by James Reason
Naturalistic studies of everyday slips and lapses (Jastrow, 1905; Reason, 1977, 1979, 1984a; Norman, 1981) clearly indicate the dependence of slips and lapses upon failures of attentional checking. For the most part, these involve inattention, omitting to make a necessary check. But a significant number of action slips are also due to overattention, making an attentional check at an inappropriate point in an automatised action sequence. Both could be termed control mode failures in that errors arise from being in the wrong control mode with respect to the current demands of the task. That is, the higher levels of the cognitive system are running open-loop (in relation to the moment-to-moment control of the actions) when they should have been closed-loop, and conversely. A more detailed consideration of these mechanisms will be given in Section 4.
Figure 3.1. Outlining the dynamics of the generic error-modelling system (GEMS).
3.2. Problem-solving failures
As indicated earlier, a problem can be defined as a situation that requires a revision of the currently instantiated programme of action. The schematic mode of control can only operate satisfactorily when the current state of the world conforms to the regularities of the past. The departures from routine demanded by these situations can range from relatively minor contingencies, swiftly dealt with by preestablished corrective procedures, to entirely novel circumstances, requiring new plans and strategies to be derived from first principles.
The problem solving elements of GEMS are based upon a recurrent theme in the psychological literature, namely that “humans, if given a choice, would prefer to act as context-specific pattern recognizers rather than attempting to calculate or optimize” (Rouse, 1981).
The key feature of GEMS is the assertion that, when confronted with a problem, human beings are strongly biased to search for and find a prepackaged solution at the RB level before resorting to the far more effortful KB level, even where the latter is demanded at the outset. In relation to Figure 3.1, this means that they are inclined to exit from the decision box (Is the pattern familiar?) along the affirmative route. They do this by matching aspects of the local state information (the problem configuration) to the situational elements of stored problem-handling rules of the kind: if (situation) then (system state), if (system state) then (remedial action).
Only when people become aware that successive cycling around this rule-based route is failing to offer a satisfactory solution will the move down to the KB level take place. And even here problem solvers are likely, at least initially, to be using ‘workspace’ processing to search for cues that remind them of previously successful rules, which could then be adapted to the present situation. As Schank (1982) has pointed out, such analogical ‘remindings’ stem from an awareness that the current situation shares a common ‘deep structure’ with something already in memory. He gives as an example the thematic similarities between the ‘scripts’ of Romeo and Juliet and West Side Story. At a purely surface level, the differences between the two are enormous: one is set in Renaissance Italy the other is a musical about gang warfare in modern-day New York. Yet both—intentionally—share the same ‘deep structure’. To note the similarity between the two, “one must be not only processing the normal complement of scripts and goals. One must also be, in a sense, summarising the overall plot to oneself, because that is where the match occurs” (Schank, 1982, p. 34). Schank (1982, p. 25) explains this phenomenon as follows: “We are reminded of a particular experience because the structures we are using to process the new experience are the same structures we are using to organize memory. We cannot help but pass through the old memories while processing a new input.”
Also relevant to this discussion is Shepard’s (1964, p. 263) summary of a very large literature on the limitations displayed by human beings when attempting to combine factors in problem solving and decision making:
At the level of the perceptual analysis of raw sensory inputs, man evinces a remarkable ability to integrate the responses of a vast number of receptive elements according to exceedingly complex nonlinear rules. Yet once the profusion and welter of this raw input has been thus reduced to a set of usefully invariant conceptual objects, properties, and attributes, there is little evidence that they can be juggled and recombined with anything like this facility. On the contrary, the contention that they can belies the obvious disparity between the effortless speed and surety of most perceptual decisions and the painful hesitation and doubt characteristic of these subsequent higher level decisions.
In short, human beings are furious pattern matchers. They are strongly disposed to exploit the parallel and automatic operations of specialized, pre-established processing units: schemata (Bartlett, 1932), frames (Minsky, 1975), scripts (Schank & Abelson, 1977) and memory organizing packets, or MOPs (Schank, 1982). These knowledge structures are capable of simplifying the problem configuration by filling in the gaps left by missing or incomprehensible data on the basis of ‘default values’. This, in turn, is derived from the memory system’s remarkable ability to keep track of the ‘frequency of encounter’ of these ‘slot’ values without conscious effort (see Chapter 5 and Hasher & Zacks, 1984). Higher-level manipulations, however, create ‘cognitive strain’ (Bruner et al, 1956) and require the slow, laborious, resource-limited involvement of the attentional control mode.
