The settlement of the Pacific is one of the great chapters of human history. This region was settled by the Austronesian people during the last 10,000 years, eventually encompassing the region from Taiwan, to Hawaii, Easter Island (Rapanui), New Zealand, and Madagascar. Along the way, these people carried with them a distinctive “Lapita” culture and one of the largest language families in the world. There are two competing scenarios for this Austronesian expansion: either a rapid tree-like spread from Taiwan beginning around 6000 BP, or an expansion from a deeper Island South-East Asia origin around 17,000 BP. Over the last few years we have built a large comparative database of linguistic information from these languages and have begun using phylogenetic methods to explore Austronesian origins. The results of some phylogenetic analyses on 400 of these languages will be presented, along with what these results tell us about Pacific prehistory.

In our Brevia, we used the example of Webster’s Dictionary–widely regarded as the inaugurating dictionary of American English–to illustrate how the desire for a distinct social identity can motivate language changes, such as spelling. Of course, some changes may have begun much earlier. We are not aware that anyone has measured how rapid or gradual these changes were by using the sorts of quantitative methods we have developed, but it would be informative to do so….

Phylogenetic tree thinking is beginning to revolutionise studies of linguistic and cultural evolution. However, linguistic and cultural traits are easily transmitted horizontally (”borrowed”) between cultures. Indeed, well over 95% of the words in the Oxford English Dictionary aren’t English. A loud and persistent debate has centered around the issue of borrowing and whether it invalidates cultural phylogenies or not. Here, we use a natural model of linguistic evolution to simulate borrowing between languages. The results show that tree topologies constructed with Bayesian phylogenetic methods are relatively robust to the effects of realistic levels of borrowing. Inferences about time depth are slightly less robust.

We review recent work on cultural phylogenetics and outline six fundamental questions that can be answered using the power and precision of quantitative phylogenetic methods. However, cultural evolution, like biological evolution, is often far from tree-like. We discuss the problems reticulate evolution can cause for phylogenetic analyses and suggest ways in which these problems can be overcome.

Our solutions involve a combination of new methods for the study of cultural evolution (network construction, reconciliation analysis, and Bayesian mixture models), and the triangulation of different lines of historical evidence. Throughout we emphasize that most debates about cultural phylogenies can only be settled by empirical research rather than armchair speculation.

Nombre de données linguistiques essentielles recueillies au fil des ans dorment dans des placards et ne sont pas accessibles à la communauté linguistique ou au public intéressé. Nous (Gray & Greenhill) exploitons ces données pour reconstruire des arbres phylogénétiques des langues à l’aide des méthodes informatiques utilisées en biologie de l’évolution et pour vérifier ainsi les hypothèses émises sur le peuplement du Pacifique. Dans cette optique, nous avons informatisé une grande quantité de données lexicales et construit une base de données à grande échelle. Nous avons débuté avec le vocabulaire fourni par les listes de Swadesh rassemblées par Blust ces dernières vingt années ; notre base s’est ensuite enrichie grâce aux données de nombreux linguistes ou de publications. On peut consulter la base de données de vocabulaire austronésien (Austronesian Basic Vocabulary Database) à l’adresse suivante: http://language.psy.auckland.ac.nz. Actuellement, cette base concerne 481 langues, pour un total de plus de 100 000 entrées. Nous présenterons quelques-unes des techniques requises pour construire une telle base et nous évoquerons l’intéret qu’il y a à mettre à disposition sur internet ces données pour des recherches en collaboration. Pour terminer, nous exposerons nos projets d’extension et de consolidation de cette base de données, en invitant les chercheurs à nous fournir de nouvelles données. Au cours de notre communication, nous mentionnerons quelques résultats issus de nos dernières analyses.

..or in English:

Much of the valuable linguistic data that has been collected over the years is languishing in filing cabinets and is not immediately available to linguists and interested members of the public. We (Gray & Greenhill) are using this data to construct phylogenetic trees with computational methods adopted from evolutionary biology to test hypotheses about Pacific settlement. As part of this project we have “computerised” a large amount of lexical data, and constructed a large scale comparative database of this vocabulary. This data began with a collection of Swadesh lists collected by Blust over the last 20 years, and has been supplemented with lists from many other linguists and published resources. This Austronesian Basic Vocabulary Database is available on the internet at http://language.psy.auckland.ac.nz, and currently has word lists from 481 languages, for a total of over 100,000 entries. We shall describe some of the technologies required to build a repository such as this, and talk about the benefits of releasing data onto the internet for collaborative purposes. Finally, we will discuss our plans for expansion and consolidation of this database and make a special plea for more data. A few results from our recent analyses will be presented along the way.

Over on Mezzoblue, Dave Shea found out that his website had been compromised subtly. The attacker had exploited some (as yet unknown) security hole and quietly modified his website to link to the standard spam sites.

Within a few hours there were tens of posts from people who'd checked their websites and found similar modifications that had been sitting their un-noticed, with people pointing the finger either at old Wordpress installations or a guesses that their hosting service had been compromised.

This shows us once again, that any software you run on your website needs to be kept up-to-date immediately, but what shocked me was that so many people out there running websites and are not watching them for file changes. They had no idea that their sites had been hacked until they went and looked for it.

So - in bold: Anyone running a website or webserver of any type needs to watch out for unexpected access and changes.

