The knowledge politics of genome editing in Africa

,

Over the past decade, the advent of CRISPR and other tools of genome editing 1 has been heralded by some as a "revolutionary" technology.This particular narrative presents the technology as cheap, easy to use, and able to circumvent supposedly restrictive regulatory regimes (Komen et al., 2020;Qaim, 2020, p. 137;Abdallah et al., 2022).It parrots the laudatory promises that accompanied previous generations of heavily hyped plant breeding technologies, including genetic modification (GM) (Glover, 2010;Rock et al., 2023). 2 However, such historical linksand lessons that might be learned from them-remain largely overlooked within accounts celebrating genome editing's transformative potential for plant breeding.Instead, commentators in both the academic literature and popular presses tend to repeat optimistic headlines about genome editing's ability to "feed the world," echoing previous attempts to elevate a new breeding technology as a silver bullet solution to increase global agricultural productivity (De Amstalden, 2023;Doudna, 2023).These narratives and their contestations bely a knowledge politics surrounding genome editing that this article aims to explore.
Africa features prominently within these promissory narratives.Anticipatory accounts have focused on the potential agricultural applications for genome editing, including the targeting of African staple crops and the uptake of genome editing within public breeding programs (Abugu, 2021;Karavolias et al., 2021a).The latter expectation-that genome editing, unlike GM, will be widely available to smaller labs and scientists in the Global South due to supposed lower costs-animates much of this literature.Others have positioned the continent as ideally suited to benefit from these technologies: "the African continent is uniquely poised to benefit from advances in gene editing technology" because it was "largely overlooked by the Green Revolution that delivered higher yielding grain crop varieties to Asia and Central/South America" (Karavolias et al., 2021b, para. 6, emphasis added).This portrayal of Africa having missed out on the original Green Revolution, thus requiring a surge in innovation and investment into new breeding technologies, has been the defining feature of the continent's agricultural development over the past 20 years (Moseley et al., 2015;Bergius and Buseth, 2019).
Claims of Africa being "uniquely poised to benefit" from this latest breeding technology are worth scrutinizing.This article shines a light on the opinions, realities, and goals of those working with genome editing techniques on the African continent to better understand the knowledge politics undergirding this new wave of technological optimism.The perspectives of scientists who have embraced genome editing, those of donors who are funding such work, and of biotechnology policy experts working to shape public and political perceptions of the technology, matter in assessing (if and) how the technology is being both accessed and deployed (Middelveld and Macnaghten, 2021).Thus, in this article we ask: how do stakeholders view the genome editing of crops and its agricultural applications in sub-Saharan Africa?
In what follows, we present data from interviews with 13 scientists, donors, and biotechnology policy experts working in the fields of agricultural development, plant breeding, and biotechnology in sub-Saharan Africa.These interviews, undertaken on Zoom during the height of the COVID-19 pandemic, shed light on how knowledge politics might shape the trajectory of a potentially powerful tool in an uncertain time.Overall, while most of our interlocutors were enthused about the technological potential genome editing might offer, many also spoke to structural challenges to accessing the technology and housing it within African research institutions.Following these interlocutors, this article approaches genome editing not as an exceptional technology but rather as another tool, entangled within social, political, economic, and historical webs.
The article proceeds as follows.First, we explore the literature surrounding the knowledge politics of agricultural research to situate the nascent but growing social science scholarship on genome editing and its agricultural applications.Immediately following, we discuss the methods that drive our analysis.Then, we present an overview of the nascent experimental pipeline of genome-edited crops and assess 3 themes that arose from our interviews: the question of regulation, the importance of infrastructure, and the perils of simplistic narratives.We conclude with thoughts on the challenges facing those wishing to access and deploy genome editing.

