“Late Autumn’s Amaranth”: Coleridge’s Late Poems

Not to correct others except under provocation; not to admonish or reprove but to explore and illuminate: this, I take it, is my duty as a scholar.  I shall refrain therefore from pointing out, with ironic quotations, that nobody has ever before dealt at large with the subject of Coleridge’s late poems, or that if they have they have made a muddle of it.  Yet as far as I know, there has been no full-scale discussion of these poems and of their place in the development of Coleridge’s poetic art.  We pay respect to the canonical poems, and even sometimes read them; but we neglect the rest.  This is no Coleridgean way of proceeding.  Coleridge said, giving a history of the development of his mind: “I followed Socrates’ Adage respecting Heraclitus – All I understand so far is excellent; and I am bound to presume that the rest is at least worth the trouble of trying whether it be not equally so.”[1]

To begin with, I should like to lay out a few landmarks, biographical, chronological, and poetical, if only to determine after what date a Coleridge poem may justly be called “late.”  In March 1801, Coleridge wrote to William Godwin to say: “The Poet is dead in me – my imagination … lies, like a Cold Snuff on the circular Rim of a Brass Candle-stick, without even a stink of Tallow to remind you that it was once cloathed & mitred with Flame.”[2]  He said the same sort of thing to Southey and others at that time, representing himself as “weak – apt to faint away inwardly, self-deserted & bereft of the confidence in my own powers.”[3]  This same view of himself he repeated in an image taken from Swift, who had spoken of himself prophetically as a pine-tree blasted at the top: “the Chestnut [is] a fine shewy tree and its wood excellent but that, alas! it dies away at the heart first = poor me!”[4]  This note was written in October-November 1806 after the return from Malta, but it endorses what had been haunting his mind for some time.  In October 1800 Wordsworth had removed Christabel (still unfinished) from the proofs of Lyrical Ballads; the necessary changes in the Preface hid the wound from the world but did nothing to assuage Coleridge’s growing self-distrust.  His self-confidence, in no robust state of health anyway, was being sapped and paralysed by the burgeoning of Wordsworth’s genius in the months approaching his marriage to Mary Hutchinson; and in any case Coleridge’s own love for Mary’s sister, Sara, seemed unlikely, in his married state, to be fulfilled.  On the night of 4 April 1802 Coleridge wrote – evidently at a single impulse – the 340-line poem addressed to Sara Hutchinson, now known as the Letter to Asra.  This poem, purged of its explicit and most painful personal references, became Dejection: An Ode.  First read in its original form to the Wordsworths on 20 April and sent to Sara Hutchinson on 6 May, the poem was published on 4 October 1802 to celebrate Wordsworth’s wedding day.  In strophe VI certain lines that are essential to both versions tell how the weight of his afflictions

Suspends what nature gave me at my birth,

My shaping spirit of Imagination. 

For not to think of what I needs must feel,

But to be still and patient, all I can;

And haply by abstruse research to steal

From my own nature all the natural man –

This was my sole resource, my only plan:

Till that which suits a part infects the whole,

And now is almost grown the habit of my soul.

Because of these lines and because of much other contemporary biographical matter – the frequent repetition of this passage in his letters together with many sad self-accusations, his sudden preoccupation with German philosophy and with psychological investigations, and other apparent fulfilments of the allegorical and prophetic killing of the albatross of his poetic imagination – there has been a tendency to see the Dejection ode as marking the end of Coleridge’s poetic career, and to say that after this he turned to arid philosophizings and unrewarding theological speculation.  The direct evidence of the poem itself is otherwise.  Wordsworth’s companion piece, the Immortality ode, rationalized the loss of poetic vigour into a desolate stoicism not devoid of nobility.  Coleridge’s ode is a drama of renunciation, which proceeds to the nadir of apathy and turns outwards in the fruitful resolution of self-neglect, in a gesture of prayer for another person.  In its form and rhythms, below the reach of thought, will, or desire, this is the form the poem traces, showing that in fact the poet was not dead.  Poets are vulnerable, being born not made; but a poet is not easily silenced if he is as richly endowed as Coleridge was.  Dejection occupies pages 362-68 of Ernest Hartley Coleridge’s Oxford edition of the first volume of the Poetical Works.  The volume continues in its chronological arrangement to end with the Epitaph on page 492.  In the interval Coleridge seems not to have been silent.

But what had happened in the fifteen years between Dejection in 1802 and the first collective volume in 1817 – Sibylline Leaves?  Coleridge had sailed to Malta in April 1804 and returned ill and shamefully changed in August 1806, bringing no completed poems with him.  But now that all the surviving Mediterranean notebooks are in print we can see that Coleridge himself did not assume that the poet was altogether dead in him.  And if we look carefully, backward rather than forward, through the records of the months after his return from Malta ­– the long half-bemused delay in London, the hesitation to meet the Wordsworths and even Sara Hutchinson, the painful plans for separating from his wife, the Christmas meeting at Coleorton – we find nearly four years after the Dejection ode the true dividing point at which the river of Coleridge’s pristine poetic impulse runs darkly underground – the point at which the “late poems” begin.

Late in December 1806, Coleridge with his eldest son Hartley joined the Wordsworths at the Beaumonts’ Leicestershire farmhouse, Coleorton Hall.  William and Dorothy, Mary and Sara were together to celebrate their friend’s return from the dead.  Wordsworth had completed his long poem addressed to Coleridge, “concerning the growth and history of his own mind,” the philosophical poem that had been fostered, perhaps initiated, by Coleridge.  Wordsworth was also already preparing for the press his Poems in Two Volumes, his first collection to follow the four editions of Lyrical Ballads.  Here at Coleorton Coleridge tried to pick up his poetic work where he had left off: looking through The Ancient Mariner (did he start the prose gloss then?), working over Christabel to try to set in motion a life already twice interrupted.  He drafted, in blank verse, some lines that might almost have been the starting-point for the long poem Wordsworth started at that time – The White Doe of Rylstone.[5]  As he worked he “felt that sort of stirring warmth about the Heart, which is with me the robe of incarnation of my genius, such as it is.”[6]  But something went seriously wrong at Coleorton, something disastrous to do with Coleridge and Sara Hutchinson and Wordsworth: we shall never know exactly what.  Coleridge drifted away, irresolute and broken-spirited.  Although he was not yet able to tell himself this in plain words that his heart could grasp, the nourishing and tormenting bond with the Wordsworths – with William and Dorothy and Sara – was broken.  For some three years, sometimes alone or wandering, sometimes living with the Wordsworths in their new house in Grasmere, Coleridge hovered in the resuscitated penumbra of their early friendship; but an openbreak was inevitable.  With a prescience surpassed only by his prophetic instinct in The Ancient Mariner, Coleridge had already written his threnody upon that loss, while he was still at Coleorton, in the long poem To William Wordsworth composed (as the title tells us) “for the greater part on the Night, on which he [Wordsworth] finished the recitation of his Poem (in thirteen Books) concerning the growth and history of his own Mind.”  That poem stands as a massive tribute to Wordsworth’s genius and to Coleridge’s appreciation of it, to the friendship between them and the interweaving of their poetic genius; it is also a monument raised to the sorrowful unfulfilment of Coleridge’s poetic promise and endowments.

