Social Science is Explanation or it is Nothing | LSE Event

[@mariomarais5644]

Very interesting! Would love to social scientists being more involved in the domain of ICT for Development.

[@Samgurney88]

My background was in analytic philosophy. It’s a shame logical positivism became so unfashionable in the wider academic culture, largely owing to rather technical difficulties involved in precisely formulating its doctrines. There is nothing fundamentally wrong with the project of attempting to unify the sciences or to use ‘the’ scientific method as the gold-standard form of knowledge. There are a variety of philosophically sophisticated options available - Lakatosian ‘research programs’, Popperian critical realism, Quinean neopragmatism, etc. Basically there is nothing wrong with the general trend of physicalism and so-called ‘reductionism’, and the youngest scientific disciplines would generally benefit from researchers thinking more strictly along these lines rather than less. It is not healthy for the social sciences to be unreflective cargo-cult sciences, aping scientific forms without providing any real explanatory success. If it would be premature to go completely for the scientific ideal then perhaps it is best to be honest about that fact. There will always be a respectable place for good scholarship in the humanities, but, by definition, this is not what the social sciences aspire to. And it is unduly pessimistic for social scientists to abandon their mission as hopeless. Empirical psychology and economics, for example, are already more successful than many of their critics allow. Whilst humility is certainly in order, it is not healthy for social scientists to veer to the opposite extreme of sinking themselves into a methodological quagmire. There is no need to reinvent the wheel - the tools of scientific inquiry are well-developed, and social scientists can and should always look up to the more mature sciences for guidance and material. What I would like to see as an outsider is more robust empirical testing in the social sciences. Preferably I would like to see more well-defined models, forecasting of measurable variables, and objective ranking of models by forecasting success. If models in, say, International Relations cannot be compared in this way, then what do the models even mean? There are too many theories in the social sciences that are ‘not even wrong’, or which explain everything but predict nothing. The phenomena with which the social sciences deal are extremely complex ones, about which we can only ever be very imperfectly informed. Any workable models will necessarily involve great simplification, and therefore will be incapable of great predictive accuracy or precision. But none of this should allow social scientists to forget that predictive success is the hallmark, the sine qua non, of the scientific method. If social scientists cannot show predictive success greater than that which can be attained using informed common sense, then it is not science but political commentary.

[@mrsothil8589]

Such very insightful comment. My discipline is in physiological Ecology : we trying hard to explain natural phenomena using reductionist approaches and apply our science to shaping our world becoming more and more mono culture. This ia the root cause of ecological tragedy at global scale (de facto we only busy with producing scientific explanation for whatever interest of any political economy motif)

[@rayturchie7038]

Passengers...