Contesting the deviant other: Discursive strategies for the production of homeless subjectivities

Abstract

This article analyzes the discursive strategies deployed in the linguistic production of homelessness and homeless persons in the context of city-level policies on urban camping in a major urban center in the USA, and outlines the ways that homeless activists contested their use in the public sphere. Homelessness has historically been defined as a deviant form of behavior and the subjectivity of ‘homeless’ has functioned as a social stigma alongside other forms of deviance. I build on analysis of racism and anti-Semitism by examining three linguistic mechanisms deployed by speakers to produce and contest homelessness as a deviant subjectivity: the use of metonyms in processes of synechdochization, the use of metaphors in describing groups and actions, and how processes of lexicalization attach meaning to places and social categories across three categories of deviance prevalent in descriptions of homelessness and urban camping: dirtiness, drugs, and danger.

Keywords

corpus linguistics deviance homeless lexicalization metaphor metonym semantic network analysis social movement subjectivity

Introduction

This article seeks to understand the discursive mechanisms that serve to justify exclusionary social policy on homelessness and marginalize efforts for change by homeless activists and their allies. To answer this question, I examine a cross-genre corpus of activist emails and blog posts, governmental committee and public hearing minutes, mainstream news stories, and alternative journalistic texts about three overlapping campaigns of importance to the local Seattle homeless community, along with filed notes and interviews with participants, by combining critical discourse analysis, corpus linguistics, and participant observation.

I suggest that people experiencing homelessness were politically marginalized through their semantic association with three major categories of social deviance: dirtiness, drugs, and danger. When the City provided vivid descriptions of ‘vile’ and ‘putrid’ camps to justify the ‘clearing’ and ‘cleaning out’ of tents and personal belongings, homeless activists argued that the City was dumping ‘survival gear’ and driving journalistic stereotyping of homeless campers and campsites, and that they needed to ‘stop treating people like trash’. Journalists and city officials talked about how ‘drunks’ and ‘druggies’ were invading neighborhoods, and activists argued that the prohibition of drug and alcohol use in activist tent-cities created a ‘safe, sober space’. When direct-action camping was seen as a ‘breeding ground’ for ‘criminals’, and organized camping efforts resulted in ‘police raids’ and ‘arrests’, those experiencing homelessness argued that they were constantly ‘in fear’ of ‘death’ ‘through weather or violence’ and that tent-cities helped people stay ‘together and safe’. Drawing on the critical discourse analysis of racism and anti-Semitism, I explain how these three semantic fields of deviance were attached to homelessness through the pragmatic use of metonyms in processes of synechdochization and the use of metaphors in describing groups and actions, and how processes of lexicalization attached meaning to places and social categories. By combining qualitative discourse analysis with the quantification of lexical markers across the corpus and an ethnographic grounding in the local social context, I point to the opportunities and limits of how grassroots organizations can foster anti-oppressive meanings among participants. I suggest that the normative reframing strategies explored here were effective in gaining a wider audience for activist interpretations by responding to existing dominant discourses, but often failed to move the focus of debate away from the relative deviance of homeless subjects and towards structural explanations and solutions for homelessness. Broader implications and future directions for an engaged approach to the critical discourse analysis of deviant social categories are explored.

Discursively producing the other: valuing categorical inequality through language use

Critical discourse analysts routinely justify their work by drawing on the social constructivist paradigm (Berger and Luckmann, 1967), suggesting that social inequality is a result of the reproduction of labeling practices that attach social value to identity categories (Reisigl and Wodak, 2001). Many sociologists agree, describing this process as revolving around the production and maintenance of categorical inequality (Massey, 2008). Given a material context of extreme economic inequality in the USA (National Law Center on Homelessness & Poverty, 2008), I would argue that the valuing of categorical attributes is one of the most important social issues of our time, resulting in persistent inequality around racialized, gendered, and classed categories, among others. These cultural routines of representation are largely defined in dominant organizational contexts like government or journalistic fields, where we have seen a historical pattern of representation that attaches negative deviant behavior to those subject positions that hold less privilege in a society (Fairclough, 1992a; Foucault, 1978, 1980; Hall, 1980). This article aims to examine the linguistic practices of importance to the valuing of classed subject positions, and how those practices are employed across institutional contexts.

Three linguistic strategies have been highlighted in research on the production of dominant and subordinate identities. First, metonyms have been shown to be important linguistic mechanisms in processes of synechdochization (Reisigl and Wodak, 2001; Van Leeuwen, 1996). In their work on racism and anti-Semitism, Reisigl and Wodak (2001) found that the representation of a person or group by relying on a part of the whole (and sometimes a part of an associated behavior or geographic residence) was an important rhetorical strategy in the dehumanization of subject positions in the public imagination. By focusing on factors other than race or religion, institutional actors were able to openly condemn entire populations by attaching them to culturally deviant behavior and highlighting that deviance as the focus of criticism and public policy.

Second, Santa Ana (1999) has suggested that metaphors provide a conceptual map through language use, allowing speakers to connect narratives ‘from one semantic source domain to a different semantic target domain’ (p. 193). Drawing on Lakoff (1987), she points to the importance of source domains (areas of social life where we often have direct experience or can otherwise readily draw on easily understood cultural narratives) in providing a metaphorical framework for understanding target domains (often areas where we have little or no direct experience, and where popular cultural narratives are not as easily accessed). In his ‘invariance hypothesis’, Lakoff (1987, 1993) suggested that topic domains, when connected through language, are cognitively summative on the target domain, meaning that the ways in which the proliferation of lexical items in a body of texts is collocated with particular people, places, or actions has long-term cultural importance for what those people, places, or actions mean.

Third, lexicalization – often described as ‘over-lexicalization’ – has been noted as an important strategy in the linguistic production of deviance among groups or populations. Fowler (1991) describes over-lexicalization as ‘an excess of quasi-synonymous terms for entities and ideas that are a particular preoccupation or problem in the culture’s discourse’, noting the ‘proliferation of (often pejorative) words for designating women’ (p. 85). Patterned use of particular qualifiers can create a sense of ‘over-completeness’ (Van Dijk, 1991) for the meaning of an object, person, or activity, as in the lexicalization of a Vietnamese gang as ‘young’, ‘angry’, ‘violent’, ‘evil’, and ‘ruthless’ in the Australian press (Teo, 2000: 21–22). Lazar and Lazar (2004) have supported this pattern, showing that over-lexicalization in US presidential speeches served to establish a sense of ‘moral degeneracy’ in the proliferation of synonymous terms like ‘cruelty, ruthlessness, mercilessness, brutality, and absence of conscience and human decency’, alongside ‘terrorists’ or ‘dictators’ like Osama bin Laden and Saddam Hussein (p. 234). Koller and Davidson (2008) suggest that these lexical patterns can work to represent metaphors, with the ‘lexical realization’ of the ‘exclusion’ metaphor in items like ‘shut out’, ‘shut off’, ‘cut off’, and ‘closure’ (p. 322). In practices of Othering (Hall, 1997) – in the linguistic production of Us and Them – lexicalization is a process whereby deviant meanings can be systematically attached to materially subordinate subjectivities, helping to ‘reinforce the cognitive models that govern discourse, while underlying conceptual metaphors may shape the surface structure of texts’ (Koller and Davidson, 2008: 311).

Work in corpus linguistics suggests that words or word forms build on each other to form fields of meaning, and that, over time, they take on meaning through their relationships to each other in a body of texts. These ‘lexical fields’ then serve as a menu of cognitive associations for participants, privileging some meanings (‘semantic sets’) over others as they are more or less prominent in a given cultural sphere, and allowing analysis of what Salama (2011) calls ‘ideological collocation’ (p. 320). Researchers in the field have sought to operationalize this by examining semantic prosody and discourse prosody (the trend when two lexical items are close to each other in a body of texts) (Louw, 1993; Stubbs, 2001), concluding that the definition of a word is a product of those that keep it company through what Sinclair calls semantic reversal (Sinclair, 1991, 1998). Examples of this work can be found in Tekin’s (2008) analysis of French political discourse, where he suggests that Turkish officials were attached to lexical markers that kept Turkey in a historically deviant role, perpetuating ‘centuries-old stereotypical images regarding Turkey and the Turk’ (p. 738). Similarly, Thurlow and Jaworski (2006) have pointed to the importance of the lexicalization of cultural phenomena in their work on the stylization of frequent flyer programs, where ‘luxury’, ‘gold’, ‘diamonds’, and other culturally resonant lexical terms were important for producing a cultural experience of luxury for consumers.

