General Presentation

1Event history analysis is a demographic technique used for quantitative research in the social sciences. A specific feature of the event-history approach is the introduction of the time factor into the analysis: individual and collective time, changes in society and in historical eras. From birth to death, individuals experience a range of demographic events: parenthood, marriage, changes of residence etc.

2As observation methods improve and questions become more specific, the trajectories recorded become more detailed. Researchers are confronted by complex sociodemographic events, but also by changes of state which are not simply marked by a specific event, since a transition from one state to another may take the form of a process not necessarily associated with a single date. The life-event histories of individuals recorded in the form of a continuous sequence of diverse events (GRAB, 1999) are the result of interactions between these events. Transition is seen increasingly as a process spread over time, no longer fixed in a single instant, and event history surveys shed light on this complex reality. This creates a paradox for event history demographers: defining an event by means of a single date is over-simplistic, yet how can we capture the various intermediate stages, the fuzzy threshold between two situations, whereby the individual shifts from a “start” status to an “end” status?

3This problem can arise even in standard demographic events (births, deaths). Though the birth of a child is marked by the moment of delivery and its date of birth, its existence as an individual depends upon religious and personal attitudes, with some people believing that the person already exists in utero. With progress in medicine and life support technologies, there is heated debate about the status of patients in a vegetative state, and the transition from life to death is no longer as instantaneous as once believed.

4These are extreme illustrations of situations frequently encountered by the GRAB researchers, in which no distinct, unique markers of transition could be recorded. For example, the period of pregnancy precedes the birth of a child. This period, which stretches from the day the pregnancy becomes known to the day the child is born, varies greatly in length from one woman to another depending on her age, experience, family environment, level of education, etc. And the period of “anticipation” may even precede conception. Entry into union takes different forms in different societies and covers a period of varying length: engagement, dowry payment, cohabitation, religious or civil marriage ceremony, etc. The choice of potential markers is therefore wide. In the study of mobility, certain surveys no longer simply record residential changes, but the entire “residential system” of respondents over a given period to better capture residential and occupational turbulence (GRAB, 1999). Lastly, when the transition studied, such as entry into adult life, results from a combination of changes in a variety of areas, the problems of defining event markers become especially acute, to the point where researchers may question the utility of an approach whereby only a limited sub-population is recognized as adult. This is indeed a conceptual problem: can we say that a young man aged 28, with no steady job, no partner and living in the family home has not yet reached adulthood? This problem is combined with a measurement difficulty: what are we observing? What information should we collect, and for what purpose?

5The development of questioning aiming to collect quantitative data on trajectories provides an innovative means to observe these periods in more detail. What have been the outcomes of this development – its advances and its limitations? What do interviews reveal about the perception of these transitions by individuals and their contact circle? What lessons can be learnt to improve the understanding of these transitions? The question of the treatment of fuzzy time in the analysis models is then raised. Beyond duration models, what approaches should we attempt, via what type of observation?

6With the aim of addressing these questions, the Groupe de Réflexion sur l’Approche Biographique (GRAB) working group brought together around 30 researchers from a range of disciplines in November 2003 to take part in three study workshops organized jointly by CEPED, INED (“event history and multilevel analysis” methodological unit) and IRD (UR 47-DIAL).

7Picking up on the key questions debated in the workshops, a number of groups were set up around particular thematic and/or methodological aspects, each with the task of producing a chapter of this book. These groups worked together using an innovative system of collaboration, often between distant geographical locations and different disciplinary standpoints, to address these questions through a collective endeavour. Such exercises are extremely rare in the social sciences: there are few successful examples of collective capitalization of ideas produced through more personal research, or specific to a small team. In fact, despite the very varied range of fields and topics, the texts produced display a remarkable degree of consistency, testifying to the pertinence and the value of the exercise. The book thus comprises six thematic chapters with collective authorship.

8Chapter 1 covers the topic of data collection and of how to deal with respondents’ perceptions, both in general and in relation to their life-event history in particular. This key question, mentioned again in the following chapters, is central to the analysis of fuzzy states. Chapter 2 demonstrates the importance of capturing the entire process leading from union to disunion, institutional time being distinct, moreover, from relational time. Chapter 3 covers the question of sexuality and fertility via two specific examples: unwanted pregnancies in France and prevention of mother-to-child HIV transmission in Côte d’Ivoire. On the basis of these two case studies, the authors examine the temporal aspects of relations between sexual partners/parent partners. Chapter 4 addresses fuzzy states via the complex notion of hébergement (being housed). Taking account of the transitions described earlier, Chapter 5 examines the pertinence of markers and the existence of founding events in the transition to adulthood. Chapter 6 deals with the question of individual mobility in relation to the context in which it occurs, introducing a necessarily multilevel approach.

9This book targets a much broader audience than the community of demographers: it concerns all social scientists who, at one moment or another, need to collect data in the field, to undertake a critical assessment of the data obtained, of the analysis categories, of the event history timeframe, and of the gradual or complex transitions observed. Fortunately, individuals do not remain frozen in a single state throughout their life. Analysis of the diversity of states experienced (residential, occupational, conjugal and familial) raises new questions about the social categories used by social scientists. Throughout this book we invite readers to rethink these concepts and categories.

10We would like to extend our thanks to Harriet Coleman for her proficient translation of the original French texts, and to Duncan Fulton for his careful rereading of the English manuscript.