“Blood Has No Colour”: Racialized Donor In/Ex-clusion in the South African National Blood Service

ABSTRACT Accounting for challenges with HIV transmission and testing, the South African National Blood Service (SANBS) transitioned toward eliminating race as a risk categorization in 2005 and actively recruiting black donors. I trace the racialization and nationalization of blood through an analysis of this transition, outreach efforts, and data from fieldwork with blood donors and SANBS staff. I examine indexicality as a semiotic means of in/ex-clusion in blood donation. Due to the sociocultural and medical significance of blood, an ethnographic account of blood services provides insights into biological citizenship and the dynamics of justice and reconciliation in post-apartheid South Africa.

in/ex-clusion, as do the purported relations between biology and racial identity. Which individual singulars (e.g. biological citizens) these pronouns include and their relationships to the collective plural (e.g. the nation), depend on contextual factors, especially race (of speakers/audience), place (e.g. township), and language (e.g.isiZulu). Though examples of indexicality I demonstrate that "biological citizenship is both individualizing and collectivizing," not only through material means such as physical donation, epidemiologic and genetic methods, and blood supply distribution, but through semiotic negotiations (Rose and Novas 2005:441). This semiotic specificity supplements broader explorations of the metaphoric, symbolic, and expressive power of blood (Carsten 2013). The use and circulation of linguistic features, and their effectiveness to convey meaning(s) to (intended) audiences in outreach campaigns provide insights into the blood service's communicative control over the boundaries of bio-civic belonging.
Although the terms of in/ex-clusion have shifted through South Africa's political transitions, in synthesizing semiotic, relational, and material analyses, I show that racial difference remains central to determinations of biological citizenship. More so than in Euro-American contexts where blood service exclusions often targeted immigrants and ethnic minorities, significant SANBS policy shifts were tethered to a crucial moment of black majority enfranchisement in which racial reconciliation was a state priority. I argue that given the evocative qualities of blood, its ability to both unify and divide people in pivotal political transitions, and how it is nationalized through a blood service, make it a crucial testing ground for the limits and possibilities of racial justice and reconciliation in the wake of apartheid.

Public narratives, complex challenges
Shifts in blood services' donor criteria and recruitment were tied to political transitions in South Africa at the end of the 20th century. In addition to disenfranchising the majority of South Africans based on racial classification, the apartheid regime criminalized homosexuality and regulated sexuality. The dissolution of apartheid and its juridical system of racial segregation began in 1989, and enfranchisement of all South African citizens led to the election of Nelson Mandela as president in 1994. At the time, this was seen as a triumph for democracy and human rights, as was the Truth and Reconciliation Commission (TRC) which provided public testimony and amnesty for those who had committed atrocities under the apartheid regime. This transitional period included health sector hearings in the TRC process, desegregation of the health care services, and the establishment of national and provincial health departments (De Gruchy et al. 1998). The new constitution enshrined the right to access health care services, and protections against discrimination on the bases of race and sexual orientation (Steyn and van Zyl 2015). During this era, the challenge of the HIV epidemic loomed. Globally, the transmission of HIV in the blood supply became a major concern. For South Africa, the epidemic was even more severe.
In the decade following South Africa's transition to democracy, SANBS struggled with two "blood wars" that tested post-apartheid juridical principles. Media, political actors, activists, and ordinary people weighed in on SANBS's respective restrictions on black and men who have sex with men (MSM) 3 donors based on HIV prevalence (King 2012). In these controversies, SANBS representatives maintained that the limited use of black donors' blood, and the deferral of MSM, were criteria based on risk calculations, international standards, and epidemiological imperatives to ensure the safety of blood for transfusion (Heyns et al. 2006). MSM deferral, which many blood services adopted globally, was based on high infection rates, and reinforced widespread use of the term MSM itself as a behavior rather than an identification as gay, but also resulted in critiques from queer activists and allies (Boellstorff 2011;Martucci 2010). But as I will show with the case of black donors, bioscientific or rights-based constitutional rationales for exclusionary donor criteria do not satisfy communities who demand redress.
In 1998, responding to the transmission of HIV in the blood supply, the blood services in South Africa established a "hierarchical model" of risk management. The model was based on statistical analysis of infection prevalence rates as they corresponded to demographics that blood services used in data collection. The analysis determined race, gender, and new/lapsed/repeat donation as the significant markers. Blood service officials classified repeat Indian and white donors of all genders (<0.005% risk of infection) as the safest and new and lapsed African donors of all genders (>1% risk of infection) as the highest risk (Jeffery 2012:221). This hierarchical model further stratified the risk cohorts into low (all Asian and white donors), intermediate (all black returning donors, all mixed race male new and lapsed, and mixed race female lapsed donors), and high (new female mixed race and all new and lapsed black donors) (Heyns et al. 2006). To better implement the risk management strategy, several blood services amalgamated into SANBS. In 1999 the government designated blood as a "national resource" taking responsibility for blood safety in a national policy whilst devolving daily operations to blood services (Jeffery 2012).
