Harman, Gilbert. 1975. “Moral Relativism Defended.” The Philosophical Review, Vol. 84, No. 1, pp. 3-22.
Harman argues that a speciﬁc sort of moral judgments – “inner judgments” – are relative to motivating attitudes. Further, he argues that such motivating attitudes are themselves derived from agreement. He is therefore arguing for a form of moral relativism. Harman concludes by tackling a cluster of objections to his theory.
Harman presents his idea as a “soberly logical thesis”: he says that inner judgments are only possible when an agent is assumed to have reasons for her actions, and further that the speaker (the person judging the agent) agrees with such reasons. By ’inner judgment,’ Harman means a speciﬁc use of the word ’ought’: there is the ’ought’ of expectation (“It ought not rain today”), the ’ought’ of rationality (“You ought not jump oﬀ that cliﬀ if you want to live”), the ’ought’ of normativity (“We ought not invade foreign countries”), and the ’ought’ of morality (“You ought to honor your promise”). Harman is concerned here with the latter ought.
Harman gives an example in an attempt at clariﬁcation: it is not logical to make an inner judgment about Hitler, since the man was beyond any motivating considerations we might have. He is “beyond the pale,” in the sense that he never shared our motivating attitudes; therefore, we can call him evil, but we cannot say “He ought not to have exterminated the Jews” in the moral sense of ought, since he and us never shared any reasons for action.
Harman then formalizes his thesis into a four-place predicate: “Ought(A,D,C,M),” where A is an agent, D is a type of action, C are relevant considerations, and M are motivating attitudes. It is best to think of this as a sort of function whose outputs are various ’ought’ statements.
Next, Harman asserts that the underlying, motivating attitudes which produce inner judgments are reached via moral bargaining among groups. Harman thinks that we should accept this thesis because it best explains a previously mysterious aspect of our morality: the fact that we place a premium on not harming others, as opposed to helping those in need. Harman says that this is reasonable given his theory, since everyone could agree to not harm others; the rich and powerful, however, would be far less likely to agree that we ought to help all who are in need, since this would cost them the most. Therefore, bargaining results in widespread acceptance of the former
and a de-emphasis on the latter.
Harman next attempts to deal with a large group of common objections
to implicit agreement theories (the sort of theory he has just presented). The ﬁrst three argue that: the fact that no one remembers reaching any agreement suggests that no such agreement exists; since no one can clearly spell out any agreement they are supposedly a part of, there probably is no such agreement; and ﬁnally, just because one agrees does not mean one has the motivation to fulﬁll an agreement. Harman believes that all three of these objections derive their force from equivocation over the word ’agreement’: for Harman, an agreement is merely sharing intentions with others; one need not actively sit down and discuss anything in order to do that. Further, just because one cannot spell out an agreement does not mean it is not there. Many other sorts of understanding are like this: we adhere to their precepts without being able to clearly deﬁne them. Finally, since agreement is merely shared intention, and intention entails motivation, motivation necessarily goes along with agreement.
The next objection Harman deals with is the idea that since we can judge those who are forced to agree diﬀerently than those who agree willingly, there must be some criterion of moral judgment before agreement is reached. Harman responds that, to the contrary, the argument that there are moral principles prior to agreement is itself only meaningful in relation to some agreement in intentions; therefore, the ’Ought(A,D,C,M)’ thesis stands.
The ﬁnal objection Harman tackles is the idea that since there is disagreement about what is right, the ’Ought(A,D,C,M)’ thesis cannot be correct, since if it was, there be no room for disagreement, since judgments are all produced from implicit agreement. Harman says that his thesis does allow for disagreement; what the disagreement is about, however, are the implications for an agreement in relevant cases. It is the application of the agreement which can be called into question. Further, disagreement and even the gradual revision of agreements can occur internally where some part of the agreement contradicts another. So, disagreement and change are possible under Harman’s theory.