The symbolic meaning of homelessness has a long history in US popular culture, with media portrayals and academic theory reinforcing the binary divides of deserving and undeserving, normal and deviant. As Wright (1997) has noted, ‘This socially shared image contains other images that convey social distinctions about housed and unhoused people, and judgments about who is deserving of receiving benefits and who is undeserving’ (p. 17). Tropes like ‘welfare mom’, ‘lazy bum’, or even ‘transient’ can reference shared cultural narratives about homeless people as filthy carriers of disease with mental instability and substance abuse problems who engage in criminal behaviors that threaten general social safety (Wagner, 1993). These stereotypes are grounded in real social phenomena that have been created and re-created over time through professionalized media texts and informal communicative settings. The dominant social practices that we have developed as a culture for dealing with homelessness reproduce the social stigmas, or symbolic meanings, associated with those experiencing a lack of adequate housing. Thus, my first research question asks: What dominant discourses were important to the socially shared image of homelessness? Based on the extant literature on media coverage of homelessness, we might expect discourses around ‘drug use’ or ‘alcoholism’ to be important, or descriptions of homeless persons as ‘filthy’ or ‘mentally ill’ to be prominent.

Communication in social movements: symbolic action and the contestation of codes

The production (and contestation) of inequality through language use can also be understood through the lens of social movement theory. The importance of communication in modern social movements has led to a growth in the study of how language and discourse are implicated in social movement processes. Research on ‘new social movements’ has emphasized the turn from material revolutions to symbolic ones (Chesters and Welsh, 2004; Garrido and Halavais, 2003; Swords, 2007), and from reformist tactics to the contestation of codes in highly mediated societies (Melucci, 1996; Rambukkana, 2007). This body of research has drawn significantly on the postmodern turn towards a social constructionist perspective on social research, emphasizing the role that discursive processes play in producing our understandings of people, issues, and events (Berger and Luckmann, 1967; Edelman, 1988). Thus, no way of defining an issue or of describing an event is neutral or somehow objectively ‘true’. Instead, competing interpretations exist simultaneously (Skillington, 1997) in a ‘regime of truth’ structured by ‘the status of those who are charged with saying what counts as true’ (Foucault, 1980: 131). As Melucci (1996) has reminded us, the struggle to define is central to the struggle for material change, and while language has always been a crucial convergence space where interpretations are created and contested, the changing economic and political contexts of information and cultural production in wealthier nations has elevated communication to a prominent location in movement organizing and research (Polletta, 2006; Snow and Benford, 1988; Snow et al., 1986; Steinberg, 1999). These discursive constructions inform how we in turn describe issues and events to others. Viewed at the level of discourse-as-action, ‘Each sign-maker (never merely a sign-user) is therefore a [potential] transformer of the historically shaped resources for representation available in their culture in the light of their interests’ (Kress, 2001: 37). The struggle over which signs are used and what signs mean has become an increasingly important field of resistance, with organizers expending great energy on activities that are, at their core, about changing the way people think about the social world (Eyerman and Jamison, 1991).

Recognizing the importance of language in social change, homeless activists and academics have attempted to redefine popular understandings of what homelessness means (Blau, 1992; Hopper and Baumohl, 1994; Radar, 1986; Wright, 1997). Scholars examining the dominant linguistic production of homelessness have shown a decided trend in mainstream media accounts towards stigmatization and dehumanization (Bawarshi et al., 2008), be they in print (Klodawsky et al., 2002) or television (Lind and Danowski, 1999; Whang and Min, 1999a, 1999b), and activists have sought to change the discourse of homelessness as they seek to change the social practices that surround it, by refocusing public debate on the institutional structures and practices that produce and reproduce homelessness. Street papers – newspapers sold, and sometimes written by, people experiencing homelessness (Van Lier, 1999) – have perhaps been the most prominent venue for this popular critique (Howley, 2003, 2005). However, even in discursive platforms established for the betterment of those experiencing extreme poverty and homelessness, linguistic behavior can still reinforce the negative social ethos associated with mainstream media accounts (Torck, 2001). Thus, my second research question asks: What linguistic strategies did movement participants use to contest the meaning of homelessness? We might expect that those areas of meaning of most interest to homeless activists would be focused on the provision of ‘shelter’ or ‘services’ as they argue for policy changes at the local or national level. This may mean changing the debate from deviant stigmas attached to homeless persons to substantive policy issues.

Methods and case

This project draws on three discrete methodological fields for theory development and empirical measurement: engaged ethnography, critical discourse analysis, and corpus linguistics. These three perspectives were combined with an interpretivist approach to the study as I sought to develop exploratory understandings of the social phenomenon that were true to the situations I experienced as a participant observer, while also drawing from and contributing to existing research in each area examined. Tying the inductive production of theory (Lindlof and Taylor, 2002) to a research design derived from the existing literature was accomplished by using retroductive reasoning (Ragin, 1994) for coding and interpretation.

In 2008–2009, I conducted nine months of formal ethnographic fieldwork, negotiating ‘the boundary between “field” and “fieldworker”’ as a participant observer and critical researcher (Emerson and Pollner, 2001: 241) with a particular interest in communication dynamics. Using a sampling procedure that was responsive to the social context (Gustavsen, 2003), I participated in two overlapping campaigns and one direct-action tent-city project of importance to the local homeless organizing community. First, the campaign to ‘stop the sweeps’ was a response to an increase in the practice by city parks officials of disposing of people’s belongings who were living in area parks, under freeways, and in other refuse spaces. Second, a group of homeless organizers started a new ‘self-managed’ tent city in the region called Nickelsville that combined direct-action camping with long-term shelter provision. Finally, the campaign to stop the construction of a new misdemeanant jail framed homeless people as the primary target for the facility and argued for alternatives to the criminalization of poverty. Participant observation in these three campaigns involved participation in organizing activities, attending events and meetings, informal discussions with participants, jotting down and writing up field notes, and conducting semi-structured one-to-one and small group interviews (Emerson et al., 1995, 2007; see also Toft, 2011).

Along with face-to-face interactions, I also engaged as a participant in the mediated communication of importance to each campaign. My role as participant observer within these campaigns provided qualitative contextual knowledge that helped me to select cases, structure my corpus sampling strategies, and learn from participants what people were talking about and where they were saying it. Mediated texts were identified and collected both in the field (at demonstrations, meetings, public hearings, in conversation) and online, using keyword searches of regional news media, government sites, and advocacy organizations, and through the news.google.com search site. Materials collected through these methods included flyers, newsletters, newspaper articles, radio stories or programs, email and mail correspondence, meeting minutes, web pages, testimonies to city councils, laws or resolutions, petitions, photos, and videos. As I engaged in organizing activities and online information gathering, I built a corpus of 1116 electronic texts (596,545 words) drawn from a broader archive of 1669 digital communication texts. This smaller corpus represents a sub-sample derived by filtering out all texts that did not explicitly mention any of the three issue areas: the Sweeps, Nickelsville and the Proposed Jail. I employed critical discourse analysis as an ethnographic participant in the public dialog, and by examining aggregated samples of texts, seeking out discourses and linguistic patterns at the micro level that had larger policy implications (Cameron, 2001; Fairclough, 2003; Gee, 1999). This method allowed me to identify what stories were being told about homelessness in the texts (Polletta, 1998, 2006), and to parse out the linguistic referents that served as definitional constructs in these stories (Sandberg, 2006).

As I used critical discourse analysis tools to analyze the texts, I also drew on tools from corpus linguistics (Adolphs, 2006; Baker, 2006; Sinclair, 2004; Stubbs, 2001) and semantic network analysis (Doerfel, 1998; Schultz et al., 2012) by employing frequency counts, collocation clustering and mapping, and concordances for analyzing the patterned use of key lexical items in semantic groupings around areas of meaning that I developed from my reading of the texts. The resulting analysis was organized around several qualitatively constructed semantic fields (fields of meaning) that were operationalized through groups of lexical items collected from the corpus (words, phrases, lemmas, and rules), forming dictionaries (groups of lexical items that combine to create an area of meaning). Finally, I returned to critical discourse analysis tools to refine the themes and analyze how they were actualized in the texts in a process that combined qualitative and quantitative strategies for the critical analysis of discourse corpora (Baker et al., 2008).

Analysis: from material poverty to cultural deviance in a local homeless movement

All contests over meaning happen within and between social structures that encourage some ways of speaking and discourage others. The combination of ethnography and discourse analysis offers the researcher an opportunity to examine the contexts of use. I draw on my ethnographic involvement in three overlapping homeless campaigns to present the discursive context within which people experiencing homelessness made sense of their experiences and argued for policy changes. I suggest that the local ‘self-managed’ shelter system created an organizational environment in which people experiencing homelessness were encouraged to talk with each other about contextual explanations for their poverty and homelessness, but that narratives about homeless people struggling to ‘stay together and safe’ in the face of low wages and high rent proved difficult to move into mainstream popular discourse. Instead, professional journalistic spaces proved to be structurally aligned with popular cultural myths about personal responsibility and social deviance, and city-level framing of homelessness policy re-enforced those meanings through press releases and journalistic sourcing patterns. I then focus on a combination of quantitative and qualitative discourse analysis tools for making sense of how social deviance was perpetuated as a salient cultural trope through the semantic association of homelessness with dirtiness, drugs, and danger across genres and organizations. I place my analysis of deviance in the context of grassroots efforts to contest these dominant social meanings through symbolic action, and point towards future strategies for both analysis and intervention in the social production of deviant social categories.

