Abbott – Methods of Discovery – Chapter 1
Chapter 1: Explanation
- A. begins by noting that traditional social science is much like a monologue . . . it is patterned and consistent.[1. This is much the same territory that Latour stakes out when he claims that his ANT provides a fix from the problem of frames and tired methods.]
- A. notes that this work is a heuristic. . . a book “of aids to the social scientific imagination” (4). A. also notes that explanation is the purpose of social science (again, Latourian inferences). This perspective on social science method requires the researcher to give up a lot of the questions of why and instead concentrate on the what (7).
- A. claims that there are three things that are required for “explanation” in the social sciences: 1) explanation allows us to intervene in what we are explaining (i.e., we can explain economy when we manage it, we can explain poverty when we eradicate it, etc.). This is a pragmatic approach that works well with things that have “a narrow neck of causality” . . . something that doesn’t happen often in social science; 2) an explanation is an explanation when we stop looking for further accounts of something (this is different from Latour . . . or is it? ANT is about extending networks; however, you have to draw boundaries at some point). This is explanation as semantic – it translated one problem of analysis into another until we are satisfied (10). These sorts of reductions aren’t particularly useful as they preclude emergence; 3) an explanation is an explanation if we’ve made an argument about it (9). This is explanation as predicated on structure . . . structure that’s valued by particular communities of scholars. In this way, the third account of explanation is syntactic in nature as it emphasized the syntax of an argument.
- Methods are procedures attached to particular scholarly communities (13).
- Method by way of data gathering:
- ethnography: gathering data by personal interaction
- survey: gathering data by submitting questionnaires to respondents or formally interveiwing them
- record-based analysis: gathering data from formal organizational records
- history: using old records, surveys, and even ethnographies
- · Methods of data analysis:
o direct interpretation: analysis by an individual’s reflection and synthesis (narration)
o quantitative analysis: analysis using one of the standard methods of statistics to reason about causes
o formal modeling: analysis by creating a formal system mimicking the world and then using it to simulate reality
- · Methods based on question posing:
o cast-study analysis: studying a unique example in great detail
o small-N analysis: seeking similarities and contrasts in a small number of cases
o large-N analysis: emphasizing generalizability by studying large numbers of cases, usually randomly selected
- Methods of data analysis:
- direct interpretation: analysis by an individual’s reflection and synthesis (narration)
- quantitative analysis: analysis using one of the standard methods of statistics to reason about causes
- formal modeling: analysis by creating a formal system mimicking the world and then using it to simulate reality
- · Taken collectively, these breakdowns of method yield 36 possible subtypes of method production. Yet, as A. notes, a methodology is a social phenomena. . . and as such, often evades characterization (15).
- · Ethnography: living inside the social situation one is studying and becoming to some extent a participant in it.
- · Historical Narration: descriptive historical work that attempts to examine the question of “what really was” at any given time.
- · Standard Causal Analysis: takes a large number of cases, measures aspects of them, and employs statistics to draw inferences about the relationships among those measurements (19). Use these inferences to consider the “causal factors” that produced the patterns observed.
- · Small-N Comparison: this investigates a handful of cases from 3 to 12ish. Using ethnography and historical narration, this method tries to improve generalizations by invoking more cases to demonstrate points. Attempts to mitigate the ungeneralizability of one case and the oversimplification of many cases by striking a middle ground (22).
- · Formalization: this uses “modes of reasoning about social reality that require some quasi-factual input” (24). This is used most in economics but was also used by Levi-Strauss.
- · A. notes that different methods actually are not suited to different situations but accomplish different types of explanations. Here are the three “exploratory programs” that accomplish different types of explanations:
· Ethnography = semantic (translation), historical narrative = syntactic (narrative structures).
· As A. notes, the semantic assumptions of a syntactic program are quite worrisome. For example, what’s the point of providing 10 game theory models for any given situation. . . the semantic viability then takes a place, despite the structural explanation. In other words, a the believability of a structural accounts meaning is important and determines some of the politics of the method (35). A. notes that, “the syntactic program buys elegance and breadth at the price of semantic indeterminacy and limitation” (35).
· A. notes that “network analysis” is a prime example of “abstract semantic explanation”; however, networks have a difficult time accounting for temporality (36).
· SCA (standard causal analysis) paradigm: methods that work by taking apart complex particulars of data (cases) and treating them as intersections of abstract, universal properties (variables) (37) (what?). SCA was developed for pragmatic reasons but has been used to explain syntactical structure.
Chapter Two: Basic Debates and Methodological Practices
· In this chapter A. goes over the traditional debates concerning method before explaining the methods with which each debate is associated. He does so to demonstrate that method debates aren’t as cut and dry as you might imagine.
· Positivism vs. Interpretivism: things can be measured independent of context and replicable vs. social life measurement is not possible in quantitative ways.
· Analysis vs. Narration: Narration tells a story while Analysis attempts to use abstract concepts (theory) to explain said story. Analysis usually looks to explain or emphasize causality.
· Behavorialism vs. Culturalism: Ontological debate about the distinction between social structure (regular patterns of behavior – demographics) and culture (symbolic systems with which people make meaning). These map nicely onto the positivism vs. interpretism debate (too nicely actually).
· Individualism vs. Emergentism: Individualism states that the only real social entities are individuals. This line of thinking is traced to the idea that individual self-interests drive and produce the social world we inhabit. (Mandeville – Fable of the Bees). Emergents believe the social is real. This can be traced to Durkheim’s thought and argues that certain social forces aren’t reducible to the actions of individuals.
· Realism vs. Constructionism: Obvious.
· Contextualism vs. Noncontextualism: Contextuals argue that context is absolutely necessary to understand anything while the opposite argue a positivist, realist perspective on meaning (48).
· When you look to explain things researchers often focus on choice or constraint (49). This is a question of free-will vs. determinism or at least a question about the ability to act and the constraints of power on acting.
· Conflict and Consensus: Consensus argues that social organizations and institutions keep people from destroying themselves (Hobbes). Conflict research (Marx/Rousseau) asks why there is so much conflict considering that people are inherently good and only cloud their reasoning and judgment with oppressive social institutions that make them act is socially destructive ways.
Metacritiques, Critiques, & Responses