site

I'm announcing today the new availability of a bibliography section here on the weblog. At present this database includes more than 11,500 entries. These represent a large fraction of the historical and contemporary literature in human evolution.

The database as it exists today owes to the work of many, many people. Foremost among these is Milford Wolpoff, who compiled and has curated an immense bibliography as a flat text file over many years. It includes entries that have been cited in many papers by Milford, his students, and many coauthors.

Milford says that these folks are too numerous to remember. That has the ring of truth. Just by looking through the entries you get a picture of an active group of people over more than forty years.

To these "legacy" references, I have begun to add fuller bibliographic information during the last few years. This began with Digital Object Identifier (DOI) tags for new, and later for older entries. In the last few months, I have begun to archive further information, including abstract.

I continue to update and backfill references that belong in the database. There are many cited here on the blog that haven't been properly archived, and I will be adding these along with new citations.

What you'll find

The bibliography has a search filter, search terms will match author, keyword, title or abstract (where present). With more than 11,000 entries, you want to be a little selective about how you search. Author names work really well, and yield a list separated by year of publication.

You'll find each reference preceded by a unique citation key in brackets. I did this purely for my own convenience, but for those who may want to download lists of citations, it may also prove useful.

A list of search results can be exported to BibTeX or RTF format for download.

There is also a "filter" tab that allows keyword, author, and year filtering of the list. This is really not very useful; the size of the database makes it much simpler to search than to filter all entries. I have entered keywords for only very recent entries, so keyword filtering is not especially helpful.

Each citation in the reference list comes accompanied by several links. You can click on the title, taking you to a reference page with full bibliographic information. For recent entries, the reference page includes DOI, direct links to journals, and the abstract. For all entries, the links include Google Scholar lookups and citation downloads (in BibTeX or RTF format). Author links give a list of all my database entries from that author. These require exact name matching, so don't rely too much on them -- a search of the database for an author's last name is usually more complete.

My bibliography section has its own feed, which lists new entries in reverse chronological order. I tend to upload new references in batches, so you'll frequently see a dozen or so show up at a time. I've put a short excerpt of the feed in a block on the left side of every page, just above my blogroll.

Snafus

On the topic of author names: I know that many of you will immediately search for your own name. You'll find some of your publications there, but you almost certainly will see that many are missing. A few of you may not find a single paper!

Please don't be offended. Remember that the entire list was typed by somebody. If a paper isn't there, it's because nobody typed it into the list. Really, only a few of mine are in there!

I'm not volunteering to add another 11,000 entries to the list right away. It will take me 5 years or so to manage that, at the current rate. But if you find yourself wanting to pitch in, to help add some of your references to the database, then let me know and I'll work with you.

There are countless typos. I do not intend to fix these, unless I cite the paper moving forward. It would literally take me weeks of work to fix the typos in these entries. Please don't contact me to fix these, because I won't do it!

Some errors in the database have resulted from a script gone wrong. I scripted them all into BibTeX. When I did this, my script langage was Pascal, which was no picnic. That's a lot of pattern matching on citations. To the credit of all the people who entered data, the overwhelming majority of citations fell into a few common patterns. But there were oddballs, which didn't get translated properly. Over the years I've fixed many of these script-induced errors, but not all. Some of these are ugly and weird. I'm sorry!

A few quirks come from the software that presents the database. I haven't begun hacking on this, and I've decided to leave the defaults for awhile to see how well they handle the requirements.

Most critically -- inline links to the references get mixed up when there are more than one reference list on the page. Since this is usually true of my front page, that's a problem! This is a minor annoyance, since the inline links only save a bit of scrolling, but it may not be easy to fix.

Google Scholar lookups work pretty well, but I've run into some issues. For some older entries, the links fail to find papers that Google actually knows about. The problem is that the system passes the full title on to Google as a "quote", requiring an exact match. These often fail. Removing the quotes from these will often recover the citation in Google Scholar.

If you do a search, the biblio page will continue to filter on your search phrase until you reset it. There's always a "reset search" link right up at the top of the page.

Keyword searching works quite well with recent entries to the database, but not with the legacy references. If you are working on a research paper and want to do a lot of keyword searches, you'll be much better off starting with Google Scholar or Web of Science.

