Bryn Mawr Classical Review 2001.03.23
Graham Anderson, Fairytale in the Ancient World. London/New York: Routledge, 2000. Pp. 240. ISBN 0-415-23702-5. $80.00. ISBN 0-415-23703-3. $27.99.
Reviewed by Christine Goldberg, Los Angeles
Word count: 2081 words
The subject of this book is neatly summed up in the blurb on the back cover: "Did the familiar children's fairytales of today exist in the ancient world?" The answer, according to the author, is Yes. The early chapters are devoted to children's favorites (Cinderella, Snow White, Beauty and the Beast, Little Red Riding Hood) and to less-known, more adult-oriented folktale complexes (the Magic Flight, the "innocent slandered maid," Bluebeard, and the Sorcerer's Apprentice, plus others). Then come Homeric tales, The Cyclops, which is the earliest of a large number of examples of its type, and Ares and Aphrodite, for which Anderson gives an ancient Egyptian analog. Finally there are moral tales (The Pied Piper, The Three Wishes, and The Singing Bone). At the end come chapters on romance and on the relation of ancient tales to their society. The fact that the book begins with Cinderella rather than with Cupid and Psyche establishes its direction.
The author begins his discussion of each tale by summarizing the modern tale type from a convenient source such as the Grimms' version. He enhances this summary with additional material in the form of details from or summaries of other modern versions of the tale, often from the Mediterranean region, sometimes from an early modern source (sixteenth to eighteenth centuries). Then comes the ancient evidence, in the form of summaries, generally beginning with the material most similar to the modern examples.
Whether the ancient evidence adduced for each tale is convincing is something readers must judge for themselves. The evidence is certainly interesting, and relevant in the sense that most of it contains some of the same motifs or themes found in the corresponding folktales. However, I am for the most part not convinced that the stories proposed as analogs represent the tale type in question.
Classicists are used to dealing with single texts, most often (at least until very recently) by identifiable authors, but folklorists require redundancy to establish traditionality. A tale type is an abstraction derived from (or, alternatively, embodied in) the sum of its texts (called variants). The tale types that correspond to Snow White, Rumpelstiltskin, etc., are manifested in dozens or hundreds of variants. In such a large corpus, strong similarities stand out against the confusion caused by the inevitable variations of detail and of structure. Even if a few of the texts seem marginal or perhaps not related to the others, the bulk of the material is convincing. Subtypes are defined within each tale type to clarify the nature of the type's variations. Particular detail motifs and stereotyped episodes are important in this analysis. While it may sometimes seem that two texts assigned to the same type number are very different from each other, the study that defined the type positions all of the variants in relation to each other (the original model for this was the stemmatic model of textual criticism). In practice, of course, some folktale studies are more thorough than others. While not all of the tale types have been subjected to detailed study, many of those discussed in this book have, and the rest, based as they are on Grimm tales, have at least been annotated by Bolte and Poli\vka,1 which means that a preliminary analysis of the tale type can easily be constructed.
The fewer texts there are, the more similar to each other they must be to justify calling them variants of a type. Single texts are useless for proving traditionality (thus Apuleius's Cupid and Psyche has generated controversy), and two or three versions, unless they resemble each other quite closely, are unreliable evidence for traditionality. In contrast to modern tales, which are often abundant, ancient sources seldom include even a few versions of a single tale type. Furthermore, ancient myths and stories are seldom as similar to modern ones as the modern ones are to each other. Anderson retells ancient stories to make them appear more similar to modern tales than they otherwise might. With regard to Snow White, he offers the story of Niobe, who refuses to acknowledge the superiority of Apollo and Artemis. Artemis shoots her children, and then "Zeus turns all the inhabitants of Thebes, including Chione [Snowgirl] to stone" (p. 47). To me, this represents a couple of components (motifs) of the modern tale, not a full-fledged variant. The romance of Apollonius of Tyre is also cited as a partial parallel to Snow White (pp. 147-8). The following myth is supposed to be Sumerian evidence for Jack and the Beanstalk: "The goddess Inanna plants a willow tree: it acquires a monstrous infestation by a snake, a Thunder-bird and a Demon-woman. Inanna calls on the hero Bilgames to free the tree, and it is cut down to provide precious objects, while the enemies are driven off" (p. 187). While this is certainly an interesting story, in my opinion it is not Jack and the Beanstalk, nor is it a close relative thereof. An early tale mentioned in connection with Aladdin seems perhaps more compelling: "Gyges...is a shepherd serving a royal master; he descends underground after a landslip and recovers a ring from the body of a dead giant inside the doors of the belly of a bronze horse. The ring confers powers of invisibility. He sleeps with the queen and kills the king, whom he supplants" (p. 107, from Plato, Rep. 359C-360B).
