Recently, a number of substructural approaches to the semantic paradoxes have been proposed, which identify one or another structural rule as the ‘culprit’ giving rise to paradox. This is often taken to be sufficient motivation to restrict or altogether ban the rule in question. However, it seems that independent, i.e. not paradox-related, motivations would be required for such restrictions to be truly convincing. In other words, we need independent arguments on what is wrong with this or that structural rule, and to this end, we need to discuss the (putative) reasons why the rule was thought to be plausible in the first place. In this talk, I adopt a dialogical conception of deductive proofs in order to discuss the rationale behind weakening and contraction in particular. I argue that, while they are both prima facie plausible dialogical principles, there are also purely dialogical reasons to restrict both weakening and contraction.