Friday, 15 August 2025

Truth, According to Analytic Philosophers, and as Seen By a Historian

 

This isn’t a commentary on the theories of truth as they’re expounded by analytic philosophers: it’s a commentary on Felipe Fernández- Armesto’s own take on this subject (as found in his book *Truth: A History*). He tackles the views of Wittgenstein, Tarski, Davidson, and Paul Horwich, and ties them all to the denial of reality and to relativism. So although relativism itself isn’t tackled in my own essay, it’s a still a prelude to Fernández-Armesto’s central target. (In his book, Fernández-Armesto does confront Saussure, Derrida, and Rorty.)

Press enter or click to view image in full size

“Felipe Fernández-Armesto (born 1950) is a British professor of history and author of several popular works, notably on cultural and environmental history. He was born in London; his father was the Spanish journalist Felipe Fernández Armesto [ ] and his mother was Betty Millan, a British-born journalist and co-founder [ ] and editor of The Diplomatist (whose current title is Diplomat), the in-house journal of the diplomatic corps in London.

“Felipe Fernandez-Armesto joined the history department at the University of Notre Dame in 2009, after occupying chairs at Tufts University and at Queen Mary College, University of London. He had spent most of his career teaching at Oxford, where he was an undergraduate and doctoral student. [ ]”

— See source here.

Genesis

It must be borne in mind that what I shall tackle isn’t Fernández-Armesto putting his own philosophical positions in his book, but what he believes his philosophical opponents believe (i.e., that they deny reality and play up language).

It’s interesting that many laypeople and philosophers believe that postmodernists, post-structuralists, etc. deny reality and propagate relativism, whereas Filipe Fernández-Armesto ties both postmodernist, etc. philosophers and analytic philosophers to these things.

Thus, if you emphasise language’s strong relation to what many people call reality, then that can be deemed to be a philosophically old-fashioned and naïve view. Or at least Fernández-Armesto tells us that

“[t]raditionally, language has been presented according to the model of Genesis: God creates things, then names then”.

Here “things” comes first, and only then are they named. Thus, things are originally unpolluted by names or by language… (But what were these things before they were named? Tell me something about them.)

So we had things then names. What about the possibility of names then things? Or as Fernández-Armesto puts it:

“But suppose it were the other way round? Suppose language precedes experience and we model the world to fit it?”

This isn’t a precise inversion in that name then thing isn’t the inversion of language then experience.

So language precedes experience… for whom? For babies? For adults? Or for communities?

It’s hard to imagine language preceding all experience — that’s even if language colours or shapes experience. Unless, on the broad scale, Fernández-Armesto is putting the position that, collectively, we come at experience with our given language/s. But even then this position is hard to grasp.

Isn’t all this what happens when a prior position is almost literally inverted? That is, why is language then experience any better than experience then language (or thing then name)? Again, perhaps there is no temporal binary opposition between experience and language… except in the case of pre-language babies.

This Genesis story is similar to a 20th century way of putting things.

According to John Heil, philosophers have believed that “words are connected to things [] by ‘outgoing’ chains of significance guided by the agents’ thoughts (‘noetic rays’)”. [See ‘Noetic’.)

Fernández-Armesto then states the obvious when he tells us that “there is no means of representing claimed truths outside language [or by] other means of expression”. But some people still believe that although “claimed truths” obviously require a language, reality itself doesn’t require a language.

No, reality may not. But where does that get us?

The Biblical naming-story is but a prelude to Fernández-Armesto’s later questions.

Fernández-Armesto on Wittgenstein

As probably noted, Fernández-Armesto focuses on language and its relation to reality. Simply put, some philosophers have focussed on language, others on reality. Wittgenstein focused on language, and thereby (so some argue) seemed to threaten reality.

Fernández-Armesto writes:

“His argument that we understand language not because it corresponds to reality but because it obeys rules of usage seems unanswerable.”

What can that mean?

He continues:

“Therefore, when we understand language, we do not necessarily know what it means or (to use a term a philosopher might prefer) refers to, except its own terms.”

These themes are now familiar, and the ripostes are too.

