I'd like to add something to these reflections, because when discussing about morality, it seems to me that there are a lot of confusions and misdirections in intellectual circles and even in journals (many of which are intentionally caused by atheists who don't want accept the implication of their worldview).
I think it is important to understand that Hume's is-ought dichotomy (also known as Hume's guillotine) is a purely logical argument.
Hume's argument is that you cannot derive (logically)
ought statements from
is statements. In other words, from "is" premises, you cannot derive logically any normative conclusion.
Philosopher Kennett Merrill (who's a contemporary Hume scholar), in his
Dictionary of Hume's philosophy, explains Hume's argument:
This “inconceivable deduction” is often described as the impossibility of inferring a normative (e.g., a moral or ethical) conclusion from wholly factual premises. It is not necessary that the words is and ought or their negatives literally occur in the argument.Thus, the following argument illustrates the sort of inference that Hume is taken to proscribe: “Stalin was responsible for the deaths of millions of persons who had committed no serious crime or no crime at all, and certainly no capital crime. Further, Stalin knew that these people were innocent. Therefore, Stalin was an evil man.” First, a point about terminology. When Hume speaks of a deduction, he means any sort of ratiocinative inference, whether it be deductive (in the contemporary sense of logically necessary) or inductive (= probabilistic).It is a mistake to interpret Hume as restricting what he calls deduction to arguments whose conclusions follow (or are claimed to follow) necessarily from their premises by strict entailment. He clearly means to include arguments based on causal reasoning, all of which fall short of demonstration. He first argues at some length that moral distinctions do not consist in relations that are “the objects of science” (or, alternatively, “can be the objects of knowledge and certainty”); namely, resemblance, contrariety, degrees in quality, and proportions in quantity or number (THN, 70 and 468; 1.3.1.2 and 3.1.1.26). He goes on to “the second part of [his] argument” (THN, 468; 3.1.1.26; italics are in Hume’s text), which is to show that morality does not consist in any matter of fact that can be discovered by the understanding (causal reason, in this case). Taken together, the two parts of Hume’s argument purport to prove that morality is not an object of reason, either demonstrative reason or (probabilistic) causal reason.Since reason “exerts itself” in only the two ways just mentioned—i.e., from demonstration or probability; from the abstract relations of our IS/OUGHT ideas or the relations of objects revealed in experience—it follows that moral distinctions are not based on rational inference at all". (
Dictionary of Hume's Philosophy, pp.155-156)
Merrill interprets Hume as including both logically necessary inferences (deductive arguments) and probabilistic inferences (inductive arguments). Also, his example tends to cause confusion because the conclusion "Stalin was an evil man" is a moral claim, not a moral obligation. A moral claim has pretentions of being true or false (i.e. provided it refers or not to the objective realm of values), while a moral obligation has pretentions of being followed and respected, not pretentions of truth.
But Merrill's point is that from purely factual premises, you cannot derive any moral statement (whether a moral claim or a moral prescription). In both cases, an ought statament doesn't follow from purely "descriptive" or "is" premises.
Perhaps with the example of a deductive argument, we can understand the point better:
For example: From these premises 1)All men are mortal, and 2) Jime is a man, you can only derive as the only logically valid conclusion "Jime is mortal". You cannot derive "Jime ought be mortal" or anything like that, because the conclusion doesn't follow.
The reason is that the conclusion of a logically valid argument only makes explicit what is implicit in the conjuntion of the premises.
Another important point to note here was Kant's distinction between "hypothetical imperatives" and "categorical imperatives".
Hypothetical imperatives are hypothetical (or conditional) because they depend on the agent's ends or preferences.
They're generally expressed in the form of a conditional statement (IF x, THEN Z...) where the antecedent of the conditional (If...) is the agent's particular end.
For example, IF I want to write in my blog, then I must have internet.
Note that the imperative "I must have internet" is conditional on my preference or end of writing in this blog. And such "must" expresses only the recognition of the factual adecuacy between the means (e.g. the internet connection) and ends (e.g. writing in a blog), but the end itself are purely subjective (i.e. conditioned by the agent's subjetivity, namely: purposes, desires, wishes, preferences and so forth).
This kind of hypothetical imperatives are known also as prudential rules or technical rules (in jurisprudence).
