Lawlor argues that social psychological studies present a challenge to the authorship account of first-person authority. Taking the deliberative stance does not guarantee that self- ascriptions are authoritative, for self-ascriptions might be based on elusive reasons and thus lack agential authority (i.e. they are no guide to the subject's future conduct). I argue that Lawlor's challenge is not successful. I claim that we can make sense of the nature and importance of agential authority only within the framework of the authorship account. Agential authority is part of the regulative ideal of the deliberative stance, but its lack does not undermine the first-person authority of self-ascriptions, since first-person authority is primarily a matter of deliberative authorship.