The inner workings

I got to talking with some researchers from Microsoft a few months ago, who were interested in the ways that I curate information. During our conversation, I came to realize that maintaining bibliographic data is the worst bottleneck in my work.

I use BibTeX in my research work, but I've never had a good workflow that would encompass both research articles and blog posts under the same bibliography system. Often, I'd blog about a paper but fail to get it into my research database. If I wanted to blog about a historical paper, I'd have to go looking for it in my bibliography.

Once I talked this through, I saw what I needed to do. I had to find a better way.

Drupal has really significantly added to its bibliographic capabilities in the last couple of years. I found that the Biblio module now has almost everything I need to import and share references. With this module, Drupal can provide inline citations for blog posts and automatically compile formatted bibliographies. It also provides all those cool links to Google Scholar and other online sources.

One more sed script to replace some macros, and I had all my BibTeX database ready to upload.

But that didn't solve my problems with data input. I found that CiteULike could take on this task pretty well. The site automatically scrapes bibliography information from the websites of journals. Increasingly, everything with a Digital Object Identifier link has some way for CiteULike to get its information --- even many edited volumes.

I've tried CiteULike and many other bibliographic tools in the past. I know that other folks like Zotero, or Papers, or Mendeley, or RefWorks, or CiteSeerX. I've tried them all -- every few months I've given them another look to see if they've reached that critical point where they're useful. I never had much luck. There was always some snag keeping me from getting back value for the effort I put in.

I can't say whether CiteULike is the best of them today, but it did what I needed done -- crucially, accurate BibTeX import of my whole reference file, including citation keys, easy setting citation keys for new references, and bulk export so I can update both my local file and Drupal without extra typing. Plus, the recommendation list is really useful.

Final thoughts

Most of the database is also available under my CiteULike profile, so if you're a user of that service, you may find that useful. It's more searchable in some respects, less so in others. CiteULike also offers a feed of my new entries, which run with abstracts. It's like a whole abstracts blog!

Meanwhile, I'll be filling in references very quickly for the next few months. It remains a work in progress, especially as I continue to merge blog citations with the database. If you find it useful, please drop me a note so I'll know how you're using it!

I've had a tradition throughout the blog of making annual predictions at the beginning of each year. In 2006 and 2007, I did pretty well: upward of 50 percent accurate each time.

But last year, I got busy during the winter break and didn't manage to make new predictions for 2009.

It didn't help that my 2008 percentage looked pretty miserable. In view of my past success, I wrote that I was "going to have to pick stranger predictions". I was no slouch on that score -- I figure only two or three had come true by the end of 2008. Who would want to return at the end of the year and review that disgraceful performance? Not me!

But then a magical thing happened: In 2009 a whole bunch of them came true! Including the far-out ones. So now I can hold my head up high. Just look at the 2008-2009 review:

• 10. A dramatic development in the Sahelanthropus story. Well, 2009 was the year the rumors came out to play. We got the whole soap opera of the missing femur, not to mention the published suggestion that Toumaï's bones had been found and reburied.
• 9. Both major-party candidates for the 2008 U.S. Presidential election will accept evolution. True, and true. This was one of the rare ones I got right by the end of 2008.
• 8. This year's featured piece of anatomy: the femur. The featured piece of anatomy for the past two years has clearly been the pelvis. Oh, there've been a couple of important femora, but nothing like the significance of Ardi's pelvis combined with the pelvis from Busidiya, and to these I think it's fair to add the CT reconstruction of Tabun 1.
• 7. No new hobbits, at least, not from Flores. That was an easy one -- and notice that in this case my two-year span has handicapped me -- another year might have refuted the "it won't happen" predictions. We've had an abundance of description, but no
• 6. An incisive example of introgression in East Asia. Nothing there, at least not yet.
• 5. A viral insertion in the human genome will tell us about a disease of the australopithecines. I'm still hoping this one will come true, but nothing yet.
• 4. Another language gene joins FoxP2. No word on whether Neandertals have the human version. We got one this fall, although we can't yet read all the details. No word on whether Neandertals have the human version.
• 3. Homo habilis: an endangered species? Far from sunk, Homo habilis may be headed for a golden age. Or not.
• 2. This year, something new from three A's: A. afarensis. A. africanus. Atapuerca. There was no significant A. africanus announcement in the last two years, although who knows in the next year what might surface? From A. afarensis, we had Woranso-Mille this fall. The Atapuerca craniosynostosis case was newsworthy enough to make the year's biggest science stories list. The really new Atapuerca news, of course, was the 2008 publication of the Sima del Elefante remains, at 1.2 million years old.
• 1. Oh, and one more A. Ardipithecus. Now, the beauty of a two-year span is that the most unlikely-sounding thing might just happen. And boy, did it ever.
• BONUS: A big, big year for Neandertals. I mean, besides the election. I'll claim this one, for both 2008 and 2009. 2010 will be another. Neandertals are chic.

