Information Technology Reference
In-Depth Information
Self-plagiarism
Authors who re-use their own text may well be plagiarizing. Using the same text in
two papers is a step in the direction of publishing the same work twice.
Some scientists feel that it is acceptable to re-use their own background material
from paper to paper. A series of papers may be based on the same ideas or previous
work, and—it might be argued—rewriting the background each time is pointless.
However, there are both principled and pragmatic arguments against this practice.
First, if an author is in the habit of copying the background in each paper, the material
is likely to rapidly become stale, and authors who adopt this practice often seem
unwilling to adapt the material even for papers on a different topic; in contrast, the
discipline of writing new text each time helps to keep the material fresh. Second,
a high-quality discussion of background material or of competing proposals adds
weight to a paper, and increases the chance of it being accepted; by copying, the
author is obtaining credit for old work. Third, some scientists view any significant
re-use as improper, and authors presumably do not wish even a minority of their
colleagues to view them as lazy or unethical. Fourth, most researchers work in teams
of shifting membership. The authors of a paper collectively own its text; for some of
the authors to take text and re-use it is inappropriate. The safe approach is to write
fresh text for each new paper.
Publication of more than one paper based on the same results is prohibited under
the standard scientific codes of conduct. An exception is when there is explicit cross-
referencing, such as by reference to a preliminary publication from a more complete
article that is a later outcome of the same research. (This is the one instance in which
significant re-use of text can be acceptable.) Simultaneous submission to more than
one journal or conference of papers based on the same results should be disclosed at
the time of submission; the usual response to such a disclosure is to reject the paper. In
this context, “the same results” does not necessarily mean a particular experimental
run; if an experiment has been tried on some data, running the same experiment on
other data is not new work unless it leads to new conclusions.
In the context of plagiarism and self-plagiarism, remember that publications are
a permanent record. It may well be that a researcher successfully publishes the same
results twice, or publishes a series of papers with figures and text in common, and
in so doing rapidly develops an impressive publication list. But as time passes it is
increasingly likely that such abuses of the system will be noticed, and there is no
statute of limitations on plagiarism. The zeal of young researchers to publish should
not blind them to the possibility of disciplinary action years or decades in the future.
Self-plagiarism can also be considered from the point of view of copyright. As
noted earlier, when you publish a research paper the copyright is usually assigned to
the publisher, who thus owns, not the ideas, but your expression of them. An author
who re-uses more than a couple of paragraphs or a figure requires the publisher's
permission to do so.
Once a paper has been assigned to a publisher, permission may also be required,
in principle at least, to self-publish your work on theWeb. However, such publication
