Governing gene-edited crops: risks, regulations, and responsibilities as perceived by agricultural genomics experts in Canada

ABSTRACT

This paper explores the role and responsibilities of agricultural genomics experts in governing gene editing (GE) for food and agriculture, engaging with the frameworks of technological determinism and Responsible Research and Innovation (RRI). We interview agricultural genomics experts in Canada to study expert views on risks, benefits, and regulatory challenges of GE crops and the extent to which agricultural genomics experts exercise the RRI principles of anticipation, reflexivity, deliberative inclusion, and responsiveness. Agricultural genomics experts wield power in food systems both in shaping the applications of technology and as advisers influencing policy and governance. Their resistance to RRI principles, especially deliberative inclusion and responsiveness, and exercises of discursive closure are challenges for responsible governance of GE crops. The study offers empirical and theoretical contributions, working across Science and Technology Studies and food systems research.

KEYWORDS:

• Gene editing
• governance
• agriculture
• RRI
• responsibility
• technological determinism

Introduction

The global food system is characterized by increasing ⁠industrialization, land consolidation, corporate concentration, financialization, and specialization or uniformity on farms (Clapp and Moseley 2020; IPES-FOOD 2016). At the same time, farmers and other food system actors face pressures and challenges associated with climate change and environmental problems, systemic risks and economic shocks, and the persistent challenges of food insecurity (Bowness et al. 2020; Mehrabi et al. 2022). Both the health of the planet and of its inhabitants are at stake in global food systems interventions.

Gene editing (GE) applications to food crops are the latest in a long line of biotechnology innovations proposed to promote food security and sustainability by improving agronomic performance, following the Green Revolution’s package of high-yielding seeds, pesticide, and fertilizer and the Gene Revolution’s advancement of genetically modified (GM) transgenic crops. Yet, the role of agricultural biotechnologies in addressing challenges facing global food systems is open to question – as are the risks of perpetuating existing inequities and causing new problems. Debates over the potential and implications of agricultural biotechnology, including GE, are not strictly technical assessments of risks and benefits; they are political in their constitutions and contestations of power, influenced by values, assumptions, and governance systems, with material implications for the future of food (Bain, Lindberg, and Selfa 2020; Hartley et al. 2016; Montenegro de Wit 2020; Motta 2014; van Oudheusden 2014).

Nonetheless, GE techniques – especially, clustered regularly inter-spaced short palindromic repeats sequences and associated enzymes (CRISPR) – are making revolutionary contributions in many domains, including disease diagnosis and treatment, pharmacogenetics, and modifying the genomes of plants and animals (Hartung and Schiemann 2014; The Nobel Prize 2020). In 2020, Emmanuelle Charpentier and Jennifer Doudna jointly won the Nobel Prize in Chemistry for their research on CRISPR-Cas9, even if GE innovation is the result of international and interdisciplinary research since the 1990s (The Nobel Prize 2020). GE techniques enable ‘edits’ of deoxyribonucleic acid (DNA) or ⁠ribonucleic acid (RNA) that delete, insert, or modify specific genes or sequences. When compared to earlier techniques, GE is more precise, more versatile, more efficient, less expensive, and easier to execute (Xu, Hua, and Lang 2019). These tools can modify genomes without the introduction of exogenous DNA – as opposed to transgenic insertions of genes from one organism to another to produce GM crops (Jones 2015). Without ‘foreign’ DNA, GE modifications are cis-genic and can be traceless, indistinguishable from natural mutations (Hartung and Schiemann 2014).

According to developers of the technologies, GE can produce more climate resilient plants, improve control of zoonotic disease, and increase yields, among other possibilities (Nature Genetics 2019; Xu, Hua, and Lang 2019; Zhang et al. 2019). However, debates in public, policy, and academic contexts are polarized in terms of GE crops’ intended and unintended consequences (Bierbaum et al. 2020; Clapp and Ruder 2020; Macnaghten and Habets 2020; Nawaz and Satterfield 2022). Further, given the complexity of global food systems, innovations may ‘solve’ one problem while worsening another or present new risks (Barnhill-Dilling and Delborne 2021; Macnaghten and Habets 2020; Regan 2021).

Since GM seeds were approved for commercial use in the 1990s, their risks and benefits have been hotly debated, even among experts within and, especially, across academic disciplines (Böschen et al. 2006; Motta 2014; Pavone, Goven, and Guarino 2011). As a general overview of a complex and multi-sided debate: molecular biologists, plant geneticists, and economists usually focus on GM crops’ agronomic performance, highlighting increased yields and profits for some farmers, while finding health and environmental risks to be minimal (Klümper and Qaim 2014; Qaim 2020); social scientists call attention to environmental risks associated with integrating genetically modified organisms (GMOs) with other agrochemical inputs and intensive farming practices (e.g. producing glyphosate-resistant weeds), uneven distribution of risks and benefits among farmers, and contributions to increasing corporate concentration and corporate power (Clapp 2021; Stone 2007; Tourangeau and Smith 2015).

Across public, private, and academic debates, there are shared concerns about the extent of corporate concentration in the agricultural inputs sector, even if priorities and worldviews may conflict. Although there are different ways to define and measure ‘corporate concentration’ (e.g. Herfindahl-Hirschman Index or Concentration Ratios), it is generally understood as a market characteristic: the degree to which a small number of companies control a large portion of the market. For example, corporate concentration generally increases with mergers and acquisitions. For food scholars, corporate concentration is a feature, outcome, and driver of increasing the power of corporations and their ability to control the global food system (Clapp 2016, 2018; Howard 2016). Beyond the study of corporate concentration, political economy and sociology literature on food and agriculture identifies capitalism and the growing power of corporate actors, including multi-national agricultural input and food companies, as constitutive elements of the global food system (Bernstein 2016; Bowness et al. 2020; Clapp 2018; McMichael 2013; Motta 2014).

Publics and governments do not experience GE’s advances in a vacuum; public perception of GMOs influence public acceptance, regulations, and governance of agricultural biotechnology more broadly (Macnaghten and Habets 2020; Middelveld, Macnaghten, and Meijboom 2022; Montenegro de Wit 2022; Nawaz and Satterfield 2022). Scientists, science communicators, and regulators are thus eager to change their approach to public engagement and governance to avoid perpetuating the existing polarization around agricultural biotechnology (Hartley et al. 2016; Kuzma 2018; Roberts, Herkert, and Kuzma 2020). Beyond motivations for avoiding public rejection, involving diverse actors in the development and governance decisions on technologies can advance responsibility, democracy, and justice (Barnhill-Dilling and Delborne 2021; Gordon et al. 2021; Owen et al. 2013).

Conceptual framework: responsible research and innovation

Broadly, Science and Technologies Studies (STS) explores the place, role, and impact of science and technology with(in) society. Our study employs a co-productionist interpretive framework, based in STS (Jasanoff 2004). Co-production is grounded in the assumptions that ‘in most exercises of world-making, neither science nor society begins with a clean slate’ but are instead inextricable from ‘an extant order’ (Jasanoff 2004, 19). Co-production does not foreclose competing interpretations and instead embraces how politics, knowledge, science, and technology are contested and continually in flux (Jasanoff 2004).

A co-productionist conceptualization of technology goes beyond the material artifacts to include the ways of knowing and being they engender. Our study understands technologies as complex and multi-directional interfaces of ecological, technical, social, political, and economic systems. The implications of technologies are not pre-determined or inherent to artifacts. In other words, this research challenges technological determinism – the pervasive assumptions that technology is autonomous and outside of society, technologists are merely applying science through a linear relationship, and technological change causes social change unidirectionally (MacKenzie and Wajcman 1985). We use technological determinism as a analytical tool to illuminate the assumptions and values influencing understandings of technology and the extent to which technological change is viewed as autonomous and inevitable or impacting and impacted by social systems (Dafoe 2015; Markham 2021).

