The Autopoietic Turn in Habermas’ Legal Philosophy

Since the publication of Between Facts and Norms, it appears to have gone largely unremarked by crit‐ ical theorists that Habermas has wholly abandoned the neo‐Parsonian systems‐theoretic account of so‐ cial systems developed in his middle period. In this paper, I argue not only that Habermas has developed a new vocabulary and theoretical structure for dealing with social systems such as law, but also that this new approach to social systems is incompatible with the neo‐Parsonianism of social sys‐ tems developed in that early work. This is not to say that Habermas’ work in Between Facts and Norms should be set aside, but rather that the Parsonianism of The Theory of Communicative Ac‐ tion must be abandoned if Habermas’ jurisprudential project is to go ahead. In these pages, Poul Kjaer argued that not only did the Habermas/Luhmann debate con‐ tinue past the publication of their joint book in the 1970s, but that important elements of Luhmann’s theory survived into the architectonics of Habermas’ Between Facts and Norms (BFN).1 In particular, Kjaer asserts that: “one outcome of the Habermas/Luhmann debate is that the late Habermas’ discourse theory can be regarded as a normative superstructure to Luhmann’s descriptive theory of society.”2 Following Kjaer, I will ask what the conse‐ quences of this might be for Habermas’ earlier Parsonian‐inspired systems theory. I will argue in this article that BFN is inconsistent with the binary model of society Habermas developed in his earlier work. In particular, I contend that Habermas has changed the system‐lifeworld model in his theory of law, even if he is loath to admit it. Ultimately, my interest is to show that the original system‐lifeworld model is unsuitable for Habermas’ legal theory and that it has, in BFN, been tacitly abandoned, and marks the end of the para‐ digm developed in The Theory of Communicative Action (TCA). This is not to condemn BFN, but to argue that Habermas has turned to an autopoietic account of social systems. Many theorists have noted the introduction of some autopoietic terminology into Habermasian critical theory.3 I will argue not only has autopoietic terminology been introduced, but that the two versions of systems theory are fundamentally incompatible and that a choice must be made.4 The system‐lifeworld model is relatively well‐known to social theorists. However, as the purpose of this paper is to criticize the use of the systems component of that model in Haber‐ 1* The author would like to thank his research assistant Kafumu Kalyalya. In addition, earlier versions of this article benefited from helpful comments from David Lea, Thomas Simon, Amelie Wirts, Ken Baynes, Jeff Flynn, and David Strecker, as well as from two anonymous reviewers. Jürgen Habermas generously com‐ mented on an early version of this argument. Any mistakes, of course, remain the authorʹs. 1 The collection of articles was published as: Jürgen Habermas/Niklas Luhmann, Theorie der Gesellschaft oder Sozialtechnologie – Was leistet die Systemforschung? (Frankfurt am Main 1972). 2 Poul Kjaer, Systems in Context: On the outcome of the Habermas/Luhmann debate, Ancilla Iuris (2006), 66– 77, 77. 3 See, for instance, Hugh Baxter, Habermas: The Discourse Theory of Law and Democracy (Stanford 2011). 4 While I personally favor an autopoietic model of social systems, I will not make that argument here. Oth‐ ers have argued against the introduction of neo‐Parsonianism, albeit for other reasons. For a representative critique of the neo‐Parsonian model, see the essays by Thomas McCarthy/Hans Joas and others, in: Axel Honneth/Hans Joas (eds.), Communicative Action: Essays on Jürgen Habermas’ The Theory of Communi‐ cative Action (Cambridge 1991). by Kevin W. Gray* Kevin W. Gray – The Autopoietic Turn in Habermas‘ Legal Philosophy 42 ANCILLA IURIS (anci.ch) 2014: 41 – Article mas’ reconstruction of law, I will draw out what I take to be the salient points of Habermas’ systems theory in the first section by discussing Habermas’ reception of Parsons, before mov‐ ing onto Habermas’ theory of law in both TCA and BFN. The model Habermas develops is idiosyncratic in two ways: unlike Parsons, he relegates systems theory to only one part of action coordination. Similarly, Habermas’ use of the term lifeworld has little in common with the better‐known Husserlian or Schutzian models of the lifeworld. For Habermas, the life‐ world is the collection of background norms and structures (including modern learning pro‐ cesses which occur inside detranscendentalized morality, in the social and natural sciences, etc.) that guide communicative action in the modern world. Structures in which communica‐ tion predominates are lifeworld structures; conversely, social systems are those regions where action coordination through replacement media dominates. I. Social Systems in The Theory of Communicative Action TCA sets out to develop a model of the structural differentiation of society which, while building on the Frankfurt School’s appropriation of Weber’s work on the rationalization and bureaucratization of society, conceives of both a realm of communicative freedom in modern society and a region of modern society integrated by instrumental action. That realm of instrumental action gives rise to the system theoretic account of social activity that draws heavily on the mature work of Talcott Parsons.5 From Parsons, Habermas draws the conceptual architecture of the action system. A social system, in Parsons’ theory, is quite simply any pattern of action which has an internal analyt‐ ical structure.6 Each social system must be integrated by a steering media which regulates exchange between that system and its environment. For instance, money regulates exchange between the economy and other social systems. Particularly in his later writings, Parsons argues that it is, at least in principle, possible to find similar media which integrate all social systems. Money serves as the model for all such media. According to Parsons, the general problem is to decide whether such mechanisms of exchange “could be generalized beyond the case of money to that of other media.”7 What holds true of money holds true, mutatis mutandis, for other media. 5 In several places in his work, Habermas acknowledges his debt to Talcott Parsons. He remarks, for instance, that: “any theoretical work in sociology today that failed to take account of Parsons could not be taken seri‐ ously,” Jürgen Habermas, The Theory of Communicative Action, vol. 2 (Boston 1984), 199. Over the course of his career, Talcott Parsons, like Habermas himself, transitioned from an earlier Weberian model of social action to a later embrace of systems theory. It is Parsons’ mature systems theory that comes into play in TCA. 6 Broadly, by an internal analytic structure, Parsons intends any social system which makes use, in order for its continued existence and function, of structures which provide for four essential properties: adaptation (i.e. responses to the environment and acquisition of sufficient resources), goal attainment (i.e. setting and obtaining different goals), integration (i.e. how different sections of the system are integrated and how soli‐ darity is maintained) and latency (which he sometimes calls pattern maintenance – i.e. the process by which institutions, organizations, and roles allow the system to continue to function by furnishing and maintaining the necessary moral and cultural preconditions). Collectively, these form Parsons’ famous AGIL paradigm. See Talcott Parsons, The Social System (New York 1951). 