Three Systems, and What a Scan Cannot Say

Helen Fisher proposed that mating behaviour in humans and many other species runs on three distinct but interacting systems, each of which evolved to solve a different problem.

Lust, associated primarily with testosterone, evolved to initiate the process with any appropriate partner. Attraction, associated with dopamine pathways, evolved to focus attention and effort on one candidate rather than dissipating both. Attachment, associated with oxytocin and vasopressin, evolved to keep a pair together long enough to complete whatever parental work the species requires.

Three problems, three systems, and the argument is that they are separable.

That separability is the model's most useful and most uncomfortable prediction. If the systems are genuinely independent, they can be aimed at different people simultaneously. Wanting one person, being fascinated by a second and being deeply bonded to a third is then an expected output of the architecture rather than a moral aberration or a sign that something has failed.

That prediction matches what people report, which is a point in the model's favour, and it is not a permission slip, which is a point worth making explicitly. A mechanism explaining why a state occurs says nothing about what anybody should do about it.

The framework is widely used and it has been criticised, both as an account of the evolution and as an account of the mechanisms. The neat assignment of one neurochemical to each system is considerably looser in the underlying literature than the summary suggests, and the same molecules turn up doing many other jobs.

Which brings us to the methodological warning, and it belongs in this chapter because this is where popular science about love goes most reliably wrong.

Fisher's work includes functional imaging of people reporting intense romantic love. Studies of this kind show differences in blood flow between experimental conditions, averaged across subjects, in regions that participate in a great many different tasks. From that you cannot read off what an individual felt. You cannot conclude that a region is love, and you cannot conclude that love has been located.

The pattern of the overclaim is always the same. A region is more active in one condition than another; the region is known to be involved in reward; reward is associated with pleasure; therefore the scan shows love lighting up the pleasure centre. Each step is a small slide and the destination is not supported by the first step.

The same criticism applies to every claim in this course with a molecule in it, including the voles in the previous book. It is not an argument that biology is irrelevant, which would be silly.

What this research genuinely establishes is that love is implemented in ordinary biological machinery, shared to a considerable extent with other animals, and not made of anything exotic. That is worth knowing. It is also not the same as having explained it, and the distance between those two statements is where most of the popular writing on this subject lives.