The knowledge politics of genome editing
The shaping of new biotechnologies has been analyzed through the lenses of expectations (Hedgecoe and Martin, 2003;Tutton, 2011) and imaginaries (Jasanoff, 2004;Bain et al., 2020)-the identification of which help us move beyond simple interest-based approaches in understanding emergence.In this article, we focus our attention on the narratives adopted by different constituencies in the ways they describe expectations and imaginaries associated with gene editing.Narratives have been used in the analysis of politics and policy for decades (Roe, 1994) and have been described as "simple stories with beginnings defining the problem, middles elaborating its consequences and ends outlining the solutions" (Leach et al., 2010, p. 45).In the case of seeds in Africa, for example, an example narrative might be "growing food deficits require massive boosts to agricultural productivity and only GM crops will provide the answer" (Leach et al., 2010).By mapping some of the extant narratives more specifically related to genome editing and questioning their assumptions based on interview data, we interrogate the politics of knowledge (Leach et al., 2010, p. 75) surrounding the emergence of genome editing in Africa.
Africa has long been important in imaginaries around agricultural biotechnology, both as a location for technological advancement and as a narrative device for conjuring what the technology can (or cannot) achieve.Academics, officials, and biotech firms alike argued that GM crops would "feed the world" (Borlaug, 2000) and enable Africa to "leapfrog into more sustainable methods of agricultural production" (Juma, 2015, p. 64).These techno-optimist assessments are rooted in a narrative that understands low productivity rates in Africa as being a function of outmoded technology, tools, and agricultural knowledge (Logan, 2020).This narrative has a long historical arc: it emerged from and was codified during colonial rule and persisted within postcolonial regimes, portraying Africa and African farmers as apart from the world, with little to offer, and in need of "improvement" (Rodney, 1982;Yapa, 1993;Canfield, 2022).
Narratives of technologies and geographies are, in one regard, a matter of knowledge politics, that is, how "individuals and groups selectively generate and/or use knowledge to establish, maintain or enhance their vested interests" (Andersson and Sumberg, 2017, p. 6).Different from "technicist conception[s] of technology," wherein "technologies are thought to have fixed functional characteristics that produce predictable effects" (Glover et al., 2017, p. 15), critical analyses of knowledge politics are rooted in the assumption that ideas, aspirations, and biases shape decision-making around technologies and research programs (Andersson and Sumberg, 2017, p. 7).In other words, technology is understood not as a static package delivered to end users but rather as something that is molded by socioeconomic conditions, prevailing ideologies, geopolitics, and imaginaries (Winner, 1980;Akullo et al., 2018;Middelveld and Macnaghten, 2021).
The politics of knowledge-especially that related to technology and development-has been of great focus within science and technology studies, critical agrarian studies, anthropology, and beyond.For example, Middelveld and Macnaghten (2021) show how different sociotechnical imaginaries of Dutch scientists are shaping conversations, research, and governance around the genome editing of livestock in the Netherlands.Andersson and Sumberg (2017) draw attention to how larger political-economic forces shape agricultural research programs.They argue, for example, that the neoliberal turn of the 1980s profoundly affected agronomy programs, in that it divorced agronomy from the state, and moved agronomic knowledge, projects, and outputs away from the public domain.Writing on what he terms the "ideology of innovation," Canfield suggests that the forces driving sustainable intensification on the continent are deeply embedded in the neoliberal project of "reframe[ing] political problems as market opportunities" which are in turn used to "catalyze technological development" and "justify the dispossession of peasant's and Indigenous people's knowledge" (2022, p. 2, 4).Others have similarly highlighted the ways in which narratives surrounding tools of development mask historical and structural inequities and thus serve to decontextualize and depoliticize complex structural challenges such as poverty (Benton, 2015;Pierre, 2020).
Questions around knowledge politics animate the social scientific scholarship on genome editing.Some critical social scientists have argued that promoters "scientize" narratives around genome editing-for instance, emphasizing (supposed) efficiency-which allows them to circumvent discussing stickier social and regulatory issues (Helliwell et al., 2019, p. 781;Mueller and Flachs, 2022).Others have zeroed in on the key battle ground of regulation as countries debate whether to regulate genome-edited crops through process-or product-based approaches (Ely et al., 2022).Kuzma (2018, p. 81) argues that some stakeholders "[play] the name game," using terms that decouple the technology from GM, in hopes of avoiding similar regulatory regimes, which they contend served to stifle the uptake of GM crops.Others have called for regulators to move beyond thresholds of what is "safe enough," and toward regulations that foreground transparent public debate and considerations of biodiversity, farmers' rights, and other socioeconomic concerns (Kjeldaas et al., 2022).
