Is lossless information transfer possible?

by kirpi2 min read8th Aug 201211 comments


Personal Blog

I am trying to establish what (if anything) makes human beings superior to other organisms.

I have a hypothesis that, the only thing at which human beings are "superior" to other organisms is that we can transfer information without a loss to other human beings.

This difference may already be well established. I couldn't find a good read on this, so I wanted to ask your opinion.

Many organisms seem to have superior capabilities than human beings; strength, speed, agility, vision, hearing, regeneration etc. And even high IQ (at least on a hardware level on dolphins etc) may not be unique to humankind.

So, my first suspect, high IQ alone does not seem to be a differentiator of our species. (It does not even seem to be predictor of success within the species)

Then I remember the famous experiment of hosing down of gorillas trying to reach bananas. (To which I can't find the original citation) Shortly;

- Some gorillas are hosed with cold water when they try to reach bananas.

- Then they learn to stop trying to eat these bananas.

- The gorillas are replaced with other gorillas one by one.

- The old gorillas prevent new comers from reaching the banana even though they are not hosed anymore.

- When all of the gorillas are replaced, they still stop each other from reaching the banana.

It seems like the information is partially transferred. They can't transfer the cause. But human beings can transfer the cause. So, are human beings the only species that can transfer information without a loss?

The primary assumption I made is that, human beings can transfer infomation without loss. This turns out to be the major discussion topic. Is lossless information transfer is even possible? There seems to be opposition against this idea also.

For example, isn't this a lossless transfer to the reader;

"The sunlight seems yellow to human beings who are at this point on earth when earth is positioned like this with respect to sun"

By the way, by information, I don't mean the representation of it but the information itself. (i.e. Digitizing, wording or syntax for short does not matter)

If lossless transfer wasn't possible, it looks like we couldn't advance (at least) technology at all (like the gorilla example) Or there may be countermeasures to this loss too. (Like various people attacking one problem over and over again independently and finding a combined solution of the problem at an acceptable level)

To sum up, are the following true assertations?

- Information can be transferred within a species without loss.

- Human beings are the only species that can transfer information without loss.

- Capability to transfer information without loss is what makes human beings superior to other organisms.

p.s. For this is my first discussion post, please don't beat this too hard :)

p.p.s. Distinguished does not mean superior.


11 comments, sorted by Highlighting new comments since Today at 10:49 AM
New Comment

I have a hypothesis that, the only thing at which human beings are "superior" to other organisms is that we can transfer information without a loss to other human beings.

Don't forget our baldness, two legs, and broad nails. Highly superior.

Two relevant LW posts: One Two

Thank you for the direction. These articles seem to be the ones I were looking for. (Plato's work and Diogenes' answer is worth further investigation)

Human beings may not be that special. This search of mine may be futile, it being a remnant of my believer times (and I being raised in Eurasian soils).

Even so, "what is a human being" seems to be the right question to further investigate. I guess this will immediately be useful (if not already is) at information security: "is perfect authentication possible?" (with protection against rubber hose technique, or clones etc.)

Noiseless communication has not yet been demonstrated. There is some change (or, in information terms, loss) in even this communication.

The original citation seems to be this:

Stephenson, G. R. (1967). Cultural acquisition of a specific learned response among rhesus monkeys. In: Starek, D., Schneider, R., and Kuhn, H. J. (eds.), Progress in Primatology, Stuttgart: Fischer, pp. 279-288.

Can anyone find a PDF?

It's on scribd here. It's rather different from the urban legend version, and crucially does not contain the step of monkeys who have never experienced the hosing (actually an air blast) stopping the new monkey from reaching for the banana (actually just described as "an object").

I think your hypothesis might be improved if you tried to better define (formalise) the concept of lossless information transfer. It is not obvious to me that losslessness is the best descpription of the difference between gorilla and human communication.

Also, you apparently generalise from one example and even that isn't particularly clear. As you describe it, the gorillas learn that trying to reach the banana causes hosing and they prefer not having the banana to being hosed. When the causal link between banana and hosing is broken, the gorillas still act like nothing has changed. But:

  1. Your explanation by imperfect information transfer doesn't work. The gorillas clearly communicate the cause, since new gorillas know that it is eating the banana which is forbidden. It is not clear whether the effect is communicated since the new gorillas clearly know that they shouldn't eat the banana but needn't to know what happens if they break the rule unless they try to, which they presumably never do.
  2. The observed evidence (and the assumed difference between gorillas and humans, about whom you seem to think that they would discover that banana doesn't correlate with hose anymore) can be explained in several ways that have nothing to do with information loss: perhaps the gorillas simply don't discover that banana does no more cause hosing. Or they are more hosing risk averse and once they discover that the banana might be associated with the unpleasant experience they avoid it even if they have reasons to think that the risk is relatively low.
  3. Beware thoughts along the lines "there is one single thing which causes this important difference and we should only discover which it is". Why do you suppose that there is a single strict and simple principle that distinguishes humans from other animals?

Lossless information transfer between humans may be possible, but it's certainly not free with respect to work and time, and it's certainly not the default.

For instance: Whenever I want to communicate a thought or an idea, for instance verbally or on paper, I find that I must first apply some work-intensive, lossy compression which outputs bad English. The output invariably looks or sounds much worse in comparison to the uncompressed idea I have in my head. Throughput is abysmal. A few minutes of thinking can sometimes require a few hours of writing in order to be communicated with some lucidity. In order to restore some similarity to the uncompressed idea as it appears to me, I need to apply further work-intensive error correction, and repeatedly compare the revised output to my internal model.

"Lossless communication" doesn't quite seem to capture what is special about human communication. The difference is more about complexity of communication than losslessness. Being able to communicate using a fully recursive grammar is the key difference. I think that's the concept you're heading towards with "losslessness".

But which came first: being able to communicate using a fully recursive grammar, or participating in a viciously competitive political/social hierarchy and having the resulting selection pressure dramatically increase the size and complexity of our brains over what is, in geological terms, a ridiculously short period of time?

Yes, humans win over animals because we have language. We're the only species that acquires it automatically, and we have a big pile of adaptations that make language work better and that use language to support other functions. Why are we acting surprised about this?

Sorry, I couldn't get what do you mean by "We're the only species that acquires it automatically". Does it mean human beings have language skills by birth and other animals don't?

What I mean is that humans reliably acquire language skills just by being around people who're talking, without needing an explicit effort to train them. We are, in effect, primed to learn language; and while it's not quite accurate to say that we have language skills by birth, it is true that many of its structural aspects come automatically just from the way human brains develop.