Skip to content

ABSTRACT: Considering the question of whether haiku can be deconstructed, two readers offer dueling close readings of two classic Japanese haiku and two contemporary American haiku. Though haiku, built out of pure image, seem to avoid making or privileging any definitive claims, they are also made out of the slippery element itself, language, and they test and call into question the very conventions that seem to govern the form. Ultimately, we find common ground in the notion that haiku are provocative and suggestive enough — and self-aware enough about the nature of language — to make the genre itself a kind of deconstructive practice. Even (or especially) in a genre that uses so few words, language repays close attention to the ways in which every utterance communicates more than it, or we, can ever know.



by Ian Marshall and Megan Simpson

Deciphering, normalizing, or tautological, the ways of interpretation, intended in the West to pierce meaning, i.e., to get into it by breaking and entering — and not to shake it, to make it fall like the tooth of that ruminant-of-the-absurd which the Zen apprentice must be, confronting his koan  — cannot help failing the haiku; for the work of reading which is attached to it is to suspend language, not to provoke it. . . . 
          — Roland Barthes, Empire of Signs (72)

There we were, two literary critics out walking in the woods, and somehow the conversation made its way from haiku to deconstruction. Understandable, perhaps — haiku is a poetic form that generally concerns itself with the world outside the self, the world of nature, and the possibility of experiencing oneness with all that. And of course that’s where we were, making observations about that world outside our selves in quick image-laden descriptions. Trail blaze on an oak trunk. Chickadee lilt. Leaf shadow. Cool of a stream through hemlocks. Check out this beetle.

And the deconstruction half of the discussion? Well, we’re literary critics, after all, prone to over-analyzing language (and things like the nature of reality), and so there we were, talking about whether haiku can be deconstructed.

Of course they can, I said, they’re made of language, sparse though it might be in any given haiku, and there are certain assumptions inherent in the form and built into every haiku — assumptions, say, about things like “oneness” or the “haiku moment.”

No, they can’t, she said, because the idea of haiku is to get beyond language, to achieve a transparency of language inherent in wordlessness, and the whole point is to be suggestive — to avoid overt statement or claims that can be pinned down. If it doesn’t make a specific claim, then how can you demonstrate through deconstruction that the language of the claim undoes itself? Haiku accepts the premise of deconstruction, that language is slippery and cannot be relied upon, can never be pinned down to a particular meaning.


Okay, but wait a minute; before we go any further let me clarify — if not “pin down” — my point a bit. My argument that haiku can’t be deconstructed was based on my idea of haiku as primarily, or ideally, image. Showing something isn’t the same as saying something about it, right? And to have something to deconstruct, there needs to be a statement of some kind to begin with. But I didn’t mean to imply that a statement has to be overt or singular and fixed in its meaning in order to be deconstruct-able. Certainly suggestions, implications, and assumptions lend themselves to deconstruction, as they often are expressed in terms of oppositions within a text. And deconstruction “works” by showing how a text seems to privilege one term of a binary, then showing that this privileging is arbitrary, and the hierarchy can be reversed to reveal that the text works against itself, is incoherent.

My thought was that the pure image that haiku relies on can’t privilege anything, since it isn’t thought, and thus, not subject to the patterns of binary thinking central to Western thought. Maybe that’s not surprising, given that haiku is originally a Japanese literary form bound up with a tradition of Zen thought. But, now that I’ve said all this, I should confess I’m no longer so sure that image can’t be deconstructed, as it’s a linguistic structure and certain words will be used to render it, others not — each element of the image necessarily implying something about its impact, tone, or character. In fact, I wonder if it is even possible to create a “pure” image in language.


Time for a case in point.

the dog out —
the stars out
          — Gary Hotham (in van den Heuvel 87)

This seems to defy deconstruction. So much suggested — so little stated. What is it that is suggested? The parallel structure in lines two and three suggests, zeugmatically, that the speaker is letting the stars out as well as the dog. In the assumption of control implied there, isn’t there a touch of arrogance, an assertion of dominion over the very forces of nature? That assertion calls into question one of the implicit claims of haiku. In its privileging of a “haiku moment” lies the assumption that there is a dissolution of the boundary between self and world, an absorption of self in the present moment and in the world outside the self. But that verb “letting” presumes agency, control over something, and that means there is still a perceiving ego here — a perceiving ego separate from the world.

This is not criticism. “Letting” may be the perfect word choice here to suggest the momentary release of consciousness as it opens itself out to a sky full of stars. The point here is that even in a genre as suggestive as haiku, favoring words as clear and concrete as haiku does, there is still a message — and that message will inevitably be undercut. Oneness? No — there is still an ego present here, conscious of itself “letting” the self off the leash and out the door along with the dog.

Of course there’s another possibility here — that despite the grammatical links of lines two and three, the two observations are juxtaposed but unrelated, or related only by the perception of the moment. That would be the way of haiku, and in this haiku it would also point out the misleading nature of language. The syntax leads us to see lines two and three as things sharing in the letting out — but they’re not. And so we may be led to see the unreliability of language. Which means the burden of deconstruction is already accomplished for us, without our even needing to read against the grain of the poem.

