[paul.fr]

Has it been measured?

[dentstudent]

Guess - Isn't it as fast as electricity can pass in the nervous system, so about 250 mph?

[another_someone]

What does the question mean?

Even in systems that we understand better, like man made computer systems, although we can produce metrics (such as mips and flops), but these are wholly inadequate to measure real world performance.

The human brain is a massively parallel processing unit with lots of specialised processing capabilities, and lots of predictive (speculative) processing.  The problem with measuring processing capabilities of a processor that is doing predictive processing is that if it gets the prediction right, then it can be superfast, but if it gets it wrong, it may seem very slow.

Then again, what do you mean by 'thought'?  Are you referring only to concious problem solving, or are you referring the time to respond to any stimuli?

[another_someone]

Guess - Isn't it as fast as electricity can pass in the nervous system, so about 250 mph?

It is not enough for the signal to reach its destination, but you need a long enough sample of signal to sensibly demodulate the signal (my understanding is that much of the electrical signals in the nervous system rely on a kind of pulse width modulation, and so you would need to receive a certain number of pulses (theoretical minimum of two, but practically probably more than that) to properly demodulate the signal).

[Bored chemist]

I don't know the answer, but I know it's a lot slower late on a Friday night.

[chris]

People have made measurements of neural activity both at the level of whole populations of nerve cells (this is the EEG - electroencephalogram, which you measure by sticking electrodes to the scalp), and individual nerve cells by single unit recording (intracerebral electrode) techniques.

In the latter case the basic idea is to record from a nerve cell that is that target of a neurological process - for instance vision - and time how long it takes for the cell to alter its activity following the presentation of an appropriate stimulus.

If you do this for the visual system you find that the retina takes about 50ms to process what you see, and the brain doesn't become conscious of a visual stimulus for at least 150ms. (This is far too slow for tennis players and international cricket batsmen to be playing under "conscious" control" - ie their reactions are all automatic and based on previous motor learning).

Taking the alternative approach, experiments were done in the past where subjects were asked to make various movements in response to the presentation of various signals whilst their EEGs were recorded. These experiments showed that the EEG altered activity up to one third of a second before the movement they made was executed.

In other words what we call the present actually happened nearly half a second ago in terms of the reactions we make and our stream of consciousness.

Chris

[Heronumber0]

Is this evidence of a quantum effect Chris?

[chris]

My answer has nothing to do with quantum effects, merely recorded activities of nerve cells in relation to the time at which a stimulus was presented.

Chris

[Cut Chemist]

I guess the speed of "thought" is rather slow ... if it takes half a second to respond to an unknown stimulus!!  Far slower than computers!!

So what is the speed of impulse or instinct??  And how does that differ from "thought," really?? 

(Impulse or instinct is technically a "programmed response," more similar to computer programing.)

[chris]

It's related to which part of the brain co-ordinates and exectutes the response. A reflex - such as a knee jerk when you tap the knee with a tendon hammer - takes a matter of less than 50 milliseconds, because it's hard-wired. Many other "spinal" reflexes are the same, such as jerking your hand away from a hot surface - these movements are hard-wired so they occur very rapidly, because they can often save your life, keep you upright, or both.

But reactions involving the brain tend to be slower because more processing is involved. It's a bit like the extra delay you notice when listening to digital radio compared with an analogue signal. Conscious thought is the slowest - it takes a good 300ms before information is presented to your consciousness. Sub-conscious thought and reaction, however, is much faster - probably twice the speed.

Tennis players at Wimbledon are certainly not consciously reacting to the ball coming over the net - it all happens too quickly. Instead their visual system is sending the information about the trajectory of the ball through a paralel processing pathway involving the brain's cerebellum. This is the "mini-brain" that sits at the back of the head and controls movement. It's involved in the smooth execution of learned movements. This pathway also involves very few cell to cell links (synapses), so it's very swift.

How it works is that when a person makes a conscious movement in response to a stimulus the pattern of moves they make is imprinted in the cerebellum. Next time the same stimulus crops up the cerebellum can pre-programme most of the response without conscious input. As it does so it compares what does happen with what should have happened; if the two don't match then the cerebellum tweaks the strength of some if its connections to tune up the process and improve the accuracy of the movement next time. This is why we all speed up and improve with practice.


Chris

[another_someone]

Tennis players at Wimbledon are certainly not consciously reacting to the ball coming over the net - it all happens too quickly. Instead their visual system is sending the information about the trajectory of the ball through a paralel processing pathway involving the brain's cerebellum. This is the "mini-brain" that sits at the back of the head and controls movement. It's involved in the smooth execution of learned movements. This pathway also involves very few cell to cell links (synapses), so it's very swift.

