Multiplicity of Perspectives on Sustainable Food: Moving Beyond Discursive Path Dependency in Food Policy

: The idea that a sustainable transformation of the food system is urgently needed is gaining ground throughout Europe. Yet, opinions di ﬀ er substantially on what a sustainable food future exactly entails, and on how this future may be achieved. This article argues that recognising this multiplicity of opinions and perspectives in policy making is productive because it creates attentiveness to innovative ideas and initiatives, and may contribute to a broad social support base for policy choices. However, food policy makers may overlook the diversity in perspectives by unreﬂexively adopting understandings of problems and solutions that are historically dominant in their organisations. In this article, we reveal the usefulness of triggering reﬂection on such discursive path dependencies amongst policy makers. We do so by presenting a three-fold case study that we conducted in the Netherlands. First, we analytically distinguish ﬁve perspectives on sustainable food that feature prominently in the Dutch public debate. Subsequently, we show that only two out of these ﬁve perspectives predominantly informed a Dutch food policy—despite intentions to devise a more integrated policy approach. Finally, we discuss the ﬁndings of two focus groups in which we discussed our analyses with Dutch civil servants who have been involved in drafting the Dutch food policy. These focus groups triggered reﬂection among the civil servants on their own perspectival biases as well as on discursive path dependencies in Dutch food policy making. We conclude by discussing the implications of our ﬁndings for the understanding of the discursive politics of sustainable agro-food transformations in Europe.


Introduction
The notion that a transformation towards a more sustainable food system is urgently needed is gaining ground throughout Europe (e.g., [1,2]). Yet, opinions differ substantially on what sustainable food production and consumption exactly entail, and on how the transformation to such sustainable production and consumption should be managed [3,4]. Should we, for instance, move towards more localised food systems, or improve the eco-efficiency of global food production and trade? Should we principally invest in technological innovations, or rather in social innovations? Additionally, should we rely on green market dynamics, or instead on a strong environmental state? Profound differences of view permeate public debates on sustainable food and agriculture, but these differences tend to be placed in the background when policy decisions are being made [5][6][7]. The Dutch government, for instance, has repeatedly stated that it aspires to make the Netherlands the world leader in sustainable food production and consumption-thereby suggesting that the definition of sustainable food is not contested, but instead universal and given [8]. In this article, we argue that such a discursive closure is unproductive because it reinforces existing dominant perspectives on sustainable food and agriculture, and overlooks the transformative potential of alternative perspectives. Furthermore, we contend that such a discursive closure is not always a conscious decision made by policy makers but that it may also be a result of a "discursive path dependency": that is, the result of a more or less non-reflexive adoption of a perspective that is historically dominant in a specific organisational context such as a policy department. We aim to contribute to uncovering such discursive path dependencies in current agro-food policy making, and to explore the usefulness of triggering reflection on these discursive path dependencies amongst policy makers.
To meet these aims, we will focus on the public debate and food policy in the Netherlands as a case study [9]. Drawing conceptually on insights from the literature on (food) paradigms, discourses, and framing (Section 2), we will begin by analytically distinguishing five perspectives on sustainable food and agriculture that feature prominently in the Dutch public debate (Section 3). These perspectives tell fundamentally different, yet internally coherent stories on the causes of the current agro-food system's unsustainability, solutions to overcome this unsustainability, and the types of knowledge and actions that are required to adequately assess and address these problems and solutions. Subsequently, we will analyse which of these perspectives informed the Dutch food policy (with a focus on policy documents issued between 2015, when the Dutch government introduced a food policy framework, and 2017). We will show that two out of the five perspectives were dominant in guiding the Dutch food policy (Section 4). In a final empirical section (Section 5), we will discuss the results of two focus groups that we have organised with civil servants who have been involved in drafting the Dutch food policy, so as to observe the type of discussion and reflection that our presentation of the five perspectives and their position in the Dutch policy elicited. We will conclude the article (Section 6) by discussing the implications of our findings for the understanding of the discursive politics of sustainable agro-food transformations.

