ABSTRACT: Defending and testing hypotheses: a response to John Shaw's paper In defence of the meltwater (megaflood) hypothesis for the formation of subglacial bedform fields.

CGRG Bibliography of Canadian Geomorphology
Search Results

Author :

Date :

Title :

Publication :

Issue :

Page(s) :

Abstract

John Shaw’s recent paper in the Journal of Quaternary Science (Shaw, 2010) discusses at length the details of two of our papers (Evans et al., 2006, 2008) on the glacial geomorphology of western Canada. The interpretations and counter-arguments ensuing from this research have been rehearsed in a number of previous publications (Clarke et al. 2005; Benn and Evans, 2006; Shaw and Munro-Stasiuk, 2006; O´ Cofaigh et al., 2009) and therefore I will not enter into more reviews of the details in this reply. It is simply my intention to restate here, for the final time, some fundamental problems with Shaw’sapproach to the research problem. According to Shaw (2010), the detailed glacial geomorphology presented by Evans et al. (2006, 2008) is merely a ‘backcloth’, because it does not pertain to the most recent event (i.e., the flood). This implies that glacial geomorphologists, specifically those working in western Canada, must view the glacial geomorphological evidence at the outset in terms of being created by floodwater at the close of glaciation. Shaw proposes that we miss the point by studying the internal structures of glaciallandforms, because they have all been truncated by the flood. Moreover, Shaw clearly revokes any notion that subglacial deformation operates at the base of glaciers despite widespread and consistent measurement of such a process (see Clarke, 2005, for review). The role of subglacial meltwater in the dynamics of glaciers and the formation of subglacial landforms is widely recognised, but is a body of knowledge that parallels and interfaces with that of subglacial deformation; why does Shaw champion one and totally exclude the otherin his interpretations of glacial landforms? With respect to previous debates on scientific methodology, Shaw (2010) states: ‘Evans and others argued that the meltwater model is not a proper hypothesis because it cannot be falsified . . . seem to have forgotten this argument when they claimed that cross-cutting fluting negates the hypothesis’ (Shaw, 2010, p. 6). It is important to again reiterate that an unfalsifiable hypothesis is not one for which others cannot find potential tests, but one that is ‘defended’ by its proponent by constantly creating ad hoc protection devices. Science moves forward by increments, but ‘outrageous’ geological hypotheses occasionally shock the system into new and more vigorous avenues of enquiry. The subglacial megaflood hypothesis is clearly regarded by its proponents as a prime example, but it is counter-productive to ‘defend’ any hypothesis. As glacial geomorphology relies heavily on process–form relationships it is crucial that we always adhere to the axiomthat the simplest interpretation is best. Additionally, the interpretation of Quaternary glacial landscapes is soundly based in the identification of modern analogues and therefore the application of modern Icelandic analogues to Canadian Quaternary landscapes is merely an example of identifying the simplest and most appropriate interpretation. We could also cite the emerging literature that reports observations of subglacial bedforms being created, without the assistance of a megaflood, at the base of the Antarctic Ice Sheet (King et al., 2007; Smith et al., 2007). In his conclusions, Shaw states that ‘Testing a hypothesis amounts to assessing how well observations and interpretations fit its proposals’. To the contrary, testing a hypothesis involvesthe identification of observations that have the potential to prove it wrong. The hypothesis then survives or is rejected based upon the outcome of the observations. Advocates of the subglacial flood hypothesis, rather than hypothesise, repeatedly champion an explanation that is unquestionablyconsistent with their chosen observations. More significantly, John Shaw and colleagues have promoted the use of ad hoc protection devices to continually modify their hypothesis so that it is always consistent with new sets of observations. I argue that one cannot selectively ignore observations and interpretations made by others that are considered irrelevant (i.e. ‘determining the environment of deposition of glacilacustrine sediments; assessing the genesis of stratified tills; interpreting glaciotectonic structures as syndepositional or postdepositional, subglacial or ice-marginal’; Shaw, 2010, pp. 249– 250), simply because they do not fit with John Shaw’s particularexplanation. This type of (outrageous) science is, to my view, best described as hypothesis defending, not testing.

Bibliography of Canadian Geomorphology