Bah, I seem to have grown a conscience over the night, so I was unable to watch much more than a couple of minutes of Heroes before realising that I should rather be typing out a set of lecture notes. (But I managed to play “Championship Manager Season 97-89″ before surrendering to this crap, so I’m not giving up hope; maybe I can kill this conscience before it grows any stronger.) So, here we go, the epistemology and methodology of political science. You lucky bastards; I hope you enjoy it as much as I do. Oh, and there might be some overlap from the previous lecture, at least in the very beginning.

————————————————————————————————————————–

What do we compare?
The main object of political scientific comparisons, is the often very elusive area we refer to as “politics”. This term can be very broad, but it is [almost?] always related to institutions of governance and the state. When it comes to how we compare, it should be pretty obvious that it’s not merely spouting one’s opinions we’re talking about here. Nor are we refering to dealing with current events, even though current events [should?] have more relevance for the study than e.g. opinions. Rather, political scientists, like all scientists, study facts based on systematic analysis, conducted in a disciplined and intelligent way. In other words. “how?” is a matter of method, and the method we choose should depend on the “what?” and the “why?” — i.e. the question you’re seeking an answer to, and you’ve come a long way when you’ve phrased that single, clear question. Of course, this all sounds bloody easy, but it’s not as easy as it sounds. [Duh.]

Comparative Politics.
CP is a huge, diverse field — even though the various theories aren’t as differentiated as in IP — in which the big, collective project is to advance our collective knowledge about politics and how it works. Not all studies within CP do the same thing, but all contribute in some way to the bigger comparative project. Additionally, not everything that is put under the tag of “Comparative Politics” is directly comparative; it can e.g. be informative about one country. This is achieved by CP providing data for other researchers, building or testing theory, or focusing on the theory or methodology of the field.

The “scientific” study of politics.
A scietific nature is always a goal in political scientific studies, but it is seldom achieved. As to what “science” is, it can be called a means of investigating a question, it is in other words the method, not the result. Political science as a whole has an empirical focus — it is fact-based — and CP is no exception. One feature of this “scientific nature”, is that all data and theory should be evaluated fairly, or objectively, if you will. “Science” should also be open, or transparent, which a scientific study can be said to be if it offers a possibility for others to check facts, and examine methods and logical conclusions. This is in many ways the essence of science, that other scientists must be able to check your research.

Unlike history, which seeks to produce singularised statements about the world, political science attempts to generalise. Political scientists are looking for similar traits in similar events, rather than focusing on singular events. The ultimate goal is to produce generalisable statements about the way the world works. (Does “x” cause “y”?)

Theory and facts: a dialectical relationship.
Facts, data, information, or whatever you want to call it, combines into the empirical basis, which we are forced to relate to, always. However, the world consists of such a multitude of facts, that we need some method of weeding out the irrelevant data. Usually, we can use theory as a basis for a question which we intend to ask and find an answer to, and through this question we can determine the relevance of data. A theory is a kind of explanation for how things work. This explanation, however, must be backed by more than just correlation; some kind of causuality must be established, and this causuality is found through the examination of empirical facts. This, then, gives us a dialectical relationship between facts and theory: We have massive amounts of facts, and we can use theory to select the ones that are relevant to us in any given situation. Teen we can apply these data to theory, and adjust the theory to the facts, combining into an ongoing dialogue between factual incidents and theories. This is a process of trying and failing, which is governed by method, and method helps us to compare systematically and understand what we have when we have something. CP as a field is interested in all three, but an individual study or researcher might focus on only one or two of these (data, theory and method).

Political science and “the Truth”.
Theories and even political science look at only small slices of “reality”, and political science as a science is not about “truth”. Rather, it strives to produce knowledge with credentials, strong reasons to accept claims to knowledge. To quote Indiana Jones: “If it’s Truth you’re looking for, Dr. Tyree’s philosophy class is right down the hall.”

