Clustering algorithms are motivated by biology in that they offer the ability for learning through classification. We routinely learn new concepts by relating them to existing knowledge. We classify the new knowledge by initially trying to cluster it with something we already know (as a basis for our understanding). If we cannot relate it to something we already know, then we must create a new structure for understanding that is different from our existing patterns. This new pattern may then form the basis for understanding new knowledge.
By clustering new concepts together with analogous old ones, and creating new clusters when we encounter new knowledge, we solve what Grossberg coined the stability-plasticity dilemma. The problem is how to adapt (learn new things) without the new knowledge destroying what we've already learned. The ART1 algorithm includes the necessary elements to not only create new clusters when sufficiently different data is encountered , but also to reorganize clusters based upon the changes.