Being Open-Minded about Open-Mindedness

Abstract

Within the field of virtue and vice epistemology open-mindedness is usually considered an archetypal virtue. Nevertheless, there is ongoing disagreement over how exactly it should be defined. In this paper I propose a novel definition of open-mindedness as a process of impartial belief revision and use it to argue that we should shift our normative assessments away from the trait itself to the context in which it is exercised. My definition works by three sequential stages: not screening new claims, impartially weighing the evidential strength of claims, and updating beliefs accordingly. Using this definition I argue for a focus on agents’ particular circumstances to determine what degrees of credulity, open-, or closed-mindedness are appropriate in any given situation. As well as providing conceptual clarity regarding the concept of open-mindedness this paper indicates the benefits of this contextual approach for our everyday epistemic attitudes. In particular it enables us to recognise, without stigma, when ourselves or others deviate from open-mindedness for appropriate reasons.

1. Introduction

Everything is what it is: liberty is liberty, not equality or fairness or justice or culture, or human happiness or a quiet conscience. (Berlin, 1958, p. 172)

At first glance, open-mindedness occupies a role as an unassumingly obvious good in modern liberal democratic societies. It supports the liberal idea of progress – the human condition constantly improving through open-minded discourse and expanding the scope of human knowledge (Berlin, 1958, p. 173). This logic is often twinned with the idea of advancing the frontiers of scientific discovery, underpinned by a willingness to revisit and revise claims (Popper, 1945). Virtue epistemologists who study character traits and their relationship with human knowledge frequently place open-mindedness atop lists of intellectual virtues (Riggs, 2010, p. 173; Baehr, 2011, p. 140). According to an influential account of virtue, virtuous actions are correct actions (Hursthouse, 1999, p. 28) and therefore if open-mindedness is a virtue, it is necessarily correct to act in an open-minded manner.

In this paper I want to argue for a more complex picture – in particular, for the idea that open-mindedness is not always good or appropriate, depending on the situation. To do so, I propose a normatively thin process account of open-mindedness which I term ‘open-mindedness as impartial belief revision’. By firmly separating out the concept of open-mindedness from the reasons one might hold for adopting it, I create the theoretical space and critical distance to evaluate when open-mindedness may be appropriate, or when alternative cognitive approaches may instead be justified. My argument forms part of a larger movement within studies of open-mindedness and virtue ethics and epistemology more generally, emphasising the importance of contextual considerations when assessing the normative value of character traits. The literature on open-mindedness itself purports to provide various examples of appropriate closed-mindedness, including wartime propaganda (Hare, 1985, p. 4), the arguments of racists (Fantl, 2018, p. 147), or when oppressed minorities face epistemically polluted environments (Battaly, 2018b, pp. 39–44). In the field of virtue ethics more broadly authors such as Curzer have, due to the importance of contextual considerations, pushed back on the thesis that virtuous actions are necessarily right actions (Curzer, 2017, pp. 62–67; Curzer, 2023, pp. 50–64). Similarly, Kidd has recently argued for both the importance of specific contexts and of considering a wide range of values, not just epistemic ones, when analysing epistemic virtues and vices (Kidd, 2021, pp. 80–83). Kidd concludes these reflections by noting that, ‘The trick, of course, will be to develop accounts of epistemic character traits and dispositions that are properly neutral, in the sense of not prejudging their normative status’ (Kidd, 2021, p. 82). The bulk of this paper will be taken up with Kidd's proposed task by developing an appropriately neutral account of open-mindedness. In the penultimate section I will illustrate some examples of its application in weighing both epistemic and non-epistemic considerations when assessing the normative values of traits. Because my account explicitly considers both epistemic and non-epistemic factors and goals, it does not directly speak to whether open-mindedness is an epistemic virtue per se, but instead whether it is a virtue all things considered.

To develop my normatively thin process account of open-mindedness I draw on the concept of impartiality. Discussion over the nature and implications of impartiality ranges far and wide from foundational questions of political and moral philosophy (Barry, 1995; Sen, 2002; Shapiro, 2016) to the principles of journalism (Kieran, 1998; Wahl-Jorgensen et al., 2017; Ojala, 2021) and even the role of punditry (Damazer, 2023). Rather than using impartiality to ground fundamental moral principles or guide institutions, here the concept is used to elucidate a feature of an individual's mental process and attitude (O'Brien, 2017, p. 142). In using impartiality to help define open-mindedness I am in part inspired by Hare's account of open-mindedness as, ‘[…] a willingness to form and revise one's views as impartially and objectively as possible in the light of available evidence and argument’ (Hare, 1985, p. 3).¹ Although he disagrees with defining open-mindedness through impartiality, Baehr provides a useful description of the conception as an honest and impartial judge preparing to hear opening arguments (Baehr, 2011, pp. 143–44). Within legal scholarship impartiality is often said to either equate to, or incorporate, open-mindedness (Lucy, 2005, p. 15; Kramer, 2007, p. 57) and this is reflected in qualitative studies of how judicial officers understand impartiality (Anleu and Mack, 2019, pp. 255–57). When I use impartiality to conceptualise open-mindedness, I mean it in a way that is similar to an idealised judge approaching a new case – giving new claims a chance to be incorporated into the agent's belief structure after being subjected to their judgement without fear or favour. In particular, impartiality characterises a person willing to listen to all sides of an issue and not prejudge claims, but instead to follow and think through the relevant claims wherever they may lead. While impartiality and open-mindedness are not identical, the linkages in both the aforementioned legal literature and scholarship of open-mindedness help indicate a degree of overlap. Both impartiality and open-mindedness are best understood as the absence of partiality, a parallel drawn explicitly by Dewey (1916, pp. 164–79).