7 Talcott Parsons, Review of Harold J. Bershady’s Ideology and Social Knowledge, in: Talcott Parsons (ed.), Social Systems and the Evolution of Action theory (New York 1977), 128; cited in: Jürgen Habermas, The The‐ ory of Communicative Action, vol. 2 (Boston 1984), 257. Of the four AGIL components, only some are ‘left’ to social systems in Habermas’ system‐lifeworld model. The IL components are left to the lifeworld in Habermas’ model. Kevin W. Gray – The Autopoietic Turn in Habermas‘ Legal Philosophy ANCILLA IURIS (anci.ch) 2014: 41 – Article 43 In Parsonian systems theory, systems regulated by money possess a number of important features: (1) Money is only good in specific situations (i.e. the economic system is limited in scope); (2) Action orientations are defined by clear interest positions including the fact that mo‐ ney possesses a generalizable value; (3) The alter (the other party) in an exchange can offer one of only two responses (he or she can accept/reject the offer); (4) The ego (the actor) can steer the alter’s responses through offers; (5) Actors are oriented strategically, that is, to the success of the transaction; and, (6) The media itself can be measured and alienated in amounts of any size, and can be stored.8 Money acts in social systems by replacing the specific action‐coordinating effects of lan‐ guage which occur in the lifeworld. It is a substitute for the “special functions of language.”9 Even though some interactions based on monetary exchange reproduce elements of lan‐ guage, those interactions set aside other aspects of language use (condensing and simplifying the full demands of communication). For instance, actors in situations regulated by money do not raise criticizable validity claims that are anchored in those lifeworld contexts which ground all communicative action; however, they do engage in strategic action (albeit strate‐ gic action of a different sort). Parsons attempts to generalize from the theory of the economic system to a general the‐ ory of social systems. He first attempts to generalize the model to account for power dynam‐ ics inside the political system, and to show that power exhibits the same structural features as money.10 Superficially at least, money and power appear to share several commonalities. Like money, power represents the symbolic embodiment of value “without itself possessing any intrinsic value.”11 The value of money in exchange situations corresponds, in the field of power, to power’s use value to realize collective goals. For instance, utility in money corre‐ sponds to goal attainment through power. Similarly, both money and power are anchored in structures in the lifeworld that allow them to detach from structures of communication and anchor social systems. Money is backed, at least in the first instance, by gold and the promise of interconvertibility; power, on the other hand, is anchored (weakly) in the lifeworld through the existence of uniforms, emblems of authority, official seats, etc., all of which serve to legitimate the exercise of power in the administrative system.12 Although he ultimately includes the administrative system (qua system) in his systems‐ theoretic account of society, Habermas thinks that Parson’s attempt to generalize the theory of media is stretched almost to the breaking point by the inclusion of power. First, while there is a system of signs in the lifeworld which anchors power (e.g., insignia, seals, etc.), it is 8 Jürgen Habermas, The Theory of Communicative Action, vol. 2 (Boston 1984), 264–265. 9 Ibid., 263. 10 Ibid., 257. After this, he will also try to analyze influence (in the system of social integration) and value commitment within the same theoretical framework. 11 Ibid., 268. In his definition of the administrative system, Habermas borrows this definition of power from Parsons. 12 Ibid., 269. Kevin W. Gray – The Autopoietic Turn in Habermas‘ Legal Philosophy 44 ANCILLA IURIS (anci.ch) 2014: 41 – Article less well‐organized than the system of prices; power is not as readily circulatable as money; and it is not as well‐institutionalized as money. Second, legitimation is a larger problem with power than with money. Originally, money was secured through gold; now, it is backed by the full faith and credit of the state. Power, however, is legitimated in a less obvious way. In his writings, Parsons is content to say that power is legitimated through the exercise of public office. In a modern state, Habermas responds, this is not enough. Parsons misses the fact that power is legitimated not through de facto compliance, but must also be legitimated through the recognition of mutually‐valid obligations: It is “a duty based on the recognition of norma‐ tive validity claims” which are grounded in the lifeworld.13 A cursory examination of the typical exchange situation highlights the problem. When monetary exchange is institutionalized, both parties arrive as agents freely able to agree to, or disagree with, a proposed transaction. Ultimately, there must be reciprocity in exchange (or the possibility thereof) or else a medium’s exchange value cannot remain constant. For power to be a suitable systems media, there must be some similar exchange situation wherein par‐ ties agree to the exchange of power. If not, then the exchange of power would cease to be stra‐ tegic, and would instead have to be backed by normative agreement. Clearly, however, both modes of power exist in the modern world – sometimes power is part of the lifeworld (and is backed by normative agreement anchored in communication) and sometimes it is a systems media (backed by strategic action). Habermas concludes, contrary to Parsons, that “[i]t is only the reference to legitimizable collective goals that establishes the balance in a power relation built into the ideal‐typical exchange relation from the start.”14 The only place where power truly functions as a systems media is in the administrative system (and not in the broader political arena). Although he is skeptical of the parallels between money and power, Habermas does in the end accept that power is a systems media. However, it is with Parsons’ attempt to broaden the theory of media further (i.e. beyond the administrative system) that Habermas becomes most critical. Beyond power and money, Parsons wants to include, among the list of systems media, influence and value commitment.15 At first, it appears that treating influence and value commitment (which is effectively moral authority) as systems media is plausible. One could imagine influence being something possessed by an agent that he or she can use to change the convictions of others. A powerful speaker, a demagogue, or someone held in esteem by others might possess sufficient moral influence to steer group decision‐making, etc. Similarly, for value commitment, respected members of the community may be able to influence members to make decisions based on their moral standing. However, modern society provides no structural equivalents in the lifeworld to anchor influence or value commitment as it does for the economy or the administrative system.16 Moreover, systems theory predicts that as modern society becomes structurally differenti‐ ated, and there is a growing institutionalization of separate action systems in modern society, clearly demarcated realms integrated by influence or value commitment should grow or become more prominent. With respect to influence or value commitment, that does not seem to have happened. In fact, Habermas argues, as societies develop, influence (prestige) and 13 Ibid., 271. 14 Ibid., 271. 15 In Parsons, influence is the media for the societal community, value commitment for the fiduciary system. 