Again, why would there be “rules of usage” without a reality which can, indeed, be referred to? Do the rules of grammar exist just for the hell of it? It can, of course, be said that they exist simply so that people can communicate with each other. But communicate about what? About rules of grammar? About language more broadly? About words and parts of speech?

Fernández-Armesto concludes that “when we understand language, we do not necessarily know what it means”. That is, we don’t necessarily know what it means because reality has effectively been erased (i.e., by philosophers like Wittgenstein). Thus, if language isn’t tied to reality, then we can’t know what it means. But Wittgenstein might have argued that this conception of language’s relation to reality isn’t required… Then again, he probably wouldn’t have said that.

What’s more, in Fernández-Armesto’s picture of relativists and/or irrealists, the terms of language refer to other terms of language. This sounds suspiciously like Ferdinand de Saussure’s position (long before Wittgenstein), who’s also covered by Fernández-Armesto. This isn’t to to say that Wittgenstein was influenced by Saussure, or that Fernández-Armesto is conflating Wittgenstein and Saussure. However, Saussure best expressed this position in one of its simplest forms.

But we can go back beyond Saussure, as Fernández-Armesto does. We can go back to medieval nominalism. (Fernández-Armesto doesn’t actually use the word “nominalism”.) Fernández-Armesto writes:

“When I last wrote about Wittgenstein, I likened his work to that of medieval philosophers who assumed ‘that words signified only themselves, not any independent reality, and statements referred only to their own terms’”.

Where does all this lead?

It “left intellectuals of the late twentieth century peering into the dark glass of self-referential language”. Of course, Fernández-Armesto continues by saying that “some philosophers of my acquaintance have told me that I have misrepresented Wittgenstein, others that I have understood him correctly”…

What a surprise!

As it is, I can hardly understand what the phrases “words signified only themselves” and “statements referred only to their own terms” even mean. This isn’t because I have a mindless or naïve commitment to the importance of Reality. Regardless of that, they’re still very odd locutions. And, as already stated, this is to take Fernández-Armesto’s words as they are given, and not to gloss over them with inputs from various philosophers he himself doesn’t mention.

Truth, Reality, and Linguistic Expressions

According to Fernández-Armesto, the American philosopher Donald Davidson tried to “separate the truth-question from the reality-question”. Among certain types of philosopher, this was part of a popular effort to (to use an ugly word) linguistify truth, as well as philosophical problems generally. Yet Davidson’s claim may seem odd, and not just to laypeople.

Surely, there is no “truth-question” unless there’s also a “reality-question”. After all, if truth is only about linguistic expressions, then won’t other linguistic expressions make a given statement true (or false)? Put differently, what is truth without reality? Yet that simply raises the question: What is reality?

So which question shall we ask first? What is truth? Or: What is reality? Perhaps this is a package deal.

We need to know at least something (philosophically) about how we express our truth claims in language. And truths can only be expressed via linguistic expressions. However, does that necessarily factor out the reality-question — or plain reality? Again, aren’t truths about reality? Or, at the very least, aren’t truths true because they’re about… things that are non-linguistic?…

Perhaps not.

Well, if you believe that “truth is a property of an expression rather than a declaration of what is really there”, then reality surely must drop out of the picture.

Or does it?

What does it even mean to state that “truth is a property of an expression”? Is truth a property of an expression simply because it’s communicated by an expression? And isn’t the claim that truth is a property of an expression to already have a position on what truth is?

Perhaps that phrase means that we give expressions truth values because it doesn’t make any sense to give reality (or a chunk of reality) a truth value. In other words, it is linguistic statements (whether written or spoken) which are either true or false…

But linguistic expressions are true or false because of things outside those expressions. The expressions, after all, aren’t self-referential (despite the quoted claim earlier).

According to Fernández-Armesto, Davidson held his position on truth and reality because

“the evidence is equivocal that there is any reality outside the mind — or even, to an uncompromisingly rigorous, outside language — for the truth of expressions to fit”

Everything stated about reality must be said by a mind — or by a human person with a mind. And everything stated about reality must be communicated by a linguistic expression. However, do these facts factor out reality?

