Choice,blind spots and free will

Introduction

Commentators have often focused on two central aspects of Berlin’s thought: his conception(s) of freedom and his value pluralism. This article, however, focuses on choice. Choice in Berlin’s thought is divided into two notions: agonizing choices and intolerable choices. The exploring of the role that choice plays within Berlin’s social and political thought means that a conception of free will becomes apparent; an area that is often sidelined by theorists who have no time for the murky waters of metaphysics. Yet Berlin had his own metaphysics, based on a belief in free will, which he takes for granted in his analyses. This presumption of free will, as a metaphysics of choice, helps bridge the gap between various aspects of his thought, in particular it explains the link between value pluralism and liberalism – between the necessity of choice and the valuing of freedom of choice. In the second part, this article outlines the theory of autopoiesis, used in biology and sociology, and shows that it applies to conscious systems; most particularly to sets of values. Using Luhmann’s criteria, this article shows that value pluralism is a conscious autopoietic system, and that autopoietic theory is a better basis for value pluralism than liberalism is. Finally, this article outlines the consequences of autopoietic theory for political and social theory.

I: Berlin’s conception of choice

Berlin is famous for his views on freedom and value pluralism, but the role of choice in his Weltanschauung is less well known, despite its being clear to many observers that choice ‘is a key concept in … Berlin’s intellectual outlook’, ¹ and an analysis of this concept in his thought is long overdue. For Berlin we always choose, being able to follow one’s choices is freedom, and being coerced into not following one’s choices is, as it were, not freedom at all. This forms the backbone of his critique of positive liberty, as ‘[i]t is one thing to say that I know what is good for X, while he does not; and even to ignore his wishes for its – and his – sake; and a very different one to say that he has eo ipso chosen it’. ² There is no freedom, in other words, if one has not chosen one’s own path, and Berlin’s defence of negative liberty over positive liberty is centred on the importance of choice.

Choice does not equal freedom, however, and the relationship between the two concepts is not necessarily linear. Berlin had also claimed that ‘it is the actual doors that are open that determine the extent of someone’s freedom, and not his own preferences’. ³ Berlin here draws a difference between freedom, defined as a ‘range of objectively open possibilities’, ⁴ being able to do what one chooses to do. If you are lucky enough to do what you have chosen to do, it does not mean others enjoy the same freedom. This is where the importance of value pluralism comes into play in Berlin’s thought. The existence of plural and irreconcilable sets of values means that it is not enough to be free from doing one particular thing or from living one particular way of life; but a choice between genuine alternatives is necessary for all to be able to live by their own set of values.

In order to show the centrality of choice in Berlin’s thought, it is important to make explicit a distinction that he makes only implicitly when he speaks of choices. Some choices may be relatively straightforward and pose little problem for the individual, but there are choices that are, by their very nature, difficult to make. Berlin expresses his belief that ‘[t]here is no avoiding choices between ultimate human values, ends in themselves. Choices can be agonizing, but unavoidable in any world we can conceive of.’ ⁵ Berlin highlights an example of this in his essay on Machiavelli, where he shows that for the Florentine thinker, the virtues of the Christian and the political virtù of Romans are incompatible sets of ethics. ‘One can save one’s soul, or one can found or maintain or serve a great and glorious state; but not always both at once.’ ⁶ This kind of choice, between Christian and political life, is an example of agonizing choice. Choice need not be agonizing all the time, but given Berlin’s value pluralism, it is inevitable that choice will end up being agonizing some of the time.

Agonizing choice is all the more inevitable given that in Berlin’s thought, there is always choice. ‘Action is choice; choice is free commitment to this or that way of behaving, living, and so on; the possibilities are never fewer than two: to do or not to do; to be or not to be.’ ⁷ If action is choice, then behind every form of action there is some form of intention. This intention might be limited to agonizing choices (or even to intolerable ones), but there is always a choice made somewhere. Choice, here conceived in a paradoxical way as ‘free commitment’ to a way of life, is fundamental to any action – whether historical or moral. Berlin’s political theory is concerned with the most important choices we make, the ones that we agonize about, choices between irreconcilable and incompatible alternatives. A society cannot completely alleviate the tragedy of choice, yet it can prevent intolerable choices, by having few obstacles in the way of an individual’s most important and defining choices. It ‘is the first requirement for a decent society’, Berlin claims, to ‘prevent the occurrence of desperate situations, of intolerable choices’. ⁸ The avoidability of these intolerable choices is essential in Berlin’s preference for negative liberty. An individual is better off being left alone to choose his or her own ultimate values rather than be forced to accept that the values imposed upon him or her from above are actually his or hers – or those of the ‘higher self’. As Hatier has noted, the lack of clear engagement with ‘totalitarianism’ as a concept in Berlin’s thought often obscures the fact that Berlin put forward a critique of the totalitarian state alongside his critique of monism. ⁹ Similarly, choice is central to his critique of totalitarianism. Since there is always a choice, it cannot be taken away even in an unfree society, yet in those circumstances it becomes a choice between intolerable alternatives. The choice between dying while resisting arrest and dying in a concentration camp in Nazi Germany or Stalinist Russia is not an agonizing choice, it is an intolerable one. In Berlin’s words: ‘[t]o be free is to be able to make an unforced choice; and choice entails competing possibilities – at the very least two “open”, unimpeded alternatives.’ ¹⁰ Where there is the possibility of action, choice exists. The best a society can do is to limit the extent to which choices become intolerable for the individual.

