Writing gender into Chinese script: The feminine third-person pronoun in modern Chinese language reform

Introduction

In 1923, the influential musician Li Jinhui (黎锦晖, 1891–1967)—younger brother of the linguist Li Jinxi (黎锦熙, 1890–1978)—published a song titled “He. She. It.” (“Ta. Yi. Tuo.” 他. 她. 牠.) in the children's magazine Little Friends (Xiao pengyou 小朋友) (Li, 1923). Intended as a pedagogical tool, the song introduced young readers to yi 她, the newly coined feminine third-person pronoun, and to a gender-differentiated pronominal system then emerging in the nascent national language. The necessity of explicitly teaching these pronouns, however, reveals a stark disconnect between the proposed national standard and the local speech of its learners. As both the feminine yi 她 and the neuter tuo 牠 later underwent shifts in writing and pronunciation, they came to embody a broader struggle: the difficulty of aligning reformist ideals with the realities of language in use. By examining the written evolution of the feminine third-person pronoun and the early debates it provoked, this article treats these reform efforts as a distinctive site through which the structural tensions of China's linguistic nationalization were articulated. The history of the character thus serves as a critical diagnostic of the broader challenges of modern Chinese language reform.

The early history of the feminine third-person pronoun unfolded within precisely these structural tensions. Historically, written Chinese employed a gender-undifferentiated pronominal system. ¹ Beginning in 1916, reformers (see, for instance, Liu, 1920a,1920b,1934; Zhou, 1918a) advocated a singular feminine third-person pronoun. The first documented orthographic attempt appeared in 1918, when Zhou Zuoren (周作人, 1885–1967) introduced the compound “他_女” in his translation of the Swedish story “Reformförsök.” In 1920, as curriculum reforms replaced “classical literature” (guowen 国文) with “national language” (guoyu 国语), debates (see, for instance, The New People, 1920, Special section: the debate over the feminine pronoun “她”) intensified (Xinren, 1920) over how—or whether—to incorporate such characters into vernacular textbooks. Although a provisional consensus favored the character 她, its pronunciation remained contested. Disagreements over 她 and other proposed forms persisted for decades. Not until the late 1940s did the pronoun stabilize in its present form: written as 她 and pronounced ta. Consequently, while modern Mandarin retains the graphic distinctions envisioned by early reformers, it has abandoned the phonetic differentiation that Li Jinhui's song once sought to instill. ²

Modern Mandarin speakers now take the character ta 她 for granted. Yet, what made its formation so protracted and contentious? Classical Chinese has long relied on a gender-neutral pronominal system, while Sinitic dialects possess their own lexical means of referring to female subjects. What need, then, did a new singular feminine third-person pronoun serve? The pronoun, coined during the New Culture Movement, clearly functioned to establish equivalence between Chinese and Indo-European languages in translation. ³ How, however, did this translational expedient become a perceived necessity for the national language? Why was a feminine ta 她 eventually regarded as indispensable to a modern Chinese script?

In his richly documented study, A Cultural History of the Chinese Character “Ta” (She): The Invention and Social Recognition of a New Feminine Pronoun (“Ta” zi de wenhuashi: nüxing xin daici de faming yu rentong yanjiu “她”字的文化史：女性新代词的发明与认同研究), Huang Xingtao (2015a) presents a thoughtful explanation for the character's eventual acceptance. Huang (2015a: 167–204) argues that the adoption of the character was not, in essence, a byproduct of western linguistic hegemony; rather, it was “fundamentally linked to the demands for modernity—or, put differently, the needs of modernization—that were newly activated in the language in the new era” (2015a: 174). In this view, the character's “Westernness” and its modernity merely “happened to coincide” (2015a: 174). By these “demands,” Huang (2015a: 173–182) invokes concerns such as conceptual precision, practical convenience, vernacularization, and gender equality. While illuminating, this framework remains too expansive to account for the specific historical dynamics that shaped the pronoun's formation. Here, “modernity” functions less as a precise analytical category than as an umbrella term under which heterogeneous—and at times contradictory—reformist goals are subsumed. ⁴ Such a framing risks obscuring the concrete tensions among script, speech, and language pedagogy that rendered the feminine pronoun a site of such prolonged debate in the first place.

