Cognition and computation

Cognition and computation

An article written by Professor Colin Blakemore FRS.

With one hundred thousand million (1011 ) neurons and one thousand million million (1015)  connections between them, the human brain is often dubbed the most complex object in the known universe.

Revolutionary advances in genomics and functional genetics over the past few decades have impacted on our understanding of the brain, just as they have in all areas of biology.  If we take into account not only the shape and size of nerve cells and the cocktail of molecules that they contain, but also the specific targets to which they send all the branches of their axons (projections of the nerve cells), and the particular array of axons that contacts them, the number of differentiated categories of neurons must surely be in the thousands. We are now starting to unravel how roughly 25,000 genes control the development, differentiation, function and adaptability of human brain cells. 

Professor Nicola Clayton FRS on cognition and what makes the human brain special.(4 mins, requires Flash Player).

 In addition to, and presumably because of, its extraordinary size, the human brain is remarkable for its cognitive capacity. The archaeological record shows very gradual increase in brain volume over the 4 million years of hominid evolution, until a more rapid enlargement of the brain with the emergence of Homo sapiens in Africa about 200,000 years ago. The existence of genes that regulate the rate of production and death of cells during development makes it easy to imagine how the brain suddenly exploded in size. However, understanding how that big, metabolically-costly brain was retained by natural selection is more difficult and is a fundamental question for scientists today.

    Audio
    Professor Patrick Haggard on whether intentions can be studied scientifically
    (3 mins).
  • Listen here.

Much has been made of the distinctive cognitive capacities of human beings.  However, apart from the use of fully syntactical language, parallels to all the other ‘unique’ human abilities have been found in the behavioural skills of apparently much simpler creatures (with much smaller brains). Scrub jays, closely related to crows, which specialise in hiding thousands of items of food, even seem to be capable of ‘mental time travel’ – reflecting on the past and planning for the future. On the other hand, humans, partly through the sharing of intelligence, have proved extraordinarily versatile in their ability to adapt to a vast range of environments, from deep in the ocean to the surface of the moon.

Measuring brain activity (credit: NASA).

There is still a huge gulf between defining cognitive skills and explaining them in terms of brain function. One approach to the latter is what some call ‘connectomics’ – defining exhaustively the connectional architecture and functional properties of a small element of the brain (for instance a ‘column’ of neurons in the cerebral cortex of the rat) and incorporating the data into a computational model of that region.  This is complicated by the fact that connections between nerve cells can be modified rapidly as a result of the nerve impulses passing through them, a process that is thought to underlie memories of place and personal experience. 

    Audio
    Professor Richard Morris FRS on the mechanisms of memory (3 mins).
  • Listen here.

Another powerful approach to understanding cognitive function is to record electrical impulses from individual neurons in the cortex of monkeys while they conduct tasks. Such research has already taught us a great deal about the processing of visual information. Now, recordings from neurons in the parietal lobe are revealing changes of impulse firing that correlate with the animal’s decisions in quite complex choice tasks.

The flood of evidence from remarkable techniques for visualising activity in the living human brain (especially functional Magnetic Resonance Imaging) shows that the human cerebral cortex, like that of monkeys, consists of a patchwork of functional areas. Research work has indicated correspondence between discrete regions and high-level cognitive functions, as well as sensory processes, but this interpretation is by no means uncontroversial.

Brain of Gauss, the mathematician and astronomer, from The Brain as an organ of the mind, by H. Charlton Bastian FRS, 1880. From the Royal Society archive.

The analytical techniques of brain imaging are also illuminating the issue of emotion in cognitive processes, showing the effects of emotional states on activity and interconnectivity even in basic sensory areas of the cortex. One small region, the amygdala, which is exceptionally widely connected to the rest of the forebrain, appears to lie at the heart of the translation of sensory experience into emotional responses and of the consequent modulation of brain function.

Finally there remains the question of consciousness, once the realm of philosophy alone. Francis Crick’s 1995 book, The astonishing hypothesis, which encouraged research on the ‘neural correlate of consciousness’, helped to bring subjectivity into the fold of neuroscience. Inference from the behavioural consequences of brain damage and from the relationship between the timing of subjective decision-making and brain activity preceding action has led to the idea that the brain contains an “executive” mechanism, associated with the frontal lobe of the dominant (usually left) hemisphere, from which some aspects of the sense of self spring. But deep mysteries remain. What is the relationship between subjective feelings of self, thoughts and intentions, and the vivid conscious experiences of sensory perception, which some philosophers call ‘phenomenal’ consciousness? Why is so much of what the brain does inaccessible to consciousness? And if conscious states are simply generated by neural activity, what do they do, and why have they evolved?

Banner image: Thalamocortical handshake (credit: Zoltán Molnár and Colin Blakemore).

Discussion meeting

This article is based on the discussion meeting 'Cognition, computation and consciousness' which was held on 20-21 September 2010.

From the Royal Society's archives

The complexities of brain science - and the link between brain organisation and behaviour - were addressed by one of the earliest Fellows of the Royal Society, the Oxford physician Thomas Willis. His great work of 1664, The anatomy of the brain with a description of the nerves and their function, is one of the first scientific accounts of the human brain – “the chief seat of the Rational Soul” as Willis termed it – and the rest of the nervous system. The book, superbly illustrated by Christopher Wren, was a seminal step in what we now call neuroscience.

Willis moved in intellectually brilliant circles in 17th century Oxford. In addition to Wren, he befriended John Locke and Robert Hooke (who briefly worked as his laboratory assistant). Both of these men would go on to theorise on brain functions. In Hooke’s case it was on the action of memory, while Locke formulated a more general philosophy of mind in An essay onhuman understanding1.

Alan Turing FRS famously raised the question of how one might identify a “thinking” machine in 1950. In the journal Mind2 he proposed what is now known as the Turing test, which uses questions answered by a machine to determine how closely a machine might emulate a human under blind interrogation conditions.

Turing’s thought experiment has proved hugely influential. It clearly inspired analogous tests of consciousness (and humanity) in fictional works such as the movies Blade Runner (1982) and I, Robot (2004). However, the term “cognitive science” was coined far more recently (in the 1970s) by Christopher Longuet-Higgins FRS. The coming together of neuroscience and cognitive science – the biology of the brain and the psychology of the mind – leads us into a new era of discovery.


[1] J. Locke (1721) An essay concerning human understanding, in four books written by John LOCKE, printed by A. Churchill & A. Manship; sold by W. Taylor, London, 1721 (first published 1690)

[2] A. Turing (1950) Computing machinery and intelligence, Mind, Vol. 59, No. 236. pp. 433-460, October 1950