STS offers a diversity of approaches to study the social and political dimensions of science and technology. Since the early 2000s, STS has catalyzed attention to responsibility and ethics in public and academic debates over technology and how it ought to be governed (Jasanoff 2016; Owen, von Schomberg, and Macnaghten 2021; von Schomberg 2019). There is also growing enthusiasm for involving society ‘upstream’ in research and innovation. Yet Responsible Research and Innovation (RRI) is one of the only frameworks proposing practical ways forward (Guston et al. 2014; RRI Tools Consortium 2021).

Originating from the European policy context, René von Schomberg (2013, 63) offered an early definition of RRI that remains in use today:

Responsible Research and Innovation is a transparent, interactive process by which societal actors and innovators become mutually responsive to each other with a view to the (ethical) acceptability, sustainability and societal desirability of the innovation process and its marketable products.

RRI aims to establish collective responsibility for social and environmental goals rather than the dominant prioritization of maximizing profits through innovation (Barnhill-Dilling and Delborne 2021; von Schomberg 2019). As a counterfactual, von Schomberg called Monsanto’s role in developing and commercializing GM seeds ‘irresponsible innovation’ – pointing to their pushing GM soybeans in the EU in the late-1990s as an inciting incident for polarization (2013, 60). A co-productionist perspective undergirds RRI, where technologies are ‘not merely about achieving ends that we already foresee but an open door to an uncharted, often uncertain future where current social understandings and practices may be fundamentally transformed’ (Jasanoff 2016, 213).

Jack Stilgoe, Phil Macnaghten, and Richard Owen developed four integrated dimensions of RRI: Anticipation, Reflexivity, Inclusion, and Responsiveness (Owen, Macnaghten, and Stilgoe 2012, 2013; Stilgoe, Owen, and Macnaghten 2013). These principles do not seek to ‘instrumentally legitimise any particular framing or commitment,’ but rather ‘open up space for essential governance discussions aimed at supporting, but not dictating, decisions about the framing, direction, pace and trajectory of contentious and innovative research’ (Stilgoe, Owen, and Macnaghten 2013, 1576). Following Florin (2022), Klerkx and Rose (2020), Regan (2021) and others, we build upon the principles of RRI to ask framing questions to guide our interview study and elucidate perceptions of agricultural genomics experts in Canada. (Table 1).

Table 1. RRI principles and framing questions.

RRI Principles (Drawn from: Stilgoe, Owen, and Macnaghten 2013, 38)	Framing Questions: for interviews with agricultural genomics experts in Canada
Anticipation Asks ‘what if’ questions about innovation and its complex implications; considering both intended and unintended consequences; explores tensions between the reification of prediction exercises and diversity of possible futures	What risks and benefits of GE in agriculture do experts anticipate? How does GE relate to the (un)anticipated impacts of earlier biotechnology?
Reflexivity Reflecting on underlying assumptions, worldviews, and ignorance; considering what is known and unknown; thinking critically about the foundations and limitations of perspectives, including one’s own	How are experts exercising reflexivity, if at all? What could/does reflexivity look like in the agri-food innovation context?
Deliberative Inclusion Seeking out a wide range of perspectives on visions and purposes; normative and substantive deliberation; representing diverse knowledges, values, and meanings	How do experts perceive ‘the public’? What are their opinions on public engagement? What assumptions/frames do they espouse regarding expertise and knowledge? How do experts envision the inclusion of different knowledges?
Responsiveness Acting on new insights developed through reflexivity and deliberative inclusion: ‘an iterative, inclusive, and open process of adaptive learning, with dynamic capability’	Is the current governance framework set up for technological change and innovation? What about changing social values and norms? Who is responsible for responsiveness?

The questions that RRI addresses grapple with power relations, policy, and governance: ‘responsible innovation cannot, and should not, be decoupled from its political and economic context’ (Owen et al. 2013, 37). However, some find that RRI does not yet meaningfully or sufficiently examine power relations and politics (van Oudheusden 2014). Although RRI does not necessarily interrogate systemic oppression or power, there is potential to expand the concept of responsibility to engage with the historical and contemporary entanglements of ‘research and innovation’ with colonialism, racism, and hierarchical knowledge politics. Some scholars push to expand the concept of responsibility to address more-than-human ethics, care and reciprocity, and supporting Indigenous movements (Barnhill-Dilling and Delborne 2021; Szymanski, Smith, and Calvert 2021).

Agricultural genomics experts in Canada

This study explores the role and responsibilities of scientists and their institutions in governing gene-edited crops (GEC). While GM innovation was industry-led, academic researchers drove forward GE developments (Brinegar et al. 2017). Still, the innovation pathway for GE is marked by entrepreneurial ventures and spin-off commercialization of university projects (Lander 2016).

We focus on the population of agricultural genomics experts in Canada with practical knowledge of GE applications in plant crops, as key governance actors in shaping agricultural futures. Agricultural genomics – the field of biology focusing on the structure, function, evolution, mapping, and editing of genomes for agri-food applications – is gaining traction globally, with the ‘promise [of] ⁠improv[ing] the productivity and sustainability in crop and livestock production’ (Nature Genetics 2019; Wang et al. 2017, 1). Agricultural genomics experts are usually trained in non-medical molecular biology or genetics. Agricultural genomics research on GE includes experiments in lab settings (e.g. plant breeding, genome editing) and computational work (e.g. genotype sequencing platforms, predictive modelling to direct edits). Given global interconnections of food systems and Canada’s large export contributions and leadership in agribiotechnology innovation, the influence of Canadian agricultural genomics experts extends beyond the national scale.

Governance approaches within and across jurisdictions will shape the conditions under which GE can be applied to agriculture and influence public perceptions and the complex impacts of the technology. There is a wide range of regulatory approaches employed internationally, whether by reviewing regulatory systems that predate GE or introducing new regulations (Eriksson et al. 2019; Friedrichs et al. 2019). In Canada, GECs are included in the existing Novel Food Regulations [Division 28, Part B, of the Food and Drug Regulations], first published in 1999, which state that ‘plants with novel traits’ are the trigger for regulation and assessment case-by-case, rather than the techniques and processes employed: ‘It is Health Canada's scientific opinion that gene-edited plant products should be regulated like all other products of plant breeding, by focusing on their final characteristics and not the method of product development’ (Health Canada 2022, 3). In other words, the Canadian regulations are the same for GM, GE, and conventional plant breeding. The United States (US) similarly uses existing ‘product-triggered regulation’ for GE in agriculture (Friedrichs et al. 2019; Genetic Literacy Project 2021). In contrast, the European Union (EU) is much stricter and more precautionary (Genetic Literacy Project 2021). The European Court of Justice ruling in 2018 stated that all directed mutagenesis (e.g. CRISPR, TALEN, etc.) will be regulated as GMOs, meaning that novel GE applications will face the same strong regulatory hurdles as earlier biotechnology in Europe (Callaway 2018).

Agricultural genomics experts constitute a powerful actor in GE governance both through direct (e.g. conducting GE) and indirect means (e.g. the privileging of scientific expertise in policymaking). According to existing empirical research, most domain experts (e.g. molecular biologists) find that GECs pose few environmental, societal, and health risks (Kato-Nitta et al. 2019; Lassoued et al. 2021; Qaim 2020). Further, Lassoued and colleagues’ multi-year international survey project with domain experts reports that ‘pressure from stakeholders with distinct cultural, environmental and societal concerns, divergent levels of risk tolerance and political machinations complicate governance of these novel technologies’ (2021, 1107).

