Atkinson: "The Philosophical Transactions of the Royal Society of London"

Atkinson, D. (1996). The Philosophical Transactions of the Royal Society of London, 1675–1975: A sociohistorical discourse analysis. Language in Society, 25(03), 333. doi:10.1017/S0047404500019205


Article Link







Atkinson''s analysis provides an insight into the development of scientific research writing by looking specifically at the Philosophical Transactions of the Royal Society of London (PT) from the years 1675-1975. The PT for quite a while monopolized the British and American scientific journal niche; it is still considered to be one of the, if not the, most influential scientific journals.


My interest in this article stems from having read much about the members of the society (e.g., John Wilkins); however, in addition to being referenced by Henze in "Scientific Rhetorics in the Emergence of British Ethnology, 1808-1848: Discourses, Disciplines, and Institutions," this work has also been cited close to thirty times.






Atkinson uses two methods to achieve his goal of "gain[ing] a fuller understanding of the development of English-language research writing over the last 300 years" (p. 337): sociohistorical register analysis (the Multidimensional approach or MD) and rhetorical text analysis. The author acknowledges that these are from different areas of language research, but argues for it as a means of triangulation.




For the rhetorical analysis, Atkinson defined a corpus of 202 research articles, collected at seven 50-year intervals (almost the entire span of the journal)These texts represented all original research articles in the first bound volumes of the journal.


For the MD analysis, Atkinson used a smaller corpus of 70 articles-- 10 per sampling interval-- selected using stratified random sampling procedures. The 70 articles "were then converted into computer-readable files for MD analysis" (p. 338). He compares the results with "those made by Biber 1988 for a large corpus of modern-day texts from 23 different genres of writing and speech" (p. 350).




Rhetorical Analysis:


Classifies results in areas of "cross-time textual development" (p. 338):


  • Author representation or place within texts. Atkinson notes that the use of first person pronouns and active-verb constructions steadily decrease from 1675-1975. In addition, he pays special attention to the use of language as an author''s way of providing insight to his/her mindset (e.g., use of terms like "expect," "trust," etc.) as demonstrated in the 1675 edition; by 1925, authors are effectively distanced in the text.
  • Preferred genres or discourse structures: Letters decrease from 51% of content to 0% by 1875. Experimental reports increase from 16% of content in 1675 to 59% in 1975. The structures change as well from predominantly narrative to more introduction -> experiment -> explanation to more theoretical pieces.
  • The discourse communities in which the texts are situated: Charts the changing structure of the community, from more dialogic to (in 1725) less cooperative to more cooperative to generalizing impact and more theoretical.


Atkinson notes that one consistent outlier was the 1725 edition, at which point the Society had a Newtonian in charge, who affected the vision and content of the publication.


MD Analysis:


Broken into five dimensions:


  • Dimension 1: differentiating involved vs. informational production. First and second person pronoun use, cognition verbs, present-tense verbs, etc., produce higher involved numbers while nouns, longer words, etc., produce lower (informational) numbers. From 1675-1975 Dimension 1 steadily dropped from 1.1 to -17.2.
  • Dimension 2: narrative vs. non-narrative concerns. High positives in this range would be "romantic, mystery, and science fiction" (p. 353); high negative scores would include "professional letters, academic prose, and official documents" (p. 353). From 1675-1975 Dimension 2 steadily dropped from -.08 to -3.3.
  • Dimension 3: situation-dependent vs. explicit-reference. Situation-dependent reference is "marked by the co-occurrence of place, time, and an ''other'' category of adverbs" (p. 355); greatest in radio broadcasts and conversations. Explicit reference "is indexed by three kinds of wh-relative clauses, phrasal coordination, and nominalizations" (p. 355); highest rates in official documents, technical communication, and the like. From 1675-1975 Dimension 3 (explicit references) steadily rose from 2.3 to 4.4.
  • Dimension 4: overt expression of persuasion. While there is a general movement towards a less-persuasive form (from -.3 in 1675 to -3.0 in 1975), both 1725 and 1825 are anamalous.
  • Dimension 5: Abstract vs. non-abstract information. Abstract information often means passive style. Generally, the move is towards the abstract, from 4.7 in 1675 to 7.7 in 1975. However, once again, two outliers: 1725 (perhaps expected) and 1975.




