“I make no secret of changing my mind on one or two important issues.” It’s a pretty unexceptional admission, but for the speaker of these words the issue of changing one’s mind has a peculiar force. Hilary Putnam has written on a wide range of topics, encompassing metaphysics, the philosophy of language and the philosophy of mind. His output has become essential reading for anyone serious about contemporary debate in all these areas. Yet what he has become most renowned for is changing his mind.

It’s something that baffles Putnam, who can reel off a long list of books and papers he wrote a quarter of a decade ago which he still stands by. He also points to colleagues, like Jerry Fodor, who have made some pretty big U-turns themselves without being saddled with a reputation for inconstancy.

Fixating on this aspect of Putnam’s work is misplaced for two reasons. First of all, not changing one’s mind is hardly a virtue in itself. “I’ve never thought it a virtue to adopt a position and try to get famous as a person who defends that position,” says Putnam, “like a purveyor of a brand name, like you’re selling corn flakes.” Putnam recalls Carnap, with whom he worked at Princeton for a year. “I remember how often he said ‘I used to think so and so, now I think so and so’. I remember admiring that very much.”

More importantly, however, is that readers who focus too much on where Putnam has changed his mind are in danger of missing the constants. Putnam himself says, “Much of the apparatus that I depend on in my own reasoning has not changed.” This apparatus is most evident when one looks at the backbone of his philosophy in his work on language and meaning.

Putnam’s philosophy of language rests on two key claims, known as semantic externalism and semantic holism. The former is most closely associated with him and is summed up in the famous line from ‘The Meaning of “Meaning”‘, “Cut the pie any way you like, ‘meanings’ just ain’t in the head!” I asked Putnam just what semantic externalism adds up to.

“I would say that to be an externalist about meaning is to say that the world external to the brain or mind plays a much bigger part in deciding what our words mean than the tradition emphasises. For example, the fact that there is only one liquid that looks like water and freezes at anywhere near that freezing point and boils at anywhere near that boiling point and satisfies thirst and so on – the world provides exactly what liquid does all that and that has a lot to do with fixing what we mean by the word ‘water’. I call that the contribution of the environment.

“The other thing is that philosophers have for a long time philosophised as if language were a tool that one person could use in isolation. I think it’s a tool like a battleship or a factory that takes a whole lot of people to use. So as an externalist I can say that I can get by using the word ‘gold’ all right even though I would be pretty unreliable if I were on a dessert island and there were pieces of shiny, yellow metal and the question was, are they really gold? I couldn’t tell you, but normally I can rely on an expert who can say.”

This second characteristic of semantic externalism is what Putnam calls “the division of linguistic labour”. Although this is a philosophical thesis, Putnam believes there is a lot of scope for empirical research to flesh out just how this works.

“I said at the beginning and still think that exactly how that works is not something to be settled a priori,” he explains. “People ought to go out and look and I have some, not as many as I would like, conversations with anthropologists who find something like that in even the most primitive society. One example I remember being given by an anthropologist is that in the tribe this anthropologist was working with, men and women had different competences in the use of names for birds. The men who hunt the birds distinguish many more species, which makes perfect sense. Presumably a woman could refer to one of those birds, even though she can’t herself identify them, because she can use her husband or some other male hunter as an expert.”

In sum, semantic holism is about seeing how important facts external to us – such as the way the world actually is and the way the linguistic community works – are in fixing the meanings of words.

The other doctrine concerning language Putnam adheres to is semantic holism. Putnam has a problem with this phrase because, as he puts is, “One must remember that the term was invented, as far as I know, by an enemy.” That enemy is Michael Dummett who used the phrase to describe Quine. “We’ve sort of at times used it as a badge of honour,” says Putnam.

However, the badge is tarnished by a characterisation of semantic holism given by Dummett which Putnam rejects. “I always have to say that if it were really defined the way Dummett defined it – if it’s the view that any change in your beliefs is a change in the meaning of all your words – that is certainly not a view that I’ve ever held or that Quine, when he allowed himself to talk about meaning, ever held.”

