Religion's pretty pervasive in humans. And why it's pervasive in humans is debated a lot. There are indications of things that look like religion in other animals, like chimps doing rain dances, and that sort of thing. Actually, I say that, but there's that and not much else.

I don't like to be dismissive of whales speaking possibilities. There is a pretty New Age-y, poorly thought set of ideas about this going around. I also think these animals benefit most from us staying out of their way. I think we learn the most by being as passive as we can, so those are my biases, and I'm sort of against it from that perspective.

When I'm out in the ocean and I watch a seabird, I think, "Oh, I probably I know what that seabird is doing" - and I don't study seabirds. And the same for the turtles and on and on. But with the whales, a lot of the time, I have no idea what they're doing, and they're the ones I study.

A bunch of individuals do the same thing because they learned from watching each other or somebody else. That's what we would call culture.

The actual communicative value of what we say is usually quite small. I've lived for times in small, isolated fishing villages, where everyone knows everyone each other and everyone knows what's going on and everyone's watched the same TV programs and, really, there's not a whole lot of new information to convey. But there's still a lot of talking. What's said doesn't seem to matter; that you say it, and who you say it to, and how you say it is what matters.

What I mean by it, and roughly what most biologists who talk about culture mean by it, is either behavior itself, or information that leads to behavior. Information that is picked up through social learning - so, from being with, watching, being taught by others. It's a way that individuals behave or get information about how they will behave that comes directly from the behavior of others.

Studying the behavior of large whales has been likened to astronomy. The observer glimpses his subjects, often at long range; he cannot do experiments, and he must continually try to infer from data that are usually inadequate.