Perspectives on Sustainable Food: Analytical Framework
This research builds on the interpretive tradition in the social sciences, which "assumes the existence of multiple socially constructed realities instead of a single reality, governed by immutable natural laws" [10] (p. 176). Accordingly, interpretative research focuses on the way in which social groups make sense of social and environmental phenomena, rather than on the characteristics these phenomena themselves. To study such sense-making frameworks, various concepts have been developed, among which "paradigms", "discourses" and "frames" are arguably the most well-known [11]. Below, we briefly discuss these concepts. Subsequently, we clarify how they feed into our notion of a "perspective on sustainable food".
The concept "paradigm" [3,12] denotes a set of fundamental cognitive assumptions on the basis of which problems and solutions are conceived and intellectually approached. As one paradigm tends to dominate thinking in domains of science [12] and food policy [3], it takes a "revolution" to establish a "paradigm-shift". The concept "discourse" [13,14] signifies an ensemble of ideas and concepts through which meaning is ascribed to phenomena, "and that is produced in and reproduces in turn an identifiable set of practices" [14] (p. 60). Discourses affect practices when they become dominant and are subsequently internalised (by individual agents) or institutionalised (e.g., in policies). At the same time, discourses are (re)produced in practices when actor-coalitions actively or tacitly support them in their sayings and doings. The concept "frame" [15][16][17], finally, refers to a selection of aspects of a phenomenon that shapes the way in which agents understand and aim to overcome problems in relation to this phenomenon. Studies into framing have shown that frames are often tacit rather than discursive and, hence, escape explicit reflection [15,17].
Drawing selectively on the above conceptual insights, in this article we work with the notion of a "perspective on sustainable food". We define such a perspective as a more or less coherent set of answers on the questions: to what degree and why is the current agro-food system unsustainable? What type of knowledge is needed to adequately assess and address agro-food problems of sustainability?
Additionally, what does a sustainable agro-food system look like and how can this future be attained? Importantly, like paradigms and discourses, perspectives are not just representations; they more or less strongly shape agro-food policies and practices, depending on their dominance in public and political debates and the concomitant degree to which they influence stakeholders' personal and policy decisions. As the perspectives inform stakeholder positions in debates but are often left implicit (like frames), we consider it important to reveal these perspectives and enable a reflection on them.
We have chosen to work with the concept "perspective" instead of "discourse", "frame" or "paradigm", for the following reasons. First, the principal reason to analytically distinguish perspectives on sustainable food was to present these to policy makers in order to trigger reflection on the sense-making frameworks that they do (and do not) adopt while devising food policies. We anticipated that using the common-sensical term "perspective" in our work and communicative efforts with policy makers would better suit our aim than using the theory-laden and more complex terms of "paradigm", "discourse" and "frame". Second, we do not intend to study which actor coalitions advocate particular positions, nor wish to focus on particular food-related issues or controversies, as is characteristic for discourse and frame analyses (e.g., [5,18,19]). Instead, we leave open the possibility that food policy makers subscribe to-and, hence, recognise themselves in-different perspectives in relation to different issues. Third, we have chosen not to work with the concept of "paradigm", because studies into paradigms tend to juxtapose a limited number of fundamentally opposing sense-making frameworks (e.g., [3]). In this study, we aim to reveal, also, more nuanced, yet no less relevant differences in perspectives.
In line with this latter aim, we have sought to analytically distinguish perspectives on sustainable food according to multiple themes that feature in the scientific literature and the Dutch public debate on sustainable food and agriculture. We adopted these multiple themes to move beyond binary readings of issues of food and sustainability that tend to dominate the scholarly debate (e.g., agro-industry vs. agro-ecology, expert knowledge vs. lay knowledge, global vs. local, and fast vs. slow food) [7]. The themes that we have selected are:

•
Knowledge and innovation: What type of (scientific and/or lay) knowledge is key to gaining a proper understanding of sustainability problems? Additionally, what type of (technological and social) innovation is needed to achieve a more sustainable food future (e.g., [20,21])? • Governance philosophies and key agents of change: To what extent have public and/or private governance arrangements contributed to sustainability problems? Additionally, what role is ascribed to governmental agents, market actors and members of civil society in establishing a more sustainable food future (e.g., [22,23])? • Geographical scale: What role do (global, regional, local) geographies of food production, trade and consumption play in causing and overcoming food system unsustainability (e.g., [24,25])? • Food characteristics: What does sustainable food look, smell and taste like? Is it "natural" and unprocessed, or rather processed to improve aspects like nutritional value and shelf-life? Additionally, is the sustainability of food measured according to the biophysical attributes of the production process and the food product itself, or are socio-cultural and economic relations of food also taken into account (e.g., [26,27])? • The nature of changes necessary: To what degree do current agro-food policies and practices need to be changed to achieve sustainable development? Are moderate reforms required, or can only a radical transformation of the agro-food system lead the way to a sustainable future (e.g., [28])?
Drawing on these thematic distinctions, in the following section we will analytically distinguish five perspectives on sustainable food that feature prominently in the Dutch public debate on sustainable food and agriculture.

Five Perspectives on Sustainable Food in the Dutch Public Debate
We analytically distinguished five perspectives on sustainable food based on our expert reading of the Dutch public debate of the last decade, which we followed non-systematically yet consistently out of professional interest via popular and specialised media. To substantiate the perspectives and put them in a wider context, we drew on Dutch and international agro-food studies literature. To enhance the robustness of our analysis, three independent Dutch scientific experts reviewed our categorisation (dr. Hans Dagevos from Wageningen Economic Research, dr. Jeroen Candel from Wageningen University), and prof. dr. Peter Oosterveer from Wageningen University), on the basis of which we refined our analytical distinctions between the five perspectives.
By distinguishing the five perspectives, we did not intend to give the definite answer on the question of which visions on food and sustainability exist in Dutch society. Instead, we aimed to reveal the existence of nuanced yet fundamental differences between multiple perspectives on sustainable food, which can serve as a basis for reflection amongst policy makers on the perspectives that they and others subscribe to.
In the following five sub-sections, we briefly discuss the five perspectives. Table 1 provides an overview of the analytical distinctions between the five perspectives according to the themes that we have set forth in our analytical framework.

Business-As-Usual: Producer-Led Incremental Change
This perspective stipulates that the current global agri-food system is successful in producing plenty of affordable food, but does face a number of environmental and human health problems. These problems are not systemic but can largely be addressed by global market forces-which have made the current agri-food system successful in the first place. Global specialisation and trade will facilitate improved eco-efficiency [29]. Residual ecological and human health problems may be addressed by governmental interventions, as long as these do not disturb the level playing field in international trade (e.g., by pricing externalities or enhancing legal minimum standards for food production) [24]. Food is principally valued according to its taste, price and convenience and consumers are, accordingly, not willing to pay a premium for sustainable food [30].

Technological Optimism: Promoting Techno-Scientific Solutions
This perspective centres on the question of how to sustainably feed a rapidly growing and increasingly affluent world population. In answering this question, it focuses on technological innovations that may carry the seeds of a "revolution comparable to the introduction of the tractor and chemical products in the 1950s" [31] (p. 15), such as smart farming and precision farming technologies, nano-technologies and gene-editing techniques [31,32]. Agro-technological innovations have been key to the post-World War II food system and its aim of fighting hunger. Facilitating technological innovations that help to sustainably feed the growing world population does, accordingly, not require radical systemic changes. Rather, it necessitates continued governmental and corporate support for establishing successful agro-technological innovation systems [33]. Food is principally approached as "fuel" for the world population and valued for its nutritional value. Consumers are satisfied with the food they find in stores, which has often been processed to enhance their nutritional value, safety, and shelf life.

Alternative Food Politics: Re-Connecting Producers and Consumers in Local Networks
This perspective holds that the current "faceless and essentially placeless food system" entails a myriad of fundamental problems, including environmental pollution, low animal welfare, and low farmer income [34] (p. 391). Overcoming these systemic problems requires a move away from anonymous global agri-food chains that are insensitive to local social and ecological realities, towards "alternative" local or regional food networks in which producers and consumers can be re-connected [35]. By re-connecting food production and consumption at a local or regional scale, farmers and consumers re-gain the possibility to act according to local ecological and cultural needs. Producers and consumers are key agents of change as they are to establish and support alternative food initiatives. Governments are asked to actively promote local or regional food sovereignty by public regulation of production methods and markets [5]. Food is far more than its biophysical properties: it is a matter of culture and belonging. Buying food in alternative food networks is not just an economic transaction but an expression of social engagement and ecological citizenship [36].

