skip to Main Content

Fluence (1): Part 2

Fluences is a section of troutswirl devoted to studying haiku, and haiku-like work, by 20th and 21st century western poets. Each installment will take a closer look at a poem, or a group of poems, by a poet who has either dabbled in haiku, been influenced by haiku, or whose work has had an influence, in some way or another, on 20th and 21st century English-language haiku.

Fluences is overseen by Nick Avis.

Fluence (1): Part 2

BY Nick Avis

Ezra Pound and In a Station of the Metro — Part Two

                                In a Station of the Metro

                                The apparition of these faces in the crowd;
                                Petals on a wet, black bough.

an exercise in fabricated metaphor

                The metaphors are
                gone, and so is my faith . . .
                sun over a moor

                     Nakamura Kusatao (24)

The fabrication of an image is achieved through appropriation and/or use of the imagination. There are three perspectives regarding fabricated or imagined images in haiku: they must be real or actually have happened; they can be fabricated but still must be capable of having happened or being real; reality is not necessary.

As for metaphor, the haiku community has been led to believe over the years that the use of metaphor in haiku, like all western poetic devices, is inappropriate if not forbidden.

Consider the following haiku by Basho (1644-1694) from his haibun, Oku-no-Hosomichi, The Narrow Road to the Deep North (1689 to 1691). (25) It is one of Basho’s best known and best poems:

summer grasses
where stalwart soldiers
once dreamed a dream

Makoto Ueda (26)

summer grasses—
traces of dreams
of ancient warriors

Haruo Shirane

natsukusu ya tsuwamonodomo ga yume no ato (27)

The adjectives “stalwart” and “ancient” are not found in the original and are added by the translators, yet both are implied in the prose preceding the haiku. “Stalwart,” meaning brave or resolute (Webster’s), though accurate, seems redundant. “Ancient” reflects the fact that Basho was thinking of the distant past, although it does not add very much to the poem.

According to Ueda and Shirane, “warrior” is the literal translation for tsuwamono to which the plural suffix domo is added. (28) The word yume (dream or dreams in the above translations) is singular and allows the translator to pluralize it although the singular would be the norm. (29)

There are two words in the original Japanese that have multiple meanings, a technique Basho often employed. According to Shirane, ato can mean site, aftermath, trace or track; and yume, can mean dream, ambition or glory. (30)

These multiple meanings can rarely be translated with one word in English and the resulting variations in the translations can be substantial. As a consequence, the variations in the interpretation of the poem can also be substantial.

Shirane translates ato as “traces;” Keene, as “aftermath” (31); Stryk, as “remains” (32); Earl Miner, as “vestiges” (33). Ueda, using the word “where,” interprets ato as place, the most benign of its meanings and the most literal. All of these translations, except Ueda’s, tend to be metaphorical.

The English words used in translation also have multiple meanings and each one brings something different to the poem. “Traces,” “vestiges” and “remains” are or can be synonymous.

In addition to its obvious meanings “traces” also means: “The path or way which anything takes.” (OED) This suggests fate or simply the way things are. It can be metaphorical in the sense that the soldiers’ dreams led them, inevitably, along a path of destruction.

“Aftermath” means: “the crop of grass which springs up after the mowing in early summer” (OED); “a result or consequence, esp. an unpleasant one.” (Webster’s) The first meaning is metaphorical and the second one implies that the soldiers’ dream was the cause of the tragic outcome.

“Remains” can also mean: “Those left, surviving, or remaining out of a number of persons” (OED); “to continue to exist, endure, persist; traces of the past; a dead body, corpse.” (Webster’s) These multiple meanings add depth to the interpretation of the poem, whether metaphorical or otherwise.

Translating yume as dream or dreams seems universal and by implication a soldier’s dream would include in most cultures glory and ambition. In Ueda’s translation the soldiers all have the same dream; in Shirane’s, each has his own dream. Mizuho (1876-1955), whom Ueda calls one of Basho’s “interpreters,” says “It is as though each soldier’s dream were lingering on each blade of grass.” (34)

The word “apparition” in Pound’s Metro Poem also has multiple meanings; and the multiplicity of meanings in both poems enriches them.

Two interesting observations regarding the season word are that the leaves of the summer grasses “are scorched at their tips under the flaming sun,” and “Basho’s summer grass is ‘warm’ with blood.” The summer grasses would be thick and deep. (35)

An “ancient battlefield is a sacred place . . . a kind of purgatory . . . where the souls of the slain soldiers, still retaining their anger and resentment, utter war cries day and night.” This is a fairly common theme in Noh drama. (36)

Basho’s haiku alludes to a famous battle at Takadachi castle and the tragic history leading up to its destruction (c. 1190.) Basho’s prose speaks of the fleeting nature of (military) glory; how it vanishes in the space of a dream; how it quickly became this grass or this grass is all that remains of it or this grass now covers it, depending on the translation. (37) He also alludes to these lines by Tu Fu: “Countries may fall, but their rivers and mountains remain. When spring comes to the ruined castle, the grass is green again.” (38)

In the prose preceding the poem Basho also alludes to a Noh play in which a man dreamed a lifetime of glory and defeat while taking a nap before supper. (39)

These literary and cultural allusions, which necessarily result in some appropriation, do not raise any concerns with Basho scholars, translators or interpreters. Such allusions are considered to add depth to a haiku and form part of what Haruo Shirane calls “cultural memory.”

Knowledge of such allusions is not always essential to interpret a haiku although it can be, but it is essential for a fuller understanding of it. Allusion is a key element or technique in many of Basho’s haiku and has been part of Japanese haiku ever since. In English haiku it is generally overlooked.

As for the soldiers’ dreams, Basho lived during a period of peace in Japan and he lived, by all accounts, a very peaceful life. (40) At best he can only imagine the battle, let alone the dreams of the soldiers who fought in it. Having never gone to war and never been in or seen one himself, he cannot even relate to the experience.

