The Modal Era in Analytic Philosophy

An attempt at a periodization of recent Anglophone philosophy

I spent last week in Tilburg, at two excellent conferences organised by the Exiled Empiricists project. One conference was on new narratives in the history of 20th century philosophy, and the other was on quantitative methods in the history of philosophy. I learned a lot from each, and they helped clarify a view I’ve been thinking about for a bit. The view is about what periods there are in analytic philosophy after the empiricist/ordinary language periods.

The short version of the view is that there is an identifiable era in philosophy from roughly 1970 to 2010, which could be called the modal era. What’s distinctive about the modal era is the use of modal concepts, understood in terms of possible worlds, and of counterfactual conditionals, as central tools in understanding philosophical problems.

It’s worth breaking down this view into three distinct claims, which are somewhat independent from each other.

1. There is an era that starts roughly in 1970.

2. Modality is central to this era.

3. This era ends in roughly 2010.

The first claim is I think the easiest to defend. Around 1970 there are field-defining works by Lewis (several times over), Kripke, Putnam, Rawls, Thomson, Singer, Frankfurt, and many others. In a longer version of this I’d back up this claim with citation data, but actually I think this is the easiest of the claims to defend.

Impressionistically, what’s striking about the period around 1970 is how many of the works from it feel contemporary. I can’t read the great works by Wittgenstein, Austin, Quine, Goodman or (perhaps) Anscombe without feeling that I’m looking at work from a prior time; work that needs care to put it in contact with present concerns. Obviously those works are older, but this isn’t just a function of age. I would have made the same distinction 15 years ago, when Word and Object was younger than Counterfactuals is now.

If all I had was this impression, it could just be a function of my age. Perhaps everyone thinks that ‘contemporary’ philosophy starts with the works their first professors thought were foundational to how philosophy was done at the time. One thing that I mean to spend a bit more time working on is seeing whether the citation data backs up my impression that there really was a break here, and it’s not just something that someone my age would think. My earliest look at the data suggests that perhaps the era starts earlier than I’d thought, perhaps more like 1967 than 1972. But I think there is going to be evidence for a break point around 1970, plus or minus a handful or years.

Onto the second claim. What I think in retrospect is most striking about this era is how much modal understandings of various philosophical claims were simply presupposed. The easiest illustration of this involves discussions of essentialism after Kripke. Here are two claims that are often attributed to Kripke:

• Origin essentialism

• The necessity of origins

I’ve said these are two claims attributed to Kripke, but that’s already a little anachronistic. The standard way philosophers from this era talked about Kripke on the metaphysics of origins was that these were two names for the same view. And, indeed, both of these views could be expressed as claims within first-order quantified modal logic, and (at least for most philosophers) those claims could in turn be helpfully understood using the ideology of possible worlds. So Kripke’s thesis about essentialism could be reasonably paraphrased as the view that for any object, and any possible world it exists, the object has the same origins that it actually has. Of course philosophers did not, with the exception of Lewis and a handful of his followers, take a particularly realist attitude towards possible worlds. But the broad idea, that origin essentialism just is a claim about necessity of origins, and that just is a claim about what possible worlds a thing can exist in, was very widespread.

The point is not that origin essentialism is universally accepted, because of course it wasn’t. Rather, the point is that this understanding of what origin essentialism is was (a) more or less universally accepted, and (b) something one could assume without comment or argument in philosophical works.

That should be surprising, because these are pretty clearly not the same view. This is easy enough to see from a contemporary perspective where essentialist claims are typically taken to be claims about grounding, and grounding is a hyperintensional notion. Even if singleton Socrates is necessarily a set with a member born from Sophroniscus and Phaenarete, it’s a further claim that it’s part of the essence of that set that it has a member born from Sophroniscus and Phaenarete.

It could also have been surprising from a perspective more familiar to people in the 1970s. At the new narratives conference, Rachael Wiseman argued that one of Anscombe’s aims in Intention was to understand the essence of intensions, in a way that takes seriously Wittgenstein’s dictum that “Essence is expressed by grammar.” Just exactly what that means is (like most things to do with Wittgenstein), a tricky question. But one thing it definitely does not mean, either for Wittgenstein or Anscombe, is that the properties common to a thing across all the possible worlds in which it exists are (somehow) expressed by grammar.

