Muti-lation at the end of the line

In Broken Hierarchies: Precursor to a Variorum? I noticed major revisions to a poem in Geoffrey Hill’s Broken Hierarchies: Poems 1952-2012, which I argued amounted to a totally altered poem following a new and different philosophical development. Because the revisions in BH are widespread and significant, I suggested that critics writing on Hill are now and for the foreseeable future faced with a problem of the authoritativeness of different versions.

Some people have been cataloguing variant versions of poems in BH here and there, especially on the Facebook group Geoffrey Hill Exchange. Perhaps soon we will start seeing such analysis in the scholarly literature.

Yesterday, looking up something I had written before, I came across a small digression of mine in Defending Poetry, on two line breaks in Poem XI of Mercian Hymns. In a prose poem* like those of Mercian Hymns  the line “ending”, if it is that, can appear more accidental, or incidental, than in a verse poem. The typesetting, the margins, etc., may be actively in control of the poet, or he may have relinquished that control. As I was writing on the poem chapter several years ago, I noticed a very small change in different editions of that poem, which nonetheless affected my reading.

The poem in the edition I was working with–Geoffrey Hill Collected Poems (GHCP – 1985)–had the first two verse paragraphs of the poem set like this:


Noting the two hyphenated “enjambments”, I wrote the following:

The word ‘muti-|lation’, illustrating its own dismemberment on the page,[1] also calls attention to its special application to books or texts (OED3, n. 3), as well as its etymological Old French sense of a ‘partial destruction of a work of art’ (OED3, n. etym). Similarly, the visual deconstruction of ‘account‑|able’ leaves us with the three distinct sense units which the poem has joined together: ‘account’ as in the verb meaning ‘to reckon for moneys given or received’ (OED2, v. II.3), ‘able’ as in ‘showing … skill; talented, clever’ (OED2, a. 7), and ‘accountable’ as in ‘answerable, responsible’ (OED2, a. 1).

Defending Poetry, p. 164

 But, because I had seen the poem in the recently published Penguin Selected, I entered this note to my reading:

[1] GHCP is faithful to the original lineation in Mercian Hymns, whereas the Penguin Selected Poems (2006) relineates the poem, doing away with the hyphenation. Whether this is an authorised change or another of the many defects of that edition will be confirmed on publication of Hill’s projected Collected Poems in 2012.

This is how the poem appeared in the 2006 Penguin Selected:


While I might have managed a similar account of the poem using that typesetting, I think it unlikely to have been anywhere near as persuasive. The hyphenation of the poem’s key words not only suggested the reading in the first place, they underwrote it.

Now that BH is out, on re-reading this note of mine it occurred to me to verify my intuition about the poem. Sure enough, we have the hyphenated words restored:


So that’s that. The 2006 Penguin is rubbish. But you will notice one change in the setting from the original. It’s the preposition “of” in the last line of the second paragraph, which in the original had been at the end of the previous line.

In a verse poem, this would have created an enjambment, the prepositional phrase divided over the line break, so that the relation announced by “of” (of what?) is delayed in time (real time and metrical time). In the verse-paragraph poem the effect is not as strong, but it is there.

Is this change significant? Will it alter readings? That depends on who is reading, of course. Nothing comes to mind right now in terms of critical treatments of prepositions at the end of the line in Mercian Hymns, but there may be something out there. Hill himself is wont to read at the end of the line. In a recent Oxford lecture [“A deep dynastic wound,” 30 April 2013], he says (of Paradise Lost):

What I’ve called the “animus” of these lines […] is an energy at once etymological and rhythmic, the rhythm achieved through the masterly enjambment of grammatical clauses across the line endings.

So, for now, here’s a little catalogue of changed line endings in the Mercian Hymns of Broken Hierarchies, vs. the original [forgetting, justly, the 2006 Penguin], for reference by future variorumists. All the other lines end exactly as they did before, including the frequent hyphenation of the kind displayed in XI.

Poem NumberEnding in MC/GHCPEnding in BH
IIIwhere a | bonfire of beer-crateswhere | a bonfire of beer-crates
VDreamy, smug-faced, | sick on outingsDreamy, smug- | faced, sick on outings
XIscrapers of | salt-pansscrapers | of salt-pans
XIIIout of | England's wellout | of England's well.
XIVand | a snail sugaredand a | snail sugared
XXIby lakesides where | all might fancyby lakesides | where all might fancy
XXVSparks had furred | its low roof.Sparks had furred its | low roof.

Now, perhaps you still think this small beans: a preposition here, an article there. For an illustration of the difference this can make, have a look at how the final poem of the collection fares in the 2006 Penguin (it is restored in BH):

Here is the original:


And here the mutilated, unmutilated Penguin typeset:


Could you make the same reading out of those two poems? Would you?

*[I am reminded by Tom Day’s article in PN Review that Hill calls these poems “versets of rhythmical prose”, and so not “prose poems” exactly].


