The legislative effect of ignorance: institutional risk and the regulation of new genomic techniques

Abstract

Recent innovations in biotechnology, above all the use of genome editing like CRISPR/Cas in plant breeding, have led authorities across the globe to question the adequacy of regulatory policies. Focusing on the situation in the European Union, this article claims that the inability to establish whether new genomic techniques (NGTs) were used in breeding a particular crop functions as the main driver behind the dawning policy changes concerning the use of NGTs in plant breeding. This ignorance takes the form of an institutional risk: as authorities are unable to execute the regulation, they are entering a regulatory impasse. Drawing a comparison to the writings of German sociologist Heinrich Popitz, who spoke of the ‘preventive effect of ignorance’, this article explores the ‘legislative effect of ignorance’ as the main driver of policy change in the NGT case and attempts to draw general conclusions for similar situations in science and innovation policy.

1. Introduction

This article contributes to the intellectual debate on institutional risk, i.e. events that do not pose a danger to human health, safety, or the environment (societal risk), but to those organizations entrusted with the management of societal risk. Taking inspiration from the work of German sociologist Heinrich Popitz, it explores a particular strategy for dealing with institutional risk. Popitz described how ignorance is strategically used by authorities to maintain the stability of the normative order they are in charge of—the preventive effect of ignorance, as Popitz called it. Against this background, this article describes a second preventive effect of ignorance. In situations where the institutional risk is caused by an inability to determine whether or not a given thing is the result of a regulated technology, and the regulator thus factually lacks the means to execute its regulations or laws, regulators are pressured into safeguarding the norm system by adapting the legislation. This second preventive effect can thus also be called the legislative effect of ignorance.

To explore the legislative effect of ignorance, this article uses the ongoing policy debate in the European Union (EU) around what came to be called new genomic techniques (NGTs), i.e. the use of genome editing (GE) techniques in plant breeding. Technology and innovation policies in this field, and the accompanying scholarly debates, have conceived of ignorance mainly in terms of societal risk. In particular, this perspective is encapsulated in the precautionary principle of both the Rio Declaration on Environment and Development adopted in 1992 and later the Cartagena Protocol on Biosafety adopted in 2000, which provide important reference points for most national and supranational regulatory approaches.¹ While the precise formulation of the precautionary principle has changed over the years (Bernauer 2003; Magnus 2008), all versions alike posit that regulators should proceed in consideration of possible negative effects of (technology) policies on health, safety, and the environment, even when such effects are not established with definite scientific certainty.

Most discussions on how to regulate NGTs have continued the focus on societal risks that had already dominated the debates on genetic modification (GM). In contrast, the argument developed in this article suggests that the currently dawning policy changes in the EU are mainly motivated by institutional risks and thus are a manifestation of the legislative effect of ignorance. This has to do with the widespread consensus among plant scientists that changes in the genome achieved by using GE-based breeding technologies could also be achieved by conventional breeding methods—the advantage of GE being only the increased speed of producing results. Accepting this view, it must be concluded that there are no particular societal risks involved in the use of those NGTs that go beyond those of conventional, and hence legally allowed breeding techniques.

The claim put forth in this paper is that the current efforts in the EU to revise the regulations of genetic engineering in plant breeding are motivated not by societal, but by institutional risks—that these efforts are taken in order to avoid a situation where EU regulators are not able to reach their organizational and administrative objectives. This situation is brought about by a specific ignorance: NGTs allow for relatively precise modifications in the genome, and most of them leave no traces, which is why it is by and large impossible to detect in a given plant or product whether NGTs were used in breeding.

To capture theoretically how this form of ignorance motivates legislative efforts, it is useful to look for synergies between the sociology of ignorance and the literature on risk. Section 2 traces back the theoretical debate on risk and ignorance that led to the differentiation between societal and institutional risk, and Section 3 introduces Heinrich Popitz’s idea on the (first) preventive effect of ignorance. The ensuing Section 4 briefly describes the precautionary principle and how its focus on societal risk has informed the legislative treatment of genetically modified organisms (GMOs). However, the advent of techniques like TALEN or CRISPR/Cas² challenged the premise of these regulations that the use of biotechnologies in plant breeding can be clearly and objectively determined. As described in Section 5, these advances in knowledge and technology provided new energy to the policy debate on agrobiotechnology in the EU. While the transformation of the regulation is not yet finalized, it can still be argued that the main motivation behind the expected legislation change is to move out of what appears to be a major institutional risk, i.e. a situation where the regulator lacks the means to execute its own rules. Based on these considerations, Section 6 takes a few steps towards theorizing what can be called the legislative effect of ignorance.³