The strategy involved SANBS closing donor clinics in areas with high HIV prevalence. They prioritized blood products from low risk cohorts and triaged the limited use of blood products from intermediate and high risk cohorts, depending on stock levels (Heyns et al. 2006). The blood tests used at this time still had a month-long window period between the time of HIV transmission and detection. Blood service officials therefore used this strategy of demographic risk profiling to further ensure the safety of blood against this window period. 4 In a 2001 publicity ploy, the black president Thabo Mbeki staged a blood donation, but had not filled out the donor questionnaire and consent form, and for this reason SANBS did not use his blood. In 2004 the donor profiling policies and Mbeki's staged donation came to media attention, which caused fierce debate (King 2012;Peters 2004). 5 A dominant public narrative ensued: that the blood service threw away black people's blood, and that even the president's blood was not safe.
Blood service officials tried to dispel this narrative. They took issue with how the public mischaracterized the risk management as an outright ban based on race and/or sexuality. Their calculus included race, age, gender, and sexual practices, but also other risk categories to inform how they used blood components and donor deferrals. Presently, depending on supply levels, blood services still implement product triage policies based solely on number of donations given. Irrespective of race, new donors' plasma may be quarantined and only used after a second donation tests negative. Red blood cells and platelets do not last as long as plasma and may only be used from subsequent donations that test negative. Platelets are only made from donors who have donated 4 times in the preceding 24 months. Per SANBS's 1999 policy, cellular components from the highest risk categories of donors (i.e. first time donors) would be placed in quarantine and only used if no other supply was available. This supply would be incinerated if they expired.
Even among those who were aware of the complexities of the risk assessment procedures, the classification of black donors as high risk remained a concern. Akani, who now works in the labs which test for donor infections, recalled his experience as a black man working in the component processing department of the blood service during this time: 6 Our stats started picking up . . . HIV . . . in the township. One . . . was in Soweto [a black township]. . . . we had more HIV from that clinic . . . because of that the SANBS director decided not to use black people's blood. Even if you have donated 10 times. They will use your plasma but not your red cells.
. . . I was among those people that were discarding those red cells from black people. I couldn't believe that the company could go through such a wastage and just collect the blood and destroy it because it's from a black person. I remember I went to an extent checking the red cell[s] from the person who donated. How many times he has been donating? And I realized . . . he has been donating for plus/minus 10 times. But because he was black, his red cell was discarded. I was not happy, I was not impressed. But with an understanding also of the challenge that we were facing . . ..If the company didn't have the equipment that can at least cut short the window period of HIV, . . . it was going to put other people's lives at risk . . . and now I was in between that confusion.
Akani, like other black staff who handled the blood components of donors, understood the seriousness of rising disease rates, but it was not until the policy was changed that the "confusion" was resolved.
Accounts that components were "discarded" because a donor is black, are procedurally inaccurate, but there is an affective accuracy in their sentiments of racial injustice.
SANBS emphasized that the donor criteria were not enforced with the intent of discrimination, but rather because government-imposed financial constraints, epidemiological evidence, and the safety of transfusion patients necessitated the risk management strategy. SANBS pointed out that the government stipulation to eliminate racial profiling was more a response to the public upheaval rather than prior ignorance of the policies. The Department of Health had approved the risk management strategy which reflected their demands to balance cost efficiency and safety standards, justifying racial risk profiling as a complement for tests. SANBS made a point of distinguishing the intentionality of illwilled discrimination from scientific "objectivity." The risk management strategy proved effective in reducing prevalence of HIV-1 in blood donors by 50% within a year (Heyns et al. 2006:522). The crisis of HIV transmission through transfusion and the priority that patients (>80% of whom were black) received the safest blood possible within testing cost constraints superseded resultant exclusions of donors (Heyns et al. 2006). Exclusion, but not hate-based prejudice, as many blood service representatives emphatically differentiate, is framed as a necessary evil. In this view, profiling must remain in place until there are other means to offset the risks of high-prevalence populations, or new evidence proves their risks as a population are acceptably lower. Anton Heyns, the CEO of SANBS at this time, framed this donor exclusion-patient safety tradeoff in terms of constitutional rights: I am not a racist and this is not racist organization . . . If we . . . [were], we would say, you are a group with bad blood and we do not want your blood . . . That would be unacceptable and immoral. I believe that the patient has rights, and . . . that the patient has a right to quality blood . . . this may in fact be a constitutional right . . . .What we have done is utilising [sic] race as a tool to guarantee the patient the constitutional right to safe blood. (Hollemans 2004) In Heyns' estimation, the rights of the patients exceeded the rights of donors. SANBS officials reasoned that black donors still contributed to the blood supply. The constitutionality of racialized risk profiling rested on its utility for ensuring access to the health care service of the safest possible blood products, which, as officials argued, were still equally and proportionally distributed across races (Heyns et al. 2006).
