Law’s performativities: Shaping the emergence of regenerative medicine through European Union legislation

Methodology: Texts and documents; content and performativity

Qualitative, empirical socio-legal research has increasingly taken up methods of discourse analysis. One variant of discourse analysis is a strategy for examining how policy discourses congeal to make some perspectives more prominent than others. In STS studies of genetics, for example, discourse analytic techniques have been used to analyse governance initiatives, addressing ‘what sorts of social relations (these) policy documents are a part of; the authorial identity that they assume; and the dominant forms of representation of science, the economy and patients that they embody’ (Kerr, 2003). Concepts from literary theory are also sometimes applied to understanding the rhetorical construction of implied authorship and readership in texts produced in formal organisations, which can support particular sorts of claims to authority or factuality (Atkinson and Coffey, 2004). In addition, the concept of ‘genre’ from cultural and literary studies is relevant for elucidating consistent structural and formal features of texts. Conceptions of ‘genre’ have undergone a ‘social turn’, away from a focus purely on matters of formal and textual regularities to a view of textual documents as agencies of social action situated in enduring social structures and cultural contexts (Bakhtin, 1979). The legal texts produced by the EU’s political process can be understood as a genre in this sense, displaying highly consistent features of style, format and terminology, and being deeply embedded in and constitutive of an established politico-legal system.

Drawing on these approaches, we can approach a legislative text and document as entities whose content and methods of composition can be interrogated. We can ask questions such as: What social and political forces are inscribed in it, and how does it act as their vehicle? Which techniques are deployed in it, such as reliance on genre and use of narrative forms? What subjects and objects does a text assume, conjure up, or ignore? And, in reference to the case examined here, which socioeconomic, medical, ethical and technological aspects of regenerative medicine does it highlight and downplay?

Beyond its use as a vehicle for discourse, a document can also be examined for its position as an actor performing actions in the social world. This perspective leads us to ask, how does such a document accomplish its actions? In other words, what is its performativity? To approach this question, I revisit the famous (and amusing) lectures of philosopher J.L. Austin (1962) on How to Do Things with Words, and some of his later commentators. Austin’s focus was on the ‘ordinary utterances’ of everyday life – what later writers called ‘speech acts’ – in other words, on what speech does as opposed to what it says in a referential sense. This conception of language as action is well established in sociology and STS, having been developed through ethnomethodology and conversation/discourse analysis. Austin distinguished between three types of utterance: locutionary (for example ‘I feel approval for this wine’ – an act of stating something); illocutionary (for example, ‘I name this ship Endeavour’ – the utterance accomplishes the naming); and perlocutionary (for example, ‘I warn you not to disobey’ – the utterance not only produces the warning it mentions, it also conveys a sense of a future consequence to the recipient). The latter two are performatives. Austin discusses these in terms of what we might call power relations, though he uses the idea that the illocutionary and perlocutionary modes ‘derive their force’ in the former case from being located in a set of societal conventions, and in the latter from achieving an immediate interactional effect by reference to a future consequence. It is worth noting at this point that while speech act theory in general has been criticised for focusing too much on context-free, isolated sentences or textual fragments (see, for example, Heritage and Atkinson, 1984), Austin himself was not guilty of such an error. He emphasised that in order for linguistic practice to be understood it is necessary to consider ‘the total speech act in the total speech situation’ (p. 147, Lecture XII). This significance of sociolinguistic context is of obvious importance: if one is not authorised in a society to name ships, one cannot accomplish the action. It also applies to ‘situations’ such as those that occur in connection with the EU polity – a polity that produced the legislative document examined in this paper.

Austin’s focus on speech can be and has been extended to a broader range of other linguistic forms. Commenting on Austin’s approach to language, for example, Derrida argued that ‘writing extends the field and the powers of locutory or gestural communication’ (Derrida, 1988: 3; emphasis in original). Derrida developed Austin’s notion of the performative to apply it to the written word, and summarised his own analysis of the performative as language that ‘produces or transforms a situation’ (p. 13). In the discussion that follows, I shall focus on this elusive distinction between ‘producing’ and ‘transforming’ situations.

