The Divine Ideas in Coleridge’s Opus Maximum:

The Rhetoric of the Indemonstrable

Murray J. Evans

(Coleridge Bulletin, New Series 22, Winter 2003, pp39-47)



In the course of his exposition of the Trinity late in the Victoria College Library MS Say 29 of the Opus Maximum, Coleridge cryptically mentions his notion of the divine ideas or ‘living, uncreated truths,… derived from the transcendent unity of God’.[1]  Elsewhere, on Coleridge’s preferred formulation of the Trinity as Tetractys—Absolute Will, Father, Son, and Spirit—Coleridge includes a fifth category, again ‘the Eternal IDEAS’.[2]  Happily for our understanding of divine ideas, one whole manuscript of the extant Opus Maximum (OM), Huntington Library MS HM 8195, entitled ‘On the Divine Ideas’ (ODI) and likely dated 1822-3 (OM 214), is devoted to the topic. The little scholarship on this manuscript sticks to its main lines or cites it selectively for individual purposes.[3]  Much of the substance of ODI remains unaddressed.  The same holds true, moreover, for the rhetoric of its arguments.  These gaps I mean to address by focusing on one such neglected passage, the last section of ODI, which explores the status of one of the divine ideas typical of the rest, the idea of God (264-90, 223).  Of particular note in this passage is its rhetoric, as it relates to Coleridge’s insistence on the doubleness of divine ideas, and also their unprovability.  To argue for the indemonstrable seems fruitless, but Coleridge’s rhetoric is much more satisfactory here than in some other discussions of ‘idea’ elsewhere in his work.  The last, puzzling ‘Essay on Method’ in The Friend is a case in point.  In reference to it, then, my essay will also show that ODI can clarify and expand our understanding of ‘idea’ in Coleridge’s more widely discussed, earlier work.  I will conclude by revisiting the issue of whether or not there is system in Coleridge’s thought.

       The kernel of what a divine idea is, is relatively clear in ODI.  Generally speaking, a divine idea is ‘a representation of the universal under the eminence of some form in particular’ (OM 236).  As Coleridge puts it elsewhere, divine ideas are ‘God’s ideas of finite things, the finite things—which originate in him but acquire separate existence’.[4]  Perkins helpfully contrasts Coleridge’s ‘idea’ with his ‘symbol’: whereas ‘the symbol reveals the universal in the particular,




 the Idea is a particular form within the universal, the Absolute’ (174). While divine ideas are not clearly angels,[5] Coleridge stipulates that we may call them ‘living spiritual and substantive in the eternal idea,… uncreated forms and eternal truths, powers, and intelligences’, speaking ‘not only truly but scripturally’ (OM 233).  Beyond these starting-points, however, there arises the problem of the doubleness of a divine idea, which Coleridge himself forecasts in his search for ‘an intelligible though not comprehensible Idea of the possibility of that which in some sense or other is, yet is not God nor One with God’ (216). The doubleness is itself twofold, both in the subject and in the object.  As for the first, we are to find the idea ‘intelligible though not comprehensible’, a familiar Coleridgean distinction between having some intellectual grasp versus an encircling understanding (cf. OM 98, 211, 216).  That a divine idea, objectively speaking, also possesses a ‘potential duplicity of being’ (OM 246)—‘in some sense or other is, yet is not God’ (216)—derives from its complex relationship to the Godhead.  On the one hand, ‘in the absolute Will, which abideth in the Father, the Word and the Spirit, totally and absolutely in each, one and the same in all, the ground of all reality is contained, even of that which is only possible and conditionally possible alone’.  But the Absolute Will self-realized, i.e. the Father, Son, and Spirit, must include in itself only the ‘absolutely real, that is, as far as the reality is actual and not merely possible’.  The ‘possible… or potential, as contradistinguished from the actual… cannot be in God’ – as the Trinity (221-22).  So, depending on vantage point, divine ideas share variously in the life of the divine tetractys.  Each idea is real to God, since ‘it is the necessary offspring of a power, the essence of which is to be causative of reality’, i.e., the Absolute Will; each idea is also actual to God, ‘because and as far as it is one with the Will of the Father’.  As for the idea itself, it is real ‘inasmuch as a Will is its essence’; and the idea is also actual, but only inasmuch as it wills to be one with the will of God (236).  As a will, then, a divine idea may choose its potential to be otherwise, may will ‘its actuality in its Self and not in God’ (as the fact of the existence of evil proves) (237-38).[6]  Divine ideas thus inhabit a set of relationships in which their nature is multiple according to our varying vantage points, given the limits of our understanding.  What this multiplicity prevents, Coleridge argues, is any misconception that divine ideas are ‘actual as particular forms’ per se (236).  Hence no argument can prove their existence as discrete objects or mere objects of the understanding.  What, then, can argument do?