Finally, my account is a process one, as I describe open-mindedness as characterising the process of how an agent engages with claims rather than as a straightforward – potentially binary – character ‘trait’ per se. By adopting a process account of open-mindedness my aim is to approximate how people engage with claims – which requires a sense of chronology and the passage of time. This account can be deployed to scrutinise individual instances where we might want to query whether an agent is being more or less open-minded, rather than focusing on open-mindedness as a settled or persisting state of character (Zagzebski, 1996, p. 137; Adams, 2006, p. 6; Baehr, 2011, p. 21).² My intention is not to rule out the possibility of an enduring character trait of open-mindedness, or to deny that people can in general terms be more or less open-minded. Nevertheless, it stands to reason that in order to understand open-mindedness, even as a settled state of character arrived at through repeated actions, one should have a conception of what an instance of it might look like (Baehr, 2011, p. 21). My proposed process constitutes three consecutive steps – how an agent selects which claims they engage with, assessing said claims, and then updating their beliefs accordingly. To be open-minded is:

1. 1) not actively screening claims due to their conclusion or source,

2. 2) considering the merits of claims impartially, and

3. 3) updating one's beliefs in accordance with one's assessment at (2).

It is possible for an agent to deviate from open-mindedness at any of the three stages. Throughout this paper I will contrast open-mindedness with partial attitudes – credulity as being partial towards claims and closed-mindedness as being partial against claims. While there is a flourishing literature within vice epistemology investigating the nuances of such traits (Battaly, 2017; Cassam, 2019; Kidd et al., 2021), my usage of the terms here is simply intended to indicate deviations from open-mindedness in terms of partiality. In accordance with my overarching normatively thin approach I do not intend to convey any pejorative meaning through using these terms. In fact, it is my contention that we can be appropriately credulous, open-minded, or closed-minded according to the particular situation.

The argument in this paper proceeds as follows. In ‘First Step: Not Screening’ I discuss not screening new claims due to their content or source. I differentiate open-mindedness from proactive attributes such as intellectual curiosity to explain why the first step is a reactive one. In ‘Second Step: Impartial Assessing’ I explain how impartial assessing is narrower than straightforward rational thought, but suitably broad to encompass what we should understand by open-mindedness. In ‘Third Step: Updating’ I discuss considering the outputs of one's assessment of the new claim and reconciling them to update one's beliefs. In ‘Exercising Open-Mindedness – A Matter of Judgement’ I briefly survey a number of reasons presented in the literature for deviating from open-mindedness and emphasise the importance of considering whether or not to be open-minded on a case-by-case basis. Finally, in ‘Conclusion’ I summarise my arguments and conclude with an area for future research, developing guidelines for when to be – or not to be – open-minded.

2. First Step: Not Screening

The first stage of my definition relates to how people identify the claims they are going to cognitively engage with. This is necessary to consider because this selection process shapes the assessment and belief revision processes which occur ‘downstream’ of engaging with specific claims. An open-minded person is someone who does not preselect claims before analysing them because they either disagree or agree with the claim's conclusions or source. This is part of what it means to treat a matter as not yet decided. The point is to subject the substance of claims to one's own judgement to determine their correctness, instead of relying on pre-existing beliefs regarding their conclusions or source to filter them in or out. In the words of Dewey it means a ‘[…] disposition to welcome points of view hitherto alien; an active desire to entertain considerations which modify existing purposes […]’ (Dewey, 1916, p. 175). There are two important elements to this stage. The first is not selecting or dismissing claims before analysing them, the second is such action taking place because of views about their conclusions or source.

Addressing the first element, this definition is framed as refraining from a type of positive action. This is because open-mindedness is primarily concerned with receptiveness to claims. While an open-minded person might actively seek out new claims, they may equally be open-minded but without any particular motivation for learning new things so long as they approach what they do consider in an open-minded way. This distinguishes open-mindedness from characteristics such as intellectual curiosity or a love of knowledge which are necessarily motivational for their holder.³ That said, despite open-mindedness not being identical with a proactive search strategy, it is possible for people to shape what claims they are exposed to. This is captured by the meaning of screening – filtering prior to full engagement. Therefore, on my account, a conscious or unconscious strategy to avoid or favour particular claims or types of claims would fail to constitute open-mindedness; but a person who did not engage in a proactive search could still be open-minded.

To some readers this may seem to have counterintuitive results – the person who makes a good faith effort to learn a domain, and in so doing screens out various sources of claims, is somehow more closed-minded than a person who does not seek to inform themselves at all. However, I believe this is the correct result. One should be considered closed or open-minded regarding an enquiry or domain of knowledge only once one is engaged with it – even if only at the limited level of engaging sufficiently to screen it. There are practically infinite realms of knowledge to learn about, and so a person who has not actively closed themselves off to something, but who has also never inquired into any of the particular areas within this multitude, is not by this fact alone closed-minded.⁴ As mentioned above, this distinction allows us to distinguish open-mindedness from curiosity and other motivational traits.

Open-mindedness is also separate from whether a person has either reasons of prudence or duties to engage in a proactive search. For example, imagine the following scenario:

Hiking Holiday

I have agreed to research and book a group hiking holiday with friends. In doing so I have incurred a duty of due diligence to carry out this research well. I then fail to perform this duty in one of two alternate ways. The first is that I leave the task so late that due to time pressures I book the first route that I come across, without making further enquiries, and therefore without any supported belief that it is appropriate or well suited to our holiday. The second is that, before starting my research properly, I feel wedded to a particular route because it is very pretty and I have already pitched it as ideal to my friends. As a result of this I consciously or unconsciously avoid reading negative or critical reviews when doing my research.

