Semantic Finegrainedness and Poetic Value

1 This paper was originally published in John Gibson (ed.). The Philosophy of Poetry. Oxford: Oxford University Press, 2015. 18–36. It is reproduced with permission of Oxford University Press. 2 This choice of example is fairly arbitrary but it seems suitably typical of a certain style of modern poetry, it well illustrates the kind of difficulty that interests me, and it was selected as among the ›best British poetry‹ for 2011. This paper – first published with Oxford University Press in 2015 – 1 argues that poetry is constituted by a practice, which is grounded in convention-governed expectations among poets and readers. To write a poem is to engage the practice and invite (one hopes also reward) certain kinds of interests and responses among readers; to read a poem ›poetically‹, seeking its poetic value, is to deploy the relevant interests and responses thereby making appropriate demands and one hopes achieving the valued experience on offer. It is part of the poetry game that in poetry we attend to the finegrainedness of language, its textures and intricacies, its opacity, in conveying thought processes, and we find value in the experience that affords, in precedence over the more humdrum norms of communication, such as transparency, the imparting of information, and the assumption of paraphrasability.


1.
It is characteristic of a certain kind of poem -notably the short lyric -to compress language in such a way that resists initial easy comprehension.Examples abound.Consider the two opening stanzas of a recent poem, Waves, by Giles Goodland (Lumsden 2011, 36): 2 The sea is a misunderstanding we have to go through in order to make sense, like the word for a loss of a word.It leaves a sense of having left, Through which silence leaks.
What waves are reopening nightly is the senseless apparatus of an eye.If you live in a house made of thought you nod in the silence the sea makes.This compression of meaning is undoubtedly a mark of, but is not peculiar to, modern or modernist poetry.It can also be found in other periods including the nineteenth century.This is the beginning of Algernon Charles Swinburne's poem Hertha (Quiller-Couch 1931, 974): I am that which began; Out of me the years roll; Out of me God and man; I am equal and whole; God changes, and man, and the form of them bodily; I am the soul.
On the face of it writing of this kind is puzzling.The words themselves are not technical or obscure, the subject matter not especially recondite, at least in the former poem, yet the meaning is difficult to grasp, the syntax irregular, there is no obvious point (to inform, to advance knowledge), and standard discursive norms of clarity are bypassed.Nevertheless, such writing is immediately recognizable as poetry and often admired, not in spite of, but because of the semantic complexity it exhibits.
My principal question is this: why should writing of this kind that is dense, complex and resistant to ready understanding have value, specifically poetic value, when in other contexts it might be censured for its wanton lack of clarity?Why does poetry license, even encourage, such compression of meaning?
There are of course certain stock answers to these questions and it might be thought they are sufficient: for example, that the complexity of poetic language simply reflects the complexity of the thoughts and emotions being expressed; or that the complexity rests on the fact that poems characteristically use metaphor and figurative speech; or simply that poets like to play with language, push it to its limits, explore its possibilities, that's what poets do.There's truth in all of these answers, without doubt, but none of them quite explains why such uses of language should be highly valued.Perhaps poets do push language to its limits -and use metaphor more than usual -but what is the value in that?Why should readers get pleasure from having to struggle to understand poetry?And is it really the case that complex language in poetry is always the expression of complex thoughts?Giles Goodland, quoted above, writes this about his poem Waves: Staying at a beach chalet with the kids, I was playing with the idea that the humours in the eyeball have a similar composition to seawater, and I read somewhere that this is a relic from our very distant evolutionary past.We all carry a little bit of ocean with us, and use it to look with.This combined with being on a beach with my children and thinking about the various meanings of waves.Sound also is very important in the poem and I allowed myself to get carried away with some slightly outrageous sound-alike words and phrases [...] Something about the sea always seems to loosen me up in my use of words [...] So I let the words play around, as the children were doing, in my notebook.Language is like the sea, as a whole it is formless, but it can form localised shapes called sentences or even poems.The poem ends with the children, and that feeling of having the ground pulled under you when a wave rolls back into the surf.Sitting on the beach watching them, I got the idea for the poem: we are like waves, we carry sea with us, generation after generation.(Lumsden 2011, 126-127) The thought itself -about eyes and sea water -is intriguing but hardly complex and the circumstance of children playing on a beach seems not to demand semantic contortion.The stock answers, then, seem only to reinforce the urgency of the original question.For Waves is a pleasing poem, fun to read and repeat, not overly profound, not perhaps one of the great works, but typical in its complex word-play.Here are the last few lines: Poe-lipped polyps lens and tense to sense the same body inside this one.Their homeland is brine, the gull-lulling greys of its waves.
