Wordsworth · three printings · his sister's journal
A Host of Dancing Daffodills
In 1807 Wordsworth printed his daffodil poem with no title, in 18 lines, spelling the flower two ways; by 1815 the host was golden, not dancing, and a new stanza made it 24 lines. Set beside his sister Dorothy's journal entry for 15 April 1802, as William Knight printed it in 1897, the poem shares more of its content words with that passage than with any stretch of the same length in his text of her Grasmere Journal outside that day's entry, though only just, and the revised text widens the gap. The first five lines of the last stanza share none of her content words; they hold the couplet Wordsworth singled out, which the 1851 Memoirs gave to his wife Mary and which Coleridge quoted in 1817 as an instance of 'thoughts and images too great for the subject'.
The title of this page is the fourth line of the poem as it was first printed, in 1807: A host of dancing Daffodills;
The same printing spells the flower differently in its last line, And dances with the Daffodils.
Three printings from Wordsworth's lifetime are set out below, each typed from the page images of one scanned copy, with every line linked to the page it is printed on: the first, in 1807; the first with the added stanza, in 1815; and one from 1849. Pick a printing to see what changed since the one before.
Three printings
The poem as printed in 1807, 1815 and 1849
The printings load from texts.json; if they do not appear, the page images are linked in the Sources below.
Choose a printing.
a changed worda new spelling of the same worda capital loweredpunctuationan added linea word whose stem is in Dorothy's passage (press it)
The diff is computed in your browser from the three transcriptions: lines are paired first, by the share of words they have in common, then the words and marks inside each pair are compared. A word here is a run of letters with any apostrophe or hyphen inside it, so gaz'd and Out-did are one word each. A respelling is a pair that is the same once 'd is read as ed and case, apostrophes, hyphens and doubled letters are set aside (Daffodills and Daffodils, gaz'd and gazed); anything else is a changed word.
From 1807 to 1815 the poem grew from 18 lines to 24, and from 117 words to 153. Four words or pairs of words were changed: dancing
became golden
, Along
became Beside
, Ten thousand
became Fluttering and
, and laughing
became jocund
. One stanza of six lines and 36 words was added, and the ten thousand moved into it: Ten thousand saw I at a glance,
Four words were respelled (Daffodills, Outdid, and gaz'd twice) and one comma was added. From 1815 to 1849 not a line was added: one word changed, shew to show, seven capitals were lowered, and six punctuation marks were added, dropped or replaced.
None of the three printings gives the poem a title. In 1807 it is number 7 of the group Moods of my own Mind, in the second volume of Poems, in Two Volumes; in 1815 it is number XIII of the Poems of the Imagination, in the first volume of Poems; in 1849 it is number X in the same class, and the running head over it reads I WANDERED LONELY.
Beside its last line the 1849 page prints a date, 1804.
, the year Wordsworth also gives in the note on the poem he dictated to Isabella Fenwick. Knight's edition of 1896 heads the poem Composed 1804.
; this page has fetched no other evidence of when it was written.
Wordsworth did have a name for it. Writing to Lady Beaumont in a letter headed Coleorton, May 21, 1807, as Alexander Grosart printed it in 1876, he says of a reader: I collect this from her having singled out the two little poems, 'The Daffodils,' and 'The Rock crowned with Snowdrops.'
In a letter to his friend Wrangham, headed only Grasmere, Nov. 4.
, he tells how a man called Butler happened to glance his eye upon the very poem of 'The Daffodils.'
That is Grosart's text; the 1851 Memoirs prints the same words as to glance his eye upon the very poem of the "Daffodils."
And the Fenwick note begins Town-End, 1804. 'The Daffodils.'
The two anthologies checked for this page use his name: Francis Turner Palgrave's Golden Treasury (1861) heads it THE DAFFODILS
, and Arthur Quiller-Couch's Oxford Book of English Verse (first edition 1900; the text checked is the impression of 1931) heads it Daffodils
. Both begin I wander'd lonely as a cloud
, a spelling none of the three Wordsworth printings uses.
An instrument of the same kind runs one stanza of FitzGerald's Rubáiyát through five editions in A Leg of Mutton. This page's question is a different one: where the poem's words came from.
Her entry, and his poem
Dorothy Wordsworth kept a journal at Grasmere, and William Knight printed it, with omissions, in 1897. Her entry headed Thursday, 15th.
, in April 1802, describes a walk by a lake in a furious wind, and a passage of 210 words in it describes daffodils. Her we includes her brother: the same entry ends that evening at Dobson's, where William was sitting by a good fire when I came downstairs.
