Adorno and logic
‘Adorno and logic’ might seem to be a combination as unpromising as ‘Nietzsche and democracy’ or ‘Sartre and Hinduism’. Not only does Adorno have no logic in the sense that he has no theory of the valid forms of argument and inference, he is also deeply hostile to any attempt to formalise thinking. In his view, formal thinking actually impedes genuine reflection, because it disguises the complexities and ambiguities that are inherent in any subject-matter, making it impossible for us to reflect on those complexities. In an effort to encourage genuinely reflective thought, Adorno writes in a fragmentary and allusive style which is far removed from the logically formalised style of argument found in much twentieth century analytic philosophy.
Despite his hostility to formal logic, Adorno does engage with an alternative tradition in logic which Kant and Hegel developed. Kant’s ‘transcendental logic’ studies the basic concepts – such as reality and causality – by which (Kant thinks) we must structure our experience. Hegel transformed this ‘transcendental logic’ into ‘dialectical logic’. He tries to show how our basic concepts emerge out of one another through a process in which each concept turns into its antithesis, then third concepts emerge which reconcile the first two but which then, in turn, spawn their own antitheses. Hegel’s dialectical logic deeply influenced Adorno’s approach to the study of socio-cultural and historical phenomena, especially his account of how ‘enlightenment’ turns into its opposite, ‘myth’. But Adorno also criticises Hegel’s dialectical logic and transforms it into what he calls ‘negative dialectic’. In its most general form, Adorno sees ‘negative dialectic’ as applying to the relations between concepts and objects, or between what he calls ‘identity thinking’ and ‘the non-identical’ (das Nicht-Identische).
To understand Adorno’s thinking with regard to logic (in the Kantian-Hegelian sense), we need to examine this cluster of concepts – the concepts of negative dialectic, of concept and object, of identity and the non-identical – as well as Adorno’s concept of constellations, which forms part of his account of negative dialectic. To make sense of these concepts of Adorno’s, we must first reconstruct Kant’s and Hegel’s conceptions of ‘transcendental’ and ‘dialectical’ logic. Let us start with Kant.
1. Kant to Hegel: From transcendental to dialectical logic
In his Critique of Pure Reason (1781), Kant distinguishes between ‘general’ and ‘transcendental’ logic. ‘General’ logic, he says, sets out the ‘rules of thought’, without ‘regard to … the objects to which the understanding may be directed’. That is, general logic is purely formal and sets out the rules for how one must think – regardless of what one is thinking about – if one’s thinking is to be valid. Thus by general logic, Kant means what we ordinarily mean by logic.
‘Transcendental logic’, on the other hand, only has meaning within the framework of Kant’s ‘transcendental idealist’ philosophy. Here I understand Kant’s transcendental idealism to be his view that we cannot know objects as they are in themselves, but only as they appear to us when subjected to the forms of representation that we must necessarily bring to experience. These forms are (1) the forms of ‘pure intuition’ (space and time) by which we must structure our sensory impressions, and (2) the concepts of our understanding by which we must organise the disparate materials of sensation into structured objects that interrelate in orderly ways.
Kant calls forms of inquiry ‘transcendental’ if they investigate the conceptual and intuitive conditions that we must necessarily bring to experience. ‘Transcendental logic’ focuses on the conceptual conditions. It sets out to identify which concepts must necessarily structure any thinking at all that we do about objects (or any thinking in which we order the materials of sensation). Kant refers to these necessary concepts as the ‘pure concepts of the understanding’, or the ‘categories’. Saying that they are ‘pure’ means that they are not derived from experience, since we need to employ these concepts in order even to have any experience (of objects) in the first place.
So which are the concepts that must necessarily structure our experience? Kant argues that all concepts function to unify manifold impressions that are given to us in sensation – concepts are ‘functions of unity’. He claims that judgements (such as ‘the rose is red’) also unify – they always unite a subject (‘rose’) with a predicate (‘red’). For Kant, different kinds of judgement unite subjects and predicates in different ways: ‘the rose is red’ unites them by affirming that the predicate belongs to the subject, while ‘the rose is not red’ unites its terms by affirming that the predicate is ‘opposed to’ the subject. In fact, according to Aristotle’s logic – which was canonical in Kant’s day – there are exactly twelve kinds of judgement: twelve different ways of uniting subjects with predicates. Kant claims that this fact provides a ‘clue’ to what the categories are. Since there are twelve ways in which judgements can impose unity (on subjects and predicates), and since concepts are also functions by which we impose unity (on sensory materials), there must be twelve pure concepts each corresponding to a particular kind of judgement. For instance, the category of ‘reality’ corresponds to affirmative judgements such as ‘the rose is red’, while the category of ‘negation’ corresponds to negative judgements such as ‘the rose is not red’. Other categories that Kant lists include unity and plurality, causality, necessity and impossibility.
Kant takes himself not to have produced a merely ‘haphazard’ list of pure concepts, but to have worked out systematically, from a sound principle, what the categories must be. But Hegel criticises Kant for relying on traditional Aristotelian logic. Hegel finds it unphilosophical for Kant simply to assume that that logic is sound. Hegel holds that our basic categories must instead ‘be exhibited in their necessity, and … it is essential for them to be deduced’. Thus, Hegel accepts Kant’s view that we must actively bring structure to our experience, and that there must be certain basic categories by which we structure any experience whatever. But unlike Kant, Hegel thinks that the philosopher must work out what these categories are by ‘deducing’ each of them.
What must such deduction involve? Hegel argues as follows. First, we show that some particular category is necessary for any thought. Then we show that this category has limitations, such that we need an additional category which provides the only possible – or, at least, the best available – corrective to those limitations. Then we show that this new category is also limited, such that yet another category is required – and so on until we have deduced a complete chain of categories. Hegel sets out this chain of categories, or ‘thought-determinations’ (Gedankenbestimmungen) as he also calls them, in his logic (of which two versions exist: the longer Science of Logic from 1812-16, and the shorter Encyclopaedia Logic first published in 1817).
