Hacker News Reader

Comment by cwmma | Hacker News Reader

This seems to say that forests were gone, so all tree dwelling birds died out and thus all modern birds are descended from ground dwelling birds. Looking at the paper [1] it seems to only say 'reduced flight capacity' which makes more sense, considering we only know of 4 times flight has only evolved so this would be a big deal if birds had evolved it twice.

1. https://www.cell.com/current-biology/fulltext/S0960-9822(18)...

View on HN

Basically, you can't take any science reporting major media at face value. The headlines are written as click-bait, and typically in a non-scientific fashion. Some scientists are also prone to releasing catchy PR that misinterprets their results.

Most of us natively write anti-clickbait titles.

Though it is rarely cited, my favorite paper that I've co-authored is:

Indirect Evidence for Lévy Walks in Squeeze Film Damping: https://arxiv.org/abs/1005.4926

It could have alternatively been titled: "Holy crap! Did you know that high-vacuum gas particles trapped between two walls follow fractal paths?!"

"You won't believe where these high vacuum particles go when they move, or how they get there!"

“The walls hate these particles, they can move in fractal ways, you can learn the secret, too”

Bait me with how this affects my car insurance and you've got a click.

These four tech giants are upending this centuries old industry... with quantum entanglement.

A light wave got sucked into a black hole, you won't believe what happened next! I predict a new meme theme of taboola/outbrain style headlines that are actually semi-accurate sciencey clickbait titles. Genius!

http://oneweirdkerneltrick.com/

I wonder if this in any way related to (causes, or shares a cause with) Mandelbrot's discovery that electrical noise on a signal line follows a Cantor set pattern? Or perhaps the motion of the particle is due to the Heisenberg uncertainty principle? (If the exact position and velocity of a particle is not knowable in detail, then its path would have to have a fractal dimension overall.)

There's unlikely to be any direct link. The Levy-walk behavior is a simple consequence of geometry interacting with the distribution of particle directions that a particle might select each time it comes off the wall.

The result turns out to be scale-invariant, and was a surprise for us, as we had expected a nice integer power-law in the scaling of the gas dynamics, and it turned out to be fractional. The upshot is that when computing gas damping in high-precision force/displacement sensors, one must resort to numerical simulation.

One consequence of this work, in concert with the groups at Trento and MIT is that the LIGO gravitational-wave detectors opened up their gaps between the detector masses and nearby "reaction masses". The easy gaps to open happened before Advanced LIGO turned on, and I believe that the more difficult/expensive gaps to open are being upgraded roughly now.

It's fun how little table-top measurements on the minutiae of force-sensing can have impacts on whether or not we observe cataclysms billions of light-years away....

REVEALED: How a high-vacuum gas particle learned to follow fractal paths using this one weird trick!

"Indirect Evidence for Lévy Walks in Squeeze Film Damping" is one of the most revered Vogon poems[0].

[0] https://en.wikipedia.org/wiki/Vogon#Poetry

To be fair to scientists, it's usually the university public relations offices that write the press releases.

they are written in collaboration with the scientists and the scientists typically have input and limited editorial powers. When I was a PI I would never let the university PR publish anything that I thought was crap.

> Basically, you can't take any science reporting major media at face value.

Or much of anything that comes out of The Guardian in general. Mainstream journalism is racing toward tabloidism, but The Guardian seems almost proud to be leading the pack.

I'm not sure what you are comparing it to. I think there are a lot of worse sources.

I think you're taking an unnecessarily hard line. Flightless birds would, indeed, have to re-learn flight. It's not an exaggeration to say that. Sure, the basic hardware is still there, since they are, after all, birds. But they're not flying anymore.

Of course, they would not be starting from scratch.

Here's the press release:

https://www.eurekalert.org/pub_releases/2018-05/fm-wtd051718...

It's a reasonable PR. Note that it focuses on an entirely different issue than "relearning flight".

But this time they wouldn't have to evolve it from scratch. It's not hard to imagine a bird such as an ostrich evolving into a flight capable bird, especially if there were no existing flighted birds to compete with. They have elaborate feathered wings, so the biggest change needed is probably just to be smaller. Probably 99.9% of the DNA necessary for flight is already there, it just needs to be regulated a bit differently. This would even be true in birds like kiwis or emus, whose wings are much smaller and vestigial.

Yes.

I know nothing about the evolutionary origin/path of the chicken except that it's a "flightless dinosaur". It certainly has "reduced flight capacity", but given evolutionary time and pressure, it certainly doesn't seem like a controversial idea that chickens -- given the right selective pressure -- could evolve in a direction towards being fully flighted dinos again...?

Chickens can perform short flights, but from my limited observations it seems like the stamina is the main limiting factor for them. They've however clearly evolved from fully flight-capable birds, but the now-extinct Moa however might have been a different story: https://en.wikipedia.org/wiki/Moa

Oh, yes, there are evolutionary dead ends, but e.g. the Moa (or even the Dodo[1]) basically evolved for really long without any evolutionary pressure towards avoiding "enemies" and thus was stuck in a local optimum with no way out.

Still ~25 seconds of sustained flight, just about the record for a chicken, I think seems like not-evolutionary-dead-end-enough to my untrained eye. :). YMMV.

[1] AFAIUI, it was humans that ultimately drove it to extiction, but it could have been any invasive species that would kill bird-like thigs. Maybe I'm wrong on that.

To be fair, the headline is not that they reevolved the ability, but that they had to relearn flight. I was expecting it to say more that many of the strategies that were dominate for flight died off. I was guessing dominate air currents higher up, possibly something crazy like a change in dominate gasses.

Point being, same basic birds. Just new advantages steering dominate flight patterns. Not a new mechanism of flight.

(Which, yeah, I guess that is all part and parcel of evolution...)

I'm not a native English speaker, but AFAIUI "learning" is typically used for individuals and not a species? I think I've seen it used before about species, but it seems to imply some sort of teleology which is decidedly not what happens in evolution. (Though it sometimes might seem like it does!)

This is more of a semantic discussion: after a while, there will be individuals with better strategies to survive and reproduce (this is the implicit teleology of evolution: living things live on, dead things go extinct). You don't even need someone giving this orders because we define living things as those who live (research communities can't even agree on the definition of life - see viruses). Our words imply specific meaning.

I guess you mean that there is another meaning in the word "learning" in that context: That the species follows a specific purpose. But what is the species? The aggregation of all individuals of some kind (the research community can't even agree on a definition for "species").

Some individuals will do stuff or become something (due to genetic modifications) that gives them more resources or better reproduction rates - sometimes individuals will learn and teach it to their offspring. If it's a good strategy, the species will contain a lot of individuals who acquired this strategy.

So I guess "learning" is accurate enough for common usage.

Yeah, I guess I was just a bit surprised by "learning" because it does sort of imply "learning by experience" which isn't really a thing when it comes to evolution.

I do agree, it's probably just a bit "too semantics", though I do feel it pretty important in science reporting to be as accurate as possible and to try to avoid fostering misconceptions as far as possible. (Let's face it, popular science reporting doesn't have a great record on this. Hell, even university PR departments[1] are culpable.)

[1] It's amazing that there even is such a thing, but I guess we just have to deal.

I’m not sure I really understand the difference between learning, adaptation, intelligence, and evolution.

I feel like people choose which one to use based on which timescales and action substrates they want to privilege, but I don’t know... is there a technical distinction?

Maybe this is more than semantic discrepancies here. There is evidence for "learning by experience" that can be inherited. I suggest reading about epigenetics and how the experiences of your ancestors might be encoded using switches on your DNA (which turn areas on or off and gives the organism a way to pass on knowledge - hence "learning by experience" supported by epigenetics).

See e.g. [1] or [2] - even simpler organisms like worms have this ability, it is very likely that we also have those capabilities (because we are descended from simpler organisms).

I hope that you find something interesting in the current research of biology and psychology, some things are not like what we used to think.

[1] http://discovermagazine.com/2013/may/13-grandmas-experiences... [2] https://futurism.com/memories-can-inherited-scientists-may-j...

I think this is right. I just don't encounter that much speaking of species, if that makes sense. Most of the speech about species is often anecdotal, at best.

I think you both squared it correctly calling it a semantic debate. I wasn't trying to call you out as wrong. My caveat at the end was supposed to be how I realized this is shifted when talking about evolution. Rather, I was attempting the principal of charity.

I do fully agree that scientific writing should be more precise. However, I also think scientific readers should accept that there is a lot of context that goes into making some terms specific. To the point that it is best to have a readership that is attempting to reach a common understanding, not one that is expecting to be given vanilla facts.

Learning is used in more general ways. Buildings learn, according to Stewart Brand. Evolution learns. Populations learn. Police forces learn techniques for dealing with new types of crime. Criminal gangs learn new ways to evade the police. Physics has learned from chemistry, and vice versa. etc., etc.

It would certainly be a fantastic discovery, but would it be that surprising? The amount of knowledge we can infer from the fossil record is infinitesimal compared to all that actually happened; and while going from 0 instances of an event to 1 is notable (it tells us that “it can happen”), so is 1 to 2 (it tells us that “it can happen more than once”), but is 4 to 5 that crazy?

Under the mostly accepted principle of parsimony, yes.

Flight has been evolved independently 4 times, gliding much more - how does that square with the principle of parsimony? (ie, if we hadn’t discovered pterosaurs yet, we could make the argument that flight evolved 3 times and probably not more due to the principle of parsimony, but we would be wrong)

the principle of parsimony is just that, given equal evidence one way or the other, the simpler explanation (flight only evolved 1 time in birds) is more likely then the more complex one (birds gained flight, lost it, then gained it again).

The fact that there are only 4 times it's happened that we know about is just evidence that it's a fairly rare thing to evolve.

But given that it evolved once in birds, after an event that wiped out all of the flying animals, wouldn't the principle of parsimony imply that birds have the lowest barrier to gaining it again to take advantage of that now empty niche?

that depends on how you think loss of function and gain of function works at the phenotype level. For example, I choose to interpret these sorts of situations in the way you do: birds that lost flight still had 99% of the flight capability, being maintained under functional selection for other uses, and so only a small amount of change would be required to regain the ability.

What are the 4 times that flight has evolved?

Insects, pterosaurs, birds, and bats. https://en.wikipedia.org/wiki/Flying_and_gliding_animals

Here's a good picture that highlights the differences: https://en.wikipedia.org/wiki/File:Homology.jpg

That's powered flight, btw. Gliding evolves much more easily.

I liked the article's statement that "there were no more perches." Is that to say, then, that even birds that "perch" refused to do so on cliffs? Certainly, not all of the trees perished as well.

Reply on news.ycombinator.com