The question of sound introduces another layer of complexity. If the goal of reform was to modernize the language, why did the phonetic differentiation so central to Li Jinhui's song eventually disappear? How might the linguistic function of ta 她 have differed had it remained phonetically distinct as yi? Huang (2015a: 168–173) claims that this loss of phonetic distinction was a strategic response to the demands of vernacularization. In his account, the homophony of the three third-person pronouns—all pronounced ta in modern Mandarin—mirrored the phonological patterns of most Sinitic varieties, thereby promoting a closer alignment between speech and writing. ⁵ This interpretation might appear suggestive, yet it depends on a series of conceptual conflations. For instance, Huang (2015a: 168–173) uses baihua hanyu 白话汉语 (vernacular Chinese) and kouyu 口语 (the spoken tongue) interchangeably—a slippage that effaces the critical distinction between dialectal speech and a national spoken standard. This imprecision, in turn, obscures the boundaries between the Chinese writing system, the written representation of dialectal speech, and the modern written vernacular. At the level of terminology, then, Huang's analysis blurs linguistic domains that reformers themselves sought to differentiate. A second challenge concerns Huang's understanding of vernacularization. For reformers such as Hu Shih (胡适, 1891–1962) and Zhou Zuoren, the objective of the “vernacular movement” (baihuawen yundong 白话文运动) was not to anchor writing in local dialectal speech, but to establish greater consistency between the modern written vernacular and an emerging national spoken standard. By interpreting phonological convergence of the three pronouns as evidence of successful vernacularization, Huang misidentifies the reformers’ target of alignment, thereby attributing to them a linguistic agenda they did not, in fact, pursue. More fundamentally, however, Huang's argument collapses the distinction among script, word, and concept. He (2015a: 168–169) asserts that, in the colloquial speech of most regions in China, the third-person singular pronoun is 他, whose pronunciation is inherently ta. This statement treats written characters as if they directly encode dialectal lexical items. Yet, Sinitic varieties do not operate by “reading aloud” characters from a shared writing system. Rather, these varieties possess their own deictic forms for third-person reference. While this deictic form may be pronounced ta and written as 他 in certain Mandarin-speaking regions, it is phonetically and orthographically distinct in other dialectal areas, such as being pronounced /hhi³/ and written as 伊 in Shanghai and parts of Zhejiang province. Such deictic words in dialect speech exist as lexical entities independent of the writing system. By overlooking this basic distinction between written characters and spoken morphemes, Huang's account ultimately obscures the structural tensions that made the formation of ta 她 into such a persistent site of debate.

This article moves beyond established interpretations that frame the pronoun in terms of linguistic Europeanization or broader processes of modernization. It situates the formation of the feminine third-person pronoun within the project of linguistic nationalization in early twentieth-century China. In particular, I examine three pivotal moments in the early history of the pronoun, episodes that have surfaced in prior scholarship primarily as archival references, but have not received sustained analytical attention: Zhou Zuoren's 1918 experiment with “他_女,” contested proposals regarding pronunciation, and Li Jinhui's 1923 pedagogical song “He. She. It.” In 1918, when Zhou first employed 他_女, the modern written vernacular was still in an early stage of development, lacking both a stable literary form and a standardized spoken counterpart. I argue that, for reformers such as Zhou Zuoren and Liu Bannong (刘半农, 1891–1934; also known as Liu Fu 刘复), the new pronoun served not only as a lexical innovation but also as a practical opportunity to explore the stylistic possibilities of the modern written vernacular. I further contend that these efforts revealed—and were constrained by—entrenched tensions between national writing and local speech. In practice, the reformist vision was complicated by the discrepancy between intellectual ideals and the uncertainty faced by readers navigating these unfamiliar forms.

Taken together, this article treats the feminine third-person pronoun as a distinctive vantage point from which to analyze the evolving relationship among the Chinese writing system, diverse Sinitic speech forms, the national spoken language, and the modern written vernacular. Seen in this light, the emergence of ta 她 brings into focus the structural negotiations required to align script, speech, and linguistic standardization. Ultimately, the case of the feminine third-person pronoun reveals that China's linguistic nationalization was not merely a matter of institutional decree, but a process shaped by the material and functional negotiations embedded in writing itself.

Zhou Zuoren's experiment with “他_女”

In 1918, Zhou Zuoren designed—or, in his own words, “fabricated” (duzhuan 杜撰)—a curious graphic form for the singular feminine third-person pronoun: “他_女.” With its hybrid structure and no specified pronunciation, this symbol emerged in Zhou's translation of August Strindberg's (1849–1912) short story “Reform” (“Reformförsök” 改革). ⁶ Anticipating readers’ questions, Zhou explained both his choice of text and his unusual orthographic invention in the preface to the translation:

This piece is also from the short story collection Getting Married. It reminds me of a similar story, “Her Daily Life,” by the Japanese writer Toshiko Tamura. Tamura, however, is a “new woman” and treats this issue with great intensity, whereas Strindberg, a Misogynist, naturally conveys a rather different sensibility. I have chosen to translate this story not because I agree with the author's attitude, nor because he is more famous than Tamura, but simply because this story is shorter. Yet the problem it raises is essentially the same and is well worth examining, so I have translated it.

Published in 1886, “Reform” portrays an unsuccessful attempt by a Parisian couple, Louisa and Hugo, to establish gender equality within marriage. The couple experiment with separate rooms, independent careers, and a shared division of domestic labor. Their effort nearly succeeds until an unexpected pregnancy forces Louisa to relinquish her independence and become dependent—financially and emotionally—on her husband.

To Louisa's frustration and at the expense of the couple's abortive venture, Strindberg's narrative stance appears detached, even unsympathetic. For much of the story, the omniscient narrator withholds the protagonists’ names, referring to them instead almost exclusively through the Swedish pronouns han (he) and hon (she). This persistent reliance on minimally specified references produces a generalized effect: the characters emerge less as individuals than as types, and their failed experiment appears emblematic of the limits of marriage reform. The narrator further intensifies this effect by frequently shifting perspective between the two protagonists while beginning most sentences with han or hon. The repeated foregrounding of pronominal contrast thus embeds gender difference directly into the narrative structure. As the story concludes with the failure of marital reform, sustained attention to gendered roles throughout the narration appears less a critique of inequality than a reassertion of gender disparity.

