All I see are eyes!

Reflections on Sam Harris’s latest episode on morality

[Edited to add, 5/9/2024]: Critics in the comments here and on reddit have convinced me that the objections raised in sections 2.0 and 3.0 are partially or wholly mistaken, depending on the claim in question. I have moved these sections to a retraction section, section 8.0, at the end of this post.]

[Edited to add, 5/9/2024: I will be updating portions of this post, so it is a work in progress]

1.0 Introduction

I recently reviewed the publicly available portion of Sam Harris’s recent podcast episode, where he sought to clarify his views on morality along with Nathan from Digital Gnosis. I don’t think Harris was successful in clarifying his views, and only managed to reiterate what he’s said in the past with little refinement or responsiveness to critics. Here, I want to reflect on Harris’s remarks and some of the remarks that arose in my discussions with others about the video. 

My main concern is that Harris’s views on morality are so muddled and poorly articulated that they actively foster confusion. Coupled with his repeated declaration that metaethical terminology is boring, Harris also poisons the well by discouraging his audience from investigating what philosophers have to say about the topic and giving the false impression that metaethics isn’t worth investigating. This one-two punch risks misleading his audience into believing the academy has failed but that he’s provided a clear path forward. He has not. Harris’s account is grossly underdeveloped, unclear, and poorly argued. I’ll conclude on more general observations about Harris’s views on morality, but for now, I want to focus on the video.

2.0 A new critique of Harris

[Under construction]

I am certainly not claiming that moral truths exist independent of the experience of conscious beings—like the Platonic Form of the Good—or that certain actions are intrinsically wrong. I am simply saying that, given that there are facts— real facts—to be known about how conscious creatures can experience the worst possible misery and the greatest possible well-being, it is objectively true to say that there are right and wrong answers to moral questions, whether or not we can always answer these questions in practice.

This is a very strange remark. Harris still appears to be mixed up about what moral realism is. Moral realism is the view that there are stance-independent moral facts. On such a view, moral claims like “murder is wrong” are not made true by our attitudes or preferences. But the way in which moral claims aren’t made true is very specific. Moral claims aren’t made true by our preferences or desires. But the specific means of acting in accord with the stance-independent moral facts in any given situation may depend on people’s stances or preferences. 

For instance, moral realism is true, and utilitarianism is the correct moral theory, then we have a stance-independent moral obligation to maximize utility. Utility may be roughly cashed out as increases in happiness and decreases in suffering. In any given situation, a realist utilitarian has an obligation to maximize happiness and decrease suffering. That they have an obligation to do so is not made true by anyone’s stances or preferences. However, in a particular situation, it may be that the way to maximize utility would be to satisfy people’s preferences. In this case, the moral facts do depend on people’s stances in one way: their stances determine what would or wouldn’t maximize utility. But they also don’t depend on their stances in another way: that one had an obligation to satisfy those preferences isn’t made true by their preferences (or anyone else’s).

Harris does not appear to clearly distinguish these senses of stance-independence and stance-dependence. As such, it’s not clear whether he understands the distinction, and understands moral realism to only be concerned about what serves as the truthmaker for first-order moral claims and whether those truthmakers turn on stances or not. Harris appears to reject certain types of moral realism, e.g., moral platonism, though it’s unclear what he means by “intrinsically wrong.” He then goes on to say that he’s “simply:” saying that there are “facts,” “real facts” of a presumably moral kind. What does he mean by “real”? I don’t know, because he doesn’t tell us. However, this remark is accompanied by two endnotes, so I had a look at those.

The first is endnote 4. Harris presents a classic remark remark from Mackie:

If there were objective values, then they would be entities or qualities or relations of a very strange sort, utterly different from anything else in the universe. Correspondingly, if we were aware of them, it would have to be by some special faculty of moral perception or intuition, utterly different from our ordinary ways of knowing everything else (Mackie 1977, p. 38)

Harris reacts to this by saying:

Clearly, Mackie has conflated the two senses of the term “objective.”

This is nonsense. Where does this supposed conflation occur? Mackie states that if there were objective values, they’d have metaphysically questionable properties, and that it’d be difficult to discover these facts because of these properties. This in no way involves conflating two types of objectivity. Mackie is not talking about epistemological objectivity at all. He’s only talking about ontological objectivity and about the epistemological problems one would encounter were moral facts to have the ontological characteristics Mackie has in mind. Those epistemological problems have nothing to do with subjective biases or whatever else might be involved in epistemological objectivity and subjectivity. Simply because this distinction is salient to Harris doesn’t mean it had any relevance at all to this passage, because it doesn’t. Mackie is referencing a standard epistemic challenge for non-naturalist moral realist accounts.

It would appear that Harris has fundamentally misunderstood this passage, possibly because it happens to mention both an epistemological claim and an ontological claim in regards to moral objectivity. Maybe Harris thought that because Mackie was making an ontological claim and an epistemological claim that he was mixing up the distinction, but even this wouldn’t make sense because Mackie explicitly distinguishes ontology and epistemology in the passage, and prefaces by saying he’s going to do so:

Even more important, however, and certainly more generally applicable, is the argument from queerness. This has two parts, one metaphysical, the other epistemological. If there were objective values, then they would be entities or qualities or relations of a very strange sort, utterly  different from anything else in the universe. Correspondingly, if we were aware of them, it would have to be by some special faculty of moral perception or intuition, utterly different from our ordinary ways of knowing everything else. These points were recognized by Moore when he spoke of non-natural qualities, and by the intuitionists in their talk about a ‘faculty of moral intuition.’

As you can see, Mackie is explicit and clear about the metaphysical and epistemological considerations at stake, and they have nothing at all to do with epistemological objectivity/subjectivity.

I included the passages that follow to add some more context, too. The passages preceding and following these remarks include a careful and concerted effort to clarify the form of objectivity that Mackie is denying. These passages allude to Mackie’s opposition to a non-natural conception of morality whereby the moral facts are construed in objectively prescriptive terms that render them highly unusual and unlike other sorts of facts. Mackie’s point is twofold: that moral facts of this kind would be metaphysically unlike the sort of things we’re already familiar with, and that this creates special epistemic challenges for anyone who purports to have knowledge of such facts because such facts would not be natural facts and could not be accessed through ordinary empirical, scientific, sensory, or philosophical standardly employed in other domains of inquiry. Here’s Mackie making this point:

When we ask the awkward question, how we can be aware of this authoritative prescriptivity, of the truth of these distinctively ethical premises or of the cogency of this distinctively ethical pattern of reasoning, none of our ordinary accounts of sensory perception or introspection or the framing and confirming of explanatory hypotheses or inference or logical construction or conceptual analysis, or any combination of these, will provide a satisfactory answer; ‘a special sort of intuition’ is a lame answer, but it is the one to which the clear-headed objectivist is compelled to resort. (pp. 38-39)

Mackie’s objection, in other words, turns on characterizing moral facts in non-naturalist realist terms, then raising both metaphysical and epistemological concerns about these facts. It has absolutely nothing to do with the distinction between epistemological and ontological objectivity. 

Insofar as Harris is presenting a naturalist account of morality, Harris’s conception of moral facts is not Mackie’s target. The epistemological point he makes turns on the difficulties of acquiring knowledge of non-natural facts. This has no relation to epistemological objectivity. This is why Mackie immediately invokes Moore following these remarks, and explicitly references “non-natural qualities.” Prior to this remark, Mackie says:

No doubt was an extravagance of Moore to say that ‘good’ is the name of a non-naturalist quality, but it would not be so far wrong to say that in moral contexts it is used as if it were the name of a supposed non-natural quality, where the description ‘non-natural’ leaves room for the peculiar evaluative, prescriptive, intrinsically action-guiding aspects of this supposed quality. (pp. 31-32)

Shortly after introducing the argument from queerness, Mackie makes it much more explicit his target is a non-naturalist conception of morality, in particular:

Plato’s Forms give a dramatic picture of what objective values would have to be. The Form of the Good is such that knowledge of it provides the knower with both a direction and an overriding motive; something’s being good both tells the person who knows this to pursue it and makes him pursue it. An objective good would be sought by anyone who was acquainted with it, not because of any contingent fact that this person, or every person, is so constituted that he desires this end, but just because the end has to-be-pursuedness somehow built into it. Similarly, if there were objective principles of right and wrong, any wrong (possible) course of action would have not-to-be-doneness somehow built into it. Or we should have something like [Samuel] Clarke’s necessary relations of fitness between situations and actions, so that a situation would have a demand for such-and-such an action somehow built into it

I am rarely this blunt, but Harris has no idea what he’s talking about. This is why superficial engagement with academic philosophy won’t cut it. Harris absolutely butchered his interpretation of this passage, and his poor scholarship on the topic of metaethics bleeds into how he frames and discusses this topic. Worse, his errors give the impression that academic philosophers are incompetent and bumbling idiots who make mistakes that are so clear and obvious that a non-specialist like Harris can readily spot them. Instead, Harris’s blunder, buried away in this passage, reveals his own lack of understanding. This is exacerbated by Harris’s pivot to a on sequitur that simply reiterates his main talking points even though they’re irrelevant to the supposed conflation:

Clearly, Mackie has conflated the two senses of the term “objective.”We need not discuss “entities or qualities or relations of a very strange sort, utterly different from anything else in the universe” in order to speak about moral truth. We need only admit that the experiences of conscious creatures are lawfully dependent upon states of the universe—and, therefore, that actions can cause more harm than good, more good than harm, or be morally neutral. Good and evil need only consist in this, and it makes no sense whatsoever to claim that an action that harms everyone affected by it (even its perpetrator) might still be “good.” We do not require a metaphysical repository of right and wrong, or actions that are mysteriously right or wrong in themselves, for there to be right and wrong answers to moral questions; we simply need a landscape of possible experiences that can be traversed in some orderly way in light of how the universe actually is.