Sir Edmund Chambers, for whose biography of Coleridge (somebody has noted) all depreciators of Coleridge will be eternally indebted, states that the lines To William Wordsworth represent Coleridge’s “last poem of any substantial importance.”[7]  Certainly that is the traditional view, even among witnesses more sympathetic than Chambers.  My intention in looking at the late poems is not primarily to assess or establish their value, although one cannot renounce that incidental responsibility – clearly we are not dealing with anything like Yeats’s Last Poems.  I should like to see whether, by looking carefully at those poems, we can know something about Coleridge’s art and his view of his art that we should not otherwise know.  It should be said emphatically at once that the common view of Coleridge as a poet who squandered and neglected his poetic faculties and took to philosophy and theology faute de mieux is a false and ill-informed view.  Coleridge had been fascinated by philosophy and theology almost from childhood.  As his precocity settled into mature interest and concern, the philosophy and theology were still important to him though changed.  His poetic imagination, specialized at times in the production of verse, commanded and unified the whole ambience of his mind; contrary to what a few lines in Dejection seem to say, philosophy and theology, being radical to his thought, were correlate to, and not hostile to, poetry.  What Coleridge was in the end, he was from the beginning; what happened to him, as he unconsciously knew when he wrote The Ancient Mariner, had to happen to him.  The poetry he made is the poetry of Coleridge, a profoundly different affair, let us say, from the poetry of Wordsworth.  I should not like to see the understanding and enjoyment of Coleridge’s poetry diminished or deflected by the blur of generalizations about Romanticism.

*

As long ago as 1950, Dr. I. A. Richards made an unprecedented remark about Coleridge’s poetry.

His poetry succeeds in more modes than most poets have attempted.  In several, moreover, it is highly distinctive.  No one else has anything like Kubla Khan, anything like Phantom or Fact, anything like Constancy to an Ideal Object.  As to amount, it seems more impressive now than in the years when it was being measured against the output levels of Tennyson, Browning, or Swinburne.  The general impression that he dried up comes in part from his own self-upbraidings, in part from his readers’ refusal to let any but Christabels count.  In spite of all the celebrated weakness of his personality something in him kept him from imitating his own poetry.  Few of his contemporaries or of their descendants showed equal integrity; the temptation to which this will-less man could not yield overcame them.[8]

I should like to start all over again from that point.

After the Coleorton dissolution, Coleridge settled for a time with the Wordsworths in Grasmere to write his periodical The Friend in 1809 and 1810.  There is little poetry from the years 1807-10: seven poems and about fifteen fragments.  In late 1810 the open break with the Wordsworths occurred, and the years 1811-14 become dark, almost unrecorded; but about ten poems survive (of which three are revisions of earlier work) and half a dozen fragments, beyond which he had reissued The Friend in 1812, and reworked Osorio into Remorse in 1813.  In 1814 Coleridge’s friends rescued him from inanition and set him to work collecting his poems and writing what was to become the Biographia Literaria.  In April 1816 Coleridge took up residence in Highgate with a surgeon named James Gillman and was to remain there for the rest of his life.  The impetus of the Bristol beginnings continued through the early Highgate years to make that time his most productive in publication: a Christmas play in verse called Zapolya, Sibylline Leaves (a collection of his poems up to 1815) and the Biographia, the two Lay Sermons, the Christabel volume, and the expanded version of The Friend – all this by 1818; at this time he gave two extensive sets of lectures, one of which has been reconstructed into the Philosophical Lectures, and wrote his characteristic prefatory study of Method for the Encyclopaedia Metropolitana.  With all this going on, not many poems got written.  The year 1811 had been a lecturing year and was moderately productive for poems; so was 1815 and 1817: but the score for each year is only five or six poems.  The rewriting of Osorio and the composition of Zapolya came before 1816: these represent an impressive sustained poetic effort which makes them of more than archaeological interest.  After 1818 there was not much prose work to be published though much was in progress – Aids to Reflection in 1825, and Church and State in 1830; but there was no marked increase in poetic composition.  The Poetical Works in three volumes (including dramas) came out in a small edition in 1828, enlarged in 1829.  The final edition of Poetical Works, in which Henry Nelson Coleridge had an important hand, was published in 1834 shortly before Coleridge died.  Only four of the late years show a moderate output (five or six poems a year): 1825, 1826, 1828, 1830.  Seven years show one or two poems each and a scatter of fragments: 1820, 1823, 1824, 1827, 1829, 1832, 1833.  For seven years we have no poems and only two fragments: 1816 (fragments), 1818, 1819, 1821, 1822, 1831, 1834.

An exact count, however, is scarcely possible yet.  The standard scholarly edition of the poems, edited by Ernest Hartley Coleridge (2 vols., Oxford, 1912), is now in many respects out of date.  But the poems are arranged chronologically, as far as E.H.C. could determine the dates; and the cumulative if not always methodical critical apparatus helps to trace successive publication and textual changes.  E.H.C. has separated out from the main collection of poems three groups of occasional poems and has placed them in Vol. II under the headings of “Epigrams,” “Jeux d’Esprit,” and “Fragments” – more than 160 short pieces (not counting the forty-nine verse lines and fragments he took from the Gutch Memorandum Book), a great proportion of these being quite early in date.  Over the years, excessive zeal and affection have drawn into the canon some poems and fragments that are not Coleridge’s; and some of E.H.C.’s conjectural dating of later material can now with confidence be revised.  This removes several of E.H.C.’s “late poems” into an earlier period.  Again, the number of poetic fragments can only be approximate until the Notebooks are completely published.  But for convenience in trying to establish something like “a known and familiar landscape” among the late poems, I have worked from the Oxford edition, excluding from attention only those poems or fragments which I know definitely to be wrongly ascribed to Coleridge or of date earlier than “late.”

*

The poems, other than juvenilia, fall into three clearly defined periods: from Lyrical Ballads to Dejection (1797-1802), from Dejection to To William Wordsworth (April 1802 to January 1807), and from January 1807 to 1834.