Unfortunately, this emphasis on patients' rights superseding donors, foreclosed solidarity between the concerns of patient recipients and donors' excluded/limited in donation. The argument belied the rights of another set of patients, those already suffering from or more vulnerable to infections. Racial differences in infection risk are not intrinsically biological, they are produced through social, economic, and political conditions. Blood service officials saw risk profiling as the most epidemiologically and financially prudent measure to respond to the transfusion transmitted infections given institutional constraints. Echoing the case of Israeli blood service restrictions on Ethiopian migrants donating blood, justifications of greater infection risk as an objective biomedical fact obfuscated the broader state neglect of racialized migrants that contributed to such high prevalence (Seeman 2009). Similarly, SANBS's defense of the risk management strategy and its racialized exclusions did not convey a concomitant emphasis on confronting the systems of inequality that disproportionately distributed disease prevalence along racial lines. In this case, state failures to provide adequate HIV treatment in South Africa violated a constitutional right to health care and led to other forms of patient-centered bio-civic mobilization such as the Treatment Action Campaign (TAC) (Robins 2004). Public pressure from a critical mass created political will for state institutions to change tack.
SANBS executives had not anticipated how little the justifications for racial profiling could be distinguished from apartheid-era segregationist blood service policies in the public's view. Moreover, the justification of consistency with international standards does not account for the ways discriminatory principles function on a global scale (Beaudevin and Pordié 2016;Benton et al. 2017;Fassin 2007).
As other cases have shown, a claim to medical neutrality, especially in contentious contexts, is a political gambit (Hamdy and Bayoumi 2016). "Evidence-based" risks and regulatory mandates seem to arrogate risk models above politics. This positioning functions as a "moral prophylactic" against having to do the harder work of overturning the very prejudices that shape epidemiological thinking and population-based risk modeling -and that sometimes underlie inclusion efforts in donor recruitment (Benjamin 2016:150).
The critique of neutrality and the characterization of medico-scientific imperatives as apolitical is not based in an anti-science stance or in ignorance of epidemiological and physiological realities of transfusion transmitted infections. This would be a grave mistake in South Africa, a place where AIDS denialism -a stance Mbeki was responsible for perpetuating -has had catastrophic effects (Mbali 2004). Instead, it is an invitation to wrestle with the ramifications of the risk management system.
Although blood service practitioners want to overturn donor exclusion policies that racialize and/ or govern sexual mores in order to widen bio-civic participation in blood donation, they also insist safety must not be compromised. I aim to shift the resolution of past controversies and the implementation of present initiatives away from a framing of the biomedical imperatives as oppositional or above negotiations of social justice and solidarity, toward a frame in which they are crucial to postapartheid national transformation efforts. I urge us to train our focus on the social and political stakes of blood service work and wrestle with these stakes as inherent to that work. I was heartened to engage with many blood service staff who are already oriented toward these considerations. They are hopeful while acknowledging that there is still more work that needs to be done.

"We need black blood"
Four years after the controversy around racial profiling, a national newspaper ran a story with the headline "We need black blood" (Tshabalala 2008): More than a quarter of a million South Africans are regular blood donors, but only 10% of them are black . . . .
[Yet] black people constitute 80% of those needing transfusions.
South African National Blood Services (SANBS) has found that black people still believe their blood is discarded because of the high HIV infection rate among black people.
Black donors are desperately needed not only to boost the supply of blood, but because genetic differences mean that better matches between donors and recipients tend to occur within ethnic groups.
