1. INTRODUCTION
Gravitation, Einstein said, cannot be held responsible for people falling in love.Footnote 1 But what precisely is the ambit of scientific responsibility for changes no less profound in the world in which we live? On 16 and 17 May 2011 in Brussels, Belgium, the European Commission, in the form of its Directorate General Research,Footnote 2 held a workshop to consider a ‘Responsible Research and Innovation Framework for the European Research Area’Footnote 3 and announced various soft law initiatives to promote responsible research and innovation (rri). The tag of rri is in some ways curious. Has research and innovation been irresponsible to date? And, if so, how? Also, the linking of responsibility to innovation might seem to suggest that innovation is the responsible response to problems, so that activities such as geo-engineering become legitimate in the face of climate change. In contrast, responsibility in the context of sustainable development might lie in foregoing the opportunities allowed to us by innovation. Responsibility will need to allow for a maturity of choice in which some possibilities are closed off as others are opened up.Footnote 4
At least one review of responsibility has been established in the context of a commitment to develop nanotechnology, while looking to minimize adverse effects or unintended consequences of that development.Footnote 5 Such approaches have been criticized as deploying responsibility as the ‘balancing calculus’ between the perceived benefits of innovation and the risks to which it might give rise.Footnote 6 This is suggestive of supporting a utilitarian endeavour, whereas more deontological approachesFootnote 7 might indicate that there are certain things that we should not do, even though we can.Footnote 8 A single framework for both research and innovation may also prove problematic. Research may be seen as advancing knowledge whereas technological innovation is concerned with application, often of different sets of knowledge.Footnote 9 It may be difficult to frame responsibilities for research when applications of research findings may be diverse and manifold. So, the idea of rri is not unproblematic. Nonetheless, the pursuit by the European Union (eu) of a soft law framework for rri is remarkable and ambitious. In this paper, I wish to explore three central questions relating to this initiative: those of objective (why research and innovation?), scale (why the eu?), and governance (why soft law?)
2. OBJECTIVE
The debate about responsibility in relation to innovation is ignited by the transformative power of science, particularly through converging technologies: the ability of different systems to join in a common endeavour. There is a ‘constellation of modern technologies’Footnote 10 (including biotechnologies, nanotechnologies, information, and cognition technologies) that involve the manipulation of both biological and non-biological materials and, thus, blur the borders not only of scientific disciplines but also of what might constitute life – the difference, say, between a brain and a computer. There is something radically different not only in the transformative capacity of such technologies but also in their aspiration. The goals of these technologies include manipulation of atomic matter, of genes, and of the earth’s climate. Their remit is sweeping and breathtaking and it relates not only to biomedical interventions at the beginning or end of life, but also to the environment in which, and human conditionsFootnote 11 under which, we live that life. While nature is inevitably interpreted and technically constructed through science, it is now also shaped by its deliberate intrusion.
Add to this the ubiquity of modern technology and the magnitude of its impact. Nanoscience, for example, does not generate a technology of single application. Rather, the engineering of functioning systems at the nanoscale produces a general purpose, enabling technology capable of countless applications in many different fields. Whereas the regulation of technologies given over to a single purpose might be amenable to regulatory oversight by a single agency, this sort of solution is foreclosed by the pervasive nature of nanotechnologies. Moreover, not only do the technologies set out to modify physical processes but they also do so on an industrial basis. This has led to Krohn and Weyer suggesting that society may be used as a laboratory in instances where, to take their example, the impact of the release of genetically modified organisms (gmos) into the environment is only finally determined by trying it out. They suggest that, as society at large is exposed to the dangers of scientific error in such circumstances, ‘this requires a redistribution of responsibility’.Footnote 12
If we place the emphasis here on responsibility, the notion of rri is evocative of a concern that modernization or innovation can be morally problematic. It could indicate that doubts about scientific progress are no longer confined to the field of technological application, but are increasingly raised at the stage of research. The call is for ethical choice to become embedded in procedures and processes which attach to the research itself to allow a determination of ‘what types of decision may be made, how they may be made, by whom, and with the assistance of what resources’.Footnote 13 This represents a move towards the institutionalization of decision-making and is well represented by recent moves of the United Kingdom (uk) Engineering and Physical Sciences Research Council (epsrc) to demand this type of foresight at the grant application stage as it did with the third nanotechnology grand challenge on climate change.Footnote 14 This is a fascinating development since research grant income is generally generated by an appeal to future worth, whereas the epsrc now demands the completion of a risk register to catalogue potential downsides. This is a seismic shift from earlier science-driven models of innovation. The focus on societal, environmental and economic outcomes of research and development in this context might be seen alongside calls for demonstration of the ‘impact’ of research undertaken. In other words, the return on the investment in science is open to scrutiny.
