736 T.SHALLICE AND P.W.BURGESS and left to work through them on his own.He scored at the 50th and 80th percentiles respectively,the total of the task times being 2 h.Both Cases 2 and 3 carried out the long form of Raven's Matrices working through the test on their own.This took them 31 and 42 min,respectively-less than the 45 min that Raven et al.(1977)give as the upper limit of the normal range and yet longer than the 2 multiple-subgoal scheduling tests would normally take.They scored at above the 95th and 80th percentiles, respectively,for their ages on the Matrices. For the Six Element task a motivational explanation has in any case little immediate face validity.It requires little effort to switch from one simple task to another.More critically,Cases 2 and 3,the subjects for whom a motivational explanation is most plausible on clinical grounds,made errors analogous to the ones made on the Six Element task on an even simpler timing task.When they had to stop carrying out a perceptuomotor task after 2.5 min with no external timing signal but a stopwatch available,they continued after the time limit.Lack of perseverance-the motivational explanation being considered-should presumably not lead to overshoot!Thus motivational factors do not seem likely to be central to their deficit. Standard retrospective memory processes are also relevant,at least for carrying out the Multiple Errands test where retaining both the layout of the street and the instructions without using the card,would be helpful.The patients differed in this respect.Case I had no difficulties.Case 2 was entirely normal on verbal memory tasks but had visual memory problems which might have affected his performance of the Multiple Errands test;however,they would seem most unlikely to contribute to his difficulty with the Six Element task where there is no spatial memory load and where the instructions are continuously clearly visible on the desk.Case 3 had problems on some memory tasks but these were all in tasks where active organization of the input is required for adequate memory performance and her difficulties appear secondary to her frontal lobe problems (see Signoret and Lhermitte,1976).She was able to carry out memory tasks where the organization was provided for her and in particular she retained text normally,as indicated by her satisfactory performance on Logical Memory where the amount she retained showed no decline over an hour's delay.Like the other patients,she was able to recite correctly the instructions for the tasks in her own words after they had been presented to her.This makes it most unlikely that retrospective memory difficulties on April 8 lie at the root of her problems,at least on the Six Element test given that the problems these 3 patients exhibit on these multiple subgoal tests do not fit well with an ass mption of basic motivational impairments on the one hand or of special purpose cognitive or retrospective memory processes on the other,do they fit with impairments to the bridge processes that lie between the two?To assess this it is necessary to consider the bridge proce sses in more detail These process es lie in the domain of the creation and maintenance of goals and intentions of their realization at a oriate times(prospective memory(Meacham and Leiman.1982:Kvavilashvili.1987))and of plan ing.This is an area where well developed theories are lacking.How er a nosition develor ed hy one of us in collaboration with D.A.Norman is that where a task cannot be adequately carried out through the application of well-learned action or thought routines alone,it requires the use of a Supe rvisory System,which is anteriorly located in the cortex (Nor rman and Shallice,1980,1986: also Shallice,1988).We will provisionally follow this approach736 T. SHALLICE AND P. W. BURGESS and left to work through them on his own. He scored at the 50th and 80th percentiles, respectively, the total of the task times being 2 h. Both Cases 2 and 3 carried out the long form of Raven's Matrices working through the test on their own. This took them 31 and 42 min, respectively—less than the 45 min that Raven et al. (1977) give as the upper limit of the normal range and yet longer than the 2 multiple-subgoal scheduling tests would normally take. They scored at above the 95th and 80th percentiles, respectively, for their ages on the Matrices. For the Six Element task a motivational explanation has in any case little immediate face validity. It requires little effort to switch from one simple task to another. More critically, Cases 2 and 3, the subjects for whom a motivational explanation is most plausible on clinical grounds, made errors analogous to the ones made on the Six Element task on an even simpler timing task. When they had to stop carrying out a perceptuomotor task after 2.5 min with no external timing signal but a stopwatch available, they continued after the time limit. Lack of perseverance—the motivational explanation being considered—should presumably not lead to overshoot! Thus motivational factors do not seem likely to be central to their deficit. Standard retrospective memory processes are also relevant, at least for carrying out the Multiple Errands test where retaining both the layout of the street and the instructions, without using the card, would be helpful. The patients differed in this respect. Case 1 had no difficulties. Case 2 was entirely normal on verbal memory tasks but had visual memory problems which might have affected his performance of the Multiple Errands test; however, they would seem most unlikely to contribute to his difficulty with the Six Element task where there is no spatial memory load and where the instructions are continuously clearly visible on the desk. Case 3 had problems on some memory tasks but these were all in tasks where active organization of the input is required for adequate memory performance and her difficulties appear secondary to her frontal lobe problems {see Signoret and Lhermitte, 1976). She was able to carry out memory tasks where the organization was provided for her and in particular she retained text normally, as indicated by her satisfactory performance on Logical Memory where the amount she retained showed no decline over an hour's delay. Like the other patients, she was able to recite correctly the instructions for the tasks in her own words after they had been presented to her. This makes it most unlikely that retrospective memory difficulties lie at the root of her problems, at least on the Six Element test. Given that the problems these 3 patients exhibit on these multiple subgoal tests do not fit well with an assumption of basic motivational impairments on the one hand or of special purpose cognitive or retrospective memory processes on the other, do they fit with impairments to the bridge processes that lie between the two? To assess this it is necessary to consider the bridge processes in more detail. These processes lie in the domain of the creation and maintenance of goals and intentions, of their realization at appropriate times (prospective memory (Meacham and Leiman, 1982; Kvavilashvili, 1987)) and of planning. This is an area where well developed theories are lacking. However, a position developed by one of us in collaboration with D. A. Norman is that where a task cannot be adequately carried out through the application of well-learned action or thought routines alone, it requires the use of a Supervisory System, which is anteriorly located in the cortex (Norman and Shallice, 1980, 1986; see also Shallice, 1988). We will provisionally follow this approach by guest on April 8, 2016 http://brain.oxfordjournals.org/ Downloaded from