A Priori

Edwin Mares

• Publisher: Routledge
• Year: 2014
• Print ISBN: 9781844652846
• eBook ISBN: 9781317547853
• Topic: Philosophy
• Subtopic: Philosophy History & Theory

In recent years many influential philosophers have advocated that philosophy is an a priori science. Yet very few epistemology textbooks discuss a priori knowledge at any length, focusing instead on empirical knowledge and empirical justification. As a priori knowledge has moved centre stage, the literature remains either too technical or too out of date to make up a reasonable component of an undergraduate course. Edwin Mares book aims to rectify this. This book seeks to make accessible to students the standard topics and current debates within a priori knowledge, including necessity and certainty, rationalism, empiricism and analyticity, Quine's attack on the a priori, Kantianism, Aristotelianism, mathematical knowledge, moral knowledge, logical knowledge and philosophical knowledge.

1 Introduction

---

1.1 What is a priori knowledge?

Consider the statement “Every coloured thing is extended in space”. This seems obviously true. If we think about some things that have no length or breadth or depth, we see that they have no colours. Examples of such things are individual points in space and, at least according to traditional philosophy, thoughts. A point cannot have any colour and neither can a thought (although, of course, we can think about colours). But surveying such objects seems unnecessary. We do not have to check everything that has a colour to discover whether it is extended in space. It seems that we can just know this by thinking about it. Traditionally, philosophers have said that this is something that we can know a priori, that is, we can know it independently of experience. Thus, some philosophers say that we can know a priori that every coloured thing is extended. Our knowledge in this case is supposedly independent of experience.

Let us look at another example. Think of a geometrical figure with three straight sides. Let it also be closed: do not allow any gaps in it. How many angles does this figure have? You know the answer: it has three angles. You do not need to check real three-sided figures to make sure they have three angles. From a consideration of the nature of three-sided figures alone, you can know that they all have three angles. This is a priori reasoning.

Here is a partial list of the sorts of things that philosophers have held that we can know a priori:

• Analytic sentences: for example, “All bachelors are unmarried”, “all vixens are foxes”.

• Mathematical truths: for example, “2 + 2 = 4”; “for any natural number x greater than or equal to 3 there are no numbers w, y and z such that w^x + y^x = z^x“.

• Moral principles: for example, “it is wrong to harm innocent people”.

• Conceptual analyses: for example, “red is a colour and not a shape”; “nothing can be both red and green all over at the same time”; “every coloured thing is extended in space”.

• Logical principles: for example, “S or not-S”; “not- (S and not-S)”.

In this book we shall look at the issue of how we know these sorts of things. In Chapters 3–8 we shall examine theories of a priori knowledge and the view that there is no a priori knowledge. In Chapters 9–12 we shall look at various topics that are supposedly the object of a priori knowledge – morality, logic, mathematics, modality and philosophy itself – and apply the theories from the first half of the book to see how well these theories work.

1.2 What is so important about a priori knowledge?

Some philosophers deny that we can know anything a priori. In this book, I call these philosophers “radical empiricists” because they think that all knowledge is at least in some way empirical. But it is difficult to explain how we know some sorts of propositions empirically. Two categories of propositions stand out in this regard:

• Propositions that state necessary truths: for example, the proposition that it is necessary that 2 + 2 = 4.

• Propositions that state norms of certain sorts: for example, the proposition that it is morally wrong to harm innocent people.

With regard to the first sort of proposition, my claim is not that it is difficult to explain how we can know empirically that 2 + 2 = 4 is true, but rather that it is difficult to understand how we could know empirically that it is necessarily true. According to many philosophers, our experience of the world tells us what happens to be true; experience does not tell us what is necessary. These philosophers hold that there must be an a priori element in our knowing what is necessarily true. I discuss the topic of our knowing what is necessary or possible in Chapters 2, 10, 11 and 12, in my discussions of logical, mathematical and modal knowledge.

Similarly, many philosophers hold that we can know empirically what is the case; but experience alone does not tell us what should be the case. With regard to moral norms, it is difficult to see how they could be purely empirical. Perhaps, as J. S. Mill tells us, we all desire to be happy. But should we want to be happy? How could we find this out empirically?

Epistemology too has norms. It tells us what sorts of propositions we should accept and what sorts of justifications we should use. Clearly, empirical evidence can be relevant to epistemology, but what sort of empirical evidence can be used to determine whether (as reliabilists claim) we should only have beliefs that are produced by reliable processes or (as coherentists claim) we should only have beliefs that are supported by and support other beliefs? Empirical evidence may help us find out which beliefs are produced reliably and empirical investigation may help us find support for our beliefs, but the norms that tell us to find this evidence are not themselves purely empirical. And epistemological theories that we shall discuss contain norms such as these.

Of course there are other norms – legal norms, the rules of games and so on – that are accessible to us empirically. But to many philosophers morality and epistemology seem to be different. I discuss these topics in detail in Chapters 6 and 9.