  • (via FB) Really interesting. Had never before considered how typesetting impacts the reading of poetry. With dead authors, this seems to create real issues regarding authorial intentionality. How can we know which, if any, version is authentic, rather than the modified creation of the publisher / typesetter?

  • (via FB) Thanks for posting this, David. Fascinating changes. The last ‘line’ of Mercian Hymns has to be ‘red mud’ on its own, surely. Has always reminded me of the Beddoes fragment ‘Like the red outline of beginning Adam’ – playing on the etymology of ‘Adam’.

  • (via FB) Have always been suspicious of “Selected Poems”. They seem to be a way of appeasing the “mass” & “populist” poetry-reader who just wants a mere ‘greatest hits’ sample rather than focusing their full attention and time upon the authoritative sequencing and artistry of single collections within a corpus. Doesn’t surprise me at all that the Penguin GH Selected has been careless in typesetting. It goes with the “populist” terrain of “Selected Poems” in general.

  • NLiu wrote:

    What do you think of Nicholas Howe’s suggestion that the true lineation of the Hymns is achieved through caesura, as in Anglo-Saxon poetry?

    ‘Viewed in this way, Hill’s poems establish their lineation . . . through their use of the caesura. If one reads aloud [Hymn XI] . . . one can hear the echoing shape of an Old English line, two phrases joined by a telling pause in the middle: “Coins handsome as Nero’s; / of good substance and weight” or “Offa Rex resonant in silver, / and the names of his moneyers”. In many of the Mercian Hymns, Hill marks the caesura with a colon or semicolon, our conventional signs for punctuating a significant, audible pause between two linked and balanced syntactical units like the a and b halves of an Old English line.’ (1)

    One logical implication of Howe’s analysis (though he doesn’t state it) is that the visual lineation really is just an accident of type-setting, much as the visual lineation of OE manuscripts is an accident of scribesmanship.

    To me, the more interesting effect is not that of any particular line-break, but the general awareness that the organisation of words which the reader sees is a partitioning imposed by the exigencies of publication, not a reflection of the poem’s ideal arrangement. Christopher Ricks has observed that Hill’s hyphens, when used to form compound words, serve as markers both of intent to unify and of impotence to realise that intent. (2) I think that applies as well to the hyphens which, in Mercian Hymns, cleave apart/together those words which stretch between lines. The poem requires an at-one-ment between “what can be printedly read” and “what can be said” (Ricks’s words, used in the context of Hill’s parentheses), but this at-one-ment can only ever occur somewhere beyond the printed page. The visual lineation is a method by which the poem gestures toward an impossible completion.

    (1) “Praise and Lament: The Afterlife of Old English Poetry in Auden, Hill, and Gunn”. Words and Works: Studies in Medieval English Language and Literature in Honour of Fred C. Robinson. Eds. Peter S. Baker and Nicholas Howe. Toronto: University of Toronto Press, 1998.

    (2) The Force of Poetry. Guernsey: Oxford University Press, 1987.

  • D-AW wrote:

    Two relevant quotations – thanks for those. I guess I don’t hear the Mercian Hymns quite like Howe generally, though I’ll grant these two examples in particular support his view. There are lots of colons and semicolons and heavy commas in MC, but I think only some achieve the caesura effect. Consider I: “the summer hermitage in Holy Cross: contractor to the desirable new estates: saltmaster: money changer: commissioner of oaths” etc. On the other hand, consider the line breaks in, e.g., II: “curt | graffito”; “A specious | gift”. To me these and many others in MC function much more like Hill’s metrical enjambments in other collections (e.g. “Imagine it great | unavoidable work” … “God | Is distant, difficult”), or like the enjambments he discusses in other English poetry (a feature of Hopkinsian “bidding”, I think). The number of restored hyphenated end-words in BH encourages me in this view.

  • (via FB) good stuff. reminds me that tom day wrote this piece in pn review a few years back about line-breaks in mercian hymns (paywall) >>>

  • Oh yeah.. I must have read that at the time. Too bad PNR couldn’t fit the quoted line from the Selected onto one line! Kind of spoils the effect [while also underlining the importance of lineation]. Here’s a pic:

    eh… it’s not that bad….

  • D-AW wrote:

    I think Tom takes the Selected too seriously in that piece, but this is well put:

    A crucial distinction between poetry and prose, it could be argued, is that the latter lacks the precise spatial coordinates of the former. There isn’t a meaningful discrepancy between, say, a large-print edition of Pride and Prejudice which cuts the first line off at ‘It is a truth universally acknowl-‘, budging ‘edged, that a single man in possession of a’ down to the second and ‘good fortune must be in want of a wife’ onto the third, and an edition which fits the whole sentence onto one line, because in prose the sentence or clause is the unit of meaning, not the line.

Leave a Reply