What Happens if Ethics is Legitimate?
There is a longstanding debate regarding the status of ethics. Above and beyond the many disputes that have occurred within ethics, such as those over what sorts of actions are morally justified, there have been disputes over whether concepts such as moral justification are themselves rationally or practically justifiable. Debates of the latter type are often said to be metaethical debates, since they inquire into the fundamental basis of ethics as a whole. I take the central question of metaethics to be whether or not ethics is an ultimately legitimate domain of inquiry. The goal of this paper will not be to defend a particular answer to the central question, but rather to argue that if we answer the central question in the affirmative, there are concomitant obligations regarding how we should go about doing ethics- namely, that ethics should follow a model of inquiry closely associated with the natural sciences. The approach developed below for ethics also helps develop a moral for dealing with contrasting ethical intuitions.
2. What is an ‘Ultimately Legitimate Domain of Inquiry’?
Over time, humans have developed various fields of inquiry that each attempt to answers questions regarding a related subset of problems. For instance, biology attempts to answer questions related to organisms, cryptography questions related to codes, and astronomy questions about the cosmos. Any field that investigates a unified subject through the posing and answering of questions about that subject is a domain of inquiry. In the case of biology, the vast majority of organisms can be said to fall under a basic common definition, and exhibit some predictably similar properties and some predictably variant properties. Mathematics examines numbers, sets, and other mathematical objects, which similarly exhibit some predictably similar properties and some predictably variant properties. However, it seems that not all domains of inquiry are legitimate. For instance, it may be the case that a domain of inquiry focuses on some subject that doesn’t exist in the way it is described as existing within that domain. Astrologists might claim that the position of stars in the sky at the moment of one’s birth can have a causal impact on an individual’s temperament, but it seems as if celestial phenomena do not possess this sort of causal power. Another way a domain of inquiry can be seen as illegitimate is by assuming an inflated ontology. A polytheistic religion may be seen as illegitimate by a monotheist if the god of the monotheist has the same explanatory power as the gods of the polytheist. Self-contradiction also seems to indicate illegitimacy for a domain of inquiry- a mathematician researching problems in naïve set theory, while assuming unrestricted set composition, would be seen as operating within an illegitimate domain of inquiry.
A legitimate domain of inquiry must investigate a subject which does in fact exist in a manner at least minimally close to how it is described in the domain of inquiry as existing, it must have a justified subject of inquiry, and it must not be self-contradictory. The demarcation problem, which asks when science ends and nonscience begins, is similar to the problem of deciding which domains of inquiry are legitimate and which are not- and both problems share the strength of being intuitively straightforward and the difficulty of being formally elusive. This problem is perhaps more challenging in the present case because mathematical and logical domains of inquiry can be described as legitimate even by those holding antirealist views about the status of abstract objects. Since we are seeking to maintain the legitimacy of these fields while remaining non-committal about abstract realism, we cannot resort to saying legitimate fields are only those whose subjects exist in reality. The sieve we develop to sort between legitimate and non-legitimate domains must therefore focus more on the internal structure of a domain of inquiry than on the strict universal veracity of its existence claims. Economics might discuss the fluctuations in a ‘market’ and psychology of events in the ‘mind,’ and we do not want discussions about the existence of emergent entities to dominate our discussion of domain legitimacy for these fields. Our approach to demarcation is still limited, however, and at present we are dependent mostly on methodological heuristics, such as noticing that in legitimate domains of inquiry claims are made such that they could be defeated by claims within the same domain of inquiry, and also that over time there are fewer radical breaks from the bulk of knowledge in legitimate domains. The constraints on legitimate domains of inquiry sketched above are necessary but not sufficient conditions, and at this time a complete enumeration of all requirements for a domain of inquiry to be considered legitimate is beyond the scope of this paper. What is important is to understand that we do have some intuitive examples of legitimate domains of inquiry, and that they tend to have many internal structural similarities. If ethics is to be considered an ultimately legitimate domain of human inquiry, it must end up having theoretical purchase about whatever its subject is the way other legitimate domains of inquiry have theoretical purchase regarding their subjects.
