Social bonding is not an emotion — it is a neurobiological program executed by a conserved mammalian circuit. The two principal molecular mediators are oxytocin and vasopressin, acting on brain regions including the nucleus accumbens, ventral tegmental area, amygdala, and prefrontal cortex to produce selective attachment, affiliative motivation, and reduced social anxiety. This circuit did not evolve for love — it evolved because social group membership was the primary survival strategy of our ancestors. Modern humans inherit this ancient machinery and run it through the complexity of language, culture, and extended consciousness.
The prairie vole (Microtus ochrogaster) is one of the rare mammalian species (~5%) that forms long-term pair bonds. Insel & Young (2001) demonstrated that the difference between monogamous prairie voles and promiscuous meadow voles lies not in oxytocin and vasopressin production but in the distribution of oxytocin receptors (OXTR) and vasopressin 1a receptors (AVPR1a) in the nucleus accumbens and ventral pallidum. Prairie voles have dense receptor fields in reward circuits; meadow voles do not. Mating in prairie voles triggers dopamine release in the nucleus accumbens simultaneously with oxytocin — this co-activation forges a conditioned bond between the partner and reward. This mechanism is conserved in humans.
Human neuroimaging confirms the vole model. Bartels & Zeki (2000, 2004) demonstrated that viewing a romantic partner activates the VTA, caudate nucleus, and nucleus accumbens — the core reward circuit — in a pattern distinct from viewing close friends or attractive strangers. Simultaneously, viewing a romantic partner deactivates the lateral PFC (critical evaluation) and the amygdala (threat detection) — the neural signature of "trust." This combination — elevated reward + reduced vigilance — is the neurobiological definition of attachment.