Counting Blades of Grass

Twitter Tuesday #28

1.0 Introduction

This week’s Twitter Tuesday features the “performative inconsistency argument against value nihilism” from @SecularOutpost. Here it is:

As you can see, this argument is attributed to Philip Goff. Note that SecularOutpost is quick to clarify that this is a restatement of Goff’s views, not a direct quote. I asked Goff about this characterization. Goff clarified that this argument should be attributed to Bart Streumer, and expressed doubt that the rendering of this argument is quite correct (note: I’m paraphrasing Goff here; if he’d like me to add a quote or comment I’d be happy to edit this section to include a statement from Goff). 

So we’ll call this the Outpost Objection, rather than Goff’s own objections. It’s worth emphasizing that SecularOutpost is very clear it’s a restatement and that Goff may not endorse it, so I think everything stated here is in good faith, even if it isn’t ultimately what Goff would endorse.

It might make sense to evaluate the Outpost Objection first, however, there’s a very good reason not to do this: I don’t know how the term “value nihilism” is being used. This objection also got me interested in what Goff says in the chapter this argument is drawn from, so I went ahead and looked that up. It’s the first chapter of Goff’s (2023) Why? The purpose of the universe. Let’s take a detour to assess what Goff has to say, then return to the Outpost Objection. That should hopefully give us a sense of what value nihilism is.

2.0 Goff’s views on value nihilism

Goff describes value nihilism as follows:

There are no facts about better or worse, or about what ought to be done. In other words, value is an illusion. (p. 11)

It’s not the view that we can all make our own meaning and live meaningful lives on that basis. That’s the subjectivist position I came to reject. Value nihilism is the view that every human activity is as pointless as counting blades of grass. (p. 11)

When described in this way, I reject value nihilism. I do think there are facts about what’s better or worse, and what ought to be done. I don’t think value is an illusion. And I do think my life and other people’s lives have meaning, and that such meaning more or less turns on our attitudes or desires. Since I don’t endorse value nihilism, objections to it may pose little or no threat to my own views.

Instead, I endorse Goff’s characterization of subjectivism. Goff is careful to note that “the word ‘subjectivism’ is used in a bewildering variety of ways in philosophy” (p. 5). As long as we’re clear, I am fine with “subjectivism” as described by Goff, and not some other form of subjectivism, this shouldn’t pose any serious issue. I remain wary of people insisting I claimed to “be a subjectivist,” then infer that I’m a “subjectivist” in some other sense than that I think my life and other people’s lives can have meaning. 

Since I endorse subjectivism of this form, it makes sense to address Goff’s discussion on “subjectivism” as described in the chapter. I’ll return to nihilism and to the Outpost argument after that.

2.1 Goff’s Subjectivism

Goff says that he decided he was a subjectivist at the age of fifteen, but doubts eventually emerged:

The first thing that raised doubts about subjectivism was realizing that subjectivism about meaning has some pretty counterintuitive implications. (p. 6)

Goff goes on to describe a person whose life is solely dedicated to counting blades of grass:

Imagine a hypothetical person Susan (Susan is the start of all of my books), who has a single overriding goal that she bases her life around: counting blades of grass. (pp. 6-7)

Goff goes on to clarify that Susan values counting blades of grass, but doesn’t enjoy doing so, adding that there’s “no incoherence here.” I agree. Someone can value something without enjoying it. Goff asks us to imagine that Susan “sweats, toils, cheats, and lies” in the service of counting blades of grass, and “manages to count a huge number of blades of grass during the span of her life” (p. 7). What’s the issue? According to Goff:

If subjectivism about meaning is correct, we ought to say that Susan had a deeply meaningful life, given that she managed to spend a great deal of time engaged in an activity she valued highly. (p. 7)

I have some concerns with this remark. What I would say about this scenario is that if Susan valued doing something and did it her whole life, then she did something she found valuable her whole life: more or less a tautology. 

Goff doesn’t preserve the language of “value” and use it consistently across cases, though. That is, Goff could have said that if subjectivism is correct, we ought to say that Susan had a life in which she did something she found very valuable, given that she managed to do something she valued highly. This would be trivial, of course, and perhaps not worth saying. But I think that’s about all I could say about the scenario.

Nothing else follows from Susan valuing her life, though. I have yet to encounter, in this scenario or any other, anything I found counterintuitive about subjectivism. This is probably because I deny subjectivism has any substantive implications in the first place.

Consider this scenario. We’ve been told Susan didn’t enjoy the activity, so that’s been ruled out by stipulation. But we don’t know anything else. Yet Goff presents another term, deeply meaningful life. I don’t know what Goff means by this. If “deeply meaningful” just means “a life in which one does that which they consider highly valuable,” then it’d be trivially true that Susan lived a “deeply meaningful” life, in which case it wouldn’t make much sense to ask whether Susan had a “deeply meaningful” life: she would have had one, because this would be a logical entailment of what was stipulated in the scenario. So presumably Goff doesn’t mean “lived a life she considered valuable” by “deeply meaningful.” But if “deeply meaningful,” doesn’t mean that then what does it mean? 