When I say ethics must be an ultimately legitimate domain of inquiry, I mean just that ethics could one day hold the status of a legitimate domain of inquiry, and that there are no internal facts about ethics that would bar it, notwithstanding contingent problems, from so being. This indicates that the legitimacy of a domain of inquiry is not a pragmatic indicator of its effectiveness in reality, but rather a determination based on its subject matter and internal consistency. We would not call a domain illegitimate simply because those engaged in it had failed to produce positive results or develop a consensus, the same way we would not call a domain legitimate simply because it had able practitioners. While it is true that repeated failure to produce results or arrive at consensus oftentimes indicates a failure in the structure of a domain of inquiry, it does not by itself indicate this, and it is possible that a domain that has hitherto failed could ultimately succeed and demonstrate its legitimacy.
3. Two Disjunctions About Ethics
Once we have established the legitimate/illegitimate domain dichotomy, we must commit ourselves to describing ethics as falling into one of the two camps. Ethics, given its subject matter and its internal methodology, will in ideal conditions be able to sustain itself and build up a body of knowledge, or it will in all cases ultimately fail to do so. In this paper I will remain neutral as to which of these two options we should choose, but rather see what consequences might follow from assuming ethics is a legitimate domain of inquiry. Before I begin on that front, however, I should offer up a point of clarification regarding the nature of what I said was the central question of metaethics.
Metaethics is often said to focus on the realism/anti-realism debate, which asks whether or not moral properties exist in the world with mind-independence. However, as I mentioned above, claiming legitimacy for a domain of inquiry is not the same as being a strict realist about that domain. Many practicing mathematicians are anti-realists about abstract objects, but still operate in a domain that has a unified subject that becomes increasingly well understood as the field makes progress. Similarly, one could be a moral nihilist while thinking ethics is a legitimate domain of inquiry- this could be the case if the moral nihilist’s view of ethics was that of a field dedicated to the study of clarifying the strict consequences of certain actions and the subsequent ordering of our intuitions. Conversely, a moral realist could describe ethics as illegitimate- this could be the case if a fideist held that the word of god determined the one true ethics, but that any human effort to discover god’s will could never form a unified body of knowledge over time. While the legitimacy/illegitimacy debate is not identical to the realist/anti-realist debate, they are not orthogonal to one another with regards to ethics. I want to suggest that if ethics is to be viewed as a legitimate domain of inquiry, it will be better approached from a realist perspective. The argument for this rests on another distinction, which is also used to separate domains of inquiry. This distinction is between formal and empirical domain of inquiry.
4. Formal and Empirical Domains
All domains of human inquiry fall into two camps, being either formal or empirical. A formal domain of inquiry is one in which formal systems, each with its own set of constraints, have their properties investigated and contrasted. All properties in these domains follow with logical necessity from certain initial definitions, which are themselves established on the basis of certain pre-theoretical desiderata (or tradition). For instance, logic, as the study of validity, takes as its subject matter how conclusions can be said to follow from premises. To this end, different logical schemas are created by balancing out different considerations. For some logicians dealing with the liar paradox is of paramount importance, and so they might reject bivalence as an axiom. Others might be interested in truth-value parsimony, and so may accept bivalence to maintain a duality of truth values. The salient feature of formal domains is that they have purely internal standards of truth/acceptability, and, strictly speaking, no one system is more fundamentally correct than any other. Even Platonists about abstract objects would hesitate to deny this, since they typically do not claim that only ‘one true math’ or ‘one true logic’ exists, but rather that there is an infinite plurality of abstract objects from which different formal systems can pick and choose. A mathematical Platonist who (for whatever reason) did not allow the number i in their ontology would have a weak case against a mathematical system that did, since the Platonist has already permitted the existence of some abstract mathematical objects and would have to invent new and fanciful restrictions to disallow i. A thought experiment helps reveal the indeterminacy of ultimate truth for schemas in formal domains of inquiry. We can imagine an omniscient being who knows all facts about the world. If we were to meet such a being, we could still sensibly disagree with them about which schemas to adopt in a formal domain of inquiry, and their choice, while perhaps fulfilling more pre-theoretical desiderata (simplicity, consistency, explanatory power) would not be objectively ‘more correct.’ This is not the case with empirical domains of inquiry.
