When I worked in New York my colleagues and I would ride the Bronx 12 bus from the university to the subway station, and on one of those trips we were interrupted by a question from one of the people on the bus: “What's that word you just used?” – “Which one, “epistemic”?” – “Yeah, what does that mean?” It is that bus passenger's question that I want to try to answer in this paper. At the time I said that “epistemic” meant “related to knowledge.” That answer strikes me as more or less right – but I have come to think that it is at least partly wrong, and that this simple answer masks some complexities that deserve scrutiny. My aim here is to provide a more precise and complete account of the meaning of “epistemic” – and to discuss some related metaphilosophical issues, including the prospects for eliminating “epistemic” from our philosophical lexicon.
1. METHODOLOGY
My question is: what does “epistemic” mean? This form of words – “what does “epistemic” mean?” – is ambiguous, and I have a particular disambiguation in mind. There is a way of disambiguating this form of words, one that owes its possibility to the philosophical “linguistic turn,” on which it indicates a question that would be answered by a metaphysical or conceptual analysis of the epistemic. Consider the question: what does “right” mean? You might hear this as indicating a question that would be answered by a metaphysical or conceptual analysis of the right; answering this question would require an inquiry in normative ethics. Or consider the question: what does “intentional action” mean? You might hear this as indicating a question that would be answered by a metaphysical or conceptual analysis of intentional action; answering this question would require an inquiry in action theory or moral psychology. My question is the bus passenger's question: I just want to know what the word “epistemic” means.
A natural place to begin is with a dictionary. The latest edition of the Oxford English Dictionary defines “epistemic” with: “Of or relating to knowledge, or to its extent, linguistic expression, or degree of validation.” This jibes with the initial answer I suggested above – “relating to knowledge.”
It seems to me that “epistemic” is a piece of academic jargon. The OED notes that “epistemic” is a term of philosophy and linguistics, and the majority of the uses that it lists are from academic journals and books. In philosophy, “epistemic” seems to be a relatively recent addition to our lexicon. (The uses offered by the OED are from the 20th century.) In both of these respects, “epistemic” is unlike “moral,” “aesthetic,” and “prudential,” which are both more familiar in ordinary language and older in academic usage. In my view, this puts some pressure on philosophers to account for the meaning of “epistemic” – pressure that is not there in the case of terms like “moral,” “aesthetic,” and “prudential.” The present inquiry is motivated by a concern with articulating the meaning of “epistemic” in as much as it is a piece of academic jargon – and in particular, in as much as it is a piece of philosophical jargon.
Another natural place to look here is to etymology. “Epistemic” derives from the Ancient Greek word ἐπιστήμη, which the Liddel-Scott-Jones Greek-English Lexicon defines with: “acquaintance with a matter, understanding, skill,” and “generally, knowledge.” This takes us not much further than the OED definition, although LSJ goes on to note that ἐπιστήμη sometimes referred specifically to professional skill or expertise and sometimes to scientific knowledge.
It is well-known that Aristotle used ἐπιστήμη to refer specifically to knowledge of causes (i.e. understanding), contrasting this with mere propositional knowledge, and contrasting mere ἐπιστήμη with knowledge of fundamental causes, which he called σοφία. Might this shed light on the meaning of “epistemic”? Perhaps, although the relatively recent appearance of “epistemic” in English may problematize any attempt to connect the contemporary English word with Aristotle's philosophy.
My strategy here will be to focus neither on dictionary definitions nor on etymology, but instead to focus on contemporary uses of “epistemic” in philosophy. Even if how “epistemic” is used in contemporary philosophy does not entirely determine its meaning, this is at least the right place to begin in attempting to explain its meaning.
2. TWO USES OF “EPISTEMIC”
I shall describe two different ways in which “epistemic” is used in contemporary philosophy. It is sometimes used to mean (roughly) “of or relating to knowledge” (§2.1) – these uses jibe with our initial answer (above) to the question of the meaning of “epistemic.” However, “epistemic” is sometimes used to mean (roughly) “of or relating to belief” (§2.2) – these uses do not jibe with that initial answer.
