The Poetics of Craquelure: Behind the Poem "青 / Qing Ghazitsu"

Nicole W. Lee introduces her readers to the ghazitsu, an invented form derived from the ghazal, the zuihitsu, and the Chinese language. Lee narrates her creation of the ghazitsu, transforming dissatisfaction with conventional forms and grief into poiesis, blending literary and linguistic traditions—both distant and intimate—into a poem with new aesthetic possibilities. At once a manifesto for the form and an exploration of Lee's relationship to Chinese languages and cultures, this essay imagines one of poetry's possible futures through attention to its many histories, voices, and meanings. The ghazitsu is a reminder that there are always new poems to be written. Lee's poem "青 / Qing Ghazitsu" was featured In Issue 75.1.

It seems a small mercy that the gap—that dynamic emptiness arising from absence—has always been a fertile space for me. Having spent most of my life trying to fit into predetermined boxes, I’ve long learned that the unfilled gaps in the corners are often sites of revelation. After years of suffering from the fatigue and other dysregulations of long covid, a second miscarriage seemed to be yet another distance between what I wanted and what I couldn’t have. Circling the park, the sky’s white nerves pressing down on me, I found myself at an emptiness at once calm and unsettling. Again my body had failed me; again, I could not fit my poems into the ghazal and zuihitsu, whose strict containers I had for months been trying to fill. But perhaps that was the point. In my study, surrounded by the ephemera of my life—compression socks, a half-assembled manuscript, a still unused baby carrier—I was reminded that their seeming non-importance was precisely the stuff of form. That craquelure is its own exquisite texture. It was from this rupture—this gap—that “青 / Qing Ghazitsu,” and its form, the ghazitsu, was born.

The ghazitsu is a hybrid form that arises from three traditions: the ghazal, the zuihitsu, and the Chinese language. The ghazal—whose root meaning is “spinning”—is a poetic form that consists of autonomous couplets linked by a strict rhyme and refrain structure, often about the beloved, desire, or metaphysics. Originating in 7th century Arabia from the songs of Bedouin tribes, it was codified into the structure we know today between the 10th and 12th centuries in Persia. It’s from these codifications that the ghazitsu takes its autonomous couplets, as well as the radif (the repeated refrain at the end of each couplet), the qafia (the repeated rhyme before the radif), the matla (the opening couplet with repeated qafia and radif at the end of each line), and the maqta (the final couplet with the poet’s pen name). The zuihitsu is a hybrid genre of prose and poetry, characterized by a “sense of spontaneity,” irregularity, and fragmentation. It’s these unique qualities—asymmetry, anecdote, digression, and an accommodation of lists and stray observations—that the ghazitsu maintains. And from the Chinese language, the ghazitsu draws on the dissonance between the visual and sonic rhymes of its written and spoken traditions by transforming the repetition of a homophone—in this case, qīng 青—into a radif-like refrain that changes meaning with each couplet.

While it seems now inevitable to me that these elements would come together in this way, like many creative accidents, it was not preordained. For several years I had been deep in the gap between writing into the strictness of these receptacles and my ability to satisfactorily do so. The ghazal, for example, is an intimate form for me. An amatory prelude to the Arabic qasida that arose as its own structure, I’ve come to associate its yearning voice with my beloved, who can frequently be found reciting Mowlavi—Rumi—in the shower. In Persian, its hypnotic rhythms, governed by aruz, a system of quantitative prosody, are entirely different to English rhythms, whose meters are quantitative, or stress-timed ^#. While I adored these rhythms, which often seemed to resound deep in my body, my attempts to replicate them seemed false. There was also something about the voice of a ghazal—its ravishing intensity, its terseness, its ecstasy!—that seemed to refuse sublimation in my own writing. All these anxieties resulted in my constant feeling of being intimidated by it—that I, a person of Chinese descent but raised in the West by parents born in British-controlled Malaya, was too Western. The unfinished attempts remain in a drawer.

