MORAL REALISM AND ITS ALTERNATIVES

In this paper I shall try to classify what kind of metaphysical entity morality could be, and discuss the strength and weaknesses of each alternative. In order to do so, I shall start with a definition of Moral Realism (Section 1) before considering the anti-realist alternatives. My first claim will be that there is not one Anti-Realism but a variety of different theories (Section 2). I shall then argue that one important theory is commonly overlooked, and offer an amended classification of moral theories (Section 3). I shall then contrast my classification with alternative ones in order to sharpen the distinctions (Section 4). Finally, I shall discuss the main strengths and weaknesses of each moral theory (Section 5). My conclusion will be that the neglected alternative is worth serious consideration, but that we should also reflect upon the more general question of why we have not found the one and only theory yet.

What is the metaphysical basis of morality?What kind of entity is a moral value or the origin of a principle?If one says, for instance, that the environment has a moral value and should be preserved, what kind of entity is this value?Value could be a property of the environment, in the sense that the environment has this value property even if there is no being that could value it.Alternatively, the value of something might merely be what an observer values, and nothing more.Finally, there might not be such a thing as value anywhere; we might just be in error about value statements.Do these options exhaust all the possibilities?
In this paper I shall try to classify what kind of metaphysical entity morality could be, and discuss the strength and weaknesses of each alternative.In order to do so, I shall start with a definition of Moral Realism (Section 1) before considering the anti-realist alternatives.My first claim will be that there is not one Anti-Realism but a variety of different theories (Section 2).I shall then argue that one important theory is commonly overlooked, and offer an amended classification of moral theories (Section 3).I shall then contrast my classification with alternative ones in order to sharpen the distinctions (Section 4).Finally, I shall discuss the main strengths and weaknesses of each moral theory (Section 5).My conclusion will be that the neglected alternative is worth serious consideration, but that we should also reflect upon the more general question of why we have not found the one and only theory yet.

Section 1: Moral Realism
What is moral realism?Before I shall try to argue for one suggestion, it is important to specify the question a little further.There are different levels of moral inquiry.One can deliberate about what the right thing to do is in a particular situation, try to find a general principle with which one can decide any moral question, or think about the nature of moral values and prescriptions in general.For instance, it is one thing to ask whether one should be a vegetarian, another to postulate a principle that states which beings are to be respected and how, and a third to ask about the nature of this prescription.The middle level, where one tries to find a general principle about what is right and good, is usually called 'normative ethics.'This field is the host to the discussions between Utilitarianism, Deontology, and Virtue Ethics.One might argue that the right principle is to maximize happiness, for instance, or the Kantian principle that one should never treat anyone merely as a means.While both of these principles are about the question of what one should do, Virtue Ethics asks the more general questions of what kind of being one should become.The right principle might then be to develop an excellent character.
A second level of moral inquiry has usually been called 'applied ethics.' The idea was that one could settle the question of normative ethics, and then just apply the right principle to particular situations.But since there is neither agreement on the normative level, nor does one principle seem to fit the complexity of moral life with its very different situations, the level is now more properly called 'practical ethics.'It addresses such questions as famine relief, protection of the environment, euthanasia, vegetarianism, or whether it could be allowed to torture a terrorist who knows the location of a hidden bomb.
The topic of Moral Realism is addressed on a third level.Even if one has figured out what is the right thing to do in practical ethics, and even if one could subsume all the answers under one principle in normative ethics, the question still remains what is the nature of this principle, and what kind of entity morality is.For instance, does this principle describe an objective reality 'out there,' or is it merely an expression of one's preferences?This third level is usually called 'meta-ethics.'In meta-ethics one applies the fields of theoretical philosophy to the subject matter of morality.One can ask about the metaphysical nature of moral entities, one can ask the epistemic question of how one can know moral propositions, one can ask questions about the logic or proper language of morality, or one can address issues in moral psychology, and ask about the proper moral motivation, among others.