One of my projects is to look at how much the debates about materialism/physicalism have the same structure. Obviously in the period after 1970 there were debates about whether physicalism was even true, and about how to best express it. After Crane and Mellor’s 1990 paper there was even a debate about whether it could be expressed. But my sense (prior to having done all the relevant reading!) was that the following was a very widely shared presupposition: If physicalism is anything at all, it is some kind of necessitation claim. The case isn’t quite like essentialism, because there is a huge debate, especially in the 1980s, about just which necessitation claim it is. Still, I think the thought that physicalism must be some kind of necessitation claim was very widespread, often assumed without much comment or argument, and the acceptability of that assumption is another key characteristic of the modal era.

It isn’t just the use of possible worlds that characterises the modal era. At least as important is the use of counterfactual conditionals. After the development of the Stalnaker-Lewis semantics for counterfactuals at the start of the era, they became a standard part of the toolkit of philosophers. As well as being used in by then familiar theories of dispositions and free action, they are used to develop theories of rational choice (Gibbard and Harper), of knowledge (Nozick and, later, Sosa), and of mental content (Fodor).

The point is not that these theories received widespread acceptance. Rather, the point is that philosophers with otherwise very different backgrounds felt comfortable reaching for them as a tool of analysis. And, more importantly, objections to these theories tended to not be objections to the very idea of using counterfactuals. Instead, the focus was simply on whether the theories got the cases right. The relevant distinction here is even clearer when we look at perhaps the most important use of counterfactuals: Lewis’s various analyses of causation.

Lewis put forward three theories of causation, first in 1973, then 1986, then finally in 2000. Every theory was met with a flurry of objections and counterexamples. But there were three kinds of objection that, I think, were not particularly common. (To be clear, the point of this post is to announce a research program; I could simply be missing high profile versions of the objections I’m listing here.)

1. One might object, following Russell, to the very idea that we should be in the business of rehabilitating causation.

2. One might object that since causation is such a central feature of this world, it cannot be understood in something as otherworldly as counterfactual conditionals.

3. One might object that something as important to physical science as causation can’t be reduced to counterfactual conditionals, when they in turn are so sensitive to delicate features of linguistic content. Colloquially, you don’t have to take a graduate level semantics class to tell whether a physical explanation is successful, but if causation is understood in terms of counterfactual conditionals, you would have to.

My sense is that for most of the modal era, these kinds of objections were not particularly prominent. Rather, philosophers tended to accept that counterfactuals were the right kind of thing to use in a theory of causation, and then got onto the more detailed questions about whether Lewis had used them in the right way. And it was that more internal kind of objection to Lewis that led to so many discussions of stone throwing and outranking wizards and the like.

So that’s what I think is really distinctive of the modal era. There are these related tools, possible worlds semantics for modals and conditionals, which are (a) widely used, and (b) even more widely accepted as legitimate to use, so that objections to uses of them are mostly about questions of detail, not about the starting points.

Onto the third big claim then, that this era ends around 2010. I use this date as the end point because two things happen around this time.

The first is that, largely due I think to the publication of the Metametaphysics volume, Kit Fine’s objections to the modal understanding of essentialist claims become more widely accepted, and much more widely known. These objections were not new; Fine published "Essence and Modality" in 1994. But that paper was barely even cited for many years after its publication. It was only in the late 2000s that it became part of the wider conversation. And when it did, the grounding literature that it launched became central to metaphysics. At that point, the presupposition that one could simply equate essentialist claims with modal claims was gone, presumably forever. Of course one can still argue that an essentialist claim just is a modal claim, but now one has to argue for it, not simply assume it.

The attraction of using counterfactual conditionals had been fading for some time before that. There were I think two big forces at work here. One was that the various counterfactual analyses seemed simply not to be working. Even dispositional claims seemed to not be susceptible to counterfactual analysis. But the other was that there seemed to be a systematic reason why they wouldn’t work. This was something Robert Shope had pointed out right at the start of the modal era, and which Timothy Williamson stressed in Knowledge and Its Limits.