Strindberg's apparent misogyny drew criticism from many of his contemporaries, yet it intrigued Zhou Zuoren. ⁷ Writing amid the New Culture Movement, Zhou (1918a) regarded Strindberg's bleak portrayal of marriage reform not as a doctrine but, rather, as an invitation to reflect on the practical difficulties of translating ideals into lived experience. In a biographical sketch accompanying his translation of “Unnatural Selection or the Origin of Race” (“Onaturligt urval eller Rasens uppkomst” 不自然选择), another story from Getting Married, Zhou (1918b) presented Strindberg as one of modern Sweden's most accomplished and erudite intellectuals. ⁸ There, Zhou contends that Strindberg's writings, though seemingly misogynistic, testify to the gap between the ideal of gender equality and the obstacles to realizing it. As discouraging as it may seem, Zhou (1918a) suggests, the story “Reform” reveals the practical challenges faced by advocates of marriage reform. As Zhou observes in the cited preface to “Gaige,” this realism could be intellectually productive even for readers sympathetic to reform, including Japanese feminist writer Toshiko Tamura (1884–1945; pen name Satō Toshi 佐藤 とし).

It was in this spirit that Zhou juxtaposed Strindberg with Tamura, despite their reputations as “misogynist” and “new woman.” In “Her Daily Life” (“Kanojo no seikatsu” 彼女の生活), a novella that caught Zhou's attention soon after its publication in 1917, Tamura expanded on the thematic concerns of “Reform” to portray a similarly pessimistic picture of women losing freedom within marriage. ⁹ Both stories, in Zhou's view, exposed the fragility of women's independence under marital conditions. Around the time “Gaige” appeared, progressive journals such as New Youth (Xin qingnian 新青年) were publishing many impassionate essays advocating women's liberation. Zhou, by contrast, translated “Reform” to invite further discussions, declining to take a categorical position. On questions of gender politics, he remained cautious. ¹⁰ Read against this narrative evidence, the fabrication of 他_女 should therefore be understood less as an ideological gesture than as evidence of a different, emerging line of concern.

This underlying orientation becomes legible in the cited preface itself. In its second paragraph, Zhou turns from marriage reform to language, and his tone becomes conspicuously firm. After the circuitous discussion of Strindberg and Tamura that opens the preface, he abruptly addresses the symbol 他_女 in a forthright and definitive manner. The Chinese third-person pronominal system, he asserts, “has no gender distinction” and is therefore “truly inconvenient.” Here, Zhou's caution gives way to certainty. What demands attention for him is not the ideological debate over gender equality but, rather, a practical linguistic problem: the limitations of written Chinese in representing gendered reference. What kind of “inconvenience,” then, was so pressing that Zhou not only endorsed Liu Bannong's proposal of 她 but also devised 他_女 as a temporary substitute?

The narrative structure of “Reform” provides an immediate answer. As the story relies so heavily on the repeated contrast between Swedish pronouns han and hon, a translation that lacked gender-differentiated pronouns would flatten this essential stylistic feature. By using the symbol 他_女, Zhou was thus able to approximate the story's pronominal patterning and to preserve the story's gendered narrative texture. In this sense, 他_女 first functioned as a translational device: one that translated not only the content but also the style of the story.

Beyond this immediate motivation, however, Zhou's preface suggests a broader ambition. In the second paragraph of the cited preface, he makes a move that appears somewhat out of place: he describes the fabrication of the symbol in detail while saying little about its specific necessity for translating the story. The emphasis falls not on a faithful rendering of Strindberg's plot but rather on the structural inadequacy of Chinese pronouns. In particular, he invites readers to consider alternative orthographic possibilities with a view to the future. He intended 他_女—or a refined alternative—to become part of the writing system and thus improve the functional capacity of the Chinese language. If the task of translating “Reform” provided an impetus for Zhou to design 他_女, the preface makes clear that a broader linguistic concern had already preoccupied him. Here, “inconvenience” names more than a momentary difficulty of translation; it signals a perceived limitation in the expressive resources of written Chinese as it confronted modern vernacular writing. Under what circumstances, then, did the notion of linguistic “inconvenience” emerge as such a pressing concern?

This broad concern must be situated within Zhou's evolving view of translation. Since the early 1900s, Zhou had actively participated in introducing foreign literature to Chinese readers. He and his fellow translators typically employed existing Chinese characters to translate—and sometimes transliterate—foreign words, creating new characters primarily for scientific concepts that lacked corresponding written forms in Chinese. For the purpose of rendering the linguistic function of referring to a female person or animal previously mentioned, most translators found the existing resources of Chinese adequate. Veteran translators of European literatures such as Lin Shu (林纾, 1852–1924), for instance, upheld Classical Chinese—including classical pronouns that are more often implied than marked—as an expressive medium. Hu Shih (1919a, 1919b), by contrast, proposed colloquial phrases like nage nüren 那个女人 (that woman) as workable equivalents for “she/her.” In his earlier translations, Zhou also relied on Classical Chinese. Yet, from the mid-1910s onward, as he became more deeply involved in the “vernacular movement,” he increasingly rejected both approaches. For Zhou, reliance on inherited forms meant foregoing an opportunity to refine the emerging written vernacular. ¹¹ In his vernacular translation, then, Zhou embraced what he later termed “hard translation” (yingyi 硬译; literal translation): a practice that sought not only semantic equivalence but also structural and stylistic correspondence. Translation, in this view, was not merely transmission; rather, it became a process of linguistic construction. ¹² The introduction of a gendered pronoun thus constituted one small yet significant step in expanding the formal possibility of the modern written vernacular.

