Ebach Malte C, Michael Michaelis S
Palaeontology, Geobiology and Earth Archives Research Centre, School of Biological, Earth and Environmental Sciences, UNSW, Kensington, NSW, 2052, Australia.
School of Humanities and Languages, UNSW, Kensington, NSW, 2052, Australia.
Acta Biotheor. 2016 Sep;64(3):241-62. doi: 10.1007/s10441-016-9282-3. Epub 2016 Jun 30.
Changes in the methodology of the historical sciences make them more vulnerable to unjustifiable speculations being passed off as scientific results. The integrity of historical science is in peril due the way speculative and often unexamined causal assumptions are being used to generate data and underpin the identification of correlations in such data. A step toward a solution is to distinguish between plausible and speculative assumptions that facilitate the inference from measured and observed data to causal claims. One way to do that is by comparing these assumptions against a well-attested set of aspects of causation, such as the so-called "Bradford Hill Criteria" (BHC). The BHC do not provide a test for causation or necessary and sufficient conditions for causation but do indicate grounds for further investigation. By revising the BHC to reflect the needs and focus of historical sciences, it will be possible to assess the cogency of methods of investigation. These will be the Historical Sciences Bradford Hill Criteria (HSBHC). An application to one area in historical science is used to demonstrate the effectiveness of the HSBHC, namely biogeography. Four methods are assessed in order to show how the HSBHC can be used to examine the assumptions between our data and the causal biogeographical processes we infer.
历史科学方法论的变化使其更容易受到被当作科学成果的不合理推测的影响。由于推测性且往往未经检验的因果假设被用于生成数据并支撑此类数据中相关性的识别方式，历史科学的完整性正受到威胁。迈向解决方案的一步是区分合理假设和推测性假设，这些假设有助于从测量和观察到的数据推断因果关系。做到这一点的一种方法是将这些假设与一组经过充分验证的因果关系方面进行比较，比如所谓的“布拉德福德·希尔标准”（BHC）。BHC 并非提供因果关系测试或因果关系的必要充分条件，而是指出进一步调查的依据。通过修订 BHC 以反映历史科学的需求和重点，将有可能评估调查方法的说服力。这些将是历史科学布拉德福德·希尔标准（HSBHC）。在历史科学的一个领域的应用被用来证明 HSBHC 的有效性，即生物地理学。评估了四种方法，以展示 HSBHC 如何用于检验我们的数据与我们推断的因果生物地理过程之间的假设。
