1. Introduction

Reporting failures is not a major concern in science. Research leading to failed or inconclusive results ends up most often unpublished, known only to those who did the work and quickly forgotten even by them. A notable exception is 'The Journal of Negative Results in Biomedicine'. In Quaternary palynology, while it is not unusual that the analyst is unable to extract pollen from sediments, the great bulk of unsuccessful work never sees the light of day and, as a consequence, the available palynological literature exhibits a misleadingly positive impression of success. In an interdisciplinary context, where the results of pollen analysis are often of interest to investigators who are not themselves palaeobotanists, such as archaeologists, palaeoclimatologists, palaeontologists and environmentalists, it is especially important that an understanding of the limitations of pollen analysis is widely disseminated. However, few non-palaeoecologists seem to be sufficiently aware of something as obvious as the fact that not all sediments are suitable for pollen analysis (Bryant and Holloway 1983; Carrión et al. 1999a; López-Sáez et al. 2003; González-Sampériz 2004a). Often they may require or hope for palaeobotanical information from locations or deposits that are ill-suited to the preservation of pollen, or where the interpretation of the results is complicated by issues of differential preservation and taphonomy. Conversely, palaeobotanists may ignore deposits of potential interest or importance on the grounds that they are unlikely to produce adequate pollen, even though the results could be of great significance at a local scale or in relation to the problems investigated by other disciplines. Even palynologists who concentrate on the 'good' sites are at risk of wasting precious time and resources repeating pollen analyses of unproductive materials, since sites assumed a priori to be good sites may turn out to be sterile or partially so, just as sites assumed to be 'bad' may turn out to offer useful information. Applications for Quaternary palaeoecological research projects are often rejected on the grounds that they ignore earlier endeavours, but if the earlier endeavours are not published, everyone is at risk of wasting time and effort reinventing the same wheel and nothing is learned from earlier failures.

This article originates from an explicit commitment to report failures with pollen analyses for the Quaternary of the Iberian Peninsula. 'Failure' is understood here as the inability or impossibility of obtaining palynomorphs after following the usual extraction methods. A few cases include pollen spectra where the absolute number of pollen recovered from a sample was not sufficient for statistical treatment and interpretation. It is worth emphasising that we do not refer to unexpected or conflicting evidence or to results that are negative in the sense that they do not support hypotheses from designed experiments, nor do we mean those results unable to disprove a null hypothesis. Our viewpoint is that, in spite of the difficulties of interpretation, failed pollen analyses will, sooner or later, be incorporated explicitly into the concepts of the discipline and its research procedures (e.g. Leroy 2008). In fact, a well-designed project should never produce a completely negative result, since there is always the opportunity to learn something. Learning about failures as well as positive results can be instrumental in providing the context for the development of new research strategies and so lead to a better return for public and private funding. Conversely, hearing only about the successes is equivalent to throwing away half of the information, and may give a misleading impression of the opportunities and limitations of pollen analysis. This issue of failure, or the production of unexpected results, is therefore of importance both to the non-specialist consumer of pollen results and the specialist palynologist, and our review is aimed at both types of reader.

PREVIOUS NEXT CONTENTS HOME