This is an Author's Accepted Manuscript of an article published as

Habermas, T. (2007). How to tell a life: The development of the cultural concept of biography across
the lifespan. Journal of Cognition and Development, 8, 1-31. 10.1080/15248370709336991

© Taylor & Francis, available online at: http://www.tandfonline.com/doi/full/10.1080/15248370709336991

How to tell a life:

The development of the cultural concept of biography

Tilmann Habermas
Goethe University Frankfurt a. M., Germany

Acknowledgments. Study 1 was supported by the German Research Council (DFG) as part of a study on the
development of life narratives #HA 2077. Thanks go to David Rubin and Susan Bluck for discussions of the
original ideas, to Cybèle de Silveira, Verena Diel, and Martha Havenith for data collection, entry and coding, to
Dorthe Berntsen, Rebekka Messinesis, and Susanne Döll for helpful suggestions on the manuscript, and to Anna
Pate for great help with language and style.

Address: Please send correspondence to the Institute of Psychoanalysis, Department of Psychology, Goethe
University, PO Box 111932, 60054 Frankfurt a.M., Germany, or via e-mail to [email protected]
frankfurt.de . Fax +49 69 798 28584

Abstract
Extending research on age norms in adults, the development of the knowledge of two components of
the cultural concept of biography, biographical salience of and age norms for life events, was studied
from late childhood to early adulthood in Study 1 and across adulthood in Study 2. The largest
increase in knowledge was found between ages 8 and 12, with knowledge reaching its maximum at
age 16. Across adulthood knowledge was relatively stable, with a small decline in older adults. In
addition, across adolescence personal memories increasingly corresponded to norms of biographical
salience, although idiosyncratic events continued to dominate. The acquisition of knowledge of the
cultural concept of biography may parallel developments in autobiographical memory, reminiscing,
and life narratives.
T. Habermas How to tell a life 2

Everyday conceptions of the life course are used to predict, control and understand development in others and
oneself. When narrating one’s life or constructing others’ biographies, cultural conventions of how to construct
a biography are used. We have introduced the term cultural concept of biography to cover the whole set of
heterogeneous cultural biographical conventions (Habermas & Bluck, 2000). These conventions were grouped
into a concept because they generalize across specific biographies, defining essential features of any account or
narrative that may be considered a biography. It is nominated a cultural concept to underline the cultural and
historical relativity of many of its aspects. The term biography covers all linguistic acts and products that refer
to a temporally extended human life as a frame of reference, such as life narratives, curricula vitae, written
biographies and autobiographies, and also more partial biographical accounts.
The cultural concept of biography comprises, for example, conventions of how to introduce and end a life
narrative (Habermas, in press), how to deal with the less well remembered parts of childhood, and what to
accomplish with a biography (e.g., demonstrating that one has lead a virtuous life or being able to reconstruct a
coherent personal development). In this paper two further aspects of the cultural concept of biography will be
studied which relate to the events included in biographies, conventional biographical salience of and age
norms for life events. A methodological advantage to studying these aspects is that they may be studied
independently from actual biographies
Linde (1993) defines a life story by characterizing the events included in it as making a point about the
narrator and as having extended reportability. To convey the narrator’s individuality and to be reportable an
event must deviate from the normal and expected, that is, it must be idiosyncratic. Characteristics of significant
personal memories described in the literature, such as turning points (Pillemer, 1998), highly emotional events,
and events linked to unresolved conflicts (Singer & Salovey, 1993) and to enduring personal concerns (Conway,
Singer, & Tagini, 2004) appear also to support the contention that they are rather individual events.
However, Linde (1993) adds that there is also a conventional set of highly reportable (biographically
salient) landmark events, such as normative transitions and major nonnormative events, which are used to
structure life stories. Normative transitions have age norms (Neugarten, Moore, & Lowe, 1965) which define
developmental deadlines (Heckhausen, Wrosch, & Fleeson, 2001) and the social clock (Helson, Mitchell, &
Moane, 1984) by creating a sequential order among normative life events. This skeletal life course may be used
for organizing life narratives by constraining the narrator to discuss progress through life phases and to justify
or excuse any major deviations (Schütze, 1982). Rubin and Berntsen (2003) term this sequence of normative
transitions life script. They extend the organizing function of the cultural concept of biography from life stories
to autobiographical memory by claiming that most important memories are selected not primarily on the basis
of personal significance but are strongly influenced by the life script. They argue that because most events in
the life script have age norms that fall into late adolescence and young adulthood, this explains the
reminiscence bump in the years 15 to 30 in autobiographical retention curves of people over 40 years old. Also,
as transitional events are positively valued, this explains the typical bump of memories for the most positive
memories, whereas the age-independent distribution of nonnormative events explains different curves for
memories with negative emotions attached to them.
Biographical salience will be used to denote the conventional biographical salience, and not individual
biographical salience. Research on age norms and biographical salience in adults and especially on their
development is scarce.
Adults of a given culture and society largely agree on age norms for specific major role transitions. In their
classic study Neugarten, Moore and Lowe (1965) asked middle-aged to elderly adults to indicate the “best age”
for a series of transitional or peak events from the work and family sphere, such as finishing school, starting to
work, settling on a career, marriage, and reaching the peak of one’s career, resulting in varying but generally
high degrees of agreement. Later studies also confirmed the existence of age norms with different items (cf.
Settersten & Mayer, 1997), even when the question was rephrased to ask not for personal opinion on best ages
or for typical ages, but for normative ages which imply prescriptive social expectations and related sanctions
(Settersten & Hägestad, 1996a, 1996b). Age norms for role transitions most probably vary across societies, as
may be inferred from differences in the factual timing of transitions, for example, in institutional transitions
between East and West Germany (Silbereisen & Wiesner, 2000) or in historical shifts such as in the normative
T. Habermas How to tell a life 3

timing of familial transitions (Hareven, 1986).

Besides role transitions, age norms may also regard the development of skills, which is treated in a
separate body of literature, usually speaking of developmental expectations (D’Alessio, 1990). Normative ages
for the emergence of skills such as walking, swimming, reading and writing are agreed upon within cultures,
but vary between cultures (Willemsen & van de Vijver, 1997). These motor, language and cognitive skills are
relevant mostly for the childhood period, whereas normative developmental expectations for adolescents
focus more on issues of autonomy and romantic relationships (Dekovic, Noom, & Meeus, 1997).
Only four studies have attempted to describe the development of knowledge about age norms. Greene
(1990) asked 104 16- to 21-year-old high school and college students about the events they expected to
happen in their personal future. Results were contradictory, in that college students named more normative
transitions to adulthood, but agreed less on their typical ages. In a second study, Greene, Wheatley, and Aldava
(1992) presented six transitional events to 80 high school and 80 college students aged 17 and 20 years. This
closed format produced the expected results: The college group agreed more and gave later average ages than
the adolescents. Not only may the latter difference result from later actual occurrence of transitional events in
college-educated segments of the population, but it is also not clear whether greater agreement regarding
average ages for transitions in college students indeed reflects the acquisition of knowledge of age norms or is
the effect of greater social homogeneity in the college group.
Another study concerned the typical sequence of life events (Strube, Gehringer, Ernst, & Knill, 1985). A
total of 46 participants, distributed over three age groups, was asked to sequentially order a stack of cards with
life events. Preadolescents (age 11) ordered the events according to the normative sequence with some
deviations. Adolescents (age 19) accomplished the task perfectly, whereas adults (35-40 years) again showed
more variation. In contrast to the preadolescents, the adults, however, did not violate established cultural or
logical conditional relationships between items, and also performed better than preadolescents at sorting life
events into five age periods in a separate task. Strube et al. (1985) concluded that whereas the preadolescents’
performance demonstrated their imperfect mastery of age norms, the adults conveyed a more realistic
variation in the life course than did adolescents, whose depiction was highly stereotyped.
This pattern of results was confirmed in a study of knowledge about age-typical psychological and social
development. A total of 236 participants, aged 4 to 17 years, assigned 48 items concerning psychological traits,
maturational characteristics, and social events to one of eight life phases. Correct attribution of events to life
phases increased up to age 13 (Schorsch, 1992).
Summing up the four studies, even preschool children show some competence in the relatively easy task
of assigning events to life phases, which was mastered at age 13 (Schorsch, 1992). The more difficult task of
sequencing life events is mastered by the end of adolescence (Strube et al., 1985). This is compatible with the
finding that on the more difficult task of estimating ages young adults agree more than late adolescents
(Greene et al., 1992). Due to disparate age groups and methods, the precise course of the development of
knowledge of typical ages of life events remains unclear. Taken together the findings are best compatible with
the hypothesis that a basic knowledge of age norms is acquired by age 13, which is further refined up to about
age 20.
For the knowledge of the conventional biographical salience of events, a relevant corpus of research is
provided by studies using lists of biographically consequential life events. In a recent study the events judged to
change life the most were the nonnormative events of death of a spouse, divorce, death of a close family
member, losing one’s job, major illness or injury, jail, and death of a close friend, with high degrees of
agreement between raters (Miller & Rahe, 1997). Normative life events such as marriage, retirement and
pregnancy were also included, but were rated as imposing less change than the events mentioned first.
Consequential events may be different for children and adolescents, including death or divorce of parents,
parent losing a job, or flunking school, as may be inferred from life event lists for children (cf. Geyer, 1992a).
Asking more directly for biographically salient events, Berntsen and Rubin (2004) had 103 students aged
21 to 54 years name the seven most important events which were to be expected in the future life of an
ordinary infant of the respondents´ own sex. Eleven of the 20 most frequently named events were institutional
and familial transitional events, whereas five were nonnormative events and one was developmental in nature,
T. Habermas How to tell a life 4