Empirical domains of inquiry take as their subject some specific set of phenomena in the world and attempt to provide a framework within which the existence and behavior of that phenomena is better understood. Physics, for example, will study the effects and causes of gravitational fields, and increasing knowledge in physics will progressively add greater apparent regularity to the behavior of gravitational fields. Just as the Platonist seemed to lack a deep argument against the mutually admissible existence of competing formal schemas, so an anti-realist about certain emergent entities lacks deep arguments against the mutually admissible existence of various empirical domains of inquiry. An anti-realist about the mental, for example, could strictly speaking deny the existence of the mind, or claim that its existence was strictly grounded in the physical, but would still not have provided a basis for thinking psychology was illegitimate, since it would still have some phenomena in nature for which it supplied greater regularity (even if this could be reduced to the behavior of other phenomena). Empirical domains must refer to things that exist in a way at least minimally close to the way they are described as existing within the domain of inquiry, and when different schemas compete within an empirical domain, they cannot both be right. To reuse the thought experiment from earlier, if we were to approach a being who knew all the facts about the world, we could not sensibly disagree with them about the existence of electrons, or who the victor was in an ancient battle. We could disagree definitionally, but so long as we were using the same terms to refer to the same world, disagreement would be hopeless.
5. Is Ethics Formal?
Let us take stock. So far, we have decided that a domain of inquiry can be either legitimate or illegitimate. Since the latter is simply defined as the negation of the former, (a domain is illegitimate just in case it is not legitimate), any domain must be one or the other. Furthermore, we have given the two models that legitimate domains of inquiry always follow- they are either strictly formal, in which case fundamental disagreement is always necessarily unresolvable, or they are empirical, in which case fundamental disagreement is always necessarily resolvable (at least in principle). So if we take it that ethics is a legitimate domain of inquiry (which we are simply assuming for the purposes of this paper) then it is either a formal domain or an empirical domain. Therefore, if we can demonstrate that is best not to view ethics as formal, we can show that if ethics is legitimate, it is empirical. In this section I want to present an argument for why we should not think of ethics as following the formal model.
A picture of ethics according to the formal model constitutes a fairly common approach to the domain. As was mentioned earlier, the schemas used in formal domains are established on the basis of pre-theoretical desiderata. In the case of ethics, the pre-theoretical desiderata are formed by our basic moral intuitions, and our additional intuitions regarding how to build an appropriate theory of ethics. One might believe that murder is never justified, and that an ethical theory is best built without any ad hoc exceptions- this would probably lead to an ethics that was a form of primitive deontology, since this would be an exceptionles prohibitive ethics in keeping with the pre-ethical desiderata. This process, that of building from a set of desiderata to a particular ethics, seems to be the most common formal model in ethics. On this view, various ethical schemas are usually proposed, and whichever best meet our various pre-theoretical desiderata (in this case, ethical and theoretical intuitions) “sticks,” and subsequently lays claim to being our best ethics. Negative utilitarians might argue with standard utilitarians by pointing out that more of the standard utilitarian’s intuitions can be explained by negative utilitarianism, thus providing more explanatory power (the desire for which is a common theoretical desideratum) and more ethical satisfaction (the desire for which is a common ethical desideratum).
