During the 1930s, historical actors discussed phenomena that made the concept of a virtual particle, contrasted to the concept of a virtual state, rather suggestive. The divergences of QED and phenomena of nuclear physics rested – when put into the framing of Dirac’s verbal model – on specific particles being present only in intermediate or “virtual” states. Studying these phenomena in the necessary depth will reveal how a verbal description of mathematical structures was only slowly established when physicist considered the phenomena to be essentially static, how specific mathematical formulations made conceptualizations more or less suggestive, and how basic pillars of the conceptual framework, like the concept of a particle or the vacuum, were slowly altered in connection to the concepts under study. In the epistemically uncertain terrain of quantum field theory in the late 1930s, the verbal structures which I called a “narrative” gathered prominence for the qualitative evaluation of theory and diagrams, Feynman-like but without calculational purpose, were proposed. Pulling the strings of conceptual development together in the concluding analysis, we can see that by the late 1930s a conceptual overlap, potentially tension between virtual particles and virtual processes had emerged and that it manifested itself in the different representational formats.