Words Beget Words

I love words. Used in the right combination and order they can do all sorts of wonderful or horrible things to my brain. However I've recently been thinking about the greater power they have, the idea that, words don't just allow us to explain our thoughts, they allow us to think them. That they act like little maps, allowing us to link up things and people with ideas and concepts, creating a more complex web than we would ever have arrived at without them.

The first place I encountered this idea was in George Orwell's 1984. He wrote an entire essay detailing how the totalitarian government created the new language of Newspeak specifically to be limited so that not only was there no word for rebellion, but there was no way of even constructing that concept from the existing words. This is essentially the same idea that I am talking about, that without the word for “rebel”, without a way of saying it out loud, we wont even be able to think it internally.

However, I think Orwell has chosen a singularly negative way of interpreting this idea. Indeed the more recent catalyst for this entry was an excellent podcast: Radiolab – Words, which delves into the science behind the idea. This week, using that science as my starting point, I'm going to attempt to show a much more redemptive side to the power of words

The power of new words is that they allow us to express new things. To be able to say “that is a car” rather than “what is this strange horseless chariot of yours?”. However there is real evidence that without the words to express them, we simply aren't able to think certain ideas. The wonderful example from the podcast I linked above can be found here, but in case you don't fancy listening to it, I'll try to put the gist here (if you do listen to it, just skip the next paragraph... they explain it far better than I expect I will).

It started when some researchers found a school with a large number of deaf children, where the teachers didn't know how to sign. Obviously this meant the kids didn't learn much, but it also meant that, because there were so many of them together, they created their own sign language. At first it was really basic, and this is the interesting part, because access to such a limited language allowed the researchers to do some tests with respect to that. They showed a video in which a boy puts his toy in a chest then leaves the room and, while he is gone, his brother takes the toy and hides it under the bed. The children were then asked, when the boy comes back, where will he look for the toy and, surprisingly, they said that he would look under the bed. The exciting implication is that this is because they had no words for think and know, they simply weren't able to infer that the boy didn't know what they now knew.*

It's really fascinating, we can think, albeit a little fuzzily, about what this might mean for someone with very few words. For instance, if the only word I have is “cookie”, then I might point to my mouth and use this word, expecting to taste a cookie soon. When you leave the room to fetch me my cookie, I will have no idea where you have gone. If I have a word for fetch/bring, then I might be able to infer that a cookie is being brought to me, but then, with no word for distance, or even time, I wont know how far the cookie is coming from or how long it will take. Obviously I am taking this to extremes, humans without words aren't simply unable to interact with the world, but looking into the evidence it is shocking how close we are to that state in some ways.

In this study (warning that's a pdf link, so may take a while to load) they demonstrate that even in something as utterly basic and universal as spatial reasoning (in this case figuring out which corner of a box is which) language makes an appreciable difference. Having the words to say this is to the left of that, allows you to recognise these conditions better. Another example, this study demonstrates that there are invariant factors in our understanding of geometry (which aren't effected by language), but they also reference some examples which definitely are. For instance comparing a couple of primitive tribes which were given pictures of shapes to sort, only the tribe which had words for the number of edges on an object ever sorted the pictures with that in mind, the other ignored shape entirely in their organisation.**

It surprises me that these things make such a difference, I would have thought it was reasonable to assume that shape recognition was totally independent of language. It's also worth noting that these studies look at the things which are easy to test, that is, very concrete and easily presentable ideas. More complex concepts may work in the same way or they may not, but it's certainly a lot more interesting to assume that they do.

In the previous section I presented rather a lot of actual hard science to demonstrate that this is happening, what's missing however is any mention of what is causing it, how words effect our thoughts. Science hasn't yet started to describe this, but that doesn't stop me, for the sake of intrigue, making my own attempt to do so.

I choose to think of it as though we have a lot of different nodes in our head, not individual neurons, but perhaps collections of them which make up various concepts or images. In this way we might have a concept of “car” which is a vague idea of what constitutes that type of thing. That would naturally be linked to a collection of images which show us what cars look like, then probably also to a very particular image which we know refers to our own car. What I'm getting at is that, it is these connections which allow us, when thinking (in language or otherwise) to jump between various ideas in an agile and context dependant way (to immediately go from a description 'Andy was in a car' to a reasonably accurate mental image of that scene, with no further description needed).

In this sense what is happening when we learn a new word like “time” is that first we have all of our individual concepts of the passage of time: the sun moving across the sky, the space between setting out to get somewhere and arriving there, the gap between dropping a stone and it hitting the ground. This new word then allows us to link them, to match them up so that rather than having a vague instinctive understanding of each, we can see that in fact they are all the same. I knew what time was before I had the word for it, but learning the word allowed me to better marshal my thoughts, to use the idea in a much more plastic and informative way.

This is what I think new words are doing. They are taking different nodes in our brain and providing a link where before there was none. This then allows us to much more easily leap between those nodes, making that connection not just a vague possibility but a hard wired known fact.

This whole way of understanding language seems to me hugely important. Contrary to Orwell however, I am going to choose to look for the possible benefits to be found here and that is what, over the next few weeks, I will be looking at. Next week I'll go over the changing of language and whether I think it is possible to move it (and so humanity) in a specific direction. The week after I may, if the subject still interests me, look at one particular word and the harm which I think it has done to our society.

*[The scientists also found, when they went back years later, that the language had developed. That the children now knew that the boy would, mistakenly, look for the toy in the chest where he left it. They even found the adults who had contact with these children and their new words, also now understood this fact.]

**[As I was finishing off this entry I came across an excellent visual representation showing how having different words for colours actually effects how you see colour. You can see it on youtube here]