I've partaken in the feud firsthand as an engineering student; when I dared venture into Psychology (a dirty medium between science and bullshit) I treated the final "research paper" as I would a true research paper, citing facts from other papers and developing new conclusions.
As it turned out, the TA didn't read any of the semantic content of the paper, but ended up crucifying me when one of the phrases I cited from a source wasn't paraphrased "enough". It was an inline quote, of the form "According to Attwood (2000)... blah blah blah". My sentence structure was entirely different from Attwood's, but four words were the same: Asperger, syndrome, imitate, and model. The first two were the name of the condition which was the topic of the paper; I suppose I could've replaced the latter two with synonyms, but I felt there was no point anyway; I cited the appropriate author, the full citation was correctly placed at the end of my paper, and I was simply using Attwood's fact to support my entirely separate argument.
Apparently, there's some cutoff line to how much you're supposed to mangle a paraphrase to not be plagiarism. I wasn't using Attwood's words to attempt to take credit for some beautiful literary artwork, I was using it for a cold hard fact. The apparent unhealthy fixation on having paraphrases conform to some stupid style guide defeats the purpose of the plagiarism policy, which is intended to protect IDEAS foremost. Furthermore, requiring roundabout ways of saying things only adds to the bullshit-level already stigmatizing the human sciences.
If a scientist publishes a finding saying "we measured 3.14 to 1% precision", and you cite that paper, you better "paraphrase" that finding as "they measured 3.14 to 1% precision", with no bullshit creative mangling like "they measured around three" or "they measured pi". We cite for semantics, not sentence structure. Taking extra time to mangle paraphrases only belittles the original author's contribution, and inevitably degrades the semantic content of the phrase.
On a side note, I was extra careful in my last physics-related mention in my psychology class; I couldn't afford to be punished for plagiarizing Newton's
"force equals mass times acceleration".
I suppose, this equally concise paraphrase in my own words was more appropriate:
"the quantity which has a quotidian manifestation as a displacing exertion can be shown to be non-relativistically equivalent to the quantity resulting from a product-under the standard definition of multiplication which defines an abelian group over the field of real numbers-of the manifestation of baryonic interactions with the space-time manifold and the abstracted compound quantity which is commonly computed as the second analytic derivative of the spatial components of any state-vector defined by a one-to-one and onto invertible transformation from unique plank-scale discretizations to disjoint numerical sets."
Also, I'm a fugitive because I jaywalked the other day.