Self-managed shelters as participatory communication spaces

Homeless organizing in Seattle has benefited from a rich history of participatory activism. In response to the often disempowering individualized care that professionalized (and volunteer) shelters and services typically provide (Lyon-Callo, 2004; Villadsen, 2008), a network of organizations has developed that put people experiencing homelessness at the front of a poor people’s movement. By the time of my fieldwork, these participatory structures had become institutionalized resources, facilitating discursive practices that highlighted participant contributions and emphasized structural determinants as responsible for the individual poverty that participants experienced.

Starting with a series of building occupations and direct-action camps in 1980s urban Seattle, SHARE/WHEEL (the Seattle Housing and Resource Effort and their sister organization, the Women’s Housing, Equality and Enhancement League) had developed into a directly democratic shelter network in what they call a ‘self-managed’ housing model for around 500 residents. Each shelter – from tent cities to overnight use of church basements – was democratically run by participants, providing both a material resource and a participatory structure that empowered residents as workers in the community and as citizens in a democracy (Evans, 2008; Krabanow, 2004). As Linda Kenny, a WHEEL participant, put it, ‘we, we all know this experience, um, many of us first hand, um, so we really are in a good position to know what it’s about, and what that population needs’ (Linda Kenny, interview with author, 28 October 2008). Born out of the SHARE/WHEEL self-managed model but organizationally distinct for political reasons, Nickelsville emerged alongside other tent cities nationwide (Herring, 2013) as a direct-action camp in direct violation of the Seattle ‘consent decree’ that specified that only one tent city could exist at a time within city limits (Consent Decree, 2002). This camp was a major site in my fieldwork, as the project became a collaborative focus for the regional movement during ‘super secret’ moves in the wee hours of the morning and calls for solidarity as supplies ran low or police action loomed.

In these self-managed spaces – what Melucci calls ‘free spaces’ (1996) – participants experiencing homelessness were encouraged to think about city policies, economic trends, and organizing strategies, with participation fostering structural interpretations of poverty over individualized failure narratives (Krabanow, 2004). Augustin Lee was a Nickelsville resident during the first few weeks and months of the project. He participated in nightly camp meetings, and dealt with internal group dynamics as well as external relations with allies, the City, and the State. I spoke with him in the immediate aftermath of the first police ‘sweep’ of Nickelsville, when they had just moved their community across a dirt berm from City of Seattle land to a small parking lot owned by the State of Washington. As news helicopters hovered over the camp and TV anchors updated their viewers on the latest news about ‘72-hour notices’ and the recent ‘arrests’ of those who refused to leave, he was quick to point to context, and often in direct disagreement with dominant narratives that blame the poor for their own poverty.

Times have changed ya know I mean you cannot find apartment, ya know a reasonable apartment, ya know for reasonable money no more, you can’t do that ya know … I got about 30, 30 something years of labor experience, ya know, cleaned up, but, um, they still don’t want uh, only pays me, ya know 8 or 9 dollars an hour. How can you live off that? You can’t live off that. I mean you can’t you can’t get an apartment today, a, well a studio today, for under 500 dollars. So, I mean how can, how can you live? You can’t live… (Augustin Lee, interview with the author, 30 September 2008)

Active participation by Nickelodeons (as residents of Nickelsville called themselves) in conversation with allies, nightly and weekly camp meetings, and interactions with journalists and policy makers helped to draw connections between their collective experience of poverty and the larger structural forces that they were trying to change. Talk of low wages, high rent, or getting ‘laid off’ offered powerful narrative structures to participants like Augustin Lee as they risked arrest for their direct-action survival activities.

Moving grassroots meanings into mainstream discourse

The collective sense-making that SHARE/WHEEL and Nickelsville encouraged through face-to-face organizing work became solidified for external audiences in media texts produced by participants and allies. One example is a letter to the editor of the Seattle Times that was collectively authored during one of the weekly Wednesday morning Nickelsville organizing meetings that I attended as a housed ally. It responded to an editorial published by the paper the previous week, by refuting personal responsibility as the dominant cause of poverty, and pointing to low wages and high rents instead:

The wages from many Seattle jobs just don’t pay enough for housing. Lieutenant Gracy should thank his lucky stars that – thanks to taxpayers – he’s got a job with a fair wage. It isn’t just mental illness or drug addiction that causes layoffs and loan defaults – as the good Lieutenant could confirm if he checked around his Precinct Station sometime. (Nickelsville email, 20 December 2008, CTA188)

As collectively authored texts, these letters and emails represent the democratically agreed-upon interpretation of the situation by an organization structured so that people experiencing homelessness were driving the narrative. As was common in these texts, Nickelodeons used ‘wages’/‘jobs’ and ‘layoffs’/‘loan defaults’ as active participants in clauses (Fairclough, 2003), removing personal responsibility or ‘mental illness’/‘drug addiction’ as the primary causes of homelessness (Van De Mieroop, 2011). The letter was never published by the Seattle Times, but it was distributed to supporters through the Nickelsville email list-serve and printed out for distribution to residents and visitors at area tent cities. While activist spaces pointed to structural determinants, dominant discursive spaces like the mainstream daily paper continued to blame the poor for their own poverty.

This focus on social deviance as the dominant journalistic and policy narrative placed the homeless community in a difficult situation: they could continue to talk about structural poverty and systemic causes, or focus on counteracting what they saw as a set of dehumanizing linguistic practices. While the SHARE/WHEEL/Nickelsville self-managed shelter network did actively work on confronting the social deviance frame, I was also a direct participant in this work. As a housed ally and a graduate student, I was approached by the local activist street newspaper Real Change News with a request to examine area news coverage and its relationship to a set of email correspondences between city officials that had been obtained through a Freedom of Information Act (FOIA) request filed by the paper. Real Change News developed out of the vibrant homeless organizing community in Seattle as a professionally run non-profit organization that provided a flexible source of income for low-income and homeless vendors through street sales of a weekly social justice-oriented ‘street paper’ (Howley, 2003). They had published a series of editorials criticizing area journalists for their coverage of the ‘sweeps’ policy, and felt that an academic team of trained analysts might offer less political baggage and more institutional credibility to the conversation.

Armed with the documents secured through the Real Change News FOIA request, and our own growing collection of government documents and media texts, I worked with several professors in English and three other graduate students to craft a multifaceted report that critically examined the discourses surrounding the City’s ‘sweeps’ policy (Bawarshi et al., 2008). We pointed out the close relationship between the City’s careful framing efforts and the patterns of representation that were picked up by area journalists as they reported on the policy and the protests that developed around it. Sections of the report focused on qualitative and quantitative analysis of newspaper and TV coverage, and the use of images in representations of homelessness. We issued a press release, posted the report online, and presented hard copies at a press conference at an overnight campout at City Hall. Even when presented from a position of institutional credibility, our well researched and carefully worded report resulted in no more than a threatening phone call from the City accusing us of biased and shoddy work, and a short mention in a blog post for a local daily paper. Much like the Nickelsville email cited above earlier, our rebuttal was not seen as newsworthy or resonant with the journalistic consensus on representing homelessness.

Self-managed spaces were valuable opportunities for re-framing the homeless experience as rooted in socio-economic structures, but our collective ability to interject this interpretation into the dominant public dialog was limited by a series of institutional constraints that systematically excluded alternative explanations. While they were able to communicate directly to housed allies like myself through their email list-serve and face-to-face visits to camps, the voices and perspectives of people experiencing homelessness were rarely found in the mainstream press, or in the discursive spaces of Seattle City government. Instead, when homelessness was on the news agenda, city officials and journalists would rely on dominant narratives that blamed the poor for their own poverty, and focused on criminalization or individual redemption as appropriate policy solutions. In situations where substantive policy debates are mediated by the perceived legitimacy or deviance of participants in the debate (i.e. mental health, drug and alcohol addiction, unwillingness to accept offers for help), the materiality of inequality can easily become lost in a cacophony of class-based stereotypes.

In the following section, I present key themes on how this occurred in the discursive environment surrounding these three overlapping homeless campaigns. I will do this by presenting several of those dominant linguistic strategies that were most damaging to the public dialog in this case, while also pointing towards some of the strategies that activists used in confronting these discourses. Understanding exactly how deviance is attached to homelessness and homeless activism might (a) help academics understand how language serves to perpetuate structural inequality and (b) aid engaged scholars and homeless activists in constructing effective strategies for changing the structure of public debate.

Contesting deviant subjectivities

Despite the proliferation of alternative explanations in the self-managed and ally community, the dominant production of homelessness and homeless persons in the local communication space occurred through the systematic labeling of those people, spaces, and activities associated with homelessness as deviant. This occurred through the deployment of discourses of filth and disease, addiction and drug use, and of violence and illegality – or in three semantic fields I call dirty, drugs, and danger. Social actors were often produced through language in ways that played on, and extended, existing stereotypes associated with homelessness, poverty, and public safety, resulting in a set of deviant subjectivities that could easily be disregarded or demonized. These deviant subjectivities were accomplished by attaching behaviors and attributes to people and spaces through processes of self- and other-presentation. Similar to other subordinate social subjectivities like race (Van Dijk, 2000a), religion (Reisigl and Wodak, 2001), and citizenship (Santa Ana, 1999), homeless persons were often represented using debilitating labels, delegitimizing groups of people as ‘drunks and druggies’, or as ‘dirty’, ‘scary homeless people’.

