Update: Stemming “ignorance creep” in paleoecology and biogeography

Abstract

ISSN 1948‐6596 news and update update Stemming “ignorance creep” in paleoecology and biogeography The continued success and relevance of any scien‐ tific field depends on critical examination of its foundational knowledge. A recent perspective in Quaternary Science Reviews by Steve Jackson (U. of Wyoming) highlights the different ways that knowledge is formed, recognized, and then lost in paleoecology (Jackson 2012). Although he was writing for a specific audience—paleoecologists— his call for greater efforts to acknowledge and combat “ignorance creep” applies to all research‐ ers, including biogeographers. Jackson (2012) uses an unlikely pairing of quotes by Donald Rumsfeld and Henry David Tho‐ reau to highlight the scope of our knowledge about any given discipline. He presents an episte‐ mological framework for scientific understanding: knowledge versus uncertainty on the one hand, and cognizance versus ignorance on the other. These contrasts generate four categories: known knowns and known unknowns, i.e., knowledge and uncertainty of which we are cognizant; and un‐ known knowns and unknown unknowns, knowl‐ edge and uncertainty of which we are ignorant. The focus of Jackson (2012) is on the often‐ ignored category that arises from this classifica‐ tion scheme—unknown knowns. Unknown knowns include “the hidden and unquestioned assumptions that underlie a discipline, the things so seemingly obvious that they are beyond ques‐ tion or reflection, all the things that are routinely taken for granted...” (p. 3). Jackson is particularly concerned about “ignorance creep”, the process that converts “knowns” into “unknowns” and ar‐ gues for increasing attention to forward models to identify and combat ignorance creep within pa‐ leoecology. Forward models trace the mechanis‐ tic processes along the path from the target vari‐ able to the proxy used to measure it. Jackson ar‐ gues that the construction of forward models serves two purposes: 1) to make processes and assumptions explicit, minimizing ignorance creep, and 2) to allow quantification of uncertainty, thereby assessing the strength of resulting infer‐ ences. Jackson uses woodrat middens as a study system to provide a worked example of forward model construction for paleoecological inference. The forward model for this system starts with a variable of interest, regional vegetation, and out‐ lines the path through data collection and analysis to the resulting inferences a researcher might make about regional vegetation. In between those two points lie a set of assumptions, observa‐ tions, and models, dealing with the ultimate source of the data (the regional vegetation), the animal vector bringing information about the source into the middens (decisions and assump‐ tions made by or about the woodrats), issues with diagenesis (physical, chemical, and biological proc‐ esses that transform the samples over time), and finally the set of analytical and inferential issues that arise in field, lab, and computer work. This forward model (and most forward models that paleoecological studies would generate) essen‐ tially recapitulates the field of taphonomy. Ta‐ phonomy is a component of paleoecology that studies how the processes of preservation affect the information found in the fossil record (Behrensmeyer et al. 2000). However, when de‐ fined more broadly, every field grapples with “taphonomic” processes, whether in the field or lab, that affect inferences made from the raw data. Every field has, at the very least, source is‐ sues and analytical and inferential issues; they may or may not have vector or diagenesis issues. Jackson argues that forward models are important when data, originally collected for one purpose, are used for another. Distribution mod‐ eling is a good example of this. Often, the original data (e.g., occurrences of individuals at particular localities) were collected for a very specific pur‐ pose: to determine some aspect of the ecology and evolution of a particular group at a particular location. Today, however, data collected for dis‐ parate purposes, with different collection meth‐ odologies, are aggregated to understand a new set of questions, at much broader spatial and tem‐ poral scales (Boakes et al. 2010): What was the frontiers of biogeography 4.3, 2012 — © 2012 the authors; journal compilation © 2012 The International Biogeography Society