1. Introduction: Publishing without belief

Imagine you are reviewing an article for a journal. It's a well-written article that presents an original and compelling argument in support of a philosophical claim. The article merits publication and it even has the potential to be a game changer: many philosophers will likely end up believing the claim upon reading the arguments, and even you are on the verge of being convinced. It also appears possible that the article will have real-life ramifications beyond academic philosophical circles: the author defends the claim that implementing lockdowns to curb a pandemic is morally and politically unjustified, and it will be published in a journal that is known to be followed by influential public intellectuals. Just as you decide to send a positive report to the editors, you notice that there is an odd footnote at the very end of the manuscript: ‘I do not believe the claim I have defended in this article.’

Would you still recommend the article for publication, or reject it on the grounds that the author does not believe the claim she is defending? Or perhaps you don't have sufficient information to decide on the matter. Should you rather inquire into the reason why the author does not believe the claim she defends, and make your decision on the basis of this further information?

In a recent, thought-provoking article, Alexandra Plakias (Reference Plakias2019) argues for the permissibility of publishing without belief (hereafter ‘PWB’), including cases where the author's lack of belief is not mentioned anywhere in the published work. She argues that an author's attitude towards the claims she defends in a publication is irrelevant to the merit of those claims, and therefore not permitting PWB would deprive us of a significant amount of philosophical work that has merit. She also argues that PWB doesn't have any undesirable consequences, while not permitting PWB would have an unfair outcome: some philosophers, such as junior researchers with unsettled views, or philosophers who do not believe any philosophical claims due to certain metaphilosophical convictions (such as those, motivated by the conciliationist stance on disagreement (Christensen Reference Christensen2007; Bogardus Reference Bogardus2009), who think that persistent peer disagreement among philosophers should make us suspend belief in all philosophical claims), would be able to publish less than others if PWB is not permitted. She limits the scope of the argument to academic philosophical publications, leaving out philosophical assertion outside publications, public philosophy, empirical sciences and journalism.

In a reply to Plakias, Will Fleisher (Reference Fleisher2020a) offers a distinction between two types of claims, which he calls ‘advocacy role claims’ and ‘evidential role claims’. The former are claims which ‘aim to promote productive debate and disagreement’ (p. 8), while the latter are those that ‘aim to add to the common stock of evidence’ (p. 9). Fleisher argues that publishing evidential role claims without belief is impermissible, and some philosophical publications do involve such claims.

The current article will introduce the following argument into this discussion:

1. (1) When an author does not believe a philosophical claim she defends, she often has a reason for not believing it.

2. (2) Some of the reasons for not believing a philosophical claim are trivial, but some are substantive: they have something to do with the shortcomings of the philosophical evidence that supports the claim (or they have something to do with the availability of equally compelling evidence that supports an alternative one).

3. (3) PWB involves withholding philosophical evidence if one's substantive reasons for disbelief are not presented within the publication.

4. (4) Withholding one's reasons for disbelief (and therefore withholding relevant evidence), is impermissible.

   Therefore,

5. (5) PWB is impermissible when it involves withholding substantive reasons for disbelief.

In the argument above and in what follows, the term ‘disbelief’ should be understood as referring not only to the doxastic attitude of believing that something is false, but also to cases where judgement is suspended and cases that involve a significant amount of doubt.

In section 1, I will introduce and discuss Plakias’ argument in more detail. In section 2, I will discuss Fleisher's reply to Plakias, and unpack the concept of philosophical evidence. In section 3, I will expand on the distinction between trivial and substantive reasons for disbelief. In the final section I will conclude the discussion with a practical suggestion: we can do more good for researchers and to the field of philosophy if we tolerate PWB only when it is strongly justified and if we try to find ways to eliminate academic factors that pressure philosophers to publish their unsettled views or to withhold their reasons for disbelief.

