Why can’t we grasp gentrification? Or: Gentrification as a moving target

Abstract

Since the term “gentrification” was coined by Ruth Glass in 1964, this concept and the phenomenon it referred to have been subject to change. This paper reviews the literature and employs Ian Hacking’s work to investigate how two types of changes—that is, changes of the concept and of the phenomenon—are implicated by each other. By investigating the interaction between a classification and its class, it becomes possible to understand gentrification as, in Hacking’s terms, a “moving target.” This paper argues that gentrification can be conceptualized as such and explores the consequences of this for gentrification research.

1 Introduction

Gentrification research faces a paradox. This paradox is reminiscent of a paradox described by Plato in his Meno. It concerns the endeavor to discover the truth. Meno asks how one can even begin to investigate something if it is not clear what one is looking for. This leads Socrates to the formulation of the paradox:

The claim is that it’s impossible for a man to search either for what he knows, since he knows it and that makes the search unnecessary, and he can’t search for what he doesn’t know either, since he doesn’t even know what it is he’s going to search for. (Plato, 2005: 80e)

In order to start an inquiry, the object of it has to be known and unknown at the same time; and this, of course, is an absurdity (cf. Polanyi, 1966: 22).

We can discern a similar problem in gentrification research. The term gentrification was coined by Ruth Glass to refer to the influx of the gentry to the city (Glass, 2010 [1964]). Since then, however, much has changed in cities. Hence, Slater (2006) asks whether scientists should remain faithful to the original concept and notes:

it is worth reminding ourselves that we are over forty years beyond Ruth Glass’ coinage! So much has happened to city economies (especially labour and housing markets), cultures and landscapes since then that it makes no sense to focus on this narrow version of the process anymore, and to insist that gentrification must remain faithful to the fine empirical details of her geographically and historically contingent definition. (Slater, 2006: 744, emphasis in original)

If the phenomenon of gentrification has changed—and it has now been nearly 60 years—we should no longer hold on to the original concept; the concept, so it seems, no longer describes the phenomenon. However, if the phenomenon does not fit the definition of gentrification, then how can we identify it as such? That does not seem to be possible. But if we cannot identify it as gentrification, then how can we conclude that gentrification research should concern itself with this new phenomenon? Hence, gentrification research is also faced with an apparent paradox: the new phenomenon has to be known as gentrification (in order to become an object of gentrification research) and as not-gentrification (for it does not fit the definition of gentrification) at one and the same time.

This paper proposes a framework for understanding the relation between, on the one hand, the fact that gentrification is a changing phenomenon and, on the other hand, the ongoing debate on how gentrification should be defined. It contributes to these discussions by describing how two types of changes—that is, changes at the level of the concept and at the level of the phenomenon described by it—are implicated by each other. It will do this by using the ideas of what philosopher of science Ian Hacking calls “dynamic nominalism” (Hacking, 2007). Drawing upon this perspective, it will be argued that gentrification can be fruitfully conceptualized as a “moving target.” This means that there is an interaction between a concept and the phenomenon described by it and, consequently, that both the concept and the phenomenon are always on the move. This paper thus approaches the persistence of the debate around the correct definition of gentrification as a research puzzle that needs an explanation.

By using the insight that gentrification is one of the most politically loaded words in geography (Lees et al., 2008: 155), I will argue that the way in which scholars understand gentrification has an effect on the social world; gentrification research is not “outside” of gentrification, but becomes part of how the phenomenon is understood by social actors. This implies that processes of urban renewal can change by virtue of being identified as gentrification. Taking up such a perspective provides the tools to study why gentrification sometimes invokes more powerful reactions than other processes of neighborhood change (Brown-Saracino, 2016: 223). I will elaborate on all this in the remainder of this paper.

Beyond its particular goal, this paper tries to show how dynamic nominalism can be useful for geography. According to Hacking (2007), his ideas are applicable to many social phenomena. Gentrification, because of its contentious nature, may be a paradigmatic case of a moving target in geography; it is nonetheless to be expected that the interaction between concept and phenomenon can also be discovered with respect to other geographical topics.

This paper is organized as follows. First, it discusses several definitions of gentrification, followed by an examination of how we can think of gentrification as a changing process. I will then introduce the work of Ian Hacking and discuss dynamic nominalism. After this, I will argue that gentrification can be understood along these lines and consider what the implications of this new conceptualization are for gentrification research.

II The concept of gentrification

This section will be concerned with different definitions of gentrification. The goal of this section is not to give an exhaustive overview of the definitional issues surrounding gentrification, but rather to show how the concept has changed since it first appeared in print. In this section, it will also be considered why the (changing) definitions of gentrification matter.

The concept “gentrification” was coined by Ruth Glass in 1964, in order to describe the changes that were going on in the working-class quarters in London at that time. These quarters had “been invaded by the middle classes—upper and lower … Once this process of ‘gentrification’ starts in a district, it goes on rapidly until all or most of the original working-class occupiers are displaced, and the whole social character of the district is changed” (Glass, 2010 [1964]: 7). As the term “gentrification” describes the displacement of working-class occupiers as a result of the influx of the “gentry” to the city of London, the concept clearly bears the traces of the place and time in which it originated.

This origin, however, has not been an obstacle for the employment of the term in other contexts and at other times. As Beauregard (2010 [1986]) notes, several different (though related) phenomena have received the label of gentrification. In this respect, the marginal gentrifiers described by Rose (1984) are a case in point. Marginal gentrifiers are people with moderate incomes who buy old dwellings and renovate them before they start living there. They do not possess the wealth that is often associated with gentrifiers; instead, they use their “sweat equity” to make living in inner-city neighborhoods a viable option for themselves (Rose, 1984: 61). A more recent example of marginal gentrifiers is provided by the urban gentrifiers who—in search of an authentic urban lifestyle—want to live in cities, but lack the money to do so without sharing apartments and living ascetic lives (Zukin, 2008: 727). Not all gentrifiers are wealthy and thus, Rose argues, gentrification researchers should keep in mind that there are different “forms of ‘gentrification’ and some types of ‘gentrifiers’” (Rose, 1984: 62). Hence, Beauregard concludes, “‘gentrification’ must be recognized as a ‘chaotic concept’ connoting many diverse if interrelated events and processes; these have been aggregated under a single (ideological) label and have been assumed to require a single causal explanation” (2010 [1986]: 13).

The gentrification literature exhibits two strategies to deal with this ‘chaos'. The first strategy is to opt for a very narrow, precise definition of gentrification, so as to be able to distinguish gentrification from other, distinct processes of urban change. This option is chosen by Van Criekingen and Decroly (2003). They argue that one can only speak of gentrification when the process starts with a decayed and impoverished neighborhood; when this neighborhood undergoes transformations via improvements to the built environment, social status growth, and a change of the population; and when the outcome of these transformations is a wealthy neighborhood (Van Criekingen and Decroly, 2003: 2454). Hence, according to these authors, marginal gentrification does not fit this description (Van Criekingen and Decroly, 2003: 2454).