Moral Realism is an issue in meta-ethics.However, there is not one clear definition of what should count as Moral Realism.A popular definition was given by Geoffrey Sayre-McCord when he stated moral realism as consisting in the following two theses: "(1) the claims in question, when literally construed, are literally true or false (cognitivism), and (2) some are literally true" (SAYRE-MCCORD, 1988, p. 5).According to this definition, the distinction between Moral Realism and Anti-Realism is a semantic one.Moral realism would be a form of cognitivism, and Anti-Realism would be confined to non-cognitivism.However, recently there has been a trend to abandon the semantic definition in favor of an ontological or metaphysical one (cf. FITZPATRICK, 2009).The problem with the semantic definition, as it seems to me, is that too much would count as moral realism.Imagine a view according to which moral judgments report one's inner state of feelings.For instance, to say that something is morally wrong would describe that one has a feeling of disapproval.Such a description is truth-apt, and could be true or false.However, to call such a view 'Moral Realism' might conceal what most people find attractive about a realist position: that there are moral entities 'out there,' which are true independently of our feelings.So one needs to specify how one uses the label 'Moral Realism,' and one has to analyze carefully how another uses it before engaging in a debate.
In this paper, I shall follow the recent trend and refer to 'Moral Realism' as a distinct metaphysical position about the nature of morality.On this view, a Moral Realist is someone who holds that morality exists 'out there,' independently of any human perspective.This moral entity could be a substance, a property of a substance, or a separate moral realm which is neither physical nor psychological.For instance, if one says that a tree has a moral value, and should not be cut down, one might hold the view that a tree -in addition to its cells and fluids that make up the biological treethere is also another entity or property called 'moral value.'It could be that there is a precious entity in the tree, or the biological tree might have, in addition to its natural properties (such as density, color etc.), also a value property.It is no objection against this view that we do not see the value of the tree if we cut it open.Value might not be a natural property that one can discover with the five senses, but it might be a non-natural property that for which one needs a special faculty of intuition to discover it.1Finally, it might be stated in a moral realm that is different from the physical and the psychological, that trees are valuable or should not be cut down.
In addition, it is not necessarily the case that this value exists 'out there,' independently of human beings.It could be that only human beings have an absolute moral value.So if there would be no human being, there would be no value on such an account, but it does not mean that value is just what we value or desire.It could be that every human being really has such a value independently of one's feelings.A better metaphysical definition of Moral Realism is therefore that morality exist independently of any stance or perspective a human being might take.This is Russ Shafer-Landau's definition, and in this paper, I will endorse this definition of Moral Realism as "stance-independence" (SHAFER-LANDAU, 2003, p. 15).

Section 2: Varieties of Anti-Realism
If one rejects Moral Realism, as I have defined it above, there is more than one alternative.It is tempting to work with a simple dichotomy, such as Realism against Anti-Realism, or -as Mackie did -Objectivism against Subjectivism (cf.MACKIE, 1977).But in doing so, one is likely to gloss over important differences.For instance, if one does not believe that there is an objective moral reality in the sense of my version of Moral Realism, morality could be subjective in the sense that it is constructed by human beings.But one might not want to hold that morality is subjective in the sense that it is constructed by one human being, in contrast to being created by a group, and therefore there is room for another view, which one might call Inter-Subjectivism.But since there is more than one view of Inter-Subjectivism, I shall for now merely call this view 'Constructivism,' and specify the different forms later.In addition to Realism and Constructivism there is at least a third view.One could deny that there is any such thing as morality.This view can be called 'Nihilism.' The three-fold distinction, between Moral Realism, Constructivism and Nihilism has been suggested by Russ Shafer-Landau (cf.SHAFER-LANDAU, 2003, p. 13-8).Realism, as noted above, is the view that morality is stance-independent.For a Constructivist, on the other hand, morality depends on a certain stance or perspective of a particular being or class of beings.Finally, Nihilism holds that there is no morality, as either stanceindependent or constructed by (human) beings.Within Constructivism Shafer-Landau offers another three-fold distinction between, what he calls, Subjectivism, Relativism, and (Kantian) Contract Theories (cf.SHAFER-LANDAU, 2003, p. 39-41).Subjectivism is the view that morality is constructed by the attitude of one human being, whereas Relativism locates morality in the actual agreement of a group.Kantian Contract theories (such as the ones by Rawls and Scanlon) view morality as constructed not by the actual agreement of a group, but by what one would choose under ideal circumstances, e.g., if one would adopt an impartial standpoint, or if one would have to design a system, but does not know which place one will occupy in that system.The hope is that -under these circumstanceseveryone would choose the same moral principles, even if they make the choice based upon their own interests.