2.1 Uses that jibe with the initial answer
In a wide range of familiar cases, “epistemic” is used to indicate a teleological connection to knowledge – or to “epistemic goods” more broadly construed to include true belief, understanding, and the like (in as much as these are not species of knowledge). Consider:
[I]f … a practice has epistemic value, it would be one that somehow furthers the goal of acquiring true beliefs and avoiding false ones. (Lackey Reference Lackey2008: 226)
An epistemic virtue is a personal quality conducive to the discovery of truth, the avoidance of error, or some other intellectually valuable goal. (Fairweather and Flanagan Reference Fairweather and Flanagan2014, dust jacket)
Let's call these teleological uses of “epistemic.” When used in this way, “epistemic” contrasts with “moral,” “aesthetic,” and “prudential.” Think here of the familiar move of contrasting “epistemic reasons for belief” or “epistemic obligations to believe” with “prudential reasons for belief” or “prudential obligations to believe,” as when considering evidentially unsupported false beliefs that are all-things-considered best.
In addition to the above uses of “epistemic,” other common teleological uses include “epistemic goods,” “epistemic goals,” and “epistemic evaluation.”
There are other cases in which “epistemic” is used to mean (roughly) “of or relating to knowledge,” but where no teleological connection to knowledge is suggested. Consider the idea of an “epistemic norm of assertion,” e.g. that you ought assert only what you know. This is an “epistemic norm” in the sense that it is a norm that invokes knowledge – it is like the following “epistemic norm of dance”: dance only the dance you know. Or consider the notions of “epistemic possibility” and “epistemic necessity” – the “epistemic modalities.” Here there is a connection to knowledge, but not a teleological connection.
One way we can tell that “epistemic” is being used to mean “of or relating to knowledge” is by looking at what other words can be substituted for it without loss of meaning. In particular cases, if someone switches, without comment, between “epistemic” and some other word or words, we can see how she intends her uses of “epistemic.” So, for example, people who are using “epistemic” teleologically often switch, without comment, between “epistemic” and “justificatory” or “evidential.”
In most cases, the semantic contribution of a word is determined by the context in which it is uttered. And in the case of most (if not all) adjectives, their meaning on a particular occasion of use depends on the noun phrase with which they are conjoined. This is a familiar phenomenon. As Grice argued (Reference Grice1988: 193), many adjectives, although they are univocal, have a plurality of “meanings-in-context,” depending on the expressions in which they appear. His example is “French,” which means “of or relating to France.” Consider a French person, a French city, a French verb, a French restaurant, and a French cheese. The relationship that each of these bear to France is different – in each case, there is a different way in which the relevant thing is “of or relating to France.” The same, we can safely conclude, when it comes to “epistemic.” For among the various cases in which “epistemic” is used to mean “of or relating to knowledge,” there are a variety of different ways in which the thing that is said to be epistemic is “of or relating to knowledge.” This can be confusing to the uninitiated, but in this respect “epistemic” does not differ from other adjectives.
2.2 Uses that don't jibe with the initial answer
In other cases, “epistemic” is used to indicate a connection to belief – or, more broadly, to cognition, representation, or intellectual activity. Consider:
Sextus directs us to understand every statement he makes … as a record of his experience … telling us how things appear to him[.] If he means ‘appear’ in its non-epistemic sense, [he] implies that the sceptic's assent is restricted to experiental reports like ‘It feels warm to me here’, ‘This arguments strikes me as persuasive’. […] If, on the other hand, ‘appears’ carries its epistemic sense, to talk about how things appear is simply to talk in a non-dogmatic way about how things are in the world. (Barnes Reference Barnes1982: 98)
I shall argue here that friendship involves not just affective or motivational partiality but epistemic partiality. Friendship places demands not just on our feelings or our motivations, but on our beliefs and our methods of forming beliefs. (Stroud Reference Stroud2006: 499)
Let's call these energetic uses of “epistemic.” This way of using “epistemic” is less familiar to philosophers, and in particular epistemologists are disposed to see it as a kind of mis-use or loose speaking. It is more common in discussions of the history of philosophy and in psychology. When used energetically, “epistemic” contrasts with “affective,” “volitional,” and “practical.”
We can tell that “epistemic” is being used energetically by looking at what other words can be substituted for it without loss of meaning – so, for example, people who are using “epistemic” energetically sometimes switch, without comment, between “epistemic” and “cognitive” or “intellectual.”
In addition to the above uses of “epistemic,” other energetic uses include “epistemic bias,” “epistemic perception,” “epistemic agency,” “epistemic responsibility,” and “epistemic authority.” In these cases, “epistemic” is used to indicate a contrast with non-doxastic, non-cognitive, or non-intellectual alternatives: an “epistemic bias” is one that is manifested in a person's beliefs, rather than in her motivations; “epistemic agency” is agency manifested in what a person believes, rather than in how she acts; “epistemic authority” is authority in matters of belief, rather than in matters of action.
