Thoughts on Publishing the Research Article over the Centuries

: The ﬁrst academic periodical was the Journal des Sçavans , which ﬁrst appeared in January 1665. It was followed two months later by the Philosophical Transactions . The Journal des Sçavans was sponsored by the state and was made up mainly of book reviews and covered all the known disciplines of the time. The Philosophical Transactions was a private venture based on Oldenburg’s correspondence and was restricted to science and technology. Scientiﬁc writers were motivated by personal reputation, the desire to improve the human condition, and, sometimes, priority. The “publish or perish” syndrome is a recent development. Among the factors that have inﬂuenced it are the increasing professionalization of science, the development of the peer-review system, and, towards the end of the twentieth century, a desire for rapid publication. The fact that English has (recently) become the lingua franca of scientiﬁc publishing creates additional difﬁculties for non-Anglophone scientists, which their Anglophone colleagues do not have to face. Scientiﬁc language, similar to all languages, evolves constantly. One area that seems to be changing at the moment is that of passive use, which is the subject of ongoing research. Cultural differences may also have a role to play. For example, French scientists may have to overcome a basically Cartesian education.


Introduction
Publications have become an essential facet of academic activity in the contemporary world.Publication implies writing and the language in which academic research is couched has its specificities.Within that general area, different subgenres will also have their own particular traits.The academic, writing for the purpose of communicating his research, has to adhere to the linguistic features expected or face the possibility of seeing his articles refused by the gatekeepers of the journals in which he wishes to publish.While it would be impossible to be exhaustive on such a vast subject in the space of an article, in this paper I would like to consider certain aspects of this question.
I shall start by looking at how the academic article started in the late seventeenth century and what motivated the men involved.I shall do this in some detail since this is the basis from which the present situation has developed.I shall look, rather more briefly, at how this has evolved up to the present day with the relatively recent development of publication as a requirement for career advancement.In view of the current, though also fairly recent, development of English as an academic lingua franca, I shall consider the implications for non-Anglophone academics in terms of both the linguistic problems involved and the point of view of cultural differences resulting from their philosophical stance.As an example of the linguistic problems involved, I will consider the use of the passive voice.While this has long been recognized as a significant feature of scientific writing, recent research seems to show that changes in its use are in progress.In attempting to cover such a wide area, it is inevitable that some aspects will be treated in greater depth than others and some may be glossed over altogether.same basic objective.This can even be traced to differences in the linguistic features found in the two publications.For example, analysis of the themes (in the linguistic sense of a clausal structure made up of a theme and a rheme) shows that the Journal des Sçavans has a thematic bias toward people (43% of themes) and to a lesser extent texts (16% of themes), whereas the object of study accounts for only 22% of themes.This can be linked to the fact that the Journal des Sçavans was made up mainly of book reviews.The thematic focus of the Philosophical Transactions, on the other hand, is the object of study (45% of themes) with people accounting for 27% and texts 7%.This derives from the decision to restrict the content of the Philosophical Transactions to science and technology.Similarly, an analysis of the use of modality shows that while examples of epistemic modality are similar in the two publications (23% of examples of modality in the Journal des Sçavans and 20% in the Philosophical Transactions), there are more examples of dynamic modality in the Philosophical Transactions (76% as opposed to 59%) and more examples of deontic modality in the Journal des Sçavans (18% as opposed to 4%).This results from the greater interest in physical actions and events in the Philosophical Transactions, and the inclusion of disciplines where moral issues might be at stake, such as theology and law, in the Journal des Sçavans [1].

