Control and Cognition 145
Model Constituents
The Contextual Control Model (COCOM) is a minimal model in the sense
that it focuses on the functions deemed necessary to explain orderly
performance and is intended to be applicable to a range of JCSs from
individuals to organisations. Since such systems may be very different in
their substance, the descriptions must necessarily refrain from going into
details about the possible underlying structures, except where such an
association is obvious. As a minimal model, the COCOM has only three main
constituents, called competence, control, and constructs.
• Competence represents the set of possible actions or responses that a JCS
can apply to a situation to meet recognised needs and demands. The
extent of this set depends on the level of detail or the granularity of the
analysis, and it is not necessarily denumerable. Furthermore, in terms of
modelling, the JCS cannot do something that either is not available as a
possible action or that cannot be constructed or aggregated from the
available possible actions.
• Control characterises the orderliness of performance and the way in
which competence is applied. The COCOM deliberately simplifies the
description of control to a set of four control modes representing
characteristic regions on a continuum going from no control at all to
completely deterministic performance. One issue of control has to do with
the conditions under which it is either lost or regained, in modelling terms
described as changes from one control mode to another. A second issue
has to do with the characteristic performance in a given mode – i.e., what
determines how actions are chosen and carried out. Both issues are
addressed by the COCOM, and define the requirements to the internal
functions of the model (Hollnagel, 2000).
• Constructs refer to the description of the situation used by the system to
evaluate events and select actions. The term is intended to emphasise that
this description is a construction or reconstruction of salient aspects of the
situation, and that it is usually temporary. Constructs are similar to the
schemata of Neisser (1976) in the sense that they are the basis for
selecting actions and interpreting information.
An essential part of control is planning what to do in the short-term,
within the system's time horizon. This planning is influenced by the context,
by knowledge or experience of dependencies between actions, and by
expectations about how the situation is going to develop – in particular about
which resources are and will be available to the person. The resulting plan
describes a sequence of the possible actions, which can either be constructed
or predefined. Frequently occurring plans or patterns therefore reflect the
146 Joint Cognitive Systems
relative constancy (regularity) of the environment rather than the constraints
of the performance model.
If we refer to the basic principles of the cyclical model (cf., Figure 7.1
above), two essential dependencies can be found. One concerns the revision
or development of the construct, i.e., maintaining a correct understanding of
the situation. Using the letters E, C, and A to represent events, construct, and
actions, respectively, construct maintenance can be described as:
1−
←
ttt
CEC
which means that the construct at time t, the current understanding, is
determined by the event given the construct at time t-1, i.e., the previous
understanding. This relation provides a link from the past to the present and
represents the reactive aspects of the model. The time it takes to do this is
denoted as T
E
or the time needed for feedback evaluation.
The other dependency concerns the selection of the next action – although
the term selection does not mean that this in any way is an explicit decision.
The selection can be described as follows:
1+
←
ttt
ECA
which means that the action at time t is determined by the current construct
given the expected outcome of the action (E
t+1
). This relation provides a link
from the present to the future and represents the proactive aspects of the
model. The time it takes to do this is denoted as T
S
or the time needed to
select an action. The temporal relations described by the model will be
treated in Chapter 8.
Control Modes
A primary feature of the COCOM is the control modes, which correspond to
characteristic differences in the orderliness or regularity of performance.
Although the control that a JCS can have over a situation may vary
continuously, it is useful to make a distinction between the following four
control modes:
• In the scrambled control mode, the choice of next action is basically
random. For humans there is little, if any, reflection or thinking involved
but rather a blind trial-and-error type of performance. This is typically the
case when situation assessment is deficient or paralysed and there
accordingly is little or no correspondence between the situation and the
actions. The scrambled control mode includes the extreme situation of
zero control. It can lead to a vicious circle of failed attempts, which is
broken when one of the attempts succeeds. In model terms this
corresponds to a transition in control mode.
Control and Cognition 147
• In the opportunistic control mode, the salient features of the current
context determine the next action. Planning or anticipation is limited,
perhaps because the situation is not clearly understood or because time is
limited. An action may be tried if it is associated with the desired
outcome, but without considering whether the conditions for carrying it
out are met. Opportunistic control is a heuristic that is applied when the
constructs are inadequate, either due to lack of competence, an unusual
state of the environment, or detrimental working conditions. The resulting
choice of actions is often inefficient, leading to many useless attempts
being made. Success is determined by the immediate outcome,
disregarding possible delayed effects.
• The tactical control mode corresponds to situations where performance
more or less follows a known procedure or rule. The time horizon goes
beyond the dominant needs of the present, but planning is of limited
scope or range and the needs taken into account may sometimes be ad
hoc. If an action cannot be carried out because the pre-conditions are not
fulfilled, establishing the pre-conditions may become a new goal (cf. also
the description of the goals-means analysis in Chapter 6). The
determination of whether an action was successful will take delayed
effects into account.