Which means my argument just deconstructed itself, and I’m proving the opposite of what I set out to prove. Let’s try this again.


Yes. But before we do, I’d like to say a few things about Hotham’s haiku. Basically, it is — perhaps like all or most haiku — already ambiguous, even intentionally so. Therefore, there’s no deconstruction work for us to do. If there isn’t a norm or convention (if we can’t identify a likely or apparent reading/meaning of the thing in the first place), then my question is not so much “can it be deconstructed?” (sure it can — just pick any one of the possible meanings/implications and read against it, as you just did), so much as “why bother?”

What’s the point of deconstructing an already overtly ambiguous text? Of course, there are plenty out there who would ask what’s the point of deconstructing any text. To them I would say it usually serves one of the following purposes, among others: (1) to illustrate something about language itself — its fallibility, its multiplicity, its inability, finally, to get outside itself at all; and (2) to “open up” an otherwise apparently closed text, giving us multiple meanings where before there seemed to be only one. And of course either of these functions can be put to more socially engaged purposes as well, to expose ideological assumptions or bias — in individual texts as well as broader cultural discourses — as poststructuralist feminist, New Historical, and postcolonial critics have done. But it seems to me, and your discussion of Hotham’s haiku shows this, that haiku is already “open,” inherently so, and it depends upon the same features of language that deconstruction does in order to achieve this openness. So what’s to gain by deconstructing it?

To read “against the grain” of this already open text, or, as Jonathan Culler puts it (86), to “show how it undermines the philosophy it asserts,” could lead us to the perverse position of trying to prove that Hotham’s haiku is really quite restrictive and closed, not open at all; rather, it’s coherent, unified. But that would be New Criticism, not deconstruction, right? Because then we’d be reconciling the oppositions and tensions in the poem, not merely recognizing them. (We could say, for example, that both control and release of control are necessary features of the speaker’s experience, and not, therefore, at odds with one another, since opening-up of consciousness and the very act of letting go requires agency in the first place — intention, certainly, maybe practice and effort, as well.)

Okay, but all this doesn’t mean that I’m so convinced haiku can’t be deconstructed, after all. In fact, it’s the evidence you present in your “another possibility,” above, the stuff that starts to convince you to reverse your position and agree with my initial argument that haiku can’t be deconstructed, that gets me thinking it probably usually can be deconstructed. If we take into account not just the textual details of an individual poem but also elements of the form and tradition of haiku, as you’ve been doing all along, then it seems to me that we will always have access to a convention, a poetic intention, a privileged way of reading, that can be read against or problematized.

So maybe according to the “way of haiku,” the second and third lines of Hotham’s poem are not related, but merely juxtaposed. Still, if it’s the syntax that causes us to see them as related, then that’s an element of the text itself that turns the text against itself, if we take as a central feature of “itself” its intention — as haiku — to merely juxtapose, to relate the lines in question only in the sense that they are co-occupants of the same perception of a moment. The fact is, as you point out, the text itself implies they are related thematically as well. Part of what you’re saying, though, I think, is that the poem (or the poet) wants to tempt us to look for “real” semantic connections between lines two and three, so that it can make its point, then, of reminding us that language is unreliable. But if you mean “deceptive,” then that isn’t the same as saying the poem already deconstructs, because it too implies a “real” underlying meaning (or a “real” lack of meaningful connection between the ideas in lines two and three) that exists outside of language. In deconstruction, there is no outside of language.

But this seems too easy and not wholly satisfying. I’m starting to see that it’s not by reading against the grain of the haiku that we gain our most significant insights. Rather, it’s the process of exploring the question itself — can haiku be deconstructed — that might get us to the richer stuff. After all, what’s at stake in that question, since haiku purports to be about the natural world, are more fundamental questions about our (human) relationship to nature.


I like your summary of the purposes of deconstruction — “to illustrate something about language” and “to open up an otherwise closed text.” I think both apply to deconstructing haiku, but for now let me expand on your description of what haiku are about. Besides focusing on the natural world, a typical haiku says, in essence, that it deals in pure image (or juxtaposed images) with the present moment, that the separation of self and world is overcome or dissolved in a moment of intense awareness, and that language can be so minimalist and concrete as to be not there at all interposing itself between us and the world. If deconstruction says there is no outside of language, haiku makes the opposite claim — that its language is no language at all, that its image-stuff is pure reality, or at least an invisible medium we see right through to the reality it presents. My point is that many (or most — or all) haiku, while accepting these claims and assumptions, at least on the surface, at the same time can be opened up to expose multiple meanings, some of them subverting the surface claims.