These three semantic fields were operationalized around a series of 134 lexical items (words, lemmas, phrases, and rules). Figure 1 shows these lexical items in a collocational network that highlights how these areas of meaning clustered together in use. As we can see, the three campaigns (on the far right) were most closely collocated with danger. We can also see the bridging function that drug discourses (in the center of the graph) played by connecting discussions of dirtiness (upper left) to those of danger (far right). Visualizing collocational networks in this way highlights the overall semantic structure of the public debate on homelessness as it was experienced by participants across genres and publication venues.

Figure 1.

Collocation between lexical items by deviance category. Graph produced using WordStat (Péladeau, 2013) and Gephi (Bastian et al., 2009). Node symbol = semantic field ( = danger, = drugs, = dirty, = homelessness, = campaign). Node and label size = weighted degree. Edges = normalized collocation similarity in a window of 6:6. Edges with a value below .076 excluded and 12 resulting isolates removed for clarity of graph visualization.

While I make no claim that some organizations encouraged only ‘good’ ways of speaking and others only ‘bad’ (Lamb, 2013), the use of deviance as a representational strategy was facilitated by a series of organizational forms that encouraged particular ways of speaking about homelessness. Table 1 shows the correlation between the normalized frequency of the semantic fields of deviance and the four major categories of publications present in the corpus. While mainstream media texts were far more likely to employ discourses of deviance when writing and speaking about homelessness, advocacy organizations were far less likely to draw on these meanings so frequently – even as they also contributed to the reproduction of deviance in their responses to dominant meanings. In struggles over the meaning of persons and activities, homeless activists engaged in active resistance by debating the relevancy of deviant labels, highlighting them as evidence of injustice in the broader society through what Raymond et al. (2014) call normative reframing. But in arguing about the deviance or normalcy of homeless persons, they were at a significant disadvantage, as the discursive terrain remained imbedded in dominant forms of meaning that have strong socio-historical roots in public perceptions of poor and homeless persons (Wright, 1997).

Table 1.

Distribution of deviance discourses across organizational spaces.

		Advocacy	Alt. Media	Main Media	Government
DANGER	Correlation	−0.137**	0.022	0.078*	0.053

	% of total words	0.687	0.676	1.029	0.967
	Word count	201	1197	1659	784

DIRTY	Correlation	−0.082*	0.002	0.095*	−0.035

	% of total words	0.161	0.142	0.216	0.127
	Word count	47	252	348	103

DRUGS	Correlation	−0.085*	−0.045	0.125**	0.010

	% of total words	0.055	0.076	0.152	0.070
	Word count	16	135	245	57
	Total %	0.903	0.894	1.397	1.164

Total	Total word count	264	1584	2252	944

Notes: Pearson correlation reporting 2-tailed p < .01 = *, p < .001 = **.

In what follows, I outline some of the more significant discursive strategies employed in the production of homelessness and homeless persons as deviant, as well as some of the more prominent strategies that homeless activists used in contesting these meanings in the public sphere. I have marked those lexical items that populate the semantic fields examined here by underlining them in excerpts that illustrate trends in the corpus.

Dirty

Homelessness has long been associated with dirtiness. People with few material resources have been portrayed in popular media as wearing filthy clothes, looking through garbage cans for food, peeing in public, as carriers of disease, and as requiring cleaning up. The production of homelessness as dirty in Seattle was achieved through a series of more localized linguistic strategies, focusing on spaces as filthy, people as trash, and the need to clean up those spaces and people who present such a significant health hazard to the housed public. The definition of homelessness as something other than dirty is no simple task, but people associated with the homeless movement in Seattle tackled the language of filth and refuse with a clear strategy of highlighting injustice and renegotiating the meaning of homeless camps as ‘clean’, orderly spaces. One such strategy was the re-labeling by homeless activists of the City’s ‘encampment clearance policy’ as a policy of homeless ‘sweeps’, a phrase with significant enough polysemous potential to be read as both common sense and evidence of injustice.

Spaces as filthy

Homeless camps were attached to filth with a discursive strategy that backgrounded (or completely suppressed) those people who lived in them (Fairclough, 2003) while foregrounding the places where they reside. Reisigl and Wodak (2001) describe this linguistic pattern as a process of ‘spacialization’, whereby toponyms (i.e. homeless camps as ‘not a good health place’) were used as metonymic personifications of a group of people classified as ‘homeless’ (Reisigl and Wodak, 2001: 47, 52).

Descriptions of camps included a growing count of the ‘tons’ of ‘debris’ that city crews had removed so far, their presence often serving as a metonym for campers themselves.

Seattle Parks and Recreation workers remove debris Wednesday from homeless encampments. (Seattle Times, 29 May 2008)

Trash. Rats. Filthy conditions. It is what dozens of homeless people used to live in until this morning. (KOMO TV, 28 May 2008)

They’re full of human waste, and drugs, and I mean, and alcohol bottles and rotting waste, uhh … you, you just have to see it. (Don Harper, on KIRO TV, 28 May 2008)

[C]ity crews had trouble determining what was salvageable among items soaked with rain, dirt or urine … But a crew chief in charge of determining what was salvageable ‘found they were so contaminated with urine she had to back away, and the bags were tossed… (Seattle Times, 14 June 2008)

Backgrounding people by foregrounding filth and human waste spatialized the homeless subject as a ‘public health hazard’, and the audience was asked to align with the workers who were ‘cleaning’ it up over the campers themselves. Local TV news and mainstream newspaper coverage served to fetishize spaces as vile and putrid, so repulsive that ‘she had to back away’, ‘you, you just have to see it’.

People as trash

The assumed value that spacialization placed on homeless persons was further specified through the use of the ‘people as trash’ metaphor. Gibbs (1994) claims that metaphors allow us to see the common-sense understanding of a target domain, and suggests that they can be read as structured statements about the value of a social category (Lakoff, 1987, 1993; Santa Ana, 1999). Speakers contributing to the discourse of homelessness in this corpus used garbage as a metaphor to describe people experiencing homelessness, saying that ‘The base of the Space Needle was littered with men in sleeping bags’ (The Daily, 1 October 2008), or tied an increase in homelessness to an increase in garbage:

There are many names for this fledgling city, where Old Glory flies from improvised flagpoles and trash heaps rise and fall with the wavering population. (High Country News, 16 March 2009)

The use of metaphor to connect a ‘camp’ to a ‘garbage dump’, or ‘trash heaps’ to the ‘wavering population’, moved persons from residents to trash, further spacializing the homeless subject. This contributed to the general understanding of homeless bodies as things that could be ‘littered’ around the base of the Space Needle, and to requests by housed neighbors to ‘clear those people out’ (Matthew Hardesty, quoted in the Seattle PI, 28 January 2008).

Cleaning and clearing

‘City officials’ and ‘Parks employees’ were described as enacting social processes of ‘cleaning’, whereby participants engaged in ‘clearing’ the spaces of ‘debris’ in ‘clean sweeps’ of urban campsites (Fairclough, 2003). This was an important legitimation strategy in representing the ‘sweeps’ policy and the enforcement responses to ‘Nickelsville’ (Van Leeuwen, 2007):

…that’s why these guys are here. To clean out everything left behind. (Anchor, KOMO TV, 28 May 2008)

Again, about 15–20 camp sites to clean up here. (Q 13 TV, 5 June 2008)

The repeated description of the activities of city parks employees as they ‘clean up’ and ‘clean out’ ‘garbage’ that was ‘left behind’ re-enforced the image of camps as dirty spaces and justified editorial headlines that proclaimed ‘City right to clean up squatter encampments’ (Seattle Times, 2 June 2008).

This continued with Nickelsville, but instead of campsites being cleared, there was greater attention paid to the people inhabiting them due to the combination of arrests and the removal of people’s homes:

Four days after more than 100 homeless people established a camp on city property in West Seattle, police cleared the site Friday afternoon, arresting 22 men and women who refused to leave. (Seattle PI, 26 September 2008)

The parallel clause relation of hypotaxis that this excerpt illustrates creates a situation whereby the language that was used to describe how ‘public property is cleared’ was re-purposed in the context of Nickelsville to describe the ‘22 men and women’ who were ‘cleared’ by police. This built on the people as trash metaphor to justify the ‘clearing’ of people from their homes.