So what is the correct characterisation of semantic holism, the one Putnam would agree with? “We holists say that interpretation is a holistic matter. That is to say, what you reasonably take a word to mean can always be changed when you see more text, more what linguists call ‘corpus’. In that sense, no finite amount of corpus – not that the fact the word is used here, and here and here and here – can infallibly show what the word means. Unless you know that it’s never used in some additional way that you haven’t taken account of, you can’t rule out the possibility of having to redefine it.

“There are very simple examples of that. Someone who heard a German speaker use the word stuhl, or a French speaker use the word chaisse, would naturally say, ‘Oh I see what it means, it exactly means chair’. And that’s what most dictionaries give. But eventually you will discover, if you’re American – I’m not sure about British usage – in American usage an armchair is a chair, but it’s not chaisse in French, it’s a fauteuil and it’s not a stuhl in German it’s a sessel. That is, I think, a very good example of the way in which meaning assignments are fallible.”

The connection between both externalism and holism is the importance to both of contextuality. “Semantic externalism is one kind of context dependence,” explains Putnam. “It says to know the meaning of a word you have to see the context in which it’s used, not just the speaker’s brain or mental images. That’s a kind of contextuality. Meaning holism says that more context can flip the best hypothesis about what a word means. So they’re all connected by their mutual connection to contextuality, to the fact that the meaning of a word is something that it has, Frege says, in the context of a sentence and Wittgenstein says in a couple of places ‘in the stream of life’, which is my favourite context principle. I call it Wittgenstein’s context principle.”

Both these views have remained constants in Putnam’s thought for over a quarter of a century. But that is not to say his thinking has stood still. In his more recent writing he has put forward the idea that “meaning is in part a normative notion”. What exactly does this mean?

Putnam expresses doubts about whether saying “meaning is a normative notion” is “a happy formulation”. Nonetheless, in essence, what Putnam means when he does employ this terminology is, “What we say a word means or what we take the sense of a sentence to be on a given occasion, that’s very often a judgement as to what it’s most reasonably taken to mean, and that is essentially normative.”

The crucial point being made here is that if meaning is normative – that is to say, it relies to a certain extent on judgement, rather than just ‘facts’ – several popular theories about meaning are ruled out. For example, Putnam recalls the early seventies. “At that time people were very optimistic about characterising the meaning of words, especially nouns, in terms of what they’re causally connected to. There the normativity cuts strongly against that.”

Putnam has a favourite story to illustrate the normativity of meaning, which he borrows from a joke told to him by the British philosopher, Hugh Mellor. (Like most jokes shared by philosophers, there is an element of the ‘you had to be there’ about it.).

The joke centres around the notion of phlogiston, a substance once believed to be present in all combustible materials but which has long since passed into the chemistry’s dustbin. Putnam recalls Mellor joking that “there is such a thing as phlogiston, it’s valence electrons. The phenomena which both phlogiston and oxygen were invoked to explain are phenomena that involve valence electrons. Nevertheless, we do not say that phlogiston, even though the phenomena which phlogiston was introduced to explain is causally connected to valence electrons, we don’t say that say phlogiston refers to valence electrons, we say that there isn’t such a thing. That has to do with a normative appraisal of whether it’s better to say that the theory was approximately right although mistaken about bla-bla, or to say it’s just wrong, there is no such thing. You can’t just look at what causes you to use the word. You also have to look at the whole context in which it’s used.”

Reading and hearing Putnam talk about such issues, one becomes aware of a general trend in his thought away both from views that treat language and meaning as quasiscientific phenomenon and attempts to over-systemise the philosophy of language. The extent to which one can provide a fully systematic account in various aspects of philosophy is a possibility which seems less plausible in Putnam’s work as times goes by.

“I don’t think meanings are scientific objects,” confirms Putnam. “There’s a big question then whether that means that in some sense that they don’t exist. In a sense, that’s Quine’s conclusion – that we can go on talking about meanings but they don’t really exist at all. That leads to a much wider question, whether the world can be described simply using the vocabulary of first-class physics or first-class rigorous science. I don’t think it can be, but that’s a very large question.”

Once again, at the core of his thinking in these regards is contextuality. In the early part of the century, the logical positivists, such as Carnap, thought you could have context-independent meanings because, as Putnam puts it, “there was this privileged class of observation terms – his examples in the last paper he wrote on this were blue, touches and warm – which had what he called complete meaning, and then you could explain the meaning of any sentence in the language if it were formalised a la logical positivism by some way relating it to these observation sentences.”

