Why Do We Publish?

If a tree falls in a forest, and no one is there to hear it, does it make a sound?

1Day (1998) suggests this question as a starting point for students to understand the importance of publication in science. It is a good starting point, and most students’ first response is: “Yes, it does”. However, this immediate answer happens to be wrong. Few people notice first (although realise on reflection) that, by definition, sound is not equal to “pressure waves in the air”. When a tree falls, it generates air pressure waves. This is not a sound. This is the physical phenomenon that is a necessary condition for a sound — but the other half, the receiver, is missing. Sound is a sensory experience that is generated in an organism’s receptor organ by the air pressure wave. The pressure waves in the air become sound when someone is present who does sense (hear) it. A “sound” assumes a perception of those waves by an organism.

2Following a similar line of reasoning, we can ask: if someone discovers a law of nature, proves it in well-designed experiments, the results of which are carefully documented, and then puts this document into her drawer, never to publish or show anyone, does this information exist? The law of nature in question existed before the experiments were carried out, and continues to exist independently (provided you believe in objective reality), but it only becomes “information” when we know about it, when we become informed about it, typically through the publication of a peer-reviewed scientific manuscript. Knowledge (or information) that is not published does not, effectively, exist, because there is no one who knows about it, and can use it. If it never comes to light (i.e. the information is not published), then doing the experiments, and (however carefully) documenting the results, was pointless and of little use — the law must be rediscovered by others. It only becomes existing information if there is a realistic chance that others can know about it. Publication is therefore a necessary, integral part of scientific research, not something that comes after the experiments (the “science”) were finished. The more one can integrate the experiments with their description for publication, the easier the task, and the more convincing the results will be.

3When something is published, we refer colloquially to it as “a known fact”; however, if we reflect, we realise that there is no piece of information that is universally known. Even putting aside the practical difficulties of access, knowledge of language, etc., what about people who are illiterate? What about infants, the intellectually disabled, those who, perhaps, due to an accident, suffered brain damage, or the dead or the unborn? Do we exclude them as members of humanity? No, we do not. Thinking along these lines leads us to realise that the conditions according to which we generally consider something “known” have always been a matter of agreement. If a publication (in the form required by that tradition) fulfils the criteria posed by the agreement, we generally consider the discovery as having been “validly published”, and, thus, existing knowledge, irrespective of how many people actually know it, or understand it. Einstein’s theory of relativity is probably not understood by most of us, yet there is no doubt about attributing the concept to his name. Today, most scientific discoveries are published in scientific journals. The chief criterion for such publication is that experts evaluate the reliability and novelty of the findings described. This “peer review” is a necessary step to valid publication of a new discovery.

4This tradition of publishing new scientific discoveries/knowledge grew out of the practice of scientific societies, established at the time of the Enlightenment, when scientific research in Europe started to blossom. Members of those societies met regularly to discuss their discoveries. The first scientific journals started as the written official records of the meetings of such societies. The Proceedings of the Royal Society of London, one of the oldest scientific journals, is one such journal, and it occasionally still publishes a brief record of a discussion, when the article is based on an oral presentation. This is a remnant of the old practice when an oral presentation was followed by a discussion, during which fellow scientists present at the talk discussed the reliability of the discovery presented. Once the consensus emerged that the reported facts could be trusted, the presenter was credited with the discovery, and this was, from now on, considered “existing knowledge”. Since then, scientific publishing has changed considerably, and many different types of scientific publications have emerged.

Types of Publications

5Publications take many different forms, but types of scientific publication fall into fewer categories. I do not attempt an overall, definitive taxonomy of scientific communication types, and only restrict this to a few important distinctions.

6The different types of theses, connected to higher education, are ones that most scientists must get acquainted with. This is often the first type of publication they themselves will have to write. Another common written publication type is the scientific article, most commonly meaning the “proper” (primary) article. In addition, there is the short communication, review article, invited article, note, comment, letter, and so on. Journals often publish book and software reviews. Scientists also write papers for conference proceedings, reports for different organisations, book chapters and books. I will briefly survey these later, but the main intention of this book is to help the reader to write a so-called primary scientific paper (hereafter “paper”).

7Primary scientific papers are published by specialist journals. The three general types of these are: the “society” journals, the commercial scientific journals, and the small, specialist journals. These are similar in that all of them publish peer-reviewed, primary scientific papers. This similarity of their shared primary purpose hides significant differences.

