I find it hard to believe that Zipf's law is at all meaningful for non-human species. We have evolved in a very specific way, such that we are limited in the bandwidth of our transmitted communications, as well as in our ability to memorize large vocabularies. There is no reason that this is true for other species, including dolphins; they may be able to distinguish and produce significantly more nuanced "words" (if that concept even exists); they may be born with, or be able to quickly learn and commit to memory, the "distance graph" between a huge number of distinct signals that represent nuanced concepts. Consider an organism that can quantize a dense signal, treat it as a hash table lookup, and reference a specific concept in their memory; these transmissions would look like pure noise to us, and any error correction might be written off as periodicity. We'd never know.
A few people speculate that sperm whales might be transmitting sonar "images" in their communications [1]. At first it seems like a crazy idea, and maybe it is---but it isn't impossible since the whales perceive their environment via sonar. If we could directly and efficiently communicate synthetic images to each other, then "words" might never arise.
Very interesting thought, whether in turns out to be true or not. Seems plausible, at least. If not whales, maybe some other species.
The existence of sonar as a highly evolved sensory experience already tells us about scale of possibility within the creative repertoire of biology. Our brain design is clearly influenced a lot by the sensory and communication tasks that are important to us.
For example, human smell is something unimpressive, by mammalian standards. We have smell and it affects us occasionally. It says 'don't eat that!' or somemsuch, sometimes. We also have some pheromones that are a basic scent communication method. Many mammals make 90% of mating decisions based on smell. Some species can react to a 6 day old pee stain the way humans react to strippers. Some find food from long distances away. They use highly developed scent based communication to transmit information about themselves (mostly) to one another. Something can smell scary, in a "Shit, that gave me a fright!" kind of way. It's not a stretch to imagine they dream or think in smells the way we do in words, sort of.
We communicate with sound in amazing ways. Right now we are communicating using sight (most of us) in a way that piggybacks on our brain's special design to understand words, which are sound based. This is fairly common in our biological branch, words... or so it seems. Anyway, the design is flexible to be used in various ways. You can learn language without having hearing. There is obviously a lot of potential for nature to be novel.
To be fair, most people massively underuse their sense of smell. Feynman described being able to distinguish who touched long untouched books by smell - sounds a bit farfetched until you try, it's not that hard for most. Then again, it's also not very useful in today's society.
If they are projecting sonar images at each other, this would be much easier for us to determine than if they are using serial, syntactical language, right?
They may not be projecting shapes of actual objects, but if they are projecting shapes at all, it would probably be easy to record and reconstruct them visually for us to see, no?
I'm thinking if they so use 3-D sonar communication, they probably have a vocabulary of geometric symbols.
For instance, with our vocal apparatus, we can imitate sounds like "bark bark" or "buzz", but instead we say "dog" and "bee". This allows us to talk about things that don't make sound, including abstract ideas.
Perhaps they have 3-D geometric symbols, analogous to our words. Instead of imitating the sonar reflection of a school of fish, they might just have a shape that means "school of fish".
If they are projecting sonar images at each other, this would be much easier for us to determine than if they are using serial, syntactical language, right?
They may not be projecting shapes of actual objects, but if they are projecting shapes at all, it would probably be easy to record and reconstruct them visually for us to see, no?
Theres an IEEE paper by H. Zhang from 2008 where the author was surprised random open source java source code followed zipf law.
It would be interesting to combine that paper's idea of analyzing source code with your idea of peculiar modes of thought and see if zipf still holds under those peculiar mental models for large enough code samples. For example does intermediate OCR API code follow zipf the same way as numerical analysis code, or some weird string theory code?
Going into the hard sci fi (or weird fantasy) arena of fictional writing, I think an elaborate yarn along the lines of ye olde fortune teller or program correctness prover or traveling salesman router, based on zipf and compression algos would make an entertaining story... "So back in the 22nd century when they proved 11 dimensional string theory was true because it was the only geometry where the DSL describing it followed zipf-law statistics, and everyone knows only real things that are true follow zipf therefore the simulation must be accurate..." I suppose the plot twist would be something along the lines of outright falsehood documents also follow zipf...