Method.
Method sounds really dull, but it is what tells us the best way to answer a question, thus making itself indispensable. Which method to apply is always depends on the question we ask, but it helps us to evaluate how good our work is, all the while providing a tool to evaluating the work of others — and thereby equipping us with a defence against untrustworthy polls and research reports. Furthermore, it is central in all branches of political science, as it is how our work “gets its credentials”, as well as for all the reasons listed above in this paragraph. Method is one of the most hotly debated topics in CP and political science. [Even though the lecturer didn't go any closer into this topic, I assume that also political science has been haunted by the ghost of postmodernism, with all its annoying (yet relevant, if you ask me) questions about method and objectivity. But that's just my guess, and I suppose there are plenty of quarrels over methods between rationalist scientists as well. I mean, academicians don't exactly need an excuse to bicker, do they? :P]

Inference.
Inference, or the connecting of prior knowledge (what you already know) to a new situation, in order to to create meaning beyond what you directly “know”, is the cornerstone of generalisable statements. And in order to make these kinds of statements, you have to know your methodology. [According to Sociology professor P.M. Schiefloe,a field's methodology = its method + its theory.] Let’s have a look at some methods, shall we?

The CP methods.
The most basic method of political science in general and CP in particular, is the case study, and there are two primary types of case studies. The 1-N studies are single case studies (the “N” refers to the number of cases in a study), while area studies are single case studies of a single area. [Or so I presume; I always hate it when lecturers use these kinds of terms without elaborating on them, especially when I'm too lazy to look them up myself.] Another basic method, along with the case study, is the focused-case comparisons, or small N studies. These are comparative historical and qualitative studies, which doesn’t include too many cases — usually 2 or 3, as the qualitative nature of this kind of study makes it hard to examine too many cases — and their case- or variable-orientation [their focus on qualitative questions?] renders it impossible to apply methods meant for huge amounts of cases. [Don't ask me why, though. :P] This is the traditional backbone of field. Finally, you have the large N studies with lots and lots of cases (e.g. “voters in Norway”). These are almost always statistical, frequently cross-national, and variable-oriented, and they are very important in the field today.

Cases and variables: cases.
A “case” is a case of something — an instance of a more general category — and a project turns into a case study only when it becomes clear what the study is a case of. Case-oriented approaches specialize in in-depth knowledge about one or a few instances of something. Case studies are holistic and contextual, with goal of understanding how things happen, and the word “case” is used in both Large and small N studies. In political science, cases are chosen because they seem like good representations of the phenomenon the researcher wants to understand.

Cases and variables: variables.
Variables focus on an abstract quality or factor, rather than on context of some event (or entity). The dependent variable (Y) is the factor or quality we seek to explain, eg. Democracy, stability, peace, while the independent variable (X) — X⇨Y — is the factor believed to influence or “cause” the (values of) the dependent variable, eg. wealth, level of education, social cohesion, or democracy. Finally, there’s the intervening variable (Z) — X⇨Z⇨Y: A factor through which the independent variable influences the dependent variable.

A variable-focused study is a study in which the researcher examine a quality or factor shared by several contries or similar.

The large N study.
The large N study have become an increasingly important part of study of comparative politics today, and there has been a huge increase in information available, e.g. through the creation of many large databases — many of them avaliable on the Net. A few examples includes the Eurobarometer, the World Values Survey and the European Values Survey, the International Social Survey program, the Comparative Studies of Electoral Systems, the European Social Survey, the Pew Global Attitudes Survey, the Gallup International Voice of the People, and various UN agencies’ databases, dealing with such things as development, population, education, and resources (e.g. Human Development Report). A Norwegian source of such statistics is the SSB, and the departments. Thus, much is at your finger tips, or is accessible with a little effort (and occasionally by paying a little fee). However, with so much data out there, you need to know more than ever what information you need and what to do with it. If you don’t, you will drown in data.

The small N study.
The number and richness of context make these distinctive, both in relation to each other and to the huge N studies. They can be variable-oriented study of rare phenomena, but even if they study more mundane phenomena, case-oriented studies depend on in-depth knowledge about the individual case. How then do you compare with other cases in a case-oriented study? Sometimes you might examine parallel, while at other times it might be just as effectiveto look at the most similar systems, and compare to these. A researcher might compare systems that are most alike to see what difference a single variation makes.

At issue: the essence of knowledge.
Because the two main methodological approaches each have its own understanding of knowledge, it can be said that the essence of knowledge is under debate here. Variable approach studies emphasizes objective knowledge criteria, and argues for the importance of broad, generalisable conclusions with a high degree of confidence. The case study approach, on the other hand, suggests that knowledge lies in really understanding what happens in a specific context, such as a country, or an election. They also point to the problems of comparison, the problems of meaning (a concern for “numerological nonsense”), and that these problems might cast doubt on the “scientific project”.

