Hacker News Reader

Comment by flareback | Hacker News Reader

This is interesting but I think the word 'communicate' is bit of a stretch. Why do people always try to anthropomorphize things?

View on HN

Physicist use this term, but always with the understanding of "communication" in the sense of Shannon information, i.e., two initially uncorrelated systems can become correlated only if they "talk" to each other. It's never intended to suggest planning/agency, and it's a useful intuition pump, properly used.

What's happening here is that the journalist is keying on that phrase used by physicists and repeating it in a popular setting where the reader doesn't know the more precise meaning. It functions as effective clickbait, but the journalist can maintain plausible deniability by quoting the researcher. (Sometimes, the physicists is guilty of doing this purposefully and the journalist is duped.) This is actually a pretty big source of popular-physics clickbait.

But why not use the word 'interact' instead?

If I were to speculate, and say it's probably because (1) they might not be interacting directly, and saying "(possibly indirectly) interacting" is a lot more syllables than "talking"; and (2) "talk"/"communicate" emphasizes the information-theoretic content, whereas "interact" is somewhat more likely to be interpreted in terms of interaction strength. (If two systems are "barely communicating", they are transferring a small number of bits, whereas if two systems are "barely interacting", the influence in displacement/impulse/temp/whatever is small.)

Shannon's original paper was "A Mathematical Theory of Communication," which actually studied radio communication, it's since been generalized, and there's no reason to rename jargon so outsiders can better understand it.

It was actually "The mathematical Theory of Communication," where communication was studied over any channel not just radio.

The paper(s) revolutionized several fields!

That's because that's about the level of "communication" that most physicists can manage with other humans.

Sorry. It's a lazy dunk. And an unfair one. But when someone tosses it up, I need to slam home the lob.

I suppose for the same reason classical fairy-tales do the same. My best guess would be some desire to relate the external world to internal experiences.

It's also an important didactic method in pedagogy. You might call it a form of "Student-centered learning", where a significant effort is made to meet the student where they are. Their desires and experiences should guide them to understand the external world through themselves.

Djikstra famously hated this tendency.

https://www.cs.utexas.edu/users/EWD/transcriptions/EWD09xx/E...

metaphors considered harmful?

It's exactly the reason people think particles know they're being observed

As another commenter notes, the dunes are behaving as cellular automata, which appears to be an apt description. Can automatons communicate? Our computers sure do... is that also anthropomorphization?

Agree, they only communicate as much as waves communicate to one another. I guess if you consider wave propogation information exchange, sure, but not really communication.

Communicate: 1. Share or exchange information, news or ideas. 2. Impart or pass on (information, news or ideas)

I think it can be argued that to say sand dunes communicate with each other is not anthropomorphization at all. I didn't take it that way and I doubt very many people took it to mean that sand dunes were exhibiting sentience and chit-chatting about the weather.

but the article state "This interaction is controlled by turbulent swirls from the upstream dune". Are the sand dunes themselves willfully creating these swirls? Could they choose not to create the swirls?

Nowhere in the definition is intent specified. This article is hardly the only place the word "communicate" is used to describe the behaviour of inanimate objects. For example, "communicate" is a handy descriptor in gas dynamics to explain the differences between supersonic and subsonic flows. In subsonic flow, the particles can communicate with upstream particles, but in supersonic flows they cannot. This is a useful way to conceptualize the differences between the two flow types, and can help a student gain an intuitive understanding of why a wing flap that works in subsonic flow may not work in a supersonic flow without a radical redesign.

In this situation, the use of the word communicate is accurate to the definition, and also a useful tool for gaining an understanding of what is happening. The word is used because it is useful. It serves a purpose beyond anthropomorphizing for the sake of anthropomorphizing.

I was about to remark on the same thing. These dunes are clearly not ‘communicating’: rather, the physics of dunes has some counterintuitive (but not entirely surprising) non-local effects... which in turn are purely an artefact of our flawed conception of them as being ‘objects’, rather than emergent properties.

My computer communicates with various remote services all day. I guess I should give it a name.

There's a system called "DNS" just for that purpose. But you can't just "give" it a name. You have to rent the name.

If we accept that to use the word "communicate" to describe something is to anthropomorphize it, then I absolutely can just give it a name. I shall name it Clippy. However it does not, then my computer will not receive a name.

But in all seriousness I'm not sure if you're joking or just didn't read the comment I replied to. I don't think there's a time against asking if someone read a comment!

I give all of my variables and functions and files good names. Does that mean I anthropomorphize them? (I do, but that's beside the point!)

Well now that you've got me thinking about, maybe it does. Perhaps I should be more considerate of how I use my variables. They may have feelings, after all.

clippy?

It does make it sound like something first hypothesized at Burning Man. And thanks to the research being described this way, numerous people can misunderstand and go on to tell others about science proving sand dunes can talk to each other.

> Why do people always try to anthropomorphize things?

https://en.wikipedia.org/wiki/Pareidolia

> Why do people always try to anthropomorphize things?

Metaphor is the method by which we have any understanding at all. Anthropomorphism comes naturally and can make complex things easy to understand.

Why do people always try to separate certain qualities or behaviors as being "human only"?

I don’t think anyone’s doing that. No one here would argue, for instance, that animals don’t communicate with each other.

The difference is intent: when we use the term “communicate“, there is an implied intention to convey meaning that simply can’t exist for an inanimate object like a sand dune (as far as we know, anyway).

Reply on news.ycombinator.com