The note Wordsworth dictated to Isabella Fenwick, as the 1851 Memoirs prints it, says the daffodils grew, and still grow, on the margin of Ullswater
.
The dismissal comes in two forms, and the first is old. The 1851 Memoirs, two pages before it prints this passage, draws a moral from her account behind another poem, Beggars: his poems are sometimes little more than poetical versions of her descriptions of the objects which she had seen; and he treated them as seen by himself.
The second runs the other way: two accounts of one windy walk by a lake will share words anyway, lake and trees and waves. Both can be tested with the same count. Both of them were there, so a shared word shows the two texts overlap; it cannot show who took what from whom, and this page does not say that either took anything.
There is a catch before any counting. The manuscript of the journal is not what is measured here. Knight's prefatory note says that where he left sentences or paragraphs out, the omission is indicated by means of asterisks
, and his text also breaks off inside entries with runs of dots, as it does just after this passage. The Memoirs had printed the passage in 1851 in different words: below Gowbarrow Park
where Knight has beyond Gowbarrow Park
, they looked so gay and glancing.
where Knight has they looked so gay, ever glancing, ever changing.
The two printings are below, side by side, and the count runs on either. This page has not seen the manuscript and cannot say which is nearer to it.
Shared words, and every stretch of her journal outside that day's entry
As William Knight printed it in 1897
When we were in the woods beyond Gowbarrow Park we saw a few daffodils close to the water-side. We fancied that the sea had floated the seeds ashore, and that the little colony had so sprung up. But as we went along there were more and yet more; and at last, under the boughs of the trees, we saw that there was a long belt of them along the shore, about the breadth of a country turnpike road. I never saw daffodils so beautiful. They grew among the mossy stones about and above them; some rested their heads upon these stones, as on a pillow, for weariness; and the rest tossed and reeled and danced, and seemed as if they verily laughed with the wind, that blew upon them over the lake; they looked so gay, ever glancing, ever changing. This wind blew directly over the lake to them. There was here and there a little knot, and a few stragglers higher up; but they were so few as not to disturb the simplicity, unity, and life of that one busy highway. We rested again and again. The bays were stormy, and we heard the waves at different distances, and in the middle of the water, like the sea
As the Memoirs printed it in 1851
When we were in the woods below Gowbarrow Park, we saw a few daffodils close to the water-side. . . As we went along, there were more and yet more; and at last, under the boughs of the trees, we saw there was a long belt of them along the shore. I never saw daffodils so beautiful. They grew among the mossy stones about them: some rested their heads on these stones as on a pillow; the rest tossed, and reeled, and danced, and seemed as if they verily laughed with the wind, they looked so gay and glancing.
The number at the end of each line is how many of the shared stems it holds. Lines with none are in blue.
The shared stems, with the poem's word and hers. Press one to strike it as a coincidence; press again to restore it.
Loading the journal.
The rule, in full
A word is a letter followed by letters or apostrophes; a hyphen splits a word, so never-ending counts as never and ending. A word is skipped if it has one letter or is in the stop list below. The rest, which this page calls content words, are cut to a stem: lower case; a final 's dropped; 'd read as ed; daffodill read as daffodil; the first of the endings -ing, -ed, -es, -s, -ly removed if at least three letters are left; then a final e removed if at least three letters are left. So dancing, danced, dances and dance are all danc, and laughing and laughed are both laugh. Two texts share a stem if each has at least one word with it.
The stop list is the function words of English plus o'er, oft, upon, thus, yet, shall, may, might, must, one, two, us, and four abbreviations the journal uses (Mr., Mrs., Wm., C.). The full list is STOP_BASE in engine.mjs.
The journal is Knight's text of the four Grasmere sections of his volume (from 14 May 1800 to 11 January 1803), with his footnotes, footnote numbers, section headings, three bracketed insertions of his own and the rows of asterisks that mark his omissions taken out: grasmere-knight-1897.txt. Knight's whole entry for that Thursday is set aside, and a window is every run of consecutive words as long as the passage being measured, before or after that entry, starting one word after the last. Windows that reach a score and overlap the one before (sharing at least one word with it) are counted as one place.
Under that rule the 1807 text shares 11 stems with Knight's passage: daffodils, dancing, waves, lake, trees, along, gay, laughing, saw, floats and little. There are 39,274 windows of 210 words in Knight's text of the Grasmere Journal outside that day's entry, and the best of them shares 10. The 1815 text shares 15 stems with the passage, and the best other window shares 12. So the daffodil passage does come first for both, by one stem in 1807 and by three in 1815.
