Edmund Gettier’s paper “Is Justified True Belief Knowledge?” first appeared in an issue of Analysis (Vol. 23, No. 6), dated June of 1963, and although it’s tempting (and common) to wax hyperbolic when discussing the paper’s importance and influence, it is fair to say that its impact on contemporary philosophy has been substantial and wide-ranging. Epistemology has benefited from 50 years of sincere and rigorous discussion of issues arising from the paper, and Gettier’s conclusion that knowledge is not justified true belief is sometimes offered as an example of the reality of philosophical progress. (The idea that one short paper could be so important continues to fascinate philosophy students.) However, what can be called the Gettier problem has little to do with the text of the famous paper itself. The importance of the Gettier problem does not depend on the attribution of the tripartite theory of knowledge to Plato, Chisholm, and Ayer, nor on the psychological plausibility of the particular examples that Gettier describes, nor on the novelty of those cases (cf. Russell 1948/2000, p. 140). The history of the Gettier problem does provide us with a case of study in conceptual analysis, a paradigm of that distinctively twentieth-century form of philosophical inquiry, but it also includes epistemology’s twenty-first century movement away from conceptual analysis, and a suggestion that the problems of philosophy are not, after all, merely conceptual.

The papers in this volume were presented in June of 2014 at a conference at the University of Edinburgh that celebrated the 50th anniversary of the publication of Gettier’s paper. In this introduction I shall briefly trace the history of the Gettier problem, from its basic form to its status in contemporary epistemology, and describe some of its apparent future directions.

1 Fifty years of research

You might think that knowledge is justified true belief, i.e. that S knows that p if and only if it is true that p, and S believes that p, and S is justified in believing that p. This elegant account is known as the tripartite theory of knowledge, and it captures several plausible ideas about knowledge, including that falsehoods cannot be known and that unreasonable or irrational beliefs, even if true, do not amount to knowledge. However, it is subject to a family of counterexamples, known as Gettier cases. We can divide these into three types. First, there are cases in which someone reasonably infers a true conclusion from a false premise that she believes with justification. The cases from Gettier’s paper (1963, pp. 122–123) are of this type. Second, there are cases in which someone believes some true proposition, and is justified in so believing, but in which her belief is caused by something other than the truth of that proposition. Roderick Chisholm’s (1966) case of the sheep in the field (p. 23n) is of this type. Third, there are cases in which someone believes some true proposition on some basis, and is justified in so believing, but in which an unusual or abnormal environmental condition makes it such that she would easily have believed something false on the same (or a similar) basis. Carl Ginet’s (1988) fake barn case (p. 106) is of this type.

In these cases, it seems that the person who believes that p does not know that p, despite its being the case that it is true that p, that she believes that p, and that she is justified in believing that p. Justified true belief is therefore insufficient for knowledge, and the tripartite theory of knowledge is false. The premise that, in Gettier cases, the person who believes that p does not know that p, has been widely endorsed by epistemologists. The basic form of the Gettier problem, therefore, has consisted of a challenge for theorists of knowledge: amend or replace the tripartite theory of knowledge with a theory of knowledge that is immune to (at least this kind of) counterexample.

Several decades (at least) of post-Gettier research were addressed explicitly to solving this problem. (Shope 1983 provides the definitive account of this period.) Some (Clark 1963) argued that knowledge cannot be derived from a false premise; others (Lehrer and Paxson 1969) argued that knowledge requires indefeasible justification; others (Goldman 1967) argued that knowledge must be caused by the truth of the proposition known; others (Stine 1976; Goldman 1976; Dretske 1981, Chapter 4) argued that knowledge requires the elimination of relevant alternatives; others (Nozick 1981, Chap. 3; Sosa 1999; Williamson 2000) argued that knowledge requires sensitivity (that you would not believe that p, were it not true that p) or safety (that you would not easily believe falsely that p). Externalist theories of knowledge flourished during this period—where these are (roughly) those that allow necessary conditions on knowledge (apart from the truth condition) the obtaining of which may be (in some sense) inaccessible to the knower. Early contextualist discussions of the Gettier problem (Cohen 1988) did not develop further, and by contrast with the problem of philosophical skepticism, there has been little interest in appealing to claims about knowledge attributions in discussions of the Gettier problem. For the most part, epistemologists agreed with Gettier’s conclusion, and took it to represent an important insight into the nature of knowledge.

Towards the end of the twentieth century, a different sort of discussion of the Gettier problem emerged, focusing on questions about the source and solubility of the problem. Some argued that the problem is, in some sense, unsolvable (Zagzebski 1994), which assumption, at least on some suitable articulation, is a central premise of “knowledge-first epistemology” (Williamson 2000), and some epistemologists now offer accounts of knowledge that appeal to the notion of knowledge (see Lisa Miracchi’s contribution to this volume). Many epistemologists agree that the Gettier problem has something to do with luck—it is argued that all Gettier cases have a “double luck” structure (Zagzebski 1966, pp. 288–289; cf. Zagzebski 1994) and that they reveal the centrality of “epistemic luck” to the nature of knowledge (Pritchard 2005). Given the difficulty of providing a straightforward solution to the Gettier problem, it is tempting to think that it reveals some previously undetected complexity in the nature of knowledge. However, many now see the Gettier problem as arising as a result of a few simple and seemingly innocuous assumptions about knowledge (see Duncan Pritchard and Timothy Williamson’s contributions to this volume).

