If you ask:

“Is it not an established principle that additional information transmitted by a trustworthy narrator is accepted, and that precedence is given to the narrator who attributes a report to the Prophet over one who attributes it only to a Companion, because the former provides additional information?”

I reply:

We hold this to be true when the narrators reporting it are of equal standing and comparable reliability. But when those who do not transmit the additional information are more numerous, our predecessors differed over the matter, and it requires consideration. Moreover, that principle applies only when the two versions do not conflict. When they do conflict, the version transmitted by the smaller number is undoubtedly rejected.

Here, the report cannot be reconciled by supposing that the Prophet—may Allah bless him and grant him peace—said these words, and that his Companion subsequently repeated them, sometimes speaking in his own right and sometimes quoting him. Rather, it recounts a particular incident: a man sought a ruling in a specific manner, using particular words that have been transmitted. The narrators state that the person seeking the ruling was Ibn ‘Abbās, whereas this version attributes it to the Prophet—may Allah bless him and grant him peace. There was only one incident. Had there been more than one, the trustworthy narrators of firmly established accuracy would not have neglected to mention this, given their well-known attention to such matters. Furthermore, the scholars versed in hadith criticism are better qualified to assess this, and they have no doubt that this version is an error.

The fourth proof:

The basic presumption concerning substances is purity. Semen must therefore be judged pure until evidence reaches us that requires us to declare it impure. We have investigated and examined the matter, but have not found for that view