The margin is thin, and the instrument lets you push on it. Two of the 1807 matches are plain accidents of the rule: the cloud floats
and the sea had floated the seeds ashore
; the poet little thought
and the daffodils were the little colony
. Strike those two and the 1807 passage scores 9, level with 91 other windows in 3 separate places. The 1815 passage keeps first place, 13 against 11. Strike instead every stem of the poem found in at least a quarter of all the windows (little, hill, tree and lake: little turns up in more than half of them) and the margins are 8 against 7 for 1807 and 12 against 9 for 1815; lower the bar to a fifth, so that saw goes too, and the 1807 passage is level with 30 other windows. The 1849 text gives the same counts as 1815 throughout.
On the Memoirs' 98 words the passage does not stand out for 1807 at all. The 1807 text shares 7 stems with it, and 220 of the 39,498 windows of 98 words in Knight's text outside that day's entry reach 7 or more. The 1815 text shares 10, and 6 of those windows reach 10, all in one place; none passes it. Five of Knight's matches for 1815 (floated, little, lake, waves, bays) come from words the Memoirs does not print.
The best rivals are worth reading, because they show what the count is up against. For 1807 the first of the 3 places that reach 10 opens on Sunday, June 1st, 1800, and runs into the next day's walk, where The waves round about the little Island seemed like a dance of spirits that rose out of the water
. For 1815 the first of the 2 places that reach 12 is a Thursday in 1802: Rydale vale was full of life and motion.
That window shares 12 stems with the poem, among them lake, waves, vale, hills, cloud, sparkling and twinkle. Her journal has other windy days by water. On this count the daffodil passage comes first, but not by much.
What the 1815 revision did
The revision moved the poem's words toward hers, and one step away. Against Knight's passage the 1815 text gains five shared stems the 1807 text did not have, bay, glance, heads, tossing and never (from never-ending, which you may well strike), and it loses one, laughing, which matched her laughed
. Of the new stanza's 21 content stems, 8 are also in her passage: along, bay, dance, glance, heads, never, saw and tossing. Its last line is Tossing their heads in sprightly dance.
Three of its four content words have a stem in one sentence of hers:
They grew among the mossy stones about and above them; some rested their heads upon these stones, as on a pillow, for weariness; and the rest tossed and reeled and danced, and seemed as if they verily laughed with the wind, that blew upon them over the lake; they looked so gay, ever glancing, ever changing.
This page cannot say why he made these changes, or whether the journal was in front of him when he did. It can say that all five of the new shared stems are in the stanza he added, and that the gap between the passage and every stretch of her journal outside that day's entry is three stems in the revised text and one in the first.
The two best lines
Now read the poem from the other end. The last stanza begins with five lines that share no stem with her passage at all, in Knight's text or the Memoirs': in 1807 they are lines 13 to 17, and in 1815 lines 19 to 23. Against Knight's text the line before them, What wealth the shew to me had brought:
, shares none either, so that is six lines in a row, the longest such run in either printing. Only the last line comes back to her words, dances and daffodils. The five lines are about memory:
For oft when on my couch I lie
1807, vol. II, p. 50
In vacant or in pensive mood,
They flash upon that inward eye
Which is the bliss of solitude,
And then my heart with pleasure fills,
Two of them have a history of their own, and it is a chain of documents, each saying less than the story usually told about them. Here are the links: first the letter, which has no year printed, then the others in order of date.
- Grasmere, Nov. 4 (no year printed)Wordsworth to Wrangham. Butler had looked at the poem and said
a fine morsel this for the Reviewers.
When this was told him, Wordsworth writes, he observed thatthere were two lines in that little poem which, if thoroughly felt, would annihilate nine-tenths of the reviews of the kingdom, as they would find no readers; the lines I alluded to were these:
and he quotesThey flash upon that inward eye Which is the bliss of solitude.
The letter names the two lines. It says nothing of who wrote them. - 1815In the Poems of 1815 an asterisk on
They flash upon that inward eye
keys a footnote:The subject of these Stanzas is rather an elementary feeling and simple impression (approaching to the nature of an ocular spectrum) upon the imaginative faculty, than an exertion of it.