Although the RRI framework emerged largely because of controversies around GMOs (Owen, von Schomberg, and Macnaghten 2021), academic case studies and policy proposals for RRI predominantly focus on nanotechnology and geoengineering (Guston et al. 2014; Owen, von Schomberg, and Macnaghten 2021; Parkhill et al. 2013). Research on synthetic biology and GMOs in the US (Roberts, Herkert, and Kuzma 2020) and United Kingdom (Marris 2015) points to relevant actors’ resistance to RRI frameworks, despite the clear consequences of public acceptability and polarization as central challenges to those developing and regulating these technologies. Others are applying RRI to digital agriculture and big data (e.g. Rose and Chilvers 2018; Regan 2021; Eastwood et al. 2019; Duncan et al. 2022).

We conducted semi-structured interviews with agricultural genomics experts in Canada who have practical knowledge of GE to answer the following research questions:

1. How do agricultural genomics experts in Canada view the risks and benefits of GECs?

2. What are the regulatory challenges for GE in agriculture, given the history of GMOs, according to these experts?

3. How do agricultural genomics experts in Canada view the role of publics in the development and governance of GE?

4. To what extent do these experts exercise principles of RRI, and what are the implications for the governance of GECs?

We present the empirical results on participants’ perceptions of risks and benefits, governance and regulatory challenges, and public engagement regarding GECs (Findings), then deductively apply the frameworks of technological determinism and RRI to direct our analysis (Discussion). The study offers empirical and theoretical contributions, working across STS and food systems scholarship.

We argue that RRI is a useful framework to critically examine the governance and implications of GE in agriculture. Although non-governmental organizations (NGOs) remark that application of RRI to GE in agriculture is lacking (Gordon et al. 2021), there is a growing body of RRI research examining GE applications for agriculture, especially in Europe (Agapito-Tenfen et al. 2018; Ludwig, Leeuwis, and Boogaard 2022; Macnaghten, Shah, and Ludwig 2021; Macnaghten and Habets 2020; Middelveld, Macnaghten, and Meijboom 2022; Middelveld and Macnaghten 2021). In particular, we recognize Phil Macnaghten, ⁠Senna Middelveld, and colleagues’ contributions to RRI scholarship in critically examining the development and governance of GE in Europe, as well as the narratives or imaginaries at play in debates about GE applications to crop plants and livestock.

Methodology

RRI and co-production coincide with social constructivist epistemology (Creswell and Poth 2018; Pinch and Bijker 1989; van Oudheusden 2014). For social constructivists, knowledge cannot be divorced from the knower(s) and their contexts – even in the natural sciences (Bijker 2010; Boon and Van Baalen 2019; Jasanoff 2004). Similarly, we exercise a social constructivist understanding of technology and technological change (Bijker, Hughes, and Pinch 2012; MacKenzie and Wajcman 1985; van Oudheusden 2014; Winner 1989).

The empirical foundations for this study are eight in-depth semi-structured interviews with agricultural genomics experts in Canada with practical knowledge of GE. This expert population is quite small (participants confirmed this to be true during the interviews). Expert interviews are increasingly common in qualitative research (Bogner, Littig, and Menz 2009; Döringer 2021). Usually, expert interviews focus specifically on areas of participant expertise (e.g. posing policy questions to policymakers). However, in line with the RRI framework and its theoretical expansion of scientists' responsibilities (Roberts, Herkert, and Kuzma 2020; Smallman 2020), we asked questions about perceptions of risks, benefits, and governance challenges. This approach to interviewing ‘experts’ also gives space for thinking critically about knowledge and expertise (Reid and Sieber 2021; Smallman 2020).

We identified and recruited participants by searching a nationally comprehensive list provided by the Genome BC (a public research funding agency focused on genomics) that included experts on GE across academia, public, and private research organizations, as well as faculty listings in relevant university departments and Canada Research Chair appointments. With a purposeful sampling strategy (Creswell and Poth 2018; Lynch 2013), we recruited a diverse group of participants to represent the population. All participants are from different universities. Several participants conducted agricultural genomics research in the private sector at some point during their careers. Some participants worked with government bodies or trade associations. With one exception, all participants are currently working at a university. Five of the eight participants are tenured professors. All participants were in Canada at the time of the interview, but one participant’s research lab is in Europe. While the population of GE experts in Canada is male dominated, it is a limitation that all participants identified as men.

The semi-structured interview approach offered guidance for participants on topics they may not be in the habit of discussing (e.g. public engagement in governance), while mitigating the interviewer’s control over the conversation (Brinkmann and Kvale 2015; Schensul, Schensul, and LeCompte 1999). Much like Regan’s (2021) study of experts’ readiness to engage in RRI for digital agriculture, we decided not to explicitly use the RRI framework in the interview questions, but used similar language. We piloted the interview with a post-doctoral researcher working on gene-edited crops. No substantive changes were necessary. While we do not include those results, that interview did not substantially diverge from the rest.

The first interview took place in person in February 2020, but all other interviews were conducted in video-calls due to the COVID-19 pandemic. The transition to virtual interviews offered more advantages than compromises; virtual qualitative data collection methods enabled continued data collection during the pandemic, access to diverse participants when travel was not possible or appropriate, and maintained interactive access to body language and visual cues to build rapport (Keen, Lomeli-Rodriguez, and Joffe 2022). All interviews were audio-recorded for verbatim transcription. The interviews varied in length to follow the pace set by participants, ranging from 50 to 90 min. Transcripts were uploaded to NVivo 12 for content and thematic coding. Data analysis deductively answered the research questions and assessed RRI principles, but also iteratively and inductively developed new insights to describe and understand the values, motivations, and experiences of participants in a specific context.

Findings

Agricultural genomics experts in Canada provided rich descriptions of their perceptions of risks and benefits associated to GE applications in agriculture, as well as conceptualizations of good governance of agricultural biotechnology. We present the findings in three categories: perceptions of risks and benefits, governance and regulatory challenges, and public engagement.

Evaluating risks and benefits of GE in agriculture

The agricultural genomics experts brought up advantages of GE before being asked about benefits. Following an introductory question, we asked each participant to describe GE applications for agriculture in lay terms. Most of the definitions and attributes of GE provided focused on benefits, especially when comparing GE with GM (transgenics) or conventional breeding. Even with a focus on technical details, participants defined GE with enthusiasm and supportive language.

Participants explained that GE techniques (especially CRISPR-CAS9) enable unprecedented precision and accuracy in modifying plant genomes using a simpler, more efficient, more accessible, and more versatile procedure as compared to transgenics and conventional plant breeding. GE expands what is genetically and biologically possible in plant breeding, such as editing multiple genomic targets at once (multiplexing) and making changes without integrating the enzyme in the plant genome – and without requiring random mutagenesis or DNA from another species (although it remains possible to integrate exogenous genes). GE also permits the elimination of unwanted or maladaptive traits that come along with the desirable traits in the breeding process (linkage drag), a common problem with transgenics.

Prioritizing benefits and new possibilities

When asked about benefits or opportunities that GE presents for agriculture, each of the participants outlined specific agronomic benefits that are possible with GE, without prompting. This is unsurprising given their direct engagement in applying GE in crop plants. The responses reached beyond existing GE crops to anticipate future applications. Within the agronomic context, participant responses can be organized in six main benefit categories: efficiency (e.g. more efficient water use, nitrogen use, photosynthesis), productivity (e.g. increase yields, root architecture), herbicide and pesticide resistance (e.g. resistance to glyphosate), biotic stress resistance (e.g. resistance to disease, fungus, pests), climate change and extreme weather tolerance (e.g. frost or drought tolerant), and new possibilities (e.g. perennials varieties of annual crops like corn, increasing genetic diversity). Participants usually mentioned benefits in all categories. Often agronomic benefits were connected to the environmental or economic benefits. For example, if crops are more tolerant to diseases, farmers can apply less pesticide and fungicides and drive tractors across the field less frequently, which would reduce the environmental footprint of farms. With more efficient crops, farming could take place on less land, which could be a benefit to biodiversity.