The Motivation of Early Scientific Authors
The question then arises as to what motivated those involved in the scientific activity of the early modern period and who wrote the articles that appeared in the Journal des Sçavans and the Philosophical Transactions.Almost all of these men were virtuosi, that is, they were gentlemen of private means for whom science was a leisure activity, even if for some it took up a great deal of their time [15,16].Moreover, they rarely held university posts.At that time, it was generally thought that the development of humanity was similar to that of the human body and that humanity had been in its prime in the era of ancient Greece and Rome, and that by the seventeenth century it was already in its old age.It was thus thought that the "Ancients", notably, Aristotle in physics, Galen in medicine, and Ptolemy in astronomy, had reached the pinnacle of human achievement.Consequently, trying to outdo them was pointless; therefore, university endeavors were restricted to argumentation based on ancient texts and explaining their significance in the modern world.Experimentation and research had no place in such a scheme of things [17].There were, of course, exceptions: Newton held a university post, although it was in mathematics, not physics.A little earlier, some future members of the Royal Society had been given university posts when Cromwell came to power.However, these were more or less political appointments and most members lost their post during the Restoration.Therefore, most of the members of the Royal Society and like-minded people worked outside the university system.There were a few who were not virtuosi, such as Henry Oldenburg and Robert Hooke [18,19], but they were the exception rather than the rule.
The situation in France was a little different due to the creation of the Académie Royale des Sciences in 1666.Unlike the Royal Society of London, which was "Royal" in name but received no financial support from the crown, the Académie Royale was established and maintained by the state.It was provided with premises in which to carry out its scientific activities and its members were paid a retainer by the state [20,21].
It is clear that in the English situation, publishing research had no bearing on advancement in a career since for the vast majority of those involved there was no such thing as a scientific career.It seems probable that for most, the personal reputation and being recognized for one's achievements in what they believed to be a field that was new, experimental, and extending the bounds of human knowledge, was a major motivation.This was done very much in the spirit of Bacon who saw scientific endeavor as a way of improving the lot of humanity.Bacon was the maître à penser of the members of the early Royal Society, and this was enshrined in Sprat's 1667 account and defence of the Royal Society [22].Thus, personal kudos could be combined with the altruistic idea of helping humanity and it is all too human that this should be attractive.For a smaller number, questions of priority, particularly if commercial interests were involved, were of importance.Priority disputes were fairly frequent, one of the most significant being that between Robert Hooke and Christiaan Huygens on the invention of the spring-balance watch.In 1675, the Philosophical Transactions printed an item, translated from the Journal des Sçavans, in which Huygens described a spring-balance watch that he had invented [23].Hooke riposted vigorously, claiming that he had invented just such a watch much earlier, and a bitter dispute developed.It seems that Hooke had indeed put forward the principles of this type of watch but that he had never carried it through to the production stage [18,19].Other disputes have occurred, such as that with blood transfusion [24] and the well-known disagreement between Newton and Leibniz about the invention of calculus [25].Although Newton manipulated the Royal Society, of which he was president, to underwrite his claims to priority [26], it is the symbolic system of Leibniz that has survived.
In France, the fact that the members of the Académie Royale were paid meant that the situation was significantly different.They might be considered to be the first professional scientists and since they were frequently called upon to advise the government on scientific matters, the first scientific civil servants.Obtaining an appointment at the Académie Royale was much sought-after and there was certainly rivalry among the prospective candidates.However, publishing does not appear to have played a role in this system.Although the work of the members of the Académie was published in its Mémoires, these publications were peculiar in two ways.First of all, the publications of the Académie, in its early years, were anonymous.Their work was considered to be collegial and was presented as the work of the group as a whole rather than of any particular individual [2,20,21].As a consequence, those who did want personal recognition for their work turned to the Journal des Sçavans or, more occasionally, the Philosophical Transactions for publication.The second peculiarity was that the publications of the Académie Royale were not widely distributed.Since the Académie had been set up and was paid for by the crown, whatever it produced belonged to the King.Its purpose was to add to the glory of the Sun King through its scientific endeavour and this in itself was considered sufficient to legitimize its activities and results.The necessity of replication and the witnessing of experiments, so important to the English scientists, were superfluous for the members of the Académie and dissemination of their work was, thus, deemed unnecessary [2,27].Although the Mémoires were published, they appeared only in limited luxurious editions, printed on fine paper, and bound in morocco leather.The volumes were considered to be the property of the King who used them as gifts for illustrious visitors.The members of the Royal Society found it difficult to understand why copies of the Mémoires were so difficult to obtain [21].This situation lasted until 1699 when the reorganization of the Académie Royale included a decision to publish on a wider basis.

