Is sociology a real science?

Look, it depends on what you mean by "science".

"Science" comes from the Latin scientia, which just means "knowledge" or "the things that are known". And this gives rise to a very wide concept of "science", in which the word can refer to any organised, systemised field of study. This was the orginal sense of the word in English. In this sense, almost anything could be (and frequently was) called a science - music, mathematics, logic, rhetoric, theology, philosophy, astronomy, geometry, ethics - anything. What made something a "science" was not its subject matter, but the degree of system and organisation you put into studying it.

In time there grew up a broad distinction between the sciences on the one hand and the arts on the other. A science was something you pursued by reflecting, observing, theorising, classifying, describing; an art was something you pursued by practice, engagement, activity. But this wasn't an entirely satisfactory division; there was a large overlap. Music theory was a science, but music performance an art; studying the grammar of a language was a science but studying the literature of the same language was an art; etc. And learning to do something, including something quite technological, was an art. So even an expert clockmaker, for example, was an artisan; the design of clockwork, no matter how much mathematics and measurement was involved, was an art, not a science. An apothecary making and administering drugs, or a surgeon dressing a wound, was not a scientist, however learned either of them might be. The application of knowledge to do something useful, whether or not the knowledge was scientifically-derived, was not a science; it was an art.

So, in time, thinking about science became further refined again; it wasn't just a matter of reflecting, theorising, describing, etc, but of using a particular formal method do do so; the "scientific method". The scientific method involves (a) making empirical observations; (b) forming hypotheses to explain those observations; (c) making predictions based on those hypotheses; and (d) designing and implementing experiments which will test those hypotheses, the outcome of which will either refute or tend to confirm the correctness of the hypothesis. In this scheme of things, technological fields of study (like pharmacy, medicine or engineering) can be sciences if pursued through the scientific method.

This gives rise to a sense of "science" in which the stereotypical sciences are things like physics, chemistry and biology.

Then you have the social sciences which deal with human behaviour - sociology, economics, pyschology, etc. These are seen as less "sciencey" because (a) it is usually not posssible to make empirical observations with the same degree of precision as is possible in physics or chemistry, and (b) it is difficult for a variety of reasons (some of them ethical) to design and implement experiments which will test the hypotheses you form. Obviously these are fields in which you can pursue your studies with more or less reliance on the scientific method, but the main point is that the scientific method is less useful in these fields than it is in physics, chemistry, etc; it delivers less certain results.

And then you have fields that study abstract or imaginary concepts - logic, mathematics, ethics - which (in this scheme of things) cannot be considered science at all, because it's simply impossible to make emperical observations of non-empirical entities, or to devise and implement experiments on them. Thus whatever mathematicians are doing, it's not science.

Depends if you are using qualitative or quantitative research.

Plus what about subjects like legal science, is that a science.

Peregrinus wrote: »

It's a social science, in that it studies how people behave in the real world, it hypothesises about why they behave like that, and then it tests the hypothesis by trying to alter real-world conditions (specifically, laws and the policies or actions of government) in a way that will lead them to behave differently.

And, at a higher level, it tries to devise and maintain a consistent and coherent mechanism for doing this regularly and effectively.

It's a practical application of the scientific method.

well then how to account for the likes of https://en.wikipedia.org/wiki/Ronald_Dworkin or https://en.wikipedia.org/wiki/John_Rawls

Both highly influence in ideas around using the law to achieve social justice.

Legal studies is open to multiple interpretations like all social science.

Plus how do you define fairness?

Peregrinus wrote: »

It's a social science, in that it studies how people behave in the real world, it hypothesises about why they behave like that, and then it tests the hypothesis by trying to alter real-world conditions (specifically, laws and the policies or actions of government) in a way that will lead them to behave differently.

It's a practical application of the scientific method.

The problem with the real world is that you can't isolate all the initial conditions. Or even most of them. You can try, but only at at the expense of overpowering tyranny.

Let's say that an initiative to combat violence at sports grounds was undertaken recently after numerous studies of historic data, focus groups, data analysis, complex modelling of various parameters using computationally intensive mathematics that would require a hefty computer to process, findings calculated, new hypotheses to test all set and ready to run in say May this year and then.......Covid lockdown!!!!

Who could have predicted that?

All of a sudden we have to prepare for other examples of antisocial behaviour/crowd control which is likely to emerge as the year inches on with lockdown restrictions still in place and young people getting highly frustrated.

All previous models highly suspect because initial conditions are utterly changed.

A social scientist will ALWAYS have to explain tomorrow why the predictions they made yesterday didn't come true today.

