Scientists use language and logic to translate monkey sounds into English | Hacker News Reader

Scientists use language and logic to translate monkey sounds into English

scientificamerican.com

141 pointspepys28 comments

> ...some monkey dialects can be just as sophisticated as human language.

These kinds of throw-away claims are a form of perennial nonsense. Nothing in the article suggests anything close to this being true. He looks at the shades of meaning of three sounds and yet asserts a similar sophistication to human speech.

It's a regular meme in animal articles that this group (monkeys here, other times dolphins or whales) have language skills that rival humans. It's a kind of warm fuzzy feel-good claim asserted regularly despite supporting evidence.

I don't think it's true but remember that decoding animal language is ridiculously hard. Our understanding of it is basically "play sound - see what animals say". And even that is nontrivial as the article points out.

Relevant farside comic: http://i.imgur.com/QSR2AfI.gif

Whenever I see claims of animal language, I remember the Far Side cartoon where everything the dogs say translates to "hey!"

Sounds to me like the Yo app is the equivalent of dogs barking. And look how many articles are written about its usefulness :))

On another comic note : https://www.youtube.com/watch?v=3YKvtX8E-v8

... and then I realize that the dolphins in the comic are repeating: "hey, do you speak spanish?" and "what's up?" in spanish to the humans... brilliant.

I can't help but notice the uncanny resemblance between the sound of a cafeteria busy with many conversations, and the sound of a large tree full of birds chirping in the morning as the sun rises.

Honestly, I feel like most of the article is throw-away material. They're claiming to have "derived" syntax and grammar from a simple lexicon, which as far as I can tell consists only of referents. In reality, they are just artificially imposing their own ideas of language on top of animal calls. Yes, the calls do convey meaning, but that does not make a language.

It's like gluing a bunch of toothpicks to steel girders, then saying you can build a bridge out of toothpicks.

It's already been done for dogs:

http://languagelog.ldc.upenn.edu/myl/llog/FarsideDogTranslat...

Here's a non-paywalled link to the paper, though I'm not positive that it is the final version:

http://ling.auf.net/lingbuzz/001792

They give a game-theoretic formalization of their theory but I'm not sure that the formalization adds much beyond the summary in the article.

I'd like to know if these sounds are learned, instinctual, or a mix of both. IOW, if you placed a monkey raised elsewhere into this environment would it know and/or adopt these sounds.

Isn't assigning meaning to otherwise-arbitrary symbols/sounds a key aspect of language?

For humans, because of the diversity of sounds we can make, apparently the only thing approaching a universal word is "huh?" [0].

It'd be interesting to see how many sounds a monkey can make. If it's a very low number there'd probably be more universally used sounds but I'd imagine it's probably greater than most animals. I couldn't find a useful way to search for that. Unfortunately a lot of the results tended towards the "What sound does a monkey make" type of response and I'm not versed enough in linguistics or monkeys to query more efficiently.

You can see on the map [1] that the Ivory Coast and Tiwai Island are more than close enough for the monkey's languages to have split off at a much earlier stage and evolved differently. I'd assume this is more likely the case.

So I'm going to assume no, a monkey taken and raised elsewhere probably wouldn't instinctually jump into the trees upon hearing a "krak". But even the article states there are more experiments that need to be performed (although that particular experiment is a little insidious considering the intelligence of the animals you're kidnapping from (maybe if you saved one whose parents were incapacitated in some way)).

Of course, you could just go to the zoo and yell "krak!" at some of the monkeys and see what happens! Might get you some weird stares! ;-)

[0] - http://www.newstatesman.com/martha-gill/2013/11/what-one-wor... [1] -https://www.google.com/maps/place/Tiwai+Island,+Sierra+Leone...

I'm suddenly very curious about the universality of the similar "uh-huh".

Well, it seems strange to me. The article you link seems to suggest "huh" might be a universal word, but in the details, they show that it's pronounced differently all over the place. For example, it tends to be closer to "ah?" in Mandarin.

I think that depends on what you call 'meaning'. Is meaning an intuitive, instinctive representation of something?

