1Research into turning points in people’s lives is a longstanding preoccupation in life course studies, whether qualitative or quantitative (Elder, 1977, 1995; Ricœur, 1985; Hareven, 1986; Leclerc-Olive, 1997; Thomson et al., 2002; Lelièvre et al., 2006). Depending on the approach adopted and the type of material in hand, this kind of research may be based on purely factual data, or use subjects’ more personal interpretations or, more rarely, combine the two. Quantitative life course surveys, which often include observation of qualitative factors obtained through multidisciplinary collaboration (GRAB, 1999), are one way to explore the subtle links between factual and perceptual data. These quantitative surveys take a retrospective life history approach, calling for the reconstruction of a “history” which, although structured by the questionnaire, is nonetheless largely narrated by the respondent (Bertaux, 1997). The data gathering procedure uses a semi-structured interview to reconstruct, step by step, the main events and stages in the respondent’s residential, family/conjugal and occupational history (including training and any periods of unemployment).

2The Biographies et entourage (event histories and contact circle) survey, in addition to reconstructing the factual events punctuating respondents’ family, residential and occupational histories, also recorded facts about the life courses of close friends and relatives (see Box) and asked respondents for a more personal summary of their life course. With this approach it was possible to take into account how respondents viewed their lives in retrospect and record whatever they regarded as outstanding personal or historical events. Each respondent was asked to divide their life into periods, identifying, explaining and interpreting the significant periods, turning points and general tone of their life course and what they viewed as landmark events (Lelièvre and Vivier, 2001; Tichit and Lelièvre, 2006; Laborde and Lelièvre, 2006). In this way, the questioning could be extended to aspects of the respondents’ lives and concerns that were not addressed in the rest of the questionnaire, thereby enriching the analysis of individual trajectories.

3The data gathered in this way represent a unique resource and open up new and particularly stimulating research channels. They can be used to compare systematically gathered factual data with the perceptual data recounted and interpreted by the respondent. This paper first summarizes the data collection methods used and then gives a preliminary analysis of perceived trajectories and turning points in comparison with the factual events recorded.

Recording perceptions of the life course

4The study does not just link two approaches, two data collection methods and two different types of material (qualitative and quantitative). It incorporates qualitative and quantitative approaches in the same interview and the same data gathering instrument (Figure 1). First the event histories of respondents and the members of their close contact circle are recorded without asking for explanations, motivations or meanings. Then, at the end of the interview, respondents are asked to divide their life into periods and explain them in terms of the following guideline questions:

5Now, using the grid we have filled in together, we’d like to know how you interpret its contents (Synthesis 1, Sy1). To do this, we suggest the following approach:

6Sy1 • Can you divide your life into different periods? Characterize these periods, identifying what differentiates them from each other and what they represent in your life.

7At this stage of the interview, the respondents have already evoked the stages of their life and those of their close contact circle, so all the milestone events can be readily brought to mind.

8Once they have divided their life into periods and have differentiated and characterized these, they are asked to grade each period on a scale of appreciation (Sy2):

9Sy2 •For each period, were these

10VG: very good years

11G: good years

12PF: problem-free years

13D: difficult years

14VD: very difficult years

15Lastly, a record of outstanding events (Sy3) adds more contextual dimensions (war, for example) or personal factors (such as an illness) that were not recorded in the rest of the questionnaire:

16Sy3 • Are there any outstanding personal or historical events that have marked your life?

17The questionnaire extract in Figure 1 illustrates and synthesizes the information collected. The factual milestones (family, residential and occupational) of Paulette’s life [1] are shown in the columns on the left, and her personal “syntheses” (Sy1 and Sy2) and what she regards as outstanding events in her life (Sy3) are shown in the columns on the right.

