Refurbishing Epistemology

Dominique Kuenzle

• Publisher: De Gruyter
• Year: 2017
• ISBN:
  9783110524659
• Edition: 1
• Topic: Philosophy
• Subtopic: Epistemology in Philosophy

Even though important developments within 20th and 21st century philosophy have widened the scope of epistemology, this has not yet resulted in a systematic meta-epistemological debate about epistemology's aims, methods, and criteria of success. Ideas such as the methodology of reflective equilibrium, the proposal to "naturalize" epistemology, constructivist impulses fuelling the "sociology of scientific knowledge", pragmatist calls for taking into account the practical point of epistemic evaluations, as well as feminist criticism of the abstract and individualist assumptions built into traditional epistemology are widely discussed, but they have not typically resulted in the call for, let alone the construction of, a suitable meta-epistemological framework.
This book motivates and elaborates such a new meta-epistemology. It provides a pragmatist, social and functionalist account of epistemic states that offers the conceptual space for revised or even replaced epistemic concepts. This is what it means to "refurbish epistemology": The book assesses conceptual tools in relation to epistemology's functionally defined conceptual space, responsive to both intra-epistemic considerations and political and moral values.

1Towards a New Meta-Epistemology

There are easy and appealing ways to get introduced to topics within epistemology. For example, the movie The Matrix (1999) makes vivid the possibility that our bodies lie cultivated in nutrient solution, or that we are brains in vats; Christopher Nolan’s Inception (2010) refers directly to Descartes’ idea that we can never be entirely sure that we are not dreaming. These scenarios can then be exploited by sceptical arguments: “Well, if it is possible that you are a brain in a vat, then you do not know that you are walking through the park now. Consequently you do not know that you’re walking through the park now.” If all goes well, the resulting puzzlement will trigger interest for the philosophy to come.

Alternatively, we start by asking what it is for somebody to know something, suggesting that what we are looking for is an analysis or definition of the concept of knowledge. Following the steps of Plato’s Theaetetus, we hope to bring out the requirements that what we know must be true, that we take it to be true, and that we do not just correctly guess it to be true. We may even put forward the idea that to know that p is to have a justified true belief that p – which can then be shown to run into Gettier problems. Again, the ensuing puzzlement can motivate reflection on the definition of “knowledge”, along the lines of contemporary attempts to define or analyse key epistemic concepts.

A third option for engaging people with epistemology is provided by the epistemic regress problem: one may think that we only really know what we have theoretical reasons for. Do we need to know these reasons, and if so, do we have to be able to give reasons for these reasons? Are there any basic, “foundational” reasons that do not themselves stand in need of further reasons?

All three problems just discussed provide pre-theoretically, intuitively graspable, attractive puzzles that, if all goes well, lead to a kind of reflecting and theorising that tidily fits under the label of “contemporary epistemology”. Moreover, much of the research conducted within contemporary epistemology can be presented as dealing with just these initial problems: do we really know what we think we know? How to deal with sceptical arguments? How to analyse knowledge, and what is the structure of our body of knowledge? Start wondering about The Matrix and you’ll find yourself doing epistemology!

These are undoubtedly fruitful ways of engaging with epistemology, and they have led to a wealth of important research. But it seems that the sheer force with which the sceptical argument, the persisting problems arising from the analysis of knowledge, and questions about the structure of knowledge strike us as problematic has not helped giving urgency to the question of what, exactly, is the epistemological nature of these problems. Epistemology, after all, is an important philosophical research discipline; we would expect it to be equipped with its own explicit aims, methods, standards and criteria of adequacy. We would expect a rich, diverse meta-epistemological dialogue about proposals regarding the aims, methods and standards of epistemology.

This, however, is not really happening, and to say that epistemology is the philosophical discipline that deals with sceptical arguments, the analysis of knowledge and the structure of our body of knowledge is not more than a start. The common dictionary references to the “scope, origin and nature” of knowledge capture much of what is actually happening under the heading of “epistemology”, but they do so only in virtue of their excessive generality on the one hand, and thanks to the tendency to ride the definition backwards, treating as (related to) knowledge just everything epistemologists care about.

In order to develop a suitable conception of epistemology’s nature, goals, methods and criteria of adequacy, this opening chapter presents five developments from the second half of the twentieth century that have shaken epistemology’s foundations and continue to demand major meta-epistemological adjustments. In each case, I attempt to do two things: firstly, I will sketch the full, radical impact of these developments. This will, in each case, yield a context for an investigation of one or two key arguments. Secondly, the results of these investigations will be incorporated into the new conception of epistemology – a new “meta-epistemological framework”, as it will be called – to be developed in the following chapters.