It’s often been said that reality outside the mind cannot be demonstrated. And anything known and said about reality must be communicated by a linguistic expression. So even if (or even though) linguistic expressions may “fit” reality, there is nothing said about reality without both minds and linguistic expressions. At a minimum, truth and knowledge depend on linguistic expressions — that’s even if reality if still allowed a place in such a philosophical scheme.

Fernández-Armesto even hints that we may not need reality. (Again, he’s putting the — supposed — position of his philosophical opponents.) Or at least he raises the point that “we do not necessarily demand a ‘morality-world’ or a ‘beauty-world’ to authenticate our moral or aesthetic judgements”. So perhaps we don’t need a reality-world either. Still, moral statements are about things which aren’t themselves moral statements. Similarly, aesthetic statements are about things which aren’t themselves aesthetic statements.

Do we need to mention worlds at all here?

It depends on what Fernández-Armesto means by “world”. Does he (or his philosophical opponents) mean the actions, events, people, etc. that are the subjects of moral claims? Or is the “morality-world” something more Platonic, rarefied and/or abstract?

Banal Philosophy About Propositions and Facts

“A proposition is true if there is a fact which fits it.”

The subject of the relation of propositions to facts has been debated to death in analytic philosophy. Indeed, statements like the above have also been analysed into oblivion.

That said, readers will still need to know what a proposition is, and what a fact is. Of course, many readers will already have a position on this. Yet it may not entirely coincide with what philosophers — any philosopher — has said on this subject… However, I’ll leave it there and simply tackle what Fernández-Armesto himself says.

He says that the “first formula is circular because it is impossible to say what a fact is without saying that it is true”. That’s not true. It is possible to say what you take a fact to be without also making a claim as to the truth of a particular factual claim. When it’s said that x is a fact, then truth enters the picture. But that isn’t about the nature of facts: it’s a claim about a particular fact and whether it’s true or false.

Fernández-Armesto may also mean (ironically) that facts are (as Peter Strawson once put it) “sentence-shaped objects”. That is, facts are as linguistic as propositions. It’s true that most laypersons believe that facts are chunks of concrete reality. However, those concrete chunks of reality are chunks because they’ve already been linguistified. So the claim is “circular” because the fact is the mirror image of the proposition. In other words, it’s as if the formulator of the proposition (although propositions are deemed to be abstract) designed the fact in the very process of formulating his proposition…

“‘Snow is white’ is true only if snow is white.”

That may be the most banal statement ever. Yet it’s been cogitated over for decades. Indeed, perhaps it needs to be banal.

It can be called the linguistic version of the correspondence theory of truth. However, it need not be so connected. As with “a proposition is true if there is a fact which fits it” earlier, I’ll only note what Fernández-Armesto himself has to say on this subject.

He calls it a “tautology”. Strictly speaking, it can’t be a tautology because the left-hand side is a linguistic statement, and the right-hand side isn’t… meant to be. On the surface, then, it seems like a sentence and the chunk of reality it refers to. That said, many readers will still see Fernández-Armesto’s point!

Fernández-Armesto claims that Alfred Tarski’s rendition “avoids commitment to belief in reality”. Yet there is no explicit commitment within Tarski’s meta-statement, and no use of the word “reality”. However, isn’t the right-hand side “about” a chunk of reality? What else could it be about?

What about the supposed sidestepping of any “correspondence between words and reality”? It depends on what that means.

For example, no correspondence theorist has ever argued that a statement needs to be like a chunk of reality in order to correspond to it. Statements don’t literally mirror reality. (In Wittgenstein’s “picture theory”, propositions only “mirror the ‘logical structure’ of facts”.) The correspondence is abstract. The correspondence also needs to be philosophically interpreted and analysed. In other words, there is no algorithmic “fit” between a statement and a chunk of reality either.

How does Tarski’s trick of making “propositions correspond to the conditions which make them true” sidestep reality? Is it simply because it uses the word “conditions” rather than the word “reality”? Conditions of what? How do conditions make statements true? (How do chunks of reality make statements true?)

Fernández-Armesto on Paul Horwich’s Minimalism

Fernández-Armesto seems to believe that Tarski sidestepped any commitment to reality. And he says the same thing about the British philosopher Paul Horwich.