2. Reasons for disbelief

Plakias (Reference Plakias2019: 638) introduces three cases to serve as working examples in the discussion of PWB. The first involves an author publishing a defence of a claim that he finds obviously false in order to expose the bad editorial practices of the journal that publishes it. In the second case, an author publishes what he takes to be a compelling response to certain objections to a view. Even though he finds the view implausible for independent reasons, he frames his article as a defence of the view (titling it ‘A Defence of Nonnaturalism’). In the third case, an author publishes an argument despite having some suspicion that the argument might turn out to be unsound and the position which she criticizes might turn out to be correct. The authors do not mention anywhere in the published work that they don't believe the claims they defend.

As Plakias hints at (p. 639), the first case seems to be inspired by the recent Grievance Studies Affair (Lindsay et al. Reference Lindsay, Boghossian and Pluckrose2018), and it also brings to mind the similar Sokal Hoax (Sokal Reference Sokal1996) and hoaxes that target predatory journals. I will exclude these kinds of cases from my discussion of the permissibility of PWB, because the very point of these publications is to defame the relevant journals by attempting to expose the fact that they are ready to publish material that shouldn't be published, that is, the authors deliberately publish articles the publication of which they find impermissible. There is, of course, a question as to whether such an act of well-intentioned fraud is permissible, but this is orthogonal to the question of the permissibility of PWB in other cases which will be the focus of this article.

Plakias argues that all three cases are permissible: if we find a claim interesting and well-defended, the fact that the author herself does not believe it ought not affect this evaluation. I agree with Plakias that facts about an author's attitude towards the claims she defends may have no bearing on the merits of a publication, but I will argue that this is the case only to the extent that such attitudes are taken as mere psychological facts, as opposed to attitudes that come with a package of epistemically relevant reasons for having the attitude.

When is an author's disbelief a mere psychological fact? Here is an idealized example:

A philosopher pens an article in defence of a claim, and it gets accepted for publication. The night before the article's publication, an omniscient neuroscientist plants something in her brain that invokes in her a very strong but inexplicable suspicion towards the claim she has defended. She wakes up finding herself not believing the claim. She tries and fails to find new evidence to reclaim her belief, but she is also unable to think of a relevant reason as to why the claim should be false. For extra-philosophical reasons, she decides not to interfere with the publication despite her disbelief. This is a case where we can agree that the author's disbelief does not have a bearing on the merit of the article and she has not done anything wrong by choosing not to retract it.

So what is the problem with PWB? The problem with PWB is not about the merit of the arguments that are eventually published, but about the overall merit of the article that should be evaluated by taking into consideration what is withheld from publication. Most real-life cases of PWB involve an author defending a claim despite having philosophically relevant reasons not to (fully) believe it. In some cases, these reasons might be trivial, but in some cases they are substantive, and it is impermissible to publish claims while withholding the philosophical evidence that constitute substantive reasons to doubt those claims.

Before explicating the trivial-substantive distinction, I will turn to the issue of philosophical evidence.

3. Philosophical evidence

In his reply to Plakias, Fleisher argues that PWB is permissible when it involves advocacy role claims, but impermissible when it involves evidential role claims, as in the following three cases he provides: the case of a prankster fooling the editors and referees of a journal with fabricated empirical evidence presented in support of a philosophical view; the case of an experimental philosopher who publishes an article despite having suspicions about the reliability of the way she has gathered her survey data; and the case of an author who publishes a state-of-the-art article despite doubting the truth of some of what he has written because he has not read many of the articles that he cites (Fleisher Reference Fleisher2020a: 4). Fleisher argues that these cases are impermissible as they run the risk of disseminating falsehoods.

I agree with Fleisher that publishing evidential role claims without belief is impermissible. However, Fleisher's distinction between the two types of claims is problematic. As I will soon argue, the two categories are not mutually exclusive.