Looking beyond laboratories and farmers’ fields, participants outlined social and economic benefits of GE applications that can permeate the food system. Participants articulated a ‘big picture’ notion that GE will contribute to food security by increasing productivity and yields. Crops could be gene-edited to last longer which could reduce food waste. They suggest that GE could improve nutritional quality and flavour with potential health benefits for plant crops as well as plant-based processed foods – though some were critical of such ‘gimmicky’ applications (Participant #8). Moreover, because GE is more efficient, less costly, and easier to execute, they believe any gains from successful GE applications are more widely accessible and can yield benefits more quickly when compared to transgenics and conventional plant breeding. Participant #6 explains, ‘The benefits aren’t new. They’re the benefits we’ve always strived for in plant breeding. We’re just going to be able to attain them in a meaningful timeframe.’

When prompted to anticipate the impacts of GE for the food system, participants were enthusiastic about new possible trajectories, especially in improving the likelihood of small companies being able to participate in developing GECs and more diverse applications:

Genetic engineering [GMOs] should have been able to help vegetable crops, small acreage crops […] but it didn't because the regulations smothered it out. And the big companies don't work in small crops, they work in big crops. So, if we don't smother this technology [GE] out, the people that are the most excited in Canada are barley, oats, and vegetable growers, because they lost out on the last round of technology, because we didn't do anything in their space. (Participant #6)

Likewise, another participant pointed to the potential GE applications to optimize seeds for greenhouses and other controlled growing environments, which they suggest could offer food sovereignty benefits in Canada’s Northern communities with short growing seasons and challenges for accessing fresh produce.

Participants anticipated that conventional industrial farmers would be more likely to benefit from GE technologies, not because of technological barriers, but because farmers using certified organic or other environmental practices (diversified agriculture, multi-cropping, etc.) are likely to ‘have huge suspicions of any biotechnology’ and reject GECs. Indeed, research in Canada and the US suggests that organic producers and organizations, strongholds of GMO resistance, are likely to reaffirm boundaries to reject GE as well (Nawaz, Klassen, and Lyon 2020). Otherwise, participants’ anticipation exercises largely omitted the distribution of risks and benefits and other power relations in the food system.

Dismissing risks and responsibility: ‘Who bears the risks? It depends on if you think there is any.’

Overall, participants believe that GECs pose minimal environmental, societal, and health risks, and that any existing risks apply in any plant breeding (i.e. not unique to GE). Two participants identified environmental risks, namely increasing the photosynthetic rate of a crop could strengthen weeds if the transgenes escape to wild populations. Two other participants explained that the ability of GE to change the nutritional qualities of foods could cause unanticipated interactions or health risks for those who eat them. In addition, a few participants consider the abuse of the technology. Participant #8 explains:

The greatest risk is in terms of abuse of the technology. And so, if growers are not careful in how they use useful technology, then they're bound to come up against situations where, whatever pest they're trying to overcome – whether it be a weed or an insect, or a virus, or whatever else might be out there – they [the pests] are going to no longer be controlled by that particular tool. […] It's the story of agriculture. So, whenever you talk to a plant pathologist, they will talk in terms of an ‘arms race.’ […] In a way, it's mostly the growers, that risk losing tools that are beneficial to them for the time being, and then having to maybe go back to using more chemical means to restore productivity in their fields. (Participant #8, emphasis added)

In response, Participant #8 and others view farmers as responsible for using technology in a sustainable way and adapting to change. Other than abiding by the required risk assessments in their jurisdictions, researchers and developers of GE applications for agriculture were not assigned responsibility over the social, political, and environmental implications of their research and products.

When prompted to respond to common perceived risks of GE (e.g. possible off-target effects, health risks of eating GE foods, furthering corporate power over the food system, risks for farmers), participants voiced criticism and questioned the legitimacy of the arguments. For example, participant #4 says, ‘unfortunately, I find that when I talk to people about what it is that makes them uneasy, it’s oftentimes based on arguments that have very little sound foundation or aren’t grounded in truth.’ In contrast, the participants never expressed doubt that their own perspectives were ‘grounded in truth’ – even when speaking about value-laden topics beyond their scientific practice and area of expertise.

It was also common for participants to emphasize how GE techniques could reduce existing risks, including unintended changes to DNA, when compared to transgenics or conventional breeding. GE, in coordination with improved understandings of plant genomics and DNA sequencing, permit a reliable way to check for off-target effects. According to participants, it is the responsibility of the government to assess and mitigate any risks of GE and oversee the fair distribution of risks and benefits. When asked about the distribution of risks across the food system, participant #6 asserts, ‘So, who bears the risks? It depends on if you think there is any. And if there are risks, it's the government's job to be the arbiter in between to ensure those risks are acceptable.’

Governance and regulatory challenges

Participants illustrated an accurate understanding of how GE and other plant breeding technologies are regulated in Canada and the EU, where the most important distinction in regulatory approaches is what triggers the regulatory framework and advantages of product- vs process-base regulation (Health Canada 2022; Nawaz and Kandlikar 2021). Overall, participants view Canada as distinct from other jurisdictions.

There’s two camps: Canada and everyone else. Everyone else is process based. We're product based. […] And I think this is where Canada was a little bit ahead of the curve with its novelty approach, or at least the philosophy of ‘doesn't matter what technique you use, it only matters what you created at the end of the day.’ And that is absolutely the right way to look at plant breeding science. (Participant #6)

Canada is the only country in the world that has a reasonable regulatory system, focusing on novel traits. […] The rest of the world focuses on technology, which makes no sense. (Participant #4)

Although product-based regulations are not unique to Canada (Argentina, Japan, the US, and several other countries also use product-based approaches), it is the only country that uses case-by-case assessment for plants with novel traits regardless of the GE technique used, whereas other product-based regulations still regulate certain processes (See: Nawaz and Kandlikar 2021). Overall, participants consistently express that there is a ‘right way’ to govern GECs, and Canada is on track. Participants were especially critical of stricter regulatory frameworks in Europe. Still, there are important problems with the Canadian regulatory framework, and GE governance challenges more broadly, which are top of mind for participants.

Multi-directional regulatory challenges for GECs

Regulatory challenges were conceptualized multi-directionally in the interviews: technological particulars of GE bring up new challenges for regulating processes and products; conversely, regulatory frameworks pose challenges for developing GECs; and, there are challenges within the science-policy sphere, independent of GE.

Each participant brought up traceability as a regulatory challenge. GE techniques can make changes to genomes without leaving a trace that would set it apart from changes achieved through traditional plant breeding or spontaneous modifications without human intervention. Advances in genome sequencing greatly improve the ability to precisely identify edits, but there is still no way of confirming the process through which the changes are enacted. In contrast, there are tools that can determine the percentage of transgenes present in a product or collection of GMOs, which is essential to regulations and labelling. Without a way to recognize GE, the products could circumvent any GE-specific regulations on what is approved within jurisdictions or permitted in international trade. Participants are concerned about a lack of accountability for developers and their products, with particular distrust for smaller companies in places whose governments they viewed as less accountable. However, the lack of traceability is much less troublesome for a product-based approach as opposed to a process-based approach since the former is agnostic about how a product or seed is created.

Participants also identified that the lack of traceability poses challenges for intellectual property rights (IPR) and patenting GE products. Some anticipated that IPR for GE will shift to sequencing and identifying the specific genes that are responsible for relevant (and profitable) traits:

The real value will be in the hands of those that hold the genes. Because if I identify a gene that when modified improves nitrogen use efficiency, I can patent that information and then I can licence that to a public company if they want to use it in breeding through editing. (Participant #1)

It is interesting to note that this perspective on knowledge politics and the future of IPR aligns with the Canadian product-based regulation. Individual genes might be patentable, but ‘higher life’ forms are not. Still, the question of whether products of GE are patentable in Canada is not entirely resolved and globally key jurisdictions – including the Australia, Canada, China, the European Union, the US – are exhibiting divergence in legal approaches on patents (Nicol et al. 2019).

