First, I would like to apologize for the long time between posts, and thank everyone for the fun comments on my introductory post. My excuse is: between 4th of July holidays and trying to set up a new electrophysiology rig in my new lab (soldering wires at 2 AM is not fun-anyone notice how it takes twice as long as you think it will to set up one of these rigs…even with all the experience?), and moving, I have been really busy.
I want to bring up a discussion about what I perceive is a dangerous trend in neuroscience (this may be applicable to other areas of science as well), and that is what I will term “scientific collectivism.” I am going to split this into two separate posts because it is so long. This first post is the weaker arguments, and what I see are the less interesting aspects of scientific collectivism-however, they deserve a discussion.
You may be already saying, “Whoa Mike isn’t that the nature of modern science, peer review, isn’t that inherently collectivist?” Yes, and I will attempt an admittedly brief discussion starter about why I believe the very nature of the modern convergence of scientific collectivism and the internet may be leading us into a disaster (a bit of hyperbole here). I will say at the outset that I do not think that collectivism is always bad, so please read the whole thought before you make up your mind about what I am going to say in this post.
1. Casual Scientific Collectivism
I began thinking about this idea about 2 years ago at the annual Society for Neuroscience meeting. I noticed two major things at that meeting: 1) there was an ever increasing emphasis on global science cooperativity; 2) the students in many of the seminars with laptops were looking stuff up on Wikipedia when there was a concept they did not understand or wanted to read quickly about. This is not a debate about the merits of Wikipedia (that can be found elsewhere) and I do not think Wikipedia is a bad thing. However, this year, when I noticed two high-school interns looking up the action potential on Wiki, I began to really think hard about this. This convergence of scientific minds at SFN and the convergence of scientific minds on Wiki are “casual” forms of scientific collectivism. The meeting is a classic form, the Wiki is a modern.
If we have an ever expanding emphasis on global scientific cooperativity (again, that may be up for debate in later posts), including more and more of the countries that are beginning to become scientific/economic powerhouses and we combine that with the increasing costs of fuel and food, then more and more of our casual scientific collectivism is going to have to be on the Wiki (or webs like it), as opposed to in meetings at San Diego, or Washington, DC.
The idea of the Wiki is simply that multiple minds are better than one, and the collective efforts of mostly anonymous people will create content that is better than an individual source. This is epitomized in the 2004 James Surowiecki book: The Wisdom of Crowds: Why the Many Are Smarter Than the Few and How Collective Wisdom Shapes Business, Economies, Societies and Nations.
But for Wiki-science, this is not true and does not work. Why? Because Wikis take time, and often a long time to correct mistruths. Often scientific topics are controversial (see: pre-versus post-synaptic locus of LTP) and the Wiki may only reflect the prevailing view. Other dissenting views (more on this in a later post) may be stamped out in the discussion of the science-Wiki.
With no clear editor, anyone can change information about scientific topics and affect the thought-process of millions of people, if only for a few hours. The scientific and policy consequences of the ever shifting sands of Wiki-science can have unforeseen consequences.
For example, there is a Wiki about me. Someone took a quote from a newspaper interview I did about music completely out of context. A casual scientific collectivist who has bought into the Surowiecki dogma may hear me give a seminar and do a quick check of my Wiki to learn that I said something that is not at all what I believe. But at that moment, and until fixed otherwise, those who view that Wiki have implanted in their minds this misinterpretation. Who cares about someone like me? No one really (well, maybe my dog…)
What is really important is something like the function of a potassium channel (I am about to use an extreme and overly simplistic example to illustrate a point-not a straw man, just an extreme example).
If there is a mistake in the Wiki, and there is no editor-and the mistake is continually missed, or, more likely, an informational cascade has begun, millions of people (especially non-biologists) may be walking around thinking that the potassium channel passes uranium ions. “Who cares about that? They can be re-educated by experts”, you say.
What if that misunderstanding of channel function by millions of casual observers is not corrected by millions more casual scientific collectivists who have bought into the Surowiecki model, because they believe, with the noblest intentions that we all should be more connected scientifically-or because they are not sure that the Wiki is completely wrong (or both)? What if the casual scientific collectivists think that potassium channels somehow involve potassium ions-but because of the discussion section of the Wiki where many people are yelling “uranium!”, and stamping out small discussions about the possibility that it is potassium, they doubt themselves enough to correct the Wiki? What if the corrections by a small minority of casual scientific collectivists are “reverted” and they lose the will to keep fixing the page? OK, you get the point.
What if the “experts” and the “editors” begin to spread the word that the Wiki is wrong, that its potassium and not uranium, and the Wiki is corrected. “Ah”, you say, “the model worked, the truth did rise out of the collective”. The direct students of the experts and editor may get the information quickly and realize the mistake, but the “myth of uranium” may prevail long enough in the non-scientific ethos to affect science and science policy. For example, how much money is given to researchers who study potassium may be affected. It may take the academic community years to right the problem and re-educate people about potassium channels.
When is casual scientific collectivism a problem? When it involves a convergence of mass distributed networks of people, being channeled and bottlenecked into faceless, anonymous, and editor-free depositories of information. This leads to massive scientific groupthink, homogeneity, imitation, centralization and a herd mentality. Don’t believe it is a problem? Go to Google (do you have many other search engine choices?) and search for something-it probably has a Wiki that you instinctively go to. Don’t think scientists are susceptible to groupthink? Ask graduate students what the important problems in neuroscience are and you will see a statistical convergence on some major categories. Who cares what grad students think? Look at what programs the funding goes to and ask yourself, why we have “trends” in scientific thought…are these trends right? Should there even be trends?
These and more serious scientific collectivism will be discussed in my next post. Thanks for reading!
Excellent post, Michael, and an interesting argument. I guess errors can become established ‘fact’ regardless of medium. Errors creep into traditional, edited textbooks too and thence can become ‘truth’ to the casual reader. It all boils down to how much trust you can put in the source. I think people (scientists especially) are pretty good at weighing up the relative merits of information presented on many-editor (wikis) and few-editor (books/journals) sources. But further help might be needed. I sat in on a talk from a guy from CrossRef a couple of days back. They’re working on digital ‘kitemarks’ that would give the reader some assurance that the article in question had been reviewed/edited in some way. I think the intention is to make this work for peer-reviewed articles, but a similar concept could be adopted for wikis. The kitemark, when clicked, could give a rating on how trustworthy the article is deemed to be (although I’m not sure what metrics you’d use). Sorry if I’ve gone off topic a bit.
Interesting post, and if that is the boring bit, I am very eager to read the next one!
The July editorial in Nature Physics is about trustworthiness of online encyclopaedias, I wrote a Peer to Peer post about it.
I thought the Scholarpedia initiative discussed in that article seemes quite interesting. (Links to these online resources are in the Peer to Peer post.)
As a student in the pre-internet days, I can certainly relate to what Matt writes about out-of date, incorrect textbooks. I think that librarians have a role in training students (of any age) about how to use and trust resources. I have read many blogs and articles along these lines by librarians, and have been impressed with what a lot of them are doing in this regard.
Collectivism meets granularity, quite a problem. An interesting one.
The scholarpedia idea is an interesting one, and even the idea of “kitemarks.” My main point is that something has to be done now, to prevent a sticky situation for science in the future.
Maybe Nature could devote a set of full-time editors to an online science encyclopedia like Wiki that was a bit more science/citation-intensive than scholarpedia and FREE. I would be willing to bet that that would solve the problem!
NPG did have a reference division but alas it is no more. One of the publications it launched was the Encyclopaedia of Life Sciences, which it sold (to Wiley, I believe), but it also published other online dictionaries and resources.