Figure 1

Example of a respondent’s summary of her life, recorded in the life course grid opposite the family, residential and occupational event history (Biographies et entourage survey, 2001)

Example of a respondent’s summary of her life, recorded in the life course grid opposite the family, residential and occupational event history (Biographies et entourage survey, 2001)

18Paulette was born in Montreuil in 1944. She grew up there until she was 12, when she left to be apprenticed to her uncle, a dry cleaner, in the 12th arrondissement of Paris. At 14 she started work as a cashier in the Prisunic department store. At 16 she married a workmate at Prisunic and they settled in Rosny-sous-Bois. Her husband took over a grocery shop there and she worked with him. Their first two children were born in 1961 and 1962, a third five years later. In 1970 she separated from her husband, returned to Paris and got a typing job. The divorce was pronounced nine years later. The same year, Paulette set up home with André; they married in 1985. A serious accident forced Paulette to stop work. She did not find work after that and took early retirement at the age of 52.

19Paulette distinguishes eight periods in her life. First her childhood, in which she points out the absence of her mother; then her adolescence, marked by her early entry into the labour force. She describes both these periods as “difficult” (Sy2). Her marriage at a young age and the arrival of the children was a happy time, but the conjugal problems that led to separation from her husband marked further difficult years (Sy2). On meeting her second partner André (Sy3), “the start of my life as a woman” (Sy1), was a good period followed by a very happy time when they were living together. Unfortunately the accident, stopping work and the resulting sense of disablement marked difficult years, though the situation improved after André’s own retirement.

20This interpretative reading of one’s own life course and the analysis that can be made of it are delicate matters. The “important” events in a life are important at a particular moment; the division into periods is only valid in the light of the years already lived, i.e. at a given date. The responses of a thirty-year-old just starting a family cannot be compared with those given later, on the eve of retirement. Different individuals’ retrospective evaluations can therefore only be compared (a necessary condition for quantitative analysis) if they are made at an equivalent moment in their respective life courses. Here, all the respondents were questioned at the time of their retirement or not long before or after, so there was some temporal homogeneity among their retrospective evaluations.

21Reports can be influenced by other factors too, especially particularities of the trajectory. We will examine how the repetition of some stages or events (several children, several unions etc.) or the occurrence of some outstanding event (such as international migration) impact the respondents’ life course reconstructions.

22More generally, in this paper we explore the phases that structure the life course from the respondent’s own point of view. Do the turning points they describe correspond to factual life events recorded? If so, which ones? What outstanding events do people mention?

Life course phases and their assessment

On average, four distinct periods

23Respondents defined the various periods of their lives in very different ways. Some identified 12 distinct periods in their lives, while others found only one, regarding their lives as “a long quiet river” (the latter were a small minority, however) (Table 1). A quarter of trajectories were divided into four periods, 20% into 3 periods and 20% into 5. So altogether two-thirds of respondents divided their lives into either 3, 4 or 5 periods.

Table 1

Respondents’ trajectories by number of periods mentioned

Number of periods cited 0, NR 1 2 3 4 5 6 7 8 9 10 or more All Number 31 103 259 593 705 570 305 144 66 30 24 2,830 % 1.1 3.6 9.2 21.0 24.9 20.1 10.8 5.1 2.3 1.1 0.8 100.0 NR: non-responsePopulation: all survey respondents (N = 2,830 male and female Île-de-France residents aged 50-70).Interpretation: 24.9% of Île-de-France residents aged 50-70 distinguish 4 periods in their lives.Source: INED, Biographies et entourage survey, 2001.

Respondents’ trajectories by number of periods mentioned

24To identify some simple determinants that may influence the division into periods, we modelled, separately by gender (Table 2), the probability of a person structuring their trajectory into more periods than the median of 4, according to various socio-demographic characteristics (such as birth cohort, socioeconomic group, educational level etc.) and variables that may correspond to a shift to a new period (numbers of different occupations, dwelling changes, children or deaths).

