This paper highlights the underutilized analytical potential of compounds and neologisms as indicators of discursive change in text mining applications, particularly in the study of parliamentary discourse and conceptual transformation. Drawing on results from two research projects, this project-wide paper discusses how compound neologisms function as markers of discursive change through case studies focused on the formation, frequency, and productivity of compounds related to the key concepts of 'market' and 'terrorism' in the Swedish Parliament. The analysis combines distant reading techniques to identify large-scale trends and close reading to examine the specific contexts of these compounds. By focusing on compound formation, we emphasize the analytical potential of basic linguistic features often overlooked in Digital Humanities research, offering a fresh perspective on large parliamentary datasets and their role in tracing conceptual transformations over time.