While I agree we've no way to know if Zipf's law applies to alien firms of communication, it seems extremely unlikely that it would apply to random or natural signal profiles that were not communicating information.
So failing to conform to Zipf's law doesn't prove a signal is not meaningful, but confirming to it is likely to be very strong evidence that a signal is a form of communication.
> So failing to conform to Zipf's law doesn't prove a signal is not meaningful, but [conforming] to it is likely to be very strong evidence that a signal is a form of communication.
Then again, Zipf's law is known to describe many non-communicative, "natural" distributions like the sizes of grains of sand, the sizes of asteroids, and the populations of cities (one of Herbert Zipf's original objects of study). And conversely, many levels of representation of human communication have non-Zipfian distributions, like phoneme distributions. (See my comment history for references.)
It's unlikely for evolution to produce such a communication mechanism as it involves non-local refactorings. Under that constraint, and assuming efficiency is selected for, they will reserve shorter symbols for more common concepts and end up with something like Zipf's Law (but again, not a global optimum).
Now, if they were intelligent, they would (as we do) come up with highly compressed encoding schemes that would have no apparent patterns to an outsider, except perhaps control signals at the start and end. But if they wanted to be heard by aliens, they would avoid using obscuring mechanisms like compression.
What if they were smart enough to see that contacting aliens wasn't in their best interests? What if they understood that continually seeking to harness ever more energy was ultimately futile and just decided to enjoy their existence without disturbing others or being known to anyone else?
Someone read Douglas Adams?
(if not, you probably should)
Anyway, I would argue, that wanting to allways remain in the same state of being, is not very smart, since the enviroment is going to change anyway ... so you got to evolve to adopt to new enviroment changes (ultimately the collapying of our sun)
A big part of human communication, especially spoken communication, is redundant, to allow error correction. Repetition and redundant context clues allow us to reconstruct meaning from lossy signals; language is a funny balancing act between compression and interpretability.
>We have evolved in a very specific way, such that we are limited in the bandwidth of our transmitted communications, as well as in our ability to memorize large vocabularies.
Maybe one of the reasons we can't quite bridge the communication gap and "do inter-species communication" is that we feel that we are above the target species in question, evolution-wise.
Maybe this is a huge mistake to make to think that we have evolved to handle complexity, rather than devolved to avoid simplicity.. I don't know, but if I could talk to a dolphin about it, I really wonder what they'd say.
(Seriously, I've thought about this often, since I came to understand Zipfs Law in the 80's and listened to too many Devo lyrics..)
Its quite possible that we're surrounded by alien communication, but are too arrogant, as a species, to be able to recognize it. I'm fairly certain we have some significant evolutionary steps to take before we can 'work it out' ..
> such that we are limited in the bandwidth of our transmitted communications
Every "creature" will have to use energy. Conservation of the energy means a linear optimization to minimal bandwidth. The channels (chemical, sonar, infrared, etc) and patterns are the most varied parts as different conditions produce different optimal expression.
Right. Any series of bits can carry information in some encoding. Testing if it matches Zipf's law is an interesting sign that it might be a human like language, not proof.
It's certainly interesting that their language has statistical properties closer to human languages than that of other animals tested.
"quantize a dense signal, treat it as a hash table lookup, and reference a specific concept in their memory"
Did you just describe a word in a foreign language?
Let's take string of letters:" t e r v e - m a a n - a s u k k a a t "
Let's assume that I can somehow quantize that signal into syllables: ter-ve-maan-a-suk-kaat. Then hash table lookup against vocabulary: terve = hello, maan = earths, asukkaat = inhabitants.
Yes it looks like pure noise if you can't separate words from each other and message length is short compared to complexity of message. Also if you only see a pitch graph of me saying that stuff, it's even more indistinguishable of random noise. Exactly the same problems as in the article.
