I read about some research today that is both inspiring and exciting. I’ve always been hugely interested in how humans deal with words in their heads and finally we have some interesting answers. This made me play with keyword research and think about how (un)intuitive our methods are right now. A team at Carnegie Mellon have used functional magnetic resonance imaging (fMRI) technology and machine learning techniques to demonstrate how human brains categorise objects. The research was published in Plos One and is called “A Neurosemantic Theory of Concrete Noun Representation Based on the Underlying Brain Codes“. The scientists are neuroscientists Marcel Just and Vladimir Cherkassky and computer scientists Tom Mitchell and Sandesh Aryal.
From the article:
“We use factor analysis of fMRI brain imaging data to reveal the biological representation of individual concrete nouns like apple, in the absence of any pictorial stimuli. From this analysis emerge three main semantic factors underpinning the neural representation of nouns naming physical objects, which we label manipulation, shelter, and eating. Each factor is neurally represented in 3–4 different brain locations that correspond to a cortical network that co-activates in non-linguistic tasks, such as tool use pantomime for the manipulation factor. Several converging methods, such as the use of behavioral ratings of word meaning and text corpus characteristics, provide independent evidence of the centrality of these factors to the representations. The factors are then used with machine learning classifier techniques to show that the fMRI-measured brain representation of an individual concrete noun like apple can be identified with good accuracy from among 60 candidate words, using only the fMRI activity in the 16 locations associated with these factors.”
Now that might just have made our entire approach to keyword research, keyword suggestion, keyword anything…seem naive and simplistic. It is. But this is an entirely new dimension! Actually there are 3 dimensions to consider:
Their research shows that basically the human brain categorises nouns according to 3 dimensions: eating, shelter, and the way the object is used. This means that when we see something or process an object, our brains ask 3 questions:
- Can I eat this?
- Can I use it to protect myself from the elements?
- What do I do with it? Do I hold it, kick it, prod it…?
They also found that they could also predict which parts of the brain would be activated when new words were introduced. They also found that they could tell how many objects were being thought about.
Mitchell says about their research:
“This result demonstrates that when two people think about the word ‘hammer’ or ‘house,’ their brain activation patterns are very similar. But beyond that, our results show that these three discovered brain codes capture key building blocks also shared across people”
If meaning is based on brain function, then it’s possible to predict what somebody is thinking about. There are 25 verbs of perception and action associated with sensory-motor functions, including see, hear, listen, taste, smell, eat, and more.
I wonder how far it is possible to influence what people think about something or how they perceive something based on a particular combination of words. Beyond what we know from poetry and literature of course. I’m thinking banner ads for example and poster ads in the street. Even blog post titles. The significance of the research is beyond this of course, I mean you could potentially communicate with devices by thought alone.
Some say that you could do a lot of evil with such a thing, but then you could do a lot of evil with a pencil. These amazing findings impress us and scare us as well which is only natural. Personally I find this inspiring and its interesting to think about the world I work in and how this changes it. Is there a better way for me to write copy for example?
CMU have a really nice and easy to digest write-up here.
There’s demonstrations available here.