General Discussions

If your contribution is relevant to more than one of the workshop questions, this may be the right place for it.

14 thoughts on “General Discussions

  1. On Granovetter’s “Economic Institutions as Social Constructions: A Framework for Analysis”

    In his 1992 paper, Mark Granovetter analyses the ways institutions are approached in neoclassical economics. He is critical of the narrowing perspective of conventional economic analysis (around mathematical methods, assumptions of rationality and methodological individualism) despite the broadening scope of subject matters (covering issues such as family, crime and education). According to the author, these simultaneous narrowing and broadening seem to impose a ‘narrow and fragile base’ (p.4) as the ‘universal grammar’ (referring to Hirshleifer, 1985) for social sciences. Granovetter’s critique involves arguments that are relevant to the questions we will tackle in the first workshop. I will try to summarise these arguments here.

    Question 1
    Regarding our first question, whether all institutions are social constructs, we find a relatively straightforward answer in the paper. While referring to economic institutions, Granovetter notes that all institutions have the same property that they “do not arise automatically in some form made inevitable by external circumstances, but are ‘socially constructed’” (p. 4), referring to Berger and Luckman’s (1966) well-known book The Social Construction of Reality.

    Question 3
    Since the paper analyses neoclassical approaches, it refers more than once to rationality of institutions that we tackle in our third question (are institutions rational or rationalised?). Granovetter points out that the New Institutional Economics diverged from the old institutional approaches by its focus on and/or assumption of the economic efficiency of observed institutions. Against a background of empirical observations of seemingly irrational phenomena such as significant wage discrepancies between categories of workers, or structures and patterns that are irrelevant to atomistic perceptions of markets such as firms and economic networks, neoclassical economics searches for theoretical explanations that fit the assumptions of rationality of agents and efficiency of markets. Institutions are thus rationalised particularly in what Granovetter calls functionalist arguments that “argue backwards from the characteristics of institutions to the reason why they must be present” (p.5). The author’s discussions on the over- and undersocialised representations of human action are also relevant to our question. The former is allegedly the approach of sociology (if we refer to Wrong, 1961) where individuals are assumed to obey social norms almost automatically. The latter is, in Granovetter’s view, the approach of utilitarian tradition in neoclassical economics where agents are interchangeable and their social relations are irrelevant. One can link the rational-rationalised distinction we raise in our question to these visions of human action where the role of institutions is either dominant or non-existent.

    Question 4
    Granovetter also discusses the temporal dimension of institutions in this paper and this relates to our fourth question; how do institutions relate to social structures and how do they relate to social change? Part of Granovetter’s criticism for conventional economics is that it puts comparative statics of equilibrium states at the centre, while institutions can only be studied and explained through a dynamic perspective. By neglecting all the socioeconomic processes and contexts in-between, the comparative statics approach assumes the automatic emergence of efficient outcomes. When it is pointed out that such efficient equilibrium states are not universal, neoclassical economics holds that imperfections related to underdevelopment are irrelevant to economic theory because they will disappear as markets become more ‘sophisticated’ (Leff, 1979). When similar cases are observed in advanced economies, neoclassical thought returns to the functionalist search for the set of conditions that make these cases efficient.

  2. Pingback: Our workshop questions and a paper by Granovetter | Constructed Complexities

  3. Our task is difficult because it involves the intersection of three major domains–complexity, social construction, and institution–each of which has literatures reporting conceptual ambiguity–i.e. multiple meanings–and vagueness–i.e. lack of clarity in application to specific cases. Just as a thought experiment, imagine that there are five meanings for each of the three concepts, and imagine that all of them are vague. That would generate a space of 125 combinations of meanings each of which would require a three-dimensional prototype semantic model.

    That’s why I thought of taking some stripped down, extraordinarily clear frameworks for each of the three concepts and starting with them. Same way I build computer simulations. If I can’t model the simple case, there’s something wrong with how I’m going at it, or else the case isn’t “modelable.”

    Say we call the research object in which we are interested “X.” Later, for our specific project, we can instantiate X with “institution” and “inherit” the simple schema as a basis for adding complications to suit that particular domain. For now, ignore that part of the strategy. Then the two questions would be, what does it mean to say that X is “complex?” Arthur’s introduction looked like one good schema to start with. The next question would be, what does it mean to say that X is a “social construction?” I’d like to experiment with an answer to that question now, drawing on the helpful comments on the web pages plus a couple of other things I looked at. I haven’t yet found an Arthur equivalent for the social construction concept.

