A moment from the well-known confessional memoir of China’s “Last Emperor,” titled Wo de qianbansheng 我的前半生 in Chinese and From Emperor to Citizen in English translation, is often cited in popular discourses (and occasionally, academic writings) surrounding the state of the Manchu language near the end of the Qing empire. The intrigue which Puyi’s (1906–1967) submission that he knew no more than one expression in Manchu — ili “rise!”—holds for China’s cultural memory endures, and the analysis of this endurance—which perhaps speaks as much to the specters that haunt China’s present as it does to its past per se—is yet to be undertaken.² This research note, in the meantime, takes up the comparatively simpler question lurking behind the culturally-cathexized symbol of a linguistically sinicized emperor: How much Manchu did he actually know, and how did he learn it?

In the following, I will introduce three sets of manuscripts upon which my subsequent discussion centers. By closely examining the running dates in the top margin of these manuscripts, I demonstrate that these manuscripts were used by one of Puyi’s three fellow-students—Aisin Gioro Ioi Cung (1903–1965)—between Xuangtong 5 and 11 (1913–1919), during their Manchu lessons in the Forbidden City after the emperor’s abdication. After the analysis of the manuscripts themselves and by way of conclusion, I will contextualize the palace Manchu curriculum as glimpsed through these manuscripts and other transmitted sources within late-Qing Manchu pedagogical practices at large, across different strata of bannermen society.

Three manuscripts within the Harvard Yenching Library’s Manchu-Mongolian collection contain the name “Ioi Cung 毓崇” on their front covers. They are:

Holboho manju gisun | Lian qing yu 聯清語. 2 volumes. 22 cm. TMA 5806.08 1300. A Chinese-Manchu lexicon, with the Chinese word given first and followed by the Manchu below. Organized roughly by category, but without explicit category titles.

All of this demonstrates that these manuscripts belonged to the same person. The color of the cover and the quality of the paper, furthermore, suggest that these manuscripts were likely produced within the Forbidden City itself. And since no one by the name of Ioi Cung 毓崇 can be found within the Renming quanwei: renwu zhuanji ziliaoku 人名權威：人物傳記資料庫 database of the Academia Sinica, it would seem reasonable to suspect that these manuscripts once belonged to Aisin Gioro Ioi Cung (1903–1965), the son of Aisin Gioro Pulun (1874–1927) who once represented the Qing at the St. Louis World’s Fair. This Ioi Cung, still young upon the fall of the Qing (hence his absence in the official biographies) nevertheless lived a notable life, being best known as having studied alongside Puyi for some time during the strange period between the emperor’s abdication in 1912 and his expulsion from the Forbidden City in 1924.

As the former emperor recounts in his memoir, “When I was five, the empress dowager Lung Yu chose a tutor for me and ordered an astrologer to select an auspicious day for me to begin my studies. This day was September 10, 1911.”³ The lessons, which eventually moved to Yuqing Hall 毓慶宮 in the Forbidden City, fell into three categories: Sinitic, Manchu, and (from when he was fourteen sui) English. And he recalls the following regarding his “fellow-students”: Figure 1This is an image of the yellow cover of a handwritten book. The title, in Manchu and Chinese, is written on the title slip in the upper right corner of the cover, while the name of the owner is written on the bottom of the cover.

When I was eight [nine sui] they thought of another way of making me study better: I was provided with some fellow-students. Each of them received a stipend of the equivalent of eighty taels of silver a month and was granted the coveted privilege of being allowed to ride a horse in the Forbidden City. Although it was the time of the Republic this was still regarded as a great honour by the young men of the imperial clan. The three recipients of these favours were my brother Pu Chieh, Yu Chung (a son of my cousin Pu Lun), and Pu Chia, the son of my uncle Tsai Tao. Yet another honour conferred on these three was that of being scolded on behalf of their emperor in the schoolroom: when I made a mistake in reading out my lessons the teacher would tell one of my fellow-students off. As Pu Chieh was my brother the victim was nearly always Yu Chung, whose studies naturally suffered when he found himself being scolded whether he read well or badly.⁴

Before turning to the manuscripts themselves, I will briefly survey what is known about Puyi’s Manchu education.

