The Internet as a Medium for Science Communication

Henry S. Rzepa

Department of Chemistry, Imperial College, London, SW7 2AY.


In modern times, the first individual to foresee how technology could help people communicate on a global scale was Samuel Butler. A contemporary of Darwin, he wrote in 1863;[1]
"I venture to suggest that ... the general development of the human race to be well and effectually completed when all men, in all places, without any loss of time, at a low rate of charge, are cognizant through their senses, of all that they desire to be cognizant of in all other places. ... This is the grand annihilation of time and place which we are all striving for, and which in one small part we have been permitted to see actually realised"

He was actually thinking of how the electronic telegraph might be used, then a new invention, but arguably in talking about the senses, he was also the first define a need for what is now called "multi-media" comunication. Furthermore, his phrase "desire to be cognizant of" expresses succinctly the need to be able to index and search for information, and to avoid being overloaded with useless information! Butler was followed by other prophetic scientists, such as H. G. Wells in 1937 imagining a world brain based on a permanent world encyclopaedia, or Vannevar Bush in 1945 proposing a technological solution to the problem in the growth of knowledge in the form of MEMEX, a multimedia personal computer.

Electronic Mail as a Science Communication Tool

So, are we seeing in the late 1990s a realisation of these scientific aspirations? Two fundamental breakthroughs are bringing us close to these ideals. The first can be traced back to the very origins of what is now called the Internet. In 1969,[2] three university sites in California and one in Utah managed to connect their computers together as part of a military-funded project in computer science. As part of the development process, they needed to exchange program code and numerical information in a reliably error-free manner. They soon realised that the very network they had built was perfectly suited for this purpose, and the first "electronic mail" programs to implement this were written. Thirty years later, the system has evolved to the point that a very high proportion of all the computers on the planet (estimated at around 200 million in 1997) are capable of being inter-connected, and of being used to send such e-mail. It is also reasonably certain that most contemporary scientists now use this tool entirely routinely, although few use it as productively as they might might for scientific communcation. Why is this so? I suggest that it suffers from two problems, one more to do with human nature, and the other a limitation of the technology, for which recent solutions have become available.

The first problem relates to the fact that e-mail is predominantly a "push" mechanism, in that only the sender has control of any message and its content, and the recipient plays an essentially passive role in the process unless they choose to actively respond and become a sender themselves. Typically, an scientist might expect to receive anywhere between 5-50 e-mail messages a day, and in reality will probably choose to carefully read only perhaps 10% of these, and to respond to even fewer. Much of the remainder is often discarded unread, particularly the "spam" or junk e-mails that have became a serious problem in the late 1990s. After a few months, in and out (received and sent) mailboxes more often resemble "information graveyards", and few people keep old messages more than a year or two. In reality, e-mail has (with one important exception) changed very little since its invention in the 1960s, and has become perceived as a very noisy communication medium, more akin to the informal telephone call, than a scholarly and permanent communication method.

The second problem is that most e-mail has traditionally been based on free-form text, which has little pre-defined or standard internal structure and hence where the semantic content of the message cannot be easily indexed and hence searched for. To paraphrase Samuel Butler, much of it is what we do NOT desire to be cognisant of. Perhaps a specific example can serve to illustrate how difficult a communication medium it can be. If two chemists wish to communicate to each other the detail of a molecular structure such as taxol (Scheme 1), with as little risk of error as possible, they would probably NOT try to achieve this using the telephone. They might succumb to exchanging fax messages, but they would soon realise that to usefully apply the received structures, they would have to redraw the diagrams using a suitable computer program, with all the risk of error that does introduce.


To describe the molecule using only a simple text-based e-mail system is not going to succeed either, unless they are both experts in the use of text-descriptors of molecules known as SMILES strings. Few chemists are! That for taxol for example is (CC(OC1(C(CC2O)OC1)C(C2(C4=O)C)C(C(C3(C)C)(CC(OC(C(O)C(C6=CC=CC=C6)NC(C7=CC=CC=C7)=O)=O)C(C)=C3C4OC(C)=O)O)OC(C5=CC=CC=C5)=O)=O)
which would stretch pretty much most chemists! So, by the late 1990s, scientific e-mail was largely restricted to the communication of information that could be expressed using simple text and ASCII characters, and this excluded much of what a very visually expressed scientific subject such as chemistry would need.