The image is rather abstract and can only reside in the imagination although most would have some appreciation of what is meant by soldiers’ dreams. No doubt in Basho’s time associating soldiers’ dreams with ambition and glory was firmly rooted in the cultural memory. Why else would the word yume mean all three? Today most cultures would make the same associations Basho did although there would likely be a great deal more ambivalence concerning the glory of war and soldiers’ dreams would have to include nightmares. Still, the soldiers’ dream of ambition and glory is more of an idea than a concrete image however well or universally understood.

Basho appropriated the image of grasses (surrounding ancient ruins) from a poem by Tu Fu, even though this is exactly what he saw, and he abstracted the idea of soldiers’ dreams from “cultural memory.” The interpretations of Tu Fu’s and Basho’s poems, on one level, are also very similar and metaphorical: the impermanence and futility of human endeavor, and the permanence and complete indifference of nature. Pound’s description of hokku as “one idea set on top of another” seems to apply quite well to this haiku by Basho.

Konishi (b. 1915), another of Basho’s interpreters, noted that: “For the first time in the history of haikai, an idea has become the subject of a poem.” (41) Ideation, generally thought of as forbidden in haiku, is at the root of fabricated images.

Shiki’s views regarding the use of imagination to create images in haiku seem to cover the full spectrum. He began by severely criticizing Basho for the poverty of his imagination and excessively praising Buson for the extraordinary range of his. He said “the fact that [Basho] discarded scenes which arise from imagination and are outside observation, as well as human affairs he had not experienced, shows that Basho’s realm was rather small.” Of Buson he said his imagination soars and “ranges beyond his country’s borders.” Later, he said that the imagination is “shallow” and that “many of the works which rely on [the imaginative method] are often bad.” (42)

Notwithstanding these contradictory statements, Shiki stressed realism and “direct, individual observation of the external world,” and his poems reflect this emphasis. (43) His observation that Buson used his imagination to create images much more so than Basho did is quite obvious from reading their poems, even in translation.

After Shiki, images from the imagination became more widespread and the more radical Modern Japanese haiku poets would often abandon reality altogether. Buson was, however, their predecessor, as these two examples clearly show:

About to bloom,
And exhale a rainbow,
The peony (44)

plum blossom’s scent—
has it risen so high?
a halo round the moon (45)

Bruce Ross says that Pound’s Metro Poem is “not really a haiku, which demands, for one thing, objectively real images.” (46) This reflects the traditional view, although the word “objectively” is redundant since the word “real” means something that has “an objective existence.” (OED)

Both images in Pound’s Metro Poem are not real based on Pound’s account in his Essay of what actually happened. He describes the faces as appearing one by one, but the poem has them all appearing at once in a singular apparition. This is consistent with the second image in which the petals are on the black bough, not landing on it. The second image is appropriated and likely imagined by Pound who took a year to come up with it. He never did say he actually saw the second image and it would not have concerned him in the least that he had fabricated and imagined either or both images.

Notwithstanding all the background information known about a poem, the (language of) the poem must speak for itself. Both images are capable of being real and there is no way of knowing from the language alone if they were made up. So both images are real in the poem itself. Appropriated images are usually easier to detect for obvious reasons.

In the following haiku by Buson, it is not possible to know that the poem is fabricated from the language alone:

this piercing cold—
in the bedroom, I have stepped
on my dead wife’s comb


Yet Buson’s wife outlived him by more than thirty years. (47) No one seriously suggests it is not a haiku for this reason except perhaps Blyth, who also says it is not a haiku because “Haiku has nothing to do with bedrooms or dead wives or treading on this or that thing with its emotional associations.” (48)

For the two images in Pound’s Metro poem to occur simultaneously in the Paris Metro; for the poem to reflect something that could have happened and does happen in the language of the poem, the poet must either recall his earlier experience of the real image of “Petals on a wet, black bough,” or imagine this image when he sees the faces in the crowd. Basho’s haiku is similar in this respect.

In Basho’s haiku, the poet visits the site of a battle, obviously a well known one, and sees the ruins and the summer grasses. Upon seeing them, he recalls (from cultural memory) and imagines the battle itself and the dreams of the soldiers who fought in it.

In all the translations referred to in this article, except Shirane’s because of the word “ancient,” the language of the translations also allows for the poet to be visiting the site of a battle he actually fought in.

Ueda states unequivocally that in The Narrow Road to the Deep North, Basho “changed the facts as he saw fit;” and that he did so “to present his theme more effectively.” He points out that Basho was not original in this and that “Ancient Japanese court diaries . . . were fictional to varying degrees” and “seemed to be based on the assumption that it was more important to record inner experience than outward events.” (49)

Shirane says that the season word in haiku can function as “a complex literary and cultural sign” that is “often highly fictional . . . .” (50)

Basho, like Pound, fabricated through appropriation and imagination the haiku itself or some part of it and probably some of the prose preceding the haiku as well. The vast majority of Basho’s haiku are, however, real or capable of being real.

The role of metaphor in haiku has been distorted by Henderson, Yasuda and Blyth, the three most influential translators and interpreters to introduce haiku to the English-speaking world. They simply ban metaphor and metaphorical interpretation in haiku as they do with all figures of speech. They also insist that haiku must have a season word.

At the other end of the spectrum are the views of Raymond Roseliep, one of the more radical poets of the North American haiku movement:

                  Metaphor I especially promote because it is the imagination’s pet tool.
                  To deny the poet either that tool or his creative mind in haiku is to re-
                  duce him to a mere poetaster. (51)

George Swede, one of haiku’s foremost critics and poets, in the late 1970s/ early 1980s took the view that metaphor was perfectly acceptable since “haiku are, after all, poetry,” noting that there are many examples in both classic and modern haiku of the restrained use of figures of speech, including metaphor. (52) The Modern Japanese poets were, however, not that restrained in their use of figures of speech.

Rod Willmot, another of haiku’s foremost critics and poets, also argued in the late 1970s/early 1980s that a “haiku contains two fundamental parts, which interact with each other metaphorically.” (53)

Shirane says that haikai, which includes haiku, “like all poetry, is highly metaphorical.” He points out, as one example, that the season word can be seen as “an implicit metaphor or extension of the poet’s inner state,” which “tends to be highly subjective.” (54)

Ueda’s translation of Basho’s haiku does not read as a metaphor, whereas Shirane’s does. How a translator perceives the role of metaphor in haiku can greatly influence their translation. Ueda very rarely translates Basho’s haiku as a metaphor or interprets them metaphorically. Shirane, as just noted, sees haiku as highly metaphorical.

