Lanell M. Mason

This article seeks to make an ontological distinction, of ethical and theological importance, between artificial intelligence and human beings. It will be demonstrated that humans possess phenomenal consciousness and therefore possess selves, while artificial intelligence at most satisfies a functionalist account of consciousness and therefore does not possess a self. That human consciousness is phenomenal will be argued for by examining how subjects learn and apply abstract concepts. Moreover, it will be argued that phenomenal experience entails a self, which is not emergent. Because humans have selves and AI systems do not, AI is not identical to human beings.

Gilbert Harman, in a well-known thought experiment, evokes the intuition that moral value can be perceptually seen. However, Harman dismisses the intuition, contending that moral concepts and judgments are the products of agent psychology and do not map onto mind-independent objects. Robert Audi, attempting to account for moral perception himself, fails to meet Harman’s challenge since his own ontological commitments do not allow for objects that moral concepts can map onto. This paper will offer an alternate theory of moral perception that maps moral concepts onto mind-independent entities, thereby meeting Harman’s challenge. To accomplish this, I offer that moral properties are not supervenient but are relational properties which possess their own non-reducible phenomenology.

The purpose of this thesis is to provide a metaphysic for moral realism and moral perception. This thesis is in two parts. The first is concerned with basic ontology. I begin in chapter 1 with an analysis of causation, demonstrating that substance theory is superior to Humeanism at accounting for our observations; thus I defend a substance ontology. In chapter 2, I address human agency, demonstrating that reasons internalism does not allow for incompatibilist freedom; hence, I affirm reasons are states of affairs. I move on to qualia in chapter 3, demonstrating that arguments both for and against qualia realism involve question begging. In response, I argue that qualia realism is just as reasonable as anti-realism, and affirm realism. Given this, I address the mind-body problem in chapter 4, demonstrating the failure of physicalists to account for mental causation. I conclude that panpsychism, idealism, and substance dualism are equal reasonable alternatives: each allows for intrinsic properties, which are integral to substance ontology. The second part builds a moral ontology, along with an account of moral perceptual knowledge, compatible with the above basic ontology. In chapter 5, I reject moral properties as monadic since this would bar mapping moral concepts onto mind-independent reality. Instead, I argue that moral properties are relations. More specifically, in chapter 6, I make the case that they are features of relations, relations themselves being the total complex states of affairs between one or more substances. This allows moral properties to have their own nonreducible phenomenal quality. To account for moral normativity, I argue that moral features are actually a class of causal features, and briefly argue that final causes determine the normative causal trajectory of all substances. In chapter 7, I conclude this thesis with addressing possible objections to my theory of perception and perceptual knowledge.