It is one of our defining characteristics, as humans, to be able to choose. In this, choice is intrinsically linked to freedom, as Berlin agrees with the common use of language which states that ‘freedom is the principal characteristic that distinguishes man from all that is non-human’ and that ‘degrees of freedom [are] constituted by the absence of obstacles to the exercise of choice’. ¹¹ One is truly human when one is free, and one is free only insofar as one can exercise choice. In fact, a ‘lack of freedom of choice means dehumanization’. ¹² Value pluralism, coupled with this essential human capacity for choice, means, for Berlin, that it is better to defend negative liberty (freedom from interference by others) as it is the best way to avoid forcing individuals to make a choice between intolerable alternatives. Furthermore, Berlin believes that we do not merely choose from pre-determined causes, but that we have a capacity for choice that is – at least partially – uncaused by physical, physiological, or psychological causes. This implies that we ultimately choose ourselves, or in other words that we possess free will.

Berlin refuses to admit this in his introduction to the Four Essays on Liberty, where he claims that he has ‘made no systematic attempt to discuss the problem of free will as such, but principally its relevance to the idea of causality in history’. ¹³ In particular, it is in the relationship between determinism and freedom that Berlin is interested. Despite claiming that he concludes nothing towards the proclaimed truth of determinism, he nonetheless says ‘that the arguments in favour of [determinism] are not conclusive’. ¹⁴ Berlin’s ‘agnosticism’ towards the truth or untruth of determinism is unconvincing, as he formulates a clear challenge to determinist arguments. He further develops his attack by arguing that the belief that ‘responsibility and determinism are never incompatible with one another [is] mistaken’. ¹⁵ One of Berlin’s key concerns during his career was to interpret history, and understanding the role of the agent in the historical process is essential. The possibilities of moral responsibility and of determinism of individuals in the historical process ‘are mutually exclusive: both beliefs may be groundless, but both cannot be true’. ¹⁶ In the same manner that Berlin had expressed doubt about the commensurability of values, he extends his scepticism towards the commensurability of determinism and of individual moral responsibility. Reluctant to place himself firmly in the debate on free will, Berlin admits being, at the very least, an incompatibilist.

Berlin’s incompatibilism pushes him further in his critique of determinism. For if determinists are correct, this goes against what we mean in everyday language when we talk of moral praise and blame and of moral responsibility. This argument may not provide sufficient evidence against determinism, as Berlin is fully aware, but it nonetheless places the onus of proof on determinists. Given the drastic legal, political and moral consequences of the – possible – truth of determinism, it is essential that we are convinced of it beyond doubt before we entirely rethink the way humankind has interpreted and used the concept of responsibility for millennia. ‘[W]e should be unwise’, Berlin argues, ‘to underestimate the effect of robbing praise, blame, a good many counterfactual propositions, and the entire network of concepts concerned with freedom, choice, responsibility, of much of their present function and meaning.’ ¹⁷ Against determinists, Berlin sides with Kant who had already critiqued the ‘miserable subterfuge’ ¹⁸ of self-determinism. Berlin’s self-proclaimed incompatibilism puts him at odds with many within the liberal tradition, who have often held the view that determinism and free will can be reconciled, as well as with many outside the liberal tradition; it is the root of his ‘disagreements with Professor Ernest Nagel, Professor Morton White, Mr E. H. Carr, the classical determinists, and their modern disciples’. ¹⁹ In history, Berlin does not believe ‘that you can see a pattern enabling you to predict’, ²⁰ and against the influential movement of historical determinism, Berlin maintains that ‘there are moments in history when individuals or groups can freely alter the direction of things. Not everything is predictable … there is space for choice.’ ²¹

Berlin’s challenge to determinism is here directly linked to the importance that he places on choice as a characteristic feature of our human nature. For if determinism is true, free choice cannot exist. In a footnote to his essay on ‘Historical Inevitability’, Berlin says that ‘if my choice is itself the result of antecedent causes, I am, in the relevant sense, not free’. ²² The important emphasis is, of course, on the relevant sense of freedom. Berlin’s argument is that we are not free, in the sense in which we understand freedom, if in addition to what we call ‘fatal obstacles to action’ – i.e. physical and biological – there are other obstacles such as our characters and habits which wholly determine us. For us to be free there must remain ‘some area, however narrow, within which choice is not completely determined’. ²³ Choice is not in itself completely incompatible with determinism. But if we mean that ‘choices are regarded as being themselves no less determined than other phenomena’, as a determinist would argue, we are not talking about the ‘basic sense of the term about which controversy has been boiling for twenty-two centuries’. ²⁴ Determinism challenges choice, and freedom of choice, as we understand them in common language.

Berlin’s agnosticism in the face of the question of determinism and free will is not consistent. He maintains the possibility of determinism’s being proved correct, yet he assumes that it is not true in his historical approach, as well as in his wider political analysis. When Berlin defends a conception of freedom grounded in choice, as I have shown above, he does not mean choice in the deterministic, deflated definition of the word. There is some area within which causality does not fully apply. We cannot act completely independently of causality, as Berlin believes one must ‘remember there are limits … we are confined by the nature of things. We haven’t much choice. Let us say one per cent. But that one per cent can make all the difference.’ ²⁵

Berlin’s belief in free will clearly derives from his interpretation of Kant. ²⁶ Reacting against the views of French Enlightenment philosophes such as Helvétius and Holbach, Kant – in Berlin’s words – argues that ‘[u]nless a creature can determine itself, it is not a moral being: whether it is causally determined, or floats about at random, or is subject to statistical laws, it is not a moral agent’. ²⁷ In Berlin’s interpretation of Kant, determinism is incompatible with moral agency and responsibility. Furthermore, ‘[n]o act can be described in moral terms, indeed it can scarcely be described as an act at all, unless it is freely chosen by me’. ²⁸ This, Berlin says, highlights the centrality in Kant’s thought of the chooser – or the choosing self – in the individual. The corollary for Kant, according to Berlin, is that he ‘attributes to the moral will a transcendent status outside space and time’ and, most importantly for this article, ‘outside the lower realm of blind, causal necessity’. ²⁹

The moral agent is placed outside of necessity, outside the rule of determinism that applies to the ‘lower realm’, which characterizes both Kant’s and Berlin’s conceptions of the individual. Berlin himself says that ‘Kant’s argument that … where there is no independence of causes there is no responsibility … carries conviction’; ³⁰ and again that he believes in ‘Kant’s idea of human personality and human freedom’. ³¹ Berlin accepts the rationale of the Königsberg philosopher on matters of free will. There is a fundamental incompatibility between determinism and freedom of choice, and there are good grounds to believe that there is freedom of choice – and in any case the onus of refutation rests with determinists. His thoughts on determinism and free will resemble his religious attitudes in general. As Michael Ignatieff points out, Berlin’s scepticism about the existence of God ‘lived happily with religious observance’. ³² Equally, one might say that his doubts about the truth of determinism never fundamentally challenged his belief in the necessity of free will.

It is a metaphysics of choice, in a conception of human nature as being at least minimally free from some area of causality, that Berlin admits to believing in. Glen Newey ³³ has critiqued value pluralists for enshrining their thought in metaphysical concepts. It is not clear, however, what he means by ‘metaphysical’ as it seems to be taken as a synonym for the meta-ethical claims of pluralists; i.e. as a synonym for claims about the truth of how values interact with each other. Let me clarify what I mean here by metaphysical. By a metaphysical choice, I understand a choice that lies beyond the realm of the natural sciences [ta meta ta physika], in the sense that natural causes do not fully apply to human action. As in Plato’s myth of Er, where the souls get to choose their lives before they drink from the river of forgetfulness and start a new life, ³⁴ Berlin’s conception of human nature implies a fundamental, existential and metaphysical choice between life-defining alternatives. This definition of metaphysical choice is much narrower than that used by Newey, and should allow me to make a clear link between this metaphysics of choice and a belief in free will.

Berlin, it is well known, is a fervent critic of metaphysics. For him, the ‘thinkers of a more metaphysical type – Plato, Berkeley, Hegel, Marx’ were those whose ‘essential purpose is to expound an all-embracing conception of the world and man’s place and experience within it … in terms of a new pattern in which what had earlier seemed to be a casual amalgam of elements is presented as a systematic, interrelated unity’. ³⁵ Berlin’s criticism against rationalists of the 17^th century is that they ‘thought that the answers could be found by a species of metaphysical insight, a special application of the light of reason with which all men were endowed’. ³⁶ Metaphysics, in Berlin’s thought, is equivalent to the belief in an all-embracing conception of the world, a unity à la Plato, and/or a faith in the power of reason. Berlin’s critique is weak here. As Gray rightly points out, ‘Berlin’s thought does not satisfy, or seek to appease, this human need for metaphysical consolation’, ³⁷ and in fact attempts to avoid all metaphysical preoccupations thanks to a certain tragedy in human affairs. Berlin here, by identifying metaphysics as a search for ultimate unity, by equating it with the Ionian fallacy, ³⁸ is guilty of a logical leap between the two concepts. The view that metaphysics always involves a search for divine unity is surely a very narrow conception of metaphysics. This view eo ipso rules out Berlin’s thought from being metaphysical, but it is too exclusive to convince.

Berlin’s thought on Maistre illustrates an area where common sense defines metaphysics in a different way than he does. Berlin explains that in Maistre’s works,

… [i]n place of science he preached the primacy of instinct, Christian wisdom, prejudice (which is but the fruit of the experience of generations), blind faith; in place of optimism, pessimism; in place of eternal harmony and eternal peace, the necessity – the divine necessity – of conflict and suffering, sin and retribution, bloodshed and war. ³⁹

Berlin does not qualify Maistre’s thought as metaphysical because he does not believe in a divine or rational unity. Berlin’s use of the word metaphysics seems unnecessarily narrow, for it excludes any metaphysics that would be based on a plurality of goods and values. His exclusion of his own metaphysics from the metaphysical realm, in other words, cannot be sustained. His value pluralism is coupled with a belief that the individual necessarily chooses between irreconcilable values; and that this choice is ultimately free from causality – all of this sounds very metaphysical, even to the most sympathetic of readers.

Critics have often highlighted a missing link between value pluralism and liberalism in Berlin’s thought. ⁴⁰ In essence, Berlin fails to disclose precisely how one can move from the necessity of choice entailed under value pluralism, to the valuing of choice – particularly freedom of choice – in liberalism. Others have tried to bridge this gap through differing accounts of Berlin’s thought, but they remain unconvincing in many respects.

Jonathan Riley ⁴¹ paints a picture of Berlin as a liberal rationalist, despite Berlin’s insistence on the limits of reason. Riley’s typology of Berlin’s thought rests, first, on the priority of liberal values, and only then on the categories of value pluralism and radical choice. ⁴² However, it seems clear that value pluralism, as a conceptual ‘truth’ that Berlin wants to put forward, is of foremost and primordial importance in his thought. ⁴³ Value pluralism hence comes first, in Berlin’s justification, even before his defence of liberalism. This seems clear enough, considering that Berlin does not speak of the ‘truth’ of liberalism, but merely as a preference of our modern societies.

George Crowder ⁴⁴ in particular attacks ‘choice’ – or more precisely the value of freedom of choice – as this link between pluralism and liberalism. For Crowder, Berlin’s claims in this respect are ‘indirect, ambitious and unconvincing’. ⁴⁵ Crowder then argues that choices may be defended because they are symbolic – i.e. as representing ‘something else to which we attach value’. ⁴⁶ Crowder’s criticism of Berlin is potent, but I wish to show that there is another way – more consistent with what I have said above about Berlin’s belief in free will – to explain the importance of choice in the relationship between pluralism and liberalism.

In addition to the truth of value pluralism, Berlin takes free will to be constitutive of human beings. Galston’s claim that value pluralists recognize that ‘preference for individual choice [is] characteristic of advanced industrial societies’ ⁴⁷ does not fully apply to Berlin. Since we have free will, it is not a matter of personal or societal preference, but it constitutes us as human beings; Berlin is making an ontological claim here. Human beings choose themselves fundamentally, and make choices that are in line with their choice of themselves throughout their lives; they live by the moral standards that they have created for themselves. If the truth of value pluralism is primary in Berlin, the secondary claim is that we have free will as human beings and that as such we cannot escape making choices that define who we are, that we create ourselves as individuals.

Further drawing on the distinction that Berlin makes between agonizing and intolerable choices, one can link value pluralism with liberalism in his thought. Value pluralism, first and foremost, shows us that we cannot avoid choices and that at times these choices will be agonizing, that is, in the words of Galston, choices between good and good. ⁴⁸ Some regimes restrict choices for some of their citizens so much so that these are no longer faced with agonizing choices (between good and good) but with intolerable choices – between bad and bad. Because human beings have free will, and have chosen their sets of values that define them, depriving them of their choices so that they can no longer choose their ‘good’ makes these choices intolerable, and eventually ends up treating them inhumanely. Crowder is correct in pointing out that humaneness provides a link between pluralism and liberalism, ⁴⁹ but he is not quite in tune with the spirit of Berlin’s thought when he interprets humaneness as a value. For Berlin, humaneness is not a value per se, but rather an ideal which is in accordance with human nature, ⁵⁰ and it cannot be a contingent value. Any political system denying us the freedom to act on our free will goes against human nature as such, not merely against this or that community’s value. Berlin’s critique of regimes that make their citizens face intolerable choices is grounded in his belief in free will. Denying us the option of acting on the good that we have chosen is inhumane in the sense that it is acting against human nature, depriving us of something that we have qua humans. Liberalism, Berlin claims, with its strong emphasis on the protection of individual choice through negative liberty, acts in a way that is closer to human nature, and is a better system than those who put their subjects in front of intolerable choices. This forms part of Berlin’s ‘blind spot’, as his defence of negative liberty is a reaction to ‘totalitarian’ systems.