I'm reminded by an idea from deGrasse Tyson, not sure if it's his... on worms, who share 70% of their DNA with us. Yet they're completely oblivious even to our presence as we walk by, forget being able to communicate with us, or grasp the intricacies of human civilisation and its history.
What are the odds then, that there may be a species not even insanely different from ourselves, which we simply cannot even observe or be cognisant of. If true, it'd make all our efforts to seek out, let alone communicate, with aliens, naively cute.
It's a fascinating idea, although I've never explored whether it has any merit. It may just be one of those ideas that sounds amazing but falls apart if more closely inspected.
I don't think "amount of shared machine code" (DNA) is a good metric for functional similarity? What percentage of machine code do two small UNIX binaries (statically) linked against libc share? Potentially a lot. But obviously they would have next to nothing in common, functionally?
That we share so much DNA with primitive creatures is just another way of saying that amount of shared DNA is not closely related to functionality similarity.
I'll try. It's a lot easier to be hidden from the perception of something that isn't even capable of perceiving itself or its surroundings, only reacting to simple stimuli.
A lot of animals are smart enough to at least get a pretty decent picture of us.
It could maybe follow that there are beings that we would be unable to really comprehend the way we comprehend other humans. But it would also be pretty likely that we could develop models, heuristics and statistical analysis to extract meaning about these theoretically confusing aliens.
We understand quite a bit about our society- which is basically a hyper complicated animal. We deal with the gaps in absolute knowledge pretty well.
Not only that, but without careful study, we really have no idea what is going on in the worms' heads, and we can't even easily tell them apart. Even with lots of study, we still don't understand the worm's perspective.
If aliens visited us, we probably would not even recognize that they arrived, or we might see their arrival as some natural phenomena.
We best start with dogs, who at least, we can bond with.
We have a very good idea of what is going on in the heads of at least one worm species: C. Elegans. It has 302 neurons and we have mapped how they are connected.
Yet we have no clue what it's going to do next. Which is a serious indictment of the whole obsession with "brain mapping" and localization in neuroscience, IMO.
I think Dr. Who and Star Trek both looked at those types of aliens. I'd guess the only type of alien we would really find hard to understand would be one that lived on a vastly different scale to us, either planet sized or nano sized. Anything from about 05.m to 20m we'd understand. Especially if it turned up in a spaceship. It would have engineering and technology and language. Work in teams, have team members etc etc.
Not necessarily. What if its a singular entity, or a hive mentality without language? What if they use a medium for communication that we do not physically possess (or even have access to via technology)?
The majority of the aliens in Star Trek always fit into the humanoid model, treating them as basic re-skinned humans with the same motivations and ideas, which is a dangerous assumption out here in the real world.
I can't imagine (maybe you can and you can comment here) that a species that can make a spaceship and cross the universe can do so without a language that we could understand.
Just to mine minerals and manufacture metal, computers, fuel, etc etc would mean co-operation on a massive scale -- millions of individuals.
Nan-bots sure, they would be talking alien wifi and machine code. Giant creatures (AIs?) that live slower than us and manufacture each component themselves without the need of help.... like spaceflight from first principles.
I've always gotta throw in a plug for my favorite sci fi novel Blindsight (available to read free on the internet: http://www.rifters.com/real/Blindsight.htm) when I see threads like this.
Blindsight is a very compelling exploration of the idea that there may be different roles of consciousness in advanced intelligence. We know that a conscious decision on say the visual processing loop ends up costing 120ms or so. Why is this a good thing? What if you can have advanced intelligence without consciousness? Some work in AI (and some worries people have about AI) suggests this may be a real possibility as well.
The end result for me ends up being a much more intense feeling of the kinds of varieties of intelligence that may exist in the galaxy, or that we may end up creating ourselves.
The closest thing to "alien" in my worldview is late industrial technology.
Including this very thing I have sitting in my lap, which consumes energy like a living creature and responds dynamically to its environment in a way that keeps it around (it's my trusty laptop). I have opted to keep it around rather than throwing it in the trash (evolutionary selection).
