of Eliot’s Selected Essays, see Donald Gallup, T. S. Eliot: A Bibliography, rev. ed.
(New York: Harcourt, 1969), 47, A21.a.
53. The Baudelaire essay stands conspicuously as the first in a section treating
modern authors, a position it still occupies today. The sentence by Eliot
quoted here is found in his Selected Essays (New York: Harcourt, 1950), 380.
54. Cleanth Brooks, “The Waste Land: Critique of the Myth,” in Modern Poetry and the Great Tradition (Chapel Hill: University of North Carolina Press,
1939), 136–172; “unified whole” appears on 136, the references to Weston
and the Baudelaire essay on 137.
55. T. S. Eliot, “The Frontiers of Criticism,” in On Poetry and Poets (London: Faber, 1957), 109–110, ellipsis mine. Ezra Pound, writing in 1939, remembered the genesis of the notes in similar terms, and he highlighted their
e¤ect on the poem’s reception: “The bearing of this poem was not over-
estimated, nevertheless the immediate reception of it even by second rate
reviewers was due to the purely fortuitous publication of the notes, and
not to the text itself. Liveright wanted a longer volume and the notes were
the only available unpublished matter.” Ezra Pound, “T. S. Eliot,” in Frances
Stelo¤ and Kay Steele, eds., We Moderns (New York: Gotham Book Mart,
1939), 24.
56. T. S. Eliot, “The Art of Poetry, I: T. S. Eliot” (see n. 15), 96, 105.
57. Jonathan Culler, Structuralist Poetics (Ithaca, N.Y.: Cornell University Press, 1975).
58. Wayne Koestenbaum, “The Waste Land: T. S. Eliot’s and Ezra Pound’s Col-
laboration on Hysteria,” in Double Talk: The Erotics of Male Literary Collabo-
ration (London: Routledge, 1989), 112, 113.
59. Christine Froula, “Corpse, Monument, Hypocrite Lecteur: Text and Transference in the Reception of The Waste Land, ” Text: An Interdisciplinary Annual of Textual Studies 9 (1996): 304–314.
A Note on the Text
s i n c e t h e s t o r y o f The Waste Land’ s publication has already been recounted in the Introduction, we can turn directly to the implications of
that publication history in assessing the poem’s text. The Waste Land ap-
peared in three more or less contemporaneous versions: first, on 16 Octo-
ber, without notes, in the October issue of the Criterion (the English journal edited by Eliot himself ); then around 20 October, again without notes, in
the November issue of the Dial (an American journal co-owned by Scofield
Thayer and James Sibley Watson, Jr.); and finally around 1 December, now
with notes, in a small book issued by the American publisher Boni and
Liveright.
Recounting the publication in this way seemingly assigns priority to
the Criterion’s version of the poem—the first to be published and the one
that Eliot could most directly supervise. But in fact the situation was more
complicated. The first manuscript which Eliot sent to press was the one
which, on 19 July, he posted to John Quinn to consign to Liveright when
he signed the revised contract that Quinn was then drawing up. (“I only
hope the printers are not allowed to bitch the punctuation and the spacing,
as that is very important for the sense,” Eliot had added [ LOTSE, 547].)
True, the manuscript sent to Quinn still lacked the notes. Those Eliot finally
completed and forwarded to Liveright sometime before 15 August, the day
he told James Sibley Watson: “I suppose that the poem is now going to
4 5
4 6
a n o t e o n t h e t e x t
press” ( LOTSE, 560). The combined manuscript, text and notes, underwent
a brisk production process, for only one month later, on 15 September,
Eliot could tell Ezra Pound his assessment of Liveright’s work: “Liveright’s
proof is excellent” ( LOTSE, 570). Thus, though Liveright’s version of The Waste Land was the last to be published, it was the first to be produced.
Its production process was so swift because, throughout August and into
the first week of September, Liveright assumed he would be the poem’s
only publisher in the United States, and by contract he was obliged to is-
sue it in his autumn list, or in October or November at the latest. It was
only in the very last days of August that Liveright received Watson’s proposal
that the poem first be published (without notes) in the Dial, then be issued as a book (with notes) by Liveright. Moreover, the letter of agreement that
sanctioned this arrangement was not signed till 7 September.
The date is important. Normally during this period, transatlantic mail
between New York and London required nine days. In e¤ect, the contract
between the Dial and Liveright was signed so late (7 September) that the
Dial did not have time to request a setting copy from Eliot. To request a
setting copy (nine days), await its arrival (nine days), produce and send
o¤ a proof of it (at least nine days), and then receive and execute corrections
(another nine days) would have required a minimum of thirty-six days.1
In fact, the Dial’s publication of the poem, about 20 October 1922, took
place only forty-three days after the journal finalized its letter of agree-
ment with Liveright on 7 September. The Dial, in other words, lacked the
time to request, or produce a proof based on, an authoritative setting text
of the poem. It had no choice except to use a version of the text furnished
by Liveright. True, Eliot had sent a manuscript of the text to James Sibley
Watson when he was still in Paris, back in mid-August; but that was a
reading copy, not one meant to serve as setting copy.2 Moreover, in the co-
pious records of the Dial, no document mentions a setting copy furnished
by Eliot or proofs overseen by him. Instead, the poem’s hurried production
at the Dial and the lack of any documents that register Eliot’s involvement combine to confirm evidence gleaned by collating the two texts: the Dial
text is derived from that of Liveright and has no independent authority.
