Andrew Brook

Section I. Opening Remarks

The remarks that follow are made in the spirit of Kant's notion that nature only answers the questions we ask her, not any questions whatsoever.(1) The importance of this notion, which has become a common-place of philosophy of science, lies in its suggestion that empirical research will yield worthwhile answers only to clear and well-formed questions. Clear and well-formed questions, in turn, require a clear, detailed, conceptually coherent picture of the alternatives that might be found in the material being studied. Such a picture and only such a picture can yield clearly delineated and unambiguous hypotheses. With such hypotheses, we will know the results we need, to confirm or disconfirm our hypotheses, and we can make sense of the results we actually get. Kant made his remarks in the context of the natural sciences, of course, but they seem to me to be just as applicable to the social sciences, including projects in the sociology of ethics such as we are discussing this weekend.

As I understand it, the aim of the project we are considering is to discover how Northern Ontarians from various walks of life feel about resource extraction, and more specifically to uncover the ethical positions people hold on this issue. Before we can find out what ethical positions people hold on an issue, we first need to have a clear picture of the alternative pictures they could hold. Otherwise, we will not even know what questions to ask. That is to say, we first need to map the theoretical terrain. In my view, the theoretical terrain in this case consists of two things: the various different ethical positions which could be adopted on an issue like resource extraction; and the various dimensions on which these theories differ. (What makes different ethical positions different is that they adopt differing positions on various ethical dimensions.) The former identifies the possible alternative positions; the latter can give us a model of how they really differ.

I identified the project before us as a project in the sociology of ethics. That is to say, it aims to discover the ethical positions that a particular social group (in this case, northern Ontarians) actually hold on an issue. This could be called a project in descriptive ethics, but I find this label more confusing than helpful. The problem is that the label 'descriptive ethics' is used for two totally different things. As well as being used for this sort of empirical study, it also used as a label for meta-ethics, the study of the nature, concepts and structure of ethics and ethical positions as a whole. Because of this duality, I find the term 'sociology of ethics' more informative.

To get a system of concepts clear and coherent enough to support successful sociology of ethics, we need to clarify a number of different concepts. The first is the concept of ethics itself. The domain of ethics is always contrasted with the domain of facts. However, the contrast is made in at least two different ways. 'Ethics is concerned not with what is, but with ...' -- and two alternatives present themselves. Historically, the usual alternative is 'what ought to be': ethics is not concerned with is's, it is concerned with oughts. The other alternative is to say that ethics is concerned with values, not facts. As a way to characterize what ethics is about, I prefer the second alternative.

Before I can say why, I need to address a potential ambiguity. The term 'values' can mean 'the values people hold', or it can mean 'the value something has'. Construed in the former way, the 'values' reading is not an alternative to the 'ought's' account, but comes to much the same thing. If we construe 'values' in the second way, however, that is not the case. Construed this way, a values account is markedly different from an ought's account. The reason is this. We can only apply the notions of ought's and ought not's to beings with some power of moral choice. However, the notion of having value can be sensibly applied to a vastly wider range of beings. (Sensibly applied is not the same as correctly applied, of course.) If I am right about this, an 'ought' reading of ethics comes very close to begging the question against many important positions in environmental ethics, since many important positions in environmental ethics hold that vast ranges of beings who do not have the power of moral choice still have ethical standing. Such a claim is an important part of the bio- or ecocentric alternative to traditional, anthropocentric Western ethics.

Here an important objection sometimes appears. Something having value, the objection runs, is a matter of that thing holding ethical interest, being worthy of our respect, having features worth preserving, or something similar. So far as we know, however, we (i.e. human beings) are the only beings capable of recognizing or respecting respect-worthy values. If so, how can there be ethical value outside of human beings? Put another way, how could anything broader than anthropocentrism even get off the ground? Despite its wide appeal, this objection contains a serious confusion.

The confusion consists of failing to distinguish recognizing and guiding one's action by ethical value -- being a subject and agent of ethical concern --, which is indeed arguably true only of human beings (and perhaps some other higher mammals), from having ethical value -- being an object of ethical concern, which could be true of vastly wider ranges of beings than that, up to and including the ecosphere as a whole. Once this distinction is made, many of the purely conceptual attempts to dismiss bio- or ecocentric ethical positions out of hand immediately collapse.

If the concern of ethics in general is value, what is distinctive about environmental ethics? In my view, nothing but subject-matter. Environmental ethics is simply ethics, all of it, applied to environmental issues.

