Self-Interpreting Animals

Charles Taylor​ has had a long, fêted career as a philosopher with a particular interest in bringing the French and German traditions into productive conversation with their Anglo-American counterparts, and in bringing political and ethical theorising to bear on contemporary politics both in the UK and in his country of birth, Canada: he was an active participant in the British New Left, and stood more than once as a New Democratic Party candidate for the Canadian parliament. So it isn’t entirely surprising that, even in his 94th year, he might want to make a contribution to current conversations about the climate crisis, and more generally about how best to acknowledge the depth of our dependence on the natural world. What is surprising, even disconcerting, is that it should take the form of a 600-page analysis of Romantic and post-Romantic Western European poetry.

Of course, sometimes the most stimulating contributions to a conversation are the most oblique, but it wouldn’t be difficult to mount a case against taking this one seriously. The book’s title has an archaic resonance, but also suggests we may be in for an elevated form of spiritual self-help. The main discussion is highly selective in a highly predictable way, focusing on ten or eleven poets from a 200-year historical span, all of them dead white men and long-canonised monsters of fame (Wordsworth and Hölderlin, Baudelaire and Mallarmé, Eliot and Miłosz); and it engages with them in a humanistic manner that is essentially unmarked by developments in literary theory and criticism after structuralism. Its interpretations depend on a range of interlocking philosophical views (about language, ethics and history) that Taylor just about has room to summarise within the confines of the book’s opening and closing sections, but not to justify. More amplitude in that dimension has been sacrificed in favour of extensive quotation from the featured poets – and I do mean extensive. Since Taylor quite rightly offers the original German or French as well as an English translation, there are many stretches of the book in which his own prose commentary becomes little more than a self-effacing setting for exceptionally generous, duplicated excerpts from the poetry itself. In this sense, Cosmic Connections is far shorter than it may seem.

The book also bears the marks of its long gestation and late arrival. In his eagerness to guide the reader through vast and tangled territory, Taylor often repeats material, in both the main text and the footnotes, and distinguishes the themes, facets and phases of his account so often that the resulting chains of subheadings proliferate and overlap bewilderingly. The style of the writing veers abruptly from elegantly polished paragraphs to staccato sentences that more closely resemble annotations scribbled in the margins of the verse, as if offering a glimpse of early drafts. The whole thing is crying out for a good editor; but that element of publishing has been in decline for a long while now, and here editorial restraint was no doubt reinforced by the author’s eminence. In this sense, Cosmic Connections is far longer than it needs to be.

All that said, there is something enthralling as well as endearing about this late work. Taylor’s passionate admiration for and responsiveness to his chosen poets, and his sense of the profound importance of the issues they confront, come through essentially undiminished. The problem is that, for any reader unfamiliar with Taylor’s other writings, this book doesn’t provide the tools needed properly to understand these intimations of deeper significance. Without some appreciation of its contexts, it will be hard to see that Taylor’s basic concern here is not really or primarily literary: it is with the nature of Western European modernity, and its immense, traumatic impact on our sense of ourselves and our place in the universe.

The most proximate of those contexts is what Taylor refers to as this book’s companion volume. The Language Animal (2016) defends a philosophical account of language and its role in human life that was first developed by thinkers on the philosophical wing of the Romantic movement (primarily J.G. Hamann, J.G. Herder and Alexander von Humboldt), and was later elaborated and refined by Heidegger, Merleau-Ponty and Wittgenstein. The core idea here is that human beings are self-interpreting animals. What Taylor means by this is that the identity of any human being is determined by their interpretation of that identity, which is itself bound up with their interpretation of the objects and situations they encounter.

All animals react to the world they inhabit as having import for them and their wellbeing, as they seek sustenance and avoid threats; but with human animals, those proto-interpretations are themselves interpreted and so shaped by the language humans inherit, which already informs their shared social world. This intimacy between self, words and world is evident in the fact that the language through which we characterise the objects and situations we encounter is also used to characterise the desires, emotions and goals elicited by them. An emotion such as shame, for example, is tied to a certain sort of situation (a shameful or humiliating one) and to a certain sort of response to it (covering up, hiding and so on); and we couldn’t grasp the meaning of any one of these elements without relating it to the other two. The kind of self-concealment elicited by shame is not the kind appropriate to hiding from an enemy: fig leaves make poor camouflage.

