Introduction
Three teams of researchers from three different universities visit a village to study the factors determining hairstyle differences between men and women. All of them were motivated by the fact that men are much more likely to have shorter hair than women in almost all cultural settings. Their research question is same and they have the same research objectives.
But each of the team has their own kind of background. One team is stronger in regression modeling and surveys and another in case studies, participatory observation, discourse analysis and group discussions. The third team has a mixed expertise on both quantitative and qualitative methods.
Each of the team use different models to view the reality of hairstyle in this village. Obviously, they come to three different conclusions. They share their ideas in a conference. Each team claims that their own result is a bit better explanation of the hairstyle difference observed. Nonetheless, they all agree to take each others findings as partial truth and the audience thinks that the average of the three findings is closer to the truth than any one of them. All of them get a slap on their back.
This hypothetical case is a typical scenario of how research is done in social science. In this context, I will try to deal with two basic types of questions in this article:
- Is the truth about human-social reality (really) subjective? Isn’t there an ultimate truth in this field?
- Is this approach to research in social science leading us closer to the truth? Or, does an optimum research design exist to answer a pre-specified research question?
My answer to these questions are respectively:
- Ontology: Truth about human social reality is objective (“objective” does not mean static!).
- Epistemology: There exist an optimum research design to solve the pre-specified research question though the current technology may not allow one to apply this design.
In this article, an attempt is made to argue that the foundation questions of ontology and epistemology are same for both natural science and social science. In doing so, I assume that learning is possible only when the phenomenon (natural or social) is observable using current technology, at least indirectly. Questions about the unobservable (relative to time) phenomenon (such as Gods or ghosts) do not satisfy the Popperian criteria of falsification and will not be considered here. I will write more about it in my next piece (Does God Hypothesis satisfy Popperian criteria?).
Truth about human social reality is objective
What there is? Do numbers exist? Do numbers like 10/3 exist? Do we exist? Are we just the result of an algorithm, currently being run in a super computer? These are the types of questions that have been haunting mankind for centuries. Hundreds of millions of hours have been spent on thoughts, observation and discussions to know the fundamental principles of natural world. Philosophical discourse is unlikely to end despite the explosion of technology and “new” methods to measure discrepancy between myth and truth.
After the rise of post-modernist ideology in the mid twentieth century, objectivity and scientific certainty has been completely discarded in social sciences, unfortunately. The situation is even more obvious in the fields where technology is not advance enough to allow for measurement of important constructs and variables. Instead of incorporating measurement technology which is squeezing the scope of “interpretative” and “multiple-reality” paradigm, social science has been departing itself from it.
I argue that social world is the summation (huge and complex one!) of natural world. Once the law governing a single hydrogen atom falls apart, everything around us will instantaneously fall apart. Social world looks much more complex than natural world due to the rules of permutations and combinations among trillions and trillions of atoms that make the world we are in. Natural phenomenon are “simple” enough to be explained by “simple” models that have far fewer parameters. But the phenomenon in human-social reality is very complex and needs much more complex models to adequately explain them thereby giving us many more parameters to estimate and many more variables to measure.
There is only one (maybe dynamic and probabilistic) truth to be uncovered in human social science and natural science alike and some methods take us closer to the truth than others. Just because one is doing social science does not guarantee that the qualitative method is the best method to apply. Qualitative research methods are the necessity of the time when measurement technology is not advanced enough. One day, hopefully, we can measure love and perception in numbers through some kind of scanner and the need for so called qualitative method will disappear. Just a hope though!
Sam Harris says that there are two types of answers to any correctly asked question. Answer in practice and answer in principle. In principle, there is an answer for every research question that a social scientist can/may ask. But we may never have the answer in practice. For example: the question “How many cells do I have in my body right now?” has an answer in principle, but not in practice.
I would use an example of a shape to explain the true (if any?) difference between natural science and social science.
Figure : Models to explain natural and human-social world.
There is no difference between physical reality and human-social reality at fundamental level. Human-social reality is the summation of physical reality as all human and social reality will cease to exist once the physical reality is distorted. We are just clumps of atoms and energy. Social reality is just another level (maybe we can call it “higher level”) of physical reality.
Figure 1 is an example of how a simple model of circle can be used to approximate the round natural shapes such as planets found in the natural world. Just two parameters (center and radius) are enough to adequately compute the area of the irregular circular shapes. But the social world (like in Figure 2) is much more complex and simple mathematical models might not be able to adequately explain the shape and area of the phenomenon. We may never be able to adequately explain such phenomenon (answer not in practice, but answer in principle). But this should not make fundamental ontological and epistemological questions of social science different from natural science. Of course, the measurement technology gives us the methods and the appropriate methods can be different in explaining different observable phenomenon of our interest. To measure the length, we might have a “good enough” scale (all scales are imperfect). But to measure perception, we might only have a recorded interview or a Likert scale, which are going to be vague at best.
Human-social world is a part of natural world and deserves the same level of analytical and methodological rigor as natural science is getting. Provided that the technology exists to make measurements of human-social constructs, the same tools that have been proven to work in natural science and engineering, should be applied to human-social science.
What about the observation effect in social science?
Many claim that social world is so fragile that the presence of a scientist or observer changes the social world itself. Imagine doing an ethnography where an outsider researcher spends time with a community to understand their behavior. They are likely to adjust/adapt their behavior to account for the presence of an outsider, however clever approach the researcher applies to remain separate/indifferent. But we should not forget that the same kind of effect is present in natural science as well. Things change just because an observer is there. A human body radiates heat, reflects lights and rebounds photons thereby affecting its surroundings. But the natural scientists have been able to devise excellent labs to control for such effects. So this is again a matter of measurement technology. If we have good enough lab in social science, we can control for the observation effect and use the same methods that natural scientists use.
Another source of subjectivity in social science, as claimed by many, comes from the very nature of tools used in social science. Unlike in natural science where the measurement tools (such as enumerator or a scale) and researcher can be totally independent, in social science researcher herself is being used as measurement tool. But again this is the result of lack of technology in social science.
Finally, the debate should not be whether to use qualitative method or quantitative method. The method that gives the most adequate explanation of the phenomenon of interest should be the choice, which depends on the availability of the measurement technology.
Conclusion
There is only one truth (maybe dynamic and involves many parameters) to be known in both natural and social science. Due to the measurement technology we have and limitations of our observation tools, we may be viewing the world around us differently. That does not mean there are multiple reality. Reality is one. It just means that we are using different methods and we have different paradigms to understand the reality. We are just using different maps to navigate the same geography. We are just looking at the same sky using sun-glasses colored differently.