Introduction
The motif of a craftsman who falls in love with a non-human woman has long been common in literature, and this trend continues to this day in written fiction and visual arts. One of the most famous literary depictions of this motif is the Pygmalion myth, familiar from Ovid's Metamorphoses. This urtext has yielded numerous re-thematisations throughout the Western European tradition, such as Pygmalion by George Bernard Shaw, Le Chef-d'oeuvre inconnu by Honoré de Balzac, The Birthmark by Nathaniel Hawthorne, and Der Sandmann (hereafter, The Sandman) by E. T. A. Hoffmann, to name just a few.
Less familiar to scholars outside of the Western context is a tale that exhibits remarkable similarities to the Pygmalion myth. To my knowledge, this tale, which I will refer to as The Painter and the Mechanical Maiden, has remained unacknowledged in Western literary scholarship to date. Although examined to various degrees by philologists of the languages in which the tale is attested―most prominently by Tocharian philologists George S. Lane, Emil Sieg, Werner Winter, Hiän Lin Dschi and Georges-Jean Pinault―the story has not received attention from literary scholarship. This tale, known in English as The Painter and the Mechanical Maiden, circulated in the ancient Silk Road area and is found in early Buddhist texts from roughly the fourth to the ninth centuries C.E. According to Hiän Lin Dschi (323), the tale was common in Buddhist societies, which is confirmed by an array of preserved versions in four languages employed in Buddhist cultural contexts. There are two original Sanskrit versions of the tale (translated into Western languages by Hofinger 185–187 after Dutt Gilgit 3: 166–168, also in Pinault Chrestomathie 252–253; Degener 47–48), one Tibetan version (Schiefner 17–18, also in Davids et al. 361–362), one TocharianFootnote 1 version (Malzahn A5-A9, Lane 33–53, Sieg 8–13, Pinault Chrestomathie 254–267), and two different Chinese versions (Chavannes 2: 12–13 for T 207; Dschi 323–324 for T 1448).
The first part of the article presents the philological facts about the tale and sets forth the relationships between the different versions. I examine how specific versions have been adapted to suit their particular cultural contexts. The Tocharian version, the longest and literarily the most attractive, is analysed along with the older Chinese version, Taishō 207 (T 207 in the following), which I argue to be related to the Tocharian version. On the basis of the Tocharian tale's formal characteristics, I discuss related genres in the Silk Road area, chu-kung-tiao [諸宮調] and pien-wen [變文]. This analysis sheds light on the role of Central Asian literatures in the circulation of originally Indian tales to East Asia. Finally, the content of the tale is analysed in a comparative perspective by tracing the resonance of the main motifs with two literary works external to the Silk Road cultural context: Ovid's Pygmalion and E. T. A. Hoffmann's The Sandman. Footnote 2
The story
A brief outline of the story of The Painter and the Mechanical Maiden, common to all versions, goes as follows (based on Tocharian and Chinese original versions and English, French and German translations in Dutt Gilgit 166–168, Schiefner 17–18, Lane 33–53, Malzahn A5-A9, Pinault Chrestomathie 254–267, Sieg 8–13, Chavannes 2: 12–13, Davids et al. 361–362, Dschi 323–324, Hofinger 185–187, Degener 47–48):
A foreigner, a painter by profession, comes to a distant land where he is hosted by a mechanic. In order to serve his guest at his best, the mechanic puts a wooden mechanical maiden on the painter's bed. The painter falls madly in love with her and tries to communicate with her, but she does not respond. He knows he should not touch her, because she belongs to his wonderful host. However, he cannot help himself and reaches for her, and as soon as he holds her hand, she falls to pieces/he realises she is wooden. The painter is ashamed when he realises that he was tricked by the mechanic and he decides to trick his host in turn. He paints himself on the door/the wall as if he has been hanged and hides himself. The next day, the mechanic sees that his guest has killed himself because of his prank and starts crying over such a tragic denouement. (The royal servants come/are called to confirm the suicide and they all cry, not being able to see that the painter in the painting is not flesh and blood.) Just as mechanic tries to cut rope with an axe/knife, the painter comes out of his hiding place and victoriously announces his trick.
The genre and the frame
The tale The Painter and the Mechanical Maiden presents itself in six preserved versions and four different cultural contexts: the original Sanskrit, Tibetan, Tocharian, and Chinese. All versions are incorporated into Buddhist scriptural compendia. In what follows, I discuss the significant distinctions among the different versions with respect to their cultural and historical context and adaptations. I begin by briefly surveying previous scholarship on the tale and evaluating the merit of arguments for its origins in folklore.
In Buddhism, jātakas are stories of Buddha's former births that “illustrate the long path to buddhahood and the acquisition by the bodhisatt(v)a (buddha-to-be) of the perfections required for that attainment” (Appleton ‘Jātaka’). Nonetheless, many jātaka stories “began life outside the genre of jātaka” as fables (ib.), an aspect which also interested scholars of this genre (Thomas William Rhys-Davids, Joseph Jacobs, Merlin Peris) and literary scholars (Ruth Cline, Octavio Paz).Footnote 3
Regarding ideological features of jātaka tales, The Painter and the Mechanical Maiden is reminiscent of this genre given the way the tale displays a strong moral message (Zhu 68). Jātakas narrate about “the acquisition of the many qualities required for Buddhahood” (Appleton Jātaka 21), which is, in this tale, the virtue of wisdom. “An essential feature of jātakas [is] the presence of the Bodhisatta” (ib., also 1–2), and in this particular tale Buddha is not the protagonist but the narrator (5).Footnote 4 Naomi Appleton suggests that in jātakas, “it is the Bodhisatta's wisdom and insight, rather than his action, that matter, and the audience can be instructed by the actions of other protagonists” (Jātaka 24). This view fits the type of tales, such as The Painter and the Mechanical Maiden, where the Buddha as the narrator focuses on the past lives of other people, which he testified as the Boddhisatta. From the formal perspective, The Painter and the Mechanical Maiden resembles jātakas in beginning and ending with a brief story set in the present and focusing on the central story from the past (Dash 41, Appleton Jātaka 6, Zhu 66). The third part of this composition offers a moral in “a juncture in which the narrator, always the Buddha or another enlightened saint, identifies characters in the past as former births of characters in the present” (Tatelman 36). This structure is typical for avadānasFootnote 5 and its subgenre of jātakas.
Several aspects of the tale suggest that The Painter and the Mechanical Maiden was initially non-Buddhist and was only later framed as a jātaka tale about the Buddha's disciples, Śāriputra and Maudgalyāyana. First, the tale's anecdotal character—namely, its concise form in the original version—makes it easy to remember and transmit the story as well as elaborate upon it. Moreover, the characters are anecdotally plain, which would have facilitated their recasting as Buddha's disciples. While most versions have been adapted to convey a Buddhist message, as I show in the discussion below, the message could also be read in a non-Buddhist context. The story does not lose its essence if read without the Buddhist implications; in other words, the story is not inherently Buddhist but it also does not contradict Buddhist ideals.
