A New Look at the Left Wing

More generally, qualitative methods such as interactive interviews and ethnography are necessary to abstract the causal mechanisms of which quantitative/statistical methods are oblivious. It should not be expected that these abstract causal mechanisms can explain events directly without any need for empirical research into the contingency of the concrete. To do so is to commit the error of ‘pseudo-concrete research’ that is common in radical structuralism such as Marxism (Sayer, 1992). Quantitative methods, on the other hand, are particularly useful to establish the empirical regularities between objects. Although these concrete regularities are not causal relations, they can inform the abstraction of causal mechanisms. Quantitative methods are also useful in drawing attention to the external and contingent relations between objects. Inferential statistical analysis can throw light on, for instance, the external relations between objects (e.g., employment and poverty) in society from a sample. One should bear in mind that these statistical generalizations are only ‘universal’ at a specific temporal-spatial intersection. A serious problem of reductionism is incurred if one attempts to treat these contingent generalizations as necessary causal mechanisms.