Take, for example, this one by J. W. Hackett:

Half of the minnows
within this sunlit shallow
are not really there  
          — (in van den Heuvel 61)

Even if we assume language to be transparent here, there is still some misdirection — maybe even deception. An apparent reality is shown to be a fiction, or at least a partial fiction, and the haiku shows that the statements of language (there are minnows there) are unreliable. Or maybe it’s not language but perception itself that misleads us.

But isn’t this haiku out of the moment of perception, out of the senses? There’s the impression of a moment, but built in here is the reconsideration, rationality stepping in to correct the impressions of the senses, challenging the perception of the moment. That “haiku moment” turns out to be a fiction.

Perhaps “shallow” is the word of undoing, or unraveling here — with its freight of “shallow” understanding. The perception of all the minnows being there is shallow, corrected by rationality. But haiku is supposed to privilege the moment, the sensory perception of the here and now — that’s supposedly where the depths of experience and, yes, the meaning of the haiku lie. But here it turns out that the perception of the moment is shallow. And understanding, the rationality that sees the Truth of the Matter here, exposes the shallowness of the moment’s perception. Doesn’t this, then, undercut haiku’s privileging of intuition, finding real perception, or at least accurate perception, in rational reconsideration of sensory information? In that sense, this seems to undermine the idea of a haiku moment. It deconstructs. Unless the point is to suggest that all matter is half-illusion. But then that, too, would undercut one of the premises of haiku — that there is an actual world there that haiku should concern itself with.

Again, the point is not to show that this or any other haiku is flawed because it deconstructs. It is to point out that even here — in the most wordless and concrete and at the same time deliberately suggestive and impressionistic and interestingly ambiguous of all poetic forms — even here, we are working with the fluid element itself, language. Inherent in language, at least language artfully shaped, is the attempt to convey meaning — and that attempt always undercuts itself. Often in haiku the meaning has to do with adhering to, and then conforming to, a codified definition of what haiku is and what it accomplishes. There is an implied message in all haiku inherent in the form, emerging from a shared theme, a shared world view, and a shared aesthetic — but oneness breaks in two, there is revealed a perceiver positioned outside the world being perceived, and the language inevitably reveals its reliance on metaphor and calls attention to itself as language, itself something other than the actual world.


Well that’s pretty clever, what you’re doing with “shallow” — very Derridian, even deManian, in its playfulness. But I wonder if either of those guys would arrive at a similar conclusion. They might ask why undermining the idea of a “haiku moment” necessarily adds up to a deconstruction of the haiku. (Isn’t it then, merely, a failed or flawed haiku?) According to this way of thinking, wouldn’t any haiku that is not true to the idea of the “haiku moment” deconstruct, rather than fail, since the assertion of there being a haiku moment is inherent in any text presenting itself as haiku? Or is a failed haiku the same as a deconstructed one?

For the sake of argument, though, let me see if I can rescue this one from the deconstruction as you have presented it. Why must the initial (mis)perception referred to in the poem (seeing the shadows as additional fish), which is then corrected in the third line with the observation that they “aren’t really there,” constitute the poem’s “haiku moment”? Why isn’t the haiku moment the recognition of the initial misperception, as part of what seems to me a very accurate perception about the actual world that is also the gist of the poem’s apparent central statement? Seen this way, the poem is not implying that matter is half-illusion, but rather, that our perceptions are imperfect. This seems a very different point. Is making a statement about our relationship to the natural world — how we see it, or fail to see it fully — antithetical to haiku?

But, to go in another direction, even if I have successfully restored to the poem its “haiku moment,” it still seems imminently deconstruct-able. Although the first two lines are image, the haiku is anything but “pure” image, since it does make a statement, an assertion. In fact, grammatically, the whole poem is a simple and complete declarative sentence. So we have something to deconstruct. And that is the assertion that something is absent. But to make that assertion, the poem makes that thing (the whole “of the minnows,” including those that “are not really there”) very present. And here’s where the image comes in: in order for me to “see” the poem’s truth, its statement, I have to “see” all the minnows in order to then imagine, in my mind’s eye, the process of separating the shadow minnows from the fleshy minnows. And that the whole poem is located “within this sunlit shallow” is pretty rich, too: the haiku’s surprising revelation has a “wow man, that’s deep” quality to it that belies the idea of shallowness. So, in terms of binary oppositions, the haiku seems to privilege absence over presence, shallowness over deepness, and illusion over reality. I’ve already reversed the first two, and I would refer back to my initial point about the poem to reverse the last pair: if the poem’s “haiku moment” is in the perception that some of what were first thought to be minnows are shadows, then it’s this truth about perception (or reality) that’s privileged.

Ironically, then, the haiku’s success as a haiku, according to my argument, depends upon this deconstruction — we have to follow this unraveling thread in order to “get” the haiku in the first place. But is that the same as saying that it already deconstructs, so we are left with nothing to do as deconstructionists? Or does this leave us somewhere else?