Unlike Kant, though, Hegel thinks that the logical categories are both forms of our thought which organise objects as we experience them, and basic structures or principles that organise things as they exist independently of our minds. For example, when Hegel deduces the category of causality, he takes causality to be not only a category in terms of which we necessarily think but also a basic principle which structures all mind-independent things so that all they stand in causal relations to one another. Indeed, for Hegel, it is because these basic principles organise all things that they also constrain the shape of human thought and experience. Thus by a ‘category’ Hegel means not merely (as Kant did) a basic form of thought but also a basic ordering principle that obtains in mind-independent reality. Hence Hegel claims that he, unlike Kant, sees the categories not as merely subjective but also as ‘the truth, objectivity, and actual being of … things themselves. [The categories] resemble the platonic ideas … which exist in individual things as substantial genera’.
I noted that Hegel derives his chain of categories by trying to show that each one in turn is limited and must be followed by a further category. In particular, he argues that the categories always follow one another according to a three-stage ‘dialectical’ process. At the first stage – the stage of ‘abstraction’ or ‘understanding’ – we begin with a particular category. For instance, Hegel’s logic begins with the category of being, defined as the simplest and most inescapable category. At the second stage – the stage of ‘dialectic’ proper – the initial category either generates or turns into its opposite. Hence being turns into nothingness, just because the category of being is so simple that being is entirely indeterminate and featureless. At the third, ‘speculative’, stage the first two categories become reconciled with one another. Being and nothingness become reconciled as follows. Nothingness does not really differ in content from being, since both are equally featureless; in that sense, nothingness turns back into being. But being and nothingness remain distinct in that they prove to be the same from opposite starting-points: being goes from being to nothingness while nothingness goes from nothingness to being. It follows, for Hegel, that being and nothingness are (1) distinct but (2) interdependent, since each exists only inasmuch as the other constantly turns into it. Moreover, Hegel concludes, this makes them aspects of the third, overarching, category of becoming.
While Hegel’s logic derives a series of basic categories, then, a higher-level logical structure governs how these categories follow one another. This structure consists of the three stages of the ‘dialectic’: abstraction, dialectic proper, and speculation. This dialectical structure regulates both the series of concepts with which we think and the series of metaphysical principles that govern reality.
We see, then, that Hegel has transformed Kant’s ‘transcendental’ logic into a ‘dialectical’ logic (a kind of logic, which, again, Hegel distinguishes from formal logic) by (1) trying to derive the categories according to this three-stage dialectical structure and (2) by interpreting the categories as not merely ‘subjective’ but as metaphysical principles as well. We may now see how Adorno draws on and criticises Hegel’s dialectical logic.
2. Hegel to Adorno: From dialectical logic to negative dialectic
Adorno often constructs his socio-cultural and historical analyses by applying Hegel’s dialectical method – the method of tracing how categories turn into their opposites – to social or cultural phenomena. A notable example is Adorno’s (and Horkheimer’s) Dialectic of Enlightenment (1944), which centrally claims that ‘enlightenment reverts to mythology’. For Adorno and Horkheimer, the intellectual phenomenon of enlightenment is subject to a dialectic whereby it turns into its opposite, myth.
By ‘enlightenment’, Simon Jarvis notes, Adorno and Horkheimer do not mean the 18th-century intellectual and political movement. Rather, they take enlightenment to be a gradual process that has taken place over all of human history, a process of ‘demythologizing, secularizing or disenchanting … mythical, religious, or magical representation[s] of the world’. Thus, according to Adorno, humanity has repeatedly distanced itself from each of its previous systems of thought by criticising those systems for being merely mythical. This progression has been fuelled by humanity’s desire to gain increased practical control over nature. ‘The Enlightenment has always aimed at … establishing [humanity’s] sovereignty’; ‘the understanding, which overcomes superstition, is to hold sway over a disenchanted nature’. Human beings have hoped that, by freeing themselves from merely mythical views of nature and by thereby gaining increased insight into the real workings of nature, they will enhance their ability to intervene into these natural processes for their own benefit. As Adorno and Horkheimer see it, the climax of this process of enlightenment, or of progress away from myth, has been the rise of modern science. Modern science has enabled humanity to exercise unprecedented control over nature by representing nature as devoid of qualities, made up purely of extensive magnitudes which can be understood using mathematical formulae.
Yet, Adorno and Horkheimer claim, the more humans try to become enlightened and to distance their thought from myth, the more they fall back into mythic modes of thinking. This happens in a range of ways. Let us consider just one of these. Enlightenment thinkers (in Adorno and Horkheimer’s special sense of ‘enlightenment’) aim to avoid appealing to mythic beliefs – in gods, supernatural forces, etc. – by sticking to the facts. These include the facts of how society is currently organised and of the chain of historical events that has led to society being organised in this way. But enlightenment thinkers cannot ask whether the course of history could have been different, as this is not a simple factual matter. This failure to ask whether things might have turned out differently conduces to an unthinking acceptance that current social arrangements are inevitable. Hence, enlightenment thinkers tend to regard these arrangements as an unalterable given, a ‘fate’ that no-one can escape. But fatalism of this kind is a typically mythical form of belief.
Ultimately, underlying the dialectic whereby enlightenment reverts to myth is a dialectic whereby enlightenment reverts to nature. As we have seen, the ultimate goal of enlightenment is to increase human control over nature, which distances humans from nature both because it defines human beings as masters over rather than part of nature and because it requires that human beings develop the ability to use abstract concepts. Nonetheless, we humans remain partially natural. We each have what Adorno calls an ‘inner nature’, consisting of our natural bodily impulses. So part of the enlightenment goal is to put humans in control of their own inner natures – to enable them to exercise self-control, to resist their own instinctual impulses, and (by doing so) to free themselves to engage in dispassionate conceptual thought.