‘Stop treating people like trash’

Homeless activists responded to the City’s ‘encampment clearance policy’ early on by highlighting how the practice was unjust and inappropriate, drawing on the metaphor of ‘sweeping’ problems under the rug in news articles and emails about the policy. The following editorial excerpt was published in the activist street newspaper Real Change News:

Just as sweeping dirt under the rug doesn’t really clean the floor, sweeping disabled and homeless people from public view or into jail doesn’t really address homelessness. They are still disabled and homeless when they are released. It is ineffective as hell, but local governments keep sweeping and we keep letting them. (Real Change News, 19 March 2008)

The ‘sweeps’ language was widely adopted by journalists who covered the issue – likely owing to the cultural resonance of the ‘sweeps’ metaphor – but its use was rarely contextualized in the ways that activists would have liked. In part, this was because ‘sweeps’ could often be read as a simile to the need for parks officials to ‘clean up’ those ‘filthy’ homeless people by ‘sweeping’ the camps.

A related strategy employed by homeless activists helped to produce a public perception of Nickelsville as a ‘clean’ place (both visually and rhetorically). Camp residents emphasized cleanliness in discussions with visitors like this area blogger, saying: ‘We want to be a clean society. We’re part of this society. We’re part of Seattle. We don’t want to be a nuisance’ (Rico Miles, ‘head of security’, quoted in Doorstep Politics, 3 March 2009). These efforts were noted in neighborhood blog posts from area residents who visited the camp: ‘There was not a single piece of garbage on the ground, not so much as a gum wrapper’ (Blogging Georgetown, 25 September 2008).

This strategy had mixed results, however, serving to contest the level of filth associated with the Nickelsville community, but showing little evidence of moving the public dialog on homelessness at large away from the relative dirtiness of a socially stigmatized group of people (Lyon-Callo, 2004).

Drugs

The second major area of meaning that was closely associated with the linguistic production of homelessness involved discussions of drugs and alcohol. Two linguistic strategies were prominent in the ways that this issue was attached to homelessness: the lexicalization of substances and drug-related paraphernalia, and the valuing of addicts by the presentation of addiction-as-sin and addiction-as-disease metaphors. The proliferation of substances (as stand-ins for ‘drug use’ or ‘drug users’ in descriptions of camps, green spaces, and community parks) worked alongside discussions of homeless persons as addicts (‘drunks’, ‘druggies’, and ‘addicts’), tying public perceptions of homeless persons to socially outcast deviant behaviors at odds with ‘safe’ communities and ‘improving’ neighborhoods. These associations were directly confronted by homeless activists, constructing Nickelsville as a ‘sober’ and ‘drug-free’ environment, and with some success. But even as participants succeeded in popularizing understandings of Nickelsville as a ‘clean’ and ‘sober’ community, this was accomplished only through the re-inscription of deviant drug use as the norm among homeless populations, setting Nickelsville apart as an exception to the rule. While necessary contests, these discourses were extremely difficult to engage in for participants, playing on deeply held assumptions about personal choice, drug and alcohol use, and the experience of poverty and homelessness (Tupper, 2012).

Substances

Homelessness was attached to drugs through the (over-)lexicalization of homeless encampments as littered with drug paraphernalia (Fowler, 1991; Van Dijk, 1991). This listing of drug paraphernalia often co-occurred with lists of ‘trash’ and ‘human waste’, presenting a vivid image of outdoor spaces filled with deviant substances and inhabited by people doing deviant things:

[C]ity workers have found human waste, syringes, other drug paraphernalia… (Seattle PI, 19 December 2007)

The camp attracts drugs. It may start out a non-violent place but it will not stay one… (Seattle PI, 19 December 2007)

Parks Department employees began their cleanup Wednesday by removing platforms, tarps, old clammy mattresses, open cans of food, bottles of urine, used syringes and mounds of high-alcohol beer cans. (Seattle Times, 2 June 2008)

As with the excerpts above, the drugs themselves often served as metonymic stand-ins for consumption, with the assumption that their presence equaled deviant behavior in their use. These images were constructed as even more troubling by reports that focused on the discomfort that they caused Parks employees, who ‘even with the special suits they were provided for the spring sweep of Kinnear Park, said they weren’t too hot on cleaning up syringes and moldy cans of food’ (Seattle Weekly Blog, 11 September 2008).

Addicts

Those people inferred from the lexicalization of drugs and paraphernalia were also constructed as labeled subject positions. Speakers deployed a strategy of ‘somatisation’, whereby ‘anthroponyms denoting an artificially produced alteration of bodily, sensual or mental capacities’ (i.e. drunks), or ‘negative habitonyms’ (i.e. drug users, addicts) were used as abstractions that represented social actors ‘by means of a quality assigned to them’ (Reisigl and Wodak, 2001: 54; Van Leeuwen, 1996). These representational practices served to depersonalize participants by Othering ‘druggies’, ‘drunks’, and ‘addicts’ as people not like ‘us’.

News articles talked about how ‘the jungle’ (an expanse of city-owned land abutting the freeway) was ‘a haven for transients and drug users who ignore “No Trespassing” signs and clamber over a fence to get in’ (Seattle PI, 26 February 2009). One author profiled a man who camped outdoors, noting that a criminal background check resulted in ‘a screenful of Seattle Municipal Court cases bearing his name, three of them open. One was for disobeying police, and two others were for public drinking. Petty stuff, but they added up to a life where lessons weren’t learned’ (Seattle Times, 30 May 2008). These articles present ‘drinking’ and ‘drug use’ as individual choices (‘lessons weren’t learned’) that were associated with criminal activity (‘a screenful of Seattle Municipal Court cases bearing his name’) and a disregard for the rules (‘ignore “No Trespassing” signs’, ‘disobeying police’).

Conrad and Schneider (1992) describe this approach to the valuing of alcohol use as arising during prohibition, when ‘drunkenness was considered a sin’ (National Temperance Society, quoted in Levine, 1978: 157). In the late 1800s – at a time when opiates were heavily prescribed for a range of medical ailments – ‘Physicians believed that “enslavement” to opiates was caused more by the user’s weak character than the drug itself and considered the lower classes to be particularly vulnerable to it’ (Conrad and Schneider, 1992: 114). Drawing on the addiction-as-sin approach, drinkers/drug users were considered morally corrupt and responsibility for their deviant behavior was placed on the individual (Tupper, 2012). When behavior is labeled as deviant (‘drinking, drug use’) and attributed to the individual’s moral failing (‘lessons weren’t learned’), people can easily be categorized as deserving of their own criminalization.

Nickelsville as ‘safe, sober space’

In responding to the proliferation of drug discourses as associated with homelessness, participants in homeless communities like Nickelsville actively constructed the camp as a ‘safe, sober space’. This process of legitimation through binary categories constructed Nickelsville as exceptional, and the rules that the camp established as a reluctant necessity were responsible for the creation of a space that was capable of keeping ‘alcohol’, ‘liquor’, and ‘drugs’ out. A journalist who often reported on homelessness at a local community radio station talked about meeting ‘a former heroin addict named Billy who was able to kick the drug cold turkey during his first week at Nickelsville’ (KBCS, 8 January 2009). An editorial written by a local pastor at a Methodist church and published by Real Change News talked about the relationship between self-managed communities and ‘sober space’:

SHARE embodies the idealism that those who are homeless can manage their own lives … In my own church we provide space for a SHARE shelter. It works. Folks are given basic human decency, and they are offered safe, sober space so that they can be strengthened to move from homelessness to housing and from survival to hope. (Real Change News, 19 November 2008)

In presenting self-managed communities like SHARE shelters and Nickelsville as spaces where ‘Folks are given basic human decency’, these reports connect ‘those who are homeless’ to opportunities for them to ‘manage their own lives’ in a ‘sober space’, moving people from ‘survival to hope’. But in the context of stereotypes promulgated by the City and area journalists, Nickelsville could occupy no middle ground: either the camp was an exceptional space characterized by strict rules that kept residents in check, or it was a reaffirmation that homeless people are drunks and druggies, unable to control themselves or others.

Danger

Homelessness was associated with danger as individuals and groups were criminalized through the use of criminonyms (Reisigl and Wodak, 2001) in their representation as social actors (‘thieves’, ‘criminals’, ‘felons’), and as recipient participants in ‘illegal’ activities alongside ‘police’. First, survival activity was criminalized through the implementation of city policies and police raids on camps, resulting in a representational strategy that presented camping as criminal. Second, as homeless activists tried to build bridges with community groups and criminal justice activists, definitions of homelessness became more closely tied to criminal identities, supporting stereotypes of the homeless as dangerous. This second strategy was most closely tied to the Proposed Jail campaign, suggesting that the issue provided a particularly difficult set of semantic meanings that participants were faced with in the course of organizing. But descriptions of homelessness also made use of danger in the representation of the homeless experience as one filled with danger ‘through weather or violence’, as characterized by a large probability of ‘death’, and where ‘survival activities’ were ‘criminalized’ through the implementation of the ‘mayor’s sweeps’ policy. In this context of fear and danger, self-managed tent cities like Nickelsville were presented as places where people had ‘no fear’, creating an environment where people experiencing homelessness could stay ‘together and safe’.