Table 2

Factors influencing the probability of citing a larger number of life course periods than the median value (odds ratios of logistic regression model)

Men Women Age 49-54 0.79 0.86 55-59 (Ref.) 1.00 1.00 60-64 1.14 1.09 65-71 1.20 1.41 * Mobility 1-5 dwellings occupied for at least a year (Ref.) 1.00 1.00 6-10 dwellings occupied for at least a year 1.49 * 1.52 ** 11 dwellings or more occupied for at least a year 1.99 *** 2.01 *** Education No schooling 0.78 0.81 Primary school or middle school 1.20 1.14 Less than baccalauréat (Ref.) 1.00 1.00 General or technology baccalauréat 1.50 * 1.22 University degree or grande école diploma 1.69 *** 2.00 *** Marital status Single 1.21 0.88 Married (Ref.) 1.00 1.00 Divorced or separated 2.36 *** 1.58 ** Remarried 1.28 1.48 Widowed 3.21 ** 1.72 ** Number of jobs 5 or less 0.71 * 0.62 ** 6-8 (Ref.) 1.00 1.00 9 or more 1.44 ** 1.37 * Numbers Persons citing fewer periods than the median value 756 904 Persons citing more periods than the median value 571 571 Total 1,324 1,475 Statistical significance: ***: p<0.001; **: p<0.01; *: p<0.05Population: 2,799 persons (out of total 2,830), i.e. those who answered question Sy1 (see above).Source: INED, Biographies et entourage survey, 2001.

Factors influencing the probability of citing a larger number of life course periods than the median value (odds ratios of logistic regression model)

25For example, one might think that the number of children respondents have or the number of deaths in their close contact circle would affect the number of periods they describe. In fact they do not [2]. The odds ratios show that the number of periods defined in men’s and women’s summaries increases according to their degree of geographical mobility and the frequency of changes in their occupational history. Among women, educational level also has a strong effect. But the most surprising finding is the marital situation effect: being divorced or widowed increases the likelihood of dividing one’s life into more periods than the median number, whereas remarrying, even though it introduces a new family event, does not produce an increase in the number of periods in the summary. This observation is particularly marked for men.

Definite, positive qualifiers

26As described above, the tone of the life course was assessed using the synthetic indicator (Sy2) which respondents attributed to each life period [3].

27All in all, respondents’ appreciations of different periods of their lives are fairly definite (Table 3). The most extreme grades (very good / very difficult years) account for a third of appreciations while the most neutral indicator (problem-free years) was much less commonly attributed (15%). The crude distribution of indicators shows that more than half of all appreciations are positive (57%). On average, these “good” and “very good” years cover 37 years of life ? two-thirds of the average duration of respondents’ trajectories ? and they are not preferentially concentrated on any specific period of life. In particular, the positive appreciations are not centred on the childhood years. This finding points to the credibility of these subjective retrospective assessments: they are not stereotyped, more-or-less predictable or nostalgic reconstructions of a romanticized childhood, for example. Slightly more than a quarter of appreciations are negative, classifying periods as “difficult” or “very difficult”.

Table 3

Respondents’ appreciations of their life course periods

Appreciations by order of frequency G VG D PF VD Overall Numbers 3,897 2,943 2,403 1,788 920 11,951 % 32.6 24.6 20.1 15.0 7.7 100.0 Mean total duration of periods (years) 21.1 15.9 8.7 9.5 3.0 58.2 Population: all respondents who answered the question (N = 2,799 male and female Île-de-France residents aged 50-70). Interpretation: 24.6% of periods were classed as “very good” (VG) and cover a mean total of 15.9 years. Source: INED, Biographies et entourage survey, 2001.

Respondents’ appreciations of their life course periods

28To analyse all these subjective trajectories, we considered each one as a “word” whose syllables are the initials of the appreciations made of each period. For example Michelle, aged 60, divided her life into three periods: the first was “problem-free” (PF), the second “good” (G) and the third “very difficult” (VD). So her whole trajectory can be summed up in the “word” PF-G-VD. In this case, each period has a different tone. This is the case for half the respondents’ trajectories, each change of period being marked by a change of tone (Table 4). In other cases, successive periods were given the same assessment. For example Lucy, aged 64, had also divided her life into three periods, but described them as VG-VG-VD. To establish the tone for the whole life course, we compacted these appreciations, only taking account of the assessments which differ. Where the same “syllable” was applied to successive periods we considered it as one, so Lucy’s life course is summed up in the “word” VG-VD. This rule had no effect on Roger’s trajectory, in which two periods assessed as PF were not consecutive: PF-G-PF remains as such.