It goes beyond this, however, because the concern that generates the call for rri is a product of our enhanced awareness of the power of modern science, which leads to two strands of thinking: one about risk and the other about responsibility. In relation to risk, we have heeded Beck’s caution, first written 25 years ago, about the unregulated, and therefore uncontrolled, development of science and technology.Footnote 15 The exponential capacity of scientific research to prise open possibility makes more formal modes of control practically impossible. This capacity, well represented by Moore’s Law,Footnote 16 also makes it likely that even feasible controls will be resisted in favour of seemingly benign promises or the appeal of possibility. Beck spoke of the changing relationship of ‘humankind to itself’ and the blurring boundaries between the natural and the man-made under such forces. The notion of ‘risk’ in risk society is wider then than hazard and becomes, as Giddens has observed, a preoccupation with a diversity of possible futures, which we explore in an effort to normalize and control.Footnote 17
Risk society, understood in this way, is intertwined with the increasing influence of science and technology and its capacity to manufacture risk. No longer an unambiguous source of discovery, scientific research is seen as generating rather than negating uncertainties, the resolution of which cannot be found simply in scientific evidence.Footnote 18 Indeed the fear is not, or not only, one of anticipated untoward consequences of innovation but (also) of the unforeseen and unforeseeable consequences. While, then, greater attention is given to the environmental and human health consequences of technological development,Footnote 19 given the uncertainty and knowledge shortfalls that are emerging, this development is outstripped by the rapidly growing need for anticipatory governance, which includes participatory debate on the sorts of future that innovations may herald. At this point, questions of risk begin to shade into those of responsibility.
Hans Jonas – coming from a different direction from Beck, but also writing in the mid-1980s and in an analysis no less compelling – is concerned with the power of technology to penetrate the boundaries of nature to the point that this generates a responsibility that both correlates with that power and is commensurate with it.Footnote 20 In other words, it is the scope of the enterprise in the exercise of this power that shapes the nature of the responsibility. Though not one of Jonas’ examples, one might say that if synthetic biology seeks to produce tools to redesign the natural world, then it is this very ambition and aspiration that engender and structure the degree of accountability. This is an extensive – perhaps all-embracing – and, for Jonas, metaphysical responsibility in so far as it extends beyond a single generation or species. For Jonas, the identification of this responsibility is crucial in discriminating between ‘legitimate and illegitimate goal settings to our Promethean power’.Footnote 21 This suggests a scientific futurology, a ‘lengthened foresight’, which may lead to self-denial of opportunity, and the preservation of values and traditions at the expense of other potentiality.