Publish or Perish
We have become so used to the "publish or perish" syndrome that it is perhaps useful to remember how recent the development is.It is probably no surprise to see that it played no part in the early development of the academic research article.Therefore, we may reasonably ask when it came into play.It is not easy to piece together the ways in which this has come about but the following are probably some of the pieces in this jigsaw puzzle.It is no doubt related to the gradual professionalization of science.The early Royal Society was not made up exclusively of men with a deep interest in scientific matters.Since it had no form of financial patronage, it relied almost exclusively on the fees paid by its members for its financial survival.Consequently, some members were elected more for their ability to pay than for any scientific prowess or even interest [14].It has been estimated that as many as 30% of those who were nominally members took no active part in the scientific activities of the early Royal Society [28].It was only in the mid-nineteenth century that the Royal Society finally became a totally scientific institution [14].
Another factor is the development of the peer review of papers submitted to scientific journals.Initially, the content of the early journals such as the Journal des Sçavans and the Philosophical Transactions were the personal decisions of their editors.In the case of the Philosophical Transactions, this remained the case until 1752 when it became the official organ of the Royal Society, and from then on its content was decided on by a committee.In 1832, this was put on a more formal footing and outside referees began to be used.In the case of the Académie Royale des Sciences, a form of peer review was built into its 1699 statutes [29].These developed into the full peer-review system as we know it today [30,31].
Also relevant is the fact that as far as we can tell, there was little pressure to publish and there are several notorious cases of delayed publication by scientists such as Newton and Darwin.Newton's Opticks [32] was written over a fairly long period.He first published an article on a theory of light and colors in the Philosophical Transactions in 1672.This raised a controversy, which at times became fairly bitter, particularly on the part of Hooke, resulting in Newton, a natural recluse, going into his shell and refusing to communicate on the subject for several decades.The writing of Opticks was virtually complete before 1690, and even then Newton delayed publishing it.It was only in 1704, after Newton had become president of the Royal Society and after the death of Hooke (in 1703), that the book was finally published [33][34][35].Curiously, the first edition in 1704 does not have Newton's name on the title page, though the preface does bear the initials I.N.[36].Opticks ran to its fourth edition by 1730, which is the edition usually referred to today.
A second example is the notorious case of Darwin's The Origin of Species [30], which he started writing in 1836 but did not publish until 1859.It used to be widely believed that the long delay before publication was due to his fear of the repercussions it might have, including those of scientific colleagues and the possibility of religious persecution.However, recent studies suggest that he was determined to finish other work in hand, particularly work relating to the Beagle voyage and his major work on barnacles, before really getting down to writing his theory on natural selection [37,38].It is also difficult to estimate to what extent the possibility of the imminent publication by Wallace of an almost identical theory played a part in precipitating the publication in 1859.The question of how much he knew, and when, has been the subject of recent controversy [39][40][41].At all events, Darwin knew that Wallace was working along the same lines as he was and, in any case, the question of evolution was "in the air" at that time; we know that in 1859 he published the single volume The Origin of Species that we now have, rather than taking the time to write the multi-volume work that he had originally intended [37].
Therefore, from the seventeenth century to the nineteenth getting things published rapidly does not seem to have been an important issue and this is despite the fact that in the course of the nineteenth century there was an explosion in the number of scientific periodicals, from about 100 at the beginning of the century to around 10,000 at the end [42]!This situation seems to have continued.Even as recently as the early 1970s, a sociologist could write in the abstract to one of his articles: A large proportion of the scientific community publishes very little.The elite, prolific scientists are neither dependent on nor rewarded for the publication of their research, and most work by average scientists is hardly ever acknowledged [ . . .]. [43] (p.1119) He goes on to argue that there is a charisma attached to scientific endeavor and it is the attraction of participating in this charismatic activity that motivates most scientists.Gupta [44] contends that in academia over the last 20 years there has been a move from a hierarchical communitarian situation to an egalitarian individualistic one; this he claims has produced (i) an emphasis on a 'finished' product of obvious immediate utility, (ii) intense competition among peers, resulting in vicious, irrational reviewing of scientific papers and the promotion of 'cartels' among the scientific community.[44] (p.391) No doubt, the "publish or perish" syndrome fits into this scheme of things.

