Is Deontic Logic Possible?

Abstract

This thesis is an inquiry into the philosophical foundations of deontic logic. Since its inception modern deontic logic has been a very controversial subject matter for two reasons: (i) it has been plagued by a spectrum of paradoxes, and more importantly, (ii) its very foundation is tottering due to so-called Jrgensen's Dilemma. Many philosophers think that genuine ought-sentences lack truth-values. Most philosophers conceive logic as dealing exclusively with truth-bearers. Thus a nasty dilemma arises: is a logic of genuine ought-sentences possible at all? Surprisingly, this question has been eschewed by most deontic logicians who have concentrated their efforts on formal problems. We attempt to solve, or at least clarify, the questions arising from the aforementioned fundamental difficulty. Chapter One explains the basic formal machinery of deontic logic and the paradoxes it leads to. We also argue for the importance of intuitive semantics as a prelude to Chapters 3 and 4. Chapter Two considers the problem of conditional obligation. Many logicians have argued that the Chisholm Paradox forces us to introduce a dyadic obligation operator. We present a model-theoretic argument to the effect that this new operator is a genuine primitive, i.e. its semantics cannot be expressed by any combination of a monadic operator and a conditional. We argue that temporal deontic logic is an improvement and that its underlying logic of time should be classical in order to have sufficient expressive power. Concerning our main question, this chapter shows that from the logician's point of view deontic logic is no different from other systems of philosophical logic. Chapter Three confronts the DLI-argument (deontic logic is impossible) and considers various responses by non-cognitivists, notably von Wright, and cognitivists like Feldman. Von Wright's so-called Final View turns out to be the best available attempt to reconcile non-cognitivism with the existence of deontic logic. However, Feldman's cognitivism offers an attractive way out of Jørgensen's Dilemma by way of axiological semantics. Chapter Four presents our case against von Wright in detail (Section 4.1) but its main purpose is to elaborate, examine and defend the implications of Feldman's approach. Against value constructivism we argue that objective values are needed (Section 4.2). Von Wright's norm-authority model, which must take general moral principles as sources and justification of particular obligations, runs into deep trouble with instantiation. It is more plausible to account for them as generalisations from particular obligations (Section 4.3). We show that cognitivism a la Feldman does not violate Hume's Guillotine (Section 4.4). We argue against moral dilemmas and for the non-neutrality of deontic logic (Section 4.5). This means that deontic logic is inextricably tied to substantive ethics. Hence, though actual, deontic logic will remain controversial.