We are accustomed to viewing responsibility as a moral foundation of legal relationships between citizens under Kantian influence.Footnote 22 The norms that shape our morality rest on the status of individuals, who should be treated as ends and not means. Arguably, placing future generations in jeopardy or depleting resources on which they would otherwise depend may be to treat such generations as a means to satisfy present desires. However, as Groves has pointed out,Footnote 23 such moral imperatives become much more difficult to apply where the citizens in question do not live in spatial and temporal proximity to each other. We therefore lose some of the guidance of moral philosophy in seeking to make determinations about what might be considered ‘responsible’ in relation to future technology. Added to the difficulties of time and space are the limitations on foresight, especially in relation to the unidentified and ambiguous. This is not merely a matter of the limits of scientific knowledge, or the existence of a territory of the yet undiscovered lying outside the boundaries of present scientific understandings. Technologies transportFootnote 24 such knowledgeFootnote 25 into a huge variety of social settings and practices.Footnote 26 It is our ignorance and the impossibility of knowing about these that pose the real challenge. This constitutes the reflexive element of modernizationFootnote 27 and its presence produces a need to act responsibly but in conditions of true uncertainty of eventualities. What then of decision-making in favour of responsible innovation under conditions of ignorance and uncertainty?
One rather obvious point is that if such conditions of contingency pertain, then there can be no guarantee that the responsibility will eliminate the risk. However, if we accept, as suggested by Jonas, that there is a correlation between the power of technology and the responsibility for it, then this does begin to suggest collective responsibility on those seeking to share in the technological offerings. The collectivism is highlighted by Jonas,Footnote 28 who offers a moral imperative that stresses the weight of the responsibility. This is that the existence and essence of humankind should never be made a stake in the hazards of action. Shared responsibility suggests an opening up of the decision-making processes beyond expert determination to allow the expression of many voices and values about what sort of world we would have and people we would be.
This section of the paper has suggested that the power and scope of modern technologies infringe upon the natural world in such a way that there is a commensurate responsibility in the deployment of the technology. This is not (only) to do with risk as narrowly defined in terms of hazard, but is more to do with the type of future that we might endow. It has been suggested that rri can never be truly responsible until full account is taken of the social context in which technologies will operate.Footnote 29 If this is so, responsibility needs to be shared widely in processes of upstream engagement and in settings that go beyond expert groups.Footnote 30 Beyleveld and Brownsword have depicted a ‘community of rights’, which engages in a continuously reflective process on the best interpretations of its commitments.Footnote 31 Responsibility sharing offers no guarantee of a safe or salubrious future but it may negate extreme opposition to innovation. Meanwhile, the process may enhance anticipatory governance by conferring greater legitimacy. Since the turn of the millennium, there has already been a marked rise in the governance focus on economic and societal outcomes together with a greater preparedness for new technologies. This is explored below, but first it is necessary to consider at what scale such governance might be pursued.
3. SCALE
Science policy has traditionally been a matter for nation states, albeit at times in liaison with international organizations.Footnote 32 This pattern is changing. There is much more supranational collaboration such as that pursued by the eu Research Framework Programmes. At the same time, significant attention is given to the regenerative capacity of science and technology, creating strong regional interestFootnote 33 in the importance of knowledge and knowledge spillover for economic development.Footnote 34 Because disruptive technologies break through traditional disciplinary boundaries and call for fresh alliances, they may be seen as opening up opportunities for new clusters of research and development activity. Supranational and regional initiatives may be related: eu science policy stresses the contribution of regions of Europe in mobilizing research and development.Footnote 35
As science policy is expanding beyond and within the realm of the nation state, governance too will follow this trend. Converging technologies can give rise to a range of goods, infrastructure, and even services. The end product of technological development will influence the scale on which it might be regulated. For example, nanoscience has already given rise to a wide variety of products in a range of market sectors. The Woodrow Wilson Institute lists over 1,000 products which it has assessed as employing nanotechnological innovation.Footnote 36 These are not necessarily whole products but can include product components such as coatings for cars and clothing. There is strong representation of nanomaterials in the sectors of medicines, cosmetics, (agro)chemicals, and electronics, which were early starters in the field.Footnote 37
