Once again in Russell's History of Western Philosophy, here where the thoughts of George Berkeley are being examined regarding reality and perception and, skipping the preliminaries but just giving the following which Russell imagines refutes whatever Berkeley is asserting.
"There is a house which noone perceives." Whether this proposition is true or false, I do not know; but I am sure that it cannot be shown to be self-contradictory.
So here it is stated as an objective fact that there is a house which exists but is unperceived. This, to emphasise, as a piece of language, is a statement of fact, a definite assertion, and not any kind of hypothesis. And for it to be declared a fact it must be known to be so, and how is the existence of this house known to be a fact but only through observation. Without this observation or perception it cannot be a known fact. And so it makes no sense to say there is a house which is unperceived. To assert that the house is is to say it is a fact based on observation; so by stating that there is a house that is not perceived, one is stating as a fact that which one is simultaneously stating cannot be a known fact.
To say: "There is a house" isn't to make a hypothetical statement; it is to state a fact about the world of observable phenomena and which can be proven, otherwise it is not a sensible claim. For someone else to then demand proof of the existence of this house that is unperceived would in response require perception of it to attempt to make the statement of its existence rational, but then if such proof is furnished, this would then falsify the statement that the house is unperceived.
You cannot attempt to make statements of facts about observable data which simultaneously deny themselves the necessary foundation to be such statements of fact, and so, unfortunately for Russell, despite his sureness that the given line is not self-contradictory, it is precisely so - an absurd and incorrect use of language.