1.1First Development: From Conceptual Analysis to Reflective Equilibrium

One common reaction to the vast literature on Gettier cases and the deep disagreement between internalist and externalist views of epistemic justification has been the diagnosis and criticism of the underlying method of conceptual analysis. This local scepticism with regards to the suitability of the method of conceptual analysis to certain epistemic problems is boosted by more general concerns relating to conceptual analysis as a philosophical methodology.

At the heart of many attacks on conceptual analysis lies Quine’s (1952) rejection of the concepts of analyticity and a priority. The epistemologist engaged in the project of illuminating the concept of knowledge (or epistemic concepts, or indeed concepts in general) has two ways to react to Quine: either by defending conceptual analysis against criticism, or by adjusting their methodology accordingly. One way to illuminate the concept of knowledge whilst avoiding the method of conceptual analysis is Carnapian explication, a methodology often combined with reflective equilibrium as a source for criteria of adequacy.

Conceptual Analysis

Conceptual analysis is a reflective activity whereby the content of some concept is clarified or expressed in some vocabulary, typically seen as conceptually more basic, or at least philosophically less problematic, than the target concept. Just how this is meant to happen depends on one’s take on the nature of concepts, as well as on the kind of clarification one is interested in. In epistemological contexts, conceptual analysis is often reductive, attempting to give necessary and sufficient conditions for knowledge and justified belief. Target concepts are typically approached by examining our use of the natural language expressions that express them. Thus, we analyse the concept of knowledge by scrutinising our use of the noun “knowledge”, the verb “to know”, etc. Relevant for what is generally seen as the content, as opposed to the non-semantic or pragmatic features, of the concept are (a) the inferential relations between the target and other concepts (e. g. between “S knows that p” and “S believes that p”), and (b) the applicability of the target concept to the world. The commitment to analyse concepts by analysing language has long been seen as a defining trait of contemporary analytic philosophy.

Epistemological applications of conceptual analysis seek to illuminate knowledge by determining its content, perhaps jointly with the content of other epistemic terms. “I put forward my account of perceptual knowledge”, Alvin Goldman writes, “as a more accurate rendering of what the term ‘know’ actually means” (1976: 790–1). Such “analytic” epistemologists in a narrow sense analyse the concept of knowledge by reflecting on our use of linguistic expressions like “knowledge” or “knowing that…” in English.1 Corresponding to the two kinds of content-relevant relations mentioned above, this is done by (a) exploring the inferential relations between sentences containing the expression, and/or (b) by determining to what kinds of states of affairs (and, if we are epistemological contextualists, in what kinds of contexts of use) we would be prepared to apply the expression. Inferential relations (a) are exploited when we ask whether the propositions expressed by sentences of the form “S knows that p” entail the propositions expressed by sentences of the form “It is epistemically rational for S to believe that p”. Applicability (b) is examined when we design thought experiments like Gettier’s, testing whether in these scenarios we are prepared to assert sentences of the form “S knows that p”.

In the standard versions of conceptual analysis we attempt to clarify some target concept as expressed by a linguistic expression by providing a (possibly complex) alternative expression, utterances of which express conceptual content that bears some kind of semantic equivalence relation to the target concept. In order for this to count as an analysis of the target concept, the so-called “analysans” must in some way be clearer, simpler, more basic, better embedded in some accepted theory, etc. The main criteria to decide whether such an analysis is successful are our intuitions regarding the applicability of both the target and the base linguistic expressions in all sorts of inferential contexts and scenarios.

As applied to epistemology, this means that our intuitions regarding the applicability of locutions of the form “S knows that p” and “S has the justified true belief that p” in specific circumstances co-determine the adequacy of the justified-true-belief-analysis of propositional knowledge. Consider the importance given to possible circumstances of application as criteria of adequacy of conceptual analysis in the following passage from Frank Jackson’s defence of conceptual analysis:

When Roderick Chisholm and A. J. Ayer analysed knowledge as true justified belief, they were offering an account of what makes an account of how things are told using the word “knowledge” true in terms of an account using the terms “true”, “justified”, and “belief”. It counted as a piece of conceptual analysis because it was intended to survive the method of possible cases. (Jackson 1998: 28)

Conceptual analysis has been the subject of widespread criticism, resulting in it being “currently out of favour”, as Jackson concedes in the preface to his From Metaphysics to Ethics (Jackson 1998: vii). Perhaps the most important line of criticism is that conceptual analysis aims at a priori results, a project that has been discredited by Quine’s (1952) rejection of the distinction between analytic and synthetic judgements. The value of this kind of criticism, as well as Jackson’s defence against it, partly depends on one’s take on the nature of concepts, a discussion of which is outside the scope of the present investigation. Chapters 4 and 5 of this book will, however, develop an account of the epistemologist’s task and methods, as well as of some structural properties of epistemic concepts (as well as other normative and evaluative concepts) that casts doubt on the prospects of conceptual analysis as an epistemological tool.

