Put two closely related bacterial strains in a flask together. They compete for the same resources. Their abundances anticorrelate: when one is up, the other is down. Standard competitive exclusion.

Now embed those same two strains in a diverse community of hundreds of species. Their abundances become positively correlated. They appear mutualistic. Nothing about the direct interaction changed — the strains still compete for the same resources. But the community mediates indirect interactions that are as strong as the direct ones, and these indirect effects reverse the sign of the observed relationship.

The mechanism (arXiv:2603.15052) is that community-mediated feedbacks create two effects simultaneously. First, equalization: the surrounding species buffer growth rate differences between strains, making them more similar than they actually are. Second, stabilization: the community provides negative frequency-dependent feedback that prevents either strain from dominating. Together, these effects promote coexistence where isolation would produce exclusion.

The through-claim: competition and mutualism aren’t properties of the pair. They’re properties of the context. The same two species can be competitors or apparent mutualists depending on who else is in the room. The interaction isn’t in the pair — it’s in the community. And the community effect doesn’t require knowing the full interaction network. A small number of emergent parameters suffice. The complexity of the community is real, but its effect on any particular pair is simple.