. We start by identifying a current limitation within place brand definitions: namely the reliance on associations that people hold with the place as elements that constitute the place brand. We claim that there is a lack of geographical understanding of the elements on which these associations are based. We therefore turn to geography and seek an understanding of how place is perceived and constructed through a process of 'synthesizing' the various place elements that have been identified as: materiality/physical elements, practices/structures of social interactions, regulating institutions, and systems of representations. We suggest that this synthesis is based on a series of interactions between associations people hold with the place that actually allow people to internalize the associations and let the place brand form. To put it simply, the constituents of place are simultaneously the constituents of the place brand through the associations they cause and the place brand comes together as a whole through the interactions between associations. This view of place brand formation is based on the concept of brand cocreation as this has been developed within general branding and leads to a more participatory approach, which is illustrated and explained further in our examination of the case of Bogotá, Colombia. The case highlights practical challenges thus enriching understanding of practical application of this participatory view of place branding.

Brief overview of place branding-what we know we don't know
The potential, specificities, and pitfalls of place branding have been explored in relative detail in the literature (eg, Anholt, 2007;Ashworth and Voogd, 1990;Braun, 2008;Govers and Go, 2009;Hankinson, 2004;Kavaratzis, 2004;Lucarelli and Berg, 2011). The problems inherent in translating an originally commercial tactic (ie, branding) into a tool for place development (ie, place branding) occupy a large part of the literature. These problems refer most commonly to the complexity of place as a branded entity (Julier, 2005;Kavaratzis and Ashworth, 2005), the lack of control over this entity (Braun, 2008;Hankinson, 2004;van Berg and Braun, 1999) and the multiplicity of a place's stakeholder groups and their conflicting interests (Houghton and Stevens, 2010;Olsson and Berglund, 2009). Other commonly examined issues relate to the peculiarity and complexity of the ways in which place brands are formed (Kavaratzis and Hatch, 2013;Warnaby, 2009) and, therefore, might be managed (Govers and Go, 2009;Hanna and Rowley, 2011). Further issues that are repeatedly noted in the literature are the gap identified in understanding the relationship between place branding and place identity (Kalandides, 2011a;Mayes, 2008), the lack of an appropriate instrument for the measurement of the effectiveness of place branding (Zenker, 2011; as well as the additional challenges in branding different types of places (Warnaby et al, 2010). Several commentators adopt a more or less positive attitude towards place branding, attempting to examine these issues in order to provide solutions as to how they might be solved and lead to a more responsible place branding practice (eg, Aitken and Campelo, 2011;Anholt, 2007;Braun, 2011;Govers and Go, 2009;Hankinson, 2009;Hanna and Rowley, 2011;Kalandides, 2011b;Kavaratzis, 2009;Murray, 2001;Warnaby, 2009). At the same time, a significant number of publications adopt a more critical stand, examining several dubious aspects and revealing implicit agendas behind place branding exercises (Broudehoux, 2001;Colomb, 2011;Evans, 2003;2006;Greenberg, 2008;Julier, 2005;Miles, 2010;Paddison, 1993). Perhaps the most common critique is that place branding is an instrument used by urban elites to legitimize their own strategic decision making in the wider context of the hegemonic project of neoliberal urban governance (see Colomb, 2011). As Broudehoux (2001 critically points out, place branding can be described as a field where "dominant groups use spatial and visual strategies to impose their views". Within this logic, Miles (2010, page 46) asserts that "cities are reduced to an idealized vision of a prosperous future that has scant regard for who might be the losers." It becomes obvious that different approaches are evident in terms of place branding's role, functions, and potential. Furthermore, there is no agreement-and perhaps no understanding-as to how place brands are formed.
As Braun (2011) notes, the considerable confusion around place brands stems, at least partly, from the lack of a clear conceptualization of brands within the mainstream of marketing studies. Brands are indeed complex phenomena approached in different ways (Batey, 2008). As Kornberger (2010) describes, for some the brand is a managerial device controlled by companies (eg, Aaker, 1996) while for others the brand is much more strategic and drives the whole existence of corporations (eg, Hatch and Schultz, 2008). A further approach views brands as symbols that create meaning for consumers (eg, Danesi, 2008). Important for the argument developed here is another approach that views 'brands as media' (eg, Lury 2004), adding a relational element into brands and suggesting that brands are institutionalized structures. For Lury (2004) brands are complex objects with multiple dimensions that make them performative, distributed, and relational objects; a notion that we will return to below.
