Abstract: "I'll tell you what I'll do," said the smith. "I'll fix your sword for you tomorrow, if you tell me a story while I'm doing it." The speaker was an Irish storyteller in 1935, framing one story in another (O'Sullivan 75, 264). The moment recalls the Thousand and One Nights, where the story of "The Envier and the Envied" is enclosed in the larger story told by the second Kalandar (Burton 1: 113-39), and many stories are enclosed in others. It was quite traditional for the Irish storyteller, historically disconnected from Arabian tradition, to use a frame-story. Folktale scholars label it "Story-teller Interrupted by Woman" and number it AT 1376A*. The Thousand and One Nights shows the literary imitation of that orally invented device; it standardizes the movement from one story into the next. So too "in the Sanskrit Five Books [the Panchatantra] the tales are neatly bound together by multiple use of framing" (Edmonson 143). Frame-stories in such collections are frequent enough for scholars to designate several as standing alone and establish a genre (Thompson, Folktale 415; Blackburn 496). All this is common knowledge to students of the Thousand and One Nights. I draw my examples from two stratified societies. Ancient Ireland surrounded its kings with a cattle-owning aristocracy whose dependants were firmly kept in lower social grades. Ancient India invented a system of caste too well known to need description (Dumont). Such societies foster the habit of subordinating one plot to another, creating that affinity that Theodor W Adorno perceived, "between the formal configuration of the artwork and the structure of the social system" (Hohendahl 172). Stratified societies favor frame-stories. In literary collections, and the criticism they have provoked, framing becomes either a genre or "a narrative mechanism for the linkage of possibly unrelated tales" (Belcher 1). In oral performance, framing is one device among many. What is the mechanism but a formal stylization of people's habit of interrupting their discourses, of going to another level? Framing is more than a mechanism; it is a human habit (Goffman) and a cultural universal. Therefore it sustains a variety of critical approaches. Modes of criticism, which intend to make sense of what is said, themselves function as interpretive frames (Bauman and Briggs 231). I discuss half a dozen kinds of framing; I attach some modes of criticism to them; I draw some examples from islands of the Southwest Indian Ocean (Madagascar, Mauritius, Reunion, Comoros, and Seychelles).' Frame-Story as a Genre If each tale, like the Irish one 1 began with, is a thing, an autonomous whole, then a genre becomes a thing as well, and the frame-story is a genre (though not only that). Still, because it requires other genres to live on, this one is parasitic. Perhaps it isn't an oral genre all over Africa (Belcher), but African performers do link (heir pieces. For instance, they often tell trickster tales in clusters or chains, so neatly that when collectors reproduce the cluster in the translations they publish, a reader can deduce principles of sequencing (Fontoynont and Raomandahy 83-86). One principle is to alternate trickster's success with his defeat and lead the audience toward a sense of cosmic order (Paulme, "Quelques procedes"). The audience's memory supplies another sort of frame: their familiarity with trickster's predictable behavior. Afghanistan and Ireland show plenty of examples of oral framing (Mills 123), the latter perhaps under literary influence (Belcher 16-18). But orality knows them too. In the story I began with, the smith requires Cuchulainn to tell him a story whilst mending his sword. When the smiths wife violates Cuchulainn's interdiction against eavesdropping, Cuchulainn breaks off the story of his adventure at its most suspenseful point, where he was in serious danger from a giant. The framed story is incomplete; the smith, his helper, and his wife are punished by the curtailment of Cuchulainn's performance. …