States of affairs as distinct from propositions
John Schellenberg argues that when one believes something, she is not directing her mind towards a proposition. That is the object of belief is not a proposition. Rather, belief is a disposition to respond to the world as if some state of affairs is true. The mind is “directed to… a possible arrangement of things, not a proposition indicating or representing or reporting that arrangement.” He refers to this as a “phenomenological” approach to belief analysis because the description of the object of belief is done not in terms of propositions but in terms of what the subject experiences. A belief, he claims, is “simply thinking of the world.”
For Schellenberg, belief simpliciter is not propositional. Thoughts, he says, are mental states consisting of a possible state of affairs being represented to an individual. Propositions enter the picture when a person has a thought that some state of affairs is possible. This act is the coming to mind, in one way or another, and however briefly, of a state of affairs. It is the same act as what is generally referred to as "entertaining a proposition." One can entertain a proposition as possibly true without believing that the state of affairs actually obtains. This he calls a "propositional thought." The important distinction here is that one can entertain a proposition without believing the proposition is true. Thinking that your cat is chasing a mouse is to entertain this proposition with no commitment that about its truth conditions. Believing that your cat is chasing a mouse is to have a disposition to act as if it’s true though the object of the belief is the state of affairs: the cat is chasing a mouse.
"Thinking of a state of affairs s and having the thought that p, where p reports s, amounts to the same thing." The two ideas are identical as long as we don’t hold that the object of the thought is a proposition. In this case, the object of the thought is a possible state of affairs. "What one is thinking of here, what one's mind is directed to, is a possible arrangement of things, not a proposition indicating or representing that arrangement."  A proposition, then, is an indicator or representation of a state of affairs. A proposition expresses the content of a thought but is not the object of thought. "When upon hearing the dog bark, I have the thought that a car is turning into my driveway, I am thinking of a car turning into my driveway, not of the proposition ‘A car is turning into my driveway.’” The propositional content is identical with the object of the belief but not with the statement that expresses the state of affairs.
He further claims that it is natural to tend to think that the psychological state of having a thought is a proposition because we tend to “work backwards” from the verbal expression, “I thought that” to the belief itself. The proposition, however, just allows us to express or represent the content of the thought but is not a part of the subjective phenomenology (unless, of course, one is thinking about a specific proposition qua expression of a state of affairs). One can have a propositional thought but one can also consider its propositional content. These are different activities. The proposition expresses the thought but this is different than thinking of the state of affairs that the proposition expresses.
Belief itself is not a thought but a disposition “to have a certain kind of thought.”  He defines “S believes that p” as follows:
S is disposed to apprehend the state of affairs reported by p, when that state of affairs comes to mind, under the concept reality.
The disposition is to act as if the apprehension of the state of affairs is the way the world actually is. The state of affairs “comes to mind” when one experiences a belief. (This is to be contrasted with believing simpliciter). “We may say that my experiencing a belief amounts to my having a thought concerning what is objectively a way things might be (or might be said to be) in which that item is apprehended under the concept reality: that is, having a propositional thought so colored by the concept of reality, one whose subjective phenomenology is so completely determined by the concept of reality, that it does not occur to the thinker at the time that the state of affairs with which she is communing in thought could be anything but realized.”
I think there is some ambiguity in Schellenberg’s account but he isolates some key features of believing that will be important to us here. The ambiguity surrounds his use of the term proposition. In some places he speaks of a proposition as other philosophers speak of sentences: an exemplification of a state of affairs that can be understood as a fact about the world. In other places, a proposition seems to be just a representation of the way the world is. I think his exegesis boils down to this. A belief is a disposition to apprehend the world as it possibly is and to hold that this possibility is actual. Propositional content is the state of affairs the belief is about. That is, propositional content is a possible way the world is as apprehended by a believer. Finally, a proposition is a fact that can be stated linguistically and which has propositional content. It’s instructive to point out that the ambiguity in Schellenberg’s exegesis creeps in due to a conflation of propositions and sentences which is a common problem when discussing propositions. Propositions are not sentences nor do they function like sentences in doxastic contexts. As we’ll see, this error is a source of some of the confusion around Gettier cases. A collapse of these states of affairs and propositions would allow him to retain his general analysis of belief while aligning more with a traditional view of propositions. We’ll address this later. The value of Schellenberg’s account rests on two key concepts. The first is that beliefs are wholly and irreducibly subjective. This implies that the following proposition is true.
P1 - the ontology of both the experience of a belief and the belief itself are essentially phenomenological and not linguistic or propositional.
This has implications for how one ought to understand descriptions about beliefs particularly in Gettier cases. If a description of a belief does not include relevant facts about the content in the mind of the believer, including the believer’s intentions, rationality, doxastic environment and the like, the description may not be of belief at all. That is, if the truth conditions described in Gettier cases are about propositions and not propositional content, the cases do not constitute a counterexample to JTB. Belief descriptions must include relevant descriptions about the believers who have them.
When it comes to Gettier cases, propositions play an important role because they play such an important role in logic. So a theory of belief should be able to account for propositions not merely as expressions of belief or states of affairs. Further, belief is a key element in the standard conception of knowledge and it will play the lead role in my response to Gettier. It is to that phenomenon we must now turn. In order to adequately show why Gettier cases do not present a sufficient rejoinder to the tripartite theory, and in order to adequately account for most of the relevant and difficult cases, a thorough examination of belief is required.
 Schellenberg rightly notes that analyzing belief in terms of dispositions gives us the latitude to talk about persons holding beliefs even when one may not be consciously thinking of the things they believe. If beliefs were understood as items of conscious thought only, we would have to say that when a person wasn’t thinking of a state of affairs, she doesn’t believe it which is, as Schellenberg says, absurd. J.L. Schellenberg, Prolegomena to a Philosophy of Religion (Ithaca, NY: Cornell University Press, 2005), 41
 Ibid., 42
 Ibid., 47
 Ibid., 42
 Ibid., 50
 Ibid., 49