Problems with Modal Realism

By Michael Gakuran | | Philosophy | 3 Comments |

owlOkay, so we’re well under way. We have a theory of Possible Worlds, which looks nice and really helps us (okay, perhaps just Philosophers) talk about the world more clearly with one another. Sounds good, but it comes at a heavy price. Here are some of the big objections to Lewis’s Genuine Modal Realism.


It doesn’t fit with common sense!! As we saw last time, if we accept Lewis’s GMR, we are forced into believing such absurd entities as pixies and vampire ducks with fins and fur that go moo in the night really exist in another world. And the price we’re paying? A lucid account of modality and an easier way to think about possibility. Lewis heads this argument off by sticking his ground and saying that the benefits are worthwhile and the cost not so great as to discard the theory. Furthermore, he attacks the alternative theories that I will examine later by saying they yield the same, if not worse conclusions than his own theory. Thus, we should accept his theory as the most plausible.


I quite imagine you may be thinking (as am I), *exactly* why Lewis’s account helps us so much. I have repeatedly mentioned clear and lucid accounts of modality and things such as helping us to explain possibility, yet failed to clearly elaborate on this. The explanation lies deep in the heart of modal logic and the way philosophers try to systemise our language and grammar when we make claims such as:

‘There could be a world exactly as ours but where people are like zombies’

or similar claims that are heated debates in Philosophy. (The problem being, do mental properties [such as conciousness] supervene – depend on – the physical properties [such as the structure of our brains]?) Unfortunately I haven’t the background knowledge nor time to learn modal logic right now. You may find more information here however:

Possible Worlds: What they are good for and what they are

Considering how widely used Possible Worlds have become to explain difficult concepts in modern Philosophy, I think it is safe to say that the idea of Possible Worlds holds real value to us in organising our thoughts and language. We would like to hold onto it, but are perhaps reluctant to admit such bizarre consequences.



One of Lewis’s main tenets in his thesis was that of ‘Plenitude’, roughly that every possibility is realised at some world. Furthermore, we can take any parts of these worlds and put duplicates of them into another world (providing they fit together). For example, we have a horse, horn, fangs, pink all over, red all over and a dragon. I could take any number of these things, recombine them and there would be a possible world containing duplicates of them. E.g. (World 1) A pink all over horse with a horn (i.e. a unicorn) and a red all over horse. (W2) A red all over unicorn and pink all over dragon with a horn and a red all over dragon. (W3) A pink all over horse with fangs, 3 pink all over dragons with horns and fangs. (W4) 27 red all over horses with fangs, 5 red all over dragons, 2 pink all over dragons with horns and 17 pink all over unicorns with fangs. (etc). You get the idea. Of course, I’m assuming for simplicity that each thing cannot be further broken down, but we can easily imagine worlds where there are horse-dragon hybrids and such.

A impossible world would be one such as a pink all over and red all over horse. (One thing cannot be both completely pink and completely red!)

Hopefully you should now start to see the problem. We can imagine many different possible worlds just from these few parts recombined, but now remember how immensely varied our world is and how many things could be recombined to create new things in any possible combination (bar impossible combinations, things shaped too awkwardly to fit – etc)

The argument then, is that according to Ockham’s Razor, modal realism postulates too many entities. (Ockham’s Razor is the idea that one should not multiply entities beyond what can be expressed more simply. I.e. “All other things being equal, the simplest solution is the best.”) In general, this a a good principle to stick to, as we prefer explanations that are simpler and more concise. This is not to say the simplest explanation is always right however.

In this case, Lewis’s theory suggests there are an extreme amount of possible worlds and recombinations. Clearly this new theory of possible worlds is asking us to believe in a lot of new, bizarre things… According to Ockham’s razor, we shouldn’t accept it.

Lewis replies that these things are merely a great deal more of what we already believe in. In other words, they are the same *kind* of thing as what we find the actual (our) world. It’s perhaps a little tricky to understand, but because Lewis is arguing the other worlds are all real and concrete (i.e. not abstract), they are like ours in terms of the *kinds* of individuals they possess. An abstract world, in contrast, would contain different *kinds* of thing, since they are not things we can readily relate to. (For a very crude image, try imagining a physical object in this world and a physical object in another world, then contrast that to a physical object in this world and some ghostly object in another world.)

What this amounts to is the controversial difference between quantitative and qualitative parsimony (simplest assumption). Quantitative parsimony would say not to multiply the number of new entities (again crudely, 1000 electrons is better than 10,000 electrons), whereas qualitative parsimony would say not to multiply the number of new kinds of entities (concrete and abstract entities, for example). Thus, what Lewis argues is that his theory is only quantitatively unparsimonious and he says quantitative parsimony is controversial and not a problem.

Even if we are to accept this however, it can be shown that Lewis’s theory is also qualitatively unparsimonious. Because of the principle of Plenitude and recombination, Lewis ends up postulating new kinds of entities too, and many, many of them. For example, there is a possible world where crows philosophise. This is arguably a different *kind* of entity to anything we have in the actual world.

It is also possible that these other possible worlds contain alien properties (2003, Melia). Such alien properties are beyond what we can imagine, but in Modality, Melia offers a good example to illustrate this point. Imagine a world just consisting of only bodies that had a mass and obeyed Newton’s laws of motion and gravity. In such a world, our properties of, for example, charm or charge would be utterly alien and something that could not be constructed or analysed by rearranging the Newtonian bodies. It could be argued that possible worlds can contain alien properties (new kinds of thing), further rendering Lewis’s theory guilty of being qualitatively unparsimonious.