How did SANBS transition from the classification of black donors as high risk to being "desperately needed"? Public pressure and the Department of Health stipulation to eliminate race as a risk profiling criterion spurred South Africa to become the first country in the world where blood services used individual nucleic acid testing (ID-NAT) for blood donations in 2005. This molecular test would complement less sensitive virology-based tests and screen individual donations rather than batches. Previously, the Department of Health had approved the risk management policy that included race as a factor rather than subsidize ID-NAT because the latter was too expensive. The Department of Health funded the change as part of their response to controversy. SANBS transformed their donor recruitment agenda toward a more equitable, inclusive policy related to donors, and attributed this shift to the adoption of the latest available technology. As Bekker and Wood put it: Differentiation on the basis of race may constitute illegal discrimination in terms of the South African constitution unless it can be established that such discrimination is fair. Establishing fairness requires a balancing of the nature and extent of right infringed in an open democratic society based on human dignity, equality, and freedom and the purpose of the limitation. In this case, alternative nondiscriminatory means to achieve a safe blood supply (nucleic acid technology) were available and have been incorporated into the 2005 policy albeit at increased cost. (Bekker and Wood 2006:557-558) In this interpretation, the transition to ID-NAT permitted SANBS to be more inclusive. Both the initial discrimination and its elimination could be justified by technical and economic reasons which citizens should understand as fair. Since SANBS implemented ID-NAT they have reduced transfusiontransmissible risk from 24 to 13 per million transfusions while black donors have steadily increased (Vermeulen et al. 2019). But this justification sidesteps a reflection on how implicit racial biases in medicine and histories of scientific racism might be recognized and dealt with, especially had no "alternative nondiscriminatory means" been available.
Although public pressure was successful to some degree in questioning how the language of "risk" and "infectious disease" had figured into a racial hierarchy, the revaluation of black donors as a focus of donor recruitment signals the enduring question of who can and should donate. Indicative of "individualizing and collectivizing" dimensions of biological citizenship, Simpson points to a tension between the "universalizing intentions of the international blood donation movement" and particularities of national and ethno-religious affinity and exclusion (Simpson 2009:112). A similar tension arises in the South African case between universal claims to color-blindness and the continued racialization of donation. Criticism of the racial profiling policy produced the theme that "Despite skin color, the color of everyone's blood is red (King 2012:31-32). A recent blood service campaign proclaims "Blood Has No Colour" on billboards and branded gifts. In contrast, the newspaper article by Tshabalala cites a linkage between race and blood type as the major reason for needing more black donors. By this logic, race does make a biological difference in the blood. In blood typing for transfusion compatibility, race again figures as a proxy for biological difference, albeit framed as a desirable difference and a means to greater inclusion rather than a reason for exclusion.
Although racial difference is a socially contingent categorization, such practices lend credence to the essentialization of race. Despite (pre-colonial) migration and linguistic and cultural exchange in southern Africa, 19th to mid-20th century anthropologists and linguists imposed static, primordialist (and hierarchical) distinctions between languages and cultures. These led to the formation of racial and ethnic population groups that are still used as units of analysis for contemporary biological and medical research (Braun and Hammonds 2008). Statistical methods afford legitimacy to these populations as stable correspondences between racial and ethnic identities, linguistic features, cultural beliefs, and/or biological/genetic markers, rather than socio-politically negotiated frameworks. Although the late 2000's "postapartheid genome" initiative intended to deconstruct colonial and apartheid racial categories through genetic ancestry testing, racio-ethnic categories were still used to index biogenetic materiality even as they pointed to greater heterogeneity than apartheid classifications (Foster 2016). The project "simultaneously marshaled difference and sameness to craft new terms of belonging for the nation-state predicated upon a shared African [genetic ancestry]" (Foster 2016(Foster :1022. In privileging the universalist claims of "We are all African" or that "Blood Has No Colour" to reconcile differences, Foster cautions that this approach furthers nonracialism, a form of unification that "seeks to end racial references and enable individual rights but does not address the structures and effects of racism" (Foster 2016(Foster :1024. This larger context of bioscientific racial projects reverberates in the linkages between race and blood group in South Africa (Walters 2018). Such linkages were reinforced in 1962, which brought the segregation of donor clinics and racial labeling of blood. Although there was no official prohibition of cross-racial transfusion, the labeling impacted how the blood was distributed. 7 The apartheid regime justified procedures as consistent with international practice, as reflected in US racial segregation of the blood supply, for example (Guglielmo 2010;Lederer 2008). There were efforts to recruit black donors, even prior to the advent of HIV. However, there was far more recruitment and representation of white donors whom officials viewed as easier to recruit, as likely to donate altruistically rather than through financial motives, and supposedly having less cultural objections. 8 As a result, a blood supply meeting its "own [racial] group needs" was a perennial concern in South African blood services (Jeffery 2012:205).
