Dialogues with Darwin: An Exhibition of Historical Documents and Contemporary Art

Would Darwin Blog?

April 17, 2009

Darwin depended on the British Imperial postal system to communicate with colleagues around the world. This written dialogue was essential to the development of his theory of natural selection. It is estimated that Darwin wrote over 1,500 letters a year at his peak, and in 1877, he spent the equivalent of his butler’s salary on stationery and postage alone. Today we have the internet, not to mention Twitter and Facebook. How are new communication technologies—and the attendant social behaviors that arise from them—transforming the nature of scientific work and communication?

Diablog Guest Participants
Every new Diablog begins with comments by invited experts on the topic. “Would Darwin Blog” guest participants are the following.

Cecelia Brown is a professor at the School of Library and Information Studies at the University of Oklahoma, where she works on the information-seeking behavior of scientists in the electronic environment as well as the information literacy of science students.

Sinéad Collins is a junior research fellow at the Institute of Evolutionary Biology at the University of Edinburgh, which Darwin attended, and does a regular blog podcast on current issues in science called “Ask Science Dude.”

James Valentine is an evolutionary paleobiologist and Emeritus Professor in the Department of Integrative Biology at the University of California Berkeley. He recently donated his 4500-volume collection of Darwin books to the APS, a number of which are featured in the Dialogues with Darwin exhibition.

---

James Valentine

Chris Anderson, editor-in-chief of Wired, recently wrote a piece entitled “The end of theory”, in which he argued that the ongoing accumulation of truly large amounts of data, having just risen to the level of petabytes, would eventually (reasonably soon?) render theory obsolete.  Underpinned by massive datasets and evaluated by statistical tools, Anderson speculated that correlations would eventually reveal the story about relationships without recourse to theories of causation, in biological as well as physical sciences.  Experimental confirmations of hypotheses would no longer be necessary.  Well, good luck with all that.  This argument brings to mind another notion, attribution unknown to me, to the effect that if we knew all of the laws of the universe and its initial conditions, we should be able to calculate everything that has happened and will happen; it is data that would no longer be necessary, as theory would be sufficient.  Good luck with that too. 

It is easy to see that data accumulation alone might suffice for, say, marketing.  If the correlations were strong and indicated a successful strategy, no one would care about psychological or sociological explanations for why sales climbed, so long as they did.  But scientists are faced with the task – -indeed it is their duty—of explaining why their data say what they say.  For evolutionists the problem of amassing data is particularly vexing.  They are faced with the task of explaining the biosphere, which, since evolution is a contingent process, involves understanding billions of years of history both of life and of the physical environment.  The only direct evidence of ancient organisms comes from the fossil record, which is quite incomplete and biased, as Darwin emphasized in the Origin of Species :  a lot more of the record is missing than is present.  Today, indirect but very useful evidence about extinct lineages also comes from genetics, especially from molecular data on the evolution of development (evo-devo), which permits us for the first time to reconstruct, not only the morphology of extinct forms from fossils, but also something of their genomes, based on the genomes of their living descendants.  Unfortunately there is sometimes not a one-to-one correspondence between the regulatory genes involved in development and the sorts of morphologies they function to produce, but a relationship certainly does exist, and evo-devo is a most welcome and powerful tool with which to probe important aspects of evolutionary history.  Although our fossils and molecules have told us a great deal about evolution, the fossil record will always be at least as incomplete as it is now, and the gaps will remain no matter how detailed our data become for the points between them.

Darwin was faced with even greater gaps in knowledge than we are, both of the fossil record and of living organisms.  He famously made many observations himself, but as is often pointed out, he also solicited information, massive amounts for his day and age, from a vast array of knowledgeable correspondents, known and unknown to him personally –- information on points germane to his hypotheses, or at least to his interests.  If Darwin were alive he might not be a blogger, but he would most certainly be a Googler; essentially he Googled the naturalists and the communities of cultivators and breeders of his time.  But no amount of data will do away with gaps in the record; to achieve an explanation Darwin needed to construct a theory.  In order to explain the data, gathered from such heterogeneous sources, Darwin formulated many hypotheses, weaving them together into that formidable body of evolutionary and ecological theory, based on natural selection, which so beautifully informs our basic understanding of the biosphere