I don’t know, because I don’t know what Goff means by “deeply meaningful.” Did she consider her own life “deeply meaningful”? If so, what did that amount to? If a deeply meaningful life is one that involves at least some degree of enjoyment, then she didn’t have a meaningful life. Goff adds that there’s nothing incoherent about this, and I agree this is not incoherent (though Goff might still consider it counterintuitive or implausible or irrational, even if coherent, whereas I would not). Yet this would also be a trivial observation about the Susan scenario, because the fact that she didn’t enjoy her life was stipulated. All it would amount to is that it is logically possible for a person to value activities they don’t enjoy. And if we stipulate a scenario in which a person did something they valued but didn’t enjoy, and “deeply meaningful” just means “enjoyed,” then it’s true via stipulation that they didn’t live a deeply meaningful life. So presumably Goff doesn’t mean, by “deeply meaningful,” a life that the person living it valued, nor a life that is enjoyable, since that the former is true and the latter false was stipulated.

So people can value doing things they don’t enjoy, and we’re asked to imagine someone spending their whole life doing something they value but don’t enjoy. Goff adds an additional remark in the endnotes (Chapter 1, endnote 11):

Perhaps the subjectivist could say that Susan’s life is meaningful relative to her own valuations but meaningless relative to the valuations of others who don’t value counting blades of grass (I am grateful to Bart Streumer for raising this point). This is a more radical form of subjectivism, as there is no longer a fact of the matter as to whether Susan has a meaningful life.

I think a subjectivist could say this. Goff calls this a “more radical form of subjectivism,” but I don’t see anything radical about this. I’m also wary of the claim that there would no longer be a “fact of the matter” as to whether Susan has a meaningful life. Presumably this means a non-relative fact, but that doesn’t strike me as a terribly troubling concern. So a person’s life is meaningful relative to their standards but not ours. That seems perfectly sensible to me. Goff then continues

Moreover, the counterintuitive implication being pressed in this thought experiment merely needs to be put a little differently. Rather than implying that Susan’s grass counting is just as meaningful as advancing scientific knowledge, this form of subjectivism implies that Susan’s conception of what constitutes meaning is just as reasonable as that of someone who values advancing scientific knowledge.

Hold up. Counterintuitive to who? I don’t find anything counterintuitive about this. This is not only not counterintuitive to me, it strikes me as an intuitive and sensible way to construe value. I also do not agree that this form of subjectivism implies that “Susan’s conception of what constitutes meaning is just as reasonable as that of someone who values advancing scientific knowledge.” A subjectivist might think this, but they don’t have to, and, I suspect, a thoughtful subjectivist wouldn’t. Goff’s remarks pragmatically imply that the subjectivist thinks there is some independent standard for evaluating how reasonable competing desires and preferences are, and judges that all sets of desires are equally reasonable. But reasonable relative to what standard? If you're a subjectivist, you may (and, I think, a reflective subjectivist probably would) think there are no nonrelative standards from which to judge the reasonableness of competing sets of preferences or values. 

Indeed, you may think that the very notion of doing so doesn’t make sense. One’s values set the standard for what’s good or bad relative to those values, but there is no further fact about whether any set of values is any more or less reasonable than any other set of values. The subjectivist may think that Susan’s conception of what constitutes meaning is “just as reasonable” on this view only in a trivial sense: none of these views is “reasonable” or “unreasonable” at all. However, I think to say that competing views are “just as reasonable” pragmatically implies that sets of values are appropriate subjects of evaluation for reasonability, and the subjectivist judges them all to be “reasonable.” I think this would be false, at least for less-objectionable forms of subjectivism that don’t smuggle in realist presumptions. And on this view, it would be a category mistake to evaluate Susan or anyone else’s values as “reasonable” or not. One’s values are descriptive facts about the agent in question. They aren’t appropriate subjects of evaluation for reasonability, any more than a person’s birthday would be. I think Goff brushes up against what I call the halfway fallacy: I think Goff is imposing conceptions of normative evaluation on this characterization of subjectivism that a subjectivist would be free to reject.

Note, too, that Goff describes the implication so-described as “counterintuitive.” I am a fan of Goff’s enthusiasm and candor, but I must be consistent: this strikes me as another case where a philosopher will state that something is “counterintuitive” without being clear about whose intuitions they’re referring to. I don’t grant that any claim is intrinsically intuitive or counterintuitive (and if someone invoking the term does, it’d be helpful to know!); I think that whether something is intuitive or not turns on the intuitions of whatever agent is evaluating the claim, and people don’t all share the same intuitions, nor do I think they would necessarily share the same intuitions with respect to any given issue on reflection. This is another case where I’d encourage philosophers who claim that something is “counterintuitive” or “intuitive,” to clarify what they mean.

2.2 Blades of grass and ideally coherent eccentrics

The activity Susan values is one I personally consider a total waste of time, and would never engage in, nor do I know anyone who engages in such an activity. If a human were to engage in this activity, many people would think there was something “wrong” with that person, much as we might worry about someone who spent all day counting or staring at the ceiling or screaming at tables. In such cases, we may worry that this person is actually suffering or living an unfulfilling life, and in many cases, we’d probably be right. Intervention, in such cases, may help that person from their own point of view, and with the benefit of hindsight. 

This is just to say that we build up, over the course of our lives, a repertoire of knowledge about human psychology, and we have a sense of what is and isn’t “normal” behavior. In some cases, we may judge unconventional behavior to be benign, whereas in others we’d be very concerned. But such concerns involve, in part, making a host of inferences predicated on our assumptions about the psychology of the person we’re evaluating. This might change if we encountered artificial intelligences or aliens with radically different psychologies. If we encountered a cosmic plasma-based alien that spent its time traveling from star to nebula to catalog rare isotopes, would we immediately presume this isn’t a life worth living? What if it informed us that it found this actively absolutely delightful and the most fulfilling action it could imagine? I wouldn’t be so quick to infer that it was making any kind of mistake. 

