2. Distinguishing the Problems of Creeping Minimalism

This simple response to the problem may seem tempting,

To explain: first, we distinguish between non-minimal TRUTH (say, truth on a correspondence theory) and truth exhaustively understood as (A), and do this for all the other notions. Second, we suggest moral realism is defined by appeal to non-minimal versions of the relevant notions: moral judgements can be TRUE, express BELIEFS, inter alia. Expressivism, instead, is defined with only the minimalist notions: moral judgements can be true, express beliefs, inter alia.

But most expressivists who accept minimalism cannot accept this.Footnote ⁵ For (A)–(D) are adopted by minimalist expressivists as forms of monism about the relevant notions they concern. (A), for example, is not just a way to understand what truth is. It is, rather, the only way to do so, exhausting the meaning of ‘true’ and the nature of truth, and denying truth any important explanatory role in metaphysics or the theory of meaning. Mutatis mutandis for the other relevant notions.

Pluralism about truth and related notions remains surprisingly underdiscussed in the PCM literature. Quasi-realist expressivists, however, clearly do accept monist forms of the above-discussed minimalism.Footnote ⁶ The PCM is thus taken to involve the expressivist adopting monist minimalism as above, with moral realists accepting monist non-minimalism in turn (Golub, Reference Golub2017, pp. 1389–90).

Moreover, and importantly for our purposes, the monism of many minimalist expressivists involves a denial of the intelligibility of non-minimalist forms of truth, belief, and other key notions. It is not just that, say, correspondence theories of truth are false. It is that such views do not offer an intelligible account of what truth is or what ‘true’ means. Since such minimalists deny the intelligibility of non-minimalist versions of such notions, they cannot accept a way to distinguish their views from their opponents that assumes there is a viable non-minimalist understanding of them. So, Bifurcation fails. (Asay, Reference Asay2013, p. 216; Golub, Reference Golub2017, p. 1390).Footnote ⁷

Dreier presents the PCM as a serious problem. This assessment is widely shared. Some, however, have recently voiced doubts over whether it is a problem and, if so, how serious it is. These doubts concern, first, the problem arising only if we grant minimalism about (inter alia) truth, and, second, suspicion that the problem threatens descent into a ‘game’ of repeatedly raising the possibility of minimalism about any notion employed to halt minimalism's march.Footnote ⁸

These doubts are legitimate concerns about the PCM. As Dreier, Golub, and others grant, the problem as presented depends upon minimalism being plausible and consistent with expressivism. And many realists, for example, will deny either claim.Footnote ⁹ The above doubts, however, reveal more than has yet been appreciated.

For notice that the PCM looked solvable if we could distinguish between minimal and non-minimal notions of (inter alia) truth, and then could characterise realism by appeal to the latter and expressivism to the former. That is, the PCM seemed solvable if Bifurcation worked. It fails, it seems, because Bifurcation requires us to think that there is an intelligible non-minimal notion of truth, belief, inter alia, and this is denied by many expressivists.

But why should Bifurcation requiring us to grant that there are intelligible, non-minimal forms of our relevant notions give us reason to think that Bifurcation fails as a solution to the PCM? This question reveals that there are in fact two problems of creeping minimalism:

1. (1) Assuming (monistic) minimalism, what is the distinction (if any) between realism and quasi-realist expressivism?

2. (2) Given that one may adopt minimalism, what is the distinction (if any) between realism and quasi-realist expressivism, if the latter adopts minimalism and the former does not?

For notice that once (1) and (2) are distinguished, Bifurcation clearly cannot solve (1). (1) assumes minimalism and asks how to preserve the distinction between realism and expressivism in light of it. Bifurcation would require a denial of the form of minimalism that rejects the intelligibility of non-minimal views. But (2) makes no assumption of minimalism. It only asks, given that one may adopt minimalism, how to draw the distinction between realism and expressivism if the latter adopts minimalism and the former does not. This means that, unlike being a non-starter in answer to (1), Bifurcation, absent further argument, may provide an answer to (2).