The most prominent concerns for the science-policy interface in agricultural biotechnology governance were public opposition and distrust. Participants viewed public perceptions as powerful in determining governance and influencing innovation, especially given the history of GMOs. These concerns featured prominently when discussing the regulatory approaches in Europe. In 2018, the EU Court of Justice decided to treat GECs as GMOs under the original directive (Callaway 2018). In line with the participants’ consensus that product-based approaches are the ‘right way’ to govern GE, participants found the EU ruling to be a ‘catastrophic’ setback for GE innovation and the possible benefits the technology affords. Participants viewed fear and misinformation, mobilized by the media and advocacy groups, as key forces shaping the Court’s ruling and governance more broadly. In comparing the trade-offs and implications of different regulatory approaches, participants were highly concerned about repeating the ‘mistakes’ of GMO governance and resulting public opposition.

Regulatory decisions also impact innovation pathways and the political economy of technologies. Participants explained that economic motivations within the regulatory context shape what GE developers pursue: ‘A clear regulatory pathway is the crux of all of it. Okay? Because if you can't raise capital, you can't get investors and you can't get an idea off the ground. If you can't demonstrate the cost to market, you can't quantify the regulatory cost.’ (Participant #6). In addition to what developers pursue, regulatory approaches in different jurisdictions will also influence where the innovation takes place. Participants identified the European governance of GMOs and GECs as having a direct impact on research and commercialization of agricultural biotechnology in the EU, with global implications for nations that rely on trade with European countries. Given the global interconnections of the food system and technological innovation, international inconsistencies and ambiguity of GE regulations will be a challenge for innovation and trade.

Responding to the unintended consequences of ‘overregulating’

When shifting their attention to anticipated impacts of GE and its governance, participants applied lessons from the past to anticipate possible future scenarios. There is a consistent message across the interviews: based on their perception of GMO governance since the 1990s, participants share a belief that strict regulatory approaches lead to expensive and prohibitive innovation and further corporate concentration.

According to participants, the cost of navigating regulatory approval limits innovation to those who can afford it, namely large corporate actors, even with product-based regulations. Consequently, GMOs (and profits they generate) became increasingly concentrated with a small group of large agricultural input companies – Bayer-Monsanto, ChemChina-Syngenta, and the like. Despite the initial proliferation of companies and labs working on transgenics, smaller companies and universities were excluded. Even though they are in favour of the product-based approach, participants believe that Canada’s current regulatory approach is overly burdensome and will reward innovations that maximize profits for these large companies, rather than those prioritizing social or ecological aims. For example, Participant #3 explained:

If we have a regulatory system like the current regulatory system, especially in Canada, then you're gonna have the traits that can result in the big companies being able to make a lot of money. And you’re not necessarily going to have the traits that are of benefit to society. So, if we relax those regulations and make it a lot easier for the little guys to get in and actually develop new varieties that, you know, might not necessarily have a big economic payback but allow the development of niche markets, then I think there's going to be more a greater societal benefit.

Likewise, Participant #2 explains: ‘[with] stronger regulations, there’s more corporate power because stronger regulation will impose higher cost and not everybody can afford it.’ This narrative was consistent across all interviews.

Learning from the unintended consequences of regulations for GMOs, participants argued for more realistic and simple regulations and approval processes for GE applications in agriculture, without compromising rigorous evaluation of each innovation. Engagement with responsiveness focused more on regulatory approaches than GE research and innovation. Participant #1 asserts, ‘you cannot make regulations based on the current state of technology, you need to write a set of rules that can also be applied as technology develops.’ Responsiveness to technological change should be built into the regulatory approaches; this belief also aligns with the preference for product-based regulation.

There is a consistent critique of overly burdensome risk assessment and regulations for plants with novel traits. It was also common for participants to ask for governments to respond and adapt based on scientific evidence and the results of earlier regulation:

You know, you present 100 million dollars’ worth of data and you have 100% success of being registered. So, it seems that after a while you should realize that, well, maybe it's [regulatory approval process] not increasing our safety. Maybe we need to pare down exactly what we're asking for in terms of regulatory. (Participant #3)

Think about it: 25 years, 40 countries all do independent environmental reviews of genetically engineered [GM] crops, no country has ever denied an approval. Nor has ever any country ever even found a risk to be managed. […] It shows you that we've overgrown that regulatory regime by an enormous amount. […] But the good thing is we do have small [GE] companies now. We have Yield10, we have Calyxt, we have Pairwise. There's a whole bunch of these really cool smaller companies that are doing a lot of consumer-based plant breeding, to try to get to that market, like high-fiber wheats, healthier oil. (Participant #6)

Because the developers that were able to afford the testing for GMOs were all able to register their products, the participants take this as evidence that the regulatory approach was needlessly burdensome and ineffective. As the regulatory approval process for GECs evolves, participants ask for governments to be responsive to keep GE innovation open and accessible to smaller developers.

The RRI principle of responsiveness is intended to prompt meaningful action on the learning based on deliberative inclusion of diverse actors. Participants want governments to be responsive to their needs and priorities. Unfortunately, the participants did not place value in including actors beyond scientists, industry developers, and governments. Without this inclusion of other perspectives, the conceptualization of responsiveness in the interviews falls short of the RRI ideal.

Public perceptions and engagement

Genomics technology in agriculture is notoriously contentious. Publics are powerful governance actors with complex impacts as diverse groups containing conflicting perspectives. There was more variation across the interviews on the questions of public engagement in GE governance than any other topic. When asked if and how the public should be consulted with respect to how GE techniques for agri-food are applied and regulated, we received responses at both extremes and in an ambivalent middle:

I don’t think the public should ever be consulted on anything. […] Doing what people generally and the public want to have happen is not necessarily in the public’s best interest, which is weird. (Participant #7)

I’m not really sure what the benefit would be (Participant #3)

I think it’s absolutely necessary that that people be involved in in deciding what sort of framework needs to be in place to make them feel comfortable. (Participant #8)

Participants identify ignorance and polarity permeating public perceptions as risks of public engagement in GE governance.

Imagining publics: weak understandings and strong convictions

While publics have strong opinions, scientists believe that the public has ‘a very poor understanding’ of plant breeding and agricultural biotechnologies. Some participants acknowledged that the public is not a homogenous group, but most assumed a collective identity, and expressed very little confidence in consumers’ abilities to understand the differences between GMOs and GECs. They view most consumers as having little care or ability to be accurately informed about how their food is produced. Further, participants consistently expressed frustration about how publics do not recognize the limits of their understanding:

What they [publics] know about it [genomics technology] is false. It's like, ‘What do you know about soil microbiology?’ And they’re [publics] like, ‘I know nothing.’ But at least they don't pretend to know something! And that's our issue, that people think they know what they don't know. (Participant #7)

Despite the lack of knowledge, they explain, publics are often devoutly entrenched in their perspectives and those of their communities.

Participants remark that public understanding and acceptance of agricultural biotechnologies are influenced by other governance actors. They attributed legitimacy and value to scientific expertise and were skeptical of non-governmental and non-academic actors providing education on biotechnology and its implications. Participants tended to hold Greenpeace and similar organizations using ‘the anti-GMO playbook’ (which Participant #6 views as linked to the ‘anti-vax playbook’) in contempt, seeing them as lacking legitimacy and getting in the way of ‘progress.’ Several participants believed that public interest and advocacy groups motivate and benefit from public distrust:

A lot of the environmental organizations that I otherwise support very much have used it [GMOs] as a fundraising tool. So, I can't imagine that the NGOs that have been so successful fundraising by demonizing transgenics won't do the same thing with gene editing (Participant #4)

A lot of misconceptions are often fueled by … [hesitates] organizations that are trying to benefit from promoting that misunderstanding. […] A big one is the organic industry. (Participant #3)

The Greenpeace and the ‘organics industry’ were singled out in the interviews as threats to GECs reaching their potential. For example, they viewed non-GMO labelling as a marketing ploy for the organic industry to accumulate profits. Participants were also consistently concerned about GE labelling because it would likely confuse consumers. If there are no additional risks for GE, which the participants largely agree that there are not, then participants do not think GE foods need to be labelled.