Lip-read the sea rolling in pain.See such children it sucks like a sweet.
The pebbles are frantic under them.
We hear and enjoy what the poet calls the »slightly outrageous sound-alike words and phrases«.But more needs to be said to explain poetic value than appeal to word play alone.

2.
To get more deeply into the discussion it will help to reflect on four supposed commonplaces about poetry.These provide a clearer focus, at least at a philosophical level, into the nature of poetic meaning and its value.
1.The Experiential thesis.The core value of a poem lies in the experience(s) the poem affords when read as a poem.2. The Heresy of Paraphrase thesis.The precise meaning of a poem is unparaphrasable.
3. The Form-Content Unity thesis.The form of a poem is indivisible from its content.4. The Semantic Density thesis.Poetic language affords a peculiar kind of ›semantic density‹.
Each of these theses is closely connected to the others; in a sense they are mutually supportive.Furthermore, as often repeated commonplaces, they seem to possess more than a grain of truth.But of course they need a great deal of unpacking.All are puzzling from the point of view of standard theories of meaning.What should be so peculiar about poetic language that, unlike nearly any other usage, it should be resistant to paraphrase?Part of the answer no doubt lies in the Experiential thesis but that itself is problematic because it is not obvious why any piece of discourse should be valued primarily for the ›experience‹ it affords rather than, say, the proposition(s) it expresses or the information it conveys.And what kind of experience is at issue?The supposed indivisibility of form and content will only make sense when the ideas of ›form‹ and ›content‹ are amplified but it would be a marked peculiarity of poetic usage -again setting poetic language apart from other usages -if it should be the case that in poetry how something is said should be so integrally related to what is said that there should be only one possible way in which a certain content could be expressed.If the Semantic Density thesis is right -that poetic language somehow generates a depth of meaning not found in other linguistic modes -then this too needs explanation.After all, most of the ›devices‹ characteristic of poetry are familiar elsewhere: metaphor, simile, imagery, alliteration, onomatopoeia, rhyme, repetition, etc.We need to fill out the theses and explore how they are related.The Experiential thesis, that the value of a poem lies in the experience(s) the poem affords, is defended by A. C. Bradley in his 1901 inaugural lecture Poetry for Poetry's Sake.Bradley argued that »an actual poem is the succession of experiences -sounds, images, thoughts, emotions -through which we pass when we are reading as poetically as we can« (Bradley 1926, 4).Taken literally, that is a strong claim as it appears to identify a poem with a succession of experiences; if it is a genuine identity claim, an ontological claim, it seems on the face of it implausible.3However, Bradley qualifies his remark by saying he is not »aiming here at accuracy« or offering a »definition of poetry« and the Experiential thesis as I present it is not a thesis in ontology, rather a thesis about value.This too is the core of Bradley's view, for the experience of reading poetry, he states, »is an end in itself, is worth having on its own account, has an intrinsic value … poetic value is this intrinsic worth alone« (Bradley 1926, 4).So the value of a poem, its poetic value, is for Bradley the intrinsic value of the experience the poem affords.A similar view is expounded by Malcolm Budd: »what matters in poetry is the imaginative experience you undergo in reading the poem, not merely the thoughts expressed by the words of the poem« (Budd 1995, 83).
