The Speed of Thought.
In 1796 Neville Maskelyne, then Astronomer Royal of Great Britain, sacked his assistant David Kinnebrook because his timing of the moments when stars crossed fixed reference points difered from his own. Maskelyne is now remembered for his attempts to undermine the achievement of John Harrison whose innovative clocks were the first practical solution to the problem of determining longitude. Maskelyne also had a vested interest in solving the longitude problem and his behaviour to Harrison suggests that he was self-centred and autocratic as well as grasping. Nevertheless he was also a capable observer and a dedicated scientist and in touch with the wider science of his time. His harsh treatment of Kinnebrook could be justified in terms of the little that was then known about human brains and minds. Of course, over thousands of years, humans had recognised that some of us, particularly the young, are much faster than others, especially the old. But, until we invented ways to precisely measure very tiny time differences we could not realise that, no matter how hard we try or how long we practice, these differences are also quite remarkably consistent. The story of how this was worked out neatly shows that Science progresses, not just by sudden brilliant insights but, more often, by gradual realisations that cherished assumptions must be wrong.
Twenty years later, another Astronomer Royal at the Prussian Court, the illustrious Friedrich Wilhelm Bessel, made the next step by noting that even reliable and experienced astronomers often widely differed when timing the same events. Archived records showed that Kinnebrook’s observations did indeed differ from Maskelyne’s but the interesting thing was that he was consistently slower by 8 tenths of a second over scores of different observations. Bessel saw that the point is not to identify and depend on ideal “gold standard”, observers but to recognise that large individual differences will always occur but, if they are consistent, they can be measured and taken into account. Bessel compared records of many astronomers and found that he could estimate for each a personal correction factor that could be added or subtracted to resolve inevitable differences from the others. Too late for poor Kinnebrook. He never regained the prestigious job at the Royal Observatory that he had lost by having a Reaction Time 800 milliseconds slower than his boss. He made a new living as a schoolmaster.
We had realised that some of us need longer than others to recognise things and do something about them and also that the older we get the slower we become. We could not know how consistently stable these differences are until ways to measure differences of a few tenths of a second were invented in the early 19th century. It was also not until then that we learned that the speed with which we can make decisions depends on how fast our nerve cells can initiate impulses and pass these on to each other. In 1849, once again at Konigsberg, another very eminent Professor, Herman von Helmholtz, arranged the recording needle of a galvanometer to reflect a light beam across a room so that its tiny movements were amplified sufficiently to be observed and recorded. This allowed him to measure the time between the start of an electrical impulse in the sciatic nerve of a frog and its later arrival at its calf-muscle. He calculated the speed of nerve transmission as between 24.6 and 38.4 meters a second. He estimated the speed of human nerve conduction by measuring differences in the times that his colleagues took to respond by moving a finger to touches on their feet and cheeks. He found that reaction times to foot-touches were measurably longer than to cheek touches, argued that this difference must occur because nerve impulses from the foot have further to travel to reach the brain and by dividing these differences by the lengths of his colleagues he estimated speeds that agreed quite well with his earlier, direct, frog observations. Also, surprisingly, with modern observations made with very precise equipment. I have many times made a fool of myself while trying to replicate Helmholtz’s gentle experiment with undergraduate students in laboratory classes. We all always tried our best but produced only impossible results. Helmholtz was, of course, a most gifted experimenter and a very great scientist.
As people grow old their reaction times become much slower. Is this because their nerve transmission rates slow ?
In the mid 1950’s Jim Birren, an extraordinarily warm and generous personal mentor, made this question a main theme of his distinguished research career. He noticed that age differences in nerve transmission rates are too small to account for differences of 50 thousandths of a second (msec) or more between the average Reaction Times of old and young adults . More experiments showed that these differences do not just reflect lags in muscles and joints but must, rather, be mainly due to slowing of the “Central decision time” that the brain needs to recognise that something has happened and to organise a response to it. Many experiments have also shown that people’s reaction times do not become slower as they age just because they become more cautious. Age-differences in central decision time are real and stable: they increase from 50 milliseconds in very easy tasks to 500 msecs or much more in difficult tasks and they do not disappear with practice.
It may seem finicky to make an issue of differences of a few thousandths of a second but this does lead to a quite interesting thought. When a flash of light falls on the retina of my eye it takes at least 40 msec for the my receptor cells to generate nerve impulses that travel to the particular parts of my brain that can decide that this is a signal to which I have agreed to respond and not a meaningless random background event. It takes at least another 50 milliseconds for my brain to send a message to the muscles in my forearm and hand to begin and complete the necessary twitches of meat and joints to move a finger to press a key. When I was 26 my simple reaction time was about 150 milliseconds. So the time that my brain then took to register that a flash had occurred and to begin to organise a response to was 150 – (40 + 50) or just 60 milliseconds. But the least time in which one of my nerve cells could activate another was 1 millisecond. This must mean that the circuitry in my brain that managed the whole business of recognising signals as relevant or irrelevant, and programmed and triggered my responses to them could not include nerve pathways of more than 60 nerve-cells connected end-to-end in series. At age 79.5 my reaction times are now 40 milliseconds slower. This fits quite well with a jokey estimate made by David Madden, working in Virginia, that as we age our reaction times to simple events such as buzzes or flashes of light increase by about 1 msec a year. I try not to be depressed and continue to be puzzled that age-lagging of my individual neurones cannot account for such a big difference. I am still trying my best to work out what else must be going on.