Both alternatives are failures of my independent obligation to carry out proper enquiries, risking bed bugs or worse. However, only the latter is a matter of closed-mindedness; closed-mindedness is not synonymous with negligence.⁵ Neither is closed-mindedness synonymous with carelessness or other instances of epistemically poor decision-making. Imagine that the same screening out of negative reviews is achieved purely by accident, for example when my website settings are accidentally set to filter out the harshest reviews. Again, I have fallen below the proper standard of enquiries and the resulting decision is not properly informed. But although I may be faulted for my error, this is not closed-mindedness because the screening was not purposively carried out at the direction of my own mental processes – whether conscious or unconscious.⁶ This distinction allows us to distinguish closed-mindedness from mistakes which might produce similar epistemic results.

The second element of my definition is that such screening occurs because of the source or conclusions of the new claims. Being open-minded here entails being open to engaging with all kinds of claims. Let us first consider screening on grounds of conclusions. This is perhaps the paradigmatic case of closed-mindedness – people rejecting claims because they disagree with where they lead. I set out here an illustration of an agent's closed-mindedness on my first criteria:⁷

Democratic Senator

A Senator from the US Senate representing the Democratic Party receives a public report entitled: ‘Former President Trump better for the country than many believe’. The Senator holds unfavourable views about President Trump and so does not read the article on the basis of these views.

One might think that the Senator has good reasons for her unfavourable views and that she is therefore in a good epistemic position not to read this article. She does not believe that she will learn anything from it, or perhaps she has better uses for her time as a legislator. One might even go further and argue that President Trump is morally problematic and therefore that one is under an obligation not to engage with reports supporting him. I rattle through these potential justifications for the Senator's decision not to comment on their potential validity. Instead, I want to draw attention to the fact that an all things considered judgement on whether to be open-minded in a particular instance will depend on a variety of agent-level factors. These will include the agent's goals, costs such as limited time or energy, their own expertise in the relevant domain and relative to the complexity of the information, and the downside risks of being misled. Not engaging with claims because one disagrees with the conclusions they endorse is quintessential closed-mindedness, but it may be justifiable. One upshot of this analysis is that as we become more convinced of something – perhaps it is true! – we may also become more closed-minded to competing claims.

The above considers disagreement with the conclusion of the claim, but my definition also considers disagreement with the source of the claim – believing it to be an incorrect or unreliable source.⁸ It might strike readers as highly unorthodox or incorrect that not engaging with claims believed to have poor source validity constitutes closed-mindedness. They might have in mind a person dutifully seeking to learn about a topic screening out sources from non-experts. Yet, they are in fact treating part of the matter as decided if, before engaging with the substance of claims, they – informed by their pre-existing beliefs – determine which sources are worthy of consideration and others not. To illustrate this, imagine a highly stylised example of two contrasting belief systems:

Theist and Atheist

Consider a believer in Abrahamic theism and a believer in naturalistic atheism. Each position includes beliefs regarding what constitutes reliable sources of knowledge. In one view God and their divine revelations are sources of truth. The other view considers metaphysics a non-starter and empiricist enquiry the basis of truthful enquiry. Each worldview comes with a community-supported set of beliefs about what are valid or reliable sources of knowledge.

In order for either believer to countenance being converted by the other, they must be able to set aside these meta-beliefs about source validity in order to even begin to process the contrary claims on their own terms. To be open-minded to an alternative way of thinking requires setting aside one's pre-existing paradigm-informed views of source validity. This may strike some readers as contrary to what a functioning social epistemology requires – the ability to discriminate between sources and identify experts (Goldman, 2001). If someone screens for expert opinions or avoids sources lacking in credentials, it seems manifest to some that this should not be seen as deviating from open-mindedness. The point I want to make here echoes the discussion of conclusion-discrimination above, namely, that there is a distinction between demonstrating the trait of open-mindedness and what we might think is better or worse epistemological practice or an all things justified epistemological attitude in a given instance.

The same point can be illustrated by the example of iconoclasts and their rejection of community standards for matters such as source validity. Iconoclasts are individuals who reject accepted beliefs or reasons – whether shared by experts or their communities (Berns, 2008). Instead, they apply their own standards of analysis to a given domain, irrespective of what constitutes a good reason or source according to the wider community. Now it may be the case that many iconoclasts go awry in their enquiries.⁹ Iconoclasts engage in unusual risky epistemic behaviour because their reasons are doubted by their community or peers. The problem here is that disagreement at a deep level can preclude clear identification of what constitutes good reasons or sources by intersubjective standards. As indicated by the Theist and Atheist example above, one feature of open-mindedness is that it – potentially – enables one to cross between such radically different worldviews. This is even so when what each might consider a good reason or source is radically divergent due to strong axiomatic and metaphysical disagreement. Open-mindedness enables enquiry to reach outside of one's pre-existing or community-endorsed paradigms. Therefore it is a mistake to make its operationalisation dependent on the paradigms it holds the potential to transcend. This is true even if social learning and accepting community-driven insights are required for human progress and development (Sterelny, 2012).

Despite the foregoing, a critic may still find the lack of inbuilt substantive limitations on my account of not screening to be problematic.¹⁰ There are certainly practical limits to how many potential sources a person can consider, and they will inevitably want to use various measures to filter these sources. Other accounts of open-mindedness adopt general language about having considered ‘relevant’ evidence or options (Hare, 2003, p. 76; Battaly, 2018a, pp. 267–72; Fantl, 2018, p. 5) or a willingness to revise beliefs in a ‘reasonable’ manner (Arpaly, 2011, p. 75). These qualifiers make open-mindedness contingent on substantive determinations regarding the subject matter which the agent themselves are investigating – raising the prospect of a problematic regress. They suggest that I can only approach something ‘open-mindedly’ once it is determined what are relevant options or what is reasonable to consider. However, these determinations are themselves the product of potentially open-minded considerations into the subject and should not be necessary for its definition. If someone considers a purportedly irrelevant or unreliable source and discovers its epistemic value, as iconoclasts aim to do, they have demonstrated more – not less – open-mindedness. Answering the substantive questions of when and how to narrow one's focus – for example, which sources and how many to consider – are not, on my account, definitional to open-mindedness. They do, of course, bear on how open-minded one is likely to be in practice. As stated towards the end of the Democratic Senator discussion above, as one understands a topic better one may have better grounds for closing one's mind on the topic or being partial towards preferred sources. This is where my account rubs up against those who seek to import substantive criteria into defining open-mindedness. On these competing accounts, closing one's mind off to certain sources can be classed as open-mindedness on grounds of relevance, reasonableness, or other substantive criteria. On my account, it is simply closing one's mind, albeit on perfectly good and justifiable grounds such as when one thinks they are of limited value or not having unlimited time to consider potentially infinite sources of information and so needing to bring one's search to a close.