Translations of some of the prose in The Narrow Road to the Deep North also tend to be metaphorical. As Ueda notes, the title itself is “more metaphorical than literal;” and even though there was an actual journey involved, Basho’s spiritual quest functions on a metaphorical level. Ueda concedes that some of the poems can be interpreted both literally and metaphorically, although he does not specifically refer to this haiku by Basho. (55) It is apparent that around this time Basho was at least thinking metaphorically.

Pound’s Metro Poem is written as a metaphor as is Basho’s haiku in all but one of the translations referred to in this article, and both can be interpreted metaphorically. However, what most, if not all of the critics referred to agree on, explicitly or implicitly, is that the juxtaposition of the two parts or two images of a haiku goes beyond metaphor. Bill Higginson argues that Pound achieved this intentionally when he changed the colon to a semi-colon. (56)

a haiku Shiki would have been proud to write

                  Amid the summer grass
                  the wheels of a steam engine
                  come to a standstill

                                    Yamaguchi Seishi (57)

Shiki advocated the modernization of haiku including its subject matter, but he was unequivocal that modern civilization was not a fit subject for haiku. (58) He specifically prohibited trains and, by implication, train stations. (59)

Seishi obviously alludes here to Basho’s “summer grasses” haiku and he appears to be the first to have blatantly violated Shiki’s particular admonition against trains. This highlights the role of allusion in haiku to other works of art, particularly to other haiku. Seishi’s haiku stands on its own, updates Basho’s haiku in a thoroughly modern context, and serves as a kind of commentary on it and Shiki’s position on modern civilization.

Higginson says that Pound’s Metro Poem, with the colon, is simply “a sentimental metaphor” because “one thing restates another in a different way, or that the first simply introduces the second.” Whereas:

                  A semicolon shows that the two statements are independent of each other,
                  though they may be related. So that “both ‘faces’ and “petals’ should be un-
                  derstood as real, physical objects, each a core image that stands out against
                  its own background. (60)

Higginson permits the second image to have been imagined. He does not insist that it be real only that it be understood as real. He also said earlier at a Haiku Society of America meeting in 1973 that the semicolon makes the images independent and equal or “coequal,” which results in “a third thing” from their juxtaposition; and that Pound’s revision of the poem allows but does not force the reader to interpret the poem as a haiku. (61) Later, in The Haiku Handbook (1985), he says it turns “an otherwise sentimental metaphor into a genuine haiku.” (62)

Ross, on the other hand, sees Pound’s Metro Poem with the semicolon as a metaphor; Stryk, as a simile.

Regardless, Higginson, like most, if not all of, the critics referred to in the previous section, requires the juxtaposition of images in a haiku to result in something more than a metaphor, and he does not interpret the poem metaphorically.

In dictionary and encyclopedia definitions of metaphor it is usually distinguished from a simile in two principal ways: it is implied rather than explicitly stated and it creates a sense of identity rather than a mere comparison between two images. The metaphor, like the simile, however, is one image expressed in terms of another; the images are related through their similarities and not their differences; and it juxtaposes a concrete image with an abstract thought, idea or interpretation. Metaphors are rarely intended to be taken literally. (63)

Pound never did describe the interaction of the two images in his Metro Poem as a metaphor, and does not even use that word in his Essay, or very often, when discussing Imagism or the Image.

Pound called the useful technique he relied on “a form of super-position.” The word “superpose” means: “Geom. To make (one figure) coincide with another in all parts, by or as if placing one on top of the other,” and if things coincide, an identity is created. (Webster’s) Pound does not appear to be describing a metaphor and if he was, why not simply say so?

In the first version of Pound’s Metro poem, the first two phrases of the first line, “The apparition      of these faces” is exactly the same length as the second line “Petals      on a wet, black bough;” one is set one on top of the other and they physically coincide.

In traditional haiku, the principal technique of juxtaposition is attributable to Basho, and his famous crow poem is considered the model (c.1679):

On a bare branch
A crow is perched—
Autumn evening.

kareeda ni karasu no tomarikeri aki no kure (64)

A number of Basho’s critics, including Shiki, Blyth, and Ueda (65), suggest he appropriated the image of a crow on a bare branch from classical Chinese art and poetry, to which he simply added the autumn evening, and maybe not even that. Yamamoto (1907 – 1988), a Basho scholar and interpreter, delicately implies that Basho may have appropriated the image from a renga verse printed in 1563—over a century before Basho’s crow poem (66):

evening crow—
in a bare tree on the mountain peak
a voice

Again, the critics do not seem too alarmed, although Shiki feels the haiku is trite, and Yamamoto is concerned with the similarities of Basho’s haiku with the one he quotes. In the poetry competitions that Basho judged, in keeping with the rules at the time, he would have had to disqualify his crow poem because it was too close in theme to the other haiku, which came earlier. (67) Implied in this is that a haiku poet needed to be familiar with what has been and is being written.

In Basho’s crow poem there is some debate as to whether the branch is bare, withered or dead because the translation of the word kareeda in the original permits all three. (68) This is another example of Basho using words with multiple meanings.

The poem is a model for two reasons: the new technique, which Henderson calls internal comparison; and the simple, objective, ordinary images it contains. (69) It is, however, a model for technique only.

In addition to the haiku-like qualities already mentioned in the previous section, there are a number of interesting parallels that can be made regarding Pound and Basho as founders of poetry movements and the two poems under consideration. (70)

Basho was weary of the artificiality of Japanese court poetry and looked to classical Chinese verse for inspiration. He set out to reform the poetry of his day and had for years consciously been working on a new technique. This poem is seen as the culmination of those efforts.