The first sentence claims that Mackie has conflated the two senses of objectivity that Harris describes. The rest of the paragraph, however, is a bizarre non sequitur. It does absolutely nothing to demonstrate that Mackie has made this conflation. Mackie’s target is a non-naturalist conception of moral facts. Claiming that we don’t need to discuss moral facts of the sort Mackie finds objectionable because we can offer a naturalistic account of moral facts is…simply irrelevant to the objection. Mackie and other antirealists would simply have a different objection to naturalist accounts. And Mackie certainly hasn’t mixed up naturalism with non-naturalism.

Harris also simply asserts that we don’t need to construe morality in the terms Mackie has to “speak about moral truth.” Harris presents no argument for why we “need not discuss” the conception of morality Mackie provides. Why don’t we need to? 

Instead, we need only “admit” what Harris goes on to describe. This choice of “admit” is very strange. To “admit” something pragmatically implies that one is reluctant to acknowledge some truth they themselves already know or believe to be true. This choice of words subtly implies that Harris’s opponents are disingenuous. It’s a bizarre and subtle form of rhetoric, and it is the kind of slippage in language I genuinely dislike. The accumulation of remarks like these can subtly wheedle one’s audience into complacent support for one’s own views by creating an us-them dynamic that depicts one’s opposition is “the bad guys.” I worry Harris does this throughout The Moral Landscape, and that much of the way he addresses morality and other topics likewise involves the consistent deployment of subtle rhetoric of this kind.

And what does this have to do with Mackie’s supposed conflation between the two senses of objectivity? As far as I can tell, nothing. So why did the paragraph begin with the accusation of a conflation, then simply deny that we need to characterize morality in the way Mackie has? What does that have to do with his supposed conflation? Harris’s line of inference here, if there is one, is opaque, seemingly confused, and appears to rely on a misreading of Mackie. What a mess.

Harris himself also doesn’t appear to understand what moral realism is about. Moral realism roughly maps onto Harris’s invocation of ontological objectivity. But these days, many philosophers would describe moral realism as the view that there are stance-independent moral facts. According to this view, moral truth isn’t determined by our stances, attitudes, or beliefs about the first-order moral claims themselves. Moral facts can still be “made true” or “depend on” our attitudes or preferences, only in a somewhat different way. For instance, suppose utilitarianism is true, and that therefore it is morally obligatory to maximize utility. Suppose helping strangers would maximize utility. And suppose the way to help a particular stranger would be to give them the food that they prefer. As such, one would have a moral obligation to satisfy this person’s preferences. Here’s how a moral realist would handle this situation:

1. The fact that one ought to maximize utility is stance-independently true. It is not made true by our attitude or preference that we ought to maximize utility

2. The fact that satisfying a person’s preference in this specific case would maximize utility is stance-independently true.

3. It is therefore true that one ought to satisfy this person’s preference, and this is not made true by anyone’s preferences.

4. That satisfying this person’s preferences would 

Harris calls himself a moral realist, and says that there are moral facts...but it is not clear what he means by this. In the above remark, he alludes to the distinction between epistemological and ontological objectivity. And it seems that because Sam thinks we can address questions about human wellbeing in an epistemologically objective way, that this provides us with all the objectivity we need to yield “moral truth.” Yet if he dispenses with ontological objectivity, and this is roughly an analogue to stance-independence in the sense employed in mainstream conceptions of moral realism, then Harris does not appear to be affirming that there are stance-independent moral facts, and thus it is no longer clear whether he is, in fact, a “moral realist,” in the way the term is typically employed. The problem is, it’s also not clear he isn’t. Sam’s remarks on this matter are so muddled it’s simply unclear whether one can even definitively identify him with the standard labels employed in academic philosophy.

4.0 Harris is way too proud of his “boring” joke

Harris references his joke from The Moral Landscape about how many of the terms moral philosophers use are boring. The remark appears in the endnotes, and is the first note to Chapter 1. Here, Sam says:

Many of my critics fault me for not engaging more directly with the academic literature on moral philosophy. There are two reasons why I haven’t done this: First, while I have read a fair amount of this literature, I did not arrive at my position on the relationship between human values and the rest of human knowledge by reading the work of moral philosophers; I came to it by considering the logical implications of our making continued progress in the sciences of mind. Second, I am convinced that every appearance of terms like “metaethics,” “deontology,” “noncognitivism,” “antirealism,” “emotivism,” etc., directly increases the amount of boredom in the universe.

Harris goes on to add that he wants to engage a wider audience and that “Few things would make this goal harder to achieve than for me to speak and write like an academic philosopher.”

This is a good goal. I don’t think Harris should speak like an academic philosopher. But the problem isn’t that Harris isn’t employing these terms. It’s that he doesn’t demonstrate an understanding of or engagement with the distinctions these terms refer to, the issues that they are intended to address, or the numerous problems and considerations that moral philosophers discuss in the literature. It’s one thing to eschew the terminology; it’s quite another to dismiss an entire corpus of work even if you think its focus were misguided or its solutions were unappealing.

I do not think, for instance, that Harris is obligated to call himself a naturalist or non-naturalist, or risk censure from the robed elites in their vaunted towers. I am an extremely vocal, even aggressive critic of the insularity, elitism, and gatekeeping that characteristics mainstream analytic philosophy. I have gone out of my way to criticize deference to the majority view among academic philosophers, express doubts about the “expertise” of academic philosophers, criticize academic moral antirealists more broadly, voice dissatisfaction with standard labels in metaethics, criticize the poor dialectical and rhetorical practices of philosophers, and criticize philosophers for a widespread tendency to making sweeping empirical claims without adequately engaging with the relevant empirical data, and explain why mainstream terms and categories are inadequate, and why my own views fall outside their scope.

Where Sam and I differ is that I show my receipts. I read what philosophers have to say, quote them, and raise direct critiques of their work. If I have a problem with the terms used in moral philosophy, I explain what those terms mean, and I usually don’t do so in a way that is so slipshod that I immediately contradict myself as Sam Harris has. I then go on to explain why I think those terms are a problem: I will point out how they are ambiguous, provide examples of ways in which this ambiguity can be misleading, talk about polysemy and ambiguity and pragmatics and give detailed explanations of why these terms are a problem. What I don’t do is just dismiss it all as “boring.” 

If Harris thinks this is too boring to include a popular book intended to engage a wider audience, that makes sense, and I would totally support such efforts. But if Harris is capable of addressing academic philosophy, why not do so in blog format, or in podcast episodes, or publish a paper, or collaborate with a moral philosopher and produce a book intended for an academic audience or at least engage with or address philosophers in some way that signals competence in their field? If you want to be a maverick, great. But the most convincing maverick is one who shows mastery of the fields they want to be a maverick about. Had Harris done this, and done it well, he could have potentially garnered the support or at least some degree of respect and approval from the academic community. How would that have hurt his reputation or career or efforts to engage a wider audience? It wouldn’t have. It would only add to his credibility, and made his efforts to reach a wider audience more effective, possibly much more effective.

There has never been any barrier to Harris addressing a wider audience in books and podcasts and TED talks and engaging with academic moral philosophy. Many academics can and have done so themselves, translating complex topics in e.g., theoretical physics, or mathematics, or medicine, or philosophy to the broader public. Often mastery of these topics precedes and is a necessary condition for adequately relaying the contents of those ideas to the public. An expert in physics is usually the best person to explain complex topics in simple form. Take, for example, this example of Sean Carroll explaining dimensions at five levels of complexity, ranging from explaining the concept of a dimension to a child through intermediate levels and ultimate at the expert level. Absolutely nothing about one’s ability to explain or address a topic at the expert level impairs or inhibits one from explaining it in simpler terms. Quite the opposite.

Harris’s obstinate lack of engagement with academic moral philosophy, after well over a decade, rings hollow to me. His stated intention to reach a wider audience makes very little sense. Yes, it makes sense to eschew technical terminology and engagement with technical topics in a book written for a popular audience. But Harris seems to lack knowledge and familiarity of basic distinctions in a more general respect. For instance, consider this exchange between Alex O’Connor and Sam Harris:

O’connor: “Maybe you can tell me what I’m missing from my picture then because like I say I’m broadly speaking an ethical emotivist [...]”

Harris: “Say more about the consequences of that…what does that mean?”

