Reason and Explanation

A Defense of Explanatory Coherentism

T. Poston

• Publisher: Palgrave Macmillan
• Year: 2014
• ISBN:
  9781137012265
• Topic: Philosophy
• Subtopic: Epistemology in Philosophy

In this new explanationist account of epistemic justification, Poston argues that the explanatory virtues provide all the materials necessary for a plausible account of justified belief. There are no purely autonomous reasons. Rather reasons occur only within an explanatory coherent set of beliefs.

1

Introduction

A coherent Thinker, and a strict Reasoner, is not to be made at once by a Set of Rules.

I. Watts Logick 1725

Be plain and coherent.

Dickens Dombey and Son 1848

It is reason’s proper work to provide us with a maximum of explanatory coherence and comprehensiveness

Sosa Knowledge in Perspective 1991

Epistemic coherentism has not merited much attention throughout the history of philosophy. Two explanations for this neglect stand out. First, when a statement is challenged one defends it by citing another statement. This aspect of how to respond to challenges in specific contexts leads to the thought that the structure of epistemic justification corresponds to the structure of responding to challenges in a specific context. Socrates’ persistent search for particular reasons or Descartes’ method of doubt manifests this natural tendency. Both search for reason’s proper ground in some unchallengeable propositions whose epistemic merits is independent of any particular perspective. It is natural to extend this pragmatic feature of the activity of justifying belief to the state of being justified in one’s beliefs. This natural search for reason’s proper grounding obscures the plausibility of epistemic coherentism.

The second reason coherentism has been neglected lies in the assumptions undergirding the traditional regress argument. The regress problem presented by Aristotle and subsequent philosophers offers three non-skeptical options to the general question how is any claim justified?: circles, regresses, or foundations. Do we start with some controversial claim and defend it by reasoning through many propositions which eventually loop back to the original claim? Does reason continue forever? Or do we reason to some claim which provides a reason without itself requiring a reason? Of these non-skeptical options foundationalism seems the most plausible; a claim is justified by being basic or inferred from basic beliefs. But these reasons support a foundationalist epistemology only if they exhaust the plausible options. Yet as many coherentists have pointed out there is a fourth non-skeptical option: holism, the support of any particular claim is a matter of how well the claim is supported by everything else within the relevant system. Justification is not foundational because any specific claim requires a reason. Justification does not require an infinite regress because holistic support does not require an infinite number of non-repeating claims. And justification is not circular because holistic support is different from circular arguments. It is up to the specific holist view to say more precisely how justification works. My task in this book is to defend an explanatory coherentist form of holism. I contend that a subject’s propositional justification for any claim is a matter of how that claim fits into a virtuous explanatory system that beats relevant competitors.

1.1 A brief history of coherentism

Holism is a historically curious view. Metaphysical holistic views arose alongside idealism. The idealists resist a dualism of mind and world, insisting that reality is fundamentally mental. Idealists reject the correspondence theory of truth according to which a true statement corresponds to some fact. On idealist views the truth of a statement consists in its coherence with other statements. This metaphysical form of holism, though, has dubious connections with confirmational holism, also known as epistemic coherentism. That epistemological view arose within the pragmatist tradition and upon reflection on the revolutions in the formal and physical sciences of the late nineteenth and early twentieth centuries.¹

Epistemic coherentism is the view that a statement’s justification—its acceptability for the purposes of getting at the truth and avoiding error—consists in its overall coherence or fit with the rest of accepted statements. Otto Neurath expresses this by the simile of rebuilding a raft at sea. The process of epistemic improvement is like rebuilding a ship at sea. One can improve parts of the ship only by relying on other parts of the ship; there is no dry dock to rebuild on firm ground. The simile suggests that one can improve one’s stock of beliefs only by relying on other beliefs one has; there are no properly basic beliefs whose justification is independent of other justified beliefs. All belief revision takes place within an assumed background of accepted beliefs. Neurath’s raft simile contrasts with Moritz Schlick’s pyramid metaphor. For Schlick, another member of the Vienna Circle, epistemic justification has a pyramid structure; the justification for some beliefs rests upon the justification for other beliefs whose justification does not depend on the justification of other beliefs. The development of logical empiricism up to Quine and Sellars favored Schlick’s foundationalism over Neurath’s holism.²

W.V.O. Quine and Wilfrid Sellars each forcefully argued against the logical empiricist project and advanced a coherence view of justification. Quine argued that the backbone of the logical empiricist attempted reconstruction of acceptable statements depended on the myth of a sharp distinction between analytic statements and synthetic statements. According to logical empiricism, each statement could be determined to be analytic or synthetic solely in virtue of its form. The analytic statements were justified by themselves and the synthetic statements required experience to be justified. The class of synthetic statements itself was divided into statements which report pure experience (the experiential statements) and those which were logical constructions of the experiential statements. Theoretical statements of science were to be treated as logical constructions of experiential statements. These theoretical statements are then analyzed into analytic equivalences of synthetic statements. Quine’s famous paper ‘Two Dogmas of Empiricism’ forcefully attacked the principled analytic-synthetic distinction presupposed by the logical empiricist program. Quine’s contrasting picture was a holistic view in which the justification of any statement depends on its role in the web of belief.