Atkinson maintains that the results of the rhetorical analysis and the MD analysis appear to correlate (classifying much of this as a shift from namely, the movement from "human -> acting on -> nature" [p. 360] to the nature/object taking prominence [even though acted on]. Atkinson then considers possible motivations for and perspectives on this change, arguing that author-centered discourse was a possible outcome of its "genteel-scientific linkage" (p. 363), and that "the ''I'' of early modern scientific reporting was thus not just the ''I'' of the author; it was the ''I'' of a whole social system" (p. 363). Starting in the early 19th century, however, one finds an "appeal to the systematic use of well-specified methods and materials" (p. 364) as well as (with the growth of specialist scientific disciplines) "depersonalized ''research problems''" (p. 365). By the early twentieth century, the shift away from methods and more to theory begins. Atkinson questions why this shift occurs, whether it be that methods were thought to be ineffective, or that they were ubiquitous and therefore unnecessary to describe.




Atkinson provides a great deal of detail in explaining his findings, and he provides many examples of the texts in order to make his connections clear. Atkinson is more interested in the "what" of his research than the "why"-- he does little theorizing as to the conditions and situations that may have promoted the changes in the text, using a very restricted definition  (in my opinion) of "rhetorical analysis." He confines his focus to the history of the Royal Society, and does not hypothesize as to other historical influences. As far as his research findings, Atkinson''s quantitative approach actually proves quite persuasive; once one can decode his table data, one can see a statistical correlation between time and rhetorical approaches. He addresses quite persuasively how the 1725 sample would be create some outliers, but the 1825 findings, which contain some anomalies, are not contextualized. That being said, this type of historical analysis is important to the technical communications field, as it reinforces that expectations or forms of science writing is not so much "natural" to the science but very much an episteme.





  • statement>thesis
  • purpose>contextualize
  • setting>archival>non-profit organization
  • design>case study
  • design>historical
  • design>interpretive
  • collection>artifact>written text> Philosophical Transactions of the Royal Society of London
  • analysis>rhetorical analysis
  • analysis>content analysis
  • analysis>discourse analysis
  • checking>triangulation
  • theory>rhetorical
  • theory>social/social constructivist





This study traces the evolution of scientific research writing in English from 1675 to 1975. Two separate methods of discourse analysis - rhetorcal analysis focusing on broad genre characteristics, and sociolinguistic register analysis --are applied to a large corpus of articles from the Philosophical Transactions of the Royal Society of London. The two sets of results are then interpreted vis-a-vis the Royal Society''s social history to yield an integrated description. Findings indicate that: (a) research writing in the 17th-18th centuries was substantially influenced by communicative norms of author-centered genteel conduct; (b) greater attention to methodology and precision in the interest of scientific specialization brought about pronounced textual changes in the 19th century, although gentlemanly norms were still in evidence; and (c) by the late 20th century, expanded theoretical descriptions/discussions appear to have replaced experiments and methods as the rhetorical centerpiece of the research article. (Discourse analysis, rhetorical analysis, register, social studies of science, scientific writing, corpus linguistics)*





Random notes...


The "Philosophical Transactions of the Royal Society of London," 1675-1975: A SociohistoricalDiscourse AnalysisAuthor(s): Dwight AtkinsonSource: Language in Society, Vol. 25, No. 3 (Sep., 1996), pp. 333-371Published by: Cambridge University PressStable URL:






place of author: central (frequent use of 1st person pronouns and active verbs)→ distanced (agentless passive). Rare occasions 1st person used to indicate uncertainty. Third person references to "the authors"


Discourse Community: 17th/18th C friendly "Dialogic"→ 1725 less friendly dialogic/more oppositional (outlier) → 1775 cooperative tone returns → 1825 new dynamic (stressing "importance" or "interest" →1925 on greater concern for theoretical rather than empirical.