Its variants are of interest to the extent that they represent a well-intentioned
typesetter’s e¤orts to make sense of the Liveright text, but they have no
independent authority. Eliot never saw them until publication and was
never consulted about them.
a n o t e o n t h e t e x t
4 7
On the other side of the Atlantic, Eliot’s involvement was more active
in the production of the Criterion’s version of the text. But it was also decid-edly negative. To Richard Cobden-Sanderson, the publisher of the Criterion,
Eliot wrote on 27 September (twelve days after he had told Pound that
Liveright’s proof was “excellent”), “I am also sending you the manuscript
and the proof of the first part of my poem, so that you may have a record
of the undesired alterations made by the printers” ( LOTSE, 574).3 And on
3 October, Eliot wrote Cobden-Sanderson again: “You will see that I am
enclosing the corrected proof of the rest of The Waste Land. I shall ring
you up tomorrow morning at about eleven and will explain why I have
done so” ( LOTSE, 576). Eliot’s consternation is palpable in these comments, and readily understandable if we collate the Criterion text with that of Boni and Liveright. Consider a sample passage, the first two verse-paragraphs
which form the opening to part III and contain thirty lines (ll. 173??
?202).
The Criterion version introduces nine variants in spelling, spacing, punctuation, and font: in four places it adds commas (ll. 187, 188, 200, 201), in
another it changes the font and in yet another it adds a blank line (202,
198–199), in still two more it alters punctuation (180, 192), and in one
last it emends a spelling (“gashouse” to “gas-house” at 190). All minor
changes, it is true, but eight of the nine are changes of precisely the sort
that Eliot had hoped to avoid (“the punctuation and the spacing . . . very
important for the sense” [ LOTSE, 547]). The result is an overall change
in the flow and pace of the text. The printer who set the Criterion text was oªcious, trying to make an unruly Waste Land conform much more closely
to conventional usage. In comparison, the Dial printing of the same pas-
sage introduces only two changes, minor alterations in punctuation which
are merely matters of house style (the period, or full stop, is eliminated
after “Mrs.” in lines 198 and 199, consistent with similar changes in lines
57 and 209).4
Nor was it just in 1922 that Eliot considered the Liveright printing
“excellent” and dismissed the Criterion one as filled with “undesired alterations.” In 1923, when the poem was issued in book form for the first time
in England by the Hogarth Press, Eliot chose the Liveright text as setting
copy. Still more tellingly, two years later, when the poem was collected for
the first time in Eliot’s Poems, 1909–1925 (London: Faber and Gwyer, 1925),
he again chose the Liveright as setting copy. That he preferred it over the
Dial, Criterion, and Hogarth printings is beyond doubt, and the present
4 8
a n o t e o n t h e t e x t
edition follows him in adopting the Liveright as its base text. (It was also
in this 1925 edition, it should be noted, that Eliot added the dedication to
Ezra Pound which has appeared with the poem ever since, a dedication
he had first written in the inscribed copy of the Boni and Liveright edition
of the poem which he gave to Pound in 1923.)
Adopting the Boni and Liveright text (B) as setting copy for both the
1923 Hogarth (H) and the 1925 Faber edition of Poems, 1909–1925 (F),
as Eliot did, entailed an obvious though not insurmountable problem. It
meant that any corrections which he made in the 1923 Hogarth would
automatically disappear unless he actively intervened to make them a sec-
ond time in 1925. Or to put it di¤erently, any real or manifest error which
B contained would automatically recur not just once in H but again in F, unless Eliot actively noted and corrected it. How extensive was the problem? Not very. The Boni and Liveright text contained eight errors and one
potential variant, a¤ecting a total of seventeen lines. (The reason for the
discrepancy between these figures [nine and seventeen] is that two of the
errors in B recurred five times each; if we subtract the four “repeats” of
the two errors [eight “repeats”] from the seventeen lines, then our two
figures coincide.)
If we set aside questions of font changes for the moment, the eight
errors were:
1. l. 42 “Od’” instead of the correct “Öd’” or “Oed’”
2. l. 111 “tonight” instead of what was then current English usage,
“to-night”
3. l. 112 “Why do you never speak.” instead of the obviously correct “Why
do you never speak?”
4. l. 131 “‘What shall I do now? What shall I do?’” instead of the correct
reading, “‘What shall I do now? What shall I do?” (no closing quotation
mark)
5. ll. 141, 152, 165, 168, 169 missing apostrophes in the words “it’s”
6. ll. 149, 153, 154, 163, 164 missing apostrophes in the words “won’t” or
“don’t”
7. l. 161 “alright” instead of “all right”
8. l. 259 “O City city” instead of “O City City”
Of these eight errors, Eliot noted and corrected five in H (2, 3, 5, 6, and 8
above). But of course his corrections were automatically undone in F be-
a n o t e o n t h e t e x t
4 9
cause it reverted to B as setting text. This time, in F, he remembered to execute only two of the five corrections he had earlier made (2 and 6 above),
and he added one more (7 above), in e¤ect adding a sixth correction to B.
