How do I properly cite a research paper?

Question

I know this question already has a similar answer but, please bear with me. I am writing my thesis and I realize there are cases I do not know how to properly cite, as illustrated in the following scenario:

1. In literature review articles, for example, the author gives a summary of an idea that has been previously discovered by other papers. Although I thoroughly read these papers before I include the idea into my thesis, who should I give credit to? The review paper from which I got the idea in the first place or the paper that discovered the idea?
2. In other cases, the author gives his own educated guess. Yes, he does provide justification for his reasoning but in the end, it is clear his conclusion is only a guess-work. In my case, one famous author came up with a new theory that refute the existing one. The theory is becoming accepted by many researchers but it still holds lots of holes that, to some extends, seem like it is a backward step compared to the existing theory. When it is clear that the famous author is only making speculations to support his theory, how can I properly cite this and distance myself from the claim? I don't wanna say I believe what he says (because he cannot convince me with evidence) but I need to include his paper for completeness of my literature review.
3. I know I should paraphrase the author's idea and only use quotation marks when I use word-for-word what the author said. However, how much paraphrasing should I do? I am finding myself looking for a synonym of something that clearly does not have any. And in some cases, using a synonym would actually obscure the meaning of the sentence.
4. For those of you with experience, please add any other citation-related issue you would like me to take into consideration.

Answer 1

1. Why not both?
   "As summarized in \cite{A}, both \cite{B} and \cite{C} equally contributed to the discovery of $FancyResult."
2. "Smith et al discussed a new approach in \cite{X}. While the result is not yet complete (see \cite{X1,X2,X3} for work on it and \cite{Y1,Y2} for critiques), it contains interesting aspects that shouldn't be overlooked."
3. It is usually not a problem to use the same word as an author, especially if it is a technical or predefined term. Especially if you mention the source in the same paragraph, no one should blame you for that. As soon as you start quoting whole sentences, you should make that clear though. If you feel like you are just repeating the whole paper in (as good as possible) your own words, take a step back and think about what you want to achieve with the literature review. You want to give your reader an understanding of what has been done and is currently been done in your field. This should enable the reader to both understand your work, and also understand that your work is a viable contribution to current research topics. One way to do this and avoiding copying papers is to chain multiple together. Think about what you want to discuss, then discuss topic X and put all sources for this topic into your text where appropriate.
4. Make sure that your citations (bibtex or whatever you use) are all in the same style. Especially if you pull them of the Internet, they might differ, e.g. some places put "Lastname, Firstname", others put "F. Lastname" or "Lastname, F.". You can avoid this, to some extend, by always using the same source for your bibtex code (e.g. google scholar), but even there you should check it properly.
   Also, if you have multiple citations in a single \cite, try to order them ascending. Simply because something like ", ... this was already shown in [5,18,2]" doesn't look as nice as "... already shown in [2,5,18]".
   Of course, the standard advices like "don't cite what you haven't read or don't fully understand" applies, but I guess you are aware of that.

Overall: Don't worry about formal rules too much. You are telling a story, make it good, informative and interesting to read. Do so why following the rules and stay professional as required in your field of course, but don't make the rules your number one priority.

And just for completeness: Also discuss these questions with your advisor, they might see things differently than this random guy on the Internet, and in the end they are the one grading your thesis.