II: Autopoiesis

The theory of autopoiesis, as it was developed in biology by Maturana and Varela and then in sociology by Luhmann, provides a critical tool to analyse Berlin’s claim about the superiority of liberalism, with its emphasis on negative liberty and the avoidance of intolerable choices, over other systems. When applied to conscious systems, as this article does, autopoiesis explains the self-referential nature of sets of values – including liberalism – and allows one to take value pluralism seriously without positing the necessity of liberalism – as Berlin does.

Autopoiesis, derived from the Greek that means ‘self-creation’, was not originally a theory about human agency, but a biological theory. Humberto Maturana and Francisco Varela, who first established the theory, ⁵¹ describe autopoiesis as a criterion for all living beings, allowing for differing structures. Two main features characterize all autopoietic beings: their operational closure and their capacity to self-reproduce.

The concept of operational closure states that everything that goes on around the autopoietic entity is understood by it according to its own structure and internal logic. Maturana and Varela appeal to the experiment of the ‘blind spot’ in our field of vision to describe this operational closure. What the blind spot (this insensivity to light in our eye where the optic nerve leaves the retina) illustrates is that we do not see that we do not see. We do not walk around with a black spot in our field of vision; because we actually do not see the outside world, but we live our field of vision. All living beings, from the simplest unicellular entities to more complex living beings, are organized in this manner. All autopoietic systems do not ‘see’ the outside world but interpret the perturbations of the environment around them in terms of their own internal logic. Maturana and Varela had already drawn some of the social consequences of their biological theory. Interaction and communication which already take place at the level of simple autopoietic entities continue between complex autopoietic entities in the form of social interaction. When interaction between complex autopoietic entities recurs, there is the creation of a history of interaction; a co-ontogeny. Social interaction is observable in a variety of animal species, while human interaction takes this process one step further by communicating about the communicative process itself; what Maturana and Varela call the linguistic domain.

Niklas Luhmann expands on this theory of autopoiesis by analysing interaction in what he calls social systems. ⁵² Moving from biological autopoietic entities to social systems, Luhmann highlights both the differences and the similarities between autopoietic theories. At the very least, they all share the two basic features discussed above: self-reference and self-creation. ⁵³ Luhmann, in his sociological theories, does not hesitate to use biological analogies to describe social systems, saying that the social system of politics operates ‘just as in the brain’ ⁵⁴ but also distinguishes between living, conscious and social systems, as the three types of autopoietic systems, and although he says little about conscious systems, he opens the door for further study.

All autopoietic systems operate on the basis of their operational closure. This closure of the system, however, is never absolute and refers only to the system’s organizational structure. Systems still operate within the world – autopoiesis is an immanent theory rather than a transcendental one ⁵⁵ – but, as with the blind spot experiment, they are unable to have direct access to the outside world and instead interpret it within their operational closure. Self-reference is an activity of a system that selects information – in the form of perturbations – from the environment, constructing its own order to reduce complexity. ⁵⁶ An autopoietic system also engages in the self-reproduction of its elements. Living systems physically produce their own elements from material received from the outside world. Similarly, social systems are comprised of and reproduce themselves through communications. Luhmann’s social systems are autopoietic because the communications that constitute each system have a capacity to self-reproduce the unity of the system as a whole. Finally, for Luhmann, autopoietic systems are also paradoxical. He is adamant that paradoxes are unavoidable, since ‘a paradox is the logical equivalent of self-reference’. ⁵⁷ The creation of a closed identity within the world is paradoxical, as operational closure is never true closure but requires openness to the outside world. Similarly, Maturana had paradoxically accepted the existence of the objective world and put objectivity (i.e. objective knowledge of this world) in parenthesis. ⁵⁸

Maturana and Varela had detailed the functioning of autopoietic living systems; Luhmann has detailed social systems. All mention, at some point or another, the importance of conscious autopoietic systems in their works – but none of them develops this point. Yet sets of values are ideal candidates for an autopoietic theory of conscious systems. Sets of values share the characteristic features of all autopoietic systems, their operational closure and their capacity to self-reproduce. Sets of values are operationally closed in the sense that they interact with the outside world only through a set of rules that is internal to the system.

