This paper represents four years of field research into residual neural patterning in inherited brain-computer interfaces, conducted in the Shelf district of GLMZ between 2221 and 2225. Three peer-reviewed journals declined to publish these findings. The Journal of Neural Interface Studies, the Meridian Neuroscience Quarterly, and Cognitive Architecture Review all rejected the manuscript within days of submission. All three journals receive primary funding from Lazarus Pharmaceuticals. Dr. Osei-Lindqvist elected to release the paper through independent channels.

The core finding: BCIs develop physical channel formations in their neural substrate that are specific to each user. These formations are created by sustained electrochemical interaction between the device and the host brain. They are not data. They are topology — physical grooves, ridges, and crystalline microstructures in the substrate material. A factory reset, which operates at the software level, cannot alter physical structure any more than reformatting a hard drive can change the shape of the disk. The formations persist indefinitely. They are the ghost weight.

The study's most significant and most controversial finding concerns the maternal line. In cases where BCIs pass from mother to daughter, the residual patterning exhibits a selectivity that current neuroscience cannot explain. Maternal memories — experiences belonging to the mother — persist at measurably higher fidelity than paternal memories in cases where both parents used the device. Emotional memories persist more strongly than factual ones. And within emotional memory, three categories dominate with overwhelming statistical significance: memories related to childbirth, memories related to child-rearing, and memories related to survival threat. The BCI preferentially retains the experiences of being a mother and staying alive.

The substrate material is silicon-carbide lattice. It is not DNA. It has no biological mechanism for sex-linked information retention. It should not — by any known principle of materials science — preferentially encode maternal experience. And yet across 200 inherited devices studied, the pattern held with 94% consistency. Dr. Osei-Lindqvist's hypothesis: the BCI interface is shaped by the host's hormonal environment. Estrogen-dominant neurochemistry creates deeper, more durable substrate impressions. The device is not selecting for maternal memory. It is simply recording women's experiences more deeply because of the biochemical context in which those experiences occur.

The paper's final line has been widely quoted in the Shelf since its independent release: "The BCI is teaching her to be someone who died before she was born. We built a machine to augment cognition. Instead, we built a vessel for the dead."