This is a hermeneutic circle: a structure of meaning that operates holistically, and whose inner complexity allows us to distinguish appropriate from inappropriate responses to a given situation. We can feel ashamed of things that are not genuinely shameful, and respond shamelessly to things that are. Within the circle, then, there is room to criticise our responses; but the significance of the circle itself is also holistic – shaped by the larger context into which it fits. Encountering a situation as ‘shameful’ will mean something different according to which modes of response are available to us – whether we can contrast shame with guilt, for example, or only with fear, or also with terror and disgust. The wider the field of available contrasts, the finer the discriminations we can make in characterising our responses (and the more various the ways in which we might be criticised for responding inappropriately). The significance of both the response and the situation are thus conditioned by this broader semantic field. Introducing new terms into that field will change the meaning of the existing terms, and thereby change the experiential meaning of the situations they characterise, and the feelings or purposes they elicit.

In short, new linguistic articulations can reconfigure the way we make sense of our own feelings, thoughts and responses – our internalised self-interpretations. And what is true of our most basic feelings and emotions also applies to more complex aspects of self-interpretation: our conception of our individual character, of what a good life for us might be, even of what it is to be human. No self-interpretation, at any level, is beyond question. Saying doesn’t make it so, because all self-interpretations can be subject to critical evaluation, but any such criticism will itself require a particular vocabulary, and so will presuppose a specific interpretation of the self in its world. There is no escaping this hermeneutic circle, because there is no such thing as a genuinely human life outside the horizons of meaning, no human life that is not a life of language (and of the broader cultural and social forms that languages inform and are informed by).

These horizons make up what Taylor calls ‘the interspace’. Within them, aspects of the world are made available, but in ways that can be grasped only by relating them to the human capacity for self-interpretation. In this kind of space, the capacities of distinctively human animals do not inevitably impose anthropomorphic distortions or delusions; they can sometimes mislead us, but then so can any of our capacities for grasping reality, and the appropriate response to such risks is to deploy them with care and critical self-awareness. And if we dispensed with them altogether, we would render ourselves incapable of grasping central dimensions of the real significance of things.

This is the philosophical anthropology of Romanticism. And it naturally tends to give art a fundamental role in shaping the way we conceive of ourselves and what it is possible for us to be, because the poets’ use of language at full stretch is a primary way in which we reconfigure the semantic fields through which we orient ourselves in the world. This is why, when Taylor identifies the central theme of his book as the human desire for a connection with the surrounding world, he immediately emphasises that the forms this desire takes have varied significantly between different phases of human history.

So Taylor’s foregrounding of the post-Romantic poets has a double justification. According to the philosophical account he draws on, the articulations of poetry are as fully capable of disclosing the existential significance of human life as the words of philosophers or saints. And the articulations provided by those poets who aim to inherit the example of Wordsworth and Hölderlin repeatedly reconfigure the desire for cosmic connection, the array of self-interpretative possibilities it informs, and so the broader horizons of human meaningfulness that make up the modern world.

According to Taylor, the Romantic poets and their successors were reacting to the way that the scientific advances we associate with Western European modernity disenchanted our world. Before the Enlightenment, human beings interpreted themselves as inhabiting a cosmos whose significance was anchored in a higher reality, an objective metaphysical order that structured the universe and determined what constituted a good life for its human inhabitants. Because the materialism of modern science seemed to offer a comprehensive way of understanding the world without relating it to any such order, it threatened to disrupt those enchanted self-interpretations, and in particular to frustrate the human desire for cosmic connection. So these poets tried to rearticulate that desire in ways that would allow us to continue to seek its satisfaction without simply denying natural science.

Romantics such as Wordsworth continued to use metaphysical articulations of cosmic order, but they presented them primarily as convincing ways of accounting for certain experiences of openness to a larger whole. The authority of those reinterpretations resides in the willingness of individual readers to regard them as compelling articulations of their own responses to the poem, and to analogous experiences of their own, rather than as metaphysical doctrines with an independent rational grounding. But the early generations of post-Romantic poets shifted the balance even further away from the metaphysical. Baudelaire retained his conviction in the reality of something beyond the world of everyday experience, but lost faith in our ability to articulate it conceptually, accepting that it can only be marked by a symbolic register whose meaning is essentially contestable. And by the time of Mallarmé, even the idea of such a beyond loses its significance, or rather its significance is tied ever more tightly to the interspace between subjects and objects.