Yet humans have pursued this enlightenment project of controlling (inner and outer) nature because they hope that this control will enhance their ability to preserve themselves. The whole enlightenment effort to distance humans from nature has been fuelled by a natural desire for, or impulse towards, self-preservation (Selbsterhaltung). Thus, Adorno writes, ‘nature is reflected and persists’ in thinking just insofar as the latter serves as a ‘mechanism of compulsion’, i.e. of domination. Ironically, the more earnestly people pursue the enlightenment project, and try to control and distance themselves from nature, the more they are in fact submitting to their natural impulses. Hence, Adorno sums up, ‘human history, the history of the progressive mastery of nature, continues the unconscious history of nature’.
We see, then, how Adorno traces both the dialectic whereby enlightenment inverts into myth just when it tries to separate itself from myth, and the underlying dialectic whereby enlightened humanity submits to nature just when it tries to stand above nature. This parallels the way in which, for Hegel, being turns into nothingness just when being assumes the form of a distinct category. Adorno thinks that many other social phenomena become subject to a similar dialectic. Minima Moralia gives many examples: The intellectual who withdraws from society, because s/he is critical of its calculating and commercial nature, ends up becoming sanctimonious and so just as hard-hearted as the calculating people from whom s/he wants to differ. What counts today as being ‘healthy’ and well-adjusted rests on a pathological level of repression (of awareness of one’s own suffering), in contrast to which those who are sick or eccentric are actually ‘healing cells’. ‘Indiscriminate kindness to all [actually expresses] indifference and remoteness to each’. And so on.
To return to the dialectic of enlightenment, Adorno also seems to follow Hegel in hinting that the solution to this dialectic is a reconciliation (Versöhnung) between enlightenment and myth and, above all, between culture and nature. What form does Adorno think this latter reconciliation should take? He hints that there could be a ‘remembrance of nature in the subject’. If we could acknowledge, or ‘remember’, that our pursuit of enlightenment has been driven by natural impulses, then we could begin to free ourselves from subjection to those impulses. Having acknowledged the hold that these impulses have had over us, we could then ask whether we think it right to pursue these impulses or whether we hold other goals to be more desirable. In other words, we would become able to deliberate about what values to adopt. With this, our thinking would have gained some independence from our impulses. ‘Through the decision in which spirit [i.e. the human mind] acknowledges itself to be domination and retreats into nature, it abandons the claim to domination which makes it a slave of nature’.
Adorno, then, supports the enlightenment project to the extent that he wants humanity to become genuinely liberated from nature. Yet the paradox is that, for Adorno, we could only truly free ourselves from our natural impulses if we acknowledged that our thinking depends on those impulses in the first place, and if we admitted that the enlightenment project has been an expression and outgrowth of those impulses. Steven Vogel concludes that Adorno’s ideal of reconciliation is hopelessly ambiguous between wanting us to be truly free from nature and wanting us to admit our dependence on – and, therefore, lack of freedom from – nature.
Plausibly, though, Adorno’s model of reconciliation is coherent. It seems to follow the Hegelian model, on which two opposed items get reconciled by proving to be distinct but interdependent (as being and nothingness proved to be distinct but interdependent aspects of becoming). If we acknowledged that our thinking is not separate from but depends on – grows out of – nature, then our thinking could truly distinguish itself from nature for the first time (as we ask whether it is desirable to follow the natural impulses that we have acknowledged). If, on the other hand, we continue to insist that our thinking is separate from nature, then that thinking will actually remain driven by – and fused with – the natural impulses whose influence we are refusing to acknowledge.
But we should not overstate how faithful Adorno is to Hegel. A first obvious difference is that Adorno is studying dialectical processes that he takes to be happening within the historical social world rather than within basic categories. As such, when Adorno suggests solutions to these dialectical processes, such as his suggested culture/nature reconciliation, he sees these solutions only as possibilities, which it is up to humanity to bring about in practice – and, as he admits, humanity may well never do so.
There are two further differences between Adorno’s and Hegel’s models of reconciliation. Secondly, in Hegel, the reconciliation of two opposed categories generally takes place because the second category proves to be essentially the same as the first, and to be distinct from the first only at a relatively superficial level. For instance, it is because nothingness proves to be essentially just as featureless as being that nothingness and being turn out to be merely two aspects of becoming. In accordance with this model, Hegel claims that culture and nature become reconciled when modern scientific and philosophical modes of inquiry establish that nature is in itself essentially rational. This claim implies that when humans use rational thought to control and transform natural things, humans are doing no wrong, because they are actually helping nature to realise its inner rational essence. Adorno would object that this Hegelian view disguises the fact that human domination over nature does wrong nature. So the culture/nature reconciliation which Adorno envisages takes a different, non-Hegelian form. As Adorno sees it, what we need to acknowledge is that our thinking depends on and grows out of natural impulses that are not wholly rational and so which cannot be exhaustively understood in rational terms. Even if nature and natural impulses can be partly understood through reason, they always have an additional non-rational aspect that resists human understanding. Thus, in recognising that our thinking depends on natural impulses, we need to recognise that our thinking depends on a nature that differs irreducibly from rational thought (and which, therefore, is damaged when it is moulded in line with human reasoning).
A third difference between Hegel and Adorno is that, for Hegel, two items are reconciled when they prove to be interdependent. For Hegel, being and nothingness depend on one another, and likewise, rationality ultimately proves to depend on the nature out of which it grows, while nature proves to depend on the rational principles which provide its inner structure. Adorno, on the other hand, stresses that culture, rationality and enlightenment depend on nature to a greater extent than nature depends on reason. He encapsulates this point by speaking of the ‘primacy [Vorrang] of the object’ (i.e. nature) over thinking: any thinking, reasoning subject is always a particular object (a particular body, brain and set of impulses) but not all objects are reasoning subjects. Even though all objects have an intelligible structure and in that respect may be said to ‘depend’ on reason, since objects also have a non-rational element they do not wholly depend on reason, whereas reasoning activity does wholly depend on objects (or as Adorno says, ‘concepts … are moments of the reality that requires their formation’).
