Human behavior; no more models please!!! : Gene Expression

Search this blog

Q & A

Books

Human behavior; no more models please!!! permlink

Find more posts in: Brain & Behavior

Share this: Facebook Twitter Stumbleupon Reddit Email + More

Posted on: October 9, 2008 4:07 PM, by Razib Khan

There's a new paper out which models human behavioral ecology, Dynamics of Alliance Formation and the Egalitarian Revolution. Anthropology.net has a good review, so I'll just point you there. I was going to read this paper, and a few others on models of human group dynamics...but lately, I've been wondering, aren't there enough models now??? It seems like there is a large sample space of models which can generate a set of testable predictions. Perhaps it is time to catch up on data and experiment and hold off on more model generation? I'll probably keep reading these papers...but of late I've started to get the feeling that without more data the task of understanding human behavioral history is Sisyphean.

Find more posts in: Brain & Behavior

Share this: Facebook Twitter Stumbleupon Reddit Email + More

Comments

Is it time for a meta-model, a model-of-models?

Posted by: Coturnix | October 9, 2008 5:07 PM

My feeling is the opposite -- there's too much data and no modelers trying to see what it says. That's not incompatible with what you're saying -- basically, the two types work independently of each other.

Posted by: agnostic | October 9, 2008 8:24 PM

yeah, you both are on to something. mostly i'm just tired of a priori models.

Posted by: razib | October 9, 2008 8:38 PM

I guess "models" is just another Thing That Scientists Like... ;)

Posted by: toto | October 10, 2008 11:57 AM

I guess "models" is just another Thing That Scientists Like

Some of us... and some hate them. Just collect a bunch of data and publish it -- no hypotheses tested, no models to organize the mess.

Posted by: agnostic | October 10, 2008 3:28 PM

Some like to test narrow hypotheses and focus more on experimental models, others are more interested in directing their investigations (and experiments) into how the topic fits together. That interest can change depending on the state of the field. Many social science topics aren't at a point where the overarching models come into contact with basic experimental findings. I think it's for good reasons among the bad.

Posted by: quidnunc | October 10, 2008 5:00 PM