In addition to these specific arguments, some general observations also point to the conclusion that this tale was originally non-Buddhist, as in Chavannes (1: xvii).Footnote 6 Other collections of Buddhist folk tales, some of which are also jātakas,Footnote 7 also consider these tales as folk tales with a secondary sacralisation. It is probable that The Painter and the Mechanical Maiden is a part of this tradition. In any case, the specific arguments pertaining to this tale's form and content suggest that it can be read as a Buddhist as well as a folkloristic tale.
The tale itself occurs in three different corpora of texts:
[a] the original frame, found in the two Sanskrit corpora, as well as the Tibetan Kanjur and Taishō Tripiṭaka (the newer Chinese version, T 1448);
[b] the frame where individual tales are isolated and decontextualised, as in the corpus of Taishō Tripiṭaka (the older Chinese version, T 207);
[c] the unique Tocharian frame in Puṇyavanta Jātaka (possibly related to the T 207).
The Sanskrit sources are found in two collections, Mūlasarvāstivāda Vinaya (a massive collection of texts and genres, which largely overlaps with other collections of avadāna and was likely compiled in the first or second century C.E.),Footnote 8 and Kaṭhināvadāna (a smaller collection of a later date from the avadāna genre).Footnote 9 In both collections, the tale is included in the introductory chapter to the ‘Bhaiṣajyavastu’ or ‘Treatise about remedies’ section. The context in which the story takes place is an assembly by the mystical lake Anavatapta, where Buddha recalls the past lives of his two main disciples, Śāriputra and Maudgalyāyana. Śāriputra is known for his wisdom and Maudgalyāyana is famous for his supernatural powers (Dutt Early 29). In Buddha's story, the two disciples, presented as artisans, test each other's abilities. The tale The Painter and the Mechanical Maiden constitutes the first example of the disciples’ trial. In all of the exemplary tales, Śāriputra's wisdom overcomes Maudgalyāyana's magic, thus accentuating the important religious point that wisdom is superior to supernatural powers (Hofinger 16).
As Buddhism spread beyond the Indian subcontinent, the Sanskrit version of The Painter and the Mechanical Maiden was further introduced to Tibetans, Chinese and Tocharians. According to Tibetan historians, it began to be translated to Tibetan from the Indian sacred texts in the early seventh century C.E. upon the order of the emperor Songtsen Gampo [Tib. Srong btsan sgam po] (Davids et al. ix). The Mūlasarvāstivāda Vinaya was translated into Tibetan in the ninth century by a translation team and is “the most complete and accurate form of this Vinaya” (Prebish 84). Due to its completeness and accuracy, the Tibetan version of the Mūlasarvāstivāda Vinaya serves as a main source for translations into Western languages. In the Tibetan language, it is a part of the KanjurFootnote 10 (II, 283). The tale is found in the section of the Kanjur devoted to the vinaya, the monastic discipline or ‘dul ba in Tibetan.
There are also two Chinese translations of The Painter and the Mechanical Maiden, both found in versions of the Chinese Tripiṭaka. The newer version is found in Taishō Shinshū Daizōkyō 1448 (24.19) under the original frame [a] of the Mūlasarvāstivāda Vinaya. It was translated from the Mūlasarvāstivāda Vinaya in the first decade of the eighth century C.E. by “Yi-Tsing” [Yìjìng 義凈] (Prebish 84, Hofinger 13). In comparison to the Tibetan translation, this corpus of translated texts is much smaller and the translations are “mediocre and incomplete” [trans. author] (Lamotte, 187).
The other, lesser-known Chinese translation is three centuries older. Taishō Shinshū Daizōkyō 207 4.8 is found among tales about Buddha without a major frame or context (and thus under frame [b]) under the section “Tsa P'i Yü King” [Zá pìyù jīng 雜譬喩經] or “Book of various apologues” (Chavannes 2: i) in Samyuktāvadāna-sūtra. The genre of the tale is avadāna (pìyù 譬喻), which narrates about past lives of other people than Buddha (Appleton ‘Jātaka’),Footnote 11 and it would have to be compiled by the beginning of the fifth century C.E. No translator is given for this version, however, the Korean Tripitaka (Koryŏ taejanggyŏng) mentions the “other version of the text”, T 208,Footnote 12 that was also compiled by bhikṣu “Tao-Lio” [Dàolüè 道畧] and translated by the Kuchean monk Kumārajīva in the year 401 (Chavannes 1: i).Footnote 13 Kumārajīva is generally believed to be the translator of all tales in this corpus.Footnote 14
Kumārajīva (344–412 or 415) was a famous Buddhist monk who lived and worked in Kucha. Kucha was an ancient Buddhist kingdom in today's Xinjang Uyghur Autonomous Region in northwest China “from which missionaries travelled to China proper” (Foltz 50) and where many Tocharian documents in the Tocharian B language were found. Tocharian The Painter and the Mechanical Maiden was found close to Kucha, in Šorčuq or the Qigexing Temple, which was a major Buddhist centre along the Silk Road in the second half of the first millennium. Kumārajīva being the translator of T 207 version is another argument for my hypothesis—built on the comparison of the additions to the respective translations—that the older Chinese and the Tocharian versions are related.
The Tocharian version of the tale belongs to the Tocharian Puṇyavanta Jātaka. This jātaka is known as a part of the Pāli Mahāvastu,Footnote 15 however, the Tocharian version adds more sub-narratives to the Puṇyavanta Jātaka and shifts the focus towards these sub-narratives.Footnote 16 In the Mahāvastu, “actual feats of the five companions […] make up the content of the Jātaka” (Lane 33), and in the Tocharian Puṇyavanta Jātaka, the adventure stories of the five companions are limited to a few lines. The narrative about five princesFootnote 17 is used as a frame for the embedded tales in the Tocharian Puṇyavanta Jātaka. Each of the five princes embodies a virtue and illustrates this virtue's superiority by giving an example in a tale. The Painter and the Mechanical Maiden serves as the first exemplary tale for the virtue of wisdom. This is the third, [c], context in which The Painter and the Mechanical Maiden is preserved.
Sanskrit, Tibetan, and the newer Chinese tales that are enclosed to the original frame of the Buddha narrating the episode at the lake Anavatapta as well as the Buddha narrating the Tocharian Puṇyavanta Jātaka render Buddhist interpretation of the text with mere framing. Some of the translations and adaptations, however, lost minor Buddhist elements. For example, the Tocharian version does not identify the craftsmen as the Buddha's disciples and the Tibetan version does not make a direct appeal to Buddhism in the conclusion. Despite this, none of the versions lost the Buddhist point completely; e.g. the decontextualised older Chinese tale nevertheless identifies the artisans as Śāriputra and Maudgalyāyana and concludes with them joining the religious order of monks.
The tale has circulated on the Silk Road as inherently mutable form that was widely adapted by cultures centred in the Buddhist religion. The tale originated from Indic folklore tradition until Buddhism housed it in several Buddhist corpora and embellished what we could now see as the pared down, even skeletal, version of the six tellings that we have at our disposal today. These added details of Buddhist values and characters transformed the essence of the tale into a religious exemplary lecture on the nature of knowledge and skill—a topic of interest outside the Buddhist religion and the Silk Road area.