Radical empiricists have not given in to this appeal to modal, moral and logical knowledge, although they have heard it many times. Often they deny that there are necessary truths or facts about what we should do. Sometimes they argue, with regard to both ethics and epistemology, that norms are facts about social institutions. An institution is something that people create, like a government or a club. These philosophers think of morality as a social institution, one that we perhaps do not consciously create, but a human creation nonetheless. On this view, moral norms can be discovered empirically by examining society and epistemological norms can be discovered by examining science. Thus they reject the jobs that a priori reasoning traditionally has been given. In Chapter 6, I look at radical empiricism and the attempt to do epistemology without the a priori. In Chapter 9 I look at radical empiricist views about ethics.

1.3 A note on “knowledge”

One problem with the topic of a priori knowledge is that it seems to require us to talk about knowledge. But the very definition of “knowledge” is one of the most controversial topics in epistemology, and it would be nice to avoid getting embroiled in this debate. Some philosophers talk about a priori “justification” instead of a priori knowledge. I talk a great deal in this book about justification, but we cannot just replace the notion of knowledge with that of justification. As we shall see in §1.4, there are two distinct notions of a priori knowledge and only one of them has to do with a priori justification. I use the term “knowledge” quite often in this book, but only in an intuitive sense. When there is reason to be more careful, I use “a priori belief” to indicate a belief that has a priori status.

As we shall see, a claim that a belief is a priori can mean one of two things. First, it may mean that its justification is in some way independent of experience. In other words, there is support for the claim that does not contain any data taken from experience. Second, the claim that a belief is a priori may mean that the belief cannot be refuted empirically. In such cases, we cannot use empirical data to prove that the belief is false.¹

Note that these two notions of apriority are not exclusive. We could maintain that a belief is a priori if and only if it is justified independently of experience and cannot be falsified by experience. But I discuss the two basic notions independently of one another. In what follows, when I examine various conceptions of apriority I ask which sense of apriority is meant.

Both of these conceptions need further exploration, so let us look at them in more depth.

1.4 A priori justification

The first conception of the a priori holds that an a priori belief is one that is justified or at least capable of being justified independently of empirical evidence. But what sort of independence is this? We rarely, if ever, know anything completely independently of sense experience. Consider, for example, the analytic sentence “Every bachelor is unmarried”. This is one of our paradigm examples of something that can be known a priori (if anything can be). But one cannot know that this sentence is true without knowing the meanings of the words contained in it. We learn the meanings of words by experience: we read them in a dictionary or we are told them by others, or by some similar experience.

Still, many philosophers hold that we can know the truth of this sentence a priori. One way of understanding this position is to make a distinction between the enabling and justificatory roles of experience. Having experience of a certain sort enables us to have the belief that this sentence is true and to justify this belief. This is the experience in which we learn the meanings of the words in the sentence. But, on this view, the justification of the belief that this sentence is true depends only on our understanding of the sentence.²

Any epistemology that allows for a priori knowledge, apart from strong forms of nativism (see Chapter 4), needs to distinguish between enabling and justifying. Otherwise, any claim about a belief being a priori might be undermined by the claim that we learn the concepts involved empirically.

There are two approaches that one can take to defining “a priori justification”. First, one can give a general definition of what it is for a justification to be a priori, and then examine various theories of a priori justification to see whether they are adequate and, in particular, whether their form of justification satisfies the definition. Second, one can examine various theories of a priori justification and then extract from them particular definitions of what it means for a justification to be a priori. I take the second route.

The reason that I do not want to give a general definition is that it is very difficult to give a definition that is adequately general. Contrast, for example, a belief that the analytic sentence “All bachelors are unmarried” is true with an innate belief. The innate belief is independent of experience because it is unlearned.³ The analytic judgement is independent of experience because grasping the meaning of the sentence is enough to see that it is true; we do not need to check any bachelors to confirm what it says. But its meaning is learned. In these two examples, we see two very different rationales for claiming that a belief is a priori. Now, I am not claiming that it is impossible to find a definition of a priori justification that adequately covers both of these cases, but I am saying that I do not think it is necessary to do so. If we decide that we want to take the term “a priori justification” as expressing several closely related concepts, it seems to do no serious harm to my project. So I leave it open as to whether there is a single definition of “a priori justification”.

In this book, I look at specific theories of apriority. According to each of these (except radical empiricism), there are a priori beliefs. But the reasons why each of these theories claims that certain beliefs are a priori differ from one another. In each chapter I comment on these reasons and I consider whether they are adequate warrant to call a belief a priori.

1.5 Theories of justification

In this book I discuss various theories of a priori justification, such as Aristotelianism, analytic justification and self-evidence. Epistemologists usually think not just in terms of what an a priori justification needs to be like, but rather what a justification of any sort needs to be like. There are many general theories of justification, but the three that I discuss are foundationalism, coherentism and reliabilism.

According to foundationalism, there are basic beliefs. These are beliefs that either cannot or need not be justified. They are epistemologically basic in the sense that they are the foundation for the justification of other b...

Table of contents

• Cover
• Half Title
• Central Problems of Philosophy
• Title Page
• Copyright Page
• Dedication
• Table of Contents
• Preface
• Part I
• Part II
• Part III
• Notes
• Bibliography
• Index