In terms of research materials, the paper relies, first, on an analysis of the core policy documents relating to GMOs and NGTs, and the accompanying debate in the scientific literature. Secondly, it is based on conversations that happened over a series of interviews and workshops designed to explore the positions of European stakeholders—both at the national and EU levels—along the value chain of a particular crop. The theoretical interpretation developed in this paper is supported by the experiences collected over this 4-year process of stakeholder involvement.⁴

2. Ignorance and risk

The role of ignorance in society has been of continued interest to science policy scholars and practitioners.⁵ The debate received a major push in the publication, in the years 1986 (German) and 1992 (English), of Ulrich Beck’s Risikogesellschaft [Risk Society]. The book fuelled the debate for two reasons: first, Beck’s core idea that risk was not just a problem for regulators, but was about to become the shaping principle of a new form of society, turned out to be highly attractive to many contemporary writers. While class society, Beck argued, had been oriented towards the shared ideal of equality, risk society’s ‘basis and motive force […] is safety’ (Beck 1992: 49, emphasis in original). This resonated with authors who critically observed the proliferation of risk management techniques into ever more areas of social practice (Jasanoff 1990; Power 1997, 2004; Rothstein, Huber, and Gaskell 2006). These authors shared Beck’s view that safety was a complicated concept, encompassing both ‘objective’ and ‘subjective’ layers of meaning. Debates around safety and risk, they observed, quickly turned into debates about the social construction of ‘objective’ and ‘subjective’ forms of knowledge and about the procedural power of risk definitions. ‘[W]hat becomes clear in risk discussions are the fissures and gaps between scientific and social rationality in dealing with the hazardous potential of civilization’ (Beck 1992: 30, emphasis in original). The incompatibilities between scientific and social rationalities relativized the alleged objectivity of the sciences. In a paradoxical twist, this led to even higher demand for risk management techniques and their continued spread to all kinds of social institutions, a process that Rothstein, Huber, and Gaskell (2006) acutely described as ‘risk colonisation’.

A second reason why Risk Society had such an enormous—and ultimately fruitful—impact on the scientific discourse on risk can ironically be found in Beck’s own neglect of the then-extant literature on risk. His first interest was in the role of knowledge, and non-knowing, in society. As he recalled, it was only after the publication of his book that he became aware of how the concept of risk was used in neighbouring scientific fields. ‘[T]he concept of risk in academic literature’, Beck said in an interview, ‘basically means that which I specifically would not express with the word […]. Risk means calculable uncertainty. While with “risk society” I was specifically getting at non-calculable uncertainty’ (cited in Sørensen 2018: 6). Beck’s lack of awareness of the state of the art in risk studies opened up a space for an intensified exchange between scholars of risk from various disciplinary backgrounds on the one hand and sociologists and philosophers interested in ignorance on the other. This space allowed all sides to question the other’s assumptions as well as their own, which provided a major push in terms of conceptual and theoretical sophistication.

In the course of this debate, one direction of analysis focused on the social, cultural, or political consequences a lack of knowledge can have in terms of risk. Ignorance, as these works showed, could be strategically used by various political and economic actors who aim to hide or manage risks (Proctor and Schiebinger 2008; Oreskes and Conway 2010; McGoey 2019). Here, but in other works as well, many conceptual and theoretical innovations have been proposed, e.g. by discerning different forms of ignorance.⁶ For this paper, the most important of these conceptual innovations is the differentiation between societal and institutional risk (Rothstein, Huber, and Gaskell 2006). The insight motivating this differentiation is that the widespread currency of risk as organizing principle within society also puts into a vulnerable place those institutions charged with the regulation and management of risks. Whereas societal risks concern the health and safety of humans as well as the environment, an institutional risk has been defined as ‘the risk that the regulator will not meet its organizational and policy objectives’ (Black 2010: 328; see also Merkelsen 2018). This second form of risk is of particular relevance for the case explored in this paper, the pressure on the EU legislators to provide a (new) regulation for NGTs.