Categorical imperatives (which Kant considered as the only moral imperatives) are not conditional, but inconditional (i.e. they command a certain course of action, regardless of the agent's personal motives, wishes or preferences, and even regardless of any consequence at all).
Another important distinction is between moral values and moral duties.
Moral values refer to what is good or bad (e.g. compasion, tolerance, justice, objectivity, responsability, love, etc. are considered typically as moral
values). Moral duties or prescriptions refer to moral
obligations (e.g. they have as referent persons with free will and intelligence, capable of understanding the prescriptions in questions and obey them), for example "
You ought to love your parents" is a moral duty based on a moral value (love).
For example, Kant's categorical imperative is not a moral value, but a moral duty or law (in fact, it is not even a moral duty but a meta-ethical prescription: it prescribes the criterion according to which we can to know what kind of moral maxims we should obey).
Another distinction is between objective moral values and duties, and subjective moral values and duties. This is an
ontological distinction.
Objective moral values and duties are the one whose existence is independent of our subjectivity. They are not moral values and duties which exist as part of the extra-mental world (in the same way that the sun exists regardless of the subjectivity of any person at all). By subjective moral values and duties, we refer to values and duties which depends on the subjectivity of people (not necessarily on people's arbitrary wishes, because even if such values and duties depend on a particular instrinsic and non-arbitrary constitution of our consciousness or our mind, they're subjective in the ontological sense).
As an analogy, Kant thought that causality is a property of the human mind. The rational human mind includes, as an essential category of it, the notion of causality. Now, is Kant's conception of causality an objective or a subjective one? It's clear that, ontologically, Kant's conception is subjective, because causality is not a property of the external world, but a property of our consciousness. It is our mind that projects on the external world the category of causality, but causality as such is not a property of the mind-independent real world. (This is why Kant's philosophy is ontologically idealistic, in contrast with the objectivistic and realistic metaphysics, like scientific naturalism).
Likewise, if moral values and duties were intrinsic and non-arbitrary properties of our mind or consciousness, they would be ontologically subjective, in contrast to moral values and duties which exist independently of our minds. The proof of this is that if consciousness disappears, such moral properties would disappear too. They don't exist in the world "out there".
Secular humanists, atheists, naturalists and materialists typically reject the existence of objective moral values (and duties) because in general naturalism only accepts the entities postulated by natural science. And natural science is morally neutral.
No natural science posits any moral value or duty in order to explain physical phenomena. This is why naturalists (if logically coherent) cannot accept the objective existence of moral value or moral duty in the naturalist ontology.
Keith Augustine has defended
one of the best arguments in favor of the subjectivity of moral values and duties if naturalism were true.
I consider that, if naturalism were true, Keith's argument is virtually irrefutable.
If scientific naturalism is true, then it is more plausible than not that moral values and duties are subjective, and it implies (as Keith correctly notes) that moral laws and obligations are not instrinsically obligatory at all. They're man-made and largely arbitrary.
Some naturalists have tried to avoid this implication (the only plausible one for naturalists) appealing to an objective Platonic realm of moral value and duties. They argue that there could exist a purely abstract realm of value and duty which is independent of any source, human or divine.
The problem with this view, given naturalism, is obvious:
1-Epistemologically, how are we know the existence of such purely abstract realm which is causally disconnected from and parallel to the physical world? (Note that claiming that such abstract realm is causally efficacious on our minds would violate the basic principle of naturalism known as the
causal closure of the physical world, which only accepts physical causes for physical effects, and since that abstract objects are not physical objects, they cannot be causally efficacious on our physical brain. So, how the hell are we going to know such abstract objects?).
2-But let's concede for the argument's sake that, despite point 1 and given naturalism, we do actually know such abstract values and duties. This question arises: is it
likely that, given naturalism and the largely random evolutionary process, have human beings evolved precisely to grasp such abstract values? It seems unlikely. (It would seem that our evolution was monitored in the direction of such values and this assumption cannot be accepted by any consistent naturalist as a likely one).