       Coleridge faces this difficulty less candidly, and rhetorically less




successfully, in his last ‘Essay on Method’ in The Friend (1818),[7] where his discourse is more lyrical than persuasive in a number of ways.  First, there is often little explicit coherence in argument, from one paragraph to another.  After the opening discussion in the essay on the two directions in humankind, ‘Trade and Literature’ (F I 507), Coleridge begins a loose series of paragraphs with the following non sequitur:  ‘Hast thou ever raised thy mind to the consideration of EXISTENCE, in and by itself, as the mere act of existing?’ (514).  A note on the preceding discussion, later added by Coleridge’s hand in Copy D of The Friend, diagnoses the problem thus: ‘This page + 10 lines may be justly complained of, for the obscurity of the connection with the preceding.—The method = transition, or pass-over, is for the Reader what the fording of a Stream over loose, oozy Stepping-stones at irregular distances would be to a foot-traveller:  I try in vain to recall my own purpose in these sentences.  S.T.C.  Omission will be the easiest cure’ (511 n. 3). 

       Besides looseness, related problems in persuasiveness remain in this same essay.  There is Coleridge’s penchant for isolated conclusions—for example, concerning a representative human’s discovery of outness: ‘that the reality, the objective truth, of the objects he has been adoring, derives its whole and sole evidence from an obscure sensation, which he is alike unable to resist or to comprehend, which compels him to contemplate as without and independent of himself what yet he could not contemplate at all, were it not a modification of his own being’ (F I 509).[8]  Besides such unsubstantiated statements, Coleridge also falls back on a kind of vague lyricism.  Defining method as ‘the self-unravelling clue’ and the ‘universal law’ of polarity is fairly clear in context; but clarity fails in talk of the gradual comprehension of ‘the relation of each to the other, of each to all, and of all to each’ (511).  Coleridge also uses rhetorical questions whose answers are not inevitable (‘By what name then canst thou call a truth so manifested?  Is it not REVELATION?’); and his terminology for method remains unarticulated:  ‘the principle of religion’, not ‘a sort of knowledge’ but ‘a form of BEING’ (516, 523-4).  And into this suggestive but vague rhetoric comes Coleridge’s discourse of ‘idea’.  That ‘intuition of absolute existence’ (that he is dancing around) ‘manifests itself’ for humans ‘in its adequate idea’ (514-5); that ‘principle of religion’ or ‘form of BEING’, he also calls ‘this life in the idea’ (524).  What these vague rhetorical strategies in this last Essay on Method lack is the broader rhetorical appeal of more logical sequential argument and a manifested coherent system.[9]  It is in these directions that Coleridge moves in ODI’s long, last section (OM 264-90), which




sets out to answer the question, ‘Is there a science of God?’[10]

       First, in ODI Coleridge uses extended argument.  Any science, he begins, start with first principles or the ‘datum’, ‘[t]hat without which we cannot reason’ (270), and concludes with results.  We can distinguish between the two: if the whole ‘universe [were] the subject which we proposed to understand… whatever the suppositions were… , the result… would be the science of the universe.  The supposition itself could not possibly be the result’ (272).  The idea of God, however, seems to be a peculiar case for a science, because an ‘idea… implies the reality of that to which it corresponds as well as its own formal truth… being thus contradistinguished alike from the forms of the sense, the conceptions of the understanding, and the principles of the speculative reason’ (270-1).  It is difficult finding a science of God, ‘unless the idea was taken as the… datum, and the result anticipated and pre-contained in the premise’ (270).  In this case, though, the resulting knowledge could not be called ‘a scientific proof’, but rather only ‘a series of exemplifications of the same truth, as if a man should demonstrate the essential properties of a triangle in a vast succession of diagrams, and in all imaginable varieties of size, and colour, and relative position: each would have the force of all’.  In summary, while sciences ‘either respect the formal truth alone’ (as in geometrical circles) ‘or else assume the reality in their premises as a datum’ (as in astronomy), ‘the very thing to be proved in the present instance is the reality [of God], that is, the positive existence, self-subsistently and not merely in and for the mind… If not from the idea, from what other source?  And, vice versâ, if the reality be not presumed[,] from whence shall we derive the conception?’ (271).