The homeless danger

The positioning of the audience in reports about homelessness often juxtaposed the general population (‘people’) with a deviant group that threatened general public safety by reinforcing fears of the homeless danger. This strategy can be understood as a practice of Othering, whereby the in-group is understood through the discursive production of the out-group (Hall, 1997; Said, 1978). By positioning the reader/viewer as part of the in-group, and then attaching deviant behaviors to the out-group, this representational strategy served to dehumanize homeless participants and justify policing activity.

The presentation of fear as something experienced by ‘people’ or ‘residents’ assumes a housed audience and negates an unhoused one by drawing on journalistic sourcing patterns that privilege housed perspectives (Silberstein, 2008; Whang and Min, 1999a). A Q 13 TV spot starts with an image of a homeless encampment in Seattle’s expensive Queen Anne neighborhood, along with a voice-over saying that ‘people are fearful to come in here’ (Q 13 TV, 28 May 2008). One article in the Seattle Weekly specifically lists the many fears that (housed) neighbors might have about experiencing a tent city in their neighborhood, noting that ‘Eastside residents protested the move, claiming that they feared for everything from the safety of schoolchildren to their own property values’, but went on to say that ‘At each site, the group has largely disproved the worst fears of timid citizens’, providing a list of attributes intended to show how the camp had done this:

The encampments are clean and orderly. SHARE/WHEEL bans drunks and druggies from its tent cities and shelters. It runs warrant checks on potential clients. Twenty-five percent of SHARE/WHEEL clients are employed, according to the group, with many working day labor jobs in Seattle and elsewhere. The group claims that 46 percent of its clients leave the SHARE/WHEEL system within a year and find permanent housing. (Seattle Weekly, 22 March 2006)

The myriad ‘security’ measures that tent cities operate with was a frequent explanation for why people should not be fearful, with statements like ‘Typically, those fears don’t pan out, in part because of the tight reins the camp organizers keep on those who live in the camps’ (Renton Reporter, re-posted in the local Save Feral Human Habitat blog, 4 March 2009).

The ‘super secret’ direct-action camping strategy that Nickelsville employed also resulted in expressions of fear over a community not having access to the same kind of pre-move public education process that neighborhoods who host permitted Tent City 3 and 4 stays would typically benefit from. Local television coverage of Nickelsville gave this a stage by privileging housed neighbors as sources and profiles in news reports:

Dan Mullins, who lives in Highland Park within a mile of the tents, said some in the neighborhood may be angry or fearful after being caught by surprise. (KOMO TV, 25 September 2008)

Blair and Dina Johnson have seen the area around their Highland Park neighborhood improve dramatically over the years. But now they fear [an] unwanted new neighbor could change all that. (King 5, 25 May 2008)

This loss of control (‘being caught by surprise’) and fear that their neighborhood would revert to a time before it ‘improve[d] dramatically’ due to the arrival of Nickelsville positioned the viewer as housed, and called on assumptions about the behavior of homeless people and the degradation of poor neighborhoods that went unmentioned in these reports.

Camping as criminal

Homeless participants were labeled as criminals via the predicative strategy of criminonyms (i.e. felons, criminals) (Reisigl and Wodak, 2001: 52). But the semantic category of criminality was also attached to participants via the activities they engaged in (processes) and the spaces they inhabited (circumstances) (Fairclough, 1992b). This close semantic connection between criminality and fear supported emotional responses that were more likely to foreground stereotypes and discourage the recognition of similarities and shared grievances between poor neighborhoods and homeless ‘campers’ (Harris and Fiske, 2006).

Examples of these strategies can be seen in editorials such as one about encampments in the ‘jungle’ that talked about how ‘Crime boiled into neighborhoods’, writing that ‘It took almost a year to nab dozens of criminals and reclaim the woods, but we did’ (Seattle PI, 7 December 2007). This trend was extended with Nickelsville when direct-action camping was described as ‘illegal’ by city officials who regulated their activity through ‘arrests’ and ‘fines’, leading to the production of campers as ‘criminal’ for their occupation of public property:

The city warned them they were trespassing and would be evicted or arrested if they didn’t leave within 72 hours, a standard deadline for the city when dealing with the homeless. (King 5, 27 September 2008)

Nickels … [is] … cracking down on illegal homeless encampments. (NWCN, 1 October 2008)

Nickelsville … [is an] … illegal homeless camp. (King 5, 29 September 2008, CTA1550)

[P]olice … [came to] … sweep away Nickelsville, its 150 illegal fuchsia tents and its 100 or so residents. (Seattle Times, 26 September 2008)

Positioning Nickelsville as within the realm of the city’s Sweeps policy meant that the ‘homeless camp’ and its ‘fuchsia tents’ could be presented as ‘illegal’, and created a situation where individuals were criminalized, with area blogs noting that ‘As of 12:35 p.m., the homeless residents of Nickelsville were officially considered criminal trespassers and subject to arrest’ (Seattlest, 26 September 2008). The referential identification of homeless campers as ‘criminal trespassers’ illustrates how the Sweeps policy encouraged depersonalizing characterizations of people experiencing homelessness through the use of ‘criminonyms’ in journalistic accounts (Reisigl and Wodak, 2001: 52).

‘Staying together and safe’

Homeless activists and housed allies shifted the subject position on issues of danger by describing the experience of homelessness as marked by threats of ‘violence’ and the Sweeps as placing people ‘in danger’ from the elements by confiscating their ‘survival gear’:

Some evicted wind up sleeping on sidewalks with no tent or sleeping bag, in danger of freezing to death on icy winter nights. (Local blogger Eugene Cho, 2 October 2008, CTA1378)

They’re lucky to have a tent and a place to put it to survive. When you take away their means to survive, you’re going to kill a percentage of them – guaranteed. So as far as I’m concerned, the Sweeps policy is about killing people. (Housed ally Mike Smith, quoted in Seattle PI, 9 June 2008)

This strategy was supported by groups like Women in Black who bore public witness to homeless people who had died ‘outside, by weather or violence’, creating a semantic relationship between the Sweeps policy and death, and connecting encampments like Nickelsville with ‘survival’:

In 2007 forty-six homeless deaths occurred outside or through violence in King County, and so far this year there have been eleven. The creation of Nickelsville is a matter of survival. (Eat the State, 1 May 2008)

The focus on Nickelsville as a ‘survival mechanism’ was articulated in emails and website posts from people in the homeless community that characterized the community as a project that was intended to keep people ‘together and safe’. Action calls prior to the camp’s first location called for ‘People who are willing to take direct action to help us stay together and safe’ (Nickelsville email, 12 September 2008). While such actions were covered in the activist street newspaper Real Change News with headlines like ‘Nickelsville is coming; you can tell by the dead people’ (Real Change News, 27 August 2008), they also provided news hooks for coverage in daily papers, free weeklies, and on local TV news programs, providing an opening for a shift in the preferred positionality of the reader on the experience of homelessness.

Discussion

I opened this article by asking about both dominant and contested strategies for representing homelessness in the public sphere. By looking at the dominant discourses surrounding a local homeless movement, we have seen how those experiencing homelessness were attached to deviant attributes through the deployment of powerful cultural tropes. Descriptions of camps and the people who live in them revolved around how dirty those spaces were, the proliferation of drugs, and the danger that the assumed housed audience experienced by their very presence in the neoliberal city. Those representational practices often associated with racism (Van Dijk, 2000a), anti-Semitism (Reisigl and Wodak, 2001), and anti-immigration discourses (Santa Ana, 1999; Van Der Valk, 2003; Van Dijk, 2000b) were also of central importance in the discursive dehumanization of homeless persons, serving to further the reproduction of the categorical inequalities that structure much of modern institutional life (Massey, 2008). Those linguistic strategies that held particular resonance when producing homeless subjectivities fell under three basic analytic categories: the use of metonyms and processes of synechdochization (Reisigl and Wodak, 2001; Van Leeuwen, 1996), the use of metaphors (Lakoff, 1987), and the (over-)lexicalization of deviant attributes (Fowler, 1991; Stubbs, 2001; Thurlow and Jaworski, 2006; Van Dijk, 1991). Looking across categorical domains at linguistic strategies in use may help to inform cross-issue organizing efforts that highlight common problems and an integrated critique of dominant social institutions.

To best understand how groups of people were characterized as deviant ‘drunks and druggies’, or as ‘dirty’, ‘scary homeless people’, we need to look beyond those dominant discourses that serve to marginalize and stratify, and examine also those strategies that participants employed to change the stories that structured their lives (Eyerman and Jamison, 1991; Reinsborough and Canning, 2008). I described how the SHARE/WHEEL and Nickelsville ‘self-managed’ shelter community allowed participants to speak to each other about their collective experiences and work to change social policy through symbolic actions (Krabanow, 2004; Melucci, 1996). This focus on structural determinants and contextual factors was hard to move into the larger public dialog, however. Instead, participants were largely stuck contesting their relative deviance, drawing primarily on two rhetorical strategies. First, those deviance labels that were attached to homelessness were highlighted as an injustice in their own right, and they were replicated in blogs, editorials, and academic reports as evidence of how people living outside were being dehumanized by ‘housed society’ and figuratively ‘swept’ under the rug (Bawarshi et al., 2008). Second, activists and camp residents at places like Nickelsville constructed definitions of themselves and their homes that directly countered dominant stereotypes, relying on binary oppositions to claim that the camp was ‘clean’ and ‘sober’, and that they had strict rules about ‘no violence’, ensuring a ‘safe’ environment. Raymond et al. (2014) suggest that this strategy is a form of normative reframing, in which activists ‘highlight the unconvincing “fit” of the existing norm to the relevant issue and then “reframe” the issue in terms of an alternative norm that they argue is more appropriately applied to this context’ (p. 200). Attention to the contested space of social meanings helps to move our analysis from one that privileges powerful social institutions to one that begins to recognize the value of peripheral/subordinate socio-linguistic contexts in facilitating social change.

