Belief and pluralistic ignorance

Pluralistic ignorance is a recurrent topic in Sociology and Psychology, and it is also treated in Cognitive Sciences and Philosophy of Social Sciences. Firstly mentioned as such by Katz and Allport (1931), it refers to the est ablishment of a social norm or behavior when every agent privately refuses such norm or behavior but she believes that most other agents assume and follow it. Many studies about this phenomenon have been developed, most of them useful applications to different attitudes and behaviors: teenagers drinking alcohol, classroom habits, top managements attitudes, racist attitudes, revenge and infidelity behaviors, etc. Nevertheless, theoretical accurate approaches to the phenomenon, its definition and treatment are scarce. Although this is ABSTRACT


Introduction
Pluralistic ignorance is a recurrent topic in Sociology and Psychology, and it is also treated in Cognitive Sciences and Philosophy of Social Sciences. Firstly mentioned as such by Katz and Allport (1931), it refers to the est ablishment of a social norm or behavior when every agent privately refuses such norm or behavior but she believes that most other agents assume and follow it. Many studies about this phenomenon have been developed, most of them useful applications to different attitudes and behaviors: teenagers drinking alcohol, classroom habits, top managements attitudes, racist attitudes, revenge and infidelity behaviors, etc. Nevertheless, theoretical accurate approaches to the phenomenon, its definition and treatment are scarce. Although this is 261 a theoretical approach, I exemplify such theoretical approach with the classroom case, one of the most popular cases in the Pluralistic Ignorance literature.
In the second section I consider the definition of pluralistic ignorance given by Bjerring, Hansen and Pedersen (2014). This definition explains pluralistic ignorance in terms of beliefs, act ions and evidence and it is going to be the basis for the rest of the study. In the third section I consider a broad epistemic approach to belief -for an agent S and a proposition p, S believes p if and only if S considers p to be true -and I apply such approach to the analysis of pluralistic ignorance according to the definition given by Bjerring et al. (2014). In the fourth section I consider a broad pragmatic approach to belief -for an agent S and a proposition p, S believes p only if to believe p is useful for S -and I apply such approach to the analysis of pluralistic ignorance according to the definition given by Bjerring et al. (2014). Furthermore, I consider a mixed epistemic-pragmatic approach to believe -for an agent S and a proposition p, (i) S believes p if and only if S considers p to be true and (ii) to consider p to be true is useful for S -and I apply such approach to the analysis of pluralistic ignorance according to the definition given by Bjerring et al. (2014). Finally, in the fifth section I defend that a theoretical study of pluralistic ignorance is useful to model such phenomena. Moreover, that pluralistic ignorance phenomena can offer some clues and arguments in the debate between pragmatic and epistemic accounts of belief. I conclude that pluralistic ignorance can be easily understood if we take a pure epistemic position about beliefs, accepting that final behaviors and act ions depend not only on beliefs but on other pragmatic elements and attitudes. Nevertheless, the pragmatic position may offer a coherent complex analysis that does not need of the concept of truth. Finally, I state that a mixed epistemic-pragmatic position is harder to apply but it can explain pluralistic ignorance more accurately.

What is Pluralistic Ignorance?
Roughly put, a social situation is a situation of pluralistic ignorance when a group of individuals all have the same attitude towards some proposition or norm, all act contrary to this attitude, and all wrongly believe that everyone else in the group has a certain conflicting attitude to the proposition or norm (Bjerring et al., 2014(Bjerring et al., , p. 2446).
Literature about pluralistic ignorance usually focuses on some popular examples: teenagers drinking alcohol (Prentice;Miller, 1993), classroom habits, top managements attitudes, racist attitudes, revenge and infidelity behaviors (Lambert;Kahn;Apple, 2003), etc. Bjerring et al. (2014) actually refer to three of these cases: the classroom case, the college drinking case and the Emperor's case (Bjerring et al., 2014(Bjerring et al., , p. 2448. Although this is a theoretical approach, I will refer to the classroom case in order to exemplify how the epistemic, the pragmatic and the mixed approaches to belief can apply to pluralistic ignorance: (The classroom case) A professor asks her students if they have any doubt when finishing a -particularly difficult-lesson. Everybody doubts but nobody raises hands: each student believes the rest of students have understood the lesson and as nobody wants to be publicly displayed as the only ignorant, nobody asks (Katz;Allport, 1931).
