Socrates’ argument against the possibility of akrasia in the Protagoras has as one of its conclusions the claim that knowledge is our savior in life. But the precise nature of the knowledge in question is a topic of controversy. Most interpreters have taken a literal approach and assumed that Socrates is speaking in earnest when he claims that the relevant knowledge is a capacity to measure the various quantities of some monolithic value (be it pleasure or not) that are instantiated in competing practical alternatives. According to this interpretation, identifying the general nature of the human good is a relatively straightforward task; the problem that plagues most people is the local one of implementation: since our intuitive (i.e. non-technical) faculty for judging the quantity of good instantiated in a given action tends to provide inconsistent results due to unavoidable fluctuations in our perspective, we have need of some non-intuitive apparatus which will supply us with dependable consistent true judgments regarding how valuable each action is. And it knowledge, conceived as an art of value measurement, that fulfils this task.

The literalist interpretation has been challenged by scholars who claim that the argument in the Protagoras contains sufficient evidence for thinking that Socrates does not in fact take the error that underlies akratic behavior as merely local, and that, consequently, he does not subscribe to the idea that the knowledge which is sufficient to save our lives is an art of value measurement. These scholars claim instead that the true purpose of Socrates’ analysis of akrasia is to show that akratic behavior implies the lack of a correct “global” conception of the human good. But interpretations of this kind are at odds and, I shall argue, insufficiently clear, about how a proper global understanding of the human good will, according to the Protagoras, guarantee we do not succumb to acting in ways we recognize as bad for us.

In the present paper, I offer a new reading of Socrates’ argument within the general framework of the non-literal interpretation. I will begin by arguing that some features of the phenomenon of akratic behavior, as Socrates and the many conceive of it, have hitherto not been properly elucidated. Once the phenomenon has received an adequate description, I shall provide an account of the global error about value which lies at its heart. And I shall offer a corresponding account of what the relevant knowledge of the good consists in, and why it is incompatible with this error.