3. Second Step: Impartial Assessing

Once an agent has decided to engage with a claim, there follows the question of what open-mindedness entails they do with it. In order to revise beliefs, one needs to comprehend the implications for pre-existing beliefs, if any, of new claims. This comprehension requires assessing them and hence this forms the second stage of my definition. When considering a new claim, being open-minded is equivalent to giving it an impartial hearing – treating it as a candidate for the truth of the matter until it is assessed to be otherwise. One of the oft-cited issues with Hare's definition of open-mindedness mentioned in the Introduction is that it is too broad and akin to a definition of rationality itself (Riggs, 2010, p. 179; Baehr, 2011, p. 152; Kwong, 2016a, p. 407; Fantl, 2018, p. 6). Given my own definition is inspired by Hare's, it is important to clarify its boundaries which delimit it from straightforward rational assessment.¹¹ When I say open-mindedness requires impartial assessment, I mean that open-minded individuals assess claims with the same cognitive tools and approach that they would use to assess other claims of that type – without fear or favour of the claim being true. To illustrate approaching a claim with fear, consider the following example of hearing negative stories about a good friend:

[…] we tend to devote more energy to defeating or minimizing the impact of unfavorable data than we otherwise would. To start with, we are more liable to scrutinize and to question the evidence being presented than we otherwise would be […] we are more likely to ask ourselves various questions about the person telling the story, the answers to which could discredit the evidence being presented. […] We will spend more mental energy generating and assessing such possible discrediting factors than we typically do when we hear gossip about someone who is not a friend (just think how rarely we do these things in those cases). Furthermore, we will go to greater lengths in the case of a friend to construct and to entertain alternative and less damning interpretations of the reported conduct than we would for a nonfriend […]. In the case of a nonfriend, we would be unlikely even to devote the time and energy necessary to develop these other options and put them on the table. (Stroud, 2006, pp. 505–6)

Given the above example, one can conversely imagine exercising one's judgement in a way designed to favour a claim – skimming over weaknesses and looking for reasons or ways to interpret it in as positive a light as possible.¹² To be open-minded to a particular claim is to adopt a middle path, to give it an opportunity to be assessed for incorporation into the agent's belief structure.¹³ However, the content of this opportunity will necessarily be dependent on the agent's particular capacities. Humans have only a bounded rationality and therefore any agent may fall short of rationally assessing new information (Simon, 1955, p. 99). Consider the following example:

Lost Student

A student is out of their depth in a class. They deploy their available cognitive tools to the learning task but end up making mistakes in revising their understanding – for example, logical errors and being improperly moved by relevant reasons. They leave the class more confused and mistaken than when they started, even though the information conveyed to them is true.

A failure to conduct a proper rational assessment here is not necessarily caused by insufficient open-mindedness, but by a lack of other cognitive capabilities.¹⁴ In this way, open-mindedness as impartiality is narrower than rational thought – open-mindedness to new claims is no guarantee that they will be assessed according to any substantive standard of rationality.¹⁵ Impartiality here constitutes scrutiny of claims absent particular motivations to try and accept or defeat them. It is reasoning absent particular partial motivations. While this might form part of a definition of rationality, it is not synonymous with rationality.

As well as being critiqued for allegedly being overly broad, the impartiality approach has also been criticised as overly narrow by one of the most cited accounts of open-mindedness (Carter and Gordon, 2014, pp. 211–12; Kwong, 2016a, pp. 407–10; Kwong, 2017, p. 1614; Fantl, 2018, p. 3; Battaly, 2018a, pp. 263–65; Song, 2018, p. 70) put forward by Baehr (Baehr, 2011, pp. 140–61). Baehr defines open-mindedness as transcending a default cognitive standpoint to take up or take seriously the merits of a distinct standpoint (Baehr, 2011, p. 152). Baehr distinguishes his account from an impartial assessment approach to open-mindedness, which he terms the ‘adjudication model’ of open-mindedness, because he believes the latter is too restrictive (Baehr, 2011, p. 145). In particular, he argues that the model fails to encompass both situations devoid of conflict or disagreement and intellectual activities outside of rational assessment or evaluation. My own account of open-mindedness requires assessing and evaluating new claims and so it is worth distinguishing how Baehr's account goes awry here. I will address Baehr's first criticism here, and the second in the next section.

Baehr defines the adjudication model as assessing ‘one or more sides of an intellectual dispute in a fair and impartial way’ and so he reasons that if there is no dispute or intellectual conflict there is nothing to be impartial between (Baehr, 2011, p. 145). What Baehr misses is that there is always the potential for intellectual dispute or conflict with respect to any claim – whether there are clear sides or not. This is because an agent always has the option to believe or disbelieve a claim, and always stands in an attitude of partiality or impartiality in how they arrive at this belief or disbelief. Consider the following example:

Anti-Astronomer

A person explains that through new scientific methods it has been identified that there is a higher oxygen content than previously understood on the planet Proxima Centauri b. The listener is not an astronomer and has no knowledge of alternative hypotheses or theories regarding oxygen content on Proxima Centauri b.