Different cases are argued in order to explain pluralistic ignorance (Brennnan et al., 2013;Bicchieri, 2006). The main one considered is a self-other difference while interpreting personal and others' beliefs and actions. There can be also an encoding difference in which actions feed wrong beliefs. Minorities -e.g. clever students in the classroom case -can influence the rest of people and in that way they can feed pluralistic ignorance. There can be a desire to maintain social identity even if accepting and acting according to falsehoods. And there can also be a lack of transparent communication.
Many of the definitions and theoretical approaches to pluralistic ignorance treats the phenomenon in terms of social norms (Prentice;Miller, 1993;Bicchieri, 2006;Brennan et al., 2013). Nevertheless, the definition given by Bjerring et al. (2014), after a detailed analyses of the main definitions, already given treats pluralistic ignorance as a system of beliefs, act ions and evidence: 'Pluralistic ignorance' refers to a situation where the individual members of a group (i) all privately believe some proposition p; (ii) all believe that everyone else believes ¬p; (iii) all act contrary to their private belief that p (i.e. act as if they believe ¬p); and where (iv) all take the actions of the others as strong evidence for their private beliefs about p (Bjerring et al, 2014(Bjerring et al, , p. 2458. Interestingly, there is a circular connection between steps (ii), (iii) and (iv) that helps to explain pluralistic ignorance persistence. Every agent S -who believes p -believes that the rest believe ¬p (ii), so S acts contrary to her initial belief p (i.e. act as if she believes ¬p) (iii), and in that way S provides strong evidence for the rest (iv) to believe that S believes ¬p (ii). In the classroom case, every student believes that the rest of students believe that they do understand the lesson (ii), so every student acts contrary to her private belief that she does not understand the lesson (i.e. act as if she believes that she understands the lesson) (iii), and in that way every student provides strong evidence for the rest (iv) to believe that she believes that she understands the lesson.
Two question arises from this definition. First, it can be argued that the qualification of the agent's attitude as a belief -S believes that she does not understand the lesson; every 262 student believes that the rest of students believe that they do understand the lesson -is odd -it is more usual to state that 'S does not understand the lesson' and that 'all believe that everyone else understands the lesson' . The authors realize this criticism and they state that pluralistic ignorance cases can be reformulated as involving beliefs -e.g. "each student believes that everyone but him believes that the material was not difficult" (Bjerring et al., 2014(Bjerring et al., , p. 2448. I accept this assumption 2 . Second, the circularity between (ii), (iii) and (iv) helps to explain pluralistic ignorance persistence. Nevertheless, it does not fully explain pluralistic ignorance emergence. It does not say which step comes first, that is, it does not est ablish if first particular agents develop their beliefs about the rest of agents (ii), if their particular act ions motivate the rest to develop the rest beliefs (iii) or if the act ions of the rest motivate the particular agent to believe that the rest understand the lesson (iv). In other words, this charact erization of pluralistic ignorance does not fully explain what comes first: beliefs (ii) or evidence to further develop beliefs (iii & iv). In the classroom case, it can be easily defended that first agents do not raise their hands to solve their doubts (iii & iv) and then they privately develop their beliefs about of the rest of students' beliefs (ii). Nevertheless, in other cases the priority of evidence is less clear 3 . This question opens an interesting ramification for the study of evidentialism on beliefs, but this is a topic for another occasion.
I defend that pluralistic ignorance is an empirical phenomenon useful for the analysis of belief´s nature. The goal of this paper is to analyze and apply different conceptions of belief to the pluralistic ignorance definition in terms of beliefs given by Bjerring et al. (2014): a broad epistemic approach (third section: for an agent S and a proposition p, S believes p if and only if S considers p to be true), a broad pragmatic approach (fourth section: for an agent S and a proposition p, S believes p only if to believe p is useful for S) and a mixed epistemic-pragmatic approach to belief (fifth section: for an agent S and a proposition p, (i) S believes p if and only if S considers p to be true and (ii) to consider p to be true is useful for S). I determine that all the previous approaches to belief may apply to the pluralistic ignorance definition in terms of belief given by Bjerring et al. (2014), although the mixed epistemic-pragmatic approach offers a more detailed and sp ecific explanation of the phenome-non. At the same time, the theoretical study of belief is useful to model pluralistic ignorance itself.