Liberalism, the set of values being put forward by Berlin, considers outside events only as perturbations of its inner workings. Berlin’s thought was largely influenced by the advent of totalitarian states in Europe in the 20th century. The image of the totalitarian state certainly influenced Berlin’s key distinction, described above, between agonizing and intolerable choices. Choices are unavoidable, but a political regime that forces intolerable choice on its population – as a totalitarian state does – is, for Berlin, dehumanizing as it denies its citizens their inherent capacity to choose for themselves. In autopoietic terms, Berlin’s liberalism can be understood and critiqued as a conscious system that understands ‘outside’ events only as perturbations of its own inner workings; in Berlin’s case the basic units being freedom, choice and free will. Equally important in Berlin’s critique of positive liberty, the totalitarian notion of freedom is perceived as a freedom for groups (the working class, the nation) with a specific teleology (the end of history, the expansion of the nation to a new Lebensraum). Berlin had to distinguish between individual and collective freedom, and between positive and negative liberty, because of the challenge totalitarianism posed to liberal societies. Yet these choices are conceived as internal developments within liberal theory. They are self-referential in that they hardly refer to the outside events that caused reflection on these issues, but refer extensively to the canon of liberal thought. Thus, liberalism as a set of values has the first feature of an autopoietic system: operational closure.

Sets of values, as conscious systems, also reproduce themselves. It is beyond the scope of this article to establish exactly what composes conscious systems, and what is being reproduced here. One may look, however, at meme theory as a candidate for explanation. Suffice to say, at this stage, that sets of values reproduce themselves in a manner analogous to how social systems reproduce themselves. The distinction between social and conscious systems is that while the former reproduce communications, the latter reproduce thoughts. These are then exchanged, taught, learned, discussed, and lead to the creation of new ideas and values. Where ‘communication is the social system’s only guarantee of reality’, ⁵⁹ thoughts are the condition of reality of conscious systems.

Luhmann ⁶⁰ summarizes in 7 points all characteristics of self-referential systems, and they all apply to sets of values as conscious systems. (1) They are empirical not transcendent, meaning that the ideas out of which conscious systems are composed are not Platonic forms, but are thoughts formulated, discussed and challenged by empirical human beings through various media (speech, written form, art, etc.). (2) They are multiple and include life, conscious and social systems. (3) They are constituted of the basic elements of the system, which can be produced and reproduced by self-referential units; in the case of conscious systems, they are composed of thoughts. (4) They are autopoietic in the sense that they produce their own elements and interpret them with reference to their own structure, not that of their environment, as I discussed in the case of liberalism’s treatment of totalitarianism. (5) The production of units reduces complexity and creates an identity for the system. This is crucial for conscious systems above all others. The main function of thoughts is to simplify interpretation of a tremendously complex reality. In the same way that a cell needs to interpret changes in its close environment to survive, a conscious system needs to provide interpretations that explain the world in order to be accepted as valid, to acquire an identity (‘liberalism’; ‘socialism’, etc.) and to perpetuate itself. (6) Autopoietic systems are composed of events, thus limited in time. As Luhmann notes, this applies particularly clearly to ‘conscious systems and social systems [that] consist of events only, of thoughts, for example, or of actions’. ⁶¹ (7) They are paradoxical by their very existence, as their unity is a product of their pre-supposed unity, and their difference a product of sameness. This interplay of unity and multiplicity is summarized by Luhmann as ‘the elaboration, by existing, of sufficient identity’. ⁶² Conscious systems are also paradoxical. Berlin’s insistence that his thought makes no metaphysical claims while he defends free will is a clear paradox, as I have shown. The main paradox in Berlin’s thought is what lies at the essence of his defence of the superiority of liberalism over other political systems. Berlin’s pluralism had insisted that values are plural and incommensurable, a statement that is in clear contradiction with his preference for liberalism. Based on liberalism’s respect for choice, this preference is not only weak in Berlin’s thought, it is outright paradoxical. Any conscious system is bound by its own paradoxes, and Berlin’s is no exception. Liberalism as a conscious autopoietic system is characterized by paradoxical axioms from which it starts; in Berlin’s case this paradox revolves around the choice and its articulation within a value pluralist framework.

Value pluralism, as Berlin discussed it, is largely compatible with this theory of autopoietic conscious systems. Sets of values are specific conscious systems, and value pluralism is a meta-theory that is itself a second-order autopoietic conscious system. Berlin had already shown this when he described sets of values that are incommensurable. By acknowledging a complex world where sets of values clash, Berlin had put forward an interpretation of conscious systems, based on agonistic choices. Yet there are several shortcomings of his theory, most notably with respect to his concepts of choice, metaphysics and free will.

Choice is central to Berlin’s outlook. He draws a sharp and fascinating distinction between agonizing and intolerable choices, and concludes that choices are unavoidable, in particular when it comes down to deciding between sets of values. Berlin is convincing in his analysis of the inhumane treatment that accompanies intolerable choices. Being forced to choose between several bads makes a quality life impossible – whatever one defines as a good life. Basing his defence of liberalism on such an attack on forced choice is not as convincing. What Berlin establishes is the superiority of liberalism over totalitarian regimes – a meagre achievement at best. Berlin’s liberalism cannot answer the questions that are central to political theorizing in other respects; for example, how to deal with a community or political regime that imposes one good over others. The choice, in this respect, is no longer between bad and bad but between good and bad; choices are no longer intolerable but restricted. Berlin has been widely critiqued for his inability to deal with ‘communitarian’ critiques. An autopoietic understanding of value pluralism can avoid this pitfall.

