Nature has published a Worldview entitled, "Not all plagiarism requires a retraction" (Not all plagiarism requires a retraction : Nature News & Comment). The author, Praveen Chaddah, delineates three forms of plagiarism and takes the stand that, in some cases, retraction does not serve the relevant field. He presents the viewpoint that, in some cases, science is better served by attachment of a published correction, acknowledging the plagiarism, to the original PDF file.
I disagree. Plagiarism in any form reduces the credibility of the research and the researcher and does not deserve the reward of publication; not requiring retraction would condone the unacceptable behavior. Allowing continued presence of the research in the published literature is little more than a slap on the hand to the plagiarizing author, who in the first place was not much affected by the prospect of being found out and likely is not shamed by the correction.