Political Consumerism: Consumers as Key Agents of Change
This perspectives stresses that a series of food scares (from mad cow disease to horsemeat scandals) has triggered consumer awareness of downsides of the productivist food system and its focus on enhancing economic efficiency through intensification and agronomic rationalisation. Consumers are no longer a priori satisfied with the food that the system delivers, but demand-and are willing to pay for-premium food qualities and attributes. A "consumerist turn" has occurred: consumer choices now steer food production practices rather than the other way around [37]. Transparency about food attributes is key in empowering consumers to make reflexive food choices. Labelling schemes are established examples of informational governance tools that enable consumers to include social and ecological considerations in their purchasing decisions. New ICT-developments (including blockchain technologies) radicalise the possibility to provide food transparency and contribute to a fragmentation of consumer concerns [38]. In view of the growing amount of food labels and claims, governments have a role to play in ensuring the correctness of information provided. In the fragmented foodscape, food choice becomes an expression of one's identity and lifestyle politics [39].

Integrated Food Politics: Food Systems Thinking
This perspective starts from the observation that the current food system entails a range of fundamental problems, including local and global environmental deterioration, "unfair" trade and malnutrition. To adequately address these problems, it is crucial to consider their interdependence and to anticipate synergies and trade-offs between possible solutions. Doing so requires the involvement of representatives of all parties that collectively constitute the food system, such as in round tables and multi-stakeholder dialogues and agreements. In this perspective, governmental agents are the key agents of change as they are in the position to bring the various parties to the table, and to adjust the regulatory and political context in which food production, trade and consumption practices are enacted. To govern effectively, governments need to operate at multiple levels, from cities and regions to the EU. Foodstuffs and their attributes are not valued in isolation but as part of a diet. Diets that are simultaneously healthy, animal friendly and socially and ecologically sustainable become the new social norm [1, 3,40]. Business-as-usual Environmental and human health problems are side-effects of overall efficient production methods. Sustainability problems are not systemic but signal a lack of full market efficiency.
Market selection pressures will incite food producers to increase their eco-efficiency and sustain their supply base. Governments should address residual problems and protect a global level playing field.
Continuous market-induced improvements in established food production processes will suffice to attain a sustainable food system.
Global specialisation and trade facilitate improved eco-efficiency, cost reduction and lower prices.
Globally operating agro-companies; transnational food corporations.
Corporate research and development enable the necessary, predominantly incremental improvements in eco-efficiency.
Food is readily available and affordable; key food attributes are price, taste and convenience.

Technological optimism
Current food production practices do no suffice to sustainably feed the rapidly growing-and increasingly affluent-world population in the future.
Technological innovations (e.g., precision farming, gene-editing) will sustainably increase world food production. Governments should actively support technological innovation with targeted policies.
Radical technological innovation is needed. Yet, no radical food system transformation is needed as technological innovation has already been the driver of food system productivity for decades.
Sustainable technologies are not specific to a particular geographic organisation of the agro-food system. Techno-scientific knowledge itself is more or less universal and may travel across the globe.
Techno-scientific knowledge is key to achieving long-term food system sustainability.
Food is "fuel"; the key food attribute is nutritional value. Products may be processed to increase its nutritional value, shelf life and safety.
Alternative food politics The current food system suffers from multiple fundamental problems that are inherent to its global, industrial, and essentially homogenous and anonymous operation.
Food producers and consumers should be re-connected in local and regional networks.
Governments and non-governmental organisations should actively promote the rise of alternative food networks.
A fundamental transformation of the current food system is required, from global agri-food chains towards food networks that incorporate local ecological and cultural needs. Local or regional food networks-such as community supported agriculture, box schemes and farmers' markets-enable the re-connection of producers and consumers.
Local food producers and consumers, which may collaboratively set up and support alternative food networks.
Appropriate knowledge is context-specific. Besides natural and social scientific knowledge, local knowledge is highly valued.
Food is part of local culture; food quality and authenticity are highly valued. Food is more than a product: it connects consumers with producers and ecological regions. consumer demand reforms the conventional productivist food system. Providing transparency through labelling arrangements and ICT applications is key in facilitating reflexive consumer choice.
The food system is being reformed as consumer demand is now the key guiding principle in organising food supply chains (rather than the wish to optimise agricultural production based on agronomic principles).
Transparency tools such as labelling schemes and ICT applications bridge large distances (in place and time) between producers and consumers.
Consumers, whose concerns inform the design of food production practices.
Access to knowledge and information is key to reflexive food choices. The type of knowledge and information that is considered most relevant differs according to different consumer concerns and sub-cultures.
Food is valued according to a large variety of qualities, which depend on consumers' specific values. In the fragmented foodscape, food choices are an expressional of one's identity and political preferences.