When Austin developed the notion of the ‘illocutionary force’ of utterances, he tentatively distinguished five types, of which two are germane to the analysis developed here: the ‘Exercitive’ and the ‘Commissive’ modes. The Exercitive mode Austin characterised broadly as ‘an assertion of influence or an exercising of power’ (Austin, 1962: 150). Examples of exercitive words, admittedly ‘a wide class’, that Austin gave include: appoint, dismiss, name, recommend, proclaim, enact, dedicate, claim, announce, authorise, sanction, repeal, veto, declare and consent. Commissive words include: permit, authorise, sanction, consent, and promise, undertake, plan, contract, propose, shall, engage, guarantee, adopt, espouse, oppose. Austin discusses the fragile difference between exercitive and commissive modes, noting that exercitives also ‘commit us to the consequences of an act’ (p. 158). So, many exercitives are also commissives because they commit one to a future course of action. Applying this analysis to documents, Cooren (2004) has argued, helpfully for the analysis that I develop, that illocutionary acts such as asking, forbidding, banning, allowing, committing, guaranteeing or authorising can be performed by documents that have ‘the force of law’ (p. 383), in other words constitutional, legal or official documents produced in a defined politico-legal system.

In summary, by distinguishing illocutionary (ship-naming) from perlocutionary (consequential) speech acts, Austin was clear in the way he distinguished the generative production of a performative effect from the constitutive accomplishment of a performative action ‘taking effect’. The latter can be understood to include acts such as the establishment of social principles such as rights, commitments, permissions, prohibitions and obligations, as in the ‘enactment’ of a new law.

The methodology used here thus treats the new EU regenerative medicine law in two ways: first, as a textual vehicle of discourse and conveyor of meanings, which is susceptible to detailed interpretation by means of various forms of structural, stylistic and content analysis; and second, as a material document that can be analyzed by developing a variant of the approach recently advocated by Lindsay Prior (2003, 2008), who draws on actor-network theory and proposes that documents can be treated not only as ‘containers of content’, but also as ‘active agents’ (Prior, 2008: 824). This point will become clear as the discussion progresses, but it should be emphasised that although I treat the legislative text and document as performing certain social and political actions, I do not claim that it does so autonomously, in isolation from either the political-legal or the technoscientific process that produces it, nor from the social, political and scientific actors who subsequently interpret and ‘act on’ it. My position here is similar to Cooren’s: ‘texts are not foundational; however, they participate, like other agents, in the daily production of organizational life’; and: ‘Ascribing agency to documents does not mean that texts are severed from their human creators’ (Cooren, 2004: 382). Nevertheless, I argue that texts and published documents carry a quantum of agency, and in what follows I show that a legislative text/document has several forms of agency.

Legislating for the field of regenerative medicine

The legislation that I examine here is a ‘Regulation’ for ‘Advanced Therapy Medicinal Products’ (ATMP) produced in the EU (European Parliament and Council of the European Union, 2007). It has undoubted major significance for the regenerative medicine field globally. A large number of international scientific conferences, meetings between product developers and regulators, and insider commentaries refer to the ATMP Regulation. For example: ‘Recent policy developments in the United States and Europe have created new regulatory regimes for advanced and complex treatments such as cell therapies, tissue engineering, and gene therapies that are of particular relevance for regenerative medicine’ (Sethe, 2010). Similarly, an oral presentation at a Commercial Translation for Regenerative Medicine conference held in London was entitled ‘Analysing the new advanced therapies legislation and understanding the changes to European regenerative medicine’ (Mackay, 2007); the presenter was the regulatory affairs manager for one of the largest biotech companies active in the field. Similarly, the British Parliamentary Office of Science and Technology issued a ‘note’ in 2009 entitled ‘Regenerative medicine’ which identified four sets of potentially relevant regulation, namely that for medical devices, transplantation, human tissues/cells and ATMP (Parliamentary Office of Science and Technology, 2009).

Having established the salience of regenerative medicine as the overarching framework in which actors in the field understand the ATMP Regulation, I next outline the political background from which the legislation has emerged.