Section 3: The neglected alternative Shafer-Landau's classification is an extremely helpful overview over contemporary positions.But is it complete?I shall argue that it does not address Kant's position.2My aim is not to do Kant exegesis here (for that cf.SENSEN, 2013), but imagine a position according to which morality is a constitutive principle of one's own reason.By this I mean that the principle describes how a faculty necessarily functions.The principle is an operating principle of human reason, and whenever one reasons, one has to do so in accordance with that principle.A parallel would be a principle of contradiction which states that something cannot be the case and not be the case at the same time and in the same respect.Whenever one reasons, one has to employ this principle.One cannot reason in contradictions, and the principle describes how one has to reason.As such it is constitutive of what it means to employ theoretical reason.Similarly, the moral law might be a constitutive principle of practical reason.It is a dictate of one's own reason.By this, however, is not meant that one creates this law in one's reasoning, as one would adopt a New Year's Resolution.The law is not a product of one's awareness, but it is there already prior to one's awareness, and guides one's awareness in thinking practically.
How should one classify such a position?It is not a form of Nihilism, since it holds that there are moral commands.It is also not a form of Realism, since morality does not exist independently of the stance of human reason.If there were no human reason, there would be no morality.Therefore morality is dependent upon a particular human stance.Is it Constructivism?It does not fit with any of the three versions of Constructivism mentioned so far.It is not a Subjectivism in the sense that it is the product of an individual's desires, and that different individuals would come up with different moralities.The moral law, on Kant's account, is unconditioned, universal and not a product of one's desires.But Kant's position is, second, also not a form of Relativism.The moral law is not the product of an actual agreement of a group, and does not differ between different people or groups.Each individual is already fitted with the same law.Finally, Kant's position as I have sketched it above is also not a Contract theory.For the moral law, according to Kant, does not depend on the conscious choosing of an individual under ideal circumstances.Rather the moral law is already given prior to any choosing, and is that which automatically figures into our choosing (although we might not act on it).
How, then, should one classify this position?I think that it should be a sub-form of constructivism,3 but a new form.It is a necessary, preconscious construction on the part of pure practical reason.I therefore call it Constitutivism (cf.REATH, 2006, p. 176-180 andKORSGAARD, 2009), but it is not an empirical form in the sense that one constitutes one's agency by acting and choosing.Rather it is a position according to which there might be an a priori, necessary law prior to one's awareness that guides one's moral reasoning.One could therefore call it 'A Priori' or 'Transcendental Constitutivism.'In order to make room for this new position, it will be easier if one not just adds another position to Shafer-Landau's versions of Constructivism, but if one replaces the three-part distinctions with a string of dichotomies.The model I want to suggest is a tree structure where at each junction one branch ends, and the other branch splits up into another dichotomy.For each dichotomy I shall pose one question.If one answers the question with 'no,' the branch will end there.If one answers the question with 'yes,' one should move to the next question.One can classify moral theories according to the following structure: (1) Does morality exist?Is there such a thing as morality?If one answers 'no,' one is a Nihilist.If one answers 'yes,' one should move to the next question (2).
(2) Is morality a construct of (human or similar) beings?If 'no,' one is a Moral Realist; if 'yes' go to (3).
