A simple drink of water is actually a complex neurological action
Catherine Falls/Getty Images
The Neural Mind George Lakoff and Srini Narayanan, University of Chicago Press
This is a book review in two parts. The first is about the ideas presented in The Neural Mind: How brains think, which are fascinating. The second is about the actual experience of reading it.
The book tackles one of the biggest questions in neuroscience: how do neurons perform all the different kinds of human thought possible, from planning motor actions to composing sentences and musing about philosophy?
The authors have very different perspectives. George Lakoff is a linguist and cognitive scientist, based, until his retirement, at the University of California, Berkeley. He studied the role of metaphors in thought. Srini Narayanan is a senior research director at the AI company Google DeepMind in Zurich, Switzerland. His work focuses on how artificial intelligences learn language.
The book’s central idea is that the brain uses the same processes for motor functions, language and abstract thought. Similar neuronal circuits and pathways, Lakoff and Narayanan argue, have been co-opted by evolution to perform all these types of thought – which seem radically different on the surface, but have profound core commonalities.
This is easiest to understand if we think about human babies, or other animals without language. While each animal’s experiences are different, there are concepts they will almost inevitably learn: ideas like up and down, motion and rest, force and resistance. Somehow, these must be represented in the brain.
In books like Metaphors We Live By (co-written with his then colleague Mark Johnson in 1980), Lakoff argued that these concepts recur in metaphors we use to convey ideas. Happiness and success are “up”, metaphorically, while sadness and failure are “down”. We use this up-down construction to describe musical notes, even though pitch is determined by the frequency of sound waves and has nothing to do with altitude. Likewise, communication is often described as a physical transfer, in phrases like “getting through to you”.
“ In the first animals, brains were mostly for motor control. Things like language are recent innovations “
The trivial reading of this is that physical metaphors help us grasp tricky abstract concepts. But Lakoff and Narayanan are arguing something deeper: these physical metaphors are literally how we think. This makes sense, they write, if you consider how brains evolved. In the first animals, they were mostly for motor control. Things like language and abstract thought are recent innovations. Since evolution is naturally thrifty, often reusing existing structures in new ways, it is reasonable to imagine neuronal circuits that evolved for motor control were co-opted for language and thought.
Suppose you want to drink from a glass of water. Most of us can do this with little difficulty, but it is a strikingly complicated action. You must reach out with your arm, then use your hand to grasp the glass. Next, you must move the glass to your mouth, and drink. You must decide how many sips or swallows to take, iterating until your thirst is quenched. Finally, you must put the glass down.
This, say Lakoff and Narayanan, is mirrored in our language and grammar. We break complex behaviours and language into chunks. Think about sentences, with their words and syllables, nouns and verbs. A subject performs an action on an object. Or think about past, present and future tenses, reflecting whether we did something, are doing something or will do something.
These physical metaphors also shape abstract thoughts. Lovers “drift apart”; regimes “fall”. If we apply the same metaphorical framing to a phenomenon, we can get stuck – and we often make creative leaps by applying a new metaphor. Instead of that regime “falling”, maybe it is “swept aside” to make way for something new.
It is hard to know how to test all of this. Lakoff and Narayanan propose circuit models that might exist in the brain and underpin these patterns of thought. But we are nowhere near a neuron-by-neuron map of the human brain, so I think true tests of their hypothesis are many years away.
Still, Lakoff and Narayanan do enough to convince me their ideas ought to be taken seriously. What they didn’t do, however, is write a readable book. The Neural Mind, I am sorry to say, is painful to read. It is repetitive and disjointed, leaping from one thought to the next in a way that is exhausting. Ideas that need careful unpacking are dispatched in a paragraph and trivial concepts are expounded at length. And there is no excuse for ending chapter 2 with a sentence spanning 130 words. Basically, I read this so you don’t have to.