What more can be said about this experience?Can anything substantial be said about it, in abstraction from particular cases?The first point to make is that the experience of each poem is unique to the poem: no two poems afford qualitatively the same experience, at least at the level of specificity which the relevant experience demands.No doubt there are shared generic experiences associated with poems on broadly similar themes or in similar poetic styles.But that is not the level of generality at which poetic value lies.
Second, the experience in question is the experience of readers, not the experience of poets themselves.Poetic value does not reside in the quality of any psychologically real state of mind manifested in a poet as either the cause or subject of a poem.Where a poem expresses an emotion it is the expressed emotion itself -expressed in just this form -that is the subject of an attentive reader's experience and thus the source of the poem's value.
Third, the experience is normative: not any experience that a poem elicits in a reader can be considered an appropriate measure of the poem's value.For Budd the relevant experience involves »interacting with [the work] in whatever way it demands if it is to be understood« and must be »imbued with an awareness of (all) the aesthetically relevant properties of the work« (Budd 1995, 4).No doubt, quite legitimately so, there will also be more purely subjective experiences (responses) leading to subjective judgments of value.However rational or dispassionate we are in our judgements it would be absurd to expect that we all enjoy the same poems.But a judgment of the poem itself should, as far as possible, be grounded in the experiences that the work elicits on its own terms, rooted in its own objective properties.
Fourth, the experience is multi-faceted; it is not just, if at all, a sensation or feeling of pleasure, nor is it to be classed as a sui generis ›aesthetic experience‹.What we shall see later is that it is an experience of a subject through a mode of presentation.It involves, as Bradley remarked, such elements as »sounds, images, thoughts, emotions« and it is a temporal process, not a state.It is a process of thought, constrained by the linguistic medium that gives it both its character and its identity.
We shall come back to the Experiential thesis when we have explored the three other theses.It is significant that in highlighting experience in relation to the value of poetry the thesis places poetry naturally in the context of other arts.That experience should be prominent in accounts of music, painting, film, sculpture or dance is not especially remarkable.But poetry is a language based art and language is indissolubly associated with meaning, so an important shift has occurred when a thesis on the value of poetry makes no explicit reference to meaning and indeed highlights experience over meaning.Our other theses, on paraphrase, form and content, and semantic density, do seem more directly connected to meaning but it will be a theme of the argument to follow that an overemphasis on meaning can be misleading in capturing what is of value about poetry.
The Heresy of Paraphrase thesis, that the precise meaning of a poem is not paraphrasable (i.e.expressible in other terms), is a commonplace about poetry, in the sense that lip-service is paid to it in general reflections on the subject, but is nevertheless controversial under closer examination.There are those who simply deny that resistance to paraphrase is an integral or even interesting feature of poetry.Peter Kivy, for example, has argued that for the most part paraphrasing poetry does not present an insuperable problem once we are clear on a »criterion of success« (Kivy 1997, 104).We must not, he insists, set such a criterion unreasonably high, so high in fact that it couldn't possibly be fulfilled.Significantly, he suggests that one such criterion is too high and unfulfillable, namely that paraphrase capture the experience that the poem affords.
