A land transaction can become a court file for reasons that are adjacent to title but not the same as a title decision. That is the careful lesson from Clifford Okello Rachuonyo t/a Rachuonyo & Rachuonyo Advocates v Mohamed Yusuf Soroya & 3 others, Court of Appeal at Nairobi, Civil Appeal No. 224 of 2018.
The dispute arose from H.C.C.C No. 118 of 2010 and a conveyancing transaction over Nairobi/Block/94/184 in Nyari, Nairobi. The Court of Appeal delivered judgment on 22 May 2020. Its result was narrow: the advocate's libel appeal failed, the purchaser's cross-appeal on fraud and professional negligence also failed, and the court made no order as to costs.
What the court was deciding
The judgment records that the advocate had acted in a sale transaction for the Nyari parcel. The later conflict included competing accounts about the transfer documents and the price figures appearing in them. The purchaser wrote a letter criticizing the advocate's handling of the matter and copied others. The advocate then sued for libel, saying the letter injured his professional reputation.
The purchaser resisted the claim and brought a counterclaim against several parties. The counterclaim sought recovery and damages on allegations connected with the same conveyance, including fraud, collusion and professional negligence. Those allegations were before the court as pleaded claims to be proved, not as established facts.
Why the libel appeal failed
The Court of Appeal treated the defamation issue as a proof problem, not as an endorsement of the letter's language. Libel required proof that the impugned words were published to a third party and that the publication caused reputational injury.
On the record before it, the appellant had not called a recipient of the copied letter to prove that the letter actually reached and was read by a third party. Nor was there third-party evidence showing that his reputation had been lowered because of the publication. The court therefore upheld the High Court's dismissal of the libel claim. The hypothetical damages discussion in the High Court remained hypothetical; it did not become an award.
Why the cross-appeal also failed
The cross-appeal required a different analysis. For fraud, the court repeated the familiar rule that fraud must be specifically pleaded and strictly proved by evidence stronger than ordinary balance-of-probabilities suspicion. It then re-evaluated the transaction record and agreed that fraud and collusion had not been proved against the advocate and the other respondents to the required standard.
The professional-negligence claim also failed on this record. That should not be turned into a careless shortcut that every negligence claim requires the fraud standard. The better reading is narrower: where a party pleads fraud, fraud needs strict proof; where professional negligence is alleged, the claimant still needs evidence establishing the duty breach and loss asserted. The Court of Appeal was not satisfied that this evidentiary burden had been met.
What this does not prove about the parcel
This is not a title-rectification judgment. It did not cancel a registration, order the register changed, declare the conveyance clean, or make an affirmative finding that every step in the transaction was risk-free. It resolved appellate questions about libel publication, reputational injury, fraud proof and professional-negligence proof.
For buyers, lenders and advisers, the practical lesson is about classification. A parcel-linked case may be important risk context even when it is not a final ownership ruling. The right follow-up questions are specific: what did the court actually decide, what issues were left undecided, what documents were proved, and what current official verification is still needed?
Source: Judiciary Court of Appeal decisions, civil appeals and applications delivered on May 22, 2020, archive member locator PDF CIVIL APPEALS & APPLICATIONS/CIVIL APPEAL NO 224 OF 2018.pdf.
ArdhiScore helps organise parcel-linked judgment signals without turning them into guarantees. See how ArdhiScore works.