Pheromone-cologne marketing leans hard on a single image: a moth flying kilometers toward a female across a forest, drawn by a few molecules in the air. The implication is that a spritz of the right cologne does something similar at a bar. The biology underneath that comparison is real, but it does not transfer to humans the way the ads imply. The strict scientific definition of a pheromone has a high bar, and almost nothing in the human candidate literature clears it. That is the gap worth understanding before you spend money.
The original 1959 definition
Peter Karlson and Martin Lüscher coined the word in a short Nature paper in 1959. Their definition was specific: a pheromone is a chemical signal secreted by one member of a species and received by another member of the same species, where the receiver shows a defined innate behavioral or physiological response. Innate, not learned. Species-specific. Predictable enough that you can describe the trigger and the response in one sentence.
That bar matters. A scent that makes someone feel mildly more relaxed, or that correlates weakly with one rating on a questionnaire, does not meet it. A molecule that triggers a stereotyped, reliable response in nearly every receiver does. The distinction is the entire reason the word exists. Karlson and Lüscher were not trying to name 'interesting smells animals make at each other.' They were naming a category of signal that functions like a key in a lock — molecule arrives, behavior runs.
Over the following decades the term has been stretched, especially in popular writing, to cover almost any chemical exchange between members of a species. That stretch is the source of most of the confusion in pheromone-perfume marketing. If you use the loose definition, humans clearly have pheromones. If you use the original strict definition, the case is far weaker, and the evidence for any single human molecule meeting all three criteria — species-specific, innate, predictable trigger — is thin.
Examples that meet the original definition (all non-human)
Bombykol — the silk moth sex pheromone
A single molecule, identified by Adolf Butenandt in 1959, released by female silk moths. Males detect it in vanishingly small quantities and fly upwind toward the source from kilometers away. One compound, one behavior, near-total reliability across the species. This is the textbook example for a reason.
Boar androstenone — the pig sex pheromone
Boars produce androstenone in their saliva. An estrous sow exposed to it adopts the lordosis posture — back arched, ready to be mounted — within seconds. Pig farmers literally buy androstenone in aerosol form for artificial insemination. Same molecule shows up in human sweat at much lower levels, which is why it ended up in pheromone colognes. The behavioral effect in pigs is dramatic. The effect in humans is not. See the breakdown on androstenone for what actually transfers.
Honeybee queen mandibular pheromone
A blend of five compounds released by the queen. It suppresses worker ovary development, attracts a retinue of attendant workers, and organizes the colony's foraging and brood care. Remove the queen and the chemical signal disappears within hours, and worker behavior shifts. The colony runs on this signal.
Ant trail pheromones
A scout ant finds food and lays a chemical trail back to the nest. Nestmates encounter the trail and follow it, reinforcing it with their own secretions if the food is still there. The compound varies by species but the logic is the same: one molecule, one behavior, written into the nervous system at the developmental level.
Why these examples create misleading expectations for humans
Look at what these cases share. A single identified compound (or small blend). A single receiver behavior that runs on near-autopilot. A hardwired response that does not require learning, context, or interpretation. The moth does not consider whether it likes the female. The sow does not consider whether the boar is her type. The behavior is the chemistry.
Nothing in the human candidate molecule literature comes close to that. The most-studied human candidates — androstadienone, estratetraenol, copulins — produce, at best, small statistical shifts in mood, attention, or hormone levels in some subjects in some studies, often with failed replications. There is no human equivalent of pressing the bombykol button and getting a fly-upwind response. Anyone selling a cologne as if there is, is leaning on a comparison the science does not support. See the breakdown of the most-cited pheromone studies for how mixed the actual numbers look.
What humans likely have instead
The honest read of the human evidence is that we probably do have some form of chemical communication, but it sits in a different category. It looks graded rather than binary. It looks context-dependent rather than automatic. It overlaps with conscious olfactory processing rather than bypassing it. The molecules in question — androstadienone , copulins , and a few others — appear to modulate mood, attention, and sometimes hormone levels in receivers. They do not appear to trigger lock-step behavior the way a true pheromone does in a moth or a pig.