The second strategy is to use a broad and open definition of gentrification (Smith and Williams, 2010 [1986]; Clark, 2010 [2005]). Smith and Williams (2010 [1986]: 10) provide a complex definition which mentions the different aspects of gentrification. They argue that gentrification does not merely affect housing; instead, they contend that gentrification can only be fully understood if the development of recreational spaces, hotels, and inner-city offices are also considered. Clark’s (2010 [2005]) influential and often-used definition of gentrification is equally broad, but much simpler. He builds upon Beauregard (2010 [1986]) and Rose (1984), but emphasizes instead that the chaos of the concept is a problem to be solved. The concept’s different elements should be disentangled in order to hold on to the essential, and remove the accidental, elements of gentrification. Clark therefore argues that gentrification consists of the combination of investments in the built environment and the displacement of original land-users as a result of the influx of wealthier people:

Gentrification is a process involving a change in the population of land-users such that the new users are of a higher socio-economic status than the previous users, together with an associated change in the built environment through a reinvestment in fixed capital. The greater the difference in socio-economic status, the more noticeable the process, not least because the more powerful the new users are, the more marked will be concomitant change in the built environment. It does not matter where, and it does not matter when. Any process of change fitting this description is, to my understanding, gentrification. (Clark, 2010 [2005]: 25)

As it is only implicit in this description, it should be added that for Clark, gentrification is by definition the result of what Neil Smith described as the rent gap (see Clark, 1992: 360, 2010 [2005]: 27). The influx of land-users of a higher socio-economic status thus results from an earlier “disparity between the potential ground rent level and the actual ground rent capitalized under the present land use” (Smith, 1979: 545).

Despite the many attempts to define gentrification, there is no consensus on what should be meant by the concept (cf. Lees et al., 2010: 3–6). Nonetheless, the debate around the definition of gentrification remains important, because how gentrification is defined influences how its effects are measured empirically, as well as how it is evaluated morally (Marcuse, 1999). Given the different definitions of gentrification—and the different conceptions of displacement, which is often considered to be the most severe consequence of the process (see, for example, Atkinson, 2015; Baeten et al., 2017; Davidson, 2009; Elliott-Cooper et al., 2020; Hyra, 2015; Marcuse, 2010 [1986])—different studies have come to diverging conclusions about the magnitude of displacement caused by gentrification (e.g., Atkinson, 2000; Freeman and Braconi, 2004; Newman and Wyly, 2006). With the different findings on the effects of gentrification (see Brown-Saracino, 2017), it is no surprise that gentrification has been fiercely criticized by some (e.g., Slater, 2006; Wacquant, 2008), as well as supported by others (e.g., Byrne, 2003; Whyte, 2010 [1988]).

A clarifying remark is needed at this point. Given the different definitions in the literature, it might have become unclear what the term gentrification is taken to mean here. In the remainder of this paper, gentrification will be understood according to Clark’s (2010 [2005) broad definition (unless stated otherwise). This definition is used as a heuristic device, because this broad definition can be used to discuss the different phenomena that are subsumed under the term “gentrification.” I will reflect on the usefulness of this definition in the discussion. At that place, I will also discuss how Clark’s definition can be adapted in order to be able to capture recent manifestations of gentrification. In the next section, though, it will first be considered how gentrification as a phenomenon has been subject to change.

III Gentrification: A process in process

As Doucet notices, gentrification is “a process of change and a changing process” (Doucet, 2014: 125). And indeed, several studies have provided insight in the changing nature of the phenomenon of gentrification. A first branch of literature in this respect is provided by the writings on stage models of gentrification. One of the first of these models is offered by Clay (2010 [1979]). He argues that gentrification starts when pioneers take the risk to fix up their houses. This stage is followed by the influx of new gentrifiers and accompanied by promotional activities that highlight the perks of the developing neighborhood. While the early stages are not accompanied by much displacement, eventually the middle-class enters the neighborhood and, consequently, displacement occurs at a rather large scale. Clay envisions the possibility of this process to go on until an end stage of “completed gentrification” (2010 [1979]: 38) has been reached.

Berry (2010 [1985]) and Bourne (2010 [1993]) share Clay’s view that gentrification takes place in consecutive waves, but argue against descriptions that are overly general. Rather, they assert that one should consider the historical and geographical context in which gentrification unfolds. Berry argues that one important factor contributing to gentrification was given by “excessive scrappage of inner-city housing” (2010 [1985]: 50) in the 1970s. Bourne (2010 [1993]) argues that gentrification has resulted from a complex of historical factors, like demographic trends and economic booms and busts. The factors responsible for stimulating gentrification were on the verge of disappearing in the 1980s; hence, Bourne predicted that the importance of gentrification might be significantly reduced (2010 [1993]: 63).

With the benefit of hindsight, we can say that this latter prediction has not become true. Gentrification has not vanished. What has happened, however, is that it has changed its form. Hackworth and Smith (2000; see also Aalbers, 2019a, 2020) argue that the role of the state has grown more important in the course of the history of gentrification. This is, inter alia, due to the fact that the neighborhoods that were relatively easy to gentrify had been “fully reinvested” (Hackworth and Smith, 2000: 469). The consequence of earlier gentrification was that “gentrifiers and outside investors have begun to roam into economically risky neighborhoods … which are difficult for individual gentrifiers to make profitable without state assistance” (Hackworth and Smith, 2000: 469). Because of earlier waves of gentrification, the process of gentrification could thus only continue if it transfigured itself, so as to fit the different environment.

Another example of how gentrification has transfigured is given by the phenomenon of “super-gentrification” (Lees, 2003; Butler and Lees, 2006). This concept refers to “the transformation of already gentrified, prosperous, and solidly upper-middle-class neighborhoods into much more exclusive and expensive enclaves” (Lees, 2003: 2487). Super-gentrification also presents a challenge to the idea that neighborhoods can be completely gentrified (Lees, 2003: 2506). Even in already prosperous neighborhoods that appear to be completely gentrified, gentrification can occur again. However, it then appears in another form, for it has to build on the neighborhood that is already gentrified, though not yet super-gentrified. And of course, the super-gentrified neighborhood can, at least in theory, be gentrified again (super-super-gentrification), and so on and so forth (cf. Lees, 2003: 2506).

Earlier forms of gentrification can also have a more indirect effect on the conditions wherein new forms of gentrification take place. Wyly and Hammel (1999) argue that earlier processes of gentrification can affect new occurrences of gentrification via changes in housing policy. For, on the one hand, “three decades of gentrification have altered the context for certain facets of housing policy” (Wyly and Hammel, 1999: 715) and on the other hand, revisions of public policy and modifications in housing finance “have also transformed the environment in which gentrification, quite literally, takes place” (Wyly and Hammel, 1999: 721). The effect of earlier gentrification can thus be mediated by public policy and housing finance (cf. Hochstenbach, 2017a; Van Weesep, 1994).