I'm claiming 7 out of 10 -- remember, according to my strange way of scoring, the BONUS prediction counts as extra credit. Now, I had two years on this set, and some of them were pretty easy, so I make no claims. If I'd predicted a Neandertal dredged out of the North Sea, well that would have been something. Still, you don't see anybody else out there calling down lightning from the sky, so take it or leave it.

Now, what about 2010? One of the really appealing aspects of making predictions is that I can write about the most outlandish things happening, like significant discoveries of the earliest hominins. So when people actually do discover early hominins, it makes it a lot harder to think of too-good-to-be-believed ideas for predictions.

But I'm not easily stopped by a failure of imagination. So here is the list for 2010. As usual, I've reserved the top five for the unlikeliest. That's not to say you should put any money on the bottom five. But they're, shall we say, more anthropological id than superego.

I think I'll go with cryptic sentences -- it's more crystal ball than Oracle of Delphi that way.

10. Get ready for the Ardipithecus backlash.

9. Make way for expensive tissue.

8. 1000 Genomes won't find much of the "missing heritability."

7. A half million years ago, in East Africa....

6. Neandertals get even more colorful. And no, the current pigment story doesn't count.

5. The headline is "most complete".

4. Two words that you don't see together much: Pleistocene pharmacopeia.

3. Two point eight is the date.

2. A. anamensis gets a lot more interesting.

1. And you thought Toba was cataclysmic.

BONUS: Did A. afarensis really make those?

Just in case you're wondering about the crypticity -- these are separate predictions.

Maybe next year I'll get started sooner and manage Nostradamus quatrains. Although, I don't want to get involved with debates over nonsense words. Ooh, ooh! I could always claim that the nonsense word that just happens to resemble a new site name was a real prediction. You know, like "Hister" and Hitler.

Oh, I'll probably be too embarrassed to review these at the end of the year anyway...

I've just returned from a week in Leiden, the old university city of the Netherlands. I was a guest of the archaeology faculty, in particular Wil Roebroeks and his stable of students and postdocs, and they were fantastic hosts. I can't say enough about the new friends I have in Leiden.

Except maybe that they set an awfully high bar for the next place I get to visit!

There was excitement in the whole country as the Naturalis museum opened the first exhibition outside Georgia of the D2700 skull from Dmanisi. The TV news covered David Lordkipanidze arriving with the skull, and followed his entourage from the airport. The daily newspapers carried huge broadsheet stories about the fossils and the exhibition. It was pretty cool.

I only wish Lucy had gotten anything like that kind of reception in the States.

I played a small part opening the exhibit by participating in the public lectures at Naturalis on Saturday. There was a very energetic crowd of ticketholders, eager to hear about the science of early humans and to attend the exhibit.

The skull and its mandible D3735 are displayed in the "Treasure Room" of the museum:

The museum houses the original Dubois fossil collections from Trinil, Java, including the Pithecanthropus skull and femur. If you visit, you can see the originals on display:

I sat down alone with them for a while during the gala reception and did what comes naturally:

Unfortunately, spending a week in the Netherlands meant that I had to miss our Thanksgiving at home. Gretchen thinks we should have turkey in the next week or two to make up for it, and I'm not complaining. On the date, however, I got a real authentic Pilgrim experience, as I stayed just above the American Pilgrim Museum in Leiden:

Such a unique place, with incredibly nice proprietors!

So, blogging has been slow as I was soaking in the surroundings, and giving my hosts a preview of some of the research that will be coming out in the next year or two. They've told me that they'll feel paid amply if I keep doing what I do here. So let's get back to it!