I worry that when philosophers create scenarios like the “Susan” scenario, they surreptitiously smuggle in and then make a kind of intuitive performance error: they create scenarios that prompt them to draw on knowledge, heuristics, and assumptions about familiar real-world cases, then mistakenly insert these into their consideration of a hypothetical in which such contingent real-world considerations should be bracketed and held in abeyance. To give an extremely simple example, suppose I asked someone to imagine a person who “disliked all sweet and savory food and only enjoyed the taste of rusty spoons,” then asked whether this was “irrational.” There is no reason in principle why there couldn’t be an alien with a different physiology from humans that couldn’t eat conventional human food. Its taste receptors, nervous system, and metabolism could all be geared around motivating it to eat rusty spoons. For such a being, there’d be nothing “irrational” about it enjoying rusty spoons. Yet if a person were to imagine a being that enjoyed rusty spoons, then smuggle in something like “…and that being has standard human physiology,” they may judge our rusty spoon lover to be “irrational.” Yet this judgment only yielded this sense of “irrationality” because the person imagining it (perhaps unknowingly) introduced covert “phantom facts” into the hypothetical. I think philosophers do this sort of thing all the time, don’t realize they do it, and then make a bunch of judgments about various cases that are completely worthless, but nevertheless figure prominently in their philosophical accounts.

This is a subtle form of (at least what I consider to be something like)  imaginative resistance: 

The phenomenon of “imaginative resistance” refers to psychological difficulties otherwise competent imaginers experience when engaging in particular imaginative activities prompted by works of fiction. Usually, we seem to have no trouble engaging with time-travel or space-exploration stories, superhero movies, or talking non-human animal fables. At other times, we do not seem to be able to play along that easily [...]

When we’re given the case of Susan, I think we should really try to push ourselves to imagine that Susan really truly as a matter of fact genuinely values counting blades of grass in just the way I value my family or eating pastries. But I think many times philosophers partially fail at doing so: they, to put it roughly, “picture” a normal human with normal human psychology doing something super weird, and then unconsciously bring in all the assumptions they’d rely on if they encountered someone like Susan in the real world. In the real world, it might be totally reasonable to infer that if you found someone counting blades of grass, that something has “gone wrong,” and that this person may need help. I think Sharon Street does a superb job expanding on the kind of work one could (and, I think, should) do to more robustly imagine people with weird values in her article, “In defense of future tuesday indifference: Ideally coherent eccentrics and the contingency of what matters.”

Street (2009) describes “ideally coherent eccentrics” or “ICEs”: hypothetical beings that commonly figure in discussions of metaethics that (a) have bizarre or repugnant values and (b) hold these values in a logically coherent way. Susan is one such ICE. What Street does in the article is to put a lot of work into fleshing out the background context and state of mind of ICEs, such that, once achieved, our intuitions (if we have them; I don’t) that the person’s actions are some kind of irrationality or mistake diminishes or dissolves. Street draws several lessons from this process. Here is the first:

Lesson 1: When discussing ICEs, one is under a strict obligation to make sure one is imagining these characters in full and accurate depth, otherwise one’s intuitions regarding them are of no philosophical relevance. (p. 281)

Street goes on in Lesson 2 to note that we should avoid imagining that ICEs are basically like normal people with a change or two. This, to me, points to the importance of contextualizing the agent in question. If I imagine a normal human, but change one thing: they only value screaming at tables, this person may need help: I may imagine they feel a compulsion they don’t want and are miserable screaming at tables but feel compelled to do so. But if I was given enough context to understand why a non-human being may wish to scream at tables all day, and I could be assured that this was something it wanted to enjoy in a way for which there was some rationale in light of its values and experiences and psychology, I suspect most of my sense that there’s something “wrong” with its decisions or values would evaporate.

I suspect, more generally, that the sense of “wrongness” or “offness” philosophers often intuit about ICEs like Susan is actual, at least in many cases, a residual artifact of their failure to fully and richly imagine the scenarios with a plausible countextual foundation that would ground and make sense of the ICE’s desires or actions. 

Street concludes in Lesson 3 that we should “avoid underestimating the resources of an attitude-dependent conception of normative reasons to yield the intuitive result about the reasons of real-life human beings” (p. 281). I have stressed a similar point many times: realists often imagine toy versions of antirealism, but never push the limits of what proponents of such views might say in defense, instead projecting partial realist-presumptions onto the antirealist and depicting us as hopelessly tangled up in contradictions or mistakes. In other cases, we’re depicted as helpless in the face of the profound lacunae riddling our muddled and forlorn worldviews. One can’t fault realists too much: unfortunately, antirealists are often concessive in their responses. Street isn’t, though, and this awesome article provides some of the groundwork for showing how proponents of particular antirealist perspectives (such as those defending stance-dependent conceptions) have the conceptual resources to address concerns raised by realists. Realists simply do not give antirealists enough credit for what we can accomplish without appealing to the notions they endorse and that we reject. This summary of Street’s lessons doesn’t do justice to the article, which is insightful and had a significant influence on how I think about these issues.

Goff goes on to consider a few objections. The first is just a misunderstanding of what’s been stipulated, so we’ll ignore that. The second challenges the hypothetical:

Another common objection to my Susan thought experiment is just to deny that anybody could have something so pointless as counting blades of grass as a fundamental life goal. I kind of agree with that, but it effectively concedes the inadequacy of subjectivism. (p. 8)

This is a puzzling objection. I don’t see any good reason to think someone couldn’t have this as a life goal, and I’m puzzled as to why Goff “kind of” agrees with this. Why couldn’t they?

