L-states are a necessary component of the causal theory of perception, which generally states that a subject’s perception is of an object only if some notion of causation by the object can be satisfied. Accordingly, the L-state is the effect an object must have on the subject which perceives it, which is that of a visual state that may be described with a sentence in the form of, “It looks to the subject that…” This is to say that the subject sees an object only if it looks to the subject as though it is looking at the object. It’s easy to suggest, then, that if the subject sees an object, the subject is in an L-state. A second inference is that if the subject sees an object, the object exists. It would seem difficult to explain a non-existent object performing an act of causation, and so this theory of perception only sets out to deal with the seeing of real things.
Grice offers several compelling examples to support such a theory of perception in the face of odd perceptual scenarios. One such example is in that of a person to which it looks that they are facing a pillar, but instead there is a mirror standing between them and a pillar, and from behind is reflected another pillar offering the same visual content which instead is what they are actually seeing. It seems that although they are seeing a certain object, the L-state they are in is not being caused by that object. Rather, it looks to the subject that they are seeing the pillar behind the mirror. Grice’s response is that these are special cases in which the object being seen is not the cause of the L-state, and so the best theory of perception must rely on a causal interaction.
Snowden agrees that the causal theory and Grice’s response for this phenomenon are reasonable, but suggests that this does not necessarily rule out other valid forms of explanation. Accordingly, the act of seeing something should be taken to be the primitive relation, in the form of ‘the subject sees an object.’ In this way, we define the L-state in terms of what it means to be seeing something. It follows that it would seem to the subject that they are looking at an object only in the case that for some object, they are seeing it. However, this does not account for things such as hallucinations and illusions, in which case one is in an L-state but may not be seeing any object at all. This leads to a final disjunction to account for these occurrences, which states that it seems to the subject that they are looking at an object, only if there is something that they see to be this object, or that it is to the subject as if there is something that they see to be this object. Hallucinations are handled by the former part of the disjunction whereas the seeing of actual objects is handled by the latter. What was previously seen by Grice to be a causal effect, Snowden argues is rather more simply a closest explanation for what one happens to be looking at. He believes that this is the best theory for L-states, because the object being seen is actually constituent to the L-state itself, and is not simply an external cause of it.
He argues that Grice has confused what it means to see something with what it means to be in an L-state in the argument for the causal condition of perception. For Grice, it is ‘looking to the subject as though,’ which should be taken as primitive. All the different types of hallucinations, illusions, and actual visions, then fall under this same basic category. What differentiates each type is the underlying cause for its specific state, and not any difference to be found in the state itself. Under this view being in an L-state is a function of the causal effect of what one is looking at. Conversely, Snowdon argues that the seeing of an object and a hallucination of the same object have very little in common with one another. In his view it is seeing that is taken as primitive, it does not matter what your perception happens to be. What is differentiated is the explanation for what one is looking at, and so L-states are instead founded on the act of seeing. In this way looking at an object can be distinguished from a hallucination of the same object by an account for the state one has come to be in.
Burge thinks that perceptual psychology has given us an account of the very nature of perception itself, and not merely what prior theories detailed as to what might be the possible causes of perception. He bases this idea on findings in the field of psychology indicating the ways in which our perception works to represent the world around us. The primitive element of perception in this view is then the perceptual representational state, which is derived from the L-state in Grice’s theory, in that it details its own veridicality conditions and suggests certain causal relations. Under the causal view of perception, the veridicality condition is equivalent to the truth value of the proposition according with the L-state that they are in, that of the form, “it looks to the subject as though…” If an object is the cause of a subject’s L-state in some specific way, then the subject is seeing this object.
However, Burge believes these types of prior perceptual theories have not gone far enough in attempting to encapsulate the workings of perception. He expands upon them to suggest that perceptual states function to represent differentiable states of affairs in the world around us. Because this infers a definite reliance of perceptual content on the external environment, the direct constituents of a certain perceptual state must be located not only within the subject’s consciousness, but are also partially to some how be found from the outside. Similarly, it is insufficient to consider solely the direct effect of an object on its subject without also following the general pattern of causation that led to this specific type of perceptual state as a whole. In this way, perception is seen to be not merely propositional in the way that the causal theory offers an explanation for the origin of a perceptual state, because it also works to approximate patterns found outside of ourselves within our environment. The way we model our perception after patterns found to be in nature seems like it can be widely variable as even slight differences across the perceptual spectrum can lead to the vision of noticeably different patterns. This is what Burge means when he says that perceptual content consists of a certain representation of the external world. He explains perception not in terms of a causal explanation of what it means to see, but as a representation of a type of the surrounding state of affairs.
Perceptions contain both broad and particular sorts of content within them. Not only does one see an object, but one represents that object as being particular to itself, and containing such and such characteristics. In this way, when we see a tomato, we may think of its contents containing ideas like ‘red’, ‘sitting on a dish’, and ‘being ripe’, but they are not exactly propositions in this simple sense. To explain this, Burge utilizes the idea of perceptual constancies to suggest a theory of the content of perception in terms of which features a subject’s environment its perception works to maintain constancy in relation with. In some manner, the subject is able to shift, or at least believe itself to, from the subjective content of its primitive sensory experiences to seemingly objective statements about states of affairs and properties of its environment. This would be an example of the difference between seeing a red tomato and being able to determine that the tomato before you is red.
It is important for this theory that it be possible for one and the same type of perceptual representational state to be on one occasion in a seeing of an object and on another occasion a referential illusion. Burge argues that perceptual disjunctivism as that described by Snowdon is inconsistent with the way perceptual representational states are individuated in the psychology of vision. This is because he believes that the disjunctivists are wrong in suggesting different visions are of different types of states, and reverts to the causal belief that all visions are of the same primitive type. The thing that distinguishes illusions from perceptions of real objects is the veridicality of each state. Our mind can be quite adept at picking out illusions from their real counterparts, even in instances of incredible similarity. In the example of the Ames room, certain configurations of the heights of the men in the room are ruled out as being implausible in relation to the constancy of human height, which doesn’t reach nearly what the large man in the room might at first appear to be. Rather than weighing the truth value of varying propositions under causal relations, these limits fall in line with the constancies used by perception to represent the world around us.