Ultimately, participants are highly critical of the effectiveness and merit of public engagement in GE governance decisions, mostly due to their perceptions of insufficient technical and scientific understanding by the public. According to participants, for any effective public engagement to occur, the public must have a better knowledge of the technology to provide an informed perspective:

You have to have a good random sample of the public and see if you can educate them first, through seminars, like one or two. So, bring in volunteers for the day give the seminar in layman terms, or whatever, and then make them answer questions on how they feel. You can’t just go to the public and ask what they think. You are going to get all the wrong answers. (Participant #2, emphasis added)

There's a major responsibility of the government, when you put out a regulatory consultation, to provide the right background to that consultation to allow the public to have an informed say. (Participant #6)

Most participants struggled to articulate what good public engagement would look like, partially because it is beyond their training and expertise, but also because there are few examples to follow. Yet, participant #6 drew attention to the Australian Government’s GE regulatory renewal program as ‘the gold standard in government consultation with the public.’ Overall, participants viewed public engagement and the necessary education to be the responsibility of governments, not scientists.

Lacking trust for deliberative inclusion

The RRI principle of deliberative inclusion aims to include a diversity of actors and members of the public upstream in innovation and governance processes (Stilgoe, Owen, and Macnaghten 2013). As outlined above, participants are highly critical of the effectiveness and merit of public engagement in GE governance decisions, mostly due to insufficient understanding of genomic technologies and food system dynamics. Participants generally did not see the value of more or earlier engagement of other actors, including farmers, in the innovation process.

Across the interviews, there is a consistent theme of lack of trust and fear as obstacles for science and innovation. The lack of trust in science is linked to misinformation and public acceptance of technology:

If you don't have trust, the science doesn't matter. […] If you do have trust, the science doesn't matter. Because if they don't trust you, they won't trust your science. And if they do trust you, they don't need to see the receipts. (Participant #6)

Further, in reflecting on the public perception and politics of transgenics and earlier biotechnology, one participant stressed several times that it is essential not to ‘oversell’ or ‘over-promise’ the capabilities and benefits of GE in agriculture to maintain public trust.

Without trust in science and developers, fear and lack of knowledge are powerful forces in shaping public perceptions, especially in the context of social media:

Fear and lack of knowledge, right? It’s quite impulsive, you know. […] Snap chat or Instagram. Everything is a 10-second or 30-second story. And you can’t convince people in 30 s. No one is willing to go back to do the research to look at the problem. No one does anything. If someone comes in and say something in 30 s, then that is the truth. (Participant #2)

They argue that public opposition of technology gets in the way of progress because publics have so much power over governments and the regulations they set. Participants are highly critical of non-scientific perspectives and view the organics sector and interest groups (e.g. Greenpeace) as advancing misinformation and capitalizing on fear. However, participants did not display the same level of critical thinking and interrogation of values and assumptions (central to the RRI principle of reflexivity) when discussing science or their own views.

According to participants, responsible innovation and good governance of GE in agriculture require a ‘bridge between the science and the public.’ However, the participants reflexively view science communication as outside of their expertise and training:

Generally, scientists, we’re not trained in it [science communication, engaging with publics, etc.]. We're not good at it. Some of us are, but very few. And it's a tough battle. So, it's unclear to me whether we even have organizations set up properly to actually make those bridges, and to provide a robust framework for communication, and for making sure that the public is informed and making sure that there's feedback, and that kind of thing. If it’s left to the scientists, then we’re hopeless. […] That's not what we do. We're not a marketing and promotion firm. Right? That's essentially what it needs. You need to convince people like the Oscars. (Participant #7, emphasis added)

Returning to the idea of responsibility, participants viewed deliberative inclusion as the responsibility of governments. These scientists view themselves as ill-equipped for deliberative inclusion and do not view it as part of their job.

Discussion

Agricultural genomics experts wield power in food systems both in shaping technology and as advisers (or ‘stakeholders’) for policy and governance. Whether or not it is acknowledged, the conceptualization of technology (i.e. what it is and is not, how it is made) influences its governance and understandings of responsibility. Technological determinism, a common STS concept, describes the persistent story of technology as autonomous and outside of society (Bijker 2010; Leonardi and Jackson 2004). We view technological determinism as a continuum with varying degrees of recognizing social forces and human agency in sociotechnical systems, rather than the common binary interpretation (i.e. present or not, right or wrong) (Dafoe 2015). We are interested in the extent to which agricultural genomics experts exercise a deterministic understanding of technology and under what conditions they include interactions with social systems because of the impact on practicing RRI principles. Contextualized by the analysis of technological determinism, we critically examine the extent to which the participants exercise RRI and the implications for GEC governance (Table 2).

Table 2. Summary of interview analysis for RRI principles.

RRI Principles	Interview Findings: Agricultural Genomics Experts’ Perspectives
Anticipation	According to participants: GE offers many advantages for plant breeding and will impact the future of food – assumed inevitability. GEC benefits could be more equitably distributed than those of GMOs, but organic farmers are likely to reject GECs. There are minimal risks of GE applications in agriculture. GE tools can mitigate pre-existing plant breeding risks (e.g. more precise). ‘Overregulating’ GE will repeat the corporate concentration that occurred with GMOs. It is unlikely that existing genomic technologies (e.g. GMOs) will be supplanted or eliminated by GE. Participants exercise anticipation more than any other RRI principle, with specific insights on consequences of GE in agriculture and the regulatory context. However, their assumed inevitability of GE for future food systems drives forward a supportive-positive imaginary (foreclosing conflicting perspectives).
Reflexivity	According to participants: Interest and advocacy groups are benefit from fueling polarized discourse and misinformation. Governments are beholden to public acceptance. Science is objective and separate from politics. Participants are reflexive in stating the limitations of their knowledge and scientific expertise – but continued to assume their knowledge was objective.
Deliberative Inclusion	According to participants: The public has a very poor understanding of agri-food technology and hold their views strongly regardless. The organics sector and interest groups (e.g. Greenpeace) contribute to public misunderstanding and rejection of GM and GE technology. Science has a privileged claim to truth; largely uninterested in including other ways of knowing in GEC risk assessment and governance. Participants are highly critical of the effectiveness and merit of public engagement in GE governance decisions, mostly due to insufficient technical/scientific understanding. Participants view themselves as ill-equipped for deliberative inclusion and do not view it as part of their job.
Responsiveness	According to participants: Product-based regulatory approaches are more responsive and ‘future proof’ than process-based approaches (e.g. EU). The government is responsible for responding to changing social values and norms through public engagement (not scientists or developers). Participants did not view themselves as responsible for establishing the parameters of responsive governance, attributing responsibility to governments.

Technological determinism and discursive closure

The agricultural genomics experts interviewed called GE a technique or technology, which includes the material artefact (seeds or crops) as well as the knowledge and skills (identifying genes of interest and editing plant DNA). In this sense, interview participants exercise an understanding of technology that ‘includes both what things are made and how things are made’ (Grübler 2000, 20). Participants explain that GE techniques are linked to the improved knowledge of genomics through genome sequencing advances. However, the social-political context of the GE research and development, the specific applications of GE for agriculture, and their implications are beyond what counts as technology, for the scientists interviewed.

In anticipating the implications of GE for agri-food applications, the interviews featured a persistent sense of inevitability in technological development (i.e. GE is a natural progression of improving agricultural genomics) and as a necessary solution to food systems problems (i.e. GE must be applied to agriculture because of growing food insecurity and climate change). For example, the experts ascertained:

I would say this is probably the most revolutionary thing in the last 50 years. Absolutely. I can’t see this technology going away. […] I completely support gene editing as the future. (Participant #2)

This technology is continuously evolving, and more advances are happening to make it more precise and more sensitive, and it can be applied in a broader number of different plant species from the agriculture point of view (Participant #5)

The inevitability of technology is central to technological determinism, viewing technology as autonomously evolving and manifesting progress (Dafoe 2015). In this sense, participants are perpetuating technological determinism with important political consequences.