More recent hematological/transfusion medicine and blood service research continues to correlate blood type frequency with racialized populations; e.g. black Africans are more likely to have type B, whites are more likely to have type A or mixed race donors have higher rates of rare blood types (Govender et al. 2021;Muthivhi et al. 2015). Consequently, when talking about typing, blood service officials often refer to "diversity" and "sustainability," in which the blood of black donors is desirable. During my fieldwork, the donor pool was 41% black but roughly 80% of the population at large, 43% white, an overrepresentation compared to 8% of the population at large, with 8% Asian, 6% coloured, and 2% unknown (SANBS 2017; SANBS 2020). Blood service officials frame diverse donor representation as a civic virtue to broaden the reach to underrepresented communities and better serve the purpose of patient compatibility needs, rather than a return to apartheid-era fears that too much white donors' blood is being used to help black patients. Patients receive blood approximately proportionate to their representation in the population at large, which was also the case in 1999. However, the implicit responsibility of "own [racial] group needs" as an immunogenetic matter still positions the distributive calculus as unsustainably over-reliant on white donors.
The political economy of blood plays an important role. The financial calculus around infections testing was a major factor in both adopting and eliminating racial profiling. While the triage protocols of limited use were not the same as "throwing away" blood, this narrative reflected not only a sense of biomedical racial injustice but also one of distributional injustice and wastage. Although "own group needs" do not rest on apartheid regime motivations and instead are based on "objective" statistical measures that attribute value to black donors, the proportionality of racial groups still become an index for the distributional fairness and typological sustainability of blood stocks. Moreover, as Williams points out in the case of bone marrow stem cell transplant matching based on statistical associations between immunogenetic tissue types and race, donation becomes an "ethico-racial imperative," the duty to serve one's racialized community (Williams 2021).
Epstein points to stem cell transplantation as an example that some scholars argue demonstrates the practical "medical utility" of profiling based on social identity (2007). A similar logic lies behind both SANBS justifications for risk profiling and subsequent active recruitment of black donors based on typological frequency, following the reasoning that "if the knowledge that a person belongs to a racial group, however defined, helps us differentiate that person's diagnosis or treatment [or degree of biovalue in donation]-then why not use that information and bracket the question of why it works (let alone the question of what race 'really is')?" (Epstein 2007:219). However, Epstein warns that racial risk profiling or defining race as a fixed biogenetic inherence may more often produce harms that outweigh pragmatic benefits. Locating racial differences in genetic code obfuscates the social causes of health disparities and the historical forces that led to racial categories of analysis in population genetics. Cases of racially and ethnically focused donor recruitment both in South Africa and elsewhere already bind that utility to ethical dimensions of racial justice, whether duty, rights, fairness, diversity and/or inclusion (Avera 2009). Even when biomedical practitioners privilege or put stock in a biological basis for racial difference for inclusionary and practical purposes, this may still reproduce both biological and social inequalities.
Racial identities and histories do impact biological processes and health outcomes. However, the social dynamics are not perfectly congruent to the genetic dynamics. My participants' distinctions between biological typologies and social identifications, notions of a racially diverse society, or outlines of racial hierarchy are often commingled or contradictory, similar to ethnographic research of biomedical discussions that purport linkages between race, blood, and genetics more broadly (e.g. Foster 2016;Fullwiley 2007;Wailoo et al. 2012) The rhetoric of race, ethnicity, nationality, and ancestry offer a lexicon for making sense of biological information, but it does so imprecisely. As encapsulated in the instance of blood, the character of socio-genetic linkages shift in relation to which group distinctions matter for the people generating the research that biomedical practitioners draw upon. Erasmus calls these determinations "throwing the genes" which she compares to a sangoma (traditional healer) throwing the bones from which they make fateful interpretations (Erasmus 2013). The variation in these interpretations demonstrate how the relationship between race and biology continues to be nebulous and yet simultaneously leveraged to determine who qualifies as a suitable biological citizen and whose rights bio-civic acts should prioritize. Returning to indexicality in which a linguistic feature may shift its meaning(s) in relation to the circumstances of its usage, likewise, which particular racial or ethnic groups point to compatibilities of blood types or epidemiological risk is context-dependent. In understanding how socio-political contingencies impact which racial categories index certain typological frequencies or health outcomes, it is crucial to resist investing these associations with the confirmatory power of race's biological fixity. Donor recruitment efforts should rather attend to the contextual shifts of interconnected social identities, biological processes, and health inequalities that are not reducible to a static racial group index for disease risk or bio-genetic typology. I now turn to an example of how specific aspects of language are also instrumental to tracking the shifting associations between blood and race, and the degree of inclusion they index.