But there seems to be a problem that emerges using this method. It is not appropriate to ask whether the axioms of a certain logical schema are true in an absolute sense, since truth is a predicate defined within the logical schema, which applies to the axioms definitionally, but not in an absolute sense. So the principle of non-contradiction is, in a way, no more true than its rejection. In constructing an ethics, the predicates we are most interested in defining the use of are typically ‘ethical’ and ‘unethical,’ and these are roughly analogous to ‘true’ and ‘false’ in a logic. Just as in the case of logic, we cannot say that any set of pre-theoretical desiderata are more or less ‘ethical’ than any other, since the predicate ‘ethical’ is only meaningful within the ethics itself. But the result is that an ethics must be said to be determined by desiderata that are not themselves ethical. In the above case, a primitive deontology against murder was formed on the basis of two desiderata- the intuition against killing, and the intuition that an ethics should not be ad hoc (not granting exceptions). While we may call a murder ‘unethical’ within that ethical schema, there is nothing more ethical about that schema than any other schema, since the desiderata themselves cannot be said to be ethical. This is problematic for two reasons.
Reason 1- Ethics is often tasked with resolving conflicting intuitions. If any set of intuitions can produce an equally ‘ethical’ schema, then there is no strict way one set of intuitions can defeat another, even if the former has greater explanatory force with fewer rules. This is because even the theoretical desiderata of parsimony and explanatory power are not more objectively ‘ethical’ for an ethical theory.
Reason 2- Logic and mathematics are able to define their subject completely. For instance, a logic can completely define its logical symbols and operators, and a mathematics can define the concept of number and the concept of succession. However, ethics must take concepts from the real world, concepts such as that of a feeling subject, murder, and persons, to name a few. Since ethics cannot exhaustively delimit its own subject in an internal manner, its inability to objectively resolve disputes between various ethical schemas signifies an inability to resolve claims made about things in the real world. So given a particular action, such as a robbery committed on a specific day, ethical schemas will disagree over the ethicality of that action while being unable to conclusively justify any one verdict amongst themselves. This would leads ethics to be indecisive in making formal claims about empirical events, and so ethics would struggle to positively add to our description of physical reality- pointing towards a potential illegitimacy.
Two early attempts at avoiding this problem will fail, one in virtue of being circular, and one in virtue of inducing an infinite regress. The first attempt sees us claiming our desiderata are ethical by making them assumptions in our ethical schemas. Hence, in the case mentioned above regarding a primitive deontology against murder, the ethics would actually read: ‘One must not kill in any case, with no exceptions, and it is permissible to form an ethical theory on the basis of the intuition that it is not permissible to kill and a resistance to ad hoc theorizing.’ However, this would simply mean the desiderata are justifying themselves, which is circular. One might try another path, which would be to have metaethical predicates justifying the desiderata. However, these in turn would be motivated by metadesiderata that would need to be justified, and so a metametaethical set of predicates would be required, and on ad infinitum.
Once we have established the problem with thinking of ethics as adhering to the formal model, where it is simply a battle between intuitions serving as pre-theoretical desiderata, we can move on to considering ethics under the empirical model.
6. Empirical Ethics
We have demonstrated that, for ethics to be an ultimately legitimate domain of inquiry, it must be empirical- but what would this look like? To discover this it is important to note an important commonality that holds between empirical domains.
Empirical domains each take certain phenomena in the world, and attempt to explain their behavior so as to make them better understood. This usually takes the form of adding greater regularity to our expectations about these and other phenomena. One who has no conception of gravitational fields will have a harder time plotting the orbit of celestial bodies than one who does, all else being equal. Sufficient study of frogs will, if herpetology is legitimate, cause us to be less and less surprised by the behavior (and properties) of frogs. So then what phenomenon is explained by ethics? I tentatively suggest that it is the sense of appropriateness of evaluating specific states (axiology) and the sense of appropriateness of attributing guilt and praise to individuals. Once we have a more well developed ethics, we will have a greater degree of regularity regarding our sense of when it is proper to positively/negatively evaluate a certain state and/or praise/blame an individual.