For Mallarmé, even thoroughly disenchanted everyday objects can still exhibit their true vitality of being, but only in fleeting experiences of resonant fullness before which words fail – unless they can be marshalled to enact that failure, and so offer a negative, self-emptying linguistic transposition of such experiences. But if our intuitions of connectedness are now embodied only in experiences that resist coherent articulation, affirming them puts increasing pressure on the idea that they count as human experiences at all – as something that opens subjects to the reality of objects in a way that can bear interpretation. Mallarmé’s ‘beyond’ thus seems to lie outside the structures of self-interpretation without which human life ceases to make sense.

To articulate such a beyond accordingly requires forging a language that signifies without organising itself around the seeming polarity between subjects and objects. The indication is that our desire for connection with reality can be realised only through experiences of overcoming the conditions that make experience possible in the first place. And since one of those conditions is time – experiences are dateable events – undergoing these extraordinary experiences amounts to the interruption of ordinary time by something that at once negates it and perfects it (freeing the essentially contingent, sequential order of everyday existence from its limitations so as to exhibit its higher nature).

One of the best moments in the book’s long chapter on Mallarmé shows how his enigmatic ‘Sonnet en -X’ might be read as enacting such an experience. The poem is structured around a rhyme-scheme involving words ending in ‘x’. Taylor suggests that it asks us to imagine an empty room just after sunset, suffused with melancholy at the passing of day. ‘The last flashing of the light is reflected in a mirror, a glace de Venise, surrounded by a cadre in gold, with a unicorn motif. There is a flickering here, as though a transaction between frame and mirror, unicorns against the nymph of the lake. Then nothing.’ A constellation arises in the night sky, and is reflected in the mirror. There is no one to see either the flickering or the reflection, and the only objects involved in either transaction dematerialise, foregrounding the momentary passage and reduplication of light. And yet the poem conjures not merely an empty physical space but a definite place at a highly charged time.

This is because it invokes phenomena (a last flickering of sunset, a mirror-image of the Big Dipper) whose existence is dependent on someone’s being in the right place to see them, while removing any such observer; and yet its way of doing so enhances their reality rather than undermining it. The poem’s constellation of window, mirror and world acknowledges the interrelation of subjects and objects, and so the openness of human awareness to the cosmos, by transcending them. It invokes subjectivity without subjects and objectivity without objects, and thereby discloses that the underlying reality of both subject and world lies beyond their mutually dependent opposition, because it lies beyond the very distinction between them. The sonnet invites us to experience a higher form of connection, the kind of unity that results from transcending the apparently absolute requirement that if two things are to be connected to each other, they must be distinct from each other.

Such poetic constellations of images will inevitably risk failing to communicate at all. So there is some point in Taylor talking about a process of intensifying epistemic retreat here. By the time we get to Mallarmé, the post-Romantic poem is reconfiguring language in ways that don’t just pass over traditional philosophical and literary methods of claiming authority for one’s insights (offering no doctrines or arguments, and treating the resources of ancient myth, with its nymphs and unicorns, as lifeless ornaments), as well as dispensing with many of the basic grammatical conventions that facilitate mutual intelligibility.

Taylor’s reading of the ‘Sonnet en -X’ is thus not a summary of the objective intellectual content of Mallarmé’s poem; it is an attempt to articulate his own experiential response to the poem. It is his way of making sense of its confounding constellations of imagery, and of the vision of cosmic order those constellations seem to disclose, from which we might derive ethical empowerment (by recognising that our distinctness from the world is mere appearance, so coming to view the world’s wellbeing as part of our own). But the only way he can convince us of the rightness of his account is by finding a way of expressing his response that leads us to share it – to experience the connectedness it glimpses, and to which it aspires to orient us.

This is not, however, a complete surrender of art’s claims to provide insight or understanding. For we can still ask whether a given response to such a poem is appropriate or inappropriate, better or worse, in its sensitivity to the poem’s content and context. As I have noted already, interpretations are always subject to criticism – they don’t bottom out in a brute reaction of liking or disliking. But even if such responses are not arbitrarily subjective, they are extremely personal. The pressures under which words are placed in a Mallarmé poem open up many possible ways of relating its elements, and so create a good deal of room for reasonable disagreement about how best to do it. Each reader’s response to such a poem risks failing to resonate appropriately with their experience of the world and their inheritance of words. The most that can be hoped for is an alignment of subjectivities – an attunement of specific individuals that opens a space of mutual intelligibility, and so a community informed by a particular interpretation of poem, self and world. And if each new poem enacts a new constellation of meaning, then it will require new attempts by its readers to align their subjectivities, and thereby to maintain their lines of communication with Mallarmé and with one another.

