Based
in Switzerland, the open access publisher Frontiers was founded in 2007 by Kamila and Henry Markram,
who are both neuroscientists at the Swiss Federal
Institute of Technology in Lausanne. Henry Markram is also director of the Human Brain
Project.
 |
| Kamila Markram |
A
researcher-led initiative envisaged as being “by scientists, for scientists” the
mission of Frontiers was to create a “community-oriented
open access scholarly publisher and social networking platform for researchers.”
To
this end, Frontiers has been innovative in a number of ways, most notably with its
“collaborative peer review process”. This abjures the traditional hierarchical
approach to editorial decisions in favour of reaching “consensual” outcomes. In
addition, papers are judged in an “impact-neutral” way: while expected to meet
an objective threshold before being publicly validated as a correct scientific
contribution, their significance and impact are not assessed.
Frontiers
has also experimented with a variety of novel publication formats, created Loop
– a “research network” intended to foster and support open science – and pioneered
altmetrics before the term had been coined.
Two
other important components of the Frontiers’ concept were that it would operate
on a non-profit basis (via the Frontiers Research Foundation), and that while
it would initially levy article-processing charges (APCs) for publishing
papers, this would subsequently be replaced by a sponsored funding model.
This
latter goal has yet to be realised. “We dreamed of a zero-cost model, which was
probably too idealistic and it was obviously not possible to start that way”,
says Kamila Markram below.
Frontiers
also quickly concluded that its non-profit status would not allow it to achieve
its goals. “We realised early on that we would need more funds to make the
vision sustainable and it would not be possible to secure these funds through
purely philanthropic means,” explains Markram.
Consequently,
in 2008 Frontiers reinvented itself as a for-profit publisher called Frontiers
Media SA. It also began looking for additional sources of revenue, including
patent royalties – seeking, for instance, to patent its peer review process by
means of a controversial
business method patent.
The
patent strategy was also short-lived. “We abandoned the patent application by
not taking any action by the specific deadline given by the patent office and
deliberately let it die,” says Markram, adding, “we soon realised that it is
far better just to keep innovating than waste one’s time on a patent.” (Henry
Markram nevertheless remains an active patent applicant).
By
the time the peer review patent had died it was in any case apparent that Frontiers’
pay-to-publish model was working well. In fact, business was booming, and to
date Frontiers has published around 41,000 papers by
120,000 authors.
It has also recruited 59,000 editors, and currently publishes 54 journals. By
2011 the company had turned “cash positive” (five years
after it was founded).
Successes not unnoticed
Frontier’ successes did not go unnoticed. Not only did it quickly gain mindshare amongst researchers,
but it began to attract the attention of publishers, not least Nature
Publishing Group (NPG), which in February
2013 announced that was entering into a relationship with Frontiers.
The
exact nature of this relationship was, however, somewhat elusive. In its press release Nature described it as a “strategic
alliance”. An associated news item in Nature
reported that Frontiers
had been “snapped up” by NPG, which was taking a “majority investment” in the
company.
A
post on the Frontiers web site also talked of NPG taking a
“majority investment”, and quoted an approving Philip Campbell (Nature’s Editor-in-Chief) saying, “Frontiers
is innovating in many ways that are of interest to us and to the scientific
community”.
In
reality it was Holtzbrinck
Publishing Group
that had invested in Frontiers, not NPG, although Holtzbrinck was the owner of Macmillan
Science and Education (and thus of NPG).
It was
also unclear as to whether the money that Holtzbrinck had invested in Frontiers
could be described as a “majority investment”. Speaking to Science in 2015,
Frontier’s Executive Editor Frederick Fenter described it rather as a
“minority share”.
Either
way, the precarious nature of Frontier’s relationship with Nature became all too evident in January 2015, when it was announced that Macmillan
Science and Education (along with NPG) was merging with German science
publisher Springer. There was no mention of Frontiers, and the situation was
only clarified when Macmillan posted a tweet in response to
the enquiries it was receiving about the status of Frontiers.
Looking
back, it would appear the much-lauded relationship between NPG and Frontiers was
more wish fulfilment than substance – encapsulated perhaps by a glossy 7-minute video produced at the
time that (amongst other things) includes a clip of the CEO of Macmillan
Science and Education (and former MD of NPG) Annette Thomas welcoming
Frontiers to Macmillan’s office in London, lauding its achievements and
promise, but failing to specify what exactly Nature planned to do with Frontiers.
The
true state of affairs does not appear to have been publicly acknowledged until the
2015 Science article cited above. When
asked to clarify the situation Fenter replied: “We made the
decision about 6 months ago to make a clean separation and never to mention
again that [NPG] has some kind of involvement in Frontiers.”