16 Moreover, there could not be if Habermas intends to keep intact the role played by communication in action coordination in the lifeworld. Kevin W. Gray – The Autopoietic Turn in Habermas‘ Legal Philosophy ANCILLA IURIS (anci.ch) 2014: 41 – Article 45 value commitment (moral authority) decline. In other words, the Weberian model of mod‐ ernization, which predicts the differentiation and growth of subsystems, fails with respect to influence and value commitment (if they are to be considered social subsystems). In other words, Parsons’ claim is not generalizable.17 Further, additional problems remain with the attempt to expand the model of monetary exchange. Influence and value commitment fail many of the other tests that Parsons proposed for steering media – for instance, they are not ‘bankable’ in the same sense as money. Their value is not constant in the same way that money is. As such, Habermas believes that influence functions more broadly in modern soci‐ ety than do other systems media. Instead, it is better to speak of influence performing an inte‐ grative function only in situations where a group is manipulated strategically (by a process which is parasitic on the communicative potential of the lifeworld), for instance, when a char‐ ismatic leader makes claims and these claims are accepted because they are backed by the possibility of being cashed‐out communicatively. In general, influence is accepted when “influential persons and institutions meet with a willingness in their audience to take advice.”18 People obtain authoritative power based on the motivating power of consensus that exists inside the group. Something similar holds true, Habermas believes, when leaders’ words are given a moral force. Arguments from prominent individuals are treated as moral arguments, without detailed reasons being given and without overt demonstrations of legitimacy only if “their utterances [...] function as authoritative in virtue of their critical‐appellative force.”19 In other words, for influence to be operational, it cannot simplify social integration through the addi‐ tion of some sort of reserve media; influence will ultimately remain parasitic on communica‐ tion. At most, Habermas argues, Parsons has identified two different types of action coordina‐ tion that exist inside society, and two social systems: coordination through non‐communica‐ tive media (e.g., money and power, manifested in the economy and the administrative sys‐ tems) and coordination through communicative action (or the possibility of it). In the transition from aristocratic class societies in European feudalism to capitalist class societies of the early modern period, some symbolic structures in the lifeworld became rei‐ fied, and so modes of social reproduction become differentiated out via media. However, in his turn to systems theory, Parsons failed to develop a theory that was able to explain all types of action orientations because he possessed a non‐nuanced concept of the rationaliza‐ tion of the lifeworld. Parsons assumed that the rationalization of the lifeworld would lead to the rationalization of all forms of social organization in ways that would create social sys‐ tems; thus, he explained the development of society only as the development and increasing complexity of systems.20 Habermas believes that for a systems media to exist, it must satisfy the six conditions described above. In Section 5, I will use that argument against Habermas and argue that Habermas’ model of law in BFN also cannot meet these conditions. 17 Jürgen Habermas, The Theory of Communicative Action, vol. 2 (Boston 1984), 275. 18 Ibid., 273. 19 Ibid., 273. 20 Ibid., 284. Kevin W. Gray – The Autopoietic Turn in Habermas‘ Legal Philosophy 46 ANCILLA IURIS (anci.ch) 2014: 41 – Article II. Law in The Theory of Communicative Action In his initial discussions of law, instead of treating law in the same neo‐Parsonian frame‐ work as he does money and power, Habermas adopts a largely Weberian view of law, whereby law becomes an instrument of domination which exercises its power through a series of legal authorities and enacted regulations.21 In a thoroughly rationalized modern society, Habermas contends, law necessarily becomes the pre‐eminent social institution, tak‐ ing over the functions of social integration in a society cleaved from traditional worldviews. As Habermas was, when writing the sections on law in TCA, concerned to explain the devel‐ opment of the pathologies of the capitalist state, his treatment of law in TCA occurs as part of a theory of the development of the interventionist welfare state. Thus, Habermas’ neo‐Webe‐ rian view of the development of law argues that social integration and legitimation occur in four stages which parallel the development of the capitalist state – a process which is effec‐ tively the gradual constitutionalization of a system of rights that emerges out of the rise of the capitalist economy. These four distinct stages of state and legal development are: (1) The Bourgeois State, which emerged through the institutionalization of structurally differentiated systems organized by the steering media, codified in a system of law designed to protect the interest of the rising bourgeois (capitalist) class; (2) The Constitutional State, which emerged through efforts to codify, and control, exe‐ cutive power in the face of the demands of the bourgeoisie; (3) The Democratic Constitutional State, which emerged through the codification of pro‐ cesses of legitimation; and, (4) The Democratic Welfare State, which emerged as a result of the codification of sub‐ sistence rights (e.g., a minimum standard of living, rights to education, etc.).22 Each stage in the development of the modern state corresponded to the progressive development of the economic system and the simultaneous development of law. In the first two stages, law developed merely to check the economic and administrative systems. Only in the third stage of the development of the modern state do we begin to see the democratiza‐ tion of processes of legitimation. This tertiary stage of law ensured the democratization of the state, guaranteeing rights of political participation through law (such as occurred in the many European states in the mid‐19th century when liberal reformers successfully argued for the expansion of suffrage.23 These stages in the development of law correspond to what Weber calls Verrechtlichung (or juridification), whereby the domain of law not only becomes larger but more internally complex.24 “Given the interventionist requirements of an avowedly active government that both steers and compensates, the functions and internal structures of the legal system are altered as well. Law as a generalized medium is not only more widely utilized; the form of law also changes according to the imperatives of a new kind of requirement.”25 21 Jürgen Habermas, Law and Morality, The Tanner Lectures on Human Values, 219, http://tannerlec‐ tures.utah.edu/_documents/a‐to‐z/h/habermas88.pdf, last access: 18 December 2013. 22 Jürgen Habermas, The Theory of Communicative Action, vol. 2 (Boston 1984), 360; italics omitted. The whole discussion occurs at 358 et passim. 23 Ibid., 360. 24 And thus, for the first time, law becomes structurally differentiated from the lifeworld, allowing us to begin to speak of a true legal system. 