with three more events falling in between maturational development and familial transitions (fall in love, first
friend, puberty). Typical ages named by the participants varied widely only for the nonnormative events,
suggesting that normative events have age norms. Summing up what is known about conventional biographical
salience, we can state that it comprises both normative and nonnormative events, but we know nothing about
its development.
Conceptually, the two aspects of the cultural concept of biography are distinct, with biographical salience
defining a class of events, and age norms defining a class of events and the ages at which they should occur.
The research reviewed above suggests an empirical overlap between the two classes of events. Events
belonging into both classes are normative life events with age norms, most often social transitions, which are
biographically salient. Some normative events with age norms may not be biographically salient; other age
norms regard developmental milestones, such as learning to speak or read, which are rarely biographically
salient. On the other hand, some biographically salient events, namely nonnormative events, do not have age
norms.
Some evidence points to the development of knowledge of age norms between late childhood and young
adulthood. What we know about the development of autobiographical remembering could suggest that the
knowledge of biographical salience develops in the same age range. Early development of autobiographical
remembering focuses on sequences of events and single narratives (cf. Bauer, 2002), whereas in childhood
autobiographical remembering uses a narrative format (cf. Nelson & Fivush, 2004), which is still restricted to
single events. The ability to use calendar time to date events develops in late childhood and early adolescence
(Friedman, 1993), whereas a biographical format for remembering and narrating seems to emerge only
between late childhood and late adolescence (Habermas & Bluck, 2000). This suggests that biographical
knowledge is acquired in this age range.
The acquisition of biographical knowledge could proceed in two ways. Schorsch (1992) argued that
normative life events near the respondents’ own age are highly salient, with past events having been salient in
the past, but future events being of no practical significance. She therefore hypothesized that biographical
knowledge is greater for past and current than for future life phases. According to this model, age norms and
biographical salience are learnt for each event separately as its normative age approaches, be it by hearing
from and watching others or by personal experience. This model allows for the use of biographical knowledge
for autobiographical remembering and for the construction of one’s own biography, which ends in the
extended present, but is incongruent with an understanding of and ability to construct biographies that extend
beyond one’s own age.
The cultural concept of biography, in contrast, implies an integrated concept of the life and therefore
requires that biographical knowledge is acquired as an integral concept, with knowledge about events from all
parts of life being learnt at the same rate. In a similar vein, Berntsen and Rubin (2004) argue that a life script
requires being learnt in all its sections equally well. Whereas the former, weaker hypothesis predicts a
continuous increase of biographical knowledge, the latter, stronger hypothesis predicts a steep increase some
time in adolescence and relative stability across adulthood. The only pertinent evidence to date supports the
latter, hypothesis: between ages four and thirteen age norms are learnt equally well for all life phases
(Schorsch, 1992).
The following two cross-sectional studies were undertaken to explore how knowledge of the cultural
concept of biography develops with age using two of its components, biographical salience of and age norms
for events. The first study covers the period during which most development of biographical knowledge is
expected to take place, the time between childhood and young adulthood. The second study tests the stability
of biographical knowledge across the adult age range.
Three hypotheses will be tested with the data of both studies. The first hypothesis is that knowledge of
the cultural concept of biography increases most between childhood and adolescence to peak at age 20 and to
remain stable in the adult years, with, following Strube et al. (1985), some more variance as the reality of
individual life courses is experienced. The second hypothesis is that as a unitary cultural concept of biography is
acquired, knowledge increases for events from all life phases to the same degree. The third hypothesis is that
the set of conventionally biographically salient events influences the selection of events when remembering
T. Habermas How to tell a life 5

and narrating one’s own life, and that this influence increases as biographical knowledge is acquired.
Both studies used two measures of biographical knowledge which ask for the biographical salience of and
age norms for two sets of events. A preliminary study served to construct these tasks by selecting items and
establishing salience and age norms. Because we wanted to test knowledge, a closed format was chosen to
render answers more comparable between respondents, to test knowledge across the lifespan, and to test
both salient and nonsalient events. In addition, two free nomination tasks were used to measure selection of
events, one concerning seven most important personal memories in Study 1, the other asking for seven
conventionally most salient events in a normal life in Study 2. This latter set of events from Study 2 was used to
judge how much personal memories (Study 1) were influenced by the cultural concept of biography. This and
further tests of the third hypothesis are reported with Study 2. In addition, results for the two knowledge tasks
will be analyzed for cross-study and cross-measure generalizability, thereby compensating for small sample
sizes.

Preliminary Study
The preliminary study served to construct two tasks of knowledge of biographical salience and age norms
to be used in the main studies by selecting items and establishing adult norms.
Method
Participants
A convenience sample of 20 medical students (mean age 21.6, SD = 1.1 years, range 20-24 years) and
eight middle-aged acquaintances of the author (mean age 42.2, SD = 4.2 years, range 37-49 years), with women
and men equally distributed, was tested either in a course for Medical Psychology or answered questionnaires
in their homes. All participants had at least 13 years of schooling. A small sample was deemed sufficient
because knowledge of norms was expected to vary only little between subjects.
Material
The questionnaire contained two sets of events. Participants were first instructed that “what is of concern
here is not your individual life, but what may be expected from a life narrative and life in general.” Judgments
concerned an imaginary average person of the participants’ own sex. The item “development of breasts” or
“development of beard” varied with the respondent’s sex.
The first task aimed at knowledge of the biographical salience of life events. Participants were “to imagine
a friend who was telling her (his) life story. Please indicate for each event, whether she (he) would talk about
this event or not. Does this event usually belong in a life narrative or not?” Fifty items expected to be salient
and 50 items not expected to be salient were selected. The 100 items included 60 age-graded events, such as
role transitions from the school and work domain, development of skills and body, and peak experiences like
having most professional success. The remaining events were 28 nonnormative life events (partly taken from
Geyer, 1992b), six family history events, and six historical events. Events from the family and general history
were included as these may play a role in the background of life narratives. All subsets of events were equally
divided between expectedly salient and nonsalient events.
Secondly, in the age estimation task participants were asked to “estimate at which age this event usually
or most frequently happens in the life of a woman (man)”. The identical 60 age-graded items from Task 1 were
presented in random order, comprising all events from a study Behnken and Zinnecker (1992) and all biological
and social items from the study by Schorsch (1992).
Results
Biographical salience. Based on the participants’ responses, 40 of the original 100 items were selected for
the biographical salience task, with 20 salient and 20 nonsalient items, including 22 age-graded events, 12
nonnormative life events, four family history and two historical events. Selection criteria were having an equal
number of biographically salient and nonsalient events, high agreement between respondents, and
heterogeneity of type of event. Only in the case of family history and historical events did we not succeed in
selecting half salient and half nonsalient events. Although only one family history event and no historical event
was judged to be biographically salient by a majority of respondents, we did include two events which
originally had been expected to be biographically salient (how parents got to know each other, fall of the Berlin
T. Habermas How to tell a life 6

wall) in order to stick to the preestablished numbers of each kind of item. There was a mean of 83.4%
agreement in the selected 40 items compared to 75.8% agreement across all 100 original events.
Age norms. Mean age estimates of when life events typically occur were taken as age norms for
subsequent studies. From the original 60 age-graded events a set of 25 was selected following the criteria of
high agreement (low standard deviation), heterogeneity of type of event, and a representation of the first nine
decades of life. Ten of the 25 items selected for the age norms task were also selected for the biographical
salience task.
Discussion
Agreement regarding biographical salience and age norms was reasonably high for most items; thus it
may be assumed that these answers are oriented by social norms. This justifies using the classification of
salience of events and mean age norms established in this preliminary study as norms by which to measure
knowledge of the cultural concept of biography in subsequent studies. Although participants’ responses
showed different degrees of variance for each event, weights were not attached to each event, as the aim of
these studies is to establish the development of the cultural concept of biography and not knowledge of
specific events. However, the differences in agreement have been taken into account when calculating
separate indicators for past and future events (hypothesis 2). To control the validity of these norms, they will
be compared with the means in both main studies in the age group included in all studies (19 to 21 years,
referred to as 20-year-olds). Furthermore, our selection of items will be compared to the free nomination of
biographically most salient events in Study 2.