This general vision of literary art and criticism as seeking such attunement is something the Romantics inherited from Kant. He argued that our responses to artworks aspire to a universality of agreement that they cannot demand on the basis of incontrovertible fact or compelling argument. And the challenge involved in realising such agreement is intensified in the context of modernism, where a loss of faith in the ability of inherited artistic conventions to facilitate genuinely valuable artworks forces artists to test which conventions might continue to sustain good work, and if necessary to construct new conventions based only on their individual conviction that work employing them measures up to the best work of the past, and on the hope that their audience might agree. Both aspects of art in the modern era reflect a shift in our conception of selfhood that has been the focus of Taylor’s work for more than thirty years, ever since the publication of Sources of the Self in 1989. This is the second main context into which Cosmic Connections must be fitted if we are to appreciate its significance. For it is the third in a sequence of large historical narratives of modernity that Taylor has written (the second being A Secular Age, published in 2007), and they all operate within the framework laid out in that immensely influential first book.

In​ Sources of the Self (and A Secular Age), the shift that interests Taylor is from a conception of the self as porous to and deeply integrated with the wider cosmos, to a conception of it as punctual or buffered, strongly demarcated from its environment. Rather than having to conform to a cosmic order of which it forms merely one element, the modern self is capable not only of controlling its own expressions of itself, but of determining the standards by reference to which it exercises that control. Because such rational self-mastery involves abstracting from our ordinary embodied ways of experiencing reality, as if taking an external perspective on ourselves as well as our world, it also involves viewing them both as subject to rational comprehension and manipulation. The vocabulary of that vision is provided by modern physics: a vocabulary of pure materiality, in which everything real can be captured in terms of matter in motion, as a network of quantifiable causal forces expressible in the language of mathematics.

In this vocabulary, there is no room for qualitative as opposed to quantitative differences, and so no room for meaning of the kind integral to the interspace that Taylor associates with distinctively human forms of animal life. So the increasing cultural authority of this buffered conception of the self imposes on us all a self-interpretation that makes it impossible not only to understand ourselves as self-interpreting animals, but also to understand what is morally attractive and empowering about the self-conception it is built around.

For rational self-mastery is the taproot of an ideal of autonomy, of controlling our own lives according to our own values, which has spread far and wide from its Western European matrix; and it is the source of many of our currently dominant, extremely demanding moral imperatives: to reduce or even end the suffering of others, to protect their freedom to choose how to live their lives. Yet for such moral demands to make sense to us, we must be able to articulate their nature and their sources, their significance or meaning, which will involve laying out and defending the interpretation of self and world that is implicit in them. But this modern interpretation of self and world denies the reality of such horizons of meaning, because they are not articulable in the language of modern physics. So it becomes impossible for those drawn to modern moral ideals properly to justify them, or even to understand their nature.

The claim that human beings are self-interpreting animals carries with it a particular conception of practical reasoning, of how to work out the right thing to think and do, as well as the best available standards against which such judgments should be measured. If people orient themselves in relation to a particular conception of who they are and what they can become, then they will evaluate themselves in terms of how close they are to realising their conception of a good life, and what changes would realise that conception more fully. In other words, they interpret themselves in narrative terms – as on a journey or trajectory from their current state to a better, realisable but as yet unrealised, state. And since the merits of the moral ideal that governs that journey are open to criticism, it is always possible that its best current expression might be improved, so anyone living by its lights must be willing to consider objections to it, and if necessary to reconfigure it.

In short, practical reasoning is a reasoning in transitions. Its aim is not to establish whether some ethical stance is absolutely correct, regardless of who we are or where we find ourselves in human history. That is a model appropriate to natural science, and so deeply attractive to us moderns, but it depends on a notion of objectivity suited to a realm of meaningless matter, not self-interpreting animals. Ethical reasoning operates in the interspace, in the realm of meanings. Its aim is to establish whether one concrete stance is superior to another concrete stance in some specific respect, and hence whether moving from the latter to the former will achieve a concrete epistemic gain – by resolving a particular confusion, by better acknowledging the value of a factor currently overlooked, and so on. And any transition held to be error-reducing could be criticised by denying that a gain would be achieved, by proposing a rival interpretation of that transition, or proposing a different transition.