The comparison of the content in different versions
Three topics were apparently of principle importance for the audiences of the time: the virtue of wisdom (as the moral of the story) and the twin problems of illusion and disgrace. The virtue of wisdom, insight and intuitive apprehension (Skt. prajña) is highly relevant to the problem of illusion in the tale because, despite succumbing to illusion and blind passion for the mechanical maiden, the painter ultimately defeats the mechanic by giving him a taste of his own medicine—the power of illusion. Although it is unclear whether the mechanic intended to trick his guest into believing that the mechanical maiden was real, the painter's intent to produce a believable illusion (a human-like artwork in place of an actual human being) is certain. The painter's triumph, therefore, derives from his ultimate success in returning the trick. In addition to demonstrating the superiority of his illusion, he manages to embarrass the mechanic in front of other people and even royal representatives. Although both artisans lose wisdom during the story by succumbing to artistic illusion, all versions of the tale conclude that the magical skill (Skt. ṛddhi) of bringing inanimate objects seemingly to life will never overcome the wisdom of a sage. Except for the Tocharian version,Footnote 18 none of the versions gives an explanation of this belief before or after the telling of the story.
The main point of difference among the versions of this tale lies in how they convey the message in the conclusion. The Sanskrit versions are identical in their endings, highlighting the superiority of Śāriputra's wisdom over Maudgalyāyana's magic powers.Footnote 19 The newer Chinese version (T 1448), much like the Sanskrit versions, addresses questions of magic and wisdom, and in it, the Buddha continues past the details of the initial contest to tell yet another story in which two painters (again prior births of Śāriputra and Maudgalyāyana) compete and Śāriputra wins again.Footnote 20 The older Chinese version (T 207) stresses the negative outcomes for both artisans, who realise that, in this world, men deceive each other and that truth is nothing but deception, and, ultimately, they join the monks, overtly appealing to readers to join Buddhism.Footnote 21 The Tibetan version, on the other hand, refrains from further comment after the painter reveals his trick.Footnote 22 The Tocharian version accentuates human wisdom and its absence in non-human entities, but never turns to Buddhist ideas.Footnote 23 Thus, we can see that some versions appeal directly to Buddhism (all Sanskrit and the newer Chinese version) while others allude to it only implicitly (Tocharian, the older Chinese) or not at all (Tibetan). Although the same themes are addressed in all versions, the ending of the original two Sanskrit versions has clearly been adjusted in Tibetan translation, the older Chinese (T 207) translation, and Tocharian adaptation―possibly because the latter also appear in different frames than the other versions, which are sub-narratives in the episode at the lake Anavatapta. In the latter versions, the ending becomes a sort of cadenza in which the translators (or rather, in this case, writers) stress the point that will affect their audience most powerfully. In T 207, the same sort of embellishment also occurs in other segments of the story, which overall overlap with the additions in the Tocharian version.
Based on this relatively free approach to translation, one in which a translator felt no hesitation in embellishing the text on their own, it appears that the translator's role in these communities was not only to transmit cultural knowledge through language, but also to influence the reception of that knowledge in a way they found most appropriate, and perhaps most salient for their respective audiences. As a result, translations not only shifted to the new cultural environment but sometimes produced new elements in order to reflect the needs of different specific audiences. For example, Chavannes reports that there are two versions of Kumārajīva's translations, proving, in his opinion, that the translator “remained open to additions and deletions that editors made of their own will” [trans. author] (1: i).Footnote 24
Complete faithfulness to the original version was clearly not a priority for translators of The Painter and the Mechanical Maiden. In addition, the availability of two versions of the original Sanskrit tales, which differed in their elaboration of the story's details, introduced more diversity to the story before it began circulating to other areas. This is particularly true of the version from the Kaṭhināvadāna, which features a longer description of the painter's attempts to court the mechanical maiden. Overall, most of the translations are quite faithful to the original Sanskrit stories and follow the storyline as described above without major modifications that would alter the narrative's significance. Two versions, T 207 and Tocharian version, however, both add a significant amount of detail to the otherwise short and relatively unembellished story. As mentioned above, the Tocharian version has been expanded significantly (it is by far the longest text of all versions) and should be considered an adaptation rather than a translation of the tale. In the following section, I make a comparison between T 207 and Tocharian tales and proceed to discuss the Tocharian version in relation to the original Sanskrit tale.
![](https://static.cambridge.org/binary/version/id/urn:cambridge.org:id:binary:20201030130426914-0789:S1356186320000152:S1356186320000152_fig1.png?pub-status=live)
Fig. 1. A detailed breakdown of the different versions in their content alterations
Taishō 207 and the Tocharian version
T 207 and Tocharian versions show a number of similarities in their alternations to the original Sanskrit versions that merit a direct comparison. While it is difficult to prove specifically any direct influence between the two versions, their content and their textual and historical contexts indicate that one of the versions influenced the other. The geographical nature of the Silk Road trade route yields a high probability that it was the Tocharian variant that influenced T 207; Buddhist monks regularly travelled from India to China via Tocharian territory in the Taklamakan desert. Kumārajīva himself—the above-mentioned Buddhist monk who translated the Buddhist texts to Chinese—is a good example of such practice in the late fourth century and at the beginning of the fifth century.Footnote 25 Tocharian and Chinese Buddhist monks were likely in contact from the mid-second century C.E. (Neelis 112, Nattier 3, Walter 5, Harrison 118)Footnote 26 but we do not have substantive knowledge about the circulation and translation of Buddhist texts between China and the Tarim Basin during that time and in the following centuries (for example, what was the role of Tocharian texts in this transmission). Local Tocharian Buddhist culture was extinct by the eighth century C.E. (Mallory and Adams 590–594, Mallory 55–65).
Another clue that links the Tocharian and T 207 versions of the tale appears in the introduction to the T 1448 version, which Hiän Lin Dschi [Jì Xiànlín 季羨林] remarks “mostly matches with the Tocharian version” (323). Although Dschi points out the similarities between the Tocharian and Chinese T 1448 versions, he was not familiar with the T 207 Chinese version, which is, in fact, even more similar to the Tocharian.
It is necessary to note that translations into modern Western languages often relied on “Chinese parallels [that] cleared up several difficult spots” in the Tocharian version (Lane 35), meaning that modern translations are sometimes a mixture of the two sources. In order to overcome this problem, of course, one should work with original texts and not translationsFootnote 27—but such a practice still cannot overcome the significant problem of lacunae in the original texts. Most versions of The Painter and the Mechanical Maiden have a few lacunae, including the original Sanskrit ones (Pinault Chrestomathie 252).Footnote 28 Additionally, it is likely that there once existed more versions of the tale than those we can access today, meaning that we can only speculate about possible connections between the two versions with the following textual evidence.