In 1920, as the character 她 began to replace 他_女 in writing, Liu Bannong explicitly concurred with Zhou's concerns. In his influential essay “The Problem of the Feminine Pronoun ‘Ta’” (“‘Ta’zi wenti” “她”字问题) Liu assesses the potential significance that this gendered pronoun could bear on the modern written vernacular:

I may say that in Chinese writing up to the present, there has been no necessity for the character “她” to exist. Because this character was absent, earlier writers relied on contextual cues before and after to ensure that the intended meaning would not be misunderstood. We therefore need not retroactively alter, one by one, the writings that our predecessors have already produced. As for Chinese writing from now on, however, I would not dare to claim that the character “她” is absolutely useless. At the very least, it ought to occupy a place in translated writing. Let me give a simple example:

With 她, the meaning of the first example presented by Liu—He says to her, “I have waited for her for a long time, yet she has not come; by chance he has arrived instead, and so I came together with him.” “She says, ‘It is good that he has come; but we must still wait for her.’”—is immediately clearer than that of the second one. In the third example, substituting 她 with nage nüren 那个女人 may clarify reference, yet it sacrifices the flow of narration—the succinct narrative economy achieved in the first example.

Beyond referential clarity, Liu Bannong also points to differences in “tonal weight” and “refinement of expression.” Although he does not elaborate on what he appreciates as “the weight of the tone and the refinement of the diction,” a closer look at the first example reveals the stylistic effect that Liu suggests and that Zhou had earlier intuited. In the second sentence from the first example, for instance, the distinct semantic radicals—女 in 她, 亻 in 他—guide readers to identify the respective referents of the two characters with relative ease. Meanwhile, the shared phonetic radical—也 in 她 as well as in 他—alerts attentive readers to the possibility of visual and semantic proximity, momentarily slowing the process of reference resolution, especially at the second occurrence of 她 at the end of the sentence. It is precisely at this momentary hesitation, when readers confirm the referent of the final 她 by comparing it with the earlier 他 and 她, that the structural symmetry of the two characters is not only clarified but also actively reinforced. The result is a line that is compact, visually balanced, and rhythmically controlled. Such is the stylistic effect generated by the visual contrast between 她 and 他.

Liu's remarks thus make visible what had remained implicit in Zhou's orthographic experiment: the gendered pronoun was valued not primarily for ideological reasons, but for the formal and rhythmic possibilities it opened up within written prose. The introduction of 他_女, and later 她, therefore enabled the written vernacular to develop modes of diction previously unavailable to it: modes that parallel, without imitating, the pronominal economy of English prose. These effects, in turn, mark an expansion of what modern written vernacular Chinese could look like on the page.

Seen in this light, Zhou's earlier fabrication of 他_女 appears not as an isolated improvisation but as an initial, more radical exploration of the same formal problem. For Zhou, the “inconvenience” he identified in 1918 signaled a missed opportunity for the Chinese language to realize new formal possibilities. By fabricating the symbol 他_女, Zhou not only ensured referential clarity within the story; more importantly, its repeated use in the translation shaped readerly expectation and showcased new ways of meaning-making within the vernacular medium.

In “Reform,” Strindberg exploited the function of gendered third-person pronouns to heighten the female protagonist's frustration over the failure of marital reform. In the Chinese translation “Gaige,” Zhou amplified this narrative mechanism: his use of 他_女 intensified the emotional and structural contrast already embedded in the plot. For progressive readers of New Youth, the unfamiliar form of 他_女—marked by the character 女—could easily suggest an alignment with women's emancipation. Combined with the preface's discussion of marriage reform, it primed readers to anticipate a narrative of successful modern partnership. Yet, Strindberg's pessimistic plot disrupted that expectation. As the story unfolded, readers might have experienced confusion at the narrator's detachment while simultaneously developing greater sympathy for Louisa, whose frustrated independence echoed the visual signal of 他_女. The repeated frustration of expectation gradually revealed for readers that Zhou's invention of 他_女 did not necessarily advocate a feminist position. Instead, readers searching for another rationale for this peculiar symbol were led away from ideological resolution and toward narrative form. They encountered the limits of reform within the story while simultaneously confronting a linguistic innovation in the medium of narration itself.

This shift from ideological expectation to formal awareness is crucial. Once attention turns to the textual presentation of 他_女 in “Gaige,” an unconventional mode of vernacular writing comes into view. In accordance with the Swedish and English versions of “Reform,” Zhou's Chinese translation consisted almost entirely of single-­sentence paragraphs beginning with 他 or 他_女:

She had noticed with indignation that girls were solely brought up to be housekeepers for their future husbands.

