If I don't know, should I infer? Reasoning around ignorance in a many-valued framework

Many-valued logic allows to reason with partial truth measured by degrees on a discrete scale, but it suffers from an ambiguous interpretation of the middle truth level, considered as intermediate truth or as ignorance, i.e. inability to assess truth. The LM extension introduces an additional value, outside the truth scale, to distinguish between these two notions. This paper studies LM from a logical perspective, examining how to reason in this framework: it discusses the definition of appropriate semantics for the logical connectives and it considers an inference task, proposing a Modus Ponens variant for LM.