Candie wrote: »

The science test in simple terms, applied to math by Berkeley.

https://undsci.berkeley.edu/article/mathematics

Let us assume, for the moment, that all horses are spherical...

mariaalice wrote: »

well then how to account for the likes of https://en.wikipedia.org/wiki/Ronald_Dworkin or https://en.wikipedia.org/wiki/John_Rawls

Both highly influence in ideas around using the law to achieve social justice.

Legal studies is open to multiple interpretations like all social science.

Plus how do you define fairness?

I didn't use the term "fairness"; why are you asking me to define it?

Think for a moment about technological subejcts like engineering or pharmacology. You'd accept those as fundamentally scientific fields, right? But there's also a non-scientific dimension to them; you can apply scientific principles to build a better bridge or a beter gallows or a better atom bomb, but the decision to build the bridge, the gallows or the bomb not a scientific one, since it fundamentally rests on non-scientific value judgments. (Should we spend public money building the bridge, or improving schools, or not spend it and reduce taxes instead? Should we be hanging people at all?)

Once you move away from pure science and into technology you are inevitably muddying the waters in this way - the question of how to do X can't really be disentangled from the question of why to do X or whether to do X. Even the apparently technical question of how to do X can be subordinated to non-scientific considerations - e.g. ethical constraints on drug trials/testing in pharmacology.

But if this doesn't stop engineering or pharmacology from being science, then it equally doesn't stop the social sciences from being science. "Should we reduce car thefts?" is not a scientific question. But if you decide that you do want to reduce car thefts, then you can certainly use scientific methods to identify which legal and/or policy measures are likely to be more effective to
reduce car thefts, and which are likely to be less effective.

A Tyrant Named Miltiades! wrote: »

quote="riffmongous;114222587"]
It does.

No not wrong, with maths I can set up a space with any number of dimensions that I choose, I can choose different different types of geometry etc. [/Quoye]



No offence but you sound like one of those people who want to talk about maths and have absolutely no idea of what they're saying.

I actually cannot make sense of this. Yes I suppose there are different "types" of geometry if you want to use that langauge, but "A follows B within the logic I'm using? What? This doesn't make sense.

these are 'alternative facts'!

By any metric or definition of what a science is, or what constitutes a scientific practice, sociology is certainly the least scientific of the sciences.

There is alot of work of dubious objectivity published. For example I once read a paper which set out to find "bias against women in schools", the data presented showed that there was in fact a bias against men, but the confusion of the paper then read along the lines of "we didn't find the bias against women here, it must be somewhere else". They went out to try to confirm something they wanted to believe, and when they found evidence to the contrary they ignored it.

I also think that there is a limitation on the utility of fields like sociology which apply to our actual lived experience: if there is result from a sciology paper which flies in the face of what you can readily observe in the world about you, then it's probably wrong; if there is some result which is in accordance with what you can see about you, then it is of little use.

I think it’s indicative that the social sciences have a real problem with the repeatability of study results. Different groups can run the same experiments and come back with markedly different results. Some researchers are explicitly trying to address this e.g. the Social Sciences Replication Project. In general, though: if a Sociology study makes some specific claim about people, I tend to assume that it’s tentative.

joe40 wrote: »

I don't doubt what your saying but that is a problem with the people carrying out the research, not whether the subject is a science.
I think we would all agree modern medicine is very scientific but in the early 20th some of the practices were far from scientific. That was a failing of the practitioners not the subject.
I think rigorous, scientific sociology, with careful guarding against bias, could be useful,

I don’t exactly disagree with you, but I still think repeatability of studies is a key metric here. Using your medicine example: those unscientific practices would not lead to usable results, but they take nothing away from those scientific practices that do lead to repeatable results. If anything, those unscientific practices tended to deviate from the purely medical into the sociological e.g. phrenology, the study of the shapes of heads as an indicator of intelligence. Ditto for Nazi doctors driven by ideology rather than pure science.

I have very little respect for sociology based on the little I've been exposed to (two terms at first year undergrad, the occasional listen to Thinking Allowed on BBC R4 and the occasional presentation at work by a colleague with a PhD in the field).

It seems highly driven by the subjective perspective of the researcher. In general, it seems to have very little ability to produce falsifiable statements that can be rigorously tested. The research I hear on that R4 often seems pretty low quality and of minimal social use.

Maybe I shouldn't rush to judgement on the little I've heard, but I'm yet to hear anything that has altered my impression.

Richard Feynman giving a very concise explanation of science.

https://www.google.ie/url?sa=t&source=web&rct=j&url=https://fs.blog/2009/12/mental-model-scientific-method/&ved=2ahUKEwizs9qd_4jrAhVkoXEKHeecBss4FBCjtAEwB3oECAAQAQ&usg=AOvVaw1G7QDL8tjGBDOiznhoncT8