Lecture One The Debate Over the Theory of Reference for Proper Names

1. Where we will make our start

2. The supervenience of reference on nature

3. The availability issue

4. What is required of a user of a name?

5. The issue about personal level knowledge

6. The demand for precise and explicit specifications

7. What is a theory of reference a theory of?

8. In the mouths of the very young

9. The description theory and interchangeability

10. What is to come and a final objection noted

Lecture Two Understanding, Representation, Information

1. Some stage setting on the value of understanding words and plans

2. Agreement and s hared u nderstandings

3. Davidson’s challenge to representation

4. Are we confusing semantics and pragmatics?

5. Why we need possible worlds

6. Voyages through logicals pace

7. How to finesse the issue in analytic ontology

8. The need for centered worlds

9. Getting information from sentences with centered content

11. Where to now?

Lecture Three Ir-content and the Set of Worlds Where a Sentence is True

1. Preamble

2. The case of proper names

3. The difference principle

4. The ‘within a world’ version of the argument u sing the difference principle

5. Sentences containing “actual” and “actually”

6. Demonstrative adjectives

7. Natural kind terms

8. A passing comment on centering

9. Where to from here?

Lecture Four Two Spaceism

1. One spaceism versus two spaceism: setting the scene

2. Two spaceism and ir-content

3. Which label: “epistemic” or “conceptual”?

4. Which possibilities, precisely, are the ones two spaceism holds are conceptually possible but metaphysically impossible?

5. How working with the bigger canvass raises some of the same questions over a gain

6. Why two spaceism is not a happy home for anti-reductionists

7. Where to from here?

Lecture Five The Informational Value of Names

1. Where we are

2. When truth at a world depends on more than how that world is

3. A diagram to give the key idea

4. A language where truth at a world is, by stipulation, a function of which world is actual

5. On looking for examples of two-dimensional sentences in the English of the folk

6. How should we approach questions like, How do we use the word “water”? and, How do we use the word “Gödel” ?

7. More on the evidential role of intuitions about possible cases

8. What information do we impart with the word “water” – first pass

9. With enemies like these, who needs friends?16

10. What information do we impart with the word “water” – s econd pass

11. The issue about narrow content18

12. Proper names and information (I)

13. Proper names and information (II)

14. Coda



The Blackwell/Brown Lectures in Philosophy

The Blackwell/Brown Lectures in Philosophy present compact books distilling cutting-edge research from across the discipline. Based on public lectures presented at Brown University, the books in the series are by established scholars of the highest caliber, presenting their work in a clear and concise format.

1. Semantic Relationism by Kit Fine

2. The Philosophy of Philosophy by Timothy Williamson

3. Language, Names, and Information by Frank Jackson

Forthcoming books by Philip Pettit, and John Broome


For Charlotte and Stella


When I read Saul Kripke’s Naming and Necessity and Hilary Putnam’s “The Meaning of ‘Meaning’ ”, they seemed to me, as they did to so many, to be enormously insightful contributions to the philosophy of language.1 However, although I had much company in coming to this judgment, I found that I was in the minority when it came to articulating the important lessons we should take away from these publications. I saw (after a good deal of reflection) Kripke as teaching us that proper names are typically sources of information about what they name, a message that fits nicely with the description theory of reference for proper names, provided that the descriptions in question are spelt out in terms of causal connections that carry information. The majority saw Kripke as refuting any version of the description theory. The one thing we did seem to be in agreement about was that Kripke had taught us that proper names are rigid designators.

I saw Putnam as teaching us that the reference of names of kinds often goes by underlying nature, especially scientifically significant underlying nature, not by the properties that first led us to postulate the kinds in question. The message was that it is a mistake to hold, for example, that “acid” refers to any substance that has the properties that led chemists to introduce the term into chemical theory. “Acid” refers, rather, to the kind that typically explains those properties, or properties like those properties.2 This does not mean that the reference of “acid” doesn’t go by known properties. We know that there is a kind underlying and explaining the properties that led chemists to introduce the category acid. Being such a kind is, therefore, a known property. It means, rather, that some substance counts as an acid if it is of the kind, or a kind, that typically underlies the manifest properties, even if it does not itself have the manifest properties.3 Equally, something can have all the properties that led to the introduction of the category but fail to be an acid by virtue of not being of the kind that typically underlies the properties. Moreover, although this is an important fact about the reference of many terms in science, it isn’t an essential feature. The concept of a vitamin is a significant one in the science of nutrition but it is not true that “vitamin” refers to a unified explanatory kind or kinds. However, the lesson the majority seemed to take away was that the reference doesn’t go by known properties at all. But my biggest disagreement with the majority was over the significance of the Twin Earth thought experiment that Putnam used to cement his ‘reference goes by kinds’ message. The majority, and Putnam himself, took the thought experiment to provide a compelling argument for externalism about linguistic content, and, in subsequent developments, for externalism about mental content. I dissented, and in addition was puzzled by the insouciance with which many were prepared to abandon narrow content. As we will see later (in Lecture Five), we need narrow content.