While I’m not sure where it leaves us, this train of thought leads me right back to the question of Western epistemology, whose binary thinking patterns are of such interest to the deconstructionist. If haiku is an Eastern form (but is it, when practiced or translated by Westerners?), does it rely upon binary thinking to the extent to which all Western thought, according to Derrida, necessarily does? Gloria Anzaldúa gives us a different way to think about oppositions in her critique of rigid gender distinctions and other harmful “dualities” of Western thought that underlie the devaluation of and violence against women, tribal peoples, gays and lesbians (53-54, 101-02). She offers instead what she calls “balanced oppositions” — a kind of difference according to which a thing’s meaning does not derive from what it isn’t (its opposite), so each term of a pair is in balanced harmony with the other, not a hierarchical struggle. (Yin and yang come to mind as an example, though Anzaldúa does not mention this pair in her analysis.) What if what we’ve been seeing as binaries in haiku are actually “balanced oppositions”?

I wonder if this concept of balanced oppositions might in fact give us a way to account for a number of haiku’s traditional features, and even help us see how they work together: juxtaposition/contrast, revelation/insight, “oneness” and the “haiku moment” of interconnectedness between self and world. In other words, haiku’s oppositions are inherently, as implied by the very form, balanced. This seems, in fact, to be the case with Hackett’s minnow haiku, as it seems to rely equally on both terms of its oppositions in order to “work” as a haiku and as a statement. So maybe that’s where my attempted deconstruction leaves us: with the recognition that the pairs are easily “reversible” because one side is not actually privileged; there is no hierarchy. It’s about perception of reality, the presence-in-absence, the depths of the shallows.


If we’re going to talk about oppositions, whether balanced or binary, then we’ve got to talk about this one by Kobayashi Issa:

The world of dew 
is the world of dew. 
And yet, and yet —
          — (in Hass 191)

This can almost be read as a poem about deconstructing haiku, calling into question as it does the ostensible universal messages of haiku. Dew is a symbol of the Buddhist idea of impermanence — Robert Hass, in his note on the poem, points out the allusion to the Diamond Sutra: “all conditioned things are like a dream, a phantom, a drop of dew, a lightning flash” (289). Haiku, as an expression of Zen spirit, typically glories in its acceptance of impermanence. Life — matter — the world — it comes and goes, and all we can do is grasp each moment of its hereness, smell the coffee, gather rosebuds while we may, notice the dew, seize the day. Life is a succession of moments (or really, more an assemblage of moments, not necessarily sequential — the sequence is illusory), and to be fully in the haiku moment is satisfaction enough for a life, even if, especially if, that moment includes our awareness of impermanence.

But Issa here — the context is that his daughter had just died of smallpox — essentially says, yes, I accept that this world is indeed impermanent, but somehow that insight offers no solace or consolation. You know what, when your child dies, recognizing that all life is fleeting really doesn’t seem to help matters. We care anyway, and care intensely. So the haiku philosophy of life doesn’t hold water; it doesn’t even hold the sound of water.

The word “dew” carries other freight as well. It’s a kigo (season word) for autumn, which seems apt here, though in fact Issa’s daughter’s death took place on the summer solstice (as recorded in his 1819 haibun A Year of My Life) (Hass 228). There’s a cruel irony in his daughter dying as the year turns to the season of full flowering, of life at its lushest, and there’s an irony in planting a metonym of fall there, too. Dew, of course, also carries connotations of freshness and beauty — but all these word associations are called into question in the haiku, which is to say that the poem deconstructs haiku, subverts its customary meanings. Again, though, I’m reading the individual haiku as a deconstruction of the idea of haiku, as passed on in the haiku tradition. That is, a good haiku plays by the rules of haiku but at the same time deconstructs those rules. (Okay, I know, they’re really more like guidelines than actual rules.) So I’ve been deconstructing by reading individual haiku as a kind of counterfriction against the conventions, and you’ve been saying that no, we have to look at the statement made (or not made) within the text itself. My point is that haiku are always very aware of the concerns of the genre and are actively engaged with and in dialogue with the genre, and so their meaning is in large part a product of the individual haiku’s inescapable place within a tradition. And isn’t this the nature of language itself? Any word exists only as part of a system and is meaningful only because of its difference from other words in that system. Every statement relies on the perceiver’s knowledge of the system that the individual statement exists within and has meaning within. Kigo words in the Japanese haiku tradition are only one way of reminding us of the haiku system of meaning. The form itself, the image-laden content, the syntactic abbreviation, all indicate an acceptance of the conventions of the genre, with all its expectations about all kinds of meaning — how perception works, how oneness can be achieved, what the nature of existence is. We can no more separate an individual haiku from the tradition it is part of than we can separate a word from the system in which it exists. Come to think of it, maybe that’s the point of haiku. This moment, this perception, this flower, this river, this rock — they’re all part of a system, and every moment where we perceive the hereness and nowness of such things serves as a recognition of how they all fit together in the system.