The author assembles composite ancient tales using details drawn from several different versions of the same myth or story. While on the one hand this might seem to be a reasonable thing to do, on the other, such composites are naturally designed to best support the author's argument. Furthermore, summaries even of a single text may leave out parts of the story that do not fit the model. For example, a chart on p. 40 compares a Sumerian myth, a seventeenth-century tale, and a modern folktale. The full text of the modern tale includes a section in which the heroine repeatedly descends to an underground garden. This detail, which perhaps seemed not to contribute to the Cinderella parallel, is not mentioned either in the text or in the chart.2
Anderson is aware of the tools of modern folktale scholarship,3 and even includes a convenient index of tale type numbers (pp. 228-229). However, he credits Mediterranean tale types with more individuality and more stability than I believe is warranted. In the case of the chart on p. 40, the first six of its ten rows describe an episode that belongs to several tale types. Here it is supposed to identify AT 510A, Cinderella, but it also appears (more reliably than it does in AT 510A) in AT 432, The Prince as Bird, AT 425 (especially in subtype C, Beauty and the Beast), and AT 894, The Stone of Patience.4 An episode that is this promiscuous in modern tradition is poor evidence for the existence of any particular tale type in the ancient world. The usual interpretation of such a situation is that the promiscuous episode was probably traditional before any or most of the tale types to which it was later appended became stabilized. Moreover, just because a similar story can be assembled from several ancient Sumerian fragments does not mean that the fabricated story itself was Sumerian. Humorous tales and thriller-like legends are also included. For example, King Midas's wish for everything he touches to turn to gold is named in connection with The Three Wishes, where a foolish man has to use his last wish to disconnect a sausage from his nose (p. 138). Soul-lamps parodied by Lucian (and now in the tale of Godfather Death) constitute a deeply symbolic motif around which narratives are built (p. 115-6). Some robber legends (pp. 100-102) in modern tradition form types but others are too fluid to categorize that way; this is surely a sub-genre for which it would be fruitful to consider ancient accounts systematically along with modern ones. The author admits that the romances he discusses conform to fairy tales "with varying degrees of closeness," and the nature of this correspondence between genres also warrants systematic investigation.
Even if most of Anderson's material should be considered components (motifs and themes) rather than full folktale variants, it is helpful to have it laid out. In the past, two approaches to the problem of ancient analogs to modern folktales have yielded trustworthy results. J. G. Frazer's extensive folklore commentaries on, for example, the Fasti and the Biblioteca, began from ancient material and listed modern analogs. In contrast, H. J. Rose, in his Handbook of Greek Mythology (1928, chapter 10), worked back from a nineteenth-century list of modern tale types. Either way, the results of this endeavor have been consistent for nearly 200 years. A good number of short, generally single-episode, tales (not what we usually think of as fairy tales) are indeed represented in ancient as well as modern sources. These include fables, where literary tradition is definitely a significant factor, as well as oral tales. The situation is different for complex (multi-episode) tales. It is not uncommon to find ancient examples of their component motifs or episodes, but whole tales are rare. This work still needs to be updated at intervals as more ancient material becomes available. In particular, now that ancient Near Eastern material is being covered in Classical mythology textbooks, more, older analogs are discovered. Conversely, folktales from Africa and from western Asia sometimes correspond to Classical material better than do tales from Europe. A well-organized, comprehensive, list of ancient parallels to modern tales would be a tremendous help to the understanding of the history of folktales.
Anderson is not interested in the history of scholarship regarding parallels between classical sources and modern folktales. Readers interested in the history of the identification of these correspondences must look elsewhere.5 Occasionally, Anderson mentions in connection with an ancient tale that no one seems to have noted it as an analog to the modern tale, but he rarely credits earlier scholars for their recognition of such analogs. This makes it difficult to tell how much of what is here is new. Nor is it easy to tell how much of what is already known is not here. For example, Anderson does not include much about The Dragon-Slayer (AT 300), which was annotated extensively by Hartland in The Legend of Perseus.6 One of the difficulties with interdisciplinary work is that it is not easy to teach yourself the history of an academic discipline. Many of the most thorough folktale monographs were originally Ph.D. dissertations, so comparative folktale scholarship is often hyper-conscious of its origins. In spite of our wish that truth once revealed is obvious and needs no rhetoric, Anderson's arguments might have seemed stronger if he had named earlier authorities who concurred with his judgements.