Go have a listen. A number of things could be going on here. Harris could be familiar with emotivism but not familiar with O’Connor’s specific view. After all, if I say I’m a moral antirealist, it may not be obvious to someone what exactly my position is, and so it would make sense to ask. It’s also possible Harris knows O’Connor’s position but his response was intended to create an opportunity for O’connor to explain his views to the audience. However, it’s also possible Harris asked this question because he genuinely doesn’t know what emotivism is.

I don’t know if Harris is familiar with emotivism or not. Maybe he is. But his conversation with O’Connor doesn’t provide compelling evidence one way or another. It’s also possible Harris has displayed familiarity with the position elsewhere. The only use of the term I could find in The Moral Landscape was the aforementioned footnote, with “emotivism” listed as one of the examples of boring terms.

There is some evidence in this discussion that Harris does not understand what emotivism is. Emotivism is a form of noncognitivism according to which prototypical moral claims like “murder is wrong,” are not propositional; that is, moral claims do not express claims that are capable of  being true or false. Instead, they express the speaker’s emotions. One can “translate” moral claims into the emotivist interpretation as follows:

1. “Murder is wrong” = “Murder?! Boo!”

2. “It’s good to be nice to people” = “Being nice? Hurrah!”

Note that booing and hurrahing can’t be true or false, any more than “ow!” or “yum!” can be true or false. While moral claims may appear to express propositions, since they’re structured as declarative sentences, they in fact don’t express propositions at all, and are thus not “truth-apt.”

This is emotivism. Now, I didn’t consult any extraneous texts or outside sources when writing this, and I did so to make a point: if you go and check these descriptions, there may be minor discrepancies, but you’ll find that my characterization is more or less accurate. It is accurate because I am familiar with what “emotivism” means, as a technical term in academic metaethics. Is Harris? It’s not so clear. Look at this exchange that occurs shortly after O’Connor describes emotivism:

O’Connor: “When we make ethical statements or pronouncements we’re essentially just expressing emotions at each other, so when I say “murder is wrong,” I’m sort of expressing some kind of distaste or I’m expressing something akin to saying “don’t do that,” which is just like a command…it doesn’t have any truth value…it’s not even subjectively true or false, it’s just like a command or an expression, and so if that’s the case then this sort of ethical talk is completely different to the picture that you’re painting

Harris: “Not really. I can translate some of that into the picture that I’m painting because again it sounds like [...] to say that it’s just a statement about the way we feel or might feel…”

O’Connor: “Well to be clear it’s not a statement about how we feel, it is the expression of the feeling.”

Here, Harris reveals a basic misconception about emotivism, immediately appearing to misconstrue it as a cognitivist position according to which one expresses propositions about their moral stances, rather than nonpropositional expressions, and O’Connor has to step in to correct Harris. Nothing about Harris’s response indicates that he was already familiar with this distinction. So, while Harris could have said that he’s aware of this distinction and his reference to “it” above when talking about “it’s just a statement about…” wasn’t about first-order moral claims, or that his use of “statement” wasn’t intended to refer to propositional claims alone. After all, there are lots of terms we could use to describe “utterances” like “murder is wrong”: statements, claims, propositions, sentences, assertions, utterances, and so on. And it’s not obvious that “statement” must be used to mean “a propositional assertion.” However, Harris doesn't say this, and in no way contravenes O’Connor’s correction. Maybe Harris was simply being polite, but I think the most plausible take on this conversation is that Harris simply doesn’t understand what emotivism is. It’s one thing to not use the term emotivism when writing a book on metaethics. It’s quite another to not even understand a basic antirealist position that undergraduates would encounter in an introductory ethics course.

Suppose for a moment that I’m correct that Harris does not know what emotivism means, despite publishing a book on metaethics. Imagine a similar situation but in the sciences. A person comes along and declares that terms like “mitochondria” and “homeostasis” maximize boredom in the universe, so they’re going to offer an account of the nature of life on earth that eschews the technical terms and distinctions employed in contemporary biology. They don’t just avoid these terms, they avoid the very concepts to which these terms refer. They avoid not just all the labels we slap onto the parts of cells, e.g., “golgi bodies,” but they make very little reference to the parts of cells at all. This person then goes on to present an account that is widely panned by biologists as ignorant and misguided. In debates over a decade later, a biologist mentions “ribosomes” to this person, and they appear to know what this means, and, when given an explanation, immediately misunderstand the explanation.

What would you make of this situation? I would conclude that this person is probably out of their depth and is unprepared to have a serious conversation about biology. Why shouldn’t we conclude the same in Sam Harris’s case with respect to morality? That is, why shouldn’t we conclude that Harris simply hasn’t studied the topic of morality seriously enough to be in a position to dismiss the technical terms used in the academic moral philosophy? 

(As an aside, O’Connor draws on an emotivist account typically associated with Ayer and a prescriptivist form of noncognitivism typically associated with Hare, since O’connor mentions both expressions of emotion and issuing of imperatives. Both are classical forms of noncognitivism, and it’d be helpful if O’connor explicitly drew attention to the distinction, but there’s no mistake in presenting both, and I think it’s a good thing that O’Connor mentioned both. O’Connor did his homework. It doesn’t look like Harris did).

These considerations are just one small sign that Harris is unprepared to have an informed and serious discussion about morality. I suspect if I looked for more instances of errors, confusions, and misunderstandings, this list would  grow quite a bit.

Harris’s dismissal of mainstream moral philosophy is superficial and inadequately justified. If you want to dismiss it, great. I’m right there with you. But if you’re going to do so, you should show your receipts. Demonstrate to your audience that you understand it, and that you’re dismissing it because you understand it and because you can present a good account of its deficiencies. This is what I meant in the video when I said that Harris would need to earn the right to criticize the fields he dismisses. I don’t mean that literally. Of course anyone can criticize or dismiss any field they want. What I am gesturing at is that if you’re not going to take a field’s contributions to a topic seriously, you should be able to explain why. Harris has not offered us a satisfactory explanation. By relegating his engagement to a snarky footnote, and reaffirming this over a decade later despite vociferous pushback, Harris reveals that he has little interest in taking moral philosophy seriously. I believe the proper response is to return the favor, and not take Harris very seriously on this topic unless and until he engages with moral philosophy in a substantive way.

5.0 Only conscious states matter

Harris makes several strange remarks about the value of consciousness. Towards the end of the video, he says:

I’m claiming that consciousness is the only context in which we can talk about morality and human values. Why is consciousness not arbitrary starting point?

The audio cuts out at this point, and the rest is behind a paywall. So for now, I want to turn to Harris’s remarks elsewhere. So let’s turn to what Harris says in The Moral Landscape. Harris appears to believe that the only intelligible subject of our values are conscious states. That is, all value is reducible to concerns about the impact our actions have on states of consciousness.

A few people in the comments argued that this is not Harris’s position. Instead, they maintained, Harris simply held that consciousness is required for value because we must be conscious in order to value things. Harris may have said things like this, but this is at best incomplete, and, in any case, would be a trivial claim to make. There is a big difference between the claim that we must be conscious in order to value things and the claim that the only things we can value (non-instrumentally, i.e., for their own sake) are conscious states.

However, Harris does make it clear that we can only intelligibly care about conscious states. In TML, Harris says:

Meaning, values, morality, and the good life must relate to facts about the well-being of conscious creatures—and, in our case, must lawfully depend upon events in the world and upon states of the human brain.

[...] whatever can be known about maximizing the well-being of conscious creatures—which is, I will argue, the only thing we can reasonably value—must at some point translate into facts about brains and their interaction with the world at large [...]

Defining goodness in this way does not resolve all questions of value; it merely directs our attention to what values actually are—the set of attitudes, choices, and behaviors that potentially affect our well-being, as well as that of other conscious minds.

The bizarre thing about this claim is that Harris does not present any arguments for this claim at all. Instead, he simply asserts that it’s impossible to think of any alternative. Harris addresses the question of why we should believe moral value is exclusively about conscious states:

And here is where the real controversy begins, for many people strongly object to my claim that morality and values relate to facts about the well-being of conscious creatures.

Harris goes on to present his critics as asking the following:

But you haven’t said why the well-being of conscious beings ought to matter to us. If someone wants to torture all conscious beings to the point of madness, what is to say that he isn’t just as “moral” as you are?

While I do not think anyone sincerely believes that this kind of moral skepticism makes sense, there is no shortage of people who will press this point with a ferocity that often passes for sincerity.

There are two serious problems with this remark. First, Harris depicts his critics as asking why consciousness matters at all. While someone could pose this question, it’s not the same thing as asking why only consciousness matters. In other words, Harris seems to subtly conflate:

1. Whether consciousness is part of what we are morally concerned with

2. Whether consciousness is the only thing we are morally concerned with

Nevertheless, I still think even this skeptic’s question is a legitimate one. As a matter of fact, consciousness does matter to me, and it probably matters to everyone reading this. But that’s a separate question from the question of whether it ought to matter. And Harris has not provided anything by way of an argument for why consciousness “ought to matter” to us. I don’t think there are any facts about what ought to matter to us. And Harris certainly hasn’t given us anything amounting to an argument that anything ought to matter to us. Even if Harris were correct that it’s the only thing that could matter, that still wouldn’t show that it ought to matter to us. 