It may initially appear that there is no intellectual dispute to assess. Yet, if the listener happens to have a deep antipathy towards the explainer they can still take a closed-minded attitude towards her claim, even without any ‘dispute’ between competing hypotheses, by trying to pick holes in her explanation. For example, they could demand that she justify her terms and the methodology, interrogate the credibility of her sources, or even challenge her motives for supporting the claim. Baehr's criticism is an artefact of how he defines the adjudication model in terms of intellectual dispute, but it does not defeat it substantively once we consider the possibility of potential conflicts with respect to any claim.

Still, even if one acknowledges the potential for conflict and partiality, some may question what standards apply to determining a claim in an impartial manner. What this constitutes will vary from person to person according to their capacities, but the central criterion is reasoning in the absence of particular motivations to try and accept or defeat the claims.¹⁶ To avoid open-mindedness collapsing into some broader substantive account of epistemic diligence, the judging standard of open-mindedness as impartiality is to treat like cases alike. To return to the example of the Democratic Senator, let us say the Senator decided to review the article commenting on President Trump and it contained an economic analysis of President Trump's foreign trade policies. To assess this open-mindedly the Senator would have to apply the same approach to reviewing this evidence as they would to any other economic policy assessment. For example, if the Senator happened to review all economic evidence presented to them in a thoughtless and ineffective manner and thus formed their resulting beliefs haphazardly, this would make them incompetent at economic analysis but not closed-minded per se. In particular, when considering alternatives to current existing beliefs, open-mindedness as impartiality requires devoting similar levels of scrutiny to the new claims as they did to the claims which caused their existing beliefs to arise in the first place.

It could be argued that this standard remains insufficiently clear, as it raises the further question of which claims are suitably alike so as to receive similar treatment.¹⁷ Ultimately this will be a matter of judgement – as it is in the judicial arena. Further guidance is provided by the motivational point I made, namely, that people making impartial judgements as to which cases are alike should not be motivated by a desire to accept or reject the claim at hand. Beyond this it is difficult to specify in advance what impartial assessment entails. Like the concept of open-mindedness itself, impartiality is in some sense a passive or responsive feature. It is primarily characterised by what is absent – partiality – rather than necessarily what it contains. Stipulating impartial assessment as a necessary step in open-mindedness also falls short of describing what these assessments may entail. As stated above, there are many elements which affect how assessments are performed, and open-mindedness is only one of these.

4. Third Step: Updating

Once a claim has been engaged with and assessed, the third stage in my definition involves changing beliefs according to one's assessment, again without fear or favour towards the new claim or one's pre-existing beliefs.¹⁸ For an agent to be open-minded it is necessary but not sufficient to engage with and assess new claims – the agent must update their beliefs in accordance with their analysis.¹⁹ In most cases claims will have implications for existing beliefs, requiring the assessment to weigh the reasons for pre-existing belief against the ones underpinning the new claims.²⁰ This genuine opportunity for belief revision to occur is in some ways the core of open-mindedness – open-mindedness is an opening of one's mind to the outside world and new inputs (Hare, 2004, p. 118; Cohen, 2009, p. 56; Arpaly, 2011, p. 75). It is not enough for mere engagement or listening to constitute open-mindedness (cf. Song, 2018). Ultimately, open-mindedness means that the continuing existence of existing beliefs are ‘on the line’. Genuine open-mindedness can therefore feel unnerving due to the uncertainty it entails in how one understands the world. Despite the centrality of the belief revision stage to open-mindedness, it is nevertheless linked to the previous two steps as one's attitude ‘upstream’ of belief revision – in screening and assessing new claims – can significantly pre-empt the updating process if one happens to be partial either for or against new claims.

It is not possible here to set out a full-blown account of belief revision (Hansson, 2021), but a rough intuitive understanding should suffice. Say I believe in a proposition because I have a number of reasons supporting it. I then determine that a claim constitutes evidence opposing the proposition. As a result I update my beliefs to accommodate the apparently contradictory information (Harman, 1986, pp. 56–57). Unless for some reason the new claim entirely eliminates the force of the pre-existing reasons, I will probably decrease my degree of certainty regarding the proposition. A number of authors have argued against this type of belief updating as a component of open-mindedness. Baehr argues open-mindedness applies in situations without rational assessment or evaluation, and Riggs and Adler argue that open-mindedness ought to be compatible with what they call ‘strong belief’ which holds no possibility of revision. Reviewing their arguments will help cast light on why this third step is necessary.

As I flagged in the previous section, Baehr's second criticism of open-mindedness as impartiality is that open-mindedness should apply in situations other than rational assessment or evaluation. To demonstrate how his conception of open-mindedness goes beyond the ‘adjudication model’ Baehr proposes an example of students who need to ‘open their minds’ to learn challenging new information – Einstein's General Theory (Baehr, 2011, p. 146). For Baehr, ‘[…] there is a clear sense in which they [the students] are not attempting to think for themselves. There is a fixed subject matter before them and their aim is to wrap their minds around it – to grasp it’ (Baehr, 2011, p. 146). Working through Baehr's example will help us understand why assessment and evaluation alongside belief updating and revision are in fact necessary components of open-mindedness.

I agree with Baehr that there is something in this ability to engage with new information – to grasp it – which is important for open-mindedness.²¹ However, Baehr's argument misses that to constitute open-mindedness any such engagement must be followed by rational assessment or evaluation to understand how this new information interacts with their pre-existing beliefs. Part of what is doing the work in Baehr's example in trying to avoid rational assessment and evaluation is that these students are receiving highly complex information from a trusted epistemic source – their teacher – which implies that the students should not be assessing the information themselves. They are essentially trying to take as a given truth what their teacher is telling them, as opposed to impartially assessing whether it is true, because it is beyond their current abilities to properly assess the new propositions. In my terminology they are demonstrating credulousness at the stage of belief revision – a partiality towards the new claim. Nevertheless, the point remains that the students must revise their understanding of the world in order for Baehr's example to make sense and this will necessarily require assessment and evaluation, albeit weighted towards credulously accepting what they are being told. To understand this distinction imagine the following:

Closed-Minded Explainer

A closed-minded person is not willing to entertain the truth of Einstein's General Theory. Nevertheless, they are required by circumstances to explain it to someone else, and so they embark on the journey of Baehr's students and seek to follow the theory and grasp its internal logic. Because they are closed-minded, they do so with no intention of allowing it to influence their own beliefs; they just want to be able to faithfully explain the concepts to a third party.