    I looked up “construction” in a couple of dictionaries and thought about everyday usage as well as reading a few academic sources. “Construct” is a change of state verb where some agency makes something out of other things over time, that final something being the “construction.” At this general level the fit with complexity is clear–emergent order based on a dynamic process.

    So, to ask if the phenomenon of interest X is a construction is to ask about the materials and environment and agency and the process where all those things interacted and caused the construction to emerge.

    “Social” means that at least some part of the agency is human. That part could be a single individual, or a collection of them, or an organized group of them. Even a single individual may be considered “social,” in the sense of Vygotsky or G.H. Mead.

    So to ask if the phenomenon of interest X is a social construction is to ask about the human agency part of the process by which X was constructed.

    Finally, as noted by some, the social construction question is meant to enable counterfactual questions that illuminate possible changes in X. In other words, If a social construction is a dynamic process involving human agency, then plausibly human agency could have constructed–and might now construct–X in a different way, “different” often meaning “better” according to some evaluation. The notions of path dependency and exploration of a space of possible outcomes makes sense here.

    Two things. As I think of this schema in ABM terms, that being the way I try and clarify my own thinking about complex adaptive systems, the schema seems to work as a way to guide questions to ask of a model of some institution as a social construction.

    Second, as I think about the two institutional domains I know something about–substance use and water–I can see how the role of human agency will be variable, complicated, and different in terms of how and in what ways it constitutes “leverage points” in the nonlinear dynamic process that “constructs.” This is as it should be, I think, and perhaps reflects the argument I found in the few recent sources on social construction that I looked at that end with a call for a blend of “realist” and “constructionist” approaches in contemporary human social research.

    As always, my goal in writing this is to make a useful mistake. As Box said in his famous quote, all models are wrong but some are useful.

    • i am a newbie … so i would frame my question from a lay person perspective. i am wondering about the performativity of meanings and how it can be captured in complexity research. if the words ‘complexity, institution, social construction’ have multiple meanings and the combination of each meaning creates around 125 possibility, i am wondering how this different meanings correlate with each other. taking the definition from Granovetter for instance, institutions have the same property that they “do not arise automatically in some form made inevitable by external circumstances, but are ‘socially constructed’” (p. 4) (i copied from the previous comment), this show a positive correlation between institutions and social construction. but what is the implication of this positive correlation, will it attract more words such as intervention (as a form of social construction for changing an institutional configuration)? and probably attract less word of intervention if it has a negative correlation?

  4. Pingback: Stripping down our concepts | Constructed Complexities

  5. @Yuti: I am not sure we can capture relationships between concepts or words with the metaphor of correlation because i) it is not easy to reduce words to magnitudes ii) even if we envisage or assign such magnitudes, we would not expect them to increase or decrease together with changes in the ‘sizes’ of other words following some stable relations. Words can continuously evolve, diverging from or converging with each other.

    I think sets and intersections between sets might be a better mathematical abstraction for words. Hence, for example, referring to our first debate question, if we assume all institutions are social constructs, we would expect institutions (or different understandings of them) to be subsets of social constructs. For multiplicity of meanings of the same concept, have a look at Wittgenstein’s metaphor of family resemblance. This can be represented with many intersecting sets with no universal subset.

    I am not sure I understand what exactly you mean by attraction of more words such as intervention (do you mean defining a word employs more words) and negative correlation (do you refer to a contrast between meanings?).

    The way multiplicity of meanings is captured in complexity research would be contingent on the methods that are employed. In agent-based modelling, for instance, one can create a world where each agent has a different understanding or expectations of the same word or phenomenon.

  6. I agree that the term correlation might mislead to regression, while what I am interested to know is the understand the genealogy (or family resemblance) of a term and how it is performative in co-shaping other concepts. I have a background in mathematics, done new institutional economics couple of years ago and now, I focus on science, technology and society, so my understanding of social construction came from social construction of technology. During my journey of moving from one discipline to another, I found that is extremely frustrating for a mathematician to deal with messiness in social science. As a mathematician, I always start with definition and stick with my own definition to build an argument, whereas in social science, scholars have tendency to create new definition, although it resonates with other definition. So when I found this discussion, I am excited to see whether investigating concepts that have intersections such as institution and social construction may help to understand the degree of nearness for concepts such as change (if we assume that institution is a configuration that holds certain routine). So what I have in mind is a comparison of an evolution network between institution-change, social construction-change and institution-social construction-change. By comparing how these concepts creates set and intersect, I am interested to see whether a term is performative or not. In this work, the words ‘institution’, ‘social construction’ and ‘change’ function as the pivot for constructing multiple worlds (or sets). This what I mean when saying attracting.