The first resemblance between the Tocharian and the T 207 version is that, unlike the other versions, these two accounts do not identify the painter and the mechanic as two of the Buddha's disciples. The artisans are only two characters in the tale, which is one example of how these two versions do not contain certain significant details or show changes in their meaning. The omission of a particular ‘policy of the land’, which dictates that the king must verify every suicide, loosens the story's connection to any particular kingdom, and, as Pinault points out, diffuses its “ethnographical background” [trans. author] (Chrestomatie 263). In other versions, the inclusion of this detail augments the mechanic's shame by making that shame not only publicly, but royally witnessed.
The Tocharian and the older Chinese versions also take pains to accentuate the mechanic's disgrace but do so in a manner that differs from the other versions. In all versions of the tale, there is a crowd of busybodies (neighbours or officials) who come to see the dead body and act as “a sort of choir” (Pinault Chrestomatie 263). Pinault adds that the naïve observers of the painting act as “a mirror image to the supposed audience of the story” [trans. author] (Chrestomatie 263), which is unlikely. The readers/hearers of the story are rather situated as observers who can see the illusions and its perils, particularly since the audience of The Painter and the Mechanical Maiden has likely seen/heard this and similar stories before. This knowledge that only the external audience holds creates a humorous perspective on the deception of the characters and their consequent disgrace in front of the crowd/royals. The Tocharian version confirms this interpretation: after the painter reveals his trick, “All people that have seen this being amazed they laugh” (A9 b5-b6). They laugh at their own naivety, just like the audience of the tale was able to laugh at the delusion of both artisans. In all but the Tocharian and the older Chinese versions, cutting the rope with an axe is suggested by the crowd. In the Tocharian and the older Chinese tale this suggestion is given by the mechanic himself—a point emphasised twice in the Tocharian version (Malzahn A9 b1 and b2) and slightly varied in the older Chinese version by the change of weapon: instead of an axe he uses a knife.
Aside lacking the same information present in all other versions, the Tocharian and T 207 version also share certain common embellishments. For instance, both tales include a description of the mechanical maiden and her irresistible beauty and a detailed description of the hanged body, neither of which is found in the original or any other versions. In this regard, T 207 version is especially picturesque, despite its overall short and concise form. Georges-Jean Pinault remarks that the description of the hanged body is very Villonesque, a reference to François Villon's Ballade des pendus, also known as Épitaphe or Frères humains (Chrestomathie 263). The following excerpt from T 207 illuminates this analogy well:
Thinking so, he immediately painted his own image on the wall, covered it with clothes identical to the ones on his body, moreover, he drew a rope squeezing his neck as if he was hanged, he also added some flies on the lips and some birds pecking his perished mouth [trans. author].Footnote 29, Footnote 30
The Tocharian description is no less illustrative:
The head tilted a little –
the eyes set, with the protruding lips (lit. stretched towards the front) the [last] sighs going out of the throat, hands and feet hanging down,
with the lascivious lower abdomen, a rope around his neck, hanging on a nail (like one) killed,
the golden skin of this body having become pale
thus he painted himself as if real (Malzahn A8 a4-b1).Footnote 31
As these short excerpts show, the two versions are artistically refined and make use of rich descriptions and direct speech in order to make the story vivid and literarily attractive. Most of the other versions use direct speech, but the Tocharian version is notably more elaborated, deploying versification for direct speech and tense shifts for dramatical effect. In his paper on ‘Tocharian drama’, Werner Winter argues that the Tocharian version was performed as drama, although not in the Western sense of drama as a staged production. Although the Tocharian tale lacks main criteria for drama—the “designation”, “stage directions”, “mention of a typical stage character”, and “fast change of the action” (Winter 27)—Winter nevertheless argues that it could be considered performable due to the tense shifts and the interchange of prose and verse (28), the so-called prosimetrum, a combination of prose and different types of meters, which is known from Indian sources as the campū style (Gunkel 82, Pinault Chrestomatie 407). On the latter point, Winter notes that all verbs in the present tense in the tale “denote an action that can easily be enacted on stage”, for example, speaking, laughing, yawning, etc. (30), although certain other verbs that could also be acted out (painting, weeping, shouting, etc.) remain in the past tense. Winter explains this contrast by positing that events which could not readily be performed on a stage were, instead, narrated in the past tense. He concludes that “Tocharian dramatic performance is done on two levels—that of action proper, and that of narration. The range of enacting is very small; no stage implements are used” (33). Moreover, he cautiously suggests that the performance element of the drama might even have been danced, based on “terms used to label the tunes” (33).
The type of performance postulated by Winter strongly resonates with the Chinese genre of chu-kung-tiao [zhūgōngdiào 諸宮調], “a ramification of story-telling, set to musical tunes” that started in the early eleventh century and flourished in the twelfth and thirteenth centuries in Shanxi province (Nienhauser 332, Chen 124). The similarities between chu-kung-tiao and the Tocharian tale are strongest in the prosimetric pattern: the melodies in chu-kung-tiao that belong “to the same mode are arranged in suites [and] the different suites are connected by prose passages. As a rule, each suite belongs to a mode different from those of the suites preceding and following it” (Nienhauser 332). This description exactly parallels what is found in the Tocharian tale, where the main ideas are versified in various types of meters. Additionally, the “suites” are similarly as “short and concise” as the Tocharian verses (Nienhauser 332).
Another Chinese genre from that same time, called pien-wen [biànwén 變文] (literally, transformation texts), also shares some characteristics with the Tocharian tale―the most obvious again being the prosimetric form (Mair 90). It is generally accepted that the pien-wen genre was brought into China with Buddhist narratives from India (Mair 90–93, see also Pinault ‘Buddhist’ 100–101)Footnote 32 in the Tang Dynasty and Five Dynasties periods (618–960 C.E.). According to Victor H. Mair, Chinese “transformation texts were a form of Buddhist-influenced prosimetric storytelling (normally associated with pictures) that enjoyed broad currency, particuarly among the lower strata of society, from the middle of the T'ang period to its end” (170). Mair defined the following characteristics of the genre: 1) pien-wen are narrative texts, 2) in vernacular language, 3) with prosimetric structure, 4) related explicitly or implicitly with images, 5) using specific pre-verse formulas in front of versified parts (15, 27, 73–105, see also Pinault ‘Une version’ 210).
Noting that the characteristics and time frame of the pien-wen genre align with the Tocharian Buddhist materials, Pinault concludes that Tocharian narratives “provide the missing link between the Indic prosimetric genre and the Chinese pien-wen” (‘Buddhist’ 101, ‘Une version’ 212). All five characteristics of pien-wen are also present in the Tocharian tale; though illustrations have not been found for this particular tale, however many jātakas and avadānas were illustrated in Kizil caves in Kucha of the fifth to seventh centuries (Zhu 58). The pre-verse formula in pien-wen, according to Mair (6, 27–28, also in Pinault ‘Une version’ 210), goes as follows: “(Please look for a moment at the) place [where] X [occurs]. How [should I] present [it]? Or: How does it go?” This formula is also found in the Tocharian tale. For example: “She, that with his [the mechanic's] reverence held in her hand, as it were, beauty and reverence, attended to him. But how so? || in the S.-tune:|| Like one ashamed casting her glance to the ground a little, she looked lovely” (A5 b1-b3).Footnote 33 “He painted himself opposite (?) the door. But how so? || in the Ṣ.-tune:|| The head tilted a little – the eyes set” (A7 a4-a5).Footnote 34
These versified segments present “oral ‘performance’, as opposed to the prose segments that are visualised by the reader (Pinault ‘Une version’ 210). Pien-wen is always labelled as a narrative and not performative genre, but this explanation by Pinault seems to suggest—judging from the Tocharian sources—that there might have been a performance. If the two genres, Chinese pien-wen and the Tocharian tale, are indeed related, then Winter's theory on Tocharian “drama” should be tested on pien-wen texts as well; such an exploration would hopefully shed some light on these narratives and the way they were shared among and within the communities.