This unusual narrative layout turned the Chinese text into a demonstration of sentence construction: a sequence of model sentences showing how to deploy the new character 他_女 alongside 他 and the plural form 他们. At one level, this arrangement serves a clear pedagogical function: by isolating each clause as a discrete visual unit, the text implicitly instructs readers in how the new pronoun operates syntactically and referentially within vernacular prose. At another level, however, the layout can now be seen to stage the very stylistic effect that Liu Bannong would later single out in his discussion of 她. Each short paragraph forms a compact visual block in which the alternation between 他_女 and 他 produces a graphic rhythm of contrast and symmetry across the page. What Liu described at the level of a single line—the weight of the tone and the refinement of the diction—here expands to the scale of the paragraph and even the page itself.

In this way, Zhou's translation did more than transmit a foreign narrative. It staged the emerging written vernacular as a site of formal experimentation, where new graphic forms reshaped narrative rhythm, perspective, and readerly perception. “Gaige” functioned as both linguistic instruction and an enactment of new prose aesthetics: readers learned how to decode the pronoun while experiencing the visual and rhythmic discipline it introduced into vernacular narrative. Zhou's translation, therefore, constituted a laboratory for the formation of modern prose style.

A few writers quickly followed Zhou's example, experimenting with 他_女 in various contexts of vernacular composition. Ye Shaojun (叶绍钧, 1894–1988; also known as Ye Shengtao 叶圣陶) (1919a, 1919b), for instance, adopted the symbol in several short stories written in 1919. Zhou Zuoren (1919a, 1919b) himself continued to use 他_女 in translations of foreign literature, including Hans Christian Andersen's “The Little Match Girl” and Anton Chekhov's “Excellent People.” Between 1918 and 1919, 他_女 appeared frequently in the pages of New Youth and The Renaissance (Xinchao 新潮). Although originally intended as a temporary substitute for 她, the symbol 他_女 briefly entered circulation as a resource for stylistic experimentation within the emerging written vernacular.

Negotiating the pronunciation

For all the promising possibilities that 他_女 and 她 opened for the written language, a crucial question remained unresolved: how should the new character be pronounced? Zhou Zuoren did not specify a reading for 他_女. Even after some scholars began replacing 他_女 with 她 in 1920, the issue of pronunciation remained unsettled. For Zhou, Ye Shaojun, and other writers, the visual novelty of these new forms appears to have been sufficient justification. In the preface to “Gaige,” Zhou expressed little concern about popularizing 他_女 among Chinese readers. By drawing an analogy with the Japanese language, he suggested that familiarity would naturally follow from widespread use. Nevertheless, the question of pronunciation could not be so easily deferred. How would a graphically distinct but phonetically indeterminate feminine pronoun not confuse students of the modern written vernacular, when its lack of a stable pronunciation risked widening the gap between writing and speech? For experienced language reformers like Zhou Zuoren and Liu Bannong, what prevented them from addressing the question of pronunciation with the same decisiveness that they applied to the orthographic structure?

In 1919, Qian Xuantong (钱玄同, 1887–1939) wrote to Zhou Zuoren about 他_女 and its pronunciation. After deliberating on various proposals, they settled on 伊, a long-standing written form for third-person pronouns. Zhou (Qian and Zhou, 1919: 239) wrote:

I was also reminded that in classical literature there is the character “伊,” which, apart from appearing in names such as Yi Yin and Sun Hongyi, is rarely used today. However, in various dialects, many remnants of its pronunciation still exist. Why not designate this “伊” as the third-person female pronoun? It would require no new invention by the printing house, and the pronunciation would differ from that of “他,” seemingly achieving two goals with one stroke.

Importantly, neither Zhou nor Qian found it necessary to assign a fixed pronunciation to 伊. Instead, they believed that the local pronunciations—“many remnants of its pronunciation” in various dialects—would suffice. This solution, in their view, improved the expressive capacity of the written language without disrupting the writing system or interfering with local speech. While this minimal approach to inconvenience might have been pragmatic, it did not resolve a crucial issue: it did little to advance the consistency between the modern written language and the emerging national spoken standard.

In 1919 and 1920, while editing “Gaige” for the publication of Droplets (Diandi 点滴), Zhou Zuoren (1919c, 1920) replaced all instances of 他_女 with 伊. ¹⁴ Droplets—Zhou's first collection of translations in the modern vernacular—featured 21 short stories by 13 authors from eight countries, along with three influential essays on literary vernacularization. Zhou intended the collection to offer a glimpse into the latest foreign literary works while also serving as experimental examples of modern vernacular Chinese. However, much to Zhou's and Qian's surprise, the use of 伊 did not function as they had hoped.

Before 1920, 伊 was often used as a gender-neutral third-person pronoun in Chinese writing, particularly in traditional vernacular fiction. ¹⁵ Given this prior usage, readers of Droplets would have initially understood 伊 in its conventional sense, making its inclusion seem somewhat unusual: Zhou had used a word closely associated with traditional vernacular writing in a collection meant to showcase modern vernacular style. Lacking a semantic radical, the character 伊 would require careful contextual interpretation for readers to recognize its use as a feminine third-person pronoun. What Qian and Zhou had envisioned as a simple distinction between 伊 and 他 ultimately demanded a more attentive reading process.