I would suggest that tennis players use even more sophisticated predictive processing.

When a tennis player hits a ball, (s)he is already trying to predict the likely returns to his/her play, and is looking at the body language of his/her opponent to anticipate which of the likely returns becomes more likely, and all of this before the ball even begins its return flight.  Thus I would imagine that the tennis player has probably already started the firing sequence for a small number possible movements (s)he needs to make to intercept the ball, and then only needs to suppress those actions that turn out not to match predictions once the ball starts its return flight.

[chris]

Yes George, that's right. There is certainly a priming effect - a bit like a penalty taker dummying in football, but there is also an extremely fast subconscious circuit from the visual system via the cerebellum. Interestingly it's thought that this same circuit is recruited in patients with "blind sight".

These are people who despite being "cortically" blind - their main visual cortex is damaged and so they have no conscious awareness of vision - if you ask them to walk in a straight line and then place an obstacle in their path - which they cannot "see" they nonetheless avoid it.

Similarly you can array an assortment of objects in front of these people and ask them to pick up the banana or the keys that are in amongst the mixture. They don't know how they do it, but they can.

Chris

[johnbrandy]

I cannot offer you a scientific answer to this question. I realize this is a "scientific" forum. But I suspect the question requires additional refinement. It is, as presented, open-ended. In practical terms, the speed of thought is measured in terms of human response time. Perhaps the question you are proposing might be, what is the relationship between the speed of neuronal firing assumed to be related to human thought and human response time? Thank you for allowing me to participate.

[paul.fr]

Hi John, your participation is most welcome. The question was deliberately openended to get a wider range of answers from you clever people.

[JimBob]

John, welcome to the forum. The people don't think I am a scientist, either - I am a disreputable geologist.  But I do believe you have framed the question in a much more precise manner than others. It is a measure - if possible - of the physical processes that arise and are perceived of as thought.

[coglanglab]

Chris has made some wonderful replies to this open-ended question. I would agree with him and some others that have been pointing out that there are different kinds of thoughts that probably operate at different speeds. The facts about neuronal signal transduction are interesting, but of course the relationship between thought and the firing of individual neurons isn't understood very well yet.

One interesting piece of work that may interest Paul was done by Molly Potter at MIT many, many years ago. She found that if you show somebody a picture of a scene for only about 100ms, they are able to remember and describe a great deal of the information presented in that time. This is interesting primarily because 100ms is a very short time period for neurons. A signal can only travel down a few neurons in that time. This comes back to the issue of trying to tie the speed of neuronal signals to the speed of thought. (If you don't think scene recognition is thought, consider that a great deal of interpretation must be done to go from the visual imprint of some line drawings to the representation of Don Quixote on a hill.)

--
Please take a few minutes for science by participating in experiments at http://coglanglab.org [nofollow]

[another_someone]

One interesting piece of work that may interest Paul was done by Molly Potter at MIT many, many years ago. She found that if you show somebody a picture of a scene for only about 100ms, they are able to remember and describe a great deal of the information presented in that time. This is interesting primarily because 100ms is a very short time period for neurons. A signal can only travel down a few neurons in that time. This comes back to the issue of trying to tie the speed of neuronal signals to the speed of thought. (If you don't think scene recognition is thought, consider that a great deal of interpretation must be done to go from the visual imprint of some line drawings to the representation of Don Quixote on a hill.)

I do not doubt that scene recognition requires substantial processing (what is 'thought', and whether it applies only to concious mental processes, or subconscious ones as well is another matter); but what it sounds to me like is what in computer terms is referred to as pipelining; or maybe even more crudely, the image is quickly latched, but then takes longer to fully process.

[lyner]

If you look at the second hand of a clock which  moves in jerks, once a second, you sometimes have to wait for what seems like more than one second. I heard that this is because of your delay in 'consciousness' of an event. If you just missed a 'tick' and then have to wait before you are aware of the next one, it seems more like 2 seconds.
The brain normally  manages to fool itself that there's no delay in order  to avoid feeling bad about its inadequacy and getting confused.
This delay doesn't often matter because we spend all our lives taking it into account and predicting. Then, of course, all our reflex actions are much quicker that that. Your mind just tells itself that you 'decided to do it' and that it was in charge all the time. (A bit like the boss, when a minion makes a good decision).  If we didn't do something like that, we would be confused with each bit of new information we had to deal with - we  would keep worrying  just how recent the info was and wonder how we reacted to it.