Where does preferential attachment collide?

Robert Merton's Matthew effect names the process by which scientists and works that already carry recognition accrue disproportionately more, while comparably good but less-known contributors receive less, so the gap widens over time independently of intrinsic merit. Merton drew the label from the Gospel line that to those who have, more will be given. Derek de Solla Price later modeled the same accumulation for citation growth and called it cumulative advantage. The referent is a self-reinforcing accrual of reward, and it is the same dynamic that network science would rename preferential attachment three decades later.[1]

Unit

A person's or work's stock of recognition over time

Recovery means

Not applicable

Measured by

Observed as widening gaps in citation, funding, or reward between early and late recipients of comparable merit

In network science, preferential attachment is the growth rule in which the probability that a new node links to an existing one is proportional to that node's current degree, so highly connected nodes attract still more links. Barabasi and Albert showed in 1999 that this rule generates scale-free networks with power-law degree distributions, a signature they reported across the web, citation, and biological networks. The rule has older antecedents in skew-distribution models from the early twentieth century. It is the formal, generative statement of the same rich-get-richer process that sociology had already named the Matthew effect.[2]

Unit

The probability a new link lands on a node, scaled by that node's degree

Recovery means

Not applicable

Measured by

Inferred from a power-law (scale-free) degree distribution and from the rate at which new links attach as a function of degree

W. Brian Arthur's increasing returns to adoption describes markets in which an option becomes more attractive the more it has already been adopted, so small and often accidental early advantages compound until one technology or standard locks in and competitors are shut out, even when an alternative was better. Arthur formalized how history and chance, not efficiency alone, can select the winner. The referent is a positive-feedback accrual in adoption and market share, the same self-reinforcing dynamic as the Matthew effect and preferential attachment, framed here around returns and lock-in.[3]

Unit

An option's market share or installed base over time

Recovery means

Not applicable

Measured by

Observed as path-dependent lock-in and rising returns to an option's installed base