The zuihitsu similarly felt at once closer and more distant. The word zuihitsu is the Japanese reading of the Chinese suí bǐ 随笔 “following the brush.” While the term and the spirit arose from the intellectual and philosophical essays of 10th century China of the same name, the zuihitsu’s wit and fragmentary list structure, made famous by Sei Shōnagon’s 11th century The Pillow Book, has much more in common with the earlier Chinese genre of “miscellaneous morsels,” or zá zuǎn 杂纂. This genre can be seen in the works of the 9th century Chinese poet Lǐ Shāng Yǐn 李商隱, whom I’ve long read and adored. Yet while I felt more familiar with the genre’s fusing of external scene with internal emotion, I still struggled with meeting the zuihitsu’s ironically strict demands for imperfection. Perhaps my inability to achieve what the Chinese call jù jué yì bù jué 句绝意不绝 “lines that end, but meaning that doesn’t end” meant I wasn’t Chinese enough.

However, the longer I sat with these two poetic structures from either side of Asia, the more I saw resonances between them. Both the ghazal and zuihitsu embody the spiritual and profane. Both seek to create atmosphere through the speaker’s lyric voice and associative motifs. Both feature repetition: if the repetition in a ghazal is lexical, transforming the radif with each return, the zuihitsu’s repetition is in its returning to the same environment or theme. These in turn are not so different to the Yì Jīng’s 易经 “Book of Changes” hexagrams which repeat and recur. And both forms are built on fragmented, autonomous thought, arguably arising during a time before linearity, shaped by the printing press and mass production, was prized; a time when knowledge arrived spontaneously, all at once. Again, this was of particular interest to me, as I’d also been exploring the same quality in the formation of Chinese characters.

It was these similarities that led me to wonder if combining what I loved about each tradition could perhaps loosen some of my apprehensions about each form. For example, I’d often felt uncertain about ordering lines in a zuihitsu, or even what to write about, given its emphasis on appearing spontaneous. But bringing in the ghazal’s predetermined qafia (pre-refrain rhymes) combined with a homophonic Chinese character radif (the refrain) gave me a structure—and often ideas from the radifs themselves, since they were different words—to lean on. Likewise, injecting the zuihitsu’s irregularity into the couplets allowed me to avoid my anxiety of sounding staid and repetitive. And of course, the semantic and phonetic aura of the Chinese characters gave the poem embedded themes. Putting it all together, I began to see how the three traditions, mingling in the gap, might begin to talk to one another. By the time I emerged from the study, I had come to the realization that an existing form could no longer hold my grief. That despite my body’s inability to create, I could still be generative. That despite this anguish, I could still gather the detritus of my life to form a new way of being.

[ ]

I have to confess, I do have one major gap I haven’t yet been able to fill: my lack of fluency in Chinese. As a child who wanted to go to the mall on Saturdays with her white, Doc Marten-wearing friends and not Chinese school, my rebellion was sitting in the front row of class reading fashion magazines. As an adult, teaching myself Mandarin—the standardized form of modern Chinese—has been a humbling task. As a speaker and listener, I can order from a menu if I focus intently. As a reader and writer, I’m child-like. These shortfalls also mean I have no hope of achieving spoken fluency in my parents’ Chineses—Hakka and Teochew, which have their own tones, pronunciation, and grammar. On the other hand, while vernacular differences still make the written word mutually unintelligible, there’s one thing that does unite them: a classical and modern literary language written in a common script. Largely bridging the two hundred languages that make up the Sinitic language family across China, as well as parts of Southeast Asia and the diaspora, this literary language crosses not only spoken intelligibility, but also time—three thousand years of access to poets’ hearts and minds—and is singular in its ability to connect me, however limitedly, to my linguistic lineage.