3. Heinrich Popitz and the preventive effect of ignorance

While not using the term, the phenomenon of institutional risk has already been explored by German sociologist Heinrich Popitz (1925–2002). In 1968, Popitz published a small book in which he laid down his views on the relation between strategic organizational behavior, ignorance, and what came to be called institutional risk. Entitled ‘Die Präventivwirkung des Nichtwissens’ [The preventive effect of ignorance] (Popitz 2006 [1968]), Popitz’s text exposed central assumptions of role theory, the sociology of crime, but also of the then-dominant structural-functionalist paradigm as unrealistic and ultimately erroneous. Then, as today, a widely heard argument was that negative sanctioning, i.e. punishment, of deviant behaviour was an effective means for a society to ensure that the norms it had set would be followed. Societies, it was argued, had to ensure that as many norm violations as possible were detected and punished because otherwise the validity of the norm would erode.⁷

Popitz claimed the opposite. He argued that norm systems could not be upheld if it were known how often norms were broken. This ignorance protected the norm from being challenged, thus providing what Popitz called the preventive effect of ignorance.

To corroborate his thesis, Popitz developed three arguments. First, he noted that no one will ever be able to have complete knowledge of all relevant information of even a single individual. We interact with our fellow human beings on the basis of a certain amount of information, but this is only a small part of the available information on the counterpart which is intermingled with past experiences, assumed likeliness, and unquestioned prejudices. ‘As a matter of fact, it is not simply a certain amount of information about other people that we use to orient ourselves, but psychological types, character images that we construct of each other’ (Popitz 2006: 161, translation by the author). What is claimed here for interpersonal interaction also applies to collectives: apart from the fact that we would heavily object, no state or corporation can ever have complete information about us. The desire for complete information runs up against the limits of feasibility—psychological, sociological, organizational, and technical.

Secondly, Popitz argued that complete information would not only be unattainable, but actually harmful. This is because complete information would, of course, include information about all norm-breaking. If it were known how often people violated the prevailing norms, then the willingness of norm-followers to abide by the rules would rapidly decline. A system of norms could not withstand gapless information and would erode.

Thus, if one wants to protect a norm system, it is advisable to devise mechanisms that allow actors to overlook certain cases of norm violation. On the individual level, this is often done by deliberately looking the other way. If superiors want to achieve stable compliance with norms on the part of their employees, they must not pettishly punish every instance of deviance or disregard. As Popitz pointedly puts it, ‘A superior who cannot play dumb in such situations is it’ (Popitz 2006: 167, translation by the author). Something similar is true in the education of children: not every rule violation can be punished, and the educator must decide to overlook some.

Forms of looking the other way also exist at collective levels, Popitz argued. Systems of norms and sanctions are almost inevitably rigid in our modern societies because the claim made on them is that they establish regularity and certainty of behavior and often also guarantee equality. This rigidity, however, contrasts with the vagueness and unpredictability of social reality. Therefore, a rigid system of norms and sanctions needs a relief option—mechanisms of overlooking or ignoring norm violations. Limiting information influx is one of these mechanisms. If we do not look closely, we do not know everything—and because we do not know everything, the image of our fellow human beings and our normative system is better than it should be factually. This image is crucial to our adhering to norms.

Popitz also developed a third argument against the position that rigorous sanctioning of every single norm violation would contribute to the stability of the norm system. This went beyond the organizational limits, i.e. the argument that judicial and penal administrations would have to grow immeasurably in terms of personnel in order to execute the increased number of sanctions. Rather, Popitz argued that consistent punishment of all norm violations would also have a negative effect on the validity of the norm system, an effect established convincingly within psychological learning theory. The more often a punishment occurs, the smaller its differentiating, discriminating and thus deterrent effect. In sociological terms, if most of my acquaintances have already been punished once for a certain offense, I am no longer afraid of losing their respect if I am caught committing the same offense. The protective function that comes from the threat of punishment thus depends on the punishment itself not being imposed too often.

