Hacker News Reader

Comment by raldi | Hacker News Reader

How would you define left and right based on these words?

From http://learnthesewordsfirst.com/Lesson-12F.html#12-21

12-20. right.

[X is on the right side of your body.] = X is on this side of your body: Most people write using the hand they have on this side of their body.

[I use my right hand when I draw pictures.]

12-21. left.

[X is on the left side of your body.] = X is on this side of your body: Most people do not write using the hand they have on this side of their body. They write using their other hand.

[My child held my left hand.]

I would think a true minimum set of bootstrapping words would be enough to teach the language to hypothetical aliens who might not have access to a group of humans to poll about handedness.

To tell an alien about handedness definitively, without an artifact, may require that both parties are aware of the CP-violation present in the weak interaction.

Put another way, how can we know that aliens won't reconstruct an electromagnetic message as a mirror image of what we sent?

Make use of the right-hand rule, which will suffice so long as they're not made of antimatter.

How will they know we're not made of antimatter?

Aren't matter and antimatter a matter of perspective? If so haven't we just reintroduced the left/right problem.

Precisely.

If we're not sure both sides of the conversation are made of the same particles, the situation is hopeless.

As far as astronomers can tell, there is very little antimatter in the observable universe (If there was more, we'd see it anihilating with matter more).

They didn't explode when we shook hands (/tentacles/general-graspy-bits)

GP is right. If aliens don't know about CP violation, there is no way to communicate rotational direction that preserves the absolute direction without some shared reference.

Because CP-violation is actually the only detectable way to discern between right handedness and left handedness. That is, without CPv, we would have no way to know that we aren't in a hypothetical mirror universe with all rotations reversed. Every other physical interaction behaves identically left or right.

So we have a few options to deal with our rotationally-challenged alien friends:

1. Hope they can parse far down enough into the dictionary to understand what CP is and either know it or can test it.

2. Communicate using a shared reference, like pointing out two quasars that rotate relative to each other. (Quasars are pretty good galactic reference points.)

3. As long as it doesn't affect giving directions, just don't worry about it, the physics works out the same. If we ever meet and they try to shake our left hand, well, they get their very own "oops I guess electrons are negative then" situation. "And that, kids, is why you always negate earth-radians before using them in a formula."

I don't follow. If we're communicating with aliens that we somehow know are made of matter, we tell them that right is the direction a horizontal wire carrying positive current away from us gets pushed by a vertical magnetic field.

That just shifts the problem to specifying the direction of the magnetic field, and you can't use the concept of left vs. right in that definition without creating a circular dependency.

Vertically up it's just too a higher local gravitational potential, no? Sure, we flip right-left when we look at things upside down, but that's a nuanced usage.

How do you determine whether your magnetic field is pointing up and not down?

Closer and farther can be used to explain up and down.

How do you use closer and farther to determine whether a magnetic field is pointing up or down? It's not like there are little arrows on magnetic field lines you can just look at.

Feynman suggested shining polarized light through a solution of glucose and water. This will rotate the plane of the polarization clockwise. https://en.wikipedia.org/wiki/Glucose#History

Cool, I didn't know that! The glucose molecule has a handedness though, which means it has a rotationally symmetric twin L-Glucose [1] with the same properties. The relative abundance of glucose over l-glucose is purely a property of which one life on earth happened to choose (they're not biologically compatible). So if you communicated blueprints for life on earth including glucose without an absolute rotation reference, the aliens could reconstruct everything two ways: biology with our glucose, or biology with l-glucose, and there's no way to differentiate between them since they would both work the same.

Veritasium makes a great video about symmetries that's very relevant here [2], and might suggest a slightly more practical way to correctly communicate handedness to aliens. The full CP violation experiment isn't necessary, as long as we assume that the aliens live in a universe made of mostly matter (as opposed to antimatter) like ours, and we can communicate the parity violation experiment (~measuring the preferential atomic decay direction of cobalt atoms near 0K in a magnetic field [3]), they should be able to reconstruct an absolute reference for rotation.

[1]: https://en.wikipedia.org/wiki/L-Glucose

[2]: https://www.youtube.com/watch?v=yArprk0q9eE

[3]: http://www.physics.utah.edu/~belz/phys5110/PhysRev.104.254.p...

