Hacker News new | past | comments | ask | show | jobs | submit login
There’s no homunculus in our brain who guides us (nautil.us)
56 points by dnetesn 34 days ago | hide | past | web | favorite | 32 comments

I wonder if this works differently for people who are Guugu Yimithirr or Sambals?

Guugu Yimithirr is an aboriginal tribe in Australia, and Sambals are an ethnic group in the Philippines.

The Guugu Yimithirr and Sambali languages use absolute directions rather than relative directions to specify where things are in relation to other things.

For example, imagine standing in the south end of hallway that runs north/south, telling someone where the bathroom is. We'd say something like it being the second door on the left.

A Guugu Yimithirr or Sambal would say it is the second door on the West. Sambali does have words for left, right, front, back but they are mostly just used for describing things on the body, like a left hand.

Another interesting use of absolute rather than relative direction in Guugu Yimithirr is how they visualize time. Researches showed Guugu Yimithirr pictures showing someone throughout that person's life and asked the Guugu Yimithirr to order them chronologically.

When they did this, the arranged the pictures from east to west, with youngest on the east. So if the person doing this was sitting at a table facing North, they arranged the pictures right to left. If they were facing South, they arrange the pictures left to right.

More detail on Sambali in this Reddit comment from a native speaker [1].

[1] https://www.reddit.com/r/todayilearned/comments/1m6l0b/til_t...

Because the sun rises in the east and sets in the west I guess? Very remarkable!

Inuit sign language has similar characteristics when describing places: https://en.wikipedia.org/wiki/Inuit_Sign_Language

I read years ago about an anklet that contained 16 tiny vibrating motors, and whichever one was north facing at any given moment would pulse rhythmically. Wearing this device for a few days resulted in the conscious signal fading, but a wild increase in the accuracy of the wearer’s mental mapping and navigation abilities.

Seems like a fun thing to try out, and probably simple to build.

There was an interesting experiment about this, and about the effects when you stopped wearing it (people's internal navigation systems took a serious hit). [1]

[1] - https://www.wired.com/2007/04/esp/

I knew about the feelSpace belt already but this was a very interesting read and a lot broader than I had expected it to be. Do you want to submit the article here? You would have my upvote right away.

I have submitted it myself now. I hope you don't mind.


Oof, I don’t know if I like the idea of losing my existing navigation abilities. Maybe I won’t mess with that! I wonder if it would be useful for people with Developmental Topographical Disorientation though:

> she could only take jobs that allowed her to commute entirely along straight roads (even curvature threw her off), and after-work happy hours at new bars were out of the question. Others with the condition say it’s like being a perpetual tourist in places that should be familiar. People with DTD can still theoretically understand verbal directions, read maps, and use GPS, though with difficulty; Roseman said that trying to understand GPS directions while driving is too overwhelming for her.


The one I've had on my radar is called North Paw[1] (from 2009, five years after rincebrain's article). It's been on my list of things to try for a while. But that build is 10 years old now and as a result it's huge—it looks like a house-arrest ankle bracelet.

We've been through a decade of electronics miniaturization since then (thanks to cell phones), so I feel like it should be possible to make it more discreet. Does anyone know of an updated build? I'm an amateur when it comes to electronics so I could use any help I can get.

[1]: https://sensebridge.net/projects/northpaw/

I suppose a relatively easy way to get something similar is an app that tracks your movement and causes the phone to vibrate the more directly north you are moving? Or perhaps different 'rhythms' depending on whether you're facing north or south.

It's not quite the same, but it would be easy to make such an app and I suspect our brains will still be able to use the information as we get used to it.

Not sure how it would impact battery life though...

the BBC microbit has a compass built in, and you can program it easily in python. shouldn't be too hard to hook it up to some of these motors (https://www.adafruit.com/product/1201) to do what you want.

Wait, why not just look at the sun (or more accurately, the shadows it casts) during the day and memorize the direction landmarks/geographic features face?

This is what I do and it almost always works unless I go to a new place at night.

I use Direct TV dishes as they are aligned generally south, assuming I can find them while travelling.

That’s a very good trick, I’m disappointed I didn’t think to do that.

I'm surprised scientists would ever theorize that way finding is allo centric.

I've often "stitched" together two areas that I knew were close simply by walking from one to the other for the first time. After doing that, I tend to stick to the first such route, even if I am consciously aware that there's a more efficient one. Is that not a common experience? Does anyone think about space in a way that they can plot everything absolutely and accurately speculate about routes they've never taken?

Your experience sounds to me like modular mapping. I also tend to have abstracted, modular mental maps. They are modular due to gaps in my knowledge or other practical reasons (compression?). I don't think this modularity says anything about ones propensity to plan or navigate in a detached, third-person perspective versus in a first-person perspective. Clearly, with a more abstract mental map, you cannot plan in purely geometric terms. But, that doesn't distinguish whether or not you think about your environment in a holistic, out-of-time, out-of-body perspective.