The strategic value of the preventive effect of ignorance that Popitz describes arises out of ignorance’s capability to restrict the institutional risks that would come with public knowledge of the quantity of norm breaches. Such knowledge would undermine the authority of the regulatory organization, eventually rendering it incapable of achieving its own objectives.

4. The precautionary principle and its role in GM regulation

Understandably, the debates on how to regulate and manage the risks of innovations in biotechnology that occurred over the last decades have focused on societal, not on institutional risk. In these debates, the precautionary principle has increasingly risen to become the major guiding idea. While, as described later, different versions of the precautionary principle exist (Sandin 1999; Manson 2003; Magnus 2008), it can be said that its objective is to steer political decision-making by putting weight on a worst-case scenario of risk. Being first used in German environmental legislation, the precautionary principle took inspiration from philosopher Hans Jonas’s (1903–1993) writings. In The Imperative of Responsibility, Jonas ([1979] 1985) argued that the presence of man in the world had turned from an unquestionable given into an object of obligation. Humanity had ‘to preserve the physical world in such a state that the conditions for that presence [of men] remain intact; which in turn means protecting the world’s vulnerability from what could imperil those very conditions’ (Jonas 1985: 10).

At least two elements of this statement are mirrored in later formulations of the precautionary principle. First, the matter of concern is the ‘physical world’. To Jonas, technological progress has created a world in which human-made risks, understood as damaging effects of technology use on human health and nature, are ubiquitous. Efforts to protect human health and safety as well as the environment, that is measures to control societal risks, become a moral imperative. As the human-created risks may engender negative impacts across society, they are to be addressed by public law.

The second aspect that found its way into the precautionary principle is the word ‘could’. Technology and science regulations cannot always rely on scientifically established certainties, but have to work on the basis of potentialities. Regulations must have a preventive character, avoiding harm to society and the environment without being able to determine the likelihood of the harmful event.

The first international legal document to mention the precautionary principle was the Rio Declaration on Environment and Development from 1992. The declaration states: ‘Where there are threats of serious or irreversible damage, lack of full scientific certainty shall not be used as a reason for postponing cost-effective measures to prevent environmental degradation’ (United Nations General Assembly 1992: 3). In other words, when it comes to the environment, scientific ignorance does not justify inaction in terms of risk mitigation measures.

Later pieces of legislation concerned with societal risks, however, rely on a different interpretation of the precautionary principle. The glossary of the EU law database, for instance, suggests a quite defensive understanding: ‘[I]f it is possible that a given policy or action might cause harm to the public or the environment and if there is still no scientific agreement on the issue, the policy or action in question should not be carried out.’⁸ Here, the precautionary principle encourages policymakers to wait. When environmental and human health risks cannot be excluded, lack of scientific certainty can be taken to justify inaction on behalf of the regulator and deliberately delaying the introduction of a new regulation or the adoption of a particular technology.

These examples show clearly that the precautionary principle conceptualizes scientific ignorance in the form of societal risks. Beyond this, however, it also becomes clear that it implies a ‘temporalized’ (Bauman 1991) understanding of ignorance; risks should be acknowledged as ignorance that one day can, and most likely will, be replaced by knowledge; an understanding according to which, as Beck and Wehling (2012: 36) succinctly wrote, ‘nothing in the world is held to be unknowable and hence the triumph of knowledge is only a matter of time’.

The different definitions and interpretations of the principle, however, gave rise to quite different regulatory approaches across various technological and environmental domains (Fisher, Jones, and Schomberg von 2006). This also concerned the regulation of GM crops. The Cartagena Protocol on Biosafety to the Convention on Biological Diversity, which regulated for the first time in a binding manner under international law the cross-border transportation, handling, and treatment of GMOs and came into force in 2003, explicitly referred to the Rio Declaration (Secretariat of the Convention on Biological Diversity 2000). But on the national and supra-national level, different approaches prevailed (Bernauer 2003; Magnus 2008; Sprink et al. 2016; Bogner and Torgersen 2018). Regulations that are more open towards the cultivation of GM crops can be found in the Americas, in Australia, and in certain Asian countries (Yamaguchi, 2007; Yamaguchi, et al., 2003; Yamaguchi and Suda, 2010). In these regions, the cultivation of GM crops has grown over the years, with figures from the pre-pandemic period showing an overall increase in cultivation area from 1.7 million hectares in 1996 to 190.4 million hectares in 2019 (ISAAA 2019). In contrast, Japan, New Zealand, Norway, Switzerland, and the EU agreed on more restrictive regulations and effectively counteracted a similar increase in cultivation on their territories (Spök et al. 2022; Yamaguchi, 2020).