There is another reason that I think it makes sense to describe the modal era as ending around 2010. Around that time a new era plausibly starts: The Social Era.

Think about what else was happening in philosophy around 2010. Across all sorts of fields, there is a surge of interest in broadly social questions.

In epistemology we see this with the rising interest in questions about testimony and, after 2007, questions about disagreement. Also in 2007 Miranda Fricker’s book Epistemic Injustice raises the salience in analytic philosophy about questions at the intersection of epistemology, ethics, and political philosophy.

In metaphysics we see this with the rise of questions about race and gender. It wasn't like these weren't being discussed before; Sally Haslanger’s Noûs paper was published in 2000. But like Fine's paper, it was not an overnight sensation. It didn't start accumulating masses of citations until many years later, especially after 2010.

In philosophy of language a whole batch of questions about social philosophy of language (a topic that was barely recognised 10-15 years earlier) start to be central. Again, it's not that these topics weren't discussed. Rae Langton had important papers on silencing in the 1990s. What changes around the late 2000s is that other philosophers belatedly acknowledge this work. Around the same time there is a flurry of papers on slurs, and how different they are to other words.

Even in more formal areas of philosophy, the rise in interest in agent-based models, where differences between and relations between the agents, play important explanatory roles, fits this general paradigm. (I think there has also been a small rise in interest in game theory as contrasted to decision theory, but that’s smaller than the rising interest in agent-based modeling.)

Now in 2024 it’s probably too early to say whether this is a trend that will continue long enough to really be an era, and if so what will be its characteristic features. But I think there’s enough happening to suggest that this will eventually be a recognisably distinct era.

Even in moral and political philosophy, it feels there is a change from what was happening in the late 20th century. You might think it’s hard to see how moral and political philosophy can fail to be at least somewhat social. And to some extent that’s right; without interpersonal relations there’s barely a subject matter for these topics. But if the focus in ethics is on the nature of reason, and on the role of moral attitudes in moral psychology, we get a very individualistic approach to ethics. Similarly in political philosophy, if the focus is on rights, and on the consent of the governed, i.e., on whether each individual consents to be governed, we get a very individualistic approach to political philosophy. Approaches to moral and political philosophy that foreground equality and respect, and which take race and gender seriously, are more importantly social than approaches which focus on reasons, rights, motivation, and consent. The timeline doesn’t quite match up here; the move to more social approaches is underway well before 2010. But the trend is in the same direction.

So that’s why I think we can identify an era, the modal era, running from 1970 to 2010. There is a flood of field-defining works at the start of that era, as a result a distinctive set of attitudes towards modality become extremely widespread, and those attitudes stay in place almost until a new era, one where social questions become central to philosophy, starts up.

There is one big objection to this view that I worry about: it doesn’t have a place for the changes in political philosophy or moral philosophy that also take place at the start of the era. The whole story I’ve told doesn’t include Rawls, it doesn’t include Frankfurt, and it doesn’t include the turn to taking applied ethics at least a bit more seriously, as seen in the importance of works by Thomson and Singer. Surely our story of what happened should not just include them, but center them?

I don’t have a great answer to this objection, but I do have two things to say.

First, as important as those works are they don’t, apart from Frankfurt, show up nearly as much in the philosophy journals as the works that kickstarted what I’m calling the modal era.

Second, the story I’m telling does have a role of sorts for these papers, albeit not as prominent a one as I’d like. The thing about the modal era is that it signals the final end of the empiricist era that came before it, and which to be honest was definitely running out of puff by the late 1960s. One of the characteristics of the empiricist era, the characteristic that most moved critics like Anscombe, Murdoch, and Foot, was that it did not take seriously ethical questions. Now I’m not saying that Kripke or Lewis’s work on modality was itself moral philosophy. But a philosophical paradigm where those works are central is not quite as hostile to ethics as Anscombe and others felt the empiricist era was.

Is that enough to say about the importance of Rawls, Thomson, etc to the new era in philosophy? It still feels like it’s missing something. But attempts to force Rawls into a modal paradigm feel seriously artificial; I think I have more work to do here.

Still, I think there is an interesting view here, and I’m very interested in knowing what others think about its plausibility.