The seasonal references in haiku, besides serving as genre markers, also expand the meaning of the haiku, connecting the particular (the observation of a natural object) to the general (natural cycles and systems). This suggests that haiku is the poetry of synecdoche — as Robert Frost said in reference to his own work, “always, always, a larger significance. A little thing touches a larger thing” (qtd. in Sergeant 325). Why is this compatible with deconstruction? Because meaning is never limited to the small things that are the ostensible focus of the haiku. Always the haiku’s observation of the small thing, the natural image, is meant to say something else as well, about life or nature. So haiku language is inherently doubled—at the very least. It claims to make no claims (because it avoids abstractions and generalizations), but in truth its claims are there, even if they are only implied by image.

In Issa’s poem, then, the reference to dew is meaningful because of its symbolic associations and its role as a seasonal reference, both of which remind us of the theme of impermanence, but do so by relying on evocations that extend beyond the parameters of the poem.

If Issa’s point is to express sorrow (yes, we know the world is impermanent but still we feel loss deeply), couldn’t it also be read against that grain? Something along these lines, perhaps: “The world may be impermanent, but yet there is much joy, much lasting joy, in the beauties of the world — like dew on morning grass.” In this reading, the dew is not misleading us as to the nature of existence, but offering the consolation of beauty. Or is this haiku another case where the ambiguity is already built in? Perhaps it can’t really be deconstructed because it’s already thoroughly ambiguous as to its intended statement — it already has a double meaning, opposite meanings, built into it, hence the word doublings of “the world of dew” and “and yet.” The doubling suggests that there are two sides to life, two ways of looking at things, balanced against (or with?) one another.

But note this — there is one word in the poem that is not doubled: “is.” I won’t get into questions about what the definition of is is, but I will point out that here it functions like the fulcrum of the balance, or like an equal sign, saying that yes, the world of dew is in fact the world of dew. This can be taken in several ways — for example, as a claim that the world of phenomena (fleeting though they be) is in fact the world of phenomena, or that the world of phenomena (the actual dew) truly is evanescent (dew as symbol). But either way, does that mean that it matters intensely or that it is illusory? Either conception is potentially compatible with Zen thought and a haiku world view — and of course both are called into question by the “and yets” that follow. Another possibility is that the doubling of “the world of dew” is a challenge to language — “the world of dew,” meaning those words, really do amount to “the world of dew,” meaning the phenomenal world (“the solid earth! the actual world!” as Thoreau called it). And so Issa initially grants the premise of haiku, or its ideal, that language can be so transparent that we see right through it to the thing itself. It uses, says William J. Higginson, “words more like things than like thoughts” (152). And thus the ideal of the “wordless poem” is achieved.

And yet, and yet — even with the most unadorned language, we are reminded that there is in fact language between the reader and the actual world, and for that matter between the writer and the actual world, between any literate perceiver and the world.

And if there’s language, it can be deconstructed. The seeds of its own undoing (or undewing) lie there somewhere.


Yes, I see that undewing, and how it can be read as an expression of both (or either) sorrow and joy. As you say, the ambiguity is already built in. And it’s not like we have to dig very deep (or at all) to discover the doublings in the poem; it might as well come with a warning label: “do not read reductively; find more than one meaning here.”

Take the word “is.” Not only is it not doubled, as you point out, but it’s the word that makes the central assertion of the poem: that the stuff on one side of it is equivalent in some way to the stuff on the other side of it. But as you also so effectively demonstrate, what is equal to what isn’t clear. Any of a number of possibilities seem equally — simultaneously — not only plausible but consistent with the principles of haiku. Okay, but what if that’s not the point, for the deconstructionist? What if what we’re trying to do is not pin down what “the world of dew” means in order to then deconstruct that, but show that it is not, in fact, equivalent to whatever the other “world of dew” means, despite the poem’s linking verb? Okay, this is easy: the “and yet” takes care of that by calling “is” into question, right? Hmmm . . . but then does the second “and yet” reverse it back again, like a double negative, reinstating the equal sign of the “is”? If so, then what’s being confirmed by my attempted deconstruction is not only that the haiku is ambiguous, not only that its various possible meanings proliferate, but also that part of its meaning, perhaps its primary meaning, is this ambiguity, this relentless undoing, this “and yet.”

So to deconstruct that, would we have to show that the text simultaneously supports a reading that sees it as absolutely unambiguous? But Issa’s “dew” haiku is not merely “already open” (as is Hotham’s dog and stars haiku which led me to this very question in the first place); it seems, in fact, to be about that openness. That is, Issa seems to be making a point about the nature of language itself. What both matters intensely and is illusory, according to this haiku, is what it’s saying about language.

