وَحَدَّثَنِي حَجَّاجُ بْنُ الشَّاعِرِ، حَدَّثَنَا أَحْمَدُ بْنُ يُونُسَ، قَالَ: سَمِعْتُ زُهَيْرًا يَقُولُ: قَالَ جَابِرٌ، أَوْ سَمِعْتُ جَابِرًا يَقُولُ: إِنَّ عِنْدِي لَخَمْسِينَ أَلْفَ حَدِيثٍ مَا حَدَّثْتُ مِنْهَا بِشَيْءٍ. قَالَ: ثُمَّ حَدَّثَ يَوْمًا بِحَدِيثٍ، فَقَالَ: هَذَا مِنَ الْخَمْسِينَ أَلْفًا.
Translation: Through Ḥajjāj b. al-Shāʿir, Aḥmad b. Yūnus and Zuhayr: Zuhayr said that Jābir said—or that he heard Jābir say—he possessed fifty thousand reports of which he had transmitted none. Later Jābir related a report and said it belonged to that collection.
«قَالَ جَابِرٌ، أَوْ سَمِعْتُ جَابِرًا يَقُولُ» preserves uncertainty in how the words were attributed or heard. It should not be replaced by a definite hearing formula.
The assertion that none had previously been transmitted can precede a later decision to transmit one. The later act does not, by itself, logically falsify the earlier statement.