Further, in F he also altered the spelling of “aetherial” to the more com-
mon “aethereal” in line 415, even though “aetherial” had been his own
usage in the autograph and typescript fair copies which he had shown to
Ezra Pound ( TWL:AF, 78–79, 88–89). This edition, then, follows Eliot
in adopting B as setting text, admitting the six corrections which he made
in 1923 (H) and 1925 (F), admitting also the one alteration (“aethereal”) he made in 1925, and of course admitting the other, far more significant
alteration which he made in 1925, the addition of the dedication to Ezra
Pound.
An attentive reader will have noticed that two of B’s obvious errors (1,
4 above) still await attention. In the second of these, B reads, “‘What shall I do now? What shall I do?’” The oªcious printers of the Dial and the
Criterion both noticed the unnecessary closing quotation marks (or inverted commas) in this passage and deleted them; but the marks returned in H
and F, and Eliot failed to catch the mistake. Worse still, the error persisted in all subsequent printings. The closing quotation marks are not found
in the typescript which Eliot showed Pound in Paris ( TWL:AF, 12 and 18),
however, and its authority and common sense dictate their removal.
The same is true for the last of B’s obvious mistakes (1 above), the
ridiculous “Od’” instead of “Öd’” or “Oed’.” Although the mistake was
corrected in the Dial, that text has no authority. The mistake was not cor-
rected in the Criterion, nor in H or F. Yet Eliot was certainly an adept reader and writer of German, one who knew where German words required an
umlaut. Indeed, in 1922 (when The Waste Land was going to press) he
wrote three letters in German to Hermann Hesse and Ernst Curtius which
are uniformly correct in their use of umlauts.5 Moreover, when he first
wrote this passage in his own hand, an autograph addition made to the
typescript of part I which he showed to Ezra Pound, he wrote “Öd’” (see
TWL:AF, 6). The authority for this emendation, then, derives from Eliot’s
own usage in the Waste Land manuscripts.
It should be noted that the variant reading “Oed’” appears for the first
time in Eliot’s Collected Poems, 1909–1935 (1936). It has long been known
that Eliot made a number of corrections in a proof copy of Collected Poems,
1909–1935 which is held in the library of King’s College, Cambridge—
5 0
a n o t e o n t h e t e x t
corrections which, for reasons that are not known, were not entered into
the text.6 The authority of that edition, therefore, is troubled and cannot
be admitted as an authority for this edition. The spelling “Oed’” is not
the one that Eliot himself used when he wrote the poem, and it may be
no more than the expedient of a typesetter who had no umlauts at hand.
(It has stood in all editions ever since.) But the question of the auth
ority
which lies behind Collected Poems, 1909–1935 raises one other, more impor-
tant question.
Line 428 of the Boni and Liveright edition reads, “Quando fiam ceu
chelidon—O swallow swallow.” The text reads the same in every early print-
ing: in the Dial, in the Criterion, in the 1923 Hogarth, and in the 1925
Faber edition of Poems, 1909–1925. It also reads that way in the 1932 American edition of Poems, 1909–1925. Only in 1936, in Collected Poems, 1909–
1935, does the text suddenly undergo a change, with the first words now
reading: “Quando fiam uti chelidon. ” But the authority of that edition is
deeply suspect, as we have already seen. Moreover, there can be no doubt
whatever about which version of this passage Eliot had in mind when he
wrote the poem: in both his autograph fair copy of part V and the typescript
fair copy of it which he prepared for Ezra Pound while he was in Paris in
early 1922, Eliot unequivocally wrote and typed “ceu chelidon, ” not “uti chelidon” (see TWL:AF, 80–81, 88–89).
Further, there is something fussy, even a bit pedantic, about this altera-
tion. While the change from ceu to uti makes no di¤erence in the passage’s meaning, the latter is the more widely accepted scholarly reading of the
Latin text.7 A similar change takes place many years later in the 1962 Mar-
dersteig edition of The Waste Land. Whereas line 202 in all the early edi-
tions reads, “Et O ces voix d’enfants, chantant dans la coupole! ” the Mardersteig edition, which Eliot himself supervised, emends the punctuation to
read: “Et, o ces voix d’enfants chantant dans la coupole.” This reading brings
the quotation into accord with the published readings of Verlaine’s text
but out of accord with the poet’s ear.8 Much the same takes place with the
alteration of ceu to uti, an alteration which has behind it only the dubious authority of Collected Poems, 1909–1935, even though it has been followed
by all later editions. If it was Eliot’s change, then it was Eliot acting as the
oªcious custodian of his monument, and acting against the poet’s ear.
Two more, quite minor changes must be noted, both in the poem’s
notes. The introductory headnote in both the Boni and Liveright and the