Publishing and the Non-Anglophone Scientist
The fact that publishing has become an imperative in the scientific world of today is compounded for the non-Anglophone scientist by the fact that English has become the international language of science.This poses problems for the language training of future science researchers [45] and for the linguistic help required by practicing scientists [46,47].While it would seem that English is likely to remain the international language for the foreseeable future, it is worthwhile remembering that this has not always been the case and that the elevation of English to this state is relatively recent.A study of the languages in which the natural sciences were published over the period 1880 to 2005 shows that in 1880 English accounted for less than 40% of scientific publications and that by 1910 it had actually fallen to 30%.From then on, the proportion of English publications rose but not always at the same speed: by 1930 it had risen fairly rapidly to a little less than 50%, then more slowly to the mid-50% by 1950.It was still in that range in 1970 but then rose dramatically to reach 80% before 1990 and over 90% by 2005.From 1910 to the mid-1920s, German had a greater share of scientific publications than English.In 1920, German accounted for over 40% of publications, English a little over 30%, and French just over 10%.In the social sciences, the situation is less stark but English is still the dominant language, rising from a 48% share of publications in 1951 to 76% in 2005 [48,49].Hence, although English is likely to be the international academic language for the foreseeable future, that situation has nothing inevitable or unchangeable about it.Just as the present situation has been brought about by political and economic factors, it is not impossible that unforeseen future political and economic developments should lead to changes in the current situation.
It is sometimes thought that the position of English today can be compared to that which Latin once held.Although it is true that Latin was the universal academic language during the Middle Ages and into the early modern period when scientists began using vernacular languages in their articles, and subsequently in books, there is one essential difference.During this period, no one had Latin as his mother tongue; it was not the native language of anyone and there were no monolingual Latin speakers.There is much to be said for having a universal language for scientific communication.In fact, it would probably be much more difficult without one.If every scientist wrote in his native language, all scientists would have to learn several languages to keep up to date in their field.This is obviously not realistic.The difficulty about the present situation is precisely that a large, and probably the most powerful, sector of the scientific community has the universal language as its mother tongue, thus creating a sense of unfair advantage for those who have to communicate, usually with a struggle, in a language that is not their first language.The idea of an artificial or constructed universal language is an old one [50,51].Towards the end of the nineteenth and beginning of the twentieth centuries, there were a number of attempts to create artificial universal languages.The only one of these to have survived in any form is Esperanto, which can hardly be said to have achieved universal status, but there were many others, including Volapûk, Ido, an offshoot of Esperanto, and, perhaps the most curious, Interlingua, an attempt to return to Latin, albeit in a vastly simplified form, "Latino sine flexione" [49].Consequently, English is with us for at least the immediate future and that is the situation with which non-Anglophone scientists have to contend.

Recent Changes in Scientific Writing
Scientific language, similar to all languages, is in constant evolution [52] and so quite naturally is still changing.I shall take as an example the use of the passive voice, which has been considered a hallmark of scientific writing ever since Barber's seminal article in 1962 [53].In a sample of issues of the Philosophical Transactions, the use of passive forms rose from a base of around 25% or less finite verbs in the early eighteenth century to over 30% by 1980 [52].Up to that point, there was no sign of the trend slackening but some recent work indicates that things might have altered since that date.In a series of articles [54][55][56], Seoane and her collaborators claim to have detected a decrease in passive use in scientific writing.They describe the decrease as "dramatic" [55] (p.107) and claim that it is greater in American English than in British English.This is not due to colloquialization, as might have been thought, as other markers of colloquialization have not followed the same path.In [54], she considers and excludes the possibility that a move towards a more subjective style or that scientific writing is becoming more personal and author-centered explain the change.Similar points are made in [56] where scientific writing is contrasted with legal discourse.
