In animals, a discrete class of small RNAs, the piwi-interacting RNAs (piRNAs), guard germ cell genomes against the activity of mobile genetic elements. piRNAs are generated, via an unknown mechanism, from apparently single-stranded precursors that arise from discrete genomic loci, termed piRNA clusters. The content of piRNA clusters, determines the capacity of the system to respond to a given element, in essence comprising an organism's evolving molecular definition of transposons. Presently, little is known about the signals that distinguish a locus as a source of piRNAs and about how abundant piRNAs are selected. To address these questions, we inserted new sequence information into piRNA clusters in mice and flies. In all cases, this information was incorporated into the piRNA repertoire and in one instance was shown to confer the ability to recognize and silence a corresponding element. Notably, patterns of piRNA abundance suggested that both intrinsic sequence and context with the cluster inform piRNA generation. Though piRNAs themselves are not conserved between species, the genomic location of clusters is often retained. We were able to create artificial piRNA clusters in non-native contexts in both mice and flies, indicating that the signals that define these as generative loci must lie within the clusters themselves rather than being implicit in their genomic position. Total RNA and RNA associated with Piwi proteins were isolated and size-fractionated by PAGE into 19-33nt. These were processed and sequenced on the Illumina GA2 platform.