Basho had been experimenting with form, content, and language. His crow poem had two different versions with different syllabic structures: 5-10-5 then the present one with 5-9-5 Japanese syllables, both longer than the usual 17. The poem is expressed in the simplest possible language and the k sounds imitate the crow’s caw, which means that the crow is not silent.

The historical importance of the poem in the development of Basho’s haiku, and haiku ever since, cannot be overestimated, although some critics, Ueda for example, say its quality as a poem is exaggerated. Blyth sees at as a masterpiece and a milestone of Japanese culture. (71)

The principal of internal comparison is best described by Henderson himself:

                  . . . the two parts that make up the whole are compared to each other,
                  not in simile or metaphor, but as two phenomena, each of which exists
                  in its own right . . . in which the differences are just as important as the
                  likenesses. Here it is not simply that ‘over the withered landscape the
                  autumn nightfall settles like a crow.’ It is also the contrast of the small
                  black body of the crow with the vast amorphous darkness of the nightfall—
                  and whatever else the reader may find in it. (72)

It is easy to see here Henderson’s influence on Higginson’s thinking.

Comparison and contrast between the images in a haiku is implicit in the works of Blyth, Yasuda and others. Ueda does not cite Basho’s crow poem as an example of internal comparison but he does use that very phrase when discussing haiku later in Basho’s development and in particular his “old pond” haiku. (73) Willmot says that haiku’s “metaphorical structure” requires “comparison and/or contrast.” (74) Shirane speaks of “resonance in dissonance, congruity in incon
gruity.” (75)

Aristotle said over two thousand years ago that “to make good metaphors implies an eye for resemblances.” Babette Deutsch, in her Poetry Handbook, has this response: “Twentieth century critics have shown that the making of good metaphors implies an eye for differences, too, and that the meaning of a metaphor issues from more complex interactions of perceptions, feelings, and thoughts than were dreamt of in that good Greek’s philosophy.” (76)

This suggests correctly that the metaphor has evolved over time and that the modern or contemporary metaphor is or can be much broader in scope than simply relating one image in terms of another through their similarities. Perhaps metaphorical interpretation and internal comparison have much more in common than previously recognized.

This concept of metaphor is very close to how Rod Wilmott saw metaphor functioning in haiku: a metaphorical structure involving comparison and/or contrast, in which the resulting relationships are interpreted metaphorically.

Fundamentally, however, metaphorical interpretation is still premised on ideas about things rather than the thing itself, something that many, if not most, consider essential to haiku. Generally, metaphors are also not intended to be taken literally whereas images in haiku are.

Higginson interprets the poem in the following way: “Our sense of the Paris commuters as delicate, vulnerable life builds, now that we see them come up out of the dark underground into a world of falling petals and spring mist.” (77)

Higginson’s interpretation does not preclude the possibility that the event can be interpreted metaphorically, nor does the semicolon. But the problem with his interpretation is that it alters the language of the poem. It requires the event to occur, not in the station, but on the way in, just outside. Higginson describes the poet/reader entering the Paris metro. This is not capable of having happened, nor can it actually have happened because of the word “In” in the title.

In the third and final part of this Fluence, some of Pound’s other haiku-like poems will be looked at, Pound’s Image will be compared with the haiku moment, and the discussion of whether or not Pound’s Metro Poem is or can be a haiku will be concluded.

Fluence (1): Part 1


24. Modern Japanese Haiku, Makoto Ueda, University of Toronto Press, 1976, p.198.
25. The Master Haiku Poet, MATSUO BASHO, Makoto Ueda, Kodansha International, 1970, pp. 30 and 169.
26. BASHO AND HIS INTERPRETERS, Makoto Ueda, Stanford University Press, 1991, pp. 242, 243.
27. Traces of Dreams, Haruo Shirane, Stanford University Press, 1998, p. 238.
28. Footnotes 26 and 27.
29. HAIKU MASTER BUSON, Yuki Sawa and Edith Shiffert, Heian International, 1978, p. 20.
30. Footnote 27, Shirane, p, 238.
31. Anthology of Japanese Literature, Donald Keene, Grove Press, 1955, p. 369.
32. ON LOVE AND BARLEY, Lucien Stryk, Penguin, 1985, p. 80.
33. Japanese Poetic Diaries, Earl Miner, University of California Press, 1969, pp. 176, 177.
34. Footnote 26, Ueda.
35. Footnote 26, Ueda.
36. Footnote 26, Ueda.
37. Footnote 27, Shirane; Footnote 31, Keene; Footnote 33, Miner; On the narrow road, Lesley Downer, Summit Books, 1989, p.70; Basho’s Narrow Road to a Far Province, Dorothy Britton, Kodansha International, 1980, pp. 56, 57.
38. Footnote 31, Keene.
39. Footnote 27, Shirane, p. 239
40. Footnote 25, Ueda, chapter 1.
41. Footnote 26, Ueda.
42. Masaoka Shiki, Janine Beichman, Twayne Publishers, 1982, pp. 58 to 60.
43. Footnote 24, Ueda, p. 7; Footnote 42, Beichman, p.31; Footnote 19, Henderson, p. 161; Footnote 27, Shirane, p. 38.
44. A ROSELIEP RETROSPECTIVE, David Dayton ed, Alembic Press, 1980, p. 20.
45. The Path of Flowering Thorn, The Life and Poetry of Yosa Buson, Makoto Ueda, Stanford University Press, 1998, p. 159.
46. Footnote 4, Ross.
47. Footnote 45, Ueda, p. 117.
48. A HISTORY OF HAIKU, Volume One, R. H. Blyth, Hokuseido Press, 1963, p. 255.
49. Footnote 25, Ueda, pp. 139 to 141
50. Footnote 27, Shirane, pp. 48, 49.
51. Footnote 44, Roseliep.
52. The Modern English Haiku, George Swede, Columbine Editions, 1981, pp. 28, 29.
53. A Haiku Path, The Haiku Society of America Inc., 1994, pp. 212, 213.
54. Footnote 27, Shirane, pp. 46 and 49.
55. Footnote 25, Ueda, pp. 136 to 138.
56. Footnote 3, Higginson; Footnote 53, pp. 93, 94.
57. Footnote 24, Ueda, p. 159.
58. Footnote, 42, Beichman, p. 31.
59. Modern Haiku, Poems by Seishi Yamaguchi, Takashi Kodaira and Alfred Marks, Mangajin, 1993, p. xvi.
60. Footnote 3, Higginson.
61. Footnote 53, pp. 93, 94.
62. Footnote 3.
63. The Cambridge Guide to Literature in English, Ian Ousby, Cambridge University Press, 1988, p. 658; Oxford Reference Encyclopedia, Oxford University Press, 1998, p. 896; Benet’s Reader’s Encyclopedia, third edition, Harper and Rowe, 1987, p. 643; A Glossary of Literary Terms, M. H. Abrams, Holt, third edition, Rinehart and Winston, 1971, pp. 60 to 63; Poetry Handbook, Babette Deutsch, fourth edition, Funk and Wagnalls, 1957, pp. 84 to 89; Webster’s; OED.
64. Footnote 25, Ueda, p. 44.
65. Footnote 64; Footnote 26, Ueda; HAIKU, Volume 3, Summer – Autumn, R. H. Blyth, Hokuseido Press and Heian International, 1982, p. 898.
66. Footnote 26, Ueda.
67. Footnote 25, Ueda, pp. 150, 151.
68. Footnote 65, Blyth.
69. Footnote 19, Henderson, pp. 18, 19.
70. Footnote 25, Ueda. The next thee paragraphs are taken from pp. 36 to 44.
71. Footnote 26, Ueda; Footnote 65, Blyth.
72. Footnote 19, Henderson, pp. 18, 19.
73. Footnote 25, Ueda, p. 53.
74. Footnote 53.
75. Footnote 27, Shirane, p. 108.
76. Footnote 63, Deutsch, p. 84.
77. Footnote 3, Higginson.

