A judgment must be read as a whole, so that conflicting parts may be harmonised to reveal the true ratio of the judgment. However, if this is not possible, and it is found that the internal conflicts within the judgment cannot be resolved, then the first endeavour that must be made is to see whether a ratio decidendi can be culled out without the conflicting portion. If not, then, as held by Lord Denning in Harper and Ors. v. National Coal Board (1974) 2 All ER 441, the binding nature of the precedent on the point on which there is a conflict in a judgment, comes under a cloud.
In Harper (supra), the decision in Central Asbestos Co. Ltd. v. Dodd (1972) 2 All ER 1135, a House of Lords judgment, had to be applied. It was found that two learned Law Lords decided the question of law in favour of Dodd, whereas two learned Law Lords decided the question of law against Dodd, stating that his claim was barred. As Lord Denning stated, the fifth Law Lord, Lord Pearson, was the odd man out, in that he agreed with the two learned Law Lords that the law did not support Dodd's case, but agreed with the minority judges that Dodd's claim was not barred. This being the case, Lord Denning spoke of the precedential value of Dodd's case as follows:
“How then do we stand on the law? We have listened to a most helpful discussion by Mr. McCullough on the doctrine of precedent. One thing is clear. We can only accept a line of reasoning which supports the actual decision of the House of Lords. By no possibility can we accept any reasoning which would show the decision itself to be wrong. The second proposition is that if we can discover the reasoning on which the majority based their decision, then we should accept that as binding upon us. The third proposition is that, if we can discover the reasoning on which the minority base their decision, we should reject it. It must be wrong because it led them to the wrong result. The fourth proposition is that, if we cannot discover the reasoning on which the majority based their decision, we are not bound by it. We are free to adopt any reasoning which appears to us to be correct, so long as it supports the actual decision of the House.
In support of those propositions, I would refer to the speech of Lord Dunedin in Great Western Railway Co. v. Owners of S.S. Mostyn  A.C. 57, 73–74, and of Lord MacDermott in Walsh v. Curry  N.I. 112, 124–125, and of Viscount Simonds in Midland Silicones Ltd. v. Scruttons Ltd.  A.C. 446, 468–469. Applying the propositions to Smith v. Central Asbestos Co. Ltd. [Dodd's case]  A.C. 518, the position stands thus: (1) the actual decision of the House in favour of Dodd must be accepted as correct, We cannot accept any line of reasoning which would show it to be wrong. We cannot therefore accept the reasoning of a minority of two — Lord Simon of Glaisdale and Lord Salmon — on the law. It must be wrong because it led them to the wrong result. (2) We ought to accept the reasoning of the three in the majority if we can discover it. But it is not discoverable. The three were divided. Lord Reid and Lord Morris of Borth-y-Gest took one view of the law. Lord Pearson took another. We cannot say that Lord Reid and Lard Morris of Borth-y-Gest were correct: because we know that their reasoning on the law was in conflict with the reasoning of the other three. We cannot say that Lord Pearson was correct: because we know that the reasoning which he accepted on the law led the other two (Lord Simon of Glaisdale and Lord Salmon) to a wrong conclusion. So we cannot say that any of the three in the majority was correct. (3) The result is that there is no discernible ratio among the majority of the House of Lords. In these circumstances I think we are at liberty to adopt the reasoning which appears to us to be correct.”