210 Sir Frederic Bartlett 



slowing-up begin? Once begun does it continue and perhaps 

 increase without recovery? Particularly are there certain 

 phases of complex skilled activity which, more than the others, 

 exhibit this slowing-up? 



It appears that there is some slowing-up of skilled actions 

 normally in the late twenties. This is followed by considerable 

 or complete recovery, but it reappears m the late thirties or 

 middle forties. There is again recovery, until the early or 

 middle fifties, and then the slowing up seems generally to 

 continue and perhaps to increase. But there is nothing very 

 remarkable or dramatic about it in the case of the fit person, 

 and if it is adequately countered probably it need have no 

 important practical significance. I am speaking of course of 

 skilled performance which must be done rapidly, but does not 

 require any great expenditure of muscular effort. If the latter 

 is required, and still more if work has to be done at relatively 

 rapid rates, and with accompanying overall bodily mobility, 

 the final slowing-up appears earlier, usually in the early forties, 

 and with little or no subsequent recovery. 



Very much more interesting is the question of whether 

 there are certain stages in the performance of skilled move- 

 ments of all kinds which are especially susceptible to change. 

 It rapidly became clear that overall time measures of per- 

 formance may be extremely misleading (Welford, 1951). 

 For example, in one of the early experiments a small object 

 was set into apparently random movement over a surface 

 marked off in small squares. When it came to rest its position 

 had to be accurately identified by manipulation and if this 

 was done successfully the object was set into further random 

 movement; and so the experiment proceeded. If the overall 

 time was recorded for a series of correct locations, it could, 

 as indeed anyone would expect, be exactly the same with 

 great and significant differences in the internal pattern of 

 manipulative movement. It was from this experiment that 

 we got the first suggestion that the most unstable element in 

 any series of accurate adjustments to changing stimulation 

 is not the speed of movement, and not the reaction time (if 