       Coleridge then moves on to show that the empirical sciences (which ‘assume the reality in their premises as a datum’) are not immune from a similar problem.  For in them,


the mind itself is the actual hypothesis.  I… place such and such a conception under [‘any series’ of phenomena] and in so doing make my own faculties and the reality implied in them the support of the [conception] and the pledge of their consistency… For the phaenomena I have the evidence of my outward and inward sense.  Of there being more than the forms of my own mind singly and exclusively, I have the proof of my sensibility and the comparison of my senses with each other, with my past experience, and with the experience of other men.  (OM 272)


But what grounds are there for these presuppositions themselves just enunciated—‘my own faculties and the reality implied in them’—other ‘than what the process itself expresses?’  It is easier to show that we grant phenomena outness (on the basis of what our Essay on Method calls that




‘obscure sensation’), than to justify this theoretically ‘by arguments derived… from the understanding alone’.  Establishing ‘a congruity of the phenomena with the forms of the understanding’ must presume ‘a something’, not ‘fairly deduced… from the evidence of the senses’ or ‘the abstractions which the understanding forms from its own processes, and substantiates for itself’, inferring itself as a faculty in the same process.  Coleridge concludes:  ‘And I greatly fear that whatever shall be thus presumed… will be found either to imply the idea of God, or… to presuppose and be grounded on that idea…’ (272-3).  Sceptical readers, to be consistent, now have three things to be sceptical about: the idea of God, the objects of their senses, and the forms of their understanding.  The science of things and of perceiving things cannot offer us the grounds upon which to critique—or to prove—the science of God.

       The only solution, Coleridge goes on to propose, is to examine each ‘supposed’ proof in turn (OM 273).  Having argued ‘positively’ for some 23 manuscript pages, Coleridge now turns for as many more (273-90) to negative proof, what he calls earlier in ODI ‘the absurdity’ of the contrary (221).[11]  If each so-called proof fails by its nature, then he could demonstrate ‘our position: that there is no speculative proof, no properly scientific or logical demonstration, possible.  In other words, that the idea of the Godhead is the true source and indispensable precondition of all our knowledge of God’ (273-4).  In turn, then, Coleridge dismisses, first, that we can ‘know God by the sense’ (or intuition), as we know geometrical figures (274-6); next, ‘that we see God with our eyes’, according to ‘the Brahmins’ (276-85); and then he previews what appears to be his critique of another scheme—the argument from design, God proved through the creation (286-90).[12]  He finds them all wanting.  All such proofs ‘pretend to the power of… the idea itself—or in the language of Mr. Locke, of conveying it into the mind by the force and instrumentality of the same reasonings as they believe demonstrative of a real existence in perfect correspondence to the idea’ (287-8).  In contrast, as Coleridge has already asserted, an ‘Idea is not simply knowledge or perception as distinguished from the thing perceived: it is a realizing knowledge, a knowledge causative of its own reality’ (223).  While Coleridge’s ‘point’ here is identical to the one he is trying to make in the last Essay on Method in The Friend, the sense of a long sequence here—from familiar generalizations about science, to the peculiarities of the idea of God, to the defamiliarizing of




empirical perceptions and refutation of opposing positions—lends it much more rhetorical conviction.

       If length and successive argument do this, so does repetition.  After Coleridge’s refutation of opposing positions, he returns to his doubt in the validity of unaided empirical perceptions or the understanding’s processing of them, as he begins to critique the argument from design.  He will dispute that the idea of God ‘is fairly deducible from the phaenomena that are the basis’ for ‘judgements concerning the reciprocal fitness and harmony of things’.  Instead he proposes to argue whether ‘these judgements do not presuppose the idea, whether by its light the phaenomena are not themselves first read and interpreted?’ (287).  Coleridge does not complete this argument—the manuscript apparently soon breaks off—but there are a number of clues to its contents.[13]  I will deal with two of these as they bear on the rhetoric of Coleridge’s repetition and the indemonstrability of the idea of God.