This is the main reason Taylor’s recent work has been so historically oriented. Taken together, these three volumes amount to an exercise in practical reasoning – a highly specific interpretation of the present of Western culture as the outcome of a series of conceptual transitions in the past, and so as the jumping-off point for further transitions in the future. They offer a genealogical narrative of our current form of life from a specific ethical perspective, a collective biography that will inevitably present its pivotal transitions as either error-reducing or error-enhancing, and so should itself be evaluated in the same terms. It is a full-blown moral argument, not a conceptual or historical preparation for one; and the specific spiritual perspective it exhibits is explicitly religious (more specifically, Roman Catholic).

Sources of the Self had three interrelated goals: to show that our present non-theistic self-conception has deprived itself of what it needs to articulate and defend its own value, and is unable to understand its own thoroughly interpretative status; to demonstrate that it emerged from certain developments in its theistic ancestors and competitors; and to argue that these transitions were error-enhancing rather than error-reducing. The implication is that we should give serious consideration to the possibility of transitioning from our current secular modes of self-understanding towards the best available contemporary versions of theistic thinking.

We can get a sense of the way Taylor’s narrative works to achieve its goal by examining one element of it – his account of modernity’s distinctive affirmation of ordinary life. This originates in the Reformation, as an aspect of Protestantism’s rejection of a prevalent Catholic view of the sacred, according to which sacramental rites or membership of monastic or clerical orders brought individuals closer to the divine. Protestantism held that the everyday human realms of production and reproduction were equally sacred, so long as we approached them with the correct attitude – a detachment from earthly pleasures and a refusal to treat created things as ends in themselves. But this early version of a disengaged stance towards the world mutated once it was combined with the hollowing-out and discarding of puritan conceptions of original sin; for that encouraged us to regard an increasingly wide range of human desires as good in themselves, and to regard benevolent behaviour to all as expressive of divine goodness. These notions were originally attached to the deist ideal of how one might contribute most to the world’s providential order; but utilitarians overly impressed by the disenchanted vision of nature bequeathed by modern science transposed this privileging of benevolence onto a wholly naturalistic ideal of fulfilling our animal drive for desire-satisfaction.

This is one way in which the secularisation of religious ideals overreached itself. It took disenchantment to the point of undercutting the intelligibility of secular ethical ideals, and it overlooked the points at which both Catholic and Protestant self-interpretations might have revised themselves in ways that would have prevented the transition to deism and then to naturalism from appearing to be error-reducing or insight-enhancing (by reconceiving the relation between laypeople, priests and monks, say, or by reinterpreting rather than discarding the notion of original sinfulness).

Secularisation also provides the larger argumentative context for Cosmic Connections. For one way in which the modern buffered self developed depended on interpreting moral sentiments such as benevolence as showing that one’s feelings as well as one’s reason might be an internal mode of access to what is good. And this led to Rousseau’s idea that nature speaks through us, which inaugurated the Romantic conception of nature as an independent locus of value and allowed us to imagine recovering contact with our most intimate feelings as a recovery of the spirit of nature in us.

This late turn in the modern self-understanding, together with its post-Romantic descendants, is hardly neglected in Sources of the Self. But in Cosmic Connections it is given a far more extensive examination, to the exclusion of any of its competitors. Why might that be? Given the pro-theistic orientation of Sources of the Self, I suspect it is because Taylor now thinks that the afterlife of Romanticism is the most fruitful place to look for a non-theistic mode of self-interpretation that can acknowledge the most valuable ethical aspects of the modern conception of selfhood while holding open lines of communication with, and so the possibility of an insight-enhancing transition to, a suitably reconfigured theism.

Sources of the Self contrasts the modern buffered self with an image of the pre-modern self as seamlessly taking its place in a larger cosmic order. In Cosmic Connections that image is recharacterised as one realisation of the perennial human need for connection with the larger reality in which it finds itself. But the genealogical nature of human moral understanding – history’s role in conditioning the content and range of our available vocabularies for self-understanding – means that recovering a liveable form of theistic ethical vision could not conceivably be a matter of simply returning to that pre-modern mode of religious faith. That is no longer a live existential option for Taylor or his readers; and although he argues extensively in A Secular Age against those who think that discarding religious belief is a necessarily error-reducing move, he fully acknowledges that once such faith becomes one among a range of viable but optional ethical stances (as opposed to being the primary, unchallenged and unproblematic stance), its claims on our attention must be significantly recalibrated.

