My Dissertation in 2,000 Words

I’ve been asked several times to make a video about my dissertation. I do not think it would make for a good video, but I do understand why some people are interested in the research I did before I became an internet philosopher.

This post is a short explanation of my dissertation research. You can read my full dissertation here, and you can read some portions of it that were published in Thought and the Journal of Philosophical Logic. The second linked paper has generated more discussion in the literature than I expected, including a response from Poppy Mankowitz in Mind.

Source

My dissertation explores what we mean when we say that something is true — and more precisely, what it means when ordinary speakers of English use the word ‘true.’ It is common in the literature to make a distinction between three kinds of theories of truth:

1. A theory of the word ‘true,’ also called the truth predicate.

2. A theory of the concept of truth, which I’ll call TRUTH.

3. A theory of the property of truth, which I’ll call truth.

To clarify the scope: my dissertation is largely concerned with (1) and (3). While my theory may have applications to (2), I did not address this issue in my work.

Contemporary debates still operate in the shadows of logical positivism. A.J. Ayer put forward a theory of the truth predicate in which it was logically superfluous. He wrote in Language, Truth, and Logic:

Thus, to say that a proposition is true is just to assert it, and to say that it is false is just to assert its contradictory. And this indicates that the terms ‘true’ and ‘false’ connote nothing, but function in the sentence simply as marks of assertion and denial

This is the predecessor to views that have come to be called deflationist theories of truth. Deflationism is a house with many rooms, but deflationists will commonly say that:

1. The truth predicate is a merely logical operator, like the words ‘and’ and ‘if.’

2. The concept of truth does not explain any other concepts, in whatever sense concepts explain other concepts.

3. There is no substantive property of truth.

To set the terms for the debate, my dissertation focuses on what I call Core Deflationism:

The view according to which (i) the truth predicate plays a logical, quasilogical, or merely expressive role in our target language, and (ii) the truth property and truth concept play no explanatory role in our theories.

If you deny either (i) or (ii), then I would say you are not a deflationist. This would make you a substantivist about truth. My dissertation is concerned with whether or not Core Deflationism is, in fact, true.

There is one more distinction we need to make to understand the rest of the dissertation: between descriptive and revisionary theories.

Descriptive theories of truth aim to describe how we use the concept of truth, take the property of truth to be like, or use the word true in some domain. Descriptive theories can be further divided into what I call folk descriptive theoriesand theoretical descriptive theories. A folk descriptive theory of, e.g., the property of truth would describe what the folk – non-philosophers, in their everyday lives – take truth to be or to do. A theoretical descriptive theory of the property of truth would describe what truth is taken to be in some theoretical domain.

But philosophers do not only have to describe the world or how we think about it. They can also make proposals for how we ought to think about the world. That brings us into the territory of revisionary theories.

By my lights, a revisionary theory can best be thought of something like a conditional assertion: If we hold theoretical desiderata …, then we ought to think that truth is …. Revisionary theories are like proposals, and they rely on taking for granted theoretical goodsand theoretical goals.³ I think of revisionary theorizing as a rather liberal enterprise with very few external constraints. Success is measured only relative to the theoretical goods and goals, and without agreeing on goods and goals we cannot expect agreement on a revisionary theory.

I am primarily interested in descriptive theories. That means I can help myself to things like linguistic evidence to see how people use the word ‘true.’

My dissertation consists of seven chapters:

1. Scope and Methodology, which was explained above.

2. Deflationary Semantics, which explores whether or not deflationism is compatible with truth-conditional semantics as practiced in linguistic departments. I conclude that it is, despite the overwhelming consensus amongst philosophers (both deflationists and substantivists) that it is not. This is a point in favor of deflationism.

3. Deflationism and Dependence, which explores the so-called dependence intuition, the idea that truth depends on reality. I argue that deflationists cannot accommodate the dependence intuition, so they should aim to debunk it. I call this the logicality strategy. I say that truth depends on reality in the same way conjunctions depend on conjuncts — which is such a thin sense of ‘depends’ that the deflationists should not worry about it.

4. True as a Gradable Adjective. This chapter is my primary argument against deflationism, and so I will explain it more below. In short, I argue that ‘true’ is a gradable adjective, denoting a property that comes in degrees.

5. Degrees of Correspondence, in which I argue that we can construct a correspondence theory of truth that allows for degrees of truth. I develop this by looking to how Aristotle, Aquinas, and Frege spoke of correspondence – correspondence to objects, not facts – which means that I can have a theory of correspondence that does not need an ontology of facts, too.

6. Pluralism without Absoluteness, in which I argue against the platitude that pluralists about truth (those who think there are many truth properties, often domain-specific) call ‘Absoluteness’ — that truth does not admit of a more or less. I argue against the common arguments for Absoluteness, and I say that given what they say about truth (such as comparing truth to winning), they should reject Absoluteness and admit that some truth properties may come in degrees.

