Friday, January 09, 2015

Fermi Problem Friday: Smell Ya Later - Communicating with Stink

Did I smell you say what I think I just smelled you say?
Back in July of 2014, researchers writing in the Journal PLOS ONE reported a new discovery in the study of gorillas and the odors they produce

The researchers knew that a gorilla's emotional state is often indicated by the smell it emits. Fear, anger, affection and other forms of arousal or stress caused them to release various distinct odors.

That's not terribly surprising - even humans can smell fear in other humans - but the study seemed to show that gorillas can actually control the smells they emit, apparently to intentionally communicate complex ideas to other gorillas.

They can also quickly shut off their odors when it's better to be seen and not smelled, as may be the case when in the presence of large and potentially dangerous silverback males.

Although the gorillas use smell for relatively sophisticated communication, the researchers argued that that the odors emitted by the gorillas they studied lacked the nuance necessary to be considered evidence of a true scent-based language.

So what would it take to make the leap from rudimentary communication to an actual language? And would it be possible for humans to learn to understand a language consisting of smells? Maybe a little Fermi-problem estimation can provide an answer.

To begin the Fermi problem, you have to have some idea of what a language is. Now I'm no linguist, but for the sake of argument I'm going to say a language consists of words that can be arranged in various ways to communicate information. I mostly speak American English, so I'll consider that the standard.

As it happens, most adult English speakers know between 15,000 and 20,000 words.

Speaking at a reasonable pace, people typically utter on the order of 100 words per minute.

It's tough to listen to someone speaking much slower than that (say at 10 words per minute). It's also difficult to understand someone speaking speaking much faster than 100 words per minute. If you've ever heard an auctioneer or a 350 word per minute speed debater, you know what I mean.

Now this is where I'm really reaching, but you could say that the above estimates suggest that a reasonable language would have at it's disposal

(100 words/minute)x(20,000 words) = 2,000,000 words squared per minute

I'm not sure what that means, exactly, but I'm proposing that if you had half as many words available to you but spoke twice as fast you could still have a 2,000,000 words squared per minute language that would allow you to communicate just as much information as normal English.

Assuming that a distinct smell counts as an odoriferous utterance, we decided to figure out how quickly a creature that communicates with odors could transmit a word. In my experiment, I lit a match one meter from one of my office mates, then put it out to see how long it would take for her to get the stinky message. 

As you can see in this video, it took about 15 seconds for my colleague to detect the smell one meter away (which she indicated by raising her hand), and another minute for the patient fellow four meters away to detect it. So the stink word rate is about four words per minute over one meter in still air, and one word per minute at four meters.

That would make smell communication much slower than sound, obviously. On the other, hand new research suggests that we should, at least theoretically, be able to distinguish a trillion distinct smells.

Potentially, using my ad hoc equation above, a smell based communication system could be the equivalent of a language with

(4 words/minute)x(1,000,000,000,000 words) = 4 trillion words squared per minute

That makes it four thousand times faster a communication method than speaking.

Of course, we wouldn't necessarily learn to identify all 4 trillion smells with specific ideas, but if we learned as few as 500,000 then we could effectively speak with smell at the same rate that we now communicate with words.

Unfortunately, humans - at least the ones I know - can generate very few smells on demand, so the breakdown in the whole scheme  would be on the transmission end, not the reception end.

But maybe somewhere out there in the wide, wide universe, on a planet that's too noisy for speech and to dark for sign language, there's an intelligent species that communicates with odor. Heck, if apes hadn't evolved into creatures like us with sophisticated vocal chords, maybe even on Earth gorillas that communicate with smell could have developed a true stink-based language as nuanced and versatile as the grunts and squeaks that we call English. 

No comments:

Post a Comment