These three points provide an initial indication of how Adorno transforms Hegel’s dialectic into a ‘negative’ dialectic. Hegel’s dialectic is ‘positive’ in that it always reconciles two opposed items by showing that the second is essentially the same as the first (e.g. that nature is rational, like culture; that nothingness is featureless, like being) and that the first and second items depend on one another. Adorno’s dialectic, in contrast, is ‘negative’ because: (1) it only suggests possible – not actual – forms of reconciliation; and because on its model, reconciliation occurs when a first thing (e.g. culture or enlightenment) that has been trying to separate itself from and to dominate something else (e.g. nature or myth) acknowledges both (2) that that other thing is irreducibly different from it, and (3) that it depends on that other thing (e.g. culture on nature) to a greater degree than the other thing depends on it.
Why does Adorno say that this dialectic is ‘negative’ whereas Hegel’s is ‘positive’? Adorno does so because in his dialectic, (from (2) above) two items are reconciled when they remain and are acknowledged to remain different from one another – that is, to be not the same as one another, but to each remain the negation of the other. ‘Dialectics’ as Adorno practices it, he writes, ‘is the consistent sense of non-identity’.
As Adorno sees it, this negative dialectic applies in its most general form between our concepts and the objects that we attempt to understand using these concepts. So, to better understand Adorno’s negative dialectic, we need to look at his ideas about concept/object relations, which are entwined with his ideas about the relations between ‘identity thinking’ and the ‘non-identical’.
3. Concept and object, identity and non-identity
In his magnum opus, Negative Dialectics (1966), Adorno claims that the nature of concepts is that they each apply to many different things: e.g. the concept ‘dog’ applies to many things – all those that are dogs. As such, each concept grasps the things to which it applies as instances of a universal. If I conceive something to be a dog, then I understand it as an instance of the universal kind ‘dog’. Adorno therefore says that conceptual thinking both (1) is classificatory thinking (through the use of concepts, we say ‘what [kind] something falls under, what it exemplifies’) and (2) is ‘identity thinking’. Why ‘identity’ thinking?
When I conceptualise something as an instance of a kind, I see it as identical (firstly) to all the other instances of the same kind, insofar as they too are instances of that kind. But this means that conceptual thinking gives me no knowledge about what is unique about any thing, e.g. about what is special about this dog as distinct from all the other dogs. So if one is thinking about something conceptually, then one cannot recognise what is unique in that thing and one sees it only as an instance of a kind. In that sense, when one thinks conceptually, one (secondly) ‘identifies’ things with the universal kinds under which one takes them to fall. (‘To think is to identify’, Adorno declares.)
However, Adorno thinks, it is in principle possible to recognise that things are never simply identical to these kinds (or to the other instances of a given kind) but always have a unique side as well. Adorno does not assert that things are ever wholly unique. He believes that things do ‘come under’ concepts in virtue of instantiating universal kinds. But this falling under concepts is not all there is to things. Each thing is also unique; Adorno refers to this aspect of things as the ‘non-identical’ element in them – that element in things by virtue of which they are not identical either to the kinds that they embody or to all the other instances of those kinds.
Now, Adorno has a range of objections to identity thinking. Firstly, he objects that identity thinking not only tells us nothing about what is unique in things, but also (as we saw above) disguises the very fact that things have a unique side. Secondly, Adorno objects – and this is an ethical objection – that identity thinking is linked to domination. This link has at least two aspects.
Firstly, when we conceptualise things, we dominate them in thought. Because conceptual thinking suggests that things are merely instances of universal kinds, and because we can understand universal kinds using concepts, conceptual thinking suggests that things are reducible to what we can understand of them. In other words, conceptual thinking portrays things as lying wholly within the reach of our intellects.
Secondly, according to Adorno, the whole purpose of conceptual thinking is to enable us to practically dominate the things that we have conceptually mastered. As Adorno and Horkheimer outlined in Dialectic of Enlightenment, we humans began to conceptualise things – and have sought to do so ever more accurately and ever less mythically – so that we can grasp how things work and can therefore intervene into their workings in ways that promote human self-preservation. So as Adorno sees it, ‘any conceptual thought … is a kind of instrument [of practical domination] and to that extent a form of mastery’. Even seemingly disinterested forms of inquiry such as mathematics or astrophysics belong within the project of science, which is inherently a project of dominating nature.
Adorno holds, then, that things always have a ‘non-identical’ element in addition to their universal side. But what is this ‘non-identical’ element, this element in things which makes them each unique? Medieval philosophers held that the element in things that makes them each unique is their haeccitas or ‘thisness’ – a property in things that makes each of them into this individual thing that it is. The notion of haeccitas, though, seems to mark the site of a problem – what is it that makes each thing this thing that it is? – instead of solving that problem.
One alternative approach to uniqueness – adopted by Hegel in his Encyclopaedia Logic – has it that the uniqueness of any thing consists in its distinctive way of instantiating a universal kind. Hegel argues that a thing cannot instantiate a universal unless it does so in some particular way, and this particular way of instantiating a universal is what makes the thing the ‘singular individual’ that it is – where the thing’s particular way of instantiating a universal cannot be captured by mere reference to the relevant universal.
However, this cannot be what Adorno means by the ‘non-identical’ in things. Adorno insists that he is not giving us a general concept, or definition, of what non-identity or singular individuality is. To give a concept of singularity would be to treat singularity itself as a universal kind or principle which all things, insofar as they have a singular side, instantiate. (Indeed, Hegel is quite clear in the Encyclopaedia Logic that he is treating universality, particularity and singularity as universal principles in this way.) To this Adorno objects that ‘as soon as we reflect upon the single … individual as an individual, in the form of a universal concept – as soon as we cease to mean only the present existence of this particular person [or thing] – we have already turned it [i.e. the single individual] into a universal’. But if we treat the individual in its singularity as an instance of the universal kind ‘singular individuality’, then we lose sight of the fact that the singular side of things is, precisely, that in virtue of which they are always more than mere embodiments of universals (even of the universal kind ‘singular individuality’).