Section II. Ethics and the Sociology of Ethics

If the project before us is a project in the sociology of ethics, what parts of the discipline of ethics as a branch of philosophy are relevant to it? In connection with ethical issues, philosophers tend to ask three kinds of question, and the various ethical positions held by people tend either to argue for or presuppose answers to all three kinds of question:

1. 'Meta-ethical'. Representative questions here would include, What is value? What has value? What is the internal structure of this or that ethical position like?

As these questions make clear, meta-ethics in this sense is just a branch of the discipline traditionally called metaphysics, the study of the most general and conceptual questions about the nature of things. Meta-ethics contains another branch, too, what we might call 'meta-normative-ethics'. Here the questions asked are about norms and normative reasoning. Can normative positions (positions about how people ought to be behave, about what has value, etc.) be justified? If so, what strategies of argument will successfully do so? And so on. One such question receiving a lot of attention at the moment is the question of whether we need to be able to ground ethical positions in something more basic if we are to justify them. The anti-foundational view that we do not need to do so is rapidly becoming the dominant view.

2. Epistemological. Representative questions here would include, How can we know when value is present? In what quantity? How can we know when we have an obligation? And so on.

3. Normative. Representative questions here would include, In any given context, what has value? What respect do we owe to it? What ought to be done? And so on.

The project before us being empirical, neither the normative questions nor most of the meta-ethical questions about norms have much relevance. We are trying to find out to what people do give value, not what they ought to value. The only exception I can think of is the question about foundations. Likewise, I cannot see that knowing how an ethical position would answer epistemological questions of the sort I sketched is especially important to discovering empirically whether someone holds it. However, those metaethical questions which I characterized as metaphysical are directly relevant, as we shall see. Without answers to questions like these and some others, I think it would be impossible to know what ethical positions someone holds, at any informative level of detail at any rate.

Section III. Positions in Environmental Ethics

The various positions in environmental ethics tend to argue for, or presuppose, answers to all three kinds of question, plus some others. Before we look in more detail at the questions of greatest relevance to the project in front of us, namely the questions about the nature of value and the structure of an ethical position on environmental values, I will first sketch some of the major positions currently prominent in the environmental debate.

1. Libertarianism(2) (egocentrism) -- The fundamental notion here is that each person has the right to do whatever he or she wishes (whatever 'right' means here). In addition, there is always some principle for abridging this 'right'. One is that the right can be abridged only by agreement, the core of the view called contractarianism (Hobbes, Locke). The most common since Mill has been the idea that the exercise of the right must not harm others, though contractarianism is again coming into fashion.

This position is the heart of classical liberalism. Since there are no obligations beyond either the restrictions of not harming others or of contracts, libertarianism defends the ideas that individuals have massive individual liberty to pursue their interests as they see them, with respect to how they treat the environment as elsewhere.

Libertarianism is not egotism, because it says nothing about the range of interests an individual may pursue (the good of the environment could be such an interest) and it is compatible with the values recognized by many other ethical positions, including even utilitarianism and communitarianism, though not of course with their views about what creates value. However, it has traditionally been associated with something pretty close to egotism about interests, and an atomistic hostility to the idea that communities or even collective interests (even 'the greatest good for the greatest number') have anything much to do with creating or recognizing value. It also has generally treated the interests of each as hostile to or at least in competition with the interests of other people (not to mention the rest of the natural order). Environmentally, libertarianism often advocates making the environment more resilient as the way to cope with the exercise of unbridled liberty.

2. Traditional Western morality (narrow anthropocentrism) -- Here I mean not the peculiar ethical position of classical liberalism, but the more dominant forms of western morality: communitarianism (Aristotle), natural law theory (Aquinas), general will theory (Rousseau, Hegel), utilitarianism (Bentham, Mill), etc. In all these positions, values other than liberty exist even without any agreement to recognize and enforce them. In all these positions, however, the domain of morality still remains people's treatment of people (or some privileged group of people, white males usually being the group). Nature is a resource, to be used as best suits our interests, in much the way we use tools to do a job; it is outside the domain of moral concern.

This is the ethical position of much of the planning and development community. In contrast to the shallow and deep ecology that follow, a term for this sort of position might be 'surface ecology' (or 'non-ecology').