This analysis has focused on the linguistic production of deviance, and illustrated some of the nuanced ways in which these meanings are perpetuated in public discourse by balancing a critique of dominant discourses with social change efforts across organizational spaces and textual genres. These normative reframing efforts did serve to shift some of the ways that social stigmas were described in journalistic accounts of homelessness, but they still left the semantic attention on whether people experiencing homelessness were or were not ‘filthy’, ‘drunk’ ‘criminals’. If we are to take seriously the insight that a semantic approach to issue contestation might offer, then effective use of those discourses most able to illustrate the underlying structural factors that perpetuate the material poverty of homelessness – unemployment, poverty, debt, housing costs, low wages, etc. – might be more likely to shift our collective infatuation away from personal responsibility and toward those social policies that might actually foster a more just and equal society. Instead of staying in the space of normative reframing (Raymond et al., 2014), activist spaces can be valuable resources for shifting the focus of debate through normative innovation, where advocates ‘attack and change the norm itself, rather than the norm’s fit to the issue’ (p. 201). I have illustrated, however, that even as participatory collective sense-making practices in the self-managed community clearly offered alternative explanations for participants’ poverty, the social context of dominant media environments ensured that deviance continued to be a powerful cultural trope. If our ability to speak and be heard is tied to our alignment with those discourses that are powerful in a given cultural moment, then our efforts at strategic intervention require a more complicated engagement with those discourses as we weave together alternative explanations and solutions. Future research should focus on how we can effectively connect our rebuttal of deviance as a measure of social importance to structural understandings of social stratification that might more readily translate into effective public policy.

Footnotes

Acknowledgements

An earlier version of this paper was presented at the annual meetings of the American Association for Applied Linguistics, Chicago, IL, 26–29 March 2011 and the International Communication Association, Phoenix, AZ, 24 May 2012. I would like to thank Kirsten Foot for her invaluable feedback as I developed this project, Kristin Gustafson and Lauren Berliner for acting as sounding boards as I revised the manuscript, and the many participants who took time out of their lives to think through these ideas with me. Part of this research was funded by the Peter Clark Public Scholarship Research Award, University of Washington, March, 2008.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Author biography

Amoshaun Toft is an assistant professor at the University of Washington Bothell in the School of Interdisciplinary Arts and Sciences, where he studies and teaches about language, media, and social change.

References

Adolphs

(2006) Introducing Electronic Text Analysis. London: Routledge.

Baker

(2006) Using Corpora in Discourse Analysis. London: Continuum.

Baker

Gabrielatos

Khosravinik

. (2008) A useful methodological synergy? Combining critical discourse analysis and corpus linguistics to examine discourses of refugees and asylum seekers in the UK press. Discourse & Society 19(3): 273–306.

Bastian

Heymann

Jacomy

(2009) Gephi: An open source software for exploring and manipulating networks. In: International AAAI Conference on Weblogs and Social Media, San Jose, CA, 17–20 May.

Bawarshi

Dillon

Kelly

. (2008) Media Analysis of Homeless Encampment ‘Sweeps’. Seattle, WA: University of Washington.

Berger

Luckmann

(1967) The Social Construction of Reality. Garden City, NY: Doubleday.

Blau

(1992) The Visible Poor: Homelessness in the United States. Oxford: Oxford University Press.

Cameron

(2001) Working with Spoken Discourse. London: SAGE.

Chesters

Welsh

(2004) Rebel colors: ‘Framing’ in global social movements. Sociological Review 52(3): 314–335.

10.

Conrad

Schneider

(1992) Deviance and Medicalization: From Badness to Sickness. Philadelphia, PA: Temple University Press.

11.

Consent Decree (2002) Consent Decree No. 49428-7-I: Share/Wheel and El Centro De La Raza vs. City of Seattle et al. King County Superior Court. Available at: http://seattlest.com/attachments/sarahlloyd/ConsentDecree.pdf (accessed 14 March 2014).

12.

Doerfel

(1998) What constitutes semantic network analysis? A comparison of research and methodologies. Connections 21: 16–26.

13.

Edelman

(1988) Construction of the Political Spectacle. Chicago, IL: University of Chicago Press.

14.

Emerson

Pollner

(2001) Constructing participant/observation relations. In: Emerson

(ed.) Contemporary Field Research: Perspectives and Formulations, 2nd edn. Long Grove, IL: Waveland Press, pp. 239–259.

15.

Emerson

Fretz

Shaw

(1995) Writing Ethnographic Fieldnotes. Chicago, IL: University of Chicago Press.

16.

Emerson

Fretz

Shaw

(2007) Participant observation and fieldnotes. In: Atkinson

Delamont

Coffey

. (eds) Handbook of Ethnography. London: SAGE, pp. 352–368.

17.

Evans

(2008) A City within a City: Citizenship at the Margins. Seattle, WA: University of Washington Press.

18.

Eyerman

Jamison

(1991) Social Movements: A Cognitive Approach. University Park, PA: Pennsylvania State University Press.

19.

Fairclough

(1992a) Discourse and Social Change. Cambridge: Polity Press.

20.

Fairclough

(1992b) Discourse and text: Linguistic and intertextual analysis within discourse analysis. Discourse & Society 3(2): 193–217.

21.

Fairclough

(2003) Analyzing Discourse: Textual Analysis for Social Research. London: Routledge.

22.

Foucault

(1978) The History of Sexuality, Vol. 1: An Introduction. New York: Vintage Books.

23.

Foucault

(1980) Power/Knowledge. Brighton: Harvester.

24.

Fowler

(1991) Language in the News: Discourse and Ideology in the Press. London: Routledge.

25.

Garrido

Halavais

(2003) Mapping networks of support for the Zapatista Movement: Applying social-network analysis to study contemporary social movements. In: McCaughey

Ayers

(eds) Cyberactivism: Online Activism in Theory and Practice. New York: Routledge, pp. 165–184.

26.

Gee

(1999) An Introduction to Discourse Analysis: Theory and Method. London: Routledge.

27.

Gibbs

(1994) Poetics of the Mind: Figurative Thought, Language and Understanding. Cambridge, MA: Cambridge University Press.

28.

Gustavsen

(2003) Action research and the problem of the single case. Concepts and Transformation 8(1): 93–99.

29.

Hall

(1980) Culture, Media, Language: Working Papers in Cultural Studies, 1972–79. London: Centre for Contemporary Cultural Studies, University of Birmingham.

30.

Hall

(1997) Chapter 4: The spectacle of the ‘Other’. In: Hall

(ed.) Representation: Cultural Representations and Signifying Practices (Culture, Media and Identities Series). London: SAGE/Open University, pp. 223–290.

31.

Harris

Fiske

(2006) Dehumanizing the lowest of the low: Neuroimaging responses to extreme out-groups. Psychological Science 17(10): 847–853.

32.

Herring

(2013) The new logics of homeless seclusion: A comparative study of large-scale homeless encampments in the western US. Report no. 21 (RC21). Berlin: Research Committee.

33.

Hopper

Baumohl

(1994) Held in abeyance: Rethinking homelessness and advocacy. American Behavioral Scientist 37(4): 522–552.

34.

Howley

(2003) A poverty of voices: Street papers as communicative democracy. Journalism: Theory, Practice and Criticism 4(3): 273–292.

35.

Howley

(2005) Community Media: People, Places, and Communication Technologies. Cambridge: Cambridge University Press.

36.

Klodawsky

Farrell

D’Aubry

(2002) Images of homelessness in Ottawa: Implications for local politics. Canadian Geographer 46(2): 126–143.

37.

Koller

Davidson

(2008) Social exclusion as conceptual and grammatical metaphor: A cross-genre study of British policy-making. Discourse & Society 19(3): 307–331.

38.

Krabanow

(2004) Making organizations work: Exploring characteristics of anti-oppressive organizational structures in street youth shelters. Journal of Social Work 4(1): 47–60.

39.

Kress

(2001) From Saussure to critical sociolinguistics: The turn towards a social view of language. In: Wetherell

Yates

Taylor

(eds) Discourse Theory and Practice: A Reader. London/Thousand Oaks, CA: SAGE, pp. 29–38.

40.