                                       LB to            Dejection to                              To WW

                                    Dejection            To WW        (sub-total)       to Epitaph

            Titles                    69                       25                  (94)                  63

            Pages                  200                       40                (240)                  83

Or to put it another way: the late period produced a third as many pages of verse as the whole early period 1797-1807, and two-thirds as many titles.  The late period produced only slightly fewer titles than the first period, but only two-fifths the number of pages.  Whichever way we look at the figures, the last twenty-seven years show a much smaller production of verse than the central ten years 1797-1807; and the five years that end with Dejection produced half as much verse as the last twenty-seven years.  All this confirms, on other evidence, that January 1807 marks the watershed between early and late poems.

The question of numbers and distribution arises from two considerations: did Coleridge himself continue to regard himself as a poet after 1807, and what are we to take as “poems”?

(1) As far as I can discover, Coleridge said very little in a practitioner’s way about his late poems.  Yet if he was a professional anything, he was a professional writer; and the record of what he himself decided to publish and where is itself an indication, no matter how tenuous, of his own attitude towards his own work.  (When we recall how pathetically little he ever earned by his writing we understand how little danger he ever was in of prostituting his genius for cash.) The pattern of publication across these periods is surprisingly uniform.  In the years 1799-1802 Coleridge published in newspapers a number of epigrams, mostly translations, usually as pot-boilers: these are mostly separated out of the main text by E.H.C.  But Coleridge had had wide enough reputation almost since his undergraduate days that he never seems to have had difficulty in placing his verse – especially the more occasional kinds – in periodicals: in the last years of his life his verse seems to have been in demand for annuals and commemorative volumes; for the same reason but under less formal circumstances he was sought out to write in young ladies’ autograph albums.  It is very usual for a Coleridge poem to have been first published in a periodical, and then to have been collected later into a larger volume.

Sibylline Leaves (1817, collected 1815) was the first collective edition followed by 1828, 1829, both in three volumes, and all three collections edited by Coleridge himself.  The 1834 collection is more difficult to decide about.  Henry Nelson Coleridge did most of the donkey work though he evidently consulted S.T.C. throughout; my own guess (contrary to E.H.C.’s) is that the 1834 collection is very much S.T.C.’s book, even though he was ill and perhaps knew that he was dying.  Nevertheless I think H.N.C. wanted to make the edition as complete as he could.  A few poems (such as Parliamentary Oscillators) which Coleridge removed from the collection after Sibylline Leaves have been put back; but once a poem was included in any of the collections it tended to stay, even though it might change its position in rearrangement of materials.  The 1834 edition took in nine hitherto unpublished poems from the 1797-1807 period; in addition, there were fourteen poems written since 1807 that had not been published before, and ten that had appeared before – most of them recently – only in periodicals.  On the other hand, a certain number of poems, some published in periodicals and some in manuscript remaining for posthumous publication, were not included in the collections, whether by design or neglect we cannot tell.

Recognizing the approximateness of the figures in the present state of the editions of Coleridge’s poems, a few conclusions may be attempted.  Posthumous editors have recovered, for the 1797-1807 period, only fifteen unpublished poems and twelve others that had appeared only in periodicals; and for the “late” period they have recovered five poems from periodicals, and only ten unpublished poems.  Of the sixty-three poems now collected for the “late” period, one-third (twenty-two) were first collected in the 1834 edition, and almost a fifth (twelve) have been collected from manuscripts or periodicals.  Or to put it another way, five-sixths of the surviving “late” poems were deliberately included by Coleridge in his last collected edition.  A larger proportion of poems (not counting epigrams) from the 1797-1807 period were left in manuscript or in periodicals: twenty-eight poems, or two-fifths of the total.  It may be argued that Coleridge was more fastidious in his selection in the earlier period; but it may also be pointed out that he did in fact authorize the inclusion in his last collected edition of a higher proportion of what he had written.

(2) Two recent selections of Coleridge poems confront us with the need to say what may be legitimately considered “a Coleridge poem.” Richards’s Viking Coleridge (1950), the first comprehensive selection to pay serious attention to poems after 1807, prints twenty-seven poems (plus two fragments and one epigram) from the “late” period, almost as many in number (though not in quantity) as (the thirty) from the “first” period ending with Dejection.  (For the interval between Dejection and To William Wordsworth he prints fourteen.)  John Beer’s Everyman edition of the Coleridge poems is much more inclusive, printing some sixty poems from the “late” period.  The Richards selection is almost without exception repeated by Beer; but whereas Richards includes only two fragments and one epigram, Beer includes eighteen fragments and one epigram.  The effect on the general tone and atmosphere of the “late” period is strongly modified simply by the presence of the fragments, even though only eight of the eighteen are “Asra” poems.

*

Coleridge’s canonical poetry – say, the poetry up to and including the lines To William Wordsworth written in January 1807 – can be seen to fall into four manners, each kind or mode tending to be distinctive in subject matter, tone, and conduct.

(a) Contemporary manner.  First there is a derivative semi-classical contemporary manner, usually indistinguishable from the mass of transitional late eighteenth-century verse to be found in the periodicals and poetical quartos of the time: it is ruminative and topographical, the decorative translation of sententious or sentimentalizing thought into verse, not intrinsically uninteresting or uncouth but lacking on the whole the whiplash clarity of real distinction.  This makes Coleridge’s early allegiance to Bowles understandable, and reflects (as the juvenilia of the more stubborn Wordsworth do not) Coleridge’s susceptible precocity.  The derivative contemporary manner accounts for the first ninety pages or so of the Oxford edition, but it began to evaporate quickly in 1795 (age 22-23), largely (I believe) under the influence of Charles Lamb.  It continues, if at all, only in some epigrams, translations, and album pieces – whenever the poetic energy is low or the motive perfunctory – and there are not many examples after 1807.  Wordsworth, after the decline of his original poetic impulse, reverted to the Augustan style he had cultivated as a young man, and practised it for many years with Parnassian skill and devotion in hope of a new dawn; for it had always provided the underpinning for his art.  Coleridge, however, did not return to his own early or juvenile poetic manner.  There was no reason why he should.

Another two manners may be considered together because of the illuminating distinction first drawn by Humphry House between the “Miltonizing” and “Cowperizing” strains in Coleridge’s poetry.

(b) Miltonizing manner.  The Miltonizing poems are political in impulse and matter, and at their best express (as Carl Woodring has well shown) a genuine and fervent political passion.  At school Coleridge had absorbed Milton’s writing sedulously enough, holding his work in respectful and affectionate regard; but Coleridge’s Miltonism was never as profound or as personal as Wordsworth’s.  Coleridge’s rhetorico-political poems – Religious Musings, for example, and Ode on the Departing Year, France: An Ode, and Fears in Solitude – never generate the Miltonic diapason resonance that Wordsworth had discovered quite early in the versified manifesto that was to preface The Excursion, or later in that very Coleridgean poem Tintern Abbey.  Perhaps Erasmus Darwin was too much in the background.  Coleridge’s rhetorical political manner did not survive the disappointment of his radical but unpartisan hopes for the regeneration of Europe.  The late political poems are few in number, and are more like The Devil’s Walk than Religious Musings: they are characteristically dry, quizzical, mocking rather than bitter, occasionally peppery but never either malignant or magniloquent.  (The translations of Hyman Hurwitz’s Hebrew odes commemorating two state occasions count as tokens of friendship rather than as self-generated political utterances.)

            (c) Cowperizing manner.  William Cowper’s blank verse, variably Miltonic in its own way – shy and mercurial, light-hearted when necessary, with a grave nobility when appropriate – was more influential to Coleridge than Milton’s and gave him the clue to what is probably his most distinctive and personal manner: the manner of the Conversation Poems, The Eolian Harp,

This Lime-Tree Bower, Frost at Midnight, The Nightingale.  Cowper’s writing, and Akenside’s, helped Wordsworth discover the informal prosaic blank verse that sustains the discursive sections of The Prelude.  But in Cowper’s writing Coleridge discovered something even more sensitive and personal, and from it shaped a manner that was intimate, internal, intrinsicated; supple enough in syntax, rhythm, and sound to embody the flickering interplay of reflection, feeling, and thought.  By the sheer intensity of its inner definition, Coleridge found in this manner the symbolic substance that can convert (as for Dante or Yeats) the inner life of an individual into a universal myth.  As though his own successes in this manner were not impressive enough, Coleridge produced from Wordsworth for the Lyrical Ballads volume of 1798 a poem that, with characteristic shiftings of emphasis and focus, almost matches Coleridge’s own conversation poems – Tintern Abbey.

(d) Symbolic manner.  The remaining canonical mode is, either by design or by pure achievement, symbolic and mythical: represented by The Ancient Mariner (which he never surpassed), Christabel (which he never finished), and Kubla Khan (which successfully continues to elude definition).  Despite a tradition of referring these three poems to “the ballad,” I prefer to call them symbolic.  Two of the poems, it is true, owe much to certain kinds of ballads; but the versification of Christabel is – like Hopkins’s sprung rhythm – a very special and individual music of mysterious paternity.  Among Coleridge’s papers there are relics of other ballads, some of them impressive in their way; but he never seriously tried to repeat the supreme achievement of The Ancient Mariner.  The gloss to The Ancient Mariner was finished in 1815 (all but one phrase): small revisions continued to be made to the poem itself until 1834.  That poem was not one of a family, but a single living expression that held the poet’s reflection in its magnetic field throughout the rest of his life.  Among the late poems in E.H.C.’s edition there is, apart from an undated fragment of a translation from Hölty, only one ballad-piece – Alice du Clos.  The sophistication of the verse suggests a late date; but even though the poem may have received late revision, it seems no more a late poem than Glycine’s Song from Zapolya which was substantially drafted, if not actually completed, almost ten years before Zapolya was written.  The ballad tradition, as Coleridge had it from the German and from Percy’s Reliques, suited him well for The Ancient Mariner; but too much can be made of the connexion.  In the ballad style, Coleridge failed more often than he succeeded, though one of the failures leaves us with the poem Love.

The three non-derivative manners, grafting themselves on to the habits of poetic management he had learned at Christ’s Hospital, began to emerge clearly at much the same time through 1795 when Coleridge was twenty-three, and were active from about 1796 to 1800.  By 1800, after the second part of Christabel had been written, only the conversation manner continued vigorously; and it had lost the un-self-conscious and reflective calm that it had encompassed so happily in the annus mirabilis.  The bleak rigour of Wordsworth’s lyrical ballads was transfigured after only a few months of use into the plain and cosmic impersonality of the “Lucy Poems” and a little later was brought to a wonderfully compact symbolic richness in The Solitary Reaper.  Coleridge’s ballads came from a different source than Wordsworth’s and did not suffer this transfiguring change.  But by a similar process of emotional modulation, the conversation manner, combining with the organic and rhapsodic tradition of the ode as it appears at its Pindaric highest in Kubla Khan, became the vehicle for almost all his most poignant expressions of delight, desolation, and self-inquiry.

The longest and most moving example of the modulated conversation poem is the Letter to Asra written in April 1802.  In its cut and slightly reorganized version, Dejection: An Ode, it is more like a formal ode than a conversation poem.  Yet if Dejection did not bear within itself the eloquent and intimate values evolved in the original version, that poem would not stand where it does in the canon of Coleridge’s poetry; nor would it have cross-fertilized, as it did, one of Wordsworth’s finest poems, the Immortality ode.  The same transformed conversation manner, in its most dignified and sorrowful kind, stands at the end of the “early” period and at the beginning of the “late,” in the poem To William Wordsworth.  Here Coleridge recalls the early halcyon days when Wordsworth’s Prelude was conceived; he has in his ear the actual sound of Wordsworth’s own reading hour by hour and the reverberant Miltonic rhythms alternating with a plainer movement: these carry Coleridge’s poem out of the ode form into a blank verse which is distinctly Coleridgean, yet coloured by respect, admiration, and grief, and by the groundswell of Wordsworth’s verse.

                                    Ah! as I listened with a heart forlorn,

                                    The pulses of my being beat anew:

                                    And even as Life returns upon the drowned,

                                    Life’s joy rekindling roused a throng of pains –

                                    Keen pangs of Love, awakening as a babe

                                    Turbulent, with an outcry in the heart;

                                    And fears self-willed, that shunned the eye of Hope;

                                    And Hope that scarce would know itself from Fear;

                                    Sense of past Youth, and Manhood come in vain,

                                    And Genius given, and Knowledge won in vain;

                                    And all which I had culled in wood-walks wild,

                                    And all which patient toil had reared, and all,

                                    Commune with thee had opened out – but flowers

                                    Strewed on my corse, and borne upon my bier,

                                    In the same coffin, for the self-same grave![9]

Coleridge’s personal poetry was not to reach this sustained grandeur and dignity again.  Yet the manner abides, providing (as it were) the armature or thread about which the later Asra poems crystallized, and all the most memorable and accessible late poems – Youth and Age, Work Without Hope, and the final Epitaph.

                                    All Nature seems at work.  Slugs leave their lair –

                                    The bees are stirring – birds are on the wing –

                                    And Winter slumbering in the open air,

                                    Wears on his smiling face a dream of Spring!

                                    And I the while, the sole unbusy thing,

                                    Nor honey make, nor pair, nor build, nor sing.