As Keith astutely notes: "
It is possible that moral laws have existed since the Big Bang, but that they could not manifest themselves until sentient beings arose. However, such a view implies that there is some element of purposefulness in the universe--that the universe was created with the evolution of sentient beings "in mind" (in the mind of a Creator?). To accept the existence of objective moral laws that have existed since the beginning of time is to believe that the evolution of sentient beings capable of moral reasoning (such as human beings) has somehow been predetermined or is inevitable, a belief that is contrary to naturalistic explanations of origins (such as evolution by natural selection) which maintain that sentient beings came into existence due to contingent, accidental circumstances."
In other words, if naturalism were true and even conceding the logically possible compatibility between naturalism and objective moral values and duties, it becomes UNLIKELY that the largely random, accidental and contigent evolutionary process that created human beings made them precisely capable of recognizing such purely abstract (and non-empirical!) realm of values and duties.
But another question (this time ontological) arises: If naturalism were true, would we expect that such abstract realm of values and duties do exist? If such values and duties do exist, naturalists cannot account for them. They simply have to accept them as brute (unexplained and mysterious) facts. The naturalistic ontology has not the resources to account for such moral facts.
Now, compare this naturalistic view with a theistic view (a view where God exists). In a theistic view, God (who's a moral agent) created the world with physical laws; but given that he created the world in order to create other spiritual beings (and not for the sake of nature itself), he created also spiritual and moral laws for them. In this theistic view, the following facts make sense:
1-A world with objective moral values and duties (because the world is created FOR the sake of intelligent, sentient beings).
2-Spiritual beings with consciousness capable of recognizing more or less correctly such values and duties.
3-Spiritual beings with free will, capable of freely choosing good and bad actions and being responsible for them.
Here, the question is not if you believe in God or not. The question is in which worldview (naturalistic or theistic) is more LIKELY the existence of objective moral values and duties, and of beings capable of recognizing them? I think the answer is obvious.
Many naturalists would agree with this. Others, the sophistical ones, in order to avoid any concession for the moral superiority of theism, and despite of powerful and logically consistent arguments like the one defended by Keith (or
Dawkins, or
Rosenberg, or
Baggini, or
Ruse, or
Clark, etc.) would try to argue that naturalism is compatible with such moral facts. (Note that this is a red herring: the discussion is not about compatibility, but about what is
more likely given the basic premises of naturalism such as: physicalism, determinism, the causal closure of the physical world, purely accidental and contingent biological evolution, etc.).
At most, they could show that naturalism is logically compatible with the objectivity of moral values and duties. But logical compatibility alone doesn't imply probability. And what's at stake is the the conditional propability that, given naturalism, objective moral values and duties exist.
Clearly, if naturalism is true, the probability that such moral facts exist is very, very low.
As Keith has argued: "
But given that moral subjectivism is just as logically viable as moral objectivism and that moral objectivism is implausible if a scientific naturalism is true, I think that there is a good case for the nonexistence of objective moral values"
I myself have a stronger position than Keith. In my view, the objectivity of moral values and duties is metaphysically incompatible with naturalism, because the basic premises of naturalism (specially physicalism, that is, the view that only physical things exist) precludes the existence of objective moral values and duties, because the latter are not physical things (proof: they have no energy, which is the basic and universal property of any physical object).
So, if objective moral values and duties exist, naturalism is not only plausibly false, but
necessarily false by metaphysical reasons.
But you don't need to agree with my stronger version of this argument. If you want, just consider the arguments by logically consistent naturalists like Keith's and others (for specific references to contemporary naturalist literature, see all my posts about "
The moral poverty of naturalism").
I've framed this post in terms of naturalism vs. theism, and I'm sure some readers will protest that I've left aside other worldviews such as pantheism, etc.
This objection has some value: actually, the reason why I've left aside other worldviews or philosophical alternatives is that naturalism and theism are the two leading competing worldviews in Academia today. Moreover, I think the case for naturalism and theism is scientifically and philosophically stronger than the case for pantheism, Buddhism and other philosophical positions relevant to this question (some readers will strongly disagree with this opinion!).
So I think I'm largely justified in arguing in terms of this naturalism vs theism framework.
In any case, I also want to address the contribution of other positions, but you'll have to wait for future posts...
David Hume is a kind of "hero" for many atheists and secularists, mostly because Hume was a critic of religion, miracles and the classical arguments for God's existence.