       The first clue occurs before the manuscript breaks off when he mentions, but neither cites nor identifies, a passage from Berkeley, in support of ‘the pre-existence of the Idea [of God] in the human mind’ (287)—arguably the passage found in a notebook entry of 1823 (CN IV 5096), from Berkeley’s Alciphron (OM 288, n. 223).  The argument of the excerpt, set as a humourous dialogue, is as follows: says Euphranor to the sceptic, Alciphron, if God has always been speaking to humankind universally from their infancy in a ‘language of vision’, it is not strange that they should be unaware of it, nor that they confound the connection between ‘objects of sight’ and their significance, or conflate them; this they also do in their own diverse languages. Do you really believe this?, Alciphron asks.  Yes, comes the reply, and so should you if you are consistent according to ‘your own definition of language’, since you must admit, given that language consists of ‘arbitrary signs’, that God must continually manipulate them for the sake of the communication through them that nonetheless occurs daily, ‘informing & directing men how to act with respect to things distant & future, as well as near and present’.  Euphranor concludes:  ‘In consequence, I say…, you have as much reason to think the universal agent or God speaks to your eyes, as you can have for thinking any particular person speaks to your ears’.  If you believe in human communication, you must also believe in the divine, through visible phenomena.  (Alciphron runs out of objections, but retorts that the position ‘is so odd and contrary to my way of thinking, that I shall never assent to it’.)  I think you will notice that this anecdote (an argument from authority) is a repetition in little of much of the ‘idea of God’ passage that directly precedes it in ODI.  It suggests a similar conclusion.  To believe in the idea of God and to believe in the evidence of our senses are alike innate acts insusceptible to abstract proofs.  Such repetition, from one rhetorical context to another, happens many, many times in OM at large.  Unlike the use of successive argument, here the sequence of repetition need not be logical.  Again, illustrations are not proofs but




‘expositions’ or examples of the ‘idea’—‘as if a man should demonstrate the essential properties of a triangle in a vast succession of diagrams…[:] each would have the force of all’ (OM 271).

       The second clue is another forecast, occurring early in Say 29, part ii, in a chapter entitled ‘On the existential reality of the Idea of the Supreme Being, i.e. of God’ (OM 96).  So as to enable his readers to follow the unfolding of the argument, here Coleridge predicts his ultimate conclusion—the ending of ODI—‘that the existence of the Supreme Being in any religious sense is indemonstrable’ (103).  In Say 29's version of a note that Coleridge added to Copy D of The Friend (I 522 n. 1), he asserts that the exclusively ‘dialectic intellect’ can affirm ‘an absolute being’, but no more: it cannot establish ‘the existence of a world different from Deity’ (OM 104).  (Here a condensed repetition of the Alciphron anecdote recurs: since the outness of phenomena cannot be proved, they in turn cannot prove the existence of God.)  The ensuing material in this chapter is integral to the imperfect conclusion of ODI, particularly given their shared allusion to Berkeley’s analogy of the world and the book (OM 108 and n. 116; 288-90).  For here also, Coleridge focuses on the argument from design, discoursing on its immense attractiveness but questioning its validity nonetheless (110).  He must remind the enquirer that this argument supposes the idea of God already present in the Mind, and the reality of His moral attributes to have been already established—to make the student aware that the great book of Nature, in order to be read, in order to be a book, supposes all the elements of thought not only to exist in the mind of its reader, but to exist in combination with those particular characters, so that to learn is, in fact, a process of reminiscence.  (110-11) He specifies two related motives for reminding readers of the innateness of the idea of God.  The first is that speculative minds, mistakenly believing that Nature proves the existence of God, often decline into pantheism, the identification of ‘the Divine Idea’ with Nature and the rejection of a personal God (111-16).  The assumption that Nature proves God, moreover, will lead to a further rejection of all Christian doctrines.  For both the idea of God and also Christian doctrines cannot be proved by the understanding according to ‘the universal law of cause and effect’; instead, they can only be commended by a system based on the postulate of the Will (116-8), that is, according to Coleridge’s argument in OM at large.