I am sure this varies by person as well, but I think people in grid-oriented cities or in relatively flat, open terrain are more likely to navigate and plan in a concretely geometric way. You can reason about and infer paths on a 2D plane and benefit from having a good knowledge of relative cartesian positions for different locations. You don't need detailed knowledge of other routing constraints when the paths are more interchangeable.

If limited by transportation modes, you might be biased towards a more hierarchical mental model. I.e. I know some areas abstractly by the highway, exit, and local neighborhood without a clear model for how (if at all) several such areas connect together via local roads. Users of public transit systems are likely to have similarly hierarchical knowledge of trains, buses, and transfer stations. Popular transit system maps are greatly simplified to show this abstract topology without the full spatial complexity of the real physical system.

When faced with challenging topography and circuitous routes, I think a more abstract topological model is more common. I have a rough sense of position along a route in terms of percentage completion or relative ordering of landmarks. This includes some topological knowledge such as trail junctions, water crossings, or ridges and saddles along this same abstract route. I don't retain a detailed memory of every twist or switchback in a convoluted trail. I can plan to walk around a lake or the rim of a valley without detailed memory of the shoreline or cliff edges themselves.

However, I do also have a pretty good sense of my real orientation, i.e. where is north, where is the sun, which way does the surface tilt, and which bearing would take me towards other places I know (if I could fly). I have used that to speculate and plan routes cross country without any visible trail or trail markers. But, I am not foolhardy, so I do consult maps as well!

Related article about London Black Cab drivers brains


> Maguire discovered that London taxi drivers had more gray matter in their posterior hippocampi than people who were similar in age, education and intelligence, but who did not drive taxis. In other words, taxi drivers had plumper memory centers than their peers. It seemed that the longer someone had been driving a taxi, the larger his hippocampus, as though the brain expanded to accommodate the cognitive demands of navigating London's streets.

The article is about how we navigate. I've noticed that some people are quite good at giving concise understandable directions and other are virtually unable to describe in any useful way how to drive a route they drive every day.

I knew an intelligent, successful person that would give directions like "Turn at the stop sign two streets before the yellow house that has the flag pole. Then keep going, but at the train tracks you've gone too far so before that turn at the light. After Elm street, well right there they were going to build a house on that corner, but they didn't because the husband got a different job in Ohio and I didn't know them that well when they moved away. But anyway make a turn, right no left well the road kind of goes right but stay to the left, but not at the red sign in front of the old hardware store, instead I drive to the next street.", etc.

This person navigates by visual sequences and landmarks. Others navigate by direction and distance. You may be interested to learn that there is a large gender-correllated divide between the two methods. Learning about these two methods has made me much better about communicating directions, because unfortunately people dont often grok the "other" way.

It is tempting to tell a "just so" story about this, but I do think direction / distance instructions helps two groups rendesvous or coordinate in new areas, whereas landmarks are natural ways to find things in known areas.


These two ways of navigating are formalized in pilot training (VFR).

The first one involves landmarks, and the training consists of finding good ones. They need to be highly visible, not ambiguous, and preferably referenced on maps. Following highways is a good strategy.

The second, called dead reckoning, uses the watch and compass. With knowledge of your airspeed, wind speed, and heading, you can estimate your position after a set amount of time.

Usually, a combination of the two is used: dead reckoning between landmarks.

tldr; the article you cite asserts that men encode w/ distances and directions more than women, not that they encode w/ distances and directions more than with landmarks and instructions


The article you cite specifically deals with how we encode information when we look at maps, so a few important qualifications. Men encode navigation in terms of distance and direction more than women _when dealing with maps_. I guess it's reasonable to extend the finding to say that men encode in terms of distance and direction more than women in general. Of course the study doesn't prove this, but there's so much other supporting research out there. Anyway, _more than women_ doesn't mean that men are encoding by distances and directions _more than by landmarks and instructions_. Nothing in the abstract of the article you cite supports that claim and I am personally rather skeptical of it.

One sibling comment mentions that '(far more than not) people explain the route based on things trivial to them' (I assume this means "in terms of landmarks and instructions"). If most men were out encoding the majority of their navigation in terms of distance and direction then one would expect to encounter navigation explanations in terms of these ~half the time, yet we don't.

Another comment (though this example feels a lot less persuasive to me) mentions that we use watches and compasses to navigate by distance and direction. If we're so naturally good at navigating with distances and direction why do we need the aid of these tools? In the opposite case we'd be so good at navigating with distance and direction that these tools wouldn't exist, but our visual and verbal memory would be so bad we'd need to carry around pictures and lists to help us navigate. Yet this isn'tt the case.