Chinese characters contain sonic, semantic, and visual repetition that can be exploited poetically in a multilingual practice. To demonstrate this, let’s look at a line from the famous poem “Jìng Yè Sī 静夜思,” or “Quiet Night Thoughts,” by the Tang Dynasty poet Lǐ Bái 李白. The poem traces an exiled poet thinking of home as he looks at the moon from his bedroom. The third line of the poem is presented with pīn yīn (the Romanized transliteration), traditional Chinese characters, and English word-for-word translation below:

jǔ tóu wàng míng yuè

舉 頭 望 明 月

raise head gaze bright moon

Chinese, like Egyptian and Mayan hieroglyphs, is a logographic language. This means that each symbol or character represents a meaningful unit of language. Some of these are pictograms, which directly depict the thing being described. For example, the radical (a component of a character) rì 日 arose from a drawing of the sun. Similarly, the radical yuè 月 arose from a drawing of the moon. Adding them together creates a compound ideogram míng 明 “bright”—the two brightest objects in the sky.

It would be easy to assume that Chinese characters are therefore “little moving pictures,” but these characters collectively only make up around 10-15% of Chinese characters. The remaining majority, around 80%, are mostly phonosemantic characters. In a phonosemantic character, one or more radicals represents a semantic component, while the remaining radical represents a phonetic component. This can be seen in the word wàng 望 ^# which means “to gaze/hope.” Wàng 望 is made up of two semantic components: the pictogram yuè 月, and the ideogram tǐng 𡈼, which represents a person standing; and one phonetic component, wáng 亡, which means “to die/lose.” When put together, the character illustrates a person standing looking at the moon, with a sense of distance, or loss—“to hope.”

Notice, however, that wáng 亡 doesn’t merely provide a phonetic component, but a sense of meaning as well. This is a subcategory that is growing in linguistic research: possibly up to 50% of phonosemantic characters contain phonetic components that simultaneously hold a semantic function. In modern linguistics these components are known as semantophoric phonetics. The characters they help form—where both the semantic and phonetic components contribute to meaning—were first identified over 2,000 years ago by lexicographer Xǔ Shèn 许慎 as yì shēng zì 亦声字 “also-sound characters.” These characters also naturally cluster into groups known as phonetic series: families of characters linked by a shared phonetic component and common conceptual thread. Using the wáng 亡 family as an example, we would find wàng 忘 “to forget,” whose semantic pictogram xīn 心 “heart/mind” combines with the semantophoric wáng 亡 to impart the sense of a loss of the mind; and the literary wǎng 罔, whose semantic pictogram wǎng 𦉰, representing a fishing net, an image of ensnarement, fuses with wáng 亡 to create the meaning “deception.” The convergence of these components—the sonic (the wang sound), the visual (the repetition of 亡), and the semantic (the resonance of meaning)—comes together to create a triple aura that poets have used consciously and unconsciously across centuries.

It was this awareness of the sonic and visual recurrences of some Chinese characters that led me to connect the character qīng 青, one of the largest, most semantically coherent, and therefore primary phonetic series in the Chinese language, to the ghazitsu. But before I continue, go back and have a look at the line from the poem, which ends with three characters containing the character yuè 月 “moon.” Consciously acknowledged or not, as the speaker raises his head, he gazes (望) with the most complex character towards a complex ideogram of brightness (明), finally ending on the single image of the moon (月)…

[ ]

In my grief, I wandered the streets of our suburb, thinking about qīng 青 and its relationship to blue. I had learned of the character years ago when working on a poem trying to trouble the color blue’s emotional tone. Beloved by many, including Maggie Nelson in Bluets, and Joni Mitchell in Blue for being the color of melancholy and serenity, it was surprising to discover that while in Chinese, qīng 青 could mean blue, it could also mean green, and black—the “color of nature”—while its deeper metonymic tenor was close to something like “vivid essence.” As I familiarized myself with its phonetic series, which contained words like qíng 晴 “clear sky,” qíng 情 “feeling,” and qǐng 请 “to request,” I realized I could maybe write a ghazal that would sonically and conceptually vibrate with the melancholy and clarity of this vividness.

I had long thought about writing a ghazal with a repeated Chinese character as the radif, but once I noted qīng 青’s mercurial nature, the poem began to take on a different shape. It occurred to me that using Mandarin homophones as a radif-like refrain meant that the poem could not be a true ghazal in the orthodox sense, since a ghazitsu’s refrain is a repetition of sound rather than of the same word. While the written Chinese characters are largely standard for all Chinese readers, their pronunciation can be radically different. Read aloud in Cantonese or Shanghainese for example, the poem would be a series of autonomous couplets that mostly share a common rhyme but lack the identical initial consonants required for a radif. Other languages, such as Hakka, split the words into different rhyme schemes, depending on regional variation. Changing up the couplet’s line lengths then gave the poem a zuihitsu’s breath, while still maintaining some of the ghazal’s repetition.

Reading cí 词 poetry, the classical Chinese poetry form written in fixed irregular lines whose structures arose from popular song tunes, provided another insight. While cí 词 poetry had strict rules about rhyming and tone patterns, most classical poems would only rhyme if read by the centralized lingua franca at the time: Middle Chinese. Since pronunciation of Chinese has always been a negotiation of power structures and sound shifts across borders and throughout time, it seemed evident that embracing irregularity of both line lengths and pronunciation could also honor my linguistic lineage. While we often imagine an orthodox classical China united in language and culture, the reality is that much of its literature was written in a codified system of rhymes that were determined by the power centers of the time, and whose auditory pleasures were received differently across its vast territories. These differences arose from sound shifts across literal geographical borders, as well as divergence of pronunciation over time. (After all, if you read classical Chinese poetry in Cantonese, Hakka, or the Min languages, the rhymes are much closer to Middle Chinese than modern Mandarin.) In this way, the ghazitsu, written in the current standardized Mandarin—and at this particular moment—can be seen as an extension of historical precedent. Irregularity and differences in pronunciation, which are still present today, have long been, and will continue to be a part of the Chinese tradition.

Another lineage I wanted to honor was the “crunchy”-ness of obscure words. Like Garth Greenwell, I’ve always loved strange familiarity of words that have fallen out of use. Words such as “vitrescent” are appealing precisely because you can often guess their etymology (something to do with glass?), while also indulging in its sonic pleasures. While I found most of the words within the qīng 青 phonetic series visually or conceptually poetic, many of them were obsolete. The character qīng 靘, a semantophoric compound, is deeply archaic, having fallen out of use after the Tang and Song dynasties. However, I knew I could trust a Chinese reader to guess from the ghazitsu’s structure that the character’s sound must be qing, and that they would know the character sè 色, which means “color” or “appearance.” Putting this together, they could make an educated guess that the character might mean “blue color” or “blue appearance,” which would be correct. In classical texts, qīng 靘 referred to a lustrous blue-black cosmetic pigment, reminiscent of the night sky or of indigo dye, used to paint eyebrows on a face. I hoped that this character would conjure both the color and the thinness of an eyebrow when envisaging the vein in a moment of vulnerability in the poem.

This “crunchy-ness” however, couldn’t bolster all my intentions. Through a variety of sound shifts over time, many of the words in the phonetic series had different pronunciations in Mandarin. In the word jīng 睛, for example, meaning “eyeball,” the Old Chinese initial sound was most likely very close to qing, but diverged over time. To keep a strict ghazal-ness, at least in Mandarin, I decided to include other qing homophones, with the hope that its conceptual halo would hover over the others. A close reader of the final couplet would ideally note that the rare and phonetically obscure qìng 綮, made up of the phonetic qǐ 啓 “to reveal” and semantic mì 糸 “silk thread,” which later came to mean an “essential juncture,” might amount to an image of “the revelation of an essential juncture,” or a “crux.” And of course, resonating through it all, the sense of blue, oh!—and there again, blue—

[ ]

It seems fitting for “青 / Qing Ghazitsu” to have been included in a multilingual issue of Shenandoah, a journal which came to prominence in part through its relationship with Ezra Pound. Proclaimed by T.S. Eliot as the “the inventor of Chinese for our time,” Pound was also a racist, an anti-Semite, and fascist. Many believe that through his selective editing of Ernst Fenollosa’s 1906 essay “The Chinese Written Character as a Medium for Poetry,” Pound took from the classical Chinese language what he wished to see—Fenollosa’s characterization that all Chinese characters are “continuous moving pictures” being one, Pound’s appropriation of Confucian ideals without their attention to ethic and communal care being another—and discarded the rest. Because he promoted his ideas without attribution, many of my generation think of Pound as part of the post-World War I European experimental movement, rather than a curator who shaped the entirety of Anglophone literature through his misinterpretations and appropriations of classical Chinese poetry and Confucian ideals. As a translator he has cursed English readers to think of classical Chinese poetry as stilted and end-stopped. (By contrast, read the Chinese-born Shangyang Fang’s Study of Sorrow, and feel the breath of a translator who deeply understands the rhythms, imagery, symbols, and references of both Chinese and English).

And yet many—including the Méng Lóng Shī Rén 朦胧诗人 “Misty Poets,” a group of 20th century Chinese poets who in the 1980s ironically looked to Pound while trying to revive the visual, imagistic qualities of classical Chinese poetry—also concede there were some “luminous” elements in the Chinese language that Pound and Fenollosa saw that do exist. Some characters can convey pictorial images; meaning can be created relationally through a character’s components; grammatical structures are often paratactic. But it’s not universal or without context. As Marilyn Chin once said, Pound shouldn’t have the last word on Chinese poetry. My sense is if you’re going to “make it new”—which itself derives from Pound’s interpretation of a phrase from a Confucian classic—it’s often a good idea to know who made “it” before, in what context, and why. I don’t mean to say a writer has to be an expert in these traditions—I am most certainly not. Migration, language loss, and assimilation have all widened the gaps in my own knowledge. But I think it’s in many poets’ best interests to fill these absences as best as we can—giving ourselves as much grace as possible, while acknowledging there is still much to learn about our own inheritances.

Creating the ghazitsu, for me, was one such filling. Having facilitated a form that honors what I love most about each tradition, rather than strictly fitting their codifications, I’m now able to play in the gap where, much like my own existence, these boxes intersect. In this gap, there’s no need to choose between the rigid rhymes of the ghazal, the spontaneity of the zuihitsu, and the tension between the visual and aural rhymes of the Chinese language—the ghazitsu can hold them all. It is perhaps the only container of my life so far to be able to hold my failures while still making a whole. I will never be truly “Chinese,” whatever that means. I don’t know if I will ever be able to truly capture the gazelle-like voice of a ghazal. I face a future in which I perhaps will never regain full health. But perhaps these fragments, these half-lives, these stains at the bottom of the cup—perhaps this leftover patina is, in fact, enough.

The box, after all, is an illusion. Go back far enough, and any form is a hybrid: of song, oral storytelling, formal systemization, of lexicons coming across borders. New forms lean on what has come before: they shred, rearrange, mixtape. While I hope the ghazitsu holds its shape—even as some days I feel the form is more “-itsu” than “ghaz-”—I’m aware it’s merely a meeting point along what is hopefully a long lineage. Just as the ghazal, the zuihitsu, and the Chinese language itself evolved from earlier structures, and continue to evolve, so too will this form one day give way to new ones. In time there will be new fractures, and therefore new fragments to be gathered from the litter of our increasingly asynchronous, transnational, and simultaneous existences. And at their corners will be many more gaps—gaps that not only arrive with the promise of play, texture, and creation, but the invitation to pick up the shards of a broken life, and with grace and courage, fill them.