The liquid water requirement tells us a life-bearing planet can't be too hot, but finding life in the deepest parts of the ocean suggests that the water may not need to be on the surface. The nutrition raining down from the surface ultimately links back to photosynthesis, though, so it's not entirely convincing evidence.

The "habitable zone" concept is still useful when it comes to actually searching for signs of life. I'm guessing a planet with organic life on the surface would have stronger signatures than one with a thriving ocean under 500 km of solid rock and ice.

> The nutrition raining down from the surface ultimately links back to photosynthesis

This is a key point. The trench lives on the spillover from the surface. If there was no habitable surface full of life, there would be no life in the trench.

Have a read of http://en.wikipedia.org/wiki/Whale_fall - the deep ocean floor may host ecosystems, but this does not mean that planets where the best conditions are like the deep ocean floor are "habitable".

Similarly, the presence of mosses and lichens in Antarctic dry valleys is interesting, but does not prove beyond doubt that such organisms could evolve if the best conditions were like that.

> So far as has been observed, life does require liquid water.

That's a bit like a newborn baby saying "so far as has been observed, there are only a couple people in existence".

That strikes me as a reasonable assumption to hold, pending evidence to the contrary.

A baby doesn't have evidence that there's stuff outside his/her perceptions. We do - knowing that we keep finding extremophiles in hot springs and underground rocks and water in unexpected places like Mars and Europa, we should be a lot more careful than a baby when we say "this is how it is".

But then again, this is an universal counterargument and can be used to dismiss any kind of selection as hubris. We need to act to the best the information available allows us and learn from mistakes.

Thus is the nature of knowledge and of discovery? This is why the Scientific Method was invented, correct?

Part of the scientific method is knowing when to say "we don't know yet". We don't know how to define "habitable" yet, so we should stop saying it.

You quoted "So far as has been observed". How much more hedging do you need? Basically you're arguing for something useless; that no definition be made until there is 100% certainty, regardless of caveats. Such a high bar gives us no utility.

But liquid water can be found in as far away places as Saturn's moon Enceladus and Jupiter's moon Europa.