According to ideal observer theory, moral terms and concepts are to be understood as the preferences of a hypothetical ideal observer, that is, a person (or other sentient being) who is all-knowing with respect to non-ethical facts, all-perceiving, and disinterested or impartial, dispassionate, and consistent. The most thorough proponent of this theory was Roderick Firth, who argued for it in the article “Ethical Absolutism and the Ideal Observer.”

While it may sound banal, this theory is refuted by the simple recognition that there is no evidence that there exists any being that is all-knowing with respect to non-ethical facts, all-perceiving, and disinterested or any being that is impartial, dispassionate, and consistent. Indeed, there would be no way of proving that any sentient being believed to have any of these attributes actually had them, though it is generally simple and straightforward to show that any given person with whom one interacts lacks these attributes (by, for example, asking them questions until they produce a wrong answer, asking another person about perceivable features of one’s immediate environment when that other person is not present and does not perceive the same things, etc.), and it is equally simple and straightforward to show that there is no reliable evidence for the existence of any sentient being that has any of these attributes, at least not all the time (even if it is possible to posit the existence of such a being or beings hypothetically).

As there is no evidence for the existence of any ideal observers, it is not possible to coherently speak of what an ideal observer would judge to be good and bad or evil, and thus it is not possible to make sense of such terms by reference to any so-called ideal observer. Put another way, unless one is oneself an ideal observer, one cannot understand how an ideal observer would experience the world what an ideal observer would mean when he or she used terms like good and bad or evil. There are no reliable experiential grounds for believing any person in history has been an ideal observer and there is no experiential reason to believe any person could accurately imagine what it would be like to be an ideal observer. And it makes no sense to believe that moral terms should be understood the way a non-existent, purely hypothetical, and not-at-all-possible-for-a-human-being-to-approximate-based-on-experience ideal observer would understand them.