A Theory of Human Dignity, #7–The Skinny Logic and the Fat Semantics of Moral Principles in Broadly Kantian Nonideal Dignitarian Moral Theory.
By Robert Hanna
TABLE OF CONTENTS
IV.1 The Skinny Logic and the Fat Semantics of Moral Principles in Broadly Kantian Nonideal Dignitarian Moral Theory
V. Some Hard Cases For Broadly Kantian Nonideal Dignitarian Moral Theory
VI. Enacting Human Dignity and The Mind-Body Politic
This installment contains section IV.1.
But you can also download or read a .pdf of the complete text of this essay HERE.
IV.1 The Skinny Logic and the Fat Semantics of Moral Principles in Broadly Kantian Nonideal Dignitarian Moral Theory
The background conceptions of logic and semantics that I’m using in this section and throughout this essay as a meta-ethical foundation, include
(i) Kant’s “pure general logic” (CPR A50–57/B74–79) (JL 11–20, 91–150), or what we would now think of as second-order intensional monadic logic, that is, classical truth-functional logic together with a restricted predicate logic employing quantification into and also over one place predicates only, and quantifying over individuals (= first-order monadic logic[i]), but also quantifying over the Kantian concepts or finegrained intensions expressed by one-place predicates (= second-order intensional monadic logic),[ii] and
(ii) Kant’s “transcendental logic” (CPR A55–57/B79–82), or what we would now think of as a finegrained intensional possible worlds semantics of propositions together with what I call Kantian modal dualism.
Now according to the robust semantics of Kantian modal dualism, in turn, there are two irreducibly and essentially different kinds of necessary truth:
(i) analytic necessity, which is a priori necessary truth in virtue of conceptual content, always taken together with some things in the world beyond conceptual content, although never in virtue of those worldly things, that is, the necessity that flows from concepts, and
(ii) synthetic necessity, which is a priori necessary truth in virtue of things in the world beyond conceptual content, that is, truth in virtue of pure intuitional and imaginational content representing the underlying non-empirical intrinsic spatiotemporal, causal-dynamic, and mathematical immanent structures of matter in the actual world, always taken together with some conceptual content, although never in virtue of conceptual content, that is, the necessity that flows from things in the world.[iii]
So, in other words, the background conceptions of logic and semantics I’m using as a dual logico-semantic meta-ethical foundation for broadly Kantian nonideal dignitarian moral theory jointly provide, at one and the same time, for a somewhat thinner and more minimalist conception of pure logic than standard classical logic in the Frege-Russell-Carnap-Tarski-Quine tradition (aka “elementary logic”) and also for a somewhat thicker and more robust conception of semantics than standard classical semantics in the mainstream Frege-Russell-Carnap-Tarski-Kripke tradition. I argued in detail and at length in Kant the Foundations of Analytic Philosophy and again in Cognition, Content, and the A Priori, that although Kant is almost universally criticized for having a skinnier logic and a fatter semantics than most logicians and semanticists in the mainstream Frege-Russell-Carnap-Tarski-Quine-Kripke tradition are prepared to accept, nevertheless, there are very good reasons to think that they are wrong, and Kant was right.[iv] As controversial as those claims are, in order not to overburden the present already long-ish essay, I will not attempt to re-argue those claims here and will simply assume the soundness of my earlier arguments.
Still, I do want to emphasize right from the outset that broadly Kantian nonideal dignitarian moral theory does indeed presuppose a special non-classical logic and also a special non-classical semantics, and also that if we take this skinnier logic and that fatter semantics explicitly into account, then our conception of broadly Kantian nonideal dignitarian moral theory will be significantly deepened and strengthened, as per The No-Foolish-Consistency Interpretation.
According to Kant in the Groundwork and also later in the Critique of Practical Reason (CPrR 5: 19–28), the Metaphysics of Morals (MM 6: 211–227), and Religion within the Boundaries of Mere Reason (Rel 6: 3–5, 20–50), and also according to broadly Kantian nonideal dignitarian moral theory, morality is grounded on a set of strictly and unconditionally universal a priori normative meta-principles which are categorically binding on all rational beings, and more specifically are categorically binding on all rational human animals, insofar as all rational human animals have
(i) a “will” (Wille), which is an innate psychological capacity for rational desiring, or practical justification in terms of either non-instrumental reasons (namely, the Categorical Imperative in its four or five analytically equivalent formulations) or instrumental reasons, and also
(ii) a “power of choice” (Willkür), which is an innate psychological capacity for effective desiring, or causally efficacious conscious motivation to choice and action.
