Tilly > Weblog

It Depends

The article introduces the main topics covered in Tilly’s edited volume (with R. Goodin), The Oxford Handbook of Contextual Political Analysis. The primary objective of the volume is to demonstrate that acquiring systematic political knowledge requires analysts to get context “right”. The authors distinguish between three classes of contextual effects that impact the study of political processes: (1) the level of an analysts’ understanding of the political processes they are attempting to study, (2) the nature of evidence available for the empirical examination of specific political processes, and (3) the nature and context of the political processes themselves. Additionally, the authors review alternative approaches to political analysis, discuss the necessity of explanatory stories in political analysis, and detail the strategies used by current political science researchers to either control or correct for context.

Afterword: Political Ethnography as Art and Science

Political Contention

The article briefly summarizes the benefits of using political ethnography over conventional social scientific methods. The primary benefits of this method are the researcher’s ability to closely observe social processes and develop more comprehensive insights. This benefit is critical to readers who, like Tilly, believe “how things happen is why they happen”.

Observations of Social Processes and their Formal Representations

Formal Methods

The article outlines the advantages of formalisms, defined as the "explicit representation of a set of elements and of relations among them", in historical research. Formalisms, he writes, help researchers discipline their research inquiries from the outset and make it easier to discover alternative explanations. The less frequent use of formalisms in historical research, he argues, results from historians’ confusion of formalisms with quantitative research. The quantitative/qualitative distinction, according to Tilly, is harmful to social science because it misrepresents the choice of using or avoiding formalisms with the transformation and representation of historical evidence. While all social scientists should use formalisms to guide their research, narratives are necessary for the presentation of certain types of research.

Why Read the Classics?

Social Scientific Knowledge

The article describes how sociological classics positively influence current research. Tilly’s primary argument is that the classics provide “sources of justification” for contemporary arguments. The classics also commit researchers towards the cumulative goal of answering important questions, and identify the theoretical conversations in which researchers are engaged.

Sociological Resources for the Study of International Relations

The article identifies areas of sociological analysis that can benefit current research in the field of International Relations. The areas identified are:

State formation
War and revolution
World Systems
Boundaries, identities, and related social processes
Contentious politics

Event Catalogs as Theories

Formal Methods

The article describes one of Tilly’s favorite formal methodologies, the event catalog. The benefit of using event catalogs, defined as a set of descriptions of multiple social interactions collected from a delimited number of sources using a uniform procedure, is that it focuses the researchers attention on three critical methodological questions: (1) how phenomena leave traces, (2) how phenomena can be observed, and (3) how traces of phenomena can be used to reconstruct attributes and elements.

Jumbo Speaks

Historical Knowledge

Tilly reviews “The Sociology of Philosophies” by Randall Collins, an analysis of the formation philosophical thought over humanity’s entire literate history. Tilly cites methodological issues with Collins’ analysis (i.e. data selectivity issues, lack of economy in the presentation of data) and questions the plausibility and relevance of the analysis.

The Trouble with Stories

Social Scientific Knowledge

The article suggests that sociology's strongest insights do not take the form of standard stories - sequential, explanatory accounts of self-motivated human action - which are common in historical analysis. This is due to the limits of storytelling including: (1) its limited number of interacting characters, (2) its constraints on analysis within a specific time and space, (3) its requirement of independent and conscious actions, and (4) its requirement that all actions, with the exception of externally generated accidents, result from previous actions by the characters.

To Explain Political Processes

Invariant Modeling, Political Contention, Social Mechanisms

The article critiques the method of invariant modeling of macro social processes. He claims that this practice leads researchers to focus on "improving the model" as opposed to understanding that the regularities between macrosociological processes do not operate in the form of recurrent structures and processes. Thus, the construction of invariant models of political revolution is a waste of time and the poor fit accounts for the slow accumulation of knowledge.

As an alternative to invariant modeling Tilly recommends the mechanism approach which consists of the following steps: (1) construct valid ontologies, (2) clearly specify the variation within the field, (3) break complex sequences into smaller events, and (4) form contingent predictions (i.e. "insofar as…") instead of attempting to form invariant general laws.

Formalization and Quantification in Historical Analysis

Formal Methods, Labor History

The article argues that formal methods will remain on the periphery of the field historical research until the major questions of the field change to ones that can be properly addressed by explicit models, system variation, or the comparison of many cases. This is because quantification is useful in historical analysis as a function of (1) the complexity of the explanatory model, (2) the intrinsic quantifiability of the phenomenon to be explained, (3) the importance of variation to the argument, and (4) the number of units observed.

Retrieving European Lives

European History, Formal Methods

The article argues that the central activity for the study of European social history is reconstructing ordinary people's experience of large structural changes. Formal methods for this task such as collective biographies have the disadvantage of missing critical elements of movements and being enormously time consuming, but also provide extremely precise information that can be used to do close comparisons. In theory, these methods allow the researcher the ability to retain the idiosyncrasy of personal experience while identifying uniformities and variations across many personal experiences.

Neat Analyses of Untidy Processes

Labor History, Formal Methods

The article speculates about the impact of formal methods (e.g. collective biography) on the study of labor history. Formal methods have been used to answer many of the peripheral questions of the field but the core questions, such as fluctuations in the militancy or effectiveness of worker collective action, remain properly answered through informal methods that are better equipped to describe shared states of mind and tactical replays of group interactions.

The Old New Social History and the New Old Social History

Formal Methods

The article contrasts the impact of "The New Social History" of the 1960s which was based on applying statistical analysis to historical evidence, in light of the field's return to a more narrative-based analysis of historical events in the 1970s which he terms "The New Old Social History".

A British View of American Strikes

Political Contention

A review of "Strikes in the United States 1881-1974" by P.K. Edwards. The book attempts to explain how and why the American strike picture has been altered during the enormous industrial and institutional changes of the past century. The review applauds the book for bringing more qualitative details into the study of strikes, but critiques its simple and sometimes unorthodox statistical methods. They argue that the author disputes previous work but does adequately show how his new theories are different from previous studies.

Computers in Historical Research

Formal Methods

The article describes the effect of computers on historical analysis. He argues that computers reduce the time necessary for collecting and matching historical data, but that there is also an underlying danger. The increased ease in which researchers can make post hoc statements regarding spurious results may lead to mindless empiricism.

Quantification in History, as Seen from France

Formal Methods

The article argues that the rise of quantification in historical analysis is the result of: (1) the widespread adoption of the collective biography and (2) the arrival of historians trained in outside disciplines with analytic training and a desire to answer questions about long-run changes in society.

Methods for the Study of Collective Violence

Political Contention

The article discusses five major methodological problems plaguing the study of collective violence:
(1) Problem of basic definition
(2) Problem of biased enumeration
(3) Problem of fragmentary documentation
(4) Problem of multiple units of analysis
(5) Problem of attaching violent action to its context

Anthropology on the Town


The article describes the methods used by urban sociologists for studying life in large cities. He argues that anthropological methods become important in this area after average data (i.e. census) became prevalent. Cases that deviated from the mean become more interesting and important. These methods include: observing areas in cities to see who walks by, how often, and what are they wearing. Getting people to draw maps of the city to see what areas they include and which areas are drawn disproportionately. While much of the data collection is commonsense, what is important is that it is done in a systematic fashion (i.e. sociometrics).