Critics
Like
most successful open access publishers Frontiers has attracted controversy
along the way. There have been complaints, for instance, about its peer review
process (including an oft-repeated claim that its editorial system does not
allow papers to be rejected), complaints about the level of “spam” it bombards
researchers with, and complaints that its mode of operating is inappropriately
similar to the one used by multi-level marketing company Amway. (By, for instance, requiring
editors to recruit further editors within a pyramidal editorial and journal
structure, setting editors targets for the number of papers they have to
publish in their journal each year, and requiring that they themselves publish
in the journal).
There
have also been complaints about the way that Frontiers promotes itself on its
blog. Its posts have attracted considerable attention (including from
high-profile media outlets like the Times Higher) but critics argue
that while its contributions tend to be presented as research the data is cherry-picked
in a self-serving way. See, for instance, here, here, here, here, and here.
In
addition, Frontiers has attracted criticism for publishing a number of controversial
papers (see here and here for instance),
and in 2014 it was accused of caving in to specious libel threats by retracting a legitimate
paper.
The latter led to Frontiers’ associate editor Björn Brembs publicly resigning.
A
number of other prominent researchers have publicly criticised Frontiers too.
In June, for instance, a blog critique was posted by Dorothy Bishop, Professor of
Developmental Neuropsychology at the University of Oxford, and another one a month later by Melissa Terras, Professor of
Digital Humanities in the Department of Information Studies at University
College London (UCL).
More
recently, in January, Micah Allen, a Cognitive
neuroscientist at UCL, rehearsed the various complaints against Frontiers in a blog post entitled “Is
Frontiers in Trouble”.
But
the most controversial incident occurred last May, when Frontiers sacked 31 editors amid a row over
independence. The editors complained that Frontiers’ publication practices are
designed to maximise the company's profits, not the quality of papers, and that
this could harm patients.
The wave
of criticism reached a peak last October when Jeffrey Beall added Frontiers to his list of “potential, possible, or
probable predatory scholarly open-access publishers”.
Supporters
On
the other hand, Frontiers has no shortage of fans and supporters, not least amongst
its army of editors and authors. It has also received public support from a
number of industry organisations.
In a statement posted on its web
site last year, for instance, the Committee on Publication Ethics (COPE) said, “We note
that there have been vigorous discussions about, and some editors are
uncomfortable with, the editorial processes at Frontiers. However, the
processes are declared clearly on the publisher's site and we do not believe
there is any attempt to deceive either editors or authors about these
processes. Publishing is evolving rapidly and new models are being tried out.
At this point we have no concerns about Frontiers being a COPE member and are
happy to work with them as they explore these new models.”
And in
response to questions being
asked
about the role that Frontiers’ journal manager Mirjam Curno plays at COPE the
statement added, “Frontiers has been a member of COPE since January 2015. In
the interests of complete transparency, we note here also that one of the
Frontiers staff, Mirjam Curno, is a member of COPE council – a position she was
elected to when she was employed at the Journal of the
International AIDS Society in 2012 and which continued (with the
agreement of the COPE Council and on becoming an Associate Member of COPE)
after she moved to Frontiers; she is now also a trustee of COPE.”
Around
the same time the Open Access Scholarly Publishers Association (OASPA) published this comment: “We are aware
that concerns have recently been expressed about the publisher Frontiers, which
is a member of OASPA. We have discussed the situation with Frontiers, who have
been very responsive in providing us with information on their editorial
processes and explaining their procedures. In light of these responses, the
Membership Committee remains fully satisfied that Frontiers meets the
requirements for membership of OASPA.”
(We
could note in passing that Frontiers’ Executive Editor Frederick Fenter was a candidate for OASPA’s Board
in 2015).
As
will perhaps be evident, a central focus for the complaints about Frontiers are
its editorial processes, including the claim that its online system does not
allow papers to be rejected. Markram agrees that there has been some confusion
over this. While insisting that reviewers have always been able to reject
papers, she acknowledges below that feedback indicated “it was not clear to
them how to recommend a manuscript for rejection to the handling Editor.”
This
issue, she says, has now been addressed. “Based on the feedback we have now
renamed this option withdraw from
review/recommend rejection, and the reasons, which reviewers can choose
from to indicate why, have also been split accordingly.”
Daniel
Lakens, an assistant professor at Eindhoven University of Technology, has experienced
Frontiers as author, reviewer and editor. He has published several papers, and
was for two years an associate editor for Frontiers
in Cognition, resigning last month due to a lack of time. He continues to
act as a reviewer.
Lakens
suspects that much of the criticism comes from researchers who have failed to understand,
or are not comfortable with, Frontiers’ distinctive peer review process.
“The
review process itself is much more collaborative. This is a good thing if you
find good reviewers willing to invest time in improving manuscripts. Forcing
scientist to enter a discussion, and respond to arguments from the other side,
leads to bigger improvements in manuscripts than at traditional journals, in my
opinion. But it really depends on the mind-set of the reviewers and authors.”
The
other important difference, he says, is Frontiers’ commitment to publishing
methodologically sound research, regardless of significance levels or novelty.