The propensity of such products to cross borders, not merely within Europe but on a global basis, poses the greatest challenge for their regulation.Footnote 38 In the absence of internationally agreed product standards, which is the present case with nanomaterials, restrictions on market circulation are usually justified by reference to risk assessment (risk, here, being narrowly defined by reference to hazards to human health). The wtoBiotech Products disputeFootnote 39 suggests that there may be little room for democratic self-determination when it is in seeming conflict with trade liberalization and even less acceptance that scientific risk assessment might be other than a policy or value-neutral process that will produce sound scientific decisionsFootnote 40 to which we can adhere. The world trade battleground for nanotechnology might be mandatory labelling requirements, which are actively being considered by the European Parliament as a means to inform consumers of the presence of nanomaterials, but which may constitute a technical barrier to trade.Footnote 41 The eu and the United States (us) are already pursuing divergent policies in relation to consumer labelling,Footnote 42 with real practical impact on importers in areas such as cosmetics.Footnote 43
Where technological innovation gives rise to infrastructure, one might anticipate that the location of the infrastructure will determine the point of regulation. For the most part this is so, though the location of that infrastructure may be hotly contested and the development itself might be resisted. However, because of the transboundary impacts of the infrastructure, there may be international agreements on operational issues and potential liabilities. The nuclear sector is a good example of this.Footnote 44 Concluding such agreements may be a slow and difficult process, but economic and environmental imperatives can drive agreement forward, as has been the case with carbon capture and storage.Footnote 45 Again, the eu provides an illustration of how supranational regulation may manage such development through harmonized processes of environmental impact assessment, even though ongoing supervision may be national or entail devolved responsibility, and even though the process involves continual readjustment of divergent approaches.
Technological innovation may give rise to service provision: cloud computing offers an excellent example of the move from product to service. In the case of cloud computing, the location of the infrastructure on which the service depends may be unknown to the end user though the service provision itself might be expected to be governed by national law. Innovations in biomedicine tend to be delivered as individualized services. In the case of reproductive technologies, national regulation, often driven from an ethical standpoint, might seem almost otiose because of the possibility of what Knoppers and LeBris have labelled ‘procreative tourism’.Footnote 46 Many states may not regulate for the advance determination of what should be permitted either through inability to reach a moral consensus or through an attachment to unhindered market provision of medical services. Even if one pursued concerted efforts to harmonize or approximate the law of individual states,Footnote 47 one might agree on the general principles (such as the inviolability of the human person) only to have such agreement unravel in discussions of what is meant by personhood.Footnote 48 Because of this, it has been suggested that procreative tourism forms an effective solution to the regulatory problem as well as one that leaves room for moral pluralism.Footnote 49 Note, however, that because of the notion of freedom of services, ethical determinations seemingly resolved at the national level with regard to access to biomedicine may suddenly become subject to scrutiny under supranational law. This happened in the case of Diane Blood,Footnote 50 who was allowed to travel to Belgium for assisted conception services, with the unlawfully obtained sperm of her dead husband, when the uk Court of Appeal employed ‘the trump of eu law to sweep aside the hand dealt by the uk Parliament’.Footnote 51