If Adorno’s references to the non-identical in things are not meant to give us, or to rely on, any general concept or definition of what the non-identity or singular individuality of things consists in, then what status do these references have? Perhaps Adorno believes that we can gain immediate access to any thing in its unique individuality by abstaining from conceptualising that thing and by approaching that thing solely through our senses. But this view – the view that we can ‘grasp what constitutes the unique essence of the thing as an individual only if [we do] not use concepts in knowing that individual’ – had, as Adorno is well aware, been forcefully criticised by Hegel. Hegel calls this view – which he criticises in his Phenomenology of Spirit (1807) – ‘sense-certainty’.
Hegel’s objection to ‘sense-certainty’ is that to know something one must be able to pick it out, and that, if one wants to pick something out without using any general concepts to do so, then one is reduced to referring to that thing as ‘this’, or as what is ‘here and now’, or as ‘what I mean’. But ‘this’, ‘here’, ‘now’, and ‘I’, Hegel argues, are themselves universal concepts, so that after all one is still using concepts – not just one’s senses – in trying to know things. Yet ‘this’, ‘here’, etc., are highly general concepts which can refer to any number of things or people. These concepts on their own cannot pick out the particular thing that I am claiming to know. Hegel concludes that – contrary to what ‘sense-certainty’ claims – we cannot attain knowledge of particular individual things without using concepts, but that we need finer-grained concepts than merely ‘this’, ‘here’, etc. Adorno agrees with Hegel’s conclusion that we cannot know anything without using concepts. For Adorno, objects cannot be known ‘immediately’ (i.e. just through the senses); we cannot apprehend objects without ‘mediating’ them through concepts.
Adorno seems to be in a dilemma. He wants it to be possible for us to recognise that things have a non-identical element without conceptualising that element in universal terms, and yet he denies that we can know anything without conceptualising it. Adorno has a way out of this dilemma, though. He sees the concept of the non-identical as a limit-concept. It does not give us any positive knowledge about things. All it does is indicate the place where our conceptual understanding runs up against its limits. The concept of the non-identical indicates that there is a side of things which our concepts do not cover. But the concept of the non-identical also indicates that we cannot know anything about this side of things just because it is what our concepts cannot cover. Thus, Adorno thinks that it is possible, using concepts, to recognise that there are limits to what one can understand conceptually, and specifically to recognise that concepts are not adequate to the non-identical element in things.
By recognising the limits of conceptual thought in this way, we could bring about reconciliation between concepts and objects. Following Adorno’s model of reconciliation, this would involve:
(1) us acknowledging (using concepts) that concepts depend on objects to a greater extent than objects on concepts (SO 502). Objects do ‘depend’ on concepts insofar as objects always have a conceptually intelligible form, but as this is only one aspect of objects, they remain independent of concepts in respect of their non-identical side. On the other hand, concepts are entirely dependent on objects: concepts only arise insofar as there are already objects that we seek to understand, and concept-forming activity also depends on objects of one particular kind – human brains and bodies. Thus, Adorno writes that mind originates in the ‘real life process’ and that consciousness is ‘a ramification of the energy of drives’.
(2) us also acknowledging (again using concepts – specifically the limit-concept of the non-identical) that objects are never reducible to our concepts of them because objects always retain a non-identical element. (Indeed, it is because of this non-identical element that objects only ever depend on concepts partially and not to the same extent as concepts depend on objects.)
Were concept and object reconciled in this way, their relation would have assumed a negatively dialectical form. Dialectical, in the sense that the formerly antagonistic relation of concepts to objects would have been overcome (an antagonism that has manifested itself in our efforts to dominate and wholly understand things using concepts). But negatively, rather than positively, dialectical, because in the reconciled state objects would remain and be acknowledged to remain irreducibly different from – not the same as – concepts. In sum, ‘Reconciliation would release the nonidentical, … [and would] be the remembrance of the many [i.e. of items that are different from one another] as no longer inimical’.
However, there are obstacles to our recognising the limits of conceptual thought and bringing about this condition of concept/object reconciliation. Whenever we recognise that a concept is limited, we will inevitably try to produce a new, improved, concept in order to overcome this limit. This is where Adorno introduces the concept of constellations.
Inevitably, Adorno thinks, once we recognise that our concepts are limited we will try to produce improved concepts in order to overcome those limits. This is inevitable because the whole purpose of concepts is to enable us to control things practically by giving us accurate knowledge about those things and about how they work. Thus, whenever a particular concept proves to be giving us incomplete knowledge about things – to say ‘what something falls under … and what, accordingly, it is not itself’ – then we will be compelled to try to produce a better concept that will give us fuller knowledge about things and so enable us to dominate things more effectively. Since this concept will prove limited in turn (because all concepts are inherently limited), we will be led us to produce yet another concept, itself limited as well – and eventually we will end up with a whole range of concepts. ‘The … flaw in every concept makes it necessary to cite others; in this way constellations arise’: these being clusters or ranges of concepts which gather or circle around the particular objects that they all try (but fail) to understand.
Adorno’s claim that we inevitably strive to extend our concepts when they prove limited is indebted, again, to Hegel. In his Phenomenology, Hegel argues that when the concepts ‘this’, ‘here’, etc., prove to give us no knowledge of particular things, we must turn to richer concepts (e.g. ‘dog’, ‘desk’) which can do so. Moreover, Hegel argues in the Phenomenology, to grasp any thing as the particular thing it is we must conceptualise it using a range of these relatively rich concepts. To pick out a particular dog, I must classify it not merely as a dog but also as light brown, friendly, excitable, middle-aged, etc. Since no one thing has exactly the same range of characteristics as any other, we can grasp a thing in its uniqueness by using a range of concepts to specify the complete set of characteristics that is unique to that thing. (It might seem that this account of singular individuality from Hegel’s Phenomenology differs from his account in his Logic, discussed earlier. But plausibly these two accounts are actually just two parts of one single account. On that account, a thing embodies a universal kind in a distinctive way just in that it embodies a unique range of universal properties; e.g., this dog embodies the kind dog in a distinctive way by being a light brown, friendly, (etc.) dog.)