Is this classification exhaustive?Does it cover all possible positions?Probably not, but this amended classification has three advantages over Shafer-Landau's helpful distinctions.My suggestion covers more positions, has a better chance of being exhaustive by using dichotomies, and it can easily be expanded without breaking its structure.This can be demonstrated with a test case.How should one classify Evolutionary Ethics?Under this heading I understand a theory that holds that our contemporary sense of morality is the product of evolution.Whereas in early human history different groups had different codes, the ones with a sense akin to our contemporary views on morality had a better chance of survival, and therefore these traits survived.Imagine a position according to which morality, i.e., (i) a sense of respecting others, (ii) of helping them in need, as well as (iii) a motive for doing the right thing simply because it is right.Imagine also that these three elements were useful for the survival of a species.Groups who had this trait were more likely to survive for instance famines, or attacks by wild animals than groups who did not possess these characteristics.Over time these traits might have become part of the genetic code for the descendants of the group.A modern descendant would therefore find (i)-(iii) as an innate sense. 4Where would a position like this fit into the above dichotomy structure?
It seems to me that it depends on the time slice one is focusing on. 5It might well be that at a certain point in time adopting morality was a conscious decision by a group of people.In that case, an evolutionary explanation of morality would be an instance of Relativism, as it is defined above.However, for the contemporary descendant morality is not any longer a conscious decision, but members of this group find themselves having these traits innately.As such, the theory has more similarities to Kant's Constitutivism.But it is not the same as Kant's position, since the moral elements (i)-(iii) would not have arisen out of pure reason, but from one's genetic code.Morality would not be the autonomous product of pure reason, but the learned and stored experience of human history.One therefore would have to add another dichotomy, ( 6): (6) Does morality arise pre-consciously out of one's own reason?'No': Evolutionary Ethicist -'Kant.' The question of how one should classify Evolutionary Ethics therefore depends on the time slice one is looking at.At the beginning it might have been a form of Relativism, but over time it turned into a form of (non-Kantian) Constitutivism.And there could be other versions of that are not yet captured by my classification, but one just would have to adjust the questions in order to get a more fine-grained distinction of moral theories.However, even if the classification is incomplete it has already yielded an important new insight.Constitutivism is a viable candidate, and one cannot establish Moral Realism by rejecting Subjectivism, Relativism, and Contract theories alone (cf.Section 5 below).Even if those three fail, it could be that Constitutivism is true.

Section 4: Alternative classifications
There are, of course, other ways of classifying moral theories, and it will be helpful to compare the above dichotomy structure to two others.The first takes into account Stephen Darwall's second-personal standpoint, the second Colin McGinn's DIME structure.Both classifications have distinct advantages, and together they will help us to sharpen the classification of realism and forms of anti-realism.I shall start with the second-personal standpoint.
Stephen Darwall has argued that morality contains essentially secondpersonal elements (cf.DARWALL, 2006).Morality describes the standing human beings have to (1) make claims on others, and (b) to be treated in a second-personal way.For instance, after an injustice, the victim has the standing to demand reparation and an apology, but it also can demand a special way of being treated, e.g., that the culprit looks the victim in the eye when apologizing etc.One could develop a classification of moral theories out of Darwall's insight, distinguishing between first-, second-, and thirdpersonal theories.A first-personal account would generate morality out of the subject-or the 'I'-perspective.A second-personal account would argue that morality consists in a relationship between or two more people or a 'You'perspective.A third-personal perspective, in contrast, talks about an 'It' or a feature of the world. 6Finally, it could be that there is no morality.This way of classifying moral theories does not necessarily generate different candidates of moral theories, but it would group them differently.
If morality is not tied to any of the three standpoints, it does not exist, and Nihilism would be true.If, on the other hand, morality is a fact from the third-personal standpoint, an impersonal fact 'out there,' one holds a Moral Realism.From a second-personal standpoint, one could distinguish whether someone holds morality to consist in an actual agreement of two or more people (Relativism), or an ideal agreement between people (Contract theories).Finally, morality might be said to arise from a first-personal perspective.If it arises a posteriori from one's actual preferences, this would be a Subjectivism.If it arises a priori necessarily out of one's own reason, one would hold Kant's Constitutivism.As in my earlier classification, an Evolutionary Ethics would fall into different categories, depending on the time slice one is talking about: When morality was useful for the earliest humans, it might have originated out of a second-personal agreement of a group, whereas for the descendants of these humans, morality is innate and seems to originate necessarily from a first-personal perspective.One would then need to differentiate between a necessary first-personal origin of morality, depending on whether it generates in one's genes or one's own reason.Here one could branch out and distinguish between a first-personal morality that emphasizes genes, feelings, a Freudian super-ego, or reason etc.
