Despite my firm predilection for the symbolic and abstract, I wonder about the neurophysiological mechanics of language processing as much as the next guy.
So I picked up French cognitive neuroscientist Stanislas Dehaene's
Reading in the Brain from the library last week. (Published in 2009, it's probable that it's simultaneously the local library's youngest book and science's most outdated information.) It turns out that scientists love metaphors too! Go figure. I've never thought much about how integral to an understanding of most natural phenomena and structures analogy must be. I could go on about this for quite some time, but let me get into the point of today's post: PANDEMONIUM.
In his first chapter, Dahaene is trying to answer the question "How do we read?" from a cognitive neuro perspective, i.e. very literally. He describes our tiny fovea darting spastically back and forth across a sentence to take in the elaborate detail of written characters, etc. He talks about the brain's multiple parallel pathway structure. Then he starts speaking my language: After a stimulus is recognized as a word, "
mental dictionaries open up, one after the other[...] our mind houses a reference library in several volumes." Sweet! I know what dictionaries are! Maybe I'm not totally lost!
Continuing, he writes, "The number of entries in our mental dictionaries is gigantic. The extent of human lexical knowledge is often grossly underestimated. [...] Any reader easily retrieves a single meaning out of at least 50,000 candidate words, in the space of a few tenths of a second, based on nothing more than a few strokes of light on the retina" (42).
Well, that's just fascinating, I thought as I read. But it turns out accessing lexical information is even MORE exciting than suggested by the stacks-of-dictionaries metaphor (if that's even possible). If our brains are full of dictionaries, the dictionaries themselves are full of fiery little congressional spirits. This is apparently the most accurate description the last 50 years of scientific thinking could come up with -- which I think is pretty cool. What it means is basically that all of the words in the world that we know have representatives crouched in our cortical folds, constantly at the ready to call out their names. Our brains are not like factories pumping out one product at a time or like a librarian leafing through dead pages; they contain assemblies of competitive neurons fighting each other for their chance to fire. They argue about who has the most legitimate claim, who has the best reason for doing his thing. The whole system is chaotic, aggressive, and ultimately very efficient. Every. Word. You. Read. Is. Causing. A. Little. Political. War. In. Your. Brain.
In case my brief description of Dahaene's explanation of the entire nervous system failed to satisfy you, here's some of the direct text. It feels really strange to be reading about reading. Nothing pulls you out of the magic of the moment like a detailed description of the intricate mechanics of that very moment.
"Several models of lexical access manage to imitate the performance of the human reading system under conditions close to those imposed by our nervous system. Almost all of them derive from a set of ideas first defined by Oliver Selfridge in 1959. Selfridge suggested that our lexicon works like a huge assembly of 'daemons,' or a 'pandemonium.' This lively metaphor holds that the mental lexicon can be pictured as an immense semicircle where tens of thousands of daemons compete with each other. Each daemon responds to only one word, and makes this known by yelling whenever the word is called and must be defended. When a letter string appears on the retina, all the daemons examine it simultaneously. Those that think that their word is likely to be present yell loudly. Thus when the word 'scream' appears, the daemon in charge of the response to it begins to shout, but so does its neighbor who codes for the word 'cream.' 'Scream' or 'cream'? After a brief competition, the champion of 'cream' has to yield - it is clear that his adversary has had stronger support from the stimulus string 's-c-r-e-a-m.' At this point the word is recognized and its identity can be passed on to the rest of the system.
Behind the apparent simplicity of this metaphor lie several key ideas on how the nervous system works during reading:
- Massive Parallel Processing: All the daemons work at the same time. There is thus no need to serially examine each of the 50,000 words one by one, a procedure whose duration would be proportional to the size of our mental dictionary. The massive parallelism of the pandemonium thus results in a substantial gain in time.
- Simplicity: Each daemon accomplishes an elementary task by checking to what extent the stimulus letters match its target word. Thus the pandemonium model does not succumb to the pitfall of postulating a homunculus, or the little man who according to folk psychology holds the reins of our brain. (Who controls his brain? Another even tinier homunculus?) In this respect, the pandemonium model can be compared to the philosopher Dan Dennett's motto: 'One discharges fancy homunculi from one's scheme by organizing armies of such idiots to do the work.'
- Competition and Robustness: Daemons fight for the right to represent the correct word. This competition process yields both flexibility and robustness. The pandemonium automatically adapts to the complexity of the task at hand. When there are no other competitors around, even a rare and misspelled word like 'astrqlabe' can be recognized very quickly - the daemon that represents it, even if it initially shouts softly, always ends up beating all the others by a comfortable margin. If, however, the stimulus is a word such as 'lead,' many daemons will activate (those for 'bead,' 'head,' 'read,' 'lean,' 'leaf,' 'lend'...) and there will be a fierce argument before the 'lead' daemon manages to take over.
All of these properties, in simplified form, fit with the main characteristics of our nervous system. Composed of close to one hundred billion (10^11) cells, the human brain is the archetype of a massively parallel system where all neurons compete simultaneously. The connections that link them, called synapses, bring them evidence from the external sensory stimulus. Furthermore, some of these synapses are inhibitory, which means that when the source neuron fires, the firing of other neurons is suppressed. The result has been likened by the Canadian neurophysiologist Donald Hebb to a network of 'cell assemblies,' coalitions of neurons that constantly compete. It is therefore no surprise that Selfridge's pandemonium has been a source of inspiration for many theoretical models of the nervous system, including the first neural network models of reading"(42-43).
Moral of the story: metaphor is everywhere, even underpinning scientific models! I wonder how often this happens? Does anyone know of other examples?