Adorno’s idea of ‘constellations’ at first sounds similar to Hegel’s idea that by using a range of concepts we can grasp the uniqueness of a thing. For Adorno suggests that the range of concepts that are gathered around a thing ‘illuminates’ or gives insight into that thing. ‘Setting [concepts] in constellation … illuminates what is specific to the thing, to which the classificatory procedure is indifferent’. And since he expands on this point with the metaphor of unlocking something by using a combination of numbers rather than one single number, this suggests that he endorses Hegel’s view that we can understand the uniqueness of a thing by conceptualising it in terms of a whole range of concepts.
But in fact Adorno cannot believe that constellations are like Hegelian ranges of concepts, for at least two reasons. Firstly, Hegel only thinks that the use of a whole range of concepts enables us to know an object qua singular because he thinks that an object’s singularity consists in that it embodies a unique range of universals. But, I suggested earlier, Adorno does not want to offer a general definition of what singularity is. Therefore, he cannot accept Hegel’s definition that the uniqueness of things consists in their embodying distinctive ranges of universals.
Secondly, Hegel’s definition of singularity implies that objects can be exhaustively understood if we use a rich enough palette of concepts. In contrast, the thrust of Adorno’s thought is that there are inherent, inescapable limits to conceptual understanding. For instance, he speaks of the ‘unavoidable insufficiency’ of ‘thought’ – and since for him ‘To think is to identify’, that is, to use concepts, he is taking it that all thought is conceptual and is unavoidably limited on account of being conceptual.
Yet Adorno’s very idea that constellations of concepts can illuminate a thing qua unique suggests that in fact it is only single concepts that are inherently limited but that groups of concepts are not, or not necessarily. To be consistent with his general stress on conceptualisation’s inherent limits, then, Adorno’s conception of constellations will have to include the idea that constellations can only ever give us partial, non-exhaustive, insight into things qua unique (such that even when grouped in a constellation, concepts still remain limited in their capacity to illuminate objects).
How then, if not in Hegel’s sense, do constellations of concepts illuminate the non-identical element in things while yet illuminating it only partially? Introducing a second sense of ‘constellation’, Adorno suggests that each object is itself a constellation of different past relations with other objects, all of which have shaped it. The
immanent universality of the individual thing is objective as sedimented history. … To become aware of the constellation in which a thing stands is to decipher the constellation which, as something that has become, [the thing] carries within itself … [K]nowledge of the object in its constellation is knowledge of the process stored within it.
On this second account, an object is a constellation of historical processes, and a constellation of concepts is a range of concepts each of which grasps one of the various historical relations that has become absorbed into, or has left its mark on, the object. Taken together, these concepts ‘gather around’ the unique history of the object, where this unique history makes the object the unique thing that it is.
Constructing a constellation of concepts around an object is different from grasping the object in terms of a range of universals. A constellation of concepts captures the particular historical relations that have shaped an object, rather than whatever universal kinds that object may embody. And a constellation of concepts can only ever capture some, not all, of these relations. This is because the history of the relations and influences that impact on and surround an object always continues for at least as long as that object exists. These ongoing relations and influences do not merely add to an object’s history. Rather, whenever these new relations and influences arise, they make it possible for us to see further aspects of the object’s past history which could not previously be appreciated. So it is not only the case that no constellation of concepts can anticipate all the future influences and relations that will mark an object; but also, no constellation of concepts can grasp all the relations that have previously impacted on the object, because many of these cannot be recognised in advance of the further unfolding of the object’s history. Constellations, then, can never provide complete understanding of their objects.
In effect, when we construct constellations of concepts around objects, we assume that those objects have histories of which they are the products, and we endeavour to use concepts to assemble narratives about aspects of these histories as they impact on the objects. It might seem that this approach can only apply in the understanding of human-made cultural and social artefacts, but actually the approach can also apply to natural things, insofar as they are all products of cosmic, geological, chemical, evolutionary, and other kinds of processes. An example of this way of approaching an object (a cultural object in this case) through a constellation of concepts is Hannah Arendt’s account of the ‘origins’ of totalitarian forms of socio-political organisation, or more precisely, as she says, her ‘historical account of the “elements” which “crystallized” into totalitarianism’. For Arendt, totalitarianism, as an object, was a constellation of elements – including Western imperialism and the ‘collapse of Western moral standards’ that it involved, modern anti-Semitism, and economic, industrial and social changes that had left people in modern societies in a rootless, alienated, condition. Arendt’s historical account of totalitarianism gathers together a plurality of concepts, each of which grasps one of these various elements (where these are only some, not all, of the factors shaping totalitarianism).
There are at least two potential objections to Adorno’s second account of constellations. Firstly, that this account presupposes a definition of singularity – a definition on which the singularity of an object consists in the total set of historical interactions that has marked it. Even so, this definition (unlike Hegel’s definition of singularity) implies that no object can ever be exhaustively understood, because the history of each object remains constantly ongoing. The definition implies that we can only ever approach the singularity of any object by assembling concrete narratives about some of the specific historical processes shaping that object. But this brings us to the second objection: it now seems that the reason why concepts, even when they are in constellations, can never exhaustively understand any object is because of the nature of objects, specifically the fact that those objects’ histories – which make those objects the particular objects they are – are unfinished, ever-ongoing. Concepts, it appears, are incomplete because their objects are incomplete, being embroiled in ongoing processes such that, until more of those processes have taken place, many aspects of those objects’ preceding histories are simply not available to be grasped.
Thus, Adorno’s second idea of constellations is in some tension both with his reluctance to define singularity and with his tendency to stress that concepts are limited in regard to objects (rather than stressing, as his second idea of constellations implies, that concepts are limited because objects are incomplete themselves). Despite these tensions, Adorno’s second idea of constellations is promising. It opens up the prospect that we might be able to use constellations of concepts to reveal something of the singularity of objects, without our use of concepts leading us back to the false assumption that concepts can give us exhaustive understanding of objects.
This essay has introduced two main aspects of Adorno’s thinking about dialectical logic. On the one hand, he describes the dialectical process in which enlightenment and culture revert to their opposites, myth and nature, just when they try to separate themselves from myth and nature. On the other hand, Adorno offers a model of how concepts and objects could be reconciled and could enter into a negatively dialectical relationship in which it is acknowledged that concepts depend on objects which differ irreducibly from concepts.