“Publication
bias is probably the biggest challenge that modern science faces. I think it is
important that Frontiers takes a responsibility in publishing all sound
research. Some reviewers, more used to traditional journals, just want to
reject papers they don’t like. For example, this happened when I submitted my own article to Frontiers,
where a reviewer thought there was nothing novel in my explanation of effect sizes,
and withdrew from the revision process. It would have been better if this
reviewer had instead provided some suggestions to improve it (which was no
doubt possible), because the rather substantial interest in the article (it has
been cited 200+ times) suggests his judgment about the novelty of the paper
seems to have been irrelevant.”
Lakens
is also sceptical about claims that it is not possible to reject papers. “Every
manuscript I wanted to reject as a Frontiers editor has been rejected.”
Radical when it started
Lakens
adds: “Frontiers was radical when it started and paved the way for even more
radical open access journals. The collaborative review process is still in many
ways novel and, very often, an improvement over the traditional peer review process.
But now we see even more innovative journals than Frontiers emerging. One
example is PeerJ,
which greatly reduces the cost of open access publishing, and also embraces
open reviews.”
In
truth, impact-neutral reviewing was pioneered
by PLOS ONE in 2006, a year before
Frontiers appeared on the scene. But implicit in Lakens’ statement, I think, is
a belief that while it has played an important part in promoting new types of peer
review, Frontiers now faces competition from younger, more innovative, and less
expensive publishers like PeerJ and F1000Research.
It
clearly will not help that Beall has added Frontiers to his list, which Lakens
believes could encourage researchers to shun the publisher. “Many scientists
are sensitive to prestige, and if these researchers would not be able to
evaluate the quality of science themselves, they might think twice about
submitting to Frontiers, although I would hope this group is rather small.”
Beall,
of course, is himself a controversial figure, and his list is widely criticised
by open access advocates. “I think Beall’s list is not transparent,” says
Lakens. “Inclusions are not justified, and occur on the basis of the personal
opinion of a single individual. The scientific community should ignore Beall’s
list, and pay more attention to the Directory of Open
Access Journals
(although no list will be perfect). I think Frontiers should take valid
criticisms seriously, because in science, there is always room for improvement,
but I don’t think Beall’s list falls on the category of ‘valid criticism’.”
It
is indeed remarkable that the decisions of a lone librarian sitting in a Colorado
library could have a significant (and global) impact on a publisher. Only too aware
of this, in December Frontiers dispatched Fenter and Curno
to Colorado to meet with Beall and try and persuade him to take Frontiers back
off his list – apparently without success.
Underlying
all this, of course, is the fact that the emergence of the Internet has triggered
manifold controversies within the research community. Above all, it has plunged
scholarly communication into a period of considerable upheaval, and put inherited
ways of doing things under growing pressure, not least traditional peer review.
The cost of publishing research papers is a further source of often bitter
disagreement – and open access publishing has amplified both issues.
A key
question here seems to be how publishers find an appropriate role for
themselves in the emerging new landscape. In the Q&A below Markram says
that “dumping all content on the Internet, unchecked, in multiple versions of
readiness, and as cheaply as possible, is not a service to anyone”.
Many,
if not most, would doubtless agree with this, which would seem to imply a
continuing gatekeeping role for publishers. But who these publishers should be,
exactly what kind of service they should provide, and what they should charge
for that service remains unresolved.
On
the issue of costs, Markram asserts that under the traditional subscription
system it costs $7,000 to publish an article, a figure she says that OA
publishers have reduced to around $2,000, and Frontiers to just $1,100.
I am
sure many would challenge these figures, but I will finish with two
(rhetorical) questions: First (leaving aside the issue of whether pedestrian
papers written solely in order to bulk up CVs should in fact be formally
published), if the average rejection rate at Frontiers is (as Markram says
below) just 19% (i.e. 81% are accepted), and if some of those articles turn out
not even to have met Frontiers’ lower threshold for publication (As Markram points it, “no
peer-review is bullet-proof, so problematic articles regrettably do sometimes
get through) then does $1,100 (or $2,000) per paper represent good value for
money? Second, how high does the acceptance rate need to go before simply
dumping papers on the Internet becomes a logical way for the research community
to save itself millions of dollars a year?
To read Markram’s detailed answers please click on the link below. These are in a pdf file preceded by this introduction.
Readers should be aware that the
Q&A is long. I have chosen not to edit Kamila Markram’s text and there are
some repetitions, but I was keen to allow her to address my questions in her
own words, and as fully as she felt to be appropriate. I have, however, made
ample use of pull-quotes.
####
This interview is published under a Creative Commons licence, so you are free to copy and distribute it as
you wish, so long as you credit me as the author, do not alter or transform the
text, and do not use it for any commercial purpose.
To
read the interview click HERE.
No comments:
Post a Comment