What I mean is, for every word that exists, do those words trace back to a reality origin through a pattern of substitutions? Substitutions of symbol to reality are really just an associative relation. Substitutions between symbol to symbol are functionally operative the same way in which a substitution between reality and symbol works.

So then my question is, is language really anything more than remembering that the cherry came from the tree? Once the cherry is disconnected from the tree, we have two things - cherry and tree. But before we distinguish them as parts, we recognize them as a whole. When I walk away from the tree, taking with me, the cherry - what happens if I still use the tree in my mind to represent the concept of fruit? It's a choice function. Does it matter whether I remember these things using sounds, symbols, images, experiences, or feelings? Language is interpreted and expressed across and using all of these domains. A poem carries greater meaning than the words do individually, and that is because there is emotional association that maps to the selection of words. We don't really call 'emotion' language, nor do we call 'art/music/math' language, yet these things arguably can have a strong influence across how we 'know' what language represents.

There is some non-arbitrariness to language.

http://en.wikipedia.org/wiki/Mama_and_papa

http://en.wikipedia.org/wiki/Bouba/kiki_effect

I guess it gets arbitrary at complex enough levels though.

I'd hazard a guess that hok and krak have some component of instinctual/physical nature to them. I personally think that certain sounds are related to physical experiences or expressions of emotions. Obvious ones are surprise of "Oh!" with an open mouth.

"Hmmm" whilst thinking or concentrating, frowning and closing your mouth.

I'm currently watching my son learn to speak and his verbalizing seems pretty closely tied to his emotions at the moment.

"Oishii" means delicious in Japanese and it seems something that makes sense to say whilst you are smiling at enjoying your food. The long "ii" vowel to rhyme with the "e" of "she" in English.

>"Oishii" means delicious in Japanese and it seems something that makes sense to say whilst you are smiling at enjoying your food. The long "ii" vowel to rhyme with the "e" of "she" in English.

Honestly, I don't necessarily agree. For example in Japanese 'iie' sounds very similar to 'yes' or 'yeah' but it actually means the opposite, it means 'no', whereas 'hai' means 'yes'.

If we want to talk about individual phonemes caused by emotional reactions, there might be some truth behind what you're saying, however as soon a we enter the realm of "this word sounds soft so it's positive" and "this word sounds hard hence it's negative" everything collapses.

Obviously you can find tons of examples of words that are different in different languages. I probably confused the situation by bringing Japanese in. My son is Japanese so we talk in Japanese to a baby. I wasn't trying to compare languages. I was trying to talk about baby words.

Yeah it falls apart at any level of complexity.

I just think there are certain cases in often used words and words that babies say or hear a lot at first. Like the mama/haha/papa/baba words. I'm talking about a 'language' in the same way the article talks about an animal language. Like a few often used words linked to emotional states.

I don't mind if you disagree I just happen to believe oishii may be one of these words.

all predicate adjectives in Japanese end in -i ookii, chisai, mazui, etc. the -i is the suffix indicating it is a predicate

oishii thus means "is delicious" - you don't need a "da" after it

I made a mistake by bringing Japanese into it - see my other comment. It's the language we speak at home so the one I use to talk to our son. The emphasis was supposed to be on baby-talk not foreign languages.

For those who do not know about Kanzi, I recommend the documentary “Kanzi: An Ape of Genius” (https://www.youtube.com/watch?v=QBlDGX95eys).

How fun, I'm looking forward to using "Oh krak!" as an interjection. :-)

I would like to see something like this for human babies

Deb Roy's TED talk is kind of interesting on this subject. He captured the initial 3 years of son's language development and did analysis on it.

http://www.ted.com/talks/deb_roy_the_birth_of_a_word

Yes. I saw this when it was released and the data that was collected and compiled is quite high quality compared to most related research efforts. Science needs more research on human development and change over time. Even something as simple as a photo of a person per day. Actually computers should technically be able to unobtrusively measure our heart rate and breathing via the cameras in laptops with the algorithm that amplifies subtle changes in videos such as the redness color shift denoting heart rate and the rise/fall of the abdomen to indicate the breathing rate.