Social scientific critiques of genome editing have also problematized the narrative of "democratization"-the expectation that genome editing will be more accessible than its technological predecessors (Ricroch, 2019).Writing on the large pool of genome editing start-ups, Clapp and Ruder (2020) argue that de-centralizing power in the field is possible, but not guaranteed, as the likelihood of industry consolidation always looms.The initial biotechnology rush of the 1980s, after all, was marked by several small start-ups and firms, which were gradually consolidated into larger firms (Boyd, 2013).Moreover, through interviews with genome editing experts and scientists, Montenegro de Wit points to how, despite narratives of democratic use, "sharp lay/expert boundaries persist, and overlapping accesses to knowledge networks, funding, and infrastructure mediate the difference between access to a tool and the capacity to use it" (2020, p. 24).
Other critical assessments highlight how innovation "systems"-financial and material resources, such as the funding and lab equipment, skills, infrastructures, and policies-shape the accessibility and development of agricultural technologies (Hall, 2005).Those concerned with plant breeding and agricultural development on the African continent expose the degree to which African public research organizations are broadly underfunded and reliant on donor support (Roseboom and Flaharety, 2016;Hall and Dorai, 2020).This is especially relative to institutes within the Consultative Group for International Agricultural Research (CGIAR), an international agricultural research system that focuses on agricultural development in the Global South.The historical divergence of these 2 systems-national public research organizations versus international research centers-is key for understanding the knowledge politics of plant breeding on the African continent: while both systems grew out of colonial scientific networks, donors began to view the CGIAR as a way to centralize and direct research priorities as African nations gained independence in the 1960s and 1970s, playing a central role in the "post-colonial transition to a new national order" (Byerlee and Lyman, 2020, p. 14).As a result, African public research organizations remain largely underfunded and undersupported in terms of infrastructure.One consequence of this structural inequity has been the growth of public-private partnerships (PPPs) between African research organizations and private firms.While some consider PPPs to be a key vehicle for bridging technical and funding gaps between the public and private sector, and allowing public sector researchers to access new technologies (Spielman et al., 2010;Spielman and Zambrano, 2013), the empirical record has shown that PPPs are often marked by an imbalance of power between collaborators (Schurman, 2016) and have struggled to successfully develop and commercialize GM crops for African farmers (Schnurr, 2019).
Finally, social scientific assessments of genome editing have challenged simplistic portrayals of the technology's efficiency and ability to meet the diverse needs of smallholder farmers.Shah et al. (2021) showcase how scientific debates regarding genome editing reveal disagreement within the scientific community over the technology's preciseness, and argue that these important disagreements are obfuscated by actors circulating narratives about genome editing's supposed superiority and accessibility.A recent review by Schnurr et al. (2022) underlines the complexity of genome editing for abiotic stresses such as drought tolerance, and therefore challenges the representation of genome editing as a more accessible form of plant breeding.This narrative of accessibility was empirically assessed by Beumer and de Roij (2023), who reviewed 30 research projects using genome editing to breed crops and traits explicitly for smallholder farmers.
They found that the number of initiatives-when compared to genome editing of plants globally-is relatively small, and that most initiatives were limited in their inclusion of farmers' knowledge, needs, and interests as part of the innovating process.
Taken as a whole, a review of social science assessments of genome editing reveal contentious knowledge politics at play.Regulation, infrastructure, and the complexities of technological innovation loom large.These themes also featured prominently within our interviews, to which we now turn.

Methodology
To understand how stakeholders view the genome editing of crops and its applications for agricultural development, we sought out interviews with scientists, donors, and biotechnology policy experts working in the fields of agricultural development, plant breeding, and biotechnology in sub-Saharan Africa.We identified potential interlocutors by generating a list of authors of relevant publications, scientists featured in news articles on the topic, and individuals known through our respective professional networks.
