Jamshid J. Tehrani (2013),
The Phylogeny of Little Red Riding Hood,
PlosOne, November 13, 2013.
Patrice Lajoye, Julien d'Huy* and Jean-Loïc Le Quellec**
*Centre d'Etudes des Mondes Africains, CNRS, UMR 8171.
**Centre d'Etudes des Mondes Africains, CNRS, UMR 8171 - Honorary Fellow, School of Geography, Archaeology and Environmental Studies - University of the Witwatersrand, Johannesburg.
A few weeks ago, the magazine PlosOne has published an interesting article, "The Phylogeny of Little Red Riding Hood ", by Jamshid J. Tehrani , University of Durham, on the study of folktale -type AT 333 – Le Petit chaperon rouge, The Glutton , Rotkäppchen - and AT 123 – Le Loup, la chèvre et ses chevreaux, The Wolf and the Kids. Indeed we have long sensed that both AT 333 and AT 123 have at least interfered or that one could derive from the other. The author had the good sense to apply methods borrowed from phylogenetics on these folktales. Good idea, because this is a way of approaching sources as neutral as possible, without preconceived ideas. By applying this method to both tale types, the author concluded that AT 333 and AT 123 are indeed two different types, and that African versions belong to AT 123, but that Far Eastern versions are a mixture of both.
We support this type of analysis: our readers were able to read an article by Julien d'Huy, based on a similar methodology, which was the first paper we published. But we need to emphasise that, regarding statistical analyses, it is required that the corpus be as broad and representative as possible. This is the weak point in Jamshid J. Tehrani's study.
First of all, he includes, uncritically, the Latin poem of Egbert of Liège (11th century), in the corpus of AT 333, while this is still debated1.
But the most important criticism that we can emit concerns the weakness of the corpus itself. The entire study covers only 58 variants (Egbert included). And for a good reason:« Data for the study were drawn from 58 variants of ATU 333/123 available in English translation from 33 populations ».
Of course, we are not asking anyone to be a new Jeremiah Curtin or a new Georges Dumézil, but if there is one field in which the knowledge of languages other than English is required, it is folkloristics. Two major languages are to be taken into account: French and German. Thus, significant gaps are found in the bibliography of the author, who works on less than one tenth of the AT 333 versions mentioned by Stith Thompson, which was far from complete. We do not find, for example, Marianne Rumpf thesis on AT 333, defended in 1951 in Göttingen, published in 19892. An article by Gottfried Henssen and another one - this time on AT 123 – by Marie-Louise Tenèze are also missing3.
Most importantly, the indispensable descriptive catalog of the French folktales by Paul Delarue and Marie-Louise Tenèze, yet reissued in a single volume in 2002, is even not mentionned4.
One might think that the last remark reveals our chauvinism but the Delarue-Tenèze corpus mentions 32 versions of AT 333 and 88 versions of AT 123 (not to mention the medieval versions from the Isopets and from French speaking territories overseas, yet also inventoried).
With only the French corpus, Jamshid J. Tehrani could have tripled his corpus. What about the other potential national corpora? The author mentions, for example, one Russian version of AT 123, from second hand and not localized (when you know the size of Russia, this is important). A simple overview of the collection of Alexander Afanasyev allows to discover two versions5, but Afanasyev's collection is not unique: there are many more! In the case of Portugal, the author mentions "Vasconcellos, L. (nd) "O Chapelinho Encarnado", without going further in search of his source, which is José Leite de Vasconcelos, Contos Populares e Lendas, published in 1963. However it has been shown that this version is directly demarcated from Perrault and Grimm6. And can we say, again, that this version is unique? Certainly not, as another one was published in 1984 by Zófimo Consiglieri Pedroso.