Study 1
Method
This study was designed to test the hypothesis that knowledge of biographical salience and age norms
increases across the adolescent age range and the hypothesis that this is equally true for events from all phases
of life. To start with, means will be critically compared with the preliminary study and indicators of biographical
knowledge will be constructed.
Participants
A total of 48 participants was divided equally between gender and four groups with age means of 8.7 (SD
= .29), 12.6 (SD = .39), 16.5 (SD = .47) and 20.7 (SD = .45) years. They were the first six of each sex and age who
participated in a larger study of life narratives (Habermas & de Silveira, 2005). The 12- and 16-year-olds were
sampled from three classes of a typical Frankfurt secondary school (Gymnasium). The oldest group was
sampled from former students of the same secondary school. Because in Germany students are sent to three
different levels of school at age 10, secondary school being the top level, the youngest were sampled from the
top of three classes of an elementary school in Frankfurt. In the three younger age groups participation
depended first on parental permission and then on the children themselves. The oldest participants were
invited to participate by mail twice. Participants received € 20.
Procedure
All participants were tested twice, two weeks apart, by two different (out of three) female interviewers. In
the first session they recounted a life narrative, 10 to 15 minutes long, completed the two biographical
knowledge tasks, and were interviewed about one freely selected important event in their life. In the second
session, participants again recounted their life story to see whether talking and thinking about one’s life had
improved coherence in life narratives, and answered questionnaires. Only measures pertinent to this paper will
be reported. Each session lasted between 45 and 60 minutes.
Material
Knowledge of biographical salience. To facilitate the task for the younger participants, the questionnaire
format used in the preliminary study was substituted by a card sorting task. Instructions were identical to the
preliminary study, with the last sentence slightly modified: “Please consider for each of these events, whether
you would expect it to be included in a life narrative.” Participants could sort 40 events each written on a card
either on a pile saying “not necessarily included in a life narrative” or on a pile saying “included in a life
narrative.” Participants were asked whether they had any questions and then proceeded to distribute the
T. Habermas How to tell a life 7

cards.
Knowledge of age norms. Participants were presented 25 cards each with an event written on it which
had been selected in the preliminary study. A lengthy slip of paper (109 by 10.5 cm) with nine intervals was
presented which symbolized the first nine decades of life. “Life starts with birth at the left and ends with ninety
years at the right. Please put each card with this black arrow next to the age at which you think it usually occurs
in the life of a person. If you are not sure what the card means please ask me. If you think that there is no age
at which the event usually occurs you may put the card away.” Cards could be placed freely with unlimited
overlaps allowed.
Seven most important memories. Prior to recounting their life, participants wrote their seven most
important specific memories on cards and ordered them in chronological order on the table in front of them.
They were instructed to include these seven memories in their life narratives, and date them afterwards. The
kinds of events named for the first life narrative were categorized using the 100 categories of events originally
used in the preliminary study and with additional categories when necessary.
Results
Throughout the analyses, continuous variables were first tested for deviation from a normal distribution
and for outliers, also for the multivariate case where necessary. Outliers and variables were transformed to
approach normal distribution whenever necessary. First results of the biographical knowledge tasks will be
compared to the preliminary study and indicators will be constructed. Then results for hypotheses 1 and 2 are
reported.
Construction of Biographical Knowledge Tasks
Comparison of items for biographical knowledge with preliminary study. To check whether norms
established in the preliminary study are indeed general norms which apply across samples, mean values of all
items of both tasks were compared between the 20 year-olds in this study and the 20- to 21-year-old
participants in the preliminary study (see Tables 1 and 2, columns 1 through 4). Five out of 40 salience items
which had been judged as not biographically salient by over 50% of the 20-year-olds in the preliminary study
were judged to be indeed biographically salient by over 50% of the 20-year-olds in this study. Two of these five
items were judged to be salient by a majority of participants across all four age groups in this study (winning a
sports match, being able to make decisions about one’s clothing and hair) and were therefore excluded from
further analyses. One item, in contrast, was judged to be salient by a majority of the 20-year-olds, but not by
the other three age groups and was therefore retained as an example of an event without biographical salience
(writing in a diary). Two more events (how parents got to know each other, fall of Berlin wall) were excluded,
because, contrary to initial intentions, they again failed to be judged as biographically salient. Thus the
biographical salience scores were based on 36 events, half salient and half not salient. Testing group
differences in age estimation, none of the 25 items differed significantly, when a corrected alpha of p < .002
was used, so that the indicator could be based on all 25 items.
Construction of indicators of biographical knowledge. Indicators of biographical knowledge were
constructed as negative indicators, showing the degree of deviation from the norm. Normatively correct
answers were established on the basis of the preliminary study, categorizing an event as salient if the majority
of respondents estimated it to be salient, and using the mean age estimation as the norm for ages. The
indicator for knowledge of biographical salience was calculated as the percentage of wrong salience
attributions for the 36 events. The indicator for knowledge of age norms was calculated as the mean of the
absolute deviations from the normative age in years.
Age Differences Across Adolescence
Because the two dependent variables, knowledge of biographical salience and knowledge of age norms,
correlated with r(46) = .57 and no significant overall sex differences were detected, age differences in these
two aspects of biographical knowledge were tested simultaneously with a multivariate analysis of variance
(MANOVA) with four age groups as the only factor. The two indicators of biographical knowledge did vary
significantly with age, Pillai-Spur = .81, F(6, 88) = 10.03, p < .001, partial η2 = .41. Univariate analyses revealed
significant age effects for age estimation F(3, 44) = 44.87, p = .00, partial η2 = .75, and more moderate ones for
biographical salience F(3, 38) = 8.18, p = .00, partial η2 = .36. Alpha levels were adjusted to .015 for post hoc
T. Habermas How to tell a life 8