This classification therefore can yield and accommodate the same theories as my adapted Shafer-Landau model, but it is less homogenous.Whereas my adapted model distinguishes the theories in one flow by using dichotomies, the perspective model needs further incommensurable distinctions -between actual and ideal agreements for second-personal theories, and empirical or necessary origins for first-personal accounts of morality.

The Dime structure
A third way of classifying meta-ethical theories into Realism and forms of Anti-Realism can be gleaned from Colin McGinn's DIME structure (I am following loosely MCGINN, 1993).McGinn argues that debates in philosophy always oscillate between four main positions, but that there is one neglected alternative.He uses 'DIME' as an acronym that is composed out of the first letter of the four standard positions: 'D' stands for 'Domestication.'It expresses a reductive explanation of a phenomenon in terms of something one knows better: e.g., one could hold that the mind is just a function of the brain, or love is just a flow of hormones.'I' is the abbreviation for 'Irreducibility.'One might hold that phenomena that are puzzling neither can have nor need an explanation -they are irreducible phenomena.One could argue, for instance, that the mind is the reality through which we perceive everything else.Even the brain is only discovered and observed through the mind.The mind is then a first reality without which we could not have the conception of a brain.But we cannot then refer to the brain to adequately explain the mind.
McGinn uses 'M' for 'magical,' but a more neutral term would be 'metaphysical.'One might explain a phenomenon by postulating a new entity: The mind is a non-material substance or soul, numbers are real entities in a third realm of existence that differs from the physical and psychological.Finally, 'E' stands for 'elimination.'One might simply deny that the phenomenon to be explained really exists.Maybe there is no consciousness, love, or free will.
McGinn argues that debates in philosophy -whether they are about the existence of God, free will or numbers, the nature of the mind or knowledge -always oscillate between positions that can be captured with the DIME structure.He also points out that there is a neglected alternative, and defends it for each topic.He calls the alternative 'Transcendental Naturalism,' and it works with a cognitive closure conception.The idea is that there is a solution and explanation for every philosophical problem, but that it is in principle impossible for human beings to detect the answer.Our cognitive capacities are simply not up to the job.The solution might be obvious to beings with different faculties than ours, but it is epistemically closed to us.
McGinn himself does not apply the DIME-structure to the nature of ethics.But inasmuch as one asks a metaphysical question about morality, one can take the same stances the DIME-structure describes.For instance, one could deny that there really is any such thing as morality, and eliminate the phenomenon: Nihilism ('E').On the other end of the spectrum, one could argue for an M-position and say that in addition to physical and psychological realities there is also a distinct moral reality such as nonnatural value properties are a separate moral realm: Realism.In contrast, one might adopt a middle position and argue that -although morality exists somehow -it is not a reality independently of (human) beings, but is constructed by them.There would be no morality without these beings, but morality exists with these beings.This would be a domestication of the moral phenomenon ('D'), and it can come in very different forms: One could reduce morality to the preferences of the individual (Subjectivism), to the actual agreement of a group (Relativism), to the ideal agreements of a group (Contract theories), or to a constitutive principle (Evolution or Kant's Constitutivism).
While this classification can cover the same positions as the other two classifications we considered it also offers to more possibilities to understand the moral phenomenon: It could be that it is a first reality, both incapable and not in need of any further explanation: Irreducibility (cf.PRICHARD, 1912).Alternatively, it could be that there is an explanation of the moral phenomenon, but that it is in principle inaccessible to human beings: Cognitive Closure.These two do not determine the ontological nature of morality, but the epistemic limitations of our inquiries.Irreducibility and Cognitive Closure might explain why we are still debating this question without universal agreement.
