Le Petomane performed at the Moulin Rouge in the 1890s to the great and good of the day, such as Sigmund Freud, the Prince of Wales, and the King of Belgium. His art was musical farting. He perfected musical farting into an art form. And there are still musical farting performances today. There is an Austrian duet, for example, who dress in formal evening wear with long coat and tails, but the trousers have a discrete hole through which the instrument can be played.

But what exactly is a fart? There are two ways of thinking of farts, or two concepts of farts. A fart is either a thing or an act. We fart farts. Let us begin with farts considered as things. What are the identity conditions of fart-things? Are they proper objects – reidentifiable particulars? Suppose someone emits a series of farts in rapid succession, like a staccato trumpet call. Is that many distinct farts? But in a nearby possible world in which the emission of gas was more legato with no breaks, would there only have been one long fart, with pulsations in volume (in both senses)? Is that legato fart the same fart as the series of staccato farts?

What about the origins of farts? Once one thinks of farts as objects out there in the world, it is natural to think that they have essential origins in bums. If particular farts have essential origins in particular bums, that partly constitutes them being the very farts that they are. It follows that something molecularly identical to a fart that does not have a bum origin, such as gas produced by rotting grass, is not a fart. Molecularly similar gasses might differ in that one was a fart and the other not. And two molecularly identical farts that have their origins in different bums are not the same fart. Furthermore, this yields a clear view of the relation between the two ways of thinking of farts. Farts must be farted. A fart is not a fart if it is not farted. That is, a fart-thing must have its origin in a fart-act.

However, this way of thinking of farts may be questioned. Suppose that two people fart in a lift, and the gaseous molecules mingle perfectly. How many farts are there? On the essential origin view, there are two farts occupying the same space and the same time. However, phenomenologically – which is the perspective of the other passengers in the lift – there seems to be only one fart in the lift, even though it is the product of two bums. The two fart-things combine to form one terrible compound fart-thing. (One recalls Aristotle's theory of mixtures.) But if fart-things have essential origins in bums, then there are two things occupying the same space and time, and two such farts cannot fuse to form one super-fart.

The lift bystander experience surely deserves our respect as well as our sympathy. It invites a view that is radically different from the essential-bum-origin view: it is the view that farts are phenomenologically constituted. If so, there is after all only one fart in the lift, despite its dual bum origin, in the way that two French perfumes might be mingled to produce a subtle third kind. On this view, farts are phenomenal objects (in both senses of phenomenal).

So, we have a conflict between two views of the metaphysics of farts: the essential-bum-origin view and the phenomenological view. On the essential-bum-origin view, we smell farts, whereas on the phenomenological view, we fart smells. This metaphysical difference implies a difference over the identity conditions of farts and thus over how farts should be counted. Consider again the two people who fart in a lift, and the offending molecules that mingle. On an essential-bum-origin conception, the bystanders might think that they smell one fart, but in fact they smell two farts. They are ignorant of an important fact about what they smell. By contrast, on the phenomenological conception, a fart is a smell, and however it came into being, there is just one of those in the lift.

The divergence over the metaphysics of farts bears on the question of whether joint smelling is possible. Consider again someone – a perpetrator – who farts in a lift. Everyone else in the lift – the bystanders – can smell the fart. Each of them can smell it. But do they smell it together, as a joint act? It might seem so, especially when we consider that they express their disgust to the others in facial expressions, hand-waving movements and verbal behaviour. So, it might be said that just as we can walk together or watch something together, so we can smell together, which is what the people in the lift do. But in fact, whether this is so depends on the metaphysics of farts. The conception of farts according to which they have essential bum origins, allows for joint smelling since there is a mind-independent thing – the fart – on which our olfactory attention can converge. But a phenomenological conception does not allow for that. On such a conception, smelling could not be a joint activity; there could be no smelling together, although many people might have similar olfactory experiences given a similar stimulus. Visual attention to physical aspects of the world can be joint attention precisely because there is something that is distinct from our experience on which two or more people can triangulate. Visual attention to physical features of the world is thus unlike smelling farts if the phenomenological conception is correct. On that conception, each bystander is cut off from the other suffering bystanders in the lift. There is no joint awareness of an awful presence in the lift.

In order to try to break the stand-off between these competing conceptions, let us first consider whether farts can be better or worse than they seem. A powerful intuitive thought we have is that if they smell bad then they are bad. Perhaps there are some framing phenomena in our sense of smell such that how things smell depends on surrounding smells or on our beliefs about what we smell. Even so, it seems that there is no phenomenology-independent real smell. But on the essential-bum-origin view, whether farts smell bad could come completely apart from whether they are bad, which seems absurd. Surely, this is implausible, for there could not be farts that in fact are far better or worse than they smell? This view, so one might say, is so much hot air.

A second consideration in favour of the phenomenological view is this: we all know that there can be silent farts – farts that are notoriously ‘deadly’. Those farts are very real. There is no escaping them since they come without a sonic warning. But can there really be odourless farts? On the essential-bum-origin view, there could be odourless farts. But this seems absurd. A mere emission of odourless air from a bum is no fart. Such a release of odourless air molecules is surely a fart without a fart. That is, it is a fart-act that does not yield a fart-thing. A fart-thing must proceed from a fart-act. But a fart-act does not necessarily produce a fart-thing.

Thus, while the opposition between the essential-bum-origin school and the phenomenological school is something of an antinomy, since there is something to be said for both sides, there are in fact at least two considerations that favour the phenomenological school.

We are led to an outlook similar to Descartes’s view of the mind: on the phenomenological view, the essence of a fart is given to us in our olfactory experience. A fart has no hidden essence over and above, or underlying, what it smells like. Smelling farts reveals their essences. If so, farts do not have essential bum origins. Farts just are smells. And if that is unintuitive, it is just the price to pay for retaining an intuitive connection between farts and smells. What farts are, their essence, is unfortunately revealed when we smell them. Smelling a fart is a revelation. A fart, or a series of farts, may be accidentally musical, but they are essentially smelly. Since we grasp the complete essence of farts in smelling them, there can be no empirical science of farts that corrects our grasp of their phenomenological essence.

For the unfortunate bystanders in the lift, nothing is hidden.