Society Journals

8Most fields of science have an international society. These societies are typically not-for-profit entities, and their aim is to promote their respective fields of science. One of the ways they do this is by publishing an international scientific journal. These journals are, as a rule, of high scientific quality, well-produced, inexpensive (especially in comparison to the ‘commercial’ journals) and, consequently, accessible and widely circulated. The key to a high level of science published in the papers is the good peer review process that is provided by the wide range of volunteer experts in the field, who are members of the society. Members of the society can subscribe to the journal at very favourable rates and, because these societies have thousands of members, this provides the publisher with a wide circulation, ensuring that the journal is widely available in the academic community, and many people read it. Due to this wide circulation (and high scientific prestige), many scientists will send manuscripts for publication, so the journal can afford to publish only the very best of them. As the circulation is high, the subscription prices (for members of the relevant society) can be kept relatively low, because, even at low profit margins, the journal will generate significant profit. The societies and the publisher share these profits. (Societies may periodically re-negotiate the publishing contract and may change publisher).

9These journals are usually the most coveted ones of the field to publish in, and they usually sit near the top of the citation statistics summaries. Examples of such journals include (in parentheses, the parent society): Ecology (Ecological Society of America), Journal of Animal Ecology (British Ecological Society), American Naturalist (American Society for Naturalists), Journal of Experimental Biology (Society of Experimental Biology), New Phytologist, and Weed Research (European Weed Research Society).

Commercial Journals

10The commercial success of the top scientific society journals lured several commercial publishers into the field. Several publishers (e.g. Elsevier, Springer) have aggressively pursued the policy of founding new journals. Some of these journals took root and became respected in their fields, whilst others merged, or disappeared. The principles for editing a successful scientific journal are the same as in the case of the society-run journals. The main difference, however, is that these journals are not published as a non-profit venture to promote a field of science. The peer review is usually organised by using a smaller number of experts, often dependent on the personal networks of the editorial board members and, thus, it is not always possible to find top-level experts for every manuscript. Further, these journals are published for profit. Consequently, they are very expensive, which often results in limited circulation. A good example of commercial scientific journals is the group of journals published by Springer Verlag (including Oecologia, Planta, Molecular and General Genetics, Theoretical and Applied Genetics, among many others). These journals, more often than not, occupy the middle range in citation statistics.

Small Specialist Journals

11There are numerous scientific journals published by small, specialised institutes, such as museums, research institutes, university departments, and botanical or zoological gardens. Such journals are of high quality within their specialist fields, but there are several unfavourable aspects of such publications. Their narrow specialisation means that the potential readership is small. Consequently, the circulation cannot be as high as for the more general journals and, thus, such journals must be relatively expensive. Moreover, the institute often has to fund the publication itself and, at times, this is not its highest priority. Consequently, the publication may appear infrequently, perhaps even irregularly. Thus, although the peer review in the journal is good, often very good, such journals, due to mostly logistical and economic factors, are of limited general impact. Examples include Proceedings of the Missouri Botanical Garden, Annales Musei Nationalis Hungariae (published by the Hungarian Natural History Museum), etc.

12The boundaries between these categories have started to blur. Today, many societies and small institutes entrust the production of their journals to professional publishing houses. The publishers are also “on the hunt” for journals that they can buy and take over, or secure a contract to publish on behalf of a society, because it is much easier to continue, or re-organise, an existing journal than to start a new one. The main differences still remain: that the best and widest reviewer range is available to the editors of the society journals and, thus, they can have the best quality specialist reviewers. High-quality peer review is the key to a journal’s quality. The society’s members also provide a secure subscriber base and, hence, the society can negotiate a favourable subscription rate for its journal with the publishing house. The publishing house benefits from the large pre-existing subscription base.

Internet Publishing and Internet-Based Publications

13The Internet has had a significant impact on scholarly publishing. The electronic processing of manuscripts has become near-universal. Several major publishing houses operate “electronic manuscript handling centres” that are intended to speed up the publication process. While the principles of electronic scholarly publishing are not different from the paper-based publications, there will be further changes that need to be followed by scientists aspiring to publish.