Choice is also present in autopoiesis. Systems, whether they are living, conscious, or social, ‘choose’ the elements of their environment that affect them. This ‘choice’ is also a necessary part of the autopoietic system, as it is essential in reducing complexity and creating the system’s identity. Yet this choice is not valued as such by autopoietic theory. Liberalism, as a conscious autopoietic system, necessarily chooses what it is sensitive to. For example, it ignores the theory of autopoiesis as irrelevant to itself, in the same way that a cell would ignore a minor change in its environment that does not impact on its internal functioning. Yet this situation can change. Liberalism can choose to engage with autopoiesis, when Luhmann shows how autopoiesis impacts on political theory. ⁶³ Similarly, a cell will choose to take into account a change in its environment when this change impacts on the inner functioning of its system. However, choice here is never assigned a value. It is neither good nor bad, contra Berlin, but conceived as a necessary feature of all autopoietic systems. Without this choice, unity in a complex world is impossible and there is no autopoietic system to speak of – no life, no consciousness, or no society.

Berlin’s valuing of choice led him to develop a metaphysics. Centred around choice, this metaphysics claimed that one needs to be, at least some of the time, fundamentally undetermined to make valuable choices. Autopoiesis denies this type of metaphysical claim. Mysticism may well be possible within autopoiesis, and Luhmann certainly developed this, ⁶⁴ but mysticism is not metaphysics. Luhmann’s speculation about Lucifer’s relationship to God – as a second-order observation – does not put forward a claim. It shows, by analogy with a biblical story, what autopoiesis is about. Mysticism here is a mode of interpretation, not a claim about the reality of the world. Berlin, on the other hand, feels forced to make a claim about the world to move from the necessity of choice to the valuing of choice. Either we are beyond the realm of physics, beyond causality in the realm of the metaphysical, or all our valuing of choice is wrong. Berlin opts for the former option, but autopoiesis as a theory leaves room for the latter. Our valuing of choice, in other words, is wrong. Choice itself, as shown in the previous paragraph, is neither good nor bad, it constitutes the mode of functioning of autopoietic systems.

Berlin, by focusing on choice as emanating from action, centred his analysis on the role of the individual in choosing one’s way of life and one’s actions. In Luhmann’s autopoietic systems theory, there are two ways in which social systems allocate meaning: either to themselves, or to other systems (or the environment). The first Luhmann describes as action, the second as experience. Berlin, by focusing on choice and action, largely ignores the role that experience plays in systems theory. As Luhmann notes, experience occurs when ‘the system turns to its environment to seek points of contact for further measures’. ⁶⁵ Instead, the individual as the choosing self is the point of measure to which Berlin comes back, without noting that it may be other systems (i.e. the totalitarian state) that trigger his need to rethink liberalism. The self-referential nature of systems does not entail that they are unable, as conscious systems, to engage with their environment. Liberalism can engage with the ‘outside’ events that trigger transformations, and perhaps reject the dichotomies that Berlin is so keen on; between positive and negative liberty, and between choice and determinism. Berlin could have noted that his preference for free choice is born out of a reaction against the model of totalitarian positive liberty of the 20th century, and was not itself an ontological truth that needs to be accepted through the ages.

Berlin further makes an ontological claim about the existence of objective values in the world that is difficult to accept. As Iain MacKenzie has noted, ⁶⁶ Berlin ‘does not sufficiently distinguish between the discovery of values in the world (realism) and the creation of values in the mind (nonrealism)’. Autopoiesis provides an explanatory framework that better deals with this issue. Value systems are understood as closed systems, operating within the constraints of operational closure and self-reproduction. An objective world is assumed, there is an environment to which these value systems react, albeit always according to their own structure and operational closure. The environment is value-free, but every interpretation of it is bound to make value judgements based on the autopoietic system’s identity. Against the Berlinian claims about the truth of value pluralism and the objectivity of (plural) values, autopoiesis posits an objective world, unknown and unknowable in itself, but always already being interpreted by various systems.

