Parkinson 1909 as a Source for Yorùbá Folktale

Five tales in this section come from a single journal article: John Parkinson, "Yoruba Folk-Lore," published in the Journal of the African Society in 1909 1. It is a short collection, it is in the public domain, and it has one quality that makes it more valuable per page than Ellis: Parkinson states where his tales came from and how they reached his page.

The provenance statement

The article opens with one sentence of method, and it is the reason this source is worth its weight:

The following tales, told to me for the most part by natives of Oyo, have been left as far as possible in the phrases of the interpreter, himself a Yoruba 1.

Three facts are declared here that Ellis never declares anywhere in his book: a region (Ọ̀yọ́), the existence of an interpreter, and that interpreter's identity as a Yorùbá speaker. A fourth thing is implied and matters just as much, that Parkinson knew his English was at one remove from the telling and chose not to smooth it into literary prose.

The consequence is visible on the page. Parkinson's English is often awkward in a specific way, with abrupt apposition and unidiomatic word order, and the awkwardness is evidence rather than incompetence. When Tela's spell reads "This rock must open because the owner of the house has come! I 'Prince Tela,'" the strain in the English is the closest thing the corpus has to the shape of the Yorùbá underneath it. The alternative, a collector who renders everything into fluent Victorian English, destroys that information silently.

Against this: the interpreter is not named, no individual teller is named, no dates are given, and no Yorùbá text is printed for any tale except one untranslated chorus.

What the collection contains

Parkinson prints twelve numbered items. Tales VIII to XI are the tortoise tales used in this section 1:

• VIII. How the tortoise helped the animals. alo-the-tortoise-helps-the-animals
• IX. The story of a tortoise and a man named Tela. alo-the-tortoise-and-tela
• X. The story of a dog and a tortoise. alo-the-dog-and-the-tortoise
• XI. A story of the pig and the tortoise. alo-the-pig-and-the-tortoise

Tale I, a hunter and an ape with sixteen tails, is treated in alo-the-hunter-and-the-ape-with-sixteen-tails. Tale XII is an Ifá narrative rather than an àlọ́.

The grouping note

Between tales, Parkinson inserts an editorial line of his own:

Stories VIII., IX., and X. tell how the tortoise got the marks on his back 1.

This is the single most useful sentence in the article and it is discussed at length in alo-why-the-tortoise-shell-is-marked. One collector, one region, one publication, three consecutive and mutually incompatible explanations of the same physical fact, printed without any attempt to reconcile them. It is direct evidence that Yorùbá aetiological tales were not understood as competing truth claims.

The one surviving Yorùbá text

Parkinson preserves exactly one line of Yorùbá across his tortoise tales, the chorus sung by the fishes, snails and crabs in Tale VIII:

Ti barua, ti barua, o, i gi, ti barua, ti barua 1.

He supplies no translation and no gloss. It is printed in 1909 orthography, with no tone marks and no subdots, and the corpus reproduces it in that state and declines to translate or reconstruct it, for the reasons set out in alo-the-tortoise-helps-the-animals. Without tone marks the line is genuinely ambiguous and a supplied reading would be invention.

Reading the editor's footnotes

The article carries extensive footnotes signed "Ed.," and they are a different kind of material from Parkinson's text. The editor systematically compares each Yorùbá tale to analogues elsewhere in Africa: Hottentot, Ronga, Basuto, Bemba, Tete, Chinamwanga, Nyanja, Malinke, Akra, Nago. The footnote to Tale VIII alone names Bain's "The Story of a Dam," Junod, Jacottet's "Le Chacal et la Source," and Monteil's Contes Soudanais 1.

Two cautions apply.

This is diffusionist comparative folklore in its period idiom. The impulse to locate every African tale as a variant of a pan-African type was the dominant scholarly mode of the moment, and it tends to treat the local telling as an instance of something else rather than as itself. The corpus records these comparisons as the editor's apparatus and does not treat a Bantu analogue as a variant of a Yorùbá tale.

The citations are genuinely useful anyway. The footnote identifying Bouche's Étude sur la langue Nago (Bar-le-Duc, 1880) as containing the yam-rock tale with a lizard is how this corpus knows that tale type has three independent nineteenth-century attestations rather than two (alo-ijapa-and-the-rock-of-yams). Neither Bouche nor Zimmermann has been obtained by this compiler, and both are reported strictly as the editor summarises them.

One editorial footnote is worth quoting for a different reason. Commenting on the tortoise's wife referring to the dog as "the man you sent," the editor observes that this is "due to the habit of personification in these beast-fables, which apparently makes the narrator forget that he is not dealing with human beings" 1. The judgement that the narrator has forgotten something is the editor's, and it is probably wrong: owe-tortoise-ijapa-trickster documents that these animals are anthropomorphic personages with names, households, in-laws and secret-society memberships, so calling the dog a man is not a lapse. The corpus records this as an instance of a period reader misreading a convention.

Parkinson himself

Parkinson is identified on the article as M.A., F.G.S., formerly Principal of the Mineral Survey of Southern Nigeria. He was a geologist working in the colonial administration, not an ethnographer or a linguist, and the collection is a byproduct of colonial employment in the same way Ellis's is. The corpus labels it as an early-twentieth-century colonial-period source for that reason.

What distinguishes it from Ellis is not the author's position but his method: he told the reader where the tales came from, kept the interpreter's phrasing, printed the one Yorùbá line he had without pretending to understand it, and grouped his aetiologies without harmonising them. Those are four decisions a modern corpus can work with.