Why Goodness is not a Property

Since G. E. Moore's Principia Ethica (1903) the question of whether "good" denotes a property of something has been important in the field of Ethics. Indeed, one might say it spawned the whole field of metaethics which looks at the epistemological and metaphysical underpinnings of ethical beliefs. Moore challenged the notion of hedonistic utilitarians like John Stuart Mill that we could determine the right things to do, the morally good acts we should perform and support, by a calculus of the acts' effects in terms of how much pleasure or happiness such acts added to, or subtracted from, the world. Formulating his "open question" argument, Moore pointed out that it made no logical sense to suppose that anything that could be brought about in the natural world was what we meant by "good" and so it could never be enough to presume that how much happiness or pleasure, or any other natural phenomenon, was fostered in the world could determine whether or not we should act to bring such things about. This, he noted, was simply because we can always ask a further question of the supposed good, i.e., whether it, itself, is good. That is, since we can conceive of circumstances in which pleasure or happiness are not good (think of the happy junkie or serial killer who takes delight in his pursuits), the term "good" cannot possibly be equivalent to being happy or, indeed, any other phenomenon or condition.

Moore's solution was to suppose that "good" must therefore be understood as denoting something else and, because everything natural (occurring as part of the natural world in which we stand) could be either good or not good (i.e., fell victim to the "open question"), what "good" denotes must, in fact, be understood as being non-natural. That is, Moore suggested, it must be some property of a thing that is outside the natural world. Moreover, he suggested that such a property must be bottom line in its own nature, i.e., it must be unanalyzable to anything more basic than itself. Finally, he reasoned, such a property must be known to us in a way quite different than physical phenomena (natural properties and their aggregates) are known to us. That is, he concluded, it must be apprehended, when it is present, in some direct, intuited way. Thus Moore concluded the word "good," while operating like other words for properties, such as "yellow," must name a kind of property quite different from the sort of thing "yellow" names.

Later thinkers in the twentieth century, while finding Moore's critique of hedonistic utilitarianism compelling, were less moved by his solution. The idea that moral goodness is intuited, without any explication or clarification of what the relevant intuition is about (how it works, what it stands on, etc.) left many of Moore's successors fundamentally unsatisfied. A new way of thinking about ethics grew up in Moore's shadow, best exemplified by the thesis of non-cognitivism, i.e., that "good" is not a denotative word at all, despite appearances, but an expressive one. The emotivist tradition supposed that assertions of goodness are about expressing how we feel about things, rather than naming anything, either in a merely emotional sense (the emotivists and subjectivists) or in a more sophisticated sense of approbation or recommendation (prescriptivism). Eventually more sophisticated expressivists, like Simon Blackburn, developed complex accounts of expressivist locutions in order to restore some measure of cognitive content to moral assertions, consistent with how we appear to actually use such assertions and in keeping with our expectations of them. That is, when we make moral claims, even assuming that they are merely expressive of our feelings or attitudes rather than claims about something in the world, we still tend to see them as being about things in the world with at least a degree of truth value relevant to them.

Blackburn, arguing for what he calls "quasi-realism" (moral realism is the idea that moral claims do denote things in the world, contra-Moore and contrary to the entire non-cognitive tradition which developed in the shadow of Moore's argument), holds that there is still room for granting cognitive content to moral talk by bending the rules for what we take to be real. He argues, for instance, that words like "true" and "false" are not merely used to designate that which corresponds to some objective truth measure (whatever we happen to settle on) for sometimes all we mean, by declaring a thing to be true, is that this is where we take our stand. We will entertain no further debate. By expanding and reshaping what we mean by truth claims, Blackburn makes room for expressive uses which appear to imply a kind of realism (hence his "quasi-realism").

But other efforts to accommodate moral talk to the descriptive assertoric language of common sense and science aim to go further. The naturalists, for instance, like Anscombe, Foot, Boyd and others reject Moore's non-naturalism in favor of a redefinition of "good" in more sophisticated naturalistic terms. Harking back to ancient Greek usages and especially to Aristotle, these individuals seek to lodge the notion of moral goodness in a concept or concepts of what works best for human beings. They take the solution to the intuitionist problem presented by Moore to lie in treating moral goodness as distinct, metaethically, from goodness itself, which latter term they treat as entirely uncontroversial. We all know what we mean by "good" they seem to be saying so let's just move on to the troubling matter of what moral goodness consists of. And here we get a variety of suggestions, stemming from Aristotle's ancient notion of good as being good for something. Thus, just as adequate water, nutritious soil and strong roots are good for a tree in order for it to fulfill its purpose of being successful as a tree (robust and long lived in the way of healthy trees) and fleetness of foot is good for antelopes, and sharp claws and strong jaws for their predators, so humans have the potential for certain qualities which are good for humans. What these are may be subject to quite a lot of dispute, however. Some qualities or needs which humans may have as a species, while necessary, are certainly shared with other similarly situated creatures. That is being living breathing animals, humans need nourishment and shelter in common with other creatures of similar type. But humans also have certain needs and qualities (or the potential for them) which are unique to the human species. These will include our capacity to reason, to operate in a social milieu and, perhaps, to develop certain capacities (for courage in the face of adversity, say, or humility, etc.). Of course this notion leads to ongoing debate about which of these features or potential features are relevant for the moral account and which are merely societal or survival related. More, it raises the question of what it means to give a moral account at all.