Integrated food politics
The food system is complex: Various sustainability problems interconnect, which cannot be traced back to a single cause or driving force.
Only an integrated approach of the various sustainability problems can foster a more sustainable food system. All food system parties should be involved in devising holistic solution.
A policy reform is needed, from the current fragmented towards a more integrated approach. Incremental changes in food production and consumption practices should eventually lead to radical systemic change.
Multi-level governance is key to achieving food system sustainability. Sustainability problems should be addressed at the appropriate scale.
Governments are in a key position. They can call various food system stakeholders to the table and adjust the political and regulatory context in which the food system operates.
To adequately grasp food system complexity, knowledge should be inter-and trans-disciplinary. It is important to integrate the knowledge and viewpoints of food system actors.
Food is part of a diet that meets multiple sustainability and human health criteria.

The Position of the Five Perspectives on Sustainable Food in the Dutch Food Policy
Having distinguished five perspectives that feature prominently in the Dutch public debate on sustainable food, in this section we discuss the position of these perspectives in the Dutch food policy. We do so by focussing on three policy documents [8,41,42] that have been issued between 2015, when the Dutch government introduced a new food policy framework entitled "the Food Agenda for safe, healthy and sustainable food" (this and all other quotes from Dutch policy documents have been translated by the authors) [41], and 2017, when we began to conduct our analysis.
The national government's "Food Agenda" was an initiative from the ministries of Economic Affairs (which was responsible for agriculture from 2012-2017); Public Health; Foreign Trade and Development Cooperation; and Infrastructure and Environment. It was drafted in response to a report of the Netherlands Scientific Council for Government Policy entitled "Towards a food policy" (WRR, 2014). This report argued in favour of a cross-sectoral, integrated approach to food-related issues that would not only take matters of agriculture but the entire food system (from farm to fork) into account, and that would focus holistically on three major challenges that food systems face: ecological sustainability, public health and resilience. Historically, the Dutch government has been particularly strong on supporting agricultural production and innovation, which has contributed to making the Netherlands one of the world's largest agricultural exporters (in 2018, The Netherlands was the second largest agricultural exporter in the world, after the United States [43]), as well as "world leaders in agricultural innovation" [44]. By drafting the food agenda, the government aimed to "move from an agricultural policy towards a food policy" [42] (p. 1) in line with the advice of the Netherlands Scientific Council for Government Policy. Along with this new policy orientation, it defined a new national ambition: "to become world leader in safe, healthy and sustainable food and sustainable agriculture and horticulture" [8] (p. 1).
The "integrated food politics" perspective, as advocated by the Netherlands Scientific Council for Government Policy, informed the rationale to devise the Food Agenda. This became clear from the involvement of the multiple ministries in drafting the agenda, the title reference to safe, healthy and sustainable food, and statements that purport that "the government wants to work in a more integrated way" [42] (p. 1). However, when looking into the actual policy measures and initiatives that are being proposed and financially supported in the Food Agenda, it became evident that two other perspectives that have traditionally informed Dutch agricultural policy continued to dominate in the new food policy: "business-as-usual" and "technological optimism".
In line with the "business-as-usual" perspective's focus on producer-led change, much policy support and economic resources were devoted to "Dutch companies to [help them] introduce, scale-up and market very promising new products" [42] (p. 4). Intentions to promote shifts in food consumption (such as lowering salt and sugar intake, and replacing animal proteins with plant proteins) were principally addressed by relying on voluntary business initiatives and supporting technological innovations (salt and sugar replacements, and "the development of alternative protein sources like pulses, seaweeds, algae, insects and cultured meat" [4] (p. 10)). The dominant position of technological optimism became furthermore clear in a section called "Knowledge and innovation" [42] (pp. [26][27], which was fully devoted to natural scientific and techno-scientific developments-rather than also social innovations. The continued reliance on business-as-usual thinking and technological optimism was well-captured in the following statement, which qualified the Dutch world leadership ambition: "By focussing on sustainability, health and transparency in food production . . . but also on the export of knowledge and technology, the Netherlands can continue to play a distinctive role in international markets" [42] (p. 4, our emphasis).
"Business-as-usual" and "technological optimism" were dominant but not the only perspectives that were informing the Food Agenda. As the reference to "transparency" (which is central to the "political consumerism" perspective) in the last quote already indicates, the Dutch food policy also paid some attention to the role that consumers may play in fostering a sustainable food future. Developing food transparency was considered principally important to help Dutch companies to market their "unique, premium and distinctive" sustainable food internationally. Furthermore, the government aimed to provide Dutch consumers "with good and understandable information" on the food they eat by developing information campaigns, clearer labelling requirements and a "food information app" [42] (p. 7). It should also be noted that "integrated food politics" thinking was not totally absent in the operationalisation of the Food Agenda. Most notably, it inspired the organisation of a national food summit in 2017 [8], in which various food system parties participated-although this summit has been criticised for predominantly attracting established food industry parties and thus supporting business-as-usual interests (e.g., [45]).
One perspective that was almost entirely absent in the Food Agenda is the "alternative food politics" perspective. The problem of farmers' low incomes and limited market power in the current agri-food system-which features prominently in the "alternative food politics" perspective-was recognised in the new Dutch food policy. However, this problem was not addressed by re-organising producer-consumer relations as suggested in the alternative food politics perspective, but rather by focussing on market competition legislation and farmer participation in innovation trajectories (which once again underscores the dominance of market-oriented and technologically oriented thinking) [42] (pp. [21][22]. The "alternative food politics" perspective is difficult to reconcile with the two dominant perspectives and this has arguably contributed to the virtual absence of this line of thinking in the Dutch food policy in the period under study.