25 Jürgen Habermas, Law and Morality, The Tanner Lectures on Human Values, 220, http://tannerlec‐ tures.utah.edu/_documents/a‐to‐z/h/habermas88.pdf, last access: 18 December 2013. Kevin W. Gray – The Autopoietic Turn in Habermas‘ Legal Philosophy ANCILLA IURIS (anci.ch) 2014: 41 – Article 47 While the first wave of juridification checked the power of the administrative system, it did not reduce the fundamental tensions inside the capitalist (economic) system: the proletar‐ iat remained at the mercy of the owners of the means of production. No guarantees of a right to a minimum level of subsistence, for example, emerged. Only the emergence of the welfare state changed this. While preserving class structure, it brought about welfare state protec‐ tions (e.g., a well‐defined work day, a broad prohibition on child labor, etc.).26 The development of law is linked to the development of the welfare state and progressive political changes. However, as Habermas and others (notably Gunther Teubner) have noticed, while the juridification of the lifeworld has positive consequences, it is not unambig‐ uously positive. Conceived in terms of the encroachment of systems on the lifeworld, “social modernization [occurs] at the expense of subjection to the logic of the system, and the destruction of intact social structures.”27 While juridification affects different areas of the life‐ world, in his writings Habermas focuses on the core areas modified by the welfare state: edu‐ cation, family law, and other interpersonal relations. To take one example, as family law evolves, legal changes alter the structure of the family. A system of geriatric care emerges to specify the rights of the elderly and to provide for their health in old age. However, family law also has “burdensome consequences for the self‐image of the person concerned, and for his relations with [his] spouse, friends, neighbours, and others.”28 Interpersonal relations become institutionalized through a system of economic compensation rather than through traditional social obligations and lifeworld contexts of interaction.29 The expansion of law permits the entry of systems into lifeworld contexts. In reality, juridification is actually a form of what Habermas discusses elsewhere as the colonization of the lifeworld by systems media (which means that we should be able to speak of the legal system in the appropriate neo‐Parsonian way). The colonization of the lifeworld is a broader social phenomenon that occurs in modern society, whereby actions that were formerly coordinated through purposive‐rational or communicative action become coordi‐ nated by substitution media (i.e. by money or power). The colonization of the lifeworld is normally a negative phenomenon, and here it does have negative effects: It promotes the dis‐ integration of life‐relations which are separated, through legalized social intervention, from the consensual mechanisms that coordinate action.30 Nonetheless, welfare‐state guarantees are intended to serve the goal of social integration and do counteract the extreme effects of economic expansion. In this way, the colonization of the lifeworld through law is different from its colonization by the economic and administrative systems, because legal colonization stands in opposition to system colonization as a partial defense of the lifeworld. 26 Jürgen Habermas, The Theory of Communicative Action, vol. 2 (Boston 1984), 361. 27 Gunther Teubner, Juridification – Concepts, Aspects, Limits, Solutions, in: Gunther Teubner (ed.), Juridifica‐ tion of Social Spheres: A Comparative Analysis in the Area of Labor, Corporate, Antitrust and Social Wel‐ fare Law (Berlin/New York 1987), 4. Of course, some areas of the lifeworld are not altered by juridification but by processes of rationalization that occur prior to rationalization. 28 Jürgen Habermas, The Theory of Communicative Action, vol. 2 (Boston 1984), 362. 29 For example, professionals begin to take care of children, the elderly, and the vulnerable. Doctors and care‐ givers are examples of those paid for their services; they replace the services offered by family members. One could presumably understand other changes to lifeworld reproduction in this way (for instance, the institutionalization of primary education). 30 Ibid., 364. Kevin W. Gray – The Autopoietic Turn in Habermas‘ Legal Philosophy 48 ANCILLA IURIS (anci.ch) 2014: 41 – Article Understanding juridification in this way, Habermas is forced to make a theoretical choice as to how to understand law. Juridification as lifeworld colonization is both the introduction of legal structure into the lifeworld and the instantiation of a legal system. Conceived in this way, law fulfills two broad roles in modern society: law as positivity and law as institution. Firstly, what Habermas calls law as positivity is the function of law in developing a system of administrative procedure. In modern society, law becomes a means to organize “media‐con‐ trolled subsystems that have [...] become autonomous in relation to the normative contexts of action.”31 Administrative law controls social systems as substitution media take over the reproduction of parts of society. As such, law “takes on the role of a steering media itself.”32 Secondly, law as an institution remains tied to lifeworld contexts. It includes those “legal norms [which] cannot be sufficiently legitimized through a positivistic reference to proce‐ dure.”33 Such laws would include, for example, criminal laws (e.g., the criminalization of murder, etc.). These laws are not justified by referring to a valid process of production inside the legal system; instead, they refer to underlying moral norms.34 Thus, modern law is ultimately Janus‐faced. It paradoxically appears to be both social system and lifeworld structure simultaneously. In many cases, this is advantageous as it per‐ mits law to mediate between system and lifeworld. In his later work, Habermas will refer to this function of law by saying that it serves as a “hinge between system and lifeworld.”35 In other words, although Habermas now postulates, at the end of TCA and in his subsequent work, the existence of a legal system integrated by non‐communicative steering media that is part of both system and lifeworld, the legal system is unlike other social systems. In the first instance, by referring to the hinge function of law, Habermas appears to have modified the model of radical system‐lifeworld uncoupling in the early sections of TCA. In TCA, Habermas understood “uncoupling” in a radical sense. It meant more than just that new mechanisms of societal integration had developed, allowing action to be coordinated without communicative agreement. Uncoupling meant, also, that the lifeworld’s resources were unnecessary in coordinating media‐steered interaction.36 In fact, Habermas seemed to imply that the lifeworld’s resources were not only unneces‐ sary, but perhaps even useless for coordinating media‐steered interaction. (He suggested that the lifeworld merely provided the structure which anchored media‐steered interaction, allowing it to proceed without the encumbrance of communication.) Now, the hinge function of law serves to protect domains of communicative action from media structured domains of action, while allowing the two domains to interact.37 Law has been modified such that it is no 31 Ibid., 365; italics omitted. This would include private and administrative law. 32 Ibid., 365. 33 Ibid., 365. 34 The difference between the two types of law refers “to whether juridification structures domains of com‐ municative action or whether they increase the density of legal relationships that are constitutive of system‐ atically integrated areas of action,” or to the appropriate type of legitimation (procedure or discourse) that justifies the juridification. Jürgen Habermas, The Theory of Communicative Action, vol. 2 (Boston 1984), 366. 35 Jürgen Habermas, Between Facts and Norms (Cambridge 1998), 56. 36 Hugh Baxter, System and Lifeworld in Habermas’ Theory of Communicative Action, Theory and Society 16 (1987), 39. 