Perhaps, if a naturalistic explanation is the best one on offer, moral judgment is itself just a function of our evolutionary condition as a species. If it is, then the idea of moral judgment loses its special status since there is nothing inherent in being human that calls us to hold any particular moral opinion. Those opinions of this sort which we do hold we have by virtue of the contingent effects of evolution and our current condition in the world. Our moral beliefs could then conceivably change tomorrow and those who hold different moral beliefs than we do may very well be in the vanguard of the next evolutionary development. On such a naturalistic view we can never presume that there is an authoritative moral right or wrong, leaving us in a kind of moral relativism. Such relativism may well be the case, but if it is, then the entire moral game must be undermined. If the killers now running wild in the Middle East who behead and immolate captives while selling others into slavery or worse are successful in what they are now doing, who is to say they are not the next vanguard in human moral development? That we happen to condemn and oppose what they're doing has, itself, no moral weight behind it if moral relativism is the case.

Much of the difficulty in all this comes, I think, from the notion we have, at least since Moore, that in seeking to define "good" we are looking for a property and, if we can't isolate one, all we have left are our personal attitudes or feelings about things. Both non-cognitivism in all its manifestations and naturalism lead straight to moral relativism. But perhaps there's another way? Perhaps the whole problem comes back to what we mean by "good" and to what it means to value anything at all. To arrive at a satisfactory account of moral goodness, perhaps we first have to follow Moore back to the denoting problem associated with Moore's account of "good" and start from there?

Well, if "good" doesn't denote a property, whether a natural or non-natural one, what does it do? Here's Merriam-Webster's Online Dictionary on the relevant use of "property":

1 a: a quality or trait belonging and especially peculiar to an individual or thing b: an effect that an object has on another object or on the senses c: virtue [a beneficial quality or power of a thing; a capacity to act: potency] d: an attribute common to all members of a class

So "property" in this sense is used in English to denote a thing's quality or trait. One may add here a "feature" or "characteristic" as well for these words fit the property bill in this sense, too. "Property" may, according to the definition, also pick out a thing's tendency to do something to something else (e.g., water's property of dissolving salt for instance) or, this definition suggests, it may name something beneficial, whether feature or disposition, and it may also just name some particular aspect of a thing which is common to all things of the same type. John Locke famously spoke of two kinds of properties which a physical thing might have. There were the primary properties which he supposed were inherent in the things which had them, meaning that they always had them in all possible situations because they were an intrinsic part of the things in question, and the secondary properties which only fixed to a thing because of the observer's way of perceiving it. Thus, for Locke, some properties of physical objects, like mass and extension, were always with the object that had them and, in effect, constitute those objects when taken together, while other properties, like a thing's color or, perhaps, its feel or smell, depend on other things, such as the light in which the thing is seen or the capacities of the viewer who sees it. Of course this view of physical phenomena no longer seems quite right in light of modern physics since we now suppose that every physical phenomenon, whether it has mass and extension or color and texture is just an array of various subatomic level phenomena, whether in terms of mesons and protons, quarks or branes or strings and their properties are nothing like those we encounter in the world through our perceptual apparatus since we don't have any perceptions of them at all (though, arguably, what we perceive is just the interplay of these deeper phenomena, both in terms of what is going on with the things we encounter and what's going on in our own sensing and processing apparatuses).

At least one thing jumps out at us in view of this, when considering the notion of "property," is that, like nearly all words in natural language, "property's" use is wide, variegated and somewhat indefinite in how the word is applied. As the dictionary suggests, a property may be an element or aspect or even a disposition of a thing, and it may be a fixed part of the thing or something projected onto it by the perceiving party. And it may be a physical feature or it may be some kind of potential to do something. Likening goodness to a property then, as Moore did, and expecting the word which names it to operate the way a word that names a property like color seems particularly naive. Why, in fact, should "good" work this way and why should we expect "property" to name only one sort of thing in the universe?

In a sense we may want to say of goodness that it is a property which some things have and others lack but what does that sort of statement even mean when the notion of "property" is so broad in its application? More, if we cannot find anything uniquely and always to exemplify just what we mean by "good" in the thing we choose to call "good," which also seems to be the case -- if, that is, even a common good like pleasure or happiness or contentment, is not always going to be thought good -- must we presume that our use of the term "good" ultimately names nothing at all, as the non-cognitivists would have it, or that it necessarily equates to particular things, elements, aspects or arrangements within the natural world just because some things may always be seen to be good for humans in general -- even if we can't always agree on what those things are?