    Island Universes

Another problem Lewis has is the notion of island universes. Remember the tenet of Plenitude, that duplicates of individuals can be recombined in many, many different ways. And the Isolation tenet: no individual, nor a world (collection of individuals) can be in any spacial, causal or temporal relation to any other (a world cannot somehow connect to another world).

But through Plenitude and recombination, we could imagine a world (Universe) which contains duplicates of all other worlds. But if all worlds are spatio-temporally isolated from one-another, how can we have a single world which contains duplicates of other spatio-temporally isolated worlds? Furthermore, if there is such a world that contains duplicates of all other worlds, then that huge world itself is another individual, and so it would have to contain itself to contain *all* possible worlds. This is logically impossible.

Lewis is likely to reply by pointing to the Plenitude tenet that states ‘providing they are not too large, numerous or shaped awkwardly to fit’. This would be such a scenario that wouldn’t work.


I can imagine myself passing (or failing) my philosophy exam in another world. But as Lewis argues, as I am an individual I cannot be in another world. The solution then, is to say that a ‘counterpart’ very similar to me exists in another world. If we say this though, what could I care if a counterpart of mine actually passes the philosophy exam? Thus we have the problem of concern that Kripke puts forward. Lewis’s notes that this problem is also applicable to theories alternative to his own such as Ersatz Modal Realism. In contrast with a concrete counterpart of myself as Lewis would argue I have, why would I care if an abstract version of myself passed the exam (as EMR would argue)? I’ll go into EMR soon.

The alternative view is to hold that I exist in many different worlds, that is, I have ‘transworld identity’, which clearly has a problem fitting into Lewis’s theory as it would entail the overlapping of worlds. Aside from seeming to go against common-sense, the idea of me existing in many different worlds has problems, says Lewis. For an individual to exist in some world, the whole of it must be among the parts of that world. But for a transworld individual to exist, in some sense it must have its parts distributed across many worlds, which is a contradiction.

Of course, the notions of counterparts and transworld identity is a large topic in itself and there are other arguments both for and against each one.


‘There is a possible world at which there is nothing’.

Isn’t this an inherent paradox? According to Lewis’s theory, where every possibility is realised, it would seem this should be too. But a world (Universe) with nothing in it would still have something, namely the shell of the world itself. I.e. the only thing that would exist would be the vast space-time of the world, which I would say has to count for something at least.

I imagine Lewis’s answer to this might be to again appeal to the line ‘providing they are not too large, numerous or shaped awkwardly to fit’ in the Plenitude tenet. Evidently nothingness is not possible, which, even if you had been content to allow unicorns, pixies and vampire ducks into your ontology up until now, is a further commanding move in need of much thought. If you are to accept Lewis’s theory as is, you would have to believe that the actual universe *must have been*. It is not possible that if the ‘Big Bang’ had not have happened, we would not be here. Could there have been nothing? According to Lewis’s theory, it seems not.


Another problem, and one for Philosophy at large. As the word suggests it is a central part of the sub-section of Philosophy: Epistemology. The problem is also very prominent in the philosophy of mathematics, known there as ‘Benacerraf’s Dilemma’. How can we know about these possible worlds? Where do we get our knowledge of them? Indeed, how we can know anything at all is a huge problem to sort out in itself, but Lewis gives a reasonable reply (1986, Lewis. P109):

“Our knowledge of mathematics is ever so much more secure than our knowledge of the epistemology that seeks to cast doubt on it. So mathematics will do as a precedent: if we are prepared to expand our existential beliefs for the sake of theoretical unity, and if thereby we come to believe the truth, then we can attain knowledge. In this way, we can even attain knowledge like that of the mathematicians: we can know that there exist countless objects causally isolated from us and unavailable to our inspection.”


So, with some problems outlined with a couple of responses thrown in, I’ll move onto considering variations to Lewis’s theory. I don’t pretend to have done any justice to the debate here, as there are many more subtleties, problems and responses than I can address at this time. Next then, what Lewis calls ‘Ersatz Modal Realism’ – trying to get the benefits of Genuine Modal Realism ‘on the cheap’. If an alternative theory can be shown to be more parsimonious that Lewis’s theory, we would do better to accept that one. So without further ado, let’s see what’s on offer.


Resources: Handouts from my lecturer, Dr. John Divers (writer of ‘Possible Worlds’ (2002)), ‘Realism and Anti-Realism’ (2007) by Stuart Brock and Edwin Mares, ‘Modality’ (2003) by Joseph Melia and ‘On The Plurality of Worlds’ (1986) by David Lewis

Page 1 – Realism and Anti-Realism – primer

Page 2 – Possible Worlds

Page 4 – Ersatz Modal Realism

3 comments on “Problems with Modal Realism
  1. IK says:

    Really nice summary here!
    You might enjoy reading (if you haven't already read) 'Modal Fictionalism' by Rosen (and looking at the response from Bob Hale!)

  2. IK says:

    Really nice summary here!
    You might enjoy reading (if you haven't already read) 'Modal Fictionalism' by Rosen (and looking at the response from Bob Hale!)

Leave a Reply

Your email address will not be published. Required fields are marked *


Get the Gakuranman Newsletter!

Greetings, fellow Adventurer!

For a limited time, subscribe free and get:

Just enter your name and email below: