Hacker News Reader

Comment by zopa | Hacker News Reader

I’m not yet persuaded this is relevant, at all. This isn’t a case where hierarchy fails us: we have a perfectly good and natural hierarchical model of speciation based on descent from common ancestors. That model doesn’t give us the common-sense notion of a tree, but that’s fine: so much the worse for common sense.

If instead we used a D&G rhizome that “connects any point to any other point” — we’d be dropping the notion of time (ie, ancestors precede descendants). Which when we’re talking about evolution, seems bad.

If horizontal gene transfer were a big part of the story here, that’d be different. Or if your ecology has time travel, or interstellar exchange ala Butler’s Xenogenesis, then sure, rhizome away.

View on HN

your concept of time is one which puts ancestry front and center as the obvious most important factor, but is in no certain terms the only method of interpretation. there are many other factors, and in fact, that reinterpretation is the point of rhizomatic thinking.

and that's exactly what we see in the article, right? if a species took a path A-B-C-D to exhibit features of D, and another took A-F-G-D and so also exhibits features of D, then isn't there more to the history there than simply genetic change via ancestry over time? there's a reason why convergent evolution is so weird and surprising.

epigenetics, for example, is an interesting manifestation of what sorts of new domains you can explore as you broaden the horizon for asking questions beyond "linear ancestry" or "natural heirarchy" - arborescent vs rhizomatic thinking. blasting out of these constraints is the process of deterritorialization/reterritorialization D&G put forward as necessary and essential.

To put this in a statistical paradigm; time is always an important predictor in your model of organism traits (outcome) but there also a whole lot of other predictors that sometimes, or often, are more important/correlated than time.

> and that's exactly what we see in the article, right?

I think we see the opposite! We start with a naive and non-hierarchical view of relatedness: trees are tall woody things, fish live in the water and have fins. And the piece argues that we should let go of our common-sense categories and use the descent-based ones, like the biologists do.

Anyway --- I didn't say that ancestry is the only factor. I said it was natural, and useful. Epigenetics is fascinating, really! But it's not 'blasting out' of anything, it adds a bit of detail and complexity around the edges. A sense of proportion is important; we shouldn't miss the forest for the...never mind.

> the piece argues that we should let go of our common-sense categories and use the descent-based ones, like the biologists do.

The piece has a list of conclusions at the bottom

> taxonomy is not automatically a great category for regular usage.

> (You can, at least, say that a tree is a strategy. Wood is a strategy. Fruit is a strategy. A fish is also a strategy.)

> insisting that “wood” is not a useful or comprehensible category would be the most fascinatingly obnoxious rhetorical move. Just the pinnacle of choosing the interestingly abstract over the practical whole.

Reply on news.ycombinator.com