Table 4

Distribution of appreciations of individual trajectories according to division into distinct periods

Number of distinct successive appreciations (%) Number of periods 1 2 3 4 5 6 or more 123456 or more 25.827.325.610.57.33.5 –32.534.122.98.22.3 ––44.634.716.64.1 –––52.134.713.2 ––––53.246.8 –––––100 Distribution of appreciations 14.3 16.5 26.8 20.4 12.2 9.8 Population: all survey respondents (N = 2,830 male and female Île de France residents aged 50-70, who described their lives in terms of one or more appreciation). Interpretation: 10.5% of trajectories characterized by only one appreciation were divided into 4 periods (e.g. G-G-G-G). Source: INED, Biographies et entourage survey, 2001.

Distribution of appreciations of individual trajectories according to division into distinct periods

29The distribution of trajectories summarized in this way is homogeneous, with a maximum clustered around 3 and 4 different appreciations. On average, Île de France residents describe their lives with slightly under four different appreciations. Of those respondents who gave the same appreciation throughout their trajectory (14% of respondents), nearly equal numbers divided their lives into one, two and three periods (about a quarter in each case). As this shows, the way people divide up their life courses is not only related to changes in the tone of their life.

30It should also be stressed that the most frequent subjective trajectory is positive: 18% of trajectories are described as consistently “good” and/or “very good”. This is also the predominant appreciation over the course of a subjective trajectory: on average, 64% of each respondent’s life up to the survey date is described in positive terms. Conversely, entirely negative trajectories (“difficult” or “very difficult”) are relatively rare (only 1% of life courses). Those that alternate between highs and lows (two or more different appreciations) represent, overall, 78% of subjective trajectories and involve a very large number of different combinations. But for the majority, the overall tone is positive despite the ups and downs.

31This preliminary analysis of the information shows that respondents’ overall appreciations of their life courses, even in summary, are very diverse. Even smoothing the nuances of assessment by merging “good” with “very good” and “difficult” with “very difficult” (but leaving “problem-free”, as a separate class) and only taking the more common trajectories (representing more than 1% of cases), the new classification only covers just under half of the respondents’ trajectories (46%). To look for regularities in these trajectories therefore calls for more detailed analysis, and constructing a typology will require a more rigorous method, such as harmonic analysis (Robette and Thibault, 2006) or classification by optimal matching (Lesnard, 2006) so as to process all the trajectories rather than just half of them. At a later stage, textual analysis of respondents’ explanations of each period will also provide a means to interpret these subjective trajectories more finely.

Subjective division of life course compared with factual milestones

32The second dimension of our analysis is to study the correlation between the factual events punctuating respondents’ residential, occupational or family histories and the boundaries they identify between the different life periods. In particular, do the events demographers commonly use as significant markers of the “ages of life” (such as marriage, the birth of a child or migration) actually match the personal turning-points respondents identify as meaningful for them?

33We find that, on average, individuals’ life courses include some twenty factual events (moving home, occupational changes, family events), whereas when respondents summarize their lives into periods they distinguish an average of four periods. To the nearest year [4], 35% of the factual events correspond to a change of period. In other words, only one-third of factual events coincide with a turning point mentioned by the respondent. Conversely, on average, the start of each period identified by a respondent usually does coincide with a factual event, which confirms, as expected, that there is an association between factual milestones and the subject’s identification of a passage from one life period to another.

34We will now examine more precisely the nature of the events more often associated with the start of a new period, bearing in mind that their frequency is also variable. On average, respondents had experienced seven occupational changes, slightly more than one union and two births, and had moved home eight times.