37 Jürgen Habermas, The Theory of Communicative Action, vol. 2 (Boston 1984), 373. Kevin W. Gray – The Autopoietic Turn in Habermas‘ Legal Philosophy ANCILLA IURIS (anci.ch) 2014: 41 – Article 49 longer purely a lifeworld structure or part of the system. Law, Habermas suggests, is both a social system and a communicative structure which serves as “the translator between the ordinary‐language communication of the lifeworld and the system‐specific codes of the eco‐ nomic and administrative systems.”38 It is not surprising that law must be anchored in the lifeworld if it is to be a social system. However, this concession strains the system‐lifeworld model. First and foremost, Habermas has abandoned the contrast between a normatively rich lifeworld and a realm that is inte‐ grated by non‐communicative replacement media. He must instead posit that there are some media (or at least one medium) that exist in both realms (viz. law). There are other problems, besides the problem of theoretical inconsistently that this poses. First, Habermas subdivided law into law as positivity and law as institution. I do not believe that this cannot be consis‐ tently maintained. Second, it is unlikely that Habermas’ treatment of law in BFN can be made consistent with his treatment of media integrated social systems. I will develop these objec‐ tions in the following sections. III. Law and Systems Theory In BFN, Habermas develops a reworked social theory of the legitimate functioning of political power, while making his first sustained attempt to engage with Luhmann’s mature work on autopoiesis, which occurs through a discussion of the work of both Luhmann and Teubner.39 First, Habermas no longer speaks of the administrative system in BFN. Instead, he now speaks of the political system. Moreover, no longer does Habermas speak exclusively of the system‐lifeworld model of society. Instead, he speaks of democratic societies as com‐ posed of central political systems and their peripheries (the so‐called center‐periphery model). While it might appear in the first instance that this model can simply be mapped directly onto the system‐lifeworld model, I will argue that it cannot. In the first instance, Habermas’ political system is unlike the administrative system in TCA: It comprises government, parlia‐ ment, and other institutions formerly relegated to the lifeworld.40 It is thus larger and more inclusive of communicative action. However, there is a larger problem: Habermas continues to speak of the same systems‐theoretic account that he offered before. I will argue, however, that his theory has changed dramatically between TCA and BFN. This comes through in its clearest form in Habermas’ engagement with the work of Luhmann and Teubner. The debate between Habermas and the systems theorists is important because it high‐ lights two potential questions that Habermas must answer: First, how can a structure that is both system and lifeworld exist and how can such a structure communicate with systems and lifeworld structures? Second, why is his system‐lifeworld model preferable (and is it even different, in its new form) to Luhmann’s autopoietic model of law? 38 Hugh Baxter, System and Lifeworld in Habermas’ Theory of Communicative Action, Theory and Society 16 (1987), 39. 39 This is in contradistinction to the debate that occurred, prior to Luhmann’s introduction of the concept of autopoiesis, in the volume edited by Niklas Luhmann/Jürgen Habermas, Theorie der Gesellschaft oder Sozi‐ altechnologie – Was leistet die Systemforschung? (Frankfurt am Main 1972). 40 Hugh Baxter, Habermas: The Discourse Theory of Law and Democracy (Stanford 2011), 165. In fact, it seems to include institutions that would have been part of the lifeworld in Habermas’ earlier work, as Baxter notes. Kevin W. Gray – The Autopoietic Turn in Habermas‘ Legal Philosophy 50 ANCILLA IURIS (anci.ch) 2014: 41 – Article In Habermas’ original system‐lifeworld model, social systems act super‐subjectively, adopting detached learning processes which operate outside communicative rationality. The growth of these systems parallels the growth of regions wherein functionalist reason pre‐ dominates. In their earliest stages, these social systems emerge when economically relevant action orientations are detached from the lifeworld contexts which would traditionally coor‐ dinate actions and are organized instead by replacement media (e.g., money). Habermas refines the concept of the legal system in BFN. Although he never gives a com‐ plete description of the legal system, Habermas’ comments on Teubner’s theory provide a sense of the nature of the legal system, which is now conceived of as autopoietically struc‐ tured.41 Speaking of legal systems in particular, Teubner defines an autopoietic system as fol‐ lows: “An autopoietic system [is one that] produces and reproduces its own elements by the interaction of its elements.”42 It differs from a functionalist system in that it both creates an autonomous order and identifies its own elements; the self‐definition of the system and its self‐referential reproduction defines the system itself.43 In contradistinction to functionalist systems, autopoietic systems are not anchored by communicative structures in the lifeworld. Moreover, unlike the systems portrayed in TCA which, while acting on the lifeworld, are largely immune to challenges from it, systems in BFN interact with the outside world and challenge the programs of their environments.44 The distinction between system and envi‐ ronment is crucial in Luhmann’s work. An autopoietic social system can only exist by draw‐ ing a boundary between itself and its environment. Systems communicate internally, but are continually influenced from the outside. As Luhmann writes, “communication in the social system is internal, as part of meaning formation in response to events in the environment.”45 To put it another way: “A system can reproduce itself only in an environment. If it were not continually irritated, stimulated, disturbed and faced with changes in the environment, it would after a short time [...] cease its autopoiesis.”46 Communication occurs internally as a response to external events, as systems are closed to information but respond to external events to maintain stability.47 41 In TCA, Habermas treated Luhmann’s work as essentially a refinement of Parsons’ work on media‐driven social systems. Jürgen Habermas, The Theory of Communicative Action, vol. 2 (Boston 1984), 154–155. A common misconception is that Habermas discussed autopoiesis in that work. He did not. In fact, Luhmann had yet to develop his own theory of autopoiesis by that point. 42 Gunther Teubner, Introduction to Autopoietic Law, in: Gunther Teubner (ed.), Autopoietic Law: A New Approach to Law and Society (Berlin/New York 1988), 3. 43 Niklas Luhmann, The Unity of the Legal System, in: Gunther Teubner (ed.), Autopoietic Law: A New Approach to Law and Society (Berlin/New York 1988), 16. 44 It is a curious feature of TCA that, although Habermas is working with Luhmann‘s and Parson’s system the‐ ory and thus largely treats social systems as open to exchange with their environments, there are few, if any examples, of the lifeworld influencing social systems given in TCA. While this absence can perhaps be chalked up to the goals of that project (i.e. the diagnosis of the pathologies of late capitalist society), it is nonetheless curious. 45 Niklas Luhmann, The Unity of the Legal System, in: Gunther Teubner (ed.), Autopoietic Law: A New Approach to Law and Society (Berlin/New York 1988), 18. 46 Niklas Luhmann, Closure and Openness: On Reality in the World of Law, in: Gunther Teubner (ed.), Auto‐ poietic Law: A New Approach to Law and Society (New York 1988), 335. 47 Jean‐Pierre Dupuy, On the Supposed Closure of Normative Systems, in: Gunther Teubner (ed.), Autopoietic Law: A New Approach to Law and Society (New York 1988), 55. Kevin W. Gray – The Autopoietic Turn in Habermas‘ Legal Philosophy ANCILLA IURIS (anci.ch) 2014: 41 – Article 51 This model of social systems differs from Parsons’ notion by conceiving of systems as structured by communication (and not by non‐communicative media). Systems adopt an internal coding which defines what is or is not part of the system. Luhmann and Teubner apply this model to the legal system, which they argue is operationally closed, but responds to external events.48 Law develops, by use of its binary coding, a program of what law is. It decides by responding to events on the outside and by revising that coding – for instance, by reclassifying some events as legal/illegal.49 There are several prima facie reasons to think that this model of law is incompatible with Habermas’ earlier work. In Habermas’ original model, systems do not communicate inter‐ nally, nor are they autopoietic or open to the lifeworld (except to influence it). Nonetheless, Habermas appropriates several ideas from Teubner, including the idea that all social systems adopt a binary code that separates the contents of the system from the environment.50 Ulti‐ mately, the real tension between the two positions is the problem of system interference or, alternatively, of how systems interact in the absence of communication between them. Teubner’s position in this regard is clear: When law and the economy interact, “the legal system is forced by the uproar outside, by the noise of the economic actors, to vary its inter‐ nal ‘order until relative quiet returns’.”51 By definition, communications from the economic system occur in the environment of the legal system, which in turn is forced to react to them. The legal system reacts by changing law, adapting procedure, etc. Habermas thinks that the attempt to treat system interaction in terms of interference is ultimately unsatisfactory. Instead, he contends that Teubner’s argument masks actual on‐ going (underlying) communicative processes. To illustrate his critique, Habermas considers the case of a rental agreement, which is, effectively, simultaneously part of the economic and legal systems.52 It does not appear, he argues, that Teubner can account for this peculiar result with his model of system interference. A rental agreement is not part of the economic system alone (and mere system interference to the legal system, as autopoietic theory would have it); instead, a rental agreement is drafted in specific legal language and forms part of the legal system. This exposes a central ambiguity in the theory, Habermas contends, in that an utterance seems to be simultaneously part of two different systems. 48 As Luhmann writes, “the legal system, is normatively closed but cognitively open”. Niklas Luhmann, The Unity of the Legal System, in: Gunther Teubner (ed.), Autopoietic Law: A New Approach to Law and Soci‐ ety (Berlin/New York 1988), 20. 49 One can imagine what events Luhmann is considering as outside events to which the system is required to respond. For instance, changes in economic organization produce moral outrage in other social systems to which the legal system is forced to respond by employing its internal code and modifying the legal pro‐ gram. 50 Jürgen Habermas, Between Facts and Norms (Cambridge 1998), 143. This is an intellectually peculiar move, as Habermas is borrowing here extensively from Luhmann’s description of autopoietically closed systems that separate themselves from their environment by means of this coding. Hugh Baxter, System and Life‐ world in Habermas Theory of Law, Cardozo Law Review 23 (2002), 264. 51 Cited in: Jean‐Pierre Dupuy, On the Supposed Closure of Normative Systems, in: Gunther Teubner (ed.), Autopoietic Law: A New Approach to Law and Society (Berlin/New York 1988). Although I have been una‐ ble to verify its actual source, the quote is purportedly from Gunther Teubner, Juridification – Concepts, Aspects, Limits, Solutions, in: Gunther Teubner (ed.), Juridification of Social Spheres: A Comparative Anal‐ ysis in the Area of Labor, Corporate, Antitrust and Social Welfare Law (Berlin/New York 1987). 52 Jürgen Habermas, Between Facts and Norms (Cambridge 1998), 52–53; quoting Gunther Teubner, Recht als autopoietisches System (Frankfurt am Main 1989), 109f. Kevin W. Gray – The Autopoietic Turn in Habermas‘ Legal Philosophy 52 ANCILLA IURIS (anci.ch) 2014: 41 – Article If an utterance is simultaneously part of two separate systems, however, it seems at first glance to pose a problem with regard to Luhmann’s view. After all, system communication is supposed to be internal to social systems. This assertion appears to contradict the central tenets of systems theory: Systems theory says that an utterance can be part of one system only and that for all other systems the utterance must be part of their respective environments. The only way that an utterance could be simultaneously meaningful for two different systems is if there were some underlying basis for communication. However, this would be to oversimplify Luhmann’s theory. Luhmann’s claim is not that every utterance occurs in only one system but that its meaning is different to each different social system (and that unlike in communicative action, meaning does not transfer between action contexts unproblematically, if at all). Teubner, to his credit, recognizes this potential misunderstanding of his legal theory. Systems can communicate between each other, he writes, without coordinating their actions through the lifeworld (which is ultimately Haber‐ mas’ position). As Teubner claims, “all forms of specialized communication [...] are always at the same time – literally uno actu – forms of general social communication.”53 Communica‐ tion between systems exists as system interference – a communicative utterance in one sys‐ tem becomes part of the environment to another system. In other words, specialized commu‐ nication has no need to be anchored by a more general type of communication. It is the most general type of communication inside social systems. Nevertheless, after citing this description of system communication, Habermas jumps on the argument, claiming that it shows that system communication is merely a specialized example of a more general type of communication that undergirds all system communica‐ tion. If Teubner means that there is a region of society (outside of systems) where communi‐ cation occurs, then there are two problems with this position. First, Teubner appears to be giving short shrift to the model of system interference that is supposed to explain how sys‐ tems exist in society without requiring a lifeworld. Second, we would then be justified in ask‐ ing what mode of communication Teubner is talking about. With regards to the second, Habermas argues that the only specialized code that could anchor both economics and law (and perhaps the administrative system) would be ordinary language itself (i.e. the species of communication captured by universal pragmatics), which operates in the lifeworld to allow for specialized media to operate outside of communicative action.54 It is at this point in BFN that law reappears to function as a hinge between system and life‐ world. The central idea is that ordinary language often relies on law to regulate other social systems. First, law produces communicative flows between system and lifeworld, translating commands from the lifeworld into specialized codes understandable to the specialized steer‐ ing media of money and power. Second, it is internally defined by the binary coding of legal‐ illegal which defines all autopoietic legal systems. First, according to Habermas, for coordination to occur between the newly energized political section of the lifeworld and social systems, society: 53 Gunther Teubner, Recht als autopoietisches System (Frankfurt am Main 1989), 107; here Teubner is using the word communicative in a very different sense than Habermas would ever use it. 54 Jürgen Habermas, Between Facts and Norms (Cambridge 1998), 55. I need take no position in this paper as to whether or not Habermas correctly understands Teubner’s position; all that matters is that his response to Teubner shows that Habermas has adopted the language of autopoietic systems theory. Kevin W. Gray – The Autopoietic Turn in Habermas‘ Legal Philosophy ANCILLA IURIS (anci.ch) 2014: 41 – Article 53 [R]emains dependent on the law that communicates with the steering media of money and admi‐ nistrative power. Normatively substantive messages can circulate throughout society only in the language of law. Without their translation into the complex legal code that is equally open to life‐ world and system these messages would fall on deaf ears in media‐steered spheres of action.55 Thus, while Habermas seems to have suggested earlier that “the most we could hope for was a democratic dam against the colonizing encroachment of system imperatives on areas of the lifeworld, in BFN the function of law goes beyond