3.3. What determines switching between levels?
3.3.1. Between the SB and RB levels
The SB level of GEMS relates to the performance of highly routinised activities in familiar circumstances. The RB level is engaged when an attentional check upon progress detects a deviation from the planned-for conditions. A primary feature of GEMS is that RB attempts at problem solution will always be tried first. If the deviation is minor and appropriate corrective rules are readily found, this phase will be terminated by a rapid return to the SB level. With more difficult problems, the rule-based cycle (scanning local signs and symptoms, rule implementation and evaluating the outcome) may be repeated several times.
3.3.2. Between the RB and KB levels
According to the simple logic of GEMS, the switch from the RB to the KB level occurs when the problem solver realizes that none of his or her repertoire of rule-based solutions is adequate to cope with the problem. In reality, however, the factors determining this transition are less clear-cut.
In the first place, affective factors are likely to play an important role. Duncan (personal communication) has suggested that the decision to resort to the more effortful KB consideration of structure-function relations will depend upon the complex interaction between subjective uncertainty and concern. Both will increase rapidly as successive rule-based solutions are recognized as being inadequate.
Second, even when this point has been reached, it is likely that the largely unconscious search for analogous problem-solving ‘packets’ will proceed in parallel with conscious ‘topographic’ reasoning. The discovery of such an analogy usually brings with it a set of largely preformed remedial possibilities (e.g., if [it’s like situation X] then [I should try action Y]). A well-understood analogy is likely to entail a set of remedial rules that will switch the focus of activity back to the RB level so long as that particular analogy continues to be entertained. This cycling between the KB and RB levels can be repeated several times as various similarities are explored.
3.3.3. Between the KB and the SB levels
Activity at the KB level can be stopped by finding an adequate (or apparently so) problem solution. This will constitute a new plan of action requiring the execution of a fresh set of SB routines. It is unlikely that this recovery plan will have the integrated or precompiled character of a familiar action sequence. Rather, it will probably comprise routines borrowed from a variety of activities and will require considerable closed-loop control from the conscious workspace to guide these routines. In other words, there will be rapid switching to and from the SB and KB levels until performance is back on some familiar track.
This shift from the KB to the SB level is
shown in Figure 3.1 by the affirmative route from the ‘Is the problem solved?’ decision point. There will be powerful cognitive and affective forces conspiring to encourage the problem solver to accept inadequate or incomplete solutions as being satisfactory at this point.
A plan of action represents a revised theory of the world, and confirmation biases will lead to its continued retention, even in the face of contradictory evidence. Such ‘secondary errors’ may also be rendered more likely by the reduction in anxiety that accompanies the discovery of an apparent solution.
But once this error is eventually detected (by an OK? check at the SB level), the problem solver will again switch back into the RB mode. By this time, the local state indicators will have changed as a consequence of the previous activity, allowing the possibility of new RB solutions being applied and so on. In this way, the focus of control will shift continuously between the three performance levels.
4. Failure modes at the skill-based level
As indicated earlier, most of the failure modes observed at the SB level (the top level of Fig. 3.1) can be grouped under two headings: inattention, omitting to perform the necessary attentional monitoring at critical (OK?) nodes, particularly when the current intention is to deviate from common practice; and overattention, making an attentional check at an inappropriate moment during a routine action sequence. Each of these two control-mode failures can take different forms, as discussed below. (To give an overall picture of where the remainder of this chapter is going, the main headings for the failure modes at each of the three levels of performance are listed in Table 3.3.)
4.1. Inattention (omitted checks)
4.1.1. Double-capture slips
Perhaps the commonest consequences of an omitted check (see Reason, 1979; Reason & Mycielska, 1982; Norman, 1981) are various forms of ‘double-capture’ slips. These are so named because they involve two distinct, though causally related, kinds of capture. First, the greater part of the limited attentional resource is claimed either by some internal preoccupation or by some external distractor at a time when a higher-order intervention (bringing the workspace into the control loop momentarily) is needed to set the action along the currently intended pathway. As a result, the control of action is usurped by the strongest schema leading onwards from that particular point in the sequence.
Such slips are lawful enough to permit reasonably firm predictions regarding when they will occur and what form they will take. The necessary conditions for their occurrence appear to be (a) the performance of some well-practised activity in familiar surroundings, (b) an intention to depart from custom, (c) a departure point beyond which the ‘strengths’ of the associated action schemata are markedly different, and (d) failure to make an appropriate attentional check. The outcome, generally, is a strong habit intrusion, that is, the unintended activation of the strongest (i.e., the most contextually frequent) action schema beyond the choice point.