This Post Has 27 Comments

  1. Both Michael and Allan have a great deal of useful information in their posts. George Lakoff’s book looks interesting as well.
    I think I’d rather stick to intrinsic and extrinsic for my own understanding, since symbols have come to mean concepts in graphic forms. And since I’m trying to hard to understand words, I think I’d better stick to verbal phrases and labels – but as I say this I see the wisdom in Allan’s reluctance to quantify these things.

  2. I agree with Adelaide that Michael has contributed a useful and lucid discussion (and distinction) here–thanks for it. Literary terms (e.g., “genre”, “form”, usw) are notoriously slippery in their meanings, and much debate has long surrounded the meaning of “metaphor”, but perhaps the proposed terminology could be simplified, from “intrinsic metaphor” and “extrinsic metaphor” to “metaphor” and “symbol”. At least, I would describe Williams’s classic haiku on Virgilio’s death as symbolic. The analogy it posits cannot be inferred from the poem itself but only from context or “being in the know”. Symbols often have a more arbitrary and “personal” quality than metaphors do (consider the work of Blake and Yeats, among others). The presence of a metaphor is usually clear from the “surface” of a poem or passage whereas that is not necessarily the case at all for symbols. And, of course, words themselves are relatively arbitrary symbols for what they signify; language, in addition, is littered with dead metaphors, a phenomenon Michael mentions without employing that precise term. It’s funny when misspellings reveal how dead they are–e.g., “tow the line”!

    Metaphor often seems most successful to me in haiku when it is so fresh and graphic that it hardly seems metaphorical at all. Two examples:

    the river floor–
    crayfish walking across
    a mosaic of light
    (Peter Yovu, Turn to the Earth)

    old pier . . .
    shoals of small fish
    rain-ripple the surface
    (John Barlow, Roadrunner 6.3, 2006)

    Beyond brevity, I have difficulty pinpointing any single quality that is “essential” to all haiku, including juxtaposition. And brevity alone is not a sufficient quality. So, the “essence” of haiku, despite many conventions, traditions, and norms, remains quite elusive–which is actually an aesthetic strength! Or so I opine.

  3. A very clear explanation, Michael. Seeing an implied metaphor in a haiku is another way the reader completes the poem. It will be there for some readers, but not for others. This personal interpretation of haiku is what I find so appealing, as a reader of haiku and as a writer.


  4. As Merrill comments, she has “always felt a shadow of metaphor in haiku, even as everyone denies it.” In response to the issue, I have long wondered if too many readers of haiku confuse intrinsic metaphor (occuring in the text of the poem itself) and extrinsic metaphor (occuring in our interpretation of the poem, or in its implications). In other words, there’s a difference between explicit metaphor and implicit metaphor. An explict metaphor would include tears in the eyes of fishes in Basho’s poem, or the nonliteral shadow in the phrase “the shadow of metaphor.” An implicit metaphor would include how, for example, we interpret Paul O. Williams’ “gone from the woods / the bird I knew / by song alone.”

    Robert Hass was once the guest speaker at the Yuki Teikei’s haiku retreat at Asilomar (early 1990s), and made the point that haiku employ metaphor frequently. Unfortunately, nearly all the examples he used all had IMPLICIT rather the explicit metaphor (meaning that the metaphorical meaning was IMPLIED rather than the poem itself being overtly metaphorical). In other words, even when I pressed him on this distinction, he seemed to confuse these two types of metaphors, or didn’t care about the distinction. I think this is a common confusion, and seems to complicate the discussion of metaphor in haiku.

    In case the distinction I’m trying to make may be tricky to understand, consider the phrase “it’s raining cats and dogs.” We know that’s a metaphor for a great amount of rain. No actual cats and dogs are actually harmed in the making of this metaphor. No cats and dogs actually fall from the sky, at least not in my experience. This metaphor is explicitly present in the words used. Now compare that to the Paul O. Williams poem. His poem functions first on the literal level — it’s about a BIRD that has left the woods, taking its song with it. The poem is wonderful on that level alone. But the poem has a deeper level, and it’s one of metaphor — but it’s a metaphorical INTERPRETATION, and THIS metaphor takes place not in the poem but in the READER. In Paul’s poem, we know he wrote it as a memorial to Nick Virgilio when he died. Paul had never met Nick, and knew him only by the “song” of his poems that appeared in the “woods” of various haiku journals. The delight of this masterful haiku is how it oscillates between these two levels of meaning, and how it embraces both at once.