Over the years I have defended my dissenting opinions,4 and, while being in the minority, I have been far from alone in my dissent. Indeed, sometimes in expressing my dissent I have found myself saying, in my own words and from my own take on the issues, what others have said, in one way or another. The kind and much appreciated invitation to give the Blackwell/Brown Lectures in Philosophy gave me the opportunity to put things I have said in various places into a coherent package, and that is what I sought to do in the lectures. Moreover, since giving the lectures, as a result of reflection on the discussions arising at them and subsequent discussions with colleagues and friends, and at conferences and departmental seminars, and with my graduate students, I came to realize two things that were not clear (or not clear enough) to me when I gave the lectures.

One is that there are two quite different things you might mean by the description theory of reference for proper names. On one reading, the theory is certainly false. On the other, it is, I will be arguing, true. I will say a little about these two readings near the end of the first lecture and in more detail in the final lecture (Lecture Five).

The other thing I came to realize was that, in the dissenting papers already mentioned, I had not sufficiently emphasized that in order to make good sense of the debate about proper names, over names of kinds, and the debate over narrow versus broad content, one needs to set the debate explicitly within an overall approach to meaning and reference, an overall approach to the question, What precisely is our subject when we discuss questions of meaning and reference? Consulting and swapping intuitions about what various sentences and words mean, in the absence of a view about what it is that we are debating when we discuss different views about meaning and reference, does not, I came to realize, get us very far.

The overall approach of these lectures might be called the informational-cum-representational-framework. I will say a good deal about this framework in the lectures to come, but the core idea is that a language is a system of representation that delivers putative information about how things are to those who understand the language. The most important thing about the sentence “There is a land mine one meter from your left foot” is how it represents things to be, the information it carries, in a way that is accessible to those competent in English. In philosophizing about meaning and reference, we need to keep this insight center stage, or so it seems to me. This implies, for example, that the crucial question for any account of proper names is whether or not it captures correctly how users of the name represent things to be, the putative information they deliver, when they use the name in declarative sentences.

I have just spoken of what I will do in the lectures to come rather than what I will do in the chapters that follow. This is partly to mark the origins of this book in a set of lectures, albeit a set of three rather than the set of five lectures that make up this volume, but mainly to signal that what is to come is very much in lecture format. I have deliberately kept the tone conversational and informal. Of course there are footnotes, but I have tried to minimize their number. Of course there are technicalities, but I have tried to corral them and to say things in ordinary English as much as possible. Although Kripke’s and Putnam’s seminal publications carry surprisingly little technical baggage, subsequent discussions have, understandably and properly, sometimes become quite technical. This means that I cannot avoid a certain amount of technical entanglement – we will, for example, need to talk on occasion of functions from worlds to truth values, and of functions from centered worlds to truth values, in discussing two-dimensionalism. However, my belief is that the key points can, in the main, be made in terms the folk – that’s you and me when we aren’t being professional philosophers, cognitive scientists, psychologists, or whatever – understand. This is as it should be. Our subject is natural language, the language of the folk, not quantum mechanics or the foundations of mathematics.

I have many debts. To my lecturers in Introduction to Statistics, and later to Robert Stalnaker and David Lewis in their publications, for teaching me about the possible worlds way of thinking of information and representation, and more generally of the world-directed nature of much of language.5 To Martin Davies, Lloyd Humberstone, and Pavel Tichý for teaching me about two-dimensionalism.6 To my graduate students, as already mentioned. To a reading group at ANU. To too many colleagues and friends to name, but I must list (in addition to those already mentioned): David Braddon-Mitchell, David Chalmers, Simon Cullen, Michael Devitt, Andy Egan, Fred Kroon, Philip Pettit, Brian Rabern, Denis Robinson, Michael Smith, Daniel Stoljar, and Wong Kai-yee. To the very helpful comments from two readers for Wiley-Blackwell. I am indebted to Scott Soames in a rather different way. His trenchant criticisms of the description theory of reference (2002, 2005a, 2005b) forced me to get clearer about what that theory was, and made me realize the need for these lectures. It goes without saying, but I will say it all the same, I alone am responsible.

Princeton, November 2009

1Kripke (1980); Putnam (1975).

2The properties need not be quite as we first thought of course.

3And not just for the uninteresting reason that it is highly diluted.

4Jackson (1980, 1994, 1998a, 1998b, 2003, 2005, 2007b)

5See, e.g., Stalnaker (1984) and Lewis (1986).

6Davies and Humberstone (1980 and in conversation), Tichý (1983).