Comparative study of these literatures is thus an important undertaking and should first be preceded by detailed study of the transmission of Indian materials to Tocharian areas. For instance, classical Indian drama needs to be taken into account before drawing any conclusions on dramatic elements in Tocharian texts, because “the classical drama of India has a peculiar construction, the prose being continually interrupted by stanzas in various meters” (Mair 97). The campū style is found in other Tocharian texts (e.g. text of the legend of Prince Haṃsasvara in Tocharian B, marked as IOL Toch 36) (Kim 454) and has been shown to be adopted from India (Pinault Chrestomatie 407), where it became increasingly popular in the tenth century (Sivaramamurti et al.). In addition to that, Tocharian language has recently been shown to use common Indian meters (Gunkel 82)—as opposed to Tibetan which retained indigenous meters when adapting and translating Sanskrit texts. The Tocharian tale was composed sometime between the fourth and sixth centuries C.E., and literary influence from India is a fact.
Werner Winter suggests that an additional parallel be drawn with the Tibetan drama, which “shows striking affinities to the type of performance described [in the Tocharian tale]” (35). Winter never points out these striking affinities, of which there are more than a few. It is known that the Tibetan drama (zlos gar)Footnote 35 and the genre of the Tibetan opera ((a lce) lha mo)Footnote 36 use “either jātakas, or Indian or Tibetan tales modelled on jātaka tales” (Henrion-Dourcy 198, see also Snyder 24) and are thus highly influenced by Indian culture. “Ache Lhamo […] is said to have evolved from a Buddhist storytelling genre in which … a ballad singer presented tales by unrolling picture scrolls that depicted popular narratives … based on jātaka tales” (Foley and Karter 131). Since Winter makes an argument that the Tocharian tale was sung and performed (33), it could be assumed that he saw resemblances primarily in this type of performance, which is also characteristic of Tibetan plays. The Tibetan adaptation of a text (rhnam thar) into a stage version (‘khrab gzhung) is made through versification, which “heavily cut[s]” the prose “whereas dialogues in verse are kept as such” (Henrion-Dourcy 198). Similarly, The Painter and the Mechanical Maiden was adapted into Tocharian with a strong emphasis on versified parts of the text. In Tibet, the transposition of a pre-existing text to the stage leaves the text in the centre, which makes actors “first and foremost storytellers” rather than performers (ib.), much like in the Tocharian version. Tibetan actors do not impersonate and embody the characters (203, 207) and “go in and out of their roles continuously” (201). As opposed to voice, “physical movement and demeanour … receive very little attention” (206). Physically, “actors display their characters as life-size icons, as if they were puppets … with the ‘envelope’ of the character (the costume, sometimes a mask) and stylised dance moves” (203). Another resemblance of the Tibetan drama with jātakas is in the traditional framing of jātakas and avadānas: a story from the present, a narrative from the past focusing on an important Buddhist figure, a moral juncture. A lha mo performance is divided into three main parts: the prologue (‘don), which is always the same, the libretto (gzhung), and the auspicious conclusion (bkra shis), with central stories being well-known to the audiences (197), as with jātakas.
This evidence suggests it is possible that Tibetan sources served as a link between Indian narratives and Chinese pien-wen. As Meir shows, “it is the verses that are the central, stable core of a prosimetric folk narrative in the Indian tradition, and in other Asian traditions influenced by it” (98), which is also true for the Tibetan drama (Henrion-Dourcy 198). Another connection between these genres is that pien-wen, lha mo, and jātakas originated from Buddhist texts that were accompanied with pictures. All these vernacular traditions having in common so many fundamental traits indicate a likely connection.
Despite the fact that pien-wen genre was very likely influenced by the southern cultures through Buddhism—as confirmed by Mair's “Indian hypothesis” (106–109) and substantiated by evidence from the Tocharian text discussed here and, foremost, in Pinault (‘Une version’ 209–213)—these genres were not necessarily alike in all their characteristics and have evolved in specific ways in the respective languages. From the perspective of typical content, for instance, the differences between the two genres are rather large: pien-wen recounts “heroic, epic” events while chu-kung-tiao narrates “domestic, realistic-comic” events (Chen 132); Tocharian material typically corresponds to neither type. Pinault proposes that it is perfectly possible to imagine that the Tocharian tale was composed entirely in the Tocharian milieu from numerous Indo-Buddhist sources that the author had at his disposal (Chrestomathie 268). He also reports that the precise sources of the additions in the Tocharian tale have not yet been located in Buddhist literature (Chrestomatie 263), except for the list of prohibited women, which is found in Buddhist sources (such as Mahāprajñāpāramitāśāstra or The Treatise of the Great Virtue of Wisdom of Nāgārjuna, translated by Kumārajīva) as well as Hindu sources (such as Mānava-Dharmaśāstra or The Laws of Manu). As I demonstrated at the opening of this article, the folkloristic influence of these additions is significant and very likely. Since the Tocharian tale is labeled as jātaka and included to a collection of jātakas, we can assume that the circulation was similar if not identical to standard jātaka tales, which circulated orally in the form of folk tales: as a result of this orality, the versified parts of the tales tended to be better preserved than the prose (Gokuldas De in Mair 97). The missing link between the Indian and Chinese tales (cf. Chavannes 1: xviii) may therefore be found in the connection of folkloristic materials to the propagation of Buddhism beyond the Indian subcontinent. The Painter and the Mechanical Maiden is a perfect example in which a pre-Buddhist tale may have been appropriated and adapted to meet religious ends. As my analysis of the Tocharian tale shows, the connection between India and China seems to have occurred with at least some mediation of Tocharian materials. This argument was made by Xianlin Ji [季羨, Hsien-Lin Chi], per Meir's opinion, “an important but almost wholly ignored article” (96), where he proposed that Tocharian texts have functioned as an intermediary stage in the introduction of prosimetric form to China through works such as The Mahāvastu and Pañchatantra (Chi 125).
Further support for a connection between the discussed genres and the Tocharian version of the tale comes from the fact that all of them used more or less the same content, with Śāriputra and Mahāmaudgalyāyana, Buddha's main disciples, as common protagonists. These texts, however, were not necessarily or completely Buddhist. Pinault, for instance, believes that the additional passages in the Tocharian tale “appeal to the stereotypes and citations from normative Indian literature – in a nutshell, on a ground that is not unique to Buddhism” [trans. author] (Chrestomathie 264). He adds that, since Buddhist sources for these additional materials have not yet been identified, the sources were most likely non-Buddhist (Chrestomatie 263). A non-Buddhist source seems particularly likely because the ideas discussed in these additional passages can also be found in juridical tradition and gnomic literature, such as The Laws of Manu (Chrestomatie 265). Accordingly, in the following paragraphs, I discuss the additions to the Tocharian tale (if compared to the two original Sanskrit versions of the tale) and their connections with the non-Buddhist sources.