Perhaps the whole idea of goodness as a property is what's wrong here. If by "property" we mean some feature or element or aspect (or the disposition of something to appear in a certain way, or do certain things), what particular sort of thing can "good" mean when applied to such a wide array of things, whether object or action, state or condition?

Indeed, if we can say of some properties that they are good and others that they are not, which we manifestly can (e.g., for some medications to have the property of reducing blood pressure in those who take them is certainly to have a good, as in helpful, property when the patient's pressure is too high), then are we saying of some properties that they have further properties, and one of these is the goodness we ascribe to them? Does it make sense to speak of properties of properties? Granting that language is highly flexible and that the meanings of our words are a function of how we deploy them, we can certainly acknowledge that sometimes we may want to use "property" in ways like these. After all the color yellow, a property on the usual interpretation of this word ("property"), can have the further property of reflecting light in a certain way, of being hard on some eyes (e.g., property as a disposition to affect) and so forth. So perhaps there is no reason to worry about treating goodness as a property then?

But it doesn't help us understand what we mean by the term "good" in that case. If anything can be a property in some contexts and properties can, themselves, have properties, then we are no closer to seeing what's meant by any usage which employs the term "good" in the standard way. But surely "good" means something for we recognize how to use it most of the time. The problem seems to lie in effectively explicating that usage.

Perhaps not all usages can be adequately explicated but that doesn't mean we oughtn't to try in this case. And here I want to suggest that we tend to think of goodness not so much as a property per se, at least not in any of the traditional senses, but as a different sort of attribute we may sometimes ascribe to some things. Suppose, for instance, we speak of something as being here or there. We don't, in those cases mean, the same location in every case or that being in one place instead of another is a property of the thing. The location we have in mind will be determined by the relations of the point picked out with a variety of other things, including, most especially, the person doing the picking out. That is, the place that is here or there (or the point in time that is now or then, to consider a related use) will vary with the speaker. So, too, do ascriptions of goodness. What seems good to one speaker may not seem so to another.

What seems to be meant by the assertion ascribing goodness to a thing is that that which is being picked out by the speaker offers some reason or reasons for the speaker to choose it, all other things being equal. That is, if I say of any X that it's good, what I have in mind, what I mean by saying that, is that the thing so designated has some feature or features which, in the given circumstances, and everything else being equal, presents me (or the person whose attention I am directing to it) with a reason to take it up.

Is providing a reason to do or acquire something a property of that thing? Perhaps in the broader sense of "property" we can suppose it is. If a property is a disposition to do or cause X, which on the definitions discovered it may sometimes be, so we may want to say that being the source of a reason to choose the thing it contains (or to choose the property itself) -- and here "choice" means a whole range of things from acquiring to pursuing to acting in accord with -- is to be a property of a sort, too. But this use of "property" is attenuated and seems confusing, not least because of the more familiar use of people like Locke who use the term "property" to designate physical features. If we think in such terms then we are prompted to look for the physical feature that is the good named or, failing to find it, to suppose nothing is to be found or that what we're after is a special sort of "property" which lies outside the natural world.

But something is to be found, as our usages make clear, for we certainly do think that good things have goodness and we show that when we use a term like "good." But what's to be found is less a property in this Lockean sense, the sense which Moore imagined was relevant, than it is in a more abstract sense. That is, what's denoted is relational rather than perceptual. It's more like what's denoted by words like "here" or "there" and "then" and "now," i.e., a temporary condition or status which a thing has. In this sense what we're doing by our use of "good" in a denotative way is announcing or reporting a thing's status, the condition in which the thing stands. And a status or condition is not a property of the thing because properties belong to the things that have them, either in every case in which the thing occurs (Locke's "primary" properties) or in certain precisely indentifiable cases (his "secondary" properties).

But goodness, as a status, is that by ascription, not description, i.e., it is ascribed to a thing when the thing possesses some properties (the perfectly natural sort, of course) which provide the ascriber a reason to select it. Once we divest the idea of goodness from the idea of being a property, in favor of a more relational term like "status" or "condition," the old problems which bedeviled Moore and his successors fall away. Of course we grant things a status without supposing that status is a permanent part of the thing (either because it's inherent in the thing itself or inherent in how we encounter it). Statuses are always a function of the conditions in which they occur and not of the thing in itself, or in ourselves. But statuses are no less cognitively real than properties. They may be ephemeral in the way all relations are, i.e., coming and going with the shift in related elements which determine them, but they are also objectively knowable because they represent temporary conditions in the world determined by the natural phenomena which constitute the world.

If the fear that "good" represents a mysteriously unknowable property or nothing at all -- or that it must simply be equated with other knowable things in the world -- leads us into all sorts of peregrinations to explain and account for our moral judgments, yet leaves such accountings in uncomfortable morally relative territory, is to be overcome, then we must start with the recognition that looking for a goodness property is a mistake. Jettison the idea that "property" matters in favor of a notion of goodness as a status or condition, and the rest falls more easily into place. It's easier to explain moral judgment, and show its validity, if we first rid ourselves of the notion that there's some special moral goodness property waiting to be found.