No one who sets out to say in prose the content of what a poem says in poetic form intends as the goal of the task to provide an alternative way of experiencing the poem.And to fault the interpreter for failing to do what is not the point of interpretation in the first place is plain nonsense.(Kivy 1997, 105) Three brief preliminary comments on this. 4 The first is that Kivy is readily conceding that there is indeed a relevant experience uniquely associated with the poem yet, on his view, it is not to be identified with the poem's content, nor in any essential way connected to interpretation.The second is that Kivy is associating paraphrase with interpretation, as if the goal of interpretation is to paraphrase a poem's meaning.That assumption needs to be challenged for it is far from clear that a good interpretation of a poem seeks simply to say in other words what the poem itself says.Poetic interpretation, I suggest, is better conceived as a way of encouraging, enhancing or developing precisely the experience a poem affords, as described in our discussion of the Experiential thesis.The third point is that Kivy, in setting apart paraphrasable content, on the one hand, and experience, on the other, is falling into just that dualism that Cleanth Brooks, who coined the expression Heresy of Paraphrase, warned against: a dualism between some kind of ›statement‹ that the poem makes and the eloquence, or beauty, or clarity with which it does so (cf.Brooks 1968, 160).In fact Brooks is quite happy to concede just the kinds of loose paraphrase that Kivy defends: »the point is [...] not that we cannot describe adequately enough for many purposes what the poem in general is ›about‹ and what the general effect of the poem is«.Indeed he allows that »[w]e can very properly use paraphrases as pointers and as short-hand references« as long as »we know what we are doing and that we see plainly that the paraphrase is not the real core of meaning which constitutes the essence of the poem« (Brooks 1968, 160).That so-called »essence«, for Brooks, resides only in the total »structure« of the poem and cannot itself be captured in other formulations.
Philosophers of language are rightly wary of claims of unparaphrasability.The thought that there could be only one way in which something could be said goes against basic principles of semantics, including what Ernie Lepore (after Donald Davidson) has called »semantic innocence«, the idea, at its simplest, that »what an (unambiguous) word means it means everywhere it occurs« (Lepore 2009, 181).The Heresy of Paraphrase violates semantic innocence to the extent it holds, as Lepore puts it, that »what the words of a poem mean is contingent upon where in a poem they occur« (Lepore 2009, 182).Lepore, however, ingeniously offers a defence of the Heresy of Paraphrase while retaining semantic innocence.To do so he introduces the idea of hyperintensionality, which claims there are »linguistic environments in which replacing an expression with its synonym changes meaning«.The prime example in natural language is quotation.Lepore illustrates this as follows: »bachelor« is the first word in »bachelors are unmarried men« »unmarried man« is the first word in »bachelors are unmarried men« The former is true, the latter false, yet »bachelor« and »unmarried man« are synonymous.He goes on: Likewise, try replacing »sheen« with »luster« in Coleridge's poem The Rime of the Ancient Mariner (1797): And through the drifts the snowy cliffs Did send a dismal sheen; Nor shapes of men nor beasts we ken The ice was all between.
The change sufficiently alters the rhythm and rhyme to break the bind between the lines, and thereby alters the poem itself.Since synonym substitution in a poem can change meter or rhyme, etc., and thereby change the topic, poems too create hyperintensional contexts.The right conclusion is not that expressions carry unique meanings inside and outside of poems.(Lepore 2009, 195) Behind the idea that poems create hyperintensionality, indeed a partial explanation of that idea, is a further claim that »poetry, like quotation, doesn't support substitution of synonyms because it harbours devices for being literally (partly) about their own articulations« (Lepore 2009, 195).Because a poem is »partly constituted by its own articulation [...] it is not re-articulable in another medium« (Lepore 2009, 193).Lepore's suggestion is ingenious and, if correct, has the merit of taking some of the mystery out of the semantics of poetry.However, what the account does not do is answer our initial question about the value of the compression of meaning in some lyric poetry.Why should it be of value or afford an intrinsically valuable experience that a poem should be somehow about its own articulation?If this is just another way of saying that pleasure can be had in attending to the modes in which thought in poetry is presented then so be it, for that is surely right, but there seem to be additional factors at work because there is no particular pleasure to be had in the hyperintensionality of quotation.And is it true that poems are even partially about their articulation?That seems doubtful in itself, quite apart from the worry that it doesn't explain their value.What a poem is about, I will suggest, is its finegrained content identified by, certainly, but not equivalent to, its particular mode of articulation.