2.3 Ambiguities
“Epistemic” is sometimes used to mean “of or relating to knowledge,” (§2.1) but it is sometimes used to mean “of or relating to belief” (§2.2). On its own, this is no big deal. However, it seems to me that some uses of “epistemic” in contemporary philosophy are ambiguous as between these two uses – and are therefore sources of potential confusion. Here are two examples:
-
• “Epistemic virtue” – does this pick out (i) personal qualities conducive to knowledge or (ii) intellectual personal qualities that are virtues (including, conceivably, personal qualities not conducive to knowledge)?
-
• “Epistemic reasons” – does this pick out (i) reasons relative to the goal of knowledge (including, conceivably, reasons to inquire) or (ii) reasons for belief (excluding reasons to do anything)?Footnote 1
In many cases, it is clear what a particular person means when she speaks of “epistemic virtue” or “epistemic reasons.” But in many other cases, this is not clear. In these other cases, the fact that “epistemic” has two different uses makes trouble for us – our philosophical jargon, in these cases, fails to provide the rigor and precision we need.
3. THE EPISTEMIC AS A DISJUNCTIVE CONCEPT
Given the way “epistemic” is used in contemporary philosophy (§2), and in as much as there is a single concept of the epistemic in philosophy, it is a disjunctive concept. “Epistemic” means (roughly) “of or relating to knowledge or belief.” (In connection with this, consider the fact that the previous edition of the Oxford English Dictionary defined “epistemic” with: “of or relating to knowledge or degree of acceptance.”) If this is right, why have philosophers come to use this disjunctive concept? And, more important, is our use of this concept justified?
3.1 The epistemic and the “aim of belief”
It seems to me that the explanation of our use of the disjunctive concept of the epistemic begins with the fact that there is something right in the slogan that “belief aims at truth.” There is an essential or constitutive connection between belief and truth – by which I mean a connection between belief and truth that flows from the essential nature of belief, one that (at least in part) makes belief the attitude that it is. Because of this, although the concept of the epistemic is disjunctive, the two relevant disjuncts seem to be intimately related to one another.
What do we mean when we say that “belief aims at truth”? It is evident that philosophers mean a great many different things, although not always incompatible things, when they say this. (Which, if any, of these things are true will not matter in what follows.) Some mean to identify belief's correctness condition, i.e. to articulate the fact that beliefs are correct if and only if true. Others mean to articulate the idea that belief is teleological, e.g. that beliefs are performances in which the believer tries to arrive at the truth. Others mean that being true is the function of belief; it is what beliefs are designed or evolved to do. And others mean that truth is a norm of belief, e.g. that beliefs ought to be true, or that untrue beliefs are bad. In any event, how might this explain why philosophers have come to use the disjunctive concept of the epistemic?
The “aim” of the heart is to circulate blood – in this case, “aim” is to be understood in functional terms. In medicine we have the notion of the cardiac, which (roughly) covers things having to do with the heart, and we have the notion of the circularity, which (roughly) covers things having do with circulating blood. But we also have the disjunctive concept of the cardiovascular, which (roughly) covers both the cardiac and the circularity. This disjunctive concept covers, in other words, things having to do with the heart and its “aim.” This makes sense in medicine, moreover, because our medical interest is exclusively in health, and in the case of cardiology, in heart health – i.e. in hearts that fulfill their “aim.” Thus medicine, and cardiology in particular, studies hearts that function well, along with diseases of the heart, i.e. ways in which hearts can malfunction, and other factors that can interfere with the proper functioning of the heart.
Here, then, is a proposal. Just as in medicine our interest is exclusively in hearts that fulfill their “aim,” in philosophy our interest is exclusively in beliefs that fulfill their “aim.” Just as medicine's use of the disjunctive concept of the cardiovascular makes sense in light of medicine's exclusive interest in healthy hearts (i.e. hearts that fulfill their “aim”), philosophy's use of the disjunctive concept of the epistemic makes sense in light of philosophy's exclusive interest in true beliefs (i.e. beliefs that fulfill their “aim”). Just as the medical concept of the cardiovascular covers both things having to do with the heart and things having do with the “aim” of the heart, the philosophical concept of the epistemic covers both things having do with belief and things having to do with the “aim” of belief.
