Plausible principles on truth seem to yield contradictory conclusions about paradoxical sentences such as the Strengthened Liar. Those who take the contextualist approach, such as Parsons and Burge, attempt to justify the seemingly contradictory conclusions by arguing that the natural reasoning that leads to them involves some kind of contextual shift that makes them compatible. This paper argues that one cannot take this approach to give a proper treatment of infinite descending chains of semantic attributions. It also examines a related approach taken by Gaifman and argues that it has the same problem.
"Descending Chains and the Contextualist Approach to Semantic Paradoxes." Notre Dame J. Formal Logic 40 (4) 554 - 567, Fall 1999. https://doi.org/10.1305/ndjfl/1012429719