Finally, note that some technologies may produce hybrids that are neither purely goods nor services. Biotechnology in relation to gmo crops offers a good example. The seeds are a classic economic good. However, their repeated provision to the farmer, when combined with the instrumental purpose inherent in genetic modification, takes on the form of a service. Moreover, once planted, the resultant crops form part of the landscape and local environment in the manner of an infrastructure project. The Biotech Products dispute and its aftermath are instructive.Footnote 52 The dispute resulted from a de facto moratorium on processing applications for imports of gmos into the eu. The moratorium was adopted as a result of Member State resistance to gmos. But since the European Commission itself had never supported the moratorium but rather condoned it to placate the Member States, it was unable to muster a robust defence when the moratorium was being challenged before a wto dispute panel – instead, it denied the very existence of a moratorium. It might be thought that the wto Panel finding against the eu would whip the recalcitrant states into line, instead of which the European Parliament voted in July 2011 to give them greater freedom. Under the European Commission proposal,Footnote 53 the European-wide authorization system would remain but it would be open to Member States to cite, inter alia, agricultural and environmental effects and their potential socio-economic impact to justify a ban or restriction on gmo cultivation. It will be interesting to see if devolved administrations within Member States will look to exploit such an opportunity (as Wales would wish to do within the uk)Footnote 54 in a move that is in line with the growing possibilities of autonomous regions within Europe exercising differing policy choices.Footnote 55
In all of these examples I am trying to demonstrate, from experience, the difficulty of centring the regulation of technologies at any particular scale. Certainly, controls at the level of the nation state look increasingly problematic, though they may serve an important declamatory purpose in terms of what is considered tolerable. Even regulatory attempts at the supranational level, as within the eu, may fall foul of international agreements. Moreover, with some technologies (such as geo-engineering) international agreement could prove vital. At the other end of the scale, more local, bottom-up determinations of the choices offered by science and technology might support sustainable development, yet under the forces of economic liberalization it is unlikely that such choices can fully determine the conditions for markets in goods and services resulting from innovation. However, as the gmo experience shows, the realm of the local cannot be ignored if a market is to develop. All of this suggests that a complex, multilevel governance of science and technology will be required if rri is to oversee the shift of locally based, single discipline scientific research into global multi-disciplinary technological innovations.Footnote 56 The question, according to the European Commission, is not whether to deal with ‘challenges posed by new knowledge and its applications’, but how to do so.Footnote 57
4. GOVERNANCE
Science and technology have long been subject to regulation. The desire to regulate may be driven by wide-ranging motivations, but these generally concern fears of market failure. One obvious concern is that certain technological innovations might generate externalities, in the form of social or environmental costs that may not be internalized in the pricing of any eventual product generated by the technology. There may be other factors, too, for which the market finds it difficult to account, such as ethical concerns generated by the application of technology. In such instances law may be seen to state the limits of scientific research, though this has been the subject of hostile reactions in the past on the part of the scientific community. One example of this hostile reaction might be the reaction to the Warnock proposals in 1984,Footnote 58 which sought to place limits and controls on embryology in the uk. Interestingly, the eventual resolution of the perceived need for intervention was a command and control solution. The uk Human Fertilisation and Embryology Act 1990 introduced controls through licensing and gave rise to a regulatory agency, the Human Fertilisation and Embryology Authority (hfea), which the present government appears to regard as a quango that is no longer affordable. Plans to wind down the hfea were met with dismay by the same scientific community that had once opposed the very idea of such a body. The hfea was seen as fulfilling an important role in negotiating the boundaries of embryo research and assisted conception. In a telling comment, two senior gynaecologists suggested that ‘when these researchers must deal with a broader range of opinions about whether their research should be done, they are likely to regret the loss of a specialist regulator’.Footnote 59
Sceptics might cite this as an excellent example of regulatory capture in which the body created to regulate grows so close to the regulated sector that it becomes seen as friend and not foe. But this might equally be seen as a success in working down the Ayres and Braithwaite regulatory pyramid,Footnote 60 so that formal enforcement has given way to successful, consensual processes of compliance. The winding down of the hfea will not necessarily bring an end to licensing, but it does suggest that the command and control structures are seen as overly bureaucratic, expensive and burdensome, and that future modes of scientific regulation will take a rather different form. It is worth dwelling on why this might be so.