The second issue with this is Harris’s latter remark. He says that he does not think anyone sincerely believes this kind of moral skepticism. Well, Harris is wrong about this:

I believe this kind of moral skepticism.

I don’t think a person who wants to torture everyone is any less objectively moral than me because I don’t think anyone is objectively moral or immoral at all. I really do believe that. And I’m more than a little tired of moral realists proposing that I’m lying about what I believe. It’s lazy, it’s rude, and most importantly, it’s just not true.

Simply dismissing one’s critics as insincere, without any arguments or evidence, is one of the lamest ways of rejecting the position expressed by that person that you could possibly employ. I’ll dub this dialectical telepathy. Dialectical telepathy occurs whenever a participant in a discussion rejects an argument, dismisses a claim, or otherwise makes dialectical moves that turn on speculative claims about the motivations or beliefs of their interlocutors that directly conflict with the explicit stance taken by those interlocutors. Imagine, for instance, insisting that a person who claims to believe in Christianity that they don’t “really” believe what they say they do. This isn’t just a dubious philosophical move, it’s rude. Harris does not have telepathic powers, and is in no position to simply declare that skeptics are insincere. 

Many critics of skepticism are so convinced that skeptical positions are ridiculous or stupid that they seem incredulous at the notion that anyone could be a skeptic. In this, Harris seems just like many academic philosophers who scoff at skeptical takes on morality, or science, or the external world. Scoffing isn’t a substantive objection, and those of us inclined towards skeptical, eliminativist, illusionist, and otherwise sparse ontologies and cautious epistemologies should not be cowed by attempts to pressure us into conformity through shame and ridicule. I am not impressed with the triumphal fist-pounding and foot-stamping of metaphysicians. Ironically, Harris appears to be among their ranks, and in this instance at least Harris is more like contemporary academic philosophers than I am. Skeptical positions will not be expunged by the declarations of our opponents that we’re insincere. Incredulity-fueled pretenses to telepathy are not a serious objection.

Furthermore, even if some skeptics were insincere, this would be moot. Does Harris have an answer to the insincere skeptic or not? Dismissing one’s opposition as insincere doesn’t discharge the legitimacy of their questions. But it sure can allow one to duck out on their responsibility to do so. And that’s exactly what Harris does here. This is the equivalent of saying “You say you want an argument for my position. But you don’t really want an argument because you don’t honestly disagree with me.” Okay, well even if nobody wanted an argument, one could still wonder whether you have one. One’s position doesn’t become true for lack of sincere critics.

Here’s the critical move that Harris then moves on to make:

Let us begin with the fact of consciousness: I think we can know, through reason alone, that consciousness is the only intelligible domain of value. What is the alternative? I invite you to try to think of a source of value that has absolutely nothing to do with the (actual or potential) experience of conscious beings. Take a moment to think about what this would entail: whatever this alternative is, it cannot affect the experience of any creature (in this life or in any other). Put this thing in a box, and what you have in that box is—it would seem, by definition—the least interesting thing in the universe.

Note the ambiguity and lack of clarity in this remark. Harris emphasizes here that we can’t think of any source of value that “has absolutely nothing to do with the (actual or potential) experience of conscious beings.” Has nothing to do with experiences in what respect? Harris could think consciousness is required for us to value things, that the only things we can value are conscious states, both, or something else entirely. In other words, there are at least two ways in which value could “have to do” with conscious experience:

1. The target of a value being a conscious state

2. The subject doing the valuing requiring consciousness

Harris’s remarks are ambiguous between these two readings. It may be that consciousness is necessary for any and all instances of morally relevant actions and events, but only if we include (2). It may be, for instance, that the only way in which one could exercise a good will in the Kantian sense is if they make conscious moral decisions or consciously morally deliberate. And it may be that the only way on a virtue theoretic account to develop courage or temperance that one must be conscious, either in the development or exercise of the virtues. More generally, if nobody were conscious, we may believe that people lack the requisite forms of agency for moral responsibility. We might even think that nothing could be valuable at all if nobody were conscious. But note that all this would show is that consciousness was a necessary precondition for anything to have value. It would not show that the only things conscious beings could value was consciousness itself. For comparison: even if one requires eyes (or some physical analogue) to see, it does not follow that one can only see eyes.

This ambiguity exposes Harris’s position to a dilemma. If we include (2), Harris’s position is trivial. It is trivial in the sense that insofar as it were true it would be consistent with pretty much any normative moral theory, most of which don’t reduce moral considerations to facts about wellbeing. Consciousness might be required to value things, but then one could value things other than conscious states. Note that such theories don’t require the denial that consciousness experience matters. You can be a deontologist or a virtue ethicist and think that conscious states or wellbeing matter; it’s simply that such views don’t hold that consciousness experiences and wellbeing are the only things that matter.

A person who values complying with moral duties or developing virtues doesn’t have to only care about duties or rules insofar as they have an impact on people’s wellbeing. One could favor complying with a rule even if it reduced people’s wellbeing, or even if nobody ever found out about it besides yourself. You could say “Aha! But in any case in which you made a decision that reduced wellbeing, or you kept a promise that nobody else knew about, you still had to be conscious to make that decision. So consciousness is involved!” This just conflates (1) and (2) above. Yes, you may have to be conscious to make moral decisions. But this does not mean that you have to keep a promise because it increases wellbeing. You can care about keeping a promise for its own sake. That is, you can think that at least part of what makes keeping a promise morally good is a non-instrumental good: it’s simply good to keep the promise, full stop. It may also be good because keeping promises tends to improve wellbeing, but this isn’t the only reason why one might favor keeping promises.

This is the whole point of deontology. Consequentialists place the locus of moral concern on the outcomes of our actions. Deontologists, in contrast, place the locus of moral concern on compliance with duties, rules, and principles. Even if compliance required consciousness, this does not mean that compliance with the duty or rule in question is morally right because of its impact on anyone’s conscious experiences. That is, whether an action is morally right or wrong turns on whether it is in conformity with some rule or principle. What makes the action morally good or bad is not directly or exclusively about its consequences. While a deontologist can care about consequences, this only serves to illustrate that they’re not insensitive to concerns about outcomes, including wellbeing. But again, there’s a big difference between saying “wellbeing is a really important part of morality, perhaps the most important part,” and saying “morality is only about wellbeing and only can be about wellbeing. It’s literally unintelligible for it to be about anything else.”

What are some things people can care about other than conscious experiences or wellbeing? First, just to be clear, I understand Harris to be claiming that the only intelligible locus of moral concern are changes in conscious states, presumably those associated with “wellbeing,” whatever Harris takes that to be. As such, the only non-instrumental targets of human value that make any sense are whatever set of conscious states fall within the scope of Harris’s conception of “wellbeing.” The problem with this claim is that many people think that it is morally right or wrong to obey certain rules, or comply with certain principles, in such a way that the rightness or wrongness of such compliance cannot be fully explained or justified in terms of the impact such compliance has or tends to have on people’s wellbeing; in fact, in some cases one may have moral obligations that require them to act in ways that would ignore or even reduce wellbeing. 

For example, a person may believe it is wrong to break a promise or act in a way that is misaligned with someone else’s preferences even if that person would never discover this fact (thanks to Tarn on Facebook for this point). For example, someone may believe it would be wrong to have an affair on one’s spouse even if one’s spouse never found out. Or suppose someone asked to have their ashes scattered thrown into the ocean after they die. Their family may feel they have an obligation to comply with this request even if the person in question couldn’t possibly experience this (since they’re dead). One can always insist that in such cases that people are really only motivated or only could defensibly be motivated to adhere to these rules  because of the impact such actions would have on their own or other people’s conscious states. For instance, one might think that a secret affair is wrong because one’s spouse could find out, or because the affair could cause one to feel guilt or shame or undermine the quality of one’s relationship. These are all plausible reasons why one might think a secret affair is immoral. But simply because you can think up reasons why a person’s opposition to a secret affair could be attributable to conscious states of wellbeing, it does not follow that as a matter of descriptive fact this is why any particular person is opposed to secret affairs. It’s an empirical question why they're opposed, and the best way to find out would be to ask that person. I would bet it would quickly become evident to anyone who looked into the matter that people at least claim not to only be concerned with conscious experiences. If so, then we’d need a good explanation for why people would say this, if it weren’t the case, or if there is something incoherent about doing so. For example, we’d have to provide evidence these people are lying or confused. Good luck with that.

Anyone familiar with deontology or virtue ethics could or of normative ethics in general could probably come up with numerous examples themselves. What I don’t understand is how, after a decade, Harris doesn’t seem to have budged on this point. It just is possible to assign moral value to nonconscious considerations and to not reduce the whole of morality to concerns about wellbeing. People can and do think this way all the time. And if Harris wants to claim this is unintelligible, fine: let him present the contradiction in doing so, or explain why this is impossible. Simply declaring that nobody can think of an alternative isn’t adequate. We can and we have thought of alternatives. Again, someone could insist that we can or must only care about whatever nonconscious considerations we cite due to their actual or potential impact on conscious states, but again, simply because one can come up with a possible way in which a claim that ostensibly can’t be reduced to a concern about conscious states could, in fact, actually only be valued in terms of its impact on wellbeing, doesn’t show that it is. 