5. Regulating plant breeding techniques in the EU

The situation discussed in the preceding section forms the background against which the EU has been developing its own regulatory position towards the use of GM techniques in plant breeding. As a supra-national entity, the EU strives to develop and maintain harmonized legislation across all Member States. Yet, the EU is not a centralized political system, but rather operates as a multilevel system where a considerable share of power remains with the Member States. Since the early 1990s, the process of establishing and further developing a harmonized EU legislation on GMOs has triggered and—in turn—has been strongly influenced by opposition and campaigns of an influential alliance of civil society organizations and parts of the farming community (Levidow and Marris 2001; Levidow, Murphy, and Carr 2007; Schurman and Munro 2010). In some Member States, these alliances became broader and more powerful and succeeded to ally with retailers, food producers, green parties, and media (Bernauer and Meins 2003; Seifert 2013, 2021; Weimer 2019). All this was fostered by a long-lasting narrative of GM crops that combined three claims: they would pose a risk to health, were potentially damaging the environment, and had no clear benefits for consumers.

Out of this controversial situation emerged Directive 2001/18/EC, which, in its amended form from 2015 (European Parliament and Council 2015), still is the key legal document regulating GM plants in the EU. Of course, the Directive was repeatedly challenged, as with new developments in molecular plant breeding techniques, the discussion reopened on whether or not it would fall under the Directive’s scope. This was the case with the introduction of the differentiation between transgenic and cisgenic plants (Schouten, et al., 2006; Jacobsen and Schouten, 2008), where it was proposed to exclude cisgenic plants from the Directive. The argument was that due to the fact that cisgenesis by definition involved only the transfer of genes from organisms that can be successfully crossed by conventional methods, cisgenic organisms posed less risk than transgenic organisms. While some shared this view (European Food Safety Authority 2012), others argued against it (Schubert and Williams 2006). In the end, EU authorities decided that the Directive indeed covered cisgenic plants and that they had to be regulated according to the procedures defined there (Dayé et al. 2023).

Another occasion to question the scope of the Directive came up with the use of site-directed nucleases (SDNs) in GE. CRISPR-Cas technology is presumably the most widely known of these, but there are others that offer comparable features (Ahmad et al. 2021). These techniques allow for minor modifications in the genome (e.g. small insertions or deletions of bases). Compared to both conventional breeding methods and earlier GM techniques, GE also allows for a considerably higher degree of precision, meaning that there are much fewer unintended changes in the genome (off-target effects), which are always an issue in safety assessment. Finally, with GE, the creation of new varieties can be realized in much shorter periods of time than with conventional breeding techniques. While, depending on the plant and how much scientific knowledge is already available, plant scientists usually need 15 years to create and stabilize a new trait in a plant when using conventional breeding methods, GE techniques allow for doing this in 2–3 years.

Again, proponents argued that some types of SDNs (SDN1 and SDN2) might fall out of the scope of the Directive, since they are technically similar to mutagenesis in conventional breeding, which is excluded from the scope of the Directive. On 25 July 2018, however, the European Court of Justice clarified that these new techniques do not qualify as ‘mutagenesis’ in the meaning of the regulation, mainly because unlike the type of mutagenesis the legislation is referring to, the new techniques had not been studied sufficiently to rule out possible health and environmental risks (Court of Justice of the European Union 2018; Dederer and Hamburger 2022; Gelinsky and Hilbeck 2018; a diverging interpretation is put forth in Van der Meer, 2023).

The Court of Justice of the European Union (CJEU) ruling refuelled the debate on how to regulate GE in plant breeding (cf. Eriksson et al. 2019). Reacting to the surprise and growing discontent among certain stakeholder groups with the CJEU ruling, the European Council (2019) asked the European Commission to initiate a study on the impact of the ruling. Published in Spring 2021, the EC study on new genomic techniques attempted to balance the perspectives of EU Member States and civil society, as well as economic, social, and ethical considerations against the backdrop of available scientific knowledge (European Commission 2021). The study emphasized that in the light of scientific and technological progress, the current legislation was not adequate anymore. Furthermore, it introduced a neologism into the debate: new genomic techniques, or NGTs. NGTs were to be discerned from ‘older’ GM methods, which remained covered by the Directive 2001/18/EC. The study also highlighted the potential contribution of NGTs to the transformation towards sustainable agriculture as described in the European Green Deal. Yet, it also pointed out what had been discussed immediately after the publication of the CJEU ruling: the inability to distinguish, by means of reliable laboratory tests, whether a genetic modification had been induced by NGTs (requiring safety assessment and authorization procedure according to the Directive) or had occurred naturally (allowing for being put on the market without such procedure). The lack of such tests made disruptions in international trade possible (Wasmer 2019).

Building on the study, the European Commission published a ‘Proposal for a new Regulation on plants produced by certain new genomic techniques’ on 5 July 2023.⁹ Adopting the regulation requires a series of readings and debates by both the European Parliament and the Council of the EU, which have an equal say in the process and may both agree on it after amendments. The Commission explicitly mentioned reaching an agreement on the new regulation as an objective of their 2024 work programme (European Commission 2023: 8), but that was not achieved. However, the current Polish presidency of the Council of the EU (January–June 2025) committed itself to ‘continue work on the draft Regulation on new genomic techniques […], bearing in mind the challenges of climate change and innovation in agriculture’ (Council of the European Union 2025: 44). This formulation suggests that the presidency is not overly optimistic about being able to finalize negotiations within its 6-month period, but NGTs are still on the agenda.

6. Theorizing the legislative effect of ignorance

By the time the CJEU published its 2018 ruling, important agricultural countries in the world, particularly in the Americas, had already established regulations that were more favourable towards the use of GE in plant breeding; others were in the process of moving in a similar direction (Sprink et al. 2016; Menz et al. 2020; Salt 2023). With NGTs like CRISPR/Cas systems addressing only very minor parts of the genome, the control of imported crops became impossible. Some of the GE procedures leave no traces at all in the plant genome. If they do, then these traces very often disappear over several cycles of back-crossing, which are made by plant breeders to stabilize the new traits. Even if the changes can be detected (detectability), it is technically not possible to discern them from ‘natural’ mutations (identifiability). In most cases, the tests used today to determine whether or not the genome of a particular organism has been technically modified are not sufficiently sensible. Moreover, there is a wide consensus among scientists that it is analytically impossible to identify such changes as resulting from the use of GE techniques, and not from natural mutation (Grohmann et al. 2019; Schulman, Oksman‐Caldentey and Teeri 2020).

In a world of international trade, with an increasing number of states and regions opting for less restrictive regulations of some GE crops relative to those of the EU, this put a major question mark over the claim of several EU countries to keep their fields free of GMOs. Furthermore, although less openly acknowledged, it endangered the credibility of organic farming labels, which by definition have to be GMO-free and now come under threat of ‘contaminations’ from both imported seeds and food ingredients (Gocht et al. 2021; Zimny and Sowa 2021).

It seems more plausible to assume that the pressure resulting from this ignorance—i.e., from the inability to determine whether or not a crop was bred using a particular technique—was more important in motivating the current developments within the EU than presumed interventions by lobbying groups or economic actors (Hartung 2020) or a change in public opinion—lobbying has been rather stable for several decades without pushing the EU to revise the legislation (although there are interesting recent initiatives; cf. Spök et al. 2022), and a change in public opinion also could not be observed in the EU (survey data on this issue is outdated, but see Woźniak, Tyczewska and Twardowski 2021; Spök et al. 2022).

With the 2018 CJEU judgment, the EU was heading towards a situation where ignorance—more precisely the inability to know—clashed with the knowledge requirements of the legislation. The CJEU judgment was understood to extend the regulation’s scope to include NGTs. But at the same time, the regulation assumed that regulators were able to assess for a given crop whether or not the NGTs were used in the breeding process. Now, regulators were put into a situation where they did not possess the means to independently determine into which category a particular crop belonged. German legal scholars have a name for this situation; they describe it as Vollzugsnotstand, a state of executive emergency. There is a legal norm, but it cannot be executed because the premises required for its execution do not exist.