35The more detailed analysis shows that, given their high frequency, moving home and changing jobs have a non-negligible impact on how people reconstruct their life histories: 37% of occupational changes and 33% of changes of dwelling matched the start of a new period. Half or more of conjugal events were mentioned as turning points: 59% of union formations and 50% of union dissolutions coincided with new periods in the life history summaries. Other family events were less likely to start a new phase: 27% of births and 22% of deaths. The relatively low importance of births is partly due to the fact that, on average, there are more than twice as many births as marriages in a life course, and that the impact of a family event on the life course reconstruction varies according to its order. It is mainly first order births that mark turning points ?42% of first births compared with 16% of subsequent births. Similarly, 63% of first marriages mark the start of a new life phase compared with 32% of subsequent marriages. Rather than the event itself, it is the new state initiated by the event (parenthood, the married state) that seems to have a marked structural impact on the life history reconstruction. Nonetheless, even in this perspective, entry into parenthood only marks a new period for just under half of those respondents who have had children.

36As regards deaths (which were of similar frequency to marriages), it is the relationship with the deceased that seems to make the difference. Thus 49% of deaths of a spouse, 40% of deaths of a child and 20% of deaths of a parent coincided (to within a year) with a new period in the summary.

37These observations suggest that the subjective meaning and weight of life events in respondents’ reconstructions are neither stereotyped nor predictable. In particular, one might have expected births and entry into parenthood to have a stronger effect in structuring the reconstituted life history. The singularity of each person’s view of a given life event highlights the value of collecting subjective reconstructions and moving away from purely objectifying life course analyses (Hareven and Masaoka, 1988; de Coninck and Godard, 1990).

38To examine more closely these links between factual events and subjective division into periods, we will focus on the impact of migration from one country to another ? an event one might suppose to be a potentially important turning point and the marker of a new phase in a person’s life.

39Of the people in our sample, 632 were born abroad. Examining whether or not the first arrival in France coincided with the start of a new period in the subjective summary, we find that only half these foreign-born respondents associated the start of a new life phase with their first arrival in France. Those who did identify their migration as a turning-point, or at least the start of a new life phase, tended to be men arriving at a slightly older age that the rest (at around age 24), and single. Country of origin and cultural region did not emerge as discriminating factors in the propensity to declare arrival in France as a significant moment marking a new period of life.

40Another interesting fact emerges with regard to the most highly mobile individuals, who had also made other international migrations and declared several periods in their subjective summaries. Half of those who identified a new period starting at the time of their arrival in France also did so for each move to a new country. In these cases, each move to a new country opened a new phase in their life. While it might be thought that the impact of migration would decrease with repetition (as is the case with births and unions), this is not the case. Even where the coincidence is not systematic, all the most mobile migrants matched at least one migration (other than their first arrival in France) with the start of a new period. And of those who had not mentioned the start of a new period at the time of their first arrival in France, three quarters did not mention a new period for any other migration either. So there seem to be some people for whom geographical mobility initiates biographical turning points (more or less systematically), and those for whom it never or rarely does, even though they have changed country several times. Geographical change and its structuring dimension seem to be apprehended very differently from one individual to another. Of course, the two migrant profiles would need to be determined more precisely, but these findings already show the relativity of the event in itself and of the importance attributed to it in a life history.

Outstanding events

41As well as asking respondents to divide their life trajectories into periods, the preliminary tests of the Biographies et entourage survey encouraged us to introduce an additional question about outstanding events in their lives. The division into periods assumes that the periods will have a certain homogeneity, but a specific event that is highly significant to the respondent may (artificially, so to speak) alter a period that in other respects they see as homogeneous. We therefore left a space for respondents to report outstanding events independently of their summary in terms of periods (as in the Triple biographie survey). These personal or historical events open the investigation to dimensions that have not been explored elsewhere. The question was:

42Sy3 • Are there any outstanding personal or historical events that have marked your life?

Wars leave their mark

49Let us now look at the public events which respondents mentioned as having an impact on their lives. By their nature, they are not addressed in the factual part of the questionnaire, but they form the background to a person’s life. Recall of historical events and their incorporation into the construction of personal time has been studied by psychologists, for example on the occasion of the transition to the year 2000 (Middleton, 2000). It will be necessary, of course, to examine more deeply the profiles of those respondents who mentioned particular events compared to other respondents who could have done so but did not. However, the preliminary analysis presented here shows that, overall, 22.4% of all outstanding events mentioned were public events.