Scheffé tests which yielded significant differences between 8- and 12-year-olds for both dependent variables
and between 12- and 16-year-olds for age estimation only (all p < .001). Even the answers of the 8-year-olds
differed from chance in the salience task when tested against a chance outcome of 50% wrong salience
attributions, t(11) = -5.84, p < .001. It is difficult to determine what a chance outcome for age estimation would
be, but the 8-year-olds’ mean deviation of 7.2 years (SD = 2.7) does not suggest that they answered totally at
random.
Figure 1 shows similar age distributions for deviations from both conventional salience and normative
ages. As hypothesized, the largest difference in biographical knowledge resulted between the 8- and the 12-
year-olds. Contrary to expectations, the 20-year-olds’ answers deviated more strongly, though not significantly,
from the norm than the 16-year-olds.
In addition to comparing biographical knowledge to adult norms, another way of measuring knowledge
of norms at a group level is to interpret high agreement as a sign of knowledge of norms. Figure 1 shows fairly
similar standard deviations across ages for estimates of biographical salience, but stark differences between
the youngest and the three older groups in age estimation. In the Levene test of homogeneity of variance,
variances of age estimation differed significantly between age groups F(3, 44) = 23.93, p < .001. Thus the
greatest difference in biographical knowledge is between the 8- and the 12-year-olds, whereas the most
knowledge seems to be attained by age 16.
Is biographical knowledge acquired equally for events from all life phases?
To test the hypothesis that biographical knowledge is acquired equally for all events whether they
normatively happen in the participant’s past or present on the one hand or in their future on the other hand,
two separate indicators were created each for knowledge of biographical salience and of age norms. One of the
two indicators was for knowledge regarding events with age norms younger than or up to one year older than
the participant’s own age, the other for the remaining events with older age norms. As before, the indicators of
knowledge of salience were constructed as the percentage of wrong salience attributions, and the indicators of
knowledge of age norms as the mean absolute deviation from the norm. However, these indicators are defined
separately for past or present events and for future events. Therefore the number of events on which
indicators were based varied with the age of the participants. The indicators for salience together referred to
those 20 of 36 items which are age-graded. Typical ages for salience items were taken from the preliminary
study. All four indicators were based on z-standardized scores for each event to compensate for differences
between events regarding mean deviation from the correct answer (item difficulty). Positive values again
indicate a greater deviation from the norm, or below average knowledge, and negative values indicate above
average knowledge.
If biographical knowledge was gradually built up as events come within one’s extended present,
knowledge about events already experienced should be constantly high, whereas knowledge about events
lying in the personal future should be constantly low. If, in contrast, biographical knowledge was acquired as a
unitary concept, it would be acquired equally for all phases of life. Inspection of means in Figure 2 shows that
only knowledge of salience is learned equally for events from all phases of life. This is confirmed by two
analyses of variance (ANOVAS) for repeated measurement with the two knowledge indicators for past or
present and for future events as dependent variables and age as the only factor. Whereas knowledge of
salience did not differ for past versus future parts of life (F(1, 44) = .07, p = .79, partial η2 = .00), knowledge of
age norms did, F(1, 44) = 10.55, p = .002, partial η2 = .19, with a significant interaction between past or future
part of life and age, F(3, 44) = 4.22, p = .01, partial η2 = .22. To check age differences for each indicator,
separate ANOVAS were calculated, resulting in a significant effect of age on knowledge of age norms for future
events, F(3, 44) = 7.17, p = .001, partial η2 = .33, but not so for past events F(3, 44) = .77, p = .52, partial η2 =
.05. Apparently knowledge of age norms is learned on the basis of single events as their normative ages are
lived through, until a unitary cultural concept of biography is acquired at age 12 when knowledge about events
past and future no longer differ.
Discussion
As expected, biographical knowledge differed most between ages 8 and 12, and increased between 12
and 16 years. In addition, the increase in knowledge of age norms was paralleled by an increase in agreement
T. Habermas How to tell a life 9

within age groups. Because knowledge of the norm will lead to more uniform answers, the decrease of
variance with age is an additional confirmation of the increase in knowledge of the cultural concept of
biography between ages 8 and 12, continuing through age 16. However, the development of biographical
knowledge does not start only in adolescence, because even eight-year-olds had some biographical knowledge.
Age estimation might be taken as the better indicator of biographical knowledge, because knowledge of
biographical salience did not vary in agreement across age.
Two questions were not satisfactorily resolved by this study. For one, some doubt remains as to whether
ages for life events do have a normative quality which induces a high degree of agreement, because some
items were answered differently from the preliminary study. Also, there was a reversal of the trend of an
increase of biographical knowledge with age at age 20. This discontinuity of the age curve may raise questions
as to whether there are stable biographical norms in adulthood. If there was no relative stability in adulthood,
it would make no sense to measure the development of biographical knowledge by taking adults as the
standard for children and adolescents.

Study 2
This study was designed to test the hypothesis that biographical knowledge remains stable across
adulthood, and to test the hypothesis that biographical knowledge is acquired for past and future life phases in
the adult age range. The study was also undertaken to provide the data necessary to test the hypothesis that
the cultural concept of biography influences the selection of most important personal events to be included in
a life narrative, namely a selection of the most biographically salient events. The proportion of these
conventionally most salient events among personal memories and the proportion of normative life events
among personal memories will serve as two indicators of the influence of the cultural concept of biography
Following Berntsen and Rubin (2004), the third hypothesis will also be tested indirectly by testing the
hypothesis that the distribution of the most biographically salient events, of knowledge about them and of
positive memories across remembered life phases, follows the distribution described by the reminiscence
bump of autobiographical memories. Before reporting tests of hypotheses, mean values in the biographical
knowledge tasks are again compared to the other two studies to further check the generalizability of norms
across samples.
Method
Participants and Procedure
A convenience sample of 239 students in an introductory lecture of psychoanalysis participated on a
voluntary basis. Twenty-six questionnaires were excluded due to missing values, leaving 213 remaining
participants, of whom 155 were female, 54 male, and four did not disclose their gender. Because the lecture
was also attended by adults enrolled in a program for the elderly, and because in Germany students tend to be
older, mean age was 31.6 (SD = 16.0), ranging from 19 to 75 years. Four age groups were formed, ranging from
19 to 21 (M = 20.25, SD = .75), 22 to 25 (M = 23.00, SD = 1.01), 26 to 50 (M = 33.33, SD = 6.67), and 51 to 75
years (M = 61.85, SD = 5.19), comprising 65 (30.5%), 67 (31.5%), 40 (18.8%), and 41 participants (19.2%)
respectively. The distribution of gender across age groups did not differ significantly, χ2(3, N = 209) = 3.84, p =
.39. Questionnaires were answered during regular class time.
Material
Seven biographically most salient events in a normal life. As in the preliminary study, the questionnaire
opened with the instruction to consider not one’s own life, but what might be expected from the life of a
typical person with a normal life in German culture and of the same gender as the participant. Following
Berntsen and Rubin (2004), participants were asked to imagine such a typical person at the beginning of her or
his life and to write down “the seven most important events this person most probably would experience
during her/his lifetime. Please name specific events which happen at a specific time”. Events were then dated
and three emotional qualities were rated for each event (angering, nice, and frightening) on a 5-point rating
scale, ranging from ‘extremely’ to ‘not at all’.
Events were categorized using the categories of Berntsen and Rubin (2004) plus the categories used in
Study 1. These categories were grouped into four summary categories of institutional normative life events in
T. Habermas How to tell a life 10

the domains of school and work (e.g., begin school, retirement), of developmental-biological normative life
events in the domains of body and family (e.g., puberty, marriage), of clearly negative nonnormative life events
(e.g., death of parents, divorce) in the conception of a normal life, and a residual category. Table 3 lists the 41
categories that each made up at least 1% of nominations in either Study 1 or Study 2.
Knowledge of biographical salience and age estimation tasks. These tasks contained the items from
Study 1, but were presented in a questionnaire format as in the preliminary study.
Results
Checks of Biographical Knowledge Tasks
Representativeness of task items. To measure how representative the events chosen for the two
biographical knowledge tasks are for the most biographically salient events, we calculated how many of the
seven biographically most salient events were also part of our two tasks. A mean of 52.4% (SD = 16.4) of all the
nominated events were among the 18 salient events used in the biographical salience test, and 51.1 % (SD =
16.8) were among the 25 events used for age estimation, demonstrating that the two tasks cover over half of
all most relevant events. We explored the relative frequencies of kinds of life events. On average, 34.0% of the
seven most salient events in a normal person’s life were institutional normative life events, 46.5% were
developmental-biological normative life events, whereas only 11.3% were negative nonnormative life events,
leaving 8.2% in the residual category.
Comparison of items for biographical knowledge across studies. To again test the generalizability of the
cultural concept of biography across samples, we compared salience judgments and estimated ages between
Study 2 and the two preceding studies, controlling for age by comparing the 20-year-olds only. Two ANOVAS
with study as the between-subjects factor revealed no significant differences for either deviation from
normative biographical salience (M = 20.63%, SD = 15.42% in Study 2), F(2,86) = .25, p = .78, η2 = .09, or
deviation from age norms (M = 2.41 years, SD = .67 years in Study 2 – means and standard deviations for the
other two studies are reported above), F(2,86) = .84, p = .44, η2 = .19).
Age Differences Across Adulthood
The first hypothesis, that biographical knowledge does not vary across adulthood, was tested separately
for biographical salience and age estimation, because the two did not correlate with each other (r(211) = .18).
ANOVAS revealed significant age differences for age estimation, F(3, 209) = 3.92, p = .009, η2 = .05, but not so
for biographical salience, F(3, 209) = .61, p = .61, η2 = .01. Alpha levels were adjusted to .015 for post hoc
Scheffé tests, yielding significantly worse age estimation in the oldest group when compared to the second
youngest, aged 22 to 25, p = .003 (see Figure 3).
Again high agreement was interpreted as an indicator of knowledge of norms. Figure 3 shows fairly similar
standard deviations across ages for both biographical salience and age norms. However, the Levene test of
homogeneity of variances showed significant age differences for age estimation, F(3, 209) = 3.47, p = .017.
Standard deviations of age estimation increased continuously with age (in ascending order: M = 2.63, SD = .70;
M = 2.51, SD = .73; M = 2.68, SD = .94; M = 3.05, SD = .95 years).
Is biographical knowledge acquired equally for events from all life phases?
Indicators of biographical knowledge were constructed for events already experienced and for events not
yet experienced, as described in Study 1, to test the hypothesis that across adulthood biographical knowledge
does not differ for past and future life phases. When calculating two ANOVAS for repeated measurement with
knowledge for past and future events as dependent variables and age as factor, neither differences in
knowledge for past and future life phases (salience, F(1, 208) = .33, p = .57, partial η2 = .02) nor differences in
knowledge of age norms were found, F(1, 206) = .02, p = .89, partial η2 = .00.
Does the Cultural Concept of Biography Guide Selection of Most Important Memories?
To test the hypothesis that the cultural concept of biography is used to structure autobiographical
memory and life narratives (Habermas & Bluck, 2000) and the even stronger claim that the normative set of
age-graded events (life script) has a stronger influence on autobiographical memory and life narratives than
personal significance does (Rubin & Berntsen, 2003), we used two strategies. One was to look at the proportion
of conventionally salient and normative events among personal memories, the other was to test predictions
derived from the assumption that the distribution of conventionally salient events in the free nomination task
T. Habermas How to tell a life 11