       In this regard, then, repetition in ODI does not cohere only by reiterated appeal to intuition or feeling, as we have already seen.  Repetition is also intellectually constituted by Coleridge’s postulate of the Absolute Will and his cumulative extrapolations from it across the considerable length of Say MS 29.  Coleridge defines this postulate again[14], at the beginning of ODI before he discusses Divine ideas:




The Will, the absolute Will, is that which is essentially causative of reality, essentially, and absolutely, that is, boundless from without and within.  This is our first principle.  This is the position contained in the postulate of the reality of Will at all… But in this affirmation it is involved that what [sic] is essentially causative of all possible reality must be causative [also] of its own reality.  (OM 220-1)


One subsequent example, early in ODI, of his using the postulate in his discussion of Divine ideas will suffice.  Coleridge reminds us of sitting in a chair, where certain acts or passions are ‘actually present’ while many others are potential (227).  To extrapolate in my own words, you might actually cough while reading this paragraph, but leave as potential the thought of, say, scratching your ear.  Presumably God as ‘a being limitless’ would have both ‘self-comprehension’ and also ‘a knowledge of the finite’, and of ‘finites’ knowledge of one another and other things (230).  Thus ‘not only would all things that are have actual reality in the universal mind’ (your coughing); but so also would ‘their potential being in relation to the finite’ (your thought to scratch your ear)—because, Coleridge adds, this potentiality ‘would be affirmed in the same act that gave being to the finite and be included in the product of that act’.  Then Coleridge surfaces the postulate implicit in this consideration, quoting his earlier definition just cited above: ‘But this act we have defined as That which is essentially causative of reality’.  Accordingly, the potential must be ‘a form of reality… no less, though a far lower form, than the actual’.  Returning to the doubleness of divine ideas, he concludes, even so may we predicate ‘both the actual and the potential of one and the same subject’.  What began, then, with us actually coughing and maybe scratching our ears becomes a new step in argument—one of many such steps in OM—informed by Coleridge’s postulate of the Absolute Will.  In this kind of repetition, the repetition of the postulate co-inheres with ever extending stages of unfolding argument.  The result is a formidable system.

       In this essay then, I have first examined the hitherto neglected, concluding section of ODI on the indemonstrability of the idea of God. Because this section is incomplete, I have drawn on two cognate passages.  The inferred Alciphron anecdote in ODI suggests that believing in the idea of God and in the evidence of our senses are innate acts otherwise not abstractly provable; and earlier in OM, Coleridge highlights the innateness of the idea of God as an antidote to pantheism and other departures from Christian doctrine.  Consistent with his position, neither of these passages strictly proves Coleridge’s point.  So I have focused also on the rhetoric of his arguments, particularly on how repetition functions in two ways.  On the one hand, there is repetition by intuitive appeal to the same conclusion in different contexts; on the other, repetition in association with Coleridge’s unfolding system across OM, based on his postulate of the Absolute Will.  This latter, what Coleridge calls positive proof, also appears in the ‘Idea of God’ section in a particular form of extended and successive argument, comparing science and the idea of




God and defamiliarizing empirical perceptions.  Here Coleridge also uses negative proof, the refutation of opposing ideologies, leaving his own standing by default.  Finally, my analysis has set out to resolve some of the hazy nomenclature of ‘idea’ in the last Essay on Method in The Friend.  There Coleridge’s discourse prevaricates around an undefined centre: an ‘intuition of absolute existence’, ‘the principle of religion’ or ‘form of BEING’, ‘this life in the idea’.  In contrast, ODI names this centre as the idea of God, and argues for it in extensive and varied rhetorical fashion.  I present OM’s demystfication of the last of the Essays on Method in The Friend as only one example of how two constructed extremes of Coleridge criticism might meet.  I mean, of course, the often discussed thoroughfares of Coleridge studies—his early radicalism, the big poems, the Biographia, and The Friend—and the ‘later’ or ‘religious Coleridge’, his prose and poetry of the 1820s and 30s.  Recently available editions of the Opus and late notebooks will continue to offer opportunities to de-marginalise the later Coleridge in a way that can only enrich Coleridge studies as a whole.

       In conclusion, my discussion also intervenes in the ongoing debate about whether or not there is an articulated system in Coleridge’s thought.  ‘Anyone writing about Coleridge must make a decision about coherence’, Seamus Perry says of Coleridge up to 1818, presenting us with three customary choices: complementing what unity is just about there, ‘discerning in the incomplete works a positive rhetorical strategy’, or ‘accepting his failure as just that’.[15]  I think that the Opus Maximum presents us with another choice.  Here in a set of long, albeit fragmentary texts, we may follow the elaboration of a large part of Coleridge’s system from its first postulate. In that elaboration, the divine ideas have a central place, in a way which illuminates Coleridge on ‘idea’ before and after the Opus Maximum.


