At Aristotle, Metaphysics E.1, 1026a14, Schwegler’s conjectural emendation of the manuscript reading ἀχώριστα to χωριστά has been widely adopted. The objects of physical science are therefore here ‘separate’, or ‘independently existent’. By contrast, the manuscripts make them ‘not separate’, construed by earlier commentators as dependent on matter. In this paper, I offer a new defense of the manuscript reading. I review past defenses based on the internal consistency of the chapter, explore where they have left supporters of the emendation unpersuaded, and attempt to strengthen their appeal. I challenge Schwegler’s central case, developed by Ross and others, that the construction μὲν ἀλλ’ οὐκ demands an implausible ‘logical antithesis’ between inseparability and mobility. This is arguably the fundamental obstacle to the manuscript reading, and counterexamples have not to date convinced emenders. I offer a new, systematic review of Aristotle’s use of the phrase, including relevant cases from the Rhetoric, to show how its usual meaning in Aristotle supports the transmitted text; I also reply to possible objections. Finally, I explore the implications of this defense for the classification of the sciences in E.1.