The turn from regulation to governance involves less top-down government and greater involvement of a wide range of stakeholders to achieve the desired ends, described as a move from ‘power over’ to granting ‘power to’ such stakeholders.Footnote 61 Such a model might be seen as better able to serve as a platform for more deliberative models of public engagement than have been discussed thus far.Footnote 62 Ironically, however, governance arrangements are the product of top-down modifications of the role of government in the late twentieth century whereby, rather than performing a providential, welfarist role, government is seen as a source of social intrusion and economic inefficiency.Footnote 63 In their replacement of nation states, market states are seen, in contrast, as enhancing choice, with choice becoming a source of authority in its own right.Footnote 64 The danger is, however, that technology can displace true rationality, as well as choices in favour of human rights and dignity, and replace this with techno-regulation that binds individuals to a code from which there is little or no opting out.Footnote 65
Notwithstanding such concerns, there may be few options but to invest in soft law modes of governance for a wide number of reasons. One is simply a matter of regulatory capacity for, as Stokes has observed, ‘the greater the role of private actors in the administration of regulatory measures, the lower the costs of acquiring information and reducing information deficits’.Footnote 66 A second issue of capacity is that more formal command and control regulation, coming as it does with criminal sanctions, depends on certainty in definition or metrics that may not yet be readily deliverable.Footnote 67 Soft law may be a wiser response to complex and diverse problems that are characterized by uncertainty, perhaps preceding hard law responses while leaving room for flexibility and responding to unintended consequences.Footnote 68 This is partly because of the mismatch of timescales between innovation and regulationFootnote 69 which suggest that, however soft the ‘law’, it may be preferable to a vacuum pending hard law measures. Vogel has suggested, however, that these private regulatory models must eventually be integrated into and reinforced by regulatory enforcement at an appropriate level.Footnote 70
There may also be a link between governance and scale. For example, the prospect of reaching international agreement in areas such as environmental law would appear to grow ever more difficult as its substance becomes entwined with areas such as development and trade liberalization. Hard choices might be more easily faced with soft law.Footnote 71 Soft law instruments in this context may take the form of standard setting rather than treaties. Equally, corporate social responsibility (csr) may generate voluntary subscriptions to codes of conduct, which might be (for example) local to a firm, nationally or transnationally sector-based, or covering a cross-section of industry within or across national jurisdictions. Supranational regulation may trigger global responses in what Heyvaert has described as a regulatory game of winners and losers;Footnote 72 soft law approaches are likely to be less prone to retaliatory action.
There is a relationship too between objective and governance. If deliberative, upstream processes are seen as part of the delivery mechanisms of responsibility, these are more likely to give rise to soft law instruments such as codes, voluntary agreements, action plans, resolutions, statements, certification schemes, standards and strategies. Responsibility itself suggests voluntary assumption rather than enforced action – soft rather than hard law approaches. The assumption of shared responsibility commensurate with the aspirations of many converging technologies suggests engagement outside of the democratic sphere of legislative action in more decentralized social settings: the sphere of soft rather than hard law. This is challenging and, to some degree, concerning to those who might fear the displacement of the rule of law by rule of technology.Footnote 73
5. CONCLUSION
Mapping out the contours of a responsible research and innovation framework is likely to be a protracted and problematic assignment. This paper tries to offer an overview of the terrain. It is rugged and somewhat indistinct, yet its silhouette is there. The objective of the framework must be to address broad societal concerns that stretch well beyond simple notions of hazard to include wider contingency and doubts about the very direction of some technologies, stirred by our awareness of their transformative capacities. The framework must be not only participatory but also anticipatory. It cannot be confined to balancing benefit against risk because it must concern itself not just with material wealth but with moral health.
We are all charged with this task, but the questions of the appropriate scale on which to pursue it are complex and the answers may vary in accordance with the innovations engaged. Because we can be sure that there is no single territory in which we can confront the challenges of ordering our technological futures, the exercise will be one of multilevel governance. This is because the task outstrips the remit of governments, certainly individually but probably even when they seek to act in concert. Partly as a consequence, soft law instruments must play a part but we need to ensure that they are the product of the upstream, inclusive engagement that forms the spine of responsibility. Soft law instruments are not in opposition to hard law variants, which may prove necessary with time. Great care must be taken that the use of soft law is institutionalized into research and development practices but that these mechanisms, as a result, do not become subject to capture. ‘Soft’ cannot mean ‘yielding’ if we are to leave room to rein in some of the potential of innovation, rather than have it reign over us.