Darwin was a singular author, in many respects to be sure, but here I mean that he had no co-authors at all (the Darwin-Wallace papers were not actually co-authored of course, and even when his work was assisted by his son Francis, Charles retained authorship of the results; he did put his name to some reports and recommendations with other signatories, but it’s cheating to count them).  His fellow naturalists were much the same; a few systematists wrote joint papers, but even those were relatively rare.  I’m just finishing a short manuscript (with a single co-author) in which the cited papers include 2 with over 30, 6 with over ten, and 15 with over 5 authors—nothing unusual in biology today, though in my particular field (paleobiology) more than 5 is still fairly unusual.  The papers with lots of authors are the work of teams, commonly scattered across a number of labs, which produce and evaluate large databases, such as genomes, that truly require a range of expertise and support beyond the reach of even several individual workers.  Of course, the teams work on the organisms that hypotheses indicate are the most likely to produce important results, perhaps that can be used for understanding development for medical purposes, or for understanding evolution, or ecology, or whatever.  Such teams and their labs are expensive, and are supported by large grants.  The data that they produce are indeed important, but in general they do not themselves constitute theory, although the data are certainly grist for theorizing. 

I’m sure you can see where I’m going with this.  Everyone is hungry for data, including theorists, and acquiring extraordinary information, such as a whole genome sequence, is worth a great deal.  BUT.  Let’s not neglect the single worker or very small group, who can concentrate on explaining some aspect of how the world (or universe) works, as a contribution to the theory thereof.  I theorize that most theories are hatched by individuals rather than groups, though I have not accumulated any data.  If resources for collecting data are taken from support for individual investigators and given to teams, we will end up with a lot of information, but will it be at the expense of theoretical knowledge?  Although few can succeed on the scale that Darwin did, he stands as a spectacular illustration of the potential of the individual contributor to produce good science, in which theory and data are mutually interdependent.  And okay, perhaps the most famous single contribution to biology since Darwin’s time, from Watson and Crick, had two authors, but that’s close enough to one.

---

Sinéad Collins

Who cares? But if I have to take a guess, I’d have to go with probably not. Darwin would have emailed, Skyped, used Amazon.com until Emma confiscated the credit cards and changed the password on the account, and probably illegally downloaded documentaries using shareware. He would undoubtedly have made use of virtual museum collections.  But he wouldn’t have blogged, or had a podcast, and his homepage would have been disastrously out of date until some poor graduate student was assigned the task of periodically updating it. I say this because despite the undeniable place of his work in heated public discussions over the past hundred and fifty years, Darwin himself engaged very little with the public, leaving even the initial explanations and defense of his work to others (most notably T.H. Huxley, a.k.a. Darwin’s bulldog). Furthermore, Darwin was cautious. He refined his ideas about natural selection for years before publishing them. And it appears that he would have continued working on them even longer, except for the fact that he was about to get scooped. Darwin did, however, write piles of letters and relied heavily on the post to keep him in constant contact with his large web of collaborators. I have to conclude that a more likely scenario would have been that Darwin would have had a busy LiveJournal page with high security settings and an overflowing email inbox.

My guess on whether or not Darwin would have blogged is based entirely on his reclusiveness, and is nothing but a wild guess. Who knows how Darwin would have behaved if he had access to my laptop and wireless Internet? But let’s step back here and take a look at a more general question: how does the kind of communication available affect how research is done, and who becomes a scientist? How does this same technology affect how we tell the rest of the world about it? 

Victorians were pop-science enthusiasts. Nineteenth-century London was a hotbed of public lectures and debates, and scientific breakthroughs were routinely published in books that were available to, and indeed read voraciously by, the public. Public science lectures included some rather explosive demonstrations of new discoveries in chemistry and physics, and the scientists who presented them gained a degree of notoriety. In the mid 19th century, Faraday’s Friday-night lectures, for example, frequently sold out. Can you imagine the equivalent modern scenario… “Dude! I just managed to get tickets to a sold-out lecture on environmental genomics! Cool! We’re going to have the best Friday night!” Ahem. Not likely. Unlike now, where an avid interest in pop science is rarely considered to be the height of cool, Victorians used to go out for an evening on the town centred around science lectures or demonstrations. My point here is not that I think we should explode more glass beakers around young couples on dates (though really, what could be more romantic and exciting than exploding beakers?), but that pop science hadn’t yet been long separated from art, including performance, and was part of Victorian culture, much as blogging and Youtube is part of ours. Pop science was truly popular science, both in the senses that it was a fashionable pastime, and that it was something that ordinary people were interested and involved in. Now, science seems to be more isolated. Even though there are a few efforts to reunite science and art in specific projects and exhibits, they’re hardly commonplace, and certainly are not the main way that new ideas in science are presented to the public.

Nowadays scientists are more likely to send email than letters, and though we see our colleagues from the four corners of the planet at conferences (thanks to the wonders of jet travel), collaborations can be carried out with very little face to face contact. I, at least, have a few collaborators that I’ve never met, and don’t really expect to. This being said, one generally knows one’s closest collaborators personally, and it is the rule rather than the exception to like your collaborators as much for their company as for their insights on the finer points of genetics. Science is still done between friends, and still done by people who, by and large, consider themselves scientists at some fundamental level of their being.  Research is rarely a 9-5 job that one just happens to do, and developing and testing new ideas is an intimate activity to share.  Many strong, long-term collaborations happen between close friends, or lead to close friendships. New ideas and seemingly hair-brained schemes that eventually yield interesting results may come out of contrived partnerships through funding body schemes, but they also come out of endless casual discussions at the pub or over coffee, on long walks through bogs, or at birthday parties. Frequently, they come out of one too many drinks and a conversation that went on a bit longer than was usual (or wise). 

Phones, the Internet and jet travel have all made communications more numerous and faster. But I’d argue that the basic way that science is done hasn’t actually changed all that much between Darwin’s time and mine. There are some glaring differences that I’d like to write about in the future (for example, I’m a queer woman and the inclusion of women, visible minorities, and openly gay people in science is relatively new, though academia has historically had some tolerance for misfits of a certain stripe). But for now, let’s just think about communication. Despite the fact that breakthroughs are attached to individual names, science is a very social activity, involving mentoring and nepotism early on, and constant discussion with peers over the whole of one’s career. High-tech communication has changed the speed and volume of communications, but ultimately scientific collaboration and success is still very much dependent on personal connections forged the old-fashioned way: by senior scientists introducing junior scientists to each other and into the larger scientific community. Still, the best indicator of future success early on in one’s career is the number of collaborators, one’s connections to the heavy hitters in the field, one’s gender, all of which affect one’s actual productivity (and the perception thereof) in terms of the number and quality of papers published.

Advances in technology have made collaborations easy to maintain once you have them. However, personal introductions still serve the same function that they always have: they vouch that you’re good, that you’re worth the time it takes to answer emails, schedule phone conversations, arrange flights, accommodations and visiting schedules, that you’re a good investment of time and money. Letters of introduction may now take the form of informal emails from a current mentor or collaborator to a future one, or a quick introduction at a conference, rather than actual physical letters carried on your person. But you still need them if you want to do research.

One major change in scientific communications that I think has fundamentally changed the way science is done is the rise of peer-reviewed scientific journals as the gold standard for communication of a completed idea (insofar as ideas are ever complete) or experiment. Rather than duking it out in the public sphere, most basic research is communicated mainly to other specialists through articles that have been reviewed by a small, anonymous jury of one’s peers. The reviewers know who the author of the work is, but the author doesn’t usually know who the reviewers are. This has two notable effects. The first is that it makes it difficult to publish complete quackery in a reputable journal, and to some degree ensures that a minimum level of scientific (rather than, say, religious) reasoning accompanies new ideas or arguments about older ones. This prevents pseudo-science from being presented as legitimate science, at least within the scientific community, and provides some sort of indicator for lay people and journalists about the credibility of a study. It also often improves the quality of papers when reviewers point out aspects of the study that are poorly explained, or catch errors or oversights.  The second and less positive aspect of peer review is that it is a very effective way to suppress minority opinions, even if they follow the rules of good practice, especially if they directly challenge the work of an established authority. Since no study, no matter how careful and inspired, will ever be perfect, reviewers can always find grounds to reject a paper they simply don’t like or which calls their own work into question. And since peer review is anonymous, reviewers don’t face any real consequences if they let personal interests colour (or even determine) their evaluation of a submitted manuscript. This is beginning to be mitigated by journals such as PloS One, where reviewers only evaluate methods and analysis. This maintains the quality check of peer review while making it more difficult to suppress work that may be unpopular simply because it challenges established work and egos.

Peer-reviewed journals are trade journals, aimed mainly if not exclusively at other scientists. Because of this, the public must rely heavily on science journalists to get an idea of what cutting edge research is up to. This distances the public much more from science (though I for one would be much more interested in giving public lectures if I were allowed to explode things and/or give the lecture with a martini glass in hand). However, there are a fair number of scientists who keep science blogs or do podcasts, which allow us to interact directly with non-scientists. While this lacks the glamour of an evening out at a science lecture, followed by drinks and discussion, it does allow those that are interested to stay abreast of cutting edge research, even without any professional connection to the field. It’s also a great way for scientists to keep up with what’s going on outside our specialty.