The discussion of the several studies in this section shows how gentrification has undergone changes. It has changed because of earlier processes of gentrification that have altered the conditions in which new processes unfold. Gentrification, then, appears to be an elusive phenomenon. The remainder of this paper will investigate this elusiveness by considering the following questions: how does the changing phenomenon affect the concept? And: how does the concept affect the phenomenon? The next section discusses the ideas of philosopher of science Ian Hacking, who has studied the interaction between concepts of the social sciences and the phenomena they intend to describe.

IV The objects of the social sciences as moving targets

Gentrification’s elusive nature is not a characteristic that is unique to it; rather, it is a characteristic of many social phenomena. Philosopher Ian Hacking has drawn attention to the interaction between social scientific concepts and the phenomena they purport to describe (see, for example, Hacking, 1999, 2007). Hacking mentions Nietzsche and Foucault as his predecessors in the tradition he calls “dynamic nominalism” (Hacking, 2007: 294–295). This is a species of nominalism, for it argues that social phenomena are categorized together because they share a name; and it is dynamic, because it is concerned with “how names interact with the named” (Hacking, 2007: 294).

We can see dynamic nominalism in action in Foucault’s first volume of The History of Sexuality (1978). In this work, Foucault attempts, among other things, to analyze how homosexuality came into existence in 1870. At first sight, this may appear as a surprising, or even downright false, claim; it suffices to read Plato’s Symposium (2008) to ascertain that erotic relations between men were a common practice in ancient Greece. However, Foucault’s point is that homosexuality only emerged after a new categorization saw the light of day:

We must not forget that the psychological, psychiatric, medical category of homosexuality was constituted from the moment it was characterized … less by a type of sexual relations than by a certain quality of sexual sensibility, a certain way of inverting the masculine and the feminine in oneself. (Foucault, 1978: 43)

And he adds: “Westphal’s famous article of 1870 on ‘contrary sexual sensations’ can stand as its date of birth … the homosexual was now a species” (Foucault, 1978: 43).

The invention of a category can have an effect on how people are understood, both by others and by themselves. On the one hand, with the medical category of homosexuality, it became possible to be diagnosed as such by physicians and to be understood as a patient in need of treatment. On the other hand, it became possible to identify oneself as a homosexual human being and to be proud of that.

Before the concept of homosexuality existed, it was not yet possible to understand oneself by means of this concept as a homosexual being; a homosexual person was not yet a distinct kind of person and one could thus also not yet be of this kind. This became a possibility only when there was a category available which provided the necessary description. Therefore, this is a case in which, according to Hacking, “our classifications and our classes conspire to emerge hand in hand” (2006a: 106). The self-identification as a homosexual person was not possible before the category of homosexuality existed in the space of possible self-understandings. Drawing upon Anscombe (2000), who argued that intentional actions are actions under a description, Hacking concludes that “if a description is not there, then intentional actions under that description cannot be there either: that, apparently, is a fact of logic” (Hacking, 2006a: 108). A peculiar characteristic, then, of “human action is that by and large what I am deliberately doing depends on the possibilities of description … Hence, if new modes of description come into being, new possibilities for action come into being in consequence” (Hacking, 2006a: 108). Foucault’s statement about the date of birth of the homosexual should thus be taken quite literally: the birth of the homosexual person coincided with the invention of the concept of homosexuality; this concept made it possible to be a new kind of person. As Hacking explains: “our spheres of possibility, and hence ourselves, are to some extent made up by our naming and what that entails” (Hacking, 2006a: 113).

The invention of this category made new expert knowledge and lay self-understandings possible. These effects, in turn, “flow back” in the meaning of the concept; this is the dynamic that results from the invention of concepts. Indeed, while originally invented as a medical category, homosexuality had become a concept around which people could build organizations, construct places to meet and coordinate political struggles to demand the same rights as heterosexual couples (Hacking, 1995: 38). As such, the meaning of the concept changed. The concept has made new social practices possible; thereby, it has changed the object to which the concept originally referred; and in turn, the concept had to be adapted in order to be able to describe its new referent. This interaction between concept and referent can continue to the point that it is no longer sufficient to change the meaning of the concept; instead, a new concept has to be developed, for the discrepancy between the original concept and that to which it intends to refer, has grown too large.

Because of the interaction between concept and phenomenon, Hacking (2007) describes the phenomena of the social world as ‘moving targets'. There is a “looping effect” (Hacking, 2007: 286) between the concept and its referent, which ensures that the described object transforms by virtue of being described in a certain way. This is the result of, on the one hand, the self-understandings of people, who may understand and experience themselves in a different manner as a result of the new description at hand. The new actions and behavior that result from a new self-understanding of people can in turn “loop back” and “force changes in the classifications and knowledge about them” (Hacking, 1999: 105). On the other hand, the interaction between concept and referent result from new institutions and practices that are made possible by a new category (Hacking, 1999: 105). In this case, a fully articulated new self-understanding which is formed directly by the new concept is not necessary; instead, new ways in which people experience themselves and their place in the world are embedded in an institutional apparatus and practice made possible by the new categorization. This latter point suggests that scientific categories alone are not sufficient to interfere with people’s self-understandings and experiences. As long as the scientific categories strictly remain within the realm of scientific discourse, they will not influence social practices (Hacking, 2004). Only when these concepts are taken up by the actors involved (e.g., practitioners, the categorized people, policy makers, etc.), they can affect social practices (Hacking, 2006b: 19).

The active role of all the different actors highlights how dynamic nominalism builds upon a Foucauldian conception of power (Hacking, 2006b). For Foucault, the exercise of power is “a way of acting upon an acting subject or acting subjects by virtue of their acting or being capable of action” (Foucault, 1982: 789). Using of a new concept is one way in which power can be exercised, since new concepts allow for new ways in which people can “constitute [themselves] as subjects acting on others” (Foucault, 1983: 237; cf. Hacking, 2006b: 19). New self-understandings give people new possibilities for action that would not have been available to them if the new concept had not been at their disposal (Hacking, 2006a). Such a conception of power does not imply that everyone has the same means to change a concept and alter the corresponding phenomenon, but it is to say that no passivity should be assumed on the part of any of the actors involved. Those who do the describing, but also those who are described by the new concept, have new possibilities for action as a result of the new concept.

Hacking’s ideas are influential in the philosophy of science, but they have been criticized as well. Two criticisms are of particular relevance for this paper. ¹ The first is concerned with the relation between language and intentional action. According to Hacking, intentional action is action under a description; hence, if the description of an action does not exist, one could not intentionally execute this action. Rachel Cooper (2004) questions this argument. She evokes the imaginary case of a caveman who—before language developed—made a fire. As the caveman did not have language and thus no descriptions at hand, Hacking’s argument would imply that the caveman could not have intentionally made the fire. However, as Cooper (2004: 82) contends, this is false: even though the caveman cannot express in words that he wanted to light a fire, we can still maintain that he did so intentionally. We can infer his intentions, for example, from the fact that it is cold outside or from the smile on the caveman’s face when the fire gets started. Some intentional actions are thus also possible without a description. The emergence of a new description would then not necessarily affect the phenomenon. In the case of the caveman without language, having a concept of fire making at his disposal would presumably not significantly influence his actions; he would continue making fires in the same way as he did before he knew the concept. Therefore, the coming into being of a concept does not necessarily change the phenomenon it describes; a concept is not necessarily constitutive of the phenomenon. It thus has to be shown in individual cases that the description actually affects the described phenomenon and that this change in the phenomenon occurred because of the existence of the concept (R. Cooper, 2004: 84).

The second criticism is concerned with Hacking’s argument that feedback loops mean that objects of the social sciences should be understood as moving targets. As Tsou (2007: 340) points out, some classifications may only slightly change the behavior of individuals, such that the criteria of this classification do not have to be revised as a result of this change. In those cases, it does not seem correct to speak of moving targets, since the target remains where it was. This is all the more important, because the interaction between classification and class can also proceed via feedback mechanisms that are stabilizing, rather than destabilizing (Kuorikoski and Pöyhönen, 2012: 195ff.; Laimann, 2020: 1054–1056). Laimann (2020: 1054–1056) mentions the example of gender. Due to gender categories and the social expectations that come with them, individuals can “come to fit their classifications” (Laimann, 2020: 1055): they comply to gender norms. This norm-confirming behavior, in turn, may be taken as a cue that the classifications can be specified more precisely, which can lead to even more stringent and narrow norms of what proper masculine and feminine comportment entails. Concepts can thus also bring about stability, rather than instability, of objects of the social sciences. In those cases they are not moving, but rather stationary, targets.

As we have seen in this section, Hacking has emphasized the interaction between a classification and its class. This interaction can make that objects of the social sciences change. He argues that these objects should therefore be understood as moving targets. However, as others have pointed out, this is not necessarily the case. Sometimes the concept does not, or only hardly, affect the phenomenon, such that a modification of the concept is unnecessary. In other cases, a description can stabilize, rather than destabilize, the object of description. In order to say that a phenomenon is a moving target in Hacking’s sense of the term, there are thus three conditions that need to be fulfilled. A concept needs to:

1. Alter the phenomenon and not leave it unchanged;

2. Affect the phenomenon in a destabilizing way; and

3. Change the phenomenon to such an extent that this change, in turn, necessitates a change of the concept.

We will now return our attention to gentrification, in order to see whether this process can be usefully conceptualized as a “moving target.”

V Gentrification as a moving target

This section will discuss the interaction of gentrification as a concept and as a phenomenon. Section 5.1 considers how the concept affects the phenomenon. In section 5.2 the case of social mix will be discussed. Here, it will be argued that the concept of gentrification has changed the phenomenon to such an extent that the original concept of Glass (2010 [1964]) had to be adapted in order to be applicable to the changed phenomenon. All in all, section 5 will thus argue that the three above-mentioned conditions are met in the case of gentrification and that it should hence be understood as a moving target.

1 How the concept of gentrification affects the phenomenon

Several studies show that the concept of gentrification can indeed affect how the process of gentrification unfolds—at least at some moments and at some places. In section 2, it was already argued that the moral evaluation of gentrification is dependent on how it is defined. This is not only an academic, but a sociopolitical issue as well: the concept of gentrification has also become part of the language of policy makers, tenants, activists, real estate developers, etc. Some authors argue that the concept bears the traces of the neoliberal reality of the Anglosaxon world in which the concept originated (e.g., Butler, 2007; Maloutas, 2018; Smith, 1996). One can say of the concept of gentrification what Berger and Luckmann say of language in general: “[it] is capable of becoming the objective repository of vast accumulations of meaning and experience, which it can then preserve in time and transmit to following generations” (Berger and Luckmann, 1966: 52). With the multiple layers of meaning it incorporates, what does the concept of gentrification do in practice?

One particular effect of the concept of gentrification on social practices is given by Smith’s (1996: 28–32) description of gentrification as a “dirty word.” Gentrification, he argues, had become a word that was widely used by activists, tenants, and others to make sense of what was going on in inner city neighborhoods. The word described how the changes in the built environment negatively affected their daily lives. The concept of gentrification had found its way from academia to the self-understandings of city inhabitants, and “this new word, gentrification, captured precisely the class dimensions of the transformations that were under way in the social geography of many central and inner cities” (Smith, 1996: 30). Hence, the image of gentrification deserved some polishing—at least, that was what the Real Estate Board of New York, Inc. thought in 1985. This group published an advert in the New York Times, stating that it was gentrification that would bring about a bright future: “We also believe that New York’s best hope lies with families, businesses, and lending institutions willing to commit themselves for the long haul to neighborhoods that need them. That’s gentrification” (cited by Smith, 1996: 28–29). Spending a large sum of money on an advert defending gentrification in fact proved the opposite of what the Real Estate Board aimed for: it showed that gentrification was indeed a dirty word that bred resistance.

Gentrification is, at least sometimes, still considered to be a dirty word—although it is sometimes also framed in more supportive ways in newspapers (Brown-Saracino and Rumpf, 2011; Tolfo and Doucet, 2021). Whether or not a process of urban change is called “gentrification” and how this process is framed in turn, influences if and to what extent the process invites opposition—at least in the Western world, where the concept of gentrification occupies a prominent place in urban discourse (cf. Lees, 2000). Only this prominence can explain the surprise of Ley and Teo (2014) when they find that (a translation of) the term “gentrification” is not used in the newspaper coverage of urban renewal in Hong Kong. Nonetheless, what was happening in Hong Kong could be aptly described as “gentrification”—and Ley and Teo predict that the term will be used more often, once a “dissemination of class analysis” has taken place there (2014: 1301). When the term gentrification becomes part of the urban vocabulary, it is to be expected that it will affect how the inhabitants of Hong Kong experience what is going on in their city.

Two studies show that the term gentrification has come to occupy a more prominent place in public discourse in the Global North in recent years. Hochstenbach (2017b) analyses Dutch newspapers between 2000 and 2017 and finds that the term gentrification (or its Dutch equivalent gentrificatie) had rarely been used between 2000 and 2010, but that its use has skyrocketed from 2015 onwards. The findings of Tolfo and Doucet (2021) show that a similar process happened in Canada: in the Canadian newspaper The Globe and Mail, the term gentrification was used with an increasing frequency between 1980 and 2017. These studies show that there is a growing tendency among newspapers to use the term gentrification to describe processes of urban change in the Global North. As such, the concept may, via looping effects, have a bigger impact than it had in earlier years on how the process unfolds in the Global North.