    Now, it could be argued that a phrase such as “it’s raining cats and dogs” is also interpreted by the reader for its meaning, and that’s true, but the important distinction is that the phrase itself does not have any LITERAL meaning first, whereas Paul’s poem has the intrinsic literal meaning as well as the extrinsic metaphorical one. Any discussion of metaphor in haiku needs to grasp this vital distinction.

    Indeed, my point with all of this is that when we talk about metaphor in haiku, we need to distinguish between the levels of metaphor. I generally avoid most overt/explicit metaphors in haiku (but nor am I afraid of them). And I eagerly welcome any implied/implicit metaphors if they happen to occur (but nor do I require them). The problem I frequently have with overt metaphor is that, certainly in the hands of beginners, the metaphor is clumsy, tired, or too much of a detour towards the thing rather than the thing itself. The mind too often gets in the way. More often than not, I’d rather see the subject itself rather than an intellectualized interpretation of the subject (that is, an overt metaphor about it). I hasten to add that the *right* subjective interpretation can be electric in some haiku (Basho’s peony exhaling a rainbow!) — everything depends on the poem — but more often than not, explicit metaphor can damage or weaken haiku, and not just among beginners.

    To complicate matters, a reading of George Lakoff’s book *Metaphors We Live By* will show you how deeply language itself is metaphorical. Consider, for example, something as common as saying “table leg.” Well, actually, no, it’s not a leg in the human or animal sense, which is the primary literal meaning, and where the metaphor comes from. But the metaphor is so deeply ingrained that we no longer think of it as a metaphor. Consequently, if anyone says “haiku don’t use metaphors,” not only might they be wrong, but it would also help if they specified implicit vs. explicit metaphor — or perhaps they should specify, in Lakoff’s terms, what levels of metaphor they mean.

    What is interesting to note, especially in Japanese haiku, is how they are much more often intellectualized (whether metaphorical or not) than our nature- and/or image-based haiku so often are (even while usually retaining at least some objective/imagistic element or core). In November of 2009, Jim Wilson (also known as Tundra Wind) made a useful description of the “siblings” of haiku (, identifying the following three traditions in the United States (I think it’s broader than just the U.S., mind you):

    1. Duration Tradition (brevity is the essence of haiku).
    2. Nature Tradition (seasonal reference is the essence).
    3. Syllabic Tradition (counting syllables is the essence).

    I don’t necessarily agree with these distinctions (I don’t think they’re that cut and dried, for starters; he doesn’t pay enough attention to juxtapositional structure; and the syllabic tradition is fundamentally a misunderstanding of Japanese haiku — an urban myth — and should never have happened in English in the first place), but they do help to identify what’s going on in our haiku. The question, of course, is where does metaphor fit in — or not — with these perceived traditions?

    I guess what I’d say is still “don’t use metaphor in haiku . . . unless it’s damn good.”


  5. Fascinating, Mark. Thanks!

    … and this, which I didn’t know, would seem to lend some credibility to my guess about Pound’s ‘apparitions’ in ‘In a Station of the Metro’ alluding to Yeats’s ‘The Apparitions’. Or maybe vice versa, if they were discussing these things.

    “At the time Pound was part-time secretary to Irish poet and playwright W.B. Yeats in London, and the pair became fascinated by the exotic form.”



  6. My misunderstanding. Sorry. Especially given the experimental and fine haiku you write, I was surprised by what the citation you gave (thinking it was an answer to the issue).

  7. jack, i totally agree. i was just sharing a reaction to what was being presented and discussed in this post.

  8. I think Merrill’s example is brilliant and illuminating (note the metaphors).
    I would also add, though I am repeating myself, that “nature” is never there: words are not things, they are representations of sounds that do have a signified (a thing), but the thing is not present in the word (it is implied and only exists by convention).
    So, what Coomler might more accurately say is that in his hokku conventions (symbols,metaphors, classes) are there.
    The idea of a transparency between nature and words might have existed (and probably did) in the 16th through 19th centuries. But, thereafter, the philosophy of language took off in other directions (see the weak metaphors).
    I daresay that every sentence, particularly those using the copula, are metaphoric to the degree that they join two separate things, which, it need not be said, are not equivalent but become so by being joined (this sentence for instance).
    I would also like to say that, at least for me, it is not necessary to be too subservient to sometimes quaint rules about haiku in order to write haiku.

  9. Scott, I’m at a loss. How can Basho state that summer grasses are all that’s left of warrior’s dreams and in saying this he meant it to be taken literally? How can dreams become blades of grass? Or, does warrior’s dreams mean something other than what’s literally stated?
    Really, how can it be otherwise? I think dreams transmuting into grass plots is a perfect example of metaphor.

  10. I came across this discussion today by Rob Kyff, a linguist, regarding how one word became another. Here is what he said:
    “But how did ‘firm’ become ‘farm’? Scholars say it’s because agriculture plots were made firm, either physically by walls around them or metaphorically by signatures on a lease.” It would seem that in this sentence he refers to the fact that the signature itself is a metaphor. Since I can not speak Japanese, I can only understand these things in the terms of the English language.

  11. Jack and folks,

    The entire text of Shirane’s Traces of Dreams is online and searchable at:

    And as long as Scott is pointing to Coomler’s recent post, a quick comment. I read Coomler’s stuff in the early 2000s and own a copy of his book that’s well-marked. Coomler writes, in part:

    Summer grasses –
    All that remains
    Of warrior’s dreams.


    There is nothing metaphorical about that at all. A metaphor, you will recall, is speaking of one thing as though it is another. There is none of that here.

    One problem is this definition of metaphor (wiki: “an analogy between two objects or ideas”) . Just look at the types of metaphor listed in wiki

    There’s an argument, by the way, that kigo function on one level as synecdochic metaphors: “a trope that is both a metaphor and a synecdoche in which a small part of something is chosen to represent the whole so as to highlight certain elements of the whole.”