The singleness of Coleridge as a person, and the singleness of his intelligence and imagination, account for the many-faceted transformations that a mode of poetry can discover under his hand, without loss of identity, over the years.  But Coleridge as a practising poet was not very much at the mercy of accident.  Wordsworth himself gave witness to “the extreme care and labour which he [Coleridge] applied in elaborating his metres. … When he was intent on a new experiment in metre, the time and labour he bestowed were inconceivable.”[10]  For all the talk of “spontaneous overflow” Coleridge is a conscious and deliberate craftsman, though at times a remarkably fluent and flawless one.  He was very learned in the poetry of his own language, as a glance only at the three marked copies of Anderson’s British Poets will show, and his sets of Shakespeare.  He became fluent in the poetry of German and Italian; he could read Spanish if he wanted to; his Latin was excellent, and his taste uninhibited by the literary snobbery that will acknowledge nothing worth reading after Apuleius or Minucius Felix.  Above all he was a good Greek scholar, could grasp at first hand the exquisite and concrete versatility of Plato’s mind, and dwelt continually upon Greek poetry, from Homer, Hesiod, and the tragedians to the Greek Anthology. He also studied intermittently but closely that most difficult and eloquent of all poets, Pindar, and worked away as only a poet could at the unriddled intricacies of Pindar’s versification, plunging if need be into the Latin and German exposition of conflicting theories in the ponderous edition of Erasmus Schmied.[11]  It is interesting that Gray, who was also a Greek scholar, wrote The Bard as well as Elegy in a Country Church-Yard.  It was The Bard, not the Elegy, that inflamed Coleridge’s prophetic fervour.  Coleridge respected Pope for a skill no poet could ignore; but he distrusted Dr. Samuel Johnson.  Most of the Augustan and neoclassical tradition is unrelievedly Roman, Latin, Latinistic, lacking those very values and dis-criminations that make Plato and Pindar and the tragedians an inexhaustible fountain of imaginative life.  Coleridge may have turned away from the Augustan tradition he inherited in his youth; but in a more positive way he was affected by the Greek spirit that his predecessors so clearly lacked.

*

Turning to the late poems in the Oxford edition, we may for convenience neglect the dramatic writings – Remorse and Zapolya, and the rewriting of the fragmentary Triumph of Loyalty.  Let us neglect also the prose gloss of The Ancient Mariner because it is now widely and justly admired.  Let us neglect also the purely political poems like The Delinquent Travellers and Sancti Dominici Pallium if only because the important political poems – A Character and A Tombless Epitaph – are also important for other reasons.  Let us ignore the translations – from the Italian of Guarini, from Aristophanes, Pindar, and Heraclitus; from Claudian’s Latin, the German of Hölty, and the Hebrew of Human Hurwitz.  Let us ignore also the three or four album verses; and the few late epigrams that in their relaxed elegance are like album verses; and the few late epigrams that for their blunt humour look back towards the wanderjahre and a German sensibility.  There are then left among the late poems the following: a number of personal poems, some of which contain openly or by implication a portrait of Coleridge himself; a group of what I shall call “emblem-poems”; and some gnomic and conceited verses that for convenience I call “metaphysical.”  We notice also that there are some poems, often overlapping the other categories, which are avowedly allegorical, or deal in personifications, or are in the form of a dialogue.  We may now examine these new kinds in turn.

(e) Personal Poems.  Of the twenty-two personal poems (a couple of these are indeterminate), eleven are “Asra poems”: that is (as I have explained elsewhere[12]) they deal with the transports of love and with the sorrow of an unrequited and now hopeless love, and take shape around the clusters of images that had illuminated his early love for Asra and which now haunted and vivified his recollections of her; in these poems there is much mournful play upon the word Hope.  But not all the personal poems of late date are so sombre.  Reproof and Reply is a long and gay argument in the poet’s own defence on a charge of stealing flowers of a Sunday morning from a neighbour’s garden: it was to satisfy a poet’s delight, he pleads; it was a scholar’s absent-mindedness.  This deft and charming apology ends in the imagined words of acquittal uttered by a judge of whom the poet could say confidently, “There’s no spring-gun or man-trap in that face”:

                                    “The spoons all right? the hen and chickens safe?

                                    Well, well, he shall not forfeit our regards –

                                    The Eighth Commandment was not made for Bards!”

A sonnet To Nature (?1820), written two or three years earlier, tells with less vigour and less charm how Coleridge tries to “trace in leaves and flowers that round me lie / Lessons of love and earnest piety.”  (And certainly we know that his love of flowers brought much pleasure to him in his declining years for their own sake, and for the sake of the young ladies who sometimes, admiring, brought them to him.)  Duty Surviving Self-Love (1826) is a “soliloquy” in what looks like a free kind of sonnet form; it is gnomic but not metaphysical, counselling that although old friends change and “burn dim,” one should “Love them for what they are; nor love them less, / Because to thee they are not what they were.”

The Improvisatore (1827) opens with a long section of dialogue between two young ladies and their “Friend” – Coleridge, that is – on the subject of love and friendship.  The prose is flexible and colloquial, arguing much more strongly than any of his verse drama or his journalistic writing does that Coleridge could write the vigorous plain prose that Dryden or Swift would commend.  The seriousness of the topic, however, leads Coleridge into careful and difficult monologue, an example perhaps of “that exhaustive, cyclical mode of discoursing” which, as H.N.C. said, was for the patient enthusiast “the object of their profoundest admiration, as it was the source of their most valuable instruction.”[13]  The monologue leads in turn – “after a pause of a few minutes” – into an improvised poem almost seventy lines long in the free manner of an ode, the verse dominantly tetrametric.  Here the Hope-and-Love theme suggests an Asra poem.  But if by eye the poem looks a little like the Letter to Asra, it soon discloses itself as a kind of allegory in which Fancy, Hope, and Love are examined with increasing gravity and point.  Coleridge thought well enough of some of these lines to include them in a poem of congratulation addressed to Mary Pridham, his son Derwent’s bride.  It is in this poem that we find the reference to

                        Late autumn’s Amaranth, that more fragrant blows

                        When Passion’s flowers all fall or fade.

The rhythms of the poem are dense and unmellifluous, the movement and imagery most closely akin to the “emblem-poems,” of which indeed this is the most extended example.  Yet in its affectionately noted passages of natural detail, in its shafts of half-bitter playfulness, and its ingenious argumentation, the poem seems a little metaphysical too.  It is not an attractive poem but it may well be an important one.  And by some process of poetic logic, it closes on a note of desperate conviction:

                        Now that at length by certain proof he knows,

                        That whether real or a magic show,

                        What’er it was, it is no longer so;

                        Though heart be lonesome, Hope laid low,

                        Yet, Lady! deem him not unblest:

                        The certainty that struck Hope dead,

                        Hath left Contentment in her stead:

                           And that is next to Best!

Some of the personal poems embrace portraits of Coleridge.  These have a special interest because they discover through time an increasing honesty.  A Tombless Epitaph belongs to 1809 and The Friend, and gives an account of himself as “Idoloclastes Satyrane” – perhaps a little overwritten, though that may be the fault of Chiabrera upon one of whose Italian Epitafi this is “imitated.”  “With a natural gladness,” he maintains “the citadel unconquered,” and is the “studious Poet, eloquent for truth” and Philosopher “contemning wealth and death” – “Yet docile, childlike, full of Life and Love!”  The picture is a bit self-conscious and unconvincing.  In A Character (1817-18), however, the citadel seems to have capitulated to merriment; for Coleridge begins with an account of his political vicissitudes, from the time when he was “A Bird, who for his other sins / Had liv’d amongst the Jacobins” until, for denouncing his associates (in the name of Truth, no doubt),

                                    … his old nest-mates chang’d their note

                                    To hireling, traitor, and turncoat, –

                                    A base apostate who had sold

                                    His very teeth and claws for gold; –

                                    And then his feathers! – sharp the jest –

                                    No doubt he feather’d well his nest!

He asserts again that he has not changed, that his name S.T.C. is a Greek pun for “He hath stood”; he complains of those who have stolen his ideas and claimed them for their own; and he ends by drawing attention to his poverty.  The poem has some joyous passages, but it is uneven and ambiguous, mixing a commendable pride with understandable self-pity.

Both these poems look towards a personal centre that eludes him.  In the Lines Suggested by the Last Words of Berengarius (1826) Coleridge picks up again the theme of heroic if futile constancy and finds himself at home with a man who must have reminded him of that old heterodox martyr Giordano Bruno, whose splendid and bragging verses in arrogant praise of the supremacy of genius Coleridge had copied in their intractable Latin into his notebook almost twenty-five years earlier.  The words of Berengarius look forward too to the quiet but complex simplicity of his own Epitaph.  “Lynx amid moles!” he cries; and the self-identification is strong and complete.  His version of Berengarius’ last words reads:

                                    No more ’twixt conscience staggering and the Pope

                                    Soon shall I now before my God appear,

                                    By him to be acquitted, as I hope;

                                    By him to be condemned, as I fear.

But how, he asks, could Berengarius be condemned for withdrawing from what can now be seen as only “a learned strife”?  Consider –

                                    That age how dark! congenial minds how rare!

                                    No host of friends with kindred zeal did burn!

                                    No throbbing hearts awaited his return! …

                                    And was it strange if he withdrew the ray

                                    That did but guide the night-birds to their prey?

When Coleridge first published his Berengarius poem, he added as a footnote a two-line epitaph in Latin for himself:

            Quae linquam, aut nihil, aut nihili, aut vix sunt mea.  Sordes

               Do Morti: reddo caetera, Christe! tibi.

“What I shall leave behind is either nothing, or to no purpose, or scarcely accountable to me.  My squalor I give to Death: the rest I return to thee, O Christ.”

The Garden of Boccaccio (1828) – written, as a surprising number of the late poems are, in five-footed couplets – is a sustained expression of his delight at Boccaccio’s literary landscape.  It was written apparently on request to accompany an engraving of Stothart’s – “An Idyll … framed in the silent poesy of form.”  The poem carries itself confidently and easily to more than fifty couplets, and gives a vivid glimpse of Coleridge in the last years of his life, in “one of those most weary hours, / When life seems emptied of all genial powers.”  The poem at its opening looks back towards Dejection with its echo of Samson Agonistes, and in the body of the poem the mantling pleasure, now mature and sophisticated, seems fanned and nourished by some breeze that blows with ominous poignancy “from the spice-islands of Youth and Hope.”[14]  Five years later the trajectory that passed through Chiabrera, Berengarius, and Boccaccio, was to touch on some words of Spencer’s –

                        Gently I took that which ungently came,

                        And without scorn forgave: ­– Do thou the same. –

and came to rest in the last lines of his own Epitaph.

                        Mercy for praise – to be forgiven for fame

                        He ask’d, and hoped, through Christ.  Do thou the same!

In the Asra poems we often catch the voice of a blank and terrible regret, as though Coleridge in his desolation cherished memories the more precious for their fragility.  The tone is usually soft, gentle, inward, like the quietest and most felicitous passages of the early conversation poems.  Among the later personal poems, other sounds are to be heard; for, even at the end, there are other themes for him than regret and an intolerable sense of loss.  Then his words have more often than not the peculiar intransigence of things that have a life of their own.

(f) Emblem Poems.  One group of poems I have called “Emblem Poems” because of the heraldic isolation and the heraldic specificity of their central images.  Technically these are a little like anticipations of the imagiste technique, though their implications are usually less oblique and evocative than an Imagist would like; and their intensity is much less reverberant than such a powerful organizing symbol as Coleridge’s own albatross.  Here is Psyche, an early example, if not the earliest, to which Coleridge adds a note explaining that in Greek the word ψύχη means both soul and butterfly.

                        The butterfly the ancient Grecians made

                        The soul’s fair emblem, and its only name –

                        But of the soul, escaped the slavish trade

                        Of mortal life! – For in this earthly frame

                        Ours is the reptile’s lot, much toil, much blame,

                        Manifold motions making little speed,

                        And to deform and kill the things whereon we feed.

This poem has Asra connexions, because it was copied into the manuscript called Sara Hutchinson’s Poets; and so was another emblem poem called Time Real and Imaginary – a poem which Coleridge said was “a school boy poem,” though surely touched up and perhaps expanded before part of it was printed in Sibylline Leaves.

The emblem poems seem to have had a definite heraldic or iconographic beginning in Coleridge’s commission to Mathilda Betham in 1807-1808 to make a seal for him.  We have the actual commission in a letter.

Pray would it be possible to draw the following figures for a seal.  In the centre (as a coat-of-arms), a rose or myrtle in blossom, on the right hand, a genius (or genie) holding in the right hand two torches inverted, and one at least recently extinguished; on the other side, a Love with a flaring torch and head averted, the torch in the direction of the head, as one gazing after something going away.  In the corner of the left part of the composition a large butterfly flying off; the motto under it, “Che sará sará” – What will be, will be.[15]

We also find him in his notebook once ruefully writing out an emblem of himself as “The moulting Peacock with only two of his long tail feathers remaining, & those sadly in tatters, yet proudly as ever spreading out his ruined fan in the Sun & Breeze –.”[16]

But the emblem goes farther back than that for Coleridge: to the pictures in Tooke’s Pantheon which he and Lamb studied at Christ’s Hospital and never forgot; to Quarles’s Emblems, introduced to him perhaps by Lamb and known to Coleridge at least by 1808; and to his personal discovery in Malta and Sicily of the truth of what had before been a literary tag, picturae libri pauperum.