But what, exactly, is it saying about language? I love your suggestion that on one side of the equation, “the world of dew” refers to itself linguistically, so that the poem is making a statement about the relationship between language and reality. But I wonder . . . is saying that words = world the same as saying that words (transparently) deliver us to that world? I’m not so sure. Let me retrace my thinking to show where I get stuck: if words are on a par with reality, wouldn’t that be a hyper-reification, a treating of words as things? If so, then they can’t be transparent, much less merely referential. The idea(l) that language is transparent, or the attempt to use it as if it is, is in fact what allows deconstruction to be possible in the first place. Here’s how: saying that language is transparent means that it performs such an effective and silent signification that we don’t even see it as it’s doing its job; we see only the stuff to which it refers, as if we’re seeing reality itself, directly. Hence, the illusion and the allure of the transcendental signified, which, in the case of haiku, seems to be Nature. But language’s job, understood this way, is nonetheless to signify, to refer. And it’s in the gap between signifier and signified (there’s no essential or natural semantic connection between a word and the thing to which it refers, meanings are conventional, arbitrary, and so on) that we find difference as well as endless deferrals of meaning. And of course you’re right to insist that the systems — linguistic, literary, generic — to which the language of a text belongs are, in fact, part of that text. You offer a convincing challenge to my earlier claim that to deconstruct, we must limit our analysis to “just the text, ma’am.”

In fact, it’s the very doublings and indeterminacy that are inherent to haiku (in both form and tradition) that lead me to wonder if Issa is actually rejecting the idea that language is transparent, both by calling attention to language itself in so many ways in this haiku, including making a direct statement about the nature of language, and by making a much more radical claim: that words do not refer to the world but are of the world. Whether we take this to mean that words have a physical presence, like a rock or a tree or a person, or that words actually constitute the world of reality, we are, it seems to me, beyond the reach of deconstruction. We are either already in the place where deconstruction would eventually take us without having to engage in deconstruction itself to get there, or we are in a place where deconstruction just doesn’t go, like, say, non-Western thought, the realm of “balanced oppositions.”


Or perhaps it is the realm of no oppositions, and of no-mind (mushin), the emptying of consciousness and of self that Zen practice idealizes — and that haiku strives to enact. The classic beginner’s koan has an acolyte asking, “Does a dog have a Buddha nature?” The answer is “mu,” there is nothing, neither yes nor no, a refusal to take either side of a binary opposition, an answer that says the answer is not discoverable through the workings of the mind. (Here’s another koan, perhaps: can haiku be deconstructed?) I can think of several haiku that seek to express the quality of no-mind. One is Bashō’s “Pine Islands, ah!/Oh, Pine Islands, ah!/ Pine Islands, ah!” Just the name of a place (in Japanese, Matsushima), repeated, and some vocalizations (the “oh” and the “ahs”) that take shape as emotive sound but are not quite words, released from some place other than mind — no comment or reflection or even description (Hamill 2000, xix). There is no mind getting in the way of the interaction between the place and the emotional reaction it evokes, a reaction that comes from a disembodied source, an observer we don’t see, because there is no awareness of a self that is somehow separate from the scene.

Or is there? Someone put those words down, a certain self who called himself Bashō, who achieved such fame that we take an intense interest in all his words, so that we are in fact quite aware of his presence when we read his poem. And Bashō, like all haiku writers, it seems to me, is always aware that the use of words (which come from the mind), even very few words, means stepping out of the state of no-mind. One of the goals of haiku is to approach the state of no-mind by eliminating abstraction and explanation and by reducing even the concrete words devoted to representation to a minimum — but they are never gone completely. (Not even in Cor van den Heuvel’s one-word haiku: the word “tundra” on an otherwise blank page [163].) And as soon as there’s a word, there’s something to deconstruct. In Bashō’s Pine Islands haiku, at least in its English translation, the only words, “pine” and “islands,” suggest puns that can serve as the starting point for deconstruction. To “pine” is to yearn for something absent, to desire, which suggests preoccupation with self and to a Buddhist is the source of suffering. And “islands” can be read as “I-lands,” the territory of the ego. So there is no dissolution of self in world here, no stepping into no-mind, but self-absorbed yearning — yearning perhaps to take possession of those islands, or at least to be over there (instead of “being here now”), no longer separated from them by water. I won’t pretend to be able to unravel many of the possible layers of language in the original Japanese, but I will note that the Matsu (“pine”) in Matsushima might not exactly obliterate a sense of self for someone named Matsuo Bashō.

What I’m getting at, what I’ve been getting at, is that the supposed ideal of “wordlessness” of haiku, meaning that its language can represent the natural world in such a way that it becomes fully present in language, in seventeen syllables or less, is a fiction. But the best haiku are aware of the fiction and of the difficulty or impossibility of using words to achieve no-mind, or selflessness, or wordlessness. Bringing deconstruction to bear on haiku reveals that even haiku to some extent concern themselves with the problematics of representation, and recognizing this enriches our readings of haiku.