Of the 32 individuals we approached, 13 agreed to be interviewed.This relatively small respondent pool reflects the nature of the small and emerging field of genome editing in Africa.The group, comprising 10 men and 3 women, included 8 scientists, 3 biotechnology policy experts, 1 university student, and 1 donor.Geographically, 6 were based in institutions in East Africa, 4 in West Africa, and 3 in the Americas.The scientists worked at a mix of research institutions, universities, and the CGIAR.The biotechnology policy experts worked at organizations with various mandates, including developing political and technical infrastructure for biotechnology projects and building public support for the technology.Given the small number of people working in this field, we only provide broad demographic details, and we attribute quotes with generic labels to ensure anonymity (e.g., "CGIAR Scientist 1").
Due to the COVID-19 pandemic, all interviews were conducted via Zoom in the spring of 2021.We asked open-ended questions such as: what possibilities or advantages does genome editing offer African agriculture?What are some of the potential challenges facing the adoption/ embrace of genome editing at the research level?What sorts of lessons, if any, can we draw from the experience of GM crops?With focus on the country of region of the respondent, what are some of the biggest challenges and opportunities for agricultural development and ensuring food security?
After transcribing the interviews, we analyzed them in NVivo, using an iterative approach that allowed for successive rounds of inductive coding.This process allowed us to zero in on the key themes, patterns, and divergences that emerged across the interviews (Thomas, 2006).Out of this process, we identified 3 key themes: the question of regulation, the importance of infrastructure, and the perils of simplistic narratives.We first present the interview evidence related to each of these themes, identify the (in)congruencies across respondents, and then analyze each theme with reference to the narratives in the literature outlined above.
Finally, we conducted a desk review to identify crop breeding projects that are using genome editing in their work.The desk review included surveying peer reviewed literature, project websites, and press releases.We identified the crops, traits, and actors that make up these projects, which we combined with our interviews and literature review to assess the current state of the experimental pipeline.

Genome editing in Africa
How is genome editing being used to develop crops for African farmers?A high-level view of the nascent pipeline provides some broad oversight into how the technology is being used (see Table 1).First, much of the work is taking place either directly within, or in partnership with, CGIAR institutes and African institutions, and much of this work is reliant on partnerships.In some cases, these partnerships include a private-sector actor, but in others, they do not.The projects focus on a variety of crops-including banana, cassava, cocoa, maize, millet, rice, sorghum, and sweet potato-and several traits, including diseases resistance and drought tolerance.
However, a list of current applications is only one part of the wider story.With this broad framing in mind, we now turn to interviews with our interlocutors, whose expertise provides insight into how the technology is-and isn't-being rolled out in real time.

"Not really GM": On the question of regulation
For those we interviewed, one of genome editing's strongest assets is its potential to rethink policy around biotechnology.Interlocutors were adamant that biosafety regimes across the continent remain too stringent and hamper innovation."Africa's experience with [GM] has been very bad," said one scientist at CGIAR, "you see a lot of people kicking against it . ..that is why up to now there are many countries in Africa that do not even have regulation . . .And so, now that genome-editing has come in, it's potentially good and different from GM, so what we have to do is, you have to do a lot of sensitization . . .Otherwise, we still have to go down the same road of GM" (CGIAR Scientist 4, February 6, 2021).Others we spoke with echoed this point."There is still a stigma of the transgenics," said one scientist at CGIAR, "so there is also the complexity of regulations.Especially in certain parts of Africa" (CGIAR Scientist 6, March 16, 2021).Another scientist at CGIAR said, "You have the biosafety regulation institution in our places.In some cases, they behave like people who are there to stop the technology to be adopted, instead of being the people who are just assessing and making sure that all the procedures are well followed" (CGIAR Scientist 3, February 4, 2021).
Many of our interlocutors expressed hope that African governments would regulate genome-edited products differently than their GM counterparts.An official involved with regulatory design shed light on the type of regulations their employer was advocating for: "the approach that the African nations want to use is the science-based The scientist continued, mentioning several crops they thought could benefit from genome editing, including cassava, yams, and teff (Scientist, CGIAR, 22 January 2021).A second scientist listed egusi melon, cocoyam, and taro as crops ideally suited to genome editing (CGIAR Scientist 4, February 6, 2021).While many interlocutors expressed excitement about the scientific possibilities of genome editing, this optimism was tempered by sober reflections on infrastructural realities and a desire to not "oversell" the technology.Many also spoke of structural inequities through the lens of infrastructure, and the very real material and financial struggles they face in their labs.We turn to those considerations in the next section.