    You know, cherry blossoms to spring, fireflies to summer, the moon to autumn, snow to winter, etc. “summer grasses” to summer is rather direct, but it’s still used by haikai poets to represent summer *as well as* itself.

    Even if Bashô is alluding to Du Fu (and it’s more than fair to say he is as he says so contextually), there’s still metaphor at play. Metaphor doesn’t stop being metaphor with repetition or allusion.

    More to the point, NOT reading metaphor, a problem intrinsic in the “Hokku” understanding of haiku, is another error. It’s rooted in its Blyth-centric origins, though much has been discarded from Blyth that is useful.

    Of course, my view of this is not just my view. Here’s Bill Higginson’s take on it from Modern Haiku:

    And I agree with Higginson’s advice to read the six primary volumes of Blyth. There’s so much there and not just in examples.

    Actually, I’m a big fan of Blyth.

    That said, Shirane comments on Yasuda and Blyth and metaphor and Chinese poetry on pages 45-46 of Trace of Dreams in the chapter “Bashô Myth East and West”. As it’s probably relevant to the metaphor discussion in general, here’s some of that section:

    […] This attempt [Yasuda’s] to negate the function of metaphor, which also occurs in Blyth’s writings, is misleading in that haikai, like all poetry, is highly metaphorical: the essential difference, as we shall see, is that the metaphorical function is implicit rather than stated and often encoded in a polysemous phrase or word.

    In discussing classical Chinese poetry [MB note, like Du Fu as Coomler mentioned, no?], Pauline Yu has argued that Western metaphor tends to be based on a fundamental dualism, the assumption of a disjunction between two realms, usually concrete and abstract, or physical and metaphysical. […] This kind of metaphor, which actually represents only one limited type, rarely, if ever, occurs in Bashô’s haikai. Nature exists as something concrete and living before the viewer’s eyes, as immediate, and is respected as such. At the same time, however, nature can implicitly have a semi-metaphorical function, particularly as a projection of the poet’s inner or outer state or as that of the addressee. The natural imagery functions on both the literal and figurative levels, collapsing I. A. Richards’s noted distinction between vehicle and tenor. […]


    I. A. Richards’s distinction is also covered in the wiki entry on metaphor.



  12. Mark, thank you for your informative and insightful remarks.
    I’ll have a look at Metaphors We Live By.
    I read Shirane’s work, although it was long ago and I’m afraid I didn’t remain these cogents remarks of his that you provided. I agree with you and also see the further internal forces at work in haiku. I’m grateful to you for your help.

  13. Jack,

    I agree that language is metaphoric. Actually, I’d go further and say that thought is metaphoric. Are you familiar with Metaphors We Live By (Lakoff & Johnson)? You can read about it here

    I think there can often be a “metaphoric reading” of a haiku, by which I mean an equating of the fragment and phrase across the kire. This type of reading is probably made more prominent when the kigo takes up all of one part, like in the two haiku you quote above.

    However, I believe that whenever a kigo is used there are multiple levels of meaning from the literal to the symbolic, from the innovative to the intertextual. Haruo Shirane writes in Traces of Dreams:

    “[T]he seasonal word, the requirement of every hokku, often exists simultaneously on a number of axes or in different contexts: as a reference to an external scene, as an implicit metaphor or extension of the poet’s inner state, as a complex literary and cultural sign, and as a greeting to the addressee. The first is highly objective and referential; the second tends to be highly subjective; the third is often highly fictional and intertextual, and the fourth is a performative utterance.”

    So, I read much more into “A spring night” than an equation of parts. There is setting, scene, emotion, and more. A Japanese poet would undoubtedly get more, not from being Japanese, but from being exposed to more quality haiku on each and every subject.

    The two haiku in your last comment, including yours

    A spring night
    without a boat I set sail
    to the moon’s shores

    remind me of one I dug up and posted earlier today elsewhere (twitter)

    millions of bats . . .
    even the seas of the moon
    darken at dusk

    (from A Handful of Pebbles, 2006)

    The reason I mention it here, other than topical resonance, is that I believe there is metaphor working here, but not so much between the bats (the kigo – summer) and the seas of the moon.



  14. Would it be too much to claim that language itself is metaphoric? Given the fact that words are arbitrary (Ferdinand de Saussure) and that there is no correlation or correspondence between words and the things they denote (except by convention), that language “acts” metaphorically by defining or naming an object (one category) by means of another (category).
    Of course, there is lacking a similarity and difference in this relationship, one that exists in metaphor proper, but still there is a similicrum of function.
    I think that is the problem we face as poets, isn’t it? The medium we use is made of of classes, abstractions, and they are the vehicle by which we raise things (an abstraction) into existence and meaning and even “particularity.”
    I have found that sometimes it is necessary to write off the subject (Richard Hugo’s term) in order to closer proximate it, since the sounds of the words themselves did not realize the emotion I intended in assembling them.
    I’ll give you an instance. I was wondering how best to evoke the beauty of a spring night and by merely giving a description, or placing before the reader the “things” themselves, I was failing because the words, their sounds and symmetry (or lack thereof) wasn’t working. So, I tried it like this:
    A spring night
    without a boat I set sail
    to the moon’s shores
    This is metaphoric haiku, I think, equating the spring night to what follows in the poem.
    On the other hand, I recall a poem I read (and I apologize to the poet, because I no longer have her book and don’t remember her name; all I remember is that is Japanese and an excellent poet) that achieves what I was looking for in a more traditional way. Her poem:
    A spring night
    even after dark
    the clouds are white
    It is not for me to say which poem is better (if making that decision is necessary), but while her poem is more in keeping with the traditions of haiku, I would still suggest that her poem is metaphoric, the subject spring night being compared to another category of things, i.e., white clouds.
    What do you think?

  15. An aside on Eliot and Pound and The Waste Land…

    According to the hypertext version of The Waste Land

    that particular section (two of 10 instances of “nothing” in The Waste Land) follows mention of Margate and “[i]n October 1921 Eliot was in Margate, recuperating from mental exhaustion.”