At the same time, I will also admit that the claim of deconstruction, at least in the extreme (and simplified) form that we’ve been playing around with, suggesting that language always refers only to itself and ultimately is incapable of any actual representation of reality, is also a fiction. What’s interesting in bringing these two worldviews into conjunction is that they stake out opposite extremes on the question of how language relates to the “real world.” In truth, both end up occupying some middle ground. I like how Robert Scholes puts it: “we neither capture nor create the world with our texts, but interact with it” (112). Deconstruction has been more engaged with the real world than the popular dismissal of it as mere language play would suggest; as you point out, it has been put to use in the service of the “real world” by, for example, feminist critics and new historicists, who have used deconstructive techniques to expose the power structures governing literary texts and the cultures they come from.

On the other extreme, it seems to me that haiku has been more engaged with the problematics of representation, especially as it pertains to the natural world, than it is usually given credit for. In Scholes’s terms, haiku don’t “capture” the world, they “interact” with it. Where we seem to be finding common ground is in the recognition that good haiku are provocative enough or suggestive enough — and most importantly self-aware enough about the nature of language — to resist making any claim too easy or self-evident. And what we’ve been looking at here are interesting test cases for both the power of haiku to transcend language itself and the efficacy of deconstruction. Even with so few words, words that seem to fully embrace the freewheeling possibilities of suggestion, even here, especially here, language repays close attention to the ways in which every utterance communicates more than it, or we, can ever know.

Maybe this is the time to return to a key issue you brought up earlier: what is revealed about human relations with the natural world when we bring together deconstruction and haiku? Deconstruction’s reliance upon the Sausserian conception of language as a set of signifiers in relationship with one another, interacting with one another, with each signifier meaning something only in terms of relation to other words, reminds me of descriptions of ecology, itself a web of relationships and interactions. Deconstruction essentially describes an ecosystem of words. Haiku, maybe, inhabits the ecotone between these two ecosystems, nature and language. It is the place where language and nature meet. Ecotones, we know, are perilous places, where predators appear from both sides of the margin, but they are productive places, too, abundant in diversity. Here we are, rambling out of the woods and into the words, tracing the borderline between them, following the path of haiku.

Let’s look at one more in this rich ecotone, another classic by Bashō — one that seems very much concerned with issues of representation and with the difficulty of trying to interact with the world in a state of no-mind while at the same time trying to capture that experience in any kind of art, linguistic or otherwise. Ultimately, this one seems to me more about words than the woods, but maybe you can help me find the path that connects the two:

The spring we don’t see
on the back of a hand mirror
a plum tree in flower
          — Bashō (in Hass 28)

On the surface, this haiku suggests that we are so self-absorbed we cannot see past ourselves to the actual world. Or it could be that when we look at the world, at spring, at nature waking, what we see is really just a reflection of ourselves. What we miss is something, some spring, beyond that, impenetrable because we cannot see beyond the reflection of ourselves. It doesn’t seem too much of a stretch to suggest that language might be that mirror — we can’t see a world beyond ourselves because our way of seeing and of knowing is so bound up with language. It sees the world for us, constructs it for us. And so we can’t help but see the natural world as resource or as symbol, and spring becomes a lesson in rejuvenation for us. But here’s the thing — if we could manage to see past the reflection of ourselves, what would we see in that beyond? What is the spring we don’t see? Not nature itself, says Bashō, but a representation of it, a drawing of a plum tree in flower. Bashō wants to privilege the other side of the mirror, but isn’t it as opaque as the reflective side? We still can’t see through it to an actual plum tree in flower. Even if we get beyond anthropocentrism, we are still boxed in by artistic convention or language. Maybe it’s not a prison — maybe it’s a decorative box with mirrors and pictures on the inside — but it’s still a container of mind holding us in, limiting our contact (contact!) with the actual world. It should be easy enough to put down the mirror, but all we do is endlessly turn it front to back and back again.

Haiku is the drawing on the back of the mirror. Look at the plum tree in flower there.


Yes, I see it . . . though I’m not sure if the one I see is the same one that’s on the back of the mirror. Again, I find it interesting that by reading individual haiku against the conventions of the genre, you’re revealing some deeper tensions and inconsistencies within the tradition itself — in the case of Bashō’s hand-mirror haiku, the impossibility of apprehending the actual world beyond human representations of it. Even as we’re invited to see the plum tree, we’re denied access to “the spring” in which it blossoms. But here in this liminal space, if the haiku does indeed occupy the sort of ecotone you posit, it (spring) might in fact encroach. It might steal over into the haiku from its natural habitat. Or it might have always already been here, if not readily visible.

But before I go scrambling about in the ecotone myself in an attempt to show you what I mean, I have to say I find your idea of the ecotone to account for haiku’s problematics of representation quite compelling. Though an ecotone where language and nature might meet is not exactly an ideal controlled environment for the deconstructionist, this is precisely what makes it work for me. And it explains something about our process so far. Even after reconceptualizing the apparent oppositions and contradictions in haiku as functions of either balanced opposition or no-mind (not interchangeable concepts, but perhaps compatible?) rather than binaries — and thus rendering haiku theoretically undeconstructable — our attempted deconstructions have been pretty revealing. So it’s both impossible and rewarding to deconstruct haiku. Maybe the answer to the question “can haiku be deconstructed?” is “mu.” But I wonder if another (albeit far more Western) way of answering the question might be “yes,” but with a caveat: the end result will be something other than, and often richer than, a deconstructed haiku.