I’m a little unhappy with Goff’s next remark:

To say that an activity is so pointless that nobody would centre their lives around it is to appeal to some external, objective standard of which activities are or are not worth doing. But if subjectivism is true, there is no such external standard. People’s arbitrary whims are the ultimate arbiter of what is worth doing. (p. 8)

I don’t like this use of “arbitrary.” The term “arbitrary” carries connotations of views held on a whim, readily subject to change, of little or no importance, and so on. None of this is necessarily true of one’s values. Unfortunately, critics of subjectivism and other antirealist positions often employ language that carries implications about the nature of the antirealist’s views that I don’t think are appropriate to attribute to us. Care should be taken not to unintentionally employ language that takes a rhetorical swipe at a view you’re criticizing.

2.3 Goff’s road out of subjectivism

Goff ends up concluding that:

It was these counterintuitive implications which started me on the road away from subjectivism. (p. 8)

I simply don’t find the scenario counterintuitive.

Goff goes on to describe Hume’s is/ought gap (pp. 8-9), adding that “It was because of this logical gap between facts and values that Hume decided that morality must come from feelings rather than reason” (p. 9). Goff elaborates on Hume’s view which I won’t detail here, beyond providing Hume’s famous remark that:

Reason is, and ought only to be the slave of the passions, and can never pretend to any office than to serve and obey them. (as quoted in Goff, 2023, p. 9)

As an aside, there’s a delightfully biographical approach to much of the book, which breaks up the tension of the more technical elements and adds a narrative element of discovery to the chapter. While I don’t agree with Goff’s views on metaethics, this is good writing, and it makes an enjoyable, accessible, and more personable reading than I commonly encounter in the journal articles. Here’s an excerpt, immediately following this characterization of Hume, which will also help take us where I want to go next:

This all seemed to make perfect sense to me when I learnt it as an undergraduate. All that changed during my first year as a graduate student. I can still vividly remember the moment, sat in The Queen’s Head pub in Reading, when I realized that there was a deep inconsistency at the core of the Hume-inspired subjectivist view that I had come to be so confident of. Appreciating this was one of the most important moments of my intellectual life.

Goff notes that while there are many disputes about what Hume endorsed, Goff came to adopt two positions:

1. The is/ought gap

2. Reason ought to be the slave of the passions

The first holds that you cannot make inferences about what you should do on the basis of what is descriptively true. The second holds that, all else being equal, you should do what you desire.

Goff notes that, once these two are presented side by side, “the tension was obvious.” And indeed, it is. Can you spot it? If (a) holds that we can’t go from facts about what is to facts about what ought to be, then it would follow that we cannot go from facts about what we desire to facts about what we ought to do.

I agree, and this is one reason I don’t endorse common characterizations of Humean or stance-dependent views that move from desires to ought claims, e.g., views like “if x would achieve y, and you desire y, the fact that you desire y gives you a reason to x.” To me, this reifies reasons. I think the idea that a desire could generate a reason is, at best, an obscure metaphor, and at worst, hopeless nonsense. There are no “reasons” out there existing autonomously. On my view, there are only descriptive facts. The is/ought gap is true in the trivial since logic is non-ampliative. I simply reject (b). I don’t think that it follows from the fact that you desire something that you “ought” to do anything, where “ought” is understood reference normative reasons or some type of irreducible normativity. I don’t think this makes any sense. Rather, I think there are only descriptive facts, and if there are normative facts, they are reducible to a type of descriptive fact (on my view, it would be a means-end relation).

I agree with Goff’s identification of a tension in these views. My solution would be to reject (2).

Goff considers three solutions:

1. Reject the is/ought gap

2. Value fundamentalism:

There are fundamental facts—as basic as the laws of physics or the axioms of mathematics—about what kinds of things are better and worse, and about how people ought to behave. (p. 11)

3. Value nihilism:

There are no facts about better or worse, or about what ought to be done. In other words, value is an illusion. (p. 11)

I don’t know if Goff thinks these are the only options. Goff simply says “There were basically three options open to me at this point” (p. 10). This could mean these were the three Goff considered, but it may mean Goff thinks these are the only possible responses. Goff ended up endorsing (2). However, I don’t endorse any of these. Whatever Goff may think, these are not the only options available. One has more options than nihilism or fundamentalism. Goff concludes this section by noting that “Unfortunately, value nihilism is a hard act to pull off, as we’ll find out in the next section” (p. 11).

So characterized, I think value nihilists succumb to the halfway fallacy. If the value nihilist grants the realist’s conception of value, then denies it, then they end up agreeing that there are no values. This is something like value error theory. But you could just deny the realist’s conception of value. This doesn’t require a crude, inconsistent form of subjectivism that retains the realist’s irreducible normativity. You can jettison that, too. Value nihilists as described here only fail because they don’t go far enough: they purge the realist’s metaphysics, but not the realist’s concepts. They should purge both. 

I suspect Goff’s mistake here is to continue to think of value in realist terms. A reconceptualization of value itself allows one to retain the is/ought gap (as one should) and reject both value fundamentalism and nihilism, both of which share the same misconceptions about the nature of value. Goff’s mistake, in other words, is to retain a faulty conception of value the whole time: whether to accept or reject it seems to  be the only consideration, not to discard it and adopt another conception of value altogether.

