You are free to reuse your previously written sentences in other papers. Usually you have spent time on refining the sentences, and cannot express them any better. So you just quote yourself. Today it is easy for you to find the original sentence on your computer and to cite it properly. This is research business as usual.
However, occasionally you may write down an idea very similar to something you wrote long ago, and which is not picked up your computer search or personal memory. Not citing this removes renewing your prior claim to the idea, and is self-plagiarism. Self-plagiarism is frowned upon because the new research paper is supposed to clearly indicate what is old work and what is new work, and who did it.
I would set the threshold at not citing three consecutive words as being OK. But remember that it is the idea that is being self-plagiarised, not the sentence.
A practice to be deprecated is to unnecessarily paraphrase an old idea to try to hide its origin.
You are free to reuse your previously written sentences in other papers. Usually you have spent time on refining the sentences, and cannot express them any better. So you just quote yourself. Today it is easy for you to find the original sentence on your computer and to cite it properly. This is research business as usual.
However, occasionally you may write down an idea very similar to something you wrote long ago, and which is not picked up your computer search or personal memory. Not citing this removes renewing your prior claim to the idea, and is self-plagiarism. Self-plagiarism is frowned upon because the new research paper is supposed to clearly indicate what is old work and what is new work, and who did it.
I would set the threshold at not citing three consecutive words as being OK. But remember that it is the idea that is being self-plagiarised, not the sentence.
A practice to be deprecated is to unnecessarily paraphrase an old idea to try to hide its origin.
In my own opinion, self-plagiarism could be avoided if you can paraphrase the words you have used in one article in another article so that the latter doesn't look exactly the same like former. We shouldn't be lazy in our report writing. Thanks