A recent debate revolves around the question whether the concept of gentrification should also be applied to other places than those in the Global North. Some scholars argue that capital easily crosses national borders and can thus take advantage of rent gaps everywhere. This process, they argue, is therefore best captured by the concept of gentrification—even though we should account for the different contexts in which the process unfolds (López-Morales, 2015; Slater, 2017; Smith, 2006). Moreover, urban policy that brings about gentrification is diffused from Western cities to other parts of the world; hence, it makes sense to use the same label to describe the outcome of such policies (Lees, 2012; Lees et al., 2016: ch. 5; Smith, 2002). A concept related to gentrification can also be employed, though: in Mexico, for example, the expression blanqueamiento (“whitening”) is used by activists (López-Morales, 2015: 567). This term captures three linked phenomena, namely, “the arrival of affluent white middle classes both culturally and physically displacing low-income mestizo communities, the actual whitening of buildings imposed by the new ‘middle-class’ aesthetic taste and the money laundering which real estate investment enables” (López-Morales, 2015: 567). As such, activists use and adapt gentrification theory for the purpose of understanding the transformation of space in other contexts than the Anglo-American world in which the concept finds its origins (López-Morales, 2015: 567). Gentrification, as a concept, may thus “be translated into an expression that is more useful for local populations, while retaining the core principle of gentrification in the translated version” (Shin and López-Morales, 2017: 16). With the concept gentrification (or its translation) at hand, activists in the Global South can not only understand how capitalist forces transform their neighborhoods, but the concept also helps them to politicize these processes, so that the course of these developments can be shifted (López-Morales, 2015: 567). The concept of gentrification could thus be used by activists to ensure that, through a looping effect, the process unfolds in a way that does not conform to its original description. In this way, the concept can affect the phenomenon. At the same time, the term blanqueamiento shows that the concept of gentrification was not simply adopted, but also adapted, by Mexican activists to make it useful for the local context.

Other scholars have argued instead that the concept of gentrification—or a translation retaining its core principle—is not well-suited to understand urban transformations in the Global South, where large portions of land are not commodified. Rather, as they argue, the term gentrification distorts our understanding of processes of urban change in the Global South (Ghertner, 2015; Maloutas, 2018; Smart and Smart, 2017). Moreover, in using concepts originating in the Global North to understand processes in other places, there is a risk of reproducing dominant theories of how the world works, at the cost of alternative idioms (Smart and Smart, 2017). Concepts make up people, but they may make them up for the wrong situation, thereby making them ineffective in political struggles. This is why the concept of gentrification—while useful for the description of processes in capitalist societies—should not be used in a context of non-commodified land, according to Ghertner: “The limitation of gentrification, as an analytic, is that it fails to grasp transformations in the peri-urban and outer areas of post-socialist and post-colonial cities where the most violent displacement is taking place and where non-fully privatized tenure endures” (2015: 559–560).

Whether or not it is considered useful to frame a process of urban change as gentrification is thus also a matter of political strategy: a looping effect, whereby the concept affects the phenomenon, is sometimes deliberately sought. Reframing an urban development plan as gentrification can radicalize ongoing debates on processes of urban change and thereby serve politically strategic purposes (Lees and Ferreri, 2016: 22; cf. Maeckelbergh, 2012: 670). However, the outcome of such a strategy is sometimes hard to predict. Werth and Marienthal (2016) argue that gentrification is a polyvalent “grid of meaning” and that the specific interpretation of the concept by protesters influences which groups are, and which groups are not, deemed worthy of protection during struggles against gentrification (see also Lee, 2020). It therefore does not merely matter if something is understood as gentrification, but also how gentrification is then understood. If the harshest consequences of gentrification are downplayed in public discourse, resistance to gentrification may be severely constricted (Ellis-Young, 2020).

When the term gentrification is available in one’s culture, it can be used as a symbol around which people can organize themselves (cf. Joas, 2002; Pull and Richard, 2021). Brown-Saracino (2021) reports how the loss of dyke bars in four American cities was considered to be the result of gentrification and how this explanation, in turn, was used to forge a collective which otherwise, in a situation of “post-identity politics,” may not have come about. As she writes: “activists strategically deploy ‘gentrification’ as a symbol to generate a sense of shared vulnerability” (2021: 1029). The multiple layers of meaning attached to the term gentrification make it especially useful as “a flexible symbol for calling out a sense of shared fate” (Brown-Saracino, 2021: 1053). Gentrification provided the lens through which people understood the loss of the bars they frequented in the past and, consequently, how they evaluated urban change.

Experiences of gentrification are also mediated by knowledge of the process, for example, that it may lead to displacement and can alter the ambiance of originally working-class neighborhoods. As a result of this knowledge, the process can be experienced as a problem, not only by those at risk of being displaced, but by gentrifiers as well. Brown-Saracino (2007) describes gentrifiers who act as social preservationist, trying to preserve the authentic nature of the neighborhood they have come to inhabit. They view their own presence in the neighborhood as a danger to the community and hence believe that it is their duty to fight against possible displacement. Moreover, since they are not part of the community they appreciate so much, they believe they should not interfere too much with it; their interference would only contaminate it. The result is that the preservationists keep themselves at a distance from the people they have deemed authentic, in order to protect this community. The self-understandings of social preservationists, that is, that they might endanger a community worthy of protection, thus explains that the preservationists are committed to a position of “virtuous marginality” with respect to the authentic community (Brown-Saracino, 2007: 460). How the process of gentrification unfolds and what its effects are (e.g., the levels of displacement and of social interaction between different groups) is thus influenced by how the gentrifiers understand themselves qua gentrifiers and by how they understand their role in the neighborhood (cf. Bridge, 2001). The self-understandings of social preservationists are constituted by their knowledge of gentrification and, in turn, they actively tried to act not in accordance with the standard image of the gentrifier.

As was argued in this section, the concept of gentrification has made new self-understandings and social practices possible and these, in turn, have affected how the process unfolds. Moreover, the effect of the concept on the phenomenon has to a large extent been destabilizing: activists and social preservationists deliberately tried to ensure that descriptions of gentrification (as, for example, leading to large scale displacement and a loss of authenticity of the neighborhood) became not true (cf. Clark, 2014, 2015). The first two conditions that have to be met in order to speak of a moving target are thus fulfilled in the case of gentrification.

With the predominantly negative connotations of the word “gentrification,” policy makers are well advised to avoid this term (Elliott-Cooper et al., 2020: 494). A proposal to “stimulate gentrification” seems to be a recipe for failure, for it will very likely provoke resistance. Policy makers have better chances of success if they employ a different term. An ostensibly more benign concept is that of “social mix,” which will be discussed in the next section. There it will be argued that the definition of gentrification had to be changed in order to include social mix policies.