Wilfrid Sellars’ famous paper ‘Empiricism and the Philosophy of Mind’ attacked logical empiricists supposition that there are pure experiential statements directly justified by experience. Logical empiricists claimed that the content of these statements directly match the content of the associated sense experience in such a way that provided a guarantee of their truth. Sellars argued that the existence of these statements is a myth; it is one form of “the myth of the given.” Sellars argues that sense experience either has belief content or it does not. Belief content is content that has a truth-value and can be endorsed in belief. If sense experience lacks any belief content then it is a mystery how sense experience can justify any belief that the world is one way rather than another. In this case it would be a complete mystery why an experience could justify a specific belief. Yet if sense experience has belief content then there must be some reason for thinking that that belief content is true. The claim that the content of a belief would require justification and yet the belief content of experience would not is incoherent; for the content to be justified in one case is the same as the content that doesn’t need to be justified in the other case. The distinction between the belief mode of hosting a content and the experience mode is artificial. If content requires justification and sense content is belief content then their epistemic fates stand or fall together. Sellars, thus, reasoned that the logical empiricist project of ending the regress of justification in pure experiential statement is a myth.

The fate of epistemic coherentism becomes murky after Quine and Sellars. The influence of these seminal philosophers extended from the early 1950s through much of the twentieth century. Yet epistemology took a curious turn in 1963 with the publication of Edmund Gettier’s paper ‘Is Justified True Belief Knowledge?’ Gettier argued that knowledge is not justified true belief. Knowledge requires some additional property than justified true belief. Gettier’s paper led to a flourishing of new approaches to knowledge. At the time epistemologists thought that the relationship between knowledge and true belief must be thought out anew. Hence, epistemologists turned their backs on the methodological issues of science which grew out of reflection on logical empiricism and instead focused on finding a solution to the Gettier problem. One dominant approach was the formulation of epistemic externalism, the idea that knowledge should not be conceived in terms of justification or reasons, but rather as a natural relation that one bears to one’s environment. For example, it was proposed that knowledge is the state of believing that p when the belief is caused by the fact that p. The rise of externalist approaches to knowledge led to the vigorous internalist externalist debate in epistemology.

The burst of activity on the Gettier problem shifted attention away from the methodological issues pertaining to theory construction which engaged the logical empiricists and subsequently Quine and Sellars. However, important developments on methodology still occurred. In the mid-1970s epistemologists realized that several prominent arguments against foundationalist views wrongly assumed that the foundational beliefs must be infallible. For example, D.M. Armstrong’s ‘distinct existence’ argument contended that experience is not an adequate ground for belief because experiential states are ontologically distinct from belief states; the one can exist without the other. Consequently, Armstrong reasoned, experience cannot provide an indubitable ground for belief. William Alston and Mark Pastin both independently observed that foundationalist views need not be committed to the thesis that experience is an infallible ground for belief. This defeasible foundationalist view held that basic beliefs are justified but can be undermined by additional information. Even so, Alston and Pastin contended that experience is still a proper stopping point in the regress of reasons.

Such was the current state of play when Laurence BonJour published his well-regarded defense of coherentism in 1985. BonJour’s book The Structure of Empirical Knowledge defends a coherentist account of empirical knowledge by way of two prominent arguments. First, BonJour argued that epistemic justification required more than simply a reason why the belief is likely to be true; a subject must be in cognitive possession of such a reason. BonJour argued that this requirement conflicts with externalist views of knowledge and justification. Second, BonJour argued via the Sellarsian dilemma that experience is unable to provide such a reason. This argument attacks internalist foundationalist views. The view left standing after BonJour’s arguments is a form of epistemic coherentism.³

BonJour’s arguments for epistemic coherentism are clear and forceful. However, his overall coherentist view is curious for two reasons. First, BonJour restricts his coherentist account to empirical knowledge, explicitly adopting a traditional rationalist foundationalist view of the a priori. His foundationalism about the a priori is at odds with the arguments he offers for resisting a foundationalist account of empirical knowledge. More recently, BonJour has renounced coherentism because of this problem.⁴ Second, related to his bifurcated epistemology, BonJour’s development of epistemic coherentism does not fit with the coherentism of Quine, Sellars, Harman, Goodman, and Lycan. In his Appendix B ‘A Survey of Coherence Theories’ BonJour notes that he will not discuss the views of Quine, Sellars, and Harman because they all ascribe to epistemic conservatism. Conservatism is the thesis that belief is an epistemically relevant factor and, under some conditions, is a sufficient condition for justification. BonJour argues that conservatism is a form of weak foundationalism and thus not a coherentist view. In contrast to BonJour’s argument, though, conservatism runs deep in the pragmatist tradition, and the mid-century coherentists were all pragmatists. Thus, BonJour’s coherentist view is considerably different from the confirmational holism espoused by philosophers writing in the tradition of Quine and Sellars. BonJour’s coherentist view shifted discussion away from the pragmatist coherentist view of the mid-twentieth-century epistemologists.

Even so, excellent work on coherentist themes continued. Keith Lehrer’s books Knowledge⁵ and A Theory of Knowledge⁶ argued for a novel form of epistemic coherentism based on the idea of a competing system. Lehrer argued that a belief’s justification is a matter of its membership in a subject’s veritic system of beliefs which beats all competitors. Gilbert Harman’s superb books Thought⁷ and Change in View⁸ developed the Quinean coherentist project by focusing on justified change in belief. Harman is less concerned with defending an analysis of epistemic justification and more concerned with defending a conservative, non-skeptical epistemic view that gets human inference right. William Lycan’s second half of his book Judgment and Justification⁹ defends a pragmatist explanationist view which is explicitly rooted in a broadly Quinean epistemology. Lycan advances the thesis that a belief’s justification consists in its membership in a virtuous explanatory system. Lastly, Paul Thagard has steadily developed a broadly coherentist epistemology aimed at tying that approach ...

Table of contents

• Cover
• Title
• 1  Introduction
• 2  Epistemic Conservatism
• 3  Reasons without First Philosophy
• 4  Explanation and Justification
• 5  BonJour and the Myth of the Given
• 6  Is Foundational A Priori Justification Indispensable?
• 7  Bayesian Explanationism
• Bibliography
• Index