In her study of imaginaries and critical consciousness about digital futures, Annette Markham traces technological determinism as a ‘pattern of discursive closure’ (2021, 393) linked to the assumed inevitability of technological change and the lack of control over its directions. Similarly, agricultural geonomics experts viewing GE as an inevitable part of the future of food drives forward a supportive-positive imaginary for the technology and forecloses alternative discourses and material possibilities (Leonardi and Jackson 2004; Markham 2021; Smallman 2020). The narrative in the interviews also aligns with Middelveld and Macnaghten’s ‘promissory policy imaginary’ for GE applications (2021).

In this case of agricultural genomics and biotechnology in food, the contours of technology are linked to conceptualizations of food and food systems. This was most evident in the discussion of the unintended consequences of regulatory approaches to governing agricultural genomics introduced above. Despite the importance assigned to social, political, and economic contexts in determining the impacts of innovation, there is a theme in the interviews that concerns over corporate concentration and other power imbalances in the food system had nothing to do with science and technology. For most participants, technology and food are separate, and both are viewed as independent from the social and political contexts. For example, one participant became frustrated by the line of questioning on corporate power in the food system, insisting, ‘We are talking about food. We are not talking about who gives you the food.’ and later again, ‘But you are still talking about corporate issues. You are not talking about the food!’ (Participant #2). The underlying assumptions appears to be that once GE is applied to a seed and planted it becomes food – it is no longer a technology nor part of a political system.¹ Participants maintain a strict separation of food and technology from social systems, corporate power, and politics: ‘It’s all sociology; it’s not science here.’ (Participant #4). Again, interviews demonstrate a pattern of discursive closure, which limits the scope of responsibility.

Elsewhere, the interviews reveal mild social constructivism, especially when discussing anticipated benefits of GE, which we argue presents a window of opportunity for encouraging further reflection from scientists and engagement with the RRI principles. With an interpretation of technological determinism on a continuum, we can notice the nuances within the interview narratives. Despite enthusiasm for the technology’s affordances, there is a clear sense that applying GE will not be an all-encompassing solution to environmental problems facing the global food system. Interview participants explained most environmental benefits are not intrinsic to the technology, but rather tied to agronomy-focused changes within the seed or, more often, practices at the farm-level. Consequently, there is a tendency to externalize responsibility to other food systems actors, especially farmers. While far less prevalent in the interviews, some participants included responsibility for the impacts of GE to those with agency over its applications. For instance, one scientist explained: ‘You can use this [GE] for good or for ill. You can use it to make crops that are more resilient, but you also can make them ones that work better as a monoculture.’ (Participant #4). In other words, participants view GE as not inherently good or bad for the environment. Thus, under these conditions, the agricultural genomics experts exercised a mild social constructivist understanding of technology (Dafoe 2015; Pinch and Bijker 1989).

Likewise, the potential social and economic benefits of GE have less to do with inherent characteristics of the technology and more to do with the context in which it is developed, applied, and governed. Participants point to regulatory approaches as key factors determining the future social, political, and economic impacts of GE, both internationally and in Canada. This demonstrates an appreciation that the social and political context, as well as the individual users of the technology, will influence its impacts. Although this does not constitute a full abandonment of technological determinism, the participants anticipated impacts and understood the multi-directional regulatory challenges for GECs (presented above) in ways that opened discussion of GE beyond the deterministic divides between science and society.

Evaluating anticipation, inclusion, reflexivity, and responsiveness

There are relevant insights from the interviews for each of the four RRI principles (Table 2). Of all the RRI principles, participants were most comfortable practicing anticipation. Participants have well developed ideas of the intended impacts and benefits of GE for agriculture, as well as the possible futures the new technology enables, although their lessons from past unintended consequences of agricultural biotechnology are largely limited to their comments on ‘overly burdensome’ regulation for GM in service of the supportive-positive imaginary. As illustrated above, the participants’ anticipation exercises mild social constructivism in certain contexts (e.g. impact of GE is dependent on context, applications, regulations) and perpetuates technological determinism in others (e.g. GE is inevitable, firm boundaries between technology and politics). Certainly, the enactment of anticipation in the interviews does not satisfy the RRI goals of working through ‘tension between prediction, which tends to reify particular futures, and participation, which seeks to open them up’ (Stilgoe, Owen, and Macnaghten 2013, 1571) or broader STS imperatives that ‘⁠call for an anticipatory approach, receptive to the contingencies and always ongoing uncertainties of science and technology that are shaping – and being shaped by – a complex stochastic world’ (Montenegro de Wit 2020, 23). Further, there was no engagement with the ongoing colonial legacies of science as an institution. Nonetheless, the participants’ receptiveness to social and political considerations in the development and anticipated implications of GE is a step forward worth recognizing and a possible opening for further engagement.

The RRI principle of reflexivity demands critical introspection on one’s assumptions and areas of ignorance, as well as the implications of their values and perspectives (Stilgoe, Owen, and Macnaghten 2013). Participants exercised reflexivity in identifying potential assumptions and values of other actor groups, namely motivations for public interest groups to perpetuate misinformation on agricultural biotechnology. They highlighted the power of public perception over governance decisions. Several participants viewed governments as beholden to the anxieties and opinions of the public. Participant #4 described the European Court of Justice ruling on GE as ‘emotional’ and ‘impulsive,’ whereas the North American systems have ‘a scientific basis to their decisions.’ There are important differences in GE governance in different jurisdictions, but there is an evaluative judgement that the North American system, where the participants are located, is more objective and scientific. The agricultural genomics experts with whom we spoke were reflexive in willingly and explicitly acknowledging the limits of their perspectives and expertise. Several respondents confessed to not have given much thought to specific questions or issues prior to the interview. When questioned about regulatory approaches or how GE ought to be governed, participants shared their perspectives with humility. Nonetheless, interview participants were less willing or able to be reflexive about the underlying assumptions or biases and limitations of their views and scientific knowledge generally. In other words, participants could analyze the social construction of others’ perspectives, but not of science or their own perspectives.

Deliberative inclusion for RRI involves ‘⁠inclusively opening up visions, purposes, questions, and dilemmas to broad, collective deliberation through processes of dialogue, engagement, and debate, inviting and listening to wider perspectives from publics and diverse stakeholders’ (Owen et al. 2013, 38). The topic of public engagement in GE governance spurred conflicting responses from interview participants, but there is a consistent discomfort across the interviews regarding the power of misinformation and polarization in public perceptions of biotechnology in food. Further, the assumed inevitability of GE applications to address food system challenges, and hesitancy to take responsibly for technologies’ impacts, assigns technologies as autonomous and forecloses alternative discourses and material possibilities. Agricultural genomics experts hold a powerful claim to legitimate knowledge in the debate over biotechnologies, which influence the imaginaries of other governance and food system actors, as Smallman illustrates in her study of UK-based national policymakers (2020).

Finally, there were important insights on responsiveness. Participants’ exercises of RRI’s principle of responsiveness are limited by the disinterest in deliberative inclusion. They attribute the responsibility for responding to public engagement, risks, and public interests to governments. Still, participants were quite concerned about responding to lessons from past agricultural biotechnology, both for regulation and interactions with publics. They believe that responsive governance model must accommodate technological change and the dynamic interactions with ecological, social, and political systems. If the regulatory treatment of GECs is consistent with GMOs, then participants worry that the same impacts will occur, namely corporate concentration and stifling innovation.