The publications in Fleisher's examples are partly reliant on empirical facts. They involve scientific evidence, survey data, and historical facts (about who wrote what and when). However, besides the occasional empirical component, philosophical publications also involve philosophical evidence: arguments, analyses, intuitions and phenomenological descriptions that help us to judge whether a philosophical claim is true or not, or at the very least, whether some degree of increased credence in a philosophical claim is justified.Footnote ¹

What I have just provided is a relatively weak characterization of philosophical evidence, one that can also be accepted by philosophers who hold that there is little, if any, that philosophy can prove. Fleisher's characterization of philosophy has overtones of sceptical metaphilosophical views where there are no objectively attainable philosophical truths, and therefore, except for a small part of the literature that involves empirical claims, philosophical publications cannot disseminate falsehoods. According to such metaphilosophies, philosophy is not an enterprise that progresses towards convergence on truth, and the best that can be done by philosophers is to develop multiple frameworks, which cannot be reconciled because they rest on non-negotiable basic disagreements or because there is some other factor that prevents convergence (see e.g. Rescher Reference Rescher1978; Dietrich Reference Dietrich2011; Leal Reference Leal2019; Lycan Reference Lycan2019). But even among philosophers holding such metaphilosophical views, there seems to be an agreement (perhaps not an explicit one, but one that reveals itself in practice) that defending a view, even for the sake of mere advocacy, is governed by norms that have something to do with evidence. Such philosophers, like others, also put forward arguments that attempt to provide evidence regarding how defensible some positions are, how coherent they are, how much they cohere with other things we seem to know, and what are the merits of preferring one over another. Therefore, cases of PWB that merely aim at ‘promoting productive debate and disagreement’ are not relieved of the problems that I will raise in the next section.

4. Trivial and substantive cases of PWB

Below are some examples of trivial reasons for disbelief. (‘Trivial’ reasons are trivial in relation to the claim defended in a publication, as will be clarified below. The same reasons may well be non-trivial in other contexts.)

Metaphilosophical convictions: These are cases where the author's disbelief towards a claim is rooted in a metaphilosophical stance that requires extremely high standards for belief which are practically impossible to meet, or one that precludes belief in any philosophical claim rather than being the result of a doubt towards the particular claim at hand. These stances are often motivated by epistemological problems such as the problem of persistent peer disagreement in philosophy (Fumerton Reference Fumerton, Feldman and Warfield2010; Kornblith Reference Kornblith and Machuca2013) or the problem of unconceived objections (Mizrahi Reference Mizrahi2014), or by the idea that philosophy does not or should not involve belief but some other attitude (Goldberg Reference Goldberg and Machuca2013; Carter Reference Carter2018; Barnett Reference Barnett2019; Fleisher Reference Fleisher2020b). There can also be cases of PWB that result from positions in philosophy of science, when a philosophical claim relies on a scientific theory and the author does not regard any scientific theory to be true due to her instrumentalist or anti-realist views.

Mere incredulity: These are cases where a philosopher has a strong feeling that there is something wrong with the claim she defends, without being able to figure out any philosophically relevant reason for this incredulity (as in the thought experiment described in section 1) or a reason why such incredulity should be a reliable indicator of the falsity of the claim.

These cases are trivial, as there is no further evidence-gathering the authors can do to support the claims they defend. The incredulous author will try and fail to find the source of her incredulity, and will conclude that the incredulity is either inexplicable or results from a non-intellectual and philosophically irrelevant gut feeling; and the authors’ disbelief in the metaphilosophical and philosophy-of-science cases results from general epistemological stances on philosophy or science, rather than evidential shortcomings of the claims they defend.