This account must be supplemented by an explanation of the relevant connection between the “aim of belief” and knowledge – since “epistemic,” on the present proposal, means “of or relating to knowledge or belief.” There are several ways this could go. You might argue that knowledge, and not truth, is the “aim of belief.” Or you might argue that knowledge is a species of true belief that is intimately connected to its “aim” – you might argue, for example, that knowledge is true belief that is held because it is true. Finally, you might argue that, if belief “aims” at truth, then belief also “aims” at knowledge – just as genuinely trying to hit an archery target requires trying to hit the target skillfully (and not by luck), so the argument might go, genuinely trying to form a true belief requires trying to form a true belief skillfully (and not by luck). In any of these cases, an exclusive interest in beliefs that fulfill their “aim” would vindicate the introduction of the disjunctive concept of the epistemic.
3.2 A worry about the disjunctive concept
My proposed account of our use of the disjunctive concept of the epistemic in philosophy (§3.1) assumed that in philosophy we are exclusively interested in beliefs that fulfill their “aim.” Is this right?
It seems to me that it is not. Philosophy seeks a synoptic understanding of things, including a synoptic understanding of what matters. For this reason, philosophers are interested in everything that matters. To focus exclusively on beliefs that fulfill their “aim,” however, would be to ignore or bracket or dismiss the other ways in which beliefs can be valuable. Consider, again (§2), the case of evidentially unsupported false beliefs that are all-things-considered best. It would be most unphilosophical to ignore or bracket or dismiss the all-things-considered value of such beliefs, on the grounds that they fail to fulfill their “aim.” The problem with being exclusively interested in beliefs that fulfill their “aim” is that the fulfillment of “aims,” however this is understood, is not all that matters. Indeed, it is not even clear that it matters at all. Something may fulfill its “aim,” but it may still be an open question whether anything good has occurred. The existence of a belief that fulfills its “aim” may or may not be good for the believer, or good for anyone else, or good without qualification.
Medicine, by contrast, does not seek a synoptic understanding of things, nor specifically a synoptic understanding of what matters. Its narrow focus on health – and, in particular cases, on organs that fulfill their “aim” – is entirely appropriate.
It is possible that our use of the disjunctive concept of the epistemic, in philosophy, was motivated by the assumption that our exclusive interest is beliefs that fulfill their “aim.” (There is a Platonic tradition that sees philosophy as analogous to medicine, with the former concerned with mental health and the latter concerned with bodily health.) But philosophers should not be exclusively interested in beliefs that fulfill their “aim.” Therefore, such an interest cannot vindicate our use of the disjunctive concept of the epistemic.
Indeed, you might worry that our use of the disjunctive concept implicates an exclusive interest in beliefs that fulfill their “aim.” Let me explain. The “aim” of pictures is to be accurate, veridical, or true to their subjects – in this case, “aim” is to be understood in terms of correctness. Now we have the notion of the pictorial, which (roughly) covers things having to do with pictures, and we have the notion of the accurate, which (roughly) covers things having to do with accuracy. And we could introduce a disjunctive concept to (roughly) cover both the pictorial and the accurate. However, it is easy to imagine that the introduction of such a concept would be a blunder in philosophical aesthetics. It would be a blunder if, for example, its introduction were motivated by an exclusive interest in accurate pictures, i.e. in pictures that fulfill their “aim.” For this would be to ignore or bracket or dismiss the many other ways in which pictures can be valuable – for example, those flowing from their formal or compositional or historical properties. There may be a discipline that is appropriately concerned exclusively with accurate pictures. But this narrow focus would not be appropriate in philosophical aesthetics. And you might well worry that our use of “epistemic” implicates a blunder analogous to this blunder in philosophical aesthetics.
3.3 The nature and scope of epistemology
You might concede that philosophy should not be exclusively interested in beliefs that fulfill their “aim,” but go on to argue that epistemology should be. We could shore up this idea by defining epistemology as the discipline essentially concerned with the epistemic – where the epistemic (roughly) covers things having to do with belief and its “aim.” Let's grant these claims about epistemology. I'll argue that this has consequences for how we think about epistemology's place in philosophy.
The basic normative question in philosophy is: how should I live? Somewhat artificially, we can divide this into practical and theoretical sub-questions:
-
• What should I do?