The emblem poems are marked by restraint from discourse or reflection, and by a strong gnomic or moral emphasis.  The emblem quality is seen in combination with other manners in some longer poems: in the Berengarius poem, in The Pang More Sharp, The Two Founts, The Improvisatore, and The Garden of Boccaccio.  One or two emblem-like poems are difficult to classify: Love’s Blindness and First Advent of Love.  But others, beyond Psyche and Time Real and Imaginary, are clear enough: the song “Tho’ veiled in spires,” Love’s Burial-place (a little mocking), Love, Hope, and Patience in Education, Not at Home, Phantom or Fact (a dialogue), and Love’s Apparition and Evanishment (though this last has attached to it an “Envoy” which is an outstanding example of a short Asra poem).  Some of the emblem poems are also avowedly allegorical: The Two Founts, The Pang More Sharp, and Love’s Apparition.

(g) Metaphysical Poems.  It is difficult sometimes to separate the “metaphysical” poems from (on the one hand) the epigrams, and (on the other) the emblems.  I have taken as metaphysical criteria, ingenuity of metaphor, violence of conceit, and the tendency to conduct argument through a train of images rather than by logical sequence.  For example, both the following examples seem to me to count as metaphysical conceits rather than simply epigrams or emblems.

                           Let Eagle bid the Tortoise sunward soar –

                           As vainly Strength speaks to a broken Mind.

                        With Donne, whose muse on dromedary trots,

                        Wreathe iron pokers into true-love knots;

                        Rhyme’s sturdy cripple, fancy’s maze and clue,

                        Wit’s forge and fire-blast, meaning’s press and screw.

The sonnet Human Life, though written “in imitation of Donne,” seems to me no true metaphysical poem.  But some of the most highly prized of the “late” intractable poems are metaphysical: Limbo (a surrealist nightmare arrested by horror and commanded by a very modern dread of nothingness), Ne Plus Ultra (a small unlyrical Pindaric Ode), Constancy to an Ideal Object (personal, haunted by memories of past life and echoes of distant poems), Reason, Self-knowledge, Forbearance (with its echo of The Pains of Sleep in the “keen sense of wrong,” of Christabel in the crime of “an alienated friend,” and of The Ancient Mariner in the loathesome image of a ship rotten in its timbers), and finally My Baptismal Birth-Day (a sonnet repeating Donne’s theme of the death of Death).  None of these poems could be mistaken for a poem by Donne or Cowley; nor is there, as elsewhere in Coleridge’s work, any savour of George Herbert.  Yet I call them metaphysical because of their harsh and impacted imagery, their nagging and painful emotional intensity, their rough but eloquent rhythms, and their resolute argumentativeness.

(h) Fragments.  We may deplore the fact that any poem should be left unfinished, or that if it be left incomplete it be regarded as a poem.  But since we wish to find out what Coleridge intended to do or thought he was doing in his later poetry, and since the evidence is not copious, I think we must take carefully into account the witness of the fragments.  Coleridge himself encourages this decision.  He himself published a few fragments incidentally in his own prose works and in his collected poems.  Others he published more ceremoniously: Christabel is such a fragment, and until recently most people regarded Kubla Khan as a fragment too.

The poetic fragments, as collected, identified, and dated by E.H.C., are assigned to fourteen of the “late” years.  These confirm the pattern I have just traced out: the decline of the Asra poems and of the established “conversation” manner, the transformation of the personal manner, and the rise of the emblem poem and the metaphysical poem.  Most of the Asra fragments are before 1807.  But there are at least three Asra fragments for 1808-10 and one of these – “I have experienced the worst the world can wreak on me” (1810) – has to be considered as either metaphysical or emblematic.  There is one metaphysical fragment for 1820, and another undated.  Otherwise the fragments are epigrams, or translations or sketches of natural detail.  The natural detail is a kind of material very common throughout the early period, in notebooks and as variants and drafts of poems.  In the late period this is less commonly found: two “nature sketches” in 1807 are on their way to being emblems, and one for 1811 (on the Pacific waves) is a metrical experiment.  But two others are interesting for reasons neither descriptive nor topographical.  The undated Fragment 55, an image of a harpist singing, has a dreamlike, almost obsessive brilliance.

                                    Or while she warbles with bright eyes upraised,

                                    Her fingers shoot like streams of silver light

                                    Amid the golden haze of thrilling strings.

The other, Fragment 49 dated June 1828 and sketched as we know in the field, is like a Brueghel landscape to the eye, and to my ear has premonitions of what was to happen in Eliot’s Marina.

                                    Water and windmills, greenness, Islets green; –

                                    Willows whose Trunks beside the shadows stood

                                    Of their own higher half, and willowy swamp: –

                                    Farmhouses that at anchor seem’d ­– in the inland sky

                                    The fog-transfixing Spires –

                                    Water, wide water, greenness and green banks,

                                    And water seen –

To trace out the distinctions and criteria upon which I have based the three categories of late poems would lead us more extensively into the text than time will permit or patience will now accept.  It should be pointed out that, particularly when the number of examples is relatively small, the categories may at first seem tenuous.  We are lucky to have so clear-cut an example of an emblem poem as Psyche.  Some of the poems embrace a sequence of manners: epigram – conceit – metaphysical; or emblem – epigram – personal; defying classification.  But the poems are more important than the categories, and these transformations will delight, rather than dismay, the affectionate reader.

*

What then are our inferences and conclusions?  As for the personal poems, these prove to be wider in technical and intellectual scope than we may have expected from the evidence of some of the retrospective and self-pitying love poems.  Looking among the poems not associated with Asra, we find an increasingly powerful undercurrent of impersonality, as though Coleridge were not only finding out how to look at himself “from outside,” but were also evolving the way of standing away from his self-knowledge and render­ing it in words.

The other two kinds bring even more distinct information.  The emblem poems, the metaphysical poems, the allegories and little dialogues – all these in their various ways point in one direction.  In his early poetry, to the end of the annus mirabilis and later to 1802 and Dejection, Coleridge had dis­covered and used with increasing success a method of obliquity which could externalize the inner life, and found for this purpose poetic symbols of great power in the vivid accidents of the actual world (whether real or imagined).  This method could “distance” experience in a number of different modes (most of which are used in conjunction in The Ancient Mariner): in time, in locale, in social context, in language.  Coleridge’s triumphs of obliquity – The Ancient Mariner and Kubla Khan particularly – are triumphs of sym­bolic precision and intensity, and were made possible by his studious attention to the outer world.  Paradoxically (because true poetry is oblique and struc­turally paradoxical) he had unfolded the inner world by discovering the outer world to be an infinite universe of symbolic forms.  At first with only half-conscious understanding, he had discovered that poetic symbols always partake of the reality which they make intelligible.  But like any artistic solution, the method could be workable only for a time: it had to be carried further if it was not to atrophy.  The conclusion I draw from the late poems is that Coleridge went on developing as a poet, that he knew what he was doing and why, and that much of what he wrote in the later period is not only not a repetition of what he had written before but a defiance of it.  He had to defy what he had done before or else write parodies of himself.

Early on, and perhaps through Wordsworth’s influence, Coleridge took nature to be capacious, broad-breasted, all-provident.  In the outer world he could find all his imagery; the sky and the whole world was alphabeted with symbols, written in a language he was often on the verge of being able to translate.  As he was seized upon by one image or another – bird, leaf, tree, moon, a brook running, the cock-crow, the lark’s song across a hill, the spider’s web – each took on a life of absorbing vitality and meaning, capable of generating new life and insight, and of forming into fascinating and fertile clusters as though in a dream.  For a time these images, taken up into the mind and nourished in the subconscious, could form a matrix which was at once the pattern of vision and the pattern of the world: under the condition of Joy this could happen best.

By the date of Dejection, the imagery of the natural world had begun to lose its power and meaning in his poetry.  Even though his concern had always been to render the activities of the mind, to clarify the inner life of feeling and thought, he had with almost magical sure-footedness discovered in the natural world the symbols he needed.  Once the relation between himself and his central images had changed, he would have to discover new modes of “distancing” if he were to avoid the cul-de-sac of description and establish new centres of poetic energy.  In the same way that in the beginning of his best work, archaic language and exotic settings had released his mind from the tyranny of naturalistic accident and plausibility, he turned later away from the this-world-ness of his early successful poetry towards a more literary, intellectual, even abstract framework.  Perhaps even the later Asra poems may be anachronisms, flowing on, through the tenacious possessiveness of desolation, into a poetic world where they are no longer at home.

In turning from the direct psychological revelation of the conversation poems to the metaphysical rugosities of a poem like Limbo, Coleridge dis­covers a vehicle for a strange mood in a distant mode.  The strength of the poetic need demands extreme projection towards what is strongly stylized, even grotesque in its formality, arresting and violent in its imagery, tone, and rhythm.  He turns now to relatively crude heraldic emblems, not as a symptom that his imaginative energy has fallen below the symbolic level of activity, but in order to deliver the same meditative-reflective assault on the mind that an emblem book makes: he consciously discovers metaphor as a mental process rather than as a figure of speech.  Personification is now no uncritical literary inheritance – “images imageless, small-capitals constituting themselves Personifications,” as he noted scornfully in the margins of Joan of Arc – but a deliberately chosen mode of stylization.  Metaphysical strenuousness suits the agonizing desperation of his thought now; allegory and dialogue will not allow him to be beguiled back from the impersonality he has striven so hard to achieve.  The resources he draws upon are no longer plausibility, narrative, smooth modulation and transition, but sharp meta­phorical parataxis, abrupt confrontation.  He uses the tough inventive rhythms that spring from the intransigence of language and the shape of passionate utterance: he pays no tribute to mere music.  This is the way of a man who looks with bleak and trenchant gaze at himself and at a world that had stopped being fun a long time ago.

In the emblem poems a new formality combines with a severe sense of deliberate limitation.  The device is as much tactile as visual: it is to do with cutting seals in hard small stones (as Matilda Betham did), or engraving with sharp steel (as Bewick did for his quadrupeds, fishes, and vignettes) in small blocks of boxwood.  Several things begin to happen.  The “descriptive” method is abandoned together with the historical and logical structures of anecdote and argument.  As in any movement from a naturalistic mode to a symbolic mode, he explores the more harsh and abrupt resources of poetry: allegory, dialogue in vacuo, bold collisions of any sort, personification as an active rather than a decorative principle; he has a tendency to avoid the uneconomical formalities of an Aristotelean beginning, middle, and end.  At first sight it looks as though he is turning back rather than going forward, becoming fragmentary rather than coherent; but this is also the way forward that Joyce took, and Kafka and Pound.  Coleridge’s doctrine of dramatic illusion has come to its logical conclusion: the poem, like a dream, imposes its own connexions, with the least possible appeal to the authority of any structure outside the dream.  He must release his images and symbols from the accidental connotations of the natural world.  He must not be beguiled by those famous harmonies which in the past had bound firmly together all the elements of his most successful verse.  He must overcome the temptation to imitate himself.  He must start afresh.  With that broken will of his, he often had to start afresh.

Under the pressure of these changes, the personal manner, the mode of the conversation poems, continued longest to provide a bridge between the early and the late poems, becoming the rallying-point for widespread changes.  His poetic technique, which in any case was never accidental or injudicious, becomes more learned and recondite, the ear less facile; but the effects are conscious, being the poetic technique of a deliberate and highly endowed artist.  In choosing verse forms he makes no concession to his virtuosity.  He abandons blank verse almost entirely and turns often to couplets.  He does not embark upon intricate stanzaic forms nor does he show any sign of an interest in mere artistry.  He no longer experiments.  Rather he seems to withdraw within the confines of a few plain and familiar forms, as though by suffering the discipline of an ascetic plainness he would come upon the sanctity of unadorned clarity.  Being an art of renunciation, it is a courageous art.

Coleridge in The Ancient Mariner was, in the strict philosophic sense, metaphysical: he was exploring the nature of being and knowing.  “Without Metaphysics Science could have had no language and Common Sense no materials,” he was to say in Aids to Reflection.  Wordsworth in Peter Bell, the companion piece to The Ancient Mariner, gave a clinical account of a man smitten, to the point of hallucination, by guilt.  When Wordsworth’s crisis came, he did what he had done in the Immortality ode: he turned back.  Coleridge, when he came to his crisis in 1807, did what he had done in the Dejection ode: he turned inward and then with the weary certainty of desolation he moved forward; and though there were few seamarks to guide him, and no contemporary support of perceptive readers to encourage him, and little enough steady resolution in himself, he went on developing.  It is a pity, as Wordsworth said, that Coleridge did not write more poetry in those last years, and that he did not complete more of what is now frag­mentary.  What he was looking for – as many uncanny notes, drafts, titles, and hints written in the Notebooks show – was what Yeats, Eliot, Pound, and Joyce were to discover almost a hundred years later.  The late poems seem to me to give clear if intermittent evidence of this.