To explain further. Since those who ask (1) are assuming minimalism, non-minimalists will happily treat (1) as a problem only for the minimalistically inclined. But (2) is a different problem raised simply by the fact that quasi-realist expressivists can and do adopt minimalism, allowing it to creep to any notion we might employ to distinguish their view from realism in their quest to emulate the realist.

Taking theories of truth, (2) is: given that realists accept non-minimalist theories of truth, and given that quasi-realist expressivists accept minimalism, how (if at all) can we draw a distinction between such theories? Mutatis mutandis for any other notion employed to distinguish between these theories. Note, then, that (2) does not require the truth, or even strong plausibility, of minimalism but, rather, only the much weaker (and relatively uncontroversial) claim that minimalism about (inter alia) truth is a view that an expressivist can coherently adopt.Footnote ¹⁰

The literature speaks of the PCM. This is a mistake. The distinction between (1) and (2) has been elided. We can see the distinction between them, and its conflation, by looking at disagreement over what solutions to ‘the’ PCM must look like. Camil Golub suggests,

And Matti Eklund claims,

Golub's statement is representative of what he, Dreier, Tiefensee, and others who take creeping minimalism seriously think a solution must look like.Footnote ¹¹ Eklund instead suggests that the solution to creeping minimalism will involve bifurcating the relevant notions and defining realism in terms of some non-minimal account of them.

But Golub and Eklund are discussing two different problems. Golub is addressing (1), asking how to distinguish between realism and expressivism assuming minimalism. Eklund is addressing (2), asking how to distinguish said views simply given the fact that the expressivist can and does adopt minimalism about truth (and other notions) and the realist does not.

Which of (1) or (2) has been treated as the PCM? This is unclear precisely because the distinction between (1) and (2) is often elided. We can further highlight this elision by showing that common formulations of ‘the’ PCM, as arising when expressivists go minimal, are ambiguous between (1) and (2) as follows.

On the one hand, Dreier, Golub, and others, who treat ‘the’ PCM as requiring a solution that would persist if minimalism were true, are not implausibly read as at least treating the PCM as (1).Footnote ¹² For notice that such theorists present these solutions against a minimalist background, and often a minimalist background that denies the intelligibility of non-minimalist accounts of the notions at issue.

Christine Tiefensee (Reference Tiefensee2019, p. 1907) further suggests that:

She continues:

So, Tiefensee treats ‘the’ PCM as (1): assuming minimalism, what is the distinction between realism and quasi-realist expressivism? She also appears to suggest that by eschewing appeal to non-minimalist notions of truth in his solution, Dreier too takes ‘the’ PCM to be (1).Footnote ¹³

But, on the other hand, Dreier in his seminal paper (Reference Dreier2004, p. 29), and since (Reference Dreier2018, p. 533), has suggested that Bifurcation is a candidate, though unhelpful, solution to the problem. Dreier (Reference Dreier2004, p. 31) has also suggested that ‘the’ problem is one for anyone who wants to understand the distinction between realism and expressivism once the latter go quasi (a problem in ‘meta-meta-ethics’).

But, first, we know Bifurcation could not be a solution to (1). If we read Dreier as taking the PCM as (1), and assumed realists and expressivists are monists about (inter alia) truth, it would be initially surprising how he (or anyone) could think Bifurcation even in principle could solve that problem. This might suggest that those thinking Bifurcation an even unattractive option may have – perhaps implicitly – had (2) in mind when doing so (or that they otherwise conflated (1) and (2)). Second, for ‘the’ PCM to be a problem for everyone it must be faced by someone even if they deny minimalism. (1) is not such a problem. But (2) is. For these two reasons, it may be more charitable to take at least Dreier to be treating ‘the’ PCM as (2).