An epistemic approach of belief applied to pluralistic ignorance Links between belief and truth is a recurrent topic in epistemology. Recently, some normative approaches are being defended -e.g. a belief p is correct if and only if p is true (Wedgwood, 2013), agents ought to believe p if and only if p is true (Wedgwood, 2013), agents may believe p if and only if p is true (Whiting, 2010), agents ought to want their beliefs to be true (Horwich, 2013), a virtue-theoretic account of epistemic norms of belief (Turp, 2013). Other authors (Engel, 2013a(Engel, , 2013bToribio, 2013;Papineau, 2013;Glüer;Wikforss, 2013) explain the relationship between belief and truth in non-normative terms. Here I take a broad relationship between belief and truth that most authors who consider an epistemic treatment of belief would accept, even if it may be too soft for some of them (Glüer;Wikforss, 2013).
For an agent S and a proposition p, S believes p if and only if S considers p to be true.
It must be pointed that this basic relationship between belief and truth relates to the agent' s attitude -S considers p -and not to the content nor the final result of her belief -just p. 4 The application of this broad epistemic approach of belief to pluralistic ignorance phenomena considering the definition given by Bjerring et al. (2014) results as follows,

'Pluralistic ignorance' refers to a situation where the individual members of a group (i) all privately consider true some proposition p;
(ii) all consider true that everyone else considers true ¬p; (iii) all act contrary to their private consideration that p is true (they act as if they consider true ¬p); and where (iv) all take the actions of the others as strong evidence for their private consider-2 In their footnote 6, Bjerring et al. (2014) state that: "For the purposes of this paper, we characterize situations of pluralistic ignorance in terms of a discrepancy between private and public beliefs. But note that other attitudes could in principle be used to characterize pluralistic ignorance. For instance, several characterizations of pluralistic ignorance involve attitudes towards norms rather than beliefs about propositions (…) the College Drinking Case can naturally be reformulated as involving norms rather than beliefs. However, since nothing substantial in our discussion hangs on the difference between "having a belief towards a proposition" and "having an attitude towards a norm", we will work with the first locution in what follows" (Bjerring et al., 2014(Bjerring et al., , p. 2449. 3 For instance, in the College Drinking Case nobody drinks alcohol at the very first moment. There must be some few teenagers that start drinking alcohol seeing nobody else doing so. 4 It may be argued that this characterization of belief is true only for explicit beliefs, but that cannot be the case for implicit beliefs. Let´s consider wishful thinking cases like the following one: (Kate's hated terrorist case A). Kate's son is a terrorist hated by the whole -or almost the whole -country, but she believes her son is innocent despite the reliable evidence showing that her son is a terrorist. Kate´s implicit belief is false but she still considers it to be true because of self-deception (Adler;Hicks, 2013, p. 164). The broad epistemic characterization of belief I am taking -the broadest I know -does not relate belief and truth in terms of the content itself, but in terms of the agent´s attitude.  Bjerring et al, 2014Bjerring et al, , p. 2458).
In the classroom case, (i) every student privately considers true that she does not understand the lesson, (ii) every student considers true that everyone else considers true that they do understand the lesson, (iii) every student acts contrary to their private consideration that she truly does not understand the lesson (she acts as if she considers true that she understands the lesson), and (iv) every student takes the act ions of the others as strong evidence for her private considerations about the truth of the rest of students understanding the lesson.
The application of the basic epistemic treatment of belief proposed is easy and it perfectly fits. The only premise we need to assume is that for act ing not only epistemic considerations but other considerations like pragmatic ones may come into play -e.g. a self-other difference or social attributions. In pluralistic ignorance, final act ion (iii) against the agent's particular belief (i) partly finds its explanation in the belief about the rest beliefs (ii). The differences between beliefs and act ions are already assumed by most authors that analyze the relationship between belief and truth (Whiting, 2014;Bykvist;Hattiangadi, 2013, p. 114;Unwin, 2007).

A pragmatic approach of belief applied to pluralistic ignorance
The explanation of belief in pragmatic terms is also a recurrent topic in Philosophy (Engel, 2013a;Papineau, 2013;Rinard, 2015; see also The numbers games in Reisner, 2013) 5 . In this section, I consider a broad pragmatic charact erization of belief. Such charact erization does not use the concept of truth and it is based on a general pragmatic idea about beliefs: beliefs are useful for their bearers, they always have a positive effect on their bearers.
For an agent S and a proposition p, S believes p only if to believe p is useful for S. Or, if preferred, for an agent S and a proposition p, if S believes p then to believe p is useful for S.