Argument One (Presuppositions)
- Inner judgments are either relative to something, absolute, or non-
- Inner judgements are not absolute.*
- Therefore, either inner judgments are relative to something or are non-existent. [1,2]
- Inner judgments exist.*
- Therefore, inner judgments are relative to something. [3,4]
Argument Two (Inner judgments depend on bargaining)
- Inner judgments are relative to motivating attitudes. [45L]
- Motivating attitudes are formed via implicit bargaining. [45R]
- Therefore, inner judgments depend upon implicit bargaining for their
Argument Three (Inference to best explanation)
- In our morality, the duty to not harm others takes precedence over the duty to help those in need. [45R]
- The ’Ought(A,D,C,M)’ thesis best explains premise (1). [46L]
- Therefore, the ’Ought(A,D,C,M)’ thesis is probably true. [1,2]
Argument Four (Replying to Objections 1-3)
- If objections 1-3 are correct, then an agreement must be an explicit ritual which everyone is consciously aware of. [47L,49L,49R]
- To the contrary, an agreement is an implicit intention which one presumes to share with others. [47L]
- Therefore, objections 1-3 cannot be correct. [1,2]
Argument Five (Replying to objection 4)
- If objection 4 is correct, then there is no room for disagreement if the ’Ought(A,D,C,M)’ thesis is true, since what is ’right’ is merely what one agrees is right. [47R]
- There is room for disagreement even if ’Ought(A,D,C,M) is true; disagreement merely takes the form of disagreeing over the implications
of an agreement in speciﬁc cases. [47R]
- Therefore, objection 4 is incorrect. [1,2]
Argument Six (Replying to objection 5)
- If objection 5 is correct, then the case of those compulsed to agree would suggest there are prior moral principles by which to judge (we judge the person compelled to obey diﬀerently than one who obeys willingly). [48L]
- The idea that there are prior moral principles before implicit agreement emerges is itself only meaningful in relation to an implicit agreement, thus aﬃrming the ’Ought(A,D,C,M) thesis. [48L]
- Therefore, objection 6 is incorrect. [1,2]
1. (∀x)(∀y)[Jx ∧ [(Rxy ∨ Ax) ∨ ¬Ex]]
2. (∀x)(∀y)(¬Ax ∧ Ex)…Therefore, (∀x)(∀y)(Jx ∧ Rxy)
3. (∀y)[Ja ∧ [(Ray ∨ Aa) ∨ ¬Ea]](1, UI)
4. Ja ∧ [(Rab ∨ Aa) ∨ ¬Ea](3, UI)
5. (Rab ∨ Aa) ∨ ¬Ea(4, SIMP)
6. ¬Aa ∧ Ea(2, UI)
7. Ea(6, SIMP)
9. Rab ∨ Aa(8, 5,DS)
10. ¬Aa(6, SIMP)
11. Rab(10, 9,DS)
12. Ja(4, SIMP)
13. Ja ∧ Rab(11, 12,CONJ)
14. (∀x)(Jx ∧ Rxb)(13, UG)
15. (∀x)(∀y)(Jx ∧ Rxy)(14, UG)
Where J=Inner judgment, Rxy= X is relative to y, A=Absolute, E=Exists.
1. J → M
2. M → I…Therefore, J → I
3. J → I(2, 1,HS)
Where J=Inner Judgment, M=Motivating attitude, I=Implicit bargaining
Since an ’inference to the best explanation’ is not a formal inference, it cannot be formalized.
1. (O1 ∧ O2 ∧ O3) → E
2. ¬E ∧ I…Therefore, ¬(O1 ∧ O2 ∧ O3)
3. ¬E(2, SIMP)
4. ¬(O1 ∧ O2 ∧ O3)(3, 1,MT)
Where O1=Objection 1 is correct, O2=Objection 2 is correct, O3=Objection 3 is correct, E=Agreement is explicit, I=Agreement is implicit.
1. O4→ ¬R
2. R…Therefore, ¬O4
3. ¬O4 (1, 2,MT)
Where O4=Objection 4 is correct, R=There is room for disagreement.
1. O5→ P
2. ¬P…Therefore, ¬O5
3. ¬O5 (1, 2,MT)
Where O4=Objection 5 is correct, P=There are moral principles prior to