2 The case of social mix

Given that gentrification had become a dirty word, policy makers have used the concept of social mix instead. Policy aimed at stimulating social mixing avoids associations with displacement of the lower socio-economic class. It suggests instead that the original inhabitants and the newcomers live harmoniously together. The actual results of social mix policy are, however, often the same as the results of gentrification: it can lead to the displacement of the lower income class and be detrimental to informal support networks in neighborhoods (Lees, 2008). Scholars have therefore criticized policy aimed at increasing diversity and interpreted it as gentrification “rhetorically and discursively disguised as social mixing” (Bridge et al., 2012: 1). In other words: social mix is gentrification, but just harder to detect; it is “gentrification by stealth” (Bridge et al., 2012).

Conceptualizing gentrification as a moving target, however, suggests a different interpretation of social mix policy. Social mix should not be seen as merely a rhetorical device to disguise gentrification; rather, this discourse affects what actually happens in gentrified neighborhoods. In her description of the process, Glass predicted that gentrification “goes on rapidly until all or most of the original working class occupiers are displaced, and the whole social character of the district is changed” (Glass, 2010 [1964]: 7). If a policy aims at a social mix in terms of income of inhabitants, such an outcome would not be feasible: if (almost) the whole working class would be displaced, there would simply no longer be any mix of different income groups in that neighborhood. A justification in terms of social mix would then no longer be feasible. ²

It is true that justifications can be employed pragmatically by social actors, but they cannot be used as one pleases. A justification is not convincing if it does not hold up when it is “confronted with the real world” (Boltanski and Chiapello, 2005: 167). The same holds for social mix policy: if it would not live up to its promise, it would lose all credibility and could no longer be used as a justification for urban policy. A justification of urban renewal in terms of social mix would not be credible if it would come with an (almost) complete displacement of the working class. For the same reasons, Tissot (2011, 2014) argues that gentrifiers’ taste for diversity should not be reduced to hypocrisy; even though gentrifiers try to control diversity by defining what counts as acceptable behavior in the neighborhood, this does not mean that they aim for homogeneity. Rather, their taste for diversity structures the space of their possible actions. In the same vein, we cannot understand social mix policy as gentrification—at least, when we use Glass’ original definition of gentrification. Social mix can be interpreted as gentrification, however, when gentrification is understood in a broader fashion, as Clark’s (2010 [2005]) definition intended.

We can thus describe the evolution from early gentrification to social mix policy as the history of a moving target. Since its first description, gentrification was associated with high levels of displacement of the working class. As a result, gentrification became a dirty word. This, in turn, necessitated policy makers to develop a policy that addressed critique on gentrification; social mix policy could do this job. Here, we see the interaction, theorized by Hacking (2007), between the concept and the phenomenon. When original inhabitants understood developments in their neighborhoods as gentrification, this bred resistance. As a result, the phenomenon changed; resistance to gentrification was something to be reckoned with by policy makers and project developers. Hence, the name affected the named. In response, a new policy was developed: that of social mix. With a policy aiming at social mix, a complete disappearance of the working class was no longer feasible; at least, not without the justifications for that policy losing all credibility. This suggests that we cannot understand social mix policy as an ideological masquerade to enhance gentrification by stealth; social mix policy does not (merely) distort our view of what is “really” going on in the real world, but rather influences what kind of urban policy is feasible. If social mix is still to be understood as gentrification, however, the definition of gentrification had to mutate in turn and become something like the one Clark (2010 [2005]) proposed. Here, we thus see how the phenomenon affects the concept. Social mix could only be included in the class described by the classification “gentrification” at the moment that this classification was adapted. The changed phenomenon thus also affected the concept; thereby, the third condition that has to be satisfied to speak of a moving target is also fulfilled.

From its earliest forms to its latest, both the concept and the phenomenon of gentrification have changed significantly. This is partially explained by the interaction between the concept and the phenomenon. As such, gentrification can be fruitfully conceptualized as a moving target. The next section explores what this conceptualization of gentrification implies for gentrification research and more recent discussions in the gentrification literature.

VI Discussion

As was argued above, the definition of gentrification has changed from describing a rather specific phenomenon in England (Glass, 2010 [1964]) to describing a broad range of phenomena (Clark, 2010 [2005]). The downside of such a broad definition of gentrification is that it leads to a collection of quite different phenomena, such that including all these phenomena in the class of gentrification contaminates that class, instead of bringing about a greater understanding (Maloutas, 2018; Marcuse, 1999). The benefit of this broad definition is that it makes it possible to connect different phenomena to each other that would otherwise be studied in isolation. Clark’s broad definition can also capture phenomena like social mix policy, which were not captured by Glass’ original definition. It thus allows one to say, for example, that social mix policy is “gentrification by stealth” (Bridge et al., 2012). The broadness of the definition is probably one of the reasons why the concept of gentrification is taken up by so many people and is, in this respect, so successful. But is the concept of gentrification still changing? Is gentrification still on the move?

To include all phenomena that are now labeled as gentrification, Clark’s broad (2010 [2005]) definition may still not be broad enough. This becomes clear when we look at the financialization of housing, which has brought about what Aalbers (2019a) calls a fifth wave of gentrification. However, it is difficult to call this gentrification when we use Clark’s definition, since the element of “a reinvestment in fixed capital” (2010 [2005]: 25) is often absent from it. When more capital flows to real estate, this can drive up house prices and rents—even without investments that enhance the quality of the property. Wachsmuth and Weisler (2018) argue that this is aggravated by technological innovations like Airbnb, which create an alternative and potentially more profitable use of real estate, also without significant investments in the quality of the property (see also Cocola-Gant and Gago, 2021; Grisdale, 2021; Paccoud, 2017). This gentrification through financialization (Aalbers, 2019a, 2019b) has become possible on a larger scale as a result of austerity urbanism (Peck, 2012). Through government budget cuts and neoliberal reforms, leading to welfare cuts and the dismantling of public housing, numerous dwellings have become available on the market and have been turned into objects from which profits are reaped (August and Walks, 2018; Boelhouwer and Priemus, 2014). This has contributed to higher housing prices and rents and an overall lack of affordability (Aalbers, 2019a; Forrest and Hirayama, 2015). As a result, an augmenting number of people are left vulnerable to displacement (Annunziata and Lees, 2016; V. Cooper and Paton, 2021; Mösgen et al., 2019; Walks and Soederberg, 2021).