The challenge of adapting to 伊 was further complicated for speakers of the Wu and Min dialects. In these regions, the gender-neutral third-person pronoun, pronounced /hhi3/ in Wu and /ĭ/ in Min, is written as 伊. Given that 伊 in these dialects shared both the same orthographic structure and similar pronunciation as the gendered 伊 introduced by Qian and Zhou, the character would have caused confusion for Wu and Min speakers reading Droplets. ¹⁶ Not only would these speakers have found it challenging to distinguish the gendered 伊 from 他, but also when spoken aloud, 伊 sounded very similar to the local dialect pronouns, making the text difficult to parse when switching between national and local speech.

As Qian and Zhou (1919: 237–240) envisioned, 伊 could function as both “a character for the eyes” (yan de wenzi 眼的文字), relying on its visual form to indicate meaning, and “a character for the ears” (er de wenzi 耳的文字), producing aural comprehension through distinguishable pronunciation. However, for speakers of local dialects, the adoption of 伊 exposed a deeper tension between the goals of vernacularization and nationalization. In order to learn the gendered 伊, speakers of regional dialects would have to consciously differentiate between the modern written vernacular and their local dialects. Moreover, by early 1920, the new national language had not yet been fully implemented in schools, let alone spread nationwide. How, then, could the experience of learning 伊 in written form help facilitate the acceptance of the still-unfamiliar national spoken language among speakers of dialects?

In contrast to Zhou Zuoren's and Qian Xuantong's more flexible approach to pronunciation, Liu Bannong's (1920a, 1920b, 1934) response adopts a more standardized stance. In his justification for “她” cited earlier, Liu begins by situating its use within the broader development of written Chinese, specifically comparing it with Classical Chinese. He presents an example: “He says to her, ‘I have waited for her for a long time, yet she has not come; by chance he has arrived instead, and so I came together with him.’” “She says, ‘It is good that he has come; but we must still wait for her’” (Liu, 1920a: 1). For Liu, this example illustrates how the use of a feminine third-person pronoun would improve both the accuracy and expressiveness of modern written Chinese, particularly in the translation of foreign texts. Here, Liu focuses primarily on the written language, heightening that the need for the pronoun arises from the written form's desire for precision and clarity. Only after establishing this claim does the question of the character's place in “spoken Chinese” (Zhongguo kouyu 中国口语) arise.

When Liu addresses the issue of spoken Chinese, however, his reasoning becomes notably tentative. With little more than a comma to mark the shift, he changes from emphasizing the prevalence of direct-quotation sentences in ­western languages to asserting the existence of similar constructions in spoken Chinese. In the aforementioned passage, he writes: “Sentences of this kind are almost everywhere in Western languages; and in spoken Chinese, if one listens attentively, it is not impossible to hear similar expressions” (Kiu, 1920a: 1). At the time, a standardized national spoken language had yet to be established. By “spoken Chinese,” Liu likely refers to the various regional dialects used in daily life. His rhetorical move thus appears intended to preempt objections about the relevance of the feminine third-person pronoun in actual speech. Yet, the phrasing he chooses quietly undercuts this claim. The appeal to attentive listening is immediately qualified by a double negative—“not impossible to hear”—a formulation that technically affirms existence while pragmatically implying rarity. Rather than demonstrating that such constructions are common in speech, Liu's wording suggests that they are exceptional enough to require effort to detect.

This hesitation exposes an important nuance in Liu's argument. On the one hand, Liu draws a parallel between ­western languages and Chinese speech, suggesting that just as 她 enables the written vernacular to translate foreign dialogue more effectively, it would also allow writing to render similar constructions found in Chinese dialects. The problem he foregrounds is thus a problem of how writing represents Chinese speech, not of how Chinese speakers actually speak. On the other hand, the implied rarity of the quoted direct-quotation example in dialects undermines the very premise that such a pronoun is demanded by Chinese speech itself. What emerges, then, is a displacement: the need for 她 lies less in spoken usage than in the ambitions of the modern written vernacular. By raising the question of speech only to bracket it, Liu inadvertently reveals that his primary concern remains the capacities of writing. This logic, however, immediately raises further questions: if dialect speech does not in fact require a gendered third-person pronoun, what happens when such a form is introduced into a national standard still in the making? How might this affect the already fraught relationship between national language and local speech?

If Liu's discussion of “spoken Chinese” ultimately remains subordinate to the ambitions of the written vernacular, this orientation becomes even more evident when he turns directly to the question of pronunciation. It is crucial to distinguish, however, between the two versions of his proposal. In the 1920 editions of the essay, Liu drew an analogy from English phonology, suggesting that 她 might be differentiated through a phonetic pattern resembling the strong and weak stress forms of English—an innovation modeled on spoken usage and without precedent in Chinese. ¹⁷ By the time he revised the essay in the mid-1930s, he had abandoned this comparative experiment. Instead, he grounded the distinction based on the wenbai 文白 (literary and colloquial) readings of the character 他. Liu did not explicitly mark this revision, and later scholarship has often treated the 1934 formulation as if it represented his original position. Yet, Liu's revision is analytically significant. Moving away from an imported phonological analogy toward a regulated distinction internal to Chinese reading conventions, Liu's revised proposal reflects a more pronounced concern with stabilizing the written vernacular and aligning it with a standardized mode of reading. Liu's revised strategy is detailed in the following passage from the 1934 edition of “The Problem of the Feminine Pronoun ‘Ta’”:

The greatest difficulty lies precisely in this: how is this character to be pronounced? Mr Zhou Zuoren avoided using “她” and instead used “伊” for the same reason that “她” and “他” can be distinguished by the eye but not by the ear—a view exactly matching that of Mr Han Bing. In my opinion, although the sounds of “伊” and “他” are clearly distinguishable, there are still several respects in which “伊” is inferior to “她.” First, in the spoken language, the use of “伊” as a third-person pronoun is limited to a very small region and is therefore difficult to make general. Second, in its graphic form, “伊” does not make the indication of the female gender as explicit as “她.” Third, “伊” leans rather toward Classical Chinese, and when used in the modern written vernacular it is not especially well harmonized. I believe it would be best simply to use the character “她,” while making a slight adjustment on the level of pronunciation.

While acknowledging that 伊 and 他 are phonetically distinguishable, Liu rejects 伊 on three grounds: its circulation is geographically restricted, its graphic form fails to mark gender with sufficient clarity, and its stylistic associations tie it too closely to Classical Chinese. Here, Liu is rather concerned that 伊 cannot adequately fulfill the expressive demands of the modern written vernacular. Instead, Liu proposes using 她 and assigning it the reading of t’uo, a pronunciation already available within Mandarin reading practice for 他. This solution is graphic as well as phonetic. Crucially, it draws on resources internal to the writing system itself: it is the established reading tradition of a character—not colloquial usage—that supplies the basis for standardization. For Liu, therefore, t’uo 她 would preserve phonological familiarity while securing auditory distinction.

While Zhou Zuoren and Qian Xuantong allowed for flexible pronunciations based on dialectal variations, Liu emphasized the necessity of a fixed pronunciation. By embracing standardization in both pronunciation and orthography, Liu positions the pronoun within the broader framework of national language reform. His approach, though still focused on refining the written vernacular, extends beyond orthographic innovation by addressing the gap between the written language and the emerging national spoken standard. This solution, however, raises new difficulties. For learners educated in Mandarin reading traditions, how could the literary pronunciation of 她 avoid constant confusion with 他, whose colloquial reading remained dominant in speech? For readers from non-Mandarin regions, the challenge was greater still: they were asked not only to acquire an unfamiliar written form but also to internalize a pronunciation tied to a national standard that many had yet to encounter in daily life. The pedagogical burden of reform thus fell unevenly across linguistic communities, exposing the gap between aspirations and readerly experience.

Liu's proposal, even in its revised form, did not resolve the difficulty it sought to address. Whether modeled on English stress patterns or grounded in Mandarin reading conventions, his attempts at phonetic differentiation left open the question of how a graphically distinct pronoun could operate coherently across writing, standardized reading, and heterogeneous speech communities. The controversy, therefore, was not confined to a particular pronunciation; it concerned the viability of integrating orthographic innovation into the emerging national language.

By 1923, the problem reappeared in a new register. A configuration took shape that seemed to reconcile Zhou’s and Qian's emphasis on graphic differentiation with Liu's insistence on phonetic distinction: the feminine pronoun would be written as 她 and pronounced yi. This solution found its most concrete expression in pedagogy. In the children's song “He. She. It.,” Li Jinhui translated earlier debates into instructional practice, staging an experiment in how script and sound might be coordinated within the classroom.

Teaching the new pronoun

In 1923, seeking to help students master the new pronoun yi 她, and especially to distinguish it from another recently introduced pronoun tuo 牠 and the established gender-neutral ta 他, Li Jinhui published the song “He. She. It.” The piece became a crucial site where orthography, pronunciation, and pedagogy converged. Its lyrics stage what Li and many reformers imagined as an ideal configuration of a gendered pronoun system for modern Chinese:

In Zhang Village there's an old gentleman,

His name is Ma Da Ma.

He (ta 他)—Ma Da Ma,

Ma Da Ma,

Ma Da Ma—

His name is Ma Da Ma,

He.

A defining feature of this new pronoun system is aural differentiation. Li Jinhui foregrounds pronunciation through deliberate rhyme. In each verse, the referent's name rhymes with its corresponding pronoun: “Ma Da Ma” 马大麻 with ta 他, sharing the vowel [a]; “Xiao Mo Li” 小茉莉 with yi 她, sharing [i]; “Bing Guo Guo” 冰哥哥 with tuo 牠, sharing [o]. Narrative and syntactically important words are also selected to reinforce these rhyme groups. In the second verse, words introducing the referent's household (the Li family; Li jia 李家;) and her familiar designation (Thirteenth Aunt; Shisanyi 十三姨) echo the [i] sound of yi; in the third, the copular verb changes from mingjiao 名叫 to jiaozuo 叫做 to strengthen the [o] rhyme with tuo. Singing this song thus becomes a phonetic exercise: students repeatedly articulate, hear, and memorize the contrast among the three sounds, and eventually become intuitive with the association of the sounds and their respective meaning. Li's pedagogical aim is clear: to naturalize the auditory distinction between ta 他, yi 她, and tuo 牠, much as English speakers differentiate [hiː] (he), [ʃiː] (she), and [ɪt] (it). Through melody and repetition, pronunciation is meant to become automatic rather than analytical.