"Research is expensive": On the importance of infrastructure The next theme that emerged among our interlocutors related to infrastructure.Infrastructure here refers to the in-country facilities that enable scientists to do their job, which could include lab access and supplies, electricity and Internet, and genetic material, alongside funding, a key resource in obtaining (and retaining) infrastructural resources.Taken together, questions of infrastructure are questions of accessibility, and speak to the narrative of genome editing being cheap, easy, and readily available.
While interlocutors were in broad agreement that genome editing was likely to cost less than GM, many pointed out that cost reduction was not necessarily synonymous with accessibility.Said one scientist at CGIAR, "As a powerful technology, many people talk about-'Oh, you can do it in your garage'.Well, it's not that easy.Conceptually, it's pretty simple.But, you know, as you start actually doing it, you realize that this is not that easy.But nonetheless, it's pretty inexpensive" (CGIAR Scientist 2, January 22, 2021).Another scientist at CGIAR emphasized that funding varied across institutions within Africa: "It is only in these international institutions [CGIAR] that can do this work . . .I'm not sure [national] institutions would not see these facilities because funding is limited" (CGIAR Scientist 4, February 6, 2021).Here, the scientist pointed to a lack of infrastructure at public universities and research institutes relative to the CGIAR centers.While some of our CGIAR interlocutors similarly raised issues around funding, overall, they seemed more confident in their ability to capitalize on genome editing given current resources: We don't have many challenges because genome editing doesn't require anything extra.You know, if you have the laboratory facilities that you are using anyways for transgenic [modification], with similar facilities we can do gene editing.So, apart from funding, . . .I don't see any big challenge.(CGIAR Scientist 1, January 21, 2021) It was a different story for interlocutors outside the CGIAR system.One scientist based at a national research institute emphasized that many public institutions employ scientists skilled in genome editing, but that a lack of funding inhibits their ability to purchase the tools and materials necessary to complete the breeding process: the pipeline for doing that kind of work is in place but what is a bit limited is finding additional money to implement on your own research ideas.Compared to America where there are many funding agencies, we have the opposite here.They're very limited and we also find ourselves applying for the same grants as other nationals, . . .and because of that our chances of succeeding are limited.But we keep trying.(Research Institute Scientist 1, February 5, 2021) An interlocutor at a public university echoed this point.This individual had recently returned home after spending a few years abroad working on genome editing.They affirmed that they were continuing their work at their new institution, citing the common narrative that genome editing requires few financial or material resources: "for . . .genome editing you don't really need much.You don't need super sophisticated materials.We have the basic tissue culture laboratory and some of the people we need" (University Scientist, February 25, 2021).
But, as they continued, the scientist began to enumerate several challenges: "Most of the things we would be needing, we have to order from the biotech companies . . .For supplies, like enzymes, and you know so many of the supplies, we will have to order them."Ordering supplies, the scientist explained, not only required funds but also time and patience: "You [might] have the money, but when the ordering process is so difficult.Cumbersome."They continued, listing the work that goes into clearing the supplies at customs, paying import duties, and so on.Once supplies are obtained and the work can commence, the scientist still faced issues with a lack of lab equipment: "And then also we will be doing a lot of sequencing.[There's] a lot of sequencing that we cannot do it here.Once we get our DNA, we can send the samples to the partner companies to do the sequencing for us."Later in the interview the scientist concluded, "research is expensive" (University Scientist, February 25, 2021).
The infrastructural constraints that the university scientist spoke to were vast: ordering and paying for supplies, navigating customs and import taxes, collaborating with an outside lab to sequence DNA, and so on.Overcoming these constraints was not only expensive but also required significant outsourcing: neither supplies nor sequencing could be obtained in house.The differing perspectives offered by those scientists within and outside the CGIAR space underscores the trenchant divide between plant breeders based at international organizations and those based at national-level research organizations whose relative lack of infrastructure and funding make such ambitious experimental programs feel largely out of reach.