Other philosophers have appealed to metaphor to explain unparaphrasability in poetry.Two lines of thought are evident here.The first says that it is in the nature of live or poetic metaphors to be inexhaustible in their connotations so no precise literal equivalent is possible.The second, following Davidson, denies there is any such thing as metaphorical meaning so there is nothing, as it were, to paraphrase.What a metaphor means is simply what the words mean in their literal application and the rest is merely effect.Neither view is especially illuminating about the supposed unparaphrasability of poetry.Perhaps the direction of explanation is the wrong way round.Might it not be that an adequate theory of metaphor could benefit from an account of the value of poetry rather than the value of poetry being explained through a theory of metaphor?The idea that metaphor is inexhaustible might be true in some cases but the nature of, and constraints on, that inexhaustibility, if relevant here, will only be determined in the light of some wider conception of poetry and its aims.As for the Davidsonian account of metaphor, it might seem in principle promising in the context of poetry if only for its focus away from meaning and its admission of non-propositional elements in responses to metaphor.However, there is an unruliness in relying exclusively on unconstrained effects which is not true to the kind of normative experiences demanded by poetry.
So what remains of the Heresy of Paraphrase thesis?At one level the resistance of poetry to paraphrase follows simply from the thought that it matters in poetry, more so than in other forms of language use, exactly how something is expressed.The form of expression is not distinct from what is expressed, a point we shall return to.The idea that there might be some other way of saying precisely what the poem says contradicts the very nature of poetry which is to draw attention to, give salience to, its modes of expression.That is the truth behind the hyperintensionality view.
The Form-Content Unity thesis, that the form of a poem is indivisible from its content, expands on, but is not identical to, the Heresy of Paraphrase thesis.The latter might be true but not the former.Resistance to paraphrase, in a milder version, might be judged just a contingent fact about the complexity of poetic language; the less complex a poem the more open it is to paraphrase.But if form-content unity is true then unparaphrasability is not contingent but necessary for if paraphrase captures content, and content is indivisible from form, then form too must be retained.That is hyperintensionality again.A. C. Bradley rarely mentions paraphrase but argues explicitly for form-content unity: »this identity of content and form [...] is no accident; it is of the essence of poetry in so far as it is poetry« (Bradley 1926, 15).
Clearly to assess the Form-Content Unity thesis the ideas of ›form‹ and ›content‹ must themselves be clarified.That in itself is not straightforward given the multiple ways the terms are used.However, it is fairly clear what kinds of features count as formal in poetry: rhyme, metre, rhythm, alliteration, repetition, etc.The ›etc‹ might be difficult to cash in but some preliminary points about indivisibility can still be maintained.One is that form-content unity in poetry is holistic: the content of the whole is indivisible from the form of the whole.But (holistic) form-content unity does not imply that for any individual formal feature it is legitimate to ask what feature of content it is equivalent to, even if a rough and ready account might be available for the contribution made to the whole by each of the parts.Also it would be wrong to think of ›form‹ in the relevant sense just as the sum of individual formal features.Form-content unity is stronger than that in insisting that the total surface configuration of a poem is indivisible from its total content.This is more like what Brooks means by »structure«.Total content, whatever that might mean, is in a sense supervenient on the structural base.
The idea of ›content‹ is equally problematic.In the light of the Experiential thesis it should not automatically be identified with meaning.Perhaps better is to think of it in terms of ›aboutness‹, while recognizing that aboutness is interest-relative. 5Specifications of what a poem is ›about‹ come in degrees of finegrainedness according to the interests served in making the specification.In some contexts it might be enough to say of Giles Goodland's poem that it is about waves or the ocean or a day out at the sea or a hypothesis about eyes and seawater.But these descriptions do not do justice to finegrained content.Ultimately the most finegrained specification of the content -the ›total‹ content -is that which incorporates the precise form in which the content is presented: in other words the poem itself.This level of finegrainedness is anticipated in hyperintensionality and in the Heresy of Paraphrase.The interest served at this level is an interest in the poem, as Bradley would say, ›for its own sake‹.It is at this level, but only here, that content becomes indivisible from form.