The easiest way to do this is to use some intrusion detection software (IDS). This sounds complex, but it's actually quite easy to do. All these programs do is to monitor your files and warn you when they change. This would have immediately spotted this type of attack.

Because of this, I've decided to write up an easy guide to simple file alteration monitoring - here it is.

Choose your weapon:

There are plenty of intrusion detection/file modification apps out there - some of the better known ones include AIDE, Samhain and Tripwire. These are all very cool, and highly powerful, but are also quite complex and hard to install, especially on cheap shared hosting.

Therefore, my weapon of choice, for the last few months has been a lightweight and fast application called integrit, so I'm going to tell you how to install it here.

Before we start: Do make sure that you're not compromised right now, there's no point running an IDS if you're already hacked. While you're at it, make sure everything's upgraded too.

Step 1: Make a place to store integrit:

Since you're on shared hosting, you can't install integrit properly into /usr, but you need to to put it somewhere anyway.

I decided to install it to a directory called "integrit" inside my home dir, so:

In the following commands, do remember to replace any mention of ~/integrit with the directory you used.

Step 2: Download and install integrit:

The integrit webpage is at: http://integrit.sourceforge.net/, so go there and get the latest version (currently 4.1), or you could cut'n' paste this:

Once the integrit archive file is in your ~/integrit directory, we need to decompress it and install it:

Next we need to put the integrit binary somewhere where we can get it, here I've just dumped it into the ~/integrit
directory, but you could put it in ~/bin or something nicer if you want:

Step 3: Set up integrit:

Look in the integrit-4.1/examples directory and make a config file from the example.

You need three things at least:

1. The known file database - this is where the integrit database is stored.
2. The current file database - this is where integrit stores the modified info.
3. A root directory to monitor - this is the full path to the directory we want to watch.

We can also tell integrit to ignore directories, by listing with an exclamation mark at the start of the line. We want to ignore the ~/integrit directory, and on dreamhost, we'll need to ignore the webserver log directory (because it changes a lot, and parts of it our user can't access which will cause errors).

All in all, it'll look something like this:

Note: You'll need to change "simon" to your user name, and "integrit" to where you installed integrit to in step 2.

Step 4: Create integrit database

We need to get integrit to store a list of the current files, and their vital statistics, so run this command:

...where "home.conf" is the integrit configuration file that you generated in step 3.

If all goes well, you'll see something like this:

Once that's done, copy the newly created database to the known database location:

Step 5: Test that integrit's working:

So lets make sure that integrit's working properly. To do this, we can add an empty file somewhere and see if integrit spots it.

Now we can run integrit:

If all goes well, you'll see something like this:

Notice how integrit's spotted the foo file that's not in the database? If one of the files has changed, you'll get much the same output, with "changed:" instead of "new:". So - remove the dummy file:

Step 7: Get integrit to run daily:

Now, we want to set up a cron job, so that integrit is run automatically for us. First of all, we should make a quick little shell script to run integrit and email the results to us:

Paste the above into a file called "run_integrit.sh" (a good place to put it would be in your ~/integrit directory), edit the paths to match your setup, and change the email address. Finally, make this file executable:

Now all we have to do is to add that to our crontab.

...and add a line that looks something like this:

This will run integrit at 21.59 every day, if you don't know what that means, then have a google for "crontab tutorial".

Save the file, and you're off.

Updating the database after valid changes:

When you've changed or added a file yourself, then you'll need to update your known database with these changes. To do this, just generate a current database, and copy it over the old one. The script I've got above will automatically generate a current one, so you can just use that version, or repeat Step 4.

Final considerations:

Note: If you can, you should run integrit (that is BOTH the database files and the binary files) off a "safe" partition, that's not writable. Unfortunately, most of us on shared hosting don't have that privilege, so just be aware that if a really clever attacker does get you, then they're likely to disable or modify the IDS if they can.

A good way of dealing with this is to copy your known file database off the webserver and make sure that the one on the server matches this one every so often.

--Simon

Greenhill, S.J. & R.D. Gray

Department of Psychology, University of Auckland, New Zealand;

This is the abstract of a talk I gave at Doom '07 - The Annual New Zealand Phylogenetics Meeting at Whakapapa, New Zealand.

The settlement of the Pacific is one of the greatest population movements in the last 10,000 years, and lead to the settlement of the region bounded by Taiwan, Hawaii, Easter Island (Rapanui), New Zealand, and Madagascar. This Austronesian expansion brought with it (and developed along the way) a distinctive Lapita cultural complex and what has become the largest language family in the world, with over 1,000 languages.

There are a number of scenarios describing this Austronesian expansion as either a rapid tree-like spread from Taiwan beginning around 6000 BP, or expansion from a deeper Island South-East Asia origin around 13,000 BP. Over the last few years we have built a large comparative database of linguistic information from these languages (Austronesian Basic Vocabulary Database) and have begun using phylogenetic methods on it.

We will present results from some large analyses of lexical data from over 300 languages, and demonstrate the power of this data at resolving these questions.

A picture from Google Earth, showing the Austronesian words meaning "hand". Data from the Austronesian Basic Vocabulary Database

Usage is quite simple, find a DOI somewhere, e.g. 10.1038/nature02029 (for this groundbreaking paper), and run this:

... and via the magic of webservices and XML, and with a bit of luck, you'll get something like this back:

The Code:

--Simon