My first month here at IAS is over, and it has been quite the experience. Besides natural Covid-related mental stress caused by spending most of your waking (and all non-waking) hours within the same four walls, I have begun working on the field(s) of my PhD. Now I have a broad overview over any research conducted within the field of Analytical Sociology by compiling a long reading list. While you’re reading this, I’m finishing the next one, this time on Computational Social Science.
Both Analytical Sociology and Computational Social Science are rather new fields of study, and while both have a very different provenience, they are beginning to converge, with one supplying the other with what it does not have (hence this article’s title). Analytical Sociology is concerned with explaining social mechanisms by analysing the processes that both lead to the emergence of social superstructures (which can be organisations, institutions, laws, norms – you name it) and, reversely, to altered behaviour within individual people. Computational Social Science, on the other hand, is closer to Computer Science and hence less theoretically grounded, but highly sophisticated with regard to its tools.
After the first month, I am sure that this is the right playing ground for me, equipped with everything that makes me happy: The aspiration to abstract theoretical explanations, which shall be followed meticulously, and all of that married with data. Lots of data. Imagine me looking into the next four years like a three year old in summer seeing a sandbox.
Analytical Sociology forces me out of my theoretical comfort zone – now I can’t be happy with just reading theory, I have to fully comprehend it before I have to use it in order to provide hypotheses which then need to be tested against large swathes of data. Especially during my lazy phases I tend to apply rules of thumb rather than precision – a feat which regularly caused amusement within my Chemistry class before graduation. (To this day, I do not precisely know why my teacher awarded me with the special price of the German Gesellschaft deutscher Chemiker, GDCh, for an astonishing graduation mark. I was theoretically good, yes, but my experiments can’t be described with anything other than this image that floats around the internet for approximately as long as I have been sure I won’t ever get back into Chemistry.)
In Analytical Sociology, there is no rule of thumb. There’s either a mechanism you can describe, or there isn’t. You have to be precise. And due to the tools of Computational Social Science that are slowly dripping into the field, it’s getting harder and harder to get away with pointing to some R2 value and explaining “But it’s significant! 😱” As I outlined in an earlier post, if your results can’t explain mechanisms and can’t be reproduced, the value of the work is limited only to the one case-study you made. But Sociology strives for more. And I like that aspiration. If I am able to explain just one social mechanism that we encounter not just once but across the board, imagine me as one of the happiest people on Earth. Because this means that I have found a theoretical explanation that in fact explains society.
But for now, I am very content with the results I achieved until now. I know who some of the important figures within Analytical Sociology are, and have gotten a shallow overview over some debates that have happened in the past. But before getting too long with this post, let me share just one funny observation I made. Within the field – and I suppose that’s the same in every other field as well – there are broadly speaking three types of characters. There are the preachers, the outlaws, and then there are the haters.
Among the “preachers,” I think, are the “big two,” Peter Hedström and Peter Bearman, who have compiled the most comprehensive collections of articles within the field and possess a knowledge that is matched by only a few. Then there is Paul DiMaggio, one of the inventors of the sociological institutionalism, who has a speed of publication that seems to be on par with that of Bob Jessop. Then there are a few names that continuously pop up, such as Jon Elster, John Goldthorpe, and Mark Granovetter.
To the group of “outlaws” I would count those that seem to be a little bit on the outside and are contributing valuable input, such as John Levi Martin, or Matthew Salganik and Duncan Watts (who are on my Computational Social Science reading list). Obviously, all of this is just based on names I encountered while cataloguing the relevant journals, so count it as anecdotal evidence (for now).
And then, there are the haters. I’m sure there are more, but one name struck me when I saw it the first time: Omar Lizardo. I remember having encountered him several times during my graduate studies, and the name seems to stick with me. I have cited one paper of his in my M.A. thesis, and thus the name caught my attention immediately. Even funnier is the position he seems to occupy within the field. To give you a glimpse of what his stance is, let me just give you a title of one paper he’s published in Sociologica: “Analytical Sociology’s Superfluous Revolution.” And the tone of his paper seems to be appropriate to the title.
I am very sure he has some valid critique of the field, but the harshness of the title immediately evoked a smile in my face – I do like some crispy critique, even more so when it’s coming from someone who clearly has contributed to several sociological fields so far. But for now I must leave it to that first idea I have of his contribution, because, alas, I have yet to read that paper.
Next week my plan is to finish the remaining reading lists in order to be prepped for the holidays, before having the second meeting with my supervisor. The week after that I will finally return home for Christmas. I generally don’t give much thought to Christmas, but this year I feel it is utterly necessary to prevent myself from getting crazy. Keeping a distance of two metres is fine, but over 1,200 kilometres is not – especially if you have to do both.