Computer Science as a Layered Discipline: Foundations, Systems, Intelligence, and Societal Responsibility

Computer science is simultaneously a mathematical science of computation, an engineering 
discipline for building reliable information-processing systems, and a sociotechnical practice 
whose artifacts shape institutions and everyday life. This review develops a layered account 
of the field that links mathematical foundations, representations and algorithms, computer 
systems, software and data abstractions, intelligent and interactive applications, and societal 
governance. The article uses a targeted narrative review of seminal research, current 
curricular frameworks, and authoritative standards to identify durable concepts and emerging 
priorities. Four claims organize the synthesis. First, abstraction is the field's central 
intellectual mechanism: each layer hides lower-level detail while exporting guarantees, costs, 
and failure modes. Second, computer science research is evaluated through plural forms of 
evidence, including proof, asymptotic analysis, simulation, controlled experiment, benchmark 
comparison, systems measurement, user study, and field observation. Third, modern systems 
cannot be evaluated by accuracy or speed alone; security, privacy, reliability, maintainability, 
energy use, accessibility, fairness, and human control are cross-cutting quality attributes. 
Fourth, the most consequential research frontiers arise at interfaces among subfields, 
including trustworthy artificial intelligence, quantum-safe security, edge and cyber-physical 
systems, sustainable computing, formal verification, and human-centered interaction. The 
paper contributes an original layered model of computer science, a reusable research-to
deployment lifecycle, and a research agenda centered on compositional assurance, lifecycle 
evaluation, resource-aware intelligence, secure migration, and public-interest computing. The 
synthesis is intended as a reference for researchers, educators, students, and practitioners who 
need a coherent map of computer science without reducing it to programming or to any single 
technological wave.