Provided that the way to move out of this state is to negotiate and pass a new regulation, it seems justified to understand the current situation in the EU as manifestation of a legislative effect of ignorance. The legislative effect gains its force from the perception of an institutional, not a societal risk. As mentioned earlier, there seems to be a consensus that GE crops pose more or less the same risks to human health and environment as crops bred with conventional techniques (European Commission 2021). The legislative effect emerges from a particular ignorance related to a recent technological innovation. Unlike with societal risk, where ignorance concerns possible future effects and thus is subject to the same uncertainties that come with any form of ‘foreknowledge’, the ignorance in this situation of institutional risk concerns the inability to apply the categorization on which the regulation is based. This ignorance is not caused by the general empirical unavailability of the future, but by the present lack of a stable method.

Clearly, in its intention, the precautionary principle is not addressing institutional, but societal risk. Yet, with a little stretch, the formulation provided by the EU might be interpreted as also being applicable to institutional risk: ‘[I]f it is possible that a given policy or action might cause harm to the public or the environment and if there is still no scientific agreement on the issue, the policy or action in question should not be carried out.’¹⁰ If an executive state of emergency is understood as harm to the public, and the consequence of the CJEU ruling as policy, then the legislative effect of ignorance that is described in this paper could also be said to result from the precautionary principle.

Going beyond this particular dynamic, there are several aspects of the current attempts to create a regulation for NGTs that relate well to positions developed within the sociology of knowledge, ignorance, and risk. First, we can see the case of regulating NGTs in the EU as an instance of Beck’s assertion that instead of knowledge, ignorance was the crucial driver of societal transformation in the post-industrial society: ‘Not knowing, but non-knowing is the “medium” of reflexive modernization’ (Beck 1996: 298, translation by the author). Of course, the juxtaposition of knowledge and non-knowing is not straightforward, for at least two reasons. On the one hand, Beck (1994) himself argued against the linearity implied by such theories of knowing; on the other hand, as has been repeatedly observed, the creation of new knowledge also brings with it the identification of new things we do not yet know (Wehling and Böschen 2015b). This complication aside, one mechanism through which ignorance gains the transformative force that Beck is thinking of is, of course, risk.

In societies that show a continuous and high-speed transformation of their socio-technical arrangements, new knowledge also creates new ignorance, and this ignorance can be a source of uncertainty and risks, creating a dynamic that has come to be called the ‘cascade of uncertainty’ (Gross 2010: 3). When societal risks are concerned, ignorance concerns the effects of innovation. However, innovations can also overstrain legal and regulatory frameworks, leading to institutional risks. In such cases, ignorance impacts the ability of the regulators to relate the basic categories of the law to the empirical realm. In order to know how to regulate X, there is a new to know in which category X belongs. This ignorance might be a fundamental ‘inability to know’ (Beck 1996: 302, translation by the author), a state of ignorance where we have ‘knowledge about the limits of knowledge in a certain area’ (Gross 2007b: 751). But even if we see the opportunity that one day we might be able to know, the pressure to (re-)establish the regulator’s ability to carry out its responsibility is high, sometimes making a wait-and-see perspective impossible.

This ignorance about the categorical nature of X negatively affects the norm system for which a regulator is responsible and creates pressure towards change (or adaptation), as the regulator wants to regain its ability to meet its organizational objectives. Thus, the effect of ignorance on the norm system in this situation is preventive, but not through stabilization, as Popitz described it, but through legislative adaptation.

Finally, the legislative effect can result from institutional risk alone; it creates pressure even in the absence of a societal risk. Yet, it is crucial that the institutional risk itself is clearly established in order to avoid confusion with another line of critical debate that focuses on juridification: an apparently growing tendency of political governments to ‘outsource’ decision-making to jurisdictional bodies in order to avoid being held responsible (Teubner 1998; Loick 2014, 2019). While these concerns resonate well with the debate on risk colonization—and indeed, juridification (Verrechtlichung) was seminally explored by Habermas (1985) in the context of the colonization of the life-world—it is nonetheless the contention of this article that in the case of NGTs, pressure indeed does result from a regulatory impasse (and thus, from institutional risk) and is not a discursive phenomenon strategically created by policy actors to push responsibilities away from the floors of parliament to those of the courtroom. There simply is no evidence corroborating the latter interpretation.