This is why it matters to him that there are modes of modern self-understanding which intuit the inadequacies of the buffered self and articulate ways of overcoming them, while shedding the assumption of an objective divine underpinning in favour of accommodating the ontological salience attained by the self in the modern era. For the advocates of such self-interpretations share Taylor’s sense that any viable recovery of the value of cosmic connectedness in the wake of disenchantment will have to go through the self rather than attempting to downplay or otherwise bypass its centrality.

Taylor here shows that he shares with other philosophers interested in inheriting Romanticism a sense that the Enlightenment’s reinterpretation of selfhood in terms of a strongly bounded inwardness might begin by emancipating us from unquestioning acceptance of our place in a larger order, but quickly generates a paralysing anxiety about how to reconnect with the reality that lies beyond those boundaries. Modern science kickstarts this anxiety by claiming that our perceptions systematically misidentify (as coloured, noisy and tasty) the reality that engenders them. Modern philosophy radicalises it, by picturing us as metaphysically imprisoned within our consciousnesses, restricted to the impressions the world makes on us without being able to check them against what causes them or even to check that they really are caused by something outside us. Modern politics struggles to reconcile individual autonomy with collective decision-making, and to make sense of an individual’s entry into social life as an enhancement of selfhood rather than a way of putting it in chains. And modern art struggles to reconcile its dependence on conventions with its drive to give authentic expression to an individual vision that might resonate with others.

In all these ways, the emancipation of the modern self creates a deepening sense that it is fated to get in its own way, to frustrate or subvert its own deepest impulses. That which frees us from submission to external reality – the self’s independence of its world – is precisely what prevents us from truly connecting to that world. Such scepticism is the price of discarding a conception of human consciousness as porous to reality, as inherently open to what transcends it; but since we cannot return to the pre-Enlightenment interpretation of self and world that enabled that mode of sense-making, we must find a way of recreating it from the forms of self-interpretation that supplanted it.

On the philosophical side of the modern mind, this reconstruction is traced in The Language Animal in the work of Wittgenstein and the phenomenological tradition, particularly Heidegger and Merleau-Ponty, who dismantle the idea of human consciousness as a representational medium which interposes itself between self and world. They suggest instead that the human being is always already out in the world, cohabiting with objects and events, not trapped within its body but living out an embodied comprehending engagement with its environment. And on the poetic side, Mallarmé’s modernist constellations of imagery aspire to induce fleeting experiences that disclose reality’s uncanny intimacy with us and thereby recover a sense of cosmic connectedness; but they reattune us to that revitalised desire by exploiting the artist’s and reader’s individual responsiveness rather than bypassing it. His poetic practice is built on the hope that even the most idiosyncratic reconfigurations of logic, mythology and grammar might form the basis of a community of understanding, and so activate a mode of interspatial meaning whose enhanced acknowledgement of subjectivity nevertheless allows for the convincing reformulation of a fundamental insight into the self’s unity with reality.

How does this post-Romantic poetic project relate to theism? Taylor notes Mallarmé’s early Christian faith, as well as certain resonances between his later quasi-Hegelian philosophical vision and Christianity – in particular, the way in which Christianity seeks to balance an emphasis on the infinite value of the individual soul with an ethic of unprecedented self-denial in properly acknowledging the vital reality of creation. But, most important, he follows his pivotal discussion of Mallarmé with just two further chapters of historical narrative, on T.S. Eliot and Czesław Miłosz respectively. In both, Christian faith is in dynamic struggle with doubt, and so their work acknowledges the epistemic retreat to the dimension of linguistic resonances between subjectivities that is the main connecting thread of Taylor’s exposition. Both nevertheless create a picture of a believable, theologically centred cosmic order that draws on the symbolist traditions inspired by Mallarmé but radically reconceives both their despair of everyday life and their vision of its transcendence.

It seems to me that Taylor views Eliot’s reconstruction of literary and religious traditions in Four Quartets – with its orchestration of memory and desire revitalised by the momentary irruptions of higher time – as constituting the most convincing case for making an error-reducing transition from secular post-Romanticism to a form of theism that has absorbed its lessons and refined them. The move is not in any sense presented as compulsory, as if a requirement of rational self-control; in fact, Taylor makes no explicit claims at all about how we should orient ourselves in relation to Eliot’s project. But by placing it at the climax of his long account, Taylor is at least inviting his readers to consider a particular revision of their self-understanding. For if those four constellated poems resonate as deeply in our experience as Eliot means them to, and if Taylor’s own articulation of how best to understand why they do so also resonates deeply with us, then we have as much of a justification as we can reasonably expect to begin a serious exploration of the theistic option in late modernity – and as much of a justification as we need.

