Abstract
Relative-age dating (RD) techniques rely on indices of post-depositional change; principally degree of rock surface weathering or soil development. In the chronosequence approach, where a measured parameter displays a trend in values that parallels a trend in surface age, the parameter is commonly assumed to be age-dependent. Such an assumption may be invalid. This is demonstrated with particular reference to glacier forelands, where environmental and sedimentological factors, that can influence commonly used RD parameters, may exhibit systematic variations in parallel with, but independent of, surface age. Where some independent dating control is available, regression analysis may be used to define chronofunctions, providing numerical ages based on RD parameters. However, the error limits of such 'ages' are often underestimated. The use of modal rather than mean values is particularly problematic. A more critical approach is suggested, involving deductive experiments to examine the influence of non-temporal controls and to test the assumptions and predictions implicit in a given interpretation. A procedure for standardizing error limits associated with chronofunctions is proposed. It is stressed that the influence of systematic errors can only be assessed by adopting a critical, deductive methodology and that attaching spurious accuracy to dates will only lead to confusion.
Get full access to this article
View all access options for this article.