It may be argued that the notion of being useful or practical is vague. Here I take a broad relationship between belief and useful effects -the broadest pragmatic charact erization of belief I know -for the sake of the pragmatic argument, even if it may be too soft for some of the authors. 6 Some considerations about this basic pragmatic treatment of belief should be done. First, there is no double implication in this consideration: a belief p may be useful for an agent but it does not force the agent to adopt such belief. In other words, it is not the case that (if S belief on p is useful for S, then S believes p). Many examples of useful non-possible beliefs are present in the literature on beliefs 7 .
Second, contrary to the previous basic epistemic approach, it may be not the very believer who considers her belief to be useful but another agent. The basic pragmatic approach states that 'to believe p is useful for the believer S' but that does not mean that the believer S considers her belief p to be useful. That is, it may be other agent S 2 who considers the belief p whose bearer is S to be useful for S 8 .
Third, the content of the believed proposition p does not need to be considered useful for the believer S but just the attitude of believing. In other words, it is not the case that (if S believes p then p is useful for S). Much less a useful proposi-5 Engel (2013a) accurately sketches this question: There are, after all, plenty of ways in which we can assess beliefs, besides their truth or falsity, and in a number of cases it can be beneficial to disregard or to ignore their alethic or rational dimensions. Why suppose that there is only one standard of evaluation? Why should we adopt the absolutist view that there is only one constitutive norm of belief rather than the relativist view that there is no particular privileged criterion for assessing belief which enjoys a privileged status? On this view there might be as many 'norms' for beliefs as there are dimensions of evaluation, depending upon our particular interests in particular contexts, none of which enjoy any central status (Engel, 2013a, p. 34-35) 6 Interestingly, Rinard (2015) defends 'Robust Pragmatism', the thesis that (i) a pragmatic reason for a belief is always a genuine reason and (ii) the only genuine reasons for beliefs are pragmatic reasons: C is a reason to believe p if and only if C is a pragmatic consideration in favor of believing p (Rinard, 2015, p. 218).
[t]here are of course some metaphysically possible scenarios in which evidence for p would not constitute a pragmatic consideration in favor of believing p. Such cases are rarer than one might think, however. What we have to imagine is a case in which believing the truth with respect to p does not make it even a tad bit more likely that your life -or the lives of others -will go well. (…) Insofar as I have intuitions about such cases, it seems to me that the verdict of Robust Pragmatism is exactly right. If it really is the case that true beliefs concerning p would not make it even the slightest bit more likely that my life, or that of others, will go well; and I am genuinely completely indifferent to whether or not I believe the truth with respect to p; then, in my view, evidence in favor of p does not give me any reason whatsoever to believe it (Rinard, 2015, p. 220). Rinard must admit that evidential reasons are many times pragmatic reasons for belief formation in order to save some criticism against her robust pragmatism. See Whiting, 2014 and footnote 6. 7 For instance, Whiting (2014) refers to the following case: (David Cameron's doctor's uncle case). Suppose that one knows that if one were to believe that David Cameron's doctor's uncle has 132,487 hairs on his head one would receive a generous amount of money (…) since the fact that one would receive a financial reward were one to have the relevant belief is no evidence that the belief is true, it seems that one cannot take it to justify so believing (Whiting, 2014, p. 220). 264 tion needs to form a belief -it is not the case that (if p is useful for S, then S believes p) 9 .
The basic pragmatic treatment of belief just est ablishes that once an agent S believes p, then the attitude of believing p is useful for S. The application of this account to pluralistic ignorance phenomena according to the definition in terms of beliefs, act ions and evidence provided by Bjerring et al. (2014) Bjerring et al, 2014Bjerring et al, , p. 2458).
In the classroom case, first in (i) every student privately believes that she does not understand the lesson and this belief is useful for her -e.g. the belief allows the student to further solve her doubts.
Second, in (ii) every student believes that everyone else believes that they understand the lesson. That is, it is useful for every student to believe (that everyone else believes that they understand the lesson) -i.e. in order to adopt the best social behavior -and everyone else belief that they understand the lesson is taken to be useful for them -i.e. if they really understand the lesson and they are conscious, then they may dedicate more time to study other issues.
Third, in (iii) every student acts contrary to her belief that she truly does not understands the lesson -i.e. she acts as if she believes that she understands the lesson -but at the same time such belief is taken to be useful for her. And finally, in (iv) every student takes the act ions of the others as strong evidence for her private belief, which is taken to be useful for her, that the rest of agents believe they understand the lesson, belief that is taken to be useful for them.