50The most frequently mentioned were armed conflicts: 44% of public events cited. A generation effect distinguishes the younger respondents from the older ones (Table 5): 5% of respondents aged 49-54 mentioned a war event compared to 17% of those aged 63-71. Next come references to various social movements or episodes (35%), then political events (14%), which range from the fall of Phnom Penh to the Carnation Revolution in Portugal for example.

51More precisely, World War II accounted for 19% of public events cited and social movements 18% (these included Woodstock, May ’68, strikes and demonstrations in general, France’s health insurance problems and the great social changes of the 1970s and 1980s). The Algerian war came next (13%), then technical and social progress (8%) such as the conquest of space, Concord, medical advances, the 39-hour week and the fifth week of paid leave in 1982. French political movements accounted for 5% of the total.

Table 5

Events concerning “war” as a proportion of outstanding life events, by age

Respondent’s age (yrs) 49-54 55-62 63-71 All Total number of events cited 3,189 3,328 2,674 9,191 Events connected with a war 170 261 467 898 as % of total: 5.3 7.8 17.5 9.8 Source: INED, Biographies et entourage survey, 2001.

Events concerning “war” as a proportion of outstanding life events, by age

Men and women do not cite the same events

52In line with research that has compared life course turning points by gender (e.g. Rönka et al., 2003), we investigated whether the respective male and female spheres of interest had an impact on the types of event cited as outstanding. Do women mostly cite events concerning private life (personal and more specifically family events)? Do men more often cite public and work-related events?

53As regards personal events, women do indeed seem more sensitive to family issues. Nearly 59% of events they cited are family events, compared with 50% for men (Figure 2). So men appear to attach slightly less importance to family and slightly more to other areas with 26% of men citing public events compared to 20% of women. In particular, they more often cite armed conflicts or related events. However, men and women who were children or adolescents during World War II were equally liable to cite outstanding events connected with that war. It is in relation to colonial wars (Algeria and Vietnam) that the gender difference emerges; men who experienced those wars when young cited them more often than others.

Figure 2

Frequency distribution of outstanding event cited, by gender

Frequency distribution of outstanding event cited, by gender

54These results confirm our hypothesis and show that there are two differences between men and women in this regard. Men are slightly more likely than women to mention the public sphere rather than private life, while within the private sphere they cite work-related events more often than women do, family events less often.

Conclusion

55This study has explored several ways of recounting one’s life and of analysing life courses. Whether the subjects report the factual events that have punctuated their lives or summarize the broad lines and turning points and cite outstanding events, the researcher is faced with an abundance of material that is not readily reducible. However, our first investigations show that life course trajectories are largely structured around the residential, family and work-related events that are systematically recorded. From this standpoint, conjugal life holds a key place. But respondents also subjectively use other personal or public factors (health, religion, wars etc.) to interpret and relate their life stories. A thorough analysis of the explanations respondents give in their life course summaries, beyond the mere division into periods and the outstanding events cited, will provide a better understanding of the impact of each of these different dimensions of the life course and the way they interconnect from the respondent’s point of view.

56Overall, analysis of the synthetic indicators by which each respondent qualified the different periods of their lives already shows that despite alternating between happy periods and more difficult ones, these life courses unfold for the most part on a positive note. We note, in particular, that the way an event is experienced and characterized is not a foregone conclusion. The summaries of the questionnaires show, for example, that ostensibly difficult times such as war can be experienced in singular ways and may be characterized in positive terms. For example, some respondents admitted (with some embarrassment) that they had fond memories of the great exodus of 1940, which took them from their urban home surroundings to the countryside. Similarly, the example of international migration shows that a change of environment, even a radical change, can be experienced very differently by the respondents concerned and have a varying degree of significance in their lives. A migrant may focus more on some other role or aspect of their life at the time which is more significant for them and so feel a sense of continuity where an outside eye may see a turning point or major break in the trajectory. The meanings individuals confer on events and on the unfolding of their lives is neither stereotypical nor predictable. This makes it all the more useful and informative to perform this type of data collection and comparative analysis, combining both objective and subjective dimensions. This study also shows that such an approach is possible within the framework of a quantitative survey, and redefines the frontiers between the qualitative and quantitative approaches.