Back to Bashō, then, to see if I can find that path between word and world. As I understand it, your analysis of Bashō’s haiku depends upon a reading of the poem’s syntax along these lines: If we take the prepositional phrase “on the back of a hand mirror” as modifying the noun phrase that comes before it, then “The spring we don’t see” is the one that’s on the back of the mirror. Next, if we take the final noun phrase of the poem, “a plum tree in flower,” as an appositive to the first, a renaming of that now modified noun phrase, then the flowering plum tree is on the back of the mirror. But there are other ways to read these same three lines, other relationships and meanings implied. (At least there are in this English translation. In Japanese, it’s quite possible we would find a kireji, or cutting-word, to indicate syntactic breaks.) For instance, what if the relationship between the first (two-line) noun phrase and the plum tree line is one of juxtaposition rather than equivalence? What if the spring we don’t see is the one on the back of the hand mirror (whatever that is), and the flowering plum tree, as the clearest image in the poem, is the spring we do, in contrast, see?

Here’s another, more radical, possibility (and I think I’ve followed the poem all the way into its ecotone for this one): What if it isn’t the other side of the mirror that Bashō wants to privilege, but the world that isn’t on either side of the mirror? If the prepositional phrase modifies the noun phrase that follows it, rather than the one before it, we get “the spring we don’t see” as one thing, and the plum tree painted on the hand mirror as the second thing. Then, if these two nouns juxtapose rather than rename one another (either is possible, yes?), then the plum tree is the image we do see, and it’s a painting on the hand mirror. So we see a representation. What we don’t see is “the spring we don’t see.” In fact, it’s so unseen that it doesn’t even come to us in the form of an image in the poem, only as an abstraction. (Is this its camouflage? In this ecotone, is it an alien species following an old game trail across an unmarked border?)

Though it does have this presence in the poem, Bashō doesn’t even attempt to represent the actual spring, since we won’t see it as long as we’re looking only at representations (such as the painting on a hand mirror, the words in a haiku). We do still “see” the flowering plum tree, but we’re seeing it for what it is, a representation. So the poem almost asks us to look away from it, to look elsewhere, maybe out there in the world, for the real spring.

According to this reading, it’s not the painting on the hand mirror that’s privileged, or even the flowering plum tree that the poem does show us; it’s the natural world outside of and beyond language. But I don’t mean this in the sense that would reinstate Nature on the throne of the transcendental signified. It’s been duly knocked off, not by any deconstruction of the poem performed by us, but by the haiku itself — as we discover through the process of testing its deconstructability. The poem not only decenters Nature, but also problematizes our assumptions about our ability to know the natural world in such a way that what we actually end up with is a haiku potentially far more respectful of nature than would be one claiming language as transparent.

And here’s where I find a viable social relevance in asking if haiku can be deconstructed: such inquiries might expose a range of ideological assumptions inherent in the idea(l) of the form as well as the ways it — and its authors — often subvert these very assumptions. Is Bashō on to the fact that language can never be wholly transparent, and so rather than try to use it as such, he’s warning us of the seduction of representation? Perhaps he is telling us that if we want to know the plum tree, reading a haiku about it won’t do the job; we have to go to the plum tree.

So let’s get out for another walk in the woods, shall we?


     Unseen chickadee
     leaf shadow
     the sound of breathing


The presence of absence?



Anzaldúa, Gloria. 1999. Borderlands/La Frontera: The New Mestiza. 2nd ed. San Francisco: Aunt Lute.

Barthes, Roland. [1970] 1982. Empire of Signs. 1st American ed. Trans. Richard Howard. New York: Hill and Wang.

Culler, Jonathan. 1982. On Deconstruction: Theory and Criticism after Structuralism. Ithaca: Cornell Univ. Press.

Hamill, Sam. 2000. “Translator’s Introduction.” In Narrow Road to the Interior and Other Writings: Matsuo Bashō. Trans. Sam Hamill. 1998. Reprint. Boston and London: Shambhala.

Hass, Robert, ed. and trans. 1994. The Essential Haiku: Versions of Basho, Buson, & Issa. New York: Ecco.

Higginson, William J. 1985. The Haiku Handbook: How to Write, Share, and Teach Haiku. New York: McGraw-Hill.

Sergeant, Elizabeth Shepley. 1960. Robert Frost: The Trial by Existence. New York: Holt, Rinehart and Winston.

Scholes, Robert. 1985. Textual Power: Literary Theory and the Teaching of English. New Haven: Yale Univ. Press.

van den Heuvel, Cor, ed. 1974. The Haiku Anthology. 1st ed. New York: Doubleday Anchor.

——. 1999. The Haiku Anthology. 3rd ed. New York: Norton.

Back To Top