Meanwhile, the song trains the eye. For readers of the modern written vernacular, the orthographic shape of 她, especially its visual distinction from 他 and 牠, reorganizes the semantic field of the pronoun system. Recognizing the semantic radicals—亻 (human), 女 (female), and 牜 (animal)—becomes a primary learning objective. Straightforward as it might appear, Li does not assume this process is intuitive. He portrays each figure in the lyrics with characteristics that can be readily associated with gender or animacy. “Da Ma,” a commonly used masculine name, suggests to young learners what is replaced by ta 他 is a male person. As women were often referred to via nicknames of gentle image, “Xiao Mo Li,” delicate like jasmine, aligns with yi 她; and the old camel aligns with tuo 牠.

Despite this careful design, Li remains both hopeful and cautious. He does not take for granted the ease of pedagogical reform. Anticipating objections, he was wary of probable disagreements on a series of questions. As efficient as the system appears, how could yi 她 contribute to the national language that was itself still unstable in the early 1920s? Are the visual and aural distinctions of the three pronouns truly meant to produce identical referential precision in both speech and writing? More fundamentally, must modern Chinese develop a feminine third-person pronoun in order to achieve such precision? Li addresses these concerns in a note accompanying the song:

The Chinese character “他” should be pronounced ta; it refers to a man.

The example Li provides reveals the nuance of his own proposal. In spontaneous speech, such sentences would not be ambiguous with gesture or contextual clarification; in traditional literary prose, they would normally be resolved through fuller lexical specification rather than pronominal contrast. In 1923, Li's vision, therefore, projects a future in which speakers would rely on pronouns alone for third-person reference. In such a linguistic environment, differentiating ta 他, yi 她, and tuo 牠 would produce immediate semantic clarity. The national spoken language and the modern written vernacular would thus converge in communicative efficiency. The new pronoun system, in his view, enables sentence patterns previously unavailable in either literary Chinese or regional dialects. By encouraging students to imagine and adopt unfamiliar forms of expression, Li casts the modern national language as a medium capable of new syntactic and rhetorical possibilities.

The promise, however, proved fragile. By 1924, works employing yi 她 and the full three-pronoun system generated more confusion than consensus. Even Li himself gradually retreated from emphasizing yi 她. In his (1924, 1925) later songs promoting the national language, such as “Long Live the National Language of China” (“Zhonghua guoyu wansui” 中华国语万岁)—a school anthem for the Specialized School of the National Language (Guoyu zhuanxiu xuexiao 国语专修学校)—and “A Theme Song for the First National Conference on the National Language Movement” (“Quanguo guoyu yundong dahui huige” 全国国语运动大会会歌), Li reverted to using ta 他 as a gender-unspecified pronoun. As noted earlier, it would take several more decades before advocates of the character 她 again asserted its value with confidence. Only in the late-1940s did the modern settlement emerge: 她 written with the 女 radical but pronounced ta.

Conclusion

Since the 1950s, the status of ta 她 in linguistic scholarship has itself become revealing. When analyzing the system of personal pronouns, linguists (see, for instance, Lü and Zhu, 2022[1951]: 9) consistently distinguish between written representation and spoken form. Within this framework, ta 她 is treated not as an independent morpheme but as a graphic variant of ta 他, used in writing to mark feminine reference. As Lü Shuxiang (2008[1980]: 42) observes, distinctions among ta 他, ta 她, and ta 它, like those between de 的 and de 地, are “purely matters of orthography”; in the language itself, they count as a single word. ¹⁹ Zhu Dexi (2023[1982]: 81–82) states the point even more explicitly:

Chinese personal pronouns do not distinguish gender. When 他 refers to a woman, it is written as 她; when it refers to a thing, it is written as 它. These are merely distinctions in writing—in speech, all are pronounced ta.

And in his influential studies on modern Chinese grammar, Wang Li (1984a [1944 & 1945], vol. 1: 268–269, 476–488; 1984b [1943 & 1944], vol. 2: 278–281, 505–507) similarly identifies the emergence of ta 她 as a feature of Europeanized written style rather than of spoken vernacular.

Seen from this perspective, the present status of ta 她 curiously resembles what Zhou Zuoren in 1918 considered only a provisional stage: a graphic innovation that could enrich written expression yet had not entered speech. For Zhou, however, such a division was a temporary expedient, not an endpoint. Does the subsequent history—from 他_女 to 伊, from 她read as tuo or yi to 她 pronounced ta—suggest that a pronoun system distinct both visually and aurally was an elegant but ultimately impracticable ideal? Or does the history of 她 instead reveal a deeper, unresolved tension between writing and speech that have shaped modern Chinese from the outset of language reform?

The invention of the feminine third-person pronoun was only one among many innovations that gave form to the modern Chinese language. Around the time Zhou experimented with 他_女, reformers were also introducing ­western-style punctuation and horizontal writing into the modern written vernacular. ²⁰ Read in this broader context, the history of 她 is not merely the story of a single character, but a window onto the material and sensory reconfiguration of language itself. It reveals how efforts to standardize, nationalize, and modernize Chinese unfolded in the visible and audible textures of everyday reading and writing: an unfinished process whose tensions remain embedded in the language today.