As mentioned in the comparison with T 207 version, there are two additions that the Tocharian and the older Chinese versions share: a description of the wooden maiden (Malzahn A5 b3, A6 a1-a2) and a realistic description of the hanged painter (A7 a4-b1). Other Tocharian additions to the story include a description of the mechanic's hospitality (A6 a3-a5); a description of painter's courtship with the mechanical maiden (A6 a5-b3); a list of prohibited women (A6 b4 to A7 a2); and the painter's lecture on his trick with illusion (A7 b2-b4) (Pinault Chrestomathie 263). These segments greatly contribute to the literary elaboration of the Tocharian version, not only by adding new information and perspective to the story but also in form. Most of these additional passages dramatise the narrative by changing the tense from the past into the present tense (e.g. A6 a1-a2, A9 b3-b4) and by versification (Pinault Chrestomatie 407); half of these additional passages are versified (the descriptions of the wooden maiden, of the hanged painter, and of their courtship) and were presumably sung (Winter 34).
The list of prohibited women particularly stands out among these additions, because it is not an elaborated description of a feature already present in other versions, but rather—like the painter's lecture on illusion—an emphasised point in the story. The list is inserted into the part of the story where the painter ponders on why the maiden was given to him and whether he should hold back or pursue her. I quote the entire list here:
Again he thinks: Seeing the great danger, the wise ones are not allowed here to profess love to ten kinds of women. Thus it is said: to the royal spouse, to the father's spouse, to the spouse of a general, to that of a relative, to that of the teacher, to an exceedingly adulatory woman, to a woman thinking of profit, to a woman available to many, and mainly (to a beautiful) (to a beauty)ful woman he who loves his life shall not go. Therefore this one as [she is] affiliated to my relative and mainly in her being beautiful to look at must not be made aware of the love (A6 b3 to A7 a2).Footnote 37
There are quite a few categories of women included in this list: those prohibited through kinship (“to the father's spouse”, “to that of a relative”), those prohibited through social status (“to the royal spouse”, “the spouse of a general”, “to an exceedingly adulatory woman”, “to that of the teacher”), and those prohibited through the likelihood of adultery (“to a woman thinking of profit, to a woman available to many, and mainly (to a beautiful) (to a beauty)ful woman”). Adultery is, according to many mentions in The Laws of Manu, one of the worst crimes: “When a man carries on a conversation secretly with another man's wife, he is subject to the lowest fine if he has been previously accused of similar offences” (Olivelle VII 354). In the case of adultery, everyone other than a Brahmin merits the death penalty (VI 359).
In the Tocharian version, we find that the mechanic is the painter's relative and the wooden maiden is the mechanic's affiliate: thus, respect towards one's kin makes her prohibited to the painter. Furthermore, she is too beautiful to be courted—a prohibition that comes off as a sort of warning against the femme fatale. Pinault explains that women on such lists of prohibition tend to have a protector, usually a relative, and that otherwise they fall under the protection of the king (Chrestomatie 265).Footnote 38 Thus, the prohibition arises through the fear of incest, as the identity of a woman could be lost without a protector: “A wise man must not marry a girl who has no brother or whose father is unknown, for fear that the Law of ‘female-son’Footnote 39 may be in force” (Olivelle III 11). The protector therefore—in this case, the mechanic—plays the role of the woman's father, in the process creating further resonance with the Pygmalion myth and The Sandman's story, as we shall see below.
By courting the mechanical maiden, the painter is clearly breaking at least two of the laws. Besides that, The Laws of Manu also say that “doing favours [to the wife of another], touching the ornaments or clothes, and sitting together on a bed—all this, tradition tells us constitutes adultery” (VII 357). While, in the original story, the mechanic's offer of the maiden to the painter as a servant (Degener 48, Hofinger 187) might be interpreted as either a trap of temptation for the painter or simply a hospitable gesture, the addition of the list of prohibited women makes it clear that the painter is not allowed to court the inanimate woman.
Analysis of the story and further resonances
In the following literary analysis of the tale, I discuss the notions of illusion and imitation from the Western point of view, with the help of Greco-Roman philosophy and examples from Western mythology and literature. I also compare the Silk Road tale to E. T. A. Hoffmann's The Sandman and illustrate the considerable resonance between these texts with no cultural or geographical connection.
First, let us consider why the painter is universally believed to be the winner of the involuntary contest between the two artisans who manage to deceive each other. Clear answers to this question are, in fact, few. One possibility is Buddhist convention: many texts in which the tale is embedded contain additional examples of contests between the two artisans, identified as Maudgalyāyana and Śāriputra, and Śāriputra's wisdom always prevails over other virtues. The newer Chinese version of The Painter and the Mechanical Maiden explicitly states that although the mechanic thought he would be able to win because of his skill, the painter ultimately won with the same, but even better, skill (the same conclusion is made in the subsequent story, which depicts a competition between two painters, identified as Maudgalyāyana and victorious Śāriputra).Footnote 40 This Buddhist convention, therefore, interprets the popular story of a contest between two artisans according to a Buddhist value system.
Universally affirmed by Buddhist are two (twin) virtues, compassion (karunā) and insight or wisdom (prajña) (Jackson 664, Keown ‘Mahāyāna’). In the main genre of Mahāyāna Buddhism, prajñāpāramitā or the Perfection of Insight (in older sources translated as the Perfection of Wisdom), wisdom is the last of the Six Perfections that make up the central element of the Mahāyāna path and is known—especially from the Prajñāpāramitā Sūtra corpus—as the culmination of the Bodhisattva's practice (Keown ‘prajñā-pāramitā’, Williams 51). In the other great branch of Theravāda Buddhism and other mainstream Buddhist texts, “wisdom is, with morality (śīla) and concentration (samādhi), one of the three indispensable Buddhist trainings” (Jackson 665). Although different branches of Buddhism do not conceive the virtue of wisdom in the same way, it follows from here that prajña Footnote 41 is deemed higher than ṛddhi,Footnote 42 which is the magical skill represented by the mechanic in this tale. “As earthly attainments”, abhijñā (higher knowledges), among which ṛddhi is found, “are deemed available to non-Buddhist sages” and thus mundane achievements (Pranke 8). Besides, the possession of superpowers, distinguished by the Buddha and his eminent disciples, is superior to superpowers acquired by some type of magical charm. Thus, an illusion of a magician is on a lower level than an act of the Noble Ones who can create actual, real things with their superpowers (Fiordalis Miracles 148). The frame narrative to the Tocharian version explains that the prince Prajñavān is nowadays known as Śāriputra (Malzahn A17 a1) and that his wisdom made him a king. “There is no other thing that does so much good to the word as wisdom, because ignorance is the root of the spread of all bad things and all damages” (A4 a1-a3). However, even “the wise one”, along with the other princes and their virtues of beauty, energy, and craftsmanship (A16 a1-a2), needs to strive for perfection and praise the Buddha as “wisdom without virtue does not shine” (A17 b5).
