A small spray bottle, a label printed with the word pheromone, a promise with no source attached. The word has a real scientific history, beginning with a silk moth in 1959. In humans the story stops somewhere quite different.
A Word on a Label
The word pheromone was coined by Peter Karlson and Martin Lüscher in 1959 from two Greek roots, one meaning to carry, the other to stimulate. It names a chemical signal passing between individuals of the same species that triggers a behavioural or physiological response in the receiver.
The name was deliberately set beside the word hormone. A hormone travels inside one body, from a gland to another organ. A pheromone travels outside, through air or contact, from one individual to another of the same species. Both are chemical signals; the difference is that this one crosses the boundary of a body.
That definition is stricter than the way the word is used on labels. According to Tristram Wyatt, in a 2015 review, establishing a pheromone takes three steps: demonstrate an odour-mediated behavioural or physiological response, identify and synthesise the active molecule, then confirm by bioassay that this very molecule produces the response.
Three steps sound simple but are hard to complete. Many products bearing the word pheromone have never passed the first step in humans. This post reads the word with a laboratory's attitude: what has been demonstrated, in which species, and what is still blank.
Pheromones in Insects
The female silk moth, Bombyx mori, releases a substance that draws males from a distance. Adolf Butenandt and colleagues isolated it, reported its structure around 1959 and described the bioassay and purification in detail in a 1961 paper. They named it bombykol. It was the first pheromone to be chemically well characterised.
Bombykol passed all three steps. There was a clear behavioural response in males. There was a molecule isolated, its structure determined and synthesised. And the synthetic molecule produced the same response in testing. That is the standard later research had to meet.
Since then pheromones have been established in a great many insect species and in some mammals too, such as the pig described below. The word pheromone is therefore not vague; it comes with a method, and that method has worked across many species.
The question is not whether pheromones exist. The question is what that method has produced when applied to humans.
Androstenone and Androstenol
Androstenone is an odorous steroid found in boar saliva. When sows in heat inhale it, they adopt the mating stance. In pigs androstenone meets the steps of the definition, and a commercial spray containing it is used by farmers to time artificial insemination. It is a pig pheromone.
Androstenol, a steroid of the same androstene family, is also found in boar saliva, in truffles and in human sweat. The library describes its smell as musky, lightly sandalwood-like, with urinous and skin-like facets, and notes that perception varies widely between people. Androstenol is used in traces in some perfumes for a skin and musk effect.
The variation in perception has a genetic basis. In 2007 Keller and colleagues showed that the receptor OR7D4 responds selectively to androstenone and a closely related steroid, while not responding to 64 other odours. A common variant of this receptor functions poorly. People carrying it were, as a group, less sensitive to androstenone and found it less unpleasant. One molecule is described by some as sweat or urine, by others as sweet and floral, and by others still as nothing at all.
A substance that some people cannot smell, others find unpleasant and others find pleasant makes an unlikely common signal between individuals. That is not yet proof against it, but it is a starting point worth noting.
Studies Not Replicated
For several decades four steroids were cited most often as human pheromones: androstenone, androstenol, androstadienone and estratetraenol. Wyatt's 2015 review concluded that there is no robust, bioassay-led evidence for the claim that these four are human pheromones.
Wyatt pointed out that the reasons for testing these molecules in the first place were weak, so positive results should be read with scepticism. The recurring problems include small sample sizes, overestimated effect sizes, a bias towards publishing positive results, and lack of replication.
A double-blind study in 2017 by Hare and colleagues tested androstadienone and estratetraenol directly, two substances once reported to signal gender. In the first task, 46 people judged the gender of neutral facial morphs; exposure to the two substances did not change their judgements. In the second, 94 people rated photographs of opposite-sex faces; the substances had no effect on the ratings. The authors concluded that the two are unlikely to be human pheromones.
Anatomically, the vomeronasal organ, which receives many pheromones in other mammals, is present in the human fetus but atrophied or absent in adults. Taken together, no substance has been demonstrated in peer-reviewed research to directly influence human behaviour in the way a pheromone does.
Skin Scent and Musk
The musk pod of the male musk deer sits under the abdomen; the secretion inside dries into dark brown grains. In 1906 Heinrich Walbaum isolated the main odorant of musk, muscone. In 1926 Leopold Ružička showed muscone to be a fifteen-membered ring ketone; he received the 1939 Nobel Prize in Chemistry for work on polymethylenes and terpenes.
Musk carries a name that suggests animals, and its smell is often described as warm, powdery, like skin. From there it is easy to slide into the idea that musk is a signal. But skin-like is a description of a smell, not evidence of a behavioural response. None of the research cited in this post places muscone among human pheromones.
Today deer musk is almost absent from the international fragrance trade; the entire genus Moschus is in CITES Appendix II, some populations in Appendix I. In perfumery the word musk almost always means a synthetic molecule. The closest botanical source is ambrette seed, Abelmoschus moschatus, of the mallow family, with ambrettolide as the molecule carrying its musk note.
Perfumes with pheromone on the label usually contain synthetic musks, sometimes with androstenol or androstenone in traces. They smell of those materials: musk, skin, wood. The promises about effects on other people are not in the data. The post Musk · From the Deer to the Ring Molecule follows the history of musk molecules.
What Is Not Yet Known
Humans are mammals, and many mammals use pheromones. Wyatt writes that it is possible, perhaps even probable, that humans have pheromones. What is rejected is not that possibility, but the evidence for the four familiar steroids.
To find human pheromones, Wyatt proposes returning to first principles: approach humans as a newly discovered mammal and repeat each step that has succeeded in other species. Likely sources include the sebaceous glands; comparing secretions of adults and pre-pubertal children may point to relevant molecules.
One of the most promising leads Wyatt names has nothing to do with attraction: a secretion from the areola glands of all lactating mothers, which stimulates suckling in any baby, not only the mother's own. If a human pheromone is established, it may well look like that, small, specific, and far from what labels promise.
Until then, the word pheromone on a perfume bottle is a word ahead of its evidence. The scent inside may still be beautiful. It simply has no need to borrow a promise science has not signed.
A silk moth showed the method.
Humans have not yet been taken through it.
Between the two sits a word printed on a label.