2.4 Value nihilism

Value nihilism is a very silly position. Goff says that while we have a sense that certain things are “worth doing,” that if value nihilism were true, this would be “an illusion.” Well, things are worth doing…to me. So I’m not a value nihilist. But I also don’t think these things are worth doing in some kind of primitive fact or a fundamental fact about reality. It’s a fact about my evaluative attitude towards things.

Since I don’t endorse value nihilism or find it worth defending, I don’t have much to say on its behalf. I do want to comment on some of Goff’s remarks, though. Goff says:

This philosophy [value nihilism] cannot be lived out. Almost every human action involves appeal to reasons. Sometimes a person eats because they are literally compelled to do so by their animal hunger. But 99.99 per cent of the time, we eat something because we take ourselves to have reason to eat. (p. 13)

I don’t know what Goff’s conception of a “reason” is, but there is a very good chance I disagree: I think there are facts about what we are motivated or desire to do, and facts about what we do do. I don’t think there are any further facts. If “reasons” are further facts, and refer to something like “normative reasons,” that are somehow “generated” by our desires or by the stance-independent normative facts or whatever, then I don’t think there is anything like that, since I think don’t even think that such a notion is meaningful. Nevertheless, people still eat, whether from animal hunger or otherwise. People can get things done without these kinds of reasons. Normative reasons are completely explanatorily superfluous. They do nothing, explain nothing, and make no sense of anything. They are a conceptual mirage philosophers dreamed up, and nothing more. But I don’t know if that’s the kind of “reason” Goff has in mind.

Further remarks don’t reveal (to me, at least) what Goff has in mind. Goff continues:

The reason might be that we need to eat to stay healthy, or it might just be that it’s going to be enjoyable. (p. 12)

That we need to eat to stay healthy is a reason? Why? What does that mean? That looks like a descriptive fact, (an “is”) not a normative fact (an “ought”). That we must eat to stay healthy … what? Gives us a reason? Is a reason? What does it mean for it to be a reason? The same holds for “enjoyable.” It’s simply not clear to me what a “reason” is supposed to be.

Goff’s next remark has me worried that he is invoking the baleful notion of a “normative reason” prominent among contemporary analytic philosophers:

But if there are no reasons, if nothing counts in favour of action, then the only non-delusional human actions are the ones in which we are compelled by our animal urges. (p. 12)

I don’t think the term “counts in favour,” when invoked by philosophers in nonrelative normative contexts, means anything. It’s a philosophical term of art that, barring an explanation of its meaning, strikes me as probably not meaning anything. The notion of “counting in favour,” simpliciter, without respect to any particular goal or standard, strikes me as just as vacuous as saying that some descriptive fact provides one with “points.” Counting seems like a metaphor for scales and the weighing of currency for and against an action and some contrary action, and only if the sum is greater, is one to perform the action. But if this is a metaphor, what is it a metaphor for? And if it isn’t a metaphor, is it meant literally? If so, then what does it mean? 

I can understand a fact “counting in favor” of an action if that fact is combined with my goals, e.g., if it’s a fact that a shiny rock I found is a diamond, and it’s a fact that if I sell it I’d make a lot of money, this does not, by itself “count in favor” of selling the diamond. Only if I also want to make money can I make sense of it counting in favor with respect to my goals. But it would make no sense to me to say that the fact that it’s a diamond and would sell for a lot would, by itself, “count in favor” of selling it. I think it only makes sense to think of “counting in favor” as something like “is conducive to acting in accordance with a given standard.” That is, a given set of descriptive facts can constitute a set of relational facts between means and ends: if your end is to make money, and selling a diamond would provide a means of doing so, then language for specifying the relation between this means and end would meaningfully pick out a relational fact: that selling the diamond would satisfy the goal of making money. On this view, “reasons,” are best understood as capturing relations between means and ends. This discharges what a “reason” is in a way that doesn’t treat it as some kind of primitive, unanalyzable, mysterious concept that resists analysis. 

Goff continues:

It gets worse. Value claims are pervasive not only in our discourse about action but also in our discourse about belief. We think people ought to apportion their beliefs to the evidence, that they ought not to believe contradictory statements, and so on. If there is no value, there is no reason to believe or disbelieve anything.

Again, what’s meant by “reason”? Antirealists can cash out a notion of “reasons” that makes sense of these situations and that allows the antirealist to navigate these situations without rejecting the is/ought gap or accepting value fundamentalism, so why is Goff only contrasting these with value nihilism? 

It is important not to conflate value nihilism with what I’ll call here value antirealism. Value antirealism denies a range of realist conceptions of value, according to which there are stance-independent facts about what is good or bad, right or wrong, and so on. Such realist views often construe value in terms of irreducibly normative facts, or (or in addition) construe values as reason-giving, or providing us with “normative reasons” that “count in favor” of performing a given set of actions. The latter is even consistent with anti realist views, since one could think that there are normative reasons that count in favor of actions, but that these reasons are determined by our stances. An antirealist can reject all of this, by rejecting these conceptions of value and reason altogether: there are no irreducibly normative facts, there are no normative reasons, and so on.

But proponents of these views don’t own the English language. We are not obligated to accept their account of the use of the term “reason” outside non-stipulative contexts where their conception of “reason” is in operation by fiat (note: Goff has not suggested otherwise; I’m speaking in more general terms). I don’t understand ordinarily discourse about people having “reasons” to do things in terms of a realist conceptions of “reasons,” and so I am under no obligation to accept that people have “no reason” to do anything, even on an antirealist view that denies Goff’s and others’ conception of “reasons” and “values” when speaking in everyday terms. I do think people “have a reason” to believe and do things, I just don’t understand this to entail that they have “normative reasons” of the sort realists (and some antirealists) suppose that they do.