3. Shallow Ecology (enlightened anthropocentrism) -- As with the last set of positions, here too the dominant idea is that earth and all that dwells therein (except humans) are there to serve human interests. However, shallow ecology pays a lot more attention to the role of nature in serving human interests. It both sees how essential the rest of nature is to human interests, at least in the long run, and emphasizes certain interests that are unique to our relationship to nature, for example aesthetic interests (we all enjoy beautiful scenery, hikes in the woods, sunsets) and spiritual interests (there is a peace to be found in, for example, purifying oneself by regaining one's sense that one is an animal in the natural order that is unlike that to be found in any other way).

What has come to be called shallow ecology often takes a utilitarian form, centred on human utilities, but it can be an off-shoot of any of the standard forms of Western moral theory named above. If the interest shallow ecology takes in nature is still restricted to human self-interest, it is far a more enlightened self-interest than is traditional in western morality. Shallow ecology urges that we husband the 'resources' of nature carefully and distribute them justly, not just over space but also over time, leaving a fair share for future generations. However, the reason for doing so is to maximize the satisfaction of human interests. Thus, even though shallow ecology pays a lot of attention to enhancing the instrumental value of non-human nature, in one good sense it is still anthropocentric -- only human interests have any non-instrumental value. The doctrine of stewardship is a common religious version of this position. Shallow ecology is the ethical position of old-style naturalists, conservationists, anti-pollution advocates, and most theories of sustainable development; the ethics of the Green Plan are the very essence of this position.

4. Moral Extensionism (sentientism) -- That pain is the fundamental evil in the world and pleasure the fundamental good are the usual animating principles of this position. Thus, all beings that have the capacity to feel pain and pleasure fall within the domain of moral concern, and not because of their value to us but 'in their own right'. Since many more animals appear to have the capacity to feel pain than the capacity to feel pleasure, including most or all of the insects, worms, shell-fish, etc., not inflicting pain tends to be the primal value of this position.

The animal 'rights' movement is the best-known example of this position. Fortunately, the position need not be developed in terms of an extension of rights. The notion of rights is a notion that has its primary home in the interpersonal context of creating and protecting rights in law, as a result of a process of agreement (or surrogate agreement, where legislators are the surrogates). If so, it is hard to see how beings who have no capacity to give consent or enter agreements could be said to acquire rights, except by the grace and favour of those beings who do.

5. Deep Ecology (bio- and/or ecocentrism) -- Deep ecology is centred on two fundamental ideas. One is a factual claim: (a) Far from being different from or dominant over nature, we are part of the natural order and are utterly dependent on it. The other is a claim about the range of value: (b) Value extends beyond human beings and human value both in scope and kind. Thus deep ecological positions tend to contain both claims about both facts and values.

Notice that I have said nothing about kind of value. Different forms of deep ecology are characterized by deep differences on the kind of value to be found in non-human nature, and also on the most desirable response to it. Some of the main variants of deep ecology are:

-- intrinsic value holism(3): every living thing and perhaps, if the Gaia hypothesis is right, the ecosphere as a whole has intrinsic value and deserves respect. There are different versions of this position, depending on different definitions of 'intrinsic value' (a point about which I will say something more in the next Section) and different ideas about how far intrinsic value extends (just to life, to ecosystems, to the biospheres as a whole (Gaia)).(4)

-- 'System' holism: The systems necessary for life clearly have instrumental value. However, compared to these systems, individuals have little value. However we define 'value', without these systems, there would be no individuals to have value. Thus, ecological system are the primary bearers of value of all the sorts there are. We know of no better way for these systems to protect and propagate themselves than the way they in fact do so. Thus, the way they do take care of themselves is the best model we know of for the way they should do so. In these protective and propagative activities, ecological systems are often extremely hard on individuals. Being the best way there is for them to protect themselves, there is nothing morally wrong with this.(5)

-- Eco-feminism: In addition to adhering to the idea that (a) there is value, and not just instrumental value, in everything that is, eco-feminism is animated by two further ideas. (b) Any ethics of duty and obligation -- any ethics based on rules -- cannot give us a desirable or perhaps even a coherent guide to our treatment of the environment. Only an ethics based on the notion of care and concern can do that. (c) Man's (note: man's, not human's) domination of nature parallels man's domination of women.(6)

-- Anti-ethics: Ethics of any kind is a human-invented prosthetic to replace an animal nature which we have lost. Back to Nature!(7)

(I have not treated religious positions on the environment separately in the above because they can usually be understood as religious versions of one or another of the positions described above, even though I have described the latter in secular terms.)