To return to the French case, the study of Paul Delarue is important because it allows precisely to nuance many aspects, particularly with regard to possible literary influences - which Jamshid J. Tehrani ignores. Regarding AT 333, Delarue mentions 20 versions which owe nothing to Perrault (all located in the basin of the Loire and north of the Alps, they are similar to Tyrolean and Italian versions), 12 mixed ones and 2 versions that owe everything to Perrault. He also notes that the Grimm version is the Perrault’s one, with an end taken from the German form... of AT 123!
Also in France, according to Marie-Louise Tenèze, 9 versions of AT 123 have been influenced by Grimm, while 4 others derive from an “image d'Épinal”. And even without that, one should take into account the rich tradition of medieval Isopets, which certainly derived from Aesop's fables - and were popularized in Europe.
The situation is even worse for the mentions of AT 333/123 tales in Asiatic sources: eighteen Korean versions were cited by In Huak Choi in 1979, seventy-three Japanese versions were summarised in 1971 by Hiruko Ikeda, and 241 Chinese version were analysed by Wolfram Eberhard in 1970! So there is justifiable doubt as to whether or not Tehrani’s sample is really as “representative” of the geographic distribution of these tales as he says.
The extension of the AT333 and AT123 corpus also allows some nuance about Jamshid J. Tehrani work, who mentions the location of the action as a discriminant criterion between the two tale types: in one, AT 333, the action happens outside, at the grandmother’s house, while in the other, AT 123, the action happens inside, at the goat and kids’ home. However, there are four French versions of AT 333 (22, 24, 25 and 28 from the Delarue and Tenèze corpus) in which the action takes place at the girl's home. These versions are all located in the Velay and the Dauphiné, where they probably form a particular local tradition. The mother is eaten by the wolf in her house, and the girl discovers the tragedy when she returns home. What can we think about these versions? Are they contaminated by AT123, or do they represent an unrecognized, older state of AT 333?
Conversely, one of the two Russian versions of AT 123 shows a homeless goat, ready to whelp. It passes through a forest and finds there a hut similar to Baba Yaga’s hut, and occupies it. But that is not her own home. These are all nuances that do not appear in the work of Jamshid J. Tehrani.
Many other gaps and lacunae appear in the analysis grid of the author. Thus, he lists a "Villain offers grandmother's flesh to the victim" motif, but he forgets "Villain offers grandmother's blood to the victim." Similarly he's got "Guardian Gives remains of child to the villain to eat" but he doesn't take into account "Villain Gives remains of grandmother to the child to eat."
The author's unfamiliarity with folkloristics appears clearly in the motif he defines as "Victim flees through the woods, and uses the help of the river, mountain, etc. to obstruct the villain's pursuit "... whereas it is the well known "Magic flight" motif (Stith Thompson's D670).
Among the problems associated with Tehrani’s characterisation of the motifs, we note that the No. 21 is: "The victim and villain take separate routes":  missing  take the path of needles and pins  the villain takes the shortcut. But as far as  is concerned, the choice can be between needles and pins, stones and thorns, stones and pins, road and fields, or simply between two different paths. The list given for the species of the villain (No. 7:  fox  ogre  wolf  tiger/leopard  lion  bush beast  hyena  bear  alligator  crow) is far from being complete, as he can also be a man, a werewolf or the Devil. And several other motifs could be added, like “Villain comes to victim’s house”, “Victim tries to hide”, “Villain enters into victim's house”, “Victim wanders through the path while villain runs”, etc. This is important as the method followed by the author is more effective when the number of characters is important.
Some of the statistical tools used by Jamshid J. Tehrani themselves are also questionable. Thus, when using NeigbhorNet, he forgets to mention that the versions may only appear at the network edge, which prevents to see some more subtle links. Sometimes, NeighborNet also cannot place such a given version in an intermediate position between many others around it ; that can biase its position7.