This definition of ethics helps solve two problems. The first problem is how to make sure ethics is capable of resolving conflicting intuitions, a feature of ethics we saw was lacking in the formal approach. Once ethics has as its goal the regulation of the sense of appropriateness of our evaluations and attributions, it can start to address related intuitions by seeing how these sensations stack up against various forms of argumentation. Two individuals, feeling differently about evaluating a certain state positively, can test their respective positions, and see how these fare against each other. This process, if repeated frequently, can begin to shed light on the nature of these intuitions. Even if these individuals persist in their disagreements, we will have learned something- namely that there may be a fundamental inability to resolve the dispute. This is an advance from the formal model, where for any desideratum whatsoever one ethics could be favored over another- in the empirical model there must be a demonstrable inability to ultimately get over the disagreement, in which case the possibility of resolution is defeated, if not the disagreement itself.
Another problem that is solved by this approach to ethics is whether or not we should be realists about ethical properties. In the case of the formal model, we saw that both realism and anti-realism were permitted. In the empirical case, we must side with realism, but from deflationary considerations. Since the subject of ethics is simply the sense of appropriateness of attributions and evaluations, whether our ethics is realist just depends on whether or not we admit that these sensation of appropriateness are real. Since they clearly are (we undoubtedly feel them every day) our ethics is realist. This is not because we hold that abstract moral properties exist, but simply because the explanandum of ethics is now clearly extant. The predicate ‘ethical’ is now just used to describe those states/individuals we ultimately feel it is appropriate to positively evaluate/praise. The predicate ‘unethical’ will be used in the opposite sense, and ‘non-ethical’ will be used for things where the feeling is neutral.
This ethical schema is certainly not conclusive, and it is just one example of an empirical ethics. Some of the key difficulties in constructing ethical theories are circumvented by this approach, and it opens up the possibility that ethics can follow in the footsteps of the natural sciences. Just as science can study implicit bias and motion tracking in human beings, so it could progressively gather more data on feelings related to evaluation and praise, helping us to slowly build a model of the general human ethical picture of the world. If this model of empirical ethics is taken seriously, there is a moral to be drawn regarding one sort of conflict between ethical intuitions. Oftentimes individuals will admit to principles that in local, isolated instances seem unobjectionable to them- but when the consequences of following these intuitions are shown on a larger scale, there is disagreement. We will say the focused intuition is acting in a local setting, whereas the latter setting is called global. If we were using the formal model, the conflict here would present a challenge- we would not be certain which scenario to privilege in building our theory, if not neither or both. They would seem to have equal weight, and no theory built from casting out the intuition due to global concerns could strictly defeat a theory built on the basis of local support, nor vice versa. However, once we use the empirical model, the balance tips in favor of local considerations. Science has repeatedly shown that humans have trouble extrapolating folk intuitions to settings much larger than those evolution adapted us to survive in. We can have an intuitive sense of small distances and times, but timescales appropriate for cosmology are deeply unintuitive for human beings. There is a similar phenomenon for ethics. For instance, one might hold that it is unethical to cause unnecessary suffering, and base this feeling on limited intrapersonal cases. However, when they realize that this may imply voluntary extinction, there is typically an about-face, where the massive consequences of the intuition seem to defeat it. Yet there is no way to tell how someone would feel it they were to have a deep, complete knowledge of what reality would look like given voluntary extinction. Since their local perspective is far more informed than is the global one, and since we are looking at ethics as being ultimately legitimate (meaning after more or less maximal reflection), we should imagine the local intuition as being at present more well explained and tested than the global, and thus the feelings of blame and evaluation it generates more valuable. In short, empirical ethics implies, at present, following our locally generated intuitions more assuredly than our global ones.
 This is further problematized by the fact that another desiderata, that of the intuition that we should justify our desiderata, would have to be added to the schema. This would in fact by itself induce an infinite regress.
 For this motivating thought, I am indebted to a similar line of argument found in Every Thing Must Go by Ladyman and Ross (Chapter 1). Their naturalism is focused on metaphysics; I roughly extend it to ethics.
 Including on large-scale consequences, which practically are difficult (if not impossible) to reflect on.