If, in the long run, the answer to Anderson's question really is Yes, this will alter our conception of the history of these tales. The present belief is that, after a very few early examples, the corpus of modern magical fairy tales begins to appear in Europe only in the twelfth century.7 And even as late as Straparola's (1550-53) and Basile's (1634-6) collections, when the genre of the magic tale is clearly established, episodes often combine in different ways from what we see in the tales of Perrault (1694-1697) and the Grimms (1812-1815 and later). Some difference is attributable to the fact that the types described in the Aarne-Thompson index fit northern Europe much better than they fit southern Europe (the revision which is currently being prepared of the index may help correct this bias). It is quite possible that more material from ancient sources will push back the assumed age of a number of tale types. However, if Anderson's idea of what constitutes an example of a tale type, which is much looser than that used in folktale monographs, becomes accepted, studies even of modern tales will have to cope with many more, and much more varied, texts. Tightly-defined tale types permit folktale studies to be rigorous and (relatively) objective; looser definitions of tale types would make folktale study much more subjective.
It is greatly to be hoped that this book will stimulate more research on this interesting subject. The back cover says that Anderson is editing an anthology of ancient fairy tales. Complete texts will make it easier for readers to judge whether they represent full analogs to particular modern fairy tales, or are more properly early evidence of a similar artistic impulse.
1. Johannes Bolte and Georg Poli/vka, Anmerkungen zu den Kinder- und Hausma+rchen der Bru+der Grimm. 5 vols. Leipzig 1913-1932.
2. No reference is given for the Sicilian tale. It is from Guiseppe Pitre\, Fiabe, novelle, e racconti popolari siciliani. 4 vols. Palermo 1982 (first published 1875), 1: 368-380.
3. Bolte and Poli/vka, Anmerkungen, in note 1. Antti Aarne and Stith Thompson, The Types of the Folktale. FFC 184. Helsinki 1961 (the abbreviation AT refers to this work). Enzyklopa+die des Ma+rchens, ed. Kurt Ranke et al. 10 vols. to date. Go+ttingen 1977-. In the entry "Erza+hlforschung," tale type numbers are listed along with the title under which the entry can be found. These entries update and expand upon the summaries and references in The Types of the Folktale.
3. Marion Roalfe Cox, Cinderella, London 1893, 480. Enzyklopa+die des Ma+rchens s.v. Geduldstein, Prinz als Vogel (in press). Jan O+jvind Swahn, The Tale of Cupid and Psyche, Lund 1955, 212-218, code I 11.
4. Swahn, The Tale of Cupid and Psyche (in note 3 above), supersedes previous studies. Enzyklopa+die des Ma+rchens s.v. Amor und Psyche. Detlev Fehling, Amor und Psyche, Wiesbaden 1977.
5. William F. Hansen, "Greek Mythology and the Study of the Ancient Greek Oral Story," Journal of Folklore Research 20 (1983): 101-112. Idem., "Mythology and Folktale Typology: Chronicle of a Failed Revolution," Journal of Folklore Research 34 (1997): 275-280. The Grimms were of course interested in evidence for ancient folktales as well as in ancient parallels to modern tales. The collector J. G. von Hahn, in Griechische und albanesische Ma+rchen. 2 vols. Mu+nchen und Berlin 1918 (first published 1864), listed folktale themes with their occurrences in ancient myth, German saga, German folktales, and Balkan folktales (1: LXXIV-XCIV).
6. Edwin Sidney Hartland, The Legend of Perseus. 3 vols. London 1894-1896. The Dragon-Slayer, AT 300, is embedded in AT 303, The Two Brothers. See Kurt Ranke, Die zwei Bru+der. FFC 114. Helsinki 1934. The ancient Egyptian tale of Bata and Anu, also called The Two Brothers, is generally considered not to belong to this tale type (contrary to Anderson's pp. 183-5, 240), but rather to AT 318. The right-hand column of the chart on p. 184 is another example of a composite designed to best fit its model.
7. Enzyklopa+die des Ma+rchens s.v. Altersbestimmung des Ma+rchens.