The vast scale and continuous evolution of the Linux kernel make its maintenance a complex undertaking, where code duplication remains a persistent challenge that can hinder development and introduce bugs. This paper presents a multimethod ethnographic study investigating the socio-technical dynamics of contributing to the Linux kernel by reducing identified code duplication. To support this study, we developed a command-line utility tool for detecting function-level duplications, offering a concrete entry point for new contributors. Our ethnographic approach, including complete participant observation and participant-as-observer analyses, demonstrated that addressing duplications is viable for lowering the contribution barrier. This point is evidenced by 8 of 13 (62%) accepted contributions (patches) from 24 newcomers (16 undergraduate and 8 graduate students) in the Linux kernel project, collectively removing 1.397 lines of duplicated code. Nevertheless, the study reveals a more intriguing and complex reality beyond mindlessly removing clones. An analysis of maintainer feedback on both accepted and rejected contributions highlights a nuanced understanding of technical debt due to code duplication within the Linux kernel community, where the benefits of eliminating duplications are carefully weighed against factors such as readability, the introduction of new abstractions, and the specific code context.