© Contributor 2003-2006



[1]          The Collected Works of Samuel Taylor Coleridge: Opus Maximum, ed. Thomas McFarland, with the assistance of Nicholas Halmi, Bollingen Series LXXV (Princeton: Princeton University Press, 2002), 211, hereafter cited as OM parenthetically in the body of this essay.

[2]          The Collected Works of Samuel Taylor Coleridge: Shorter Works and Fragments II, eds H.J. Jackson and J.R. de J. Jackson, Bollingen Series LXXV (Princeton: Princeton University Press, 1995), 1511.

[3]          See for example Stephen Happel, Coleridge's Religious Imagination (Salzburg: Institut für Anglistik und Amerikanistik, Universität Salzburg, 1983), 704-17 and Mary Anne Perkins, Coleridge’s Philosophy (Oxford: Clarendon Press, 1994), 157-8, 175-93.

[4]          The Collected Works of Samuel Taylor Coleridge: Shorter Works and Fragments I, eds H.J. Jackson and J.R. de J. Jackson, Bollingen Series LXXV (Princeton: Princeton University Press, 1995), 156, hereafter SWF I.

[5]          Robert Barth, S.J., Coleridge and Christian Doctrine (Cambridge, Mass.: Harvard University Press, 1969), 106 n. 4 .  On the provenance of divine ideas, Perkins (173) points out that Coleridge takes ‘classical Logos theology’ and goes ‘beyond the familiar tradition of Christian Platonism’ by adding ‘insights of German idealism, and ... advocating the primacy of Will, resulting in a  more dynamic philosophy than that of Ralph Cudworth’s True Intellectual System (1678).  

[6]          Thus, on Coleridge’s reckoning, in the Absolute Will and divine ideas, but not the Triad, ‘the potential necessarily co-exists as alternable with the actual’ (232).

[7]          The Collected Works of Samuel Taylor Coleridge: The Friend, ed. Barbara E. Rooke, Bollingen Series LXXV (London: Routledge & Kegan Paul and Princeton: Princeton University Press, 1969), I , 507-24, hereafter F I.

[8]          On ‘outness’, see, for example, Owen Barfield, What Coleridge Thought (1971; reprint, London: Oxford University Press, 1972), 59-68.

[9]          Coleridge himself apparently notices the problem, judging by other notes he added to Copy D (e.g., F I 515, notes 2 & 3, and 522, n.1), which only start filling in gaps in his material, which OM will address systematically.  Indeed, the last of these, on pantheism, is reproduced in almost identical form in Say MS 29 (OM 104-7).

[10]        OM 264-8 is a first draft on this topic: ‘If it is possible to demonstrate the existence of the Deity, there must be a corresponding science’ (265).  This attempt Coleridge apparently abandons, to begin again at p. 268: ‘The sciences are distinguished by their subjects…’

[11]        On positive and negative proof, see SWF I 786-7.

[12]        I say ‘what appears to be’ his critique of the argument from design, because he considers in these last pages of the manuscript ‘the adaptation of means to ends, and the pre-established harmony of things ... [which] so irresistibly refer the awakened soul to the perfections of the great Creator and Governor of the World’ (286).  But there is considerable textual instability in this context.  Coleridge refers to two quotations, which are not included: one from Warburton (285 and n. 218), and one from Berkeley (287, 288 and n. 223).  He also cancels two passages in the manuscript, which would take his argument in different directions: to the non-validity of Christian ‘evidences, historical and moral’ (OM 285, note a), when divorced from the revealed doctrines of orthodox Christianity; and back to material, which precedes his refutation of ‘Brahmanism’ (287, note b) and which I have already discussed in my essay above (see 00-00).  He does, in fact, continue on aspects of this second topic, as I discuss in the next paragraph.

[13]        Two, which space does not allow me discuss, occur earlier in ODI (OM 230) and in Say 29,  (OM 161-4). 

[14]        My argument implies that ODI continues the argument of the three parts of Say 29.  In this I follow the likely chronological sequence of fragments as proposed by Nicholas Halmi (OM xx).

[15]        Seamus Perry, Coleridge and the Uses of Division (Oxford: Clarendon Press, 1999), 2.