Around 1992, the first enhancement to e-mail which has the potential to address some of these limitations was proposed by Borenstein and Freed [3] in an Internet standard known as MIME (Multipurpose Internet Mail Extensions), which allow a separate datafile to be enclosed (or attached) with an e-mail message. Interestingly, the first "M" of MIME is sometimes mistakenly intepreted as standing for "Multimedia", and here again we can see Butler's vision of the use of all our senses being implemented. This datafile could in turn be processed by an additional program presumed resident on the recipients computer known as a "helper". The first applications of MIME evolved during the late 1990s to exchange generic and standard attachments such as Microsoft Word documents, or individual graphical images such as the one above for taxol. This indeed was the mechanism by which this article was sent to the editor!

A very recent extension of the MIME method to a more focused scientific area such as chemistry3 (the so-called chemical MIME types) has now permitted, at least in priciple, an error-free exchange of a visually comprehensible and instantly usable expression of e.g. the taxol molecule. Because information received in this manner can have a precisely defined semantic content and a standard or documented structure, it can be in principle easily indexed and ultimately searched for by the recipient or sender. Use of the MIME mechanism as as carrier of high value scientific information has hitherto not received the uptake it deserves amongst the community, but as better e-mail software which supports such methods becomes more routinely used, this is expected to become a standard scientific tool of the future.

The World-Wide Web used for Scientific Communcation

The second key breakthrough on a global scale was started around 1989, and was driven by financial reality as much as any utopian vision. Tim Berners-Lee[4] was working at the European laboratory for particle physics in Geneva (CERN), an organisation faced with escalating costs and reducing budgets. Such "big" science often involved large teams of scientists from all over the world, and some way needed to be found of allowing them to exchange their most recent results in an error free and productive manner, other than the expensive option of flying them at regular intervals to meet in Geneva. Berners-Lee harnessed the power of the then relatively new Internet to invent what is now known as the World-Wide Web. This was not the first time that a large scientific organisation needed to achieve effective communicaton, so what elements did Berners-Lee's invention have that allowed it to succeed, apparently so spectacularly? There are three crucial characteristics which underpin the success of the Web.

1. The first was the adoption of a so-called structured document language known as HTML (Hypertext-markup language), and this was possible because a set of guidelines on how to design such as language (known as SGML) were by 1989 reasonably mature and accepted as a de-facto global standard. The use of HTML in turn enabled documents to be easily indexed and retrieved. It is rarely realised that within a year or two of the widespread global deployment of the Web and HTML around 1994, the Internet had been almost comprehensively indexed, and a variety of so-called "search engines" such as Alta Vista were in use. HTML in turn is now evolving into XML (Extensible Markup Language),[5] which will allow scientists to express more richly than they could with HTML alone their own subjects, ultimately allow their subject to be indexed and searched for. Now at last, one can see how Samual Butler's vision of "finding all that we desire to be cognizant of" might be close to realisation for scientists.

2. Berners-Lee realised that if a document could be expressed in structured form, then finding individual components within any individual document was as important as finding the document itself. He went further, and defined the "resource" as the important principle of scientific communication. He therefore used two components of the Internet, the ability to define almost any device connected to it via an address known as the "IP" address, and the MIME extensions referred to earlier to create something called a "URL" (Uniform resource locator). This locator could be inserted into a document, or indeed into programs or into databases in the form of a hyperlink, and this could serve to establish a network of links, or context, between diverse multi-media resources. Created by say a chemist, such a rich construct of links can serve to define knowledge within a subject, rather than just simply a collection of information.

3. The final design feature is a little more arcane, but no less important. The Web was designed as a "stateless" mechanism for "pulling" resources from a server. Up to 1989, use of the Internet was predominantly as a point-to-point communication channel between two devices. The channel whilst it was open reserved vital bandwidth by preserving the state of the link (essentially a history of the entire session) , and it could often remain open for hours. Forseeing that his creation of the hyperlink meant opening one-to-many channels, Berners-Lee stipulated that each URL when invoked should close each channel after data transmission was complete, a design that has indeed allowed the 100+ fold increase in traffic that has resulted.

We have described how both e-mail and then the World-Wide Web have introduced global mechanisms which can act as links between information resources to establish the connections between different scientific disciplines, and which now form the backbone of much modern scientific communication. An overview of this modern infra-structure is shown in the scheme below, which charts the relationship between e-mail and the Web, and how the MIME standards can be used to link the two technologies. The original article describing this should be consulted for further detail.[3]


Figure 1. A scheme showing how Electronic Mail and World Wide Web can be used to exchange documents between two or more users, using MIME Mechanisms.