Furthermore, since the people who are making these value claims will deny that they only value rules or principles of virtues due to their impact on conscious states, Harris is going to need a good account of why these people lack introspective access to the real basis of their endorsement of these values (i.e., their impact on consciousness). For Harris’s view to be correct, everyone who disagrees would have to be conceptually confused or lying, and actually unconsciously agree with Harris. Again, they’d have to b e some kind of crypto-Harrisian. This is possible, but it’d take very, very good evidence to show that everyone who claims to disagree with you is lying or confused. Harris’s “I can’t think of an intelligible alternative to my view and neither can you” move is nowhere near adequate for establishing such a strong claim. Harris’s personal failure to grok other people’s moral concerns is not a legitimate basis for concluding that only Harris’s position is intelligible.

Harris joins the ranks of philosophers who operate under the presumption that people have a whole host of implicit commitments or beliefs and that we can determine what those beliefs are merely by reasoning about it. Recall that Harris claims to know that only conscious states can matter from reasoning alone, i.e., from the armchair, without any empirical investigation. This is as ironic as it is bizarre for someone whose entire thesis centers on the claim that science can answer moral questions. To employ a Harrisism, suppose I said that if we can know anything through reason alone, we can know that Harris is wrong, because we can think of sources of value that have nothing to do with conscious experiences.  A person could simply care that the world continues existing after all conscious beings died. A person could simply care that a promise is kept or an oath honored even if nobody could ever find out. A person simply could value complying with a rule even if they knew it would help no one and yield no good consequences. People appear to hold values like this all the time. Or at least they claim to. Of course, we don’t only know that people appear to care about things other than conscious states by reason alone. We have good reason to believe they care about other things because they say they do. Thus, even in this case, I wouldn’t actually claim to know that people care about nonconscious states from reason alone. It’s an empirical question. And I am open to empirical evidence that Harris is correct. Only Harris doesn’t have any decent empirical evidence that he’s right. If these people are all unconsciously motivated by a concern for changes in conscious states, the onus is on Harris to demonstrate that this is the case. Harris simply has not done so. 

More generally, if science is the best (or only) tool for answering moral questions, why isn’t it also the best (or only) tool for assessing whether people only value conscious states, or are confused or pretending if they claim to do so? 

There is a whole field of science dedicated to questions about what people believe: psychology. But Harris doesn’t even consider that perhaps psychologists would have something to say about the sorts of things people value. No. Harris can settle that question, a question about what everyone, everywhere in the world thinks by sitting down in his chair and thinking about it himself. This is an instance of generalizing from the armchair, and it is one of the worst mistakes armchair philosophers who avoid science tend to make. Harris could say that his claim is a normative one, and not a scientific one: sure, maybe people do care about nonconscious considerations, but this is unintelligible or these people are confused or mistaken. Note, however, that Harris also depicts people who hold contrary views as insincere or pretending. Whether or not this is the case is unambiguously an empirical question, and Harris is not in a position to know whether other people are confused or being insincere from the armchair.

Harris could also say that the claim that only conscious states matter is not, itself, an empirical claim, and thus isn’t subject to empirical inquiry. But then we’re dealing with a normative claim about value that falls outside the scope of scientific investigation, which would appear to conflict with Harris’s whole project. So is there a scientific or empirical way to discover that only conscious states matter? Harris inviting his readers to think of something else that could matter, then confidently concluding that they won’t be able to (notably, without any empirical evidence or engagement with what his readership thinks), is the very definition of armchair philosophy, not empirical investigation. Ironically, Harris is acting just a rationalistically inclined (as opposed to empiricist philosopher) while in the very process of laying the foundations for his scientifically grounded, naturalistic framework.

Empirical data could also put considerable pressure on Harris if we were to discover that many if not most people valued nonconscious states, and were thus subject to the claim that they had confused or unintelligible values. Of course people can be wildly and confused and mistaken about things, but all else being equal, I find it plausible that people don’t hold stupid and confused views, and that if your theory says they do and mine doesn’t, that’s a point in favor of my theory. A small point, but a point nonetheless. And when Harris has so few points going for his own view, even a small point can make a difference.

It might also put pressure on Harris if we found that most philosophers after reflecting on morality continued to or began to value nonconscious considerations. Sure, they might also be confused or mistaken, but lots of well-informed people disagreeing with you is at least some evidence that you may be mistaken. And since Harris has given us so little to go on to begin with, such evidence carries comparatively greater weight than in situations where we have more arguments and evidence for and against a position. 

I’ve sometimes encountered philosophers who insist that this or that position logically entails some other position, and that because you hold the initial view, you must be committed to its entailments. One might say people are logically committed, but it does not follow as a matter of psychological fact that you agree or endorse or implicitly believe whatever entailments others attribute to you. Even if you find someone else’s views silly, and even if you think your own views are correct, it does not follow that other people secretly agree with you. Harris’s remarks amount to declaring that skeptics who claims to disagree with him are crypto-Harrisians who covertly agree with him or else have unintelligible nonsense values. But Harris seems to lean on the suggestion that people who claim to disagree are insincere or pretending.  It’s ridiculous. I could make the same claim, and it would be just as toothless: 

Harris does not sincerely believe that consciousness is the only thing that matters. He can press this point with such ferocity that it may pass for sincerity, but he doesn’t really believe it. (disclaimer: I don’t actually think this, see the surrounding remarks for context)

Would Harris, upon reading this, simply concede “ahh, shit, you got me!”? No. Because Harris knows what he actually thinks. Likewise, as a moral antirealist, I know what I think better than Harris does. And I do not think it’s true that only consciousness can matter, nor do I think that it’s true that only conscious states “ought to matter.” 

For that matter, I don’t even believe phenomenal consciousness exists, and given Harris’s remarks on the matter, I might not even agree that the sort of consciousness Harris thinks there even exists, because it might itself be literally unintelligible. I am, after all, a qualia quietist. Or maybe I’m insincere about this, and just pretending not to believe in qualia, too?

Setting aside my own views on the matter, if it really is the case that we can only value conscious states, this presents another problem for Harris: what, exactly, are conscious states? And do nonphilosophers all share the same conception of consciousness, such that insofar as they value conscious states, they value the same sort of thing? If so, how could we know? Suppose, for instance, that Dennett and Frankish are correct, and illusionism about consciousness is correct. Harris has explicitly stated that:

Consciousness is the one thing in this universe that cannot be an illusion.

 Harris includes the following note:

While Descartes is probably the first Western philosopher to make this point, others have continued to emphasize it–notably the philosophers John Searle and David Chalmers. I do not agree with Descartes’s dualism, or with some of what Searle and Chalmers have said about the nature of consciousness, but I agree that its subjective reality is both primary and indisputable. Of course, this does not rule out the possibility that consciousness is, in fact, identical to certain brain processes.

And, again, I should say that philosophers like Daniel Dennett and Paul Churchland just don’t buy this. But I do not understand why. My not seeing how consciousness can possibly be an illusion entails my not understanding how they (or anyone else) can think that it might be one.

I also don't buy this. I think Harris is wrong. Also, for what it’s worth, I don’t think Paul Churchland is an illusionist. Furthermore, I don’t think nonphilosophers typically hold substantive philosophical views about consciousness according to which it consists of qualia or phenomenal states or indeed anything in particular. That is, I don’t think ordinary people have some shared “theory of consciousness,” that could be a shared target of value in the first place.

I think philosophers imagine some singular, shared essence or phenomena, consciousness, around which a variety of supposedly unified experiences gravitate. But I see little reason to think nonphilosophers think this way, and join in the philosopher’s metaphysical presumptions. I’m not alone in thinking this. Pete Mandik has proposed meta-illusionism, the view that philosophers are subject to the illusion that nonphilosophers are subject to an illusion, in this excellent paper. 

Harris himself may be a victim of this meta-illusion, and may mistakenly operate under the presumption that people generally think about consciousness in the way Harris does. I don’t know how Harris, specifically, thinks about consciousness, since as usual Harris eschews engagement with the philosophical literature on the topic. 

At present, there is little empirical evidence to suggest that nonphilosophers believe in qualia or phenomenal states, or endorse the hard problem of consciousness. See my discussion on some of this literature here. Maybe these studies are flawed. If so, then we still don’t have positive evidence in favor that people’s views generally accord with Harris’s in a way that is sufficiently close that whatever Harris means when he says that we can only value conscious states that we should be confident this maps onto how nonphilosophers think. 

What’s more, even if nonphilosophers did think this way, then I’d simply think illusionism was true, and that both Sam Harris and most nonphilosophers were mistaken about consciousness. In other words: either most people don’t value the sorts of “conscious states” Harris thinks are the only appropriate target of value because they don’t share a close enough conception of conscious states as Harris, or they do value those conscious states, but, since Harris holds an incorrect view about consciousness (some non-illusionist view), both Harris and nonphilosophers are mistaken about what consciousness is. And since the sort of consciousness Harris and they would think exists and has value doesn’t exist (and may not even be intelligible), it wouldn’t be an appropriate target of value either way.

Harris doesn’t effectively grapple with the question of consciousness any more than he grapples with the question of morality. Insofar as his views about morality turn on his views about consciousness, any shortcomings in his perspective on the latter will bleed into the former. In this case, I suspect the problems with Harris’s views about consciousness may be sufficient to sink his views about morality, but it’s hard to know, because he also has a muddled perspective on consciousness and has not adequately engaged with nor seem to have understood the illusionist position.

6.0 Wellbeing

One of the most frustrating elements of Harris’s moral theory is that he construes what is morally right or wrong in terms of wellbeing, but his notion of wellbeing is so broad, expansive, and ill-defined as to be practically useless. Harris characterizes well-being in a number of ways in TML, but none of them are especially clear. At one point, Harris says:

I will argue, however, that questions about values—about meaning, morality, and life’s larger purpose—are really questions about the well-being of conscious creatures. Values, therefore, translate into facts that can be scientifically understood: regarding positive and negative social emotions, retributive impulses, the effects of specific laws and social institutions on human relationships, the neurophysiology of happiness and suffering, etc.

Well-being is cashed out here with a non-exhaustive list of the sorts of things Harris has in mind. It’s not clear what relation the items on this list have to each or even how they relate to “value.” Is a retributive impulse a value itself, or is it in some way related to values? If it is a value, that’s an odd notion of value. If it’s related to values, well, how is it related to values? This may look like a substantive account, but it isn’t. The relation between various items on this list to values isn’t adequately explained. 

Later, Harris says this:

While the argument I make in this book is bound to be controversial, it rests on a very simple premise: human well-being entirely depends on events in the world and on states of the human brain. Consequently, there must be scientific truths to be known about it.

Depends how? Harris isn’t clear on the dependency relation. Let’s make a gastronomic comparison:

Human taste preferences entirely depend on events in the world and on states of the human brain. Consequently, there must be scientific truths to be known about them.

There is a sense in which this is true: our ability to taste anything at all depends on events in the world and states of the brain. Does that entail gastronomic realism? No, it does not. Harris simply isn’t clear about what dependency relation he has in mind.

Harris also says this:

Many readers might wonder how can we base our values on something as difficult to define as “well-being”? It seems to me, however, that the concept of well-being is like the concept of physical health: it resists precise definition, and yet it is indispensable.In fact, the meanings of both terms seem likely to remain perpetually open to revision as we make progress in science.

I don’t have much to say about this, other than that there are efforts to discuss what health is. Harris has a whole book to work with. This strikes me as a pretty poor excuse for not even giving an attempt at a partial and imperfect account. Later, Harris says:

If we define “good” as that which supports well-being, as I will argue we must, the regress initiated by Moore’s “open question argument” really does stop. While I agree with Moore that it is reasonable to wonder whether maximizing pleasure in any given instance is “good,” it makes no sense at all to ask whether maximizing well-being is “good.”

Here we have a problem: It’s one thing to say wellbeing is good. It’s another to say that maximizing wellbeing is good. This claim from Harris is absurd. It makes no sense to ask whether maximizing wellbeing is good? Maximizing whose wellbeing? Everyone’s? Again with the lack of clarity. Harris is simply not clear. I knock analytic philosophers a lot, but Harris’s lack of clarity is profoundly frustrating in comparison. If it’s everyone’s, how would it not make sense to wonder whether it’s good? Suppose that maximizing everyone’s wellbeing required you to donate almost all of your money and give up almost all of your belongings and live in virtual squalor, because their need is greater than yours? Does it make no sense to ask whether that would be “good”? 

Harris then says:

It seems clear that what we are really asking when we wonder whether a certain state of pleasure is “good,” is whether it is conducive to, or obstructive of, some deeper form of well-being.

Harris makes the same mistake many philosophers do: claiming something seems to be the case without specifying who it seems that way to, and using “we” without qualification. Here we’re simply told it “seems clear.” It seems clear to who? To Harris? To people in general? Who is “we”? Everyone? Again with the lack of clarity.

Note, too, the irony behind Harris’s methodology. Harris says it literally makes no sense to wonder whether maximizing wellbeing is good. Why? Because it seems clear (to Harris) that “we” (everyone?) is implicitly committed to a conception of good that is fully reducible to wellbeing.

Harris reaches a very strong conclusion: it makes no sense to ask whether it’s good to maximize wellbeing using the completely unscientific method of making a sweeping judgment about what everyone thinks based on how things personally seem to Harris. That is, Harris is making an armchair pronouncement about what people mean by particular terms. Sweeping claims to know what everyone, everywhere means, and how they think all made from the armchair without a single iota of actual empirical evidence (from e.g., comparative linguistics, cultural anthropology, psychology, history, etc.) is exactly the type of objectionable methods analytic philosophers use. Notably, it is also completely unscientific. So Harris’s whole approach to getting the ball rolling on his scientific approach to morality involves the utterly unscientific approach of armchair conceptual analysis and armchair psychology. 

Note that Harris presents no arguments, no conceptual analysis, no examples, no engagement with the literature, no anything to support this declaration about how things seem. He simply asserts it. At the very least, rationalistically inclined analytic philosophers would typically do more than this. In other words, even the least scientifically motivated analytic philosophers would employ more rigorous and scientifically respectable methods than Harris is employing here.

We get a bit more on well-being and value:

Defining goodness in this way does not resolve all questions of value; it merely directs our attention to what values actually are—the set of attitudes, choices, and behaviors that potentially affect our well-being, as well as that of other conscious minds. While this leaves the question of what constitutes well-being genuinely open, there is every reason to think that this question has a finite range of answers.

Why is there every reason to think there’s a finite range of answers? We’re not told. Again, Harris simply makes an assertion, but provides no arguments or evidence for that assertion. Note, too, the total absence of any serious effort to spell out what well-being might be. Harris’s framing here also seems to present it as a matter to be discovered, and not a matter over which people could ineliminably disagree. This is perhaps one of the most serious issues with Harris’s account. The issue is that, even if we suppose that only wellbeing is an appropriate target of value, and thereby dismiss all accounts of morality that don’t center on conscious experiences of “well-being” and “suffering” to use Harris’s terms, all this would mean that the set of nonconscious states are not appropriate candidates of value, and that the set of conscious states are appropriate candidates of value. What this does not entail is that all rational persons would all conscious states that fall within the scope of “well-being” the same as one another. To get very precise, suppose we signify wellbeing states with w1, w2, w3 (the w-set), and so on, and all nonconscious states as n1, n2, n3 (the n-set). Suppose someone thinks we could value both states of wellbeing and nonconscious states:

Possible values = {n1, n2, n3…} + {w1, w2, w3…}

Harris argues that we cannot intelligibly value the set {n1, n2, n3…}

I think Harris is mistaken, and has provided virtually nothing by way of a decent argument for thinking this, but let’s just grant it for the sake of argument. Now people can only assign value to:

Possible values = {w1, w2, w3…}

And let us suppose there are scientific facts about what is conducive to any given item in the set. If one society emphasizes w1 but not w2, while other favors the opposite, these represent two peaks on the moral landscape, two ways of living flourishing lives.

The problem is that Harris’s account at best only shows that it’s not intelligible to care about the n-set, not the w-set. Yet this is consistent with someone not caring at all about other members of the w-set. Is there any scientific fact that could show that such a person is incorrect? It’s not clear that there is, or how there could be. If, for instance, a person cares about one type of conscious state, but not another, what are they wrong about? Harris could insist they “don’t know what they’re missing” but he doesn’t present any arguments or evidence or reasons to think this is a necessary condition of not caring about those states. What if that person does know, but they just don’t care? Suppose, for instance, one society cares about w1 but not w2. They encounter a society that maximizes w2 but not w1. They think this is bad, so they eradicate this society and replace it with members of their own society, maximizing w1. Did they make a mistake? If so, what was the mistake? They increased w1! What Harris has not shown is that one must assign equal moral worth to all members of the set. He’s only shown that one’s options are restricted to members of the set, not that one must treat all members of the set equally.

To provide a concrete example: if someone argued that we could only care about fruit and could not care about vegetables, it would not follow that everyone must care about all fruit equally. They could find that they love apples and hate bananas. And they could embark on a quest to wipe out all bananas and replace them with apples. If one’s only overarching moral mandate is to “maximize fruit,” and such activities did maximize fruit, then they’d be “good” with respect to the mandate to maximize fruit. 