A study of recent scientific research articles does show that some sort of change seems to be underfoot.In the corpus that I was using for a recent study [57], I found sections written in the commonly expected passive form: Phylogenies were constructed and dated in MRBAYES v. 3.2.5 [39].All phylogenetic analyses were run for 20 million generations, sampling every 1000 generations, with four chains and four independent runs for each analysis.The heating parameter was set to 0.05 for analyses that included fossils and 0.1 for analyses that did not include fossils.Priors were set using established protocols [27] (see electronic Supplementary Material S1), and convergence was judged using inbuilt diagnostics of MRBAYES and TRACER.[58] (p. 2) [my highlighting] But I also found passages using the active voice with first-person pronoun subjects.
In this section, we illustrate the temporal evolution of the three generation-interval distributions (intrinsic, backward and forward) described by Equations (2.4), (2.11) and (2.8) with a simple epidemiological model.
In Figures 2 and 3 we use the well-known SEIR compartmental model (susceptible-exposedinfectious-recovered) where we include n E (resp.n1) exposed (resp.infectious) compartments in order to model realistic duration of latency and infectiousness with Erlang distributions (gamma distributions with integer shape parameter) [11].We will refer to this model as an Erlang REIR and details of this model are given in §3.[59]

(p. 4) [my highlighting]
These two extracts are both taken from the same issue of Proceedings of the Royal Society B for 2015.This study distinguishes between articles in the physical sciences (taken from the Proceedings of the Royal Society A, which deals with mathematical, physical, and engineering sciences) and the biological sciences (taken from the Proceedings of the Royal Society B, which deals with biological sciences).In this study of my (admittedly small) sample (and excluding papers by single authors), there does seem to be some evidence of a fall in the use of passive voice from 30% of finite verbs in 1985 to 25% in 2015 for the Proceedings A sample, though this was not reproduced in the Proceedings B sample.On the other hand, the use of active voice with first-person pronoun subjects as a percentage of passivizable verbs rises considerably from 18% in 1985 to 30% in 2015 for Proceedings A, and from 9% to 18% in Proceedings B. Therefore, subject to verification on a larger sample, there does seem to be some evidence that there is increasing use of first-person pronoun plus active form at the expense of passive, though this development is clearer in the physical than the biological sciences.Among the possible factors favoring this change is the increasing use of non-Anglophone referees, which may be leading to a more permissive attitude toward the expected linguistic features in scientific writing.Indeed, in my sample, I noted several examples of English that I would have considered as non-standard in my native variety.In single-authored papers, first-person pronouns remain rare but even here I found one single-authored paper in my sample that used them extensively.Therefore, one can only presume that gatekeepers do not find this an impediment to publication.
However, it would probably be wrong to imagine that some scientific authors are simply using we + active where they would previously have used passive voice.Closer investigation shows that in the 2015 Proceedings A sample where we + active is used, 75% of the verbs involved express a mental process (as used in Systemic Functional Linguistics [60,61] where the mental process includes cognitive verbs such as consider, estimate, and believe, and verbs of calculation or computer analysis, such as calculate, linearize, or derive).In comparison, only 4% of the we + active examples are of material process, that is, physical actions or events.In the 2015 Proceedings B sample, the corresponding figures are 54% for the mental process and 18% for the material process.Hence, the vast majority of personal pronoun subjects with active voice occur with mental process verbs and correspondingly few with material process verbs.This can, I think, be related to a phenomenon that began to occur in scientific writing towards the end of the nineteenth century in the physical sciences.Up to that time, articles in the physical sciences had been mainly experimental but from that date on mathematical modeling was introduced.This gathered pace during the twentieth century so that by the end of the century it had attained a significant, if not dominant, position [52].There was little evidence of this phenomenon in the biological sciences up to 1980 but the biological sciences now seem to be following the same route as the physical sciences.The introduction of mathematical modeling has had effects on the linguistic features of scientific articles.This is notably in the types of themes (in the sense of thematic structure seen as a distinction between theme and rheme) used to introduce clauses [60,61], which are now frequently of a mathematical nature, and the types of processes used with a strong increase in the use of mental processes of the mathematical calculation type.My hypothesis is that an author, where he is the agent, has a much closer relationship with a mental process, which is personal, cerebral and internal, than with a material process, which is external and involves entities in the external world.Hence, I would suggest that he is more likely to use a first-person pronoun with a mental process verb than with a material process verb [57].Thus, it might be thought that the situation has become superficially easier for the non-Anglophone scientist in that he is freer to use the active voice more than he would have done previously, but at the same time he still has to learn when that can be done appropriately.