Moreover, we can now see that enlightenment and culture become subject to the dialectic in which they revert to their opposites because this reversion manifests the relationship in which enlightenment and culture actually stand to myth and nature. Enlightenment and culture may attempt to deny their dependence upon myth and nature, but since that dependence actually exists it must manifest itself, and will do so all the more in proportion as it is denied. Just as for Freud repressed sexual desires must manifest themselves somehow – in the form of symptoms and mental pathologies such as hysteria – likewise for Adorno the asymmetrical relationship between concepts and objects, enlightenment and myth, culture and nature must manifest itself. It can do either benignly – if it is acknowledged – or destructively – if attempts are made to deny it.
This reveals something else. Adorno presupposes that, in reality, enlightenment and culture – and, more generally, concepts – stand to myth and nature – and, more generally, to objects – in a negatively dialectical relationship, whereby the former set of items depend asymmetrically upon the latter. Previously I have suggested that concepts and objects would enter into a negatively dialectical relationship only if they were reconciled. But in fact, were concepts and objects to be reconciled, this would only mean that we were acknowledging the negatively dialectical relationship in which concepts and objects already, ultimately, stand to one another. This is a relationship whereby concepts depend on, but can never exhaustively understand, objects – a relationship that has throughout history been denied, with the disastrous consequence that ‘the fully enlightened earth radiates disaster triumphant’.
What we see from all this is that Adorno’s thinking about dialectics forms a coherent whole, which revolves around his conception of the negative dialectic between concepts and objects. So, although Adorno has no logic in the formal sense that is generally accepted today, near the centre of his thought are his reflections on a range of topics – the concept/object relation, the limits of conceptual thinking, and the nature of dialectic – which do belong under the heading of logic in the expanded sense which Adorno inherits from Kant and Hegel. ‘Adorno and logic’ may be an unlikely combination, but it is also a surprisingly fruitful one.
 See Adorno, ‘Subject and Object’ (1969; hereafter SO), in The Essential Frankfurt School Reader, ed. Andrew Arato and Eike Gebhardt (New York: Continuum, 1982), p. 498; ‘Zu Subjekt und Objekt’ (hereafter ZSO), in Kulturkritik und Gesellschaft II (Frankfurt: Suhrkamp, 1977), pp. 741-42. Throughout, when quoting from translations of Adorno’s works I sometimes correct these without special notice.
 Here I explore only how Adorno relates to Kant’s and Hegel’s conceptions of logic and not to their thought in any other fields such as politics, morality or metaphysics.
 Kant, Critique of Pure Reason, trans. Norman Kemp Smith (Basingstoke: MacMillan, 1929), A52/B76, p. 93.
 This position is ‘idealist’ in the sense that it claims that we can know things only as we represent or form ‘ideas’ of them and not as they are in themselves.
 Kant, Critique of Pure Reason, A55-56/B79-80, pp. 95-96.
 Ibid., A68/B93, p. 105.
 Ibid., A72/B97, p. 108.
 Ibid., A81/B107, p. 114.
 Hegel, Encyclopaedia Logic (hereafter EL), trans. T. F. Geraets, W. A. Suchting, and H. S. Harris (Indianapolis: Hackett, 1991), §42R, p. 84.
 Michael Forster suggests that for Hegel a category B counts as the best available corrective to the limitations of some category A if B corrects A’s limitations while differing from A in content less than any other known category (Forster, Hegel’s Idea of a Phenomenology of Spirit (Chicago: University of Chicago Press, 1998), p. 186.
 EL, §153-§154, pp. 227-30.
 Hegel, Philosophy of Nature, trans. M. J. Petry (3 vols.; London: Allen & Unwin, 1970), §246A, vol. 1: p. 200. It might be objected that Kant does see the categories as ‘objective’ both in the sense that they are necessary for any thinking whatever and in that they structure objects as we experience them, which (for Kant) are the only objects about which we can have positive knowledge. But Hegel argues that this Kantian view of the objectivity of the categories makes their objectivity dependent on the subject who applies the categories and so ultimately reduces objectivity to something merely subjective (EL, §41-§42, pp. 81-86).
 EL, §79-§82, pp. 125-33.
 Ibid., §86, pp. 136-39.
 Ibid., §87, p. 139.
 Ibid., §88, p. 141. Thus, in terms of how I described Hegel’s logic above, being proves to be limited in that it turns into nothingness, and the category that gives the best available corrective to this limitation is becoming.
 Ibid., §19A, p. 48.
 Dialectic of Enlightenment (1944; hereafter DE), trans. John Cumming (London: Verso, 1997), p. xvi; Dialektik der Aufklärung: Philosophische Fragmente (hereafter DA), Gesammelte Schriften vol. 3, ed. Rolf Tiedemann (Frankfurt: Suhrkamp, 1981), p. 16.
 Jarvis, Adorno: A Critical Introduction (Cambridge: Polity Press, 1998), p. 24.
 DE, pp. 3, 4; DA, p. 19, 20.
 DE, pp. 24-25; DA, pp. 41-42.
 This range reflects the fact that myth is itself a complex mode of thinking with many aspects – or so Adorno and Horkheimer think; hence they do not define myth.
 Here I am expanding on an argument made in Jarvis, Adorno, pp. 25-26. See also DE, p. 27; DA, p. 44.
 DE, p. 13; DA, p. 29. See on this Deborah Cook, ‘Adorno’s Critical Materialism’, in Philosophy and Social Criticism 32: 6 (2006), p. 724.
 DE, p. 39; DA, p. 56.
 DE, p. 39; DA, p. 56.
 Negative Dialectics (1966; hereafter ND), trans. E. B. Ashton (London: Routledge, 1990), p. 355; Negative Dialektik (hereafter NDk; Frankfurt: Suhrkamp, 1975), p. 348-49.
 Adorno, Minima Moralia: Reflections from Damaged Life (1951; hereafter MM), trans. E. F. N. Jephcott (London: Verso, 1974), p. 26; Minima Moralia: Reflexionen aus dem beschädigten Leben (hereafter MMb; Frankfurt: Suhrkamp, 2003), pp. 27-28.