Harris’s arguments mostly center on attempting (not very successfully) to counter arguments from people who ask why we should care about fruit at all. This obscures the internal problems with Harris’s account, which is papered over by hand-wavy remarks about how wellbeing is an expansive concept. Sure, let’s grant that it is. Even if it is, Harris is still going to have his work cut out for him to address additional normative brushfires the arise once one admits multiple, distinct items into one’s set of possible values. How do we resolve conflicts between communities whose w-set values are calibrated differently? Are there scientific facts about how to resolve those conflicts? If so, they can’t be reducible to facts about what increases one or another member of the w-set, because all members of the conflict could simply agree:

“Yes, I know that this action would increase w2 a great deal, but I don’t care about w2, I only care about w1.”

It’s unclear how this person is, or even could be, making a mistake. Such a person does care about wellbeing, and does want to maximize it; it’s just that they want to maximize one subset of it at the expense of all the others.

Harris laments that:

I have heard from literally thousands of highly educated men and women that morality is a myth, that statements about human values are without truth conditions (and are, therefore, nonsensical), and that concepts like well-being and misery are so poorly defined, or so susceptible to personal whim and cultural influence, that it is impossible to know anything about them.

Harris should seriously consider that maybe he’s heard this from thousands of people because they’re correct, at least about some of these claims. Even if we grant that science is the best means of achieving our goals, and that our goals could only ultimately be cashed out in terms of human values, this does not mean we must assign equal moral weight to exactly the same conscious states. In short: If wellbeing is a sufficiently expansive notion, shared emphasis on wellbeing will be insufficient to settle conflicts of value, while if it is too narrow, it will exclude as a matter of descriptive fact the targets of value many people actually have. In the latter case, these people would have little reason to accept Harris’s account of morality, since it excludes what they care about, and, even if they did accept his account of morality, they’d simply have little reason to be moral, and would be better off favoring some nonmoral set of values.

Harris also fusses over the distinction between what we can know in principle vs. what we can know in practice to make a point: that there are scientific truths that could be known in principle even if in practice we’d never be able to figure them out. Harris invokes the notion that we may be unable to determine how many birds are currently flying over the earth, but there is some fact of the matter, even if we don’t know what it is.

Some people don’t think fairly clearly about this distinction and so they think one way of rejecting a scientific case for increasing wellbeing is to insist it’s really hard to figure it out. There may be cases where the empirical tools available to us simply don’t allow the level of precision one might wish for. This, too, is a common objection: if you can’t give me exact units, then I’m going to dismiss your whole case! This is silly, too. Harris is absolutely on solid ground in rejecting these sorts of objections.

However, there are two major problems that Harris doesn’t  adequately address.

First, commensurability. Two or more instances of human experience are commensurable when we could make cross-experiential comparisons according to a shared evaluative standard. Think of it as a common currency, or at the very least a standardized exchange rate that would allow us to convert the “currency” of one type of experience into another: units of goodness, so to speak. We can compare the masses of different objects because they share underlying properties in common that render such comparisons straightforward: we could compare the mass of a banana on earth to the mass of a neutron star millions of light years away, and precisely quantify just how much more mass the neutron star (or the banana, if it is a very large or dense banana) has.

At first glance, it may seem reasonable to make inter-experiential comparisons on similar grounds. There is some scientific fact of the matter about how a given action (e.g., stealing) or policy (e.g., an increase in taxes) will impact human wellbeing. However, to evaluate the impact of these actions on human wellbeing, we need some operationalization of wellbeing. This can be cashed out in operational terms: self-reported happiness, income, suicide rates, and so on. In all of these cases, it’s not as though we’re so stupid we think self-reported happiness just is wellbeing. Such a measure would be both incomplete (there is more to human wellbeing than how happy we think we are) and inaccurate (people will lie, or interpret questions differently, or assign a different number to the same degree of happiness as someone else, e.g., one person’s “5” on a 7-point scale may be equal to a “6” on another person’s scale). Such measures are thus indirect and partial measures intended to capture some latent variable. A latent variable is a variable that cannot be directly observed. Instead, its presence must be measured via observable variables which serve as proxies for the latent variable, and which in turn must be operationalized in accord with some model of the phenomenon in question.

Suppose there is a single, unified latent variable, e.g., “wellbeing,” or “happiness,” that is just as real and measurable in principle as the mass of objects. It may be extremely difficult to measure it, but it would still be measurable in principle. And we could do our best, using a variety of observable variables and models, to best approximate the facts about what is or is not conducive to wellbeing, so-understood. And, we might suppose, there are a variety of different forms wellbeing can take, none of which are qualitatively superior to one another. So, we could use “height” as a dimension for representing total quantity of wellbeing, and then map location across a plane, or landscape to represent these qualitative differences. Two peaks on that landscape could both be at equal altitudes, but represent different peaks on that landscape, because they represent two qualitatively distinct locations on the same wellbeing landscape.

This is the picture Harris appears to be offering us. The problem is that it’s not at all clear that there is any scientific fact about how to make qualitative distinctions. In other words, at any point along the landscape, there may be some scientific fact about the “altitude” of some measurable outcome, but it does not follow that there is some scientific fact about how to compare two peaks. How, for instance, do we compare one society which optimizes for liberty at the expense of pleasure to a society that optimizes pleasure at the expense of liberty? Let us suppose that both societies are well-structured and idealistic forms of the intended systems: a libertarian utopia of abundance, and a hedonic, Bacchanalian world of abundance, both with little or no crime or disease or misery or suffering. Which one “has more wellbeing”? Which peak is higher? Is there a fact of the matter that we could discover using the tools of science, not simply in practice, but in principle? I suspect the answer is a straightforward “no.” 

Harris’s conception of “wellbeing” faces a dilemma. Either it is so expansive it includes mutually incommensurable conceptions of what sorts of lives we should live, or it so restrictive that it excludes conceptions of a flourishing and good life that many people endorse and would continue to endorse on reflection, and would therefore have no incentive to construe morality in Harrisian terms. This dilemma alone may be sufficient to undermine Harris’s entire project.

There is a second, related problem, too. Even if we take ourselves to be dealing with a putatively measurable latent variable that represents a single altitude range, a single line according to which we could in principle plot conscious experiences as higher or lower with respect to, e.g., valenced conscious experiences, such that we have negatively-valenced states (suffering) through neutral and positively-valenced states (happiness), we still may encounter a problem: such a presumptive latent variable may flatter our desire for mathematical elegance, but it may, at best, be a crude metaphor for what is also, in actuality, an incommensurate array of non-fungible and qualitatively distinct conscious states.

Dennett (1995) makes a similar point with regard to conscious experiences and suffering:

Many discussions seem to assume tacitly: (1) that suffering and pain are the same thing, on a different scale; (2) that all pain is “experienced pain”; and (3) that “amount of suffering” is to be calculated (“in principle”) by just adding up all the pains (the awfulness of each of which is determined by duration-times-intensity). These assumptions, looked at dispassionately in the cold light of day—a difficult feat for some partisans—are ludicrous. (p. 708)

Dennett goes on to ask us to consider the total amount of suffering we experience in a given year, then choose between:

1. Experiencing that suffering normally, distributed as it is across the year

2. Experiencing all of that suffering compressed into five minutes

Dennett adds that we could even include a multiplier for (b), where it is double or quadruple the total amount of suffering we’d experience in (a). Like Dennett, there is at least some sense in which I would probably opt for (b) even with a very large modifier. This would seem to imply at least some of us think a world with much more suffering, indeed, much more of our own suffering, would be much better than a world with less suffering. Temporal compression of our suffering reflects just one way in which the distribution of suffering is independently relevant to our assessment of its desirability. And people may very well different in what sorts of distributions of experiential states they’d favor, and there may be no scientific fact about which of these preferences is correct. The preferences are not themselves reducible to facts about wellbeing or flourishing, nor are they quantifiable or rankable along a single spectrum, whether it be ordinal, interval, or ratio. 

In the previous paragraphs, I argued that we cannot make comparisons between qualitatively distinct conscious states and variation in how actions and policies influence our conscious states. Dennett’s point only compounds the problem. Even apparently commensurable comparisons with respect to a single dimension of wellbeing or value aren’t genuine instances of a single, quantifiable continuum of morally better and worse experiences. The appearance of such continua is an illusion, an artificial compression, an attempt to subject qualitative, or nominal differences to a single operational metric, a proposed “metric system” of conscious experience that purports to capture a genuine and scientifically verifiable (if in principle) feature of what the world is like. I agree with Dennett, who says:

It seems obvious to me that something is radically wrong with the assumptions that permit us to sum and compare suffering in any such straightforward way. But some people think otherwise; one person’s reductio ad absurdum is another’s counter-intuitive discovery. We ought to be able to sort out these differences, calmly, even if the best resolution we can reasonably hope for is a recognition that some choices of perspective are cognitively impenetrable. (p. 709)

This remark ends on a pessimistic note, but there is a flair of optimism as well. While I, like Dennett, have come to regard such attempts at summing and comparing experiences as an absurdity, others, such as Harris, may maintain that such a project is feasible. Unfortunately, Harris seems to set his position up against a host of what are likely very real but unfortunately weak opponents, and to claim victory in the wake of their defeat. Only victory cannot be secured by successfully rebutting one’s weakest opposition, but by directly engaging with the strongest opposition to one’s views. When it comes to the matter of wellbeing, Harris simply hasn’t done that. Perhaps serious engagement with the philosophy of consciousness would prove too boring.