The DIME-structure is therefore very close to my classification in that it groups the positions in a similar way: Constructivism is a form of Domestication or reduction, and can itself have different forms.Another thing we can glean from McGinn are the advantages and disadvantages each position has (again my discussion is loosely based on MCGINN, 1993).Realism or an M-position takes seriously the phenomenon one is examining.On the other hand, it has to introduce a further entity: a soul or numbers, for instance, but it is not always clear how to conceive of such an entity -it is very unlike everything else we can see and touch -and how one can discover it.In addition, it is not clear whether introducing a new entity really solves the initial puzzle one had.If it is not clear how a brain could have something like a self-consciousness and phenomenal experiences, introducing another entity does not by itself explain such experiences.Self-consciousness seems to be a first-person experience.But a soul would be a third-person objective fact, similar to the brain.D positions can account for the phenomenon as well, they do not need to introduce new entities, and they do not face the new problem how the new entity can account for the phenomenon.While these are all advantages, the disadvantage of a Dposition is that it might not be able to adequately describe the phenomenon: If, for instance, one says that love is just a play of hormones, it does not capture what it feels like to be in love.Or if one holds that having free will is just that one acts without coercion and compulsion according to one's desires, it does not adequately capture the phenomenon of free will where it seems that we can act independently of any desire.In reducing the phenomenon to something else, the distinctive quality of the experience might be destroyed.An E-position goes even further in that direction.It eliminates the phenomenon, and would hold that there is no love, free will, or mind etc.It would be like going to a dentist because one has a tooth ache, and the dentist says: 'You don't have a tooth ache.'But an E-position does not face the problem of explaining the phenomenon, and it has the disadvantages of the M-and D-positions as its advantages.These general pros and cons can be found in the more concrete discussions in moral theory.I will confine myself some of the main, exemplary lines.
Section 5: The arguments So far I have argued that there is not one clear dichotomy between Moral Realism and Anti-Realism, and tried to classify the different positions.But which of them is right?Can one demonstrate that one is the correct position and that the others are wrong?There is no consensus.What is more, it is very hard to give positive support for one view.The debate is largely to attack an opponent, point out the superiority of one's own position, and shift the burden of proof to the opponents (cf.SHAFER-LANDAU, 2003).One important criterion in the debates is whether a theory can capture the moral phenomenon.But, of course, there is also no agreement about what exactly the moral phenomenon is.One possibility is that something falls in the domain of morality if it concerns the rights of others.However, it seems that not everything that concerns the rights of others is a matter of morality (e.g., if one studies the history of the rights of others).And conversely, one can argue that not everything that is a matter of morality falls under the rights of others.There might be duties towards self.
What, then, is the moral phenomenon?The previous candidate proposed one specific content of morality (the rights of others).I shall therefore suggest that one should classify the moral phenomenon in a formal way.On this account, the moral phenomenon is that one ought to do something independently of what one wants.A familiar example of such an experience is a bad conscience.One might have acted the way one wanted to, but still one is left with the sense that it was morally wrong to do so.Prior to the action one might have the sense that one should not do the action, even though no desire speaks against it.How well can the different positions capture this moral phenomenon?
At first glance, an M-position seems to be able to explain the moral phenomenon very easily.If morality is grounded in metaphysical properties (such as a non-natural value) then there is a standard independently of our desires which could constrain them.However, Moral Realism has been subject to the same criticisms that M-positions more generally are subject to.On the one hand, the M-position has to make further assumptions about the existence and cognition of moral properties which are not always vindicated.On the other, it is not clear that these new entities can capture the phenomenon they were supposed to explain.The first kind of criticism was famously raised by Mackie.He argued that the kind of properties moral realism would postulate are 'queer ' (cf. MACKIE, 1977, p. 38-41).They would be properties unlike any other we know.In addition, one would need a special faculty to discover these properties which again would be unlike every other faculty we know.The second sort of criticism questions whether Moral Realism could even capture the moral phenomenon.If there are moral properties 'out there' what obligates me to follow them?