Some actual examples, obtained in various natural history diary studies (see Reason & Mycielska, 1982; Reason 1984c), are given below.
Table 3.3. Summarising the main headings for the failure modes at each of the three performance levels.
Skill-based performance
Inattention
Overattention
Double-capture slips
Omissions
Omissions following interruptions
Repetitions
Reduced intentionality
Reversals
Perceptual confusions
Interference errors
Rule-based performance
Misapplication of good rules
Application of bad rules
First exceptions
Encoding deficiencies
Countersigns and nonsigns
Action deficiencies
Informational overload
Wrong rules
Rule strength
Inelegant rules
General rules
Inadvisable rules
Redundancy
Rigidity
Knowledge-based performance
Selectivity
Workspace limitations
Out of sight out of mind
Confirmation bias
Overconfidence
Biased reviewing
Illusory correlation
Halo effects
Problems with causality
Problems with complexity
Problems with delayed feed-back
Insufficient consideration of processes in time
Difficulties with exponential developments
Thinking in causal series not causal nets
Thematic vagabonding
Encysting
(a) “I had decided to cut down my sugar consumption and wanted to have my cornflakes without it. But the next morning, however, I sprinkled sugar on my cereal, just as I always do.”
(b) “We now have two fridges in our kitchen, and yesterday we moved our food from one to the other. This morning, I repeatedly opened the fridge that we used to have our food in.”
(c) “On starting a letter to a friend, I headed the paper with my previous home address instead of my new one.”
(d) “I intended to stop on the way to work to buy some shoes, but ‘woke up’ to find that I had driven right past.”
(e) “I brought the milk in to make myself a cup of tea. I had put the cup and saucer out previously. But instead of putting the milk into the cup, I put the bottle straight into the fridge.”
(f) “I meant to get my car out, but as I passed through the back porch on the way to the garage, I stopped to put on my Wellington boots and gardening jacket as if to work in the garden.”
(g) “I have two mirrors on my dressing table. One I use for making up and brushing my hair, the other for inserting and removing my contact lenses. On this occasion, I intended to brush my hair, but sat down in front of the wrong mirror, and removed my contact lenses instead.”
(h) “I went to my bedroom to change into something more comfortable for the evening, and the next thing I knew I was getting into my pyjama trousers, as if to go to bed.”
(i) “I meant to take off only my shoes, but took my socks off as well.”
(j) “I was making shortbread and decided to double the amounts shown in the recipe. I doubled the first ingredient—butter—but then failed to double anything else.”
(k) “I decided to make pancakes for tea. Then I remembered we didn’t have any lemons, so I decided not to bother. Five minutes later, I started getting together the ingredients for pancakes having completely forgotten my change of mind.”
(1) “I was putting cutlery away in the drawer when my wife asked me to leave it out, as she wanted to use it. I heard her, but continued to put the cutlery away.”
Examples (a), (b) and (c) are clearly similar. Each involved a change of routine that led to an old-habit intrusion because of a checking omission. Examples (d) and (e) also show clear signs of strong-habit capture during a moment of inattention: but in these cases, the inattention led to a strong-habit exclusion, rather than to an intrusion. Examples (f) and (g) are instances of branching errors, in which an initial common action sequence leads to different outcomes, and the attentional check at the choice-point is omitted. Examples (h) and (i) are conceptually very similar except that they involved overshooting a stop rule that was not regularly imposed. Slips (j), (k) and (1) share a failure to attend to the need for change at a critical moment. In (j) and (k), this resulted in the unwanted reversion to an earlier plan; whereas in (1), it led to the continuation of an habitual sequence of actions. This last slip is interesting because it reveals something about the actor’s attentional state. He clearly heard and remembered his wife’s request, but failed to act upon it. This suggests that the wife’s request was noted and recorded by the ‘fringes’ of consciousness, but was not acted upon because the man’s focal attention was directed elsewhere.
4.1.2. Omissions associated with interruptions
In some instances, the failure to make an attentional check is compounded by some external event. For example:
(a) “I picked up my coat to go out when the phone rang. I answered it and then went out of the front door without my coat.”
(b) “I walked to my bookcase to find the dictionary. In the process of talking it off the shelf, other books fell onto the floor. I put them back and returned to my desk without the dictionary.”