Section IV. A Scheme for Distinguishing Positions in Environmental Ethics

Having laid out various ethical positions people now hold on the environment, I may surprise some people when I now say that any such list of positions seems to me to have very limited utility, by itself, as a framework for guiding empirical research into people's ethical positions. Without careful definition, these labels do not tell us much, if anything, about the nature of the ethical position they label. Nor can any short, thumb-nail description such the ones I gave above provide such a definition. That we can get only if we find out what a position's stance is on a number of different dimensions of value. In fact, not only do the labels not provide this, even the location of the position on any one dimension will not provide it. To illustrate this point, consider what I will call scope. Scope, as I introduce it below, is a matter of how many kinds of beings, distributed over how much space and time, fall within an ethical position's domain of moral concern. Clearly, this is the primary dimension used above for arranging the positions I sketched. Why does this single dimension not tell us much by itself?

The reason is this. The scope of a theory by itself does not tell us what kind of value is being given this scope. Take, for example, a position with the scope of deep ecology, namely all living beings and the systems which support them in all places and all times. For this information to be really informative, we also need to know, for example, (a) what kind of value it is that is being given this wide scope. Do all beings in all places and times have value because they all are or could be instrumental in human well-being, or is it because they have some value in their own right? And we need to know, (b), what does the position see as the implications of this value for the value of individuals? for what human beings should do? for the distribution of resources? And so on. In short, neither simply attaching a label to a position nor even locating it on a single dimension of value will tell us much about the nature of that position. What we really need to know is where the position stands on a number of ethical dimensions. No one dimension will do the job.

I have rather belaboured this example of scope for a reason. People often seem to think that holism, the extension of value to everything that is, is the defining characteristic of deep ecology. This views is seriously inadequate; knowing that an ethical position is holistic actually tells us very little about the nature of the position, for the reasons laid out in the last paragraph.

As a tool of empirical research, any list of positions will face other problems, too. One important one is this. Again referring to the list developed above, notice that, for most of those positions, holding one position may not even be incompatible with holding some of the others, depending on how the positions are defined in detail. That is to say, they may not even be mutually exclusive positions. As I sketched it, libertarianism, for example, is compatible, at least in principle, with practically any other position. Thus someone could -- again in principle -- be an intrinsic value deep ecologist and still be a libertarian. The person may recognize that her values are, as such, merely her preferences, that others may disagree with her about deep ecology, and that the only way to resolve such conflicts and justify claims about values is by agreement among the affected parties. Or, to take another example, someone could be a shallow ecologist and yet be a value holist -- the person sees us as merely part of the greater whole of nature, with no right or power to dominate the rest of nature, but thinks that the value everything has is instrumental value: value for the well-being of other beings and particularly humans. (We might be able to find a view containing something like this mix in early Callicott and perhaps even in pre-Sand County Almanac Leopold.) Yet a third example: one could give a shallow ecologist analysis of value, and yet be an eco-feminist, at least so far as points (b) and (c) of eco-feminism as I defined it are concerned. Thus, an ethical position being located under one of the labels in the list I just sketched does not even preclude its being located under others. Lists like this, by themselves, are a very limited tool of empirical analysis.

We could, of course, attempt to rectify these problems with labelled positions arranged on a single dimension. The most obvious way would be to force each position to reveal its stand -- or take a stand -- on each of a number of different dimensions. But then we would run into a different group of problems, problems stemming from the difficulties of coming up with adequate single definitions of kind words. The notion of finding a single set of necessary and sufficient conditions for the application of a label for a kind -- the notion of finding what we might call a standard definition for kind terms -- was abandoned about two decades ago. First, as such words are actually used, it is seldom, if ever, the case that all the instances of the kind share a single set of properties. Rather, they are usually linked by what Wittgenstein called family resemblances. Secondly, if we force such terms into the mold of a single definition, they are then unlikely to apply to many of the positions that people actually hold, that is to say, they are unlikely to have lose whatever use they originally had a classification scheme. These two points are closely related, of course.

In the years since the demise of necessary and sufficient conditions, two ways of attempting to capture the meaning of kind terms have been prominent. The first attempts to substitute clusters of characteristics for the old necessary and sufficient conditions, the idea being that if something has 'enough' of the items in the cluster, then it deserves the label. Among the many problems with this approach, let me mention just two. First, how much is 'enough'? Secondly, we usually cannot find any adequate cluster, either, except one that is extremely disjunctive and therefore trivial (Wittgenstein is famous for his far-ranging exploration of this difficulty). In short, it is hard to define a label with enough precision to ensure that attaching it to something will provide a useful level of information about the thing -- and harder still to devise a single ordered list of them!