Form-Content Unity now takes on a different complexion.It need not be thought a merely contingent fact about (some) poems that their content is indivisible from their form.Form-content identity is not something that one discovers in a poem but rather something one demands of a poem when bringing to it a certain kind of interest: an interest in the poem for its own sake.This is not to deny other kinds of interest in a poem which might focus on more abstracted features of content or indeed form.One might, for example, be interested in commonalities across poems, perhaps metrical or thematic similarities.
But it is not an eccentric interest to attend to form-content indivisibility in a poem. 6ttention to finegrained content -a subject-realised-in-just-this-way -is essentially attention to expression.The philosopher W. M. Urban in Language and Reality (1939) defends form-content identity like this: »The artist does not first intuit his object and then find the appropriate meaning.It is rather in and through his medium that he intuits the object« (quoted in Brooks 1968, 163).Such a view is familiar also from R. G. Collingwood for whom the very act of expressing an emotion gives a clear identity to that emotion: the expression both characterises and discovers the emotion, which would remain unknown without that mode of expression (cf.Collingwood 1947, 111, 122).So it is that the specificity with which Giles Goodland characterises his thoughts on the ocean and seaside in the poem is not a way of capturing thoughts that pre-existed the expression, except in the most general terms, but crystallised and brought into being the thoughts themselves.
Before we begin to tie our theses together and draw out some general conclusions about poetic value, we must attend briefly to the fourth thesis, Semantic Density, which is most explicitly about meaning.For the New Critics of the 1930s and 1940s what distinguished literary language, by which they primarily meant the language of poetry, from ordinary or non-literary language is a kind of ›density‹ of meaning characterised by ›ambiguity‹, ›tension‹, ›irony‹, ›implicit meaning‹, ›connotation‹, etc., these being favoured terms of art. 7It was with examples of just such density or compression of meaning that we started so it is hard to deny that such qualities can be exhibited in poetry.Again, though, caution is needed in grasping the significance of this.The trouble with semantic density is that if you look for it -anywhere -you are likely to find it.All words are rich in connotation.In conversational discourse nuances of meaning are sometimes picked up but more often than not transparency is aimed for and achieved in the imparting of information.Connotations, implied meanings, and potential ambiguities are overlooked or viewed as distractions.And that is the point.In most contexts language is treated as a transparent medium for communication.Getting an idea across in some way or other is the goal.In poetry, given form-content indivisibility, transparency is replaced with opacity.The very forms of expression draw attention to themselves.Richness of meaning is not an obstacle but a medium.Connotations, allusions, symbolic meanings, resonances, add to the pleasures of the poetic experience.Semantic density, then, like form-content unity, is not discovered in poetry but demanded of it.It is striking that Cleanth Brooks in The Well-Wrought Urn finds tension and irony not only in John Donne and W. B. Yeats, where you might expect it, but also in Gray's Elegy, Pope's Rape of the Lock, and Tennyson's Tears, Idle Tears, which seem, on the face of it, deceptively uncomplicated.

3.
Again, going back to where we started, the mere presence of semantic density or multiple meaning does not explain poetic value.Indeed its own value is what needs explaining.But I think now we have the elements to offer an explanation.It is wrong to think of the pleasure of reading poetry, especially difficult poetry, as centred on the deciphering of meaning, somewhat like doing a crossword puzzle.8Pleasure might be had in that activity but it cannot be at the heart of the value of poetry.Issues about meaning of course crop up -a word or phrase might need explication -but often where help in interpretation is needed what matters is grasping the point, a kind of contextualisation.Indeed this is what Giles Goodland offers in his comments on his poem, drawing our attention to the theme that connects eyes and sea water.But grasping the point in this way -like noting word-meaning -is not an end but a mere preliminary for a deeper appreciation.Here meaning gives way to experience.