Results of Focus Groups with Dutch Civil Servants: Fostering Reflection on Discursive Path Dependencies in Food Policy Making
In this third and final empirical section, we present the results of two focus groups in which we have discussed our findings with civil servants who have been involved in drafting the Dutch food policy documents that we have analysed in the previous section [8,41,42]. We organised the focus groups for two main reasons: to test the robustness of our analyses of the Dutch public debate and food policy; and to investigate the type of reflection that our findings elicited. As described in our analytical framework, we anticipated that perspectives inform policy makers' positions in public and policy debates, but tend to remain elusive rather than the object of discussion themselves. By organising the focus groups, we aimed to investigate whether this anticipation was correct, and to explore the added value of fostering explicit reflection on the position of perspectives in food policy making.
The two focus groups took place in May and June 2018 and involved in total nine participants. The first one contained three civil servants from the current Ministry of Agriculture (in 2017, the merger of the ministries of Economic Affairs and Agriculture was made undone); the second one contained six civil servants from the ministries of Agriculture, Economic Affairs, and Public Health. We approached these participants via email based on our own knowledge of which civil servants had been involved in drafting the food policy documents, and subsequently applied the snowball method by asking these civil servants to forward the invitation to interested colleagues.
We purposively selected these civil servants for the following reasons. First, their involvement in drafting the Dutch food policy documents under study [8,41,42] implies that they hold the expertise to assess the robustness of our analyses of the Dutch public debate and food policy. Second, in the Dutch governmental system, civil servants who work at ministries are responsible for providing their respective ministers with factual information ("speaking truth to power") on the basis of which the minister can make political decisions, and for carrying out ministerial decisions. Dutch civil servants are, accordingly, not supposed to give the minister political advice [46]. Given their formally non-political role, the civil servants are well able to assess the degree to which the different positions of the various perspectives in the Dutch food policy are the result of overt political choices (from the minsters), or the result of discursive path dependencies. Importantly, our sample was not meant to be quantitatively representative for all civil servants or policy makers who have been involved in devising food policy in the Netherlands. Rather, the sample was designed to meet our aim of exploring the role that discursive path dependencies play in Dutch food policy making, as well as the effects of triggering reflection on this role. Given that discursive path dependencies are an outcome of social interaction, focus group discussion in which social interaction can be observed are well-suitable for exploring this social phenomenon [47,48].
The two focus groups were organised and led by the two authors, and were structured as follows. To acquaint the participants with our findings, we presented our analyses of the Dutch public debate and food policy orally at the start of the focus groups (as well as in a Dutch-language report [49] that the participants had received beforehand). Following the oral presentation, we led group discussions with the use of a semi-structured topic list that centred on three main topics: the robustness of our findings, the dynamics that determine the dominance of perspectives in policy design, and the usefulness of explicit reflection on these discursive dynamics. Following the logic that "if the goal is to learn something new from the participants, then it is best to let them speak for themselves" [47] (p. 40), we only used the topic list to introduce the topics of our interest, while leaving much room for the participants to discuss their own thoughts and ideas. Both group discussions lasted for approximately 90 minutes and were recorded, transcribed, coded and analysed by the authors.