From blood wars to brotherhood
Since 2016, gaz'lam is a phrase that has circulated throughout SANBS outreach materials, as the title of its magazine for donors, and as a hashtag on SANBS social media posts. It is an abbreviated form of the Nguni (e.g., isiZulu/isiXhosa) words igazi [blood] + lami [my] which translates into "my blood." Gaz'lam indexes a host of other things. It is commonly used in the dialects of youth in black and coloured urban townships, which are characterized by a mix of languages. 9 The use of a term like gaz'lam and these racialized dialects more generally, circulate around a complex identification with township life, and with the ways that they are hard places to live, punctuated by poverty and violent crime (Jones 2013).
Simultaneously, this use of gaz'lam, especially as part of a larger repertoire of code-switching, is also meant to register a kind of coolness, and/or a sign of belonging in a township space (Slabbert and Finlayson 2000). Crucially, to call someone gaz'lam also signifies a relationship withstanding life or death circumstances in township communities. It may refer to a kinship connection that draws upon blood as a sign of biogenetic family but can and does exceed that significance. More broadly, blood has a wide range of semiotic uses in the township context, invoking a greeting, gang relations, and/or associations with violence (Aycard 2010;Bembe and Beukes 2007).
The indexicality of gaz'lam is part of why SANBS has utilized it in their public outreach efforts. Invoking the phrase calls forth these meanings that shift based on the contextual specificities of a situation in which it is used. SANBS marketing is not simply using gaz'lam to draw on its varied meanings; in the process of using it, they are also attempting to transform what the phrase indexes, particularly around whose blood and whose relationality the phrase implicates. To analyze this attempt at semiotic transformation gives us a granular account of a larger strategy of post-apartheid racial reconciliation that stakeholders wage both within SANBS and in South African society. The usage of gaz'lam through various orders of indexicality, represents the sociopolitical tensions at play in determining the efficacy of inclusivity initiatives geared toward black communities (Blommaert 2007;Silverstein 2003).
At a university in Johannesburg, I arranged with the instructor of their critical diversity literacies seminar to serve as a guest facilitator. I asked the class -a group who represented a subset of the blood service's ideal donor pool demographics: young, educated, and predominately black and colouredwhat they thought about the SANBS gaz'lam campaign and outreach materials, given the context that the majority of transfusion recipients are black patients in public hospitals. Their response was critical: Dineo: . . . Gaz'lam emerges from a township setting . . . .So the thing is especially when the SANBS doesn't even particularly . . . have a stamp in the township . . . I don't think the township people resonate with the SANBS . . . they don't know it, it's not something that features. But yet [gaz'lam] is generated and housed in the township. Because it's not like it's just something black, it's particular. It's black, but also black township. . .. To go to those depths to . . . pull out your marketing, when it is those people are [seen as] the last of the barrel to benefit from the service . . . . Shannon: . . . but let's just work with the reality of it . . . .Chances are you have a family member who may be shot . . . .We rush them to Bara [the public hospital in Soweto] and then there's no blood . . . .Something really dumb that could have saved their lives, we don't have, and yet if one of you in the community came forward this person would be alive today . . . .The campaigning just has to be specific to where we are.
For these students, gaz'lam is a phrase embedded in the township context, not easily dislodged from its particularity and sense of place. They read the blood service's use of gaz'lam as sanitizing its associations with the harsh realities of township life. The donor magazine states, "Gaz'lam/The blood that binds us together" and defines "Gaz'lam/noun-/gaz-lam/Zulu word for 'blood relatives or brotherhood.'" In this instantiation, SANBS's attempt to change the indexicality of gaz'lam is hardly complete. To Khwezi, the blood service's use of gaz'lam is linguistic minstrelsy. Rather than emerging from an authentic place, Khwezi equates this usage to a kind of racial impersonation, cosmetically indexing the black township, but not affectively or effectively connecting with it.
The students' perspective reflects their interest in analyzing the work of diversity and inclusion in South Africa. In contrast, a blood service HR coordinator told me that she thought the SANBS gaz'lam campaign was "really smart." As a young professional black woman who was working on employment equity, she felt that the use of a phrase with so much meaning in the black community was recognition rather than a "pandering" marketing ploy. The difference in their interpretations likely had something to do with a varying degree of enthusiasm about corporate branding. Workplace settings are common for donor drives which serve as a visible form of corporate social responsibility. People conceptualize the acts of blood service work and donation around their different opinions on community engagement, corporatization, and civic responsibility.
