Empiricism and model-building in stratigraphy: Around the hermeneutic circle in the pursuit of stratigraphic correlation
Journal Articles
Overview
Research
Identity
Additional Document Info
View All
Overview
abstract
The discipline of geology has varying theoretical and methodological approaches to the study of the earth and its pro- cesses. By focusing on the ways in which geological theories are constructed and tested through the use of various methodologies, it be- comes evident that research approaches compete with each other to legitimate their own constructions of scientific knowledge. It also becomes clear that uncritical adherence to theoretical and methodological approaches and assumptions may obscure rather than illumi- nate objects under study. The scientific process in geology is exemplified by the hermeneutic circle (Frodeman, after Heidegger), in which empirical obser- vation, generalization and theorizing (induction), are followed by construction of hypotheses (including models) and renewed observa- tions to test and refine or abandon a theory (deduction). Ideally, this is a continuous and circular process, whereby theoretical assumptions are put to the test, but history demonstrates that the inductive and deductive approaches have largely been followed by dif- ferent groups of stratigraphers with different objectives. Further, these stratigraphers have tended to work in isolation from each other. According to Hallam, "geologists tend to be staunchly empirical in their approach", but are also inveterate model builders, at- tempting to explain their universe by developing deductive models. Two contrasting case studies illustrate empirical and model-based approaches to dating and correlation. A synthesis by Callomon (1995) of Jurassic ammonite biostratigraphy, based on a century of data collection, and the inductive building of a biozone scheme, reveals numerous gaps and considerable local stratigraphic variability in the studied sections in southern England. By contrast, a comparison by Gale at al. (2002) of two sections in India and France using a se- quence model for correlations was interpreted by them in terms of global uniformity of sequence-generating processes and eustatic sea-level control. Modern dating methods should be rigorously empirical, including the cross-correlation of multivariate dating techniques and the use of non-events as boundary markers ("golden spike" concept). Extreme caution needs to be employed in introducing such deductive concepts as "global cycles," "event stratigraphy" and "cyclostratigraphy" into methods of high-resolution chronostratigraphy. Cur- rently, cyclostratigraphers have established the Milankovitch model of orbital forcing as the centerpiece of a research program to docu- ment climate change and to provide a basis for a high-resolution time scale. There are at least three problems with this approach: 1) Researchers downplay the probability that orbital frequencies may have differed in the geological past. 2) There is a tendency to make assumptions about stratigraphic completeness and constancy of sedimentation rate that may not be valid. 3) Independent chrono- stratigraphic calibration of cyclostratigraphic data is insufficiently precise, and cannot at present provide adequate constraints on cyclostratigraphic models based on tuning, filtering, and other statistical techniques.
