A somewhat incredible interpretation of 'that'-clauses: Russell, non-existent kings and false objectives

'The present King of France', Russell maintained in On Denoting, seems to force us into either of the following positions. Either we follow Meinong but then infringe the law of contradiction, or posit senses beyond references, but we seem never able to genuinely talk about them. Hence, Russell urged, it is worth considering 'a somewhat incredible interpretation' of definite descriptions. This is his famous theory of descriptions, now not considered to be incredible anymore. In this paper we will investigate whether and, if so, how some of On Denoting's core claims can be applied not to the present King of France, but to propositions. We will see that following Russell on his considerations concerning the present King of France but not on parallel considerations concerning propositions and phrases that seem to refer to propositions is an unstable position, which Russell himself in On Denoting, and most of us today, seem to occupy. We will then discuss how to reach a more stable position. According to the account we will end up presenting, while there is some thing we believe when we believe truly, there is no thing we believe when we believe falsely. The account might look like a non-starter and goes against what Russell himself claimed. But, we will argue, we can find in On Denoting and in its theory of descriptions the resources to render such an account at least a bit less incredible.