Ophelia, a black female lab supervisor at SANBS, did not know about the racial profiling at the time but when she first heard about it she was "shocked." She felt bad, "guilty for all those people donating for nothing." She said, "I was disappointed because I thought we were all 'in the same boat' doing the same thing. And here where the decisions are being made, they will say 'no, you can come donate.' So it was under false pretense . . . in getting people to donate without using their blood." However, she believed that "things have changed" since then. She expressed her pride in the blood service's work to reach black communities: Recently they started campaigning which is very nice to see SANBS out there. You're like yeah, that's my company. These Gaz'lam adverts, there's billboards . . . Remember in the olden days black people didn't know so much about blood donation. But they're more exposed now. There's even more clinics going out to the townships . . . SANBS has mobile clinics everywhere. It's a good feeling because you know that SANBS is now on the right path, and there's no prejudice or racism. We're all here to do the same thing, which is saving lives, at the end of the day.
The blood service's endeavor to semiotically turn gaz'lam, mirrors the pivot that the blood service aims for in black communities more generally. Ophelia identified these efforts as a form of reconciliation in which the blood service was "making things right." SANBS want to change their reputation as an organization from equating blackness with risk to treating blackness as valuable. Thus, they take gaz'lam from the township, a code that marked its speakers as marginalized but connected bodies surviving in the face of state violence, and transplant it into marketing materials that align it with blood donation, as an act of "solidarity with fellow humans," civic virtue, and salvation for the nation. This semiotic turn is an attempt to restore social equilibrium between black township residents and public-dependent institutions like the blood services, whose work relies on donors. Yet gaz'lam still holds onto its other indexical meanings while the marketing materials seek to shift township slang into an object of corporate and civic legibility. The transformative semiotic potentiality of a phrase like gaz'lam highlights the effective aspects of this kind of strategy and the challenges that this presents.

Soweto: A metonym for black life and township in transition
Contrary to the students' assumptions, SANBS is not disconnected from the township. One of the SANBS blood banks is based at Baragwanth "Bara" Hospital in Soweto. Many SANBS staff live in townships where I socialized with them, although they often work outside of those spaces. The contestation over the meaning of gaz'lam between the students and the blood services reflects the more widespread usage of township dialects in nationwide advertisements (Aycard 2010). Such linguistic adaptations illustrate the increasing cultural and commercial value of anything indexing blackness and the township. Though I venture that this valuation sometimes belies more complicated narratives about the township and black identity in South African national consciousness and in the blood services.
Soweto, a township in the Johannesburg metro, was a nationally important site in the antiapartheid struggle, where police brutalized and shot student protestors in 1976. Drawing on Soweto's significance, SANBS held a blood drive near Mandela's former home, for one of their biggest campaigns of 2018: #NewBlood. The campaign featured Thuma Mina, another resonant Nguni phrase, meaning Thuma [Send] Mina [me]; "Send me," referencing a Hugh Masekela song and a biblical passage in Isiah 6:8. In President Cyril Ramaphosa's election that year, he invoked Thuma Mina, rallying South Africans to answer the call of civic duty and service to the nation. At the time, Ramaphosa's presidency represented a turning point from the dismal state capture of the Zuma years. Building on this renewed faith in government, the blood service folded their donor recruitment into this incitement to civic duty.
Soweto's status as a national emblem for the promise and the peril of black township life imbues blood service activities there with the same significance. So too does blood donation signify discrimination and validation in black communities. Marjorie, a Sowetan donor and former student protestor, for example, contrasted her enthusiastic donation and her son's distrust of the blood service.
You know what discouraged him? There was this incident when. . .I don't know whether it was because people were unhappy with the transition of the government or whatever, because blood banks were mostly run by the whites. They used to throw people's blood away. So that's still in black minds. They're still very reluctant. They say, I'm going to give my blood and they are going to throw it away. That's what my son is telling me. I said but nowadays it's no longer like that . . . .Let bygones be bygones, like Mandela said. Leave them, it's fine. Let's give blood to others. There are those who need.
Her account of familial tensions over blood donation and the "throwing blood away" narrative not only represents the contested status that the blood service holds in the imaginary of black communities in South Africa. She also articulates the differing degrees of acceptance or rebuke of the democratic transition, reconciliation process, and transformation initiatives in the post-apartheid era. As in blood donation and national politics, the eradication of prejudice, oppression, and marginalization is partial, ongoing, and tenuous.

Conclusion
Blood service employees and donors demonstrate multilayered understandings of the medical and social stakes of their respective endeavors. They have a keen awareness of epidemiological imperatives and organizational mandates, acknowledge what SANBS is trying to do in the aftermath of racialized exclusion, and express a desire for a greater realization of racial justice. Their stories compel a continued reckoning with both the pitfalls and potentials of blood as a national resource and donation as a form of biological citizenship in South Africa.