Early in our century, a still different (although related) strain of research came to prominence. Meta-philosophical research on the nature and evidential status of philosophical intuitions (cf. DePaul and Ramsey 1998; Sosa 2011) often treated the Gettier problem as a case study of intuition-based philosophical argument, such that the rejection of the tripartite theory of knowledge is understood as based on the “Gettier intuition” i.e. an intuition that, in Gettier cases, the person who believes that p does not know that p. Some self-styled “experimental philosophers” suggested that philosophical intuitions are generally unreliable (cf. Knobe and Nicols 2008), and while some alleged that the having of the Gettier intuition more common in non-Asian men (Weinberg et al. 2001; Nicols et al. 2003; Buckwalter and Stich forthcoming), experiments suggesting this could not be replicated (Nagel 2012). More recently, some have argued against the standard meta-philosophical treatment of the Gettier problem, on the grounds that Gettier’s conclusion is not based on intuition, but rather on a principled argument (Williamson 2008, Chap. 6).

Although there has been a trend towards metaphilosophical and methodological approaches to the Gettier problem, it is worth contrasting this trend with the work of those contemporary philosophers who defend solutions to the Gettier problem. These philosophers generally appeal to some form of virtue epistemology, on which knowledge is identified (roughly) with true belief that manifests intellectual virtue (Sosa 1991, 2007; Greco 1999, 2010; Turri 2011; Pritchard 2012), thus excluding the true beliefs possessed in Gettier cases. The emergence and development of these virtue-theoretic accounts of knowledge, whose appeal outstrips their credentials as solutions to the Gettier problem, is one of the significant legacies of Gettier’s paper.

2 The future of the Gettier problem

Consider the influential fake barn case (cf. Sect. 1): Henry’s environment is full of fake barns, which he cannot tell from real barns, but he happens now to be standing before a real barn, which prompts him to believe that there is a barn before him. His belief is justified and true—does it amount to knowledge? It has seemed to many that he does not know that there is a barn before him, but this assumption is now controversial. You might argue that Henry knows by appeal to an independently appealing theory of knowledge, whose theoretical virtues outweigh the cost of violating the aforementioned “Gettier intuition” (Weatherson 2003; Sosa 2010; and see Eric Olsson’s contribution to this volume). The assumption that Henry does not know was once a fixed point, a datum that epistemologists were expected to vindicate; it is now a substantial philosophical claim, and a matter of epistemological dispute.

Recall, then, the basic form of the Gettier problem (Sect. 1). After 50 years of research, including much (ongoing) metaphilosophical and methodological reflection, epistemologists today face a different problem. Our question now is: in Gettier cases, or in the different types of Gettier case, does the person who believes that p know that p? And our challenge is to defend an answer to this question. This is the Gettier problem as it exists today.Footnote 1

The Gettier problem has also taken on new forms outside of the theory of (propositional) knowledge. Epistemologists have explored the implications of Gettier-type cases for understanding and practical knowledge (on which see Amber Riaz and Yuri Cath’s contributions to this volume). Others have sought to situate the Gettier problem in a broader context, e.g. as one aspect of a general theory of intentional action (Sosa forthcoming; and see Christian Piller’s contribution to this volume) or a general theory of luck (Pritchard 2005; Pritchard and Whittington forthcoming). These are the directions in which the Gettier problem is moving.

It is presently fashionable to denigrate early research on the Gettier problem, either as an absurd attempt at something foolish to begin with, or (as though spilled ink were a species of spilled blood) as a tragic loss of philosophical effort. The elaborate analyses of knowledge offered in the late 1960s are a Baroque offense to our minimalist sensibilities. For aiding and abetting the Gettier problem, the theory of knowledge itself is sometimes held in disrepute, as against supposedly reputable (and at least de rigueur) topics as the theory of understanding, intellectual virtue, testimony, and epistemic normativity.

The importance and legitimacy of inquiry about particular philosophical topics, it seems to me, is a matter of taste—with all the relativistic implications of that phrase. However, our exasperation with what I called the basic form of the Gettier problem (Sect. 1) threatens to obscure some interesting things. First, there is widespread agreement that this problem arises in virtue of some philosophical insight into the nature of knowledge that is not articulated in accounts of knowledge proposed before Gettier’s paper. That this insight has proven difficult to articulate is surprising only given the assumption that all philosophical insights can be easily articulated. Second, does the theory of understanding (for example) compare favorably to the theory of knowledge, on the assumption that the Gettier problem is unsolvable? Consider epistemologists’ inability, as of yet, to articulate the sense in which understanding requires a “grasp” of explanatory structure. Third, and finally, anyone who would lament early research on the Gettier problem should take a closer look at that work—and contrast it with the best that contemporary epistemology has to offer—and note well the extent to which certain virtues are manifested there: intellectual rigor, a commitment to precision and clarity, and imaginativeness, as revealed in the creative articulation of analyses and counterexamples. This research was a high-water mark for these virtues in analytic philosophy, and can serve as an example of conceptual analysis in the best tradition, among the other legacies of the Gettier problem that we presently enjoy.Footnote 2