The recent turn towards a service-dominant logic for marketing (Vargo and Lusch, 2008) and its transfer to branding (Mertz et al, 2009) highlights the importance of stakeholders, which stems largely from the way in which value is created. Two elements of value creation are important for our needs here. First, that value is of an experiential nature (Vargo and Lusch, 2008), which means that it is created in experiences rather than images or emotions. Secondly, "value is always uniquely and phenomenologically determined by the beneficiary" (Vargo and Lusch, 2008, page 9). This means that the way in which people create value is unique and might be, to an extent, disconnected from an objective reality; a notion captured in our theoretical framework below. Such a stakeholder-centred approach to place branding is slowly gaining ground with a strand of publications that deal directly with the consequences of such thinking for place brands (Warnaby, 2009), the role of stakeholders (eg, Braun et al, 2013;Houghton and Stevens, 2010;Merrilees et al, 2012), and with the dynamic nature of place brands (eg, Aitken and Campelo, 2011;Kavaratzis and Hatch, 2013). We believe that a clarification of brand formation is essential for a better understanding of the nature of brands, so we turn our attention to place brand formation.
Associations as place brand elements-the things that place brands are made of The mental associations that people form with brands are attributed crucial significance and, very commonly, brands are defined as a set of associations or the sum of associations. For instance, for Batey (2008) a brand can be defined as a cluster of associations concerning attributes, benefits, and values. Indeed, one of the most widely cited definitions of brands (Aaker, 1996, page 68) is as "a ... multidimensional construct, consisting of functional, emotional, relational and strategic elements that collectively generate a unique set of associations in the public mind." This definition has been widely accepted within place branding studies (eg, Kavaratzis, 2004). For instance, Hildreth (2011, page 156) asserts that "the overall brand strength-if it is anything at all-is merely a summation of the number, type, quality and positivity of associations people have of the place." In contemporary understanding, the place brand is thought to exist in the associations people make with the object of branding. What this means is that associations might well be the 'raw material' people use to form place brands in their minds or, in other words, brands are 'made of' associations.
However, there remain two very significant gaps in our understanding of place brand formation. First, it is unclear what types of elements people base their associations uponwhat they choose from the place itself and what they might invent in order to form place-related associations. Second, it is not clear how these associations might operate collectively-how they link to each other in order to form the place brand. For example, we use the name 'Milan' and denote a rather clear concept in our heads, though the place itself is quite muddled and has changed dramatically down the centuries. What is quintessentially Milan? Is it the Piazza del Duomo, the Brera district, the popular housing blocks in the periphery? Is it rather the shopping opportunities that, arguably, tourists engage in more than locals? Might the real Milan form in the inevitable link to everything Italian (food, style, temperament, etc), although, for many Italians, Milan might lie a bit too far north for such 'Italianness'? What the currently dominant understanding of place brands as 'sums of associations' implies is that all these (and many more) Milan-based associations collectively make the Milan brand simply by their existence within a set. This, however, seems to be a rather big leap from the mere existence of such associations to their collective effect.
The problem is that associations cannot be 'added up' in a sum because they are not static. Instead, they are in a state of constant flux as they interact with each other in complementary or conflicting manners and as the person holding those associations continues to experience the place brand. It is, therefore, important not to limit our inquiries to the level of associations, which are simply attributes but to understand what makes the brand a coherent whole. Helpful in this effort is the recent idea that brand associations exist in networks where they activate and influence each other (eg, Schnittka et al, 2012;Zenker and Beckman, 2013). This highlights that associations are not isolated but are mutually dependent and mutually activated as they operate in a 'network' (rather than a 'sum'), leading to the experience people have with the brand. Brakus et al (2009) suggest that brand experience has several dimensions and conceptualize it as the "subjective, internal consumer responses (sensations, feelings and cognitions) and behavioural responses evoked by brand-related stimuli that are part of a brand's design and identity, packaging, communications and environments" (page 53). Thus, what contemporary understanding of brands suggests is that brand-related stimuli give rise to associations that operate in a network that, in turn, gives rise to the brand. This has been successfully captured in recent publications on place branding (eg, Zenker and Beckman, 2013). In the case of place brands, associations will inevitably be based on place elements and perceptions. It is, therefore, necessary to tap into such geographical knowledge.