The attempt to find cluster definitions has recently been replaced by the study of so-called prototypes, accounts in terms of our stereotype of paradigm cases of a kind of thing. The stereotype need not contain any necessary or sufficient features of the kind. Indeed, it need not even contain features which all examples of the kind actually have. All it need contain are the features by which we usually recognize examples of the kind.(8) As a way of capturing the meanings of words, this approach has the advantage that it can at least be made to work. The thumb-nail sketches I gave of each position in Section III are examples. Each of these sketches is really a description of our stereotype of each position, and they work. Unfortunately, for our purposes, the approach also has two big draw-backs.

First, even if we could recognize in some way that the ethical position someone is expressing fits our stereotype of one of the positions on our list, we would not yet know what they have in common. We would merely have recognized some kind of resemblance between them. This outcome is even worse than what we might have hoped to gain from a standard or a cluster definition. Secondly, the ethical stance(s) that express our stereotype of a position need not actually be held by the people to whom we and even they are inclined to ascribe the position.

In sum, any of the available ways of attempting to overcome the weaknesses of arranging ethical positions on a single dimension fail to give us what we need, namely useful information about the nature of an ethical position. Whether we attempt to capture the content of a kind of ethical position in a standard definition, in a cluster account or even in a description of our stereotype of the position, we cannot generate a tool adequate for revealing the content of the actual positions people hold. Ethical theories cannot informatively be classified by any single set of labels or on any single dimension. This is a major weakness of the work by Secrest, et. al. presented to us as a possible model to follow.(9)

Rather than beat this dead (or at least very moribund) horse any further, I propose that we look in a different direction. Rather than trying to improve any list of single labels, I propose that we give up trying to fit people's ethical positions under single labels, and assess them instead on a number of ethical dimensions, not just one (one dimension or one list of labels). This sort of analysis would provide some real information about the structure of the position under investigation. Rather than merely telling us about which label applies or about the place of the position on a single dimension, this investigation would yield a model of the position's whole structure.(10)

How would such a multi-dimensional analysis of the structure of the position a person holds relate to fitting the person's position on the list of standard positions developed in Section III? The answer, I think, is not very tidily. With an analysis of the structure of a person's ethical position in place, you could, if you wished, check to see to which standard position, as captured by its stereotype description, the person's position tends to gravitate. But this check will reveal little extra information. Moreover, it is quite likely that the person's position will not fit the stereotype of any single position very closely.

There is another, somewhat related possibility, however, which strikes me as both more likely and more interesting. If we were to analyze the structure of the various positions of a whole population, it is likely that many of the differences on the various dimensions will co-vary. It is likely, for example, that someone who holds that ethical concern has narrow scope will also not grant much role to intrinsic value. It would be very interesting to see how differences of location on the various dimensions cluster, and how many distinct clusters are to be found. Each distinct cluster would identify a real ethical position, one with internal integrity and clearly distinct from any other, not a mere stereotype. It would then be interesting to see to what extent such distinct clusters coincide with the environmental positions captured in the list of labels sketched last Section. My guess is that the co-incidence would be anything but perfect.

With these very incomplete remarks about lists and single-dimensional vs. multi-dimensional analyses in place, let me now turn to the dimensions themselves. What follows are some of the dimensions that we might use to analyze the structure of ethical positions on the environment. Note: analyzing the structure of an ethical position is not the same as evaluating either its conceptual or its ethical merits. For the purposes of the project before us, it does not matter how good an ethical position is, either at the level of its conceptual structure or as an ethical position. What matters is that we describe it correctly. The first dimension has already been mentioned. It is scope.

1. Scope:

<me, here, now vs. all, everywhere, everywhen>

(i.) How many kinds of beings, distributed over how much space and how much time, should we care about? (ii.) How should costs and benefits be distributed over these beings, spaces and times? (Humans, animals, all living, all parts of ecosphere. In our vicinity vs. everyone on face of earth? Now vs. future generations? With respect to the latter two, distant beings discounted, or all beings treated equally?) As I said above, wide-scope, holistic views can still be instrumentalist and anthropocentric (also cf. 4a. and 4b. below).

Clearly scope is not really a single dimension; rather, the term labels a parallel dimension in a number of different phenomena. The only reason I am treating it as one dimension is to keep the total number of them manageably small (and perhaps also to hide some of the true complexity of what I am proposing).

The ethical positions sketched in Section III have obviously been classified in some way by their scope. However, until we know what is being classified by its scope, that does not tell us very much. The answer, as it happens, is largely: scope granted to intrinsic value. At one extreme, libertarians hold that only thing has intrinsic value for me, namely me, and that what has this value for me is my wants. At the opposite extreme, intrinsic value holism holds that everything has intrinsic value. So let us make distribution of intrinsic and instrumental value our second dimension.