If we want to understand these developments as gentrification, it is necessary to adapt Clark’s (2010 [2005]) definition, while retaining what is arguably its core. A “change in the built environment through a reinvestment in fixed capital” (Clark, 2010 [2005]: 25) would then no longer be a necessary component of the definition of gentrification. Gentrification should then rather be understood as the process involving displacement (in all its forms) of land-users as a result of market developments that create a rent gap, such that the new land-users are of a higher socio-economic status. Displacement should be understood in the broadest possible fashion, ranging from last-resident displacement and exclusionary displacement (Marcuse, 2010 [1986]), via cultural and political displacement (Hyra, 2015) to the experiential process of un-homing (Davidson, 2009; Elliott-Cooper et al., 2020). Market developments, in this definition, refer to changes in the market (e.g., higher rents or real estate prices), changes of the market (a new organization of the market, resulting from, for example, new policies and regulations or technological innovations such as Airbnb) and/or the introduction of market mechanisms in formerly non-commodified spaces. Such a definition of gentrification seems to be able to capture the broad range of phenomena that currently go by this name.

I want to draw attention to three aspects of this definition of gentrification. First of all, this definition highlights that gentrification is no longer always clearly visible in the built environment. If a rent gap can be created without investments in fixed capital, such that this latter aspect of earlier gentrification should no longer be understood as an essential component of gentrification as such, this implies that gentrification manifests itself differently in the built environment than it did before. With gentrification through financialization, it is no longer necessary that a real estate owner invests significant amounts of money in the quality of his property to make it profitable. Hence, one would expect that gentrification would no longer unfold in the manner depicted by Smith (1982): as the result of capital reinvestments in a place which follow after a period of disinvestment at the same location. In Smith’s depiction of the process, gentrification should be clearly visible in the enhanced quality of buildings; in gentrification in its new form, this is no longer necessarily the case. Through financialization of housing, new land-users can pay a lot more than former land-users, without the quality of the dwelling being enhanced. Gentrification through financialization has resulted in an increasing number of people who are faced with the problem of precarious housing—no longer only the working class, but increasingly also segments of the middle class. Ironically, though, as socioeconomic inequality has increased, gentrification—as the most visible spatial expression of this inequality—may no longer always be as clearly visible in the built environment as it was during earlier waves.

Second of all, by including struggles over space due to an introduction of market mechanisms in formerly non-commodified spaces, this adapted definition is (partially) the result of studies of gentrification in formerly non-capitalist contexts in the Global South. As some scholars have pointed out, this can help to avoid hegemonic theorization: gentrification theory would then not only be applied to, but also informed by and adapted as a result of, research in the Global South (Bernt, 2016; González, 2016; Lee, 2020; Waley, 2016). Moreover, this broadened definition would make it possible for a broader coalition of precariously housed people in different parts of the world to organize themselves around the notion of gentrification (cf. Annunziata and Lees, 2016). Through the above discussed looping effects this can, in turn, inform resistance to gentrification and affect how the process will ultimately unfold in different places.

Third of all, this proposal to adapt Clark’s (2010 [2005]) definition intends to shows that gentrification is still on the move, but this does not mean that anything goes: if gentrification no longer includes displacement and the rent gap in its definition, it seems to have lost its distinctive usefulness as a concept. Conceptualizing gentrification as a moving target thus does not preclude it from having a stable core (cf. Clark, 2010 [2005]). At the same time, though, it should be emphasized that this adapted definition must not be taken as the ultimate one. Current discussions show that the definition of gentrification is also being questioned by scholars who investigate the relation between gentrification and racial capitalism. These researchers argue that gentrification should be defined more precisely. They have emphasized that gentrification is always a racialized process and that this should be captured in its definition (Danewid, 2020; Ramírez, 2020; Rucks-Ahidiana, 2021). These scholars thus suggest that the definition of gentrification should be adapted by the addition of an essential element. Roy (2017) argues, though, that the concept of racial banishment, rather than gentrification, should be used to study evictions under racial capitalism, since “the public means of evictions as well as forms of racialized violence, such as slavery, Jim Crow, incarceration, colonialism, and apartheid … cannot be encapsulated within sanitized notions of gentrification and displacement” (Roy, 2017: A3). However, recent research that depicts gentrification as a violent process (Baeten et al., 2017; Elliott-Cooper et al., 2020; Kern, 2016), as well as Moulden’s (2021) proposal to view gentrification as a crime, suggest that gentrification is maybe not such a sanitized notion after all. In short, we can say that gentrification keeps on moving, because it keeps evoking new ideas that can, via looping effects, influence how the process unfolds.

Because the notion of gentrification keeps arousing new ideas, it is, according to Clark (2015), a useful concept and an “adventure of ideas.” Viewed through the lens of dynamic nominalism, the latter phrase suggests that gentrification is at the same time an adventure of practices. The new ideas and questions that the notion of gentrification evokes, can also give rise to new social practices (of resistance, of policy making, etc.) that are influenced by these ideas; this happens through looping effects and through the making up of people. This seems an apt way to describe the promise inherent in the concept of gentrification: by its effect on social practices, the concept of gentrification may “contribute to making rent gap theory not true” (Clark, 2014: 394, emphasis in original). It can transform the way in which land-use is distributed and thereby, in an ultimate feedback loop, help in “making the notion of gentrification irrelevant” (Clark, 2015: 455).

With the different actors (scholars, activists, policy makers, journalists, etc.) that make use of the concept of gentrification, we can also discern different registers in which it is used. The concept can be used, for example, to describe changes in a neighborhood; the concept can be used to analyze the causes and effects of these changes; or the term gentrification can be used to politicize and problematize these changes. The term gentrification serves all these purposes. Indeed, one of the reasons why gentrification changes, both as a concept and as a phenomenon, is that the concept switches between these different registers. It started out as a scholarly concept to describe changes in inner-city London; it then became a “dirty word” in public discourse and bred resistance; policy makers avoided the term and suggested social mixing as a more benign concept and process; and scholars, in turn, broadened the concept of gentrification to subsume social mixing under it, thereby broadening the class of phenomena that together constitute the class of gentrification. This is not to say that the histories in these different registers are coterminous with each other; for example, activists may not accept all the definitions that scholars propose and vice versa. It shows, though, that the history in each of these registers cannot be understood in isolation from the histories in other registers and that the multiple ways in which different actors use the term should be considered.