If Bradley is right the value of a poem as a poem lies in the intrinsic value of the experience the poem demands.What is this experience?In brief, it is the experience of a form-content unity.And what exactly is that?It is the experience of a subject matter partially defined by the very modes of expression through which it is presented.The language of a poem is not a vehicle for conveying a thought that is independently expressible; it conveys a thought that is encapsulated in the vehicle.Poems create hyperintensional contexts, with content that is unparaphrasable, not just because synonyms cannot be substituted but because, given the interests brought to the poem, the content demands the most finegrained identity conditions; it is a content given in just this way, inseparable from the form of its presentation.
Why should the experience of finegrained content be valuable?Sometimes of course it is not valuable -not all poems succeed and they can fail in many directions but always in the end because of weaknesses in the form-content unity, in diction, thought, originality, wit, and so on.To criticise a poem as a form-content unity is to criticise at one and the same time both what is expressed and how it is expressed.The poem for whatever reason doesn't work, it doesn't capture our attention, it is flat and lifeless.Maybe in the end we just abandon the attempt to read it as a poem.Where poems do succeed it is because of the pleasures to be had in adopting a unique (finegrained) perspective on a subject whether, in broad terms, familiar or unfamiliar.Of difficult poems we should not ask »why could it not have been put more simply?«because there is no reference for ›it‹, the content, such The second stanza speaks of the memories prompted by the cluttered objects: [...] They resurrect absurdly youthful passions exploring marvels of what might have been but for a word misplaced or not spoken, a touch delayed a moment too long.... The third and final stanza broadens out the reflections, as it were cashing in the symbolic weight of the loft as a store of memories and lost desires: Yet knowing so little and understanding less was all the wit I had.Besides the play was different when all these ghosts were living hands and eyes, when hopes and desires which can barely be recalled were as urgent as a trumpet.Or perhaps those scruples were not so foolish, since every season has its proper logic: sufficient unto its time the reasons thereof.[...] A key to the experience the poem affords is not the grasping of a summative proposition, the idea of memories prompted through a visit to the loft.It is rather that the thoughts are pervaded by the mood evoked -the tint of melancholy, time past, age, loss, lessons learned.The darkness of words like »dismay«, »past defeats«, »dead clothes«, »broken cases«, »broken lamp«, »flotsam«, pervade the experience; the perspective of sadness gives the content its special character, it gives shape to the process of thought as it develops.
I said earlier that form-content unity and semantic density are not discovered in a poem but demanded of poems when read »poetically« (in Bradley's term).The fundamental idea here is that of broadly marked conventions of reading or at least expectations brought by readers when approaching a piece of writing as a poem.Poetry is constituted by a practice, which is grounded in convention-governed expectations among poets and readers.To write a poem is to engage the practice and invite (one hopes also reward) certain kinds of interests and responses among readers; to read a poem »poetically«, seeking its poetic value, is to deploy the relevant interests and responses thereby making appropriate demands and one hopes achieving the valued experience on offer.A simple and familiar example of how conventional expectations in poetry can shift the focus of interest might arise from Noam Chomsky's famous sentence »Colorless green ideas sleep furiously«, which he introduced (in 1957) as an example of a sentence syntactically well-formed but semantically without meaning.In the context of semantics the sentence, with its multiple category violations, lacks evident truth conditions but there have been sought after under relevant modes of attention when the poetry game is engaged.The definitive experiences are not sensations or even emotions -there is no determinate phenomenology of poetic experience across all genres (the very idea is absurd) -but are processes of thought precisely shaped by the form-content unities that are their focus of attention.Of course there might be accompanying emotions or sensations in particular cases but these are contingent and context-dependent.So, returning to our initial question, why do we value the complexity -and difficulty -of (some) poetry when we spurn it elsewhere?The answer at its simplest is: because it is part of the poetry game that in poetry we attend to the finegrainedness of language, its textures and intricacies, its opacity, in conveying thought processes, and we find value in the experience that affords, in precedence over the more humdrum norms of communication, such as transparency, the imparting of information, and the assumption of paraphrasability.11