Key Articles:3

Her criticism of strong AI (artificial intelligence) is that its proponents are only looking at the products or results of thinking. The author, however, wants to concentrate on what the knower knows. She views this as an interactive relationship between mind and world. She argues that the process by which one acquires knowledge of particular objects is different from the process of acquiring concepts about these objects.

In this book, the acquisition of knowledge is referred to as ‘x’ and the development of the concept as ‘y’. She refers to ‘x’ as ‘object positing’ and ‘y’ as ‘property attribution’. She thinks that these processes can be analysed separately, although they often operate in tandem. The ‘x’ object positing process involves the identification or recognition of particular features. Object positing usually involves direct perception or direct experience. The ‘y’ property attributing capacity deals with sorting, qualification and quantification. The two types of mental processes, object positing and property attributing, together comprise the mental processes, by which we interpret the world.

The author’s objection to strong AI is that it lacks any object positing, and works only at the level of property attribution. The ‘x’ process incorporates the ‘passing show’ of the world, and the ‘y’ process comprises the computational structures that work on the raw data of experience. The ‘y’ process of computational structuring taken by itself does not work, because it lacks the raw data on which to work. Similarly the passing show does not mean much without the structuring work of the ‘y’ processes.

The ‘x’ and ‘y’ distinction is not new in philosophy, but it is against the grain of 20^th century philosophy, which has tended to base theories on the ‘y’ process, and to work hard to squeeze out the ‘x’ process of first person experience. The price of this has been to exclude what actually happens in the world.

The author particularly attacks the work of the behaviourist philosopher W.V. Quine, who argued that if the reports of different observers agree, their public report takes priority, and private reports are of no significance. This is his justification for ignoring the ‘x’ process. Further, Quine, as a behaviourist, sees the personal experience as inaccessible to investigation and therefore worthless. Quine thinks that the senses need to borrow concepts of objects from elsewhere, while the author argues that an object impinges directly on the senses. It would however seem that the public report favoured by Quine is in fact comprised of a series of private reports, without which it would not exist. The author notes that the philosopher, John Searle, has argued that the fact that there are difficulties with investigating private experience does not mean that private experience does not exist. Both the author and Searle take the further view that even if we had the exact biochemical correlates of experience, the experience itself would remain private. For instance, even if we had the exact correlates of the radar experience of Thomas Nagel’s bat, we would have no idea what its radar experience felt like.

The author goes on to discuss her view of causation, which she characterises as ‘kausation’ to distinguish it from other theories of causation. In this, the observer is in direct contact with the external world, and at the same time is identifying and naming various aspects of experience. The author makes a further distinction between ‘x’ as the mind side of the process, and ‘r’ as the external world or reality side of the relationship. The subject identifies some external object as the source of their experience. She sees ‘x’ and ‘r’ as effectively identical, although separable for purposes of discussion. ‘x’ is the experience of an object. The object has meaning for the experiencer. This is the intentionality, or the ‘aboutness’, or ‘ofness’ of an experience, and this is something only possible for an entity with a point of view. In speaking of something with a point of view, the author presumably refers to conscious entities.

The author goes on to provide a diagram to further clarify her reasoning. She says that an aspect of reality 'r' impinges on a person 'P' kausing the experience 'x'. The person 'p' takes the view that their experience has been 'kaused'. P recognises 'x' as an experience 'r'. P names 'r' as the kause of the experience 'x' and identifies 'r' with the notion of the experience 'x'.  The author contrasts her position with that of Searle. He sees it necessary to introduce agency to explain perception or experience, while the author sees experience as something that must be exported to agency.

Some critics consider that a percipient’s brain requires a concept of a thing before it can perceive it. The author sees the brain as more active, with perceivers focusing attention and thence creating an ‘x’. Parts of the philosophical tradition have problems with the supposed concept of action at a distance in the percipient recognising a separate object, but the author argues that sensory inputs from light and sound are in principle no different from sensations from within the body. Sensory awareness should not be seen as less direct than bodily sensation and the idea of an external/internal split in the processing of sensory inputs is criticised as artificial.

Helen Keller

The author comments on the famous case of Helen Keller. For her words were initially just a game, but she suddenly grasped that the ‘x’ experience of pressing certain symbols was linked to the experience of washing. The author’s point is that it needs an entity that has experiences, to link one experience to another. A computer that didn’t have experiences could never link the typing of particular symbols to the experience, as distinct from the knowledge of the washing function. This is seen as similar to Searle’s Chinese room where he maintains that number crunching or symbol processing does not equate to understanding.

20^th Century Philosophy

Russell and other 20^th century philosophers held that the concept of sense data was distinct from the actual sensation. Russell thought that colours and sounds were sense data that were distinct from the actual sensation of seeing a colour or hearing a sound. In his theory, the colour itself is a sense datum, not a sensation. Thus Russell saw the territory as being split three ways between the physical object, the sense datum and the final sensation. In particular, he thought it necessary to have a sensation/sense data split, to distinguish between the act of perception and the perception itself. This sense datum turns out to be a process of logical inference, by which the subject determines the nature of object placed at a distance. However, this sense datum is not required for internal bodily processes.

This approach remained unchallenged during much of the 20th Century. The author highlights three features of the 20th century orthodoxy. Firstly there was  a strong distinction between the nature of sensations from the body, and the nature of sensations from the external world that were viewed as problematic. The external world is suggested to require separate data for sensations to be created that are not necessary for the internal world. Finally, the external sensations are judged to be objectively real, whereas the internal sensations are viewed as subjective and therefore not properly real.

