I'm wondering the benefit of paraphrasing when citing in a scientific journal. The same stuff in different clothes? Some people say that it shows that the author has understood the text which he is citing. That is hardly true if one just uses synonyms and changes word order. Which then can also lead to less precise text.
Secondly, do I need to paraphrase when citing my own publications?
The purpose is to make your article read more naturally and smoothly. If you use direct quotations every time you cite, your article will sound strange and be hard to read. It will also be full of quotation marks or blockquotes.
Obviously you do need to make sure that what you write is precise. Sometimes you have to use the same phrase as in the original because it is a technical term.
Yes, even when citing your own publications, it is better to paraphrase rather than give a direct quotation. But in this case direct quotations might sound less unnatural as they are written by the same person.
The purpose of citing is to make clear that a particular statement is not your contribution, but the contribution of the person's being cited. So the question of paraphrasing is secondary. Typically, the flow of the text would require some sort of paraphrasing in order for your text to be understandable, but sometimes a direct quote would do a better job or just the statement in isolation is enough.
If I read an article, I do that because I want to know the content. I could not care less about the author. I am not judging you to see if you understand an article or not. Instead, I am judging the argument you are making. It does not help me to know if you got it right by accident or because of deep understanding; right is right and wrong is wrong.
Things are different when you are a student. It is the job of the instructor to figure out if you understood something, so now your understanding does matter.
Did you ever play the game "telephone" when you were a kid? A group of kids sits in a circle, one whispers to the child to their right, who then whispers what they heard to the child to their right, and so on. When it gets back to the first kid, it's often completely different.
You can also do this with synonyms, taking a synonym of a synonym of a synonym can lead to amusing results.
I think this is the problem with paraphrasing longer quotes. If you are using only a brief notion from the paper, fine:
In his dissertation, Flom said to use condition indexes to measure collinearity.
But if you are interested in more than a sentence or so from a source, I'd lean towards using the exact language, with quotation marks (or block quotes, as appropriate) and a cite to an exact location. Then any confusion is due to the original author, not you, and anyone who is really interested can go and find the rest of the piece.
It's true, as Toby544 said (+1) that too much of this can make your prose seem stilted. (Do you see what I did there?) But a) Academic language, compared to (say) a novel is known for placing more emphasis on precision and less on readability and b) In most contexts, you don't need to cite too extensively. I can think of some contexts where you do, but they are unusual.
