Meta-source

The concept of a meta-source is developed by A. W. Carus and Sheilagh Ogilvie in “Turning Qualitative into Quantitative Evidence: A Well-Used Method Made Explicit,” The Economic History Review, 62(4):893–925, 2009. https://doi.org/10.1111/j.1468-0289.2009.00486.x

Page 903:

The raw data extracted directly from the sources is compiled or organized into a loose classification system, a meta-source. (903) s Page 910:

… transforming qualitative sources into a community-level interlinked ‘meta-source’ which could be interrogated quantitatively over a series of decades by historians with very different theoretical preoccupations.

Page 912:

the concepts of the meta-source were sufficiently close to the data, sufficiently neutral, that they could be used to address questions radically different from those for which it was originally compiled.

Page 913:

The ‘units’ to be counted are not clearly suggested by the physical arrangement of the source itself; therefore, their identification requires conceptual intermediation—it requires interpretation.

Page 914:

If certain basic parameters from the work database matched up to what was known from the ‘naturally’ quantitative sources whose concepts require less interpretation, it could then ‘borrow’ its robustness and representativeness from those sources; its reliability would be ‘guaranteed’ by them. In itself, this kind of internal cross-check among different sources is nothing new … In compiling his meta-source, Smith reflected critically on the processes that generated the data … carefully considered whether one sort of transaction was more likely to be recorded than the other, whether the frequency of court sessions was changing over time, and whether incomplete survival of court records was likely to bias the results. Given the inherent indeterminacy of some qualitative data … Smith stated his results as lying within a range of values …

Page 915:

… the linking of data from many different sources to individual names also made it possible to check different sources against one another …

Pages 920–921:

… the meta-source compiled for a particular community can be tested for reliability and representativeness, to ensure that the quantitative cross-checks on the historian’s initial, subjective response to the sources in their qualitative aspect are as robust as possible. Of course, nothing will ever guarantee that the data are absolutely reliable, just as no particular observation in biology or physics is ironclad; we are not dealing in certainties. And in history—why deny it?—we are considerably less certain than in theoretical sciences. History is messy. But this is no more reason than in medicine not to do as well as we can. Applying quantitative cross-checks to our qualitative, hermeneutic, holistic grasp of the documents minimizes our chances of going drastically wrong.