Large-scale genomic instability is a hallmark of cancer progression. Advanced and metastatic tumors are characterized by a large number of copy-number aberrations (e.g. gains, losses) and their underlying structural variants (SVs, e.g. duplications, deletions). While the pathogenic role of single-nucleotide variants is frequently established through recurrence and functional studies, precisely recurrent SVs are rare and their mechanisms remain elusive. The discovery of position effect variegation in Drosophila provided the first link between disruptions in chromosome architecture and altered gene-expression patterns. Several genetic disorders caused by balanced translocations and inversions have been subsequently interpreted as dosage-independent positional effects. The recent availability of paired genomic and transcriptomic data from thousands of human cancers, made it possible to study position effects as an organising principle for somatic structural variation. In this talk I will concentrate on prostate cancer, and discuss the causes and consequences of large-scale genomic instability. I will present evidence that, similar to mutational signatures, aberrations in select oncogenes and tumor-suppressors cause characteristic patterns of structural variation. Somatic SVs are also associated with dosage-dependent and dosage-independent positional effects, analogous to the germline SVs underpinning genetic disorders. Indicative of synthetic dependencies between mutations and chromosomal aberrations, select SVs are recurrent, conditional on the presence of mutations in other genes. By example of recurrent rearrangements within an ultra-conserved syntenic locus I will highlight the challenges of delineating functional consequences of SVs, and the additional insights afforded by chromosome conformation capture sequencing and haplotype-resolved linked-read sequencing.