that.”56 Now law has a much more sig‐ nificant role: The legal code remains located partially in the lifeworld through the medium of everyday language. “[I]t also accepts messages that originate there and puts these into a form that is comprehensible to the special codes of the power‐steered administration and the money‐steered economy.”57 By transforming moral communication from language into a specific code, it allows moral discourse to influence systems through the specific language of law. Law becomes the only medium that circulates throughout the legal system, while being programmed by politics and political legislation. This view of law and the political system is more optimistic than the system theoretic account of the administrative system and lifeworld colonization given in TCA. While the administrative system may have colonized the lifeworld in TCA, we now have a political sys‐ tem capable of pushing back. However, it comes at the cost of increased complexity. Now, weak publics in the lifeworld (e.g., civil society) influence the strong publics of the political system (e.g., parliaments) which allow the translation of moral communication in the life‐ world into the specific language of law and are institutionalized by law.58 Here, the lifeworld is supplemented with a model of the public sphere. The legal system is programmed through interaction via public sphere, civil society, and the lifeworld. It produces administrative acts, binding (legal) commands, etc., which serve as functionally binding commands for other social systems. It is functionally integrated not by systems media, but by a specific linguistic coding which is a subspecies of language more generally. Second, law grounds systems in the lifeworld by providing legal institutions that enable abstract exchange media. “The steering media of money and administrative power are anchored in the lifeworld through the legal institutionalization of markets and bureaucratic organizations.”59 In comparison to Parsons’ model of money or power, law enjoys a special status amongst social systems, defining the coding that creates steering media. By introduc‐ ing the idea of coding, Habermas is borrowing from (while altering) Luhmann’s conception 55 Ibid., 56. 56 Jeffrey Flynn, Communicative Power in Habermas’ Theory of Democracy, European Journal of Political The‐ ory 3 (2004), 433–454. Habermas introduces the metaphor of the “democratic dam” in his article “Further Reflections on the Public Sphere,” contained in Craig Calhoun (ed.), Habermas and the Public Sphere (Cam‐ bridge 1992), 421–461. 57 Jürgen Habermas, Between Facts and Norms (Cambridge 1998), 81. 58 Here, Habermas is borrowing the discussion of weak and strong publics introduced by Nancy Fraser in her article “Rethinking the Public Sphere: A Contribution to the Critique of Actually Existing Democracy,” also contained in Craig Calhoun (ed.), Habermas and the Public Sphere (Cambridge 1992), 421–461. Weak pub‐ lics are those which are engaged in opinion formation, whereas strong publics (such as parliaments) are also engaged in decision‐making. 59 Jürgen Habermas, Between Facts and Norms (Cambridge 1998), 75. Kevin W. Gray – The Autopoietic Turn in Habermas‘ Legal Philosophy 54 ANCILLA IURIS (anci.ch) 2014: 41 – Article of autopoiesis and is appropriating the idea that all systems adopt a binary code that sepa‐ rates the contents of the system from the environment. In the case of law, the appropriate code would be legal‐illegal, and is determined by courts, bureaucrats, etc. inside the legal sys‐ tem.60 However, Habermas does not limit this claim to the legal system but says that he is mak‐ ing a general claim about social systems. That said, while he seems to think all systems pos‐ sess this method of binary coding, Habermas does not tell us what the appropriate coding would be for the economic or administrative systems (or even that he actually means to mod‐ ify his concept of system integration given in TCA, though this is what he has done). One can imagine that the appropriate coding might be something like ‘paying‐for’ versus ‘being‐paid‐ for’ in the economic system, and ‘exercising‐power’ versus ‘having‐power‐exercised‐on‐one‐ self’ in the administrative system.61 Whatever the appropriate coding, in the final two sec‐ tions, I will argue that this new model of social systems and of law in particular is incompat‐ ible with Habermas’ neo‐Parsonianism. IV. Problems in Between Facts and Norms In Section 3, I identified two potential problems in Habermas’ theory. First, I doubted if Habermas’ distinction between the two different types of law could be upheld, and, second, I doubted that there were any reasonable candidates for the appropriate steering medium nec‐ essary in neo‐Parsonian systems so that law could be treated as a social system. I will return to these problems and show how Habermas’ response to Teubner, rather than clarifying the issue, causes additional problems. In BFN, Habermas gives one more description of how we might distinguish between law as social system and law as lifeworld structure. He writes, I take the legal system [...] to include all legally regulated action systems; reflexive law constitutes a core area for the production of legal acts by private persons, whereas material legal norms imme‐ diately direct behavior in various ways. [...] A stratification exists between formally organized sphe‐ res of interaction that are legally constituted and those spheres that, primarily regulated through extralegal institutions, merely have an overlay of law. In formally organized spheres like the econo‐ my or the state apparatus, most interactions are guided by law [...] whereas in spheres like the fa‐ mily, neighborhood, or educational system it is only in cases of conflict that the law emerges from the background and enters the awareness of the actors.62 Put simply, the difference between spheres of action that are legally integrated and those that are not, is that in formally organized spheres of action we are conscious of the role of law, whereas in lifeworld contexts, law exists in the background.63 60 Ibid., 143. This is an intellectually peculiar move, as Habermas is borrowing extensively from Luhmann’s description of autopoietically closed systems that separate themselves off from their environment by means of this coding. See Hugh Baxter, System and Lifeworld in Habermas’ Theory of Law, Cardozo Law Review 23 (2002), 473. 61 The first example (i.e. economic coding) is Baxter’s. Ibid., 473. 62 Jürgen Habermas, Between Facts and Norms (Cambridge 1998), 196. 63 While it is not clear from the context of the quote if this is a hard and fast division, I believe that it would be uncharitable to interrupt Habermas’ stratification analogy other than by saying that the division between different types of law is one of gradation. Kevin W. Gray – The Autopoietic Turn in Habermas‘ Legal Philosophy ANCILLA IURIS (anci.ch) 2014: 41 – Article 55 That said, this distinction fails for two reasons. First, integration media depend on moral norms in ways that make it difficult to separate the two. Let us return to the discussion of rental agreements. On Habermas’ telling, what makes a rental agreement part of the legal system (and also part of the economic system, though that is irrelevant here) is that it is gov‐ erned primarily by law. However, it is unclear that a contract is actually governed more by law than by moral norms. Moral considerations come into play in contract law generally in such a way that it is difficult to separate the moral components of contracts from the legal components (namely that we have a moral obligation to keep promises). Marriage contracts, for instance, embody both legal and moral norms.64 The confusion between moral and legal norms holds true in many other legal contexts, where our actions seem to be governed both by moral and by legal norms. For example, rules governing