In the autor’s database we replaced “-” by “?”, we transformed the base in Jaccard (taking into account the similarities and not the differences) and we launched a « principal coordinates analysis » (PCoA, transformation exponent : c = 2) with Past 3.0 (Hammer et al. 2001). PCoA is a well-know method that tries to find the main axes through a matrix. It starts with the matrix and assigns for each item a location in a low-dimensional (here two) Euclidian space. Then it calculates many eigenvalues and eigenvectors. Coordinate1 and Coordinate 2, which are the first and second principal coordinates, accounts for as much of the variability in the data as possible (here, proportion of explained variance: C1 : 30,629% ; C2 : 13,396% of the data). Indian, and not European, versions seem placed at the center (fig. 1).
When we placed in Easting and Northing the variable that explains the most distintive features (30.62%), we also obtained a similar result where the European versions are no longer in the middle (fig.2).
This does not mean of course that the method used by the author is to be rejected: quite the contrary. But it must be applied with caution and only after establishing and criticizing a corpus as complete as possible.
Patrice Lajoye, Julien d'Huy, Jean-Loïc Le Quellec
Alexander Afanasyev, Народные русские сказки, Moscow, Nauka, 3 volumes, 1984-1985.
Jacques Berlioz, « La Petite robe rouge », in Jacques Berlioz, Claude Brémond et Catherine Velay- Vallantin, Formes médiévales du conte merveilleux, 1989, Paris, Stock.
In-Hak, Choi, A Type Index of Korean Folktales, 1979, Seoul, Myong Ji University Publications, xii-353 p.
Paul Delarue, « Les contes de Perrault et la tradition populaire », Bulletin folklorique d'Île-de-France, 1951.
Paul Delarue et Marie-Louise Tenèze, Le Conte populaire français. Catalogue raisonné des versions de France, 2002, Paris, Maisonneuve et Larose.
Wolfram Eberhard, Studies in Taiwanese Folktales, 1970, Taipei, Asian Folklore and Social Life Monographs, 193 p.
O. Hammer, D. A. T. Harper and P. D. Ryan, PAST: Paleontological statistics software package for education and data analysis, Palaeontologia Electronica, 2001, 4(1), 9 p. http://palaeo-electronica.org/2001_1/past/issue1_01.htm
Gottfried Henssen, « Deutsche Schreckmärchen und ihre europäischen Anverwandten », Zeitschrift für Volkskunde, 51, 1953, p. 84-97.
Paul Heggarty, Warren Maguire and April McMahon, « Splits or waves? Trees or webs? How divergence measures and network analysis can unravel language histories », Philosophical Transactions of the Royal Society B, 2010, 365, p. 3829-3843.
Hiroko Ikeda, A Type and Motif Index of Japanese Folk-Literature, 1971, Helsinki, Suomalainen Tiedeakatemia (Folklore Fellows Communications (vol. 209), 375 p.
Zófimo Consiglieri Pedroso, Contos Popolares Portugeses. Revista e aumentada, “Outras Obras”, 1984, Lisboa, Vega [1ª ed.: 1910] (297-301)
Marianne Rumpf, Rotkäppchen, eine vergleichende Märchenuntersuchung, (1951) 1989, Peter Lang.
Marie-Louise Tenèze, « Aperçu sur les contes d'animaux les plus fréquemment attestés dans le répertoire français », IV International Congress for Folk Narrative Research in Athens (1964), Lectures and Reports, Athènes, 1965, p. 569-575.
José Leite de Vasconcelos, Contos Populares e Lendas, coordenação de A. S. Soromenho e P. C. Soromenho, I, 1963, Coimbra, Acta Universitatis Conimbrigensis.
Francisco Vaz da Silva, «Capuchinho vermelho em Portugal.», Estudos de Literatura Oral 1, 1995, p. 38-58.
1Delarue, 1951, p. 227 ; Berlioz, 1989.
2Rumpf, (1951) 1989.
4Delarue et Tenèze, 2002.
5Af 53/23a ; Af 53/ 23b.
6Vaz da Silva, 1995.
7Heggarty and al., 2010.