What Is It?
It consumes energy and emits electromagnetic radiation. It contains a small wire filament. It is widely used all over the world. It was invented by Heinrich Göbel in 1854, though Americans often attribute its invention to Thomas Edison. What is it?
The electrical light bulb, of course.
But hold on. Is there really something that satisfies these conditions? What kind of thing would this be? It can't be any particular light bulb, say, the one in my bathroom. For this light bulb is used only in my bathroom, not all over the world, and most Americans don't even know that it exists. Nor can it be any other particular material thing. Nor can it be a mental object, something like the idea of a light bulb: ideas don't contain small wire filaments. This alleged thing, the light bulb, is a very strange kind of object. It is not a light bulb (all light bulbs are concrete, particular light bulbs), but like all light bulbs it contains a wire filament, consumes energy and emits electromagnetic radiation. It is is located in time (as it didn't exist before 1854), but presumably not at any particular location in space.
So reconsider: is there really something that satisfies those conditions?
If not, we have a problem. On the one hand, we can hardly deny that Göbel invented the light bulb, that the light bulb consumes energy, that Göbel invented something that consumes energy, that there exists something that Göbel invented which consumes energy and emits electromagnetic radiation, and so on. On the other hand, this seems to be exactly what we are denying.
Here is one way out: "the light bulb" denotes a certain type or kind. This type is an abstract entity. It doesn't really contain a wire filament, and it doesn't emit radiation. But it was indeed invented by Heinrich Göbel in 1854, and is widely used all over the world. When we say that the light bulb contains a wire filament, we're speaking loosely. What we mean is that every light bulb contains a wire filament. We don't talk about the abstract type or kind at all, but only about its concrete instances.
Once we're on this path, where we deny that the light bulb satisfies all the mentioned conditions, we might as well go further. For instance, we might say that the light bulb type is just the set of all possible light bulbs. This set satisfies yet fewer of the conditions. It wasn't in any literal sense invented by Göbel, nor is it in any literal sense used all over the world. So it hardly satisfies any of the conditions. Thus we shouldn't call it "the light bulb". A better name is "the set of all possible light bulbs". On this suggestion, expressions like "[the] light bulb", "[the] radio", "penicillin", "[the] tiger", etc. are not referring expressions. (A fortiori, they are not directly referring.)
Why not just say the guy invented the idea of the light bulb and that individual lightbulbs are applications or instantiations of the idea? It seems this avoids talking about sets and yet doesn't necessarily commit one to Platonism. Further this lets us deal with cases where some of the normal constituents of the idea aren't present. (i.e. the requirement of a wire in the fashion of a typical bulb as in a fluorescent light bulb - and yes I know there are some wires in those) The point is that ideas can have a causal relationship on new ideas and in a sense still be "present" in them and thus still instantiated.