Hidden Agenda

The author suggests that there are several assumptions concealed in this orthodox view of perception that she thinks these are false. She sees no reason to believe that sensing external objects is more problematic than sensing pains in the body, and therefore she sees no requirement for a sense datum in the former case. She does not agree that it is necessary to have a system of logical inferences in order to identify an object. The author argues that the psychological literature suggests that humans are just as well equipped to detect external objects, as they are states in their own bodies. In principle, she argues that there is no great difference between detecting an object in your locality, and detecting a blister on your toe. In both cases, attention diverted elsewhere could lead to a sensation being missed or misidentified.

Psychological research

The author goes on to examine some recent psychological research felt to support her arguments. Studies demonstrates that three operations are performed continuously by both the eyes and ears, which are segregation of objects from their background, determining distance and determining motion. The figure/background distinction applies if there are stimuli from more than one region. The figures rather than the background are the objects of interest. Interest is a crucial factor. The author sees interest in things as a prime distinction between humans and computers. Computers are much more efficient than humans for particular tasks, but they are not viewed as being interested in the task, or having any intentionality (feeling of the task being about something). Further to this some aspects of sensory performance such as the sensation of depth are directly linked to the nervous system. Another experiment, involving young children judging distance, showed that the majority of infants could judge that a larger/distant box was larger than a nearer/smaller box, despite the fact that the nearer/smaller box cast a larger image on the retina. The author argues from these examples that there is no separate sense data or logical inference machinery involved in perception. The last experiment demonstrates a built in or hard wired perspectival system in the brain. She further points out than some suggestions for logical inference structures in sensing the world rely on Euclidean space rather than perspectival structures now supported by experiment. These features suggest that perception is a spontaneous or hard wired feature, not requiring a sense datum or a process of logical inference. The author sees the perceptual act as directly tied to the nervous system. In being so tied, it is exactly the same as internal sensations of pain etc.

The author goes on to make a further attack on the notion that inferential activity is involved in sensing the external world. She regards the reasoning process involved in inferring as distinct from sense experiences. She takes the example of a belief that basket ball players are tall, and the knowledge that a particular individual plays basket ball, to infer that that individual is tall. However, this is an example of sorting to find a likely category for the individual concerned and does not involve any sense experience. Each logical step is independent of sensory experience. Sensory experience is something different and does not involve propositions.

The author discusses recent studies of brain deficits in patients with respect to the insights it gives for object recognition. In visual agnosis, patients cannot see an object, but can nevertheless catch it, if it is thrown. On the other hand, patients with what is known as optic ataxics can see an object, but cannot grasp it. Researchers, Milner and Goodale, suggest that this reflects the very different processing needed for object recognition and visuomotor activity. Thus object positing is suggested to be a distinct process from other brain activity, although most of the time, the two tend to be very closely coordinated.

The scientific paradigm

The author argues that the scientific world is based on a ‘y’ type mechanical point of view, but that by contrast, human thinking is never completely free of intentions and judgements. In her view, perception is viewed as causal on the world-to-mind side, and as a semantic identity relationship on the mind-to-world side. Kausation is seen as a matter of recognition and understanding, not of one thing making another happen. Experience gets divided into discrete and nameable units. The kausal relationship is a denoting and understanding relationship.

An important segment of 20^th century philosophy involved language analysis thinkers, who held that all philosophical problems were related to language. Symbolic logic was supposed to give language the same rigour as mathematics. The author’s approach to objects is, however, incompatible with this approach, because the nature of the actual experience is opaque to logical analysis. Moreover, the dependence of the concept of an object on experience means that only entities that have experience can have the concept of an object. This notably excludes computers.

Agency

The author takes the view that not involving agency at the perception stage, makes it easier to deal with agency later on in the process. She illustrates this with a small scene in which a man asks a female colleague to a party. The first three steps in a four step process involve the man remembering that there’s going to be a party, realising that he likes the woman, and subsequently spotting her in the cafeteria. The point is that these all involve experiences, some of them in memory, and it is only this that allows the final agency step of approaching the woman, in order to invite her to a party. What is argued to be essential, in order for an agent to act is the tying together of experiental material. This can involve ideas about experiences constituting a preferred future condition, and the agent then acting to achieve this. Agency could thus be seen as a sort projected understanding, the projection being into future conditions. Thus the ‘x’ in thought becomes one of the determinants of future action.

The author regards emotion and cognition as distinct although they interact. She uses the example of Downs syndrome, where people can show warmth socially, but are limited in cognitive capacities, and the contrasting example of patients with complete emotional dysfunction but good cognition. This is once again an example to the distinctness of the experienced from the cognitive. Similarly, the author considers there is direct experiental access to emotions, psychological needs, desires and anxieties, all rated as ‘x’ experiences and that these are a necessary basis for cognition.

The author argues that the commitment of cognitive science to syntactical structures as having a monopoly of mental language needs to be revised. Reality is found in experience but then organised and sorted by syntactically structured thinking processes. With subjective experience as the basic ground, there is direct contact with reality, but its ability to understand the experience is limited by existing knowledge. The ‘y’ type processes have the ability to generalise and makes inferences, and create possible scenarios. Knowledge involves both ‘x’ and ‘y’ processes encoding both experience and syntactical structures.