There is a circularity in what Thomas Kuhn calls ‘normal science’:
Before he could construct his equipment and make measurements with it, Coulomb had to employ electrical theory to determine how his equipment should be built. The consequence of his measurements was a refinement in that theory.
— Kuhn 1969, 33-4
If the theory has to be replaced instead of refined, then we have a scientific “revolution”; but even a revolution is (in practice as well as etymology) just a turning again or re-turn when seen from a more long-term perspective. Kuhn’s paradigm is not just a theory but an ‘achievement’ which encapsulates the whole cycle of theory-application-experiment-measurement. All parts of the cycle are renewed, including the kind of evidence that is considered relevant to the theoretical question. ‘Normal science’ aims at greater articulation and precision in applying the ‘paradigm’ to ordinary problems; but ‘extraordinary problems … emerge only on special occasions prepared by the advance of normal research’ (Kuhn 1969, 34).
Maturana and Varela (1992, 28) also explain scientific explanation as a circular process. More specifically, Varela described his ‘neurophenomenology’ as ‘the circulation between a first person and an external account of human experience, which describes the phenomenological position in fertile dialogue with cognitive science’ (Varela 1996, 333). This ‘circulation’ can also be described as a movement ‘back and forth’ – as for instance Edelman does when dealing with the connection between ‘our neural model’ and ‘the experienced properties of a conscious subject’: ‘I believe that the issue is best clarified by stressing the neural mechanisms first, and then going back and forth between phenomenal issues and these mechanisms to show their consistency with each other’ (Edelman 2004, 60). This procedure differs from neurophenomenology mainly in that Edelman, unlike Varela, gives priority to the third-person view. But of course it doesn’t matter where you start in a circular process.
Varela and his collaborators have also spoken of a ‘necessary “circulation” between everyday experience and scientific experience’:
On the one hand, everyday experience provides the sensuous, material contents from which and with which science must work. On the other hand, the scientific analyses built from these contents contribute to the formation of our life-world and provide important leading clues for phenomenological analyses of how our experience of the world is genetically and generatively constituted.
— Thompson (2007, 34)