How does social science differ from natural science?
Sunday, 30 Oct 2022, 21:47
Visible to anyone in the world
Edited by Samuel George Gaze, Saturday, 12 Nov 2022, 13:17
Thoughts from the Limitations of Social Research
By Marten Shipman, Professor of Education, Warwick University, 1981
"Science had once been seen as the progressive completion of the jigsaw puzzle that was to become a complete picture of the natural world... but in the twentieth century, science has come to be seen as interpretive, involving human explanation as well as collected facts."
Shipman's book is mostly talking to people who read about science rather than do science. But personally, I think we are all still influenced by the social representation of science as progress towards a "theory of everything." The problem with this approach becomes more acute as we turn the lens of science on ourselves.
"Social scientists have consistently opposed the attempt to use [natural science methods] for looking at human action... The natural world is not made by humans. The social world is by contrast made and remade as people interact with each other."
Part of this remaking is the interaction of people with the discourses of science. Science is part of the construction of our social world, and it is necessary for us to be able to critique the way we use science to talk about it. Obviously in such a system, there can never be a closing point where all the meanings are fixed; for a start, social science can't anticipate how it will affect the world it studies.
"Social scientists cannot merely observe behaviour, they have to find out how individuals give meaning to and organise their interaction with others ."
This is why Rom Harre (2010) argued that the social sciences must use a different 'grammar' than the one used in natural sciences. Important detail is missed if a description does not involve the construction of goal-directed behaviour in people and the social construction of their interactions.
"Theories are often explanations of the "if X, then Y, kind. In the natural sciences these usually take the form of laws. In social science there are no laws that hold across varied circumstances. But theories still aid explanation. It is likely that any relation will be stated in terms of probability: "if X then it is likely that Y."
The above points notwithstanding, it's not like the fact that social science theories cannot be treated as laws (though sometimes they are, especially in economics!) means that theories are of no value. At the very least, a theory is a simplification that shifts the conversation about a topic. Social constructionism itself is a theory that has prompted a lot of vigorous conversation and argument about identity that has changed society.
"Reliability in social science can often only be achieved at the cost of validity. Interpretation can be excluded by the rigid design of the investigation, but in doing so any relevance to everyday life is likely to be lost."
I must admit that I've never heard it put quite like this. The more attempts are made to control variables and produce generalisable results, the further away from reality it takes the conclusion. He puts it well: "The tension between generalising to approach the rigour of the natural sciences and trying to preserve natural human activity while observing them is inevitable in social science."
"Many social scientists reject this approach [of objectivity.] For them, objectivity is apparent, not real. The social scientist imposes their own preconceptions, and these may bear little resemblance to the situation as conceived by those actually involved."
Is objectivity possible in social science? Based on my very limited three year study, I'm not convinced this question has been addressed properly. There are at least two ways of doing it: one, aim to be as objective as possible by not allowing commitments to theories to shape your interpretation of the data, (an approach that is easily criticised as performing objectivity rather than being objective,) or two, identify and admit your position on a topic and biases make up part of the research (easy to do in theory, but in practice leaves a lot of space for unreflexivity, or even dishonesty).
Personally, I wonder if the point should be more about application than the abstract truth - saying from the outset what consumers of the study can expect to do with the information within. For instance, in natural sciences, the results of a paper on friction mean you can expect what has been observed to affect every applied instance of friction. Whereas a paper on the construction of male identity in fitness spaces means you can expect to see these discourses in the gym, and perhaps we can critique it and imagine different ways to be.
Thanks for reading! As an easter egg, I leave you with the most savage burn in the book:
"However much social scientists object to being categorised by the methods of the natural sciences, that is usually the basis of their claim when viewed by those who provide the money for research."
Shipman, M.D. (1981) The limitations of social research. 2nd ed. London: Longman
How does social science differ from natural science?
Thoughts from the Limitations of Social Research
By Marten Shipman, Professor of Education, Warwick University, 1981
"Science had once been seen as the progressive completion of the jigsaw puzzle that was to become a complete picture of the natural world... but in the twentieth century, science has come to be seen as interpretive, involving human explanation as well as collected facts."
Shipman's book is mostly talking to people who read about science rather than do science. But personally, I think we are all still influenced by the social representation of science as progress towards a "theory of everything." The problem with this approach becomes more acute as we turn the lens of science on ourselves.
"Social scientists have consistently opposed the attempt to use [natural science methods] for looking at human action... The natural world is not made by humans. The social world is by contrast made and remade as people interact with each other."
Part of this remaking is the interaction of people with the discourses of science. Science is part of the construction of our social world, and it is necessary for us to be able to critique the way we use science to talk about it. Obviously in such a system, there can never be a closing point where all the meanings are fixed; for a start, social science can't anticipate how it will affect the world it studies.
"Social scientists cannot merely observe behaviour, they have to find out how individuals give meaning to and organise their interaction with others ."
This is why Rom Harre (2010) argued that the social sciences must use a different 'grammar' than the one used in natural sciences. Important detail is missed if a description does not involve the construction of goal-directed behaviour in people and the social construction of their interactions.
"Theories are often explanations of the "if X, then Y, kind. In the natural sciences these usually take the form of laws. In social science there are no laws that hold across varied circumstances. But theories still aid explanation. It is likely that any relation will be stated in terms of probability: "if X then it is likely that Y."
The above points notwithstanding, it's not like the fact that social science theories cannot be treated as laws (though sometimes they are, especially in economics!) means that theories are of no value. At the very least, a theory is a simplification that shifts the conversation about a topic. Social constructionism itself is a theory that has prompted a lot of vigorous conversation and argument about identity that has changed society.
"Reliability in social science can often only be achieved at the cost of validity. Interpretation can be excluded by the rigid design of the investigation, but in doing so any relevance to everyday life is likely to be lost."
I must admit that I've never heard it put quite like this. The more attempts are made to control variables and produce generalisable results, the further away from reality it takes the conclusion. He puts it well: "The tension between generalising to approach the rigour of the natural sciences and trying to preserve natural human activity while observing them is inevitable in social science."
"Many social scientists reject this approach [of objectivity.] For them, objectivity is apparent, not real. The social scientist imposes their own preconceptions, and these may bear little resemblance to the situation as conceived by those actually involved."
Is objectivity possible in social science? Based on my very limited three year study, I'm not convinced this question has been addressed properly. There are at least two ways of doing it: one, aim to be as objective as possible by not allowing commitments to theories to shape your interpretation of the data, (an approach that is easily criticised as performing objectivity rather than being objective,) or two, identify and admit your position on a topic and biases make up part of the research (easy to do in theory, but in practice leaves a lot of space for unreflexivity, or even dishonesty).
Personally, I wonder if the point should be more about application than the abstract truth - saying from the outset what consumers of the study can expect to do with the information within. For instance, in natural sciences, the results of a paper on friction mean you can expect what has been observed to affect every applied instance of friction. Whereas a paper on the construction of male identity in fitness spaces means you can expect to see these discourses in the gym, and perhaps we can critique it and imagine different ways to be.
Thanks for reading! As an easter egg, I leave you with the most savage burn in the book:
"However much social scientists object to being categorised by the methods of the natural sciences, that is usually the basis of their claim when viewed by those who provide the money for research."
Shipman, M.D. (1981) The limitations of social research. 2nd ed. London: Longman