The definition proposed doesn't differ strongly from 1913 Webster:

Of or pertaining to that side of the body in man on which the muscular action is usually stronger than on the other side; -- opposed to left when used in reference to a part of the body; as, the right side, hand, arm. Also applied to the corresponding side of the lower animals.

http://www.dict.org/bin/Dict?Form=Dict2&Database=*&Query=rig...

The actual dictionary definition uses the normal position of the heart instead (though it assumes no situs inversus): https://www.ldoceonline.com/dictionary/left

It links to definitions for some sub-concepts (e.g. containing or heart) but oddly enough not others (e.g. body), I'm guessing because body is part of the core 360 words.

I’d actually like to see a minimum set of _concepts_ necessary to communicate among humans then choose the language that best uses them.

Such a list would be great to have for many reasons actually.

For a early super-minimal example, see: https://users.cs.cf.ac.uk/Dave.Marshall/AI2/node69.html

Later efforts (for attributed grammars, semantic networks, etc.) used many more primitives. Semantic primitive / "Interlingua" based formalisms never quite caught on for the most part, however.

the <"X---" == left> and <"---X" == right> example is an example of an analogical representation -- these have sensorimotor/perceptual groundings and mimic how we actually learn some concepts, but although there has been some research in AI in utilizing analogical representations internally, it is not typical.

I once had to describe left and right to my kids as they grew up, and the best [ * ] way to do that in my opinion is with this sentence:

"Left is on this side of the sentence, and at the end of it is right."

[ * - as in: they didn't 'get' it until I formulated it this way]

Can you come with the Arabic translation, please? (-:

اليمين على هذا الجانب من الجملة ، وفي نهايته هو اليسار.

The dictionary starts with those 60 words, like axioms in mathematics. Then it builds up a vocabulary on top of them.

Left is defined this way: [X is on the right side of your body.] = X is on this side of your body: Most people write using the hand they have on this side of their body.

So "on a side of", "people", "be/is", are included in those first 60 words. "body", "write", "hand", etc. are defined after the first 60 words, but before "right" and "left".

Sure, but "go ask a bunch of people which hand they write with" seems like a huge copout to me.

It's like if prime numbers were defined as, "The numbers most mathematicians say are prime."

So how would you propose to explain which side is which?

Use an asymmetry of the human body. Left is on the same side as your heart.

Which is again, only most of time (cf: dextrocardia).

Not as common as left-handedness (one in ten), but still in the order of one in ten thousand people.

Your heart is squarely in the middle. You feel the heart beat on the left because the ventricle is larger.

Left: Face the noon sun. Lift one arm to point at where it was first seen this morning.

Edit: works where I live. YMMV. Adapt as needed (which I thought about saying but decided wasn't needed here.)

> Left: Face the noon sun. Lift one arm to point at where it was first seen this morning.

That...fails to be generally accurate pretty badly.

As a quick example: north vs south hemisphere swaps this.

That's approximately correct, though It think it's most actually “north or south of the subsolar point at noon on the day in question”.

North and South of the tropics will give either consistently right or consistently wrong answers, but within the tropics you'll get different answers on different days.

The main point for me is that you don't only need a good definition of left (/right), but you also need to keep your audience in mind.

If you ask me "what type of person doesn't know what 'left' means", my answer would be "either a child or a foreigner who just started learning the language". For that audience, even saying "the side where your heart is" (like some other comment suggests) would require knowing what "heart" means, which might not be a good assumption for this specific audience.

lesson 12 covers left/right (360 words).

Why don't you follow the link and find out?

I asked because I followed the link and couldn't figure it out. I couldn't even find the list of 60 words, and clearly I'm not alone as evidenced by the fact that the top of this thread is the top-ranked comment.

Also, please reread https://news.ycombinator.com/newsguidelines.html

That’s an about page. If you go to the home page then you can see those words:

http://learnthesewordsfirst.com/

The site isn’t the prettiest nor most logically laid out but the link to the words themselves was pretty clearly labelled (in my opinion).

Reply on news.ycombinator.com