shows a reminiscence bump.

Proportion of conventionally salient and normative events among personal memories. To exclude
developmental and historical influences, we compared the relative frequencies only of events named by the
20-year-olds in Study 1 and Study 2. Table 3 contains all categories which were named by over one percent of
all participants in either study, but shows relative frequencies for only the groups of 20-year-olds. All sixteen
events nominated for a normal life most frequently by the 20-year-olds (Table 3, columns 3 and 4) correspond
to events among the 24 most frequently named in the study by Berntsen and Rubin (2004). Among the 20-year-
olds, these 16 most frequent categories make up 80.11% (SD = 18.14) of all the events nominated for a normal
life, but only 26.67% (SD = 13.03) of the participants’ most important personal memories. However, some of
these events could not yet have been experienced by the narrators due to their age. Taking only those
conventionally salient events which typically take place before age 20, the remaining seven of the first 16
events still make up 71.08% (SD = 22.53) of all remaining conventionally salient events in a normal life as
compared to 24.29% of the personal memories (SD = 8.91), t(75) = 7.06, η2= .40, p < .001.
Thus, only a minority of the most important personal memories corresponds to the most conventionally
salient events. Even fewer of the personal memories are made up of events with age norms. All events,
including those in the category ‘other,’ were coded as either social-institutional age-graded events, biological-
familial age-graded events, negative nonnormative events, or other nonnormative events. Of the most salient
events nominated for a normal life a mean of 80.50 % (SD = 15.49) were normative versus 18.63% (SD = 14.93)
nonnormative life events, whereas important personal memories contained a mean of only 42.62% (SD =
19.28) normative versus 55.00% (SD = 20.10) nonnormative events. The difference between the proportion of
normative and nonnormative events differed significantly between the most salient events in a normal life and
personal memories, t(75) = 6.26, p < .001, η2= .34. Thus both most conventionally salient events and events
with cultural age norms make up a substantial part of events selected for personal life narratives, but by far not
the majority of events.
If indeed there was an influence of the cultural concept of biography on the retrieval of important
personal memories, the proportion of biographically salient and normative events in life narratives should
increase across adolescence as knowledge of the cultural concept of biography is acquired. Comparing the
relative frequency of the 16 biographically most salient events among the personal memories across all age
groups in Study1 no significant differences were found, (8-year-olds M = 12.7%, SD = 15.77%, 12-year-olds M =
17.86%, SD = 16.26%, 16-year-olds M = 16.67%, SD = 10.25%, and 20-year-olds M = 26.67%, SD = 13.03%), F(3,
44) = 2.11, p < .12, partial η2 = .13. When also comparing the relative frequencies of normative versus
nonnormative life events across ages, there are significant age differences, F(3, 44) = 5.85, p < .002, partial η2 =
.29. However the proportion of normative events increased in a noncontinuous fashion with age (means were
18.45%, 22.61%, 17.86%, and 42.62%, standard deviations 16.56%, 19.70%, 17.86%, and 19.28%). The only
large increase was between the 16- and the 20-year-olds, which does not correspond to the developmental
pattern of the knowledge of the cultural concept of biography as found in Study 1.
Do conventionally salient events have characteristics of the reminiscence bump? Berntsen and Rubin
(2004) hypothesize that the life script of normative events structures retrieval of memories and is therefore
responsible for the overrepresentation of memories from ages 15 to 30. They therefore suggest that more
biographically salient events come from this period than from other periods of life, that events from this period
are mentally represented in more precise fashion, and that positive emotions are connected more to events
from this period than from other periods. We used the events nominated for a normal life to test these
predictions.
The distribution of biographically salient events across life indeed shows a ‘bump,’ which in our data peaks
in the five year period between 15 and 20 years of age (Figure 4, first diagram). However, when looking at 15-
year periods, almost as many events regarded the first 15 years (40.7%) as the second 15 years (46.7%).
To test the hypothesis that knowledge of events from the bump is better than knowledge of events from
before or after the bump, the events used in the salience and age norm tasks were grouped into events falling
within the bump (age norms 15-30 years, with seven events for salience and ten events for age norms) and
outside the bump. Again z-standardized scores were used. Mean knowledge of events from the bump period
T. Habermas How to tell a life 12

(M = .00, SD = .44 for salience, M = .00, SD = .45 for age norms) did not differ from mean knowledge for other
periods (M = .00, SD = .38. and M = .00, SD = .34 respectively), t(212) = -.003, p = .998, η2 = .00, and t(212) =
.065, p = .948, η2 = .00.
To explore whether the ratings of events as ´nice´ peaked in the bump years whereas ratings of events as
´scary´ and ´angering´ were evenly distributed, as would be expected by the rationale of Rubin and Berntsen
(2003) that positive events peak in the bump whereas negative events are roughly evenly distributed across
the ages, emotion ratings of the seven most salient events in a normal life were averaged for every five years.
The ‘nicest’ events stemmed from the period between 20 and 25 years. Events located in the first thirty years
were marked by positive valence (‘nice’), moderate anxiety and little anger, whereas events located after age
30 were rated as scary, angering and not so nice (Figure 4, diagram 2), roughly confirming expectations.

Discussion
This is a first attempt to explore the development of the cultural concept of biography by using two of its
aspects which regard everyday conceptions of the life course, biographical salience of and age norms for life
events. Two tasks were constructed to test biographical knowledge by comparing it to adult norms. Age norms
appear to be more well defined than biographical salience of events. Also, the suggested relation between
events with biographical salience and with age norms was confirmed. While the development of some skills
does have age norms, these were not listed among the biographically most salient events, in contrast to the
study by Berntsen & Rubin (2004), in which learning to walk and to talk were at positions 20 and 26. On the
other hand, biographically salient nonnormative events have no age norms.
As expected, knowledge of conventional biographical salience and of cultural age norms does indeed
increase between childhood and mid-adolescence. Confirming results by Schorsch (1992), 8-year-olds’
knowledge is already above chance level. After a peak at age 16, scores in the biographical knowledge tasks
decrease in the 20-year-olds, to remain fairly stable across adult age groups with a trend to again decrease in
the older adults. In line with Strube et al. (1985), the slight decrease in scores can be interpreted to reflect not
a decrease in knowledge but a loosening of stereotypic views of the cultural concept of biography with age.
The alternative explanation of cohort differences in the cultural concept of biography would predict an
increasing deviation from the norm with age, but a constant amount of agreement. Because standard
deviations do increase across adulthood for age estimation, the developmental interpretation of increasing
individuality is more convincing than a cohort explanation.
However, it cannot be excluded that the apparent decline in biographical knowledge in the respective
oldest groups in both studies is due to sampling biases. In Study 2, most of the older participants came through
a special program for the elderly, and all older students, it could be argued, were themselves deviating from a
standard life course by going to school at an age at which it does not serve to prepare for the job market,
thereby perhaps biasing answers towards atypical conceptions of the life course. In Study 1, former students of
the secondary school were chosen for the age 20 group to ensure compatibility with the younger groups.
Nevertheless there may have been some selection bias in that we mostly reached the less mobile participants
who were still living in Frankfurt, some of whom had not yet settled on going to college or a work arrangement,
thereby possibly biasing answers towards an alternative atypical conception of the life course. To resolve this
ambiguity, future studies require even more homogenous samples. Also, in order to differentiate between
knowledge of norms and judgments of the variability of factual life courses, which supposedly motivates more
heterogeneity in older participants, knowledge of normative and actual timing, and of conventional and
personal biographical salience should be measured separately.
The pattern of acquisition of biographical knowledge speaks to its conceptual nature: when a cultural
concept of biography is acquired, it is learnt as a unitary concept with equal amounts of information about
events from all phases of life, regardless of whether the normative age for an event has been passed or not.
The only exception regards the time before a cultural concept of biography is acquired: until then age norms
are learned one by one as their normative ages are reached. This is not so for biographical salience, maybe
because it requires comparison with all other life events and therefore a concept of the life course as a frame
T. Habermas How to tell a life 13