7.0 Conclusion

This has been quite the odyssey. I find the publication of The Moral Landscape fortune for me, but unfortunate for the world. I had been an ardent proponent of new atheism in its heyday. While cracks had begun to form, my disappointment and frustration with Harris’s seemingly cavalier attitude in The Moral Landscape played a formative role in widening those cracks to a fatal degree, and allowing me to recognize the shortcomings of new atheism. 

A single factor in what may have been an inevitable departure from that movement is hardly compensation for the amount of people who I worry have been misled by TML and by Harris more generally. Harris’s intransigence renders him seemingly impervious to correction. With time, it became apparent to me that almost any criticism directed at Harris was diagnosed by Harris as a mischaracterization or misunderstanding, as if every critic must be malicious or confused, and that few, if any, could ever make a cogent point that ran contrary to his conclusions. This podcast episode brought a wave of nostalgia with it, but a very bittersweet nostalgia. There was a time in the late 2000s when I felt I was riding the crest of a wave on the tides of history, where atheism would finally become a respected, mainstream perspective, and religion would finally recognize the terminal diagnosis that had been delivered during the enlightenment, and shuffle off to cultural equivalent of hospice care.

That didn’t happen. New atheism is now memorialized in derelict wikis and dusty blog posts. In barely a decade, it became a relic. These days, ecumenical engagement with religion is the more productive ethos among internet atheists. TML may have aspired to offer a secular alternative to religious narratives on the matter of morality. In that respect, it failed. Far more robust secular accounts of morality exist and have existed in the philosophical literature that better address virtually every aspect of Harris’s account, from metaethics, to normative ethics, to applied ethics. I say this as an inveterate critic of academic philosophy. 

Almost all of my ire is directed against the academy, and against moral philosophy in particular. I have metaphilosophical and methodological qualms with virtually every aspect of mainstream academic philosophy, along with a critical attitude towards the profession’s insularity and with what often manifests as a contemptuous attitude towards other fields and towards input into philosophical matters from outside the profession. Philosophy has a gatekeeping problem. This very blog is itself an effort to open the gates and let others in. As a critic of such gatekeeping, I want popular books, blogs, podcasts, and so on to address topics in moral philosophy and psychology. I just want them to do it well. 

My criticism of Harris is thus not motivated by a hostility towards an “outsider” stepping on philosophical turf. It is motivated by frustration with someone doing a bad job. If you’re going to step on philosophy, great. I’m all for it. But when you do, wear a better pair of boots.

8.0 Retractions and corrections

2.0 Harris contradicts himself almost immediately.

As I noted in our response to the video, Harris contradicts himself almost immediately. From 2:07 to  3:20, he says: 

John Searle once pointed out that we should distinguish between epistemological and ontological senses of objectivity [...] Ontology relates to questions about what exists. For instance, is there only one type of stuff in the universe? Are there only physical things, or are there really existent things which are not physical? For instance, do numbers exist beyond their physical representations, and if so, how? Science is fully committed to epistemological objectivity. That is, to analyzing evidence and argument without subjective bias. But it is in no sense committed to ontological objectivity. It isn’t limited to studying objects, that is, purely physical things and processes. We can study human subjectivity, the mind as experienced from the first person point of view objectively, that is, without bias and other sources of cognitive error.

Note the two bolded remarks. Harris first characterizes ontology to be concerned with what exists, including both physical and nonphysical considerations. He then says that science isn’t committed to ontological objectivity because it isn’t only concerned with physical things. But ontological objectivity isn’t, by his own characterization, concerned with physical things. This is a serious error. He managed to completely contradict himself with barely a sentence or two in between the conflicting remarks. 

What’s more remarkable is that Harris probably has a whole team who could review his work, and at the very least has extraordinary access to numerous people that could check his work before he made it publicly available. If I had those resources available to me, I’d certainly avail myself of them. Didn’t anyone proofread his script? How does he butcher basic distinctions this badly? This isn’t a minor error buried an hour into the video. It occurs right at the start. This is not an auspicious sign for an endeavor ostensibly intended to clarify his stance for his audience.

[Edited to add on 5/9/2024]

Harris did not contradict himself. My thinking was that Harris described what “ontology” was with the goal of employing the term in his subsequent remarks. Since ontology is concerned with both physical and non-physical things, it looked to me like his invocation of the term “ontological objectivity,” involved the presumption that ontology was only concerned with physical objects. However, this was based on my misreading of Harris’s claim that science is “in no sense” committed to ontological objectivity to mean that Harris thinks science rejects ontological objectivity as a position. Ontological objectivity is not a position that is exclusive with respect to ontological subjectivity, but a distinction used to distinguish different “modes of existence.” The distinction is between that which is ontologically objective (mountains, tables, chairs) and that which is ontologically subjective (pains, pleasures, tickles).

Harris was thus not suggesting in the latter remark that ontology is only concerned with that which is physical, but was instead saying that science isn’t restricted to the study of ontologically objective entities.

This does raise some questions. First, does Harris think that only ontologically objective entities are “physical,” but that ontologically subjective entities are non-physical? If so, then does Harris endorse some form of dualism? If ontologically subjective entities are non-physical, then does Harris think science can study “non-physical” things, specifically? I’m not quite sure what Harris thinks about this divide, and it may be a partial misreading of Searle’s own views (or a reflection of Harris’s own modification to those views) but there is no contradiction here.

3.0 Harris’s metaphysics are muddled

Harris claims that science is only committed to epistemological objectivity, but not ontological objectivity. This appears to be tied to his misconstrual of ontological objectivity as only concerned with the study of physical objects. But, Harris adds, “We can study human subjectivity, the mind as experienced from the first person point of view objectively, that is, without bias and other sources of cognitive error.” 

Does Harris think subjectivity and the mind aren’t purely physical things? His remarks would imply this is the case. But how would this show that science isn’t committed to ontological objectivity? If ontological objectivity were about physical objects, and consciousness wasn’t physical, then we could only study consciousness by dropping a commitment to ontological objectivity. This is why I think Harris does so, but this mistake is predicated on his misunderstanding of what ontological objectivity is.

Harris hasn’t actually provided a good rationale for abandoning ontological objectivity. If he did, it would suggest that science is in no way committed to the objective existence of tables or trees or hydrogen or oxygen. I’m a scientific antirealist, so I’m happy to go along with this. But is Harris? I’m not so sure. The problem here isn’t that Harris has convincingly committed himself to scientific realism or antirealism. If he did, he would appear to be a scientific antirealist, and, by extension, a moral antirealist. 

I’m not comfortable attributing either of these views to Harris. Instead, I think, ironically, that the most charitable stance to take is that Harris is confused about Searle’s distinction between epistemological and ontological objectivity, and any remarks downstream of this are the result of that confusion. Otherwise, Harris would appear to be either a scientific and moral antirealist, or agnostic on these positions. It’s hard to say. Maybe Harris thinks all these labels and terminology are too boring to discuss. But this is exactly the problem with dismissing such terms. A familiarity with the concepts and distinctions behind these terms can allow one to locate their position within the mainstream positions that are already out there, or to recognize those specific ways in which their position departs from mainstream positions. 

[Edited to add on 5/9/2024] These remarks are based on a misreading of Harris. As I said to a critic in the comments:

You’re right. I did misinterpret Harris. A plausible reading of Harris is that he’s using the term “committed” in the sense of being “exclusively dedicated to.” On this reading, Harris would be saying that science is exclusively dedicated to epistemological objectivity ( as opposed to epistemological subjectivity), but it is not exclusively dedicated to ontological objectivity (since it is concerned with both ontological objectivity and ontological objectivity). In other words, while science is all about epistemological objectivity alone (i.e., approaching things in as unbiased a manner as possible in such a way that we can decisively resolve scientific questions), it is not all about ontological objectivity alone, because it is not only concerned with the study of mountains and tables, but with pain, pleasure, and other states of conscious beings.

I don’t take this to be an obvious or straightforward use of the term “committed,” but it makes more sense than how I was taking his remarks. As such, I was definitely mistaken in claiming that only my interpretation was the correct one.

However, there is some indication that Harris is simply confused about the relevant terms and distinctions. Here’s an excerpt from The Moral Landscape where Harris (2010) echoes earlier remarks about the epistemological objectivity:

However, many people seem to think that because moral facts relate to our experience (and are, therefore, ontologically “subjective”), all talk of morality must be “subjective” in the epistemological sense (i.e., biased, merely personal, etc.). This is simply untrue. I hope it is clear that when I speak about “objective” moral truths, or about the “objective” causes of human well-being, I am not denying the necessarily subjective (i.e., experiential) component of the facts under discussion. 

This is fine, as far as it goes. Perhaps people do conflate the notion of something relating to our experience to the notion of it being biased. However, Harris goes on to say:

References

Dennett, D. C. (1995). Animal consciousness: What matters and why. Social Research, 691-710.

Harris, S. (2010). The moral landscape: How science can determine human values. New York, NY: Simon and Schuster.