Against this backdrop, a concise, formulaic expression of the legislative effect of ignorance is as follows:

1. Scientific or technological innovations might be used to produce objects for which it is not possible to determine in hindsight whether the particular technology was indeed involved in production.

2. When this ignorance regarding the role of innovation relates to central categories of regulation or legislation, this creates an institutional risk for regulators. Regulators lack the means to execute their regulations and risk missing their organizational or policy objectives.

3. In the face of this risk, which concerns the raison d’être of the regulating organization as well as the stability of the norm system it has to police, the regulator decides to revise or adapt the legislation. Thus, it is justified to speak of a legislative or, in reference to Popitz, of a second preventive effect of ignorance.

7. Conclusion

This article argued that institutional risk may emerge from ignorance under the condition that this ignorance relates to the categories on which the organization in charge of executing a particular regulation or risk management scheme relies. Scientific and technological innovations might lead to a situation in which the means available to the authorities to determine into which regulatory category a specimen under scrutiny belongs are not fit for the task anymore. This leaves the authority in an ‘executive state of emergency’, where it cannot meet its organizational objectives anymore. One conceivable strategy to find out of such an impasse situation is to reduce the scope of the regulation to exempt particular innovations. Another strategy would be to reformulate the regulation, introducing different categories or categorization methods. In either way, adaptation has to occur on the level of legislation. In a nod to Heinrich Popitz’s notion of a preventive effect of ignorance, where deliberate ignoring functions as a safeguard for the normative system in the face of institutional risk, the effect just described was thus called the legislative effect of ignorance.

The article further made the claim that the legislative effect of ignorance can be observed in the current struggles in the EU to regulate the use of NGTs in plant breeding. Moreover, it claims that this is the major driver behind the developments in this regard since the CJEU ruling in 2018. While far from showing a homogenous opinion on these matters, quite a few European value chain actors, ranging from farmers’ associations to plant breeding companies to the food-producing industry, see high potential in the use of NGTs. Across the globe, then, more and more economies open up towards NGTs and install new regulations for their use in plant breeding (Spök et al. 2022). However, the precision of NGTs implies that it is not possible to technically determine whether NGTs have been involved in the breeding process of a given plant or crop. The article claimed that this inability to know is the main factor driving the current processes of policy change in the EU, as it creates political pressure. The combination of (1) other world regions beginning to install less restrictive policies towards GE in plant breeding (among them important agrarian regions like South America and China), (2) the lack of a scientific method to determine whether or not GE has been used in the breeding of incoming produce, and (3) a legal situation where NTGs have to be treated as GMOs (meaning that virtually no GMOs are allowed to be marketed in the EU without permission) has put the EU into a situation where it cannot execute its own legislation.

Finally, the article also explored the generalizability of the legislative effect of ignorance observed in the NGT case. It is not too difficult to think of another situation where a technological innovation could be used to produce results that were indistinguishable from results created through other, ‘conventional’ means. For instance, whether or not an AI application like ChatGPT was used to write this paragraph can hardly be determined in hindsight, but it impacts rules governing authorship, intellectual property, copyrights, etc. This is not a societal risk, but various regulators may—and, in fact, do—worry about the institutional risks that come with this innovation. The exploration of this mechanism through which innovation creates an ignorance that puts a regulator into an institutional risk is thus not only of scholastic interest. Rather, as the major strategy to cope with this risk is to change or adapt the regulation, it creates pressure on political actors to revise legislation or regulation to make sure that the involved agencies and authorities maintain their ability to act according to their organizational objectives.

Acknowledgements

The author wishes to thank Armin Spök, Thorben Sprink, Matthew de Roode, and the two anonymous reviewers for their helpful suggestions and valuable comments. Data can be obtained from the author upon reasonable request.

Conflict of interest.

None declared.

Funding

This work was supported by the EU’s Horizon 2020 Research and Innovation Programme under grant agreement no. 760891 (H2020-NMBP-BIOTEC-07-2017: New Plant Breeding Techniques (NPBT) in molecular farming: Multipurpose crops for industrial bioproducts).

Footnotes

References