Consilience
Public.icon
In science and history, consilience (also convergence of evidence or concordance of evidence) refers to the principle that evidence from independent, unrelated sources can "converge" to strong conclusions. That is, when multiple sources of evidence are in agreement, the conclusion can be very strong even when none of the individual sources of evidence is very strong on its own.
To ensure coherence, independent methods of measurement must be used. That is, the measurement methods have few common features. That is, the mechanisms by which measurements are made are different. Each method depends on unrelated natural phenomena. For example, the accuracy of laser rangefinders is based on scientific understanding of lasers, while satellite photos and meter sticks depend on different phenomena. Because the methods are independent, if an error occurs in one of the methods, the likelihood of an error occurring in the same way as the other methods is extremely low, and the measurement value may differ. Note 1 If the scientific understanding of laser characteristics is inaccurate, laser measurements will be inaccurate, but other measurements will not be inaccurate. As a result, when multiple different methods agree, no method is wrong and the conclusion is strong evidence that it is correct. This is because the possibility of error is greatly reduced. If the consensus estimate value obtained from multiple measurements is incorrect, the error must be the same for all samples and all measurement methods, but this is very rare. Random errors tend to be offset by regression to the mean as the number of measurements increases. Systematic errors are detected by differences between measurement values (which tend to be offset by the fact that the direction of error is random). In this way, scientific theories become highly reliable by accumulating many pieces of evidence that converge to the same conclusion over time.
In reality, not all experiments are perfect, so some deviation from established knowledge is expected. However, if the convergence is strong enough, new evidence that contradicts previous conclusions is usually not strong enough to surpass that convergence. If there is not equally strong convergence for the new results, the weight of the evidence still supports the established results. In other words, new evidence is likely to be wrong.