  7. @Yuti: Thanks for these interesting thoughts. I have two suggestions for the study you have in mind. You can make a critical literature review focusing on how influential authors addressed the connections you have mentioned (e.g. institution-change) over time and in different fields. If you can access to (or collect) a large set of texts about these concepts that you think is a good representation of the variety of ways these concepts are used, you can also use automated text analysis tools to find connections between words (depending on number of times they are used together). Good luck and keep us updated on your work!

  8. And thanks for the suggestion. I think I am going to execute it by having three case studies. First I am going to do a network analysis by using data from scopus to identify the network behavior of three spaces (e.g. institution-change). I am thinking to start from the emergence of new institutional economics and make the graph periodically, probably in 10 years period. And then I am going to take a qualitative analysis by taking three papers that follow the conclusion of the network analysis for each space. By doing these steps I can have insight from network and qualitative analysis about the performativity of a concept 🙂

  9. I enjoyed your conversation. It made me smile to read, “I found that is extremely frustrating for a mathematician to deal with messiness in social science.” I’m not a mathematician and I’ve been frustrated since my first social science course (:

    If one thinks of social science concepts as part of ordinary human discourse about ordinary human worlds, then ambiguity and vagueness aren’t a surprise. For instance, just for fun I looked up NIE on Wikipedia and a section begins, “Although no single, universally accepted set of definitions has been developed, …” I always liked the famous Einstein quote, though I don’t know the context in which he used it, “As far as the propositions of mathematics refer to reality they are not certain, and so far as they are certain, they do not refer to reality.”

    Making sense of concepts in terms of how they are used makes a lot of sense. There’s several places to go with this, Malinowski’s context of situation, James’ pragmatism, Wittgenstein’s forms of life, Kuhn’s paradigms–not to mention the incentive structure of the research university, the politics of research support–all the doors that open once you think of how human social science applies to itself. Though I don’t know the field, I guess that’s what STS is all about, for all of science and technology, though it seems to me that the game changes with human social science because the researcher is also an example of the phenomenon.

    I think Yuti and Ozge propose to escape this problem in an interesting way. One escape valve is to exempt the goal of “clarity” from the Law of the Excluded Middle. It’s not the case that concept X is either clear or not clear. It can be kind of clear, clear as mud, clear enough, partly clear, and so on along some multidimensional fuzzy scale. Another escape valve is to take a lot of examples and a lot of contrasting ideas about how to translate them into some representation and put them into a blender and see what happens. If I’ve got it right, using my inelegant smoothie metaphor, it sounds interesting and I’ll be really curious to hear how it turns out.


    • I like the quotes from Einstein, “as far as the propositions of mathematics refer to reality they are not certain, and so far as they are certain, they do not refer to reality” since it reflects my work in both worlds, mathematics and social science. In mathematics, I worked with assumptions and the beauty of it was, I know what I referred when I said X and know the consequence of it. In social science, it works the other way around, I start with a concept or a formula and each time I go to the field, I need to add the limitation of the formula. The limitation appear after and not before as in mathematics. And since it is impossible to have a perfect repetition, since time is irreversible and concepts travel across spaces, concepts in social science always betray its conceptualization. I guess it means that certainty in social science only appears in paper, when the reality is on-hold.

  10. Yes, it is. My last comment was just my amazement on how well Einstein’s quotes explain my situation. Initially, I was frustrated with social science since I expected the same thing from social science as I found in mathematics, certainty in doing the analysis. The comfort that I got from mathematics was possible since protection against the real world was done at the beginning, through assumptions, whereas in social science it was at the end. After knowing this, I understand that the source of my frustration was expecting the impossible thing from social science. Mathematics and social science have different epistemology (and probably I am the last person on earth that realize it :D) and it influences their whole methodology and assumptions. Thanks Mike for pointing the quotes, now I know that my problem is eternal. To understand the reality I have to deal with uncertainty.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s