A Moral Realist would not be disheartened by these objections (for a defense see SHAFER-LANDAU, 2003 andSCHÖNECKER 2006).Regarding the argument from queerness the Moral Realist could bite the bullet or look for partners in crime.The realist could accept that morality is just a special phenomenon that needs its own kind of explanation.Or the realist could point out that that mathematical entities (numbers etc.) as well as advanced scientific concepts (such as quarks, strings, DNA, black holes, ten dimensions) are unlike everything else we know.If that does not discredit these concepts in mathematics and science, why should it discredit the existence of queer moral properties?The critic could come back and hold that scientific concepts are still thought to be natural properties, while moral properties would be non-natural.But it is not clear that this distinctions hold.There could be so-called scientific concepts which we will never be able to perceive with our five senses (maybe a black hole absorbs all light and will never be visible).In that case, we can only infer from what we see that these scientific entities do really exist.But a Moral Realist could hold that this is the same for moral properties.
Second, the moral realist could respond to the bindingness challenge by arguing that moral properties are inherently motivating and binding (contemplating the good might cause a respect for the properties), or the moral realist can separate bindingness and motivation, and admit that moral properties might not motivate the agent, even if the action would still be the moral thing to do. 7he Anti-Realist, by contrast, might hold one of several positions.He might be a Nihilist or a Constructivist.As a Constructivist, he might be a Subjectivist, Relativist, a Contract theorist, or a Constitutivist.What speaks in favor and against each of these theories?There is a spectrum for both the advantages and disadvantages.The Nihilist will make the fewest assumptions (advantage) but will be least able to capture the phenomenon (disadvantage), while the different forms of Constructivist can progressively capture more of the moral phenomenon (advantage), but can do so only at the cost of more assumptions (disadvantage).Think again about the example of the tooth ache.A nihilist about tooth aches would deny that there is any such thing as a tooth ache.He does not have to deny that the patient feels something, but it is not a tooth ache.The patient errs.The Constructivist would admit that what the patient feels is a genuine tooth ache, but he would explain the ache in term of something else, e.g., tooth aches are just the anxiety of going to a doctor etc. Similarly in morality, the Nihilist does not have to make any assumptions, but he cannot account for what the moral phenomenon feels like.
The Reductionist or Constructivist, on the other hand, can admit that there is a genuine moral phenomenon.It just is not quite what it seems.Also within Constructivism there is a spectrum regarding the assumptions a theory has to make, and how well it can account for the phenomenon.For the Subjectivist morality is just a personal preference.(This does not, of course, prevent the phenomenon from being internalized, and appearing as if it is an absolute command.However, deep down it is not.)It is like saying to the patient at the dentist office that -while he does have a tooth ache -it is merely the anxiety of going to a dentist.The phenomenon is reduced to something else seemingly less problematic.However, the Subjectivist might see this as a strength.One does not have to postulate queer entities to explain the moral phenomenon, but there is a natural, non-spooky reason for it.One does not have to assume anything beyond personal preferences and desires.The Subjectivist has the added strength that it is easy for him to explain moral motivation.Morality should be able to motivate, and bring about change, and this is easy to explain if morality is based on one's own preferences.
However, it would change what the moral phenomenon is really like.For the agent morality seems like a demand that restricts one's desires, not something that is merely another desire.But it would also change what morality is like for a group, such as having a moral argument.If one has a moral argument, it does not seem that one merely expresses one's desires and tries to influence the other (cf.SHAFER-LANDAU, 2003, p. 22-7).Rather one seems to be concerned about finding an independent truth.Subjectivism about morality therefore exhibits the same strengths and suffers from the same problems as a D-position in philosophy more generally.It can explain the phenomenon, but it seems to change it somewhat.
In contrast to the Subjectivist, a Relativist can more easily explain why morality seems to override one's desires.If moral rules are actual agreements of a group, the desire of the individual might often be against that.However, the Relativist offers a more demanding explanation than the Subjectivist -in addition to one's preferences there is also the agreement of a group -and he loses something when he tries to explain moral motivation.One needs to be motivated by the desire to benefit from the group, not any desire one individual has in isolation.In addition, Relativism has the specific problem in the way it changes the moral phenomenon.It does not seem that morality is just the agreement of any group.We admire most the dissidents against unjust regimes, who protested the existing morality in name of a higher code.