-
• What should I believe?Footnote 2
Philosophical ethics attempts to answer the first of these questions. And a natural thought is that epistemology attempts to answer the second. But if epistemology is exclusively interested in beliefs that fulfill their “aim,” then this cannot be right. If epistemology is exclusively interested in beliefs that fulfill their “aim,” then epistemology does not attempt to answer the question “What should I believe?” If epistemology is exclusively interested in beliefs that fulfill their “aim,” then epistemology is analogous to medicine – which does not attempt to answer the question “What should I do?” In philosophical ethics we plausibly seek input from medicine – it may be useful to know, for example, which sorts of actions are conducive to health. Perhaps, then, when answering the question “What should I believe?” we should similarly seek input from epistemology – it may be useful to know, for example, which sorts of beliefs fulfill their “aim.” But just as it would be a mistake to confuse medicine with philosophical ethics, it would be a mistake to confuse epistemology with the philosophical discipline that attempts to answer the question “What should I believe?”
You might object that epistemology attempts to answer questions about our epistemic obligations to believe. That is fair enough; but the philosophical question was: what should I believe? There is no qualification or relativization to “epistemic obligation.” Compare: medicine attempts to answer questions about our medical obligations, in the sense that medicine will tell us what sorts of actions are conducive to health. But this is not the same as trying to answer the philosophical question: what should I do?
You might object that there are no non-epistemic obligations to believe.Footnote 3 Questions about our obligations to believe, then, are just questions about the epistemic. And so epistemology, on the present proposal, does attempt to answer the philosophical question: what should I believe? However, this ignores the fact that “epistemic value” is not all that matters. (Indeed, it is not even clear that it matters at all.) Consider, again (§2), the case of evidentially unsupported false beliefs that are all-things-considered best. Let us grant that we ought not have such beliefs. Surely, however, that such beliefs are all-things-considered best is relevant to the basic normative question in philosophy: how should I live? It seems, then, that we were wrong to articulate our theoretical sub-question as a question about obligations to believe; we should have articulated this as (something like): what beliefs would be best? And it is clear that epistemology, on the present proposal, does not attempt to answer this question.
So: philosophical ethics attempts to answer the question “What should I do?” It may seek input from medicine, which attempts to answer questions about health. Epistemology, on the present proposal, is analogous to medicine: it may provide input to our attempts to answer the question “What should I believe?” But it does not itself attempt to answer that question.
A natural question then: what discipline does seek to answer the question “What should I believe?”? We might re-conceive of epistemology, rejecting the present proposal, on which epistemology is the discipline essentially concerned with the epistemic, such that epistemology does attempt to answer this question. Alternatively, we might hypothesize a discipline distinct from epistemology, analogous to philosophical ethics, that essentially (although perhaps among other things) attempts to answer this question.Footnote 4
4. DOING WITHOUT “EPISTEMIC”
I have raised two worries about the use of “epistemic” in contemporary philosophy. The first is that there are ambiguous uses of “epistemic” (§2.3). The second is that our use of the disjunctive concept of the epistemic might implicate a mistaken exclusive interest in beliefs that fulfill their “aim” (§3.2). Given these worries, you might wonder whether we could do without “epistemic” in philosophy. The good news is: we can.Footnote 5
There are two reasons for this. The first is that many uses of “epistemic” are superfluous. They can be stricken without any effect on the meaning of the discussions in which they occur. The second is that many uses of “epistemic” can be replaced with alternative jargon, usually with a gain in clarity and precision. In these cases, it can often be made clear whether we mean “of or relating to knowledge” or “of or relating to belief” (cf. §2). When we mean “of or relating to knowledge,” we have a number of more precise options, including “justificatory,” “evidential,” and “alethic.”Footnote 6 When we mean “of or relating to belief,” we have a number of more precise options, including “cognitive” or “intellectual,” as well as “doxastic.”
For example, consider the following passage:
To describe what it is rational for S to believe from a purely epistemic point of view is to describe what S himself from his own perspective on reflection would believe to be an effective means to his epistemic goal of now believing truths and now not believing falsehoods. (Foley Reference Foley1987: 125)
It seems to me that the first occurrence of “epistemic” in this passage can be replaced with “alethic” without loss of meaning – indeed, it makes what is being said more obviously right. (The “epistemic” point of view is here being contrasted with points of view characterized in terms of goals other than true belief.) And it seems to me that the second occurrence of “epistemic” can be stricken without changing the meaning of what is being said. (True belief is just, by definition, the “epistemic” goal, so “the epistemic goal of true belief” is redundant.) In neither case is “epistemic” needed to say what is being said.
Thus, I propose an exercise along these lines. Take any contemporary philosophical essay, and consider each use of “epistemic.” I submit that each is either superfluous – in which case to be stricken – or replaceable with alternative jargon – in which case to be so replaced. This will give you a picture of what philosophy without “epistemic” looks like.Footnote 7