Modulation is a real effect. It is just a much smaller, much more conditional one than the marketing language suggests. A signal that nudges someone's amygdala activity in a brain scan is interesting. It is not the same thing as a signal that overrides their decision making in a bar.
Context dependency is the part people underestimate. A copulin effect that shows up in a lab when male subjects rate photographs of women's faces after a 15-minute exposure is not the same effect that would survive a noisy bar, a few drinks, the receiver's existing preferences, and a brief encounter at arm's length. The lab effects are real signal. They are also delicate, and they get washed out fast by the actual conditions in which a cologne would be doing its supposed work.
The VNO point again
Most mammals with strong, reliable pheromone responses have a dedicated functional vomeronasal organ — a separate sensory structure wired directly into the emotional and hormonal centers of the brain, bypassing conscious smell. That is the architecture that makes the sow's lordosis response or the rabbit pup's nipple-search response possible.
Humans appear to have a vestigial VNO pit at best, with no clear evidence of functional wiring in adults. That is covered in detail in the VNO debate , but the short version is: without that hardware, the kind of subliminal, bypass-the-cortex pheromone signaling that drives stereotyped mammal behavior is mechanically much harder to pull off in a human. Whatever chemical signaling we do is more likely routed through the main olfactory system, which is conscious, learned, and heavily context-dependent.
What this means for pheromone perfume marketing
When a cologne ad implies the product will produce the kind of automatic attraction response that boar androstenone produces in sows — heads turning, approaches multiplying, decisions made before the target can think — that is a comparison the science does not support. The animal examples are real. The leap from those examples to a human product is the part being oversold.
The realistic human-cologne effect is, at most, a small mood and attention modulator stacked on top of a good scent profile and the confidence boost that comes from wearing something you like. That stack can absolutely change how an evening goes. It does not run on the moth-and-bombykol mechanism, and pretending it does sets buyers up for disappointment. The longer breakdown lives in do pheromone perfumes work .
The honest reframe
Humans almost certainly have some form of chemical communication. Copulins probably modulate male perception in narrow contexts. Androstadienone clearly affects brain activity and cortisol in women in controlled lab settings (Saxton 2008, with the Hare 2017 replication failure as the necessary counterweight). Mother-infant chemical signaling is real and well documented. None of this is in dispute as a category.
What is in dispute is the magnitude and reliability. The animal examples are in a different category, not a different point on the same scale. A moth pheromone response is closer to a reflex than to a mood. A human chemical-signaling response is closer to a mood than to a reflex. Once you see them as two different things, the cologne marketing reads very differently — and you can value the placebo and confidence layer for what it actually contributes, instead of expecting the moth-flight response that is not coming.
Further reading
Real references
- Karlson, P., & Lüscher, M. (1959). 'Pheromones': a new term for a class of biologically active substances. Nature 183: 55-56. The original coining of the term and its definition.
- Wyatt, T. D. (2015). The search for human pheromones: the lost decades and the necessity of returning to first principles. Proceedings of the Royal Society B. The leading skeptical review, and the one to read if you read only one.
- Schaal, B., Coureaud, G., Langlois, D., Giniès, C., Sémon, E., & Perrier, G. (2003). Chemical and behavioural characterization of the rabbit mammary pheromone. Nature 424: 68-72. A clean example of a true pheromone in a mammal — newborn rabbits search for the nipple in response to a single identified compound.
- Saxton, T. K., Lyndon, A., Little, A. C., & Roberts, S. C. (2008). Evidence that androstadienone, a putative human chemosignal, modulates women's attributions of men's attractiveness. Hormones and Behavior 54: 597-601.
- Hare, R. M., Schlatter, S., Rhodes, G., & Simmons, L. W. (2017). Putative sex-specific human pheromones do not affect gender perception, attractiveness ratings or unfaithfulness judgements of opposite sex faces. Royal Society Open Science 4: 160831. The failed replication that anyone citing Saxton has to reckon with.
No comments yet