Otherwise put, the Wille is a legislative practical capacity that generates, recognizes, and is more generally reasons-sensitive to principles and imperatives, whereas the Willkür is an executive practical capacity that enacts and implements principles and imperatives by means of reasons-sensitive conscious conations, drives, or impulses. Together, the faculties of Wille and Willkür jointly constitute a dual faculty for rational desire-based choice. In turn, the Wille has
(i) a higher proper part (pure practical reason, that is, a power for non-instrumental reasoning) that generates, recognizes, and reasons with categorical imperatives, and also
(ii) a lower proper part (what I’ll dub impure practical reason, that is, a power for instrumental reasoning) that generates, recognizes, and reasons with hypothetical imperatives.
So, to summarize, according to my broadly Kantian approach to cognition, volition, and practical reasoning, the overall structure of the human will or faculty of desire looks like this:
Human Will or Faculty of Desire (Begehrungsvermögen):
higher part = faculty of practical reason or will proper (Wille):
higher part = pure or non-instrumental reason
lower part = impure or instrumental reason
lower part = power of choice (Willkür)
Moreover, according to the broadly Kantian nonideal dignitarian theory of moral principles that I’m developing here, we need to distinguish very sharply between
(i) absolutely universal and objective moral meta-principles, which are strictly and unconditionally universal and objective a priori normative rules binding on all rational beings, including all rational human animals or real human persons,
(ii) first-order substantive ceteris paribus objective moral principles (aka “fairly universal and objective moral principles”), which tell us what we ought to do, other things being equal, and are binding on all rational human animals or real human persons in any set of circumstances, provided that certain favorable background conditions obtain, and finally
(iii) moral duties, which are first-order objective moral principles that are also agent-centered obligations.
Granting all that, then in given act-contexts, moral agents can find that other things really are equal. So moral duties are first-order substantive ceteris paribus objective moral principles with agent-centered application, under absolutely universal and objective moral meta-principles. A first-order substantive ceteris paribus objective moral principle is essentially the same as what Kant calls a “ground of obligation” (MM 6: 224). A ground of obligation is a morally sufficient reason for choosing-and-acting or for refraining, other things being equal. Similarly, to use an everyday analogy, your mother, father, or kindergarten teacher can tell you what you ought to do or not do, other things being equal, and s/he might be completely right. But because our actual natural and social world is a thoroughly nonideal world, other things really might not be equal in any given actual act-context; and, correspondingly, because the ceteris paribus condition therefore really might not be satisfied in that actual act-context, it does not automatically follow that you are obligated to do what your mother, father, or teacher rightly tells you that you ought to do — unless, in that act-context, things really are equal, and you yourself really can do it. Thus in order to be a moral duty, a first-order substantive ceteris paribus objective moral principle has to have adequate agent-centered force in an actual act-context, and this depends in part on the way the world and other people just contingently really happen to be, quite independently of the agent herself, as well as depending in part on the actual agent herself and her agential capacities in that actual act-context.
This, in turn, shows us how correctly to formulate the well-known Kantian principle that “ought implies can” (CPR A548/B576, A807/B835) (MM 6:380). As Robert Stern has pointed out, the Kantian version of “ought implies can” does not mean that “nothing can be right that we are incapable of achieving,” but instead means that “we cannot be obliged to do what is right unless we are capable of acting in that way.”[v] As we all know, there is significant contingent variability in how our basic shared agential capacities are actually realized in different real human persons, in different act-contexts. Thus moral duties obligate us to do what some moral principles tell us we ought to do, other things being equal — that is, leaving out contingent conditions in act-contexts. But if we reintroduce contingent conditions in act-contexts, then we might be morally obligated, although it is not necessarily the case that we will be morally obligated; for we do not have a duty in each actual act-context, but rather only in some actual act-contexts. I’ll come back to this point in sub-section IV.2 below.
Necessarily, every moral duty is also a first-order substantive ceteris paribus objective moral principle, but not every first-order substantive ceteris paribus objective moral principle is also a moral duty. This is because there can be real conflicts between first-order substantive ceteris paribus objective moral principles, even in cases in which an agent has one and only one moral duty:
A subject may have, in a rule he prescribes to himself, two grounds of obligation …, one or the other of which is not sufficient to put him under obligation, so that one of them is not a duty. (MM 6:224)
Indeed, while there can be real conflicts of first-order substantive ceteris paribus objective moral principles, there cannot be conflicts of moral duties, as a matter of analytic a priori necessity: “a collision of duties and obligations is inconceivable” (MM 6: 224). Thus the distinction between first-order substantive ceteris paribus objective moral principles and moral duties captures the essence of what Ross was driving at in his famous distinction in The Right and the Good between “prima facie duties” and “actual duties,” but without the strange consequence that a given moral principle can sort-of be my moral duty without its also really being my moral duty.