I have shown how various stakeholders draw lines of biological citizenship through material, relational, and semiotic processes. SANBS campaigns to increase black donors' participation evoke Soweto as part of a nationalist ethos that commemorates black liberation struggles and uses linguistic features to evoke township authenticity and the healing offered through the provision of the "national resource" of blood as forms of civic connection in a democratic South Africa. Although the "Blood Has No Colour" campaign may signify a color-blind universal enfranchisement that counters legacies of racial profiling, participation in blood donation is at other times informed by race as a proxy for blood type and proportional donor racial representation as more materially sustainable. Who someone indexes as "my"/"lami" in "my blood"/gaz'lam is based on the contextual assessment of how far they extend the relational reach of gaz'lam. Is it one's biogenetic family, the township community, "own group needs," or the nation writ large? Gaz'lam's indexicality in this negotiation of the singular self and the plural collective continues to shift. Likewise, the me/mina in Thuma Mina may index not only President Ramaphosa, or an individual, personal "me" who asks to be sent forth to serve their country, but a collective South African citizenry. Both in the semiotics of gaz'lam and Thuma Mina, and in the physicality of blood donation, the blood donor as an "embodied citizen" challenges the boundary of selfhood, making the private bounded body public (Valentine 2005). The body is particularized at the same time that it is anonymized and/or universalized.
The case of blood donation in/ex-clusion in South Africa helps us to see biological citizenship in motion; it does not cut across a single axis of a particular condition (albeit HIV exceptionalism is implicitly involved), nor does it only function on the basis of race as a biogenetic, epidemiological, or even anthropological variable, it also operates through linguistic shifts in meaning. The bio-civic dynamism of blood reveals the broader challenges of the post-apartheid state project of justice and reconciliation in its aims for recognition -and redress, to varying degrees -of the lasting legacies of inequality and segregation while also finding ways to unite citizens across social cleavages.

Notes
1. The Western Province Blood Transfusion Service (WPBTS) was renamed WCBS in 2019. 2. Donors for both blood services indicate their ethnic group on questionnaires: African, White, Coloured (a local mixed heritage/race category), and/or Indian/Asian. These categories and their capitalization roughly corresponded to the apartheid race classifications which are still used as post-apartheid South African census categories, and in blood service annual reports as: African/Black, White, Indian/Asian, Coloured, and Unknown. The question of race and ethnicity in South Africa is more complex than these categories. Many have advocated for a more encompassing view of black identity, and a disavowal of coloured identity altogether as an apartheid imposition. Following the work of Erasmus (2017), (Walters (2018:xv andKhan 2021:47), I generally use the (de)capitalization or lowercase of the original quotations and context, and as consistent with the de-capitalization of "apartheid." South African blood services discussions predominately use the racial terms "white," "coloured," and "black." However, Erasmus clarifies that although, "'Black,' capitalized, refers to the apartheid racial category, in certain contexts it also refers to the global political identification on the part of individuals and collectives who are constructed and classified as black" (2017:147). Many authors use "Black" capitalized in the South African context in alignment with Black feminism and the Black Consciousness Movement while mindful of its complications (Marco 2021:65). 3. A deferral is a waiting period of disqualification from donation due to health concerns. At SANBS deferral for sexually active MSM was the same as other "high risk" exposures (e.g. donors with multiple-partners), i.e. initially 12 months and later reduced to 6. 4. For historical accounts of earlier blood service donor policies organized on the basis of race, see Jeffery (2012) and Kesson (2009). 5. Although both blood services in South Africa used MSM profiling, only SANBS used race in their risk management. The difference was based on the epidemiological and racial dynamics of the Western Cape Province which has a smaller black population than other regions, and lower rates of HIV. Both blood services eliminated MSM deferrals in 2014 after studies determined that this was no longer an indicator of high risk (Brailsford et al. 2015;Ngcobo 2016). 6. Participant names changed for confidentiality. 7. There were cases of white donors' blood being transfused to black recipients, but not across other racial lines (Jeffery 2012). 8. These supposed challenges in recruiting black African donors (not only in South Africa but globally) locate the problems in culturally essentialist logics of superstition or ancestral rites rather than in racial biases in biomedicine that have historically privileged white, Eurocentric perspectives. This has also been the case in discussions of cultural barriers in South African bone marrow donor recruitment (Avera 2009). 9. There are shifting definitions of these dialects (Mesthrie 2008). Gaz'lam is a regular part of their lexicons (Bembe and Beukes 2007).