A legislative document in the EU polity

The legislative document to be considered is a rich and highly condensed text. Under the EU’s political ‘co-decision’ procedure, each of the different administrative units, committees and working groups of the tripartite political system contributes to the process resulting in the final document: the European Commission (which proposes legislation), the European Parliament (which debates it) and the Council of the European Union (‘Council of Ministers’, which debates and finally votes on it, representing the political positions of Member States). In this context, the text is the outcome of a massive collective effort involving hundreds of people over several years, many of whom were acting in one or another representative capacity, thus bringing even more people, institutions and interests into the process. The published document crystallises the fine-grained detail of a socio-political and technoscientific process and its outcome at a certain point in time. Many phrases, clauses or paragraphs were subjected to intense debate, conflict, proposals and counter-proposals, as amendments were made in committees and in debates in the European parliamentary institutions. Short phrases and even single words can have large significance. In general, what might be called the ‘summative quality’ of this document differs from the discursive and repetitive nature of many policy documents that are intended to discuss and persuade audiences about particular concepts, strategies and proposals in a policy domain (see, for example, Iannantuono and Eyles, 1997). It is important to note that the text/document is a material object that is written and rewritten through the legislative process involving many European institutions. For example, the European Parliament’s Environment and Public Health committee (known as ENVI), which was the lead committee for debating the Regulation, proposed and debated a series of amendments and counter-amendments to the wording of the text. Similarly, the Working Party on Pharmaceuticals and Medical Devices, constituted within the Council of the EU, produced a series of debates that were recorded in footnotes and annotations to a draft of the document that had been proposed by the European Commission. These notes and annotations recorded Member States’ differing positions about the wording of the draft’s clauses.

The text of the legislation was pored over in fine detail, not only by lawyers but also by many of the parties engaged in developing technologies, such as scientists engaged in research and development (R&D) activities, regulatory affairs managers and strategy leaders from large companies, national regulatory authorities, hospital R&D managers, tissue bankers, and regulation and business consultants who provide advice to smaller companies. Indeed, a number of such participants interacted with each other and regulatory policy-makers in attempts to influence the wording of rules of engagement for the emerging field. Eventually, these efforts were sedimented into the final text.

My analysis uses the English-language version of the original text, which had been translated by the European Commission into 21 other languages. It was published by the EU in both paper and online form in the Official Journal of the European Communities (OJ) (European Parliament and Council of the European Union, 2007). According to a (performative) declaration in the document, it ‘shall enter into force on the 20th day following its publication’. My analysis starts with a summary of the issues that the Regulation addresses, including a brief overview of the main points of the history of its negotiation, before presenting the textual and document analysis that addresses the questions above about its content and performativity.

Negotiation, structure and legislative measures of the advanced therapy Regulation

The ATMP Regulation (European Parliament and Council of the European Union, 2007) was drafted through a series of consultations and negotiations, which began seriously in 2002 following earlier failures in attempts to include human tissue products in EU medical devices regulation. The concept of ‘advanced therapy’ was adopted in large part in order to align tissue engineering, for which there was as yet no EU-level regulation, with cell therapy and gene therapy, which already had been subject to pharmaceutical regulation. Tissue-engineered products typically combine viable cells with some form of manufactured material, often composed of polymers. This ‘regulatory pharmaceuticalisation’ has been analysed in detail elsewhere (Faulkner, 2009a,b). The European Commission consultation on the issue made a case for legislation that would bring together the three types of therapeutic product into a ‘coherent ensemble’ for regulatory purposes (European Commission, 2005). The Commission stated that these three types of product were distinguished by four shared characteristics: innovative manufacturing; scarce scientific and industrial expertise; the importance of traceability and risk management; and the primary participation of small and medium-sized enterprises (SMEs), hospitals and tissue banks. The European Parliament and Council of the European Union voted on and approved the Regulation under its ‘co-decision’ procedure in 2007, with implementation of pan-European and national Member State provisions due by the end of 2008.