Puyi’s Manchu teacher from 1911 onwards was Iktan of the Gūwalgiya hala (1861–1923), a translation jinshi of 1886 from the Plain White Banner whom Reginald Johnston (1874–1938)—Puyi’s English tutor—remembered in the following terms:

I-K’o-T’an, who as I have said was one of the first three to be appointed tutors to the emperor Hsüan-T’ung, was a Manchu. The duty assigned to him was to teach the Manchu language, or rather to see that the emperor was not brought up in entire ignorance of what had been his ancestors’ native tongue. If there was nothing very striking about his personality, he had a gay and genial temperament and was popular both in the palace and elsewhere. Whether his knowledge of his ancestral tongue was profound or not I cannot say, though I think he was a more fluent speaker of Pekingese than he was of Manchu.⁶

These heuristic accounts of Puyi’s Manchu ability given in memoirs, then, can be imprecise and, as already suggested by Erling von Mende, quite possibly distorted.¹¹ Fortunately, it is possible to reconstruct Puyi’s Manchu curriculum to some degree, both by looking at his memoir more closely and by consulting some more recently published materials from the palace museum itself. Jenner’s English translation of Puyi’s memoir omits the list of titles he gives for the texts that he had studied during those years, which include the following:

The Thirteen Classics (shisan jing 十三經) which he describes as the main text, with everything else being auxiliary texts.

Read against other sources, however, it seems likely that Puyi reported some texts read in Manchu using their Sinitic titles and further omitted other materials used in his Manchu (and even Mongolian) education. In Puyi’s selected diary entries published under the editorship of Wang Qingxiang 王慶祥 in 1999, one finds the additional mention of the following texts among what he studied and reviewed during this period:

Tangshi 唐詩

A third, source for Puyi’s curriculum is a “Chongxin jiaoding chuangke zhangcheng” 重新校訂窓課章程 in the Forbidden Palace, quoted in a 1982 article by Zheng Li 鄭里 and discussed by Martin Gimm and Murata Yūjirō.¹⁶ It is unfortunately unclear when this “revised” schedule was established, or what other information may be included in it beyond the short excerpt given by Zheng Li. Nevertheless, the “Manchu-Mongolian” section of Puyi’s daily curriculum, according to this document, includes the following.

All of which was to be completed in around forty-five minutes or three ke 刻, in contrast to the over twenty ke (or around seven hours) Puyi subsequently spent studying Sinitic texts. One can only imagine the time or attention dedicated to the study of each text individually. Nevertheless, it would seem that Puyi had studied a larger range of Manchu texts than what is recorded in his memoir and diary, and that the learning materials he did employ are surprisingly standard, being a mixture of classical texts, lexicons, and dialogue books. The latter were presumably of a more vernacular register, as they often are in order to teach what was supposedly the spoken language.

The remainder of the research note adds to these scattered accounts of Puyi’s Manchu education by turning to the Harvard-Yenching Library manuscripts themselves. Through a close examination of their marginalia, it can be determined that they were annotated during the interval of Xuantong 5 to 11 (1913–1919) and that Ioi Cung’s classroom schedule followed closely what we know of Puyi’s schedule in the Yuqing Hall. It follows that Ioi Cung likely did study Manchu (as well as Sinitic) alongside Puyi and, conversely, that Puyi’s Manchu lessons would have been highly similar if not identical to those of Ioi Cung’s, which we can reconstruct to some degree from the manuscripts. Both throughout my discussions and in particular in the concluding remarks, I show that the Yuqing Hall Manchu curriculum shared significant continuity with late Qing Manchu pedagogical practice in general, as seen from other student materials both in the Harvard-Yenching collection and elsewhere.

All three items are extensively annotated in red ink. Taking the Enduringge tacihiyan be neileme badarambuha bithe | Sheng yu guang xun 聖諭廣訓 (TMA 1686 3245) as an example, we may observe that the annotations are of two kinds. There are red dots at the end of each Sinitic phrase, and there are dates written in red in the top margin, also in Sinitic. Significantly, both types of annotations run through the entire six volumes of the manuscript.

Both the position of the name (“Ioi Cung 毓崇”) on the front cover and the use of running dates in the top margin demonstrates the similarity between this manuscript and other print and manuscript Manchu materials used in pedagogical settings in the late Qing. Manju gisun justan-i bithe (Staatsbibliothek zu Berlin, Hs. or. 8454), with “Wengjy ningge” written on the cover, would be an example of such a manuscript produced in a less prestigious context.