Baehr appears to rule out this possibility when he states that open-mindedness requires an agent to be committed to ‘taking up or taking seriously’ a new standpoint, which requires giving them ‘a “serious” (i.e. fair, honest, objective) hearing or assessment’ (Baehr, 2011, p. 151). Yet this requirement does not in fact rule out Closed-Minded Explainer. Let us say they are committed to Caplan's Ideological Turing Test whereby they must be able to explain any position they oppose so fluently that they could be taken for a genuine proponent of its views (Caplan, 2011). This requires them to fairly, honestly, and objectively reconstruct the theory in order to explain it in a suitably convincing and comprehensive manner. According to Baehr's definition, it appears that they have successfully demonstrated open-mindedness with respect to Einstein's Theory, but I maintain that by keeping their own beliefs carefully segregated from the new information the explainer still qualifies as closed-minded.

The reason for this discrepancy is because Baehr's original explanation is not explicit as to what happens once a person has followed or understood a new standpoint beyond ‘taking it up or taking it seriously’. My account is clear that revision of a person's pre-existing viewpoints has to be a third step. Therefore, once a new position has been engaged with and understood, the person needs to reconcile this with their pre-existing knowledge and this requires rational assessment and evaluation. The Closed-Minded Explainer, for all their hard work learning and understanding the internal logic of Einstein's Theory, stops short of this assessment, evaluation, and reconciliation process and so avoids open-minded belief revision in favour of closed-mindedness.²²

My account entails the potential for belief revision – either in terms of an agent adjusting the certainty of their beliefs or wholesale belief change – as the third component of open-mindedness. This element is opposed by Adler and Riggs who each take issue with updating one's certainty of belief as a component of open-mindedness (Adler, 2004; Riggs, 2010). They argue that understanding oneself as potentially fallible constitutes open-mindedness, in particular awareness of one's cognitive biases or flaws. Their aim is to make open-mindedness compatible with what they call ‘strong belief’ which a person may hold without entertaining any possibility that they are wrong. My own account is sympathetic to Adler and Riggs’ focus on fallibility – open-mindedness as impartiality entails the revisability of beliefs. However, their aim in trying to fit fallibility with strong belief leads them astray. In particular, Adler and Riggs’ argument leads to contradictions. In order to make ‘strong belief’ compatible with awareness of one's own fallibility Adler and Riggs both argue for a strong separation between this meta-knowledge of fallibility and holding object-level beliefs with certainty (Adler, 2004, p. 131; Riggs, 2010, p. 180):

The possibility that I, or the method I employ, has erred in coming to believe that p is not the possibility that the proposition believed is false, given my grounds for it. (Adler, 2004, p. 130)

Adler uses the analogy of a widget factory to explain this (Adler, 2004, p. 132).²³ This widget factory has very high standards of competence but still requires random spot checks for quality control. In his view, there is no conflict or incompatibility between these two assessments being simultaneously held by the quality control inspector: ‘This widget has no defects or imperfections’ and ‘I should carefully examine widget 30 for defects or imperfections’. Adler wants to keep the strong belief in the widget's quality – the first statement – carefully segregated from the meta-level knowledge of potential fallibility – in this case the possibility, albeit of low probability, of failure. But there is an inconsistency in keeping entirely separate meta-level and object-level beliefs where the former pertain to a category that includes the latter.²⁴ This becomes apparent if we tweak the example to feature a flawed widget factory with only a 50 per cent standard of competence such that every other widget is defective. In the tweaked example it should be clear that it is incoherent for the quality control inspector to maintain their strong belief in the widget's quality while also acknowledging the meta-level of the widget factory's fallibility. Simply put, the degree of object-level certainty has to be conditional on the meta-level fallibility for agents to avoid inconsistent beliefs.

This leads me back to my point that acknowledging our own fallibility should leave us open to revising the certainty of our object-level beliefs as a component of open-mindedness. It is particularly this conclusion that Adler and Riggs are keen to avoid in their defence of strong belief. Adler and Riggs argue that we do not, and should not, hold degrees of belief as this would be too complex, lack commitment, and demonstrate either epistemic insecurity or cowardice (Adler, 2004, p. 129; Riggs, 2010, p. 180). I disagree: there does not seem to be anything particularly insecure, cowardly, or lacking in commitment for becoming less, or more, sure of a proposition as new evidence emerges.²⁵ Instead, this seems to be the appropriate response, paraphrasing an apocryphal statement of Keynes – when our information changes so should our beliefs. In the example of the Democratic Senator, if the Senator found the analysis in support of President Trump compelling, they should then update their underlying beliefs. For example, if it contained a positive analysis of President Trump's economic policies, they might reduce their certainty that President Trump was bad for the US economy. This would not necessarily impact their assessment of President Trump in other respects; for example, they may still maintain that President Trump was corrupt or otherwise pursued unwise or unjust policies.²⁶

Through the above analysis it should be clear that a plausible understanding of open-mindedness will require its adherents to review and revise their pre-existing beliefs following their assessment of new claims. Open-mindedness does not rest on simply following or understanding new information without reflecting on its implications for the person's actual beliefs. Neither can open-mindedness be compatible with holding existing beliefs so strongly that one cannot entertain doubts about them. Still, this leaves plenty unsaid about how precisely one weighs new claims in light of one's pre-existing knowledge. As before, being open-minded only says so much about one's reasoning and judgement. It entails that this reconciliation process of new claims and pre-existing knowledge be carried out without a motivation for one to prevail over the other – a form of impartiality. One cannot have consciously or unconsciously pre-decided which of the new claim and the pre-existing knowledge should have the better of the reconciliation process. Beyond the absence of this motivation, how one actually reconciles new claims which purport to conflict with or alter one's pre-existing beliefs is a matter of exercising one's reasoning faculties.