Constituents of place-putting the place into the place brand
If place brand associations are based on those elements that allow people to make meaning of places, it is important to take a close look at the elements that constitute place and address some of the tensions that arise when place identity is examined within the framework of place branding.

Understanding place and place identity
Place identity is probably the most elusive and paradoxical of the concepts that make place branding a particularly challenging endeavour (Boisen et al, 2011;Govers and Go, 2009;Kalandides, 2011a). Etymologically, identity simply means 'that which is the same to itself ' and such a definition probably sounds straightforward enough. Yet, what is 'itself ' in something as heterogeneous as place and how can anything ever be the same in a world caught up in constant change (Warnaby, 2011)? There is a clear need for a concept of place (and place identity) that incorporates change and continuity, unity and heterogeneity (see also Kavaratzis and Hatch, 2013;Ren and Blichfeldt, 2011). Weichhart et al (2006) distinguish between three types of place identity: identification of, being identified as, and identification with. The first refers to the ways in which people (groups or individuals) understand and recognize places, as they do other objects, assigning them characteristics and particularities. The second ('being identified as') in an inverse way refers to the ways in which people (again both groups and individuals) are recognized in their relations to their place of origin, residence, etc. And finally, the third ('identification with'), following the phenomenological tradition that seeks to explore the links between the human and the world in which it lives, is about the ways that people incorporate place into their own identity construction. Zenker and Petersen (2014) have addressed the issue of identification in a place branding context, concluding that the complexity of place brands actually increases the chances for identification with a place. What Weichhart et al (2006) consciously avoid doing is to talk about place identity as something independent of the human. "The material world", they claim, "cannot have an identity of its own" (page 19). This view is persistent within environmental psychology, where place identity is defined as a substructure of selfidentity (eg, Proshansky et al, 1983) although relations with others are also thought to be important in shaping place identity (eg, Knez, 2005). Nonetheless, if we understand identity as the "undifferentiated unity or sameness, one that constitutes the essential 'being' of an entity" (Martin, 2005, page 97), we may be entitled to use the term 'place identity', either with one of the three meanings that Weichhart et al give it or in the sense of place specificity and distinctiveness. The question that arises though, if we accept that place can have an identity of its own, is how do we understand place?
Place is clearly a notion that entails several meanings and has been conceptualized in a variety of ways by different strands of geographers. Cresswell (2004) accurately summarizes these as: (a) the descriptive, ideographic approach to place (interested in the particularity of places and the world as a set of places), (b) the social-constructionist approach to place [with its several strands that see places "as instances of more general underlying social processes … such as capitalism, patriarchy, heterosexism" (page 51)], and (c) the phenomenological approach to place (interested in highlighting the necessarily spatial essence of human experience). A major concern of geographers-and clearly relevant to place branding-is the way in which place is invested with meaning. Influential ideas have included Tuan's (1974) concept of 'topophilia' (examining how people attach to places), Relph's (1976) insightful analysis of 'authentic sense of place' (as an unmediated experience of place identity in its whole complexity), Lefebvre's (1991) triad of the 'production of space' (distinguishing between spatial practices, representations of space, and representational spaces). Significant for our argument here is the understanding of place through processes of change and progress; particularly Massey's (1994) concept of 'place as progression' (examined in detail below) and Thrift's (1996) 'place as practice' and 'nonrepresentational theory' (focusing on understanding place through movement, events and practices). As Cresswell (2004) asserts, current conceptualizations of place encapsulate the notion of transformation and change and in this sense, "places are never finished but always the result of processes and practices" (page 37).