The application of the basic pragmatic approach of belief to the pluralistic ignorance phenomena in terms of the definition of Bjerring et al. (2014) is harder than the application of the previous basic epistemic approach. The fuzziest step is (iii), in which the agent acts contrary to her initial belief that is taken to be useful for her (i), and it is very difficult to envisage a basic pragmatic approach that allows for conscious non-useful act ions: pragmatic accounts also est ablish that agent's act ions want to produce useful outcomes. But this difficulty may be solved appealing to step (ii), in which it is useful for every agent to believe that the rest of agents believe ¬p, belief that at the same time is considered to be useful for the rest of agents. For instance, in the classroom case every student has two confronted beliefs: the belief that she does not understand the lesson and the belief that the rest of students understand the lesson. Both beliefs present different useful outcomes and finally the agent acts following the second belief rather than the first one. A self-other difference cause may be argued to explain this behavior (Brennnan et al., 2013;Bicchieri, 2006). Every particular agent relies more on what is considered to be useful for the rest of agents than on what it is initially considered to be useful for herself, and she finally develops her act ions according to what she considers to be more useful -i.e. to be socially accepted. Such consideration is enhanced by the evidence given by the others' act ions (iv).
A basic pragmatic account of belief may deal with pluralistic ignorance phenomena and it does not take into consideration the concept of truth. On the contrary, it does not substitute nor define belief: it just est ablishes that the belief is useful for the believer. The treatment of the pluralistic phenomena is more difficult and fuzzier than the offered by the previous basic epistemic approach because different pragmatic outcomes given by different beliefs confront and final action is explained by only one of these beliefs.
An epistemic-pragmatic approach of belief applied to pluralistic ignorance The third and last approach I analyzed is a mixed pragmatic-epistemic approach based on the idea that beliefs aim at truth and truth provides a useful reason for further act ion. Daniel Whiting (2014) has recently defended this idea: [b]elieving only the true because subjects aim to believe only what is a useful reason because subjects aim in action and decision to be guided only by useful reasons. The epistemic perspective is not in tension with the useful perspective but dictated and contained within it (Whiting, 2014, p. 21-22).

265
It must be pointed that this approach states that truth provides a useful reason for further act ion, but not the only one. Other non-epistemic reasons may be argued to explain further act ion.
Based on this idea and combining the previous basic epistemic and pragmatic approaches to believe, I consider the following epistemic-pragmatic approach to belief: For an agent S and a proposition p, (i) S believes p if and only if S considers p to be true and (ii) If S considers p to be true, then to consider p to be true is useful for S. Some considerations should be done. These considerations sp ecially affects (ii) and they are quite similar to the considerations stated about the basic pragmatic approach to belief.
First, there is no double implication in (ii): a consideration p may be useful for an agent S but it does not force the agent to adopt such consideration as a true one. In other words, it is not the case that (if to consider p to be true is useful for S, then S considers p to be true). Many examples of considerations that are useful if taken to be true but that cannot be adopted as they are not known to be true are present in the literature on beliefs 10 .
Second, in (ii) it may be not the agent who takes her truly consideration to be useful but another agent. This epistemic-pragmatic approach states that 'to consider p to be true is useful for S' but that does not mean that 'S takes her consideration p taken to be true to be useful for herself ' . That is, it may be other agent S 2 who considers the proposition p whose bearer is Sand that is taken to be true by S -to be useful for S 11 .
Third, the content of the proposition p considered to be true by S does not need to be useful for S but just the attitude of considering p to be true. In other words, it is not the case that (if S considers p to be true then p is useful for S). Much less a useful proposition needs to form a consideration taken to be true -it is not the case that (if p is useful for S, then S considers p to be true) 12 .
This epistemic-pragmatic treatment of belief est ablishes that an agent S believes p if and only if S considers p to be true and that if S considers p to be true, then such consideration p is useful for S. The application of this account to pluralistic ignorance phenomena according to the definition in terms of beliefs, act ions and evidence provided by Bjerring et al. (2014) results as follows, 'Pluralistic ignorance' refers to a situation where the individual members of a group (i) all privately consider true some proposition p → then, it is useful for all to privately consider p to be true; (ii) all consider true that everyone else considers true ¬p → then, it is useful for all to consider true that everyone else considers true ¬p, consideration that is useful for them; (iii) all act contrary to their private consideration that p is true (i.e. act as if they consider true ¬p), consideration that is taken to be useful for their bearers; and where (iv) all take the actions of the others as strong evidence for their private considerations about the truth of p → every agent takes the actions of the others as strong evidence for her considerations taken to be true, which is taken to be useful for the agent, that the rest of agents consider true ¬p, consideration that is taken to be useful for the rest (based on Bjerring et al, 2014Bjerring et al, , p. 2458).