The practice of the running dates, observed in these manuscripts, is associated with language pedagogy in several ways. First, in some textual objects on which the running date appears, there is explicit mention of the school setting, either in the text itself or in the marginalia. A printed edition of Ninggun baita-i targabun gisun | Liushi zhenyan 六事箴言 in the University of Chicago East Asian Collection (M1681/4917, preface dated 1815), for example, contains not only the running dates in red ink but also occasional remarks by the student about missing school (e.g., on the top margin of 1:33r, “十九日未上學” or “on the nineteenth day I did not go to school”). Another manuscript titled Justan-i bithe belonging to a Cunsiyang in the Staatsbibliothek zu Berlin (Hs. or. 10791, undated), also bearing running dates, includes dialogues about Cunsiyang’s own family background and school routines, and it is clear that this manuscript was produced in the very school mentioned in the text itself.

Second, we know that the memorization of short segments of Manchu texts was a common and important strategy in day-to-day Manchu pedagogy, as it has been discussed in contemporary print and manuscript sources dealing with Manchu literacy. For example, the nineteenth meyen of the second fiyelen of the An-i gisun de amtan be sara bithe | Yongyan zhizhi 庸言知旨 (preface 1819) contains the following utterance of a teacher to a pupil.

The verb for assigning a segment of text here is pilembi, and in context, it likely means to mark up a text in the process of studying it. The daily memorization regime suggested here, in fact, dovetails with the popular genre of meyen or huatiao 話條 (literally “a string of utterance”) in Manchu pedagogy, which are rather short dialogues or monologues, often but not always in a vernacular register. The quote above is itself an example of the meyen genre. The brevity of these segments meant that they could be studied in a single lesson, which also aligns with what the running dates on these texts, when they are present, tend to suggest. In the case of longer texts that, unlike the meyen genre, were created outside this pedagogical context, they will simply be divided into shorter segments to fit back into the daily memorization routine. While the typical assignment for a day, which is anywhere between two lines to just over a half-leaf, may seem short, it might be worth keeping in mind that in a school setting more than one text was studied every day (as suggested, e.g., by Puyi’s diary and schedule), and thus a meyen text might be studied alongside Sinitic classics or a Manchu-Sinitic lexicon, with each text having a similar daily amount of memorization.

If we now turn to the pattern of the dates in Ioi Cung’s Enduringge tacihiyan be neileme badarambuha bithe manuscript, two features are immediately worthy of notice: First, the dates are almost entirely consecutive—that is, they follow each other nearly without break, which is not usual in other pedagogical texts with running dates. Second, as mentioned before, the dates appear throughout the entire six volumes of the work, which is again unusual, especially for a multi-volume work.¹⁹ But the consistency of the annotation allows us to ascertain the duration of time necessary for the student to study each volume of the text, and to study the text as a whole. It is regrettable that, with this type of marginalia, typically only month and day are given (as more information is generally unnecessary in context), thus if the manuscript itself does not have a colophon it is difficult to establish when, exactly, it was used in a classroom. However, as we will see, the idiosyncrasies of the dates given in this and the two other manuscripts allow us to determine the exact year in which they were annotated.

The beginning and end date of each volume of the Enduringge tacihiyan be neileme badarambuha bithe manuscript are as follows. I give the dates as written on the manuscript (8.23 means “the twenty-third day of the eighth month,” and so on), without converting to the Gregorian calendar yet.

Note that at the beginning of the third volume, the year—and not just the month and the day—is specified, although not the reign name. A total of one year and four months, then, elapsed during the study of this six-volume text from cover to cover, with each volume taking anywhere between two and a half to four months. The text covered each day stays within the range of two to four lines.

I give also the dates in Holboho manju gisun (TMA 5806.08 1300):

Volume one: from 7.14 to 2.6 of the sixth year, with seven unmarked half-leaves at the end.

Finally, here are the dates in Booi tacihiyan-i ten-i gisun (TMA 1686 3214):

Volume one: from 2.24 of the eighth year to 4.21.

Together, we can ascertain that the three texts were studied in the following sequence. First, Holboho manju gisun, a basic Manchu-Sinitic word-list, was studied from 7.14 of the fifth year (of the Xuantong reign, as we shall see, hence this date is August 15, 1913 in the Gregorian calendar) to 8.1 of the sixth year (September 20, 1914).²¹ A few weeks later began the study of Enduringge tacihiyan be neileme badarambuha bithe, which started on 8.23 of the sixth year (October 12, 1914) and concluded on 2.23 of the eighth year (March 26, 1916). The study of Booi tacihiyan-i ten-i gisun immediately followed, beginning on 2.24 of the eighth year (March 27, 1916) and concluding over three years later, on 5.15 of the eleventh year (June 12, 1919).