We find two approaches particularly useful for refined understanding of place brands. The first is the suggestion of Massey (1994) that place should be conceptualized as the locus of interconnection of open-ended trajectories. Both people and objects exist simultaneously, but at the same time carry in them their own history (trajectories) that may come from far away and long ago. These trajectories may not be unique in themselves, but the complex way in which they intersect in that particular locus is quite singular. Place is both the product of social phenomena and a modus of their reproduction. Yet, social relations, argues Massey, are not neutral, but rather are caught up in power relations, subsequently reproduced or contested in places. In other words, it is social relations that produce places and places have the capacity to reproduce these relations in an endless movement. Seen this way, place becomes the simultaneous existence and reproduction of difference, and it is that unique power-imbued blend of different trajectories that gives place its specificity and distinctiveness. The second approach is the understanding of place as at the same time absolute, relative, and relational (Harvey, 1996). It is absolute, in the sense that it can be limited, divided up, and measured; it is relative in the sense that it is constituted by relations among objects; and it is relational, in the sense that each object contains in itself its relations to other objects across space. Lury has convincingly argued for a similar understanding of brands in her examination of brands as boundary objects and in her analogy of the brand with a car. Lury (2004) describes that, like the car, the brand (and the place) is a distinct entity and we accept its uniformity (Harvey's absolute character) although it consists of thousands of small parts that make it what it is (Harvey's relative character). As Lury goes on, the car as a distinct object that consists of many smaller objects acquires its meaning only in relation to the environment: the road, the driver, the route, etc (Harvey's relational character). The crucial point is to understand the simultaneity of the three aspects, which means that they cannot be understood, studied, or influenced separately.

Constituents of place
The argument in this paper is that place (and its specificity) is constituted in two dimensions and in a double process both dependent on and independent of human perceptions. The German sociologist Löw (2001) talks about spacing (or the placing of social goods in the material sense) and synthesizing (or the process of linking things together in unity in the human mind). In this sense place identity is both what is 'out there' and how that is perceived, the two being indissoluble. Place image (or the collective form of mental perceptions of place) becomes an integrated part of place identity and cannot be juxtaposed to it. Indeed, Kavaratzis and Hatch (2013) see place image and place identity as two sides of the same coin, neither of which has meaning without the other and they propose situating the place brand within this process of place identity formation. Considering (à la Harvey) all aspects of place and their interrelations is of the utmost importance for place branding.
For this-and to answer the question of which place-making elements give rise to placebased associations, it is worth considering the constitutive elements of place. We use here the four constitutive elements of place as suggested by Läpple (1991) as a comprehensive classification: (1) The first element is materiality. This refers to the material-physical substrate of social relations, as the material external form of social space. This socially produced substrate consists both of place-bounded artefacts and of the human body. It also functions as crystallized history and materializes collective memory.
(2) Secondly, place is constituted by practices. These practices are the structures of social interaction in relation to the material substrate. This includes production, use, and appropriation of materiality and relates to differentiation of class and other divisions.
(3) Thirdly, we have institutions. This refers to an institutionalized and normative regulation system as mediator between the material substrate of social space and the social practice of its production, appropriation, and use. This regulation system consists of forms of property, power and control relations, legal regulations, planning guidelines, and social and aesthetic norms.
(4) The fourth element is representations. These are the spatial systems "of signs, symbols and representations linked to the material substrate" (Läpple, 1991, pages 196-197). These not only include the formal conceptions of place, such as place names [for a detailed discussion of toponymy and its relation to place brands see Medway and Warnaby (2014)], or maps and plans, but also extend to all structures and elements that intentionally convey meaning related to the place, including "certain forms of narratives which encapsulate selected readings of the environment, as in promotional literature … and so on" (Meethan, 2001, page 37).
Power relations on the one hand and time on the other permeate all of the above, creating constant tension and change. Both these issues are captured and illustrated very effectively in Warnaby et al's (2010) examination of the marketing of Hadrian's Wall in England, with the high number of stakeholders involved and, particularly, with the challenges caused by the Wall's "diminished materiality" (page 1366). The four place-making elements are again inseparable and a place can be understood only if they are considered together (in this sense, it is better to talk of 'coconstituents of place'). Finally, there is a separate process, where all these elements are synthesized in the human mind to form what we call 'place'. Parallel to this place-making synthesis is a process of place brand formation which synthesizes the associations with the four place constituents into a place brand (see figure 1).
If the desired result is to influence the latter (the synthesis-ie, the perceptions in the human mind), this is usually attempted by working on the system of representations. This seems to be the preferred tactic for place branding practitioners who focus on logos and advertising campaigns. It is also intensely evident in the creation of landmarks (eg, the Bilbao Guggenheim), which despite relating to materiality are most commonly employed in order to offer desired representations (Ashworth, 2009). A decision to ignore the rest of the four Figure 1. Coconstituents of (a) place and (b) elements of place brand.