"Maybe . . .overselling it": On the perils of simplistic narratives
The hope that genome-edited products would not be considered GMOs-either by regulators or by the public-that remains pronounced within the literature was echoed by many interlocutors.This designation is as much discursive as it is technical.Discursively, creating distance between genome editing and GM was seen to be a way to move past a technology that has generated immense public and political scrutiny.In other words, cementing a linguistic distinction was a way to generate more public support.

Said one CGIAR scientist:
There is a lot to learn from GM when moving forward.It is more related to communication I will say.So, you know start with early communication, . . .and use very simple way of explaining the technology in a clear way.Then is public awareness, so people don't fear the new technology . . .To build the confidence in the public I think that communication is the key.(CGIAR Scientist 1, January 21, 2021) Describing differences between GM and genome editing was also a way, some believed, to generate more favorable regulatory models (explored above in the section entitled "On the Question of Regulation").Here, some interlocutors-mainly those who worked outside public institutions-again referenced the narrative of genome editing as easier, cheaper, and more accessible than GM.Said one interlocutor: "Genetic modification is already phasing out now, because gene editing is less cumbersome, is more precise, . . .and doesn't take so much money [and] doesn't take so much time" (Biotechnology Policy Expert 2, February 19, 2021).
However, some interlocutors worried that such messaging oversimplified a complex field and could be counter-productive for convincing members of the public of its value.One topic that interlocutors raised was the narrative-and technical designation-of genome editing as a significant departure from GM.At least one interlocutor called this narrative "disappointing" and "confusing": it is disappointing because many people, maybe because of vested interests, want to portray genome editing as the same as conventional [breeding].And that is causing a lot of confusion.It is true that because of the high expenses in GMO, most of what will come up as genome editing will eventually be the same as conventional.But I think the precise truth is that some of the products will come out identical as conventional from the beginning, some of the products will have a component of GMO at some stage, then that GMO part is removed.(Biotechnology Policy Expert 1, February 12, 2021) We asked our interlocutors what lessons, if any, they thought were relevant from the previous generation of GM.One scientist at CGIAR worried that portraying genome editing as being easy to implement parroted previous narratives of GM as being a quick and easy sell.The scientist worried that deploying such a narrative around genome editing might be "overselling it": Probably a big mistake of the early days of transgenic technologies was assuming that society would embrace these things.As scientists, . . .there was so much excitement about what transgenic would be able to achieve.That, you know, scientists look at this, "Wow, we're going to be heroes, everyone's going to love us for this, . . .who could possibly be against this?" .I think this [was] sort of naı ¨ve . . .And then maybe the other thing was overselling it.I mean, and this is also true of gene editing.A lot of really important traits in any crop, such as drought tolerance or heat tolerance or certainly yield, are influenced by many, many genes . . .So, you're not going to triple yields with a transgenic, or with the gene edited product . . .So that, those are some of the mistakes that we made with transgenics and I don't know if it's rescuable.(CGIAR Scientist 2, January 22, 2021) The comments from both the Biotechnology Policy Expert 1 and CGIAR Scientist 2 come down to the question of building public trust.Both interlocutors identified several variables that might dampen public trust and acceptance: misrepresenting the type of biotechnologies used to develop a plant; glossing over issues of safety and regulation; and promising mass access.One interlocutor noted that these considerations were not unique to the African continent: "In Africa, gene editing is still so new, [like] everywhere else" (Research Institute Scientist 2, March 16, 2021, emphasis added).