Cultural Difference
A further difference that might cause difficulties for the non-Anglophone scientist wishing to publish in English is that of cultural differences.That might be obvious where the differences are large, such as with the cultures of Asian countries.It might be less evident for geographical areas much closer to home.I shall take the example of Francophone scientists.This takes us back to the seventeenth century.When the Royal Society of London was founded in 1660, it was based on the ideas of Francis Bacon, that is, observation and experiment are paramount and precede any attempt to theorize [22].A short time later this was given firm philosophical foundation by John Locke [62].This empirical attitude meant that the Royal Society was deliberately distinguishing itself from the inductive methods of Descartes and his followers.One of the most striking examples of this is the very first phrase of Newton's Opticks: "My design in this Book is not to explain the Properties of Light by Hypotheses, but to propose and prove them by Reason and Experiments."[32] (p. 1) In other words, he was not going to set up a theory that he could then test against available evidence, which would have been the Cartesian approach, but he would carry out experiments on the basis of which he could build a theory.He was nailing his flag firmly to the empiricist mast.This difference between the empiricist stance of the Anglophone scientists and the Cartesian approach of those on the continent of Europe can be seen in the linguistic features of their writing [35].Therefore, in the late seventeenth century, English and French scientists were working along different lines.It was only slowly that French scientists would gradually adopt the empirical approach.Maupertuis unsuccessfully attempted to promote Newton's ideas in France [63], but it was only after Voltaire and Emilie du Châtelet had taken up the cause and after the publication of Voltaire's Elements de la philosophie de Newton in 1738 that the tide really turned [64].However, although French intellectuals in the hard sciences adopted the empirical model, this never extended to the humanities and social sciences; hence, even today, French academics in these fields tend to have a basically Cartesian approach.Consequently, when French scientists face the problem of writing in English, they not only have to overcome the language difficulties but also the fact that their education within the French system has been geared to producing a Cartesian mindset.To the best of my knowledge, this question is not addressed in the linguistic training that French scientists receive.As other academic fields move toward publication in English, French researchers in these areas will be faced with the same problems.
In terms of Holliday's [65] distinction between small culture and large culture, we might say that French and English scientists belong to the same small culture, that of the European scientific community.But this overlaps with two large cultures, French and English, and the scientists in question will, in general, belong to one or the other.Holliday claims that discourse is the product of small culture but here we see that large culture also has an effect on the discourse produced.

Final Thoughts
We have seen that the academic research article was born in 1665.The motivation of these early scientists seems to have been self-fulfillment and, particularly where commercial interests might be involved, priority.Career advancement does not appear to have played a part even in France where members of the Académie Royale were, in a sense, state employees.Even in the nineteenth century, there was little pressure to publish rapidly and publications could sometimes be delayed for long periods.The rise of the publish or perish syndrome is relatively recent and probably the increasing professionalization of scientific activity and the peer review system in the selection of articles for publication have their part to play.For the non-Anglophone scientist, the position of English as the dominant, and for many virtually exclusive, language of publication poses problems.Although there may be a certain relaxation in the use of certain linguistic features, scientific authors still need to know when it is appropriate to use them.They need also to be aware of the cultural differences that separate them from the Anglophone mindset.
All language evolves and the language of the academic research article is no exception.Linguistic research can establish the present direction of change but it is impossible to foresee how long that direction will be maintained.I am doubtful about the possibility of forcing the development of language in a particular direction.Therefore, the linguistic analysis of developing academic language has a bright future and will be essential for non-Anglophone academics wishing to publish in English.