The main features of the ATMP text, following its final ratification and publication of the document, can be summarised with the following list, which distils a range of commentators’ interpretations of its key aspects:

the ‘scope’ of tissue engineering as a technology is legally defined, alongside cell therapy and gene therapy;

EC/EU centralised authorisation for products is introduced;

a new expert scientific-technical Committee for Advanced Therapies (CAT) is created, based in the European Medicines Agency (EMA);

‘pre-certification’ of new products is introduced, allowing SME producers to obtain scientific assessments at an early stage in the regulatory process;

a substantial licensing fee-waiver for small companies is introduced;

hospital-based, individually prescribed products are excluded from the requirements of the Regulation (though they must still meet equivalent national safety and quality standards);

a requirement is introduced for 30-year traceability of each new material and patient, so that any adverse clinical event can be traced back to its source;

post-marketing follow-up surveillance rules are introduced, in order to allow the regulatory authority to monitor the safety of new products; and

‘technical requirements’ such as preclinical and clinical data requirements are delegated to ‘comitology’ (‘technical’ committees outside the parliamentary process) for future development.

The legislation shows, in abbreviated form, the position reached on each of these points. My summary list combines several types of action that are ‘performed’ by the document, including not only the formulation of a new regulatory infrastructure, including a new Committee and a substantive definition of tissue engineering, but also specifications of the way the legislation is presented in the form of rules, measures that are ‘introduced’, requirements that are ‘delegated’, and the committee which – at least in the legal realm – is ‘created’.

The Regulation is written and enacted in the format of a textual document. As Prior (2008: 822) has pointed out, many contemporary documents include audio recordings, images, diagrams and other non-verbal elements, but this document is comprised wholly of written language. The text displays a basic structure. The first half comprises 31 paragraphs, most of which set out the issues raised and the principles of justification on which the Articles of the legislation are allegedly based (I refer to these later as the ‘justificatory’ sections). Many of these paragraphs made up the ‘Explanatory Memorandum’, a typical part of European Commission draft legislation that was included in earlier versions of the Regulation. They can be regarded as outlining the EU’s worldview on regenerative medicine products in relation to industry and public health in EU member states. Although these paragraphs are part of the text, they do not themselves constitute laws. The justificatory section is followed by the legal pronouncements, consisting of thirty legal ‘Articles’, laying out the legal framework and the legislative provisions. These are followed by a performative statement that ‘[t]his Regulation shall be binding … and directly applicable in all Member States’, thus distinguishing the document from a Directive, whose text requires transposition into national legislation before it is binding. This section is followed by four Annexes, which deal, first, with types of cell and tissue ‘manipulation’; second, with a list of 30 items that must be included in written descriptions of any product characteristics; and then with lists of items of information required to be included on the outer packaging and in the package leaflet of any product.

The text as a whole can thus be regarded as having a narrative structure that includes an extensive though condensed statement of issues to be addressed and factors deemed relevant to the case in hand, followed by the legislative provisions designed to implement them. This narrative form is the established genre for EU Regulations and Directives. The Articles of the Regulation are divided into the following ‘chapters’ (a full list of the Article headings, by number, is given in Table 1).

OPEN IN VIEWER

Table 1.

Headings of the ATMP Regulation’s legal Articles

1. Subject matter

2. Definitions

3. Donation, procurement and testing

4. Clinical trials

5. Good manufacturing practice

6. Issues specific to medical devices

7. Specific requirements for advanced therapy medicinal products containing devices

8. Evaluation procedure

9. Combined advanced therapy medicinal products

10. Summary of products characteristics

11. Labelling of outer/immediate packaging

12. Special immediate packaging

13. Package leaflet

14. Post-authorisation follow-up of efficacy and adverse reactions, and risk management

15. Traceability

16. Scientific advice

17. Scientific recommendation on advanced therapy classification

18. Certification of quality and non-clinical data

19. Reduction of the fee for marketing authorisation

20. Committee for Advanced Therapies

21. Composition of the Committee for Advanced Therapies

22. Conflicts of interest

23. Tasks of the Committee for Advanced Therapies

24. Adaptation of Annexes

25. Report and Review

26. Committee procedure

27. Amendments to regulation EC 726/2004 (tissue and cells banking etc.)

28. Amendments to Directive 2001/83/EC (medicines for human use)

29. Transitional period

30. Entry into force

ATMP Regulation: Articles – Chapter headings

_________________________________________

Subject matter and definitions

Marketing authorisation requirements

Marketing authorisation procedure

Summary of product characteristics, labelling and package leaflet5. Post-authorisation requirements

Incentives

Committee for Advanced Therapies

General and final provisions

___________________________________________

As can be seen from this list, the emphasis of the Regulation is on provisions for placing new products in the healthcare marketplace, describing them to users and patients, following up and evaluating results of their clinical use, supporting innovative companies and other entrepreneurial establishments, and creating an institutional regulatory framework.

