IEEE Signal Processing - May 2018 - 7

many of them today occurring outside
the realm of the IEEE.
For example, free online repositories
for manuscript preprints have become
very popular and are being embraced
almost daily by growing communities.
Among these we may mention arXiv
(physical sciences), bioRxiv (biological
sciences), socArXiv (social sciences),
chemRxiv (chemistry), agriXiv (agriculture), The National Bureau of Economic Research (NBER), Social Science
Research Network (SSRN), Zenodo (data
sets), and others. Most of these acronyms
refer to repository platforms where manuscript drafts are posted online by their
authors for broad dissemination. ArXiv
is one of the earliest bottom-up efforts
and is closest to our field of interest. It
is widely popular in the mathematics
and physics communities and has also
gained a strong following within computer science and engineering. Many
authors in our research community have
embraced arXiv, and our Society has
recently launched a signal processing
category within arXiv. The growth of
arXiv is living proof that there is strong
interest in advancing new ways of sharing science. At times, even news outlets
like the BBC and CNN report on "fresh"
results posted on arXiv. Founded in
1991, it has crossed over 1 billion paper
downloads since then. It has also motivated several of the subsequent platforms like bioRxiv (launched in 2013),
socArXiv (2016), and chemRxiv and
agriXiv (2017). Most of these open platforms are focused on "sharing" science
online and less on exploiting the online
medium for "assessing" or "generating"
science. While more can be done, these
platforms are a positive step forward.
Others are exploring more effective
solutions, with more features and opportunities, and new tools are waiting to
happen. However, success in promoting
the "sharing," "assessment," and "generation" of science more broadly will
require directives from large institutions,
leading universities, funding agencies,
and governments. Some nontrivial changes in culture will be required. I believe
that effective progress is possible. After
all, this is the 21st century! We can send
Tesla cars into orbit and Mars Rovers to

presented his masterpiece work on the
a remote planet. And yet, we cannot sit
propagation of heat for review to the
down and explore ways to solve a more
Academy of Sciences in Paris. His work
mundane problem: how can we vet the
was reviewed by a notable committee,
ever-increasing amount of research more
which included giants such as Laplace,
thoroughly? How can we make it freeLagrange, Legendre, and Monge (the
ly available for the entire world to see?
father of differential geometry). What is
And how can we get researchers from reremarkable about this story is that we
mote corners of the world more excited
know who reviewed Fourier's work, and
about collaborating on open problems
we also know how accomplished they
online? Amazingly enough, despite the
were. We even know more. We know
huge progress we have witnessed over
that the committee rejected Fourier's
100 years, our current vetting of science
work, and we also know the reason why.
is somehow between what Galileo and
There were shortFourier experienced.
Galilei Galileo, the
amazingly enough, despite comings in the proof
noted Italian astronothe huge progress we have of convergence of
mer, looked through
witnessed over 100 years, the series representation (now known as
his crude telescope in
our current vetting of
the Fourier series),
1610 at a remote object
science is somehow
which he introduced
in the sky and discovbetween what galileo
in hissolution of the
ered some anomaheat equation. It was
lies around Saturn.
and Fourier experienced.
natu ral to wonder
He was not able to
how the sum of smooth sinusoidal funcrecognize these anomalies as the rings
tions could add up to rectangular pulses
around the planet. He instead thought
with sharp transitions. The concerns that
that he had observed a composite of
were raised by the committee evaluatthree objects and wrote "oOo" in his
ing Fourier's presentation in 1807 were
notes to refer to the shape of the object.
technically sound. Dirichlet, who was a
Galileo was a superlative scientist, and he
student of Fourier, would give the first
had a clear instinct that he had discovered
rigorous proof of the convergence of the
something very important. He wanted to
Fourier series only two decades later in
make sure that he would get the credit for
1829. The rejection in 1807 did not deter
it once it became clear what it was. So
Fourier. He knew that he was working
he sent a coded message to Johannes
on something singular. In remarks he
Kepler, another giant of astronomy, with
made about his work on the heat equagarbled letters [1]. Once decoded, these
tion, he commented on connections to
letters would result in a sentence attesting
Newton and on how "heat, like gravity,
that Galileo had discovered the phenompenetrates every substance of the unienon first. This form of communication,
verse" [2]. The reference to Newton's
through coded messages, was not unusual
gravity does not go unnoticed. Fourier
during Galileo's time. That is how some
pursued his work and resubmitted it
scientists used to share their discoveries
again four years later to the 1811 Grand
to claim priority. It took almost 50 years
Prize competition; the award commituntil, in 1656, the mysterious objects
tee this time again included Laplace,
observed by Galileo were finally disLagrange, and Legendre among others.
covered to correspond to rings around
He was awarded the prize albeit with
Saturn by Christiaan Huygens, of wide
some reservations, which delayed the
fame in the field of optics. Huygens also
publication of his work until 1822 (that
announced his discovery in a coded mesis an 11-year wait in a publication queue
sage only to be decoded a few years later!
in our modern language)! Today's FouImagine if we were to write papers today
rier's tools are at the core of the signal
in that same manner in code! Fortunately,
processing discipline. What is noteworwe have moved away from that mode of
thy about this story is how giants of
(not) sharing science.
their fields vetted this work in the open
Move forward almost two centuries
and how this openness only made the
to 1807 when Jean-Baptiste Fourier
IEEE Signal Processing Magazine

May 2018

Table of Contents for the Digital Edition of IEEE Signal Processing - May 2018