14Many journals today are published electronically as well as in printed form, but there is an increasing number of journals that are only published electronically — for example, the Public Library of Science (PLoS) range of journals. From a commercial point of view, the enthusiasm of publishers to embrace “free” electronic journals is understandable. These journals are free to readers — but not free to the authors wishing to publish in these journals. As electronic publishing still costs money, the publishers, if they publish an Open Access journal, can have a faster return on their investment, because the authors pay these costs before their article is published. Income security is also improved: the profit no longer depends on the number of readers or subscribers — while in traditional publishing, it does. That said, there are other publishing models in existence, and some of them are “truly free”; it may be worth checking our the list on www.doaj.org (read attentively, because not all journals on that list are free for aspiring authors).

15When considering a submission to an electronic -only journal, it is wise to think about similar factors as for the “traditional” journals (see above). Do not publish in a journal that is not well-known and respected in its field. No matter how quick, or cheap, the publication is, publishing in an electronic journal of no reputation is a waste of your work (and money). If you are careless, your work will not even be published. Several unscrupulous “publishers” will happily collect your money and never publish your paper. Avoid this disappointment.

On Predatory Journals

16Scientific publishing is a profitable activity, and the possibility of making large profits has lured more than a few unscrupulous players to the field. They range from publishers who run journals with “very soft”” (uncritical) peer review, to those that have little or no quality control, those that publish all submitted manuscripts unedited and also those that trick prospective authors to part with their money but never see their manuscript published. The number of these publishers (judging from the number of messages in my mailbox) does not seem to be decreasing, thus there must be enough scientists being misled who keep them afloat. Consequently, a few words of advice are in order so that the reader can avoid falling into any such traps.

17Because traps they are: if your work is not published but you have paid up, of course you are worse off. However, an article published in such a forum brings you no benefit; actually, your reputation will be tarnished. You should absolutely avoid such journals. But how can you recognise them?

18A proper scientific journal relies on declared peer review, and does not try to impress prospective authors with unrealistic claims. Predatory forums (I am reluctant to call them “journals”) advertise themselves aggressively, usually through unsolicited e-mail messages, inviting you to become their editors or to submit manuscripts that are, as a rule, promised to be published with extreme speed. Their homepages are unsophisticated, frequently contain glaring errors, display no, or very few, published articles, or you cannot find any content behind apparently “published” articles. Do not rely on their list of editors, because they often list editors’ names without their consent. Librarian Jeffrey Beall started a website listing such journals but it was discontinued due to harassment and threats. The scientific community has since revived the list and you can find more information and lists of predatory journals, hijacked journals, predatory publishers and useful related information at: https://predatoryjournals.com.

Primary and Non-Primary Literature

19The first, and most important, matter in publishing is the distinction between primary and non-primary publications. A primary publication is defined by the Council of Biology Editors (CBE, 1994) as:

[...] the first disclosure containing sufficient information to enable peers to 1) assess observations, 2) to repeat experiments, and 3) to evaluate intellectual processes; moreover, it must be susceptible to sensory perception, essentially permanent, available to the scientific community without restriction, and available for regular screening by one or more of the major recognised secondary services [...].

20This definition was carefully considered and merits some explanation. Naturally, the novelty of the scientific information is paramount. More than one publication of the same information is not possible, mainly for ethical and legal reasons. Thus, it also must be a “first disclosure”.

21The availability criterion has three main components. First is the language. For a long time in Europe, the language of science (as well as culture in general) was Latin. With the development of modern national languages and nation states, this monopoly gradually gave way to different languages. The advancement of the field often spurred the increased use of the language of the country where particular developments took place. Thus, German was an important language of taxonomy and biology in the 19^th century, as well as of physics in the first half of the 20^th century. Today, the language of international communication in biology (and many other scientific fields) is English. This may put an extra burden on non-native English speakers, but it makes international communication rather simple and (potentially) effective. Practically, papers published in a language other than English may qualify as a primary publication in the case of an intellectual property right dispute but, in “everyday” science, they have rarely been considered as such. One proof of this general perception is the existence of numerous scientific papers published in a less accessible language (usually in a small journal that publishes articles in the language of its country) and also in English, in an international journal. A word of caution, though: journal editors may now consider such articles as attempts to double-publish (Uchmanski, 2019). Perhaps the increased sophistication of automatic translation services is behind this shift — whether this becomes a trend, or the above example remains an oddity, it is too early to tell.