I now proceed to analyse the written, conceptual content of the text. I shall argue that the text exhibits a number of substantive presences and absences – both of actors and of concepts – which are evident in the field of regenerative medicine. I argue further that this selective morphology is shaping the EU regenerative medicine field during the early stage of its emergence.

Analysis of textual content

The approach to content analysis that I used was as follows. I searched the text for a wide variety of terms (such as industry, patient, public, product, risk, ethics, future), partly from my knowledge of the field acquired over a period of about 7 years of study. Passages containing such words were scrutinised for other, related terms. I also used a thesaurus program to identify possible terms, and scrutinised the text by reading it repeatedly it to locate other related and cognate concepts and terms. For interpreting the text, the frequency of occurrence of terms has some meaning but must be considered alongside their position in the narrative structure of the document as a whole and their conceptual significance in the regenerative medicine field. Conspicuously absent words, phrases, or themes also may ‘speak volumes’ about the text’s narrative profile and significance.

As noted above, the analysis is animated by questions about the relationship between techno-scientific innovation and regulation. My analysis pursues these questions with a specific focus on expectations and visions, the scope of the material technology of regenerative medicine, regulatory measures and institutions, and the ‘rules of engagement’ for participants. Content analysis can shed light on several questions about regenerative medicine. Do actors in this scientific and regulatory field orient themselves to a real or imagined regenerative medicine industry or sector? Which types of actors take, or are given, responsibility for regenerative medical technologies, and what relationships of power or accountability do they have to each other? What relationships between users, producers and regulators of these technological products are envisaged? How are the involvement of patients, the wider society, and various other publics represented?

Law’s performativities

In this section, I move from the content of the legislation to focus on its performativity, and take as a starting point the notion that legal texts and documents act, and have consequences, given the political provenance inscribed in them. These actions and consequences, I argue, are two key, distinct forms of performativity exhibited by the Regulation, which can be identified as different forms of sociopolitical and technoscientific action. It bears repeating that this focus does not deny the agency of the sociopolitical actors who produce such documents, nor does it deny their salience for bringing about practical, observable consequences. However, the document and its publication effect new actions and generate effects, given the context of a specific political system.

Up to this point, I have introduced and described the content of a new piece of legislation and outlined its formal characteristics in terms of a narrative structure. As my content analysis showed, a legislative text can enshrine expectations about a new technology and provide (or at least, provide for) social legitimation for its producers and products. It can at least partly accomplish its stakeholders’ aims of introducing a set of principles and an institutional regulatory set-up for controlling and monitoring a field. Indeed, it plays a part in constructing the field as regulatable. In order to analyse the accomplishments of the regulatory document, I shall now distinguish more carefully between the different types of performativity that are evident in it.

Performativity draws our attention to what is being done to the relationships between speakers or writers of texts/documents and their audiences or readers – what Jürgen Habermas referred to as the ‘level of intersubjectivity’ (Habermas, 1976). How are these performative actions accomplished; actions that shape the intersubjective relationships between the emergent actors and technologies of regenerative medicine? As noted earlier, Austin’s work provides some helpful distinctions. One key distinction is between generative performativity (perlocutionary, consequences) and what I shall now call ‘enactive’ performativity (illocutionary, ‘ship-naming’, effecting actions). Enactive performativity refers to a textual- or speech-act that takes effect by virtue of its own production, in this case by virtue of its publication as a vehicle of a politico-legal process, actually accomplishing ‘what it says on the label’.

