It is worth noting that the structure of scientific papers has developed towards the structure that Medawar outlines. In this respect, the original works of Galileo or Newton would look very different from a contemporary journal article. If you are interested in considering these issues further, the work of Montgomery (1999) should be useful.
For a more detailed discussion of the processes and motives of communicating science within the scientific community, see Rowland (1999b, 1999c).
The important thing to remember here is that these processes of representation are often hidden, although not deliberately, from the receivers. This is less of a problem for scientists, because they become aware of these processes as they become scientists. However, this does have important implications for how science is communicated in wider society, because much of the scientific information consumed by individuals once they have left school comes from news media. Research has shown that science journals are a key source of information for news media outlets (Nelkin, 1995). In part, this is because peer reviewed journals, particularly high-profile publications such as Nature, Science, The Lancet and the British Medical Journal, which have built up a reputation for publishing important scientific reports, are seen as trustworthy sources by media professionals (Wilkie, 1996). As such, these publications are likely to have influence both within the scientific community and, less directly, in wider society.
If we take the production of a newspaper article as an example, a journalist is likely to regularly access information from high profile sources that s/he trusts. Peer-reviewed academic journals, such as those listed above, are such sources. The information a journalist receives from these sources is likely to come in the form of a press release, delivered electronically, listing the articles to be published in the latest edition with further relevant details. This information will have been selected and then constructed into a press release by the journal's staff, in discussion with the author(s) of the paper. Taking Medawar's argument forwards, the information that the journalist receives is, therefore, already a partial representation (the press release) of a partial representation (the paper) of actual scientific practices. Furthermore, the journalist will take the information they receive and then select from this, constructing a further mediated account according to the norms and conventions of the newsroom. As a result, the reports of science that are widely available to the public not only reflect a very small selection of science, but they also appear to the public following a ‘filtering’ process involving a range of actors, including the scientist who conducted the research, the journal's editor, the peer reviewers, the journalist who wrote the story, the news editor who agreed to publish the article and the sub-editor who checked the article and produced the headline. This translates actual scientific practices into a range of partial accounts for particular audiences, whether within or outside the scientific community.
If you now think back to the initial activity in this section, where you considered several high-profile examples of science communication, you can see that the accounts you have consumed will have been highly mediated, by a wide range of actors selecting and constructing information for a variety of reasons and taking account of certain social norms and conventions. In each case a level of shared understanding was crucial for the communications to be successful, involving science communicators of reputation and influence, and who were seen as trustworthy. This does not mean, however, that the producers can assume how their audience made sense of these communications.
Of course, this assumes that you are working in an area that is likely to generate media reporting. Many scientists go through their entire careers experiencing few, if any, interactions with media professionals. What then of science communication that is unlikely to generate mass communication? What opportunities are there for scientists to communicate more directly with nonexpert audiences? To consider these issues you should read the following section.