Lakoff

(1987) Women, Fire, and Dangerous Things: What Categories Reveal about the Mind. Chicago, IL: University of Chicago Press.

41.

Lakoff

(1993) The contemporary theory of metaphor. In: Ortony

(ed.) Metaphor and Thought, 2nd edn. Cambridge, MA: Cambridge University Press, pp. 202–251.

42.

Lamb

(2013) Power and resistance: New methods for analysis across genres in critical discourse analysis. Discourse & Society 24(3): 334–360.

43.

Lazar

(2004) The discourse of the new world order: ‘Out-casting’ the double face of threat. Discourse & Society 15(2–3): 223–242.

44.

Levine

(1978) The discovery of addiction: Changing conceptions of habitual drunkeness in America. Journal for the Study of Alcohol 39(1): 143–174.

45.

Lind

Danowski

(1999) The representation of the homeless in US electronic media: A computational linguistic analysis. In: Min

(ed.) Reading the Homeless: The Media’s Image of Homeless Culture. Westport, CT: Praeger, pp. 109–120.

46.

Lindlof

Taylor

(2002) Qualitative Communication Research Methods, 2nd edn. Thousand Oaks, CA: SAGE.

47.

Louw

(1993) Irony in the text or insincerity in the writer? The diagnostic potential of semantic prosodies. In: Baker

Francis

Tognini-Bonelli

(eds) Text and Technology: In Honour of John Sinclair. Amsterdam: John Benjamins, pp. 157–176.

48.

Lyon-Callo

(2004) Inequality, Poverty, and Neoliberal Governance: Activist Ethnography in the Homeless Sheltering Industry. Peterborough, ON: Broadview Press.

49.

Massey

(2008) Categorically Unequal: The American Stratification System. New York: Russell Sage Foundation.

50.

Melucci

(1996) Challenging Codes: Collective Action in the Information Age. Cambridge: Cambridge University Press.

51.

National Law Center on Homelessness & Poverty (2008) Changing Laws, Changing Lives: 2007 Annual Report. Available at: http://www.nlchp.org/content/pubs/2007_Annual_Report2.pdf (accessed 22 October 2009).

52.

Péladeau

(2013) Wordstat: Content analysis module for QDA Miner. Vers. 6.1.19. Computer software. Provalis Research.

53.

Polletta

(1998) ‘It was like a fever’: Narrative and identity in social protest. Social Problems 45(2): 137–159.

54.

Polletta

(2006) It Was Like a Fever: Storytelling in Protest and Politics. Chicago, IL: University of Chicago Press.

55.

Radar

(1986) Signal through the Flames: Mitch Snyder and America’s Homeless. Kansas City, MI: Sheed & Ward.

56.

Ragin

(1994) Constructing Social Research: The Unity and Diversity of Method. Thousand Oaks, CA: Pine Forge Press.

57.

Rambukkana

(2007) Is slash an alternative medium? ‘Queer’ heterotopias and the role of autonomous media spaces in radical world building. Affinities: A Journal of Radical Theory, Culture and Action 1(1): 69–85.

58.

Raymond

Weldon

Kelly

. (2014) Making change: Norm-based strategies for institutional change to address intractable problems. Political Research Quarterly 67(1): 197–211.

59.

Reinsborough

Canning

(2008) Re:imagining Change: An Introduction to Story-based Strategy. Oakland, CA: PM Press.

60.

Reisigl

Wodak

(2001) Discourse and Discrimination: Rhetorics of Racism and Antisemitism. London: Routledge.

61.

Said

(1978) Orientalism. New York: Random House.

62.

Salama

AHY

(2011) Ideological collocation and the recontexualization of Wahhabi-Saudi Islam post-9/11: A synergy of corpus linguistics and critical discourse analysis. Discourse & Society 22(3): 315–342.

63.

Sandberg

(2006) Fighting neo-liberalism with neo-liberal discourse: Attac Norway, Foucault and collective action framing. Social Movement Studies 5(3): 209–227.

64.

Santa Ana

(1999) ‘Like an animal I was treated’: Anti-immigrant metaphor in US public discourse. Discourse & Society 10(2): 191–224.

65.

Schultz

Kleinnijenhuis

Oegema

. (2012) Strategic framing in the BP crisis: A semantic network analysis of associative frames. Public Relations Review 38(1): 97–107.

66.

Silberstein

(2008) How would you feel? TV coverage of the homeless camps in Seattle. In: Bawarshi

Dillon

Kelly

. (eds) Media Analysis of Homeless Encampment ‘Sweeps’. Seattle, WA: University of Washington.

67.

Sinclair

(1991) Corpus, Concordance, Collocation. Oxford: Oxford University Press.

68.

Sinclair

(1998) The lexical item. In: Weigand

(ed.) Contrastive Lexical Semantics. Amsterdam: John Benjamins, pp. 1–24.

69.

Sinclair

(2004) Trust the Text: Language, Corpus and Discourse. London: Routledge.

70.

Skillington

(1997) Politics and the struggle to define: A discourse analysis of the framing strategies of competing actors in a ‘new’ participatory forum. British Journal of Sociology 48(3): 493–513.

71.

Snow

Benford

(1988) Ideology, frame resonance, and participant mobilization. In: Klandermans

Kriesi

Torrow

(eds) From Structure to Action: Comparing Social Movements Research across Cultures (International Social Movement Research: A Research Annual), vol. 1. Greenwich, CT: JAI Press, pp. 197–217.

72.

Snow

Baker

Anderson

. (1986) The myth of pervasive mental illness among the homeless. Social Problems 33(5): 407–423.

73.

Steinberg

(1999) The talk and back talk of collective action: A dialogic analysis of repertoires of discourse among nineteenth-century English cotton spinners. American Journal of Sociology 105(3): 736–780.

74.

Stubbs

(2001) Words and Phrases: Corpus Studies of Lexical Semantics. Oxford: Blackwell.

75.

Swords

ACS

(2007) Neo-Zapatista network politics: Transforming democracy and development. Latin American Perspectives 34(2): 78–93.

76.

Tekin

BÇ

(2008) The construction of Turkey’s possible EU membership in French political discourse. Discourse & Society 19(6): 727–763.

77.

Teo

(2000) Racism in the news: A critical discourse analysis of news reporting in two Australian newspapers. Discourse & Society 11(1): 7–49.

78.

Thurlow

Jaworski

(2006) The alchemy of the upwardly mobile: Symbolic capital and the stylization of elites in frequent-flyer programmes. Discourse & Society 17(1): 99–135.

79.

Toft

(2011) Contextualizing technology use: Communication practices in a local homeless movement. Information Communication & Society 14(5): 704–725.

80.

Torck

(2001) Voices of the homeless people in street newspapers: A cross-cultural exploration. Discourse & Society 12(3): 371–392.

81.

Tupper

(2012) Psychoactive substances and the English language: ‘Drugs’, discourses, and public policy. Contemporary Drug Policy 39: 461–492.

82.

Van De Mieroop

(2011) Identity negotiations in narrative accounts about poverty. Discourse & Society 22(5): 565–591.

83.

Van Der Valk

(2003) Right-wing parliamentary discourse on immigration in France. Discourse & Society 14(3): 309–348.

84.

Van Dijk

(1991) Racism and the Press. London: Routledge.

85.

Van Dijk

(2000a) New(s) racism: A discourse analytical approach. In: Cottle

(ed.) Ethnic Minorities and the Media. Buckingham: Open University Press, pp. 33–49.

86.

Van Dijk

(2000b) On the analysis of parliamentary debates on immigration. In: Wodak

Van Dijk

(eds) Racism at the Top: Parliamentary Discourses on Ethnic Issues in Six Parliamentary States. Klagenfurt, Austria: Drava Verlag, pp. 85–104.

87.

Van Leeuwen

(1996) The representation of social actors. In: Caldas-Coulthard

Coulthard

(eds) Texts and Practices: Readings in Critical Discourse Analysis. London: Routledge, pp. 32–70.

88.

Van Leeuwen

(2007) Legitimation in discourse and communication. Discourse & Communication 1(1): 91–112.

89.

Van Lier

(1999) Selling the street beat. The Quill 87(8): 16.

90.

Villadsen

(2008) Doing without state and civil society as universals: ‘Dispotifs’ of care beyond the classic sector divide. Journal of Civil Society 4(3): 171–191.

91.

Wagner

(1993) Checkerboard Square: Culture and Resistance in a Homeless Community. Boulder, CO: Westview Press.

92.

Whang

Min

(1999a) Blaming the homeless: The populist aspect of network TV news. In: Min

(ed.) Reading the Homeless: The Media’s Image of Homeless Culture. Westport, CT: Praeger, pp. 109–120.

93.

Whang

Min

(1999b) Discourse analysis of television news on public antagonism against the homeless. In: Min

(ed.) Reading the Homeless: The Media’s Image of Homeless Culture. Westport, CT: Praeger, pp. 95–107.

94.

Wright

(1997) Out of Place: Homeless Mobilizations, Subcities, and Contested Landscapes. Albany, NY: State University of New York Press.