Explication

Facing the potential methodological consequences of Quine’s influential criticism of analyticity, theorists committed to the programme of illuminating philosophically important concepts had two options. They could either defend, or subscribe to an existing defence of, conceptual analysis as a method, as the one mentioned by Jackson (1998). Alternatively, conceptual analysis can be dropped in favour of methodologies that illuminate or explain concepts in ways other than analysing them (by saying in other words what we say when we use the analysandum).

Starting with the work of Rudolf Carnap, the notion of explication has been put forward as one important alternative to “analysis” or “definition” (Carnap 1962 [1950]). Just like conceptual analysis, explication aims at illuminating one concept or vocabulary – the “explicandum”, or more generally “target vocabulary or concept” – by means of another – the “explicatum” (Carnap’s choice), “explicans” (by analogy to “analysans” and “explanans”), or more generally “base vocabulary or concept”. Unlike conceptual analysis, however, explication explicitly allows for regulated departures from our intuitions regarding possible cases – i. e. to give up on the idea that competent speakers’ intuitions about possible cases are sacrosanct criteria of adequacy for conceptual analysis as they stand. We should, as Frank Jackson puts it, “be prepared to make sensible adjustments to folk concepts, and this may involve a certain, limited massaging of folk intuitions” (1998: 47). But what counts as “sensible” adjustments?

Where Jackson vaguely speaks of allowing for “sensible adjustments” to the target concept, Carnap attempts to specify what kinds of considerations allow for what kinds of departures from speakers’ intuitions regarding the applicability of the explicandum. Furthermore, he explicitly states that the explicandum is “replaced” by a “new, exact concept” (Carnap 1962 [1950]: 3). Intuitions about possible cases are anything but sacrosanct criteria of adequacy for “satisfying” explication. They help determine whether the explicans is similar in content to the explicandum, which is only one of four criteria. The other three are exactness, fruitfulness and simplicity (Carnap 1962 [1950]: 5). Working with the naturalkind term “fish” as an example, Carnap illustrates how fruitfulness as requiring that a concept can be used for the formulation of general laws justifies replacing the pre-scientifically used term (applying to whales) by its scientific counterpart (“piscis”, excluding whales) (Carnap 1950: 6).

Explication as a method for transforming or replacing concepts according to (theoretical or practical) goals and criteria provides a valuable alternative to conceptual analysis. Carnap himself, however, ties the criteria that determine the success of explication to scientific or formal (logical) theories. So, for anybody who does not readily see the theory of knowledge as one scientific theory among others, or even opposes such assimilation, crucial questions remain with regards to the aims and criteria behind the explication of epistemic terms. Nothing Carnap says prevents epistemic or moral concepts from being legitimate targets of explication. But since properties of the theories in which the explicans is to be fitted provide explication’s criteria of adequacy, and since ethics and epistemology (as “target theories”) potentially differ significantly from physics, biology, probability theory and logic, it is advisable to proceed with two notions of explication. On the one hand, there is Carnap’s narrow notion, as defined by his specific, scientific and/or formal criteria of adequacy (exactness, fruitfulness, simplicity). On the other hand, there is a wider, more generous notion, characterised as a methodology of relating a target concept (= explicandum) to a base concept (= explicans). In the latter case, similarity in content matters, but important criteria of adequacy are taken from properties of the “theories” in which the explicans is to fit.

Reflective Equilibrium

Within the philosophy of language there are two ways of approaching the problem of linguistic meaning. On the one hand, we can tackle our use of the expression “meaning”, for example by analysing the concept, or by explaining its use. This strategy is often seen as stated in Wittgenstein’s Philosophical Investigations, where he says that for a “large class” of uses of “meaning”, the meaning of a word is its use (Wittgenstein 1953: §43). Alternatively, we can treat meaning as a phenomenon about language use; something that would need explaining even when dealing with languages that lack any semantic vocabulary. This second strategy tends to present itself as engaged in an explanatory enterprise, with successful linguistic communication as its explanandum (“How does the speaker manage to convey information about the city of Bangalore, sitting in a Zürich café using the German language?”)

Because of their explanatory ambitions, theories that pursue the second strategy are often called “theories of meaning”, where the use of the term “theory” is meant to exclude concept...

Table of contents

• Cover
• Title Page
• Copyright
• Acknowledgements
• Contents
• Introduction: Retooling the Epistemic Workshop
• 1 Towards a New Meta-Epistemology
• 2 What Is “Meta-Epistemology”?
• 3 Epistemic States and Performances
• 4 Varieties of Norm-Talk
• 5 Epistemic Evaluations and Concepts
• Conclusion
• Bibliography
• Author Index
• Subject Index