So, near the beginning of my career, am I trying to integrate into a completely different world than Darwin was? At a very basic level, I’d have to say no. Science is still work done between friends. We can now communicate faster and more widely, but most people will admit that they’re more likely to work with someone they know and like than with a random stranger who has similar qualifications. We have more universities, more scientists, more research, but in any specialized field, a few labs tend to dominate, and there is usually a lot of collaboration between them, with graduate students and junior scientists being sent back and forth so that they are introduced “into society.” Even though I have access to email, blogs, and more conferences than I would ever want to go to, most of my collaborations are the result of having worked in the same department, or of being introduced to someone through a colleague that we had in common, and not through random meetings at conferences or “cold call” emails. Finally, and most importantly, I don’t come up with ideas in isolation. Like Darwin (and pretty much every other scientist ever), I spend a lot of time discussing ideas with friends and colleagues, which is how this whole mess of stuff jumbling around in our heads gets turned into something beautiful. And frankly, I think this is the most important ingredient of doing good science: we get by with a little help from our friends.

---

Cecelia Brown

Would Darwin Blog?  Of course he would! Not only would Darwin blog but he would post his findings freely online! Darwin was a highly curious individual who shared his work all around the world. The ability to use technology to communicate with other scientists beyond time and space boundaries would have been the ideal way for him to share his work quickly and easily. Imagine the possibilities of being able to disseminate Darwin’s theories so readily! Would his theories be advanced further than they are today? Unfortunately we cannot know the answer to this question but to give us some clues we can examine the ways scientists traditionally and currently communicate. At this point in the digital revolution it appears that not everyone in the scientific world has been as quick as I imagine Darwin would have been to embrace electronic modes of scholarly communication.

Science is a social enterprise that depends upon scientists communicating their research findings to one another. The traditional, primary vehicle for scholarly scientific communication is the peer-reviewed journal article.  In 1972 William D. Garvey and Belver C. Griffith published their model of the journal based mode of scientific communication in their classic work recounting the communication patterns of psychologists. In 1978 F.W. Lancaster incorporated electronic information dissemination into the model while Julie Hurd modernized it even further in 1996 and 2000. Hurd projected that advances in digital information sharing abilities will fundamentally transform the way scientists communicate by the year 2020, however she recognized that several behavioral, economic, and social pressures could prevent scientific communication from becoming fully electronic. Hurd’s concern is well founded in history as such pressures were at work over 60 years ago when President Roosevelt charged Vannevar Bush, the United States Director of the Office of Scientific Research and Development, to make recommendations for continuing the high level of scientific research and development stimulated by World War II. To do so, Bush advocated the development of a technology, called the “Memex,” to mechanize the storage of books, reports, and other records of scientific communication. Despite the Memex’s potential to facilitate and enhance scientific communication, scientists in the mid 1940s did not adopt Bush’s electronic system for their communication and information seeking activities.

Even without a concomitant enhancement in communication practices, scientific productivity continued to flourish over the next two decades leading Derek de Solla Price in 1963 to coin the term “Big Science” to describe the shift from the solo scientist working on “Little Science” to large, expensive, and equipment intensive collaborative experiments among several researchers dispersed around the globe. Big Science resulted in an increase in the number of multi-authored papers plus an overall increase in the number of papers published creating “the needle in a haystack” dilemma for scientists in their quest for information they want and need to participate in the scientific enterprise. Although Bush did not live to see it, his early vision began coming to fruition at the end of the 20th century with the creation of online repositories holding millions of peer-reviewed papers plus terabytes of data generated from research in genomics, proteomics, particle physics, astronomy, and meteorology trumpeting the emergence of what is frequently termed “e-Science” or “telescience.”

With the advent of e-Science removing physical, temporal, and geographical boundaries from scientific communication, all scientists potentially have the opportunity to participate in the worldwide scientific conversation beyond the usual journal article publication. Such conversations among scientists were originally confined to letters and perhaps face-to-face conferences and subsequently expanded to telephone conversations. This in effect restricted the number of individuals capable of participating within a particular network. Dubbed the “Invisible Colleges” by De Solla Prince in 1966, these select scientists formed groups that essentially charted the course of research by establishing a united front in decision and policy making which extended to the distribution of research funding and even the establishment of new avenues of research. Consequently each scientific area’s Invisible College has been responsible for the majority of papers published and the subsequent flow of information within the scientific community. Ethnographic studies of scientific knowledge production conducted in the late 1970s and early 1980s by Bruno Latour and Steve Woolgar and by Karin Knorr-Centina further illustrate the human element of scientific communication. These researchers observed first hand, in laboratories, scientists communicating not only through written texts, but also through the scientific instruments they use, the data they gather, as well as via the people who are directly and indirectly involved in the research. It seems likely, as suggested by Leah Anne Lievrouw and Kathleen Carley in the early 1990s, that the digital communication could precipitate an increase in interpersonal interactions and hence an expansion of communication beyond the elite groups forming each field’s Invisible College. Real examples of this are web based collaborations, or collaboratories, which serve to digitally connect scientists to each other, to instruments, and to data, beyond a scientist’s home laboratory. However, as illustrated by Finholt’s 2002 review of 15 collaboratories in the Annual Review of Information Science and Technology, plus additional studies of online collaborations in the first decade of the 21st century by several researchers including Diane Sonnenwald, Sanda Erdelez, Tawnya Means, Catherine Blake, Wanda Pratt, and Carole Palmer, digital communication technologies are making it easier for scientists to collaborate but they are not fundamentally changing the way scientists “do” science.