    The hypertext TWL has been around for years and years and is very useful for quick information. However, if you are interested in the extensive editing Pound did to shape Eliot’s work, I highly recommend getting a copy of

    The Waste Land: A Facsimile and Transcript of the Original Drafts Including the Annotations of Ezra Pound (Paberback ISBN 0-571-11503-9)

    According to wiki, “[t]he drafts of the poem reveal that it originally contained almost twice as much material as the final published version.”

    The interplay between the poet Eliot and poet/editor Pound on TWL in particular is fascinating. I’m traveling now or I’d look up Pound’s notes on the nothings in question…

    BTW, the dedication of Eliot’s The Waste Land is

    For Ezra Pound
    il miglior fabbro.

    [The better craftsman. (Purgatorio xxvi, 117)]


  16. “Consciousness, our sun, our dreams, would be forever mired in a “moor,” a wasteland, without the ability and understanding that things can be made to carried over into other places and meanings.” Jack Galmitz”

    The metaphors are
    gone, and so is my faith . . .
    sun over a moor

    Nakamura Kusatao

    I can connect
    Nothing with nothing.

    from T.S. Eliot, ‘The Wasteland’, (1922) which Pound had much influence on, to the point of editing it, if I recall rightly.


  17. Thank you, both Nick Avis and Jack Galmitz. I find this area of interest fascinating.

    I have been inclined to believe that implied metaphor has been acceptable in ELH, but that overt or blatant metaphor has been considered ‘iffy’ at best.

    I haven’t checked dates, but assuming that ‘In a Station at The Metro’ was conceived and published after Yeat’s ‘Last Poems’, I’d imagine that Pound would have been familiar with ‘The Apparitions’, and was bridging, with the greatest respect, traditions in English language poetry and Japanese haiku.

    ‘Fifteen apparitions have I seen;
    The worst a coat upon a coat-hanger.’

    Also, in ‘Last Poems’, ‘High Talk’, which (in the grand, overt poetic style) announces, towards the close of this most nontraditional sonnet:

    ‘All metaphor, Malachi, stilts and all.’

    and follows with a clear, natural image:

    ‘ A barnacle goose
    Far up in the stretches of night;’

    Then, as if to prove that metaphor is inextricable from our common language:

    ‘night splits and the dawn breaks loose;’

    before concluding with a fantastical/ mythical image , which has earned its place by being grounded in what has gone before,

    What has this to do with haiku? If haiku is poetry, then there must be ‘bridges’ between English language poetry and haiku. I find Pound’s much-anthologized poem such a bridge,


  18. Thank you for the lesson in Japanese, Gabi. It also strengthens my suspicion that both poets were speaking about the same subject.

  19. “both poets are deprived of cultivation in
    the “moor” (in the one case) or in
    the “withered fields” (in the other).”


    tabi ni yande yume wa kareno o kake-meguru

    hiyu morotomo shinko kiete kareno no hi

    Both Japanese poets use the same word, KARENO.
    The translators used different words.

    for example, from the yahoo dictionary
    a desolate [wintry] field

  20. Jack, I tend to agree with you. As someone who thinks in images first, the word itself is a sort of metaphor to me,and I’ve been trying to separate any kind of understanding without metaphor with no success whatsoever. Perhaps the word metaphor means separate things to separate people? But I tend to feel that Jack has put into words what I couldn’t find words for. Thanks.

  21. The metaphors are
    gone, and so is my faith . . .
    sun over a moor

    Nakamura Kusatao

    I’d like to use this above poem, incorporated in Nick Avis’s splendid essay about metaphor in haiku, to initiate further discussion on the subject.
    The poem suggests, by association, i.e., by metaphor, that the loss of metaphor is somehow similar to the loss of faith, and, further, is equivalent to the sun over a moor. Why would this be true, assuming that it is true?
    I’m reminded by this poem to Basho’s death poem: stricken while journeying/my dreams still wander about/but on withered fields.
    In both poems, consciousness, substitute “sun” in Kusatao’s poem, and “dreams” in Basho’s, when deprived of wellness (faith and metaphor in Kusatao and “dreams” in Basho) requires a place of cultivation to prosper. In both poems, both poets are deprived of cultivation in the “moor” (in the one case) or in the “withered fields” (in the other).
    Metaphor is the transfer of meaning in languages (all human arts) and allows a secondary meaning to be derived from a primary meaning through the use of a vehicle. Without this carrying over (from the original Greek for metaphor), it would be impossible for there to be faith of any kind, because things would have no possibility of standing in the place of anything else, could not represent anything else. From a Western perspective, there would not be a creation, or, more precisely, the world would not suggest anything other than itself.
    Consciousness, our sun, our dreams, would be forever mired in a “moor,” a wasteland, without the ability and understanding that things can be made to carried over into other places and meanings.
    We turn the moor of the world into cultivated fields, into villages and cities and interconnect them further into our societies by the use of metaphor, which, in its greatest sense, is the ability to transfer and carry meaning from one place to another.
    Actually, I would suggest that the unmotivated nature of language (the fact that words bear no relationship to things) requires metaphor to function at all and that we would have no interpretation of haiku or anything else without it.
    Think of any haiku (perhaps all those mentioned by Nick Avis above) and everyone requires some form of substitution of the literal sense to another symbolic sense to have any meaning at all to us.
    We are users of metaphor; we are, I believe, bound to metaphor as our most primordial inheritance.

  22. “if one does not try to get some graspable truth from the metaphor, it can be a way one’s mind consummates itself: although symbols can be redeemed only by mind, the mind does not function in a vacuum but is activated by (or as) symbols.”

    David R. Loy, “Awareness Bound and Unbound, Buddhist Essays.” He is commenting on Dogen’s use of metaphor.

  23. The link doesn’t work to this, but I’m so glad I found a way to get to it. This is marvelous. We’re going to have 3-4 days of rain here and I’ve downloaded it to contemplate this weekend.
    I have always felt a shadow of metaphor in haiku, even as everyone denies it. So trying to get to the bottom of this mystery, to me, will be quite rewarding.

Leave a Reply

Your email address will not be published. Required fields are marked *

Back To Top