You don't need his permission to write a paper. Your rationale can be that you have already written it and you want his feedback before you publish. As you have taken the initiative he will have to give some reason for not wanting you to publish from which you can work with. He might tear your work to shreds and further dent your self confidence but he might decide to include other people's work in your first name paper.
Thank you, Tudor_Queen and rewt, for coming back and giving me further advice.
I've done 5 major experiments so far. The results from Experiments 1 & 3 have been inserted into other people's papers (not submitted yet). So now I have the results from only 2, 4 & 5, but these 3 experiments can't form a complete story. Experiment 1 is the most important one - later experiments are all based on it. Therefore, in order to write my own paper, I need to ask 1 & 3 back and use 1,2,3 & 4 to write a decent paper. But I've already agreed that my supervisor can use 1 & 3 for other papers. If I tell him I need them back, then there'll be scuffles. Another option is, let him publish the two papers, and then I still use 1,2,3&4 to write my own paper. But 1&3 are already published, so will any journals still accept my paper? I doubt it.