I cannot settle this exegetical matter here. Rather, these remarks regarding how ‘the’ PCM has been understood are provided solely to highlight further that the distinction between (1) and (2) has been elided. Elided, that is, even in the interpretation of Dreier's seminal paper that birthed the creeping minimalism literature.Footnote ¹⁴

I argue that (2) is a more serious problem than (1).Footnote ¹⁵ For, first, in the face of (1) the non-minimalist has an easy way out: denying minimalism. Second, to even generate (1) we must assume very contentious minimalist theories. Third, given the non-minimalist's easy response, this suggests the PCM as (1) is only faced by those who take minimalism seriously.

But (2) cannot be so easily escaped by the non-minimalist. The non-minimalist faces the question of how to distinguish their view from that of their opponents who adopt minimalism regardless of whether minimalism is true. Second, (2) does not assume minimalism. Third, (2) is a problem for anyone trying to understand the realist/expressivist divide: it asks how anyone, realist or otherwise, could articulate a distinction between realism and expressivism if expressivists adopt minimalism whilst realists do not.

Finally, solving (1) will not solve (2). Even if we can articulate a distinction between realism and quasi-realist expressivism, assuming minimalism, this does not tell us how to distinguish between these theories without having to already assume minimalism. But, at least prima facie, that is what we need to solve (2): a distinction between realism and expressivism that does not assume minimalism.Footnote ¹⁶

(2), then, is a more serious problem than (1). This is important, since ‘the’ PCM seems to be increasingly treated or understood as (1). The most serious problem creeping minimalism raises is not the question of how to distinguish realism and expressivism assuming minimalism. Instead, it is how to do so simply given that the expressivist can and does adopt minimalism whilst the realist does not.

3. Why the Most Serious Problem Cannot be Non-Parochially Solved

(2) cannot be solved in a non-parochial way. By ‘non-parochial’ I mean a solution that all parties to a dispute can accept: in this case, a solution both realists and expressivists can accept. This is why,

First, (P1): a non-parochial solution to (2) requires the bifurcation strategy. The strategy is to distinguish between minimal and non-minimal notions, and then to characterise expressivism by the former and realism the latter. Bifurcation offers the only way to allow, whilst attempting to draw the realist/expressivist distinction, both minimalists and non-minimalists their respective understandings of the notions on which minimalism creeps.

Next, (P2). The bifurcation strategy for (2) will succeed in offering a non-parochial solution, however, only if both realists and expressivists can accept it. Bifurcation succeeds in drawing a distinction between the realist and expressivist that both can accept if it offers a distinction that they can accept without having to assume some feature of their views – for example, the minimalism of expressivism – false.

But, moving on to (P3), Bifurcation requires us to maintain that there are intelligible, non-minimalist notions of, for example, truth. This, as we've already seen, is something that the minimalist expressivist does not accept. Bringing us to (C): thus, our second problem of creeping minimalism cannot be solved in a way that all parties can accept. When faced with the more serious problem of creeping minimalism, we cannot make sense of the distinction between realism and expressivism in a way that both parties can accept.Footnote ¹⁷

It is important to see that this argument is not for the claim that Bifurcation fails as a solution to the PCM understood as (2). Rather, it is that Bifurcation turns out to be a parochial solution to (2) and, since, I maintain, it is the only way we might have hoped to provide a non-parochial solution to (2), (2) cannot be non-parochially solved.Footnote ¹⁸

An objection to (P1). Many solutions in the literature seem non-parochial because they eschew dependence upon any notion of truth, belief, inter alia. Dreier's influential solution, for example, is explanationist: realism should be understood as the view that moral facts and properties explain what it is to talk and think about them, whilst expressivism should be understood to deny this. This suggests (P1) is false: we can, at least in principle, non-parochially solve (2) without appeal to Bifurcation.

Dreier's and similar solutions do avoid notions of truth, belief, inter alia commonly used to distinguish realism and expressivism. And they are found attractive partly for this reason.Footnote ¹⁹ But, first, there is substantial debate over whether explanationist solutions to the PCM are compatible with minimalism. For example, as Taylor (Reference Taylor2020) and Golub (Reference Golub2021) have argued, representationalist explanations of meaning seem incompatible with a minimalist framework for a domain, since such frameworks deny any explanatory role for representational relations like truth and reference. If so, then explanationism will be a parochial solution to (2) and, as such, no counterexample to (P1).