Within this context, it is possible to determine the exact era name to which the years belong through the appearance of an intercalary fifth month in the sixth year in the second volume of the Holboho manju gisun (2:21v). Of all the Qing era names, only the Xuantong era (extending into the Republican era) had a month intercalated between the fifth and the sixth months in its sixth year. Furthermore, there was an intercalary second month in the ninth year of this era (during which the third and fourth volumes of Booi tacihiyan-i ten-i gisun were being studied). Throughout the Qing this had happened only twice: during the Kangxi (r. 1661–1722) and the Xuantong eras. The former, however, is very unlikely, not least because the Yongzheng Emperor (1678–1735), who was to compile the Booi tacihiyan-i ten-i gisun, was not yet born in the year Kangxi 9 (1670).

The appearance of intercalary months thus demonstrates conclusively that these manuscripts were annotated during the Xuantong period, and more specifically between 1913 and 1919. Furthermore, the dates in these manuscripts also display consistent alignment with what we know of Puyi’s schedule and school breaks during these years, which Wang Qingxiang, the editor of Puyi’s diaries, listed in a series of footnotes to Puyi’s diary entries during this period. In the interest of length, I will only give the school breaks listed within the published diary entries under the year Xuantong 7, during which Ioi Cung and Puyi had been studying the Enduringge tacihiyan be neileme badarambuha bithe. Both due to the nature of diary entries, which often mention texts he reviewed rather than newly learned on specific days, and the selective nature of the publication, these breaks are not exhaustive even for the year Xuantong 7.

Unsurprisingly, all of these breaks in Puyi’s class schedule are consistently absent also in the running dates in Ioi Cung’s copy of the Enduringge tacihiyan be neileme badarambuha bithe. For example, on 5.28 of Xuantong 7, Ioi Cung had finished the line [. . .] fafun erun de ele šar seme gūnin be baitalahabi 而於刑罰尤惓惓致意 on 3:26v, and it was not until 6.3 that he continued onto the next line, bi soorin de tehe ci ebsi 朕臨御以來 [. . .]. Because the break in the schedule is so specific—it was to celebrate the birthday of an imperial noble consort—it is difficult to explain its presence in Ioi Cung’s study unless he had been studying the text alongside Puyi.

Having now established the ownership and context of these manuscripts, I will conclude by way of some remarks on the kind of Manchu pedagogical practice they suggest. Liu Xiaomeng’s 劉小萌 study of private schooling in the family of Wanggiyan Lincing (1791–1846) divides Qing private bannermen schools into three categories: school in the Forbidden City, schools in princely mansions, and schools for the ordinary bannermen.³³ And while considerable scholarship in recent years have demonstrated that the degree of Manchu language decline in the late Qing has been greatly exaggerated in received historiography, when we turn to the extant textual artefacts used in Manchu language instruction both in the Forbidden City and in ordinary banner schools, we find that the endurance of Manchu as an administrative and literary language even in the final days of the empire was supported, at least in Beijing, by striking similarities in the format and technique of Manchu language education across social strata.³⁴

These formats and techniques of language instruction constitute what Fitzgerald observes as the “surprisingly consistent institutional practices and experiences amongst Manchu students” in the late Qing across different schools and also across time.³⁵ Presumably, Puyi and Ioi Cung, like many bannered children before them, began their study of Manchu by reciting the “Twelve Heads” and had already finished a text titled Dan qingyu 單清語, which Puyi reviewed alongside the “Twelve Heads” in 1916, before turning to the three texts discussed in this research note.³⁶ The types of annotations Ioi Cung left on these manuscripts as he progressed through the texts are so typical, in fact, that, a few material idiosyncrasies notwithstanding, they could easily have come from a humbler Manchu school such as the one attended by Wengjy, the owner of the Staatsbibliothek zu Berlin manuscript mentioned earlier. As Fitzgerald puts it, in such private schools, “primers were copied, songs were sung, dialogues were memorized, homework was corrected, and notes were taken on any available surface.”³⁷ The three Ioi Cung manuscripts demonstrate certain of these practices, although others—such as corrected homework—would perhaps be more fitting for a different part of his (and Puyi’s) curriculum. In any case, it seems that Ioi Cung and Puyi’s Manchu education involved a similarly diverse range of materials, with the particularly notable inclusion of the meyen genre, unfortunately absent in these manuscripts, which had circulated widely and often anonymously in the late Qing, and the linguistic coarseness of which their compilers had sometimes felt the need to justify.³⁸