5. Exercising Open-Mindedness: A Matter of Judgement

One of the intended consequences of my definition of open-mindedness is to shift the normative evaluation of appropriate cognitive traits to the context they are exercised in. This runs contrary to the usual lionising approach to open-mindedness. For the early Rawls, open-mindedness was a necessary characteristic of the ‘reasonable man’ (Rawls, 1951, p. 179). Hare cites a range of philosophers including Socrates, Russell, Mill, and Dewey for whom open-mindedness was central to serious philosophical enquiry, and he himself produces a slew of reasons for its importance including helping to reach justice, dealing with disagreement, and avoiding intellectual hubris (Hare, 2006, pp. 7–15). Open-mindedness's conduciveness towards discovering truth is also cited as a reason to always adopt it (Hare, 2006, p. 15; Kwong, 2017).²⁷ One might think of open-mindedness and reasons as ‘asymmetric weapons’ – unlike violence which can triumph irrespective of moral weight, open-minded reasoning should favour truth and correct outcomes (Hare, 2004, p. 117; Alexander, 2017). As cited in the Introduction, this thinking is all of a piece with the more general thesis that acting virtuously entails acting in a correct manner (Hursthouse, 1999, p. 28). Following this thesis, if open-mindedness is a virtue, then acting open-mindedly is also necessarily correct.

As the foregoing definitional analysis has hopefully highlighted, matters are not nearly so simple. Even advocates of open-mindedness acknowledge that it can sometimes lead individuals epistemically awry (Hare, 1985, p. 4; Baehr, 2011, p. 64). This seems plausible for the simple reason that updating beliefs can lead to them becoming more or less correct. What does the work in getting this updating process right is a function of our existing knowledge, specific environment, and reasoning abilities. There are therefore risks to open-minded enquiry. For example, we risk believing untruths or violating duties not to consider, believe, or express moral wrongs (Brennan and Freiman, 2020).

The standard Aristotelean virtue-theoretical way to resolve this issue is to say that the virtue of practical wisdom, phronesis, is required to determine when and how to act virtuously open-mindedly when demanded by the relevant situation (McDowell, 1979; Athanassoulis, 2018). In one sense I am sympathetic to this emphasis on practical wisdom as taking into account the wide variety of potentially relevant factors when considering whether to be open-minded. The purpose of developing a normatively thin account of open-mindedness in this paper is to shift the evaluative emphasis to judgement of contextual factors. I do note, however, that phronesis cannot be read as a catch-all solution. In the words of Curzer:

Practical wisdom is sometimes treated as a magical incantation that solves all of the problems confronting virtue ethics theorists. Discussions proceed thus.

Questioner: “How do virtuous agents do X?”

Virtue Ethicist: “Practical wisdom!”

Black box materialises.

Questioner is silenced and dumfounded. (Curzer, 2023, p. 221)

In particular, Curzer is sceptical of the ability of phronesis to adjudicate in conflicts between competing virtues or virtues and duties, describing them as incommensurable and rejecting Socratic ideas of weighing different goods to determine appropriate trade-offs (Curzer, 2023, pp. 218–19). I do not take a position here on the extent to which phronesis can solve apparent conflicts between various goods and duties. What I instead set out in the remainder of this paper are some illustrative examples of how a wide range of value considerations such as one's reasoning capabilities or knowledge, epistemically polluted environments, the value of knowledge-seeking itself, racial justice, and emotional labour, can all come to bear on whether one should be open-minded or not. Ultimately if, as Battaly (2018b) and Curzer and Gottlieb (2019) argue below, open-mindedness can be a vice under particular circumstances, then it seems that the real work in determining the normative valence of the trait is being done by this contextual assessment and not the trait itself. The labels ‘virtue’ and ‘vice’ become contingent on situational judgement in a manner congruent with my analysis.

Authors in the literature have discussed how circumstances can lead to closed-mindedness being appropriate for epistemic reasons. Fantl argues for closed-mindedness when people lack relevant expertise and are therefore susceptible to being taken in by misleading arguments from those who are more argumentatively skilled (Fantl, 2018, pp. 27–48). He does this by arguing that individuals may be misled by incorrect arguments without being able to determine why or how they are flawed. Battaly argues for closed-mindedness when faced with what she terms ‘epistemically polluted environments’ – environments filled with falsehoods and poor epistemic norms (Battaly, 2018b, pp. 37–44). Battaly is clear that she does not think this analysis characterises the current informational environment for most people, affecting perhaps just members of non-dominant groups suffering under oppressive epistemic norms (Battaly, 2018b, pp. 32 and 40). In contrast, Curzer and Gottlieb argue that the necessary conditions for positive open-mindedness are so undermined in the current world so as to make open-mindedness a vice outside of the classroom (Curzer and Gottlieb, 2019). They cite factors including overwhelming amounts of novelty, increased specialisation of knowledge, and organised campaigns of deception. In each of these discussions what is at stake is epistemic risk – the possibility of open-mindedness leading to more or less truthful beliefs and the implications for one's knowledge. There are, however, a wide array of non-epistemic considerations which also bear on whether one should be open-minded or not.