Progressing further on the spectrum, Contract theories are better able to capture the idea that morality can be against any of one's desires, whether it be the preferences one has as an individual (Subjectivism), or the ones one has in relation to a group (Relativism).If morality is based on what one would agree to under ideal circumstances, for instance, from an impartial standpoint, or if one does not know in advance which position one will occupy in the new order, morality is less grounded on one's individual desires.Again, one would still endorse what is right based on one's preferences, but under ideal circumstances the results of what people will endorse would converge.The explanation is now more demanding, but it is still a natural explanation which does not have to make queer assumptions.
However, Contract theories face two special objections which are peculiar to interest-or desire-based Constructivism.The first objection is a question about moral motivation, and the second objection is whether this is even a stable position.The first objection is the problem of a free-rider.Under ideal circumstances, one can endorse that everyone should pay for a ticket in order to use a train, for instance.So, one would judge that it is right for everyone to buy a ticket.However, why should one follow this rule if one can be sure that no one will find out that I have not paid and that my defecting will not bring down the whole system?In other words: Why should I be motivated to abide by the general rule?If the only reason why one endorses the moral principle is a self-interest, why can one not also defect out of self-interest if the situation is right (cf.SCHÖNECKER, 2006, p. 313-5)?
The second objection peculiar to interest-based Contract theories questions whether they are even a distinct and stable position.In order to formulate that objection, Shafer-Landau uses a Euthyphro-style scenario (cf.SHAFER-LANDAU, 2003, 41f).Contract theories face the dilemma whether the ideal condition (of impartiality etc.) under which one should choose one's principles is a moral demand or not.If it is a moral demand, then there is already a moral standard prior to one's choosing under ideal conditions.There is therefore a moral demand independently of Contract theories, and this moral demand must be accounted for by Moral Realism or another theory.However, if the ideal condition (of impartiality etc.) is not a moral standard, then why should one expect the procedure to yield any moral results?The procedure might yield principles that we do not recognize as moral demands at all, e.g., of not helping others.Why should this have greater authority than our moral intuitions themselves?
Finally, Kant's Constitutivism is at the other end of the spectrum.It has the advantage that it can accommodate many of the objections that have been raised against other forms of Constructivism, and might best be able to explain the moral phenomenon.If morality is a necessary law of one's own reason, then it can easily account for the idea that moral demands can be contrary to one's desires.It also does not face the objections that have been raised against Contract theories.Since Kant does not base morality on one's desires, there is an independent answer as to why one should not be a freerider.Even if the free-rider can be sure not to be found out, his own reason still commands not to abuse the system.Constitutivism can similarly answer the Euthyphro charge.The requirement to adopt an ideal standpoint, the perspective of pure reason, is a moralized demand, but unlike Moral Realism, it is not an external standard, but arises out of one's own reason.Therefore, Kant's Constitutivism has all the advantages of Constructivism, and it can better explain the phenomenon than other D-positions.
However, these advantages come at the price that Constitutivism is more demanding, and has more assumptions than other forms of Constructivism.Why should one believe that there is this constitutive law of one's own reason?What kind of metaphysical entity is this law-giving reason, and why should one follow its dictates over one's inclinations?(For an attempt at an answer see Sensen 2013 and forthcoming.)

Conclusion
My first aim in this paper has been to classify Moral Realism and Anti-Realism.I have argued that there is not one clear contrast since there are a variety of positions.I have tried to present the main competitors and analyze their strengths and weaknesses.The discussion in this paper is, of course, far from complete.There are many more moves and countermoves advocates and opponents can make.The debate which of the positions is the correct one is not settled and on-going.What I hope I have achieved is to point out that any classification that does not include Constitutivism is an incomplete one.This is an important result, because much of the debate consists in discrediting one's opponents.But if the list of opponents is incomplete, ruling out some but not all does not establish one's own position.Against this, I have tried to show that Constitutivism is worth taking seriously as an important contender, and that it deserves the same scrutiny that the other positions have received.
The other lesson which we can learn from Kant is that we should ask ourselves why these debates have not been settled.Have we just not come far enough, or might it be that the moral phenomenon is a puzzle that is hidden from our view?Maybe we cannot agree on the truth of a D-, M-, or Eposition in morality because we have a cognitive closure regarding the metaphysical nature of morality.