I see that you're getting downvotes but I wish to cast my support for your comment. I suggest those interested in the thought process please check out the book What Technology Wants by Kevin Kelly. The book basically says that technological growth is almost independent of human input.
Interesting thought, and I've been there before, but I've also come to think of technology as an extension of humans. As computers are not currently capable of independent thought, they must [generally] be instructed by humans on what to do. We've also laid out their CPUs and GPUs and busses and power circuitry. Their physical evolution is based on the business roadmaps at places like IBM, Intel, AMD, NVidia, Micron, Dell, etc driven by tasks which they can perform to further the economies of humans. Contrast this to animals, where there is a drive for survival and furthering the species. Computers don't have this sense because they don't have life. Technology is impressive, but in terms of what a computer wants, I don't see it as much different from a shovel; it's a tool created and used by humans for the benefit of humans.
Inheritance of genes separates life from chemistry.
Inheritance of knowledge separates animals from plants and microorganisms. [1]
Inheritance of capital separates humans from animals. [2]
In the same way DNA is an integral part of life, and in the same way the knowledge to live in a particular environment is an integral part of an animal, technology is an integral part of being human.
Raise a human in the wild among goats without human culture, without human technology. As long as that human is separated from culture and technology, would an alien be able to differentiate the human from animal? [3]
[1] Animal born in captivity can't be released to the wild because they didn't have the opportunity to learn from their wild parents.
[2] Hammer, houses, watches, factories, passed down from generation to generation.
Dolphins are mammals, and they are hardly "alien". I would rather look at cephalopods, who had a different evolutionary path (including reinvention of eye).
Their mimicry skills stay sharp across generations because during breeding season, while the strong males are competing physically for access to females, the smaller males with better mimicry skills are mimicking females (!) and sneaking past the boxing ring for a quickie.
The female cuttlefish can also store lots of sperm and decide later whose to use. Gross.
Ok, but in order to analyse a form of communication, you have to have communication to analyse. Do Cephalopods meaningfully communicate? If not, your investigation isn't going to get very far.
They totally do, cuttlefish and some squid have colouring that they can manipulate. They can flash colours across their body to communicate: http://acp.eugraph.com/cephal/
Dolphins (and whales, or cetaceans in general) are still mammals. Which means they have a mammalian neocortex that is very similar to that of a human.
I think crow intelligence is more interesting. They are very smart and capable of using tools. And they evolved brain structures independently from mammals that enable them to do those things.
I hope we value the lives of intelligent aliens more than we value the lives of dolphins and cephalopods. Judging from current efforts in avoiding fisheries by-catch, it's more than zero but less than $1/lb.
This is interesting...but Zipf's law was discovered (according to the article) in 1930's and the paper on the dolphin gradient (or language pattern) being similar to humans was written in 2009. So...6 years later any updates?
Seems to me they just found a fact that was interesting and decided to wrap an article around it.
It has been demonstrated that echo patterns captured from dolphins are a 2d planar projection of the 3d objects in the environment facing them. Thus, dolphins use a form of hieroglyphics where the 'words' are really pictograms. The real question is how humans went from this kind of pictogram/sign language (that our ancestors used) to the abstract mapping of sound<->writing<->picture that we use today. We attribute intelligence as the catalyst for this change. Yet intelligence , per se, evolved to facilitate survival in the face of increasingly efficient predators. Aliens with intelligence would certainly be using a similar abstract mapping. I shudder to think of their predators tho.
it doesn't seem to work for Polish (there's no articles in Polish, "of" is expressed as changed ending of noun, and in most texts I tried "is", "to", "from", "in" etc. are the most common with very similar frequencies).
I would think dolphins and aliens may be more alien than Slavs.
I don't know how big "textfile.txt" is, but try it with a much larger body of text? Also you'll almost surely want a much better tokenization algorithm than that. Zipf's Law can only hold "at the limit" of an infinite stream of text, so as a consequence you get a better approximation with a larger sample.