A second reason for the painter's victory, as underlined in most versions of the tale, is the fact that the painter's shame is revealed in private, while the mechanic's shame is public and therefore more disgraceful. A third possible explanation might be that the painter's skill is two-dimensional (and thus less life-like figure) while the mechanic's is three-dimensional (working with figures). From this perspective, the mechanic is more responsible for his trompe-l’œil as he was deceived by the painter's two-dimensional image of a corpse (believing it to be a real, three-dimensional corpse), whereas the painter is deceived by the mechanic's three-dimensional imitation of a woman (which is, as an object in itself but not necessarily as an imitation, more human-like). It is odd for the mechanic to have lost this contest, given that it took place on his own terrain—in terms of dimensionality and literally, as a host who has already tricked his guest with a similar illusion.
The Tocharian version does not distinguish between the two types of artisanship: every artist that “creates an art(istic) object, … often he has five advantages [from it]:” the object itself “comes into existence”, “the pleasure arisen from the skill”, he “receives friendship from the beings”, “obtains students” and “property beyond that” (Malzahn A2 b4-A3 a1). A good craftsman is praised as “worthy being treated with reverence and worthy to gather the adulation of humans”; material things can be “cut off” by “water, fire, kings and thieves”, but his “skill will never be lost” (A2 b2-b4). Craftsmanship is thus a valuable skill, which is not hierarchically valued in regard to the type of skill but rather the quality of the skill (and the creation) itself.
Despite these contrasts between how the artisans’ skills are perceived, there does not seem to be much practical difference between the two artisans and their reactions. T 207 is the only version that openly addresses this issue; the end of this version acknowledges that both artisans equally failed to recognise the truth: “The carpenter then said: ‘You could fool me, and I could fool you. Our host and guest relationship is ended. We owe nothing to each other’. They said to each other: ‘In this world, men deceive each other; how is this different from what has just happened?’” [trans. author].Footnote 43 Both artisans play with optical illusions that make them appear to have supernatural powers. Their power of illusion is so convincing that they both believe the other's artifice to be real without hesitation; neither notices that the object they look at is a visual representation rather than reality. The painter even mentions that the mechanical maiden's breasts lift a little, implying that she is breathing (Malzahn A6 a1-a2), and he himself paints his own image on the door as if real (Malzahn A8 a6). Both artisans thus fail to comprehend this twice-removed reality.
Looking at the contest from a Platonic perspective, the conclusion about the contest is just the opposite from the Buddhist conclusion, which proclaims the painter's wisdom as superior. Per Plato's views, the painter is the one who truly lost the battle because he should know better than anyone—and certainly better than the mechanic—how illusion works. According to Plato's The Republic, the painter is “not a craftsman of some kind”, like a mechanic would be, but “an imitator of that of which these others are craftsmen” (597d), meaning that he does not imitate the idea itself but “the works of craftsmen” (598a). In other words, painters are able to imitate without knowledge of the truth (598e-599a) and hence are twice removed from the truth. Craftsmen and mechanics, on the other hand, are only once removed from the truth, as they imitate the form itself (597e-598c).
In the discussion on imitation from the Book X of The Republic Socrates explains that there are three levels of existence: the form of a thing (the abstracted notion of a couch), the individual thing (a particular couch), and the imitation of the thing (a painting of a couch). In The Painter and the Mechanical Maiden, the painter's work clearly falls into the final category of an imitation; however, the status of the mechanic's artifact, the mechanical maiden, is less clear. Depending on interpretation, the mechanical maiden may be an instance of either the second (the individual thing) or the third category (the imitation). The mechanical maiden is either an imitation of a human form and therefore falls into the third category of existence, or she is an individual thing, which is, in addition to that, an imitation of a human form, and so falls into the second category of existence (the status of a mechanical maiden would be like that of a lamp, for example, that imitates the light of the sun). For the painter, the mechanical doll is an individual thing (of the second category) since he is delusional in his perception of this human-like object as a fleshly human. For the mechanic, however, it cannot be said whether he considers his creation solely an imitation or an individual thing.
If the mechanic's intention was indeed to compete with the painter, then his work moves from the second category—where the work of a craftsman would normally fall—into the third category, where artisanship mingles with art. Under this interpretation, like everything that imitates human form, the mechanic's maiden imitates both the human form itself and the human-like form of a doll and is twice removed from the truth. Given the other option, under which the mechanic never intended to trick his guest, we must infer that the painter's passionate desire for the mechanical maiden was so strong that it made him completely blind to reality: “Oh, such is the power of passion!” (Malzahn A7 b1);Footnote 44 “Fie, blind passion!” (Malzahn A8 a1).Footnote 45 In this case, the illusion that the painter experiences, in which a human-imitating object is taken for human, is caused not by the mechanic's skill but rather by the painter's delusional mind.
There is another twist to the Platonic analysis of The Painter and the Mechanical Maiden. In The Republic, Socrates claims that the painter's knowledge is inferior to that of the maker of other products and, furthermore, that the maker's knowledge is inferior to that of the user's (601c-602b). In the Silk Road tale, however, the user is always less knowledgeable than the maker. This implies that both the painter's and mechanic's imitations were so perfectly made that they superseded the category of imitation and entered the category of individual things. In other words, they moved from works of art to the reality that particular works of art represent.
On the topic of the reliance of representation as a criterion for success in painting, philosopher Abhinavagupta (950–1016 C.E.) writes that in “that which bears a similarity to another … its soul is worthless: for an imitation, we are aware only of what is imitated, as in a painting of a manuscript. There is no conscious awareness of the minimum and other paints [of which it is constituted]. And this fact does not lead us to prize it” (Ingalls et al. 717–18, Kachru 31).Footnote 46 For Abhinavagupta, a portrait is thus an inferior mode of realising likeness; however, he discusses portraits that are recognised as portraits and not considered as the actual objects that they depict. The latter perception of a painting is a kind of madness—the ultimate loss of insight into reality—which makes for a double loss: the living portrait is at loss with the human it depicts, the observing human is at loss with reality.
There is plenty of commentary to be found on the problem of illusion throughout Western philosophy and fiction. For instance, a contest between two artisans is a standard theme in the Greco-Roman tradition; a famous example comes from Pliny's Naturalis Historia in which he recounts the contest of two renowned painters, Zeuxis and Parrhasius, who also trick each other with illusions. Likewise, statues treated like living women and statues undergoing an actual metamorphosis are common in many mythologies. A well-known example of the first type comes from a story in which Zeus wants to make Hera jealous, so he dresses up a new bride—a wooden doll. Angry Hera undresses the doll, recognises that it is an artificial object, and requires it to be destroyed. An illustrative example of the second type is the story of Pygmalion, famously depicted in Ovid's Metamorphoses: Pygmalion sculpts a perfect ivory woman and asks Venus to give him a woman like his ivory statue.Footnote 47 The statue truly becomes a woman of flesh and blood, and she and Pygmalion, her maker, marry and have a child. In other texts, the statue may be replaced by some other work of art, such as a painting depicting a human, or by a scientifically modelled human-like creature, like an automaton. E. T. A. Hoffmann's Olympia, from his short story The Sandman, is a particularly powerful example of the latter.
When we consider these canonical Western examples alongside the Tocharian tale, we can see how a cluster of similar motifs occurs across the non-metamorphic variants of the story. This allows for a comparison of the Silk Road tale with two texts, one from Greek Antiquity, the other from German Romanticism, each of which have been influential in the Western literary tradition and beyond—E. T. A. Hoffmann's The Sandman and Ovid's Pygmalion.
All three stories are related through a main motif of a man falling in love with an artificial woman—a motif that grew into a major theme in the Western literary tradition. The most important difference among the three stories is that, in The Painter and the Mechanical Maiden and The Sandman, the inanimate woman's creator is a different character from her naïve suitor. In Pygmalion's story, by contrast, the suitor and the creator are one and the same character, Pygmalion himself. Most other major differences between the three stories stem from this central fact. For instance, since Pygmalion is the statue's creator and is therefore aware of her artificiality, he cannot be deluded to the extent that the painter and Nathanael are in The Painter and the Mechanical Maiden and The Sandman. What is more, Pygmalion's ivory girl undergoes an actual metamorphosis as her “flesh / grew soft, its ivory hardness vanishing” (Ovid 233), following which his relationship ends in a happy marriage and offspring—quite a different scenario from the disappointing experience of the other two suitors.
The cluster of motifs that connects the Silk Road tale with Hoffmann's tale is rather rich. Besides the main motif, three other motifs are surprisingly identical, as already point out by Pinault (Chrestomatie 251), although they lead to very different endings in the two stories. First is the motif of blind passion, which causes the two suitors to lose their minds and their judgment by falling for the trick of illusion. Their minds (and their eyes, which play a very important role in The Sandman) completely fail them and let their hearts prevail. Passion leads Nathanael into failure, increases his delusion, and ends ultimately in fatal heartbreak. The painter, on the other hand, is saved from his delusion and, disgraced, seeks revenge. Despite significant differences in the unravelling of the two stories, the same motifs appear consistently across the tales: the madness of blind passion, the loss of rational judgment, and, finally, the motif of suicide as a plot twist. Pygmalion does not deal with any of these issues: his animated woman only breaks in the sense of relinquishing her human-like ivory armour in favour of actual, human flesh, rendering her ready to be his wife forever after.
Pygmalionesque stories of animating woman-like creations and falling in love with them are common also in the East. Kālidāsa, a Classical Sanskrit writer who lived around the time when the Silk Road tale was translated into Tocharian, dramatised the story of Śakuntalā from Mahābhārata, and a scene involves a mad, amnesiac king who turns his beloved woman, who used to be a painting, back into a painting (Kachru 35). In India, such stories were present in folklore, as testified in The Kathāsaritsāgara (Ocean of the Streams of Story), a famous collection of Indian tales from the eleventh century, attributed to Somadeva, which includes a story aptly titled The Merchant who fell in Love with a Painting. In this tale, a merchant's son falls in love with the painting of a princess, thinking she is an actual woman. A hermit, Boddhisattva in disguise, uses his powers to paint a cobra next to the beloved princess in the painting. The cobra bites the princess and the merchant's son—just like Nathaniel in The Sandman and, by illusion, the painter from the Silk Road tale—decides to commit suicide. He is stopped by Boddhisattva who asks him: “what is this delusion of attributing reality to the creation of your own desire that has taken possession of your passionate heart?” (Somadeva The Ocean 91–92). Instead, Boddhisattva suggests, he should “investigate the truth with equal intensity in contemplation, in order that [he] may not again become the victim of such sorrows” (92)—a message close to the Silk Road tale's variant conclusions.
In the same edition, a tale The Three Young Brāhmans who restore a Dead Lady to Life (Somadeva The Ocean 179–181) presents a variant from Mongolia, in which a girl is carved out of wood by four assemblers and comes to life (264). Each of the assemblers contributed a different part and each claims the girl for himself. A similar, and better known, tale of assembling and animating a living being—a lion—is also a part of the Tocharian Puṇyavanta Jātaka, entitled The Foolish Lionmakers, and is also found in the Sanskrit Pañcatantra collection (a pre-Buddhist source).
Another edition of the Kathāsaritsāgara collection presents another tale focusing on a Pygmalionesque relationship—Niścyadatta Meets a Vidhyādharī—in which (again) a young merchant falls in love with an upper-class woman, carved by a painter and a sculptor from a pillar, who possesses supernatural powers (Somadeva Tales 91–104). Delusion is not at work in these two tales, as the women are actually animated. Nonetheless, the mere act of creating a woman-like entity and acting as if she was one's property and a woman of flesh and blood is Pygmalionisque in its essence.
In China, this motif was popularised in Pu Songling's collection Liaozhai Zhiyi (聊齋誌, Strange Tales from a Chinese Studio) from the eighteenth century; Songling liked to use characters of poor and young scholars who fall in love with depictions of beautiful women. The eighteenth century is also the time when Pygmalionism became a prominent theme in Western literatures and visual arts, where it persists as a prominent trope.
The dominance of this motif in the earliest texts from India to China tends to be disregarded, however. Besides European and East and South Asian cultures, I have located it, among others, in Native American (Swinton 181–82, Boas 744–46, McIlwraith 356–57), North African (Frobenius 129–33), and Balto-Slavic (Kurrik 108–11, Terseglav et al. 118–19) mythology and folklore, which testifies that Pygmalionism runs deeply in human character. Motif-indexes of folk literature, such as the famous Stith Thompson's index, connect some of the above-mentioned tales with traditions from all over the world (e.g. under T11.2.1.1, the motif of a youth who makes a statue of a girl and seeks a girl like the statue is traced in a Buddhist, Chinese (in Chavannes), and Icelandic folklore). Although many tales from Chavannes's collection are indexed, The Painter and the Mechanical Maiden has not yet been put into relation to other tales of near and far—which is one of the objectives of this particular study.
Conclusion
The discussion here has focused on two facets of the Silk Road tale The Painter and the Mechanical Maiden: its philology, and its various parallels with Western reflections of the same motif. In the first section of this article, I provided a necessary overview of the context of the tale and contributed new insights to its interpretation. The second part identified and expanded on a basis for the inclusion of the Silk Road tale within the Pygmalion paradigm. A more detailed analysis of these tales is warranted but lies outside the scope of the present article. Connecting texts with a Pygmalion-like motif broadens the scope of the study of animated human-like figures from the predominant Western literary tradition to literary sources emanating from the rest of the world. Beyond this effort, further studies of the Silk Road tales could explore the circulation of tales from the trade routes in the East (such as Buddhist jātakas or non-Buddhist Pañcatantra collection) to their Western counterparts (Aesop's Fables via, for example, the Thousand-and-One-Nights).