The need for a model of consciousness has been felt among scientists. The evidence for this is the fact that two such models have been proposed in the recent scientific literature. We shall quote from the authors’ summaries of their models and point out that they suffer from a common drawback. We shall then go on to develop a new model which we believe achieves the objectives for which these models were designed but which avoids their drawbacks.

A pathfinding model has been proposed by the psychologist Charles Tart. He suggests seeing consciousness as a system.

While the components of consciousness can be studied in isolation, they exist as parts of a complex system, consciousness, and can be fully understood only when we see this function in the overall system. ³

The components of the system are introduced as follows.

To understand the constructed system we call a state of consciousness, we begin with some theoretical postulates based on human experience. The first postulate is the existence of a basic awareness. Because some volitional control of the focus of awareness is possible, we generally refer to it as attention/awareness. We must also recognize the existence of self-awareness, the awareness of being aware. Further basic postulates deal with structures, those relatively permanent structures/functions/subsystems of the mind/brain that act on information to transform it in various ways. ⁴

Eleven subsystems are identified by experiential criteria, and the model diagrams feedback and feedforward connections between all these subsystems.. The attention/awareness subsystem has a central place in the system receiving input from every other subsystem.

There is enough information here about Tart's model to compare it with that proposed by the neurophysiologist E. Roy John to which we now turn.

Consciousness is a process in which information about multiple individual modalities of sensation and perception is combined into a unified, multidimensional representation of the state of the system and its environment and is integrated with information about memories and the needs of the organism, generating emotional reactions and programs of behavior to adjust the organism to its environment. Consciousness is third-order information. Many levels of consciousness can exist, in which these dimensions are present in variable amounts. The content of consciousness is the momentary constellation of these different types of information.

At the same time that consciousness is the product of an integration of preconscious sensations and perceptions structured in the light of previous experience and reflecting emotional state, drive level, and behavioral plans, feedback from consciousness to these more fundamental levels must take place. Memories are activated, attention is focused, perceptions influenced, emotions aroused, drive priories altered, and plans of behavior revised as a result of this feedback, producing a continuous reorganization of basic processes because of the influence of higher-level integrative and analytical functions. ⁵

The reference to consciousness as third-order information needs to be place in the context in which sensation is categorized as first- order information, and perception as second-order information. About these two lower orders John says,

We choose to define perception, as well as sensation provisionally as preconscious or unfelt categories of information processing. ⁶

John's model identifies a fourth, fifth, and sixth order of information which corresponds with subjective experience, the self, and self-awareness, respectively. While much needs to be said about these higher orders we have enough information for the purposes in hand.

The drawback shared by both the models of Tart and John is that they run together questions about the theoretical properties of a living system which are taken to be necessary conditions for the emergence of consciousness with the question of the nature of consciousness as a phenomenon experienced by such systems. We shall try to establish this point in the case of both models.

There is explicit recognition in Tart's model of these two perspectives in the distinction it makes between the postulate of attention/awareness and postulates of structure. We can understand this distinction to be a

distinction between the phenomenological content of consciousness and the variety of structurings which this content undergoes. However, if we so understand it we cannot go on to entertain the possibility that phenomenological content and structures are subsystems which interact.

Tart himself recognizes that the postulate of attention/awareness has a very different status from the postulates of structures as the following remark reveals.

The systems diagram...shows awareness in a distinct place, but it really spreads out through the various subsystems and so becomes consciousness. ⁷

On the interpretation here being offered Tart is forced to say this, since it is attention/awareness which is structured by the subsystems. Accordingly, a model of consciousness should be modelling attention/awareness itself, and not the structures which are responsible for changes from one state of attention/awareness to another. We may read the following passage from Tart as calling for just this.

The systems approach stresses the importance of attention/awareness as an activating energy within any d-SoC [discrete state of consciousness]. Yet if we ask what awareness is or how we direct it and so call it attention, we cannot supply satisfactory answers.

We may deal with this problem simply by taking basic awareness for granted, as we are forced to do at this level even though we do not know what it is. After all, we do not really know what gravity is in any ultimate sense, but we can measure what it does and from that information develop, for example, a science of ballistics. We can learn much about d- Socs [discrete states of consciousness] in the systems approach if we just take basic awareness and attention/awareness energy for granted, but we must eventually focus on questions about the nature of awareness. ⁸

A model of consciousness, we are maintaining, is precisely a device to aid our understanding of what Tart calls 'the nature of awareness'. The model

should be a model of the phenomenological properties of human experiencing and should not incorporate the structures which constitute a scientific explanation of those properties. This conclusion should be even more firmly borne in on us when we take note of the fact that our knowledge of the structures/subsystems to which Tart is referring is derived from our study of the characteristics of attention/awareness itself. As Tart himself acknowledges,

We acquire data about structures when the structures are functioning utilizing attention/awareness energy or other kinds of psychological energies. ⁹

Turning now to John's model, the first thing to note is that consciousness is identified with a certain kind of information. Here again we need to distinguish questions about the origins of information and scientific theories about information from questions about the experience of information as that information enters a human mind. After all, information can exist in libraries and such information has no direct connection with consciousness. This distinction is recognized in the passage quoted from John's work in that he equates consciousness with the transformation of information into a 'unified, multidimensional representation of the state of the system'. The key word here is the word 'representation'. Information is represented insofar as it is experienced, and it is this experiential factor which constitutes consciousness. Thus a key sentence in John's passage is the one which says that the content of consciousness is the

momentary constellation of these different types of information. However, although John makes a clear distinction between information and its representation or 'momentary constellation', he does not have the idea that a model of consciousness needs to model this representation as distinct from modelling a system in which this representation is itself a component interacting with other components which are identified only by their scientific properties. To make this crucial difference clear the following analogy used by Bateson is decisive.