2. Kind of Value:

<instrumental vs. intrinsic>

This dimension clearly cross-classifies the previous one. We can discuss, for example, the instrumental value of something for me, for us or for other kinds of being, and the same is true, of course, for intrinsic value. I will return to the vexed question of the meaning of 'intrinsic value' at the end of this Section.

3. Kinds of states and processes valued:

<like ours vs. what thing values vs. of thing>

The first point on this scale consists of what humans value (so is either ego- or anthropocentric or extensionist). The middle point is teleological intrinsic value. The last point is harm/benefit intrinsic value.

4a. Place of the individual as object of the theory (a.):

<good of the individual vs. treatment of all>

What balance is struck between the good of the individual as worthy of respect and the requirement that all individuals in the scope of the theory be treated equally?

4b. Place of the individual as object (b.):

<good of the individual vs. of the ecological system>

What balance is struck between the good of the individual and the good of eco-systems and the ecosphere as a whole?

5. Place of the individual as subject of the theory:

<individual liberty vs. to ecological well-being ('equality')>

Where does the theory fall on the scale of individual liberty vs. contribution to equal treatment of others or the ecosphere as a whole?

6. Basis of ethical concern:

<duty to respect vs. care and concern>

This is the scale defined by the debate between traditional moral theory (including 'traditional' rights- and even value-based deep ecology -- any theory that bases ethical concern on duties), at one extreme, and care-based eco-feminism, at the other.

7. Kind of reasoning used:

<foundational vs. non-foundational>

Does the position hold that ethical positions need to be justified via syllogistic reasoning starting from foundational principles?

8. The way values are recognized:

<moral authority of some kind vs. by mutual agreement>

The first pole resembles the value-creating and -recognizing mechanism advocated by contractarianism, namely interpersonal agreement. The moral authority of the other end of the scale could be religious authority, philosophical inference from distribution of utilities or some other first principle, or whatever. Dimension (7) and dimension (8) are related. Contractarians about recognition of moral values must in consistency be anti-foundationalists about their basis. The reverse is not true, however. An anti-foundational ethic, for example an ethic of care and concern, could still hold that the value of care and concern does not need any agreement to become a generally-justified value to adhere to and to insist that others adhere to.

In case it is not already obvious, I should note that an ethical position could not be located on the kind of dimensions I have been identifying by its place on a line, in the way a physical object could be located on a scale running from large to small by finding where it would fall on a line running from one term to the other. To locate an ethical position on dimensions of the sort we have been discussing, it would have to be located by a description. For example, Rolston's position might be located on dimension (2), intrinsic/instrumental value, with the description, 'everything has both, but the higher something is on the evolutionary ladder, the more intrinsic and the less instrumental value it has, in general'.(11)

Since the notion of 'intrinsic value' is far from clear, I will close this Section with a few remarks on this topic. There are two prominent ways of defining the term, teleologically, and in terms of capacity to be benefitted or harmed. The argument for the harm-benefit definition is straight-forward: if something can be benefitted or harmed, then it has a good, an interest or whatever, quite independent of its value for anyone or anything else. An argument for a teleological definition might run as follows: if something has goals, then it seeks its good -- its good, quite independent of its value for anything else. The case is even clearer if something can distinguish between the state it is in and the state it would be good for it to be in (as for example even plants do when they seek the light). In such cases, the thing recognizes a good for itself, and therefore takes itself to have value (to be worth protecting), quite independent of its value for anything else.(12) Because of its simplicity, I prefer the harm/benefit account of intrinsic value. It is also far broader than the one in terms of teleology, obviously, since virtually anything can be benefitted or harmed, but only entities with goals can have a telos.

The notion of intrinsic value has often been criticized, especially by the more hard-headed, ontologically parsimonious kind of philosopher, as requiring a special ontological category, a special and rather unusual kind of property. I do not think this criticism can be made to stick, not on all forms of the notion at any rate. In fact, in its less extreme forms, the notion is perfectly benign. It says no more than that everything has some value, interest, worth, importance in its own right, quite independent of its value to anything else. There is nothing radical in that.(13) Of course, developed this way, the notion is also pretty trivial. For if we say that everything has some interest and worth, the notion of intrinsic worth instantly loses all power to guide our actions. To do that, we need in addition some mechanism for rank-ordering worth. The notion of intrinsic worth might become no more than the admonition 'Consider the costs and benefits to everything involved before making decisions about what and how to use things'. To be sure, even this simple admonition would often be far from useless.