There is extensive social science research on the consequences of corporate power over the food system, especially for agricultural input companies (Clapp 2018; Gibson 2019; Weis 2007; Magnan 2014). Critical food systems scholarship and social movements tend to call for stricter regulation for a more democratic distribution of power and capital associated with agritech innovation. Yet, according to agricultural genomics experts, strict ‘overly burdensome’ regulations for geonomics technologies in plant breeding resulted in a concentration of technological capacity and profits in the hands of a small group of powerful corporations for GMOs, and the same is likely to occur if GE governance takes does not change its approach. For participants, it is not the particulars of either technology that determine its risks and political economic impact, but rather the regulatory context in which they exist, with an exercise of mild social constructivism. In addition, participants feel that engagements with publics on the topic of GE in agriculture should learn from the challenges of GMOs, but they do not view themselves as capable of or responsible for educating and including publics in decision-making and governance.

Conclusion

RRI approaches iteratively connect diverse social actors in the purpose, process, and products of research and innovation to prioritize ethical, ecological, and social goals. Practical and institutional operationalizations of RRI principles, united by a vision of collective responsibility, require a clear understanding how those researching and developing the technology understand the complex implications of their products. Agricultural genomics experts’ views on GE for agriculture will influence governance and public discourse.

Responsible GE innovation and governance demands integration of all four principles (See Tables 1 and 2). Participants exercise anticipation more than any other principle, although they do not satisfy the RRI ideal. Participants anticipate that benefits of GECs outweigh any risks and predict that strict regulatory approaches, with costly approval processes, will limit innovation and further corporate concentration. For these scientists, risks and benefits of GE for agriculture have little to do with the technical characteristics of the GECs – the area of participants’ expertise – and more to do with social, political, and economic context. Thus, participants demonstrate a conditional rejection of technological determinism in some contexts but maintain a view of science and technology as separate from social and political systems and assume the inevitability of GE. Participants identify public misunderstanding and the influence NGOs and other actor group on public perceptions and government decisions as challenges for GEC governance. Our findings correspond with the broader literature on expert views on gene editing (Kato-Nitta et al. 2019; Lassoued et al. 2021; Qaim 2020).

Participants exercised reflexivity when they identified how values and assumptions influence other actors’ perspectives and decisions but were less inclined to think critically about the foundations and limitations of their own perspectives or those of scientific knowledge more broadly. Generally, participants did not see the value of deliberative inclusion of diverse groups or knowledges (e.g. consumers, farmers, etc.) and were therefore limited in their ability to exercise responsiveness. They demonstrated a pattern of ‘discursive closure’ elsewhere, such as the assumed inevitability of GE, which limits the scope of responsibility. Likewise, the conditional technological determinism that undergirds agricultural genomics experts’ perceptions acts as a barrier to dialogue and distributed responsibility when there are conflicting assumptions of actor groups.

Overall, participants exhibited some degree of ‘RRI readiness’ (Eastwood et al. 2019; Regan 2021), but do not feel responsible for or capable of meeting RRI expectations, especially for deliberative inclusion or responsiveness. Participants view public engagement and risk mitigation as government responsibilities. This poses a real challenge to enacting the collective responsibility envisioned by RRI (Owen et al. 2013, 2021). Moreover, the power dynamics and legitimacy assigned to the views of scientists (even on matters beyond the scope of their expertise, such as regulation and politics), as well as publics’ distrust of science and scientists’ distrust in publics, are obstacles to deliberative inclusion for the governance of GE.

Scientists have a valuable point of view in understanding the history of agricultural biotechnology. Despite a growing lack of trust in science from publics, technoscientific viewpoints are powerful forces for policymakers and government operations (Smallman 2020). Yet, understanding the risks and benefits of GE for agriculture should not be limited to technical assessments from developers, whether they are in academic, public, or private institutions. As Hartley and colleagues explain, ‘when the debate about agricultural biotechnology is confined to a technical assessment of risks to human and environmental health, it limits who can legitimately participate in decision-making processes, privileging technical experts’ (2016, 4). For instance, NGOs provide important insights on the social and ethical dimensions of agricultural biotechnology (Gordon et al. 2021; Helliwell, Hartley, and Pearce 2019). An RRI approach demands the inclusion of a diversity of actors and ways of knowing.

RRI avoids linear ‘knowledge deficit’ education from experts and expands the responsibility of scientists beyond novel contributions to knowledge, research ethics, and academic integrity (Barnhill-Dilling and Delborne 2021). Those studying and innovating technologies have an obligation to consider and respond to societal impacts of their developments (Roberts, Herkert, and Kuzma 2020; Smallman 2020). For example, Glenn Stone suggests that scientists ought to ‘act as ‘honest brokers’ to help educate, enrich debate, and inform policy’ (2017, 1) and that their knowledge, skills, and resources position them well to take this role. However, based on our interviews with agricultural genomics experts, we are reminded that the responsibilities that Stone (2017) and others (e.g. Georges and Ray 2017) ask of scientists will not come easily to most – it is not their training, nor are they culturally adapted to doing such work. Nonetheless, as we argue below, changing the role and responsibilities of science and scientists is an idea worth pursuing.

While there can be a role for individual scientists, our interest here is to advocate for collective responsibility to ground the entire structure of science (e.g. universities, graduate and post-doctoral training, funding agencies, and regulatory bodies) to engage in deliberative inclusion and research that responds to social and environmental goals. One example is the Innovative Genomics Institute (innovativegenomics.org), an academic research institute that produces educational resources for a diversity of audiences (e.g. sharing protocols and reagents for other researchers, CRISPR Made Simple accessible for children, the practical CRISPRpedia textbook) and aims to guide society in the ethical use of GE through new evidence-based approaches in science communication and public engagement (Doudna 2021). Or, instead of redefining the job of a scientist, it may be useful to create a new type of research professionals who convene deliberative inclusion and can serve as interlocutors between the scientific community, publics, and policy, like the professionalization of ‘science communication.’

Ultimately, responsible governance of GE will require proactive decisions on regulatory frameworks that satisfy the requirements for human safety and environmental sustainability, while maintaining space for decentralized and democratic technology development and equitable distribution of risks and benefits. RRI offers theoretical and practical tools to engage with the governance of science and technology, acknowledging the uncertainties prompted by changes in socio-technical systems and the importance of co-constructing more socially just and environmentally sustainable futures.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by Genome British Columbia: [Grant Number Societal Issues Grant SOC 0009]; Social Sciences and Humanities Research Council of Canada: Joseph-Armand Bombardier Canada Graduate Scholarships - Doctoral [Grant Number].

Notes on contributors

Sarah-Louise Ruder

Sarah-Louise Ruder is a PhD Candidate at the University of British Columbia's Institute for Resources, Environment, and Sustainability. Interdisciplinary by training, Sarah-Louise is an environmental social scientist studying food and agriculture in Canada. She earned a Master of Environmental Studies degree and an Honours Bachelor of Science in Environmental Science from the University of Waterloo and Queen's University, respectively. Her research explores transitions to more sustainable, food secure, and just food systems and the politics of novel agri-food technologies.

Milind Kandlikar

Milind Kandlikar(PhD Carnegie Mellon) is a Professor at the Institute for Resources, Environment and Sustainability, and the School of Public Policy and Global Affairs at the University of British Columbia. His work focuses on the intersection of technology innovation, human development, and the global environment. Dr. Kandlikar's current projects include the regulation of agricultural biotechnology including implications for food security; the global energy transition and its equity implications; and the impacts of climate change in the developing world. He has also published extensively on the science and policy of climate change.

Notes

1 Despite this deterministic assumption in the conceptualization of what is ‘food’ and its independence from political, economic, and social systems, participants did not express ‘genetic determinism,’ which reduces organisms to their genetic code – see Kitcher’s ‘Battling the undead: How (and how not) to resist genetic determinism’ (2001). The participants discussed GECs with an appreciation of the differences between genotype and phenotype and other enactments mild social constructivism.