‘Free will’, in the strong sense of the term, is the ultimate metaphysical concept upon which rests Berlin’s thought. Yet objections to free will are plentiful, and too numerous to be described here. From the perspective of autopoiesis, free will is completely superfluous. Already in the work of Maturana and Varela there were observations about free will. ‘We cherish our free will and want to be beyond determinism. But at the same time, we want the doctor to cure our diseases by treating us as structurally determined systems.’ ⁶⁷ Without accepting that we ‘choose ourselves’, Maturana and Varela agree we are nonetheless self-created. We may not have the freedom that free will requires, in that we are structurally determined, but we have created and re-created our own structures, so to speak, and are responsible for them at this level. In terms of sets of values, considered as autopoietic entities, we may not be able metaphysically to choose between them – we did not choose our place of birth, upbringing, social background and so on – but they nonetheless require a constant creation and re-creation, as thoughts, which we make our own. Causality is not sidestepped altogether in human choice, but it is no longer conceived in a linear manner; the same cause, at one point in time, leads to different effects on different individuals. But the process of creating an individual, of acquiring an identity, is not itself free from determination. What has made us who we are may be outside our control, but its influence on our identity is always self-created. This is where autopoiesis parallels the ‘communitarian’ critique of liberalism with reference to the creation of an atomistic individual.

Autopoiesis acts as a critique of Berlin’s liberalism, even though it accepts his value pluralism which remains, as Gray pointed out, a ‘single idea of enormous subversive force’. ⁶⁸ As Luhmann noted, there are historical reasons to be sceptical of liberalism’s claims to finding foundations for a universal value of negative liberty. First, because liberalism’s conception of the individual rests on specific historical circumstances arising in the 17^th and 18^th centuries, ⁶⁹ weakening the universality of the claim. Second, because modern society itself is based on the increasing differentiation of its constituent parts, leading not only to different systems, but to different subsystems within systems. Liberalism itself, as a conscious system, is divided into various subsystems which include, among many others, economic (free-market) liberalism, feminist liberalism and value pluralist liberalism. The unity of the system itself is challenged from within, as increasing differentiation – a mark of modernity for Luhmann – leads to what he calls a tendency for the system to short-circuit. ⁷⁰ This tendency is clearly illustrated in the works of John Gray, who has accepted value pluralism, but dropped the liberal conclusions of Berlin to aim for a more neutral conception of modus vivendi. ⁷¹ The preferability of negative liberty, the valuing of choice itself, and the metaphysical account of free will can be dropped altogether in favour of an autopoietic vision where objectivity is put in parenthesis.

Maturana had faced objections that if you put objectivity in parenthesis, political resistance becomes impossible to justify. This is not true. Maturana, who had to flee his native Chile following Pinochet’s rise to power, is adamant that all ways of living are equally legitimate, from a biological perspective. Yet they are not equally desirable. But we cannot justify our preferences, in Chile’s case a democratically elected government over military rule, on objective or scientific grounds. I would add, against Berlin, that nor can we justify our preferences on metaphysical grounds that claim universal ontological validity. Berlin’s liberalism is weak in that it claims that choice must be valued, since it is essential to our free will. But there are no objective reasons to prefer the value of choice over another value, say economic development. Berlin’s error is to have stopped his value pluralism too early, claiming that choice was an over-arching value for anyone accepting value pluralism. But autopoiesis shows us that choice is merely a mode of organization of autopoietic systems, and is value-neutral. In a larger sense, one cannot exclude positive freedom – as Berlin sometimes does – as a valid value system. Value pluralism, as an autopoietic conscious system, merely shows us that there are subsystems that clash with each other in incommensurable ways, but it cannot itself claim to be an objective standpoint from which to observe reality. Autopoiesis abstains from objective claims, as it remains a second-order observation, capable of itself being observed from another standpoint. As Luhmann noted, ‘political theory is an attempt to co-ordinate processes of self-observation and to equip them with the possibilities of self-critique’, but it also ‘would have to give up the belief that, as a science, it could grasp, explain, and improve reality better, more completely and concretely’. ⁷² The task of social and political theory is not to exclude entire sets of values based on ontological claims about metaphysical concepts; but to provide critical tools with which a system’s blind spots can be understood and problematized, instead of being accentuated.

Conclusion

Every system, autopoiesis tells us, has its blind spots. Berlin’s liberalism is no exception. As a conscious autopoietic system, his liberalism draws on the necessity of choice to establish a metaphysical claim about free will. This claim is unconvincing. Autopoietic theory, on the other hand, proposes an analysis of choice that allows for a more nuanced understanding of the role it plays in political theory. Choice here is no longer a value, but merely a necessity for the functioning of a system. By drawing on biological and social autopoiesis, this article has shown that autopoiesis – as it was suggested by Maturana, Varela and Luhmann – clearly applies to conscious systems. By focusing on Berlin, whose value pluralism lends itself so well to autopoietic theory, this article has highlighted shortcomings in his theory, and strengthened the case for value pluralism as an autopoietic meta-theory – by sacrificing the link with liberalism. This short-circuiting of liberalism is not fatal to it; value pluralist liberalism can reformulate itself to accommodate for autopoiesis, but it will need to do so in a more modest way than Berlin did, and be more tolerant of alternative value systems.