The first author is in some sense the moral owner of the text. Coauthors should correct and come up with suggestions to improve the text. If in the end there is a discussion about language variants that differ only in taste, the opinion of the first author prevails. Senior authors should restrain themselves and not impose their own taste. First authors should be persistent in not giving in.

Obligatory items

Title

Start with a working title.

Titles are short. No definite articles. Be precise. If your paper is 100% experimental, do not use a title that might give the impression that you have also done theory.

Authors

First author

The first author writes the manuscript. Other authors comment and give suggestions. Never should a senior scientist write part of the paper himself. He should keep on discussing and suggesting modifications until he can live with the manuscript. A senior scientist, not being the first author, who writes part of the manuscript himself, will seriously depress and indeed insult the first author.

Many first authors

Group leaders with a big ego organize the writing of a paper differently. They split up the content of the paper in a number of parts. Each prospective coauthor gets to write such a part. They all have to turn in their sections and the group leader will glue it all together. Junior scientists in such a group will have a hard time growing up.

Coauthors

Who will be coauthor? If your mother did not teach you any decency it is too late now anyway. The question of coauthorship only arises when the text is mature enough to be a concept for a paper. Or when there is agreement on what material should be in the paper. Discussing authorship earlier might give unnecessary friction, because the question who is going to be an author depends heavily on what is finally going to be in the manuscript.

Sending a concept with authors A, B, and C to colleague D with the question whether or not he wants to be a coauthor is an insult to colleague D.