conflicts of interest in banking refer to both types of norms. It is not the case that bankers are normatively free to disobey banking regulations if they are prepared to pay a penalty. In the case of banking, while it is true that legal details of these contexts of interac‐ tions are governed by laws and regulations (which specify what constitutes a conflict of interest), it seems that there is no good way to separate those actions that Habermas wants to call part of the lifeworld from those that are part of the legal system. Even for a reader who is sympathetic to Habermas’ attempt to draw a boundary between morally and legally regulated spheres of interaction, another more important problem threatens the entire attempt to reconcile law with the system‐lifeworld model. In TCA, Habermas reinterpreted Parsons’ discussion of substitution media inside social systems. Par‐ sons, in order to show that all social systems were integrated by substitution media, began with the specific case of money, which he then generalized into a general theory of function‐ alist media. Money remains, in Habermas’ work, the paradigmatic case of a substitution medium. As we saw, Parsons identified six important properties that money possesses that all other media must possess.65 In TCA, Habermas explicitly rejected much of Parsons’ attempt to generalize the theory of social systems, arguing that the only other potential mode of action integration in the modern world is power. Habermas argues against Parsons’ attempt to treat influence and value com‐ mitment as replacement media, arguing that there are “no institutions that, in analogy to property and offices, would permit a well‐circumscribed normative anchoring of influence or value commitment.”66 Similarly, influence and value‐commitment fail several of the six conditions proposed by Parsons. 64 It appears that Habermas does not recognize how radical a position this actually is, nor does he engage with Joseph Raz, whose work might buttress the argument. Even positivists such as H.L.A. Hart, who one might superficially expect to be sympathetic to the position that rules are not based on morality, turn out not to believe this in many cases. Herbert L. A. Hart, The Concept of Law (Oxford 1994). In fact, Hart’s response to Dworkin’s critique commits him to the inclusivist position. For this argument to work, Habermas would have to believe that adherence to contracts is optional – which he does not appear to do – in order to separate apart moral obligations from legal obligations (e.g., the obli‐ gation to keep one’s contract would have to be reducible to the claim that if you do not keep your contract you will have to pay a penalty). This seems unlikely, as law carries moral weight both for many legal posi‐ tivists and in BFN. 65 Jürgen Habermas, The Theory of Communicative Action, vol. 2 (Boston 1984), 264–265. See supra Section 1. 66 Ibid., 275. Kevin W. Gray – The Autopoietic Turn in Habermas‘ Legal Philosophy 56 ANCILLA IURIS (anci.ch) 2014: 41 – Article I will proceed in the same vein as Habermas did in his rejection of Parsons’ attempt to generalize the theory of systems media, but by asking if there is a structurally similar medium to money or power that integrates the legal system. It appears that the only available candidate for such a social medium is law as positivity, which is, after all, part of the legal system. Law as positivity seems, on first reading, to be a good candidate as it encodes a yes‐ no binary that defines all legal interactions. Something is either legal or illegal. Thus, law seems to satisfy Condition 3 (namely that the alter in an exchange can offer one of only two responses). Similarly, in the legal transaction actors do not coordinate their actions communi‐ catively but strategically (satisfying Condition 5, i.e. that actors in legal systems are oriented strategically to the success of the transaction). They advance positions based on the desire to ‘win’ and not necessarily based on moral norms they hold to be true (satisfying Condition 2). At any moment, they can agree to a decision and/or choose to compromise (satisfying Condi‐ tion 4). It is less clear, however, whether law satisfies the other conditions (discussed in Section 1 above). As to Condition 1, law enjoys wider validity in modern society than either money or power. In any situation where we can evoke money or power as an appropriate medium of system integration, we could also say that law plays a role (in fact, to satisfy his conception of the hinge function of law, Habermas would seem to have to maintain this). At any moment in a monetary transaction, we could challenge the transaction and say that what actually mat‐ ters is the legality and not the economics of the situation (e.g., that the proposed transaction is illegal, etc.). Similarly, power (in Habermas’ sense of the word) is only recognized when there is a legal framework in place. Since the modern world is so thoroughly juridified, it is not clear whether there remain many areas that are not, in some way, governed by law.67 If that is the case, then it seems that law fails to be appropriately limited so as to be a good candidate to serve as a system‐integrating replacement medium. As to Condition 6, money is obviously bankable in the sense that it can be put aside for another occasion. However, is law enough like money and power to enable us to say that it can function as a replacement medium in modern society? Law can be stored in the sense that legal rights and legal obligations endure seemingly forever. However, unlike money and power, it is hard to see what it would mean to say that law can be measured and quantified. To say that law could be possessed in an amount that can be measured would be to go against the basic legal‐illegal binary that defines law as a structure. Thus, law fails Condition 6. Ultimately then, we must conclude that Habermas’ attempt to describe law as a social sys‐ tem does not succeed. Beyond the prima facie tension between the neo‐Parsonianism and the autopoietic models of social systems, law fails the tests proposed by both Parsons and Haber‐ mas for what constitutes a social system. 67 This is in many ways an alternative formulation of Habermas’ thesis of the colonization of the lifeworld; it is developed to a great extent in the second volume of The Theory of Communicative Action. Kevin W. Gray – The Autopoietic Turn in Habermas‘ Legal Philosophy ANCILLA IURIS (anci.ch) 2014: 41 – Article 57 V. Concluding Remarks It seems that Habermas has made substantial modifications to the view of social systems that he adopted in TCA. The acceptance of the existence of the legal system would require substantial changes to Habermas’ overall theory. First, Habermas would have to find a better way to distinguish between the two types of law. Second, it would require a new under‐ standing of the properties of replacement media that would, out of necessity, differ radically from the Parsonian view. It has been suggested that the incompatibility between the two versions of Habermas’ theory of law derives from the different goals of the two major works discussed here: the diagnosis of social pathologies in TCA versus the development of a theory of democracy in BFN.68 That may be true, of course, but it is secondary to the discussion at hand. If the former theory is incompatible with the latter, then a choice has to be made. One assumes that it is a theory of democracy and the public sphere which interests Habermas more at this stage in his career, and therefore that the architecture of TCA must be abandoned. What is clear, how‐ ever, is that the two models of systems theory are not reconcilable as Habermas presents them. Parsons must be abandoned to save Habermas’ theory of law and democracy. 68 Amelie Wirts made this point in her comments on an earlier draft of this paper presented at the meeting of the APA Eastern Division in December 2011.