The scientific community’s struggle to store, access, use, and understand large volumes of data and information lead to the convening of a Blue Ribbon Panel in 2003 by the Assistant Director for Computer and Information Science and Engineering of the National Science Foundation. Not unlike Roosevelt’s 1945 request of Bush, the Panel, headed by Daniel Atkins, was asked to explore trends in digital data collection and warehousing and to make strategic recommendations for the creation, dissemination, and preservation of scientific knowledge.  In what is called the “Atkin’s Report” (http://www.nsf.gov/od/oci/reports/toc.jsp) the Panel recommended the creation of a coordinated cyberinfrastructure for the support and sharing of digital collections of literature, data, instruments, and software to foster and facilitate collaborations among scientists around the world. Even though Clifford Lynch recently pointed out at the 2008 the Annual Meeting of the American Society of Information Science and Technology in Columbus, Ohio, such a cyberinfrastructure is yet to be fully realized, the currently available digital technological capabilities have greatly advanced, since both the design of Garvey and Griffith’s model of scientific communication and Hurd’s proposal of a totally electronic model of scientific communication. Combining these capabilities with the United States’ National Institute of Health’s Public Access Policy (http://publicaccess.nih.gov/policy.htm), signed by George W. Bush in late 2007, requiring all NIH funded researchers to submit an electronic version of their final, peer-reviewed manuscripts to the National Library of Medicine’s openly accessible repository, PubMed Central, there remains untapped potential for significant shifting in the ways scientists communicate their research results.

Today, in 2009, scientists are extending their reach far beyond that imaginable only a decade ago when Hurd used the term “Bigger Science” to describe the influence of technology on scientific output with the use of interactive Web 2.0 technologies which provide a global forum for the exchange of scientific information 24/7. Although the ease of access to scientific information afforded by the Internet is a boon to communication of scientists, the continued growth of the information haystack coupled with virtually unlimited availability heralds the era of what could now be termed “Mega Science.” However, studies of scientists’ communication behaviors including their citation practices, reading habits, downloading activities, dissemination of online information, and collaborative activities indicate that the digital infrastructure facilitates, but has not fundamentally altered scientific communication which remains rooted in intellectual and social ties rather than the ability to communicate electronically. Scientists still rely on the time-honored peer review journal system to disseminate and validate their work since it the way it has always been done. Not only is the way the academy works, but journal publishers handle all the time consuming aspects of publication and dissemination from peer-review to galley proofs to print pages to pdfs letting scientists do what they do best—science. So perhaps my original, confident “Of course Darwin would blog!” at the beginning of this post should be more of a hedge:
“Maybe Darwin would blog if it counts for tenure and promotion!”???

Footnote:
In preparation for this blog post, one of my former students forwarded an abstract to me from James Currie’s dissertation entitled “Greedy for Facts”: Charles Darwin’s Information Needs and Behaviors which readers of this blog may find interesting. The 2007 dissertation is the result of Currie’s work at the University of Pittsburgh and the full text is available via UMI. The work explores the question: “What were Darwin’s information needs and how did Charles Darwin seek, organize and use information?” using a unique hybrid of historical research and case study methodologies. What emerged is a five-sphere model depicting Darwin’s broad information behaviors as being comprised of 52 descriptive information behaviors where 13% of the occurrences are classified as information communicating. This includes claim staking, corresponding, dispensing, face-to-face networking, presenting, publishing, questioning, quoting, and transmitting.  The remaining 87% of Darwin’s information behaviors were consumed with managing (26%), organizing (25%), seeking (21%), and using (15%).  Again, imagine the utility of an online system to Darwin!

User Comments

Post a new comment

To comment, please login or register.

• Facebook
• Twitter
• Flickr