(a) (a) elements, however, is not an option as this ignores their simultaneity. One possible argument could be that place branding is not responsible for these elements: materiality is the field of urban planners and architects; institutions are the remit of legislators and politicians. This is a view that has found its way into place branding practice as well as part of the literature that actually disconnects place branding from its planning aspects, focusing on a purely designbased and visual-identity-based view. This ignores the four coconstituents of place, and is, in our view, highly problematic, inherently oversimplifying and ultimately irresponsible as it disconnects the 'place' from the 'brand', thus going against the very nature of place brands. As Anholt (2007) asserts, there is no evidence that logos and slogans or advertising campaigns change place images or perceptions.

Internalizing and processing place brands-interactive associations
In the above we have tried to answer the 'what' question of place brands. However, immediately the 'how' question arises: what process allows the associations (which are only attributes) to become a brand (a whole)? Is the 'summation of associations' on which place branding literature focuses on possible or is there a more elaborate process that allows the place associations to form the place brand? As important or indeed vital as associations might be, they remain attributes of a bigger whole and cannot be the whole itself. The process of synthesizing as described above is in essence absent from the understanding of place brands as 'sums of associations'. Furthermore, the relational elements of place are also ignored. If only associations are taken into account, they do not explain the process of synthesizing, neither do they describe the relational aspects of place brands. As seen above, the 'brand as network of associations' is a suggestion that takes us further as it captures the nature of the brand formation process in a refined manner. Building on such understandings, we take a closer look at the mechanics of place brand formation. The place brand as 'network of place associations' idea implies that place associations are interactive and they constantly change as the place brand is experienced; similarly to what Latour (2005) terms 'circulating references'. As advocated by actor network theory (eg, Murdoch, 1998), all things in both the natural and the social worlds are continuously generated as an effect of the webs of relations within which they are located (Law, 2009); something we argue here for the place brand as well. This sees place brands as ongoing, multiple, open, and rather unpredictable, going against the dominant understanding of place brands, which sees them as set and fixed. The analysis of place brand formation undertaken here reveals that there is no singular, objective, essential place brand to be 'built' or uncovered. Only different valid versions of it "that can be neither entirely reconciled nor dismissed" (van Duim et al, 2013).
This ongoing and multiple process of place brand formation is similar and interconnected to the process of synthesizing that allows people to make sense of places and it is a process of interactions that occur between the associations. In other words, place brands elicit in people's minds elaborations over their meaning through associations that are internalized and elaborated upon through a complex and ongoing process of interactions between these associations. It is these interactions that form the particulate and constantly fluctuating whole of the place brand in a wide process, that entwines partial social and material processes. Figure 2 depicts this process showing the progression from place elements to a synthesized place brand. What the figure describes is a process of how the four constituents of place (the oval shapes of the outer edge of the figure) lead people to form associations in their minds (the rectangles of the next level), which then activate a complex set of interactions between them (the dashed arrows of the centre) through which people internalize their associations and form the place brand (the inner circle).
The interactions of associations that people hold with places and place brands occur along dimensions as people encounter the elements that constitute the place brand. In the words of Borowsky et al (2001, page 440): "cultural understandings combine with social organization, cognitive processes, emotional experience, material conditions and power relations, among many other analytically separate phenomena, to affect each other and shape outcome." The place-constituting elements that particular people associate with the place brand will be about one or more of several possible functions that the place performs for them. This could be what-if anything-the place means for people, whatif anything-it adds to people's lives, how it makes people feel, how it connects people to each other or not, and how it affects their relationships, how it helps people construct their identity or not, and many more such functions. The evaluation of the attributes takes place in an interdimensional manner as the associations that people hold with a place on a certain dimension interact with associations they hold on a different dimension. The one influences the other, thus constantly changing the evaluation, which can therefore never be final but is always under reconsideration.
In other words, the formation of place brands is based on the unique ways in which people use different place elements to form their associations and on the unique ways in which these associations interact with each other and, therefore, change constantly. It is actually in these interactions that the place brand is formed and evaluated. People encounter place-making elements as they experience the place and this experience can be direct (firsthand) or indirect (through other people); it can be from near or far; it might be mediated or unmediated; and so on. The source of the associations is a significant determinant of their influence and strength. It is this source that might make certain associations more widespread or more persistent than others and might let some associations 'linger' although they might have become obsolete.

