In the previous article in this series, we started looking at the object of belief as a means of unpacking the belief component of JTB. I made a distinction between propositions and states of affairs primarily under the framework of John Schellenberg. I presented the idea that the object of belief is broader than the idea of a proposition typically construed and that the distinction between propositions and states of affairs is less precise than we typically may think of them. In this essay, I explore this idea further and describe how the distinction relates to the Gettier Problem.
Propositions and their relationship to states of affairs
Similar to Shope and Schellenberg, Roderick Chisholm prefers an epistemology that places states of affairs at the center. Still, Chisholm is not as ready to minimize the ontological role propositions play in belief and, instead, chooses to place propositions as a subclass of states of affairs. This allows him to talk about belief in terms of propositions while still maintaining the identity relation with states of affairs. Chisholm developed his epistemology and theory of intentionality primarily in his 1979 Person and Object and his 1981 The First Person with the latter book providing a richer development of (and, in important ways, superseding) the former. My exegesis will focus on The First Person and reference Person and Object when doing so will be necessary to fill out his ideas.
In Person and Object Chisholm provided a general ontology in which he established states of affairs as the basic doxastic element on which he based his epistemology. His view in that book is that believing is primarily propositional. A belief de re is a species of belief de dicto such that the believer accepts a proposition that implies that an object possesses a property. If a proposition implies that something has a property and is true, then necessarily something is said to possess that property. On this view, individuals have a uniquely identifying property. If there is a proposition that consists of a conjunction of your uniquely identifying properties along with another property, then I can be said to believe (de dicto) that you possess that property.
This theory works when applied to beliefs about other persons but appears to break down when applied to beliefs about oneself. It requires that an individual accept propositions that include some self-identifying property. But what would this property be? It could be an individual essence or haecceity but then it should be possible to isolate specific features of this property that distinguish it from other properties. But the only distinguishing property seems to be being identical with me which does not seem to be any property at all. His solution is radical. He concludes that the best approach to this difficulty is to deny that there are such things as first-person propositions. While there are first-person sentences, there are no such things as first-person propositions because no sense can be given to self-identifying properties that are not empty. Further, this conclusion is based on a more general theory that the normal function of sentences containing demonstrative terms or names is not to express propositions.
States of Affairs and Propositions
Before we unpack the implications of this solution, I want to survey Chisholm’s general ontology of states of affairs and propositions and then discuss the epistemology he presents in The First Person. Many of the specifics of his ontology are established in Person and Object but the ideas provide a foundation for the work he does in the later volume.
Chisholm holds that states of affairs are occurrent or non-occurrent rather than true or false; they either obtain or do not obtain. States of affairs do not stand in an identity relation with events or objects in the world per se. States of affairs are distinct abstract objects. In Person and Object, Chisholm offers the following by way of definition.
C1 – p is a state of affairs = Df It is possible that there is someone who accepts p.
He expands this definition in The First Person by addition two qualifications:
C2 – p is a state of affairs = Df p is necessarily such that (i) it is possible that there is someone who conceives it, (ii) whoever conceives it conceives something which is possible such that it obtains and (ii) it is not a property or a relation. 
Specifically, a state of affairs is anything that is not a property or a relation and that can possibly be conceived by someone. To say that a state of affairs exists does not mean that it in fact obtains or occurs. This is important as he implies that states of affairs exist apart from anyone knowing them and that states of affairs that are not "concrete events" or do not obtain also exist. This idea is very similar to Schellenberg’s notion of propositional content though Schellenberg doesn’t seem as willing to allow for these types of abstract objects in his ontology. Since belief is our primary concern, I think we can set aside the difference in metaphysics between Schellenberg and Chisholm as long as they both function the same in doxastic contexts. Even so, Chisholm’s epistemology is tightly related to his metaphysics so a broader understanding of the latter will be important for adequately understanding the former.
Chisholm further develops the ontology of states of affairs in terms of properties and relations. The occurrence of states of affairs is described in terms of the exemplification or lack of exemplification of a specific property or relation. He states,
C3 – For every property or relation G, there is a state of affairs p and there is a state of affairs q which are necessarily such that: p obtains if and only if G is exemplified and q obtains if and only if G is not exemplified
Chisholm acknowledges that this view presupposes “an extreme view of Platonism” but that it can deal with problems like states of affairs that cannot possibly obtain like being both round and square. Further, Chisholm holds that all states of affairs like these as well as those that actually do obtain or exist necessarily. Regardless of whether one accepts Platonism of this sort, it seems true enough that people do believe seemingly impossible things. Chisholm’s position is that the objects of those beliefs are state of affairs that do not obtain and these states of affairs have at their root properties and relations, though states of affairs themselves are not a property or relation (there is no property “something being red”). States of affairs, then, “are of two sorts: those that obtain (occur or take place) and those that do not."
Propositions, like states of affairs, are abstract objects existing in every possible world but have the modal property of being true or false rather than occurrent or non-occurent. Since the two are so similar, aren’t we needlessly multiplying entities beyond what we need? Do we really need to say that among the things that exist eternally in all possible worlds is the state of affairs of Socrates being mortal and also the proposition that Socrates is mortal? He doesn’t believe the latter is ontologically distinct from the former and so defines propositions essentially are states of affairs but with an accidental modal property. A proposition is a state of affairs that either always occurs or never occurs.
p is a proposition = Df p is a state of affairs, and it is impossible that there is a time t and a time t’ such that p occurs at t and does not occur at t‘
A true proposition is one that occurs and a false proposition is one that does not occur. When a person believes a proposition, he believes a state of affairs is necessarily occurring or not occurring. If I say, I believe the proposition, “Bill Gates is the richest man in the world,” I believe the state of affairs “Bill gates being the richest man in the world” occurs and is necessarily true. To understand how this idea functions in Chisholm’s system, I think it’s best to think of a proposition as an abstract object that functions in doxastic contexts like any other property. States of affairs entail not only other states of affairs but also relations and properties. What does it mean, then, to say that a person believes a state of affairs? What does it mean to say, for example, that a man believes a storm is occurring? Chisholm considers these questions in the form of an objection:
‘(i) Your theory implies that, if a man believes that a storm is occurring, then that state of affairs which is the occurrence of a storm is the object of his belief. But (ii) the sentence “He believes that a storm is occurring” is natural and clearly grammatical, whereas “He believes the occurrence of a storm” is unnatural and not clearly grammatical. Hence (iii) if a man believes that a storm is occurring something other than the occurrence of a storm is the object of his belief.
