Introduction: the sublime before and after Longinus Any search for the sublime in Greece and Rome has to begin with one work, the only treatise on the topic to have survived from antiquity, the socalled Peri hupsous. This hardly gives us a sure foothold on the problem. The treatise is riddled with unknowns, from its author, date, and title to its place of origin (Greece or Rome?) to the meaning of its core concept, which seemingly defies definition. For starters, next to nothing is known about the author, including his name. The primary Byzantine manuscript, a codex dated to the second half of the tenth century, attributes the work to "Dionysius Longinus" on the title page where the essay begins (Figure 1.1), while the table of contents that appears at the front of the codex assigns the title to "Dionysius or Longinus" (Figure 1.2). 1 Something has gone wrong here, but exactly what is harder to say. A slip was obviously made somewhere along the line, but the confusion goes deeper than quality control. Evidently the author's name was unknown. A copyist or scholar, puzzled by the problem and faced with so fine a work, must have assumed that it had to stem from one of the great critical names from the past, such as Dionysius of Halicarnassus, the Augustan literary critic and antiquarian, or Cassius Longinus, the third-century ce Neoplatonist, polymath, and critic. Neither guess is very compelling. While there are a few points of contact that might encourage the identification, the similarities are at best superficial, while other factors speak against either possibility, not least the distinctive critical styles of each of the writers in question. 2 None of this will have escaped the Byzantines to whom we owe the preservation of 1 Parisinus (P) 2036, the archetype codex of the remaining ten MSS. See Roberts (1899b) 3-4; Russell (1964) xxii-xxx; Häussler (1995) 154; Mazzucchi (1989); Mazzucchi (2010) xxxix-xliv. 2 Cf. the marginal note in P at Subl. 39.1: "NB: Dionysius wrote on composition" (202 v ; discussed by Russell (1964) xxiv n. 2), which would appear to be corroborating the identification. But Longinus' work of the same title was in two books (Subl. 39.1), while the study we have by Dionysius is in one. Cassius Longinus has not found favor, though the case has been restated most recently by Heath (1999) and Heath (2012) 11; 15-16. Kaibel (1899) lays out most of the counter-arguments against this latter's candidacy; see also Russell (1964) xxiv-xxv; Russell in Halliwell, et al. (1995) 146-7; Mazzucchi (2010) xxxiii.