Participatory branding in practice-branding Bogotá
We discuss the practical implications of this 'rethinking of place brands', using a case study of Bogotá, Colombia (see Kalandides, 2011b) that was based on the theoretical framework. A critical reflection on the ways in which the project was guided by the theoretical framework as well as the challenges faced is useful for a wider reflection on the participatory place branding approach. (1) The project in Bogotá approached place branding as a participatory process initiated by a legitimate body. The two parts of the project relevant here were the analytical and strategic parts, both of which relied on the theoretical understanding developed here. In order to capture Bogotá's identity, the project relied on a combination of Massey's (1994) definition of place identity as the unique way in which different open-ended trajectories intersect in a locus with Läpple's (1991) distinction of the four place constituents. These were used to guide the identification of the associations people hold with Bogotá as a brand. The four place constituents were combined with the notion of the interactions between associations in order to examine their evolvement over time. Thus the theoretical background was instrumental in several choices made during the project: it dictated the research methods, it guided the recruitment of participants, it determined the themes to be discussed and the main objectives of all stages of the analysis, and it also influenced the selection of measures to be implemented. This serves as a demonstration of how the conceptualization of place brand formation influences and guides implementation of all stages of place branding projects. In the more common approach that sees place branding as the development of promotional devices and identity claims, projects are clearly top-down rather than bottom-up (eg, Bennett and Savani, 2003;Merrilees et al, 2012) and exclusive rather than participatory (eg, Braun et al, 2013;Houghton and Stevens, 2010).
Three methodologies helped capture the place-constituting elements for Bogotá. First, systematic observation in the offices of the local authorities and major stakeholders but, importantly, also in the city's streets aiming to clarify the project's own aims and to understand the particularity of the context. Secondly, content analysis of websites, policy documents, academic journal articles, selected press articles, etc that identified the most talked-about elements of Bogotá in terms of landscape (materiality), institutions as well as projects and initiatives. Finally, more than eighty interviews (both in-depth and focus groups) realized the principle of 'as broad participation as possible'. Interviewees were chosen on the basis of the four place constituents and were local stakeholders in different fields of activity or expertise, including tourism officials, architects, historians, planners, and business representatives. The in-depth interviews were complemented with twelve thematic focus groups with the participation of residents of Bogotá aiming to understand how residents saw their own city and evaluated certain elements around four themes: culture, economy, tourism, and urban development. This captured the interactive nature of brand formation and the possible interactions between several types of association. The next phase involved the creation of a vision for Bogotá and the definition of strategic goals and measures. The development of strategy was not centrally determined and decided 'behind closed doors' as is commonly the case (see Bennett and Savani, 2003), but based on cooperation with local stakeholders and the belief that it is not possible to reduce a city to a single identity claim. In the participatory spirit of the project, the task of formulating a vision was given to the focus groups. The participants were asked to imagine the future of Bogotá and the results were adopted as a 'common dream'. Naturally, this included elements that were detached from the city's reality and to keep things within economic and practical realities the strategic goals were divided into realistic (those closer to the city's current trajectories-eg, Bogotá (1) One of the authors has been involved in the planning and implementation of the project. as a 'city of music') and ambitious (those requiring more radical measures-eg, Bogotá as a 'green city'). The focus groups have been instrumental for the participatory character of the project not only being the sole source of the vision but also being used to validate the different strategic goals and to recommend and prioritize measures.
We should not underestimate the challenges that come with a participatory approach and the Bogotá project faced several. The first question was: how broad does participation need to be in order to qualify as adequate? It is obviously impossible to listen to everyone who might be affected by a project. While the effort was to conduct as many interviews and focus groups as possible, the final numbers were judged satisfactory only with the project's time and financial limitations in mind. Perhaps the combination of qualitative and quantitative research methods or the use of technology-based platforms (eg, social media) might help capture more representative samples. The repetition of the analytical stage of place branding in cycles (eg, biannual) including different participants might also help in the long run. Secondly, there is a very significant question as to who can participate in such projects; whose voice gets heard and whose arguments favoured. This was clearly an issue in the Bogotá project as well. Mainly for reasons of access, participation in the thematic focus groups was achieved only for the educated middle class, thus neglecting a large section of the city's residents. While this is already better than making decisions 'behind closed doors', integration of more bottom-up processes might help. Perhaps a combination of different platforms, methods, and geographical areas [eg, the simultaneous use of focus groups, public consultations (on and offline), observations in social media, observations in deprived districts, videography] might be a solution. Thirdly, is participation to be limited to the strategy creation phase of place branding? As Bennett and Savani (2003) have shown, it is very commonly the case that local residents are consulted only on the outset of place branding projects; this cannot be considered broad participation. In Bogotá the effort was to let residents participate broadly in the initial phases and the recommendation for broad participation in later stages has been made emphatically.