- 1817Samuel Taylor Coleridge, in the second volume of the Biographia Literaria, chapter XXII, lists defects in Wordsworth's poems. The last is
Fifth and last; thoughts and images too great for the subject. This is an approximation to what might be called mental bombast, as distinguished from verbal
. His first instance is the couplet:But if we describe this in such lines, as
They flash upon that inward eye, Which is the bliss of solitude!
in what words shall we describe the joy of retrospection, when the images and virtuous actions of a whole well-spent life, pass before that conscience which is indeed the inward eye
? And then:Assuredly we seem to sink most abruptly, not to say burlesquely, and almost as in a medly from this couplet to
the poem's last two lines. The 1817 printing gives the reference asVol. I. p. 320.
; the poem is on pages 328 and 329 of the 1815 volume, and the text in Project Gutenberg readsVol. I. p. 328.
His next example beginsThe second instance is from Vol. II. page 12
. - Dictated; the copy was finished in August 1843The note Wordsworth dictated to Isabella Fenwick, in the manuscript copy Grosart printed:
Town-End, 1804. 'The Daffodils.' The two best lines in it are by Mary.
It does not say which two. Grosart prints a pencil note from the opposite page:Mrs. Wordsworth--but which?
His daughter Dora Quillinan, who wrote out about half of that copy, wrote at its end:To dearest Miss Fenwick are we obliged for these Notes, every word of which was taken down by her kind pen from my father's dictation.
The copy was finished on thetwenty-fifth day of August 1843
. Mary is Mary Wordsworth: in the autobiographical memoranda printed in the Memoirs he writes,In the year 1802 I married Mary Hutchinson
. - 1851Christopher Wordsworth's Memoirs of William Wordsworth prints both documents. Under the couplet in the Wrangham letter it puts a footnote:
These two lines were written by Mrs. Wordsworth. See below, p. 190.
On that page is its text of the Fenwick note:The two best lines in it are by M. W.
The footnote gives no other evidence. This is the join: the letter supplies which two lines, the note supplies whose. - 1876 and 1896Later editors repeat the join as fact. Grosart, printing the Wrangham letter, adds
These two lines were composed by Mrs. Wordsworth
and cites the Memoirs; Knight's edition of the poems says the same,These two lines were composed by Mrs. Wordsworth.
So the attribution is not something Wordsworth is recorded saying in any one of the documents fetched for this page. In one he picks out two lines; in the Fenwick note, in a copy finished in 1843, he says the two best lines are Mary's; the 1851 Memoirs puts the two together. The join is reasonable, and it may be right. It is still a join.
And it lands on the same five lines. The couplet Wordsworth said would annihilate nine-tenths of the reviews, the couplet the Memoirs gives to Mary, and the couplet Coleridge chose to show thoughts too great for their subject all sit in the stretch of the poem with none of Dorothy's content words. The scene is where her vocabulary is: the waves, the dancing, the lake, and after 1815 the tossing heads. The lines about what the scene does afterwards, in memory, have none of it. That is a fact about the words. It is not evidence about who wrote which lines, and this page does not offer it as such.
The check
The instruments compute every count in your browser with engine.mjs, from three files beside it: texts.json (the three printings, typed from their page images, and the Memoirs' text of the passage), grasmere-knight-1897.txt (Knight's text of the journal) and quotes.json (every quotation, with its source). The counts in the prose are the same computations, written down. One program you can run recomputes them a second way and holds every figure to the sentence that states it:
curl -sO https://artwaste.land/checks/verify-a-host-of-dancing-daffodills.mjs node verify-a-host-of-dancing-daffodills.mjs
It downloads this page and those four files and:
- diffs the three printings with its own code (a different tokenizer, line pairing and edit distance from the engine's) and checks the line, word, stanza and change counts;
- re-applies the stem rule with its own code and scores every window by brute force, checking the shared stems, the 39,274 and 39,498 windows, the best rivals, the strikes and the quarter and fifth thresholds;
- finds every figure in the prose in its own sentence, and fails on any number on the page that it has not checked or declared;
- checks every quotation against its passage in
quotes.json, and the description copies in the page head against the visible summary.
With --mutate it plants wrong figures in a copy of the page and must fail. The free choices, named: the stem rule and the stop list (crude on purpose, so you can see every match and strike the ones you think are chance); windows as long as the passage, one word apart, with Knight's whole entry for that day set aside; overlapping windows counted as one place (counted as unbroken runs instead, the 1807 ties are five places, not three); Knight's text as the journal, since the Memoirs prints only extracts; the line-pairing threshold of half the words; the passage's limits, from the woods to Knight's next run of dots. What it cannot check: that the transcriptions match the page images (each line of the poem links to its page; read it), and that each source says what its excerpt says (the links are below). The counts are for Knight's 1897 text and the Memoirs' 1851 text, not for Dorothy's manuscript.