Today’s interview comes from ScienceOpen Editorial Board member Professor Bernd Hartke, who gives us the benefit of his experience and his insights on Open Access publishing versus the traditional publication model. We are pleased to have scientists like Professor Hartke on the Science Open Editorial Board, who are willing to think innovatively about science publishing.
Professor Hartke heads the Hartke Group at the Christian-Albrechts-University in Kiel, Germany, which conducts research in Theoretical Chemistry. He did his doctoral work at the Bavarian Julius-Maximilian University, Würzburg and postdoctoral work at UCLA in Los Angeles, and taught at the University of Stuttgart before coming to Kiel in 2002. He has published over 80 articles and presented his research all over Germany and internationally.
Q. Can you describe your research area for us, and perhaps say a few words about what you’re working on currently?
A. My research area is theoretical and computational chemistry. Essentially, we use classical mechanics and quantum mechanics to simulate the behavior of molecules. In principle, we can predict the outcome of any chemical reaction on the computer, without performing a single experiment in a chemistry laboratory. And, again in principle, we can design new molecules from scratch, to perform any desired action at the sub-nanometer scale. The only obstacle is that such calculations require huge computational resources and very long times, so approximations have to be made.
My research group is currently focused on these topics:
(1) Simulation of photochemically triggered reversible molecular changes: Such molecular “switches” can be used in “smart materials,” to switch desired properties on and off, or as motors in molecular machines.
(2) Design of reactive force fields: This is difficult, since it amounts to compressing all possible results of the Schroedinger equation into a far simpler functional expression. However, when successful, this reduces the effort needed to simulate chemical reactions by about 4-5 orders of magnitude.
(3) Non-deterministic global optimization: With techniques like Genetic Algorithms, good approximative solutions to huge optimization problems can be found with a very small fraction of the effort needed for an exact solution. We use this approach not only for new reactive force fields but also to find near-optimal ways of molecular aggregation (which is important e.g. in atmospheric chemistry: ozone depletion), to fold proteins, etc.
Q. What are your thoughts on Open Access scientific publishing?
A. In typical German universities, the traditional publishing model has become a big problem: The journals we want can only be bought in big bundles together with many others we do not really want. Even more annoyingly, during the past decades, journal subscription prices have increased enormously. With limited library budgets, we have to cancel subscriptions literally every year. Of course, with Open Access, publishers still have to make money. Therefore, in some abstract sense it makes no difference if money is paid for publishing one’s own work (which we have to do to get recognition) or if money is paid for buying the works of others (which we have to do to see what others are doing). For the typical academic author at a university, these payments are made from different sub-budgets, but ultimately all of them belong to the same total university budget. Nevertheless, Open Access has the potential to be something like a “system restart.” Perhaps some of the worst aberrations of the old system can be avoided in the new system, at least at first. A true revolution of the publication system could be far more radical: The typical scientific paper of today already is publication-ready when first submitted as a “manuscript” to a publisher. All the typesetting and layout is already done, the language is as good (or as bad) as it will be in the final product, and tables and figures can be printed as they are. Peer review is done for free by my peers (as I do it for them).
So, what is keeping me from simply depositing my “manuscript” as a final paper in an online repository set up by my university for free, instead of publishing it commercially? With all the internet search technology of today my work can then be found there just as easily as other web content, and it could be downloaded for free, as part of the public service duties of my university. Seriously, what has a publisher to offer for the money we pay (be it traditionally or open-access)? A publishing executive at a recent talk at my university only had this reply to this question: “We offer reputation.” (i.e., impact factor, see below) I do not think the reputation can be so great that we can buy less and less of it every year. We should rather turn to radical alternatives like the one just indicated.
Q. What made you decide to join the Editorial Board of ScienceOpen?
A. As mentioned in my replies to the previous question and to the next one, the present system of scientific publishing has very serious flaws. Therefore, anything that is significantly different has a good chance of improving this situation. And instead of just complaining I thought I could try to take on a slightly more active role (but not much, since I still want to do science rather than publishing, as my main occupation).