The Internet Used for Scientific Communication: A Case Study

How have scientists actually used all this technology to improve how they communicate with each other? This is illustrated here with a specific example drawn from the area of chemistry, which shows how a scholarly article can be electronically enhanced with Internet-based technologies.[6] This example is drawn from the "state-of-the-art" in 1997, and given the rapid rate of progress in this area, is expected to be entirely obsoleted by new developments quite rapidly. If you are reading this more than a year or two after 1997, perhaps it already looks quite quaint!

The case study starts at the normal point of an author preparing a conventional manuscript for publication, in this case for the Royal Society of Chemistry journal ChemComm. The article describes how several new molecules have been prepared, how their structures have been determined, and how information obtained from instruments which measure their spectral properties relates to their structures.


Figure 2. Caption on Electronic Version: "Click here for 3D view"
Reproduced from Paul E. Kruger and Vickie McKee, ChemComm, 1997, 1341-1342.

You can see for yourself that expressed on paper as in Figure 1, it can be difficult for the reader to see the three dimensional perspective of the molecule, or to identify individual regions or atoms which might be obscured by other atoms. Using a chemical version of the MIME mechanism,3 the reader of the electronic form of this journal article can produce a rotatable 3D model of the molecule on their computer screen, which can be viewed by them from whatever angle and magnification they wish.


Figure 3. Caption on Electronic Version:

"Click here to acquire the JCAMP Digital Spectrum"
Reproduced from Paul E. Kruger and Vickie McKee, ChemComm, 1997, 1341-1342.

Associated with this molecule is an infrared spectrum (Figure 2), which for the reader of the printed page is "what-they-see-is-all-they-get". Using a slightly different Internet tool known as the Java applet, the e-reader can acquire the original data as recorded from the instrument, and again expand regions of the data, measure distance between peaks, or integrate the area under the peaks. The two sets of data can also be inter-linked. For example, if the reader selects a peak in the infrared spectrum, the atoms associated with the molecular vibration assigned to that peak can be highlighted in the 3D molecule display. Because the data is available in digital form to the user, it can be transferred to other programs and tools. For example,[7] the 3D coordinates of the molecule could be acquired and transferred to a spectral simulation module, where the vibrations can be modelled theoretically. The theoretically derived infrared spectrum could then be overlayed with the experimental version to see how well they compare. In effect, what has happened is that the reader of such an e-journal article can have almost immediate access to the authors' original data, and can apply that data to their own purposes and scientific needs.

Present and Future Implications for Practising Scientists.

The most conspicuous feature of the modern Internet is that can allow many forms of communication between author and reader, and that the invention of the hyperlink allows these various forms to be easily and transparently inter-linked. Thus a scholarly journal article expressed electronically can have links to the author's e-mail address, allowing them to be contacted very rapidly by e-mail. Appropriate materials can be included with the message using the MIME method. The process can in fact start at a much earlier stage than formal publication. Collaborating scientists could exchange materials and results by exchanging e-mail messages and mounting draft documents on their personal Web servers for mutual (but still private) viewing. To gain wider feedback, the results could be re-directed to say an electronic conference in the form of a poster, where the authors could interact with interested readers via an e-mail discussion list. Some ten such chemical scientific conferences have been organised during the period 1994-1997.[8] At one level, such conferences can act as a new form of collaborative peer review mechanism, having the advantage over more traditional peer review methods that experimental details can be easily linked to the article to colleagues to evaluate the results if they wish. Finally, at this stage the original authors may wish to formally submit their article to a reputable journal. An early pioneering article was in fact processed in much this manner,[9] being mounted on-line to allow anonymous referees to review it, before being ultimately printed in a conventionally bound journal.

Even in 1998, the abandoning of much of the electronic enhancements illustrated above (Figures 2, 3) in order to achieve such printability was still the norm. The advent of electronic-only journals such as the Internet Journal of Chemistry[10] suggests that this situation may change, although no-one expects the process to occur rapidly. In part, this is because the problems of long term (ie > 10 years) archival have not really yet been addressed. Some of the electronic conferences noted above have been archived on CD-ROM, preserving all the active chemical functionality, but often being tied to specific features of the Word-Wide Web (such as the browser version) which may make it difficult to avoid longer term obsolescence. Other electronic journals adopt archival on CD-ROM of generic page description formats such as Adobe Acrobat which in effect emulate paper, but which are less amenable to chemical "enhancements". One discrete trend that the Internet and electronic dissemination will undoubtably bring about is the so-called "aggregation" of individual journals into large databases of scholarly and searchable content. Commercial pressures seem more likely in the short term to threaten the existence of individual journals than to result in their scientific enhancement.