However, one scientist at CGIAR mentioned an additional factor that was more geographically, and historically, specific: "it will help to increase the trust in the technologies if people know that it's developed locally, by a fellow citizen" (CGIAR Scientist 3, February 4, 2021).They continued, drawing a connection to the COVID-19 pandemic and vaccine deployment: I don't know if you are aware of the discussion when people were in the phase of development of the vaccine for COVID-19?There were some people that were thinking of doing the trials in Africa, and there was a lot of discussion.People said " no, no, we are not going to be guinea pigs for you guys.Develop your vaccine, try it on your own people, don't bring it here" .People have the feeling . . .there are many people in our places that have the feeling that the Western guys they are plotting something . . .And this is very difficult to change.So, then, if you want to build trust, then you have to invest in . . . the continent.Otherwise, it will be very difficult.(CGIAR Scientist 3, February 4, 2021) Here, the interlocutor is speaking to specific historical and political-economic contexts through which they believe individuals may judge genome editing as citizens or consumers.In the next section, we discuss how historical contexts, structural inequalities, and ideas about technological advancement provide essential insights into the future of genome editing in Africa.These insights lead us to question narratives around the "revolutionary" potential of genome editing and to interrogate the knowledge politics at play across the continent's diverse contexts of research and innovation.

Discussion
Across our interviews, respondents expressed a range of ideas regarding the possibilities and realities of genome editing.Most of the scientists we interviewed expressed enthusiasm about the technology's potential, one that echoes the narrative that dominates within popular and academic sources.But there was a distinct difference between interlocutors within the CGIAR system, and those outside of it.In comparing the transcripts from the scientists at African universities or national research institutions-who spoke of cumbersome processes of ordering supplies, sending materials for sequencing, and obtaining funding-to that of the CGIAR scientists ("apart from funding, I don't see any big challenge"), a particular and important tension comes to light: that of the capacities enjoyed by the CGIAR centers versus national research institutions.This division reflects a half decade of divestment from national research institutions, by both governments and donors, even though repeated reviews have recommended devolving and decentralizing CGIAR's role in plant breeding to national programs who are better positioned to lead such experimental programs (Byerlee and Lynam, 2020;McIntire, n.d.).
While some scholars are optimistic about the ability of small labs to benefit from genome editing (e.g., Qaim, 2020), interlocutors outside of the CGIAR system worried about their ability to access the funding and tools they would need in order to lead their own genome editing projects.Here, concern extends beyond simply whether scientists possess the skills to utilize new agricultural technologies, but rather, whether the systems exist that enable them to apply them (Hall, 2005).
This divide in access is evident within a review of current projects that are using genome editing-either as a main instrument or a single tool within a larger suite-to address diseases, pests, and staple crops.Table 1 reveals several national institutions involved in genome editing projects.However, few serve as the lead research institution on projects; in fact, the entirety of this work is being undertaken as partnerships.To be sure, partnerships are a common component of agricultural research, especially within the portfolio of biotechnologies.But partnerships also raise important questions around availability and accessibility of technology.In other words, if a technology is only accessible through partnering with another entity-an arrangement that inevitably involves negotiations among partners endowed with different capabilities and distinct interests-how "accessible" is it?
This unmasks another dimension of the knowledge politics surrounding genome editing.The ability of scientists, whether within the CGIAR or national research institutions, to work on crops of their choosing is highly dependent on their access to materials and technologies.Partnerships are one way to access these, but they can be complicated and messy, and are usually marked by asymmetrical power relations (Beumer and de Roij, 2023).Public-private partnerships also have a mixed track record when it comes to the delivery of public goods in the realm of agricultural biotechnology development on the African continent (Schnurr, 2019).
An important aspect of the power imbalance among partners stems from patents, an aspect of research infrastructure that was not necessarily raised by our respondents, but which looms large.Both public institutions (such as University of California-Berkeley) and privatesector firms (such as Corteva) have rushed to claim genome editing patents (Egelie et al., 2016;Jefferson et al., 2021).While some of these entities are licensing their patented technology for noncommercial use (van der Oost and Fresco, 2021), patenting still serves as a method of enclosure (Ajates, 2022;Canfield, 2022).Access to patentprotected techniques and materials is at the discretion of the patent owner, and so the rapid patenting of the sector raises doubts around future accessibility by smaller labs (Kock, 2021).
Through interviews and review of the nascent experimental pipeline, an image begins to emerge of a new technology unfolding unevenly across the African continent.Despite the buoyant public narratives that surround genome editing, there are currently few instances of it being applied to crops targeted for African farmers.This tracks with recent findings from Beumer and de Roij, who argue that globally, "the use of gene editing for smallholder farmers is emerging slowly at best" (2023, p. 1).In some ways, this slow emergence and uneven unfolding reflect larger structural, historic inequities.Addressing this is not simply a matter of building knowledge and labor capacities (though those do play an important role) but also of directing funding toward the building and maintenance of infrastructure in national research institutions, universities, and other domestic entities.