Q. ScienceOpen has adopted a post-publication, transparent peer review process rather than the traditional blind pre-publication model for its publishing and networking platform. What are your thoughts the more traditional peer review model and what possibilities exist for improvement?
A. With almost 25 years of experience as author and reviewer in the traditional model, it is clear to me that this model has severe deficiencies:
(1) At least 8 out of 10 referee reports carry little scientific weight if any. Instead, they show that the referee has not understood major portions of the manuscript, has not read it sufficiently carefully, only wants his/her own work to be cited, etc. Generally, far too little effort is put into preparing referee reports — which is no surprise since the referee is not paid and is strictly anonymous.
(2) The position of the author is far too weak. Starting a serious, factual argument with a referee leaves the author almost no chance to win, and it would delay publication by many months. Therefore, the most successful strategy is to pretend that the referee is always right, and to try to appease him/her with the smallest changes possible. Typically, the referee then gives in after just one iteration. Already at this point, this is an exercise in futility.
(3) It is a common (but unchecked) assumption that the quality of refereeing is better in “better” journals (and that therefore your work is indeed better if you “manage” to publish it in a high-ranking journal). This is not true. I have been there. There is little to no difference in the quality of refereeing between different journals (unless one is willing to drop down to the bottom of perceived journal quality). What really makes a difference is the personal, subjective stance of the referee. If he/she likes your line of work (for whatever reasons), your manuscript will be published. If not, he/she can easily ask for amendments of the manuscript or of the work behind it that cannot really be fulfilled (or can simply refuse to accept the amendments as sufficient).
(4) Therefore, when an author receives a negative referee report, the best reaction is to retract this submission and to send the manuscript unchanged to another journal. Chances are good that it will be published there with far less additional work than a proper reaction to that negative report would have required. As referee, I have had dozens of manuscripts I initially criticized re-appearing on my desk later, with little or no changes, but coming from a different journal. And this then was not always a decline on the ladder of journal quality; just as frequently the perceived quality of the second (or third) journal was higher. In fact, it looked mostly like random choices. Obviously, this “journal hopping” (in which previous referee reports magically disappear in most cases) further degrades the value of referee reports. (Interestingly, several journals now require a history of previous submission attempts from the author, apparently in an effort to suppress this behavior — but there still are sufficiently many journals out there without this requirement.)
Clearly, a more transparent review process can ensure that more effort is put into refereeing, that the author has a fairer chance for a serious reply and for amendments of his/her work (and is forced to also take that more seriously), and (if all publishers adopted such a model) that initial negative reports cannot be made to disappear.
Q. What would be your advice for young and up and coming scientific researchers? In particular when they are considering publishing opportunities for their work?
A. Sadly, the impact factor craze (i.e., the assumption that if you publish in a high-impact journal you are doing good science) is so wide-spread that it really influences essential decisions: if you get funding for this grant application, or if you win the competition for this new position. Yes, scientific content also matters, but only to some degree. Too many people believe too firmly in impact factors. You can really influence a search committee simply by pointing out that candidate X has 20 papers in the famous journals Y and Z. If you want to convey the same impression for candidate A who only has 20 papers in the mediocre journals B and C, you have to get into a long-winded argument. So, if you are not so sure of your scienfitic abilities, publish in the “top journals.” If you are absolutely convinced that you are doing top science, you have some freedom to also try other publishers, including Open Access and ScienceOpen. Or you are convinced (as I am) that this publishing game is obviously seriously flawed — and that this flaws are self-perpetuating: Despite my misgivings for impact factors, I also read almost exclusively what I perceive as “top journals” in my field (not by impact factor but by the typical level of quality in their contents), simply because I cannot possibly read all those millions of journals out there. In that case, you and I have some obligation to break out of this vicious circle. This could be another reason to also try other publishers, including Open Access and ScienceOpen.