22The second availability criterion is circulation. Peer-reviewed journals are much more widely circulated than other publications, such as yearbooks, government publications, and conference proceedings. This is the main reason why papers published in the latter types, irrespective of their quality, do not qualify as primary publications. However, the rapid spread of Internet-based publishing will soon make this criterion obsolete: now even small universities are making full MSc and PhD theses freely available on the Internet via repositories.

23It is also important that a primary publication is covered by at least one, and preferably more, of the widely circulated abstracting publications and online scientific databases. Every field has at least one of these, with Index Medicus, Biological Abstracts, Web of Science, or Zoological Records being widely used. These publications make available the bibliographic information of a publication, the addresses of authors, keywords and the abstract. Originally produced in print, they are all now online, and their use has become essential in keeping up with the rapid increase of the scientific literature. As no library can afford to have all, or even most, important primary publications in any field of science within their collections, access to journal articles has become a two-step procedure: a paper spotted in one of the abstracting services or databases can subsequently be looked up and obtained by going to the original journal itself. Finding a paper directly in the original journal by searching the archives of that journal is possible but, due to the sheer number of new papers, rarely happens.

24Finally, it is very important that the new information is evaluated by peers, and this evaluation takes place before publication.

What Is “Peer-Review”?

25Peer review is the process of obtaining written evaluations of a manuscript by experts in the field. On arrival to the editor’s office, the manuscript is sent to one or more (usually at least two) experts of the editor’s choosing, who will read and comment on it. The peer reviewers should consider the presentation of the problem, its significance, the suitability of the methods chosen, evaluate the reliability and the significance of the results, the views and interpretations expressed in the discussion, the necessity and clarity of the tables and figures as well as the language. The reviewer is required to send the editor a written review that evaluates the above aspects. The reviewer is also requested to make a definite recommendation about the acceptability of the manuscript for publication (usually: accept, reject, or accept with minor/major modifications).

26This review is often (but not always) anonymous. This is the “single-blind” system: the reviewer knows who the authors are, but the authors do not know who the reviewer was. A reviewer can always sign the review, but only if she so wishes. When the peer-review system is “double-blind”, neither party knows the other: the reviewer does not know the authors’ names, because these are removed from the manuscript, and the name of the reviewers are not released to the authors.

27The system is not faultless, and various assumed biases have been long debated (Weber et al., 2002; Smith, 2006; Resnik, 2011; Sopinka et al., 2020) but no one has come up with a better system of evaluation that has been widely adopted. The reviewers are the editor’s essential helpers, who, collectively, try to make sure that the quality of science published is reliable.

28An interesting new system was started in 2011 by Finnish scientists, the “Peerage of Science”. They are trying to reinvigorate the peer review system by forcing it to conform to deadlines, a clear track record and mutual assessments (see more at www.peerageofscience.org). The reward for participating reviewers is a record of keeping deadlines, attracting a score of the quality of their review from their peers. Another attempt to document peer-reviewing activity, and to get credit for this, is the Publons initiative (www.publons.com). Others (Fox and Petchey, 2010) have suggested “privatising the commons” where only those who themselves contribute to it can enjoy the benefits of peer review. Yet another attempt is to completely remove the review process, and let the reading public, collectively, decide about quality. Unfortunately, this last system seems doomed simply because of the huge amount of information available, regardless of any other factors that make it attractive or not. The Internet is already the least reliable of information sources (see Chapter 14), precisely because of the lack of systematic quality control of its content.

Non-Primary Publications

29There are many types of publications that do not fulfil the criteria of a primary publication. These are not worthless, nor are they necessarily of low quality. Some of these are mentioned below.

Conference Proceedings

30Thousands of conferences are organised every year. These can be very useful occasions for discussion, presenting new theories, findings, and so on. Frequently, the conference will have a permanent record in the form of conference proceedings. These are, mostly, occasional publications, containing written versions of varying length, detailing the talks and posters presented at the conference. Conference proceedings are usually published in small print runs, and distributed only to conference participants plus a few libraries that seem to be selected according no particular set of criteria. It is, therefore, uncertain as to whether even the largest scientific libraries will have a copy of the proceedings of any particular conference. They do not qualify as primary publication, because they usually fail to fulfil the circulation criterion. More and more such proceedings, however, are electronically published, so one may argue that on the availability criterion alone, they cannot be classified as non-primary. Other problems, however, remain.