In one important respect, however, the remarks of the previous paragraphs are irrelevant. For the purposes of the project we are considering, the merits of the notion of intrinsic value do not matter. All that matters is whether the target population accepts that intrinsic value exists or not, and what scope if any they take intrinsic value to have. For this purpose, all that matters it that we have a notion that is clear enough for empirical research. Whether there is any intimate merit to the notion is completely irrelevant.

Section V. Distinguishing Values from Other Things in Empirical Studies

To end this presentation, I will make a few remarks on two further topics. Up to now, I have been discussing the kind of conceptual scheme we need if we are to study a population's ethical positions successfully. I now want to turn to what happens when we apply this scheme to a population. How can we map this scheme onto a person's ethical position?

Clearly this is a big question with many parts. Here I will examine just two of them. We want to uncover a person's values; but all we have available to us is their behaviour: what they say, how they behave, and how they respond to our research instruments (survey questionnaires, scenarios, budget allocation exercises, projective instruments, or whatever). Take the latter. Unfortunately, all a person's responses to a research instrument tells us for sure is what he or she thinks it is appropriate to say or do in response to this instrument. They do no even tell us what the person thinks his or her values are, let alone what they really are. On the other hand, we cannot read the real values off from their behaviour, either, because the human race is notorious for not living up to its values, whether through fear, self-interest, lethargy, or some other form of weakness of will. Thus we have four alternatives:

1. The response a person thinks to be appropriate,

2. What a person believes his or her values are,

3. The person's actual values, and,

4. What motivates the person's behaviour,

and our dilemma is this: We want to get to (3), the person's actual values; but neither working down from the responses, etc., nor up from behaviour via interferences to the motives for the behaviour seems at all likely to get us there. Due to false consciousness, self-righteousness, the influence of ideology, a desire to protect self-esteem and other factors, people often know surprisingly little about their own values -- as they often find out to their surprise in times of crisis, when they find out that what they really care about morally is rather different from what they would have said they cared about. The life-long cynic suddenly finds that she feels she ought to risk her life for a child (whether or not she then does so is a separate question). The apparently duty-driven moralist suddenly finds that all he cares about or feels anything other than a public-relations reason to care about is his own skin. And so on; we could multiply examples like this almost endlessly. How then can empirical research possibly get us to people's actual values?

Since I have written extensively on the answer to this question elsewhere,(14) I will make only a few remarks here. For something to be one of a person's values, it seems to me, at least two things are necessary. (1) The person must feel shame or guilt or other 'moral reactive attitudes', as Strawson called them(15), if the values are not acted upon, and (2) the person must be willing to bear at least some costs to see them upheld. (I deal with the former idea in the aforementioned work,(16) but not the latter one.) If I am right about these two points, an instrument that has any hope of revealing values will put people in situations where the researcher can find out how a subject would react emotionally to a value being violated, and what kind of costs the subject would bear to keep that from happening.

I have no idea what sort of instruments could meet these requirements. Moreover, criteria such as I have just sketched for what a value is are not yet criteria for determining which particular values someone holds, given their linguistic and other behaviour. However, I can make some suggestions. Such instruments must be concrete. Not only do people's ethical reactions vary from situation to situation, depending on the size of the costs, risks and benefits involved and how they are distributed, but vague, general research instruments (questionnaires, for example) make it extremely easy for subjects to slide off onto some idealized image of their own values. The instruments must also introduce as much emotional reality as possible, so that the researcher can judge whether the person is actually apt to be willing to bear costs in situations in which the value being professed should be operative. People's most sincere statements of what they feel about things are notoriously open to error and self-protective deception. Finally, the researchers must know as accurately as possible the situation to which a subject is reacting when he or she expresses a value. (Even when a situation is specified, people can imagine their own situation and express the reaction they would have to it.) Finally, people's values are often extremely context-sensitive.

A last note on yet another topic. It seems to me that it would be interesting to correlate subjects' ethical positions on various environmental situations with some other things. One is variations in the subject's attitude to risk. It seems to me likely that situations in which a person is very averse to taking risks (e.g. with one's own children) are apt to evoke values quite different from those evoked by situations in which a person is quite willing to run risks (e.g., perhaps, a tiny extra increment of damage to the ozone layer). Another is a subject's attitude to the role of values in human action. I would expect subjects who feel that values have a large role in governing people's actions to have rather different environmental values from those who cynically feel that values are essentially epiphenomenal, such motives as self-interest and a desire to be accepted and fit in being the real motivators of human action. A final interesting study might be the correlation between people's values concerning the environment and the extent to which they feel that ethical values are the result of history and community. Those who feel that history and community play a major role in the creation of sound ethical values might be expected to have environmental values quite different from those of someone who feels that values are essentially an individual, atomistic phenomenon.

