The basic reality is, (Wikipedia) is no more or less error-prone than other sources of information,” Jay Walsh, former director of communications for the Wikimedia Foundation, says.
Really, Mr. Walsh? While there are some excellent science articles on en.Wikipedia created by editors who understand how to write accurate, well-sourced, up-to-date articles with readable prose, this is not the case for a lot of bad science which appears on en.Wikipedia’s main page.
The basic reality is that the crowd-sourcing format coupled with anti-elitist hostility favors a main page that highlights inaccurate science articles written and promoted by mostly male, Western, technology-savvy editors who do little but edit en.Wikipedia. These editors use jargon without understanding, jumble up word order to create the opposite meaning, simply guess at the science they don’t know, and have no ability to write prose in a logical fashion.
Recently, in celebration of Halloween, and in the race to the WikiCup, an editor expanded an article about a bat species, Natterer’s bat (Myotis nattereri) to get it on the main page on October 30, 2013. In order to show my Halloween spirit, I blogged about the article on the 31st. Then I started clicking through the internal links (wikilinks) in the article and seeing how well-supported the information was by other en.Wikipedia articles.
The article is about a microbat, a member of the clade (or grade, possibly) of bats that are known for using echolocation to catch prey. The article generalizes that all bats use echolocation to catch prey. (“Bats emit sounds at too high a frequency for most humans to detect and they interpret the echoes created in order to build a ‘sound picture’ of their surroundings.”) This is not true; only some bats emit these high-frequency sounds to create sound pictures. Even worse, this generalization directly contradicts one of the great mysteries of evolutionary biology, the evolution of mammals that fly and those that use echolocation.
The article is standard for en.Wikipedia bad science articles on the main page: a crazy quilt, a collage, a vomitous association of buzzwords plagiarized without understanding from a random assortment of sources, some appropriate, but also including pre-plate tectonics books for geology articles, pre-PCR biology (even for taxonomies), blogs, unpublished class assignments (see below), and even an unfinished PowerPoint from a junior high school website.
But let’s look at just one article. Natterer’s bat is a microbat and this bat uses echolocation. But, what the Natterer’s bat article says is that bats use echolocation. I’m confused. Can we use en.Wikipedia articles to find the truth of the matter, get us all on the same bat-channel?
In the bat article, I learn:
Bat (order): “Microbats use echolocation; with the exception of Rousettus and its relatives, megabats do not.” (unsourced)
Okay, Rousettus and its relatives are the only megabats that use echolocation. The relatives of a genus are other genera in the same family (Pteropodidae), or possibly subfamily. It’s unsourced, so I can’t verify the accuracy of the information.
Okay, let’s go to the megabat article:
Megabat (suborder): “In contrast to the microbats, the fruit bats do not use echolocation (with one exception, the Egyptian fruit bat Rousettus egyptiacus, which uses high-pitched clicks to navigate in caves).”
Finally, something that is sourced! And the source is …
“Matti Airas.” Echolocation in bats”. HUT, Laboratory of Acoustics and Audio Signal Processing. p. 4. Retrieved July 19, 2013.”
… a paper written ten years ago for a class, a postdoc seminar, but, a class.
“Airas, M., “Echolocation in bats”, in Proceedings of Spatial sound perception and reproduction (the postgrad seminar course of HUT Acoustics Laboratory), April 2003.”
The author sources his paper; en.Wikipedia editors could have used his sources.
Let’s get help from the microbat article:
Microbat (suborder): “Microbats use echolocation, whereas megabats do not typically (The Egyptian fruit bat Rousettus egyptiacus is an exception).” (unsourced; punctuation in original)
The bat article implies a family or subfamily of megabats use echolocation; but the suborder articles suggest it is a single species.
Okay, what does the family article say?
Pteropodidae (family): no article
I’m moving on to the genus. Does the Rousettus genus article talk about echolocation? No, of course not.
Rousettus (genus): nothing about echolocation
On to the species article. For a moment, I pause, as I learn that the wings of Egyptian fruit bat “feel like pantyhose.” Although unsourced (of course), it is properly wikilinked to “pantyhose.” I don’t have anything more to say about that.
Egyptian fruit bat (species): “Egyptian fruit bats, along with other species in the genus Rousettus, are the only megachiropterid bats to use echolocation, which they accomplish by emitting a series of sharp clicks with their tongues.” (unsourced)
So it’s a family of megabats, or it’s a single species, or else it’s the genus that is the sole exception to the non-echolocating megabats.
I’m now going to add the Animal echolocation article, bat section.
Animal echolocation: “one genus of megachiropteran bats (Rousettus)” uses echolocation
Let’s run through this in order.
According to which en.Wikipedia article, —– which megabats echolocate?
- Natterer’s bat —– not megabats, but all bats (unsourced)
- Bats —– the family, genus Rousettus and relatives (unsourced)
- Pteropodidae —– no article
- Megabats —– the species Rousettus egyptiacus (unreliably sourced)
- Microbats —– the species Rousettus egyptiacus (unsourced)
- Rousettus —– no information
- Rousettus aegyptiacus —– the genus Rousettus (unsourced)
- Animal echolocation —– the genus Rousettus (unsourced)
Voting on the information runs thus:
- all bats —– 1 vote
- family Pteropodidae (or subfamily) —– 1 vote
- genus (but info not from the genus article) —– 2 votes
- species spelled “R. aegyptiacus” —– 0 votes
- species spelled “R. egyptiacus” —– 2 votes
Any truth out there? This point is important because it goes to the evolution of the bats. And, I can’t even verify that the en.Wikipedia articles are no more or less error-prone than their sources because the en.Wikipedia statements are not sourced.
The basic reality is that when an encyclopedia is getting tens of thousands of readers to an article from a Google search with something as shoddy as this bat article, or the recent article on Sea, it pushes away access to real science written by a writer who has to justify which sources her or she uses, has to use sources of the correct age (post-plate tectonics geology, often), has to read the sources, not just the chosen sentences, then has to coherently bring the information from the sources together, has to outline the article in some logical fashion, and then, finally, has to write well and cohesively.
Wikipedia editors can’t even see their errors, Mr. Walsh (such as in the vegetation section of the Wildlife of Chad article we are told that “Lake Chad in the southwestern corner of Chad …”, right next to a map that shows Lake Chad not in the southwestern corner of Chad); so, who is evaluating how error-prone Wikipedia is?
A disclaimer: I am not a scientist; I am a student just finishing up my degree. However, I have published in the life and earth sciences, and I work in the sciences. The errors I find are so blatant that any ninth-grade science student should catch them.
Image credit: Flickr/possumgirl2, licensed under Creative Commons Attribution-ShareAlike 2.0 Generic