7. Modal Measurement Theory. This is my attempt to provide a theory of truth that is more deflationary than what I developed in chapters 4-6, and I propose this as an alternative theory. This treats the truth predicate as measuring distance between possible worlds.

This dissertation was written between 2015 and 2019, so it has been seven years since I defended it. I don’t know how much of it I still believe. For instance, in Chapter 7, I write:

The two theories I have fleshed out in this dissertation – the gradable truth theory of Chapters 4 & 5 and the modal measurement theory of the present chapter – each have their own theoretical virtues. The gradable truth theory adequately analyzes the way that we ordinarily talk about truth, as it contains a semantic theory which utilizes degrees to handle gradable truth-talk. The gradable truth theory also encompasses a metaphysical theory that allows for truth to be a real property that truthbearers can have more or less of. Thus, the semantics has a metaphysics and the metaphysics has a semantics. In this sense, the gradable truth theory is a broad account of truth, encompassing both predicate and property. For those with substantivist leanings, the gradable truth theory should be a viable contender — especially if our truth theory is thought to be beholden to way we ordinarily usetrue. But to those with deflationary leanings, the gradable truth theory is another instance of ontological extravagence. The modalmeasurement theory has the benefits of other deflationary theories, int hat it posits no substantial property of truth. Some gradableconstructions–comparativesandequatives–are explained, though others arenot. The modal measurement theory has a minimal, deflationary ontology, but it buys this ontology at the cost of failing to explain some linguistic data. How are we to choose between these theories?

Put bluntly: we do not need to choose. Theory choice can only properly take place when the theories are competitors within the same project. But the way that I have put forward both theories should make it clear that there is no real competition. The gradable truth theory is a project in natural language metaphysics, a kind of folk descriptive theory in the terminology of §1.2.2. The modal measurement theory is a revisionary theory. It is not intended to describe how we ordinarily think or talk about truth, or how we take truth to be. It is a theory of how we might want to think about truth, provided that we hold certain theoretical goods and goals fixed.

These goods and goals might include: a minimal ontology, an avoidance of representational or semantic notions when possible, the ability to compare theories and claims, an explanation of seemingly alethic phenomena in terms of well-understood philosophical notions, and a straightforward way of meeting the constraints articulated in §7.2. Given these goods and goals, the modal measurement theory is an attractive theory of truth — or, perhaps better, an attractive theory of degrees of truth that strictly speaking is not committed to there being a property of truth. Thus on the modal measurement theory, there are degrees of truth (worlds, ordered by similarity to the actual world) but there need not be a property of truth.

As conclusions go, this is a bit of a wet fart. I now think that I should have spent more time on the question of whether or not MMT could be developed in such a way as to be explanatorily adequate for the ordinary usage of ‘true,’ which would mean it would be a competitor with the substantivist theory of truth I developed in Chapters 4-6.

---

Now, let’s look back at Chapter 4, where I present my argument against deflationists. This will necessarily be brief; I encourage you to consult the paper (linked above) published in the Journal of Philosophical Logic for a full explanation. (I posted a PDF prepint on PhilArchive, making it available to all.) I also won’t include formal notation, as the LaTeX editor on Substack isn’t great.

My argument hinges on the fact that folk usage of the truth predicate are more diverse than is typically admitted. We say things like:

1. That is only half true.

2. General relativity is more true than Newtonian mechanics.

3. What John said is absolutely false.

These aren’t what I would call ‘philosopher sentences.’ Philosophers find them quite weird! But if you look at corpus data, you’ll find examples. In my paper, I give examples from the Washington Post, The Atlantic, and other news sources, working on the assumption that these reflect ordinary English usage more than philosophy papers.

If these are good bits of English, this means we use the truth predicate as a gradable adjective. We treat it like we do the words ‘tall’ and ‘full.’ Linguists have identified markers of gradability, like having an antonym pair, being used in comparative constructions, and taking modifiers like ‘a little’ as evidence of being a gradable adjective. I say that the truth predicate possesses all of these markers. Thus, the truth predicate is a gradable adjective. In my dissertation, I develop the semantic theory of the truth predicate as a gradable adjective and defend the analysis against objections.

Looking back, I still think I’m right. The truth predicate is a gradable adjective, and it denotes a property that comes in degrees. This is a radical claim, but I believe it. This means that Core Deflationism is false, as the truth predicate has a role beyond its logical function.

This opens up many interesting questions. How does gradable truth relate to knowledge? Assertation? The concept of evidence? What historical theories of truth were gradable? All of this would have been fruitful areas of research — but I left academia, and then my interests shifted, so I never worked on those open questions.