This last claim​ returns us to the issue with which I began: why devote time and effort to reading such a deliberately old-fashioned book on poetry, written by a philosopher? Taylor’s affinity with Romanticism and its afterlife repeatedly leads him to present the poetic and the philosophical not as two essentially distinct cultural enterprises, but as two aspects of a single project of reconnection. Yet he doesn’t seem entirely comfortable with this idea of an internal relation, and in particular of an equality of rational standing, between poetry and philosophy. He certainly emphasises that poetry can yield a powerful experience of reconnection, but he also insists that the insight gained through such experiences is very different from the conviction gained through the force of argument, and that this is why poetic insight will often be incomplete, tentative and enigmatic, inexplicit and provisional. The implication seems to be that philosophical argument can supply us with convictions that are significantly less tentative, provisional and enigmatic than any provided by the subtler languages of poetry, and that philosophy is therefore a more robust or reliable way of gaining insight.

But two points speak against such a privileging of philosophy. The first is that providing arguments is not the only way of taking full responsibility for one’s discourse. Another way is to arrange words in a constellation whose linkages exploit various dimensions of each word’s individual history, its patterns of denotation and connotation, and its idiosyncratic associations, in such a way that each link – and so the whole which, taken together, they constitute – can be convincingly accounted for. Such articulations of the sense of a poem are certainly more figurative than deductive, more dependent on personal than impersonal modes of attunement. But that doesn’t mean they must always be incomplete, tentative or enigmatic, let alone that poetic sense-making is an arbitrarily subjective matter as opposed to a rational one. For any agreement they secure is as fully the result of a precise and robust linguistic ordering as is agreement on a conclusion secured by a suitably interlocking set of premises.

The second reason for questioning Taylor’s privileging of philosophy is that in key respects his own mode of philosophising resembles the practices of his chosen poets. For over the vast span of his published work, even in the three historical volumes I have focused on, Taylor endlessly varies and revises the vocabulary in terms of which he attempts to articulate his vision of modernity, its discontents and its possible modes of redemption. I have had to smooth out these shifts here, but anyone familiar with his writings will know that ideas taken from other philosophers are almost always conveyed in terminology of Taylor’s coining, and that when he reiterates his own philosophical ideas in later works, he keeps on looking for new ways of expressing them. It is as if he is searching for the perfectly subtle language that will resonate most precisely with his readers’ experience, and thereby convey an insight that will attune his subjectivity with theirs, and both with the reality that the words disclose. Each revised vocabulary has its strengths and weaknesses in this respect, but none is entirely perspicuous, so each proves to be provisional, always capable of specific refinements that nevertheless risk creating opacities of their own, and so ceaselessly on. Such ways with words are not at all conducive to the strict regimentations of sense that valid argument-forms demand.

Because Taylor’s readings of these poets are so deeply informed by his post-Romantic philosophical commitments, his articulations of their figurative constellations often draw on the concepts that embody those commitments. But if such poetry requires so much philosophical vocabulary in order that it might be properly interpreted, then what sense is there in maintaining that poetry’s ways of making sense of things are fundamentally distinct from those of philosophy? Taylor’s twin muses – Heidegger and Wittgenstein – both found it necessary to turn to more poetic modes of taking responsibility for their later philosophical writings. Heidegger did it by interpreting poets and calibrating their uncanny intimacy with genuine thinking, and Wittgenstein did it by coining an array of signature concepts (‘language-games’, ‘forms of life’) and aphoristic imagery through which to convey philosophical insight without resorting either to argument or to the assertion of doctrines. In my view, then, just as Taylor’s historical narratives constitute exercises of moral reasoning rather than ways of preparing for such exercises, so Cosmic Connections is best understood as an extended exercise in philosophising rather than a companion volume to such exercises. It is a book of philosophy, not an episodically entertaining but fundamentally amateur excursus into literary criticism, and it should be judged as such – whatever Taylor himself might say about it. In this respect the work transcends its author’s interpretation of it, and so of himself. And although appreciating that fact doesn’t render it immune to criticism, failing to appreciate it guarantees that the criticisms it will inevitably attract will miss their target.