The focus groups started with critical assessments of our delineation of the five perspectives. Some participants initially doubted whether the five perspectives cover the entire Dutch public debate: "the perspectives are clearly recognisable, but I lack the overview to know whether there is a sixth or seventh or eighth or ninth one" (this and all other quotes from the focus groups are the authors' translations). Others, instead, raised arguments against analytically distinguishing one or more of the perspectives from the others. Some participants, for instance, were against considering "political consumerism" a separate perspective because the perspective would rest on a notion of a "consumer that doesn't actually exist", and because consumers play an important role in each perspective. Notably, such critical remarks did not fundamentally undermine our analysis but instead tended to inspire other participants to defend our categorisation. In doing so, the participants came to a first important point of reflection: some of the perspectives represent blind spots for the national policy officials. According to one of the participants, such a blind spot explains why his colleagues considered the "political consumerism" perspective non-distinct: "My 14-year-old daughter and her friends are only into food blogs. They do not watch tv. We watch tv, but they do not. We live in separate worlds . . . There is already a huge vegan youth culture that we just do not know about".
A similar line of reasoning emerged in relation to the "alternative food politics" perspective. The following excerpt is taken from a discussion between two participants (P1 and P2): P1: "Yes, I do recognise the five perspectives . . . I do encounter all five of them. But in the national debate, I do not really see advocates of those alternative food networks." P2: "You are stuck in your own bubble when you only look at things from your national perspective. Regional sustainability is developing very strongly right now." The fact that some of the perspectives represent blind spots in national policy circles goes some way in explaining why these perspectives are less identifiable in the Dutch Food Agenda than others. Yet, participants were eager to clarify that it was not simply a matter of them being unaware of the existence of particular perspectives. Rather, they lacked attention for some of the perspectives in their role as a civil servant: "It really depends on your role. As a father and as someone who enjoys food, I approach it differently than as a policy officer". What is more, our finding that the "integrated food politics" perspective was eventually placed in the background in the drafting of the Food Agenda cannot be explained by considering this perspective a blind spot, as it was put on the national agenda by the Netherlands Scientific Council for Government Policy and had inspired the drafting of the Food Agenda in the first place.
In searching for an explanation for the dominance of the "business-as-usual" and "technological optimism" perspectives at the expense of the other perspectives, participants came to a second main point of reflection: that the dominance of these two perspectives have been firmly institutionalised in Dutch agro-food policies and in the ministries that are responsible for it. Participants contended that "the entire system, all our thinking, is based on it [the two dominant perspectives]. If you look at the regulation, that is all based on policies from the 1960s and the 1970s and we just continue to build on that"-which makes it difficult to integrate new perspectives in agro-food policy. Furthermore, participants stated that "the entire ministry [of Agriculture] has just fundamentally been built on these [dominant] perspectives". This would become evident in the fact that most employees are educated in the same school of thought "which I always call 'classical agriculture'", and in "the way the Ministry is organised. We are organised according to sectors... So that integrated approach is just not part of the DNA of the organisation." All in all, the participants of both focus groups supported our categorisation of perspectives on sustainable food and our food policy analysis, and considered the exercise of reflecting on discursive dynamics in food policy making useful. They held that their tacit ascriptions to the two dominant perspectives were understandable yet undesirable, as the new food policy was devised to be more integrated, and valuable policy input could be derived from the other non-dominant perspectives as well. When being asked what the participants took home from the focus groups, most answered that they wanted to relate more reflexively to the various perspectives in their future work and to move beyond the current discursive path dependency in their work. As one participant phrased it: "The overall conclusion is that two perspectives are highly dominant and that there's little room for the other perspectives. That is a harsh message for us. I think that the conclusion is correct and it can all be well-explained. But it is our task to break through that".