In The Right and the Good, Ross argues that we have rational, self-evident, non-inferential, infallible a priori intuitions about an irreducibly plural class of co-basic moral principles, the prima facie duties, which include the seven duties of “fidelity,” “reparation,” “gratitude,” “justice,” “beneficence,” “self-improvement,” and “non-maleficence.”[vi] These seven principles, purportedly, are knowable by any mature, reflective, rational human animal. Prima facie duties are sharply distinguished from actual duties, or duties proper, that (i) are the objectively real moral obligations binding on moral agents or persons in particular act-contexts, and (ii) are objectively determined by their being the moral principles that, in that act-context, have the greatest balance of prima facie rightness over prima facie wrongness, of all possible acts for that agent in that context, when the act is taken in “its whole nature.”[vii] At the same time, however, according to Ross, it’s not possible rationally to intuit, or authentically to know, actual duties — at best, it is possible to cognize actual duties with “right opinion,” and not sufficiently justified true belief,[viii] that is, essentially reliable justified true belief, or what elsewhere I call High-Bar justified true belief, whereby there is an intrinsic connection between the warranting evidence for belief, as delivered by our properly functioning cognitive capacities or mechanisms, and the truth.[ix]
In any case, Ross has three main goals in combining the classical theory of rational intuition with his theory of prima facie duties vs. actual duties. First, he wants to provide a secure, realistic, and a priori but also non-monistic foundation for moral theory. Second, he wants to accommodate the obvious empirical fact of conflicts of duties — moral contradictions or moral dilemmas — that seem to arise directly from the foundational fact of a plurality of basic duties together with our actual “human, all too human” existence in this thoroughly nonideal natural and social world. And third, he wants to incorporate some measure of commonsensical or real-world fallibilism about our moral judgments in particular contexts in this world. Ross’s moral intuitionism is thereby designed precisely in order to accommodate the thoroughly nonideal character of our moral lives. But his intuitionism is also philosophically notorious.
Correspondingly, here are the three classical critical objections to Ross. First, Ross’s postulation of a mysterious and “queer” (in John Mackie’s sense[x]) faculty for intuitively knowing the prima facie duties has no independent plausibility or empirical support whatsoever. Second, Ross’s infallibilism about moral intuition seems to fly in the face of the highly plausible thesis of fallibilism about a priori knowledge, as well as fallibilism about empirical knowledge. And third, the obvious empirical fact of widespread disagreement, even amongst mature, reflective rational human animals, about precisely which moral principles are true and which are false seems to undermine completely Ross’s claim that even some moral principles are known intuitively with self-evidence. In addition to these three classical worries, I also have a non-classical, fourth critical objection that in certain respects is similar to John Rawls’s main worry about Ross’s theory, to the effect that Ross cannot ultimately avoid a theory of the lexical ordering and weighting of the supposedly equally morally binding, lexically unordered, and unweighted prima facie duties.[xi] More precisely, my non-classical, Rawls-inspired critical objection can be posed as a dilemma:
Ross’s theory of prima facie duties, on the one hand, explicitly postulates an irreducible pluralism of basic moral principles; yet on the other hand, he implicitly presupposes a monistic deontological scale in explicating the advance from prima facie duties to actual duties — otherwise how could there be an objective determination of one moral principle’s being the one which, in a given context, expresses the greatest balance of prima facie rightness over prima facie wrongness, of all possible acts for that agent in that context, when the act is taken in “its whole nature”?
In Cognition, Content, and the A Priori, I respond to the three classical worries about Ross’s moral intuitionism, at least by implication, by developing a broadly Kantian theory of rational intuition that’s an extension of a broadly Kantian theory of mathematical intuition, and also includes moral intuition as a sub-case.[xii] In the present context, however, I want to respond directly only to the non-classical, fourth objection to Ross’s moral intuitionism by extending the notion of structuralism from mathematics to morality. Mathematical structuralism, as an explanatory metaphysical thesis in the philosophy of mathematics — defended, for example, by Stewart Shapiro,[xiii] and in another way by Charles Parsons[xiv] — says that mathematical entities (for example, numbers or sets) are not ontologically autonomous or substantially independent objects, but instead are, essentially, positions or roles in a mathematical structure, where a mathematical structure is a complete set of formal relations and operations that defines a mathematical system. What counts as an individual object of the system is thereby uniquely determined by the system as a whole. That is, any such individual object is identical to whatever possesses a specific set of intrinsic structural system-dependent properties. In a text quoted as one of the epigraphs for this section, it seems clear enough that Ross himself had a moral structuralist idea in mind:
The moral order expressed in these propositions is just as much a part of the fundamental nature of the universe (and, we may add, of any possible universe in which there are moral agents at all) as the spatial or numerical structure expressed in the axioms of geometry or arithmetic.[xv]
But he never systematically developed or elaborated that important thought. Nevertheless, Ross’s important thought can be unpacked and effectively deployed within the framework of broadly Kantian nonideal dignitarian moral theory . So, standing on Ross’s shoulders, here are the six basic ideas behind my nonideal broadly Kantian dignitarian version of moral structuralism.
First, there’s a three-levelled hierarchy of moral principles, not a “flat” or non-hierarchical set of moral principles, as is usually assumed to be the case. Second, moral principles are not ontologically autonomous, substantially independent, “atomic” semantic or normative objects, but instead are, essentially, positions or roles in a moral structure, where a moral structure is a complete set of semantic relations and normative forces that defines a moral system of principles. Third, the semantic content and normative force of any individual moral principle is thereby determined by the moral system as a whole — that is, any such individual principle is identical to whatever possesses a specific set of intrinsic structural system-dependent properties. Fourth, completely convincing, intrinsically compelling, or self-evident moral intuition applies only to the top level in the hierarchy, which are procedural meta-principles, and neither to intermediate-level first-order substantive ceteris paribus moral principles, nor to bottom-level actual duties. Fifth, the rational advance from the completely convincing, intrinsically compelling, or self-evidently intuited top-level meta-principles to the intermediate-level first-order substantive ceteris paribus principles to the bottom-level actual duties is a process of cognitive and practical construction. And finally, sixth, real conflicts of first-order substantive ceteris paribus moral principles at the intermediate level of the hierarchy are automatically resolved by a special set of level-theoretic structural constraints, taken together with one other moral meta-principle called The Lesser Evil Principle, which collectively fully preserve the absolutely universal objective truth and reality of the authoritatively-intuited meta-principles at the top level of the hierarchy. As the rest of this section rolls out, I’ll deploy these six moral structuralist ideas against the backdrop of broadly Kantian nonideal dignitarian moral theory, in order to capture The Kant-Sartre Insight and also express the full explanatory power of the theory.
Moral principles, whether absolutely universal and objective moral meta-principles, first-order substantive ceteris paribus objective moral principles, or moral duties, should also be sharply distinguished from moral judgments, which are constructive applications of objective moral principles in particular act-contexts. Indeed, the confusion between objective moral principles and moral judgments is perhaps the most persistent fallacy in both classical and contemporary interpretations of Kant’s ethics alike. It’s one sort of thing to determine the logico-semantic structure and normative implications of a given objective moral principle, and another very different sort of thing to figure out in the thick of things how a given objective moral principle is to be deployed or instantiated — that is, constructively applied — in a given actual act-context.
Now the thesis of constructivism, whether inside or outside moral theory, says that human minds and human agents play active, basic roles in determining and generating the content of all beliefs, truths, knowledge (especially including the knowledge of language), desires, volitions, act-intentions, and objective logical or moral principles. Corespondingly, broadly Kantian constructivism in the theory of mental content, cognition, and knowledge (aka Erkenntnistheorie) says that innately-specified rules essentially constrain the process by which human minds determine and generate mental representations of a manifest world that must also structurally conform to the formal constitution of their cognitive faculties.[xvi] And finally, broadly Kantian constructivism in moral theory says that a fundamental conception of the rational human agent essentially constrains the process by which agents determine and generate first-order substantive objective moral principles.[xvii]
Given this backdrop, every moral judgment constructively presupposes one or more objective moral principles; but the specific character and general properties of moral judgments cannot be automatically extended to objective moral principles, nor can the specific character and general properties of objective moral principles be automatically extended to moral judgments, since the mind-driven and agent-driven constructive process necessarily intervenes and mediates between the two. I’ll come back to the important distinction between objective moral principles and moral judgments in sub-section IV.3 below. In any case, assuming for the purposes of my argument at least the intelligibility of a sharp fourfold distinction between
(i) absolutely universal and objective moral meta-principles,
(ii) first-order substantive ceteris paribus objective moral principles,
(iii) moral duties, and
(iv) moral judgments,
all of which are projected into the larger theoretical frameworks of broadly Kantian constructivism and nonideal dignitarian moral theory, I want now to address the three classical problems of universalizability, rigorism, and moral dilemmas.
[i] See, for example, G. Boolos and R. Jeffrey, Computability and Logic (3rd edn.; Cambridge: Cambridge Univ. Press, 1989), chs. 10, 22, and 25, and esp. pp. 250–255.
[ii] See R. Hanna, Cognition, Content, and the A Priori: A Study in the Philosophy of Mind and Knowledge (THE RATIONAL HUMAN CONDITION, Vol. 5) (New York: Nova Science, 2018), ch. 5, also available online in preview HERE.
[iii] See Hanna, Cognition, Content, and the A Priori: A Study in the Philosophy of Mind and Knowledge (THE RATIONAL HUMAN CONDITION, Vol. 5), ch. 4.
[iv] See R. Hanna, Kant and the Foundations of Analytic Philosophy (Oxford: Oxford Univ. Press, 2001), also available online in preview HERE ; and Hanna, Cognition, Content, and the A Priori: A Study in the Philosophy of Mind and Knowledge (THE RATIONAL HUMAN CONDITION, Vol. 5), esp. chs. 2, 4, and 5. See also R. Hanna, “Kant’s Theory of Judgment,” The Stanford Encyclopedia of Philosophy (Winter 2017 Edition), E.N. Zalta (ed.), available online at URL = <https://plato.stanford.edu/archives/win2017/entries/kant-judgment/>.
[v] R. Stern, “Does ‘Ought’ Imply ‘Can’? And Did Kant Think It Does?” Utilitas 16 (2004): 42–61, at p. 59.
[vi] W.D. Ross, The Right and the Good (Oxford: Oxford Univ. Press, 2002/1930), p. 21.
[vii] Ross, The Right and the Good, pp. 20, 33, and 41.
[viii] Ross, The Right and the Good, pp. 30–31, and 145–148.
 Hanna, Cognition, Content, and the A Priori: A Study in the Philosophy of Mind and Knowledge (THE RATIONAL HUMAN CONDITION, Vol. 5), chs. 6–8.
 See J.L. Mackie, Ethics: Inventing Right and Wrong (Oxford: Oxford Univ. Press, 1977) See also section I above.
[xi] J. Rawls, A Theory of Justice (Cambridge, MA: Harvard Univ. Press, 1971), p. 41. Rawls correctly points out that unless Ross explicitly works out a monistic objective deontological lexical ordering or weighting scale for his prima facie duties in relation to actual act-contexts, then his view is a dead end. My objection is just the flip side of this, namely that Ross is clearly implicitly presupposing a monistic objective deontological lexical ordering or weighting scale for mapping his prima facie judgments to actual duties in actual contexts, without either admitting it or telling us precisely what it is.
[xii] See Hanna, Cognition, Content, and the A Priori: A Study in the Philosophy of Mind and Knowledge (THE RATIONAL HUMAN CONDITION, Vol. 5), chs. 6–8.
[xiii] See, e.g., P. Benacerraf, “What Numbers Could Not Be,” Philosophical Review 74 (1965): 47–73; S. Shapiro, Philosophy of Mathematics: Structure and Ontology (New York: Oxford Univ Press, 1997); and S. Shapiro, Thinking about Mathematics (Oxford: Oxford Univ. Press, 2000),ch. 10.
[xiv] See C. Parsons, Mathematical Thought and its Objects (Cambridge: Cambridge Univ. Press, 2008), esp. chs. 3, 5–6, and 9.
[xv] Ross, The Right and the Good, pp. 29–30.
[xvi] See, e.g., Hanna, Kant and the Foundations of Analytic Philosophy, chs. 1–2; R. Hanna, Kant, Science, and Human Nature (Oxford: Oxford Univ. Press, 2006), ch. 6; and Hanna, Rationality and Logic, chs. 4 and 6.
[xvii] See, e.g., T. Hill, “Kantian Constructivism in Ethics,” Ethics 99 (1989): 752–770; and O. O’Neill, Constructions of Reason (Cambridge: Cambridge Univ. Press, 1989).
AGAINST PROFESSIONAL PHILOSOPHY REDUX 548
Mr Nemo, W, X, Y, & Z, Monday 19 April 2021
Please consider becoming a patron!