• Finally, in the strategic control mode, the JCS has a longer time horizon
and can look ahead at higher-level goals. The dominant features of the
current situation, including demand characteristics of information and
interfaces, therefore have less influence on the choice of action. At the
strategic level the functional dependencies between task steps and the
interaction between multiple goals will also be taken into account in
planning. Outcomes are successful if goal post-conditions are achieved at
the proper time and if other goals not jeopardised.
The scrambled control mode is clearly the least efficient, while the
strategic is the most efficient – seen from the perspectives of either efficiency
or safety. In practice, normal human performance, and therefore also the
performance of JCSs in general, is likely to be a mixture of the opportunistic
and the tactical control modes. This corresponds to an equilibrium condition
or balance between feedback and feedforward and therefore to an efficient
use of available resources. Although the strategic control mode theoretically
speaking is optimal, it usually requires so much effort that it cannot be
sustained for longer periods of time. The main characteristics of the control
modes are summarised in Table 7.2.
The COCOM describes JCS performance as a mixture of feedback-
controlled and feedforward-controlled activities. This offers a way of
capturing the dynamic relationships between situation understanding (as
constructs), actions (as realised competence), and feedback or information (as
148 Joint Cognitive Systems
events). On a general level, the model shows how actions depend on the
current understanding (construct), which in turn depends on the feedback and
information (events) received by the system, which in its turn depend on the
actions that were carried out, thereby closing the circle. If control is lost, the
actions will likely be incorrect and inefficient, thereby leading to an
increasing number of unexpected events. This will necessitate a revision of
the construct, which requires both time and effort. If the revision fails, the
loss of control will remain and the situation may deteriorate even further. If,
on the other hand, the revision succeeds, control may gradually be regained.
Table 7.2: Main Control Mode Characteristics
Control
mode
Number of
goals
Subjectively
available
time
Evaluation of
outcome
Selection of action
Strategic Several Abundant Elaborate Based on
models/predictions
Tactical Several
(limited)
Adequate Detailed Based on
plans/experience
Opportunistic
One or two
(competing)
Just
adequate
Concrete Based on habits/
association
Scrambled One Inadequate Rudimentary
Random
In the beginning of this chapter, control was defined as the ability to
direct and manage the development of events, and especially to compensate
for disturbances and disruptions in a timely and effective manner. In terms of
control modes this can now be described as the ability to maintain a control
mode despite disturbing influences as well as the ability to regain a control
mode, should control have been lost. For a given domain it is possible to
develop more precise criteria for control mode transitions, and use these as
the basis for a model (e.g., Yoshida et al., 1997). The performance
characteristics of the control modes can also be used as a basis for identifying
the status of a JCS, hence as the basis of performance monitoring systems
(e.g., Hollnagel & Niwa, 2001).
Although COCOM is adequate to describe the basic dynamics of control
and to illustrate the principle of the control modes, actions can best be
understood by invoking different but co-existing layers of control. These
refer to different levels of performance rather than to different levels of
information processing, hence to a property of the JCS rather than of the
operator's internal cognition.
Control and Cognition 149
ECOM – EXTENDED CONTROL MODEL
In order to describe multiple levels of performance it is necessary to extend
the basic COCOM model, thereby turning it into an Extended Control Model
(ECOM). ECOM provides a way of describing how the performance of a JCS
takes place on several layers of control simultaneously, corresponding to
several concurrent control loops. Some of these are of a closed-loop type or
reactive, some are of an open-loop type or proactive, and some are mixed.
Additionally, it acknowledges that the overall level of control can vary, and
this variability is an essential factor with regard to the efficiency and
reliability of performance.
The idea that human behaviour comprises multiple simultaneous
processes is far from new. An early allusion to that was provided by Craik
(1943), as well as by Miller, Galanter & Pribram (1960). As noted in Chapter
4, MacKay (1968) also made the same point early on, though in a more
general manner; a later cybernetic model is the recursive hierarchy in the
Viable Systems Model (Beer, 1981; 1985). Broadbent (1977) provided an
excellent overview of the issues and pointed out that multilayered models
must comprise independent processes at each layer. Carver & Scheier (1982)
proposed a five-level system, although with feedback loops only. Other well-
know examples are Michon (1985) and Rasmussen (1986). Most of the
information processing models, however, fail on Broadbent’s criterion since
they do not involve parallel processes of a different nature but only the same
type of process repeated at different layers.
As far as the number of layers goes there is obviously no absolute
reference to be found. Since the purpose is to provide a way to describe,
analyse and model actual performance, the number of layers should be
sufficient to serve this purpose, but not so large that the descriptions become
unmanageable. Using the stages in the development of process control as an
analogy, a useful distinction can be made between four layers or loops called
tracking, regulating, monitoring, and targeting (goal-setting), respectively.
Adopting these four layers for the description given here does not in any way
preclude a revision at a later point in time. The assumption of multiple layers
of activity is crucial for the modelling approach, but the specific number of
layers or loops is not. In the following sections, each of the four layers of
activity will be described using the example of driving a car.
Tracking
The tracking loop describes the activities required to keep the JCS inside
predetermined performance boundaries, typically in terms of safety or
efficiency. In the case of driving a car this could mean maintaining an
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.