The focus of this of necessity short article has been on e-mail and the Web. There is however something of a continuum of collaborative methods available between these two extremes. Mention should perhaps be made of an interesting experiment that started in late 1997, that of the "virtual chemistry lecture", which made use of Web technologies to deliver a keynote lecture, followed by panel discussions and eventually moderated comments from a virtual audience of some 350 registrants to the event.[11] The lecture slides were themselves hyperlinked to other on-line materials, and were chemically enhanced in a manner similar to that shown in Figures 2 and 3. It seems unlikely that such events will in the near future supercede the more conventional delivered talk to a real audience. Indeed, as the author of original lecture, I followed this up by giving the same lecture to a live audience a few weeks later. Whilst the latter allows more ephemeral devices such as humour to be employed, only time will tell which of the delivery methods has the longer term impact on scientific communication.

This very brief overview of Internet-enabled scientific communication merely scratches at the surface of what is possible. Of course, communication is more than just about the technology, it is about finding methods that the scientists find helpful, and where the technology does not distract from the science. The must be the challenge we face over the next decade, in reconciling technical wizardy with genuinely enhanced scientific perceptions. We must not also forget the legacy we must bequeath future generations, for whom the current technology will seem quaint indeed, and who must not not lose access to the knowledge because of this. Other brain teasers include how to handle to issues of copyright resulting from such "active" content. There remains much to be done, but it would be fair to say that the potential for changing the some of the ways that scientists can communicate with each other is substantial.


References and Further Reading

[1]Quoted in George Dyson, "Darwin amongst the Machines, The Evolution of Global Intelligence", Addison-Wesley, N.Y., 1997. ISBN 0-201-400649-7.
[2] For details, see "The Internet: A Guide for Chemists", Ed. S. Bachrach, American Chemical Society, 1996. ISBN 0-8412-3223-7.
[3] For a detailed discussion of this, see H. S. Rzepa, P. Murray-Rust and B. J. Whitaker, J. Chem. Inf. Comp. Sci, 1998, November issue.
[4] For a detailed history of the World-Wide Web, see http://www.w3.org
[5] P. Murray-Rust, "Chemical Markup Language, A simple introduction to structured documents", in "XML, Principles, Tools and Techniques" (Ed. D. Connolly,), O'Reilly, 1997, pp 135-149. For details of all W3C (World-Wide Web Consortium) recommendations, proposed recommendations, working drafts and notes, see http://www.w3.org/TR/
[6] An example of the use of chemical MIME to integrate a variety of chemical data types into the body of an electronic journal is the CLIC Electronic Journal Project; D. James, B. J. Whitaker, C. Hildyard, H. S. Rzepa, O. Casher, J. M. Goodman, D. Riddick and P. Murray-Rust, New. Rev. Information Networking 1996, 61. For the project itself, http://chemcomm.clic.ac.uk/ For details of how a chemically enhanced article was prepared, see O. Casher and H. S. Rzepa, in Proc. E. Conf. Trends in Organomet. Chem.: ECTOC-3 (Eds H. S. Rzepa and C. Leach), Royal Society of Chemistry, 1998. ISBN (CD-ROM) 0-85404-889-8.
[7] N. Stewart, A. Tiller and Henry S. Rzepa, "Integration of Enhanced Electronic Journals and WebLab Instruments, in Proc. E. Conf. Trends in Organomet. Chem.: ECTOC-3 (Eds H. S. Rzepa and C. Leach), Royal Society of Chemistry, 1998. ISBN (CD-ROM) 0-85404-889-8.
[8] For example, the ECTOC conference series, published on CD-ROM as electronic proceedings (Eds H. S. Rzepa and C. Leach), Royal Society of Chemistry, 1996-1998.
[9] H. S. Rzepa, B. J. Whitaker and M. J. Winter, J. Chem. Soc., Chem. Commun., 1994, 1907. This article is still available as http://www.ch.ic.ac.uk/rzepa/RSC/CC/4_02963A.html
[10] S. M. Bachrach (Ed), Internet. J. Chem, 1998; http://www.ijc.com/
[11] This event was sponsored by ChemWeb.com, and involved a lecture delivered in the form of an electronic slide-show stored on a central database server. The event was archived into an electronic library, along with the discussions, which people could add to over a period of several months after the event. For more details, see http://chemweb.vei.co.uk/. For an article describing this event, see W. Warr, J. Chem. Inf. Comp. Sci, 1998, November issue.