Demands on cognitive and emotional resources may lead us to deviate from open-mindedness. For example, consider Holocaust denial. Cassam describes avoiding Holocaust denial material as ‘nothing to be proud of’ (Cassam, 2019, p. 117). For Cassam, all persons who want to genuinely know that the Holocaust occurred must engage with counterarguments to it, such as the work of Holocaust denier David Irving, and we cannot simply rely on professional historians to do so for us (Cassam, 2019, pp. 100–20). Cassam rules out resorting to any kind of Kripkeian dogmatism by which one could ignore evidence against P on the basis that we know P is true (Kripke, 2011). Instead, Cassam maintains that ‘It is incumbent on us as knowers to base our views on the evidence, and that means taking the trouble to find out what the evidence actually is and what it does (and doesn't) show’ (Cassam, 2019, p. 118). While Cassam emphasises the importance of epistemic rigour in achieving knowledge of such an important historical event, there are nevertheless competing non-epistemic factors. There are well-known unevenly distributed burdens for different groups in engaging with certain topics or debates. For example, Eddo-Lodge has spoken about the emotionally and physically draining impact of having to endlessly justify her lived experiences of structural racism (Eddo-Lodge, 2014). In a similar vein, Kittay has spoken movingly about the emotional burden of engaging in debate regarding the moral and personhood status of her cognitively disabled daughter (Kittay, 2009). While Eddo-Lodge and Kittay have themselves publicly engaged in these discussions, it seems understandable that others facing similar costs might reasonably prefer to closed-mindedly screen out potentially racist or ableist claims for reasons of self-preservation and self-care.

The stakes of open-mindedness can also be explicitly moral – some authors have argued that there can be a moral obligation to be open-minded (Arpaly, 2011; Song, 2018). It certainly seems correct that open-mindedness can form part of an attitude of respect towards another which we might owe. For example, Song describes positively the case of Daryl Davis open-mindedly listening to Klu Klux Klan members (Song, 2018). Yet this logic runs both ways – presumably we should therefore withdraw open-mindedness when we do not owe respect. Indeed, in contrast to Song Fantl argues that open-minded engagement with morally problematic arguments and speakers can constitute a failure to stand in solidarity with oppressed groups or individuals (Fantl, 2018, p. 147). He explicitly rules out giving respect to those sympathetic to the Klu Klux Klan (Fantl, 2018, p. 191). For Fantl these arguments and speakers are not due respect, in the same way that someone hurling abuse is not entitled to politeness. Song and Fantl therefore agree that there are moral stakes to attitudes of open-mindedness with interlocutors but disagree on how these are best understood.

I do not mean to resolve these apparent conflicts here, and my point is not that people should always avoid engaging with difficult or what are perceived to be morally incorrect claims. Sometimes engagement may be appropriate, or even required, particularly if the holders of wrong views are to be persuaded otherwise. I instead mean to illustrate that these are judgements, and often trade-offs, we have to make with respect to open-mindedness. Moving away from an inherently normative understanding of open-mindedness allows us to see these trade-offs more clearly in the context within which they appear. Many different cognitive attitudes might be justified depending on the situation. Sometimes we should screen out claims, or in listening to them we should not do so open-mindedly.²⁸

It could be argued that this emphasis on a range of potential options and case-by-case assessment for open-mindedness risks leaving people adrift without guidance on how to behave. I therefore conclude with three comments by way of reply. First, it is often better to be uncertain and to work out the correct approach in a given situation rather than to confidently rely on problematically broad generalisations. Second, and more importantly, characterising open-mindedness as a virtue and alternatives as vices leads us to be too hesitant to recognise ourselves as improperly open-minded and reluctant to acknowledge the appropriateness of alternatives. By my definition we are all at times closed-minded or credulous to various extents. In many instances we live by the dictum proposed by Russell: ‘As a general rule, it is good to be open-minded about whatever does not affect adversely the broad pattern of your life’ (Russell, 1950, p. 151). We treat beliefs on various matters as ‘settled’ and screen out, or do not engage with, arguments or evidence presented to the contrary. This allows us to devote our energies to the pursuit of valuable projects within our belief paradigms. At other times, for example in deferring to experts or educators, we may allow credulity to largely replace our own judgements. My definition of open-mindedness is not supposed to presuppose the answer to these questions. Nevertheless, it will help us to move beyond using the terms ‘open-mindedness’ and ‘closed-mindedness’ as simple positive or negative epithets but to instead focus discussion on the reasons which sit behind these cognitive attitudes. Third, if one was to try and derive general rules or principles from the above literature for when to practise open-mindedness or its alternatives, one option might be that credulity or open-mindedness are most appropriate when the agent identifies the circumstances as ‘benign’. Benign circumstances include trusting the good faith of potential interlocutors or reputability of sources one is likely to encounter as well as having sufficient time, energy, and resources (such as emotional and cognitive resources) to properly consider claims. The paradigmatic cases of benign circumstances include well-run classrooms or educational institutions (Curzer and Gottlieb, 2019). In contrast, closed-mindedness might be most appropriate in non-benign circumstances which demonstrate the contrary characteristics – malicious interlocutors, a misinformation-strewn environment, and a lack of personal preparedness to properly consider claims.

6. Conclusion

In this paper I set out a novel process conception of open-mindedness as impartial belief revision and in doing so distinguished it from a number of competing accounts in the field. I argued that open-mindedness is best understood as a three-step process of not screening new claims, assessing them impartially, and then updating one's beliefs accordingly. An important consequence of this approach has been to separate the question of whether to be open-minded from its definition. I have highlighted throughout this paper a variety of instances where deviations from open-mindedness might be reasonably justified, ranging from credulity in classrooms to closed-mindedness to traumatic material. I raised in the penultimate section some very tentative ideas regarding what guidelines might help us in determining when to be, or not to be, open-minded. This is a potential area for future research – such guidelines will not only depend on the relevant context, taking account of resources, duties, and so forth, but also the structure and characteristics of the beliefs and claims in question, for example how central they are to our life projects, their relationship to our other beliefs and overarching worldview, our pre-existing evidence for them, and their empirical and normative content mix.