I think intelligence is overrated in the grand scheme of life, so "alien life" would probably baffle us. Survival of the fittest on Earth is worth nothing if the fittest die as well. Evolution would be slower without intelligence but who cares, is there a race out there?
Alien life would still have to be based on the same chemical and physical laws as terrestrial life, so while I'm sure it would be challenging to understand it shouldn't be an intractable problem. In the end, it's all just chemical reactions.
If the fittest die as well everything's dead, so no more life.
Not sure what you mean by evolution being slower without intelligence. I'd expect the opposite. Intelligence means you can adapt to your environment behaviourally and technologically, instead of biologically so natural selection becomes less of a factor.
First, as far as we know, physics doesn't change depending upon universe position. So, we assume that life has to obey our standard rules of physics.
Second, the chemistry of life balances on a particular set of characteristics:
1) Water is an extremely unusual molecule. It has much more varied reactions than just about any other solvent. This puts life in a particular temperature band.
2) Carbon chemistry is very energetically favorable relative to the other options. Binding energies are close enough that swapping around carbon, nitrogen, oxygen and hydrogen atoms doesn't take excessive amounts of energy.
3) Carbon, Oxygen, Nitrogen, and Hydrogen are all relatively close to one another in size as well. While carbon chemistry does borrow from higher in the periodic table, that tends to be the exception.
While intelligent life will likely be completely different from us, it's very likely that the underlying chemistry is going to be very similar.
The interesting question will be whether we have exactly the same basis DNA pairs. There has been recent successful work to synthesize new "unnatural" base pairs which can then be replicated in organisms.
This is a most succinct and accurate explanation. Thank you - I'm sure I will use it in future conversations. It serves as an answer to the "street light argument" - https://en.wikipedia.org/wiki/Streetlight_effect
That said, have you every thought of how alien intelligence might look that does not adhere to the above chemistries?
There are some great science fiction speculations - life living in the ultra-high-gravity "crust" of a neutron star [0], intelligent Turing machines running in the convective layers of brown dwarf stars[1], just plain bootstrapping an AI environment on the substrate of the Universe.[2]
[0] Robert L. Foward, Alystaire Reynolds
[1] Stephen Baxter
[2] Greg Egan
4) "Room temps" are really comfy for using enzymes to perform energetically infeasible reactions. Too cold and life is too slow to survive and much hotter and you get thermal decomposition faster than life can grow. Its chemically universal... we have the technology to do o-chem and biochem under much weirder conditions than where it actually works best.
5) Trace elements. Despite huge evolutionary pressure and no small amount of experimenting with blood chemistry we're still stuck with red iron based blood. Imagine if you could make blood out of pure protein or even just carbohydrates... But no, every red blooded animal on the planet is stuck refining bioavailable iron, no matter how much of a PITA that is. Would not be surprised if space aliens bleed red like us.
6) Water is awesome for buffer chemistry and this feeds back into your point #1 and my point #4 but homeostasis in the blood or innards in general would be nearly impossible under any other chemistry.
I would not be surprised at all if space aliens were quite familiar with hemoglobin, chlorophyll, some of the cycles like the Krebs cycle. Krebs is very complicated and PITA but most earth life depends on it because it works, if something better were possible, it would have gone extinct half a billion years ago... It would be a fun race, seeing if human and space alien physicists or biochemists start communicating first. My money is on the biochemists, more shared background.
Your points are good, but despite their entropic advantages they remain limited by the anthropic principle. Other forms are likely possible, just unlikely under our set of circumstances.
I would still hazard a guess and say that I would expect carbon based life to be there basis of the lions share of life in the universe though.
Regarding the Krebs cycle and other biological feedback cycles, I strongly disagree.
It came to dominate at a point where it propagated through all the extant branches of life, but it is by no means the only way life could have decided to do the process of energy production.
Once evolution evolves something, it is stuck with that system - no complex biochemical system will evolve into a Krebs replacement whilst Krebs occupies that niche, and evolution isn't going to spit it out of the way!
