For example we might try to explain why the cuban revolution took place. This would be different from simply recounting the story of what happened. We would be trying to pin point what it was that coused the revolution at that time and in that form.
Descriptive social science can be likened to journalism.
Explanatory is best seen as trying to explain why a given event happened.
Natural Science is generally seen as being concerned with generating predictive laws - if two chemicals react by exploding under certain conditions we would expect that we could repeat the reaction by replicating the experiment.
Most controversially some social scientists try to equal science by formulating general laws which allow prediction. Just how controversial this is depends on what we are looking at. Economics depends heavily on laws of how markets work such as supply and demand. These are often presented as certain laws in the same vein as scientific laws.
It becomes much more controversial when we start to try to make laws about such things as crime rates and voting intentions. Obviously social science is at a great disadvantage to natural science in trying to generate such laws. In natural science it is possible to run repeated experiments in controlled laboratory conditions. In social science this is hardly ever possible.
For example you can test whether water is made up of hydrogen and oxygen by following whatever scientific method is appropriate. You can repeat this process many times. It is often argued that this ability to repeat procedures to confirm them is a hallmark of science.
You can’t restage events in social science. You can’t restage the Cuban Revolution or experiment with the causes. Even relatively simple and repeated phenomena such as people voting in an election always take place under different conditions. For this reason laws in social science are always generalisations and based on probabilites rather than certainties.
Nobody disputes that you can make educated predictions about society. However there is a split between those that believe that it is logically possible, if practically impossible, to make perfect predictions given perfect information and those that believe this is a key difference between the natural and social sciences. This is a debate that we may go into in the seminars although it is a hotly disputed and extremely complex philosophical debate.
Causation
“The relationship between two events or states of affairs such that the first brings about the second”.
If we are to fully understand an event or make predictions we need to understand its causes. If we know that event one caused event two then if event one should occur again we might expect event two to occur again, all other things being equal. Of course in social science things are rarely linked so strongly. We might say that some kind of personal problem like unemployment is often a cause of alcoholism, but you are unlikely to find something that invariably leads a person to alcoholism. Social science is multi-causal. However that doesn’t mean that causes do not exist and that they can’t be found out.
Three Causality Criteria.
1) An association between “a” and “b”.
2) A particular time order - “a” should preceded “b”.
3) Shown that the association between “a” and “b” is that they are not both caused by some third variable.
e.g. When you strike a match you see light and feel heat, but one does not cause the other, rather they are both products of a third factor - the chemical reaction on the match.
In social science we label different concepts as variables. So for example, unemployment rate, inflation, your social class, age, sex, race, etc. are all variables.
Social science often tries to explain one variable by reference to another.
For example, it might be argued high unemployment is linked to high crime rate. This is equivalent to saying unemployment causes crime.
For something to count as a cause of something else as a minimum condition it should accord with the three causality criteria given above.
An association between them mean that where one occurs the second one must tend to occur. Because social science does not demand a perfect relationship how often they occur together is variable. We use statistics to decide whether the incidence of their joint occurrance is significant or not.
A particular time order occuring between them means that we expect the cause to precede - or happen before - the cause. If we discovered a link between alcoholism and unemployment - that unemployed were more likely to be alcoholics than the employed - it wouldn’t be clear which was the cause and which the effect.
Alcoholism could have caused the person to lose their job or loss of the job could have driven the person to drink. If we examined which of the two factors or variables was occuring first this would give us an idea which one was the cause and which one the effect.
The third causality condition requires a check that the relationship between two variable is on of real cause and effect. Just because two things occur together doesn’t always mean one cause the other. Often it may be that both are effects of a third factor. For example it may be that people who have poor diets also tend to be people that do not have a car. Now obviously not having a car doesn’t cause you to have a poor diet or vice versa. But it might be that both are linked as effects of not having much money.
Similarly voting Conservative doesn’t cause you to buy the Daily Telegraph or vice versa but I’d be willing to bet that the two are statistically linked. What these examples show is that you can’t simply look at statistical patterns in data and draw results from that. You have to think about what possible causes and effects are in what you are looking at. Now sometimes these might be be self explanatory, but in other cases it will require more thinking about. Sometimes it will be up for debate.
This is a point to mention if you answer the essay question on objectivity in social science.
Even in using statistics we do not eliminate subjective human choice. However, if you think about it that argument could also be made about natural science.
Variables are what we call the different measures we are using. If you think of doing a simple questionnaire which asks you your age, your job and where you are from.
That would have three variable - age, job and location. Each person could vary on each variable. The name of the actual variables would remain the same but the answer or value on each would differ. What you would look for is whether any of the variables display any covariation. In other words whether there is a statistical link between the answers on one variable and the answers on another.
e.g If I asked everyone their age and where they were from I might find that people from the North tended to be older. Assuming I abided by certain statistical conventions to do with things like the size of the sample I asked then I justifiably argue that I had evidence to show people in the North tended to go to university younger than those in the south. It is more complicated than that but that is the basis of it.
This is shown in the graph below.
Science
The term Social Science implies that these disciplines are part of science. The main questions to ask are:
-Is this true?
-Does it matter?
There is a debate about whether social science is scientific.
If you answer the first essay question then this this is something that you’ll have to address. Scientific method is said to be objective. The opinions of the researcher do not interfere with the research. Although this is questionable for natural science. Many argue that this is beyond social science.
For example you’ll often see people claim to be using objective research to make a political point. e.g. Community Power Debate -
USA 1950s through to 1970s.
One side argued that America had become governed by an upper class. Based this analysis of the backgrounds of the top political, industrial and military figures. All had the same background and mixed socially. From this they argued that power was concentrated in a ruling elite (there is a bit more to it than this). This was taken very seriously in the USA where they were very proud of their democracy and people were probably less cynical about politics than today.
This argument was not left unchallenged. A group of political researchers - most notably Dahl and Wolfinger - decided to examine how actual political decisions were made in an American city. Rather than studying the background of decision makers they looked at how key decisions were made by observing council meetings and so on. They found that the people with most say were elected officials and groups with an interest in the issue being decided. They concluded that democracy was alive and well.
Throughout this debate the conflict centred on what was the right research approach to study where power lay. The conclusions you come to would depend on what you saw as the proper way of studying power. The problem is because things like power are absract concepts which are difficult to pin down.
Now there are two ways of looking at this debate. One is that these researchers objectively studied what the right methodology would be and then came to their conclusions on the state of American democracy. However, I doubt that is what actually happened. We all have opinions about the society we live in. I think more often than not social science researchers choose a research method to at least partially suit their opinions. So whilst there is an element of finding out in research - there is also usually an element of trying to prove a point or an argument.
Now in natural science this seems less likely. Whilst chemists might want to prove certain theories etc there is much less scope for them to choose their own research method and so on. Therefore scientists often argue social science can’t be scientific. They argue they deal in facts and soc. scientists deal in opinion.
Now it might be argued so what, why should we be scientific. However most people do accept that you have to have some objectivity in research. I mean people can’t get away with things just because they argue for them - they have to accord with accepted practices - it is just that what is accepted practice in social science is less defined than in natural science.
How did social science come about. It used to be Science and philosophy. Science was about what is and philosophy would deal with ethics and what ought to be. Social science started in the 19th century and can be seen as a mirror to industrialisation and the rise of science in society. The idea of planning society or explaining society gained popularity as the importance of economics became clear both for workers and employers. Marxism is a supposed scientific approach to explaining the condition of the workers in terms of the stage of economic / industrial development.
Soc. Science took a step forward with the advent of the questionnaire in USA. One striking use was in the post war Labour election victory. It was widely thought Churchill would win as the war leader. The first electoral survey predicted a Labour win but was ignored until Labour did win.
Since then it social science has developed. There are still those that think you should try to mirror natural science - statistics etc, But most people accept that interviews, documents, observation etc are equally valid research tools. The point is that you must be able to justify the tool / method you use - eg - community power debate.
So what is it to be scientific?
Normative - relying on intrinsic value judgements.
Science is generally seen as non-normative. It is explaining the world as it is rather than making a judgement on how it should be.
Cumulative, science is usually seen as building on previous understanding and adding to a stock of knowledge.
Science doesn’t just make generalisations from a mass of data.
Two views of how scientific theory develops.
1- Deductive, theories are developed and they are then tested against observations/empirical data. (Karl Popper)
2- Inductive, theory is developed from looking at the empirical evidence.
Either view requires that theories are subject to testing by empirical evidence (observations about the world).
It is controversial as to where the boundaries of natural science are.
However, it is generally accepted that if we can specify the causal chain accurately then this is what constitutes explanation.
How Social Science may differ from science
1) Social behaviour is rule governed, to understand it we need to do more than say “a” causes “b”, we need to understand the rules/culture of the society to fully understand what we see. (Peter Winch)
Example: If a man from mars were to observe traffic lights he would know that a red light stopped cars etc. But that wouldn’t give him a real understanding of what was going on.
If Winch’s view is accepted then the methods of natural science are incapable of giving full understanding in the social sciences.