Also note that the articles involved in these cases of PWB can be transformed into publications where there is a second-order claim which envelopes the main claim and which the author does regard as true: a conciliationist can remark, e.g. in a footnote or parenthetically, that she does not believe the claim she defends as she does not believe any philosophical claim, but she nevertheless believes that there are good reasons to endorse and advocate this claim and not its alternatives. (See Fleisher Reference Fleisher2020b for an account of endorsement and advocacy in the context of conciliationism.) After all, there should be some reason why she defends that claim rather than its alternatives. Similarly, the scientific anti-realist, who does not believe the claim she defends because she expects a paradigm change in the relevant scientific field, can nevertheless express a belief that the claim logically follows from the best theory within the current paradigm; and the incredulous philosopher can remark that the claim she defends is true on the assumption that her feeling of incredulity stems from a philosophically irrelevant source. We do not expect such remarks to be explicitly stated as they have no particular relevance to the philosophical issue discussed in an article. However, similar remarks can be relevant if an author's reasons for disbelief are substantive.

Substantive cases of PWB result from an omission of philosophical evidence that is directly relevant to the defended claim. Substantiveness is a matter of degree, and the permissibility of each case should be assessed based on how substantive it is and on real-life practical factors regarding what can or should be expected from a philosophical publication created under various pragmatic, research-related constraints.

I cannot comment conclusively on whether the two cases presented by Plakias are trivial or substantive. They are, as presented, too underdescribed to evaluate in this regard. However, I provide below some examples of substantive cases, which satisfy the merit criterion offered by Plakias and the criterion of promoting productive debate and disagreement offered by Fleisher, but which, nevertheless, represent less-than-ideal cases of academic publishing.

Insufficient literature review: A philosopher comes up with an interesting argument in support of the claim that philosophical progress is impossible. Later she suspends her belief, as she comes to think that she has not sufficiently studied some of the relevant literature (which she has seen referred to in secondary sources) and therefore might have omitted some of the available objections to her view. However, she urgently needs to publish more articles in order to boost her CV and to secure a sought-after research position. Moreover, her day job prevents her from allocating sufficient time for research. She thinks that her article's shortcomings will not be noticed, as the relevant literature that she has left unread was produced by an obscure 19th century philosophical circle that most contemporary philosophers are unfamiliar with. She goes ahead and publishes the article.

Insufficient reflection: While working on a chapter for an edited volume on medical ethics, a well-established philosopher comes up with an original and persuasive argument in further support of her pet theory. She has done extensive literature review and has not come across any ideas that could pose a challenge for her argument, but she has not attempted to come up with possible objections. For this reason, she suspends belief in her claims. She has also avoided discussing her new argument with her colleagues, worrying that someone might steal it and publish it before her. She would have preferred to reflect more on the issue and make up her mind, but the deadline for the invited contribution is fast approaching. As a seasoned philosopher, she knows that there should be some interesting objections to her argument: indeed, sketches of some objections inadvertently cross her mind, but she deliberately refrains from developing on them. She is also aware of her highly influential position in her field and the consequent potential that her article will make an impact before any objections are voiced, running the risk of prematurely swinging the field towards a certain view which has consequences for ethical directives regarding the allocation of medical care in times of scarcity of ICU beds. She nevertheless publishes the argument.

Significant incredulity: A philosopher comes up with an argument against the common-sense belief that there is a mind-independent reality. It is a belief that she has always taken to be true, and she has a deep-seated feeling that something has to be wrong with the argument she has devised. She first takes this feeling to be a mere hunch. Later, she considers the possibility that it might be of some philosophical importance: perhaps it indicates that the common-sense belief is grounded in an intuition that is more fundamental than the intuitions that support her argument, but which takes some effort to explicate; or it indicates that there is a mistake in her argument which somehow feeds into her suspicion but which is hard to identify and formulate. Not knowing what to believe, and thinking that going through the effort conclusively will take a long time, she publishes the argument, looking forward to her tenure.