Again, this isn’t a commentary on Horwich’s minimalism: it’s a commentary on Fernández-Armesto’s take on Horwich’s minimalism.

Fernández-Armesto tells us that when it comes to Horwich’s “minimalism”,

“[y]ou do not have to worry about whether the proposition in question corresponds to fact or refers to anything real”.

Really?

It’s not that I disagree with that: it’s that I don’t understand what it means.

What do these propositions do then? What is their role? Do they refer to other propositions? Are they about things which have nothing to do with “anything real”? Such as what?

Here again Fernández-Armesto says that

“[p]erhaps you do not even have to admit the most basic demand of earlier theorists, that truth is a property of certain forms of words”.

Perhaps truth isn’t a “normal” property.

Again, I’m bouncing off Fernández-Armesto’s words here, not off the mountains of words which have been spilt on these subjects by philosophers. So it’s not that there’s a problem with bypassing reality or with truth being a property of certain forms of words: it’s that Fernández-Armesto doesn’t really explain these positions (which he’s against).

Certainly saying that truth is not a normal property doesn’t give us much… or even anything. In fact, if it isn’t a normal property, then what is it — an abnormal property?

However, one can guess at what certain alternatives are.

For example, truth is not a normal property because statements are true for very different reasons. In addition, true statements are very different from each other in both form and content.

Fernández-Armesto’s final statement on Horwich doesn’t help either, except to show his readers the motivation behind his own characterisation of Horwich’s position.

To explain that. The theme of relativism is strong in Fernández-Armesto’s book Truth: A History. And he concludes by saying that Horwich “is right to claim his theory as a way of escape from relativism”. Is it? Why? It’s because “by minimalist standards, all propositions are true or false, even if one can hardly ever say which is which”.

Fernández-Armesto on the Coherence Theory of Truth

Fernández-Armesto also tackles what analytic philosophers have called the coherence theory of truth.

The Bible could be deemed to be coherent within its own parameters (or language game). So too Islam, Catholicism, Marxism and fascism. Fernández-Armesto picks up on this and says that “religious philosophers [have] valued coherence as a presumed feature of the mind”. Indeed, he continues by saying that this is “especially [true of ] Catholics, whose religion was rebuilt into a perfect system by those visionary craftsmen — the Doctors of the Church”.

Despite being a Catholic himself, Fernández-Armesto isn’t happy with coherence alone when it comes to truth. He states the following:

“In the end, it just appeals to belief. A coherent system includes whatever fits into it; there is nothing left outside it by which to test its claims.”

The is a strong way of putting it, but Fernández-Armesto doesn’t name any names (i.e., apart from “religious philosophers”). So let me rewrite that passage:

Catholicism/Islam/Marxism/Nazism is a coherent system which includes whatever fits into it; there is nothing left outside it by which to test its claims.

In everyday terms, if anyone criticises Marxism to a Marxist, the Marxist will state that he does so because he/she is a member of the “bourgeoisie” or a “capitalist” or “brainwashed by the mainstream media”. If anyone criticises Neo-Nazism, they must do so because they’ve been infected by “Jewish thought”, “Zionism”, or “the mainstream media”.

There are other calls for consensus truth. Fernández-Armesto also picks out “the general will”, but we can also include “the will of the nation”, the “will of the working class”“the language game”, etc. too.

Conclusion

It’s worth stating again that the prime motivation Felipe Fernández-Armesto has for tackling what analytic philosophers have had to say on truth (as well as what many others have said) is to emphasise his argument (or his deep worry) that they’ve threatened both truth and reality with their philosophical theories. In that respect, then, Ludwig Wittgenstein, Donald Davidson and Paul Horwich are not that dissimilar to, say, Jacques Derrida, Michel Foucault and certainly Richard Rorty. (The latter started off as a analytic philosopher… and perhaps still is.) One may wonder, then, how different Fernández-Armesto’s own position is to something he himself wrote (and I quoted at the beginning of this essay) in his book Truth: A History:

“Traditionally, language has been presented according to the model of Genesis: God creates things, then names then.”

Thus, is it that philosophers too (to use a similarly naïve language) must firstly experience things, and only then name them? Isn’t this a crude way of expressing the philosophical position known as metaphysical realism?