Anyway, I think you only have to carefully examine how you personally navigate your environments to come to the realization that we naturally navigate mostly (almost exclusively?) in terms of landmarks and instructions. If I had to guess, distances and directions (and some other intuitions such as "place") are really only a feature of spatial awareness which does not normally extend beyond the "room" you're in and the things you're looking at, and which you usually only use to navigate within these immediate spaces. Interestingly, because of this you can sometimes cause weird dissonances by engaging system 2 thinking about the real spatial relationships in you environment. In my case, the room I'm in happens to be contiguous with a bathroom and I happen to be sitting about three feet away from a toilet that others use. Usually I'm not aware of this, but now I feel like my personal space has been weirdly violated.

I travel a lot and so often (far more than not) people explain the route based on things trivial to them; go right at Jose-with-the-nose, go left at the office of Manuel etc. I have never been to the town and obviously this info is not on the map.

Basically if google maps cannot find it, calling or whatsapp with the owner almost never helps at all.

The story about the foreigner asking for directions in Ireland comes to mind. The (helpful) reply: "You want to go to Ballywhereever? Well, I wouldn't start from here..." Top that!

There's something similar associated with people from the state of Maine in the US: "You can't get there from here."

Turn left where the old Johnson house used to be, then keep going until 2 miles before the gas station, then turn right. Some Chevy Chase movie=funny farm?

I don't really understand what the argument is about here.

Different people organize their knowledge in different ways.

From cybernetics we have the Law Of Regulatory Models, to wit: every good regulator of a system must be (contain) a model of that system. (Conant and Ashby, 1982) http://pespmc1.vub.ac.be/ASC/LAW_MODEL.html (In humans this model (in the mind, there are others, e.g. homeostasis) is called the "ego".) A mobile entity situated in a physical environment and capable of navigating around it in repeatable ways is also a system so the regulator (in the brain of the entity) must contain a model of the physical environment.

I would be really surprised if, at any time, using any method, you could find anything that looked like a map inside the brain of a human, even while visualizing a map.


I don't really understand what the argument is about here.

Indeed, this "theory" seems to involve talking some particular "thing" the brain "does" very prematurely, at the level where the speaker just knows too little about the brain's internal model to make their concept even meaningful.

Article: "At the time, the dominant theory in psychology for how people find their way was the cognitive map, which posits that humans and many animals create representations of the environment in the brain that they use to navigate the world."

As you say, pretty much tautologically, an organism that navigates its environment has a model of its environment. But to make the above statement falsifiable, the theorists apparently put the theory as "a uniformly coded model existing in a particular spot in the brain", which is a big, big step.

The thing is, if one has ever worked on a large computer system, one knows that just because a program shows some functionality, that doesn't mean that there will be discreet module separable from other other modules that implement this functionality (there might be such a module but it's functions are likely wrapped with many other modules). Yet it seems like cognitive scientists reason in this constantly.

>I would be really surprised if, at any time, using any method, you could find anything that looked like a map inside the brain of a human, even while visualizing a map.

No but there's probably a specific part of the brain that does deal with visual mapping.

I read this excellent article some years back about the test that aspiring London cabbies, the ones that drive the black cars, take and the preparation required "The Knowledge" to pass this difficult test. Since London's road system is so byzantine and requires lots of insider knowledge, it is unlikely that the cabbies will in the near future ever be supplanted by companies like Uber or Lyft.

Anyway turns out the gray matter in the posterior and anterior hippocampus of some of these guys were tested before and after studying for the test and there was significant growth. Perhaps this would be the place to start to pin down a neural equivalent to a real-world map.


> there's probably a specific part of the brain that does deal with visual mapping.

Sure! In fact, you could say that the eyes are the part of the brain that has a close isomorphism to (a tiny subset of the EM manifestation of) the real world. Once you get about 2-3 cm from the retinas, though, I think that information has been encoded in ways that make it unrecognizable from the outside. I'm saying I doubt we'll ever be able to decode a model (let alone a map) of London from studying a cab driver's hippocampus.

This is probably not true. There have been experiments where faces have been reconstructed by measuring the neural responses of monkey's when seeing faces: https://www.bbc.com/news/science-environment-40131242

Wow, that's cool! Cheers!

I have to adjust me "priors", as a Bayesian might say. Still, monkeys aren't humans, and faces aren't London. Recognizing faces and their expressions has been a thing since faces first evolved, eh?

Place cells have recently made the rounds in popular science which are basically the cognitive topological maps being discussed. The fidelity of these maps remain to be seen. Some examples: https://www.nature.com/articles/ncomms14531 https://www.nature.com/articles/s41467-019-08550-1

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact