Machine Learning and Theological Traditions of Analogy

Progress in machine learning or artificial intelligence presents us with computer systems that exhibit properties at least in some way akin to aspects of human (or wider animal) cognition. In naming and thinking about these machine learning capacities, it would seem ill-advised either to suggest simple equivalence with human faculties, or complete disjunction. A rich seam of theological writing on the subject of analogy offers a middle way, explored here from the writings of Thomas Aquinas, Thomas de Vio Cajetan and Francisco Suarez. These authors offer a wide variety of accounts as to how an analogy might function, or be justified. We begin with accounts that place the emphasis on the ingenuity of the speaker, in applying a word beyond its usual domain, and move through to accounts where analogy is said to rest more objectively on a connection in reality, in this case between the human being and the computer. Such a connection may be direct, arising from the human as maker, or rest upon a joint relation to something more ultimate still. This, towards the conclusion of the article, raises questions about the origin, from a theological perspective, of all such faculties or capacities in God. We end with the observation that while much theological writing on scientific themes involves the theologian taking account of what the sciences reveal about the world, sometimes, as here, it is possible for the theologian to offer resources, often philosophical, in return.