of reference. The pattern of acquisition does not indicate what the mechanisms of learning biographical
knowledge are. Therefore possible ways of learning biographical knowledge need to be explored, such as
modelling by relatives, reading biographies or novels, or watching talk shows on television.
The expectation that biographical knowledge is used for selecting events to be included in biographies
was only partially confirmed. Only a minority of most important personal memories correspond to the
biographically most salient events, and only a minority of memories regards normative events. Therefore we
may conclude that a majority of personal memories is individual, whereas a substantial minority of personal
memories is oriented by a normative conception of the life course. The developmental evidence for an increase
of the proportion of conventionally salient and normative events with age was tentative. Possibly the
conventionality of personal memories also depends on personality traits such as social conformity.
Also the evidence for a reminiscence bump in the distribution of conventionally salient events and related
differences in knowledge and evaluation was mixed: of all life periods, the second 15-year period contains the
most biographically salient events, and events from this period are also valued most positively, suggesting that
the reminiscence bump, i.e. the preference for memories from late adolescence and early adulthood in adults
older than 40, may in part be derived from these characteristics of the set of biographically salient events. This
impression, however, is tarnished by almost similar findings for the first 15 years of life. Thus the reminiscence
bump may partially be attributed to the cultural concept of biography.
This study has implications for research on time in memory, identity in memory, and culture in memory
and self development. This study confirms the studies by Montangero (1996) that it is between late childhood
and early adolescence that a basic sense of the life course is acquired. He found, for instance, that by the age of
10 children understand that their drawing ability has evolved over the years, and by age 11-they are able to
describe a series of drawings showing milestones of the life of a character in a single sentence by using the
whole life as a frame of reference. Also Friedman’s (1993) studies of the use of calendar time in memory point
to early adolescence as the time when the calendar is used to date past events. According to Friedman (2004)
the personal past is reconstructed with the help of conventional time patterns of the calendar and of highly
individual, overlapping lifetime periods. This research, however, tentatively supports the notion that the
personal past is also reconstructed with the help of a unitary and ordered conventional set of biographically
salient events which form a skeletal structure of life. This life script also plays a role in the autobiographical
knowledge base as defined by Martin Conway, since it can be considered the conventional aspect of the
individual life story schema (Bluck & Habermas, 2000), which is conceived of as the most abstract and
overarching knowledge structure in autobiographical memory (Conway, Singer & Tagini, 2004).
Rubin and Berntsen (2003) treat the explanation of the reminiscence bump with life script and identity
accounts as rivals, since they interpret identity formation as an individual process. However, the Eriksonian
(1968) concept of psychosocial identity is not individualistic, but describes the integration of the individual into
society by way of identification with roles and their individual combination and transformation. Therefore
conventionally salient events may well serve to define the self, for instance as an adult with a mature sexual
identity or as a married person. This does not preclude individual interpretations of conventionally salient
events. Accordingly, among McLean’s (2005) self-defining memories of late adolescents, there were
conventionally salient events such as falling in love, death of someone close and parents’ divorce. Using at least
some conventionally salient events to define oneself in a life narrative may even be imperative in order not to
violate listeners’ expectations that identity is also defined with reference to an everyday conception of a
normal life.
In addition, if the overwhelming majority of most important memories may be considered to be self-
defining, then the amount of conventionally salient events among personally important memories reflects the
conventionality of an individual’s identity. This interpretation of individual differences suggests a way to test
the relative influence of the cultural set of biographically salient events versus the influence of identity
formation, which could be measured by the correlation between the reminiscence bump and the
conventionality of personal memories.
Finally, the acquisition of the cultural concept of biography in early adolescence does not contradict, but
complements Katherine Nelson’s (2003) conception of the emergence of a cultural self. She speaks of the use
T. Habermas How to tell a life 14

of cultural myths which are learned and used during the early school years as “models of the ideal life, on the
basis of which individual lives may be led or reformulated in terms of a life story that conforms to the model”
(Nelson, 2003, p. 22), intending myths of kings, queens and superheroes. This research shows that a more
realistic, conventional conception of the life course is acquired later, in early adolescence. Only this later
accomplishment may serve as a somewhat realistic guideline for living and recounting one’s life. The cultural
concept of biography is yet another, more mature form culture offers to individuals to create their individual
path in reference to what is considered normal in a culture.
This research has some limitations based on the samples studied and the methods used. Samples were
homogenous in terms of social class and culture. The cultural, and possibly social, relativity of the concept of
biography could not be studied. On the other hand, samples may not have been homogenous enough across
age groups, as discussed above. Also, the participants in the adult sample were 3/4 female and 2/3 were
between the ages of 19 and 25, the rest spreading across a span of 50 years. Therefore it cannot be excluded
that age differences may be biased by gender, and the two older age groups comprise very large age ranges.
Future studies will have to test the development of biographical knowledge with different and more
controlled samples. The relation between biographical knowledge as tested in this study and other aspects of
the cultural concept of biography remain to be explored, as does the use of biographically salient events and
age norms for structuring life narratives. In a more general vein, the emergence of a biographical
understanding of individuals, that is a biographical person and self concept remains to be studied.

References
Bauer, P. (2002). Early memory development. In U. Goswami (Ed.), Blackwell handbook of childhood cognitive
development (pp. 127-146). Malden, MA: Blackwell.
Behnken, I., & Zinnecker, J. (1992). Lebenslaufereignisse, Statuspassagen und biografische Muster in Kindheit
und Jugend [Life events, status transitions, and biographical patterns in childhood and adolescence]. In J.
Zinnecker (Ed.), Jugend ´92, Band 2 (p. 127-144). Opladen, Germany: Leske & Budrich.
Berntsen, D., & Rubin, D. C. (2004). Cultural life scripts structure recall from autobiographical memory. Memory
& Cognition, 32, 427-442.
Bluck, S., & Habermas, T. (2000). The life story schema. Motivation and Emotion, 24, 121-147.
Conway, M., Singer, A., & Tagini, A. (2004). The self and autobiographical memory: Correspondence and
coherence. Social Cognition, 22, 491-529.
D’Alessio, M. (1990). Social representations of childhood: An implicit theory of development. In G. Duveen & B.
Lloyd (Eds.), Social representations and the development of knowledge (p. 70 - 90). Cambridge, United
Kingdom: Cambridge University Press.
Dekovic, M., Noom, M., & Meeus, W. (1997). Expectations regarding development during adolescence:
Parental and adolescent perceptions. Journal of Youth and Adolescence, 26, 253-272.
Erikson, E. H. (1968). Youth and identity. New York: Norton.
Friedman, W. J. (1993). Memory for the time of past events. Psychological Bulletin, 113, 44-66.
Friedman, W. J. (2004). Time in autobiographical memory. Social Cognition, 22, 591-605.
Geyer, S. (1992a). Lebensverändernde Ereignisse, soziale Ungleichheit und Belastungen [Life-changing events,
social inequality, and stress]. In J. Zinnecker (Ed.), Jugend ´92, Band 2 (p. 221 - 237). Opladen, Germany:
Leske & Budrich.
Geyer, S. (1992b). Methodische Aspekte der Erfassung lebensverändernder Ereignisse [Methodological aspects
of measuring life-changing events]. In J. Zinnecker (Ed.), Jugend ´92, Band 4 (p.27-40). Opladen, Germany:
Leske & Budrich.
Greene, A. L. (1990). Great expectations: Constructions of the life course during adolescence. Journal of Youth
and Adolescence, 19, 289-305.
Greene, A. L., Wheatley, S. M., & Aldava, J. F. (1992). Stages on life’s way: Adolescents’ implicit theories of the
life course. Journal of Adolescent Research, 7, 364-381.
Habermas, T. (in press). „Kann ich auch ganz, ganz am Anfang anfangen, als ich noch ganz klein war?“ Wie
Jugendliche lernen, erzählte Lebensgeschichten zu eröffnen und beenden [„Can I start at the very
T. Habermas How to tell a life 15

beginning, when I was real little?“ How adolescents learn how to open and end a life narrative]. In H.
Markowitsch & H. Welzer (Eds.), Autobiographisches Gedächtnis. Stuttgart, Germany: Klett-Cotta.
Habermas, T., & Bluck, S. (2000). Getting a life: The emergence of the life story in adolescence. Psychological
Bulletin, 126, 748-769.
Habermas, T., & de Silveira, C. (2005). The development of the life story in adolescence. Unpublished
manuscript.
Hareven, T. (1986). Historical changes in the family and the life course: Implications for child development.
Monographs of the Society for Child Development, 50, 8-23.
Heckhausen, J., Wrosch, C., & Fleeson, W. (2001). Developmental regulation before and after a developmental
deadline: The sample case of „biological clock“ for childbearing. Psychology and Aging, 16, 400-413.
Helson, R., Mitchell, V., & Moane, G. (1984). Personality and patterns of adherence and nonadherence to the
social clock. Journal of Personality and Social Psychology, 46, 1079-1096.
Linde, C. (1993). Life stories -The creation of coherence. New York: Oxford University Press.
McLean, K. (2005). Late adolescent identity development: Narrative meaning making and memory telling.
Developmental Psychology, 41, 683-691.
Miller, M. A., & Rahe, R. H. (1997). Life change scaling for the 1990s. Journal of Psychosomatic Research, 43,
274-292.
Montangero, J. (1996). Understanding changes in time – The development of diachronic thinking in 7 to 12 year
old children. London: Taylor & Francis.
Nelson, K. (2003). Narrative and self, myth and memory: Emergence of the cultural self. In R. Fivush & C. A-
Hadenn (Eds.), Autobiographical memory and the construction of a narrative self (pp. 3-28). Mahwah, NJ:
Erlbaum.
Nelson, K., & Fivush, R. (2004). The emergence of autobiographical memory: A social cultural developmental
theory. Psychological Review. 111, 486-511.
Neugarten, B. L., Moore, J., W., & Lowe, K. C. (1965). Age norms, age constraints, and adult socialization.
American Journal of Sociology, 70, 710-717.
Pillemer, D. B. (1998). Momentous events, vivid memories. Cambridge, MA: Harvard University Press.
Rubin, D. C., & Berntsen, D. (2003). Life scripts help to maintain autobiographical memories of highly positive,
but not highly negative events. Memory & Cognition, 31, 1-14.
Schorsch, S. (1992). Die Entwicklung von Konzepten über das Lebensalter bei Kindern und Jugendlichen
[Development of concepts of life-phases in children and adolescents]. Münster, Germany: Waxmann.
Schütze, F. (1982). Narrative Repräsentation kollektiver Schicksalsbetroffenheit [Narrative representations of
collective experiences]. In E. Lämmert (Ed.), Erzählforschung (pp. 568 - 590). Stuttgart, Germany: Metzler.
Settersten, R. A., & Hägestad,, G. O. (1996a). What’s the latest? Cultural age deadlines for family transitions.
The Gerontologist, 36, 178-188.
Settersten, R. A., & Hägestad,, G. O. (1996b). What’s the latest? Cultural age deadlines for educational and
work transitions. The Gerontologist, 36, 602-613.
Settersten, R. A., & Mayer, K. U. (1997). The measurement of age, age structuring, and the life course. Annual
Review of Sociology, 23, 233-261.
Silbereisen, R. K., & Wiesner, M. (2000). Cohort change in adolescent developmental timetables after German
reunification: Trends and possible reasons. In J. Heckhausen (Ed.), Motivational psychology of human
development (p. 271 - 284). Amsterdam: Elsevier.
Singer, J. A., & Salovey, P. (1993). The remembered self: Emotion and memory in personality. New York: Free
Press.
Strube, G., Gehringer, M., Ernst, I., & Knill, K. (1985). Knowing what's going to happen in life. (Technical Reports
No. 13 and 14). München, Germany: Max Planck Institute for Psychological Research.
Willemsen, M. E., & van de Vijver, F. J. R. (1997). Developmental expectations of Dutch, Turkish-Dutch, and
Zambian mothers: Towards an explanation of cross-cultural differences. International Journal of Behavioral
Development, 21, 837-854.
T. Habermas How to tell a life 16

Table 1
Mean Percentages of 20-Year-Olds in all Three Studies Judging Life Event as Biographically Salient
Pre Study Study 1 Study 2
N = 10 N = 12 N = 65
M SD M SD M SD
Age-graded events
Birtha 80.00 42.16 83.33 38.92 60.00 49.37
b
Playing with a rattle 20.00 42.16 25.00 45.23 3.08 17.40
b
Looking at children’s’ books 20.00 42.16 25.00 45.23 21.54 41.43
Losing last milk-toothb 30.00 48.30 8.33 28.87 16.92 37.79
First day of schoola 70.00 48.30 75.00 45.23 81.54 39.10
First time household choresb 30.00 48.30 25.00 45.23 23.08 42.46
Start deciding about own clothingc 50.00 52.70 83.33 38.92 35.38 48.19
a
First time in love 80.00 42.16 75.00 45.23 98.46 12.40
b
Start writing diary .00 .00 58.33 51.49 29.23 45.84
First time go to discob
40.00 51.64 41.67 51.49 50.77 50.38
First boy-/girl-frienda
80.00 42.16 91.67 28.87 90.77 29.17
First time travelling on one’s owna
70.00 48.30 75.00 45.23 73.85 44.29
Leave schoola 90.00 31.62 100.00 .00 90.77 29.17
Start smoking regularlyb 20.00 42.16 58.33 51.49 46.15 50.24
Winning a sport’s competitionc
30.00 48.30 83.33 38.92 40.63 49.50
a
Leave parents’ home 90.00 31.62 100.00 .00 90.77 29.17
Marriagea 90.00 31.62 100.00 .00 95.38 21.15
Birth or adoption of childa 100.00 .00 91.67 28.87 93.85 24.22
a
Death of a parent 90.00 31.62 100.00 .00 95.38 21.15
b
Hair turning grey 20.00 42.16 58.33 51.49 46.15 50.24
Retirementa 100.00 .00 100.00 .00 76.92 42.46
Need to use walking stickb 10.00 31.62 66.67 49.24 20.31 40.55
Nonnormative events
Change of schoola 77.78 44.10 83.33 38.92 83.08 37.79
Repeating a class in schoola 60.00 51.64 83.33 38.92 70.77 45.84
Relocation to another citya 90.00 31.62 91.67 28.87 89.23 31.24
Separation of parentsa
80.00 42.16 100.00 .00 78.46 41.43
Severe accidenta 90.00 31.62 91.67 28.87 83.08 37.79
Loss of frienda
90.00 31.62 91.67 28.87 84.38 36.60
Visit of a cinemab 20.00 42.16 8.33 28.87 20.00 40.31
First time shopping by oneselfb
20.00 42.16 16.67 38.92 29.23 45.84
T. Habermas How to tell a life 17

Having measlesb 40.00 51.64 16.67 38.92 20.00 40.31

Losing mittens or hatb
20.00 42.16 .00 .00 20.00 40.31
b
Falling off a bike 20.00 42.16 33.33 49.24 23.08 42.46
Losing one’s purseb 10.00 31.62 25.00 45.23 32.31 47.13
Family facts and history
How one’s parents first metc 30.00 48.30 41.67 51.49 46.15 50.24
Parents’ age and professiona 70.00 48.30 66.67 49.24 59.38 49.50
Parents moving into shared apartmentb 10.00 31.62 16.67 38.92 10.77 31.24
b
Parents’ color of eyes and hair 10.00 31.62 8.33 28.87 20.00 40.31
Historical events
Fall of the Berlin wallc 50.00 52.70 25.00 45.23 26.98 44.74
Move German government to Berlinb 10.00 31.62 8.33 28.87 6.15 24.22
Note: Interpretation of an event as conventionally salient or not is based on all participants of the preliminary
study and Study 1: a – biographically salient event; b – biographically not salient event; c –excluded from the
indicator of knowledge of salience.
T. Habermas How to tell a life 18