1. Here is what Kant actually said:

[R]eason has insight only into that which it produces after a plan of its own, ... [I]t must not allow itself to be kept, as it were, in nature's leading-strings, but must ... constrain ... nature to give answer to questions of reason's own determining. (Critique of Pure Reason, B xiii).

2. I recognize that this hydra-headed term has a variety of very different meanings in different contexts. The sense in which I am using it here is, I think, one of its most common senses.

3. The term 'intrinsic value holism' is my invention. No single term has stabilized as a standard way of referring to this form of deep ecology. The same is equally true, and true in both respects, of the next variant, which I call 'systems holism'.

4. Paul Taylor, Respect for Nature (Princeton, 1986), and Holmes Rolston, Environmental Ethics (Temple, 1988) are two well-known examples of this positions, examples moreover that show interesting variations on the definition of intrinsic value and its extent.

5. Callicott's early work and the Earth First movement are notorious examples of this position. J. Baird Callicott, "Animal Liberation: A Triangular Affair", Environmental Ethics 2 (1980), 319-321; "Non-Anthropocentric Value Theory and Environmental Ethics", Amer. Phil. Q. 21 (1984), 299-309.

6. Karen J. Warren is a well-known exponent of eco-feminism: "Endangered Earth: An Ecofeminist Perspective", forthcoming.

7. In Canada, John Livingston is the best-known advocate of this position. Cf. for example, "Ethics as Prosthetics", and my commentary, "Ethics and Survival", both in Environmental Ethics, ed. P. Hanson (Simon Fraser Publications, 1986).

8. Ludwig Wittgenstein, Philosophical Investigations (Blackwell, 1953). Hilary Putnam, "The Meaning of 'Meaning'" (1975), in Mind, Language and Reality Philosophical Papers Vol. 2 (Cambridge, 1975). This strategy for capturing word-meaning has subsequently received a lot of attention from cognitive scientists and psycho-linguists. It has also had quite a role to play in the attempt to remove some of the most egregious weakness of single-label diagnoses of psychopathology (cf. e.g. Nancy Cantor, et. al., "Psychiatric Diagnosis as Prototype Categorization", J. Abnormal Psych. 89 (1980), 181-93.

9. D. Secrest, G. Brunk, H. Tamashiro, "The Underlying Structure of Ethical Beliefs Toward War", J. Peace Res. 26 (1989), 139-52, and "Military Views of Morality and War: An Empirical Study of the Attitudes of Retired American Officers", Intern. Studies Q., 34 (1990), 83-109. I have changed the order of the authors to reflect the role Secrest has played in the project for which I am preparing these remarks.

10. Such an approach is coming to play a bigger and bigger role in the diagnosis of psychopathology. Cf. e.g. T. Widiger and K. Kelso, "Psychodiagnosis of Axis II", Clin. Psych. Rev., 3 (1983), 491-510.

11. Needless to say, I do not intend this description to be a full or even, necessarily, an accurate description of Rolston's position. I think it is approximately accurate, but I introduce it here merely to illustrate the point I am trying to make.

12. Rolston develops an account something like this, based on his notion of normative beings (and, later, normative systems), beings which can distinguish their good from their current situation and seek it (ibid.).

13. Developed in this minimal way, the notion of intrinsic value is close to Haworth's notion of treating things objectivity, i.e. "locating the significance of one's life in ... objects [themselves], rather than in their impact on oneself or on the groups with which one identifies", so that "the ways they are affected by our actions are relevant to the appraisal of those actions -- relevant finally, and not just in view o further ramifications for [people]." (Decadence and Objectivity (Toronto, 1977), pp. x and 153). Haworth argues that this stance is the stance we have to adopt "if we wish to live in a way that reflects the continuity of each [person] with the species [human being], with other species of living things, and with nature at large." (p. 158.)

14. The Needs and Values Project (Privy Council of Canada, 1975).

15. P. F. Strawson, "Freedom and Resentment" (1962), in Studies in the Philosophy of Thought and Action (Oxford, 1968).

16. pp. 89f.