Reflection and Policy Implications
In this article we have presented a nuanced, non-binary categorisation of perspectives that feature prominently in the public debate on sustainable food in the Netherlands. We have used this categorisation to analyse recent developments in Dutch food policy and have revealed that despite intentions to renew the rationale of this policy, it continued to principally draw on two long-established perspectives ("business-as-usual" and "technological optimism"). We have discussed our findings with Dutch civil servants in focus groups and found that our analyses triggered reflection amongst these policy officers on their own perspectival biases, as well as on discursive path dependencies in Dutch food policy making. Notably, whilst our findings are specific to the Netherlands and analyses of public debates and food policies in other countries may reveal other perspectives on sustainable food, similar discursive path dependencies likely play a role in food policy making in other polities. Accordingly, similar analyses of public debates and food policies in other countries may contribute to forestalling non-reflexive discursive closure in food policy making.
The existence of multiple perspectives on sustainable food and their differentiated positions in Dutch food policy clearly show that the definition of sustainable food is not universal or given, but rather a matter of political contention. Behind a seemingly clear and unambiguous definition of sustainable food and agriculture, as implicated in the Dutch ambition to become world leader in this matter, there looms a multiplicity of perspectives and political choices. Bringing this to the attention of civil servants resulted in their willingness to more reflexively engage with the different perspectives that are at play in society, while collecting information and fulfilling their role of "speaking truth to power". This observation raises the question how policy makers can best relate to the diverse-and, at points, arguably incompatible-perspectives? While it is beyond the scope of this article to thoroughly investigate this question, we do want to finish by offering three different approaches that are brought forward in political science literature.
A first approach, which is typical for the Dutch (neo-)corporatist tradition, centres on searching for consensus between advocates of the different perspectives [50]. In line with this approach, explicit reflection on conflicting perspectives may contribute to overcoming discursive deadlocks and establishing joint understandings and solutions [17,51]. A second approach is to openly choose between options that are advanced by advocates of the different perspectives. According to this radical democratic approach, an outright battle between supporters of the different perspectives will help to sharpen ideas and, accordingly, enhance the productivity of political debate [52,53]. A third approach is to come up with so called clumsy solutions. This approach starts from the notion that all perspectives make valid knowledge claims that are overlooked by the others and, therefore, seeks to "creatively combine all opposing perspectives on what the problems are and how they should be resolved" [54] (p. 817)-even if such solutions are conflicting in terms of policy coherence or consistency [55].
Crucially, these three approaches have in common that they do not hide different perspectives behind a seeming consensus about the definition of sustainable food, but instead explicitly recognise all perspectives that exist in the public debate on sustainable food futures. Such explicit recognition may not only shed light on innovative ideas to foster sustainable food systems, but may also contribute to a broader social support base for policy choices.