Opportunistic omission: These are similar to insufficient reflection and insufficient literature review cases, with the difference that the author already has access to the evidence that weakens the defended claim. The example that follows is essentially the same as Plakias’ second case, but with more details fleshed out: a philosopher believes that she has good arguments against the idea that interrogation by torture is acceptable in certain cases, but she also thinks that a popular objection to it does not work, and writes an article demonstrating what is wrong with the objection. She considers including some remarks regarding why torture is nevertheless unacceptable and what are some better objections to it, but at the end she decides to leave them out, as she comes to think that including such remarks would inflate the article and make it more prone to objections, which would reduce its chances of publication even though it would make the article richer in content. She titles the article ‘A Defence of Interrogation by Torture’. It proves to be a very influential article and makes a new generation of philosophers develop an affinity to the idea that torture is occasionally permissible, which is made possible partly by the fact that she had omitted the better objections from the article. She finds herself cornering the market on defences of torture and makes a publication career based on defending the idea against all objections except those which she believes that do work but remains silent about.

In all four cases above, PWB results from omitting philosophical evidence that the authors readily possess or that could be acquired from further literature review, discussion with peers, or philosophical reflection. These publications would have more merit if the missing evidence was included, or even just mentioned in order to help other researchers navigate towards that evidence; even simple second-order claims such as ‘the argument presented here is sound assuming that …’ or ‘the claim should be taken with a grain of salt, because …’ would be helpful in this regard. Withholding such evidence from a publication constitutes a vice according to Cassam's characterization of epistemic vices, as it ‘systematically obstruct[s] the gaining, keeping, or sharing of knowledge in the actual world’ (Cassam Reference Cassam2019: 12).

So, the problem is not with the lack of belief per se. As I have argued, there are also trivial, unproblematic cases of PWB. But it doesn't seem possible to come up with cases where an author still has non-trivial doubts about the defended claim despite having satisfied all the relevant criteria for rigorous research and honest presentation, that is, having gone through sufficient literature review, adequately reflected on the issue, examined her incredulity, and having refrained from opportunistically omitting anything that is relevant to the defended claim; and we should note again that the inclusion of a conditional second-order claim, or presenting the article not as a defence of the truth of a claim but as a piece of research that introduces new ideas or arguments into the discussion, transforms the article from a case of PWB to a case of publishing with belief.

If an author has not fully satisfied the above conditions, does this mean that the content is philosophically unworthy? Certainly not. Imagine that a philosopher stumbles upon an interesting but sketchy argument and notes it down, and dies or retires before ever ending up believing the argument's conclusion. Her colleagues may still choose to publish the argument for others to develop on it. But this doesn't mean that the philosopher would be doing something permissible if she chose to publish it herself at that stage without having a high credence towards the arguments conclusion and presented the works as a defence of a view rather than a draft to be worked on by the philosophical community.

Why would any author choose to publish without belief? The reasons for premature or omissive publication in the fictional cases I presented above are extra-philosophical, often involving a mixture of time constraints and careeristic motivations, as it is, presumably, in many real-life cases. Indeed, it is hard to see why an author would omit relevant evidence from publications if not pressured by such extra-philosophical factors. I take it that any author who is committed to academic honesty would ordinarily prefer to steer the reader towards the truth and would defend claims that she has a high credence for. (Imagine how odd it would be if the author in the opportunistic omission case described above kept publishing academic articles in defence of interrogation by torture while being a vocal and passionate activist for the prevention of the practice in her daily life.) In the next section, I will propose that we should try to eliminate the effects of these extra-philosophical factors on academic research and publication, rather than having an overly permissive attitude towards PWB.

5. The permissibility of PWB

If substantive cases of PWB indeed constitute an epistemic vice, to what extent should we permit or tolerate them? Specific aspects of the relevant situations may result in differing evaluations: the first and third cases discussed above are arguably excusable, while the second and fourth seem to be rather irresponsible acts. Under what conditions do extra-philosophical factors justify substantive cases of PWB? A thorough treatment of the intricacies of this issue shall wait for another occasion, but I provide below some brief remarks to provoke further thought.

