The use of meta-communication makes it easier in my experience. It feels very cumbersome and artificial. But when laying out the central thesis, before the cont

The use of meta-communication makes it easier in my experience. It feels very cumbersome and artificial. But when laying out the central thesis, before the contextualization, you describe the content and (sub)conclusions of the different parts of the paper. My favourite lecture always said there must be no surprises in a research paper. You have to spoil the surprising bits from the get go, and it is better to repeat yourself than surprising anyone. edit: If I recalled correctly she said that everything of substance in a paper should be touched upon at least four times. One in the abstract, another in the the introductory parts of the paper, again where the actual substance is (results, methods, conclusions, etc.), and finally in concluding remarks.