But, second, suppose that explanationist solutions are compatible with minimalism, and that they offer attractive solutions to the problem of how to distinguish between these views assuming minimalism. Explanationism will still, for other reasons, not offer a plausible non-parochial solution to the further question of how to distinguish realism and expressivism simply if expressivists go minimalist. That is, even if explanationist solutions were attractive solutions to (1), they do not offer non-parochial solutions to (2).

For, first, recall that these solutions are always presented upon a minimalist background. Such background is perhaps dispensable. But importantly, once we remove the background of minimalism on which they are presented, we are confronted with the question of why we should appeal to the ways of distinguishing between realism and expressivism these solutions offer.

Dreier's solution, for example, makes the dispute between the realist and the expressivist over what explains the content of moral thought and talk. Many moral realists will find this suggestion strange. For whilst they disagree with the expressivist about that, many will take themselves to also be defending metaphysical theses about the existence and nature of moral facts that the expressivist denies (and not only disputing what explains moral semantic content).Footnote ²⁰ The recourse to an explanation like Dreier's seems attractive when faced with minimalism. But the attraction seems to depend in part upon assuming the minimalist background upon which it is presented. If we do not presuppose that background – and we do not when facing (2) – why take that route when it seems there is more at hand between the realist and the expressivist than what the solution identifies?

And crucially, there are realists who will not take that route. For some realists will take their commitments to non-minimalist views of, e.g., truth and properties to not just be accidental features of their theory but, rather, part of what characterises it as a realist theory as opposed to any other.Footnote ²¹ For example, realists who accept a non-minimalist correspondence theory might think taking moral truths to be those that correspond to moral reality is part of what it is to be a realist as opposed to, say, an anti-realist. Dreier's and similar solutions take this move to be unavailable to the realist.

Solutions such as Dreier's that avoid appeal to our troublesome notions, then, are in a dialectically strange position as solutions to (2). As solutions to (1) they are highly attractive, since they, we are granting, leave the background minimalism the problem assumes untouched. But any solution to (2) that ignores what realists may understandably point to, absent a minimalist background, in distinguishing realism from expressivism becomes parochial.

The point is not that such suggestions fail as solutions to (2). Rather, it is that once we remove the background minimalism that motivates turning to them to save meta-ethical distinctions, such solutions become ones that at least one party – realists of the kind outlined – will not accept (as far as those realists are concerned, such solutions are implausible, unnecessary, or incomplete). So, prominent solutions like Dreier's that eschew appeal to the notions upon which minimalism creeps do not provide non-parochial solutions to (2). Only Bifurcation allows the realist these resources whilst also granting the expressivist their minimalism. But since such minimalists deny the intelligibility of the relevant forms of non-minimalism, Bifurcation does not offer a non-parochial solution to (2).

A rejoinder: perhaps some realists will not accept a solution like Dreier's if they take their commitments regarding, say, truth to be central to what characterises them as a realist as opposed to a quasi-realist expressivist. But it suffices to show (P1) false that a realist could accept a solution other than Bifurcation which the expressivist can also accept. Dreier's solution is, at least in principle, acceptable by both realists and expressivists.Footnote ²² So, (P1) is false.

But this rejoinder misses the point. The kind of realist we are discussing is precisely someone who, given how they, as a non-minimalist realist, understandably conceive of the realist/expressivist divide, cannot accept Dreier's solution as a solution to (2). For, they think, it draws the line between such theories in the wrong place. For Dreier's solution to be non-parochial, such a realist must be able to accept it. So, Dreier's – and analogous solutions that, against a background minimalism, eschew appeal to the notions upon which minimalism creeps – are parochial solutions to (2), not being acceptable by all parties. The only solution that had any hope of being that, albeit dashed, was Bifurcation.