Well, I wasn't implying that the laws of physics or chemistry change from place to place. More that the spectrum of physics and chemistry is incredibly broad and diverse, even just the part we can actually see. More importantly, I was challenging human-centric ideas of what constitutes "aliveness" and what doesn't, within the realm of Earth-like physics and chemistry, let alone beyond it.
Is there a firm distinction between "alive" and "dead" at all? I think not so much. What we call "dead" is a body whose micro-organisms no longer organise in a common human perceptual window of space and time. With a microscope, some people have been able to see that a "dead body" is actually buzzing with life. Many of the same organisms that live symbiotically with us survive the immediate death of the eukaryote body and pass memories of our memories on to future generations of micro-organisms. They are (usually) symbiotic with, but not dependent on, as the organelles of "human cells" have become, our larger organising bodies. What happens as these wild bacteria and fungi unpack and metabolise our memories on the deepest level? Is that not somehow part of the life cycle? Isn't it possible that "we" somehow experience that in a less integrated way? Poetically, as the inside members processes the death of its larger body, so do the exterior members of the human body, other people, animals and social organisations. The locus of stable control is released, the steering wheel disintegrates. But does life somehow "disappear"? Not at all. It only shifts in, out, to scales whose feedback mechanisms elude mainstream western mythology.
Eventually, over thousands of years, all membrane bound creatures deposit their bodies. What went in (the membrane) must go out(side of the membrane). Some parts are reused immediately. Others are left for deeper cycles (including by not limited to eventual metamorphosis in the high pressures of the deeper subterranean). This part of the life cycle is much more mysterious to me, as I haven't really studied it, and it really is so much more vast. The physics and chemistry of rocks and minerals, liquids, gasses is part of the life cycle. They are part of the evolutionary story as much as any being.
Why are many mystics attracted to crystals and rocks? Because they are mineral deposits of ancestral beings. They carry, literally, the energies accumulated in some ancient, swirling ecosystem. They are a memory of things our bodies had forgotten. Same reason why a lot of scientists are attracted to them.
As for the conversation about "intelligence". The Earth has 400 billion trees. Each is connected to the others by roots, by air, by micro-organisms, by water, by chemicals, by ecological relationships. How many neurons do we each have? Maybe 80 or 90 billion.
Just because plants can't move or talk in very specific human babble noises, it doesn't mean they aren't experiencing something substantial or contributing in dynamic "intelligent" ways to their ecologies.
Anyway, I need to take a break and eat lunch, because I'm "alive" in this certain kind of way where that my fire is burns for tasty food. I'm gonna stoke it with some soup.
It's only an assumption that, existing in the same universe as us, they would be subject to the physical and chemical laws of that universe. Or are you considering aliens residing in other universes?
Edit: I can't reply to your reply, but I can edit this post.
Unless the form that mate selection takes radically diverges from evolved patterns I don;t see how that makes much difference. Since survival for intelligent species is much more loosely coupled to their biology I don't see why sexual selection strategies would mitigate against that. It seems to me they'd follow the same pattern in devaluing biological fitness criteria as not being as significant as they were pre-sentience.
(You're probably already aware of this but just in case you're not...) HN restricts the timing of replies deeper into conversations, sometimes annoying but it suppresses flame wars and what-have-yous.
>Evolution would be slower without intelligence but who cares, is there a race out there?
I'm quite certain that does not have anything to do with "speed of evolution".
I am not a biologist, so I can't come up with a better source, but I'd suggest that everyone would very carefully study at least the Wikipedia article https://en.wikipedia.org/wiki/Evolution which looks like a reasonable summary.
I always thought the closest thing to aliens to us would be plants. I imagine taking to an alien species would be closer to taking to trees than dolphins.
If subsistence is effortless and never a problem, is it really a problem that they never make tools? If humans had a matrix-like virtual reality, you could do whatever you wanted, and never to work a day in your life, I would imagine many people would never work or make tools.
