Any definition of research has to be distinct from merely reviewing what other people have said. You, as the researcher, need to have some input as well.
"This is what author A says... On the other hand, Author B says..." -- review.
"This is what author A says... On the other hand, Author B says... I agree with author A because of reasons C,D,E..." -- research.
The "reasons" need to be somehow primary, and stripped of subjectivity as far as is possible. In Humanities, that means primary evidence; in Sciences, that will mean your own experiments and/or theoretical calculations.
There's nothing wrong with a review exercise, ie an appreciation of what the various competing viewpoints on a topic are, but it still deserves to be treated distinctly from active research. It's as fundamental as the difference between being a bystander and a participant, or the difference between the audience and the actors.