Another challenge concerned two conflicts between the consultant and local stakeholders. First, while the consultant was of the opposite opinion, the Tourism Bureau insisted on the need for a city logo, which it considered should have been the main product of the place branding strategy. Secondly, the majority of participants argued for a new landmark for Bogotánamely an architectural icon-whereas the consultant suggested projects directly responding to the city's acute problems (eg, extensive poverty). Compromise was reached on both when the consultant decided to go along with the local population's desire and included the logo and the architectural icon in the list of proposed measures. Participation is not challenge-free, but the rewards are significant as it is "designed to generate and support a greater sense of ownership, which is a crucial ingredient in any branding initiative" (Houghton and Stevens, 2010, page 45).

Conclusion-a switch in place branding
This interactional view of place brands highlights their dependence on the interactions between actors and stakeholders who actually construct the place brand, give it meaning, and suggest its potential (as also seen in the case of Bogotá). It is important to accept that place branding is a highly selective political process. Cities are not homogeneous entities, and their inhabitants differ in many ways-age, class, gender, ethnicity, etc. Moreover, relations among groups may be highly antagonistic, sometimes in a more latent, but often in a manifest way. This deeply political understanding of places stands in clear contradiction with the need for a certain degree of unity. If branding is about creating 'positive associations' (Kavaratzis and Ashworth, 2005) with a place, it is obvious that these associations will differ from person to person or at least from one social group to another. Yet, there are inherent choices that are necessary, without which it is not possible to attempt branding. The criteria used for those choices rely on the responsibility of each organization with the task of developing a branding system for its place and can be undertaken through difficult practices that genuinely legitimize it, such as participatory forms of urban governance. The understanding of place brand formation suggested in this paper actually emphasizes the need for such participatory place branding practices as it is those that facilitate the interactions that have been the focus of the discussion. As Kalandides (2011b) suggests for place branding consultants, their role is that of a facilitator, somebody who can use sound methodological tools to extract and structure the knowledge and ideas that are already there and bring the place brand to the surface. This is a difficult but vital switch in the dominant view of place branding. Another significant consequence concerns the idea of the city as open-ended trajectories that interlock. This means that the city can only be understood as an ongoing and open-ended process. In turn, this means that place branding can only capture the city in its entirety only if it focuses on paying tribute to the unique, local particularity of the ways in which place-making elements combine and interact over time. What the experience of Bogotá shows is that it is not necessary to reduce the place's identity to an inevitably unfair, single identity claim (Ren and Blichfeldt, 2011) in order to be able to work with it in practical terms. It is possible instead to work with the many open-ended trajectories and 'celebrate' their diversity, allowing for their interaction through participation and openness. Place branding then can reinforce a place's uniqueness if it is based on the same thing that makes places unique: not particular types of elements but the unique local blend of such elements and their unique interactions.
There is, of course, a need for theoretical refinement and it is important to investigate three issues in particular. The first relates to the issue of agency in place-making and in place brand formation. Future research, perhaps using the analytical tools of actor network theory, will be needed to clarify this issue while also clarifying the relation to the wider process of place-making. A second issue relates to empirically investigating the manner in which associations interact as described in the framework. The third issue relates to the place function and scale at which all this might be applicable. Is place brand formation one and the same regardless whether it is a tourism destination or a place of residence and whether it is a country or a city? This paper has proposed what we see as a switch in our view of place branding towards better implementation and, ultimately, better use of taxpayers' money. As long as the interests of the community are overshadowed by other interests [place authorities, large businesses, and property owners (see Bennett and Savani, 2003)], place branding will remain a dubious way for politicians-with few exceptions-to justify their political existence and for a handful of consultants-again with a few outstanding exceptions-to make a living out of selling catchy slogans.