IEEE Systems, Man and Cybernetics Magazine - July 2018 - 40

Leadership Corner

by Vladik Kreinovich

Publications Beyond
Traditional Papers

t a recent IEEE Panel of Editors meeting, John Baillieul,
chair of the IEEE Publications Strategic Planning Committee, challenged the audience to go
beyond the traditional thinking that
papers are publication units. At first,
one may wonder why we need to go
beyond papers in the first place. But,
on second thought, this idea makes
perfect sense. The whole purpose of
writing a paper is for others to read it
and use its results. Let us, therefore,
consider the publication process from
the viewpoint of the reader.
While a few revolutionary papers
provide novice solutions to completely
new problems, most papers address
formulated problems that have existing solutions. The paper's value provides a new, better solution to the previously described problem, be it a more
efficient control, a faster algorithm, an
application to a slightly different class
of problems, or other solutions.
How can an author convey a new
idea to a publication's readers? Unless
the problem is well known, which happens rarely, the publication must contain some description of the problem.
If the problem has been discussed in
prior literature, good details are likely
provided. Currently, the researcher
who discovered a better way to solve
a problem cannot simply copy a description into his or her paper; that
Digital Object Identifier 10.1109/MSMC.2018.2833518
Date of publication: 18 July 2018

Vladik Kreinovich

would be plagiarism. As a result, the
researcher must select one of the following two options: 1) citing the description instead of formulating the
problem in detail or 2) rephrasing the
existing description.
This may sound natural to those
who are familiar with current publishing practices but, from the viewpoint
of the readers, both options come with
limitations. With the first option, the
reader must search for and download
yet another paper or even buy it. With
the second option, instead of the existing well-written and designed introduction, the reader gets an amateur
retelling. From the reader's viewpoint,
an ideal situation would combine the
original paper's introduction with the
new technical information. In other
words, from the reader's viewpoint, it
is ideal when the new author(s) do(es)
not write a completely new paper but
rather a subpaper, which could com-

IEEE SyStEmS, man, & CybErnEtICS magazInE July 2 0 1 8

bine with the already available introduction to make a perfect new paper.
So, why not allow this? Why not explicitly allow the publication of subpapers and modular papers?
If the new author has a better introduction, from the reader's viewpoint,
we should replace the original introduction with the new one. What if the
original introduction is almost perfect
but can be made even better with minor editing? From the reader's perspective, it would be better to have the edited version. Of course, since a reader
is also an author, the reader would like
to see a clear indication of who created
the original text and who modified it.
The need for such subpapers is not
limited to new introductions. If there is
an interesting theoretical paper where
something is mathematically proven
(e.g., the stability of a certain nonlinear
system) and someone comes up with a
new, better proof that is shorter, clearer, and more accessible to the general
audience, why write an entirely new
paper? Why not replace the proof-containing subpaper?
If the new proof has a somewhat
more general result, then instead of
writing everything from scratch, why
not make a minor modification to
the subpaper that contains the result
formulation? And even if there are
no modifications and we have only
one paper, why not design it in such
a way that allows different audiences
to combine the subparts in ways most
appropriate for them? For example, if
we have theoretical results with applications to practical problems from different engineering areas, a theoretician
would be more interested in the proof
of the result and the brief description
of applications while a specialist in one
of the application areas would probably not be interested in proof details
but rather in the details of the application. Why not add a tutorial subpaper

Table of Contents for the Digital Edition of IEEE Systems, Man and Cybernetics Magazine - July 2018