31One of those is the quality of editing. There are no strict rules about the review process for the inclusion of a specific contribution in a volume of conference proceedings. The editors are usually the main organisers of the conference and, often, do not have editorial experience. Additionally, the conference frequently does not have enough money to pay for the services of professional technical editors, which leads to very variable production quality and publication speed. All too often, such proceedings serve as a permanent record of the conference, a testimony of the organisational work, rather than a vehicle for scientific information—and the thicker the volume, the more important the conference is supposed to have been. This is fallacious logic.

32Occasionally, conference proceedings are published as special issues of a scientific journal — in which case they are subjected to the same review process as any other manuscript submitted to that journal. Not all contributions presented at the conference will necessarily be included, but the ones that satisfy the process should be considered as valid primary publications. In any other case, do not consider conference proceedings as primary publications. This does not mean that you should never write one — for advice, see Chapter 27.

Government and Institutional Reports

33These are usually obligatory reports written on work done using government grants, expert committees, or an output of commissioned research. Their primary purpose is to inform the funding organisation about the work carried out, often concentrating on the recommendations emerging from the project. They rarely qualify as primary publications because they are seldom peer-reviewed, have limited circulation, and are often in a local language.

Review Publications

34A special set of non-primary publications are the review journals. These exclusively publish reviews of published literature in a specific area. Because reviews are the first source of synthetic information when one wants to get quick orientation in a new field, review journals are read by many people, and can be very influential. An additional reason for their popularity is that, with the increase of the primary literature, more and more people find it difficult to follow the developments from reading the primary literature only. Review publications (Box 1) can be yearbooks such as the Annual Review series, and even monthly, like the journals in the “Trends” (Trends in Biochemistry, in Ecology and Evolution, in Genetics, etc.) series.

Reference Literature

35There are only estimates of how many scientific journals are published today (about 55,000, Gastel and Day, 2016). There is no library where all these can be found. The need to be informed about the availability of a new piece of scientific knowledge is greatly met by the existence of reference journals or publications. These publish only selected parts (sometimes only the reference) or summaries of primary scientific articles. The idea is to provide some extremely condensed information about the article, and its necessary bibliographic data, so that interested readers can find and read the original, full-length version. Many of these, until about the late 1980s, were produced by teams of reviewers who received, read, and abstracted the original papers. These “abstracting” journals have titles such as Biological Abstracts, Plant Protection Abstracts, etc., and contain summaries of varying lengths of selected original, longer papers, plus assorted indices to help interested readers to find the sources if needed. Today, probably all of them are produced by at least semi-automatic means, containing the (usually unaltered) abstracts of the papers, and various computer-generated indices of topics, authors, and affiliations.

36The most successful of these, originally named Current Contents (which became Web of Science), and published by the Institute for Scientific Information in Philadelphia, USA, also gave rise to the science of scientometrics.

37By a stroke of genius, the founder of Current Contents, Eugene Garfield (1925-2017), abandoned the manual extraction of information from articles, and only published the photographed contents pages of journals, plus indices of topics, authors, and addresses. Anyone reading the title of an article, if she was interested, could find the contact details of the author, and could ask for a full copy of the original article. The disadvantage was that a reader had to decide, after reading the title only, whether the article in question contained relevant information. However, the advantage was huge: while normal abstracting journals usually published their more detailed, but still partial, information about published papers with a delay of 6 months to -2 years following publication, Current Contents published the (admittedly more limited) information in 6 weeks — to 6 months. This was possible because most of the work could be done much faster: it only involved photographing and printing the contents pages, and generating the relevant indices. This time-saving method proved the definitive factor in the success of Current Contents.

38Current Contents was organised in sections, and the original, selected range of journals has gradually expanded to include, today, >18,500 scientific journals. The latest development has extended to the listing of conference proceedings and books. There is still a bias towards US-based scientific journals published in English. When the Canadian publishing giant Thomson Reuters became the owner, several doubtful practices emerged as they aggressively, and rather inconsiderately, extended the claims for what such a database can achieve. Now this database (Web of Science, or Web of Knowledge) is no longer part of the Thomson-Reuters media empire and, despite vigorous protests from academia (see www.sfdora.org), the current website still claims that it offers “an objective analysis of people, programs and peers”.