Episode 4 – On the Possibility of Unicorns explores concepts in the philosophy of language discussed by Saul Kripke in his famous work, Naming and Necessity. This Explainer lays out the portion of Kripke’s argument that leads to the counterintuitive conclusion that fictional creatures such as unicorns are not metaphysically possible objects.

Core Concepts

Reference

Names refer to objects. Sounds simple, right? “The Will to Power Hour” refers to this awesome podcast. “Meg” refers to my cute wittle beagle puppy. And “Richard Nixon” refers to the U.S. President who resigned in disgrace following the Watergate scandal. The point is, names pick out objects. But how exactly do they do this? And what really is the meaning of a name? Theories of reference attempt to answer these questions.

Possible Worlds

When philosophers talk about possible worlds, they’re not talking about the same thing physicists talk about when they refer to multiple or parallel universes. Philosophers are talking about counterfactual situations. It’s quite possible that instead of writing this blog post, I could have just spent my time watching a movie. In other words, there’s a possible world in which I am watching a movie and not writing this blog post. Of course, a number of other facts would differ as well such as my mental state, my location, and the electrical properties of my computer and TV. All of these facts together comprise a possible world. Possible worlds are very useful tools for analyzing counterfactual, or “modal,” statements for a number of reasons. However, it’s not entirely clear what exactly a possible world is – is it an abstract object, a set of propositions, or something else entirely? We’ll explore the nature of possible worlds in depth in the next episode. For now, you just need to know the basics, namely that a possible world is a counterfactual situation.

The Argument

Millianism and Descriptivism

John Stuart Mill said that the meaning of a name just is its referent, the object itself. In other words, the meaning of “Meg” is simply my dog herself. I’m guessing most of you probably think something along these lines, as it’s pretty intuitive. But, to use Kripke’s example, it creates problems for sentences like, “It was a scientific discovery that Hesperus is Phosphorus.” Hesperus and Phosphorus were just the old words for the planet Venus. For a while, people thought that Hesperus, which was the brightest object in the evening sky, was a different object from Phosphorus, which was the brightest object in the morning sky. But after a lot of careful observation and mathematical calculations, it was discovered that they were in fact the same object, i.e. the planet Venus. So the sentence “It was a scientific discovery that Hesperus is Phosphorus” should be true, as indeed we discovered this fact scientifically, by looking at the world. But if, as Mill claims, the meaning of each of those names is simply the planet Venus itself, then that sentence actually comes out false, since it translates to “It was a scientific discovery that Venus is Venus.” This translated version is false because it wasn’t a scientific discovery that Venus is Venus, but a logically necessary fact. But clearly the original sentence should be true! We discovered that they were the same object scientifically.

Problems like this motivated philosophers and logicians like Gottlob Frege and Bertrand Russell to posit various forms of descriptivism. Unlike Mill, who thought that the meaning of a name just is its referent, descriptivists claim that names are really synonymous with descriptions. So “Hesperus” really just means “the brightest object in the evening sky,” while Phosphorus really just means “the brightest object in the morning sky.” So our sentence about the scientific discovery actually translates to, “It was a scientific discovery that the brightest object in the evening sky is the brightest object in the morning sky.” Voila! Problem solved. When we replace names with descriptions, our previously problematic sentence is completely unproblematic: there’s nothing wrong at all with saying that we discovered scientifically that the object we see in the evening sky is really the same object we see in the morning sky.

But not so fast! What if we imagine a counterfactual situation or possible world in which Hesperus is not, in fact, the brightest object in the evening sky? This seems perfectly possible. There easily could have been another, brighter object. But in that case, how could we say something like “Hesperus is not the brightest object in the evening sky”? Remember, descriptivism tells us that sentence would really mean, “The brightest object in the evening sky is not the brightest object in the evening sky,” which is a contradiction. And if you say that now the description changes to “the second brightest object in the evening sky,” then the sentence is entirely uninformative and true by definition (of course the second brightest object isn’t the brightest object). So if neither Millianism nor descriptivism is true, then how the hell do names work?

Rigid Designation, Baptism, and Transmission

As you can see, when we think about possible worlds, descriptivism starts to fall apart. Problems like this highlight a feature of names that both descriptivism and Millianism leave out. Names are what Kripke calls “rigid designators”; the name “Hesperus” rigidly designates, or picks out, the very same object in all possible worlds, namely the planet over there yonder. Descriptions can be used to initially fix a reference, but the meaning of a name CANNOT be a description because descriptions are merely contingent features of objects that differ across possible worlds.

So how do names originate? Initially, a description might be the only thing that can fix a referent. For example, the potential Planet X is only known by its effects on the orbits of objects out in the Kuiper Belt. We have never directly observed it and can only refer to it as the thing with these gravitational properties. It’s possible that it’s not a planet at all. All we know is that the orbits are not acting as they should and that something or many things are likely causing disturbances (it is looking more and more likely it’s a specific object, but it might not be). So let’s say we find out that yes, it really is a planet (say, we catch a glimpse of it or someone develops an incontrovertible proof of its existence). Then we can point it out as an object; we can name it. Kripke calls this process “baptism.” An object is “baptized” and the name is then transmitted through a community of speakers. Once baptized, its description becomes dispensable. For example, there’s a possible world in which Planet X is bigger than it is in the actual world or in a completely different part of the solar system. But it’s still Planet X.

Natural Kind Terms, Essential Properties, and Unicorns

Now we can FINALLY talk about unicorns! Kripke says that natural kind terms – terms like “tiger” or “gold” that pick out kinds of things rather than individuals – act just like names. In short, all the thought experiments above work out just the same when we use species terms instead of names for individuals. For instance, we could imagine a possible world in which tigers are a different color, but they’re still properly called “tigers.” So the meaning of “tiger” can’t simply be a description.

Like “tiger,” “unicorn” is a putative species term. But there’s a key difference between tigers and unicorns, and it has to do with essential properties. Kripke believes that essential properties exist. And he seems to think that DNA and origin are essential properties of any species. Therefore, if a tiger-like animal existed in some possible world that had radically different DNA or origins than actual tigers, then that animal would not really be a tiger.

The problem with unicorns is two-fold. First, we lack a sufficiently detailed description of unicorn DNA and origin. All we have is an outward appearance. So we can imagine possible worlds in which multiple similar looking horse-like creatures with horns exist but have radically different DNA and origins. None of those candidate species would be the unicorn species. There simply would be no fact of the matter. This is the so-called “metaphysical thesis.” Second, even if we found fossils of unicorn-like creatures in the actual world, they would only be unicorns if the description in the myths were about THOSE creatures, i.e. if THOSE creatures were baptized as unicorns. This is the “epistemological thesis.”

Ergo, unicorns are not possible objects. There is no possible world in which they exist. There are worlds in which horse-like creatures with horns exist, but they are NOT unicorns. The most we can say about unicorns as such is that they are abstract objects and they only exist in that sense.

Lingering Questions

It seems that if we specified a unicorn’s essential properties, that should be enough to turn “unicorn” into a rigid designator. Remember, an object has its essential properties in all possible worlds. These are the properties of an object that cannot be otherwise. If you were to imagine an object with different essential properties, you’d really be imagining a different object. So if we specified a unicorn’s essential properties, then it should be the case that any object that has those properties in some possible world would rightfully be called a unicorn. This creates a problem for Kripke because it suggests that some form of descriptivism must be true and his account about how objects are baptized is either incomplete or false. Or perhaps I’m just missing something. What are your thoughts?