Value antirealism thus does not commit one to denying that there are “reasons” to do things. Rather, such a view can construe reasons in terms consistent with antirealism. In my case, the question of what “reasons” are never arises: while technically there are “no reasons,” in the sense that there are no autonomous “reasons” out there in the ether, this isn’t the only way to understand reason talk in the first place. As an antirealist, I can make sense of speaking in terms of reasons without reifying “reasons” or thinking that reasons are some special “counting in favor” property or relation. We are not required to understand ordinary “reason talk” as reifying reasons or treating them as “counting in favor” of irreducibly normative or whatever else realists might attribute to them. We don’t have to think there literally are “reasons” in some irreducible sense to make sense of reason talk.

We can think, in light of these considerations, of at least three discourses: a philosophical discourse about reasons where they are construed in realist terms (something a value nihilist agrees with, but then denies there are reasons of this kind), antirealist terms, or ordinary, colloquial terms. I don’t like the terms “realism” and “antirealism” because they imply the realist thinks there is something “real” and the antirealist doesn’t. Also, as noted earlier, some antirealists will endorse this conception of “reasons,” anyway, so the distinction I’ve proposed isn’t adequate, anyway. Either way, some antirealists simply deny the realist’s views, but this is consistent both with sharing the realist’s concepts but denying anything corresponds to them (a metaphysical thesis), and with denying both the realist’s concepts and the metaphysical facts that would correspond to them (a linguistic/conceptual and a metaphysical thesis). I adopt the latter approach, which is consistent with endorsing an alternative conception of “reasons” where reasons do exist, and thus are “real.” I’m only an “antirealist” about realist conceptions of reasons, not about my own conceptions. I think reasons are real, I just don’t think they’re what realists (and some antirealists) think they are. So let's call what realists and some antirealists endorse “irreducible reasons,” what I endorse “reducible reasons,” and the ordinary discourse on reasons “ordinary reasons.” Here’s my view on the matter:

1. I deny there are irreducible reasons

2. I affirm that there are reducible reasons

3. I endorse indeterminacy with respect to ordinary talk of reasons

With respect to 3, this is because I don’t think ordinary use of the term “reason” or “reasons” in the relevant contexts has any uniform, fixed meaning that comports with one or another competing philosophical accounts. For a similar proposal about indeterminacy with respect to the analysis of first-order moral claims in ordinary language more generally, see Gill (2009).

When Goff or others talk of value nihilism entailing that we have “no reason” to do anything or believe anything, this (on my view) at best only involves denying at least that there are irreducible reasons, and possibly that there are ordinary reasons, if one holds that ordinary reasons are or are generally used to express irreducible reasons. However, I do not deny that there are reducible reasons. So if we take some claim like:

There are at least some instances in which people have reasons to do things.

I do not affirm or deny this statement. Rather, I would require disambiguation: what is meant by this? If it means (1), I deny it. If (2), I affirm it. If (3), then I think its meaning is indeterminate, so I don’t affirm or deny it. What do I mean by that? I mean that I don’t think ordinary language determinately invokes a reducible or irreducible conception of reasons. If it does, and I am convinced of this, then insofar as its meaning matches (1) or (2) I’d deny or affirm it, respectively. What I won’t do is stop using ordinary language, and say things like “Alex had no good reason to go to the store,” because I don’t currently grant that irreducible conceptions of “reasons” have some kind of hold over ordinary language.

I think much of the force of objections to antirealist views involves identifying one’s philosophical analysis of the meaning of some terms or phrases with the ordinary usage of these phrases, then suggesting that anyone who denies one’s philosophical account denies the ordinary meaning, too. 

Yet ordinary meanings often bake in pragmatic features that aren’t exclusively a feature of the philosophical conception in question but not others, and one thereby implies anyone who denies one’s philosophical theory is rejecting the pragmatic features associated with the ordinary language analogues to one’s claims. I’m not suggesting Goff is doing this, though this might be influencing Goff’s views about these issues. What I am suggesting is that this is a common feature of philosophical argumentation, and it has been inadequately addressed and critiqued, simply because it remains a ubiquitous feature of the way philosophers argue. Either way, I think ambiguity about what’s meant by “reason” plays a significant role here.

3.0 Value nihilism and a Moorean response

Goff’s chapter has finally arrived at the point associated with the original tweet. Goff attributes this concern about reasons to Streumer, noting that:

He argues, quite plausibly, that it is part of the nature of belief that one cannot self-consciously believe something whilst at the same time thinking there is no reason to believe it. (p. 13)

This again turns on what’s meant by “reason.” I reject Streumer’s conception of a “reason,” and his conception of the conditions necessary for a belief, as described in Streumer (2013).

Streumer (2013) states that:

(B1) We cannot fail to believe what we believe to be entailed by one of our own beliefs. (p. 195)

Streumer uses the term “belief” here to mean a “full belief,” and a “full belief” is a belief about which we are “wholly confident.” Streumer goes on to say that “To see that (B1) is true, consider a claim of the following,” then provides this claim:

I believe that p, and I believe that p entails q, but I do not believe that q.

This would be a strange thing for someone to say, but how does this show that a person cannot believe p, and believe that p entails q, but not believe that q? At this very moment, many (perhaps most or nearly all of us) hold views that conflict with one another, or are in tension in some way. Our entire network of belief is not fully salient at all times. We don’t have immediate introspective access to whatever inconsistencies may be present. This is the main point of attempting to reach a state of reflective equilibrium: reflection and discussion may bring inconsistencies to light by rendering them salient at the same time. 

So the first problem with Streumer’s claim is that it purports to show that (B1) is true by presenting us with a sentence. But how does this sentence show that (B1) is true? Simple: it doesn’t. Whether a person can believe both p and that p entails q, but not believe q is a psychological question about the agent. Firstly, there is no logical contradiction in a person holding such beliefs (the beliefs themselves may be logically contradictory or face other logical challenges, though). That is, there is no a priori reason why people could not believe logically impossible things. Some people may even be dialetheists, or reject classical logic, or just be very ignorant, or…for some other reason simply not, as a matter of empirical fact,  believe q, even if they believe p and p→q. The beliefs we hold are not necessarily subordinate to classical logic. Indeed, I think they rarely are in practice.

As always, one philosopher’s modus ponens is another’s modus tollens. If Streumer or others insist on or stipulate that what they mean by a belief itself entails that one believes q if one believes p and p→q, then fine: then I don’t think people have beliefs*, I think they have schmeliefs, where a schmelief is like a belief but doesn’t carry this quality. There is simply no way around the empirical question of what people actually think. If you want to talk about “beliefs,” and you want to talk about empirical facts about what’s actually going on in actual people’s minds, there is simply no way of arriving at a conclusive answer via armchair reasoning alone.

Streumer provides an example:

For example, suppose that someone says: “I believe that Socrates was a man, and I believe that this entails that Socrates was a human being, but I do not believe that Socrates was a human being.” This person may be insincere, or may fail to understand what he is saying, or may be considering whether to give up one of these beliefs. (p. 195)

These are all possibilities. But Streumer has not shown that these are the only possibilities, and that the only possibilities available to a believer preclude believing p, p→q, but not q. Streumer simply asserts:

Alternatively, he may be sincere, may understand what he is saying, and may not be considering whether to give up one of these beliefs. But if so, he is too confused to fully believe what he says he believes. (p. 196)

Streumer appears (at least here) to simply assert that if the other three conditions aren’t met that this person is “too confused.” But Streumer does not (at least in this article) present any further arguments or reasons for concluding that this is the case, though I may have overlooked where he did so.

Next, Streumer turns to reasons:

(B2) We cannot have a belief while believing that there is no reason for this belief. (p. 196)

What does Streumer mean by “reason”? All Streumer says on this matter is:

I shall use the term “reason for belief” to mean a consideration that counts in favor of a belief. (p. 196)

What does it mean for something to be a “consideration that counts in favor of a belief”?

Who knows. Streumer doesn’t say. This is defining one undischarged use of a term in terms of another, equally undischarged turn of phrase, neither of which illuminates what is meant by the notion of a “reason.” This invocation of “counting in favor” is exactly the kind of philosophical term of art that I charge with being meaningless. Streumer doesn’t acknowledge or address this issue here, though Parfit confronts it more directly in a passage I’m fond of quoting:

When Williams argues that there are no such reasons, his main claim is that Externalists cannot explain what it could mean to say that we have some external reason. I admit that, when I say that we have some reason, or that we should or ought to act in a certain way, what I mean cannot be helpfully explained in other terms. I could say that, when some fact gives us a reason to act in some way, this fact counts in favour of this act. But this claim adds little, since ‘counts in favour of’ means, roughly, ‘gives a reason for’. Williams suggests that the phrase ‘has a reason’ does not have any such intelligible, irreducibly normative external sense. When he discusses statements about such external reasons, Williams calls these statements ‘mysterious’ and ‘obscure’, and suggests that they mean nothing. Several other writers make similar claims. (p. 272)

Williams and these several other writers are correct: this notion of “reasons” as “counting in favor” is not intelligible.

Streumer doesn’t invoke the notion of an “external” reason. So maybe Streumer has some other, intelligible notion of a “reason” in mind. Unfortunately, there’s no way to know, since “reason” is only defined in terms of the mysterious “counting in favor” notion. Parfit seems to consider these phrases to be synonymous with one another, and in light of this, suggests that what he means “cannot be helpfully explained in other terms.” This brings us to the notion of a primitive, or unanalyzable concept. Let’s call this notion of reasons primitive reasons, or p-reasons.

I don’t think there are any such concepts, and if there were, I would expect some standards or criteria that would justify the claim that any given use of a term is a legitimate instance of such a concept, rather than some kind of confused and meaningless use of a term. Unfortunately, there’s not enough to go on here for me to judge, but my suspicions are with Streumer’s use of the notion leaning towards unintelligibility. If this is what Streumer has in mind, then I think Streumer is invoking a meaningless term. I don’t think we “have reasons” of this kind.

Since I deny that there are p-reasons of this kind, I deny that I (or anyone else) has any p-reason to believe or do anything. As such, I both believe all sorts of things, and I simultaneously believe that I have no p-reason to believe those things. 

Call this a Moorean objection to Streumer (and, by extension, Goff, insofar as Goff holds a similar view):

1. I believe various things.

2. I believe I have no reason* to believe those things

*where a reason is a p-reason.

Since both (1) and (2) are true, Streumer is wrong. People can believe things and believe they have no reason to believe those things because I in fact do so.