The response to this objection is that there is no ontological difference between saying the man believes the occurrence of the storm and saying he believes that a storm is occurring. We may easily substitute either phrase with, “He believes in or suspects, or is counting on, or is mindful of, the occurrence of the storm.”  Similarly we may speak the same way of desire or emotive states (terms like fears, regrets, hopes) by substituting the event itself for a that clause. The object of the belief is the same in each case: the occurrence of the storm is what the belief is about and that is a state of affairs. Thus, descriptions of the man’s beliefs in terms of propositions really end up being descriptions in terms of states of affairs. Here again we can see similarities to Schellenberg’s account. The object of belief is a way the world possibly could be. It is not a sentence describing the world but a thought that the world actually is a certain way.
We could ask of both Schellenberg and Chisholm, “What is the distinction between thinking of a way the world possibly could be and thinking of a proposition that about the way the world possibly could be?” This is asked as a metaphysical question but we easily could ask it as an epistemological one as well: what is the distinction in terms of what a person believes. In both cases the believer appears to have as the object of his belief a way the world possibly could be. For Schellenberg, the answer is that a proposition (not propositional content) is a description of the world that is considered as an object of thought. As I stated in the previous article in this series, it turns out to look very similar to a linguistic token. Thinking about the proposition that the cat is chasing the mouse is to be thinking (quotationally): “the cat is chasing the mouse.” But the person thinking this just entertains the state of affairs as possible. Belief involves accepting this state of affairs as actual and so the believer does not merely think of the description of the way the world might be but thinks of a representation of the world itself and considers it to be actual.
For Chisholm, when one believes that a proposition is true, one believes that a state of affairs is necessarily occurrent. When one believes a state of affairs, the object of his belief is a world state (a possible way in which the world could be) in which some property or relation is exemplified, e.g. it is foggy outside. For a belief that it is foggy, the mind has as its object the state of affairs in which the property “being foggy” is true of the world outside my window as an event that possibly happens “outside the mind.” Notice that on this view, it is foggy makes no commitment about whether it now is foggy or if it has been foggy in the past or if it will be foggy in the future. It could be the case that it is never is foggy but the believer holds that it could be. But if the believer takes this state of affairs propositionally—“that it is foggy”—she holds that the state of affairs is true or occurrent. If she holds it to be true, then the description is necessarily true because she is committing to the idea that it is foggy now (or at time t). If it is foggy at time t then it is not possible that it is not foggy at time t. Believing propositionally is accepting that a state of affairs actually is occurrent (or non-occurrent) at a time.
For both Schellenberg and Chisholm, the object of the belief is a representation of world, or a possible event in the world, or a state of affairs. That is, the mind is directed to a broad set of properties and relations not to a description that can be expressed in language. While the state of affairs can be described in language, this description is not the object of thought.
In the next essay, I’ll explore Chisholm’s discussion of belief de re and de dicto as a means of further disambiguating how propositions and states of affairs function in belief contexts.
 Roderick Chisholm, Person and Object: A Metaphysical Study (La Salle, Illinois: Open Court Publishing Company, 1979), 14
 Alvin Plantinga thinks that self-identifying properties do make sense and so do not need to be treated differently than other properties. See Plantinga, The Boethian Compromise, 2003
 While we cannot explore this idea in full, I state it here because this conclusion is what drives Chisholm’s entire thesis in The First Person and is essential to understanding his epistemology in this volume.
 Roderick Chisholm, Person and Object: A Metaphysical Study (La Salle, Illinois: Open Court Publishing Company, 1979), 117
 Roderick Chisholm, First Person: Essay on Reference and Intentionality (Studies in Philosophy), (Minneapolis, Minnesota: University of Minnesota Press, 1981 by The Royal Institute of Philosophy), 9
 Person and Object, 119
 The exemplification relation, while important, is too involved to address here. For our purposes, it may be sufficient to say that a property or relation is exemplified when it is had by at least one thing; that is, when it occurs in reality.
 Ibid., 119
 It does seem odd to say that there are properties like square-roundness that cannot possibly obtain because then what does it mean to say “there are” properties like this? More to the current problem, what would it mean to say that a person is disposed to act as if there are things that are both round and square? Yet it seems correct to say that people can behave as if impossible things obtain so Chisholm’s position has at least prima facie plausibility. One perennial example concerns the existence of God. If a being like God does not exist (the properties of a divine being cannot possibly obtain), it is certainly the case that many people are disposed to act as if they in fact do obtain. So the state of affairs is accepted by these people yet they are impossible.
 Ibid., 122
 Ibid., 123
 Adding the temporal index is important here and helps make sense of the modal property. Really, what I’m saying is that Bill Gates is the richest man in the world at t. If that state of affairs obtains at t then it is not possible that the state of affairs does not obtain at t. There is debate whether statements about the future have truth value. On this theory, they would not. For “Bill Gates will be the richest man in the world” neither obtains nor does not obtain and thus could not be necessary or impossible. But these are complexities we can set aside for the present.
 Ibid., 124
 Ibid., 123