One could argue that, despite leading to immature publications, substantive cases of PWB should nevertheless be permitted because not permitting them will do more harm than good. As mentioned previously, Plakias argues that penalizing PWB would be unfair to junior researchers with unsettled views. However, the reasons that their views are not settled is rarely trivial, e.g. that the views are counter-intuitive and they just need time for these views to sink in. A more common reason is that they have not yet worked through their views sufficiently: they have to undertake further research, reflection, discussion, analysis and argumentation. It is not that publications by such philosophers have no merit at all, but they would have more merit if they were published at a more mature stage in their career. One might also think that discouraging the publication of unsettled views would delay the introduction of new ideas into the field, but leaving aside extra-philosophical motivations, there seems to be no reason why such ideas should be presented in the form of defences of a view; researchers that want to present unsettled views can incorporate explicit second-order claims into their work rather than withholding their reasons for doubt towards the conclusions of their arguments, and publishing articles in the form of well-developed defences of claims is not the only way philosophers contribute to the pool of ideas: conferences, classrooms, blogs, informal discussions and public philosophy are other venues where philosophers disseminate their ideas.

An alternative to adopting a very permissive attitude to PWB is to tolerate it only minimally, while setting ourselves the long-term goal of overcoming the conditions that lead to PWB. One way to do this would be to have a different set of publication criteria for junior researchers. The cold fact is that most philosophers rush to publish at an early age not for the purpose of contributing to the literature, but to compete better in the job market. We can create more publication outlets dedicated to junior researchers, and the work that is submitted there would be governed by norms that are different than the norms of regular journals: immature and explicitly unsettled views would be welcome, and the authors would not be expected to make contributions to philosophical literature that are on a par with those of senior researchers. The primary purpose of these publications would be to serve as a further seal of approval on the merits of the author that are relevant for a research career in philosophy: analytical and argumentative skills, knowledge of the literature, creativity, and so on. There do exist some journals dedicated to junior researchers, but articles published in these outlets are currently not regarded as significant decorations for a CV.

A more ambitious but more worthwhile long-term goal would be to eliminate publication pressure altogether. Publication pressure puts a big strain on researchers and incentivizes the proliferation of publications that are often substandard and/or redundant, and turns analytic philosophy into a field preoccupied with providing small theoretical variations on trendy puzzles (see Mulligan et al. Reference Mulligan, Simons and Smith2006 for a related discussion). Perhaps philosophers should not be expected to publish at all before they feel that their views have matured. We should come up with new ways of ascribing research-related prestige to philosophers beyond their publications and broaden the means by which philosophers can contribute to research without publishing as a single author and without being preoccupied with defending specific claims. (See Bourget Reference Bourget2010 for a similar proposal.) For instance, if philosophical research were conducted by large research teams that work on long-term projects in a systematized way and with division of labour, a researcher could contribute to making progress in philosophical problems by contributing to various sub-tasks (literature review, classification and reconstruction of arguments, constructing thought experiments, etc.) without publishing articles as a stand-alone philosopher. (There has recently been an increased interest in promoting co-authorship and team work in philosophy. See e.g. Miller and Schliesser Reference Miller and Schliesser2017; Moravec and West Reference Moravec and West2021.)

Until then, it looks like we have to tolerate some of the substantive cases of PWB, but keeping in mind that PWB is the result of a less-than-ideal state of affairs. We should also promote a culture of academic honesty where authors include second-order claims in their publications, functioning in ways similar to the ‘more research is needed in order to …’ clauses in scientific articles, which can lead others to further evidence and contribute to the growth of philosophical knowledge. Plakias considers and dismisses the idea that there should be a requirement for including such caveats in one's work, adding that it is reasonable for an author to expect that doing so ‘would negatively affect [the author's] chance of publication.’ (Reference Plakias2019: 641) Perhaps it would, but it should be resisted, and we should not let it shy us away from voicing our disbelief. ‘The profession benefits from having access to arguments’, Plakias remarks in defence of PWB (p. 644), but what about those arguments that are omitted for the sake of concealing one's disbelief? If a core and venerable requirement of being a philosopher is to raise doubt when it should be raised, why should a philosopher withhold her reasons for doubt towards one's own position?Footnote ²