Streumer does say a bit more regarding his use of the term “reason.” He adds:

If this term is used in this way, the belief that there is a reason for a belief is a normative judgment. (p. 196)

I don’t know what Streumer means by a “normative judgment,” and Streumer adds that he’ll be “neutral between different accounts of the nature of normative judgments.” This strikes me as a bit strange. My account of the nature of normative judgments is one according to which there are no p-reasons. Maybe Streumer can’t be that neutral. Once again, Streumer asks us to consider a sentence:

p, but there is no reason to believe that p.

This might sound strange in ordinary language, but if “reason” = p-reason, there’s nothing strange about it. It simply involves both a claim that one believes something and a denial of a confused conception of “reasons.” One can still believe one has some other kind of “reason” to believe p. My own view of reasons reduces them to means-end relations. Call this r-reasons, for relational reasons. One could believe the following:

p, and there is no p-reason to believe that p, but there is r-reason to believe that p.

A careful critic can simply deny Streumer’s conception of a “reason,” while endorsing their own conception of a “reason.” We’re not obligated to use Streumer’s conception of a “reason.”

Once again, Streumer also attempts to show that a person could hold to (B2) and believe that “p, but there is no reason to believe that p,” based entirely on a priori considerations. And once again, I deny that this is an appropriate way to evaluate what it is possible for people to do. Questions about what people believe are empirical questions about psychology that cannot be resolved via a priori considerations. One may stipulate their way to a conclusion, but then one’s conclusions follow trivially from those stipulations. In that case, I’d agree with Streumer but maintain Streumer’s conclusions are trivial: fine, people can’t believe that p and that there’s no reason to believe that p. In that case, then I would simply deny people have “beliefs” of the sort Streumer is talking about at all, and instead hold that they have “schmeliefs,” that do everything a belief does except don’t involve the entailment Streumer discusses here.

I can’t quite tell if Goff endorses Streumer’s views. This is something that I will hopefully have the chance to address with Goff directly. Nevertheless, Goff does close out this section by saying:

Despite all of the above, I tried to live as a value nihilist for many years. It’s impossible as a human being not to deliberate, not to weigh reasons for or against a given course of action. (p. 13)

I deny that there are reasons or values of the sort that are probably picked out by “value nihilism,” and instead endorse an alternative conception of value. Nevertheless, I am capable of deliberating and weighing r-reasons for and against various courses of action, without ever entertaining p-reasons. I am living proof that one is capable of doing what is apparently impossible, unless Goff, Streumer, or others could show that I am mistaken about my own beliefs, commitments, or concepts, or have misunderstood their positions (both definite possibilities, though of course I don’t think either is true).

I think the mistake Goff makes is to settle on a specific conception of value, then suppose that if one denies that conception of value, one denies that there is “value” altogether. The result is that the skeptic becomes someone who thinks deliberating is a “delusional activity” and that value doesn’t exist. But this is not a requirement for someone that rejects realist conceptions of value, or “value fundamentalism.” One can both reject value fundamentalism and reject that value fundamentalism offers the only or only legitimate conception of “value.” One can instead endorse alternative conceptions of value, and hold that things are valuable in these respects. I find it strange Goff does not seem to consider this possibility.

4.0 Returning to the tweet

Let’s turn our attention back to the Outpost Argument:

(P1) Necessarily, if value nihilism is true, then there are no facts about better or worse, or about what ought to be done.

(P2) The action (believing that one has reason to believe that value nihilism is true) presupposes that value nihilism is false.

(C1) If a value nihilist takes the action (believing that one has reason to believe that value nihilism is true), then it is true that value nihilism is true and it is false that value nihilism is true.

(C2) It is impossible to consistently believe that one has reason to believe that value nihilism is true.

This argument only works against a very stupid nihilist. This argument does not show in any way that there is a contradiction or any kind of error in being a value antirealist, that is, someone who denies there are stance-independent value facts. The “value nihilist” is someone who, as a matter of stipulation, holds contradictory views. Well, of course a person who you stipulate has contradictory views has contradictory views. But actual antirealists like myself are not necessarily going to endorse all the assumptions that are packed into the premises of the argument presented here. At least one possible response would be to deny (P2). A “value nihilist” could simply deny that one “has reason to believe that value nihilism is true.” Alternatively, a value nihilist could just accept the conclusions of this argument. After all, neither argument actually shows that value nihilism is incorrect or self-defeating. All it shows is:

1. That if a value nihilist thinks that they have reason to believe value nihilism is true, then they’d face some kind of contradiction.

As I just noted above, they could simply deny the antecedent. I don’t think this poses any problems, since, if “reason” here is understood to mean a p-reason, and (as I believe) p-reasons are nonsense and one doesn’t need them and shouldn’t believe one has them, then they’d be correct to deny the antecedent.

Accepting (C2) is also no problem, so long as “reasons” are p-reasons. I grant that it’s impossible to consistently believe that one has p-reasons to believe that one has no p-reasons to believe anything. I don’t think anything interesting follows from this.

References

Gill, M. B. (2009). Indeterminacy and variability in meta-ethics. Philosophical studies, 145, 215-234.

Goff, P. (2023). Why? the Purpose of the Universe. Oxford, UK: Oxford University Press.

Parfit, D. (2011). On what matters (Vol. 2). Oxford, UK: Oxford University Press.

Street, S. (2009). In defense of future tuesday indifference: Ideally coherent eccentrics and the contingency of what matters. Philosophical Issues, 19, 273-298.

Streumer, B. (2013). Can we believe the error theory?. The Journal of Philosophy, 110(4), 194-212.