So far I have described actions of the text/document, such as responsibilities being ‘created’, the public being ‘constituted’, and how the new law ‘defines the scope’ of tissue-engineered or advanced therapy products. These are all performative concepts, but of different types. Recalling the dual structure of the Regulation – divided as it is between the explanatory ‘justificatory’ text and legislative ‘Articles’ – we can see that the less complex vocabulary of the Articles indeed ‘enacts’ new law, for example by accomplishing the legal formulation of a new regulatory committee and defining rules for the material constitution of a new type of technological product (tissue-engineered products). These actions would not have had legal force if the text had not been voted on in the parliamentary process and the document had not been published as a result of that process. In Austin’s terms, such actions are exercitive, as they involve the exercise of power in a ‘conventional’ speech situation (in this case, they involve the genre of EU legislative documents produced by the EU polity). At the same time, the Regulation is partly generative and partly ‘commissive’ in the way it commits actors in the regenerative medicine field to engage in the future with, for example, the new regulatory committee that will make authorisation recommendations for new products.

In contrast to such legislative performativity, how should we understand what I called the ‘constitution of the public’ – a fairly typical formulation in social constructivist or co-constructivist STS (e.g. Braun and Schultz, 2010)? I noted earlier that ‘the public’ appear in the text, though as a distant presence. Certainly, the Regulation acts in a performative way when ‘constituting the public’. The concept of enactment appears more appropriate as a reference to the constitution of an abstract citizenry, rather than to the generation of consequences, but in this case there is no legislative act or constitution of a legal entity to consider. The justificatory preambles in the document, such as the reference to an unspecified ‘industry’ that will produce regenerative medicine products and try to get them authorised, also perform enactment. However, in these cases we have a different form of ‘enactment’ – one that is common in many current STS formulations. Thus it is useful to distinguish between a ‘legislative’ mode of enactment and what can be termed a discursive ‘social’ or societal mode of enactment. The latter includes the constitution of publics or actors – for example, patients or medical professionals, as well as an expectation that products will be brought forward by an undefined, un-named industry.

On the basis of this analysis, I propose that it can be helpful to approach the various bodies of scholarship that have taken up concepts of performativity and enactment (noted in the introduction) in terms of the different forms of performativity that I have identified. First, we can distinguish between generative and enactive forms of performativity, and second between legislative and ‘social’ forms of enactive performativity. This analysis of legal performativities may have general implications for the approach to the analysis of documents in sociology and STS, such as Prior (2008) suggests when he applies an actor-network approach to documents-as-non-human-actors.

Prior argues that ‘once a text or document is sent out into the world there is simply no predicting how it is going to circulate and how it is going to be activated in specific social and cultural contexts’ (Prior, 2008: 824). Prior’s assertion is valuable for pointing to the flexible ‘activation’ of documents. It is clear, on the one hand, that the legislative document considered here is generative of ‘schemes of organisation’ (Prior, 2008: 827): designs for future social, economic and regulatory relationships, which comprise the regulated and regulating actors of regenerative medicine. However, this formulation does not fully capture how the legislative text/document acts. To say there is ‘no predicting’ modes of activation may be an over-generalised statement about the social mobilisation to which some classes of document are susceptible. There may not be such a ‘free market’ in the active consumption of legislative texts as there is with other texts, because of the direction and ‘context’ (‘total speech situation’ in Austin’s (1962: 147) terminology) of their performativity. In a similar vein, it has been argued that texts are ‘active’ in that they predispose readers to a given interpretation (Watson, 1997). To adapt a term from innovation studies, enactive legislative documents, by virtue of their status as published vehicles in a politico-legal system, may be regarded as having a condensed and directive quality that leads to a measure of lock-in to a limited range of possible societal uses, effects and functions. Such effects are related to the document’s obvious framings of technologies and actors, and its social position at a particular point in history. The document’s legislative enactive performativity, especially, constrains the possible multiplication of different actions through its circulation in different ‘social and cultural contexts’ (Prior, 2008: 824). This form of ‘exercitive’, enactive performativity is particular to its political, legislative context. It also presents commissive ‘designs for’ future organisation. The publication of the document should thus be seen as enacting the new organisation that it legislates for in the text (though it does not, of course, implement it into the variability of social practices). It is more forceful, societally legitimate and effect-creating, than a mere proposal or plan of action, and more strongly constitutive than many other types of documents produced through lesser degrees of authoritative socio-political negotiation – domestic shopping lists, for example. As Austin put it, such acts ‘take effect’ rather than merely bringing about a certain state of affairs (Austin, 1962: 116). This formulation echoes Derrida’s (1988: 13) distinction between ‘transforming’ and ‘producing’, discussed earlier, and encapsulates the crucial distinction that I have shown between generative performativity (producing) and enactive performativity (transforming).