It is, of course, true for the TV set that a satisfactory picture on the screen is an indication that many parts of the machine are working as they should; and similar considerations apply to the 'screen' of consciousness. But what is provided is only a very indirect report of the working of all those parts. ¹⁰

Using this analogy we can make the point that the models of consciousness offered by Tart and John are comparable to a model of a TV set in which the screen is one of the parts interacting with all the others. In contrast, we are maintaining, a model of consciousness should be a model of nothing but the screen, and should not include the parts of the machine which are physically responsible for the phenomenon of the screen.

We first need a model of consciousness before we can follow a theory about the biological function of consciousness, but John offers us a model which runs all these matters together. It runs together the question of how consciousness comes about and the question of what it is. The model we shall describe also conceives of consciousness in systems terms, as both John's and Tart's do, but the components of the system will consist entirely of phenomena of experience; i.e., it will only have phenomenological properties.

We find that U. Neisser criticizes current models of consciousness very much along the lines we have just done, only not so explicitly, in such a passage as this.

The villains of the piece are the mechanistic information- processing models, which treat the mind as a fixed-capacity device for converting discrete and meaningless inputs into conscious percepts. Because recent experimental studies of attention appear to support these models, it seemed necessary to suggest another interpretation of their results. ¹¹

Neisser suggests another interpretation, but he does not fully integrate his new interpretation with a view of consciousness itself. We feel that the adoption of the model of consciousness we are developing would enable him to do this, and that the model is certainly relevant to the type of theorizing he is doing.

The model of consciousness we propose as an alternative constitutes a systems approach to consciousness. Unlike Tart, however, we do not conceive of consciousness as a system. In order to adequately compare our model with his it is important to understand how his approach models consciousness as a system.

In explicating his systems approach Tart introduces the new term 'discrete state of consciousness' or 'd-SoC' and speaks of such discrete states in the following way.

We have now defined a d-SoC for a given individual as a unique configuration or system of psychological structures or subsystems, a configuration that maintains its integrity or identity as a recognizable system in spite of various (small) changes in the subsystems. The system, the d-SoC, maintains its identity because various stabilization processes modify subsystem variations so that they do not destroy the integrity of the system. ¹²

What Tart apparently has in mind is that consciousness, as a system, is capable of assuming a variety of steady states each of which can be referred to as a 'discrete state of consciousness'. Essentially, Tart is interested in understanding the relationship between various states of consciousness such as the ordinary waking state and dreaming sleep. He has labeled such states 'discrete states' and is suggesting that it is helpful to conceive of such states as steady states of a system, consciousness. For to conceive of a particular state of consciousness as a steady state of a system entails understanding that the state is a product of a variety of contributing factors. Hence in any attempt to explain phenomena such as the transition from one state of consciousness to another a variety of stabilization processes that contribute to maintain particular states despite disruptive forces must be taken into account together with the various forces that would disrupt a state or interfere with the stabilization processes. In effect, Tart is suggesting that systems theory can be employed in the study of states of consciousness in such a way as to rule out the kind of attempt to explain a particular state of consciousness that would over-simplistically single out a factor contributing to the maintenance or induction of that state as the cause of the state.

Yet regardless of the reasons Tart has for adopting this systems approach to the study of states of consciousness it is a strategy that obliges him to conceive of consciousness as a system comprised of components related to each other as are the parts of a system. In enumerating these components and describing the relationship between them he is

constructing a model of consciousness. We have objected to the model he constructs in that it conflates a description of the phenomenological properties of human experiencing - what it is like to be aware, or attend to something - with a description of information processing structures of the organism. This confusion is the result of conceiving of both awareness and such information processing structures as components of consciousness, related to each other as parts of a system.

Insofar as Tart hedges on his decision to treat attention/awareness as a component of consciousness or subsystem on a par with the other subsystems we might entertain the possibility that his model could be corrected if the phenomenological term attention/awareness was deleted from the list enumerating subsystems, leaving as components only the subsystems of the information processing type. If such a correction were to be made the model would describe a system comprised of components or subsystems but could be understood only as modeling a system we take to be conscious or have experience. At best, in other words, Tart's model can be understood to describe a conscious system rather than describing consciousness as a system or giving a systems view of consciousness itself.

The mistake Tart makes in constructing his model of consciousness is essentially the result of an over-simplistic description of the phenomenological properties of human experiencing. He assumes that to be conscious is to be aware of something and that a simple distinction can be made between that which a person at a given time is aware of and that which

the person at that time is not aware of, as if consciousness were merely a spotlight lighting some part of the world and leaving other parts dark and unexposed. Although he allows that awareness can be under the control of the person whose awareness it is and accordingly speaks of attention, the phenomenological terms he employs - 'consciousness', 'awareness', 'attention' - are vague and hence not conducive to the construction of a model of consciousness itself. Another way of saying this is that insofar as Tart has a model of consciousness itself it models consciousness as a 'spotlight' in the way we have described. [continued]