It is perhaps for these reasons, and more, that many of our interlocutors spoke to the danger of simplistic narratives, and how their use might harm efforts to build trust and understanding among the public.For example, Biotechnology Policy Expert 1 shared that it was "disappointing" that some sought to "portray genome editing as the same as conventional [breeding]."This insight echoes Kuzma's argument: that using overly simplistic phrases or narratives can be perceived as "dishonest" and "lacking integrity" by the public (2018, p. 81).CGIAR Scientist 3 spoke to the need for transparent discussions around genome editing, and referenced COVID-19 vaccine trials and people's fear of being "guinea pigs" (see above) to make their point.As Flint (2020, p. 130) has shown, "historical antecedents of . . .medical experimentation, [and] continued structural inequalities" live on and inform views on things like vaccines.Historical precedent and structural injustices also shape how some view biotechnology (Rock, 2022).
In sum, our conversations with stakeholders actively invested in genome editing in Africa reveal the myriad challenges they face on the ground and expose some of the more trenchant obstacles-in terms of access, logistics, and control-that stakeholders perceive as potential barriers to achieving genome editing's triumphalist potential in Africa.The knowledge politics that underlay the rollout of this new technology mirror that of previous generations of agricultural technologies: a multistakeholder program that affords preferential access to those scientists within multinational bodies (like CGIAR) relative to those housed within national organs.At the same time, scientists across different institutions expressed considerable interest and expertise in using genome editing tools.Narrowing this gap between interest and accessibility will be crucial for achieving genome editing's transformative potential.

Conclusion
In this article, we interviewed scientists, donors, and biotechnology policy experts to answer the question: how do stakeholders view the genome editing of crops and its agricultural applications in sub-Saharan Africa?Some of our findings support that which is already established in the social scientific scholarship, mainly that practitioners are by and large optimistic about the potential applications of genome editing, and that many hope-and are working toward-political regimes that regulate genome-edited products as distinct than genetically modified ones.
However, our interviews provide a more sober, empirical assessment to a narrative that is positively buoyant in its expectations for genome editing.As explored here, the dominant narrative that positions genome editing as being more accessible than previous biotechnologies overlooks the vast complexities of plant breeding and genome editing, and the choices scientists must make when developing a research and development strategy: when using genome editing on a crop for the first time, scientists must first develop a proof of concept prior to executing on their actual targets; labs must be well-equipped with supplies and tools; projects must be well-funded to pay for both materials and labor; desired genetic targets may require alternative approaches alongside or instead of genome editing, potentially including transgenesis; in some cases, scientists must negotiate complex licensing agreements with third parties; and practitioners must have the support, skills, and experience to navigate international markets and regulatory agencies.
These complex yet essential factors are obscured by oversimplistic narratives that present genome editing as accessible and democratic, and those that present agriculture in Africa as monolithic and in need of technological advancement.Promoters of the technology would do well to wrestle with the institutional and structural inequities explored here.Indeed, by exploring these complex factors, we are by no means arguing against a particular technology.Instead, we aim to center the empirical realities and valuable insights offered by those who are leading experiments and policy building on the ground that shed light on the rhetoric and the realities underpinning the knowledge politics of genome editing.Such insights will play a crucial role in shaping the technology's trajectory moving forward.

Data accessibility statement
Data supporting this study were obtained through interviews and are thus not publicly available due to human subjects protection.Questions regarding methodology should be directed to the corresponding author.

Table 1 .
Select research and development pipeline of genome edited crops targeted for Africa : if the product is same as the conventional, it should be regulated as conventional.If the product is the same as (a genetically modified organism[GMO]), it should be [treated] as GMO.If the product starts off as GMO and then ends up as non-GMO, once you confirm that the GMO part has been removed, it should be regulated as conventional product" (Biotechnology Policy Expert 1, February 12, 2021).One scientist we interviewed considered such a regulatory approach a "landmark": approach