Given the identification of these different types of performativity, it is now possible to turn to some key feature of the legislative text and document under consideration, which have important generative implications and enactive effects in the regenerative medicine field, both in Europe and globally.

Conclusion

On 9 October 2009, the EMA, drawing on the advice of its newly constituted Committee for Advanced Therapies, approved for market the first product under the regime established through the publication of the ATMP legislative document. The product, called ChondroCelect®, is a knee cartilage regeneration technology produced by a Belgian company. The decision meant that the product could be marketed throughout the EU without recourse to individual member states’ national regulatory agencies. By contrast, the US Food and Drug Administration (FDA) asked the company for evidence from an additional study, which delayed the marketing and availability of the product in that country. The performativity accomplished by publication of the ATMP document thus extends to ‘downstream’ material effects in the emerging regenerative medicine field.

In summary, I have unpacked the different forms of performativity of a single legislative document, which I treated as a vehicle for a politico-legal process. My analysis identified a mixture of presences and absences, clarities and imprecisions, and enactments, provisional actions, and inactions. By focusing on a single document, I have not given as much attention to intertextual aspects and institutional, organisational ‘contextual’ agents. Likewise, I have not given much attention to questions of how a legislative document might sustain its performativity over time. Such a focus would require attention to how states implement law, among other matters. I attempted instead to highlight actions that the text and document – through its enactment and publication – did accomplish. I have argued that this document is an example of a type that is an important constituent of a polity’s toolbox for shaping, locking-in and legitimating an emerging technoscientific field. The legislative text enshrines stakeholders’ expectations for a document that generates consequences, for example in the form of participants’ rights and obligations, a regulatory body, and ultimately new regenerative medicine products (a type of material effect not countenanced by J.L. Austin). Furthermore, I have shown how the document legislatively enacts various ‘rules of the game’ in this sector-in-the-making, such as the definition of tissue engineering technology and certain patients’ rights within the geopolitical space of the EU. And, finally, it has socially enacted further, more diffuse matters, such as an open-ended vision of the future of a safe and publicly acceptable regenerative medicine industry.

In this paper, therefore, I contribute to clarifying the different types of performativity exhibited by and through a particular type of textual document, which might be useful more widely in studies of the shaping of techno-scientific governance processes, such as those mentioned in the introduction to this paper. Further, and more broadly for STS scholarship, my analysis is in line with Barad’s call for theorisation of performativity that ‘allows matter its due as an active participant in the world’s becoming’ (Barad, 2003: 803). My argument contributes to this project of improving our understanding of the relationships between materiality and discursive practices, by showing how legislation acts through both textual discourse and through published material documents.

Finally, the introduction to an inspiring analysis of ‘active texts’ presents a fascinating list of some 40 different types of text, ranging from tattoos and paper money to graffiti and railway timetables (Watson, 1997), but it does not mention legislative texts. Given the analysis that I have developed here, it is clear that legislative texts and documents could be accorded a more prominent place in theorising the emergence of new biomedical and other sociotechnological fields. It also draws attention to the key importance of the act of publication for the generation and enactment of effects, through a legitimate, authorised process. My analysis shows that the authorship of the Regulation prefers the flexibility of neutral, abstract, de-contextualised concepts to more substantive concepts, so that the boundaries of the networks of actors and actions producing regenerative products remain constructively undefined. The discussion of different forms of performativity shows, finally, that a legislative document does much more than legislate. The ATMP regulatory document brings into being – enacts – socioeconomic actors, regulatory rules and definitions of technologies, and it also generates a vision that helps constitute the future of regenerative medicine in society, whatever form that might take.