The question remains, however: Are these really possible? Or even meaningful? [What's possible or meaningful? Gods damn it; I thought you had to be clever and reflective in order to get a PhD, but anyone who's just a little more clever than me can understand that it's well-nigh impossible to understand what the hell "these" in that first question refers to. Is it refering to the problems? Or the points of view on knowlegde? Or something else entirely? Perhaps it's the lecturer's dinner? Fuck.]

The logic of inference.
Correlation, or association, of variables is not necessarily causation, and thus it is not sufficient evidence to establish a credible chain of causation. However, causation requires association, as invariable association is a way to establish cause. Thus, all one has to do in order to establish causation, is elimination of the non-causes. This can be done by first applying a method of agreement, and then a method of disagreement.

The method of agreement.
Both the method of agreement and that of disagreement were created by John Stuart Mill back in the first half of the 19th century, and they were presented in his “The Prophet”. The method of agreement can be phrased as a simple equation: If X then Y. The minimum requirement is that to be a cause, a factor must be present where the explanandum (the case that is to be explained) is present.
Table of cases (horisontal boxes) and possible causes (vertical boxes).
This is known as “Mill’s First Canon”: “If two or more instances of the phenomenon under investigation have only one circumstance in common, the circumstance in which alone all the instances agree is the cause (or effect) of the given phenomenon.”

The method of disagreement.
The method of disagreement is usually considered to be more scientific than the method of agreement, particularly under this current paradigm, where falsification is generally seen as more probable than verification. It, too, can be phrased as a simple equation: If X then Y; if no X then no Y. This is a more demanding requirement, as to be a “cause” according to this method, a factor must be present where the explanandum (phenomenon to be explained) is present, and must be absent where the explanandum is absent. This should always be the case; if not, it’s less likely to be any causuality. as it can’t be a case if it isn’t there, nor can it be if it pops up everywhere without causing anything.
Table, method of disagreement.
This means that you examine a set of cases of the phenomenon you wish to explain (”no” in the extreme left vertical column), as well as one or more cases of a phenomenon which is relatively unrelated. Then you examine them, and try to discover if the various potential causes (the top row) are present in any of the cases. If a potential cause is present in the case which weren’t of the phenomenon you wanted to explain, while also being in all of the ones you want to explain, there’s no cause, only correlatio, as it can’t be a cause if it doesn’t cause whatever it’s supposed to cause.

[This is kinda heavy stuff; I hardly understand it myself...]

General points about methodology.
First of all, it is important to remember that because people aren’t particles, all methodologies have advantages and disadvantages, and that it is quite possible to combine approaches. Many methods have the same underlying logic, but it is none the less important to keep in mind that the method you choose should suit the questions you ask, take note of other research in the field, and reflect your available resources.

[Here are some stuff about the establishment of democracy in Iraq, but I'm not in the mood to type it out, so I'll just copy/paste it:]

Example: Can The US Establish Democracy in Iraq?
•How we phrase question is critical
–Alternatives (Problem stilling)
•Can Iraq develop into a democracy?
•Can democracy emerge under occupation by a foreign power?
•Can autocracies transform into democracies?
•Can Islamic countries be democratic?
•Can countries of the Middle East become democratic?
•Factors (Independent): US, foreign invasion/occupation, civil war, specific qualities of Iraq, Islam (type of Islam), geography (Middle East)
•Factor (Dependent): Democracy. What sort of Democracy? Regular elections? Popular rule? Particularly democracy? Secular democracy? Constitutional monarchy? Participation of women on an equal basis with men?

One (very rough) Research Design:
–Makes US the active agent, focus becomes what the US does
–What are the factors that should be compared? (theory, previous study)
•What countries has the US invaded and occupied?
–For example, Germany and Japan after WWII
–What happened in those countries?
–Similarities, and differences –would another set of comparisons make more sense?
–Is this a meaningful and or useful comparison? (theory, previous studies)
–Conclusion must match design
•Might tell us some things, but not others

One (very rough) Research Design:
–Makes US the active agent, focus becomes what the US does
–What are the factors that should be compared? (theory, previous study)
•What countries has the US invaded and occupied?
–For example, Germany and Japan after WWII
–What happened in those countries?
–Similarities, and differences –would another set of comparisons make more sense?
–Is this a meaningful and or useful comparison? (theory, previous studies)
–Conclusion must match design
•Might tell us some things, but not others

Phew.