Must a solution to (2) be non-parochial? This is unclear. Perhaps (2) could be solved but only parochially. For example, if pluralism about the relevant notions was adopted, then perhaps a version of the bifurcation strategy could succeed.Footnote ²³ And if we settled whether minimalism or non-minimalism were correct, then we could solve (2) by treating it as based upon falsely assuming one view was a live option.

Denying minimalism or non-minimalism, suggesting they are inconsistent with realism or expressivism, adopting pluralism about our troublesome notions, or even employing Dreier's explanationism may offer a solution to (2). But these solutions, for the reasons I have given, fail to deliver a distinction between realism and expressivism both sides can accept. Insofar as we hope for a non-parochial solution, these suggestions leave that hope frustrated.

I end on a general objection to my paper that draws out its significance. For suppose one grants, as I've argued, that there are two problems of creeping minimalism and that the second is more serious. Suppose one grants further that Bifurcation cannot offer a solution to (1), and that, even if in principle able to offer a solution to (2), Bifurcation can do so only parochially. We might now wonder how important it is to note that there is this distinction between (1) and (2) in the first place. For the reason I have said that Bifurcation fails as a solution to (1), namely that it is inconsistent with the minimalism expressivists endorse, is the same reason that it cannot provide a non-parochial solution to (2). So, other than drawing the distinction between (1) and (2), and suggesting that the latter, more serious problem cannot be non-parochially answered, what progress has been made here?

To answer, let us take a step back. I have distinguished two ways we might draw a distinction between some theories. First, we might try to do so on certain specific background assumptions that provide a framework within which a debate occurs (such as minimalism or non-minimalism). Alternatively, we might try to do so without such assumptions, and upon merely recognising that the theories we are trying to distinguish could be held with differing background assumptions. This objection asks more generally why we should think that focussing upon this difference means we have made any progress in reflecting on the debate(s) these theories are engaged in.

My answer is this: the distinction raises important questions over how to understand and frame the debate over creeping minimalism – and debates over how to distinguish between theories generally – that require further attention. Alongside distinguishing between (1) and (2), arguing that (2) is more serious, and that (2) admits of no non-parochial solution, the progress that we have made here consists in uncovering and emphasising such questions. They are these.

First, highlighting the distinction between (1) and (2) forces us to be more sensitive to a subtle distinction which is often overlooked in philosophy. This is the above issue of how we can distinguish between theories ‘from within’ the framework some theory assumes or from without. For example, we could try to distinguish between realism and expressivism on the assumption of minimalism, or we could try to do so without that assumption. This distinction reveals that any question of how to distinguish between two theories is, in fact, ambiguous. Ambiguous, that is, between asking for a distinction on certain background assumptions that may or not may not be shared by the theories in question, or for a distinction that does not rely on such assumptions. We must ask, then, which we are intending to investigate.

Second, the distinction forces us to ask whether we first need to settle debates over general frameworks before we attempt to resolve particular debates or not. For example, in meta-ethics, we often proceed by simply adopting some highly contentious background assumptions – say, minimalism about truth – and then trying to preserve distinctions between views whilst developing our theories. But this methodology may be suspect.Footnote ²⁴ And one reason why is that it may present the illusion that standard distinctions between theories can be preserved when, in fact, such distinctions may be based upon specific background theories – which may themselves be false.

Third, we are forced to ask, in light of differing background assumptions and the distinction between (1) and (2), whether we need distinctions between theories that could be accepted by all parties – that is, whether or not we should be trying to draw non-parochial distinctions between theories in the first place. I have argued that there is no non-parochial solution to (2). I have not argued, though, that the solution to (2) – or distinctions between theories more generally – must be non-parochial. Whether it should be is an important question for future work.

Note, however, a consequence of my argument: if the solution to (2) must be non-parochial, then there is no solution to (2). If so, then there is no way to recover the realist/expressivist distinction once minimalism creeps. And if versions of (2) arise in other disputes in philosophy, this would entail the distinction between realism and its sufficiently sophisticated opponents vanishes there too. That would be, at the very least, a troubling consequence.