Table 2
Mean Estimated Ages in Age Estimation Task for 20-Year-Olds in All Three Studies
Age-Estimated Life Event Pre Study Study 1 Study 2
N = 10 N = 12 N = 65
M SD M SD M SD
Learn to talk 1.75 .72 1.63 .77 2.05 1.64
Go to nursery school 3.50 .53 3.33 .49 3.38 .80
First day of school 6.10 .32 6.33 .49 6.14 .32
First time household chores 10.89 2.52 8.67 2.19 11.56 2.77
Development of breast/beard 13.80 1.81 14.00 2.45 12.96 1.53
First time in love 13.10 1.91 13.17 2.41 13.66 1.79
First time drink glass of alcohol 15.10 .88 14.25 1.14 14.73 1.13
First time go to disco 14.67 1.22 15.42 1.51 15.19 1.22
First boy-/girl-friend 16.00 2.05 14.67 1.56 15.91 1.65
First holiday job 15.80 1.55 14.92 1.68 16.04 1.64
First time no curfew 16.60 .97 17.50 .80 16.92 1.24
First time travelling on one’s own 16.60 1.26 17.58 2.78 17.25 2.00
Driver’s licence 18.10 .32 18.08 .51 17.89 .45
Having intensive political discussions 17.50 1.58 17.92 2.87 18.21 2.71
Leave parents’ home 19.70 1.49 20.67 2.57 20.72 2.28
Finish professional training 21.90 2.02 23.58 3.65 23.32 2.38
First time work fulltime 26.50 3.14 23.25 3.65 23.92 2.99
Marriage 26.60 2.12 28.33 2.74 26.98 2.80
Buy apartment or house 32.20 4.16 37.08 3.78 32.89 6.49
Divorce 34.33 3.50 38.08 6.80 37.84 5.89
Children leave house 44.40 9.61 45.75 9.04 48.05 5.69
Death of a parent 56.60 12.07 53.17 7.84 48.62 9.13
Retirement 63.00 2.58 62.67 3.73 63.32 4.76
Move into home for the elderly 77.33 3.94 74.17 9.23 76.19 6.34
Death 83.00 6.65 77.33 4.48 81.60 7.22
T. Habermas How to tell a life 19

Table 3 Comparison of Relative Frequencies of Conventionally Salient Events in Normal Life and in Life
Narratives in 20 year-olds, with Ages of Events in Normal Life
Percent of Events Named Typical Age
Study 1 Study 2 Study 2
N = 12 N = 67 N = 67
% Events in % Events in Age of Events in
Personal Memories Normal Life Normal Life
M SD M SD M SD
a
College, job training 15.48 9.55 12.69 9.81 18.85 2.95
a
Begin school 3.57 6.46 10.67 6.45 6.28 .91
b
Marriage .00 .00 9.49 7.72 26.31 3.77
b
Having children .00 .00 8.73 7.66 28.94 3.20
b
Fall in love/first partner 1.19 4.12 7.69 9.48 15.19 4.10
a
First job .00 .00 5.93 7.53 23.23 5.23
a
Begin daycare 1.67 5.77 4.96 7.03 3.37 .75
b
First sex 1.19 4.12 4.62 6.73 15.74 1.21
b
Puberty .00 .00 2.89 5.84 12.67 1.06
b
Leave home 1.19 4.12 2.51 5.64 20.21 3.47
c
Other’s death .00 .00 2.20 6.13 29.76 15.33
a
Retirement .00 .00 1.98 4.97 64.18 2.75
c
Parent’s death .00 .00 1.98 4.97 45.23 10.37
a
Settle on career 1.19 4.12 1.58 4.58 20.42 3.40
c
Own death .00 .00 1.32 4.17 76.16 11.49
b
First kiss .00 .00 1.10 3.84 12.92 2.00
d
First friend .00 .00 .95 4.7 7.88 6.24
c
Major illness or accident 1.19 4.12 .88 3.46 35.63 25.02
d
Vacation, trip 8.33 11.33 .22 1.77 14.38 11.53
d
Relocation 5.95 12.86 .44 2.49 16.86 8.28
d
Internship, courses 5.95 9.55 .22 1.77 19.00 .
d
Leisure activity 4.05 7.46 .00 .00 34.00 43.84
b
Birth 3.57 6.46 .44 2.49 .00 .00
d
Migration 3.57 6.46 .00 .00 . .
Illness or accident of
2.86 6.78 .00 .00 . .
significant otherc
Activities with familyd 2.86 6.78 .00 .00 1.25 1.06
c
Separation of parents 2.38 5.56 .22 1.77 10.25 6.40
T. Habermas How to tell a life 20

Own room or apartmentd 2.38 5.56 .00 .00 . .

a
Finish grade school 1.19 4.12 1.10 3.84 10.43 .53
d
Quarrel 1.19 4.12 1.10 3.84 9.20 4.83
d
Major achievement 1.19 4.12 .00 .00 26.50 5.97
d
Having peers 1.19 4.12 .66 3.02 13.00 4.08
d
Acquiring or losing object 1.19 4.12 .00 .00 . .
a
School-related events 1.19 4.12 .88 4.28
d
Siblings .00 .00 .70 3.20 3.28 1.79
d
Birthday .00 .00 .88 3.46 3.00 3.16
c
Minor illness or accident .00 .02 .22 1.77 6.82 1.33
d
Sports .00 .00 .00 .00 . .
d
Celebration .00 .00 .00 .00 13.75 7.27
d
Animals .00 .00 .00 .00 . .
d
Other events 21.90 14.12 10.59 12.96 25.08 20.73
No specific event 2.38 5.56 0.88 3.46 18.38 16.16

Note: All events are shown which were nominated by at least 1% of all participants in Study 1 (personal
memory) or by all participants in Study 2 (event in normal life). Some events have zero frequencies because
none of the 20 year-olds nominated them. First events are listed which made up over 1% of important events
in a normal life in the whole sample of Study 2, then remaining events which made up over 1% of personal
memories in the whole sample of Study 1.
a – institutional normative life events in school and work; b – developmental-biological normative life events
concerning body and family; c – clearly negative nonnormative life events; d – other events.
T. Habermas How to tell a life 21

Figure Caption
Figure 1. Study 1: Mean (+/- SD) deviation from cultural concept of biography-norm across adolescence:
Proportion of wrong salience attribution and absolute deviation from age norms in years.
Figure 2. Study 1: Mean z-standardized deviation from cultural concept of biography-norm across
adolescence by having or having not experienced the life events judged for biographical salience and age
norms.
Figure 3. Study 2: Mean (+/- SD) deviation from cultural concept of biography-norm across adulthood:
Proportion of wrong salience attribution and absolute deviation from age norms in years.
Figure 4. Study 2: Events in a Normal Life by Age at the Time of Event: Relative Frequencies and Mean
Emotional Qualities. Each five-year interval is labelled by the oldest age included.
 T. Habermas

Incorrect Attribution of Biogra phical Salience

Absolute Deviation from Age Norm

10
20
30
40

2
4
6
8
10
8

8
How to tell a life

12

12
Age

Age
16

16
20

20
22
T. Habermas How to tell a life 23

events already
experienced
events not yet
experienced
Incorrect Attribution of Biographical Salience

0,4

0,2

0,0

-0,2

8 12 16 20
Age

events already
experienced
events not yet
experienced
0,4
Absolute Deviation from Age Norm

0,2

0,0

-0,2

8 12 16 20
Age
T. Habermas How to tell a life 24
Incorrect Attribution of Biogra phical Salience

40

30

20

10

19 - 21 22 - 25 26 - 50 51 - 75
Age
Absolute Deviation from Age Norm

10

18 - 21 22 - 25 26 - 50 51 - 75

Age
T. Habermas How to tell a life 25

Age
25%
19 - 25
26 - 75

20%
Relative Frequency of Events

15%

10%

5%

0%

5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85
Age at Time of Event

4
nice
Mean Rating of Events from 0 (not at all) to 4 (absolutely)

scary
angering

5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85
Age at Time of Event