However, the idea of context-independent observation terms has not proven to be very durable. Putnam cites a counter example offered by Charles Travis. “If I go to the stationer and I ask for blue ink and it looks black in the bottle, but I dip my pen in and write and sure enough the text is blue, then I would normally say, yes, he sold me blue ink. I don’t care how it looks in the bottle, I care how it looks on the paper. But in another context I might not call it blue ink. It goes back to the whole idea of a sense-datum language and all that. I think we need to rethink what meaning is. I think we expect from logical positivism that the meaning of sentences was supposed to exhibit very little context sensitivity. Words like I and this and the present tense introduce a certain element of context sensitivity, but that’s all.

“Now a number of philosophers, like Travis and myself, and I would argue Wittgenstein already, argue that sentences do not normally have context-independent truth conditions. It’s the meaning of the sentence or the words plus the context that fixes the truth conditions. We need to rethink what meaning is. That’s something I really called for in ‘The Meaning of Meaning’, by the way, where I argue that we need to rethink what a normal form for a dictionary entry for a word should ideally look like.”

However, whereas this proposal calls for more work in linguistics, semantics and lexicography, Putnam bemoans the fact that “philosophy stands almost entirely apart from that, giving much too much significance to ideal language, mathematical logic and all that.” Wittgenstein is a figure who seems to be behind a lot of the recent trends in Putnam’s work. However, it would be wrong to see Putnam as having become a fully-fledged Wittgensteinian.

“In general, I don’t like the limits Wittgenstein puts on philosophy or Wittgensteinians put on philosophy, but many of his criticisms of the traps we fall into seem to me profoundly right. Some of them come from a misplaced belief that systematicity must be possible, what he calls a ‘philosopher’s must’. As I’ve grown more aware of the importance of the notion of context sensitivity, I see that now really as running through a lot of Wittgenstein’s arguments.”

Where he does stand shoulder to shoulder with Wittgenstein is on the issue of what ‘rigorous’ philosophy is. “I think that we gave a certain metaphysical significance – and I am a mathematician too, as well as a philosopher – to mathematical logic in the twentieth century, and I think that was a mistake. I think we still suffer from the idea, and I think maybe we’re now doing that with Chomsky and linguistics. We’re still suffering from the idea that these paradigms tell us what tightening something up must be like. There I am very much with Wittgenstein. That should not be what tightening up something in philosophy means. In fact, here philosophy is not unique. It happens periodically in sociology, it happens periodically in economics, it happens periodically in all the social sciences, that we get this dream that we can take over formalisms.”

Oddly enough, Putnam believes part of the attraction of some of these formalisms is their obscurity. “I think part of the appeal of mathematical logic is that the formulas look mysterious – You write backward Es!”

On a more personal note, in the introduction to his recent Renewing Philosophy, Putnam says that he is a practicing Jew. Although the connection may not be obvious, in his review of Putnam’s The Threefold Cord, Colin McGinn’s pays Putnam the back-handed compliment that his fundamental problem is that he cannot bear to be boring.

“That could be the worst thing said about me,” laughs Putnam, whose writing is indeed a welcome antidote to the dry, humourless style of many of his contemporaries. Both the style of writing and the acknowledgement of his religious views reflect something quite important about how Putnam sees philosophy.

“I think that the philosopher should to some extent disclose himself as a human being,” he explains. “That’s something that James argues in lecture one in Pragmatism. I’m paraphrasing him, but he quotes a Walt Whitman saying, which is, ‘who touches this book touches a man’. He says in effect that’s what he wants people to say about his books. That seems right to me.

“I think there is such a thing as the authority of, not reason with a capital R, but what we call intelligence,” he adds later. “Nevertheless, I also agree with Dewey that it’s always situated, it’s never anonymous.”

And on that point, he is not about to change his mind.

This interview originally appeared in Issue 15 of The Philosophers’ Magazine.

Julian Baggini’s book, Freedom Regained: The Possibility of Free Will, is published by Granta in the UK and by Chicago University Press in North America.