If the concept of gentrification can affect processes of urban change by politicizing and problematizing them, some caution is warranted when the concept is applied in retrospect. As Massey (1995) argues, places have multiple pasts and these pasts are forged in the present. One way in which such a past can be forged, is through the use of new concepts. A new concept, like gentrification, makes a new interpretation of the past possible: it allows past events to be understood in a manner that was not possible during these events. As Hacking writes: “redescriptions may be perfectly true of the past; that is, they are truths that we now assert about the past. And yet, paradoxically, they may not have been true in the past” (Hacking, 1995: 249). Indeed, it is possible to say that a process of urban change that happened in the past can be described as gentrification—also when this concept was not yet available to the people involved. However, these people could not have understood themselves via this concept—it was, for example, not yet possible for new inhabitants to understand themselves as gentrifiers, because a description of this kind of person did not yet exist. This also means that people may interpret their own past actions differently after they come to perceive their actions as, for example, bringing about gentrification; what may in the past have been experienced as a completely innocent decision about where to live, may in hindsight be interpreted as an action that instigated gentrification. This may alter people’s sense of what they have done in the past and it can alter how they interpret their own life history (cf. Hacking, 1999: 162). Before knowledge about gentrification was widespread, gentrification was not yet a category that could make up a pressing moral problem—as it does now, for example, for Brown-Saracino’s (2007) social preservationists who actively try to avoid that their presence disturbs “authentic” communities. This new way of problematizing and politicizing processes of urban change, made possible by the availability of the concept of gentrification, may go some way in explaining differences between how gentrification unfolds in contexts where the concept is, and contexts where the concept is not, available.

If how the process of gentrification unfolds is affected by it being categorized in a certain way, it is important to study how the category is understood, both by experts and lay people (cf. Atkinson, 2003; Brown-Saracino, 2016). Hacking (2006a: 111), in this respect, makes a distinction between experts who categorize from above and categorized people who respond to this categorization from below. By their categorization, experts create a new reality for the people affected by it. In the case of gentrification, this could lead to different outcomes. If policy makers define a neighborhood as disadvantaged and in need of improvement, inhabitants may embrace this policy and welcome changes that increase available amenities (Doucet and Koenders, 2018; Uitermark et al., 2007). However, inhabitants can also interpret a need for improvement of their neighborhood as a territorial stigma, which can invoke resistance to urban policy that targets this neighborhood (Wacquant et al., 2014; Pinkster et al., 2020). This, in turn, confronts policy makers with a new reality. Do inhabitants, for example, understand this policy as tacitly provoking their displacement or not (cf. Kern, 2016; Lagendijk et al., 2014)? And if policy makers think that interventions in the neighborhood are necessary, do inhabitants feel that their contributions to the neighborhood are misrecognized (cf. Honneth, 1995)? The interpretations and experiences of what is going on in their neighborhoods and whether or not it is gentrification, affects the ways in which the policy can be implemented and thus what its ultimate results will be.

This ambivalence also holds for social mix policy (Shaw and Hagemans, 2015; Valli, 2015). Long-time residents of originally working-class neighborhoods can embrace a social mix discourse and welcome a more diverse neighborhood (Van Eck et al., 2020). When social mix is interpreted as “gentrification by stealth” (Bridge et al., 2012), however, social mix may in time become just as dirty a word as gentrification already is. It should therefore be determined via empirical research what the precise effects are of targeting a neighborhood for social mix policy, but it is to be expected that there will be an interaction between the category and the categorized. The same holds for adjacent concepts like “urban renewal” and “urban renaissance”: these concepts may also interact with the phenomena they describe, which should then be understood as moving targets as well.

VII Conclusion

This paper argued that gentrification can be fruitfully conceptualized as, in Ian Hacking’s (2007) terms, a moving target. Hacking, working in the tradition of dynamic nominalism, provides the tools to theorize how a concept and its object reciprocally influence each other. In order to argue that this is also applicable to gentrification, this paper has discussed different branches of the gentrification literature. It has discussed how the concept of gentrification has been modified; how the phenomenon of gentrification has been subject to change; how the concept affects social practices, so that the phenomenon changes; and how these changes of the phenomenon can feed back into the concept. As of yet, there has not been a theoretical framework in which these different branches of the gentrification literature were explicitly connected to one another. This paper has argued that dynamic nominalism provides a framework in which these connections can be drawn. Dynamic nominalism provides the concepts that can be used to study the interaction between the concept and phenomenon of gentrification in concrete cases. It also draws attention to the long-term evolution of the concept. As such, it can be a valuable resource for gentrification research.

The goal of this paper was not to argue that the interaction between gentrification as a concept and as a phenomenon is the only reason for its elusiveness. Rather, the perspective of dynamic nominalism shows that we cannot fully understand the changing nature of the phenomenon of gentrification if we do not take the changing meaning and evaluation of the concept of gentrification into account; and vice versa, we cannot fully understand how the different meanings and evaluations come about if we abstract from gentrification as a changing phenomenon. Dynamic nominalism provides a framework that helps to understand the interaction between the concept of gentrification and the phenomenon it refers to. This framework helps to comprehend how knowledge production and social practices are intertwined in the struggle over access to space.

As this paper’s discussion of the gentrification literature has shown, gentrification has been used more often in newspapers in recent years. It has become part of the conceptual repertoire of an increasing number of people in different parts of the world; thereby, it has come to play an increasingly important role in the shaping of people’s self-understandings. Gentrification thus becomes more and more inevitable as a moral problem that demands consideration, for example, when one decides (not) to rent an Airbnb, (not) to live in a certain neighborhood, (not) to protest against investments in the built environment, or (not) to invest in real estate. This suggests that the concept of gentrification is more important in making up people than it was ever before and that it will therefore have an increasingly important effect on how the phenomenon unfolds. Therefore, I believe, the use of the concepts of dynamic nominalism in gentrification research would be opportune.

At this point, we should return to the problem that was introduced at the beginning of this paper: how can we deal with the paradox faced by gentrification research, that is, that gentrification research has to deal with phenomena that—at least when using Glass’ (2010 [1964]) original definition—cannot be identified as gentrification? As we have seen, dynamic nominalism says that there is an interaction between social phenomena and the concepts used to describe them. This can change a phenomenon to such an extent that a new definition is necessary in order to still be able to use the same word for the changed phenomenon. Gentrification is a moving target and therefore both the concept and the phenomenon are subject to change. Hence, with reference to Wittgenstein (1953: §133), we could say that from the perspective of dynamic nominalism, the paradox of gentrification research is not solved, but rather disappears.

Now that we have seen that gentrification can be understood as a moving target, it is also clear why we cannot grasp gentrification. The reason for this is that the phenomenon changes as a result of the grasping. After all, a description of a process as gentrification influences how this process unfolds. Using the term “gentrification” makes it possible to describe what is going on in some changing neighborhoods, but at the same time, this description alters the phenomenon it refers to and thereby ensures its elusive character. The move that makes knowledge of gentrification possible is thus the same move that ensures that gentrification always eludes the researcher to a certain extent. Dynamic nominalism provides the tools to study this elusiveness. Therefore, we should investigate the moving target that gentrification is and describe the looping effects between concept and phenomenon. Approached in this way, the question of what gentrification is, should no longer only be a starting point of empirical research, but also an always provisional outcome. Attention should be paid to how terms like gentrification and social mix are used by the people involved in the process (policy makers, project developers, people who are displaced, shop owners, gentrifiers themselves, etc.), how these usages alter the phenomenon and how this interacts with the concept’s meanings.