In the classroom case, first in (i) every student privately considers true that she does not understand the lesson and this consideration is useful for her -i.e. it allows the student to further solve her doubts.
Second, in (ii) every student considers true that everyone else considers true that they understand the lesson. In other words, it is useful for every student to consider true (that everyone else considers true that they understand the lesson) -e.g. in order to fit into the group -and, furthermore, it is thought that everyone else considers true that they understand the lesson, consideration that is taken to be useful for them -e.g. if any agent really understands the lesson and she is conscious, then she may dedicate more time to study other issues.
Third, in (iii) every student acts contrary to her truly consideration that she does not understand the lesson -she acts as if she considers true that she understands the lesson -but at the same time the initial consideration taken to be true -i.e. she considers true that she does not understand the lesson -is also taken to be useful for her. And finally, in (iv) every student takes the act ions of the others as strong evidence for her private consideration taken to be true, which is also taken to be useful for her, that the rest of agents consider true that they understand the lesson, consideration that is also taken to be useful for them.
The application of this epistemic-pragmatic approach of belief to pluralistic ignorance phenomena according to Bjerring et al. (2014) definition is the hardest one of the three approaches here analyzed. Once again, the most difficult step is (iii): agents act contrary to their initial considerations taken to be true and useful, something that contradicts both the epistemic and the pragmatic approaches. Agents act following 266 the falsity and such act ion presents disadvantages. But once again, the problem may be solved appealing to step (ii): it is useful for every agent to consider true that the rest of agents consider true ¬p, consideration that at the same time is taken to be useful for the rest of agents. In the classroom case, every student has two faced considerations taken to be true: on the one hand, she considers true that she does not understand the lesson, and on the other hand, she considers true that the rest of students understand the lesson. Both considerations taken to be true present different useful results and the student finally acts following the second consideration taken to be true rather than the first one. And once again, a self-other difference cause may be argued to explain this behavior (Brennnan et al., 2013;Bicchieri, 2006). Every particular agent prioritizes what she thinks the rest of agents consider to be true -and so useful -rather than what she initially considered to be trueand so useful -for herself, and she finally develops her act ions according to what she considers to be more useful -i.e. to be socially accepted -even when she personally considers it to be false. Such priority is supported by the evidence given by the others' act ions (iv).
This epistemic-pragmatic approach of belief may deal with pluralistic ignorance phenomena but it is harder to apply. At the same time, it still needs the concept of true to run and succeed. Different pragmatic outcomes confront and the final act ion is explained according to social beliefs -i.e. beliefs about the rest beliefs or truly considerations about what the rest consider to be true. That supposes that every agent finally acts contrary to what she considers to be true -i.e. every agent finally acts contrary to her initial belief -because she considers such act ion to be more useful. Even if fuzzier, this epistemic-pragmatic approach offers a more accurate theoretical description of pluralistic ignorance.

Conclusions
Pluralistic ignorance offers some clues on the traditional debate about the nature of belief and its epistemic and pragmatic features. At the same time, the philosophical debate on belief also provides some interesting keys to better charact erize and analyze pluralistic ignorance phenomena. Here I deal the analysis of the very phenomenon once emerged, but I also showed that the study of pluralistic ignorance emergence may offer some interesting clues on the debate about the reasons for belief -e.g. if there can be non-evidential reasons for belief emergence -and such debate may also improve the study of pluralistic ignorance emergence.
In our case, I show that a pure and basic epistemic treatment of belief -believers consider their beliefs to be true -easily applies to pluralistic ignorance phenomena if we make a difference between belief and the final act ion motivated not only by epistemic reasons but also by another pragmatic reasons.
I also show that a pure and basic pragmatic charact erization of belief -beliefs are useful for believers -may apply to pluralistic ignorance phenomena without the necessity of the concept of truth. We need to consider that different personal beliefs -about ourselves and about the rest of agents -may supply different useful outcomes and final act ion depends on the useful outcomes supplied by the beliefs about the rest of agents rather than on the beliefs about ourselves.
Finally, I defend that a mixed epistemic-pragmatic characterization of belief allows for a more accurate analysis of pluralistic ignorance. We need to accept that agents may personally consider something to be true and useful but that they may act contrary to this consideration because another consideration about what the rest of people consider to be true is more useful -although the agent does not consider true what the rest of people consider true. In other words, if we assume a self-other difference, then this mixed epistemic-pragmatic approach offers a more precise charact erization of pluralistic ignorance.