 MM, p. 73, and see also pp. 62-63; MMb, p. 81, also pp. 69-70.
 MM, p. 77; MMb, p. 86.
 Adorno only hints at the possibility of a reconciliation (DE, p. 40; DA, pp. 57-58). Nonetheless, ‘[t]he central impulse of the Dialectic of Enlightenment is towards the possibility of a reconciliation of culture and nature’ (Jarvis, Adorno, p. 35).
 DE, p. 40; DA, p. 58.
 DE, pp. 39-40; DA, p. 57.
 Or so Steven Vogel notes in Against Nature (Albany: SUNY Press, 1996), p. 62.
 Ibid., p. 62.
 See my Petrified Intelligence: Nature in Hegel’s Philosophy (Albany: SUNY Press, 2004), ch. 3.
 On this view of Hegel’s, see John Passmore, ‘Attitudes to Nature’, in Environmental Ethics, ed. Robert Elliott (Oxford: Oxford University Press, 1995), p. 135.
 ND, p. 192; NDk, p. 193.
 He claims that objects do have definite properties, or ‘form’, which are intelligible by thought and in that sense constitute a ‘subjective’ element within objects (SO, p. 504; ZSO, p. 749).
 ND, p. 11; NDk, p. 23.
 In respect of myth, though, Adorno and Horkheimer do say that ‘myth is already enlightenment’ (DE, pp. xvi, 8; DA, pp. 16, 24). That is, mythic thought already tries to understand nature with a view to controlling it, albeit in defective ways (e.g. by seeing nature as peopled by gods whose moods can be influenced through rituals and sacrifices). Here Adorno might seem to be following a Hegelian model and claiming that enlightenment need not – and cannot – separate itself from myth because myth is already part of the process of enlightenment anyway. However, Adorno believes that mythic thought is in part magical thought, and that magical thought in turn draws on ‘mimesis’ (DE, pp. 9-11; DA, pp. 25-27). Mimesis is a non-rational, instinctual behaviour in which one organism imitates another organism or object. For example, in magic one might try to harm someone by harming a doll that resembles (and so ‘imitates’) that person. In respect of its mimetic element, mythic thought is entirely non-rational, and because myth always has this non-rational aspect, it remains irreducibly different from enlightenment rationality.
 ND, p. 5; NDk, p. 17.
 ND, p. 149; NDk, p. 152.
 ND, p. 5; NDk, p. 17.
 Consider again, for instance, his claim that all objects do have conceptually intelligible form (SO, p. 504; ZSO, p. 749).
 Are all things unique? What about mass-produced items? But this is precisely why Adorno objects to mass-production – on the grounds that it erases the singularity which is – or should be – proper to things.
 Gordon Finlayson, ‘Adorno on the Ethical and the Ineffable’, in Journal of European Philosophy 10 (2002), p. 4; emphasis added.
 EL, §163, pp. 240-41.
 Or as Adorno puts it, he is not giving an ‘ontology’ (a theory of what kinds of thing exist) which states what ‘non-identity’ is (ND, p. 136; NDk, p. 140).
 SO, p. 498; ZSO, p. 741.
 Robert Stern, Hegel and the Phenomenology of Spirit (London: Routledge, 2002), p. 45.
 Robert Pippin, Hegel’s Idealism: The Satisfactions of Self-Consciousness (Cambridge: Cambridge University Press, 1994), p. 119.
 ND, pp. 186, 172; NDk, pp. 187-88, 173.
 This idea of a limit-concept derives from Kant. For Kant the concept of things-in-themselves functions as a limit-concept, marking that there must be such things (otherwise there would be nothing to appear to us subject to our forms of representation) but that we cannot positively know anything about what these things are like. Adorno is sympathetic to this Kantian ‘block’ (as he calls it; ND, p. 386; NDk, p. 378) on the possibility of positively knowing about things-in-themselves. Hegel objects to Kant that as soon as we identify limits to our knowledge then we become compelled to try to go beyond them. Adorno agrees with Hegel here. (Hence, as Deborah Cook notes in her introduction to this volume, Adorno thinks that Kant gives up too quickly on the possibility of knowing things-in-themselves.) Tacking between Kant and Hegel, he will argue that these efforts to overcome our limits result in constellations which (he believes) can reveal something about the singularity of an object without reducing it to universality. See Section 4.
 ND, pp. 198, 265; NDk, p. 199, 262.
 ND, p. 6; NDk, p. 18.
 ND, p. 53; NDk, p. 62.
 Hegel calls the kind-universals that things thus embody in distinctive ways ‘concrete universals’, and the property-universals (by instantiating which things embody the kinds in distinctive ways) ‘abstract universals’; see Robert Stern, ‘Hegel, British Idealism and the Curious Case of the Concrete Universal’, in British Journal for the History of Philosophy 15: 1 (2007), sec. III.
 ND, p. 162; NDk, p. 164.
 ND, p. 163; NDk, p. 166.
 ND, p. 5; NDk, p. 17.
 On the difference between Adorno’s two conceptions of constellations (as groups of concepts and as the aspects of an object’s history), see Michael Rosen, Hegel’s Dialectic and its Criticism (Cambridge: Cambridge University Press, 1982), pp. 167-68. Adorno derives this second account of constellations from Walter Benjamin’s suggestion, in his Origins of German Tragic Drama (1928), that constellations are clusters of concepts which capture facets of the history of a unique, particular event or phenomenon; see Benjamin, Origin of German Tragic Drama, trans. John Osborne (London: New Left Books, 1977).
 ND, pp. 163; NDk, p. 165-66.
 Seyla Benhabib, The Reluctant Modernism of Hannah Arendt (London: SAGE, 1996), p. 64, quoting Arendt. Like Adorno, Arendt derived this understanding of constellations from Benjamin.
 Ibid., p. 65.
 DE, p. 3; DA, p. 19.
 This remains true even though Adorno only came to formulate that account explicitly after co-writing Dialectic of Enlightenment.